Frank D Gunstone - Lipid Technologi PDF

Lipid Technologies
and Applications
Lipid Technologies
and Applications
edited by
Frank D. Gnnstone
Myinefield Research Services, Ltd.
Scottish Crop Research Institute
Invergowrie
Dundee, Scotland
Fred B. Padley
Consultant
Loders Croklaan, b.v.
A Division of Quest International
Wormerveer, The Netherlands
MARCEL
Marcel Dekker, Inc . N ew York • Basel •H ong Kong

DEKKER
The Library of Congress Cataloging-in-Publication Data
Lipid technologies and applications / edited by Frank D. Gunstone,

Fred B. Padley.
p. cm.
Includes index.
ISBN 0-8247-9838-4 (alk. paper)
I. Lipids. I. Gunstone, F. D. II. Padley, F. B. (Fred B.)
TP453.L56L558 1997
664’.3— dc21
97-8301
CIP
The publisher offers discounts on this book when ordered in bulk quantities. For more
information, write to Special Sales/Professional Marketing at the address below.
This book is printed on acid-free paper.
Copyright © 1997 by MARCEL DEKKER, INC. All Rights Reserved.
Neither this book nor any part may be reproduced or transmitted in any form or by any
means, electronic or mechanical, including photocopying, microfilming, and recording, or by
any information storage and retrieval system, without permission in writing from the
publisher.
MARCEL DEKKER, INC.

270 Madison Avenue, New York, New York 10016
http://W W W . dekker. com
Current printing (last digit):

10 9 8 7 6 5 4 3 2 1
PRINTED IN THE UNITED STATES OF AMERICA

Preface
This book is concerned with the wide-ranging use of oils and fats. Annual production of such
materials from vegetable and animal sources will be around 90 -1 0 0 million tonnes by the end
of this century. A significant part o f these products is used in the country o f origin, while
roughly two-thirds is exported or imported in the form of seed or extracted oil. The use of
such materials is more widespread than is generally realized.
Most fat and oil products are used for human food, and some, such as milk, butter, olive
oil, and fish oil, have been part of the human diet for centuries. Others, such as spreading
fats, baking fats, frying oils, salad oils, ice cream, chocolate, and vanaspati, are largely
products o f the 20th century. About 80% o f these products is consumed in these forms and
6% is used in animal feed.
Nonfood uses account for the remaining 14%. These uses are many and varied, and fat-
based products find their way into many areas of human life. Most nonfood uses of fatty acids
and their derivatives are based on their amphiphilic nature (i.e., their ability to be involved
with both polar and nonpolar materials at the same time). Soap for cleansing and oils for their
emollient properties and as carriers o f pigments and perfumes are examples o f per
sonal care products that have been used for centuries. These are now joined by other fat-based
products for personal care, domestic and industrial cleaning, coatings, and lubricants, etc.
Today these materials have added significance because they are produced from renewable
resources in contrast to nomenewable resources such as oil, gas, and coal and they are
readily biodegradable, so that when they have served their purpose, they are easily converted
to simpler molecules for reuse in natural processes.
Biotechnology plays an increasingly important role in the production and usage of lipids.
This is already apparent in two ways. The first way is in opportunities to improve and modify
the yields and the characteristics of oils, particularly from vegetable and microbiological
sources. Second, procedures are being developed to modify natural fats using enzymes as
catalysts as an alternative to those that are now extensively employed often under harsh tem
perature and pressure conditions.
Ill
IV Preface
This book is divided into six parts, covering the nature of lipids and their major sources,
while the second part examines the processes by which these lipids are recovered from natural
sources, how they are refined, and how they are modified to broaden their range of uses: the
third, fourth, and fifth parts focus on the major food uses, and the sixth on the major nonfood
uses. Each part comprises several chapters and the authors of each chapter were asked to
prepare authoritative accounts of their topics: these accounts cover the present state of the
topic, the required starting materials, the volume of the products obtained, and indications of
future trends. We would like to express our gratitude to all the contributors to this book.
Frank D. Gunstone
Fred B. Padley
Contents
Preface III
Contributors ix
Part I Introduction
1. Fatty Acids and Lipids Structure 1
F ranko. Gunstone
2. Major Sources of Lipids 19
F ranko. Gunstone
3. Phospholipids 51
Michael Schneider
4. Lipids and Nutrition 79
Michael I. Gurr
Part II Processing
5. Extraction of Lipids from Natural Sources 113
Maurice A. Williams
6. Refining 137
Oavid A. Allen
1. Oil Storage, Transport, and Handling 169
Michael H. Milder
8. Fractionation 199
Ralph E. Timms
9. Interesterification of Oils and Fats 223
A. Rozendaal and A. R. Macrae
V
VI Contents
10. Hydrogenation of Edible Oils: Technology and Applications 265

Wicker T. Koetsier
Part III Food Emulsions

11. Butter, Margarine, Spreads, and Baking Fats 305
Eric Flack
12. Ice Cream 329
H. Douglas Goff
13. Cream Alternatives 355
Iain J. Campbell and Malcolm G. Jones
Part IV Nonaqueous Foods

14. Ghee, Vanaspati, and Special Fats in India 369
K, T. Achaya
15. Chocolate and Confectionery Fats 391
Fred B, Padley
16. Frying Oils and Salad Oils 433
Timothy L. Mounts
Part V Special Food Applications

17. Edible Coatings and Film Barriers 453
Thomas H. Shellhammer and John M. Krochta
18. Spray Processing o f Fat-Containing Foodstuffs: Spray Drying and Cooling 481
Keith Masters
19. Low Calorie Fats 501
John W, Finleyy A. McDonaldy and L. P, Klemann
20. Food Emulsifiers 521
Niels Krog
21. Lipid Emulsions for Intravenous Nutritionand Drug Delivery 535
Robert T, Lyons and Eric G. Carter
22. The Role of Lipids in Animal Feeds 557
Julian Wiseman and Phil C. Garnsworthy
Part VI Nonfood Uses

23. Anionic Detergents 579
Maurice R. Porter
24. Cationic Surfactants 609
Alan D. James
25. Nonionic Surfactants 633
Guido Bognolo
26. Lipids: Their Use in Personal Care Products 695
Keith Coupland and Julie A. Nichols
Contents VII
27. The Use o f Oils and Fatty Acids in Paints and Surface Coatings 711
John Bentley
28. Lubricants 737
Theo Mang
29. Epoxidized Oils 759
Frank D. Gunstone
30. BioFuels 771
Y. M. Choo, A. N. Ma, and A. S. H. Ong
31. Products from Castor Oil: Past, Present, and Future 787
Henri-Jean Caupin
Index 797
Contributors
K. T. Achaya CSIR Emeritus Scientist (retired), Bangalore, India

David A. Allen Technical Consultant, Partnership Solutions, Liverpool, England
John Bentley Consultant, Buckinghamshire, England
Guido Bognolo ICI Performance Chemicals, Everberg, Belgium
Iain J. Campbell Unilever Research Laboratory, Shambrook, Bedfordshire, England
Eric G. Carter Pharma Business Nutrition Department, Pharmacia & Upjohn Inc., Clayton,
North Carolina
Henri-Jean Caupin Engineering Polymers Division, Elf Atochem S-A, Paris, France
Y. M. Choo Palm Oil Research Institute of Malaysia (PORIM), Kajan, Selangor Darul
Ehsan, Malaysia
Keith Coupland Department of Research and Development, Croda Universal Ltd., Hull,
England
John W. Finley* Nabisco Foods, East Hanover, New Jersey
Eric Flackf Danisco Ingredients (UK) Limited, Bury St. Edmunds, Suffolk, England
Phil C. Garnsworthy Department of Agriculture and Horticulture, University of Nottingham,
Sutton Bonington Campus, Loughborough, England
H. Douglas Goff Department of Food Science, University of Guelph, Guelph, Ontario,
Canada
* Current affiliation: Monsanto Company, Chesterfield, Missouri
Current affiliation: Consultant, Greenewood, The Park, Great Barton, Bury St. Edmunds, Suffolk, England
IX
Contributors
Frank D. Gunstone Mylnefield Research Services, Ltd., Scottish Crop Research Institute, In-
vergowrie, Dundee, Scotland
M ichael I. Gurr Maypole Scientific Services, St. Mary’s, Isles of Scilly, England
Martin H. Hilder Manufacturing Application Group, Oil and Dairy Based Foods, Unilever
Research Laboratory, Vlaardingen, The Netherlands
Alan D. James Akzo Nobel Chemicals Ltd., Littleborough, Lancashire, England

Malcolm G. Jones Water Structuring, Foams and Emulsions, Unilever Research Laboratory,
Shambrook, Bedfordshire, England
L. P. Klemann Nabisco, Inc., East Hanover, New Jersey
W icher T. Koetsier Unichema International, Emmerich, Germany
John M . Krochta Department of Food Science and Technology, University of California-

Davis, Davis, California
Niels Krog Danisco Ingredients, Brabrand, Denmark
Robert T. Lyons Pharmaceutical Product Development, Pharmacia & Upjohn, Inc., Clayton,
North Carolina
A. N. Ma Palm Oil Research Institute of Malaysia (PORIM), Kajang, Selangor Darul Ehsan,
Malaysia
A. R. Macrae Unilever Research Laboratory, Shambrook, Bedfordshire, England
Theo M ang Fuchs Petrolub AG Oil + Chemie, Mannheim, Germany

Keith M asters NIRO A/S, Soeborg, Denmark
A. McDonald* Nabisco, Inc., East Hanover, New Jersey
Timothy L. M ounts Food Quality and Safety Research, National Center for Agricultural Uti
lization Research, Agricultural Research Service, U.S. Department of Agriculture, Peoria, Il
linois
Julie A. Nichols Product Applications Laboratory, Croda Universal Ltd., Hull, England
A. S. H. Ong Malaysian Palm Oil Promotion Council, Kuala Lumpur, Malaysia
Fred B. Padley Bedford, Bedfordshire, England, Consultant, Loders Croklaan bv., a Divi
sion of Quest International, Wormerveer, The Netherlands
Maurice R . Porter Maurice R. Porter & Associates, Sully, Vale of Glamorgan, South Wales
A. Rozendaal Fat Technology and Oil Processing Unit, Unilever Research Laboratory,
Vlaardingen, The Netherlands
Michael Schneider Department of Research & Development, Lucas Meyer GmbH & Co.,
Hamburg, Germany
Thomas H. Shellham m er Department o f Food Science and Technology, University of

Califomia-Davis, Davis, California
Ralph E. Timms Consultant, Oils and Fats, The Cottages, Swinderby, Lincoln, England
* Current affiliation: Consultant, Chicago, Illinois

Contributors XI
Maurice A. Williams Anderson International Corp., Cleveland, Ohio
Julian Wiseman Department of Agriculture and Horticulture, The University of Nottingham,

Sutton Bonington Campus, Loughborough, England
1
Fatty Acids and Lipids Structure
Frank D. Gunstone
Mylnefield Research Services Ltd., Scottish Crop Research Institute,
Invergowrie, Dundee, Scotland
L INTRODUCTION
It is almost obligatory to start any book on lipids with an account of the chemical structure of
the lipids and their component acids, and this book is not different in this respect. The follow
ing material is based on recently published fuller accounts [ 1 ,2 ], where further details may be
sought if required.
There is no generally accepted definition of the class of natural products designated as
lipids, but I adhere to the view that lipids are composed of fatty acids or closely related
components such as the corresponding alcohols and the sphingosine bases. Expressed slightly
differently, they represent those products of the acetate-malonate scheme of biosynthesis de
rived through the reductive pathway.
II. FATTY ACID NOMENCLATURE

Fatty acids are designated in several different ways. Despite the alternative descriptions set
out here, many trivial names are still widely used. These names were often given before the
chemical structure of the acid was elucidated and were frequently chosen to indicate the
source of the acid. Examples include palmitic (from palm oil), oleic (from olive oil, Olea
europea), linoleic and linolenic (from linseed oil), ricinoleic (from castor oil, Ricinus commu
nis), and arachidic acid (from groundnut oil, Arachis hypogea). A trivial name may be as
signed or may continue to be used because the systematic name is cumbersome, as for exam
ple with a-eleostearic acid, which is simpler than 9cl D13/-octadecatrienoic acid. Trivial
names are easy to use, but they are not in themselves indicative of structure.
Systematic names are based on internationally accepted rules agreed to by organic chemists
and biochemists. Those who know the rules can interconvert systematic names and structures.
As a simple example, oleic acid is cw-9-octadecenoic or Z-9-octadecenoic acid. This is a
carboxylic acid (oic) with 18 carbon atoms (octadec) and one olefinic center (en) that lies
2 Gunstone
between carbons 9 and 10 (counting from the carboxyl end) and has the cis (Z) configura
tion, i.e.,
CH3(CH2)7CH=CH (CH2)7C00H
which may also be represented by the line drawing
-COOH
Representations like this are increasingly popular. They are more useful when the number
of double bonds and/or other functional groups is larger and the saturated sections of the
molecule are short. The structure shown below is not easily recognizable as stearic acid (octa-
decanoic) because the number of carbon atoms represented in this structure is not immediately
apparent and has to be carefully counted. This difficulty is overcome in the shorter formula
tion shown to the right.
'COOH COOH
X 16
The line forms are useful because of their immediate visual impact; they are not convenient
for tabulated data or for insertion into lines of text. Because the word form can be complex
and clumsy, systematic or trivial names are sometimes abbreviated to two or three capital
letters as in the following examples:
Gamma-linolenic acid GLA

Arachidonic acid AA
Eicosapentaenoic acid EPA
Docosahexaenoic acid DHA
Another way of designating fatty acids involves the use of numbers such as 18:2. This
symbol describes an acid such as linoleic with 18 carbon atoms (assumed to be straight-
chain) and two unsaturated centers (assumed to be cw-olefinic). Since there are many isomeric
compounds that could be represented by this symbol, additional descriptors may be added
thus:
18:2(9,12), 18:2 (9c, 12c), 18:2 (9Z,12Z), 18:2 (/i-6)
All these refer to the same acid. The first indicates the position of the two unsaturated centers
in the Cjg chain with reference to COOH= 1. The second and third formulations confirm the
cis or Z configuration of the double bonds. The symbol A is sometimes added to show that
numbering is with respect to the acid function. The fourth designation introduces a further
concept in fatty acid nomenclature.
It is sometimes useful to designate double bond positions with respect to the CH 3 end
group, and this is done with symbols such as o>6 or n-6, which indicate that the first double
bond is on carbon 6, counting from the methyl group. In the absence of other information it
is assumed that all the double bonds are methylene-interrupted and have the cis (Z) configura
tion. Symbols such as c, t, and e are used to show cis, trans, and ethylenic unsaturation
(where configuration is not known or does not apply), and a (acetylenic) or y (ynoic) is used
to show a triple bond.
III. FATTY ACIDS— MAIN STRUCTURAL FEATURES

The number of known natural fatty acids exceeds 1000, though only a relatively small num
ber, perhaps 20-50, are of common concern. Based on a survey of these 1000 structures, and
noting particularly the structures of those acids produced most commonly in nature, it is
possible to make four general statements. Each of these is generally true, but there are excep
tions to all four. The exceptions are frequently trivial, but sometimes they are significant.
Though originally based on a survey of chemical structures, it is now clear that these state
ments reflect the underlying biosynthetic pathways by which the acids are produced in nature.
1. Natural fatty acids, both saturated and unsaturated, are straight-chain compounds with
an even number of carbon atoms in their molecules. This is true for the great majority of
structures and for the more abundant acids. Chain lengths range from two to over 80 carbon
atoms, though they are most commonly between Cj2 and C 22. Despite the validity of this
statement, acids with an odd number of carbon atoms (e.g., heptadecanoic, C 17) occur, as do
those with branched chains (e.g., isopalmitic, anteisononadecanoic) or with carbocyclic units
(e.g., sterculic, chaulmoogric).
2. Acids with one unsaturated center are usually olefinic compounds with cis (Z) config
uration and with the double bond in one of a limited number of preferred positions. This is
most commonly A9 (i.e., nine carbon atoms from the carboxyl group as in oleic) or n-9 (i.e.,
nine carbon atoms from the methyl group as in oleic or erucic acid). But double bonds occur
in other positions (e.g., petroselinic, 6c-18:1) or have trans configuration (e.g., elaidic 9t-
18:1) or can be replaced by an acetylenic unit (e.g., tarine 6a-18:1).
3. Polyunsaturated acids are mainly polyolefinic with a methylene-interrupted arrange
ment of double bonds having cis (Z) configuration. That is, cis double bonds are separated
from each other by one CH 2 group as in arachidonic acid:
XOOH
The 1,4 pattern of unsaturation is characteristic of natural fatty acids and differs from that
in acyclic isoprenoids, which is usually 1,3 (conjugated) or 1,5. Polyunsaturated acids occur
in biosynthetically related families. The most important are the n-6 acids based on linoleic
acid and the n-3 acids based on a-linolenic acid. In contrast to this very common pattern of
unsaturation, some acids have conjugated unsaturation, which may be cis or trans (e.g., eleo-
stearic, calendic, parinaric acids); some have mixed en/yne unsaturation, which may be con
jugated (e.g., isanic) or nonconjugated (e.g., crepenynic); and some have nonconjugated
unsaturation, which is not entirely methylene-interrupted. These are known as non-methylene-
interrupted polyenes (e.g., columbinic, pinolenic).
4. Fatty acids rarely have functional groups apart from the carboxyl group and the various
types of unsaturation already discussed. Nevertheless, acids are known that also contain a
fluoro, hydroxy, keto, or epoxy group. Two important examples are ricinoleic (12-hydroxy-
oleic) and vemolic (12,13-epoxyoleic) acid.
These generalizations have a biosynthetic basis, and even some of the exceptions are ac
commodated in the general biosynthetic scheme with only minor modifications.
Based on the annual production of commercial vegetable oilseeds it has been estimated
that eight acids account for about 97% of the total production: lauric (4%), myristic (2%),
palmitic (11%), stearic (4%), oleic (34%), linoleic (34%), a-linolenic (5%), and erucic (3%).
The level of linolenic acid would rise if all green tissue were taken into account. The major
acids in animal fats and in fish oils are myristic, palmitic, palmitoleic, stearic, oleic, eicosen-
oic, arachidonic, EPA, docosenoic, and DHA.
Gunstone
Table 1 Names and Selected Physical Properties of Some Alkanoic Acids
Acid Methyl ester
Chain Systematic Trivial mp bp mp bp Mol

length name name (°C) ccr (°C) CCY wt
4 Butanoic Butyric - 5 .3 164 — 103 8 8 .1

6 Hexanoic Caproic - 3 .2 206 -6 9 .6 151 116.2
8 Octanoic Caprylic 16.5 240 -3 6 .7 195 144.2
10 Decanoic Capric 31.6 111 - 1 2 .8 228 172.3
12 Dodecanoic Lauric 44.8 130> 5.1 262 200.3
14 Tetradecanoic Myristic 54.4 149‘ 19.1 114' 228.4
16 Hexadecanoic Palmitic 62.9 167‘ 30.7 136' 256.4
18 Octadecanoic Stearic 70.1 184^ 37.8 156' 284.5
20 Eicosanoic Arachidic 76.1 204' 46.4 188^ 312.5
22 Docosanoic Behenic 80.0 — 51.8 206^ 340.6
24 Tetracosanoic Lignoceric 84.2 — 57.4 222" 368.6
^bp at 760 mm or at 1 or 2 mm as indicated by superscript.

Source: Adapted from Ref. 1, p. 1.
IV. SATURATED ACIDS

Table 1 gives the systematic and trivial names and selected physical properties of some C4 -
C 24 acids.
A. Short- and Medium-Chain Saturated Acids (C4-C14)

Members of the C4- C 14 group of acids occur in milk fats and in the lauric group of vegetable
oils. Cow’s milk fat, for example, contains butanoic (butyric, C4 ) acid at a level of about 4%
(by weight). This may not seem very much, but because of the low molecular weight of the
C4 acid compared with the more common Cjg acids it represents about 8.5% on a molar basis,
which means that it could be present in up to 25% of milk fat glycerides. Also present are
lower levels of the C6- C 12 acids. The short-chain acids are retained in butter and in other
products made from milk fat.
Some seed oils contain very high levels of lauric acid (C 12, about 50%) with significant
levels of caprylic (Cg), capric (Cjo), and myristic (C 14) acids also. The best known of these
are coconut oil and palmkemel oil. At the present time there is interest in other sources of
these medium-chain acids. Among these are the cuphea oils, which show great promise but
still need to be fully domesticated before they produce an easily cultivated crop. A transgenic
rapeseed has also been developed that produces a lauric-rich oil.
B. Palmitic and Stearic Acid

Palmitic acid (16:0) is the most widely occurring saturated acid. It is present in fish oils (10-
30%), in milk and body fats of land animals (up to 30%), and in virtually all vegetable fats
at levels of between 5 and 50%. Useful sources of palmitic acid include cottonseed oil (15-
30%), palm oil (30-60%), Chinese vegetable tallow (60-70%), lard (20-30%), and tallow
from sheep and cattle (25-35%).
Despite the popular impression, stearic acid (18:0) is much less common than palmitic
acid. It is, however, a major component of the tallows of ruminant animals (5-40%) and a
significant component in a number of vegetable tallows (solid fats of vegetable origin) includ
Fatty Acids and Lipids Structure 5
ing cocoa butter (30-35%), Illipe or Borneo tallow (40%), and shea butter (—45%). Stearic
acid is also easily made by hydrogenation of readily available oleic, linoleic, and linolenic
acids.
Since many population groups in the world have scruples about animal products, vegetable
sources of palmitic acid and stearic acid are preferred. These are used in food and non
food products (surfactants, cosmetics, personal hygiene products) to obtain products that find
general acceptance.
C. Long-Chain Acids
Saturated acids of chain length greater than 18 carbon atoms are present at low levels in a
few seed oils and at higher levels only in a few uncommon sources. The C20- C 30 members
are often present in waxes. A convenient source of some of these acids is groundnut oil,
which contains 5-8% long-chain acids, including arachidic (20:0), behenic (22:0), and ligno-
ceric (24:0) acids. Rarer but richer sources include rambutan tallow (Nephelium lappaceum,
—35% 20:0), kusum (Schleichera trifuga, 20-30% 20:0), Lophira alata (15-30% 22:0), and
L. procera (—20% 22:0) seed fats. Long-chain acids can also be made from more readily
available shorter chain acids (Ci2-Cig) by appropriate chain-extension procedures. These reac
tion sequences are not confined to adding one or two carbon atoms at a time, as methods exist
for adding five or six carbon atoms in one cycle of reactions.
V. MONOENE ACIDS
Over 100 monoene acids have been described. These fall almost entirely in the range C iq-
C 30, with Cj 6, Cjg, and C22 members the most common. Most have cis (Z) configuration,
and the most common are either A9 or n-9, i.e., the double bond is nine carbon atoms from
the carboxyl group (A9) or the methyl group {n-9).
9-Hexadecenoic acid (palmitoleic, zoomaric) is a minor component (<1% ) of many seed
oils and animal fats. It is more significant in fish oils (up to 10%) and is a major component
of a few seed oils such as macadamia oil (around 20%).
9-Octadecenoic acid (oleic) is the most widely distributed and the most extensively pro
duced of all fatty acids. It is the prototype of all monoene acids and serves as the biological
precursor of other n-9 monoene acids and of the n-9 family of polyene acids. Olive oil (60-
80%) and almond oil (60-70%) are rich sources of this acid, and it is also present at high
levels in several nut oils (55-75%). It is a significant component in low-erucic rapeseed oil,
groundnut oil, and palm oil. High-oleic varieties of sunflower oil (80% and above) and saf
flower oil (70-75%) are now available, and the former is commonly used as a source of oleic-
rich glycerides. Oleic acid is a major component in most animal fats, where it is often accom
panied by low levels of other 18:1 isomers. This may or may not matter depending on the
end use of the acid.
Other octadecenoic acids include petroselinic (6c), elaidic {9t), and vaccenic (11c and lit).
Petroselinic acid is a major component (35-85%) of many seed oils of the families Umbellif-
erae, Araliaceae, and Garryaceae. Examples include carrot, caraway, parsnip, and parsley.
Attempts are being made to develop coriander as a new crop containing about 80% petroseli
nic acid.
Eicosenoic acid (20:1) occurs in fish oils and also accompanies erucic acid (22:1) in vegeta
ble oils rich in the C22 acid. It is present mainly as the A9 (n -ll) and/or A ll (n-9) isomers.
The most important docosenoic acid (22 : 1 ) is erucic acid (A13c, n-9), present at high
levels in seed oils of the Cruciferae. This acid has been bred out of rapeseed grown for food
use, but there is a considerable demand for erucic acid in the form of its amide. Erucic acid
6 Gunstone
is available on a commercial scale from high-erucic rapeseed oil (45-50%), mustard seed oil
(—60%), and the seed oil of Crambe abyssinica (—60%). The major docosenoic acid in fish
oils is the A ll isomer (cetoleic).
Jojoba oil (from Simmondsia chinensis) is unusual in that its major lipids are ester waxes
(esters of long-chain alcohols and long-chain acids) rather than glycerol esters. Most of the
acids and alcohols are C 20 and C22 compounds with one double bond. In meadowfoam oil
(Limnanthes alba) more than 95% of the fatty acids are C20 and C22 compounds. These are
mainly 20:1 (5c, 63%) and 22:1 (5c, 3% and 13c, 10%) along with 22:2 (5cl3c, 18%).
The major C 24 monoene acid [nervonic (A 15c, n-9)] is a significant component of many
sphingolipids. Honesty seed oil {Lunaria biennis), with C22 (43%) and C24 (25%) monoene
acids, is probably the most convenient source of nervonic acid.
VI. METHYLENE-INTERRUPTED POLYENE ACIDS

The most important polyene acids have a methylene-interrupted pattern of unsaturation with
two to six double bonds and cis (Z) configuration. They are grouped in families and are
biosynthetically related to other members of the same family. The two major groups are the
n-6 acids based on linoleic acid and the «-3 acids based on a-linolenic acid. Two minor
groups are based on oleic acid {n-9 family) and on 9-hexadecenoic acid {n-1 family). The
common and important Cjg members are present in most vegetable oils, whereas the C 16, C20,
and C22 acids are present in fish oils and in the lipids of other animals. Some of these acids
cannot be made by animals and are essential dietary requirements of vegetable origin. Struc
tures and names are given in Table 2, and some comments on the more important
n-6 and n-3 acids follow.
Linoleic acid (18:2, n-6) is the most common of all the polyene acids and serves as a
prototype for other acids in this category. It is present in all vegetable fats and is a major
component in several. Among the more common sources are soybean (45-60%), low-erucic
rapeseed (10-30%), groundnut (15-45%), rice bran (35-40%), poppy (—70%), safflower
(55-80%), sunflower (20-75%), com (maize) (35-60%), cotton (35-60%), sesame (—45%),
and tall oil fatty acids (40-50%).
y-Linolenic acid (18:3, n-6, GLA) is not a common acid, but recent interest in its use as
a dietary supplement has directed attention to its three common sources: evening primrose oil
(8-12% ), borage oil (20-25%), and blackcurrant seed oil (15-17%).
Table 2 Methylene-Interrupted Polyene Acids
Trivial name Structure Double bond position^
Linoleic 18:2 {n-6) 9,12

y-Linolenic 18:3 {n-6) 6,9,12
a-Linolenic 18:3 (n-3) 9,12,15
Stearidonic 18:4 (n-3) 6,9,12,15
Dihomo-y-linolenic 20:3 {n-6) 8,11,14
Mead’s acid 20:3 {n-9) 5,8,11
Arachidonic 20:4 (n-6 ) 5,8,11,14
Eicosapentaenoic (EPA) 20:5 (n-3) 5,8,11,14,17
Docosapentaenoic (DPA) 22:5 (n-3) 7,10,13,16,19
Docosahexaenoic (DHA) 22:6 (n-3) 4,7,10,13,16,19
®A11 double bonds have cis configuration.

Arachidonic acid (20:4, n-6, AA) is of considerable importance as the precursor of many
important C20 metabolites such as the prostaglandins, thromboxanes, and leukotrienes (mem
bers of the eicosanoid cascade). This acid is a minor component of many fish oils and attains
higher levels in animal phospholipids such as those from egg and liver. Though rare in the
plant kingdom, it has been identified in mosses, ferns, and some algae and fungi.
a-Linolenic acid (18:3, n-3) is an important component of lipids in leaves, stems, and
roots. It is a minor but significant component of soybean oil (4-11%) and of rapeseed oil (5 -
15%) and a major component of linseed oil (50-60%) and perilla oil (60-70%). Eicosapentae-
noic acid (20:5, EPA) and docosahexaenoic acid (22:6, DHA) are also important n-3 acids.
The former is a precursor of a series of eicosanoids, and the latter is present at high levels in
the lipids of sperm and the retina. The major sources of these two acids are the fish oils, up
to 30% of which may be the two acids combined. One or both are also produced by some
microorganisms.
Mead’s acid (20:3, n-9) is of interest in that it becomes significant in animal lipids only
during periods of essential fatty acid deficiency.
VII. OTHER POLYENE, ACETYLENIC, AND ALLENIC ACIDS

The best known conjugated polyene acids have 18 carbon atoms and three or four double
bonds per molecule. Two A9,11,13,15-tetraenes are derived biologically from a-linolenic acid
(A9,12,15), and the A 9,ll,13- and the A8,10,12-trienes are derived from linoleic acid
(A9,12). These are listed in Table 3 with structures, trivial names, and major sources.
Linoleic acid furnishes a mixture of conjugated dienes (9,11 and 10,12 isomers) in a num
ber of enzymatic and nonenzymatic processes. This mixture is reported to display anticarcino-
genic properties, with the 9 c llt acid being the active isomer. The mechanism of this activity
is not known, and the suggestion that it acts as an antioxidant has been questioned. Conju
gated linoleic acid (CLA) is present at low levels in many dietary sources. The highest levels
Table 3 Natural Octadecatrienoic and Octadecatetraenoic Acids with

Conjugated Unsaturation
Typical source
Configuration Trivial name (seed oils)
Trienes (18:3)
8 c 1Ori 2c Jacaric Jacaranda mimosifolia
8 rl 0 rl 2 c Calendic Calendula officinalis
8 rl 0 rl 2 r — C. officinalis
9cllrl3c Catalpic Catalpa ovata
9cllrl3r a-Eleostearic Aleurites fordii^
9rllil3r j3-Eleostearic A. fordii^
9rllrl3c Punicic Punica granatum
Tetraenes (18:4)
9cllrl3rl5c a-Parinaric*^ Impatiens balsamina
9rllrl3rl5r j8 -Parinaric —
Âlso kamlolenic, 18-hydroxy-a-eleostearic (Mallotus philippinensis).

‘’A lso licanic, 4-oxo-a-eleostearic {Licania rigida).
‘Âlso chrysobalanic, 4-oxo-a-parinaric (Chrysobalanus icaco).
‘‘Tung oil.
8 Gunstone
appear in ruminant and dairy fats. For example, lamb fat contains 5.6 mg of CLA per gram
of fat, over 90% of it the 9 c llt isomer.
Some polyene acids have incomplete methylene-interrupted unsaturation. These are found
in some seed oils, some mycobacteria, and some marine lipids, especially sponges. The addi
tional double bond is frequently A3 or A5. Examples include a diene acid (20:2 5c 13c) in
meadowfoam oil, pinolenic acid (18:3 5c9cl2c) in tall oil, columbinic acid (18:3 5/9cl2c) in
aquilegia (columbine) oil, and long-chain trienes such as 26:3 (5,9,19), 28:3 (5,9,19), and
30:3 (5,9,23) in sponge lipids (demospongic acids).
Acetylenic acids occur only rarely in seed oils and in mosses. Structures have been identi
fied that contain acetylenic unsaturation only (e.g., tariric 18:1 6a); mixed olefinic and acety
lenic unsaturation, which may be conjugated (ximenynic or santalbic, 18:2 9 c lli) or noncon-
jugated (crepenynic, 18:2 9c 12a); and yet others that also contain a hydroxyl group
(helenynolic, 9-OH 18:2 1Or12c).
A few allenic acids are also known and are probably produced by rearrangement of acetyle
nic precursors. The allenic acids are optically active by virtue of the chiral allenic group and
include laballenic (18:2 5c6c) and lamenallenic (18:3 5c6cl6r) acids.
VIII. ACIDS WITH TRANS OLEFINIC UNSATURATION

Polyene acids having trans olefinic groups in conjugated and non-methylene-interrupted sys
tems have already been discussed. Natural acids with trans unsaturation are very uncommon.
Nevertheless they are a significant component (about 5%) of human dietary intake. They arise
from two major sources: (1) from those cooking and spreading fats that have been produced
by partial hydrogenation with a heterogeneous catalyst and (2) from dairy and other ruminant
fats. These last result froni biohydrogenation of linoleic and linolenic acid in the animal’s diet.
For example, butterfat contains ten 18:1 trans acids of which about 75% is iraa^-vaccenic acid
(A ll/). The trans acids in human milk fat (2-4%) probably arise from these dietary sources.
Dietary trans acids are also present in some refined oils exposed to high temperatures (230-
250°C) during the refining process. The major acids of this type (present, e.g., in refined
soybean and rapeseed oil) are the 9c 12c 15/ and 9/12c 15c isomers from linolenic acid. Linoleic
acid undergoes a similar reaction less readily to give the 9c 12/ and 9/12c diene acids.
IX. BRANCHED-CHAIN ACIDS

There are many branched-chain acids, though these seldom occur at a significant level. Exam
ples include the iso acids (mainly with an even number of carbon atoms), the anteiso acids
(mainly with an uneven number of carbon atoms) found in wool wax and other animal fats,
and the phytol-based acids present at low levels in fish oils.
CH^ CH^
CH3CH(CH2),,C00H CH3CH2CH(CH2)„C00H
iso acids anteiso acids
COOH phytanic acid (G 2 0 )
COOH pristanic acid (C 19 )

X. CYCLIC ACIDS
The most common cyclic acids contain a cyclopropane, cyclopropene, or cyclopentene unit.
Cyclopropane acids occur in bacterial membrane phospholipids and are mainly Cj 7 or C 19
(lactobacillic) acids. The cyclopropane unit, like a cis double bond, introduces a discontinuity
in the molecule and increases fluidity in the membrane.
CH 2
C H 3(C H 2)5C H C H (C H 2)„C 00H

w = 7 or 9
The best known cyclopropene acids are malvalic and sterculic, which are present at high
levels in sterculia oils and at lower levels in kapok seed oil (—12%) and in cottonseed oil
(—1%). These acids are highly reactive and are destroyed during refining and hydrogenation
of the oils. They have attracted interest because they inhibit the biodesaturation of stearic to
oleic acid. 2-Hydroxysterculic acid has also been identified. This compound is probably an
intermediate in the bioconversion of sterculic to malvalic acid by a-oxidation.
CH 9
CH3(CH2)7C=C(CH2)„C00H
malvalic acid n = 6
sterculic acid n = l
Seed fats of the Flacourtiaceae are unique in containing several cyclopentene acids. They
range from to C20 acids and may also be unsaturated in the side chain. These acids are
optically active by virtue of the chiral center in the cyclopentene unit.
C ^ C H 2 ) „ COOH (n = 0 A 4 )
XI. OXYGENATED ACIDS

The most common of the oxygenated acids contain hydroxy, epoxy, or furanoid units. Other
systems not included here include keto (0x0) and methoxy groups.
A. Hydroxy Acids
The best known natural hydroxy acid is ricinoleic ( 1 2-hydroxy oleic) acid. This is the major
acid (—90%) in castor oil (Ricinus communis seed oil). An isomer (9-OH 18:1 12c) occurs in
Strophanthus and Wightia seed oils. Other acids of this type include densipolic, lesquerolic,
and auricolic (Table 4). Lesquerolic acid is the C20 homologue of ricinoleic acid, and attempts
are being made to develop Lesquerella fendleri as a new crop.
Castor oil and ricinoleic acid are important materials used in cosmetics, in lubricants both
before and after hydrogenation, and as a drying oil after dehydration (dehydrated castor oil,
DCO). Also important are the products of pyrolysis (heptanal and 10-undecenoic acid and
materials derived from these) and of alkali fusion (2 -octanol, 2 -octanone, 10-hydroxydecanoic
acid, and the C jq dibasic acid, sebacic acid). Of less importance are some hydroxy acids based
on 8-, 9-, or 13-hydroxystearic acid with conjugated unsaturation. Some typical examples are
isanolic (8-OH 18:3 9 a \la \le ) , helenynolic (9-OH 18:2 10fl2a), dimorphecolic (9-OH 18:2
10H2c), and coriolic (13-OH 18:2 9 c ll0 -
10 Gunstone
Table 4 Mid-Chain Hydroxy Acids Without Conjugated Unsaturation
Chain length and Position of Source

unsaturation hydroxyl group Trivial name (seed oils)
16:1 9c 12 —
Lesquerella densipila
18:1 9c 12R Ricinoleic Ricinus communis
18:1 1 2 c 9S Isoricinoleic Strophanthus and Wightia spp
18:2 9cl5c 12R Densipolic L. densipila
2 0 :1 1 1 c UR Lesquerolic L. densipila
20:2 llc l7 c 14 Auricolic L. auriculata
B. Epoxy Acids
Several natural epoxy acids are known. Vemolic {cis-12,13-epoxyoleic) was first discovered
in 1954 and is still the best known in this group. It occurs at high levels in the seed oils of
Vernonia anthelmintica (70-75%), V. galamensis (73-78%), Cephalocroton cordofanus (60-
65%), Euphorbia lagascae (60-65%), Stokes aster (65-80%), and Erlanga tomentosa (50-
55%), and attempts are being made to develop V. galamensis and E. lagascae as commer
cial crops.
C. Furanoid Acids
Natural furanoid acids with the structures shown have been identified. They are present at low
levels, often as a complete series, in fish oils. The proportion of these increases significantly
during prolonged fasting. The short-chain urofuranic acids that have been identified in animal
blood and urine are probably breakdown products of furanoid acids believed to come from
dietary vegetable sources.
HOOC COOH
CH3(CH2)m " (CH2)„C00H

A K
R (CH2)2C00H
R = CH3 or H R = CH3(CH 2)4 or CH3(CH 2)2
/w = 2 or 4 n = 6,8, or 10 or CH3CH2CHOH or CH3CH 2CO
furanoid acids urofuranic acids
XII. LIPID STRUCTURE

The fatty acids discussed in the previous section seldom occur in their free state but usually
as esters and sometimes as amides (Table 5). Esters contain acid and alcohol components,
and the most common alcohol present in lipids is glycerol (propane-1,2,3-triol). Important
glycerolipids include the acylglycerols, the phospholipids, and the glycosyIglycerides. Other
alcohols that occur less commonly in lipids in an acylated form include carbohydrates (sugar
esters), sterols (sterol esters), and long-chain alcohols (wax esters). The amides are derivatives
of long-chain acids with amines such as the sphingosine bases or some amino acids.
Before detailing these it is useful to explain some terms that occur frequently. Simple or
neutral lipids are triacylglycerols, some ether lipids, sterol esters, and wax esters. Complex
Table 5 Types of Lipids
Esters
Of glycerol Acylglycerols (glycerides)
Phosphoglycerides (PA, PC, PE, PS, PI, PG)^
Glycosylglycerides (MGDG, DGDG) ^
Of other alcohols
diols
sugars Sugar esters
long-chain alcohols^ Wax esters
sterols Sterol esters
Amides
Of long-chain bases (sphingosine, etc) Ceramides, cerebrosides, gangliosides
Of taurine
Of serine
^PA, phosphatidic acids; PC, phosphatidylcholines; PE, phosphatidylethanolamines; PS, phosphatidyIserines;

PI, phosphatidylinositols; PG, phosphatidylglycerols; MGDG, monogalactosyIdiacylglycerols; DGDG, diga-
lactosyldiacylglycerols.
*Âlso form ether lipids.
or polar lipids include phosphoglycerides, glycosy Idiacylgly cerols, and sphingolipids. Phos
pholipids are derivatives of phosphoric (or less commonly phosphonic) acid, glycolipids con
tain one or more carbohydrate units, and sulfolipids contain sulfur.
Glycerol contains a prochiral carbon atom carrying two CH2OH groups. When these differ,
as when they are acylated with different fatty acids, then the molecule is chiral and exists in
two enantiomeric forms. To designate the stereochemistry of glycerol-containing components,
the carbon atoms are stereospecifically numbered (sn). When the glycerol molecule is repre
sented by a Fischer projection with the secondary hydroxyl to the left of the central (pro
chiral) carbon atom, the carbon atoms are numbered 1, 2, and 3 from top to bottom. Mole
cules that are stereospecifically numbered in this fashion have the prefix sn- immediately
preceding the term glycerol. The prefix rac- in front of the full name shows that the
compound is an equal mixture of both enantiomers. When x- is used, the configuration is
unknown or un- specified.
CH20 H sn -1 a
HO- —— H sn-2 p
CHÔH sn-3 OL or
Any glycerol lipid will be chiral when the substituents at the sn-l and sn-3 positions are
different. Mirror image molecules or enantiomers possess opposite but equal rotations. How
ever, if both substituents are long-chain acyl groups, then the optical rotations are very small
and may be difficult to observe and measure.
XIII. ACYLGLYCEROLS
The major reserve lipids are triacylglycerols (formerly known as triglycerides). Mono- and
diacylglycerols may also be present as minor components. However, these are important inter
12 Gunstone
mediates in the biosynthesis and catabolism of triacylglycerols and other classes of lipids.
They are also prepared on a large scale for use as surface-active agents.
Monoacylglycerols (monoglycerides) are fatty acid monoesters of glycerol and exist in two
isomeric forms, the a- and j8-monoglycerides, whose structures are illustrated.
CH2OCOR CH2OH CH.OH
HO- -H HO- "H RCOO- -H
CH 2OH CH2OCOR CH 2 OH
1" and 3- monoacyl-^/i-glycerols 2-monoacyl-5/i-glycerol

a-m onoglyceride P-monoglyceride
Pure isomers readily change to a 90:10 mixture of the a and isomers. This rearrangement
is promoted by acid or alkali.
Diacylglycerols (diglycerides) are fatty acid diesters of glycerol, and they also exist in
two isomeric forms, as illustrated. They readily form an equilibrium mixture, with the 1,3-
diacylglycerol being the more stable.
CH2OCOR CH2OH CH2OCOR
R'COO "H R'COO -H HO- ■— H
CH 2OH CH2OCOR CH2OCOR'
1,2- and 2,3-diacyl-^«-glycerols 1,3-diacyl-5A2-glycerol

aP-diglycerides aa'-m onoglyceride
Triacylglycerols (triglycerides) are fully acylated derivatives of glycerol:
CH20C0R^
R^COO• ■H
CH2OCOR
It is unusual for a natural triacylglycerol to have only one acid, though this does happen,
e.g., triolein in olive oil and tripalmitin in palm oil. More usually, two or three different acids
are present. The number of possible triacylglycerols rises very quickly with the number of
Table 6 Triacyl Glycerols Containing Only Palmitic (P) and Oleic (O) Acids
ppp
POP
r oppi^
LPPoJ
OPO
r pooi^
LOOPJ
000
Ûnsymmetrical triacylglycerols in enantiomeric pairs.
Table 7 Relation Between Number of Acids and Maximum Number of Triacylglycerols That
Can Be Formed from Them
Number of acids
5 10 20 n
Triacylglycerols
All isomers distinguished 125 1000 8000
Excluding optical isomers 75 400 4200 {n^ + n^)i2
No isomers distinguished 35 220 1540 {rP + 3n'^ + 2n)l6
acids present in the fatty acid pool. With only two acids there are eight triacylglycerols (in
cluding enantiomers, Table 6). With three acids this number is 27, and it rises very rapidly
thereafter (Table 7). The triacylglycerol composition of some natural oils is discussed in
Chapter 2.
On hydrolysis with aqueous acid or alkali, the triacylglycerols are split into glycerol and a
mixture of fatty acids. Enzymatic hydrolysis usually occurs with some degree of regiospecifi-
city. Most commonly this involves deacylation of the a positions (sn-l and sn-3), with no
reaction occurring at the position, so the products are fatty acids from the a positions and
2-monoacylglycerols :
-1 "1 “OH OH
2- 2 1 20—
“T 2
L3 ÔH L3 L
‘■OH
(1, 2, and 3 represent the mixed fatty acids present at each of these positions.)
XIV. GLYCOSYLDIACYLGLYCEROLS
The glycosyldiacylglycerols form a class of glycolipids that are present in plants (especially
in the chloroplast) and in bacteria. Diacylglycerols are linked to one or more sugar units
(especially galactose) at the sn-3 position. They are usually rich in polyunsaturated Cjg acids.
Typical examples are the monogalactosyldiacylglycerols (MGDG) and the digalactosyldiacyl-
glycerols (DGDG).
sugar unit
galactose MGDG
Ô sugar galactose-galactose DGDG
XV. PHOSPHOGLYCERIDES
There are several types of phosphoglycerides (Table 8). These are based on the phosphatidic
acids, which are themselves diacyl derivatives of 3-glycerophosphoric acid. The phosphatidic
acids are monoesters of the tribasic phosphoric acid. The remaining phospholipids are diesters;
i.e., a second hydroxy compound is associated with the molecule (Table 8).
14 Gunstone
CH2OH CH20C0R^
I I
HOCH O R^COOCH O
I II
CH2OPOH
I II
CH2OPOX
OH
I
OH
3-glycerophosphoric acid phosphatidic acids (PA, X = H)
The composition of some natural phospholipids is discussed in Chapter 2. Many phospho

lipids contain saturated acids at the sn-l position and an unsaturated fatty acid at the sn-2
position, but this is not an invariable rule. The phosphoglycerides, with two lipophilic chains
and a hydrophobic head group, are important structural lipids. They are essential parts of lipid
bilayers that make up cell membranes.
Each phosphoglyceride has four ester groups, and their hydrolysis is illustrated in Table 9
for phosphatidylcholines. Complete hydrolysis occurs with aqueous acid, and the products are
glycerol, fatty acids, phosphoric acid, and choline. With alkali the acyl groups are easily
hydrolyzed, but the phosphate ester bonds react more slowly so that hydrolysis furnishes fatty
acids and glycerophosphorylcholine (GPC), which is then slowly hydrolyzed to choline and
glycerophosphoric acid. This last is a mixture of a and isomers because of the migration of
phosphoric acid among the hydroxyl groups.
Enzymatic hydrolysis is more selective and gives rise to a range of interesting products.
Phospholipase A1 causes deacylation only at the sn-l position, liberating the fatty acids from
this position and leaving a lysophosphatidylcholine. Phospholipase A2 behaves similarly at
the sn-2 position. Phospholipase B is now recognized as a mixture of the A1 and A2 enzymes
and removes fatty acids from both positions, leaving glycerophosphorylcholine. Phospholipase
C reacts at one of the phosphate ester bonds to give 1,2-diacylglycerols and phosphorylcho-
line, whereas phospholipase D cleaves the other phosphate ester bond to give choline and
phosphatidic acids (Table 9).
These stereospecific processes have been exploited in a number of synthetic procedures.
Starting with natural phosphatidylcholines, for example, it is possible to remove the fatty
acids from the sn-l and/or sn-2 positions and replace them with an acid of choice, thereby
retaining the stereochemistry of the natural product. Also, the phosphatidylcholines can be
converted to phosphatidic acids by reaction with phospholipase D and then be esterified as
required to produce other phosphoglycerides.
Table 8 The Major Glycerophospholipids
Name Abbreviation X^
Phosphatidylcholines PC CH2CH2NM^^
Phosphatidylethanolamines ^ PE CH2CH2NH3
Phosphatidylserines PS CH2CH(NH3)C0 0 -
Phosphatidylinositols PI C,H „05
Phosphatidylglycerols PG CH2CH(0 H)CH2 0 H
^See the structure o f the phosphatidic acids (X = H) in the text.

‘’Also occur as esters of phosphonic acid (H3PO3).
Table 9 Enzymatic Hydrolysis of Phosphatidylcholines ^
Phospholipase Lipolysis products’’ Source
Al Fatty acids (sn-l), lyso PC Snake venom

A2 Fatty acids (sn-2), lyso PC Snake venom
B Fatty acids (sn-l and 2), GPC —
C DAG, phosphorylcholine B acteria (e.g ., Clostridia)

D PA, choline Most plant tissues
^These enzymes are also effective with phosphatidylethanolamines.

’’PC, phosphatidylcholines; GPC, glycerophosphorylcholines; DAG, diacylglycerols; PA, phospha-
tidic acids.
A1
CH2OCOR ‘
RÔCOCH o
■ CH
' OPOCH
II 2 2 CH2 NMe 3
A2
/l\ D
XVI. ETHER LIPIDS

Four types of monoether lipids have been recognized depending on whether they are related
to triacylglycerols or phosphoglycerides and on whether the ether group does or does not
contain a Al double bond (trans). The ether group is generally in the sn-l position, but some
diethers (sn-l and sn-2) are also known.
- or‘ o c h - chr' ■o r ' OCH=CHR
I
0
R^COO - R^COO - R^COO - R^'COO -
II
-ocor’ ÔCOR ^ '■OPOZ^ OPOZ'^
0“ O”
3 4
Compounds of type 1 occur in some marine oils. When they are hydrolyzed they yield
fatty acids (2 mol) and glycerol ethers. The alkyl chain is commonly 16:0, 18:0, or 18:1, and
the glycerol ethers (which are diols) derived from these are known as chimyl alcohol (16:0),
batyl alcohol (18:0), and selachyl alcohol (9c-18:l). The vinyl ether group present in com
pounds of types 2 and 4 resists reaction with alkali and with lithium aluminum hydride but is
hydrolyzed under acidic conditions to give an aldehyde.
2 -» glycerol + fatty acids + [RiCH=CHOH] RICH 2CHO
16 Gynstone
Ethanolamine derivatives of type 4 occur in animal tissue and were originally called plasmalo-
gens to reflect the production of aldehydes. Their biochemical significance is not understood.
Compounds designated platelet-activating factor (PAF) have biological activity (aggrega
tion, inflammatory, edemic) at very low concentrations and are specifically bound to receptors
in platelets. They are 1-0-alkyl-2-acetyl-5'^-glycerol-3-phosphocholines (5).
•0(C H 2); i CH s
O
CH3 COO ”
■f
OPOCH 2 CH 2 NMC 3
o
XVII. ACYL DERIVATIVES OF OTHER ALCOHOLS
All the lipids discussed so far have been esters of glycerol. There remain to be detailed some
esters of fatty acids with other alcohols. Ester waxes are derivatives of long-chain acids and
long-chain alcohols such as occur in fish (sperm whale oil, orange roughy oil), animals (bees
wax, wool wax), and plants (camauba, jojoba). Esters (and ethers) based on ethanediol and
propane-1,3-diol have also been identified. The latter may be widely distributed at low levels
but are difficult to detect in the presence of glycerol esters. Sterols, sugars, and other natural
alcohols also occur as esters with long-chain alcohols. This acylation leads to a reduction in
water solubility and an increase in fat solubility.
Cutins and suberins are plant polymers produced from mono-, di-, and trihydroxy and
Cjg acids.
XVIII. SPHINGOLIPIDS
The component units of sphingolipids are fatty acids, often including 2-hydroxy acids, bound
as amides to long-chain amines. These amines also have two or three hydroxyl groups, one
of which is linked to one or more sugar units or to a phosphoester unit thus:
RCH(OH) CH(NH2) CUtOH
acylated with a sugar or

fatty acid phosphoester unit
The long-chain bases (sphingoids) are mainly Cjg or C 20 compounds, and the structures of
two such bases are illustrated here.
i(E) (R) (S)

sphingosine CH3(CH2)i 2CH=CHCH(0H)CH(NH2)CH20H
(4 / -sphingenine)
(R) (S) (S)
phytosphingosine CH3(CH2)i 3CH(0H )CH (0H )CH(N H 2)CH 20H
Their A^-acyl derivatives are known generally as ceramides. The structures and occurrence
of typical glycosphingolipids (gangliosides, cerebrosides) and a sphingomyelin are given in
Table 10.
Table 10 Structure and Occurrence of Some Sphingolipids

r CH=CHCH(0H)CHCH20R2
I
NHR’
Name R' Occurrence
Sphingenine H H —
Ceramides COR' H Skin

Cerebrosides COR' glu or gal Brain, spleen, ganglion cells
Gangliosides COR' glu-gal-sialic acid Myelin sheath of brain and nerve
Sphingomyelin COR' PO 3 CH2CH2N MC3 Animal membranes
There are several diseases in which sphingolipids accumulate in various organs and tissues
through enzymatic defects in sphingolipid metabolism. These are known generally as sphingo-
lipidoses and include diseases associated with the names Tay-Sachs, Fabry, Gaucher, Färber,
Niemann-Pick, and Krabbe.
Though structurally different, the phosphoglycerides and sphingolipids resemble each other
in that both contain two long chains and a polar head group and are conveniently located in
lipid bilayers.
REFERENCES
1. F. D. Gunstone, J. L. Harwood, and F. B. Padley (Eds.), Lipid Handbook, 2nd ed., Chapman and
Hall, London, 1994.
2. F. D. Gunstone, Fatty Acid and Lipid Chemistry, Blackie, London, 1996.
Major Sources of Lipids
Frank D. Gunstone
MylnefieldResearch Services Ltd., Scottish Crop Research Institute,
I. INTRODUCTION
This chapter is devoted to the major and some minor sources of lipids. The publications of
Mielke GmbH of Hamburg {Oil World, etc.) provide figures for the production, imports and
exports, and stocks of 17 major oil and fat sources [1]. These include three animal fats
(butterfat, lard, and tallow/grease), three industrial oils (castor, fish, and linseed), and 11
vegetable oils (coconut, com, cotton, groundnut, olive, palm, palmkemel, rapeseed, sesame,
soybean, and sunflower). These are discussed here along with some other significant sources
such as rice bran oil, tall oil fatty acids, and others. Sections II-X I are concerned with
general matters and Section XII contains an account of individual sources.
It is appropriate to note that among vegetable oils, which together make up about 75% of
the annual world production, some such as soy, cotton, com, and rice bran oils are by
products, and the supply of such oils is controlled by the demand for the more important
component, protein meal for animal feed from soybeans, fiber from cotton, or cereal from
com and rice. A second group, such as coconut, palm, palmkemel, and olive oils are pro
duced from tree crops. The trees have to be planted and come to maturity, after which they
generally produce annual crops for 20 years or more. Finally, some of the oils (groundnut,
rapeseed, sesame, and sunflower) come from annual crops the availability of which can vary
from year to year as they are planted annually according to their perceived profitability.
The major emphasis in this chapter is on fatty acids and triacylglycerols because these are
the major components in each lipid source, but reference is also made, where relevant, to
phospholipids, tocopherols, and other minor components.
IL NAMES
Table 1 lists the English, French, and botanical names of many lipid sources of vegetable
origin. The original reference [2] contains some additional entries and also gives Russian
names.
19
20 Gunstone
Table 1 Botanical Name of Plant and Name of OiP
Oil
Plant English French
Aleurites fordii Tung nut oil Huile de tung

Aleurites moluccana, Candlenut oil, artist’s oil, Huile de bancoul
A. triloba lumbang oil
Arachis hypogea Arachis seed oil, groundnut oil, Huile d’arachide
peanut oil
Attalea speciosa, Orbignya Babassu oil Huile de babassu
speciosa
Borago ofßcinalis Borage seed oil Huile de bourâche
Brassica juncea Brown mustardseed oil Huile de moutarde brune
Indian mustardseed oil Huile de moutarde d’Inde
Brassica napus Colza oil, rapeseed oil Huile de colzo
Brassica nigra Black mustard seed oil Huile de moutarde noire
Brassica rapa Turnip seed oil Huile de navette
Butyrospermum paradonum. Shea butter, shea nut oil Amande de karité
B. parkii
Camelina sativa Cameline seed oil, camelina seed Huile de cameline
oil, gold of pleasure
Camelina sinensis Teaseed oil Huile de thé
Cannabis sativa Cannabis oil, hempseed oil Huile de chenevis
Carthamus tinctorus Safflower oil, safflower seed oil Huile de carthame
Ceiba pentandra Kapok seed oil Huile de kapok
Cocos nucifera Coconut oil, copra oil Huile de coprah
Huile de coco
Corylus aveliana Hazelnut oil Huile de noisette
Crambe abyssinica Sea kale oil, crambe oil Huile de crambe
Cucurbita maxima Pumpkin seed oil, squash seed oil. Huile de cucurbitacéas
vegetable marrow seed oil (citrouille, courge, patiron)
Cuphea spp. Cuphea seed oil Huile de cuphéa
Elaies guinensis Palm oil Huile de palme
Palmkemel oil Huile de palmiste
Euphorbia lathyris Spurge seed oil Huile d’épurge
Eagus sylvatica Beechnut oil Huile de faîne
Glycine max Soyabean oil, soybean oil Huile de soja
Gossypium spp. Cottonseed oil Huile de coton
Guizotia abyssinica Nigerseed oil Huile de niger
Helianthus annus Sunflower seed oil, sunflower oil Huile de tournesol
Hevea brasiliensis Rubberseed oil Huile d’hévéa
Hibiscus esculentus Okra seed oil Huile de gombo
Jatropha curcas Jatropha seed oil, physicnut oil Huile de jatropha
(pignon d’Inde, pulghère,
purghère)
Juglans regia Walnut oil Huile de noix
Licania rigida Oiticica oil Huile d’oiticica
Limnanthes alba Meadowfoam seed oil —
Linum usitatissimum Linseed oil Huile de lin

Madhuca longifolia, Bassia Indian illipe butter, mowrah butter Beurre d’illipé
longifolia
Oenothera biennis, 0 . la- Evening primrose oil Huile d’onagre
markiana
Major Sources of Lipids 21
Table 1 Continued
Oil
Plant English French
Olea europeae Olive oil Huile d’olive

Orbignya oleífera Babassu nut oil Huile de Babasso
Oryza sativa Rice bran oil Huile de riz
Papaver somniferum Poppyseed oil Huile d’oeillette
Perilla frutescens Perilla seed oil Huile de périlla
Persea americana Avocado pear oil, avocado kernel Huile d’avocat
oil
Prunus armaniaca, Apricot kernel oil Huile de noyaux d’apricôt
Armeniaca vulgaris
Prunus domestica Plum kernel oil Huile de noyaux pmne
Prunus dulcís, Prunus Almond oil Huile d’amande
amygdalus
Prunus persica Peach kernel oil Huile de noyaux de pêche
Ribes nigrum Blackcurrant pip oil Huile de pépins de cassis
Ricinus communis Castor oil, castor seed oil Huile de ricin
Sapium sebiferum Chinese vegetable tallow, stillingia Huile de stillingia
oil
Sesamum indicum Sesame seed oil, sesame oil Huile de sésame
Shorea macrophylla, Borneo oil, Borneo tallow, illipe Suif de Bornéo, beurre
Shorea stenoptera butter d’illipé de Bornéo
Shorea robusta Sal fat Graisse de sal
Simmondsia chinensis Jojoba seed oil Huile de jojoba
Sinapsis alba White mustard oil, white mus- Huile de moutarde blanche
tardseed oil
Theobroma cacao Cocoa butter, theobroma oil Beurre de cacao
Triticum aestivum Wheatgerm oil Huile de germe de blé
Vernonia spp. Vernonia seed oil Huile de vemonia
Vitellaria paradoxa Shea butter Beurre de karité
Vitis vinifera Grapeseed oil Huile de pépins de raisin
Zea mays Maize oil, com oil Huile de germe de mais
®The original publication also gives the name of each oil in Russian.
Source: Ref. 2.
III. ANNUAL PRODUCTION AND FORECASTS

Table 2 co n tain s production data and forecasts for the period 1986-2012 for 17 major oils
and fats obtained from a recent Mielke publication [3]. These are average annual figures for
the five-year period 1986-1990 and predictions for the quinquennia 1993-1997 and 2008-
2012. Over these three time periods annual production is expected to rise from 76 to 91 to 132
million tonnes. Five of these oils—palm, palmkemel, soybean, rapeseed, and sunflower—
predominate and together represent 53,59, and 65% of the total supply in the three time
periods. Although the others rise in annual production, all (except com oil) are expected to
decline in terms of market share. Palm oil, which has long held second place to soybean oil,
is predicted to become the dominant oil before the end of the first decade in the next century,
and the two products of the oil palm will together represent one-fourth of the annual produc-
22 Gunstone
Table 2 Production and Forecasts of Average Annual Production

(million tonnes) for 17 Major Oils during Three Quinquennia Between
1986 and 2012^
1986-1990 1993-1997 2008-2012
Palm 9.1(12.0) 15.3(16.9) 29.8(22.6)

Soybean 15.3(20.2) 17.9(19.8) 25.1(18.9)
Rape 7.5 (9.9) 10.1(11.2) 15.6(11.8)
Sunflower 7.3 (9.7) 8.4 (9.3) 12.0 (9.1)
Tallow/grease 6.6 (8.7) 7.0 (7.7) 8.1 (6.1)
Lard 5.4 (7.1) 5.7 (6.3) 7.7 (5.8)
Butterfat 6.4 (8.5) 6.0 (6.6) 6.9 (5.2)
Cottonseed 3.6 (4.8) 4.2 (4.6) 5.9 (4.5)
Groundnut 3.7 (4.9) 4.2 (4.6) 5.7 (4.3)
Palmkemel 1.2 (1.6) 2.0 (2.2) 3.5 (2.6)
Coconut 3.0 (4.0) 3.0 (3.3) 3.3 (2.5)
Com 1.3 (1.7) 1.7 (1.9) 2.7 (2.0)
Olive 1.9 (2.5) 2.0 (2.2) 2.1 (1.6)
Fish 1.5 (2.0) 1.2 (1.3) 1.3 (1.0)
Sesame 0.66(0.9) 0.71(0.8) 0.94(0.7)
Linseed 0.72(1.0) 0.64(0.7) 0.88(0.7)
Castor 0.38(0.5) 0.45(0.5) 0.55(0.4)
Total 75.6 90.5 132.1
^The figures in parentheses represent percentage of the total.

Source: Ref. 3.
tion in the 2008-2012 quinquennium. Most of this will be produced in two countries in
southeast Asia—Indonesia and Malaysia.
IV. GEOGRAPHICAL AREAS OF PRODUCTION

Information summarized in Table 3 shows the major geographical sources of soybeans, rape-
seed, sunflower seed, cottonseed, groundnuts, palm oil, and copra (the source of coconut oil)
Table 3 Major Geographical Areas of Production of Seven Vegetable Oils/Fats Based on Average
Annual Figures for 1989-1992
Total tonnes ^
xlO^ Major geographical sources (%)
Copra 4.9 Philippines (41), Indonesia (25), India (8)

Cottonseed 33.5 China (24), USA (16), FSU (13), India (13), Pakistan (10),
Brazil (4)
Groundnut 16.4 India (34), China (25), USA (9)
Palm oil 11.6 Malaysia (55), Indonesia (23), Nigeria (5)
Rapeseed 25.2 China (26), India (20), Canada (14), EU (23), Poland (5)
Soybeans 105.9 USA (50), Brazil (17), Argentina (11), China (10)
Sunflower seed 22.3 FSU (29), Argentina (17), EU (18), China (6), USA (5),
Eastern Europe (10)
^These are not figures for oil (except for palm) but for beans, seeds, etc.
Source: Ref. 4.
Table 4A Typical Fatty Acid Composition of Some Major Oils and Fats
Fatty acid
Oil or fat 16:0 18:1 18:2 Other (including, %)
Coconut^
Com 13 31 52 4
Cotton 24 19 53 4
Groundnut 13 37 41 9 (C2o-C,4, 7)
Olive 10 78 7 5
Palm 44 40 10 6
PalmkemeU
Rape 4 56 26 14 (18:3, 10)
Sesame 9 38 45 8 (18:0, 6)
Soybean 11 22 53 14 (18:3, 8)
Sunflower 6 18 69 7 (18:0, 6)
Castor^
Fish oils^"
Linseed 6 17 14 63 (18:3, 60)
Butter''
Lard“
Tallow'
"See Table 11.

^See Section XII.A
^See Table 31.
‘‘See Table 25.
^See Table 27.
Source: Ref. 5.
[4]. These figures show the importance of the United States, Argentina, India, China, and the
countries of southeast Asia as sources of production. Export figures are somewhat different
because highly populated countries such as India and China consume much of their indigenous
oils. European Community countries growing rape and sunflower are not the same because
these are grown mainly in northern and southern regions, respectively.
V. FATTY ACID COMPOSITION

The fatty acid composition of most of the oils and fats discussed in this chapter are summa
rized in Tables 4A -4C. Most vegetable oils contain high levels of oleic and/or linoleic acid
along with palmitic acid as the major saturated acid. Sometimes the oleic acid content is very
high, sometimes linoleic acid is high, and sometimes both have moderate values. For exam
ple, oleic acid exceeds 50% in olive oil (78%), in rapeseed (56%), and in the high-oleic
varieties of safflower oil (Section XIII.M) and sunflower oil (Section XIII.P). Linoleic acid
exceeds 50% in sunflower (69%), soybean (53%), cotton (53%), and com (52%), and both
acids exceed 30% in com (31 and 52%), groundnut (37 and 41%), and sesame (38 and 45%).
These examples are taken from the oils listed in Table 4A. Other examples are given in
Table 4C.
Occasionally other acids beyond these three are present at significant levels. Examples
include medium-chain acids (C 10- C 14 and especially C 12) in coconut and palmkemel oils,
stearic acid in cocoa butter and some similar hard fats, a-linolenic acid in linseed, rapeseed.
24 Gunstone
Table 4B Typical Fatty Acid Composition of Some Major and Minor Oils and Fats
Fatty acid
Oil or fat 16:0 18:0 18:1 18:2 Other (including, %)
Cocoa butter 24 34 37 3 2
Crambe 2 1 16 8 73 (18:3, 7; 22 :1, 56)
Mustard 3 1 23 9 64 (18:3, 10; 20:1, 8; 22 :1, 43)
Rape (high-erucic) 3 1 16 14 66 (18:3, 10; 20:1, 6; 22:1, 49)
Rice bran 16 2 42 37 3
Safflower
High-oleic 6 2 74 16 2
High-linoleic 7 3 14 75 1
Sunflower
High-oleic 4 5 81 8 2
High-linoleic 6 5 20 68 1
Tall oil
American 1 2 46 36 15 (conjugated, 9; pinolenic, 2)
Scandinavian 1 2 30 45 22 (conjugated, 5; pinolenic, 9)
Tung oil 3 2 11 15 69 (eleostearic, 69)
Vemonia seed oil 3 3 5 14 75 (vemolic, 75)
Wheat germ 14 1 22 58 5 (18:3, 5)
Source: Ref. 5.
and soybean, and ricinoleic acid in castor oil. Several other examples are found throughout
this chapter.
The more variable fatty acid composition in butter, lard, and tallow, and in a range of fish
oils discussed in Sections XII.V-XILX.
VI. TRIACYLGLYCEROLS
Since seed oils and animal fats consist mainly of triacylglycerols (>98% ), the composition of
these materials can be fully defined only in terms of the individual glycerol esters present.
Such information is, however, not normally provided with traded samples, and there are
several reasons for this.
1 . The potential number of triacylglycerols is quite large. In the oversimplified case of
an oil or fat with three major acids (such as palmitic, oleic, and linoleic), ignoring the minor
acids, there could be 27 (3^) different glycerol esters. If stereoisomers are ignored, this num
ber falls to 18, and if all isomers are ignored it reduces to 10 species (PPP, POP, PLP, POO,
PLL, POL, OOO, OOL, OLL, LLL). In a more realistic case with five different acids these
numbers would be 125, 75, and 35, and the situation is even more complex with materials
such as milk fats and fish oils containing many more acids. When it becomes possible to
analyze natural material to this level of sophistication it will be difficult to understand the
significance of the results because of the large number of different triacylglycerols, many of
them at quite low levels.
2. The changes in fatty acid composition that occur in natural material and are recognized
in Codex specifications are multiplied in the triacylglycerol composition. Small differences in
fatty acid composition lead to larger changes in triacylglycerol composition. This implies that
to define a fat or oil either there will have to be a wide range of permitted values for the
major triacylglycerols or all samples will have to be analyzed.
Table 4C Fatty Acid Composition of Some Minor Oils and Fats
Major component acid
Common name Botanical name 16:0 18:1 18:2 Other
Aceituno Simarauba glauca 12 58 2 18:0, 28

Almond Prunus amygdalus 7 61 30
Apricot Prunus armeniaes 5 60 32
Argane Argania spinosa 14 47 32 18:0, 6
Avocado Parsea americana 12 71 14
Babassu Orbignya martiana, 9 13 3 12:0, 45; 14:0, 17
0 . oleifera
Brazil nut Berthrolletia myrtaceae 14 29 47 18:0, 9
Buffalo gourd Cucurbita foetidissima 9 25 64
Candlenut Aleurites moluccana. 6 -8 17-25 38-45 18:3, 25-30
A. trisperma
Cashew Anacardium occidentale 11 60 17
Cherry Prunus avium 6 35 45 a-Elst, 10^
Chinese vegetable Sapium sebiferum 62 27 1
tallow^
Gold of pleasure Camelina sativa 4 -6 15-20 16-24 20:1, 10-16;
18:3, 30-40
Grapeseed Vitis vinifera 7 16 72
Hazelnut (filbert) Corylus avellana 7 69 21
Hemp Cannabis sativa 6 12 55 18:3, 25
Illipe butter Shorea stenoptera 18 35 1 18:0, 46
(Borneo tallow)
Kapok Ceiba pentandra + others 16 21 37 mal, Ster, 13*^
Macadamia Macadamia tetraphylla 8 56 3 16:1, 22
Mango Mangifera indica 5 -8 40-50 2 -4 18:0, 40-45
Mango olein — 7-10 48-53 4 -8 18:0, 30-34
Meadowfoam Limnanthes alba — 1-3 1-3 20:1, 60-65;
22:1, 11-14;
22:2, 17-22
Mowrah Madhuca longifolia 28 14 49
Nectarine — 6 66 27
Oiticica Licania rigida 7 6 — Licanic, 78^
Passionflower Passiflora edulis 8 13 77
Peach Prunus persicae 8 59 33
Pecan Carya illipensis 6 67 22
Pistachio Pistachio vera 9 69 18
Plum Prunus domestica 6 62 30
Poppy Papaver somniferum 10 11 72 18.3, 5
Pumpkin Cucurbita maxima 11 47 34
Safflower (HO) Carthamus tinctorus 6 74 16
Sal Shorea robusta 4 -7 37-43 0 -4 18:0, 40-46
Sal olein — 6 -9 48-52 2 -6 18:0, 31-35
Shea Butyrospermum paradoxum 3 -4 44-46 6 -8 18:0, 42-45
Stillingia^ Stillingia sebiferum 5 13 23 18:3, 47; HO 8:2,
5; 10:2, 5
Sunflower (HO) Helianthus annus 4 81 8 18:0, 5
Tea Thea sansanqua 16 59 22
Tobacco Nicotiana tabacum 8 16 71
Ucuhuba Virola sufinam 7 8 2 12:0, 13; 14:0, 69
Walnut Juglans regia 8 18 58 18:3, 8
^Stillingia oil and Chinese vegetable tallow come from the same source.
*’Eleostearic, 9 c llil3 c -1 8 :3 ; malvalic, Cig-cyclopropene acid; sterculic, C i 9"Cyclopropene acid; licanic, 4-oxo-
9 c llil3 c -1 8 :3 .
Source: Refs. 5 and 7.
26 Gunstone
3. A third problem is that there is still no simple way of measuring triacylglycerol compo
sition and such analyses are usually a matter for the research laboratory rather than the quality
control laboratory. This, however, is changing in the study of some of the harder fats. The
several ways of studying this problem generally give partial information, and it is not easy to
combine the results obtained by different procedures. The most important and widely used
methods of analysis are as follows.
a. Separation of glycerol esters according to their levels of unsaturation by silver ion

chromatography. Originally achieved by thin layer chromatography, this is now gener
ally effected by HPLC using a silver ion column. The results are given in terms of
triacylglycerols but not usually for individual molecular species. For example, a figure
assigned to LLO represents the sum of three isomers depending on whether the oleic
chain is in the sn-l, -2, or -3 position. Similarly, the value for LOP represents the six
esters containing these three acids.
b. Separation according to partition number by reverse-phase HPLC. Again, the triacyl
glycerols identified are usually not individual molecular species but mixtures as de
scribed in (a).
c. Separation according to the number of carbon atoms in the three acyl chains by gas
chromatography. This has improved since its first introduction, and a lot of useful
information can be obtained in this way. Smaller separation by unsaturation is added
to the major separation by carbon number.
d. Analysis by enzymatic procedures or by HPLC of mono- or diacylglycerols resulting
from reaction with ethylmagnesium bromide. These are separated either on a chiral
column or after reaction with a chiral reagent. These procedures can give an indication
of the composition of fatty acids in each of the sn-l, -2, and -3 positions. Triacylglyc
erol composition can be calculated from the results only if it is assumed that the acids
in each position are randomly associated with each other. This is described as a 1-
random 2-random 3-random arrangement. It could be argued that the procedures that
give information about the fatty acids in each of the sn positions are developments of
fatty acid composition rather than triacylglycerol composition. Sometimes they are
combined with other separation procedures carried out in a preparative manner.
Some results obtained by these methods are included with the discussion of individual oils in
Section XIII.
VII. SAPONIFICATION VALUE, IODINE VALUE, AND TOTAL

UNSAPONIFIABLE MATERIAL
Many oil and fat specifications still include a reference to classical values such as the saponi
fication value, the iodine value, and the content of unsaponifiable material.
The saponification value (SV, the number of milligrams of potassium hydroxide required
to saponify 1 g of oil) is less immediately indicative of fatty acid composition than the saponi
fication equivalent (SE, the weight of oil saponified by 1 g-mol of potassium hydroxide). The
latter is a measure of the average chain length of all the acyl chains. The high level of Cjg
acids in most oils gives a saponification equivalent around 291-297 (sap value 189-193).
This figure is raised by the presence of unsaponifiable matter (nonacidic) and by the presence
of acids with molecular weight greater than 297 as glycerol esters (e.g., ricinoleic acid, ver-
nolic acid, C20- C 24 acids). Conversely, the saponification equivalent is reduced by the pres
ence of palmitic acid at higher levels than usual (e.g., palm oil, cottonseed oil) and for the
Table 5 Saponification Value (SV), Saponification Equivalent (SE), Iodine Value (IV), and Content
of Unsaponifiable Material in Some Vegetable Oils
Unsaponifiable
Oil SV^ SE^ IV (%)
Castor 176-187 300-319 81-91 1.0
Cocoa butter 190-200 281-295 33-40 0 . 2- 0.6
Coconut 248-265 212-226 6-11 1.5 max
Com (maize) 187-195 288-300 103-128 2 . 8 max
Cottonseed 189-198 283-297 99-119 1.5 max
Groundnut 187-196 286-300 80-106 1 . 0 max
Linseed 189-195 288-297 177 min 1.5 max
Olive 184-196 286-305 75-94 1.5
Palm 190-209 268-295 50-55 1 . 2 max
Palmkemel 230-254 221-244 13-23 1 . 0 max
Rapeseed (low emcic) 188-193 291-298 110-126 2 . 0 max
Rice bran 181-189 297-310 99-108 3-5
Safflower 186-198 283-302 135-150 1.5 max
Sesame 187-195 288-300 104-120 2 . 0 max
Soy 189-195 288-297 120-143 1.5 max
Sunflower 188-194 289-298 110-143 1.5 max
Tung 189-195 288-297 160-175 0.5
Vemonia 165-210 267-340 104-108 6.7
Wheat germ 184-185 303-305 120-130 2-6
Trilaurate 213.0 0
Tripalmitate 269.1 0
Tristearate 297.2 0
Trioleate 295.2 86.0
Trilinoleate 293.1 173.2
Trilinolenate 291.1 261.5
^SE = 56108/SV.
Source: Ref. 5.
lauric oils (coconut oil, palmkemel oil). These factors have the reverse effect on the saponifi
cation value, as the two are related by the expression SE = 56,108/SV.
Iodine values are a measure of average unsaturation. Most oils rich in palmitic, oleic, and
linoleic acid have iodine values in the range 90-130 depending on the relative levels of
saturated, monoene, and diene acids. The iodine value is higher in oils rich in triene acids
(linseed, tung) and lower when there is a high level of saturated acids such as palmitic or
stearic (palm, cocoa butter) or of the medium-chain acids (coconut, palmkemel).
Vegetable oils generally contain up to 1.5% unsaponifiable material (sterols, 4-methylste-
rols, triterpene alcohols, tocopherols, hydrocarbon, etc.; see Section IX) though some higher
values are apparent in Table 5.
VIII. PHOSPHOLIPIDS
Phospholipids of various types (Chapter 1) are present as minor components (0.5-3.0% ) in
most c m d e oils, from which they are largely removed during refining. They may be recovered
in a by-product, generally called lecithin, which is a mixture of phospholipids and triacylglyc-
28 Gunstone
Table 6 Composition of Phospholipids from Selected Natural Sources
Percent of total
Kjiai pmjapiiuiipiil -------
Source in oil (%) PC PE PI Sum PI/PA PA Other
Cottonseed 1.0 23 14 13 9 41
Groundnut 0.4 23 8 17 3 49
Palm — 36 24 22 9 9
Rapeseed 2.5 48 9 20 —
23
Soybean 3.2 39 23 20 5 13
Lecithin, liquid 43 26 — 19 — 12
Lecithin, granular 32 18 — 32 — 18
Sunflower seed 1.5 49 21 28 2 —
Egg yolk“ 25 5 tr — 70
^Composition o f total egg yolk lipid (including neutral lipids 62% and sterols 5%).
S o u rc e : Ref. 5.
erols. The material recovered from soybean oil is an important commercial product. It pro
vides a source of mixed and individual phospholipids at various levels of purity. Lecithin is
also available from rape, com, and sunflower, and egg yolk is another important source of
phospholipids. These are important surface-active compounds used extensively in the food,
pharmaceutical, and cosmetic industries.
The major phospholipids in cmde lecithin are usually phosphatidylcholines (PC), phospha-
tidylethanolomines (PE), phosphatidylinositols (PI), and phosphatidic acids (PA) accompanied
by other phospholipids at lower levels. The proportion of these vary with the source and with
the method of recovery, and some typical figures are given in Table 6. One commercial
supplier reports his soybean lecithin to contain PC (20-23%), PE (16-21%), and PI (12-
18%), with the balance being triacylglycerols and minor phospholipids. The component acids
vary with the phospholipid source and the phospholipid class, and typical results are given in
Table 7. The fatty acid composition of individual phospholipid types can be manipulated by
enzymatic processes.
Table 7 Fatty Acid Composition of Some Phospholipids
PC PE
Fatty acid Soy Rape Sunflower Egg Soy Egg
16:0 15 6 11 38 37 15
18:0 3 1 4 9 2 28
18:1 7 56 12 32 10 21
18:2 68 31 72 12 50 9
18:3 6 4 tr — 1 —
20:4 (n-6) — — — 2 — 10
22:6 (n-3) — — — 5 — 13
Other 1 2 1 2 tr 4
S o u rc e : Ref. 6 .
Table 8 Sterol Composition (Percentage of Total Sterol) of Some Vegetable Oils
Palm- Cotton- Soy- Ground Sun

kemel Coconut seed bean Com nut Palm flower Rape
Total
Range^ 0.8-1.2 0.5-1.1 2.7-S.9 1.8-4.1 7.9-22.1 0.9-2.9 0.4-0.5 2.4-4.S 4.8-11.3
Average^ 1.0 0.8 4.5 3.2 13.8 1.6 0.4 3.4 7.5
Cholesterol 2 2 1 1 tr 2 4 1 1
Brassicasterol tr 1 tr tr tr tr ND tr 10
Campesterol 10 9 8 20 21 17 23 10 34
Stigmasterol 14 12 1 18 6 8 11 9 tr
Sitosterol 67 46 84 54 63 58 57 60 49
A^-Avenasterol 6 26 3 2 6 12 3 3 5
A^-Stigmasterol tr 2 1 3 2 2 1 10 1
A^-Avenasterol tr 1 1 1 1 1 1 5 tr
Other ^ 1 1 1 1 1 tr 2 tr
tr = trace; ND = not detected.

"In grams per kilogram.
Source: Refs. 5 and 8 .
IX. STEROL COMPOSITION

Most vegetable oils contain 1000-5000 ppm (0.1-0.5% ) of combined sterols present both as
free sterols and as esters with long-chain acids. Higher levels are apparent in rapeseed oil
(range 5000-11,000 ppm, mean value about 7500) and in com oil (range 8000-22,000, mean
value 14,000). Sitosterol is generally the major phytosterol, representing 5 0 - 8 0 % of total
sterol; campesterol, stigmasterol, and A ^-avenasterol frequently attain significant levels. Bras
sicasterol is significant only in seed oils of brassica species.
Cholesterol is generally considered to be a zoosterol and is not present in plant systems at
any significant level. The normal value of 10-20 ppm is to be compared with the much higher
values reported for animal fats (up to 1000 ppm), fish oils (up to 7000 ppm), dairy fats
(2000-3000 ppm), and egg yolks (12,500 ppm = 1.25%). Details are given in Table 8 [5,8].
X. TOCOPHEROLS AND VITAMIN E

Though present at still lower levels, the tocopherols and tocotrienols attract attention because
of their vitamin E and antioxidant activities. The tocopherols are a series of benzopyranols
with one, two, or three methyl groups attached to the phenolic ring. The molecules also have
a Ci 6 side chain on the pyran ring; this is saturated in the tocopherols and has three double
bonds in the tocotrienols. There are four tocopherols and four tocotrienols designated a, j8,
y, 8. Some typical contents are given in Table 9.
The tocopherols show different vitamin E activity in the order a (1.0) > ^ (0 .5 ) > y
(0.1) > 6 (0.03). The vitamin E activity of some oils, expressed as a-tocopherol equivalents,
is listed in Table 9 in the extreme right column. These depend particularly on the level of a-
tocopherol. Antioxidant activity is in the reverse order (8> y > /3> a) so that the various oils
do not follow a similar sequence of vitamin E and antioxidant activity.
Walnut oil and wheatgerm are good sources of tocopherols. Others with high vitamin E
activity include sunflower, cottonseed, safflower, and palm oils. Antioxidant activity is high
in soybean oil, com oil, and walnut oil. Some oils have high oxidative stability despite
30 Gunstone
Table 9 Tocopherols and Vitamin E Activity in Oils and Fats^
Tocopherols Tocotrienols Vit E activity

as a-tocopherol
a equivalents
Coconut^ 5 6 7
Com 112 50 602 19 198
Cottonseed 389 387 428
Groundnut 130 214 21 152
Olive 119 7 120
Palm‘d 256 316 70 146 335
Palmkemel 62 62
Rape 210 42 tr tr 215
Safflower 342 71 349
Sesame 136 290 165
Soybean 75 15 797 266 171
Sunflower 487 51 8 492
Walnut 563 595 450 636
Wheatgerm 1330 710 260 271 26 18 1736
Cod liver 220 220
Herring 92 92
Menhaden 75 75
Lard 12 15
Tallow 27 27
Âll figures in milligrams per kilogram (i.e., ppm).

Âlso yT3 19.
Âlso yT3 286.
their low a-tocopherol levels because other antioxidants (usually phenolic compounds)
are also present. Oxidative stability will also be influenced by the level of unsaturation in
the oil.
The tocopherols are partially removed during oil refining. This reduces oxidative stability
but can furnish a concentrate that can be added back or added to other oils or fatty foods.
Much natural antioxidant comes from a by-product of soybean refining. Concentrates of to
copherols from palm oil are also produced for use as vitamins (Palm Vit E).
XII. WAXES
Waxes include a variety of long-chain compounds occurring together in plants and animals
and finding use in the food, pharmaceutical, and cosmetic industries. They are generally
water-resistant materials made up of mixtures of hydrocarbons, ketones, acids, alcohols, and
their esters and are used for the protection of surfaces of shoes, automobiles, furniture, paper,
etc. They differ from the long-chain acids in triacylglycerols in that they are usually of higher
molecular weight (up to C^q and beyond) and are frequently branched with one or more
methyl groups, and although they may be unsaturated they generally do not contain the meth
ylene-interrupted polyunsaturation characteristic of the common polyene acids [5,9]. The
more important commercial materials are discussed in Section XII.U.
XIII. INDIVIDUAL SOURCES

A. Castor Oil
The castor oil plant (Ricinus communis), which grows in tropical, subtropical, and even some
temperate regions, is cultivated mainly in India, Brazil, and China. The oil makes up about
one-half of the shelled bean. The residual meal contains several poisonous and allergenic
materials and can be used for animal feeds only after detoxification. (It is used mainly for
ruminants.) The oil differs from other commercially available oils by reason of the presence
of the hydroxy acid ricinoleic acid (Chapter 1, Section XI.A, page 9) at the 8 3 - 9 0 % level.
Compared with other vegetable oils, castor oil is more viscous, less soluble in hexane, and
more soluble in alcohol as a consequence of the presence of the hydroxy acid. Ricinoleic acid
and the esters are most commonly used after chemical modification.
1. Pyrolysis of castor methyl esters at 4 5 0 - 5 0 0 ° gives methyl 10-undecenoate (C ^) and

heptanoic acid. The C ^ compound is used to produce perfumes and polymers.
2. Alkali fusion at 250-275° converts the Cjg hydroxy acid to Cg (2-octanol or 2-octa-
none) and C jq (sebacic acid or 10-hydroxydecanoic acid) compounds.
The surface-active properties of ricinoleic acid or its esters can be enhanced by reac
tion of the midchain hydroxyl group. For example, sulfated castor oil (Turkey Red
oil) is much used in the textile industry. It can also be ethoxylated.
4. Hydrogenation affords 12-hydroxy stearic acid, which is used in the paint industry to
introduce thixotropy and as a component of many greases in the form of its lithium
salt.
5. Dehydration of castor oil by heating at 230-280° in the presence of an acidic catalyst
gives isomers of octadecadienoic acid (mainly the A 9 , l l and A 9 ,1 2 dienes). Dehy
drated castor oil is used in paints and varnishes as a drying oil.
Unlike some oils containing hydroxy acids, the hydroxyl groups in castor oil are not acy-
lated and remain free. It was shown some time ago [10] that the major triacylglycerols contain
two or three ricinoleic acid groups: RRR (68 %), RRO (8%), RRL (8%), RRSt (5%), and
others (11%). Newer information has been obtained by mass spectrometry [11].
Seed oils containing other hydroxy acids are being developed as potential new oilseed
crops. Lesquerella fendleri seeds yield an oil (25%) containing about 55% lesquerolic acid.
This is the C 20 homologue of ricinoleic acid (14 -h y d ro x y -ll-c w -e ic o s e n o ic acid) and should
show properties similar to those of the Cjg acid. Dimorphotheca oil contains dimorphecolic
acid (9-hydroxy-1 0-trans, 12 -cw-octadecadienoic acid), which is readily dehydrated to give a
Cig acid with a conjugated triene system.
B. Cocoa Butter
The cocoa bean (Theobroma cacao) is the source of both cocoa powder and cocoa butter,
which are important ingredients of chocolate. The value of cocoa butter in this connection
depends mainly on its melting behavior, which is related to the fatty acid and triacylglycerol
composition. Cocoa butter is a solid fat that melts in the mouth, producing a desirable cooling
sensation on the tongue. This is a consequence of the high level (64-77%) of symmetrical
disaturated oleo glycerol esters (SOS). The content of unsymmetrical esters (SSO) is below
1 %. The major glycerol esters are POP (18-23%), POSt (36-41%), and StOSt (23-31%)
(P = palmitic, 0 = oleic, St = stearic, S = saturated acids). The crystalline forms and melting
behavior of this fat have been studied intensively [12-14].
32 Gunstone
Because cocoa butter commands a good price that tends to fluctuate depending on the
vagaries of cocoa production, there is an interest in less expensive fats from other sources
that show similar melting properties [15]. No less than five different approaches to this prob
lem are being examined.
1. Other natural fats with similar fatty acid and triacylglycerol composition are known
(Table 10), and some of them are commercially available though at lower levels than
cocoa butter [5]. Supplies are sometimes unreliable.
Fractions of oils with high levels of SOS can be isolated, usually by crystallization.
For example, double fractionation of palm oil produces a product (palm midfraction,
PMF) containing palmitic (48-54%), stearic (5%), oleic (35-38%), and linoleic (4-
7%) acids.
3. Attempts have been made to produce a fat with the required properties by growing
yeast in the presence of stearic acid or one of its simple alkyl esters or, alternatively,
in the presence of sterculic acid, which inhibits the desaturation of stearic to oleic acid.
4. A concentrate of StOSt that is a valuable component of chocolate fat has been obtained
by regiospecific lipolysis of high-oleic sunflower oil containing triolein with stearic
acid using Mucor miehei as a source of lipase.
p 0 Stearic — 0 — St St
acid
— 0 -- 0 — 0 ------ --------- > O
Mucor
— 0 miehei ^ St L- 0
5. By genetic engineering it is possible to produce a rapeseed oil in which the level of

stearic acid has been raised to about 40%. A high-stearate (15-30%) soybean oil has
also been developed.
C. Coconut Oil
The coconut palm {Cocos nucífera) grows between the latitudes 20° north and 20° south of
the equator, in the Philippines and in Indonesia. Inside the hard shell of the coconut is a thick
endosperm layer. When dried and separated from the shell this is known as copra and is the
source of coconut oil (65-70%). This oil is used for food and non-food uses in roughly equal
amounts. For example, alcohols can be produced from coconut oil by hydrogenolysis. Coco
nut oil is an important lauric oil characterized by high levels of medium-chain acids, espe
cially lauric acid (C 12) at about 50%, along with lower levels of the C 14, Cô, Cg, and
Table 10 Natural Oils Containing High Levels of SOS Glycerides (wt %)
Source Oil mp (°C) 16:0 18:0 18:1 Other SOS
Shorea robusta Sal oil 30-35 5 44 42 9 64

S. stenoptera Illipe butter 34-38 19-20 40-42 35-38 1 79
Valeria indica Malabar tallow 34-36 12 41 45 2 74
Garcinia indica Kokum butter 38-42 4 53 40 3 74
Madhuca butry Phulwara butter 40-42 60 4 36 — 69
Madhuca latifolia. Mowrah butter — 24 19 43 14 —
M. longifolia — 28 14 49 9 —
Butyrospermum parkii Shea butter — 3 44 46 6 —
Source: Ref. 5.
Table 11 Component Acids (wt %) and Major Triacylglycerols (>5%, mol

in Coconut and Palmkemel Oils
Component acids
Acid 6:0 8:0 10:0 12:0 14:0 16:0 18:0 18:1 18:2
Coconut 0.5 8 7 47 18 9 3 6 2
Palmkemel 0.5 2.5 4 49 16 9 2 14 2
Major triacylglycerols
Saturated Number of
acyl groups isomers ^ Coconut Palmkemel
12, 12, 8 11.9 6.4

12, 12, 10 5.7 4.7
12, 12, 12 10.6 19.8
12, 12, 14 10.8 14.1
14, 12, 8 9.1 3.6
Other 51.9 51.4
^Not including stereoisomers.
saturated acids. Hydrolytic rancidity furnishes free acids that give the oil a soapy off-flavor.
The acids can be further oxidized and decarboxylated to give odd-chain methyl ketones, which
are the source of ketonic rancidity.
RCH2CH 2C 0 0 H -^ RC 0 CH 2C0 0 H ^ RCOCH 3
Triacylglycerol analysis by gas chromatography shows that the major glycerol esters have
two or three lauric acid chains (Table 11).
D. Corn Oil (Maize Oil)

Com (maize, Zea mays) is grown mainly for use as animal feed, with an increasing amount
being subjected to wet milling to produce com starch, sugar, and symp. The oil, recovered
mainly from com germ, has good oxidative stability in spite of its high level of linoleic acid
and is used as a salad oil, a frying oil, and in margarine (after partial hydrogenation). The
levels of palmitic, oleic, and linoleic acid are in the range 9-17, 20-42, and 39-63 respec
tively (Codex values). Triacylglycerol composition has been examined by stereospecific analy
sis [16], silver ion chromatography [5], and gas chromatography [5]. These studies indicate
that the major triacylglycerols include LLL (15%), LLO (21%), LLS (17%), LOO (14%),
LOS (17%), LSS (5%), 0 0 0 (6%), and OOS (4%) or, alternatively, that the TAG molecules
are mainly C52 (26%) or C54 compounds (66%).
E. Cottonseed Oil
Cotton (Gossypium spp) is a subtropical species now grown also in temperate zones. It is
produced mainly in China but also in the United States, the former Soviet Union, India,
Pakistan, and Brazil (Table 3). It is grown mainly for its fiber, but the seed is a useful by
product, yielding both oil (—30%) and protein (—30%). The seed represents only about 11-
12% of the total value of the crop. A significant amount is fed to animals as whole seed. The
oil is used as a salad or cooking oil after winterization (i.e., removal of the more solid
34 Gunstone
triacylglycerols). In the early decades of this century cottonseed oil was the major oil in the
United States, but it is now third after soybean and com. On the world scale it ranks fifth
among the vegetable oils after the four market leaders (soy, palm, rape, and sunflower).
Cottonseed oil imparts a desirable nutty flavor to crisps/chips. It also tends to form jS'
crystals. This is beneficial because it extends plastic range, aids aeration, and helps to contrib
ute to uniform appearance.
Cottonseeds contain gossypol (up to 0.4%), which is considered to be undesirable. It con
centrates in the meal rather than in the oil, from which it is finally removed during alkali
refining. Because of the gossypol the meal is suitable only for mminants. Glandless cotton
contains lower levels of gossypol.
Cottonseed oil is rich in palmitic acid (22-26%), oleic acid (15-20%), and linoleic acid
(49-58%). It also contains up to 10% C20- C 24 acids (mainly saturated) and about 1% sterculic
and malvalic acids in the c m d e oil. The cyclopropene acids are undesirable components, but
they are largely removed during refining, particularly deodorization, and also during hydroge
nation. They are not considered to present any health hazard in cottonseed oil.
The major triacylglycerols include the isomers represented by PLL, PLP, LLL, POL, LOL,
and POP, with palmitic acid present mainly in the sn-l and -3 positions and hardly at all in
the sn-2 position. The disaturated triacylglycerols (POP and PLP) are largely removed during
winterization. Bland e t al. [18] give a detailed triacylglycerol analysis indicating the presence
of PLL (26%), LLL (16%), POL (14%), OLL (13%), and PLP (9%), with seven others
making up the balance (22 %).
F. Evening Primrose Oil (Aiso Borage and Biackcurrant

Seed Oils)
Evening primrose {Oenothera biennis, O. lamarkiana) oil is one of a small number of oils
containing y -lin o le ic acid (6,9,12-18:3). This relatively rare triene acid of the n-6 polyene
family has attracted attention because of claims that it is beneficial in the treatment of a
number of diseases such as atopic eczema, premenstrual syndrome, multiple sclerosis, arthri
tis, and several others [19]. These claims are not universally accepted, but the oil is widely
used as a health supplement. y-Linolenic acid is the first intermediate in the bioconversion of
linoleic to arachidonic acid, and the first step of A6 desaturation is known to be rate-limiting.
Evening primrose oil is also very rich in linoleic acid (70-75%), so that with the y -lin o le ic
acid (8-10%) the content of n-6 essential fatty acids is very high. Other commercial sources
of y-linolenic acid include the oils from borage {Borago officinalis) and from blackcurrant.
Both these contain higher levels of y-linolenic acid than evening primrose oil but they are
used less extensively. Borage oil contains less linoleic acid and also contains C 20, C22, and
C24 monoene acids (total —8%). Blackcurrant seed oil contains significant levels of n-3 poly-
Table 12A Fatty Acid Composition of Oils Containing y-Linolenic Acid
Fatty acid (%)
Oil 16:0 18:0 18:1 18:2 18:3 (n-6) Other
Blackcurrant 7 2 11 47 17 16"
4 16 38 23 9b
Borage 10
Evening primrose 6 2 9 72 10 1
Încluding 18:3 (n-3), 13% and 18:4 (n-3), 3%.

‘’Including 20:1, 4.5%; 22:1; 2.5%; and 24:1, 1.5%.
Table 12B The Major Triacylglycerols

(> 1 wt %) in Borage and Evening Primrose
Determined by Reverse-Phase HPLC ^
Evening
TAG Borage Primrose
LG G 9.8 0.7
LLG 15.4 17.6
PGG 2.4 —
LLL 10.1 54.3

OLG 12.3 1.6
PLG 9.5 1.2
O LL 8.2 13.7
LGD 2.5 —
PLL, OOG 8.9 7.9

SLG, POG 5.0 —
PPG 1.5 —
LLD 4.1 —
OOL 4.1 0.8

POL, SLL 0.3 2.2
Other 5.9 0
Êach TAG entry includes all possible isomers.

Further information is also available in Ref. 48.
Source: Ref. 47.
ene acids (18:3 and 18:4). Analytical data are presented in Tables 12A and 12B [20]. Higher
levels of GLA are present in a variety of upgraded oils.
G. Groundnut Oil (Monkey Nut, Peanut, Arachis Oil)

The groundnut (Arachis hypogea) is an annual legume grown in tropical and semitropical
regions, especially in China and India, and in the United States at a lower level. The oil
shows excellent oxidative stability and is considered to have a desirable flavor. Its major
component acids are palmitic (8-14%), oleic (36-67%), and linoleic (14-43%) with typical
values of 13, 37, and 44%, respectively. The oil also contains some 6- 8% (total) of C 20,
C22, and C 24 saturated and monoene acids.
The major triacylglycerols have been examined by silver ion chromatography [5] and by
HPLC [17, 21] with the results detailed in Table 13.
Table 13 Triacylglycerol Composition of Groundnut Oil
LLL LLO LLS LOO LOS LSS 000 oos OSS Other Ref.
4 20 12 17 22 6 6 10 3 _ 5
6 26 8 21 13 2 5 16 1 2 21
Fatty acid composition [L (40,42), O (39, 43), S (21,15)] calculated from the above figures show that the oil samples
do not have the same fatty acid composition and therefore cannot be expected to have the same TAG composition.
36 Gunstone
Table 14 Triacylglycerol Composition of Linseed Oil
LnLnLn LnLnL LnLnO LnLO LnOO LnLnS LnLS LnOS Other Ref.
35 14 19 6 6 5 7 4 4 23
24 14 15 5 5 12 5 7 13 5
H. Linseed Oil
The flax plant (Linus usitatissimum) is grown either for its oil (linseed oil) or for its flber
(flax, linen). The oil variety is grown mainly in Argentina, India, the former Soviet Union,
the United States, and Canada, with Argentina and Canada the main exporting countries.
The seed content is 35-44% oil that is characterized by a high iodine value (>177) related
to the high level of a -lin o le n ic acid (9,12,15-18:3), which is usually 50-60% . The meal
remaining after removal of the oil is a valuable animal feed for cattle, pigs, chickens, and
fish. The whole seed finds limited use in baked goods. T h e triacylglycerol composition of this
oil is summarized in Table 14. Its stereospecific analysis is given in Refs. 5 and 22.
Because of its high unsaturation, the oil is used mainly in paints and varnishes, in the
production of linoleum, and as a sealant for concrete. Production of linseed oil has changed
little over the last 50 years because it has been replaced in some of its traditional uses by
products of the petrochemical industry. However, there is now evidence of renewed demand
for linseed oil in some of these products. For example, linoleum is being promoted on the
basis of its durability and its nonflammability and on environmental grounds in respect of
both its natural constituents (linseed oil, cork, woodflour, pine resin) and its ease and safety
of disposal. It is especially acceptable in high traffic public areas such as shops, schools,
hospitals, and airports.
Plant breeders in Australia have developed a variety of linseed with low levels of linolenic
acid (^--2%) and high levels of linoleic acid (—70%). Called linola, it yields a linoleic-rich oil
(like sunflower) and can be grown in the same temperate zones as rapeseed (canola). It is
already being grown in Australia and Canada and most recently in Europe. Table 15 compares
the fatty acid composition of linseed and linola oils.
I. Olive Oil
Olive oil is obtained from the mesocarp of olives (Olea europeae). Commercial growth of the
olive tree is confined almost entirely (—97%) to the Mediterranean countries of Italy, Greece,
Spain, Turkey, and Tunisia. The ripe mesocarp is 15-40% oil that is especially rich in oleic
Table 15 Fatty Acid Composition of Linseed Oil and of Linola
Fatty acid
Oil 16:0 18:0 18:1 18:2 18:3
Linseed 6 3 17 14 60
Linola
Glenelg 8 5 21 64 2
Croxton 8 3 18 69 2
acid. Codex ranges are palmitic 8--20%, oleic 55-83% , and linoleic 4-21% . Christie et al.
[24,30] has given the fatty acid composition of olive oil at each of the sn-l, sn-2, and sn-3
positions. In a study using silver ion HPLC it was concluded that the major triacylglycerols
are S S O (3 % ), S O O (3 1 % ), 0 0 0 (4 6 % ), S O L (6 % ), an d O O L (1 4 % ). In a n o th e r re p o rt, 99
olive oil samples were investigated by reverse-phase HPLC [23]. Six dominant glycerol esters
make up 82-93% of the whole oil; the median values are LOO, 11%; OOO, 43%; POP, 3%;
P O L , 4% ; P O O 2 2 % ; a n d S tO O , 5 % .
In addition to the usual range of minor components, olive oil contains the C 30 hydrocarbon
squalene at a higher level (150-700 mg/100 g) than most other vegetable oils (5-50 mg/
100 g).
The oil is available in several grades and commands a good price. For this reason the more
expensive grades are subject to adulteration with poorer grades of olive oil and/or with other
oils, and considerable effort is devoted to developing analytical procedures to detect this. The
oil is used mainly for human consumption as a salad or cooking oil.
J. Palm and Palmkernel Oils

The oil palm (Elais guinensis) grows in tropical regions of Asia, Africa, and America and
predominantly in Malaysia and Indonesia. In both of these latter countries there has been a
considerable increase in production during the past 15 years. This increase is expected to
continue for some years to come, and palm oil will replace soybean oil as the largest produced
oil; also, Indonesia is expected to overtake Malaysia as the leading producer. For good pro
duction the tree requires a humid tropical climate with rainfall (—2000 mm) spread through
the year. In terms of oil per hectare, the oil palm outproduces all other oil crops, yielding
about 5 -7 tonnes/hectare on well-managed plantations.
Palm trees begin to produce fruits after 3 -4 years, peak after about 10 years, and may
have a production life of up to 30 years. In later years oil yields decline and the increasing
height of the tree may present harvesting problems. The fruit bunches (4-20 kg) contain 200-
2000 individual fruits and yield palm oil (20-24%) and palmkernel oil (2-4% ). Breeding
objectives include still higher average yields, plants of lower height, higher ratios of kernel
to mesocarp, and a modified fatty acid composition of the oil.
The oil palm produces two distinct oils: palm oil from the fleshy mesocarp and palmkernel
oil from the fruit kernel. The usage of both oils is extended by fractionation. Palm oil gives
a more valuable olein (an excellent frying oil) and a less valuable stearin (used in part as a
replacement for tallow in the oleochemical industry). With palmkernel oil the stearin is the
more valuable fraction [26]. The oils and their fractions can be further modified by blending
with other oils, by partial hydrogenation, or by interesterification. Further fractionation of the
palm oil fractions yields an intermediate fraction (PMF, palm midfraction) that can be used
as a cocoa butter extender. Palm oil is characterized by high levels of carotene and of tocoph-
erols and tocotrienols (vitamin E ), which can also be isolated or concentrated as products of
considerable value. Palm oil fatty acid distillate (PAFD), a by-product of the principal refining
procedure, is an important component of animal feed.
Palm oil is widely used as a food oil with limited non-food uses also. It differs from other
commercially available oils in its fatty acid and triacylglycerol composition. It contains almost
equal amounts of saturated acids (mainly palmitic with some stearic) and unsaturated acids
(mainly oleic with some linoleic acid). The proportion of palmitic and oleic acids in the major
triacylglycerols leads to stability of the ¡3' crystals. These are very desirable in the production
of margarines and shortenings, especially those with high levels of unsaturated acids. Some
information on triacylglycerol composition is given in Table 16.
38 Gunstone
Table 16 Triacylglycerol Composition of Palm Oil
(i) Major triacylglycerols (%)
POO,
PPP POP PLP post PLSt PLO ooo Other Ref.
7 33 7 6 23 7 3 14 27
5 43 8 4 29 8 2 1 28
7 31 9 6 23 9 4 11 29
(ii) Distribution between sn-l, -2, and -3 positions^ (%)
16:0 18:0 18:1 18:2
TAG 48.4 3.7 36.3 10.0

sn-l 60.1 3.4 26.8 9.3
sn-2 13.3 0 .2 67.9 17.5
sn-3 71.9 7.6 14.4 3.2
^14:0 and 16:1 are also cited in the original paper [30].
Source: Refs. 2 7 -3 0 .
The triacylglycerol composition of palm oil has been examined by HPLC and GLC proce
dures and also by stereospecific analysis. As expected from the high levels of palmitic and
oleic acids in this oil, the triacylglycerols containing these two acids (POP and POO) are the
dominant species.
Palmkemel oil, like coconut oil, is an important lauric oil with significant food and non
food uses. Some information on triacylglycerol composition is listed in Table 11 along with
that for coconut oil.
K. Rape and Other Brassica Species (Also Crambe

and Mustard)
Several species of Brassica are available in commercial quantities including rape {Brassica
napus, B. rapa (or campestris), and B. júncea] and mustard (B. alba and other species).
Typically these oils were rich in erucic acid as is Crambe oil {Crambe abyssinica and C.
hispánico), but some have been modified by classical or modem methods of seed breeding to
give a different fatty acid composition.
Today emcic acid is available from high-emcic rapeseed oil (about 50%) or from Crambe
oil (about 60%). At present there is a demand for about 20,000 tonnes of emcic acid based
compounds derived from about 56,000 tonnes of emcic-rich oil. It is used mainly as the
amide (slip agent for polyethylene film), alcohol (emollient), esters (lubricants), and various
nitrogen derivatives (personal care products). There is also a demand for behenyl alcohol
(pour point depressant) and behenyl glycerol (food emulsifier). Behenic acid (22:0) is the
reduced form of emcic acid [31].
Low-emcic rapeseed is now the third largest source of oil (after soy and palm) and also of
meal (after soy and cotton). The seed contains over 40% oil, which represents about 80% of
the seed’s monetary value. The residual meal is 33-44% high quality protein. Seed breeders
have developed seeds that produce oil low in emcic acid (<5% or <2% ) and meal low in the
undesirable sulfur-containing glucosinolates. This low-emcic oil (known as LEAR or canola)
finds many food uses and some non-food uses (biodiesel, lubricants). The cmde oil is rich in
phospholipids (—3.5%), though these are reduced to 10-300 ppm (phosphoms) after refining
Table 17 Stereospecific Analysis of High-Erucic Rapeseed Oil
16:0 18:0 18:1 18:2 18:3 20:1 2 2:1
TAG 3.9 1.3 18.6 16.1 8.7 10.4 38.0

sn-\ 5.6 1.9 16.5 1.2 3.3 17.4 51.1
sn-2 1.6 0.4 33.4 39.3 21.3 1.9 0 .8
sn-3 4.9 1.7 5.6 7.4 1.3 12.0 62.3
Source: Ref. 32.
and are themselves a useful by-product. This oil grows in cooler agricultural regions including
China, northern Europe, and Canada as well as on the Indian subcontinent. In common with
other Brassica seeds, rapeseed oil contains brassicasterol at levels (—600 ppm) not attained
in other species.
Plants of the Brassica species lend themselves to genetic engineering, and considerable
progress has been made in changing their fatty acid composition. Oils with lauric acid (40%)
or stearic acid (40%) are already undergoing extensive field trials, and other possibilities under
active investigation include oils with ricinoleic acid, erucic acid at the 90% level, oleic acid
at a high level, vemolic acid, and y-linolenic acid and oils like jojoba oil that produce wax
esters rather than glycerol esters. The high-oleic oil (>80%) is associated with the desired
reduction in linolenic acid from 8-10% to less than 3%. Other breeding objectives include
increasing seed yield, winter hardiness, and resistance to frost, disease, insects, lodging, and
shattering and an improvement in meal quality.
Bergqvist and Kaufmann [28] used HPLC to demonstrate that the major triacylglycerols in
low-erucic rapeseed oil are LLL (5%), LLO and LnOO (19%), LOO (27%), and OOO (41%).
Takagi and Ando [32] reported a stereospecific analysis of the high-erucic oil (Table 17), and
similar data were obtained by tandem mass spectrometry [33].
L. Rice Bran Oil

Rice {Oryza sativa) is an important cereal with an annual production of about 500 million
tonnes. To produce white rice the hull is removed and the bran layer is abraded, giving about
8-10% of the rice grain. This contains the testa, cross cells, aleurone cells, part of the aleu-
rone layer, and the germ and includes almost all of the oil and the majority of the protein,
ash, vitamins, and fibre of the rice coreopsis. Less than 40% of rice is treated in this way.
This indicates a potential of 15 million tonnes of bran and over 3 million tonnes of rice bran
oil. In fact, present production is only about 0.45 million tonnes per annum, and not all of
this is food grade [34].
Lipases liberated from the testa and the cross cells promote rapid hydrolysis of the oil, and
it should therefore be extracted within a few hours of milling. The more free acid has to be
removed, the more oil will be lost at the same time. Attempts have been made to upgrade rice
bran oil with 30% free acid by reaction with glycerol and lipozyme followed by neutralization.
Refined rice bran oil is an excellent salad and frying oil with high oxidative stability. It
also finds several non-food uses. The oil lowers serum cholesterol by reducing LDL and
VLDL with no change in HLDL. This effect seems not to be related to fatty acid or triacyl-
glycerol composition but to the unsaponifiable fraction and probably to the oryzanols (1.5-
2 .0% of oil), which are ferulic acid esters of sterols and triterpene alcohols.
The major acids in rice bran oil are palmitic (12-18%, typically 16%), oleic (40-50%,
typically 42%), and linoleic (29-42%, typically 37%). The oil contains phospholipids (—5%),
a wax that can be removed and finds commercial use, and unsaponifiable material (3-5%).
40 Gunstone
This last contains sterols (—1.8%), 4-methylsterols (—0.4%), triterpene alcohols (—1.2%),
and less polar compounds (—0 . 8%) such as tocopherols, tocotrienols, and squalene.
M. Safflower Seed Oil

Safflower seed oil is considered a minor oil, for its annual production was about 250,000
tonnes in 1982-1983. Safflower (Carthamus tinctorius) is grown mainly in India and to a
lesser extent in Mexico and the United States. The plant was grown for the dye present in its
leaves long before there was any interest in the seed oil. It is normally a linoleic-rich oil
(—75% linoleic acid) with L 3 (47%), L2O (19%), and L 2S (18%) (L = linoleic, 0 = oleic,
S = saturated acid) as the major triacylglycerols. An oleic-rich safflower oil (—74% oleic acid)
has also been developed. Christie et al. [30] conducted stereospecific analysis with the results
shown in Table 18.
N. Sesame Oil
Sesame (Sesamum indicum), the source of another minor oil (annual production about 660,000
tonnes), is grown mainly in India and China but also in Burma, Sudan, and Mexico. The
seed contains 40-60% oil with almost equal levels of oleic (range 33-54% , average 40%)
and linoleic acid (range 39-59% , average 46%). The oil contains sesamin (0.5-1.1% ) and
sesamolin (0 .3 -0 . 6%) and has very high oxidative stability. The view that this is due to the
production of the phenol sesamol from sesamolin during refining has recently been ques
tioned. Early work [5] indicates that the major triacylglycerols are LOO (22%), LLO (20%),
LOS (17%), LLL (19%), LLS (9%), 0 0 0 (8%), OOS ( 8%), and other (6%).
O- Soybean Oil
Soybeans {Glycine max) have been grown for 4000-5000 years and have become an important
oil crop during the last 50 years even though the oil itself is a by-product in the production
of a valuable protein meal used worldwide as an animal feed. Soy is now the prime source of
oil, though it is likely to be overtaken by the oil palm in the next decade.
About half the world crop is grown in the United States, followed by Brazil (—17%),
Argentina (—10%), and China (—10%). The balance is produced in over 30 countries, and
Italy, Paraguay, India, and Indonesia have recorded significant increases in the last 10-15
Table 18 Stereospecific Analysis of Safflower Seed OiU
16:0 18:0 18:1 18:2
TAG 7.9 2.1 13.7 76.3

Method 1
sn-l 14.2 2.9 14.4 68.6
sn-2 2.9 0.2 15.3 81.5
sn-3 6.8 3.1 11.4 78.7
Method 2
sn-l 16.5 2.7 14.0 6 6 .8
sn-2 0.6 0.4 15.7 83.3
sn-3 9.1 3.0 11.0 76.9
^Two sets of results obtained in slightly different ways.

Source: Ref. 30.
Table 19 Stereospecific Analysis of Soybean TAG
HPLC method Enzyme method

Fatty
acid sn-l sn-l sn-3 sn -l sn-2 sn-3
16:0 16.7 2.2 16.1 16.9 1.4 16.6

18:0 5.4 0.3 4.6 5.2 0.3 4.7
18:1 24.3 23.4 24.6 22.3 23.0 26.9
18:2 46.4 68.1 47.0 47.4 69.2 45.2
18:3 6.4 5.7 7.0 7.5 5.7 5.9
Other 0.8 0.3 0.7 0.7 0.4 0.7
Source: Ref. 32.
years. The product is imported as whole beans, as oil, or as meal, and crushing is carried out
in Japan and the E U countries as well as in the producer countries.
The bean contains high quality protein (38-42%) and oil (1 8 -2 2 % ), which is usually
recovered by hexane extraction. The meal represents 60-65% of the total value of the beans,
and oil, though important, is a by-product.
The oil is used mainly in food—often after partial hydrogenation—as salad oil, cooking
oil, and frying oil and in margarines and shortening. Non-food uses are based mainly on the
high levels of unsaturation and include coatings, epoxidized oil, and dimer acids. Valuable
by-products of the refining process include lecithin (a mixture of phospholipids), which is
widely used because of its surface activity, and vitamin E (tocopherol) used as an antioxidant.
The oil also contains sterols, triterpene alcohols, chlorophyll, and hydrocarbons as minor com
ponents.
Soybean oil is a linoleic-rich oil containing 6-10% a-linolenic acid. This is considered to
be undesirable for most food purposes and has been reduced to 3-4% or even to 2% by seed
breeding methods. The oil with only 3-4% linolenic acid has enhanced oxidative stability and
improved flavor characteristics. A high stearate (15-30%) oil has also been developed in the
hope that it can be used in margarine and shortenings without partial hydrogenation.
Stereospecific analysis has given the results in Table 19, and an older silver ion chromato
graphic procedure indicates the glyceride composition given in Table 20.
P. Sunflower Seed Oil

Sunflowers {Helianthus annus) grow well in temperate zones. They are grown extensively in
the former Soviet Union (29.0%), Argentina (16.6%), EU (17.7% especially France, Spain,
and Italy), China (5.7%), the United States (5.2%), and eastern Europe (9.8%).
The oil represents 80% of the total value of the seed. It is usually 60-75% linoleic acid
and >90% oleic and linoleic acids. In this form it is widely used as a salad oil, as a cooking
Table 20 Major Triacylglycerols (>5% ) in Soybean

Oil by Silver Ion TLC
LeLL 7 LLO 16 LOS 12

LeLO 5 LLS 13 OOS 5
LLL 15 LOO 8 Other 19
Source: Ref. 5.
42 Gunstone
Table 21 Stereospecific Analysis of Sunflower Seed OiP
16:0 18:0 18:1 18:2
TAG 7.2 4.6 21.9 66.3

Method 1
sn-l 10.6 3.3 16.6 69.5
sn-2 1.3 1.1 21.5 76.0
sn-3 9.7 9.2 27.6 53.5
Method 2
sn-l 11.2 4.0 17.7 67.0
sn-2 0.7 0.4 20.4 78.5
sn-3 10.4 9.9 28.7 51.0
^Two sets o f results obtained in slightly different ways.

Source: Ref. 30.
oil, and in the production of margarine and shortening. Its value in these food uses is en
hanced by the virtual absence of linolenic acid. For the same reason it produces nonyellowing
paints. Its triacylglycerol composition is given in Tables 21 and 22.
High-oleic varieties (Sunola, Highsun) with about 85% oleic acid have been developed.
Their oil has increased oxidative stability and finds use as a good source of oleic acid in
enzymatically modified products. Recent developments of sunflower seed and its oil are dis
cussed by Haumann [35] and by Morrison et al. [36].
Q. Tall Oil Fatty Acids

The term tall oil comes from the word tallolja (the Swedish word for pine oil), which is a
mixture of fatty acids with some neutral material (—9%). Tall oil fatty acids are by-products
of the wood pulp industry in which pine wood chips are digested under pressure with an
aqueous mixture of sodium hydroxide and sodium sulfide, which converts the acids to their
sodium salts. Tall oil is produced mainly in North America ( —2 5 0 ,0 0 0 tonnes) and Scandina
via (—90,000 tonnes), but the products from these two sources differ in composition because
of the different wood species that are pulped (Table 23). The crude acid extract is distilled to
give fatty acids (with <2% of resin acids) and resin acids (with <3% of fatty acids). The
fatty acid fraction contains little or no saturated acid (<3% ). It is mainly a mixture of oleic
and linoleic acids (75-80%) with variable amounts of pinolenic acid (5 c 9 c l2 c -1 8 :3 ) and con
jugated diene acids. Also present are C20 homologues of pinolenic acid (5,11,14- and 7,11,14-
20:3) and cyclic acids, probably derived from pinolenic and related acids [37]. An example
is cyclopinolenic acid, whose structure is shown here.
Table 22 Major Triacylglycerols

(> 5%) in Sunflower Seed Oil by
Ag+ TLC
LLL 14 LOO 19
LLO 39 LOS 11
LLS 14 Other 3
Source: Ref. 5.
Table 23 Fatty Acid Composition of Tall Oil
Other ^
16:0 18:0 18:1 18:2 18:3 A B C D
American 0.5 2 46 36 0 2 9 0 .5 -2 0.5 -3

Scandinavian 0.5 2 30 45 1 9 5 0 .5 -2 0 .5 -3
Â, pinolenic acid; B, conjugated acids; C, resin acids; D , unsaponifiable.

Source: Ref. 37.
cyclopinoleiiic acid
Tall oil fatty acids contain some sulfur compounds that interfere with catalytic processes.
They are therefore not usually converted to alcohols or to nitrogen-containing compounds but
are used to produce dimer acids, alkyds, and coatings, detergents, and lubricants. New uses
being investigated include employment as solvents, in inks, and as biodiesel oils.
R. Tung Oil (China Wood Oil)

Tung oil is a minor seed oil (~50,000 tonnes in 1990) of the tung tree {Aleurites fordii and
A. montana) grown in China and the United States. It is characterized by a high content of
a-eleostearic acid (—70%, 9c,llr,13i-18:3), which contains conjugated triene unsaturation. It
dries (hardens through oxidative polymerization) more readily than linseed oil, which contains
a-linolenic acid (a nonconjugated Cjg triene acid and an isomer of eleostearic acid), and is
used in enamels and varnishes. The acid and its glycerol ester are easily isomerized to the all-
trans isomer (j8-eleostearic acid), which has a higher melting point.
Oiticica oil and kamala oil contain other conjugated triene acids, and these and some others
of this type are discussed in Chapter 1, Section VII, page 7.
S. Wheatgerm Oil
Wheatgerm oil is extracted from the bran and germ of the wheat grain {Triticum gestivum)
(see discussion of rice bran oil). The oil is rich in linoleic acid (60-65%), but its value lies
more in the high level of tocopherol [38]. It is one of the richest sources of vitamin E (see
Table 9).
T. Vernonia Oil
Vernonia galamensis seed oil and some other seed oils (see Chapter 1, Section XI.B, page
10 .) are characterized by their high levels of vemolic acid {cis-12,13-epoxyoleic acid), in the
range 50-80% [5]. This acid with multiple functionality is of considerable potential though it
has not yet been commercialized.
44 Gunstone
U. Waxes
Waxes are lipidlike materials of animal or vegetable origin. (Hydrocarbon waxes from mineral
sources are not included in this account.) The true ester waxes are esters of long-chain acids
and long-chain alcohols, and these are often, but not always, the major compounds of the
natural waxes. Also present are free acids, free alcohols, hydrocarbons, and other minor com
pounds.
The major waxes available commercially include beeswax and wool wax from animal
sources and camauba (Copernicia cerifera), candelilla {Euphorbia antisiphilitica, E. cerifera,
and Pedilanthus pavenis), ouricouri {Syagrus coronata. Cocos coronata, and A ttalea excelsa),
and jojoba oil {Simmondsia chinensis) from plant sources. Some indication of the composition
and properties of these waxes is given in Table 24. Fuller details are to be found in a review
by Hamilton [39], and Barnett [49] reported on lanolin and its many derivatives.
¥e Butter (Cow’s Milk) Fat

Practically all butter (~ 7 million tonnes/annum) is made from cow’s milk fat, though some
other milk fats such as those from goats and camels are in local demand. Butter is produced
throughout the world .and is used almost entirely for food purposes—mainly as a spread but
also for baking and frying. Butter fat can be fractionated to give harder and softer fractions
that find wide food uses [40,41]. Butterfat changes in composition between summer and win
ter, presumably reflecting different dietary intakes. It can also be modified by manipulation
of the animal’s diet [42].
Whole milk lipids are mainly triacylglycerols (97-98% ) with smaller amounts of free acids
(0.1-0.4), cholesterol (0.2-0.4% ), diacylglycerols (0 .3 -0 . 6%), and phospholipids (0.2-
1.0%). The latter are mainly phosphatidylcholines (35%), phosphatidylethanolamines (32%),
and sphingomyelin (25%) along with lower levels of phosphatidylserine (3%) and phosphati
dylinositol (5%). Butterfat is very complex in its fatty acid and triacylglycerol composition.
It contains over 15 major acids, but over 500 components have been identified. These include,
in addition to the usual range of the C 15 and Cjg acids, short- and medium-chain acids (C4-
C 14), a range of trans monoenes that are mainly 18:1 isomers with vaccenic acid predo-
Table 24 Composition and Properties of Some Commercial Waxes
Beeswax Wool wax Camauba Candelilla Ouricouri Jojoba
Composition
Ester waxes 70-80^ 84-85 28-29 -100
Acids 12-15 3 -4 7 -9 —
Alcohols — 2 -3 12-14’^
Hydrocarbons 10-15 2 -3 50-51 —
Property
mp (°C) 62-65 35-42 78-85 67-69 82-84 -7
Acid value 17-36 7-15 3-10 13-18 8-2 0 2
Sap. value 90-149 100-110 79-95 35-87 70-100 92
Ester value 64-84 85-100 — — 75-85 —
Iodine value 7-16 15-30 7-14 14-27 6-8 82
^Mainly C40 - C 50 .
‘’Including sterols.
Source: Ref. 39.
Table 25 Major Fatty Acids (wt %) in Summer and Winter

Milk Fats
Fatty acid June December Average
4:0 4.2 3.5 3.6

6:0 2.5 2.2 2.2
8:0 2.3 1.1 1.2
10:0 2.2 2.6 2.8
12:0 2.4 2.8 2.8
14:0 9.0 10.6 10 .1
15:0 1.3 1.1 1.1
16:0 22.1 26.0 25.0
17:0 0.7 1.1 1.0
18:0 14.3 11.6 12 .1
18:1 30.4 24.8 27.1
18:2 1.2 2.8 2.4
18:3 and 20.0 2.6 2.3 2.1
Other 4.8 7.5 6.5
Source: Ref. 43.
minating ( H i - 18:1), hydroxy acids that give rise to the y- and 6-lactones (important flav o r
components), and mono- and multimethyl branched acids. Quantitative information on the
major acids and triacylglycerols is given in Tables 25 and 26.
W- Lard and Tallow

Lard and tallow [45] are by-products of the rendering industry. They are rich in palmitic,
stearic, and oleic acids but contain many minor components such as odd-chain acids,
branched-chain acids, and trans acids. They contain high levels of cholesterol (lard 0.37-
0 .4 2 % , beef tallow 0.08-0.14% , and mutton tallow 0.23-0.31% ) and lower levels of natural
antioxidants than are usually found in vegetable oils. In the tallows, unsaturated acids concen
trate in the sn-2 positions and saturated acids in the sn-l and sn-3 positions. This distribution
pattern is different in lard, where, unusually, palmitic acid predominates in the sn-2 position.
Tables 27-30 present data on the composition of lard and tallow.
X. Fish Oil
The total catch of fish has risen from around 20,000 tonnes in 1950 to almost 100,000 tonnes
at the present time. This includes shellfish and fish from both marine and freshwater sources.
Table 26 Triacylglycerols (%) by Carbon Number

in Australian Milk Fat
26 0.6 36 10.5 46 6.4

28 0.5 38 13 .1 48 8.4
30 1.0 40 10.5 50 11.7
32 2.3 42 6.3 52 11.3
34 5.5 44 5.6 54 6.0
Source: Ref. 44.

46 Gunstone
Table 27 Major Fatty Acids in Lard and Beef and Mutton Tallows
14:0 16:0 16:1 18:0 18:1 18:2
Typical values
Lard 2 .0 27.1 4.0 11.0 44.4 11.4
Beef 2.5 27.0 10.8 7.4 47.5 1.7
Mutton 5.6 27.0 1.6 31.7 31.7 1.6
Range
Lard 0.5-2.5 20-32 1.7-5.0 5-24 35-62 3-16
Beef 1.4-6.3 20-37 6 .7 -8 .8 6-40 26-50 0 .5 -5
Mutton 5.5 25.8 1.5 30.5 30 1.4
Source: Ref. 5.
Table 28 Major Triacylglycerols in Lard
PPSt 2.0 POO 5.2 ooo 11.7

StPSt 2 .0 StOO 6.1 OPL 7.2
PPO 13 OPO 18.4 Other 24.6
StPO 12.8 StPL 2.1
Source: Ref. 5.
Table 29 Major Triacylglycerols in Beef Tallow
PPSt 3.6 PPO 3.9 OPO 2.1

PStSt 3.2 StPO 3.6 trans 2.3'=
POP 5.8 trans 4.5^ OOO 4.9
p o st 10.5 POO 12.1 other 32.8
StOSt 4.8 StOO 5.9
Ône double bond.

‘’Two double bonds.
Source: Ref. 5.
Table 30 Stereospecific Analysis of Tallow
TAG sn-\ sn-2 sn-3
14:0 3.4 0 .2 4.3 2.9

16:0 29.5 45.7 21.0 25.5
18:0 26.0 33.1 12.5 31.2
18:1 33.1 16.6 52.0 33.7
Other 8 .0 4.4 10.2 6.7
Source: Ref. 30.
The amount of fish produced by marine and freshwater aquacultures, now about 14,000
tonnes, has doubled during the last ten years. About one-third of the total catch, especially
the small shoaling pelagic species, is used for industrial purposes, i.e., for direct conversion
to fish oil and fish meal. Fish meal is the cheapest source of animal protein and is in high
demand [50].
It has been reported that 1 tonne of anchovies will produce about 210 kg of meal (still
containing about 10% oil) and 30 kg of oil. The same tonnage of anchovy is the source of
about 11 kg of EPA and DHA, of which about 7.5 kg is in the oil and 3.5 kg in the meal.
Of all the fish oil produced, about two-thirds is used in food products and one-third for animal
feed. This latter proportion is rising, particularly to feed farmed fish.
Fish oil and fish meal are the most convenient sources of n-3 acids. The meal is used to
feed poultry, pigs, and fish, and the n-3 acids are passed into the human food chain through
these products. Traditionally fish oil has been used only after partial hydrogenation. This is
necessary if the oil is to be used as a component of fat spreads. Partial hydrogenation gives a
product of increased oxidative stability with the required melting behavior but with a reduction
in nutritional value. The n-3 polyene acids are converted to acids of lower unsaturation with
some double bonds of trans configuration. It has been demonstrated that with appropriate
refining procedures the fish oils (without hydrogenation) can be incorporated into a range of
fat-based products. Small amounts are also incorporated into infant formulas and supplied
as health foods in capsule form. Products with enhanced levels of EPA and/or DHA are
being developed.
Fish oils make only a minor contribution (—2%) to the total production of oils and fats.
They come mainly from fish such as herring, capelin, sand eel, menhaden, sardine, and an
chovy.
Fish oils generally contain a wide range of fatty acids ranging in chain length from Cj 4 to
C26 with zero to six double bonds. Their major acids are generally saturated (14:0, 16:0, and
18:0), monounsaturated (16:1, 18:1, 20:1, and 22:1), and members of the n-3 polyene family
(18:4, 20:5, and 22:6). There are also many other acids at low levels, including branched-
chain and cyclic acids. Their typical fatty acid composition is given in Table 31.
ABBREVIATIONS
D Docosenoic acid
EU European Union countries [12]
FSU Former Soviet Union
G y-Linolenic acid or acyl group
GLA y-Linolenic acid or acyl group
L Linoleic acid or acyl group
Table 31 Component Acids of Fish Oils (Typical results, wt %)
14:0 16:0 16:1 18:0 18:1 18:2^ 18:4‘> 2 0 :1 20:5 2 2 :1 22:5'’ 22:6'^
Herring‘S 9 15 7 1 16 1 — 16 3 23 — 3
Menhaden 8 22 11 3 21 2 — 2 14 2 — 10
Capelin 8 9 16 1 8 1 1 17 9 20 1 3
Anchovy 8 20 9 3 15 1 2 3 18 2 1 11
Sardine‘s 8 16 9 4 11 1 2 3 17 4 3 13
Cod (liver) 4 14 12 3 22 1 1 12 7 11 trace 7
"n-6.
^n-3.
‘^Nova Scotia.
‘^California.
48 Gunstone
Ln a-Linolenic acid or acyl group

0 Oleic acid or acyl group
P Palmitic acid or acyl group
PC Phosphatidylcholines
PE Phosphatidylethanolamines
PI Phosphatidylinositides
R Ricinoleic acid or acyl group
S Saturated acid or acyl group
St Stearic acid or acyl group
TAG Triacylglycerols
REFERENCES
1. T. Mielke (Ed.), Oil World Annual 1991 and later volumes, ISTA Mielke GmbH, Hamburg,
Germany, 1991.
2. BSi Standards, Nomenclature for Oilseeds, BS 6406: 1993, ISO 5507; 1992, (1993).
3. T. Mielke, reported by editorial staff in INFORM 5: 715 (1994).
4. B. Howard, Oils and Oilseeds to 1996, The Economist Intelligence Unit, Special Report No.
M.703, London, 1993.
5. F. D. Gunstone, J. L. Harwood, and F. B. Padley (Eds.), The Lipid Handbook, 2nd ed.. Chapman
and Hall, London, 1994.
6. F. D. Gunstone, Fatty Acid and Lipid Chemistry, Blackie, London, 1996.
7. B. S. Kamel and Y. Kakuda (Eds.), Technological Advances in Improved and Alternative Sources
of Lipids, Blackie, London, 1994.
8. J. B. Rossell and J. L. R. Pritchard (Eds.), Analysis of Oilseeds, Fats, and Fatty Foods, Elsevier,
London, 1991.
9. T. J. Walton, Waxes, cutin and suberin, in Methods of Plant Biochemistry, Vol. 4 (J. L. Harwood
and J. R. Bowyer, Eds.), Academic, London, 1990, p. 105.
10. K. T. Achaya, B. M. Craig, and C. G. Youngs, The component fatty acids and glycerides of
castor oil, J. Am. Oil Chem. Soc. 41: 783 (1964).
11. L. R. Hogge, D. C. Taylor, D. W. Reed, and E. W. Underhill, Characterization of castor bean
neutral lipids by mass spectrometry/mass spectrometry, J. Am. Oil Chem. Soc. 68: 863 (1991).
12. V. K. S. Shukla, Cocoa butter, properties and quality. Lipid Technol. 7: 54 (1995).
13. V. K. S. Shukla, Confectionery fats, in Developments in Oils and Fats (R. J. Hamilton, Ed.)
Blackie, London, 1995, p. 6 6 .
14. I. Hachiya, T. Koyana, and K. Sato, Fat polymorphism and chocolate crystallisation. Lipid Tech
nol. 2: 34 (1990).
15. J. B. Rossell, Vegetable fats for chocolate, couyertures and coatings. Lipid Technol. 4: 106 (1992).
16. I. A. de la Roche, E. J. Weber, and D. E. Alexander, Effects of fatty acid concentration and
positional specificity on maize triglyceride structure. Lipids 6: 531 (1971).
17. G. Sempore and J. Bezard, Determination of molecular species of oil triacylglycerols by reversed-
phase and chiral-phase high performance liquid chromatography, J. Am. Oil Chem. Soc. 68: 702
(1991).
18. J. M. Bland, E. J. Conkerton, and G. Abraham, Triacylglyceride composition of cottonseed oil
by HPLC and GC, J. Am. Oil Chem. Soc. 68: 840 (1991).
19. D. F. Horrobin, Nutritional and medical importance of gamma-linolenic acid. Prog. Lipid Res.
31: 163 (1992).
20. F. D. Gunstone, Gamma linolenic acid—occurrence and physical and chemical properties. Prog.
Lipid Res. 31: 145 (1992).
21 . J. A. Singleton and H. E. Pattee, Characterisation of peanut oil triacylglycerols by HPLC, GLC,

and EIMS, /. Am. Oil Chem. Soc. 64: 534 (1987).
22 . T. Takagi, Chromatographic resolution of chiral lipid derivatives, Prog. Lipid Res. 29: 111
(1990).
23. M. H. J. Bergqvist and P. Kaufmann, A multivariate optimization of triacylglycerol analysis by
high-performance liquid chromatography, Lipids 28: 667 (1993).
24. F. Santinelli, P. Damiani, and W. W. Christie, The triacylglycerol structure of olive oil deter
mined by silver ion high-performance liquid chromatography in combination with stereospecific
analysis, J. Am. Oil Chem. Soc. 69: 552 (1992). (See also Ref. 30.)
25. R. V. Flor, L. T. Hecking, and B. D. Martin, Development of high performance liquid chroma
tography criteria for determination of grades of commercial olive oils. Part 1. The normal ranges
for the triacylglycerols, J. Am. Oil Chem. Soc. 70: 199 (1993).
26. W. Hamm, Trends in edible oil fractionation. Trends Food Sci. Technol. 6: 121 (1995).
27. S. Takano and Y. Kondoh, Triglyceride analyses by combined argentation/nonaqueous reversed
phase high performance liquid chromatography, J. Am. Oil Chem. Soc. 64: 380 (1987).
28. M. H. Bergqvist and P. Kaufmann, A multivariate optimization of triacylglycerol analysis by
high-performance liquid chromatography. Lipids 28: 667 (1993).
29. G. J. Sassano and B. S. F. Jeffrey, Gas chromatography of triacylglycerols in palm oil fractions
with medium-polarity wide-bore columns, J. Am. Oil Chem. Soc. 70: 1111 (1993).
30. W. W. Christie, B. Nikolova-Damyanova, P. Laakso, and B. Herslof, Stereospecific analysis of
triacyl-5/2-glycerols via resolution of diastereomeric diacylglycérol derivatives by high performance
liquid chromatography on silica, J. Am. Oil Chem. Soc. 68: 695 (1991).
31. C. Leonard, Sources and commercial applications of high-erucic vegetable oils. Lipid Technol. 6:
79 (1994).
32. T. Takagi and Y. Ando, Stereospecific analysis of triacylglycerols by chiral high-performance
liquid chromatography, Lipids 26: 542 (1991).
33. H. Kallio and G. Currie, Analysis of low erucic acid turnip rapeseed oil (Brassica campestris) by
negative ion chemical ionization tandem mass spectrometry. A method giving information on the
fatty acid composition in positions sn-2 and sn-H2 of triacylglycerols. Lipids 28: 207 (1993).
34. R. N. Sayre and R. M. Saunders, Rice bran and rice bran oil. Lipid Technol. 2: 72 (1990).
35. B. F. Haumann, Modified oil may be key to sunflower’s future, INFORM 5: 1198 (1994).
36. W. H. Morrison, R. J. Hamilton, and C. Kalu, Sunflower seed oil, in Developments in Oils and
Fats (R. J. Hamilton, Ed.), Blackie, London, 1995, p. 132.
37. A. Hase and S. Pajakkala, Tall oil a fatty acid source. Lipid Technol. 6: 110 (1994).
38. P. J. Barnes, Lipid composition of wheat germ and wheat germ oil, Fette Seifen Anstrichm. 84:
256 (1982).
39. R. J. Hamilton, Commercial waxes: their composition and applications, in Waxes: Chemistry,
Molecular Biology and Functions (R. J. Hamilton, Ed.), The Oily Press, Dundee, Scotland, 1995,
p. 257.
40. L. Eyres, Milk fat product development. Lipid Technol. 1: 12 (1959).
41. W. de Greyt and A. Huyghebaert, Food and non-food applications of milk fat. Lipid Technol. 5:
138 (1993).
42. W. Charteris, Monobutter: a fridge-spreadable butter. Lipid Technol. 3: 90 (1991).
43. R. G. Jensen, G. W. Gander, and J. Sampugna, Fatty acid composition of the lipids from pooled,
raw milk, J. Dairy Sci. 45: 329 (1962).
44. R. E. Timms, Detection and quantification of non-milk fat in mixtures of milk fat and non-milk
fat, J. Dairy Res. 47: 295 (1980).
45. R. R. Grummer, Production and consumption of animal by-products from the rendering industry.
Lipid Technol. 5: 114 (1993).
46. The British Nutrition Foundation Report, Unsaturated Fatty Acids: Nutritional and Physiological
Significance, Chapman and Hall, London, 1992.
47. Y. S. Huang, X. Lin, P. R. Redden, and D. F. Horrobin, In vitro hydrolysis of natural and
50 Gunstone
synthetic y-linoleic acid containing triacylglycerols by pancreatic lipase, J. Am. Oil Chem. Soc.
72: 625 (1995).
48. P. Manninen, P. Laakso, and M. Kallio, Separation of y- and a-linolenic-acid-containing triacyl
glycerols by capillary supercritical fluid chromatography, Lipids 30: 665 (1995).
49. G. Barnett, Lanolin and derivatives, Cosmet. Toiletries 101: 21 (1986).
50. R. J. Hamilton and R. D. Rice (Eds.), Fish Oil, Technology, Nutrition, and Marketing, P. J.
Barnes and Associates, High Wycombe, UK, 1995.
Phospholipids
Michael Schneider
Lucas Meyer GmbH & Co., Hamburg, Germany
I- TERMINOLOGY
In the scientific literature, rational nomenclature according to lUPAC or abbreviated lUPAC
nomenclature (e.g., 1,2-diacyl-i-n-glycero-S-phosphocholine or phosphatidylcholine) [1] and
trivial terms such as lecithin (for PC) and cephalin (for PE or PI/PA) still coexist. In industrial
and commercial understanding, the term lecithin denotes a complex mixture of neutral lipids
(predominantly triglycerides), polar lipids (phospho- and glycolipids), and carbohydrates. This
blend is described as a generally permitted food additive “lecithin” in Europe under E 322 [2]
and in the United States under the Code of Federal Regulations [3]. The two descriptions
differ to a minor extent in their specification details but not in principle. As this book is
about industrial application and technology, industrial terminology will be used throughout
this chapter.
II. OCCURRENCE
Two main classes of phospholipids occur naturally in amounts that have led to industrial
applications: phosphoglycerides and phosphosphingolipids. Table 1 gives the most commonly
used terms and abbreviations for the phospholipids found in materials used for commercial
purposes.
As phospholipids are integral structure elements of all kinds of membranes in every living
organism, they can be obtained from all types of biomass. Commercial sources, however, are
predominantly vegetable oil seeds (soybeans, rapeseed, sunflower seed). For pharmaceutical,
cosmetic, and some dietary applications, egg yolk, milk, and brain have gained some impor
tance.
Vegetable materials usually contain only small amounts of phospholipids, ranging from 0.3
to 2.5wt% maximum (on dry basis) [5-7]. Animal materials have substantially higher phos
pholipid co n ten ts: e g g , 14% [8]; b ra in , 6% [9]; an d w h o le m ilk , 2% [10].
51
52 Schneider
Table 1 Phospholipid Classes
Class Abbreviation
Phosphoglycerides
1. Ester phospholipids ^
Phosphatidylcholine PC
Phosphatidylethanolamine PE
Phosphatidylserine PS
Phosphatidylinositol PI
Phosphatidylglycerol PG
Diphosphatidylglycerol DPG
Phosphatidic acid PA
A-Acyl-phosphatidylethanolamine NAPE
2. Ether phospholipids
Choline plasmalogen
Ethanolamine plasmalogen
Phosphosphingolipids
Ceramide phosphocholine (sphingomyelin) SPM
Glycosylated inositolphosphoceramides
Phytoglycolipids PGL
Âlong with the listed diester species, small amounts of monoacyl types, the
so-called lysophospholipids, are found in all natural sources.
Source: Ref. 4.
III. COMPOSITION
The various methods of obtaining commercial phospholipid products are described later. All
production methods are based on complex mixtures of various species of phospholipids that
are by-products of vegetable oil processing or of extraction from wet or dried animal sub-
Table 2 Phospholipid Composition [wt%] as Extracted from Commercial Raw Material
Sunflower Brain ^
Soy^ Rapeseed^ seed^ Com^ Egg Milk*’ (bovine)
PC 21.9 24.6 25.4 30.4 74 27.0 32.4

PE 13.6 22.1 11.0 3.2 19 .1 36.4 23.5
PI 12.0 14.7 19.4 16.3 0.4 4.3
PS — 0.8 1.0 — 11.0
PG/DPG 2.3 1.2 1.4 0.6 2.1
PA 5.8 3.3 9.4 — 0.9
NAPE 2.8 1.0 2.6 —
SPM — 2.5 29.0 20.4

Lyso-PLs 2.9 19.4 5.4 2.9 3.0
Others 3.6 19.2 0.5 7.6
Ref. 11 12 13 14 11 10 11
^Figures are based on the acetone-insoluble constituents of commercial lecithins, which also contain certain amounts
of glycolipids.
‘’The data consider only phosphorus-containing polar lipids.
Phospholipids 53
Table 3 Fatty Acid Composition [wt%] of Phospholipid Mixtures
Sunflower
Soy Rapeseed seed Com Egg Brain
16:0 21.4 18.3 14.8 22.8 30 1

18:0 3.8 0.6 3.0 1.5 16 41
18:1 12.0 21.3 13.3 26.5 29 31
18:2 57.0 47.9 68.7 48.5 14 —
18:3 5.8 7.4 0.7 1 —
20:4 5 3
22:6 3 8
Others 2 16
Ref. 14 15 13 14 11 11
stances. In addition, vegetable extraction products contain smaller or larger amounts of phos
phorus-free polar lipids (glycolipids) and sometimes even carbohydrates. The phospholipid
composition is a “fingerprint” typical of the raw material being extracted, (Table 2), and so
also is the composition of the fatty acids linked to the phospholipid backbone (Table 3). The
data in Table 3 reflect the composition of the natural phospholipid mixtures. The fatty acid
composition of isolated phospholipids can differ considerably from the listed data. For more
details, see Ref. 15.
IV, OBTAINING PHOSPHOLIPID PRODUCTS

Methods for obtaining phospholipids on an industrial scale are completely different for vegeta
ble-derived products than for animal-derived products. Vegetable phospholipids are manufac
tured exclusively from by-products of the vegetable oil refining process. For animal phospho
lipids, specific extraction processes have been developed and are applied commercially.
A, Vegetable Phospholipids
Today the exclusive sources of vegetable phospholipids are commercial lecithins obtained in
the first step of vegetable oil refining during the degumming process. To stabilize vegetable
oils against sedimentation and off-taste development and also to enable further refining steps,
phospho- and glycolipids must be removed. During the standard batch degumming process
[16], 2% of water by volume is added to crude oil heated to approximately 70°C and agitated
for V2- I h. The addition of water causes the polar lipids contained in the oil to hydrate,
leaving them oil-insoluble. Centrifugation of the hydrated gums leads to a mixture of water,
phospho- and glycolipids, and also some triglycerides, free carbohydrates, traces of sterols,
free fatty acids, and carotenoids. Careful drying of the wet gums immediately after removal
from the crude oil leads to crude vegetable lecithin.
The literature reports dozens of variations of the degumming process [16], all with the aim
of reducing the amount of phosphorus remaining in the oil to a minimum. Degumming condi
tions have an influence on crude vegetable lecithin composition and quality.
Table 4 shows a typical composition of soybean lecithin [14]. Traditionally, crude vegeta
ble lecithins are the starting materials of choice for further fractionation and purification pro
cesses to obtain phospholipid compositions suitable for various industrial applications.
54 Schneider
Table 4 Typical
Composition of Crude
Soybean Lecithin (%)
Triglycerides 34.2
Diglycerides 0.4
Free fatty acids 0.4
Sterols 0 .8
Glycolipids 6.5
Carbohydrates 6.7
Phospholipids 50.0
B. Animal Phospholipids
In contrast to vegetable products, there is no comparable industrial animal oil processing
technology, and therefore phospholipids have to be specifically extracted using various tech
nologies. Since egg phospholipids are the sole products of commercial interest, the processing
alternatives are those shown in Fig. 1.
Accione (17)
Fig. 1 Egg yolk phospholipid processing principles.
Methods for milk phospholipid extraction are numerous [17-22]. Some of them describe
the use of solvent mixtures to use partitioning [18,19] or a combination of extraction and
adsorption chromatography [21] or extraction with chloroform-methanol mixtures [22]. Ultra
filtration processes have also been described.
Brain phospholipid extraction is usually based on modifications of the Folch procedure
[23,24]. Removal of nonlipidic impurities is a lengthy process that requires further solvent
extraction or gel chromatography. A modification of the extraction process that does not re
quire extraction or adsorption steps is described in Ref. 9.
V. INDUSTRIAL PHOSPHOLIPID PROCESSING

As this book is about industrial applications and technologies, only processes used on a com
mercial scale are described here.
Industrial processing can be split into
Phospholipids 55
1. Solvent fractionation
2. Gas extraction
3. Chromatographic purification
4. Chemical modification
5. Enzymatic modification
A. Solvent Fractionation
Commercial solvent technologies are limited to a small selection of solvents that are appro
priate to the image and use of a natural functional ingredient. These solvents are acetone,
ethanol, compressed gases, and others (e.g., hexane, methanol, isopropanol). For many in
dustrial applications, crude phospholipid products obtained from vegetable oil refining (crude
lecithins) can be used directly. During the last 30-40 years, however, solvent fractionation
has widened the range of commercial phospholipid products.
Table 5 gives an overview of today’s industrial consumption of some phospholipid pro
ducts. (Explanations of compositional details are given in the sections describing process de
tails.)
Table 5 Commercial Phospholipid Products and Their Predominant Use
Product Application^
Soybean lecithin AF/FO /TECO /PH/DI

Deoiled soybean lecithin AF/FO /TI/CO /PH/DI
Ethanol-soluble fractions FO/CO/PH/DI
Ethanol-insoluble fractions AF/FO
Deoiled soybean lecithin fractions (PC 50-80%) FO/CO/PH/DI
Purified soybean PC (>90%) PH/CO
Sunflower seed lecithin AF/FO
Rapeseed lecithin AF/FO /TI
Com lecithin A F/TI
Egg phospholipids (deoiled) FO/DFCO
Egg lipid mixes (PL-containing) DI/PH
Milk phospholipids CO/PH
Brain phospholipids PH/DI
ÂF, Animal feed; CO, cosmetics; DI, dietetic; FO, Food; PH, pharmaceutical; TI, techni
cal industry.
B. Acetone Deoiling
Vegetable lecithins contain approximately 30-40% neutral lipids, predominantly triglycerides.
To improve the handling of the highly viscous crude lecithins and to improve dispersibility
properties, industry makes use of the fact that polar (phospho- and glyco-) lipids are almost
insoluble in acetone whereas neutral lipids dissolve.
Acetone extraction leads to powdered or granulated products that contain 2-3% residual
neutral lipids. The products are slightly hygroscopic but flow freely after addition of flow
agents such as tricalcium phosphate or silica.
Because of the removal of natural tocopherols (—1000 ppm in crude soybean lecithin), the
shelf life of deoiled products is usually limited to approximately 12 months. Table 6 gives the
typical composition of deoiled soybean lecithin.
56 Schneider
Table 6 Typical Composition of a Commercially Deoiled

Soybean Lecithin
Component Percent content
Phospholipids 80
Phosphatidylcholine 24
Phosphatidylethanolamine 22
Phosphatidylinositol 14
Phosphatidic Acid 7
A-Acyl-phosphatidylethanolamine 6
Lysophospholipids 3
Glycolipids 12
Neutral lipids 2.5
Triglycerides 2
Free fatty acids 0.5
Sterols
Moisture
C. Compressed Gas Deoiling

Since the introduction of supercritical gas extraction on an industrial scale, which was first
used for decaffeination of coffee and tea [25], a number of publications and patents have
described the deoiling of lecithin [26-28]. Supercritical carbon dioxide at temperatures of
~40°C and at a pressure of 300-700 bar shows solvent characteristics like those of liquid
acetone—it dissolves neutral lipids but leaves behind the polar lipids. Advantages of such a
process are the mild temperature treatment, no oxidation during drying, no solvent residues,
and no flammability risk.
However, owing to the powder structure of deoiled lecithin, only batch processes have
been described so far—extracting the crude lecithin either in bulk or after nebulizing [29].
The major economic drawbacks of carbon dioxide extraction—batch processing and poor oil
carrying capacity—have been overcome with the development of countercurrent extraction
processes using propane-carbon dioxide mixtures at 80 bar [30]. In cases where slightly
reduced selectivity in splitting between neutral and polar lipids can be accepted, compressed
propane at —40 bar is an excellent alternative [31]. However, at present there is no commer
cial vegetable phospholipid product on the market that is being processed by compressed
gas technology.
Processing methods used for vegetable lipid mixtures have also been applied to egg lipid
blends [32]. Following the trend toward low cholesterol, low fat products, one can remove
fat and cholesterol (neutral lipids) from dried egg yolk using either compressed carbon dioxide
or propane or mixtures of the two. In the United States there is only one consumer product
that has been prepared in this manner.
D. Alcohol Fractionation
Large quantities of vegetable lecithin “fractions” are produced mainly with ethanol or etha
nol-water systems. Commercial interest in these fractions is due to the difference in their
technological functionality [soluble fraction oil-in-water (o/w) emulsifier; insoluble fraction
water-in-oil (w/o) emulsifier]. The interest of the industry in PC-enriched products for dietary
purposes is also a driving force.
Table 7 shows phosphatidylcholine concentrations after fractionation of both a crude and a
deoiled soybean lecithin [33].
Phospholipids 57
Table 7 Alcohol Fractionation of Soybean Lecithin

Triglycerides (%) Phosphatidylcholine (%) Yield (%)
Crude lecithin 39 15
Soluble fraction
Methanol 25 39 23
Ethanol (100%) 60 22 26
Ethanol (95%) 48 33 15
Ethanol (90%) 23 46 12
Isopropanol 54 19 49
Insoluble fraction
Ethanol (95%) 38 7 85
Deoiled lecithin 2.5 24
Soluble fraction
Ethanol (100%) 2.5 52 28
Insoluble fraction
Ethanol (100%) 2.5 12 72
What is described for soybean lecithin can in principle be applied also to other crude
vegetable phospholipid products (lecithins). However, although the results are similar, the
other lecithins are not used commercially.
E. Chromatographic Purification
A review of chromatographic purification is given in Ref. 33. Generally speaking, chromato
graphic separation is the method of choice for the preparation of polar lipids of high purity.
It is almost impossible to give a brief account of the numerous methods published in the
scientific and patent literature. Most of them are used for the quantitative analysis of tissue
extracts and less for the production of commercial quantities of glycolipids or phospholipids.
In this chapter the focus is on a few short descriptions only.
The best results in separation are usually achieved by combining different separation tech
niques—for example, solvent treatment of the crude lipid mixture and adsorption processes.
Among the long list of adsorbents [aluminum oxide, silicic acid, Florisil (magnesium silicate),
DEAE (diethylaminoethyl) cellulose, and ion-exchange resins], alumina and silica play an
outstanding role. Both materials are used especially for low and medium pressure chromatog
raphy to produce large quantities of soybean and egg PC products.
A typical alumina separation is as follows. Deoiled soybean lecithin is fractionated first by
treatment with ethanol to produce a soluble part with about 50% PC. This solution is then
passed down a column filled with aluminum oxide. Nearly all of the PC-accompanying phos
pholipids are bound to the alumina. The effluent consists of almost pure choline phospholipids.
One drawback of this method is the impossibility of removing lysophosphatidylcholine (LPC)
from the material. In fact, additional LPC is formed during the chromatographic process.
Attempts to purify egg phospholipids show that in an ethanolic solution sphingomyelin has
the same elution behavior as PC and is therefore included in the PC/LPC effluent.
The separation of highly purified PC as well as other phospholipids is possible using step
wise elution with a chloroform-methanol-water mixture of various compositions.
Table 8 lists some chromatographic systems used for purification of phospholipids.
In comparison with alumina separation, silicic acid chromatography seems to be more
flexible with respect to solvent systems, separation characteristics, and reconditioning of the
58 Schneider
Table 8 Chromatographic Systems for

Phospholipid Purification
Adsorbent Solvent system
Aluminum oxide Ethanol

Chloroform-methanol
Silica gel Hexane
Ethanol
Hexane-isopropanol-water
Chloroform-methanol
Acetone-chloroform-methanol
Chloroform-acetone
Florisil Chloroform-methanol
Amberlite Ethanol
Ethyl acetate-ethanol
DEAE-cellulose Chloroform-methanol-acetic acid
Source: Ref. 33.
column. Silica is also used for selective elution processes to obtain the different polar lipid
classes of a given lecithin. The preferred solvent systems again are chloroform-methanol and
methanol-acetone mixtures. Separation is mainly done stepwise but in some cases by gradient
elution. For more details about the wide range of possibilities for chromatographic separation,
refer to the literature on this subject [33].
As already mentioned, chromatography on an industrial scale today is mainly focused on
the preparation of PC of natural origin (egg, soy), which is of great commercial interest and
available in ton quantities. Preparation of the natural phospholipids such as PS and PI is of
growing interest. The separation technique of choice is large-scale HPLC, which is also used
in the production of semisynthetic and synthetic phospholipids.
Pure PC from a chromatographic process is still a mixture of various molecular species
differing in fatty acid composition. This is demonstrated impressively in Table 9, which gives
an example of the long list of molecular species of egg PC. However, so far the literature has
given little indication of the possibility of separating phospholipids on the basis of the degree
of unsaturation. One possibility for analytical applications is to use argentation chromatogra
phy, which uses silver nitrate impregnated silica as an adsorbent. The higher the degree of
unsaturation, the more firm ly the phospholipids are bound to the stationary phase and thus
separated from each other. A recent development in this area is the use of reverse-phase
HPLC. However, it is not known whether these processes are already used commercially. It
seems to make more sense to synthesize the appropriate molecule.
Table 9 Molecular Species of Egg Yolk Phosphatidylcholines
Molecular species Percent Molecular species Percent
16:0, 16:1 0.98 18:0, 18:2 11.2 2

16:0, 18:1 38.20 18 :1, 18:2 1.50
18:0, 18:1 9.30 16:0, 20:4 2.52
16 :1, 18:1 0.92 18:0, 20:4 3.36
18 :1, 18:1 2.96 16:0, 22:6 1.84
16:0, 18:2 21.78 18:0, 22:6 1.60
Source: Ref. 34.

Phospholipids se
F. Modification of Phosphoiipids
The m o le c u la r structure of phospholipids can be changed by either enzymatic or chemical
means. The aim of all these processes is to obtain tailormade technological and/or physiologi
cal properties that are somewhat different from those of the natural substances themselves.
1. Enzym atic M odification o f Com m ercial P hospholipid Products

There are different hydrolytic enzyme systems known that are able to modify the phospholipid
structure. They are all ester hydrolases and react site-specifically as follows:
P h o s p h o l i p a s e Ai
Phospholipase A j
P hospholipase C
P h o sp h o li p a s e D
It is typical of hydrolytic enzymes that their catalytic influence can be used for hydrolysis
or for transesterification. Commercial use to a large extent is known only with phospholipase
A 2 from porcine pancreas (reaction 1) [35,36].
PL A 2
Partially hydrolyzed vegetable phospholipid products (lecithins) show clearly improved

emulsifying properties up to a certain level of lysophospholipids. Above this level of conver
sion, water dispersibility increases but not the emulsifying capacity. On the other hand, with
increased hydrolysis the interaction with proteins in complex food systems seems to increase.
For another hydrolytic process using enzyme technology, cabbage-derived phospholipase
D is used (reaction 2). The reaction increases the phosphatidic acid (PA) content in lecithin
products [37]. The resulting products are reported to have improved emulsion properties and
the ability to mask the bitter taste of certain pharmaceuticals [38].
PL D
H,0
-O H
Whereas enzymatic transesterification of natural phospholipids with phospholipase A 2 (in

sertion of fatty acids with special pharmacological effects) is only of academic interest today
[39], the phospholipase D catalyzed formation of phosphatidylglycerol is already a commer
cial process in Japan. The resulting products are reported to have improved emulsion capacity
and temperature resistance. In Europe commercial interest is restricted to hydrolyzed but not
transesterified products as the European food regulation (E 322) only permits the “hydrolyzed
substances” [2],
Phospholipase Aj and phospholipase C reacted products are not on the market because of
the limited availability of phospholipase Aj (and fatty acid migration from the 2- to the 1-
position after hydrolysis) and a lack of interest in diglycerides formed by a phospholipase
C reaction.
60 Schneider
2. C hem ical M odification

There are various ways to chemically modify phospholipid molecules, but only few of them
are commercialized. The reason is that none of the resulting products has food grade status,
except for the products of hydroxylation (allowed in the United States for baking). However,
despite that problem, substantial development and application work has been done on chemi
cally modified phospholipid products.
Hydroxylation. An effective way to improve emulsion properties of vegetable lecithins
for o/w systems is hydroxylation [40]. Hydrogen peroxide reacts with the double bonds of
unsaturated phospholipid fatty acids under the catalytic action of small amounts of organic
acids of low molecular weight (e.g., lactic acid) to form dihydroxy fatty acid derivatives
(reaction 3).
L_p X
Acetylation. Another way to improve the hydrophilicity of vegetable lecithin products is

acetylation. During the reaction of vegetable phospholipid mixtures with moderate amounts of
acetic anhydride, acetylation of the free amino group of PE takes place (reaction 4). [41,42].
+A c 2O
-P - 0 - C H 2 - C H 2 - N H 2 -P-0-CH2~CH2-NHAc
Acetylated phospholipid mixtures are excellent o/w emulsifiers and also exhibit good ther
mal stability. The latter is explained by the fact that there is no primary amino group left for
Maillard-type darkening reactions with the carbohydrate contents of vegetable lecithin
products.
Hydrogenation. All natural phospholipid mixtures (lecithins) contain more or less high
levels of unsaturated fatty acids. This makes all products susceptible to oxidative processes,
which lead finally to unpleasant taste and smell (rancidity).
During catalytic hydrogenation of natural phospholipids, the unsaturated fatty acids become
saturated. As a result the products are more stable against oxidation but less soluble in fats
and oils.
As a side effect of hydrogenation, most of the natural pigments are destroyed, turning the
resulting products into white to off-white free-flowing powders.
The hydrogenation process for phospholipids requires quite drastic conditions regarding
hydrogen pressure as well as type and quantity of catalyst. Only palladium and platinum
catalysts are resistant enough to the poisoning effect of phospholipids to allow reasonable
conversion rates [43-45].
Because of the lack of food grade status and the high costs, the process is limited to
application in pharmaceuticals and cosmetic products.
Phospholipids 61
modification fractionation compounding

1
hydroxylation acetylation hydrolysis acetone ethanol
/ \
emulsifiers carriers fats and oils
hydroxilate d acetylated enzymatically deoiled | PC-enriched lecithin lecithin lecithin

le cithins iecitins modified lecithins | fractions and on ca rrie r compounds
lecithins
Fig. 2 Industrial processing of soybean lecithin (abbreviated).
To summarize this section on the industrial processing of natural phospholipid products.

Fig. 2 presents a visual impression of industrial processing of the most commonly used soy
bean products.
VI. APPLICATIONS
Phospholipid applications follow two main principles. They can be used either for their tech
nological functionality or for physiological benefits. The number of application areas has
rapidly increased over the last decades not only because of their functional properties but also
because of new scientific evidence for certain therapeutic activity.
Because of the vast range of applications this review concentrates on the industrially inter
esting ones and describes application principles in a very pragmatic way. For more scientific
background, reference is made to the already existing literature [15,46-50].
A. Phospholipids as Functional Ingredients

Phospholipids are amphiphilic molecules of moderate hydrophilicity. Even though the HLB
(hydrophilic-lipophilic balance) considerations do not fit perfectly for zwitterionic molecules,
application technologists very often raise questions regarding indications of HLB values. Ta
ble 10 gives some guiding values based on numerous emulsion tests.
62 Schneider
Table 10 Indication of HLB values for

Commercial Phospholipid Products
Product HLB
Crude soy lecithin 3

Deoiled soy lecithin 4
Deoiled egg lecithin 5
Hydrolyzed soy lecithin 7
Acetylated soy lecithin 8
Hydroxylated soy lecithin 8
Acetylated and hydrolyzed soy lecithin 9-10
Purified soy PC 7
Phospholipid products in crude or refined or modified form are used for wetting, emulsifi
cation, or dispersion of products from the technical industry, in feed and food products, and
in cosmetics and pharmaceuticals. Application principles can be reduced to a very few behav
ior patterns of phospholipids as illustrated in Fig. 3 [51].
1. Industrial Products
Paints and Varnishes. Predominantly crude or modified crude vegetable lecithin products
are used as dispersing aids for both organic and inorganic pigments. Applications range from
printing inks to dispersion paints and varnishes. As phospholipids are quite lipophilic, they
are used predominantly in systems based on organic solvents. Recent developments toward
highly hydrophilic lecithins modified by acetylation, hydrolysis, and/or hydroxylation lead to
products that seem to have attractive properties for water-based systems.
Another interesting application is in the manufacture of magnetic tapes. Here lecithins are
used as dispersants for the magnetic metal oxide pigments [52,53].
Other technical applications are found in the use of phospholipid-containing formulas for
plant pesticides [54,55] or fertilizers. For this area specific phospholipid concentrates loaded
with active ingredients that can be easily diluted with water are of interest. The attraction of
the formula is that in final dilution the product is a liposome preparation [56]. Liposomal
pesticide or fertilizer preparations are reported to have improved functionality and contribute
to reducing the use of toxic ingredients and to reducing environmental pollution [57,58]. (The
liposome principle is described in detail in the section on cosmetics.)
A nice example of the combined action for both technological and physiological principles
of phospholipids is the subject of a patent for the removal of petroleum spills [59]. In this
application lecithin serves not only as an emulsifier/dispersant for a concentrated all-natural
preparation but also as a nutritional cocktail for soil microorganisms, which are stimulated to
“digest” the spilled oil.
2. F eed Products
The attraction of phospholipid products in feed preparations is, as just mentioned, the combi
nation of the emulsifier function with nutritional benefits. Lecithins contribute to the animal’s
energy supply, they are sources of essential fatty acids and micronutrients such as choline and
inositol, and they support fat metabolism.
Phospholipids 63
- o - -
y ' q-:-:-
Protein
rnmmn Protein
iilllMU
Fig. 3 Simplified schematic physicochemical action principles of phospholipids.
(a,b) Coating of hydrophilic (a) and lipophilic (b) surfaces (wetting, dispersing);
(c,d) stabilization of o/w (c) and w/o (d) emulsions;
(e) micellar organization (solubilization);
(f) liposome (entrapment of hydrophilic and lipophilic substances);
(g,h) association of phospholipids with proteins;
(i) inclusion of lysophospholipids in amylose.
The annual world consumption of lecithins for cow’s milk replacers (refattened skim milk
products) is 40,000 metric tons. They are used in combination with synthetic emulsifiers
(ethoxylated products, sugar esters, polyglycerol derivatives, etc.).
Additionally it should be mentioned that deoiled powdered lecithins are gaining importance
as extrusion aids in the manufacture of feed pellets, especially for fish farming (better oil
uptake, better sinkability, improved extruder capacity).
More information on the use of phospholipids in feed is given in the subsection on animal
nutrition. Section VLB.5.
3. F ood A pplications
The industrial production of vegetable phospholipid products (lecithins) as by-products of the
oil refining process started in the mid-1930s. The availability of a cheap and versatile natural
emulsifier has led to a rapid increase in the use of lecithin, especially in the food industry.
Approximately 60-65% of the annual production of lecithins is still going into various food
products.
As there are so many different kinds of applications, it is impossible to be complete and
go deeply into details in this chapter. The manufacturers of various phospholipid products
offer excellent documentation and practical assistance in solving application problems.
Until recently, the application of lecithin to food products has been a predominantly empiri
cal exercise that required long years of practical and pragmatic experience. Because of the
64 Schneider
complexity of the various products offered and the complicated physicochemical behavior of
phospholipids and glycolipids, only a few basic principles have been discovered over the last
60 years of industrial application.
Food Emulsions. It has been demonstrated that purified phospholipids (commercially
available material is predominantly phosphatidylcholine from soybeans and eggs) are poor
emulsifiers, especially for o/w systems [60]. Of absolute importance to emulsion quality and
type of emulsion (o/w or w/o) is the detailed composition of the phospholipid mixture. The
more acidic types of phospholipids like PI and PA play an important role. At the pH value of
food products they contribute to a highly negative zeta potential of the dispersed phase, re
sulting in improved physicochemical stability [61].
oiL -iN -W A T E R EMULSIONS. Mayonnaise, salad dressings, and similar sauces require protein
in addition to phospholipids for stability. Natural vegetable lecithins are not very efficient
emulsifiers for these types of emulsions as they are usually prepared at a low pH value (close
to the isoelectric point of phospholipids, between pH 3 and 4) or they exhibit high ionic
strength due to salt content or they are subject to extensive heat treatment during sterilization.
All of these “stressful” conditions have a negative effect on emulsion stability. However,
recent findings describe the specific functionality of hydrolyzed phospholipids (lysophospho-
lipids) in the presence of proteins. Obviously a kind of lipoprotein structure formed during
food preparation shows a good stabilizing effect against various kinds of physicochemical
stress [62,63].
This principle of combining proteins with lysophospholipids may become a universal con
cept for complex food preparations that undergo stress treatment during manufacture such
as sausages, heat-sterilized emulsions, mayonnaise, dressings, sauces, even bread (frozen
doughs) and low fat spreads.
WATER-iN-oiL EMULSIONS (MARGARINE, SPREADS). Traditionally a typical margarine is 80%
fat; “light” products such as low fat spreads can have fat contents o f 40% or less.
One major disadvantage of margarine as a butter substitute is its poor spattering perfor
mance and the sedimentation and burning of the bulk protein in a hot pan. Combinations of
monodiglycerides and lecithin or hydrolyzed lecithins have more or less solved the problem.
Low fat spreads, however, require ethanol-fractionated lecithins with higher than natural PC
contents and a specific PC/PE ratio of approximately 4:1, quite often in combination with
gelatin or other stabilizers [64].
It is interesting, but not yet understandable, that deoiled lecithins—hydrolyzed or slightly
PC-enriched— seem to perform better in low fat spreads than oil-containing liquid products.
Sometimes the formula does not even require thickeners and stabilizers.
Separation Agents. Lecithin-oil combinations are widely used as pan release agents or
separating agents in various food applications. However, lecithin can develop color problems
when heated to temperatures of more than 80°C because of Maillard-type reactions between
the primary amino group of PE and free carbohydrates contained in vegetable lecithins. Two
strategies have been developed to overcome this problem: (1) ethanol fractionation to increase
PC and decrease PE, one of the reaction partners for the browning products, or (2) acetylation
as another way of reducing the percentage of primary amines. Ethanol-fractionated substances
show an additional benefit for the conventional release agents, as these products dissolve
clearly in vegetable oils and do not tend to sediment.
Confectionery. Lecithin has become a traditional ingredient in all types of chocolate
masses. Its function is to cut down on the requirement for cocoa butter, the most expensive
ingredient in chocolate products.
Phospholipids 65
— Viscosity (Pas)
— Yield point (Pa)
Fig. 4 Chocolate viscosity and yield point. (■) Viscosity (Pa-s); (□) yield point (Pa).
Even though the mode of action is not yet fully understood, lecithin acts because of its
surface-active properties at the interface between sugar and cocoa particles and fat. Covering
the surface of the hydrophilic ingredients, it reduces internal friction, i.e., it reduces the
viscosity and yield point of liquid chocolate masses. The flow characteristics of these masses
are dependent on many different factors including fat content, moisture, couching time, tem
pering, particle size and size distribution, and quality of cocoa butter. Accordingly, the opti
mal lecithin dosage can vary as much as the type of lecithin, whereas the latter is determined
primarily by the type of chocolate.
Modem lecithin technology offers a range of custom products and fractions that are spe
cially designed for chocolate bars, hollow articles, coating, and dietetic chocolate. Common
to all types of products and all lecithins is the optimal dosage range of 0.3-0.7% (see Fig.
4). Above a critical quantity (obviously dependent on the grade of dispersity of the mass),
viscosity again increases. The best time to add lecithin is shortly (30 min) before the end of
couching in order to achieve homogeneous distribution and yet not hinder moisture evapora
tion and optimal flavor development.
Over recent years a range of phospholipid products (from soybeans) have been developed.
These have specific properties and influence either viscosity or yield point or both, as indi
cated in Table 11.
Baking. Another industrially important yet diversified phospholipid application area is the
baking industry [63]. The main uses are in flour treatment (to standardize baking properties,
predominantly used in Germany and Belgium) and in bread improvers. Lecithins show their
effects predominantly in all kinds of wheat-based bakery goods. They do not function nearly
as well in rye products. The reason is again the specific tendency of phospholipids to interact
with proteins—in the case of wheat products, with gluten. Van der Waals forces bind phos
pholipids to the protein network, increasing its elasticity. Gases developed during baking are
kept inside the pores and are better enabled to expand the stmcture (i.e., increase the volume).
For two main reasons enzymatically hydrolyzed phospholipids are again of special interest.
(1) They can be inserted in the helix structure of the amylose part of wheat starch, thus
stabilizing it against reorganization during storage (rétrogradation). The result is an improved
shelf life. (2) They have a strong tendency to interact with hydrophobic parts of gluten,
making the bread suspension/dispersion more stress-tolerant, i.e., heat-tolerant, contributing
66 Schneider
Table 11 Lecithin Types and Properties in Chocolate
Properties
Product Viscosity Yield point
Standard soybean lecithin i 4

PC-enriched soybean
lecithin 'i 4 4 T
Combination of soybean
lecithin and PGPR^ 4 4 4
Ammonium phosphatides’’ i
Combination of lecithin and
ammonium phosphatides i i i i
^PGPR = polyglycerol polyricinoleate.
’"Ammonium phosphatides are based on partially hydrogenated rapeseed mo
nodiglycerides, which have first been reacted with phosphorus pentoxide and
then neutralized with ammonia. The products are o f food grade status and
registered under E 476 as food additives.
to an even distribution and a stable protein network. The result is greater volume with small,
even pores.
Commercial bread improvers can contain various ingredients besides lecithin (emulsifiers
like data ester, malt flour, enzymes, ascorbic acid, hydrocolloids, sugar, enzymes, etc.). In
recent years there has been a strong tendency to use all-natural bread improvers that do not
contain synthetic emulsifiers and even to avoid enzymes because of an ongoing discussion
of allergies.
Instant Food Products. Modem food technology and development concentrate on conve
nience foods, which are easy to use, cheap to transport, and have long shelf life [64]. Instant
products, i.e., dried products that have to be reconstituted by the consumer by dispersing or
dissolving them in water, milk, etc., are probably the first generation of large-scale conve
nience foods. Typical instant products include dry milk products (whole, skim, refatted),
cocoa and chocolate drinks, protein powders, and soups and sauces.
During reconstitution of the powdered products, two problems can occur: poor wetting
because of fat content and/or particle stmcture and rapid gelling at the powder surface, leading
to an impermeable gel layer surrounding the product. Two strategies have been developed to
overcome or minimize these problems:
1. Agglomeration. Capillary porous stmcture forces the liquid to penetrate the particles.
2. Lecithination. Added phospholipid products mask free surface fat.
Phospholipid products can be included during manufacture of the dried product (i.e., during
spray drying or blending) or can be applied onto the surface of the final product by spraying
either a water dispersion of a deoiled lecithin powder or a low viscosity oily solution.
Ideal phospholipid blends are either of natural soybean origin and composition (in liquid
or deoiled form) or moderately PC enriched products. The latter have the advantage of having
much less influence on taste because of relatively low dosage levels.
4. Cosmetics
Phospholipids are well recognized natural ingredients in cosmetics [65] as they combine a
long list of positive properties that are of special attraction to the cosmetic formulator. They
Phospholipids 67
are natural nonirritating emulsifiers and emollients that contain essential fatty acids and have
pH-balancing, film-forming, and moisture retention properties. These qualities have made
vegetable and egg phospholipid products traditional cosmetics ingredients. However, at the
same time they suffer some disadvantages. They are yellowish to brown in color and oxida
tion-sensitive, are poor emulsifiers by themselves, and they have a strong smell. Most of these
disadvantages (color, oxidation, smell) can be overcome by hydrogenation [66]. But the use
of hydrogenated products is still limited, probably because of their limited emulsification
power and high prices.
The discovery of their membrane-forming properties and the development of simple and
cheap liposome manufacturing methods have led to an almost revolutionary use of phospholip
ids in cosmetics.
Liposomes. Liposomes are tiny globular vesicles. An aqueous cave is entrapped inside a
structure of one to several double layers of phospholipids. In so-called multilamellar lipo
somes each double layer is separated from the other by an aqueous compartment.
In cosmetics, liposomes can be used for moisturizing as they carry a lot of water inside
the onionlike structure (up to 15 mL of water per gram of phospholipid!). But they can also
be considered a microencapsulation system for both water-soluble substances (in the aqueous
compartments) and lipophilic ingredients (inside the lipophilic domain of the membranes).
Liposomes not only protect sensitive ingredients, they also provide a long-lasting effect on
the skin (multilamellar structures) and their solubilizing effect contributes to solving formula
tion problems.
Historically, the liposome manufacturing processes have been very complicated, requiring
expensive equipment and sometimes toxic solvents, but a recently introduced method [67] has
dramatically simplified even large-scale production. The so-called pro-liposome method also
offers many formulation alternatives [68,69].
THE PRO-LIPOSOME PRINCIPLE. The pro-liposome method is based on the idea that addition
of water to a pro-liposome mixture leads to the spontaneous formation of liposomes. A pro-
liposome mixture consists of a carefully selected mixture of phospholipids solubilized in a
hydrophilic medium (usually but not necessarily an ethanol-water mixture). The phospholip
ids are in stacked bilayer form. The addition of water under moderate mechanical agitation
then triggers the formation of closed vesicles (liposomes). Active loading of water-soluble
ingredients can be done by the addition of a concentrated aqueous solution to the pro-liposome
mixture prior to final dilution. Lipophilic substances are ideally admixed with the pro-lipo
some preparation in ethanolic solution. For details of the process see Ref. 56.
5. P harm aceuticals
Phospholipids are useful tools for the formulation of pharmaceutical products. Unlike food
and feed applications, for pharmaceuticals there is a preference for egg-derived phospholipids.
To my mind the only reason for this tendency is that egg phospholipids have been used in
intravenous emulsions as safe ingredients since the 1960s. It is definitely not because of
superior technological functionality.
Finished product categories where phospholipids have found application include total par
enteral nutrition (TPN) (emulsifier), fat emulsions as drug carriers (emulsifier), aerosols (sus
pending aid), dermatological creams (emulsifier), liposomes (altered bioavailability), mixed
micelles (solubilization), suppositories (for improved absorption), and antibiotics (for im
proved dispersibility). As can be seen from this list, phospholipids are efficiently used to
solve formulation problems with lipophilic ingredients (oils, lipophilic drugs). As most newly
68 Schneider
developed drug substances are lipophilic, the perfectly biocompatible and nontoxic natural
phospholipids (and blends) have a bright future as formulation tools.
Typical products used for pharmaceutical application are
Fractionated, sometimes chromatographically purified, deoiled soybean phospholipids with

PC contents between 50 and 95%
Purified, deoiled, sometimes chromatographically purified, egg phospholipids with PC con
tents between 70 and 95%
Hydrogenated egg or soy products
Enzymatically hydrolyzed products (lysophopholipid-enriched)
Enzymatically converted products (e.g., phosphatidyIglycerol, PG)
6. P hospholipids as Antioxidants
Whether phospholipids exhibit a direct primary antioxidative effect is not completely clear
yet. There are published reports available on investigations in which purified phospholipids
have been used. Whenever liquid soy lecithins or soy lecithin fractions have been tested, the
effect observed can probably be explained by the natural content of tocopherols (—1000 ppm)
in crude products. The stabilizing effect almost completely disappeared when deoiled lecithins
were studied (after removal of neutral lipids including tocopherols with acetone) [70].
However, the synergistic effect of lecithins and phospholipids, especially in combination
with ascorbyl palmitate and tocopherol, is unquestionable [71]. Speculation about the mecha
nism of antioxidative effects leads toward the metal-binding characteristics or synergistic ef
fect of phosphatidylethanolamine [72] or even toward the prevention of oxidation because of
the formation of Maillard-type products [73]. Various commercial products are offered on
the market.
Care has to be taken that phospholipid-containing antioxidant cocktails are only used up to
a maximum of 80°C. Above that temperature, the natural phospholipids start to degrade.
B. Physiology of Phospholipids
Soon after the isolation of phospholipids from animal and vegetable sources and the identifi
cation of similar molecules in human tissue, the pharmaceutical industry started to offer phos
pholipid preparations with various therapeutic indications. Since the 1950s, scientific investi
gators have discovered both the biological role of phospholipids and their therapeutic
potential. Phospholipids are highly specialized lipids that are key constituents of cell mem
branes and are essential for the growth, maturation, maintenance, and functional capacity of
all cells of the animal and human body. In addition to their roles in cell membranes, phospho
lipids are essential elements of the lipoproteins in blood circulation and of the surfactants that
wet internal body surfaces, and they are used in the bile as emulsifiers.
The outer cell membrane can be considered the cell’s gatekeeper, whereas the internal
membranes are the metabolic work centers. Phospholipids create the functional microenviron
ment for the membrane proteins.
The crucial role of phospholipid-containing structures in human health has stimulated many
research activities and clinical investigations.
Before the therapeutic ideas are described, a few words should be given to the metabolism
of phospholipids after oral intake.
The average intake of phospholipids with normal food is 1.7-2.2 g/day [74]. These phos
pholipids, like all orally administered phospholipids, are hydrolyzed by pancreatic phospholi
Phospholipids 69
pase A 2 after micellarization with bile fluids. The resulting 1-acyl-lysophospholipids can be
taken up by the mucosa enterocytes.
Inside the mucosal cell, reesterification to intact phospholipids occurs using the pool of
fatty acids present in the cells. Finally, phospholipids are released into the chyle, where they
are transported as part of the chylomicron organization [75-77].
This brief outline of absorption seems to be true for about 50-60% of the orally taken
phospholipids, the rest are split into smaller molecules, which are metabolized in the usual
way.
Nutritional aspects of phospholipid supplementation in humans cover the following main
areas: lipid (especially cholesterol) metabolism, liver therapy, neurological disorders, and
memory improvement. In this overview only the major applications can be described, as there
is a vast literature on the potential therapeutic effects of phospholipids. The purpose of this
section is not to describe the biochemical background of their therapeutic activity but to focus
on a description of the effect of industrial phospholipid products.
1. L ipid M etabolism
As coronary heart disease is the major cause of death in most societies, interest in using a
natural supplement like soy phospholipids (most clinical studies have been carried out with
soy phospholipid preparations) for preventive and probably even therapeutic reasons is high.
Purified soy PC and even mixed soy phospholipids have a clear effect on lowering elevated
serum lipid levels. An overview reporting on 27 clinical trials [78] shows that there is indeed
such an effect. Two of the most striking sets of study results are summarized in Tables 12
and 13.
The most dramatic reduction in both cholesterol and triglycerides was seen when a fine
(liposomal) aqueous dispersion of mixed soy phospholipids was applied (9 g/day). Whether
dispersity may have had an influence on the effect has not yet been investigated. But it is
obvious that results after administration of bulk phospholipids (e.g., in granular form) are not
as good. However, the evaluation of trial results is very complicated, as study designs have
been too different to allow direct comparison.
The HDL/LDL balancing effect of soybean phospholipid preparations is obvious, in con
trast with vegetable triglycerides (soybean oil), which reduce both HDL and LDL cholesterol
to a moderate extent. Phospholipids not only lower total cholesterol but at the same time raise
HDL (the lipoprotein that washes out cholesterol) and lower LDL (the one that delivers it to
the cells), thus having a twofold positive effect [81].
Finally the existing clinical data have led to the publication of a monograph description by
the German health authorities [82]. This monograph grants the natural mixture of soybean
phospholipids drug status for the indication of moderate disorders of lipid metabolism. The
therapeutic dose is based on 3.5 g PC per day.
2. L iver Therapy
Various experimental and clinical results support the assumption that soybean phosphatidyl
choline has protective and even curative and regenerative effects on liver cell membranes. In
various animal models the hepatoprotective effect of PC against various toxic substances (eth
anol, chlorinated hydrocarbons, drugs, etc.) has been impressively demonstrated. Also, more
than 100 clinical studies on animals and humans with some type of liver disease have been
conducted with PC. The main diseases treated were cirrhosis, hepatitis (acute and chronic),
hepatic coma, fatty liver, and toxic liver damage from xenobiotics and from mushroom poi
soning. Treatment lasted for weeks to months, and dosages ranged up to 3 g PC/day (orally).
70 Schneider
Table 12 Cholesterol-Lowering Effect of a Lecithin Solution^
Total cholesterol Triglycerides HDL LDL

(mg/dL) (mg/dL) (mg/dL) (mg/dL)
Before After 4 wk Before After 4 wk Before After 4 wk Before After 4 wk
Lecithin 307± 32 211 ± 3 9 223 ± 3 9 156 ± 27 41±6 54±9 204 ±31 162 ±28
Placebo 322 ± 3 9 290 ±35 201 ± 4 2 214 ± 3 4 38±6 43±7 192 ± 3 9 186 ± 3 2
În a controlled study, 240 patients were given either 5.4 g deoiled lecithin per day in a 9% solution containing 16%
ethanol or a placebo.
Source: Ref. 79.
Besides excellent tolerance, clinical improvements included [83-85] improvement of specific

symptoms, better biochemical and histological outcomes, accelerated improvement of clinical
findings, and shortened duration of hospitalization. These impressive findings pushed the
German health authorities to publish another monograph on phospholipids [86]. Purified soy
bean PC (73-79%) has been given drug status as a liver therapeutic substance (dosage 1.5
g/day).
3. N eurology
The assumption that phosphatidylcholine may have a positive effect on degenerative neurolog
ical disorders goes back to the discovery that choline is able to pass the blood/brain barrier
and to stimulate acetylcholine synthesis [87].
Administration of choline salts raised blood choline levels only for short times and led to
a fishy body odor because of amine formation from choline. PC administration, however,
efficiently raised choline levels for a much longer period without having the mentioned side
effect [88]. Since then intensive clinical activities have begun to investigate PC effects.
The two major known neurodegenerativo diseases, Alzheimer’s and Parkinson’s, feature
significant losses of the neurons that produce acetylcholine. Dietary PC is a convenient precur
sor to make choline available for enhanced acetylcholine synthesis. But the results with both
Alzheimer’s and Parkinson’s disease patients are not very convincing. For patients with fully
developed symptoms, even extremely high doses did not lead to significant improvements of
the symptoms [89,90].
Slightly better results could be achieved in CNS-related movement disorders like tardive
dyskinesia, Friedreich’s ataxia, Tourette’s disease and Huntington’s disease. Improvements
were observed in the reduction of abnormal movements and in speaking ability [91,92].
Table 13 Lipid-Lowering Effect of Phosphatidylcholine ^
Total cholesterol Triglycerides HDL LDL

(mg/dL) (mg/dL) (mg/dL) (mg/dL)
Before After 6 wk Before After 6 wk Before After 6 wk Before After 6 wk
PC 309 251 354 249 32 44 195 130

Placebo 285 310 327 415 38 41 168 182
group of 28 patients were given 1.8g PC per day in capsules or a placebo.

Source: Ref. 80.
Phospholipids 71
In summary, phosphatidylcholine preparations for single therapeutic action are not suffi
cient but can assist ethical medication.
4. M em ory
Moderate to no improvement of cognitive abilities is reported from the use of conventional
types and forms of vegetable phospholipids. The situation after administration of bovine brain
derived phosphatidylserine is different. Significant improvements could be observed, espe
cially in all kinds of age-related cognitive decline. So far 15 clinical studies have been done,
and all of them demonstrate impressively that PS seems to be a very effective nootropic
substance. The mechanism for the PS activity is not fully understood yet. A good survey is
given in Ref. 93. The therapeutic activity seems not to be linked to the origin of the PS.
(Different origins would mean different fatty acid compositions.)
Since animal brain is not a very attractive raw material for therapeutic and dietary sub
stances like PS, a “vegetable” PS has been produced by enzymatic transphosphatidylation:
PLD
PC. PE, Serine PS Choline, Ethanolamine,
This product has been introduced into the market in Israel and the United States. A clinical
study showed that there seems to be no difference in therapeutic efficacy compared to the
animal-derived product [94].
As mentioned earlier, there are thousands of articles describing other effects of phospholip
ids on human health (especially worthy of mention is the work of Zeisel and coworkers [95-
98]). These findings have not yet led to major commercial and industrial applications, but it
is certain that more products with novel therapeutic concepts will soon appear on the market.
5. Phospholipids in A nim al N utrition

Besides their use as emulsifiers in various feed preparations (see Section VI.A.2), phospho
lipid products have gained a substantial market as nutritional ingredients in feed formulations.
Here they serve two functions: improvement of absorption of lipophilic nutrients (fat, vita
mins) and supply of nutritional substances (essential fatty acids, choline, inositol, etc.).
For terrestrial animals the main function of phospholipids is the improvement of fat digest
ibility. The biosurfactant property of phospholipids enhances emulsification of dietary lipids
in the small intestine. Since in young animals the digestive system is still immature, the
quantity of secreted bile is most probably insufficient to emulsify the dietary fat, particularly
if the levels of fat are high. Feed supplemented with lecithin therefore improves the emulsifi
cation of dietary lipids and contributes to better absorption of the fatty acids and to better
utilization of added fat as the most efficient energy source.
Phosphatidylcholine is very involved in the metabolic transport of fatty acids, which can
be transported from the liver only in the form of triglycerides within the lipoproteins. An
essential constituent of those lipoproteins is phosphatidylcholine. When adequate supplies of
phosphatidylcholine are not available, the liver is unable to export triglycerides and becomes
infiltrated with fat (fatty liver syndrome). There is an absolute dependence on PC; no other
phospholipid will suffice. As a result, fat digestibility is improved (equal to an increase in
feed conversion) and fewer problems with diarrhea are observed in farm animals.
In the following sections the nutritional benefits of lecithin and phospholipid for growing
terrestrial animals, fish, and crustaceans are described in more detail.
72 Schneider
Table 14 Effect of Lecithin on the Development of Early Weaned Pigs^

Control group Group 1 Group II Group III
Deoiled lecithin (%) — 1.2 — 1.2

Soybean oil (%) — — 6 6
Daily gain (g) 372 391 381 390
Feed conversion 1.66 1.67 1.78 1.75
^3-8 weeks o f age, test period 5 weeks.

Source: Ref. 99.
Piglets. A good quality starter feed for piglets has to meet three criteria: It must meet all
of the nutrient requirements of the young pig, it must be palatable, and it must be easy to
digest. Trials have demonstrated that piglet diets containing additional phospholipids have a
positive effect on weight gain, feed efficiency, and good health (immunology). Table 14
presents the results of one study. Figure 5 shows the improvement in fat digestibility.
Lecithin/kg feed
^ 0g
B2 g
■ 4g
EDS g
Fig. 5 Improvement in digestibility of fat and other feed nutrients in dependence on different dosages
of deoiled lecithin (From Ref. 100.)
Crustaceans. It is known that phospholipids (especially PC + PI) are needed not only to
promote the health of crustaceans but also for best performance [101,102]. Unlike terrestrial
animals, crustaceans are incapable of synthesizing phospholipids. Crustacean diets, therefore,
have to be fortified with phospholipids. The fortification of feed with phospholipids contri
butes to the inositol and choline needs of the animals due to the presence of PI and PC in the
phospholipids. The requirements of Penaeus japonicus larvae for cholesterol and soybean
phospholipids were studied [103]. The results obtained indicated that the survival rates of
larvae increased as levels of soybean PC rose from 0 to 3.0% and those of cholesterol rose
from 0 to 1.0% of the diets. However, the supplementation of higher levels of PC (6.0%) and
cholesterol (5.0%) did not result in further improvement of survival rates. (See Table 15.)
In another trial, the addition of 3.0% deoiled lecithin to the diet of prawns significantly
improved the weight gain and feed conversion rate (Fig. 6). The whole body of the prawns
receiving a lecithin-deficient diet contained a lower concentration of phospholipids than the
body of those fed lecithin-supplemented diets.
Fish. A critical situation is the nutritional requirement of larvae at the time of transition
from endogenous to exogenous feeding. The fish egg contains levels of nutrients that must
Phospholipids 73
Table 15 Survival Rate and Growth After Feeding

Increasing Levels of Soybean PC and Cholesterol to
Penaeus japonicus Larvae for 7 Days
Soybean PC in the diet [%]

Cholesterol 1.0 3.0 6.0
in the diet
(%) Survival rate (%)
0 13 25 32 26
0.5 21 32 64 59
1.0 25 69 72 80
5.0 35 42 55 68
Growth index
0 4.0 5.0 5.5 5.5

0.5 4.5 6.7 7.0 6.9
1.0 5.0 7.0 7.0 6.9
5.0 5.0 6.2 6.5 6.3
meet the demands of both energy and growth of the embryo and the resulting fry. Throughout
the period of embryogenesis, PC is the only lipid that declines in absolute terms [106].
From the results of this trial, it was concluded that PC has a range of biochemical and
physiological functions during embryogenesis in many fish species. PC may also have a role
in the transport of yolk components to the developing embryo. In addition, PC is undoubtedly
an important source of long-chain {n-3) PUFAs for developing egg and larvae. Further nutri
tional attributes of PC include the provision of phosphorus and choline. The presence of
PC in the diet of larval fish may also aid in the emulsification and absorption of dietary
neutral lipid.
Many research workers have demonstrated the need for dietary phospholipids in juvenile
salmonids [107,108]. But there is evidence that phospholipids are effective in larger salmonids
also. In a field trial, growing rainbow trout were fed a commercial feed supplemented with
1.6% deoiled lecithin but in the absence of choline. The fish had an initial mean weight of
-w ith Lecithin -w ith o u t Lecithin
Fig. 6 Weight gain (%) of prawns fed diets with and without lecithin (From Ref. 105.)
74 Schneider
Table 16 Effect of Deoiled Lecithin on the Performance of Growing Rainbow Trout

(Salmo gardneri)
Change against control

Control Trial (%)
Deoiled lecithin (%) 1.4
Choline chloride, 50% 0.5
Number of trout 100 100
Trial period (days) 99 99
Initial weight (g) 270 262 - 3 .0
Final weight (g) 504 528 + 4.8
Weight gain (g) 234 266 + 13.7
Daily weight gain (g) 2.39 2.70 + 13.0
Feed conversion 5.18 4.87 - 6.0
Losses (%) 8.7 1.2 - 86.2
Source: Ref. 109.
325 g. In the 61 day trial period, the lecithin-fed trout had 14.6% higher weight gain than
that of the control group, which received a feed containing 0.5% choline chloride [109].
Results are summarized in Table 16.
To conclude this section about phospholipids in animal nutrition, it has to be mentioned
that besides their biochemical and physiological properties, phospholipids also exhibit techno
logical functions that are of industrial importance in the manufacture of feed products. These
functions enable them to serve as emulsifiers, wetting and dispersing agents, extrusion aids,
release agents, and antioxidants.
REFERENCES
1. lUPAC-IUB Committee on Biochemical Nomenclature Recommendation, Lipids 12: 455 (1977).
2. E 322 - Lecithine, Amtsblatt Eur. Gem. L223: 30 (1978).
3. Lecithin, Fed. Reg. Vol. 49, §184.1400 (1984).
4. O. W. Thiele (Ed.), Lipide, Isoprenoide mit Steroiden Georg Thieme, Stuttgart, 1979.
5. E. Schulze and E. Steiger, Über den Lecithingehalt der Pflanzensamen, Z. Physiol. Chem. 13:
365 (1889).
6 . E. Schulze, Landwirtsch. Versuchsstationen, Landwirtschaft 49: 203 (1898).
7. E. Schulze, Über pflanzliche Phosphatide, Chem. Z. 32: 981 (1908).
8 . D. A. Gomall and A. Kuksis, Molecular species of glycero phosphatides and triglycerides of egg
yolk lipoproteins. Can. J. Biochem. 49: 51 (1971).
9. F. Phillips and O. S. Privett, A simplified procedure for the quantitative extraction of lipids from
brain tissue. Lipids 14: 590 (1979).
10. O. S. Privett et al., in Dairy Lipids and Lipid Metabolism (H. F. Brink and D. Kritchevsky, Eds.)
Avi, Westport, CT, 1968, p. 99.
1 1 . Lucas Meyer, Hamburg, Germany, data based on ^^P-NMR.
12. L. A. Appelqvist and R. Ohlson (Eds.), Rapeseed: Cultivation, Composition, Processing and
Utilization, Elsevier, Amsterdam, 1972.
13. S. S. Grewal, P. S. Sukhija, and L. S. Bhatia, Polar lipid composition during sunflower Helian
thus annuus seed development, Biochem. Physiol. Pflanzen 173: 11 (1978).
14. E. J. Weber, Compositions of commercial com and soybean lecithins, J. Am. Oil Chem. Soc. 58:
898 (1981).
Phospholipids 75
15. B. F. Szuhaj and G. R. List (Eds.), Lecithins, American Oil Chemist’s Society, Champaign, 1985.
16. F. J. Flider, The manufacture of soybean Lecithins, in Lecithins (B. F. Szuhaj and G. R. List,
Eds.), American Oil Chemist’s Society, Champaign, 1985, p. 21.
17. G. Rouser, A. J. Bauman, N. Nicolaides, and D. Heller, Paper chromatography of lipids: meth
ods, applications and interpretations, J. Am. Oil Chem. Soc. 38: 565 (1961).
18. T. Majonnier and H. C. Troy (Eds.), The Technical Control of Dairy Products, Majonnier Bros.
Co., Chicago, 1922.
19. J. C. Crane and B. E. Horrall, Phospholipids in dairy products. II. Determination of phospholipids
and lecithin in lipids extracted from dairy products, J. Dairy Sci. 26: 935 (1943).
20. P. Walstra and J. J. De Graaf, Note on the determination of the phospholipid content of milk
products, Neth. Milk Dairy J. 16: 283 (1962).
21. A. H. Duthie and S. Patton, New method for extraction of milk phospholipids, J. Dairy Sci. 48:
170 (1965).
22. J. Cerbulis, Distribution of lipids in various fractions of cow milk, J. Agric. Food Chem. 15:
784 (1967).
23. J. Folch, I. Ascoli, M. Lees, J. A. Meath, and F. N. Le Baron, Preparation of lipid extract from
brain tissue, J. Biol. Chem. 191: 833 (1951).
24. J. Folch, M. Lees, and G. H. Stanley, A simple method for the isolation and purification of total
lipids from animal tissues, J. Biol. Chem. 226: 497 (1957).
25. K. Zosel, Verfahren zur Entcoffeinierung von Rohkaffee, DP 2005293 (1970).
26. Kali-Chemi Pharma, Verfahren zur Gewinnung von Lecithin, DP 3011185 (1980).
27. Lucas Meyer, Verfahren zur Gewinnung von Lecithin, DP 3229041 (1990).
28. E. Stahl, Extraction of natural substances by dense gases, Fette, Seifen, Anstrichm. 84: 444 (1982).
29. E. Stahl and K.-W. Quirin, Deoiling of crude lecithin by jet extraction with dense carbon dioxide,
Fette, Seifen, Anstrichm. 87: 219 (1985).
30. Lucas Meyer, Process for the preparation of lecithin, by preference the continuous preparation of
pure lecithin, Eur. Patent 0156374 (1985).
31. S. Peter, Process for removing oil from raw lecithin, PCT 94/01004 (1994).
32. SKW Trostberg, Verfahren zur Herstellung von pulverformigen Eiprodukten mit einem reduzierten
Gehalt an Fetten und Cholesterinderivaten, DP 4236474 (1992).
33. M. Schneider, Fractionation and purification of lecithin, in Lecithins, Sources, Manufacture &
Uses (B. F. Szuhaj, Ed.), Americal Oil Chemists Society, 1989, p. 109.
34. A. Kuksis and L. Marai, Determination of the complete structure of natural lecithins. Lipids 2:
111 (1967).
35. Unilever, Use of self-releasing binders on an isocyanate-Zsulfonic acid basis for producing shaped
articles, DP 1900859 (1969).
36. Lucas Meyer, Verfahren zur Herstellung eines hydrolisierten Lecithins, Eur. Patent Appl.
0260573 (1987).
37. M. Heller, N. Mozes, and I. Peri, Interactions of phospholipase D with 1,2 diacyl-5 u-glycero-3 -
phosphocholine and dodecylsulfate. Lipids 11: 604 (1976).
38. Kao Corporation, Method for suppressing bitter taste of oral preparation, Eur. Patent Appl.
0631787 (1994).
39. C. S. Chen and C. J. Sih, General aspects and optimization of enantioselective biocatalysis in
organic solvents by the use of lipases, Angew. Chem. Int. Ed. Eng. 28: 695 (1989).
40. General Mills, Journal box structure, U.S. Patent 2,445,948 (1944).
41. Central Soya, Method of preparing heat resistant lecithin release agent, U.S. Patent 4,479,977
(1984).
42. Central Soya Co., Process of phosphatide preparation, U.S. Patent 3,301,881 (1963).
43. Armour & Co, Verfahren zur Herstellung von stabilisierten Emulgatoren aus Soyabohnenphospha-
tid, DP 1002607 (1953).
44. American Lecithin Co., Hydrogenation of phosphatides, U.S. Patent 2870179 (1954).
45. D. V. Sokolski and T. Omaraliev, Hydrogenation of the lecithin fraction of soybean oil phospha
tides, Chem. Abstr. 68: 96992 (1968).
76 Schneider
46. G. Cevc (Ed.), Phospholipid Handbook, Marcel Dekker, New York, 1993.
47. G. B. Ansell (Ed.), Form and Function of Phospholipids, Elsevier, Amsterdam, 1973.
48. D. M. Small (Ed.), Handbook of Lipid Research, Plenum, New York, 1980.
49. G. Ceve and F. Paltauf (Eds.), Phospholipids: Characterisation, Metabolism and N o vel Biological
Applications, American Oil Chemist’s Society, 1995.
50. B. F. Szuhaj (Ed.), Lecithins, Sources, Manufacture & Uses, 1989.
51. H. Pardun, Die Pflanzenlecithine, Verlag Chem. Ind., 1988.
52. Victor Co. of Japan Ltd., Magnetisches Aufzeichnungsmaterial, DP 2910487 (1978).
53. International Business Machines Corp., Magnetic inks containing lecithin as a surfactant, French
Patent 1548700 (1966).
54. A. Nattermann, Aqueous phospholipid solutions containing a solubilizing agent, U.S. Patent
4,874,553 (1987).
55. Newman Agrochemicals, Agrochemical formulation, PCT 89/08628 (1988).
56. Phares Pharmaceutical, Liposome forming composition, Eur. Patent 0158441 (1985).
57. Kabi Pharmacia, Lipid formulation system, PCT 92/05771 (1992).
58. Nova Pharmaceuticals, Lipospheres for controlled delivery of substanees, PCT 91/07171 (1991).
59. Cognis GmbH, Verbesserte Nährstoffgemische für die Bioremediation verschmutzter Böden und
Gewässer, DP 4218243 (1992).
60. L. Rydhag and I. Wilton, The function of phospholipids of soybean lecithin in emulsions, J. A m .
O il C hem . Soc. 58: 830 (1981).
61. L. Rydhag, Importance of the phase-behavior of phospholipids for emulsion stability, Fette,
Seifen, Anstrichm 81: 168 (1979).
62. Nestlé, Zusammensetzung für Wärmestabilisierung von Proteinen und Produkten daraus, Eur. Pa
tent 0628256 (1993).
63. Unilever N. V., Hitzesterilisierbare Wasser- und Öl-Emulsionen, Eur. Patent 0328789 (1988).
64. Unilever N. V, Antispritzmittel für Wasser enthaltende Speisefette, DP 1492952 (1963).
65. Y. Pomeranz, Lecithin in baking, in Lecithins (B. F. Szuhaj and G. R. List, Eds.), American Oil
Chemist’s Society, Champaign, 1985, p. 289.
66. Spray T ow er Philosophy, Lucas Meyer Publ. 5, 1988.
67. Lecithins in Cosmetics, Lucas Meyer Publ. 9, 1991.
68. Liposomes— The Pro-Liposome Approach, Lucas Meyer Publ. 14, 1993.
69. S. Perrett, M. Golding, and W. P. Williams, A simple method for the preparation of liposomes
for pharmaceutical applications: characterization of the liposomes, J. Pharm. Pharmacol. 43: 154
(1991).
70. G. Mieth and F. Linow, Zur fettstabilisierenden Wirkung von Phosphatiden; 2. (Mitt.) Antioxida
tive Wirkung ausgewählter Phosphatide, Nahrung 19: 577 (1975).
71. G. Pongracz, Antioxidant mixtures for use in food. Int. J. Vitamin Nutr. Res. 43: 517 (1973).
72. D. H. Hildebrand, J. Terao, and M. Kito, Phospholipids plus tocopherols increase soybean oil
stability, J. Am. Oil Chem. Soc. 61: 552 (1984).
73. C. K. Park and D. H. Kim, Relationship between fluorescence and antioxidant activity of ethanol
extracts of a Maillard browning mixture, J. Am. Oil Chem. Soc. 60: 98 (1983).
74. B. Borgström, On the binding of bile salt to panereatic lipase, in Lipid Absorption: Biochemical
and Clinical Aspects (K. Rommel et al., Eds.), MTP Press, 1976.
75. A. Nilsson and B. Borgström, Absorption and metabolism of lecithin and lysolecithin by intestinal
slices, Biochem. Biophys. Acta 137: 240 (1967).
76. R. O. Scow, Y. Stein, and O. Stein, Incorporation of dietary leeithin and lysolecithin into lymph
chylomicrons in the rat, J. Biol. Chem. 242: 4919 (1967).
77. D. Lekim and H. Betzing, Intestinal absorption of polyunsaturated phosphatidylcholine in the rat,
Z. Physiol. Chem. 357: 1321 (1976).
78. J. T. Knuiman, A. C. Bey nan, and M. B. Katan, Lecithin intake and serum-eholesterol, J. Am.
C lin. N utr. 49: 266 (1989).
79. D. Kupke, U. Schmidt, and H. J. Reimann, The influence of lecithin on hyperlipemia— a report
on prevention and treatment of hyperlipoproteinemia, Med. W elt 38: 1244 (1987).
Phospholipids 77
80. A. K., Horsch, I. Majolk, C. C. Heuck, and E. Göpfert, Einfluss von Polyenylphosphatidyl-
cholin (PPC) auf die Semmlipide bei Patienten mit Hyperlipoproteinaemie, VASA 15: 251 (1986).
81. J. N. Hawthorne, M. Hoccom, and J. E. O’Mullane, The nutritional value of linoleic acid given
either as com oil or as soya lecithin, Clin. Dietol. 8: 253 (1981).
82. BGA Monography, Lecithinum ex Soja (Sojalecithin), Bundesanzeiger, May 5, 1988, p. 2018.
83. K. J. Gundermann, Biological activity of polyunsaturated phosphatidylcholine in different dis
eases, in Phospholipids: Characterisation, Metabolism and Novel Biological Applications (G.
Cevc and F. Paltauf, eds.), American Oil Chemist’s Society, Champaign, 1995, p. 208.
84. C. S. Lieber, L. M. DeCarli, K. M. Mak, C. I. Kim, and M. A. Leo, Attenuation of alcohol-
induced hepatic fibrosis by polyunsaturated lecithin. Hepatology 12: 1390 (1990).
85. C. S. Lieber, A. J. Robins, J. Li, L. M. Decarli, K. M. Mak, M. Ki, J. M. Fasulo, and M. A.
Leo, Phosphatidylcholine protects against fibrosis and cirrhosis in the baboon. Gastroenterology
106: 152 (1994).
86 . BGA Monography, Phospholipide aus Sojabohnen mit 73-79% 3-s«-Phosphatidylcholin, Bundes
anzeiger, July 19, 1994, p. 7361.
87. E. L. Cohen and R. J. Wurtman, Brain acetylcholine: control by dietary choline. Science 191:
561 (1976).
88. R. J. Wurtman, M. J. Hirsch, and J. H. Growdon, Lecithin consumption raises semm-free-
choline levels. Lancet 2: 6 8 (1977).
89. P. Etienne, S. Gauthier, D. Dastoor, B. Collier, and J. Rainer, Alzheimer’s disease: clinical
effect of lecithin treatment, in Nutrition and the Brain (A. Barbean et al., Eds.), Raven Press,
New York, 1979, p. 389.
90. P. Etienne, D. Dastoor, A. Gauthier, R. Ludwick, and B. Collier, Lecithin in the treatment of
Alzheimer’s disease. Aging 19: 369 (1982).
91. J. H. Growdon, A. J. Gelenber, J. Doller, M. J. Hirsch, and R. H. Wurtmann, Lecithin can
suppress tardive dyskinesia, N. Engl. J. Med. 298: 1029 (1978).
92. A. Barbean, Phosphatidylcholine lecithin in neurologic disorders. Neurology 28: 358 (1978).
93. P. Kidd, Phosphatidylserine, Lucas Meyer, Hamburg, Germany, 1995.
94. J. Gindin, M. Novikov, D. Kedar, A. Walter-Gunzburg, A. Nacer, O. Karta, E. Zurand, and S.
Levi, Kaplan Hospital, Rehovot, Israel, unpublished results.
95. S. H. Zeisel and D. J. Canty, Choline phospholipids molecular mechanisms for human diseases,
J. Nutr. Biochem. 4: 258 (1993).
96. S. H. Zeisel, K. A. da Costa, P. D. Franklin, E. A. Alexander, J. T. Lamont, N. F. Sheard,
and A. Beiser, Choline, an essential nutrient for humans, FASEB J. 5: 2093 (1991).
97. S. H. Zeisel, Lecithin in health and disease, in Lecithins (B. F. Szuhaj and G. R. List, Eds.),
American Oil Chemist’s Society, Champaign, 1985, p. 323.
98. S. H. Zeisel, Lecithin in health and human nutrition, in Lecithins, Sources, Manufacture & Uses
(B. F. Szuhaj, Ed.), American Oil Chemist’s Society, Champaign, 1989, p. 225.
99. Lucas Meyer, Hamburg, Germany, unpublished results.
100. K. D. Günther, Rein-Lecithin als Wirkstoff in der Ferkelemährung, Kraftfutter 4194: 129 (1994).
101. L. R. D ’Àbramo, C. L. Bordner, and D. E. Conklin, Relationship between dietary phosphatidyl
choline and serum cholesterol in the lobster Humarus sp.. Marine Biol. 67: 231 (1982).
102. A. Kanazawa, S. Teshima, and M. Sakamoto, Effects of dietary lipids, fatty acids and phospho
lipids on growth and survival of prawn (Penaeus japonicus) larvae. Aquaculture 50: 39 (1985).
103. J. W. Hertrampf, Feeding crustaceans with lecithin. Feed Mag. 11: 12 (1989).
104. S. Teshima, A. Kanazawa, H. Sasada, and M. Kawasaki, Requirements of the larval prawn,
Penaeus japonicus, for cholesterol and soybean phospholipids, Mem. Fac. Fish Kagoshima Univ.
31: 193 (1982).
105. S. Teshima and A. Kanazawa, Role of dietary phospholipids in the transport of ['"^Cj-tripalmitin
in prawn. Bull. Jpn. Soc. Sci. Fish 52: 519 (1986).
106. A. J. Fraser, C. Gamble, and B. Sargent, Changes in lipid content, lipid class composition and
fatty acid composition of developing eggs and unfed larvae of cod (Gadus morhua). Marine Biol.
99: 307 (1988).
78 Schneider
107. H. A. Poston, Effect of body size on growth survival and chemical composition of Atlantic
salmon fed soy lecithin and choline, Prog. Fish Cult. 52: 218 (1990).
108. J. W. Hertrampf, Zur Phospholipid-Versorgung von Salmoniden, Mühle Mischfuttertechnik 128:
463 (1991).
109. H. N. Allwörden, Versuchsbericht des Institutes für Humanemährung und Lebensmittelkunde der
Univ. Kiel, Germany, University of Kiel, personal communication, 1990.
Lip id s and Nutrition
Michael I. Gurr
M a y p o le S c ie n tific Services, S t M a r y ’s, Isle s o f Scilly, E n g la n d
I. INTRODUCTION
During the past decade or more there has been intense interest among nutritionists, health
professionals, and consumers in the subject of dietary fat. Virtually all dietary recommenda
tions in industrialized countries have focused primarily on reducing the amount of fat in the
diet, and some have emphasized the need to modify its composition. This theme has been a
regular feature of national daily newspapers, magazines, and radio and television programs.
There are several reasons. Many medical reports have highlighted the high prevalence of
cardiovascular disease (CVD) in industrialized countries, and most have associated this with
high consumption of fat, especially saturated fatty acids (SFA). Thus, the most recent such
report from the U.K. Department of Health [1] clearly indicates that it considers that the
incidence and mortality from CVD in the United Kingdom could be substantially reduced if
the British population were to reduce its consumption of fat and SFA. This echoes the views
of many similar national and international reports. A major health initiative in England (The
Health of the Nation) aims to reduce average intakes of fat by 12% and of saturated fatty
acids by 35% by the year 2005 [2].
High fat intakes or the consumption of an inappropriate balance of the different fatty acids
have also been suggested as being connected with several other chronic diseases, but, apart
from heart disease, most attention had been given to obesity. In the United Kingdom, which
is not atypical of many industrialized countries, there is the paradoxical situation that whereas
total energy and total fat intakes have been steadily declining, the proportion of the population
deemed to be overweight has been rising. Professional views about the contribution of exces
sive intakes of fat (as distinct from carbohydrates or total energy) have undergone many
changes during this century. Currently, the consensus seems to be that dietary fat may make
a specific contribution to the development of obesity. The Health of the Nation aims to reduce
the number of people between the ages 16 and 64 who are obese by one-fourth in males and
at least one-third in females by the year 2005.
79
80 Gurr
The widespread desire to reduce fat in the diet as painlessly as possible has given rise to
remarkable growth in the availability of low or reduced fat products and a continuing quest
for fat substitutes that will deliver the properties of fat but not its energy content. Nevertheless
fat has its important place in the diet and should not be regarded as a dietary constituent that
is entirely dispensable.
II. LIPIDS IN THE BODY

Table 1 indicates the main types of lipids in the human body and their functions. Briefly, the
properties of biological membranes are strongly influenced by the degree of “fluidity” of the
lipid bilayer, which is in turn dependent on the nature of the phospholipid fatty acids, the
ratio of cholesterol to phospholipids, and interactions between the phospholipid polar head-
groups and membrane proteins [3].
Fluidity can be influenced by dietary lipid intakes but seems to be regulated within certain
limits by subtle changes in the proportions of phospholipids, cholesterol, and fatty acids. The
degree of fluidity of the membrane is important for the proper functioning of membrane pro
teins that act as enzymes, transporters of small molecules across the membrane, or as recep
tors for substances such as hormones, antigens, and nutrients. For this reason, the composition
of structural lipids in specific tissues is fairly conservative, in contrast to that of storage lipids,
which may vary widely.
The body’s energy reserves are mainly in the form of triacylglycerols in adipose tissue.
Milk fat triacylglycerols provide a source of energy for the newborn baby. The fatty acid
composition of human storage fat is strongly influenced by the fatty acid composition of the
dietary fat.
Table 1 Types of Lipids in the Human Body and Their Functions
Structural Lipids» These lipids are present as a bilayer in biological membranes, as a layer on the
surface of the skin, and in the lung surfactant lipid. Because of their specialized functions, their compo
sition is conservative. Although fatty acid composition can be influenced by diet, homeostatic mecha
nisms operate to maintain physical properties within a limited range. The main components are the fol
lowing.
Phospholipids. Major components. The “fluidity” of membranes is regulated by the degree of unsat
uration of the fatty acids and interactions between the polar headgroups and the membrane proteins.
Fatty acids may be highly saturated, as in the lung surfactant, or highly unsaturated, as in mitochondrial
membranes. The phospholipid fatty acids provide a store of long-chain PUFA for eicosanoid formation.
Glycolipids. Minor components except in neural membranes.
Cholesterol. Major component. Regulates “fluidity” of bilayer. Membrane function is absolutely
dependent on cholesterol. Even minor modifications to the structure of the steroid molecule lead to loss
of function. Also provides precursor for synthesis of steroid hormones (adrenal gland), bile acids (liver),
and vitamin D (skin).
Membranes also have many minor components, such as vitamin E as part of the lipid bilayer.
Storage Lipids. Triacylglycerol fatty acids provide long-term storage of fuel that can be mobilized
during food shortage, for growth, and during vigorous exercise. The main storage depot is the adipose
tissue. Milk fat provides a store of energy for growth of infants. Storage fatty acid composition is
strongly influenced by the fatty acid composition of the diet. Small amounts of other lipids are dissolved
in the storage triacylglycerols: cholesterol, cholesteryl ester, fat-soluble vitamins, and adventitious fat-
soluble compounds derived from the diet.
Lipids in Transit. Fatty acids transit between adipose stores and structural sites via the blood esterified
in lipoprotein lipids or as nonesterified fatty acids carried on blood albumin.
Lipids and Nutrition 81
Lipids in both structural and storage pools are continually being replaced as fat stores
become used as body fuels and as structural lipids become precursors for a variety of meta
bolic end products (see Sections IV and V).
III. LIPIDS IN FOOD

A. Sources of Dietary Lipids
Lipids in the human diet derive from the storage and structural lipids of plants and animals.
More than 90% of dietary lipids are triacylglycerols originating from the adipose tissue or
milk of animals or from plant seed oils, mainly in the form of manufactured products. While
much of this is “visible fat” (the fat on a joint of meat, the cream on milk, spreading fats, or
cooking oils), there is also much fat in the diet that is “hidden,” such as the fat incorporated
in baked foods, pastries, and confectionery goods and that present in processed meats,
cheeses, and desserts. The phospholipid fatty acids of the structural lipids tend to be forgotten
because they are also hidden but, although quantitatively in small amounts, they are neverthe
less nutritionally very important. Table 2 indicates current sources of fat in the U.K. diet [4].
B. Composition of Food Lipids

Figure 1 illustrates the major classes of fatty acids provided by different types of lipids used
in human foods. Detailed information on fatty acid compositions of individual food lipids can
be found in Refs. 3 and 5-7 .
Seed oils and animal storage fats provide the bulk of the lipids consumed in foods. How
ever, some foods, although contributing little total fat to the diet, are nevertheless important
sources of specific essential fatty acids. Note in particular the contrast between the n-6 PUFA-
rich lipids of animal structural lipids and the n-3 PUFA-rich lipids of plant leaves. Human
milk fat (whose composition is not illustrated in this figure) has a composition very different
from that of cow’s milk, a fact that is of great importance in human nutrition. The reader is
referred to Jensen et al. [8] for further details of the composition of human milk fat and to
Table 2 Contribution of Fat to the U.K. Diet by Various Foods
1981 1983 1985 1987 1993 1994
Fat in grams per person per day

All dairy products 32 30 27 24 19 18
All meat 28 26 25 25 21 19
Margarine and other fats 26 26 24 26 18 20
Biscuits, cakes, pastries 6 7 6 7 7 7
All other foods 12 12 14 14 19 18
All foods 104 101 96 96 84 82
Fat as percent of dietary energy
Total fat 42 43 43 42 41 39
Saturated fatty acids 19 19 18 17 16 15
Ratio of polyunsaturated to
saturated fatty acids (P/S) 0.25 0.29 0.32 0.37 0.43 0.44
Energy intake
Kilocalories 2210 2144 2016 2040 1930 1884
Megajoules 9.23 8.96 8.42 8.52 8.10 7.87
Source: Ministry o f Agriculture, Fisheries and Food: National Food Survey for 1993, HMSO, London, 1994.
82 Gurr
___ STRUCTURAL FATS

Meat Leaf
SMCFA
12-16SFA -f+ + HIGH PUFA SEED OILS
LAURIC OILS >16SFA - Sov S u n
18MÜFA -f-f- SMCFA _
SMCFA >18MUFA -
12-16SFA ++++ 12-16SFA 4 4
t-MUFA - - >16SFA
>16SFA - n-6PUFA 4-+
18MUFA 18MUFA +4 4+ 4
n-3PUFA 4+4+ >18MUFA -
>18MUFA
t~MUFA _ t-MUFA
n-6PUFA n-6PUFA ++++ ++++
n-3PUFA ” n-3PUFA 4 -
DAIRY FISH OILS

Cod Herring
SCMCFA + SMCFA - -
12-16SFA 12-16SFA 4+ 4+
>16SFA + >16SFA ».
18MUFA ++ 18MUFA 4 4
>18MUFA - >18MUFA 4+ +4+
t-MUFA t-MUFA _ -
n-6PUFA - n-6PUFA -
n-3PUFA - n-3PUFA +4+ “
ANIMAL FATS HIGH MUFÄ OILS

______Lard Beef Lamb
SMCFA - ... Raoe Olive
SMCFA
12-16SFA +4 4+ 4+
12-16SFA 4 4
>16SFA 4 4 4
PALM
>16SFA - -
18MUFA
>18MUFA
4+ 4
-
+4+
-
+4+
- SMCFA — 18MUFA ++++ ++++
12-16SFA +4+ >18MUFA
t-MUFA 4 4
t-MUFA -
n-6PUFA 4 4
>16SFA -
n-3PUFA - _ 18MUFA 4+4
»
n-6PUFA 44 4
>18MUFA n-3PUFA 4 -
t-MUFA -
n-6PUFA 4
n-3PUFA -
Fig. 1 At-a-glance indicator of the principal fatty acid classes in natural fats used in foods. The fats
and oils illustrated are natural fats and oils used as ingredients, not manufactured fats. SMCFA, short-
and medium-chain fatty acids, carbon chain lengths 4-10; 12-16SFA, saturated fatty acids with chain
lengths between 12 and 16 carbon atoms; >16SFA, saturated fatty acids with chain lengths greater than
16 carbons, mainly stearic acid; 18MUFA, monounsaturated fatty acids with chain length of 18 carbons,
almost entirely oleic acid; >18MUFA, monounsaturated fatty acids with chain lengths greater than 18
carbons, mainly 20 and 22 carbons; rMUFA, trans-nnsatmaXed fatty acids, mainly 18-carbon monounsat
urated; n-6PUFA, polyunsaturated fatty acids of the n-6 family, mainly linoleic and arachidonic acids;
n-3PUFA, polyunsaturated fatty acids of the n-3 family, mainly a-linolenic acid in plant fats and eicosa-
pentaenoic and docosahexaenoic acids in fish oils.
Goedhart and Bindels [9] for a detailed discussion of the contribution of milk fat to the
nutrition of infants. Important differences in fatty acid composition lie mainly in the content
of the long-chain PUFA (20 carbon atoms and above) of both the n-3 and n-6 families in
human milk, which are important in the development of the infant’s nervous system. Apart
from differences in the fatty acid composition of the two milks, differences in the structure of
the triacylglycerols are important to the digestion of milk fat. Human milk triacylglycerols
resemble pig storage fat in that a high proportion (70%) of palmitic acid is esterified at the
sn-2 position of glycerol, while the primary positions contain a high proportion of unsaturated
fatty acids. Cow’s milk triacylglycerols, however, have a much smaller proportion of 16:0
(40%) at position 2 and relatively large proportions of saturated fatty acids at positions 1
(1 6 :0 , 18:0) an d 3 (4 :0 , 6:0) [6].
Animal fats provide on average 400 mg cholesterol per day in the U.K. diet and small
amounts of fat-soluble vitamins: retinol (vitamin A), vitamin D, and tocopherol (vitamin E)
(Section V.E). Seed oils and plant leaves also supply carotenoids, tocopherols, tocotrienols,
and plant sterols, although the latter are not absorbed by the human gut.
IV. METABOLISM OF LIPIDS

The following is a brief account as a background to understanding the nutritional roles of
lipids. A detailed account can be found in Refs. 3 and 10.
A. Digestion and Absorption

Lipid digestion in adults takes place mainly in the upper part of the small intestine. Hydrolysis
of triacylglycerols catalyzed by the enzyme pancreatic lipase generates fatty acids and mono-
acylglycerols. These are absorbed by the enterocytes lining the small intestine and reconverted
into triacylglycerols in the enterocytes. They are packaged, together with other fat-soluble
substances absorbed from the diet, with a stabilizing coat of proteins and phospholipids, as
chylomicrons (Fig. 2).
The chylomicrons, irrespective of the fat eaten, consist mainly of triacylglycerols esterified
with fatty acids of chain length 14 carbon atoms or longer. Short-chain fatty acids (from milk
and dairy products) and medium-chain fatty acids (from milk and coconut oil) are selectively
transported into the portal blood as nonesterified fatty acids and are not incorporated into
chylomicrons. The short- and medium-chain fatty acids are rapidly metabolized in the liver.
Normally, they do not contribute to plasma lipids nor are they deposited in adipose tissue
unless consumed in very large quantities [ 1 1 ].
B. Transport
Chylomicrons contain triacylglycerols, cholesterol, cholesteryl esters, phospholipids, and
other minor lipids in different proportions as well as specific proteins known as apoproteins,
designated apo A, B, C, D, and E. The latter not only help to stabilize the particles but also
provide a means by which the different lipoproteins are recognized by receptors on the surface
of cells, thus determining their ultimate fate.
Chylomicrons, which are the main carriers of dietary lipids, are cleared rapidly from the
bloodstream as a result of hydrolysis by the enzyme lipoprotein lipase (LPL) at the surface of
adipose tissue (Fig. 2). After a meal, the secretion of insulin activates adipose tissue LPL,
releasing fatty acids from chylomicron lipids for storage in adipose tissue. When stored fatty
acids are needed as fuel, insulin levels fall, LPL is “switched off,” and a lipase in adipose
tissue is activated, releasing fatty acids into the blood. LPL in muscle then becomes active,
enabling that tissue to use fatty acids as a fuel.
Chylomicrons are not totally degraded by lipoprotein lipase. Smaller, denser lipoprotein
particles called chylomicron remnants remain and are further metabolized in the liver to low
density lipoprotein (LDL), which is depleted in triacylglycerols and enriched in cholesterol.
The uptake of cholesterol by tissues can occur by nonspecific diffusion across the cell mem
brane or by interaction with specific LDL receptors on the cell surface. After binding to the
receptor, the lipoprotein-receptor complex is taken into the cell, where the lipoprotein is
broken down. The resulting free cholesterol interacts with intracellular membranes to inhibit
a key enzyme in cholesterol biosynthesis. The intracellular production of cholesterol is thus
w
Fig. 2 The metabolism of lipoproteins. The endogenous pathway is concerned with the transport and metabolism of
fat coming from the diet. The fats are packaged as chylomicrons, which circulate and are removed from the plasma by
the enzyme lipoprotein lipase, mainly in the adipose tissue. The chylomicrons are not entirely consumed by this enzyme
but are degraded to smaller particles called remnants, which are removed by the liver. Parts of the chylomicrons also go
into making HDL, which, in conjunction with the enzyme LCAT, remove excess cholesterol from membranes and other
lipoprotein particles, converting it into cholesteryl esters. This process involves the interconversion of two forms of
HDL— HDL2 and HDL3— and cholesteryl esters are taken to the liver for further processing. The endogenous pathway
is concerned with the transport and metabolism of fats made in the body itself. The products are the VLDL, which are
metabolized in a manner analogous to the chylomicrons. Their remnants are usually called intermediate density lipopro
O
teins (IDL) and are further metabolized to LDL, which are taken up by the specific LDL receptor. (From Ref. 8 6 .)
reduced and does not become reactivated until circulating cholesterol concentrations (re
flecting, in part, dietary intake of cholesterol) begin to fall. When the concentration of choles
terol in plasma is low, the number of receptors for LDL is low and the rate of endogenous
cholesterol synthesis is high. When the concentration of cholesterol in plasma rises, the num
ber of receptors increases, bringing cholesterol into the cell and shutting down the cell’s own
production. This elegant mechanism enables the body to maintain its supply of cholesterol
within well-controlled limits depending on the supply of dietary cholesterol. This is necessary
because cholesterol is a vital compound, yet in large concentrations it may have undesirable
metabolic effects.
The liver is the main site for making new fat from carbohydrate or recycling old fat mole
cules, which it exports as very low density lipoprotein (VLDL) into the blood. This circulates
and is metabolized in a very similar manner to that described for the chylomicrons.
High density lipoprotein (HDL) plays a vital role as a scavenger for cholesterol. Choles
terol is present in plasma lipoproteins mainly as cholesteryl esters. A blood enzyme converts
unesterified cholesterol in HDL into cholesteryl esters. The HDL, which is now depleted of
unesterified cholesterol, interacts with membranes, blood vessel walls, or other lipoprotein
particles that may have an overabundance of cholesterol, removes the unesterified cholesterol,
and transports it to the liver, where it is channeled into other metabolic pathways. The HDL,
now depleted of cholesterol, is returned to plasma to continue the scavenging cycle.
C. Biosynthesis
Human beings living in developed countries generally consume diets that contain a high pro
portion of energy as fat. Under these conditions, the enzymes of fat biosynthesis are “switched
o ff’ or working at a low level and the needs for both structural and storage fats are satisfied
from dietary intake. However, the phenomenon of turnover—the building up and breaking
down of lipids—proceeds continuously in all tissues, and so a low level of enzymatic activity
may be present at all times. Current opinion is that fatty acid synthesis from carbohydrate is
an unimportant process in human beings, at least in adipose tissue, even when there is little
fat in the diet.
High rates of fatty acid biosynthesis can occur in the mammary gland during lactation.
The medium-chain fatty acids, caprylic (octanoic) and capric (decanoic) acid, are produced
specifically in the mammary gland and in no other tissue. They can therefore act as markers
of endogenous fatty acid biosynthesis. When the amounts of fatty acids in the human diet are
very low, there is a marked elevation of the milk medium-chain fatty acids compared with
concentrations in the milk of women eating a high fat diet. When the diet contains an appre
ciable amount of fat, this activity is suppressed, and much of the milk fat is derived from
circulating lipoproteins that are broken down by mammary gland lipoprotein lipase activated
by the hormone prolactin.
Diet may influence the regulation of lipid metabolism in several ways. One is to increase
or restrict the availability of molecules that are substrates for the enzymes in a metabolic
pathway. Another is the obligatory supply from the diet of a coenzyme, a small molecule that
must be attached to the enzyme before the enzyme can act as a catalyst. The enzymes of fatty
acid biosynthesis, for example, require pantothenic acid and biotin, both B-group vitamins,
as coenzymes. Deficiency of these vitamins, therefore, leads to defects in lipid biosynthesis,
which may have profound effects on health. An important effect of diet in the regulation of
lipid metabolism is to bring about changes in the concentrations of specific circulating hor
mones, insulin being among the most important, that induce or repress the biosynthesis of
some important enzymes of lipid metabolism. Currently there is much research into the ability
of different fatty acids and other lipids to regulate the expression of various genes [ 10].
86 Gurr
D. Essential Fatty Acids

1. Essential Fatty A cid Deficiency
In 1929, Burr and Burr [12] described how acute deficiency states could be produced in rats
by feeding fat-free diets and how these deficiencies could be eliminated by adding only certain
specific fatty acids to the diet. It is now recognized that the two primary essential fatty acids
(EFA) for humans are linoleic and a-linolenic acids, because the body is unable to synthe
size them.
Well-documented EFA deficiency in humans is rare but was first seen in children given
fat-free diets when they developed skin conditions similar to those produced in rats. The skin
abnormalities and other signs of EFA deficiency disappeared when linoleic acid was added to
the diet [13]. More recently, it has become apparent that EFA deficiency is secondary to a
number of disease conditions, so that part of the disease at least is amenable to dietary treat
ment [14]. Individuals who are unable to absorb long-chain fatty acids, either because of
gastrointestinal disease or as a result of surgery [15], are as close to being deficient in essential
fatty acids as any group yet studied. Other vulnerable groups are hospital patients on intrave
nous alimentation with glucose and amino acid mixtures [16].
2. EFA M etabolism
Animal tissues contain enzymes called desaturases that insert double bonds into saturated fatty
acids, normally beginning at position 9 [3]. The enzyme bringing about this transformation is
9-desaturase. Enzymes are also present that catalyze the insertion of more double bonds to
produce PUF A, but in animal tissues this can occur only at positions between the first double
bond and the carboxyl group. The enzyme responsible for this second desaturation is 6-desa-
turase, which inserts a double bond beginning at position 6. Thus, oleic acid is converted into
cis, cis-6,9-octadecadienoic acid (n-9), whereas linoleic acid (n-6) is converted into all-cA-
6 ,9 ,12-octadecatrienoic acid (n-6), and a-linolenic acid (n-3) is converted into all-cis-
First
Source Family member Desaturation Elongation Desaturation
+C,
Accumulates
Diet or 8 , 11 - 20:2 5 , 8 , 11 - 20:3 in EFA
endogenous deficiency
synthesis
8 , 11 , 14 - 20:3 5 , 8 , 11 , 14- 20:4

arachidonic 1. Components
Diet only of membranes
2. Precursors
of eicosanoids
8 , 11 , 14 , 17 - 20:4 5 , 8 . 11 , 14 , 17- 20:5
Fig. 3 The metabolism of three families of unsaturated fatty acids. The first member of the n-9 family,
oleic acid, can be taken in from the diet or can be formed in body tissues, whereas linoleic and a-
linolenic acids, the first members of the n-6 and n-3 families, respectively, can only come from the diet.
The first step in their metabolism is the introduction of a new double bond at position 6 by the enzyme
6 -desaturase. All three fatty acids compete for this enzyme. There follows a series of alternate competi
tive elongation and desaturation steps. The products formed from oleic acid accumulate only when
linoleic acid is absent from the diet or present in only a very small amount. They are, therefore, markers
of essential fatty acid deficiency. Arachidonic acid (n-6 family) is the major long-chain polyunsaturated
fatty acid in human membranes of most tissues; DHA (n-3 family) is a characteristic component of
nervous tissue. (From Ref. 8 6 .)
6,9,12,15-octadecatetraenoic acid {n-3) (Fig. 3). O leic, lin o leic, and a-lin o le n ic acids are the
simplest members of the three major families of polyunsaturated fatty acids designated as n-
9, n-6, and n-3, respectively.
An elongation by two carbon atoms and another double bond insertion at position 5 by the
5-desaturase then follow. Although the next higher PUFA contains a double bond in position
4, there is no evidence for the existence of a 4-desaturase. Instead there are two sequential
chain elongations after the 5-desaturation followed by a further 6-desaturation. In the n-3
family this results in the formation of 24:6 (ai-3), and retroconversion (chain shortening) of
this metabolite results in the formation of docosahexaenoic acid [DHA, 22:6 (n-3)].
The long-chain PUFA products of EFA metabolism are further metabolized to biologically
active compounds called eicosanoids (Section V.A. 6).
E. Breakdown of Tissue Lipids

1. Lipolysis
All tissues contain lipases, which are enzymes that catalyze the cleavage of ester bonds in
lipids [3]. As a result of the concerted activity of lipases and acyltransferases (enzymes that
catalyze esterification) there is a continual turnover, or replacement, of all parts of lipid mole
cules in cells that allows fine control of their metabolism. Especially important is the mem
brane phospholipase that liberates essential fatty acids for the biosynthesis of eicosanoids.
2. Oxidation o f F atty A cids fo r Cellular Energy

The process known as beta oxidation is a controlled step-by-step breakdown of fatty acids to
yield metabolic energy [3]. Beta oxidation is catalyzed by a group of enzymes in a specialized
compartment of the cell called the mitochondrion. The complete oxidation of one molecule of
the 16-carbon fatty acid palmitic acid yields eight molecules of acetic acid, which are further
degraded to give carbon dioxide and water. During this process, the substance adenosine
triphosphate (ATP) is generated. ATP is the most important source of chemical energy in the
cell and is required to provide the power for many biosynthetic reactions (including the bio
synthesis of lipids).
3. P eroxidation
Peroxidation of lipids leads to food spoilage and to a reduction in nutritive value [3,5,17]. It
may also occur in the living body. When it does so in an uncoordinated way as a result of
tissue disorganization or local lack of antioxidant defense, it may lead to free radical damage
to tissues and quite possibly may contribute to disease. However, carefully regulated lipid
peroxidation is a mechanism by which essential fatty acids (EFA) are converted into eicosa
noids, which are intimately involved in metabolic regulation in all parts of the body [3]
(Section V.A. 6).
V. NUTRITIONAL CONTRIBUTIONS OF VARIOUS

FOOD LIPIDS
A. T riacylglycerols
L Palatability o f Food: Flavor, A rom a, and Texture
Food is nutritious only insofar as it is good to eat, and the presence of fat contributes substan
tially to the palatability of food [5].
88 Gurr
2. R ole in Storage and Provision o f M etabolic Energy

The energy value of triacylglycerols to the body is about 38 kJ (9 kcal) per gram, compared
with about 17 kJ for proteins and 16 kJ for carbohydrates. As storage fuels, triacylglycerols
have the advantage that they can be stored in anhydrous form, representing more energy for
less bulk than complex polysaccharides such as food starches or body glycogen, which are
highly hydrated.
Although the gross energies of fatty acids with different degrees of unsaturation (as mea
sured by bomb calorimetry) differ slightly, these differences are insignificant nutritionally.
The useful energy available to the body (metabolizable energy) depends mainly on the digest
ibility of the fat and on the chain length of the fatty acids. Most common food fats are
efficiently digested by most people with insignificant differences in their digestible energies.
However, saturated (and to a lesser extent unsaturated) fatty acids with chain lengths of more
than 18 carbons are less well digested and absorbed, the efficiency falling as chain length
increases. The energy values of medium- and short-chain fatty acids are considerably lower
than those of the longer chain acids (Section VI.G.3).
3. Supply o f Structural Lipids

Dietary fats also provide a supply of fatty acids for the body’s structural lipids. These will be
considered under the appropriate fatty acid classes.
4. Saturated Fatty A cids

Saturated fatty acids (SFA) are major components of the phospholipids in the membrane lipid
bilayer, accounting for 15-50% of membrane fatty acids. The presence of saturated fatty acids
is important in achieving the appropriate degree of fluidity for each particular membrane. The
phospholipids of the lung surfactant lipid are esterified almost entirely with palmitic acid.
Without this, the lungs would collapse during expiration. Thus although saturated fatty acids
are not dietary essentials, they can be regarded as essential components of membrane struc
tures.
Nonspecialists normally think of saturated fatty acids in terms of their alleged capacity to
raise blood cholesterol, since this topic has received much publicity in relation to dietary fats
and health (see Section VII). Different SFA are by no means equivalent in their influence on
blood cholesterol, however [18].
Short- and medium-chain fatty acids (4:0, 6:0, 8:0, 10:0), particularly abundant in coconut
and milk fats, have no effect on plasma cholesterol. Bonanome and Grundy [19] confirmed
that stearic acid (18:0) has no effect on plasma cholesterol, an observation originally made in
the 1960s.
Of the many SFA in the human diet, only three are now thought to have a marked influence
on plasma cholesterol: lauric (12:0), myristic (14:0), and palmitic (16:0), and the main lipo
protein fraction affected is LDL. Hayes and Khosla [20] propose that the different hypercho-
lesterolemic fatty acids have different thresholds at which they exert an effect on plasma
cholesterol, which are dependent on the concentrations of linoleic acid (18:2, n-6) and choles
terol in the diet and the initial plasma LDL concentration. At a level of 18:2 in excess of
about 6.5% of dietary energy, the LDL receptors are fully active and are able to remove LDL
particles from plasma and maintain a relatively low plasma concentration of LDL. At lower
dietary 18:2 inclusions, 14:0 overrides the effects of 18:2 and reduces the activity of the
receptors. Palmitic acid is less potent, and more of this fatty acid is required to reduce receptor
activity and only at very low intakes of 18:2.
5. M onounsaturated A cids
Most naturally occurring fatty acids of dietary significance have double bonds in the cis geo
metrical configuration. Oleic acid (cw-9-18:l) is the only monounsaturated fatty acid (MUFA)
of quantitative nutritional significance. It is present widely in foods and is particularly abun
dant in olive and rapeseed oils. The cholesterolemic effects of MUFA have been reevaluated
during the last decade. Although not all experiments have been well controlled [21], most
found that MUFA, when substituted for SFA, lowered plasma total cholesterol as effectively
as n-6 PUFA [18]. The lowering was almost entirely associated with the LDL fraction. When
substituted for carbohydrates, MUFA resulted in a similarly low plasma LDL cholesterol but
did not elicit the rise in VLDL (and, therefore, triacylglycerols) often seen with high carbohy
drate diets. There is also a suggestion that the inclusion of MUFA in diets also helps to
maintain a high concentration of HDL in contrast with high carbohydrate diets. They also
have the advantage over PUFA that they are less prone to oxidation.
While there is little evidence that the very long chain MUFA (erucic, cetoleic) found in
older varieties of rapeseed oil or in some fish oils (e.g., herring) influence plasma cholesterol,
concern has been expressed that they may result in lipidosis and finally necrosis of heart
muscle. However, there is little evidence of harm from the quantities normally ingested in
practical human diets [22,23].
A small proportion of dietary MUFA has the double bond in the trans geometrical configu
ration. Although some trans double bonds occur naturally, many are introduced as a result of
industrial hydrogenation. Their nutritional properties are discussed in Section VI.D.
6. P olyunsaturated Fatty A cids

Roles o f Essential Fatty Acids The essentiality of linoleic and a-linolenic acids is depen
dent on their metabolism to the longer chain n-3 and n-6 PUFA, which have at least two vital
functions [3,10].
First, they play a role in stabilizing biological membranes by creating physical properties
that are optimal for the transport of substances across the membrane and for the biochemical
reactions occurring in the membrane. For example, the membranes of liver mitochondria of
EFA-deficient animals have smaller proportions of linoleic and arachidonic acids and larger
proportions of oleic and all-cis-5, 8,11 -eicosatrienoic acid than those of healthy animals and
are less efficient at oxidizing fatty acids and synthesizing ATP, which is a vital source of
chemical energy in cells.
In contrast with tissues like liver, kidney, and muscle, whose PUFA tend to be predomi
nantly of the n-6 family, nervous tissue and the retina of the eye have larger proportions of
the longer chain acids with five and six double bonds, predominantly of the n-3 family. Their
main precursor is a-linolenic acid, although the preformed long-chain PUFA may be obtained
in the diet, mainly from fish oils. Fatty acids of the n-3 family may have a specific role in
vision and brain function, and there is evidence to suggest that when the ratio of n-3 to n-6
fatty acids is modified experimentally in rat tissues some changes do occur in the electroretino-
grams and in behavioral responses. Although such a clear role has not been established in
humans, the circumstantial evidence for the essentiality of a-linolenic acid is compelling
[13,24].
Recent research suggests that the capacity of human tissues to convert 18:3 (n-3) into 22:6
(n-3) is limited, and if this is so, then a dietary supply of preformed 22:6 (n-3) may be quite
important. The competition between n-6 and n-3 described earlier may enhance the impor
tance of dietary 22: 6 (n-3) because diets in which the n-6 acids overwhelm the n-3 family
(as tends to be the case in most industrialized countries) may cause a suppression of DHA
90 Gurr
formation from 18:3 (n-3) and its replacement by 22:5 {n-6) in membranes that have a specific
n eed fo r 2 2 :6 {n-3) [10].
A second major role for EFA is related to their metabolic conversion into a whole range
of oxygenated compounds, eicosanoids, which exert a range of profound physiological activi
ties at concentrations down to 10~^ g per gram of tissue. There are two basic metabolic
pathways (Fig. 4), one regulated by an enzyme called cyclooxygenase, producing prostaglan
dins, thromboxanes, and prostacyclins; the other by an enzyme called lipoxygenase, produc
ing leukotrienes.
The main sources of fatty acids for the production of eicosanoids are the membrane phos
pholipids, from which they are released by the action of phospholipase A. Phosphatidylinosi
tol, which contains a high concentration of arachidonic acid in position 2 , provides a major
store of eicosanoid precursors. A crucial point about eicosanoids is that they are so potent in
their biological activities that they must be produced close to the site where they are required
and the excess broken down as soon as their job is done by enzymes that convert them into
inactive metabolites.
Another feature of eicosanoids is that they may have different, sometimes opposing, activi
ties when they are produced in different tissues. Broadly, eicosanoids are involved in what
biologists call tissue homeostasis. This means that certain physiological functions are kept
constantly within predetermined limits so that they do not under- or overperform. A good
example is the tendency of a type of blood cell called platelets to aggregate together to plug
LTA 4 LTA 5
LTB 4 Leukotrienes LTB 5

+++
4-f-f signifies strongly inflammatory

■f signifies weakly inflammatory
Fig. 4 Competition between the n-3 and n-6 polyunsaturated fatty acid pathways for production of
eicosanoids. (From Ref. 87.)
a wound. If they do not aggregate sufficiently, the wound will not be healed; if they aggregate
too much they may precipate a potentially fatal thrombosis. Thromboxanes, which are pro
duced by the platelets themselves, stimulate the platelets to aggregate, whereas prostacyclins,
produced by the walls of the arteries, inhibit the aggregation. It is the balance of these two
opposing activities that keeps the platelet aggregation system functioning properly. Many
more examples of these homeostatic mechanisms regulated by eicosanoids are being dis
covered. Other eicosanoid activities are those involved in regulating kidney function, in
flammatory reactions, reactions of the cellular immune system, secretion of hormones such as
insulin, reproduction, bone calcification, muscle contraction including the constriction and
dilation of blood vessels, electrical activity of the heart, and various aspects of visual func
tion [ 10].
Requirements for Essential Fatty Acids. It seems possible that there are different levels
of requirements for EFA for their different functions. For example, there may be a relatively
low level of requirement (as for most vitamins) to prevent overt signs of FFA deficiency such
as skin conditions and growth retardation. This may be around 1% of dietary energy as lino-
leic acid and about one-fifth of this amount as a-lin o len ic acid [24]. At a higher level, there
may be a larger but quantitatively less easily defined quantity required for such functions as
maintaining an optimal balance of membrane fatty acids, preserving a reservoir for eicosanoid
formation, and maintaining optimal concentrations of the different plasma lipoproteins. The
absolute amounts required at this level are likely to be influenced by the amounts and types
of other fatty acids in the diet because of the competition for metabolic pathways between
FFA and non-FFA as well as competition between the FFA families. Because primary FFA
may be only slowly converted into their long-chain PUFA derivatives, there may sometimes
be a dietary requirement for the latter.
n -6 Family. In addition to its requirement as an FFA, linoleic acid results in a lowering

of plasma cholesterol, and this may be important in reducing the risk of coronary heart disease
(CH D ) [21].
y-Lin o lenic acid (GLA), the first metabolite of linoleic acid, formed from it by 6-desatura
tion, has entered the folklore as having health-giving properties. In the human body, GLA is
merely an intermediate in the pathway from linoleic to arachidonic acid and does not accumu
late in mammalian cells.
The rate limitation of the 6-desaturase in the n-6 pathway is important in understanding the
claims for health benefits of GLA. 6-Desaturase activity decreases with age and is also de
pressed by the so-called stress hormones, adrenalin and cortisol, by alcohol, by fatty acids of
the n-3 family, and in the condition of diabetes mellitus. Bioassays based on water permeabil
ity of skin show that GLA has higher EFA potency than its precursor, LA. A commonly used
biochemical index of EFA deficiency has been the so-called triene/tetraene ratio. When the
dietary supply of linoleic acid is limited, oleic acid becomes the prime substrate for 6-desatur
ase (Fig. 3). The major product of this (n-9) pathway is the trienoic acid 5,8,11-20:3, and its
ratio to A A increases in EFA deficiency and decreases when LA is added back to the diet.
The ratio was reduced far more rapidly when GLA was substituted for LA [25].
Reports on the blood cholesterol lowering potency of GLA are inconsistent. One publica
tion claimed a large cholesterol-lowering effect that was disproportionate to the degree of
unsaturation compared with linoleic acid, while others have found no differences between the
effects of linoleic acid and GLA.
The literature is replete with claims for the therapeutic benefits of GLA. Disorders for
which its efficacy has been tested in controlled clinical trials include atopic dermatitis, rheu
92 Gurr
matoid arthritis, diabetic neuropathy, multiple sclerosis, ulcerative colitis, hypertension, pre
menstrual syndrome, schizophrenia, hyperactivity, and several cancers. In almost all catego
ries the results have been inconsistent.
The main sources of GLA are evening primrose, borage, and blackcurrant oils. There are
preliminary indications that evening primrose oil may be more effective in some of the physio
logical effects described above than the other oils in which GLA occurs. One possible expla
nation is that GLA is present in evening primrose oil almost entirely as molecular species of
triacylglycerol in which one GLA is combined with two LA molecules: dilinoleoyl-monogam-
malinolenoyl-glycerol (DLMG) [26]. Another possibility is that minor components of evening
primrose oil, not GLA, are responsible for some of the effects.
n-3 Family. The parent, and simplest, member of the n-3 family is a-linolenic acid, all-
cw-9,12,15-octadecatrienoic acid (a-18:3). It is synthesized only in higher plants, algae, and
phytoplankton and is present universally in chloroplast membranes, where it is present almost
exclusively as a constituent of galactosyl-diacylglycerols [3]. Some seed oils, however, con
tain a - 18:3 in their triacylglycerols (e.g., linseed, rapeseed, and soybean oils. Fig. 1).
When a - 18:3 is ingested by human beings, some is incorporated unchanged into lipids, a
little is oxidized, and some is metabolized by a series of desaturations and elongations to
longer chain PUFA. The predominant end products are eicosapentaenoic acid (ail-cis-
5,8,11,14,17-20:5, EPA) and docosahexaenoic acid (all-cw-4,7,10,13,16,19-22:6, DHA).
However, the extent to which a-linolenic acid is converted into longer chain derivatives is
limited, and the largest sources of these long-chain PUFA are the oils of fish that eat phyto
plankton containing a high proportion of a - 18: 3 in their lipids [6]. These fatty acids are
present in triacylglycerols that are stored in the animal’s liver (e.g., cod) or in the flesh (e.g.,
mackerel, herring).
Inclusion in the diet of four or more grams per day of n-3 polyunsaturated fatty acids from
fish oils maintains low blood triacylglycerol concentrations in people with normal levels and
significantly reduces their concentration in people with hypertriglyceridemia [27]. Unlike the
n-6 family of PUFA, however, there is no consistent effect of n-3 PUFA on blood cholesterol.
In some individuals, n-3 PUFA have even been known to raise blood total cholesterol. There
are small and inconsistent effects on HDL cholesterol. The triacylglycerol lowering effect is
due entirely to changes in plasma VLDL concentrations brought about by a reduction in
VLDL triacylglycerol biosynthesis in the liver and to some extent by reduced clearance and
catabolism of VLDL [27].
The conversion of long-chain n-3 PUFA into eicosanoids may explain a number of physio
logical effects of eating fish oils, including effects on blood clotting and on immune function.
Thus, consumption of about 200-300 g/day of fatty fish results in prolonged bleeding times
and reduced tendency for platelet aggregation. The effects of consumption on the order of 30
g/day or less that may be practical or acceptable in most western countries are less conclusive.
Nevertheless, in one study, consumption of at least two portions of fatty fish per week (200-
400 g) resulted in a significant decrease in mortality from coronary heart disease [28].
Inclusion of long-chain n-3 polyunsaturated fatty acids characteristic of many fish oils in
the diet is quickly reflected in elevated concentrations of these fatty acids in plasma and in
the membranes of red and white blood cells. The biochemical effect of an increase in the n-
3/n-6 PUFA ratio in cell membranes is to alter the composition of the eicosanoids formed
from these PUFA through the action of the enzymes cyclooxygenase and lipoxygenase. The
strongly proinflammatory eicosanoids from arachidonic acid are reduced and replaced by the
weakly inflammatory eicosanoids from eicosapentaenoic and docosahexaenoic acids [10]. This
knowledge has formed the basis for dietary treatments for inflammatory diseases, of which
rheumatoid arthritis has been most actively researched. Controlled dietary trials have demon
strated modest but potentially beneficial improvements in the severity of rheumatoid arthritis,
and longer term studies are now needed to establish whether dietary supplements of n-3 PUFA
should be standard recommended treatment [10,27,29].
7. D istribution o f F atty A cids in Triacylglycerols

The way in which fatty acids are distributed in a triacylglycerol may also influence plasma
cholesterol irrespective of the overall composition of the fatty acids, and this may explain
some apparent anomalies. Thus linoleic acid is more /zypocholesterolemic [30] and SFA more
/zyp^rcholesterolemic [31] when esterified in position 2 than in positions 1 or 3. Butterfat
is much less /zyp^rcholesterolemic when the positions of its fatty acids are randomized by
interesterification [32].
B. Phospholipids
Phospholipids, as intact molecules, have not been regarded as essential dietary components
because they are readily synthesized in the body. However, some revision of this view may
be necessary. The most abundant phospholipid, phosphatidylcholine, provides a source of
choline in the diet. Choline, in the form of acetylcholine, is a neurotransmitter involved in
the transmission of nerve impulses. The only known pathway for the formation of choline in
the human body is by méthylation of another phospholipid, phosphatidylethanolamine. There
fore, a dietary source of choline as phosphatidylcholine or of ethanolamine as phosphatidyleth
anolamine should probably be regarded as essential.
Phospholipids are also carriers of PUFA, but because phospholipid fatty acids contribute
only a small proportion of the normal dietary intake of fatty acids, it is debatable whether
their nutritional contribution is significant. An exception may be the supply of arachidonic
acid, present in the phospholipids of meats.
C. Glycolipids
The daily intake of glycolipids is very small. Their contribution of a-linolenic acid is im
portant. There is little evidence that any of the sugar moieties supplied by glycolipids has
dietary significance.
D. Sterols
1. Cholesterol
The influence of dietary cholesterol on plasma cholesterol is still a subject of debate. In the
United States, dietary guidelines clearly emphasize the importance of dietary cholesterol as a
determinant of plasma cholesterol, whereas in Europe the scientific evidence has generally
been interpreted to indicate that it plays a minor role if any. McNamara [33] analyzed and
reviewed 68 human intervention studies that examined the effect of dietary cholesterol on
plasma cholesterol and concluded that there is an effect but it is small compared with the
influence of SFA. A mean change of 100 mg/day dietary cholesterol causes a change of only
2.3 mg/dL plasma cholesterol. Somewhat larger reductions could be achieved by relatively
modest reductions in body fat.
2. P lant Sterols
The amounts of plant sterols consumed in omniverous diets in western countries are very
small but will be significantly higher in vegetarians. They inhibit the absorption of cholesterol.
94 Gurr
so that endogenous synthesis of cholesterol may be expected to be higher in vegetarians than

in omnivores.
E. Fat-Soluble Vitamins
Dietary fat is also important insofar as it contributes several fat-soluble vitamins; A, D, E,
and K. Requirements for the fat-soluble vitamins are of a different order of magnitude from
those of the essential fatty acids, being measured in micrograms rather than grams per day.
Some fat is necessary in the diet to improve the absorption and utilization of fat-soluble
vitamins. However, there is little evidence that, within the normal range of fat intakes, the
amount of dietary fat significantly affects the utilization of fat-soluble vitamins.
1. Vitamin A (Retinol) and C arotenoids

Retinol is found only in animal fats. Dark green-leaved vegetables and vegetable oils like
palm oil contain a precursor (provitamin A), ^-carotene, that is converted into retinol in the
body. Milk fat is relatively rich in retinol equivalents. Fish liver oils are by far the most
concentrated source of retinol, but mammalian storage fats such as lard contain none. The
British diet provides on average about twice the “reference nutrient intake” for vitamin A.
Reference nutrient intake (RNI) is defined by the U.K. Department of Health [24] as an intake
likely to meet the needs of 97% of the population. The vitamin A RNI for adults is currently
600-700 jLtg/day. Two-thirds of average U.K. intake comes from retinol itself, and about
one-third from carotene. About 21% of dietary retinol equivalents in the United Kingdom
comes from milk and milk products [4].
Vitamin A, if taken in excess, will accumulate in the liver and eventually destroy it.
Vitamin A is required for vision and in the maintenance of cellular differentiation. There
is mounting evidence that ^-carotene may have specialized roles (e.g., as an antioxidant) that
are independent of its function as a vitamin A precursor.
2. Vitamin D
Vitamin D is the name for a family of sterols with antirachitic properties. The only one of
significance is the one now called vitamin D 3 or cholecalciferol. It is present in only a few
foods, the richest sources being the liver oils of fish.
Cholecalciferol is produced by the ultraviolet irradiation of 7-dehydrocholesterol, a sterol
derived from cholesterol and widely distributed in animal fats including the skin surface
lipids.
Many people get little or no vitamin D from their diet but obtain all they require from the
action of the ultraviolet rays in sunlight on 7-dehydrocholesterol in the skin. For this reason,
many nutritionists have questioned whether vitamin D should be considered a vitamin at all.
However, at least two groups of people may have a special need to obtain vitamin D from
the diet. In the first group are infants, children, and pregnant and lactating women, whose
requirements are particularly high. The vitamin D content of cow’s milk and even human
breast milk is quite low, but manufactured infant feeds are fortified with this vitamin. In the
second group are people who are little exposed to sunlight, such as the housebound elderly
and people living in northern latitudes or those who wear enveloping clothes. Dark-skinned
immigrants to northern European countries are especially vulnerable. Infants and children who
obtain too little vitamin D develop rickets, with deformed bones that are too weak to support
their weight. The reason vitamin D preparations are provided for children and pregnant
women and margarine is fortified in the United Kingdom and some other countries is that
these degenerative changes soon become permanent if supplementation is not begun early
enough.
Dietary vitamin D is absorbed along with the bulk of the dietary fat and carried into the
blood in chylomicrons. In the liver, it is converted into 25-hydroxycholecalciferol, which is
further converted into 1,25-dihydroxycholecalciferol in the kidneys. The latter can be regarded
as a hormone that is primarily involved in the regulation of calcium and phosphorus metab
olism.
Like vitamin A, vitamin D is toxic if consumed in excessive quantities.
5. Vitamin E
Vitamin E activity is shared by a family of compounds known as tocopherols, the most active
of which is a-tocopherol. The richest sources are vegetable oils, cereal products, and eggs.
a-Tocopherol is present in the lipid bilayers of biological membranes and may play a struc
tural role there. It is known to be a powerful antioxidant and prevents the peroxidation of
unsaturated fatty acids. The products of peroxidation of unsaturated fatty acids can cause
damage to cells if the oxidative process is not kept in check, and such damage appears to be
exacerbated in animals given diets deficient in vitamin E. These observations have led to the
postulate that vitamin E plays a vital role in the prevention of the oxidation of membrane fatty
acids in living cells. It is a generally held view that intake should be considered in relation to
the PUFA content of the diet rather than in absolute amounts. A ratio of vitamin E to linoleic
acid of about 0.6 mg/g is generally recommended. In general vegetable oils containing high
concentrations of PUFA are sufficiently rich in vitamin E to give adequate protection. How
ever, with the growing interest in fish oils, which are even more highly unsaturated, the
question of vitamin E adequacy may well have to be reviewed. More unsaturated relatives of
the tocopherols, the tocotrienols, share some vitamin E activities with the tocopherols but
may also exhibit some independent biological activities.
VI. EFFECTS OF PROCESSING ON THE NUTRITIONAL

CONTRIBUTIONS OF FATS
A. Mixing or Homogenization
Mixing and homogenization are normally mild processes, but sometimes they can disrupt
food structure, releasing enzymes that degrade fats. Two classes of enzymes are of particular
importance. Lipases hydrolyze the complex lipids in food (mainly either triacylglycerols or
phosphoglycerides), leading to release of free fatty acids (FFA) [3]. When free fatty acid
concentrations reach a certain threshold, off-flavors and unpleasant aromas may develop, but
the FFA do not usually cause nutritional problems.
The other type of enzyme is an oxygenase or peroxidase that catalyzes the oxidative deteri
oration of fatty acids once they have been released by lipases. Some of these enzymes (such
as the lipoxygenase of plants) are involved in the formation of pleasant flavors characteristic
of particular plant foods (e.g., the characteristic smell and taste of freshly cut cucumber), but
oxidation of larger quantities of fatty acids may lead to off-flavors and toxic compounds.
These usually make the food unpalatable long before they reach concentrations that cause
nutritional or toxicological problems. Thus combined effects of lipase and oxygenase action
may sometimes be seen after milk homogenization, with accompanying rancidity problems.
The enzymes are located in the milk fat globule membrane and cause the release and oxidation
of unsaturated fatty acids present in the membrane phospholipids [34].
96 Gurr
B. Heating
The functions of heating, which may be done industrially or in the home, are to make food
palatable and also to make it safe by k illin g microorganisms. However, heating may reduce
heat-labile and oxygen-sensitive nutrients, so there is a need to achieve a balance between
microbiological safety and nutritive value.
Some changes occur during heating in which oxygen is not involved. These may occur,
for example, in the bulk of a deep fat fryer where oxygen has little access. Combination of
radicals under these circumstances gives rise to polymeric products formed from the triacyl-
glycerol molecules in the heated fat. The first substances to be formed are cyclic monomers
of triacylglycerols followed by formation of polymeric triacylglycerols.
Fats heated under normal household conditions (at, say, 180°C) generate 10 -2 0 % of poly
meric material, but at this level the functional properties of the oil are not noticeably impaired,
and feeding experiments with rats have not demonstrated any harmful effects. None is appar
ently deposited in adipose tissue [35].
The most important changes during heating involve reaction with oxygen from the air.
Oxidation can take place when the fat is stored in the presence of oxygen at room or even
refrigerator temperature but is greatly speeded up by heating and in the presence of trace
metal catalysts (e.g., iron or copper).
The major pathway is attack by oxygen at the double bonds of unsaturated fatty acids with
the initial formation of a peroxide [3,5,36]. The more highly unsaturated the fatty acid, the
greater its susceptibility to oxidation.
Free radicals produced during lipid peroxidation can reduce concentrations of essential fatty
acids (EFA), vitamin A, and vitamin E [36]. They may also cause damage to proteins, includ
ing enzymes, and DNA and may also generate carcinogens. Lipid peroxidation may be held
in check by the presence of lipid-soluble antioxidants: a-tocopherol (vitamin E) and carotenes
(both natural) and butylated hydroxytoluenene or butylated hydroxyanisole (synthetic).
Cholesterol can also autoxidize, especially when present in dried and/or heated foods (e.g.,
dried egg powder). Products include 25-dihydroxycholesterol, 7-ketocholesterol, 7-hydroxy-
cholesterol, 5- and 7 -hydroperoxides, and cholesterol-3,5,6-triol. These compounds have been
shown to cause arterial disease in rabbits at levels at which cholesterol itself is ineffective
[37].
C. Irradiation
Irradiation is also a process that may give rise to free radical formation and, therefore, lipid
deterioration. This process is useful for controlling insect infestation of foods, reducing total
microbial load, and reducing the numbers of specific pathogens in foods. It is not useful for
all foods, and in many cases it can cause deterioration in palatability through poor flavor
and texture.
Vitamins E, K, and ascorbic acid are particularly susceptible to irradiation damage. Sur
prisingly, carotene has not been shown to be much affected.
Some highly unsaturated oils may suffer extensive damage due to a combination of free
radical formation and reduction of antioxidant nutrients that would normally reduce the extent
of free radical damage. Thus, these two types of damage may reinforce one another. When
mixtures of unsaturated fatty acids and proteins have been irradiated prior to animal feeding,
the quality of the protein was found to be reduced, probably due to protein damage by peroxi-
dized lipids. In general, however, irradiation does very little direct nutritional damage.
D. Hydrogenation
Vegetable oils and, in some countries such as the United Kingdom, fish oils are often hydro
genated industrially to provide fats of more desirable physical properties, texture, and keeping
quality; these can be used in, for example, the manufacture of margarines, cooking fats, and
shortenings. During this process, there is a decrease in total unsaturation and an increase in
the concentration of saturated fatty acids, and some new isomeric fatty acids are usually
produced. Double bonds are shifted along the hydrocarbon chain, and some geometrical isom
erization from cis to trans occurs, but the amounts of isomers vary according to the processing
conditions and the starting material [38]. Hydrogenated soybean oil or fish oil may contain as
much as 50% by weight of the fatty acids as acids containing one or more trans sites. The
final concentration, however, in a blended margarine or cooking fat will be much lower.
Monoenoic acids represent the major fraction, with trans unsaturation between positions 6
and 14. The most abundant isomers are those with unsaturation at positions 9, 10, and 11
[39,40]. A small proportion of isomers with diunsaturation (cis,trans and trans,cis) is also
found, but improvements in catalysts have reduced the amounts of trans,trans isomers in
modem hydrogenated fats almost to zero. A range of positional cis isomers are also produced
during hydrogenation, but their biological properties have been largely unresearched.
In mminant fats, the proportion of the trans group of acids is usually around 5% of the
fatty acids present. The isomers present are qualitatively similar to those in industrially hydro
genated fats, but the quantities are different; the isomer with a trans bond in position 11
predominates [40].
Average consumption of trans fatty acids in the United Kingdom in the early 1990s is
estimated to have been 4 -6 g per person per day, representing about 5.5% of dietary fat
intake or 2% of dietary energy. It has been estimated, however, that some individuals may
consume up to 30 g of trans fatty acids per day and others as little as 1-2 g. A recorded
decrease in the United Kingdom in recent years may have come about partly as a result of
different (and probably better) methods of estimating intakes and partly because of a tme
decrease in consumption. Intakes of mminant fats have tended to decrease, and the trans fatty
acids content of hydrogenated fats has also fallen as a result of changes in raw materials and
processing methods. In the United States, average current intakes are based on estimates from
limited data but appear to be higher than in the United Kingdom (8-13 g per person per day).
The digestion and absorption of isomeric fatty acids, their transport around the body, their
storage as body fat, or their utilization directly as a source of energy by biological oxidation
are all similar to those processes for the more common fatty acid components of fat in the diet.
A number of studies suggest that trans fatty acids are treated more like saturated than
unsaturated fatty acids by the biochemical systems responsible for the biosynthesis of mem
brane components. This is probably because the presence of a trans unsaturated bond pro
duces an overall molecular shape at this site in the molecule similar to that of a saturated
fatty acid.
In general, the proportion of trans fatty acids incorporated into fats at the various sites in
the body is dependent on the proportion present in the diet. As with all fatty acids, however,
once incorporated the trans fatty acids are subject to turnover. There is no evidence from
animal studies or the limited data from human experiments that trans fatty acids accumulate
specifically at any site in the body.
If the level of trans fatty acids in the diet is disproportionately high, the metabolism of
EFA may be impaired and signs of essential fatty acid deficiency may develop. This may
arise from either inhibition of or competition for desaturases. However, similar competition
98 Gurr
occurs with c w -M U F A , so the effect is not a unique property of trans fatty acids. Overt signs
of EFA deficiency as a result of trans fatty acid consumption have not been recorded in
humans. Nevertheless, some scientists believe that metabolic changes may occur that are
characteristic of the early stages of EFA deficiency yet produce no overt clinical signs. Such
changes, they argue, could involve subtle impairment of membrane function or of physiologi
cal functions such as the integrity of the vasculature. Such ideas need further research, but
until such effects have been clearly demonstrated in practice it must be concluded that there
is no sound evidence for specific adverse effects of trans fatty acids on essential fatty acid me
tabolism.
There is preliminary evidence that trans fatty acids from a mother’s diet cross the placenta
during pregnancy and interfere with EFA metabolism in the newborn. There appears to be an
inverse association between the mother’s trans fatty acid intake and the baby’s birthweight
[41], and this needs further research.
Taken together, several well-conducted human studies [38,42,43] have shown that dietary
18-carbon fatty acids with a single trans double bond increase the concentration of LDL
cholesterol in the blood compared with a similar amount of dietary 18-carbon cw-monounsa-
turated fatty acid. This LDL cholesterol raising effect was directly related to the proportion
of trans fatty acids in the dietary fat and was a little lower than that of the 14- and 16-carbon
saturated fatty acids. Unlike the saturated fatty acids, however, which tend to raise HDL
cholesterol slightly, the trans fatty acids significantly lowered HDL cholesterol. There appears
to be a dose-related effect, at least at intakes between 3 and 1 1 % of total energy. Average
intakes in the United Kingdom are estimated to be about 2% of energy, and it is unlikely that
such levels of consumption contribute more than a small part of the risk of CVD in that
country. More information is required on the physiological effects of trans fatty acids from
hydrogenated fish oils. There is evidence from at least two studies that trans acids in the diet
lead to increases in the concentration of plasma lipoprotein (a) [Lp(a)] (an independent risk
factor for CVD), especially in those individuals who already have high baseline plasma con
centrations of this lipoprotein [44,45].
There is no firm evidence about the influence of trans fatty acids on thrombosis.
A large epidemiological study of 85,000 American nurses [46] reported that women con
suming the highest levels of trans fatty acids had a 50% greater risk of CVD than those
consuming the lowest amounts. The association was significant only for trans fatty acids from
hydrogenated vegetable oils, not from ruminant fats. In view of this apparent anomaly of high
risk from industrially produced trans fatty acids and no risk from ruminant trans fatty acids,
it is prudent to consider whether the trans fatty acids content of hydrogenated fats might have
been acting as a marker for socioeconomic status or for other aspects of an unhealthy lifestyle.
Two recent studies in which the concentration of trans fatty acids in adipose tissue was used
as an index of long-term intakes of trans fatty acids found no association between trans fatty
acid intakes and coronary heart disease [47,48].
E. Interesterification
The process of interesterification, which involves treatment of a fat with a catalyst at an
elevated temperature, leads to randomization of fatty acids on triacylglycerol molecules as
distinct from their natural stereospecific distributions. Interesterification causes changes in the
physical properties of the fats, such as crystal structure and melting point, and may influence
their absorption and metabolism. Thus butterfat in which the fatty acids have been randomized
by interesterification has minimal effect on plasma cholesterol [32], whereas natural butterfat
tends to raise plasma cholesterol [49].
In experiments with laboratory animals designed to induce atherosclerosis by including

high levels of fat in their diets, natural peanut oil was found to be strongly atherogenic, but
the same did not apply to randomized peanut oil [50]. The reasons for these findings have not
been established.
F. Fractionation
1. M edium -C hain Triacylglycerols
The most abundant sources of medium-chain triacylglycerols (MCT) used in the preparation
of human foods are coconut and palmkemel oils, which contain 15% and 8% respectively of
8:0 + 10:0. The most abundant acid in these oils is lauric acid (12:0), which accounts for
nearly 50% of the total fatty acids. Seed oils of Cuphea species also contain medium-chain
fatty acids and could become another potential food source [6].
Medium-chain fatty acids are rapidly absorbed in conditions in which the absorption of
long-chain fatty acids may be impaired. Failure to digest lipids may occur because a disease
has resulted in insufficient lipase or bile salt; malabsorption is generally due to damage to the
mucosal surfaces of the small intestine. Absorption may also “back up” due to congenital
failure to produce specific apoproteins that are obligatory components of plasma lipoproteins.
Patients with malabsorption are at risk of deficiencies in energy, fat-soluble vitamins, and
essential fatty acids but can usually absorb medium-chain fatty acids [51].
For such patients, MCT, which can be purchased as cooking oils or fat spreads, provide
useful sources of dietary fats, although of course they lack the essential fatty acids. MCT
have also found applications as a source of energy in the enteral feeding of preterm babies.
An alternative application for MCT that has been much discussed is in reduced calorie
diets and diets that will not elevate plasma cholesterol, because medium-chain fatty acids are
not deposited in adipose tissue and do not contribute to plasma lipoproteins (Section IV.B).
However, to my knowledge, convincing evidence for their long-term effectiveness in humans
is yet to be published. Their effectiveness would need to be substantial in view of the cost of
the product and competition from other low calorie fat substitutes now being developed.
A recent development in fractionation technology has been the production of spreadable
butters by removal of higher melting triacylglycerols.
2. R educed C holesterol Foods

In reduced cholesterol foods, most or part of the cholesterol has been removed by treatment
with cholesterol-degrading enzymes or by differential extraction, for example by supercritical
carbon dioxide extraction. This technology has been applied to eggs, fish, beef, and milk fat
[52]. There is little scientific justification for these products, since there is only a weak rela
tionship between cholesterol in the diet and cholesterol in the blood. Their production is
motivated by perceived marketing opportunities. It has been argued that in the long term,
promotion of reduced cholesterol dairy products will do little for the credibility of the dairy
industry [52,53].
G. Reduced Fat Products, Structured Fats,

and Fat Substitutes
1. R educed F at Spreads
The aqueous phase in a margarine contributes about 15% of the weight of the product. It is
now possible with skillful use of modem emulsifier technology to increase this to about 75%
100 Gurr
or even more to give low fat spreads that may have applications in energy-reduced diets.
These are often blends of vegetable and dairy fats to give a relatively high polyunsaturated
fatty acid content yet retain a butterlike flavor.
In regard to “slimming,” the concept of the low fat spread is that for the same bulk of
spread the energy content can be reduced by half, and since fats have over twice as much
energy per gram as carbohydrates and proteins, this can be a big potential saving. However,
they will only be effective in human diets if there is no tendency to compensate for the low
energy density of the food by automatically adjusting total food intake over the medium or
even the long term. This has been the subject of considerable research interest, but the results
are not yet entirely clear on the matter. One much quoted publication suggested that if the
energy content of one meal is reduced, energy intake in later meals is increased subcon
sciously to maintain overall energy consumption [54].
2. U ndigested F at A nalogues
In a low fat spread, the energy content of the actual fat phase is not reduced and the products
have limitations when it comes to being used for cooking. Therefore, the search has been on
for substances with fat-like physical properties yet without any energy value.
Several types of compounds have been tested including polyglycerol fatty acid esters and
“retrofats,” which are esters of fatty alcohols and citric acid [55]. The most recent and most
promising types are the sucrose polyesters, and one company has exhaustively tested a product
known widely as olestra [56]. Like all the compounds mentioned, olestra is not digested by
enzymes in the small intestine nor is it absorbed intact. It therefore passes down the gut and
contributes no energy to the body.
Nutritionists have been concerned that in clinical trials olestra caused abdominal cramping
and loose stools in some individuals. More importantly, olestra may inhibit the absorption of
fat-soluble vitamins (A,D,E, and K) and carotenoids from foods eaten at the same time as the
olestra-containing products. For these and other reasons, it has taken a considerable time for
the safety of olestra in foods to be evaluated. In January 1996 the US Food and Drug Admin
istration approved the use of olestra in certain savory snack foods provided that the olestra-
containing products have vitamins A,D,F, and K added to them and are specifically labeled.
It is likely to be several years before the consumer acceptability of these products and their
efficacy in reducing overall fat intakes can be properly assessed.
3. D igested R educed Energy Fats

Another approach has been to devise digestible fats (thereby avoiding problems, real or imag
ined, of having undigested fat in the gut) that have intrinsically lower metabolizable energies
than “normal” fats, contradicting the golden rule that all fats provide 9 kcal (38 kJ) per gram.
It has long been recognized that the shorter the chain length of a fatty acid (and therefore the
lower its molecular mass), the lower its gross energy per gram. Palmitic acid (16:0) consists
of 75% carbon, whereas carbon constitutes 67% of capric acid (8:0) and 55% of butyric (4:0)
acid, and the heats of combustion decrease accordingly. The metabolizable energy (MF) per
gram also decreases with diminishing molecular mass. During beta oxidation, less ATP is
produced per gram of short-chain than of long-chain fatty acids. Furthermore, the contribution
of glycerol (with an MF value of 4.3 kcal/g) to the metabolizable energy of triacylglycerols
becomes relatively more significant as the proportion of short-chain fatty acids increases and
the molecular mass of the triacylglycerol decreases.
These theoretical considerations led to the logical development of mixed acid triacylglycer
ols containing short- and long-chain fatty acids. In 1991, Peters and colleagues [57] described
Caprenin, a mixed acid glyceride of one long-chain acid (behenic acid, 22:0) and two me
dium -chain acids [capric (10:0) and caprylic (8:0)], and in three papers [58-60] published
during 1994 another triacylglycerol having similar attributes and named SALATRIM was
described. The authors cleverly combined the advantages of the reduced energy value of the
short-chain fatty acids with the poorer absorption of stearic acid (18:0) to produce a triacyl
glycerol with significantly reduced energy value (5 kcal/g; 21 kJ/g) compared with conven
tional fats (9 kcal/g; 38 kJ/g).
An important point emerging from the study was that the actual absorption of 18:0 de
creased as the proportion of 18:0 to short-chain fatty acids increased and was also lower when
it was esterified in positions 1 or 3 than in position 2. In a human clinical trial in which 17
subjects received a diet containing 22% of energy from SALATRIM for 7 days, a metaboliz
able energy value of SALATRIM of 4.9 kcal/g and absorption of stearic acid ranging from
63 to 70% were reported [60].
SALATRIM is produced by interesterifying triacylglycerols containing short-chain fatty
acids (e.g., tributyrin) with long-chain saturated triacylglycerols (e.g., hydrogenated rapeseed
oil). It is therefore simple to produce by methods that are standard in the fats and oils pro
cessing industry. This material has the potential to reduce substantially the energy value of
foods into which it is incorporated while still retaining desirable textural properties. Whether
it turns out to be a useful aid to weight maintenance remains to be seen.
VIII. HOW MUCH OF A HEALTH PROBLEM IS FAT?

The intense consumer interest in fat is mainly due to much publicized associations between
high fat intakes and ill health. It would be impossible to adequately review here the evidence
for and against these associations. Instead, I briefly summarize what seems to be the current
consensus among nutritionists, indicate where there have been recent changes in views and
where current evidence is weak, and discuss priorities for future research. The main focus is
on obesity and coronary heart disease, but other health aspects are also discussed briefly.
A. Obesity
1. B ody F at D istribution
Arguably, obesity is one of the major health problems in industrialized countries and one that
seems resistant to all public health attempts to control it. Opinions differ widely, however, on
how seriously to treat the problem. Few studies in which the degree of overweight or obesity
has been tested as a risk factor for CHD have given clear-cut answers. Recent research may
indicate why this has been so. The index of overweight frequently used in epidemiological or
experimental studies as a measure of obesity, namely body mass index (BMI = kg/m^), is an
extremely rough guide because it does not indicate how the fat is distributed in the body. Fat
distribution is now recognized as being far more important than the mere fact of being over
weight. Broadly, people divide into two main types: those who have the bulk of their fat
around the waist (mainly, but not exclusively, men) and those whose fat is mainly on the
hips, thighs, and lower limbs (more characteristic of women). A simple measurement of fat
distribution now in common use is the waist-to-hip girth ratio.
It is now recognized that central obesity is a significant risk factor for CHD whereas lower
body or overall obesity have virtually no relationship. Central obesity and the common disor
ders glucose intolerance [and its main clinical manifestation insulin-independent (maturity
onset) diabetes mellitus], hypertriacyIglycerolemia, and hypertension form what is sometimes
dubbed the “deadly quartet” [61]. The common link between the four is thought to be hyperin-
sulinemia, i.e., habitually raised blood insulin concentrations. Central obesity is characterized
by extremely enlarged fat cells in the abdominal adipose tissue. The large fat cells are insulin-
102 Gurr
resistant and need more insulin to sustain their metabolism. Th is contributes to the raised
blood insulin. They also have a greater tendency to break down their triacylglycerols, liberat
ing free fatty acids into the plasma. High concentrations of fatty acids in the blood are thought
to reduce the insulin sensitivity of muscle as well as interfering with the removal of insulin
by the liver. The excess of insulin and fatty acids causes the liver to turn the fatty acid into
VLDL, which explains the hyperlipidemia. Finally, high insulin levels are thought to act upon
the kidneys and the vascular system in such a way as to produce high blood pressure, although
the latter mechanisms are more obscure than those affecting blood lipids.
2. Is a F a t Calorie M ore Fattening Than a Carbohydrate C aloriel

It might seem that there is an obvious link between fat intake and obesity because (1) fat has,
weight for weight, over twice as much energy as carbohydrates and (2) fat contributes to
palatability and might therefore be expected to encourage overeating.
Many have argued that what matters in regard to the potential for gaining excessive weight
is total energy intake rather than whether the calories are contributed by fat or carbohydrate.
In other words, they have assumed that the “fattening potential” of 1 cal of fat in excess of
requirements is equal to the fattening value of 1 cal of carbohydrate. In recent years, this
view has been challenged [62,63].
It is clear that the body weight of adults (even if it is above what might be regarded
as ideal) remains relatively stable over long periods of time despite considerable short-term
fluctuations. The relative proportions of the major body constituents—protein, carbohydrate,
and fat—remain relatively constant during these periods even though the food we eat every
day may vary widely in composition. It is apparent that there is a balance between the accre
tion of body constituents and the oxidation of protein, carbohydrate, and fat that acts to
correct any major perturbations from what the body regards as normal.
One thing is certain: The body cannot thwart the laws of thermodynamics. Body energy
stores can remain stable only when energy intake is equivalent to energy expenditure over an
extended period. There must be some mechanism, therefore, by which the body is able to
adjust either its intake of energy as food or its expenditure of energy by oxidizing the compo
nents of that food, or both.
Body components that may be regulated are the energy stores: glycogen (small: about 1.5-
2 kg) or fat in adipose tissue (large: potentially two orders of magnitude larger than glycogen
stores). Historically, one school of thought held that energy intake was regulated by signals
originating from body fat stores, working through detectors in the brain. This was called the
lipostatic theory. When Tremblay and colleagues [64] induced fat loss in obese subjects by
means of increased physical activity, they found that there was a point beyond which they
could lose no further fat. This coincided with an inability of their fat cells to increase the rate
at which they hydrolyzed triacylglycerols compared with the fat cells of lean subjects. These
observations are consistent with some kind of lipostatic control mechanism.
Alternatively, since variations in glucose metabolism are known to affect food intake,
another school of thought proposed the presence of brain receptors that were sensitive to
differences in glucose concentration between the venous and arterial blood and stimulated
according to the rate of glucose utilization. In support of this idea, there is recent evidence
that depriving cells of glucose increases food intake and also that carbohydrate intakes are
more stable than fat intakes in human beings [65].
These ideas, developed in the 1950s and 1960s, put emphasis on controls over food intake.
In the 1970s and 1980s, the pendulum swung more toward concepts that individuals who are
prone to obesity would be characterized by defects in energy expenditure that would favor a
positive energy balance, whereas individuals who remained lean could adapt their energy
expenditure according to fluctuations in food intake or to changes in body stores. Although
such adaptive changes in energy expenditure have been well characterized in animals, there
is little evidence for their importance in humans. Basic components of the current concept are
as follows:
Equivalence of total energy intake and expenditure is not sufficient in itself to maintain
energy balance. There must also be an equilibrium between the intake and oxidation of
each macronutrient: protein, carbohydrate, and fat.
The rate of oxidation of carbohydrate matches carbohydrate intake quickly enough to pre
vent excessive exhaustion or accumulation of carbohydrate reserves.
The same is not true of fat oxidation. It appears that there is little or no ability to increase
fat oxidation in response to excessive fat intakes in the short term, and since the fat
stores are virtually unlimited, fat storage increases more readily in response to excessive
consumption of fat than of carbohydrate. In the longer term, fat oxidation adapts so that
the oxidation of the fuel mix is appropriate to the new body composition, bringing about
a new equilibrium between intake and oxidation of fuels and resulting in a new, higher,
stable weight.
Physical activity may either increase or attenuate food intake in people consuming high fat
diets, depending on each individual’s inherited ability to increase fat oxidation in re
sponse to exercise.
High fat diets are associated with appreciable short-term increases in energy intake. When
given foods with a high ratio of fat to carbohydrate, people tend to consume sufficient
carbohydrate to fulfill the expected daily carbohydrate need. This leads to an increase
in energy from fat. There is some tendency for total food intake to be reduced (sug
gesting that some kind of compensation mechanism was at work), but, because of the
high energy density of the foods, the body was not able to fully compensate, and energy
intake increased. Some epidemiological studies seem to confirm that there is a positive
statistical association between percent body fat and proportion of fat in the diet [65].
In summary, the answer to the question “Is a fat calorie more fattening than a carbohydrate
calorie?” appears to be yes. The reasons are complex, but the main cause is the body’s
inability to increase fat oxidation rapidly in response to increases in fat consumption. Instead,
fat is put into stores that are, to all intents and purposes, limitless. By contrast, when carbohy
drate intake exceeds its limited storage capacity, the excess is readily oxidized. In the longer
term, fat oxidation adjusts to the higher body fat stores, bringing intake and oxidation into
equilibrium and resulting in a new stable weight. Whereas a decade or so ago the different
views about the regulation of body weight seemed to be in conflict, the various observations
described here can now be integrated into a model that goes some way to demonstrating how
the body manages to maintain constancy of composition for relatively long periods despite
large fluctuations in intake and in physical activity.
B. Cardiovascular Disease
It is recognized that many inherited and environmental factors interact to affect predisposition
to the development of cardiovascular disease (CVD). Of the environmental factors, smoking
and diet have been given most prominence, and of the various dietary components, most
emphasis has been placed on fat. The consensus among nutritionists, epidemiologists, and
medical practitioners is that SFA are the principal dietary components contributing to CVD
through their influence on blood cholesterol, particularly LDL cholesterol. These particles are
104 Gurr
held to be a prerequisite for the development of atherosclerotic plaque [1]. Although the
adherents to this hypothesis would be the first to accept the gross oversimplicity of this view,
the unbiased reader of the U.K. Department of Health’s most recent report, “Nutritional As
pects of Cardiovascular Disease,” could surely be left in little doubt that those responsible for
public health in the United Kingdom and elsewhere regard reduction in dietary SFA as the
main plank in the strategy to reduce CVD.
I [66] and others [67-69] have argued strongly that the scientific evidence on which this
strategy is based is weak in the extreme. A summary of the arguments is presented in Table 3.
A considerable weakness in the concept of SFA involvement in CVD as illustrated by the
U.K. Department of Health report [1] is the focus on total cholesterol and LDL in the fasting
state. In practice, most human beings in developed countries spend most of their time in the
postprandial state [70]. Zilversmit [71] proposed that the most atherogenic particles were not
LDL but the remnants of chylomicrons and VLDL after part of their triacylglycerol had been
removed. This idea has now been extended and refined, and current views of the role of
lipoproteins in CVD are summarized in the concept of the atherogenic lipoprotein phenotype
(ALP) [70,72].
Briefly, the ALP describes a collection of abnormalities in lipoprotein metabolism that
predispose apparently healthy people to increased CVD risk. Its characteristics are a moder
ately raised concentration of triacylglycerol, low concentration of HDL, and a predominance
of unusually small, dense LDL and HDL particles in the postabsorptive state. The ALP is
also associated with the condition of insulin resistance, in which tissues need more of the
hormone than normal to accomplish a particular metabolic task. This causes the pancreas to
secrete more insulin and results in habitually elevated blood insulin concentrations. There is
a strong genetic basis for the ALP, which may mainly affect the enzyme lipoprotein lipase,
which removes triacylglycerols from circulating chylomicrons immediately after a meal. How
ever, it is quite clear that development of the condition involves diet, smoking, and physical
activity interacting with the predisposing genes. It is quite possible that certain fatty acids
directly influence the expression of the genes involved in the ALP, and this is currently the
subject of vigorous research.
Table 3 Deficiencies in the Lipid Hypothesis
1. There is insufficient correspondence in vascular pathology (a) between animals and humans and (b)
between general atherosclerosis and lesions seen in familial hypercholesterolemia.
2. International epidemiology is flawed by confounding factors and selection biases.
3. Within countries, epidemiology gives little support for a relationship between dietary fat and CVD.
Fat intake does not explain differences in total CVD or differences in incidence of CVD according
to region, sex, or social class.
4. Trends in CVD mortality are not coincident with changes in the amount and type of fat eaten.
5. The hypothesis cannot explain the greater risk in men compared with women, enhanced risk in
postmenopausal women, or consistent recent decreases in mortality in women compared with more
erratic changes in men in many countries. There have been no corresponding decreases in CVD inci
dence.
6 . Plasma total cholesterol is a weak predictor of CVD compared with hemostatic factors. Less than
50% of CVD risk is accounted for by known “risk factors.”
7. Extrapolations from drug trials in high risk individuals to the general population are unwarranted.
8 . Intervention trials have not demonstrated a major impact of dietary fat modification on CVD and
none on total mortality. Only the most stringent diets achieve useful plasma cholesterol reductions.
One reason this condition has been overlooked in the obsession with total and LDL choles
terol is the technical difficulties in routinely measuring lipoprotein subclasses, and another is
that values for blood TAG, total cholesterol, and LDL that are characteristic of an ALP
usually fall below the levels that clinicians regard as appropriate for taking therapeutic action
after lipid screening. A likely outcome of the appreciation of the importance of this syndrome
will be a swing away from the focus on dietary SFA and n-6 PUFA toward the n-3 PUFA.
The latter are more effective in controlling levels of triacylglycerol-rich lipoproteins and in
reversing the effects of insulin insensitivity, which lie at the heart of the ALP.
3. F a t and Thrombosis
The culmination of the series of events that lead to death from CVD is an arterial occlusion
or thrombosis. By contrast with atherosclerosis, little research has been conducted on dietary
influences on thrombosis. This has been mainly due to lack of adequate techniques for as
sessing the process of thrombus formation in the living human being. It is extremely complex
and involves three main types of events: the aggregation of cellular elements in the blood (the
platelets), the interactions of a sequence of blood clotting factors known collectively as the
coagulation cascade, and a mechanism for the dissolution of clots once formed.
Experimental studies linking dietary lipids with thrombosis are still in their infancy and
have been concerned mainly with (1) effects on platelet aggregation and (2) effects on differ
ent components of the coagulation cascade. Platelet studies rely on tests of function in vitro
whose relationship with events in vivo is uncertain. These have been inconsistent, although
some studies have shown increased potential for aggregation in diets with a high intake of
SFA [73]. In studies on the components of the coagulation cascade, reduction of fat intake as
a proportion of dietary energy reduced the plasma concentration of factor VII^, but the compo
sition of the fatty acids had no influence [73,74]. Most studies have found the concentration
of plasma fibrinogen (a powerful predictor of CVD) to be uninfluenced by dietary fat. The
few studies so far reported on the effects of dietary lipids on the fibrinolytic system (which
dissolves clots) are conflicting [74].
Future research seems likely to be concentrated on investigating the role of total fat, SFA,
and the n-3 PUFA on thrombus formation. Present indications are that total fat and SFA may
be prothrombotic and the n-3 fatty acids antithrombotic. The balance of fatty acids in the diet
will be of primary practical concern. Another important area of study will be to more clearly
distinguish the dietary influences on the events leading to a stroke from those leading to a
heart attack. Although they are both cardiovascular diseases and have ostensibly similar natu
ral histories, heart attacks and stroke clearly respond quite differently to different risk factors,
including diet [1].
<*
C. Cancer
Cancer is a disease of cell growth regulation in which nearly all control has been lost over
cell growth and the cells grow autonomously. It is generally thought to develop in three
stages: initiation, in which some agent such as a chemical carcinogen causes a permanent
alteration of the DNA of a cell; promotion, in which a “promoter” increases the chance of
misinformation inserted into the genetic code being expressed; and propagation, which in
volves any factor that stimulates uncontrolled growth.
Doll and Peto [75] estimated that about one-third of cancer deaths might be due to diet,
although their best guess ranged from 10 to 70%. Diet may influence cancer in three ways:
(1) through provision of carcinogens in foods, (2) by influencing the conversion of harmless
components in the gut into carcinogens by the gut microflora, or (3) as a result of the effect
106 Gurr
of nutrient balance (fat, proteins, carbohydrates, minerals, vitamins) on the initiation, promo
tion, or propagation of tumors. Information gained so far implies that energy, fat, and some
proteins may contribute to cancer, while fiber, vitamins A, C, and E, the carotenoids, cal
cium, and some proteins may be protective.
Animal experients indicate that high fat diets result in increased tumor incidence but that
this may be largely an effect of overall energy intake. Indeed, the most effective way to
prevent or reduce cancer is to restrict energy intake. There must be a minimal dietary level of
linoleic acid {n-6) for promotion to occur, but only in the context of a high fat diet. PUFA of
the n-3 family seems to have a protective effect, but the evidence is so far insubstantial.
Evidence for associations between fat intake and human cancer comes from international
epidemiological comparisons. The strongest relationships are with breast and colon cancer.
Theoretically, the effects of fat on breast cancer could be mediated via an effect on hormone
balance or through effects on the immune system via the control of eicosanoid synthesis,
while effects on colon cancer might be mediated via effects on the secretion of bile salts that
can be metabolized to carcinogens in the colon by the gut microflora. International epidemio
logical evidence is flawed by unreliable data on food intakes and by confounding factors such
as smoking, reproductive practices, body weight, physical activity, and especially energy
intakes, which have rarely been properly controlled. An anomaly is that fat intakes also corre
late with intakes of fat-soluble vitamins, the latter being generally regarded as protective.
Few within-country epidemiological studies, prospective studies, or case control studies have
provided any strong evidence for or against a role for fat in most cancers. The report by the
U.K. Department of Health [24] on dietary reference values was unable to make any specific
recommendations about fat in relation to cancer.
The results of several studies undertaken principally to investigate risk factors for coronary
heart disease have implied that whereas high plasma cholesterol concentrations may be re
garded as increasing the overall risk of mortality from cardiovascular disease, very low plasma
cholesterol concentrations may be associated with increased risk of cancer [76,77]. Firm con
clusions about the true importance of low plasma cholesterol as a risk factor for cancer cannot
yet be drawn, since several studies have been unable to demonstrate this association and no
plausible biological mechanism can yet be advanced to explain it.
Despite the overwhelming emphasis on the adverse effects of dietary fat on cancer develop
ment, there has been growing interest in a naturally occurring fat-soluble cancer inhibitor:
conjugated linoleic acid (CLA) [78]. CLA is mixture of isomers of linoleic acid (cis-9, cis-
12-octadecadienoic acid) the main component being cis-9, 11-octadecadienoic acid.
These are formed during the process of biohydrogenation of PUFA in the rumen of cows and
subsequently find their way into milk and beef tissues. Evidence for involvement in cancer
development has so far been confined to small laboratory animals and tests in vitro. Much
further research is needed on the effects of CLA on human cancers before any firm conclu
sions can be drawn.
D. Inflammatory Disease
There is accumulating evidence that nutrition and particularly fat in the diet may in some way
influence the functioning of the immune system. Certainly protein energy malnutrition mark
edly depresses the body’s ability to mount an immune response, and deficiencies of specific
nutrients, including zinc, vitamin A, and EFA, are also associated with immune malfunction.
Likewise, ovemutrition also seems to have an adverse effect on immune function. In particu
lar, excessive intakes of n-6 PUFA seem to be immunosuppressive, and this phenomenon has
been exploited in dietary support during clinical treatments designed to suppress the rejection
of organ transplants and in the management of multiple sclerosis, a disease in which autoim
munity is thought to play a part [79]. The mechanism probably involves the regulation by
dietary EFA of the amounts and types of eicosanoids formed, since these compounds are
known to be involved in the regulation of cell-mediated immune function [80].
Rheumatoid arthritis (RA) is an inflammatory disease in which an inappropriate immune
response plays a part. The idea that the severity of RA is related to diet, in particular SFA,
has received widespread publicity, but the scientific basis for these beliefs is not substantial.
A recent review [81] reflects the paucity of good scientific research and demonstrates over
whelming emphasis on the effects of the n-3 PUFA rather than on SFA.
E. Depression
The finding that some intervention trials that involved lowering blood cholesterol appeared to
have resulted in more deaths from suicides and other so-called violent causes led to much
speculation about whether changes in the cholesterol content of neural membranes might lead
to functional changes resulting in altered states of mind [82]. In a recent review, Hibbeln and
Salem [83] argue that focusing on cholesterol might be misleading. They suggest that in the
cholesterol lowering studies other dietary changes, particularly increases in the n-6ln-3 PUFA
ratio, might be more important. In a wide-ranging review they present a thought-provoking
case for the aggravation of depressive conditions by a relative deficiency in dietary, and hence
membrane, n-3 PUFA.
F. Postscript
This chapter has reviewed the many and varied roles of lipids in the body and in the diet. In
assessing the roles of dietary lipids in health and disease, it is important to put fat in perspec
tive with other dietary components and diet, in turn, in perspective with other environmental
factors. It has been asserted frequently and sometimes stridently that western diets have
changed drastically during the last century compared with diets to which, it is assumed, the
body became accustomed during the evolution of the human race, and that this is a major
cause of the many ills affecting developed nations [84,85]. Such arguments have to be treated
with caution for a number of reasons.
First, the human body is very adaptable. Homeostatic mechanisms ensure that its cells,
tissues, and organs adapt to changes in nutrient intakes by subtle changes in the amounts and
activities of enzymes in metabolic pathways, by changes in membrane composition, and by
changes in the numbers and activities of membrane receptors and other regulatory proteins.
These changes take place over months, weeks, hours, or even minutes, certainly not centuries.
Second, the notion that a single “primitive” diet is appropriate for all human beings fits
uneasily with the knowledge that diets, apparently consistent with good health and well-being,
vary enormously throughout the world’s communities, including severalfold differences be
tween fat intakes expressed as a proportion of total energy as well as very large differences
in the types of fats eaten.
Third, life expectancy and general health have steadily improved in the industrialized coun
tries, a fact often forgotten or dismissed in the enthusiasm for branding dietary fat as a ma
jor killer.
Although a reduction in fat intake may be a distinct advantage for many people, especially
those who find difficulty in maintaining energy balance, there is a danger that in this obsession
with dietary fat other perhaps more important lifestyle factors are forgotten, including smok
ing, too little exercise, and raised blood pressure. The role of fat must be seen in perspective,
and it should not be concluded that widespread changes in dietary fat consumption will lead
108 Gurr
inevitably to freedom from the so-called diseases of affluence. The evidence does not point
that way. The key to a good diet is variety.
ABBREVIATIONS
AA Arachidonic acid
ALP Atherogenic lipoprotein phenotype
Apo- Apoprotein
ATP Adenosine triphosphate
BMI Body mass index
CHD Coronary heart disease
CLA Conjugated linoleic acid
CVD Cardiovascular disease
DHA Docosahexaenoic acid
DLMG Dilinoleoyl-monogammalinolenoyl glycerol
DNA Deoxyribonucleic acid
EFA Essential fatty acid(s)
EPA Eicosapentaenoic acid
GLA Gamma-linolenic acid
HDL High density lipoprotein(s)
LA Linoleic acid
LCAT Lecithin-cholesterol acyltransferase
LDL Low density lipoprotein
LPL Lipoprotein lipase
LTA Leukotriene A
LTB Leukotriene B
MCT Medium-chain triacylglycerol(s)
ME Metabolizable energy
MUFA Monounsaturated fatty acid(s)
PGE Prostaglandin E
PGH Prostaglandin H
PUFA Polyunsaturated fatty acid(s)
RA Rheumatoid arthritis
RNI Reference nutrient intake
SFA Saturated fatty acid(s)
SMCFA Short- and medium-chain fatty acid(s)
VLDL Very low density lipoprotein(s)
REFERENCES
1. Department of Health, Nutritional aspects of cardiovascular disease, Report on Health and Social
Subjects, No. 46, Her Majesty’s Stationery Office, London, 1994.
2. Department of Health, The Health of the Nation: A Strategy for Health in England, HMSO,
London, 1992.
3. M. I. Gurr and J. L. Harwood, Lipid Biochemistry: An Introduction, 4th ed.. Chapman and Hall,
London, 1991.
4. Ministry of Agriculture, Fisheries and Food, Household Food Consumption and Expenditure,
Annu. Rep. Nat. Food Survey Committee, HMSO, London, 1993.
5. M. I. Gurr, Role of Fat in Food and Nutrition, 2nd ed. Elsevier, London, 1992.
F. D. Gunstone, J. L. Harwood, and F. B. Padley, The Lipid Handbook, 2nd ed., Chapman and
Hall, London, 1994.
7. B. Holland, A. A. Welch, I. D. Unwin, D. H. Buss, A. A. Paul, and D. A. T. Southgate,
McCance and Widdowson’s The Composition of Foods, 5th ed., Roy. Soc. Chem., Cambridge,
UK, 1991.
8. R. G. Jensen, A. M. Ferris, C. J. Lammi-Keefe, and R. A. Henderson, Lipids of bovine and
human milk: a comparison, J. Dairy Sci. 73: 223-240 (1990).
9. A. C. Goedhart and J. C. Bindels, The composition of human milk as a model for the design of
infant formulas: recent findings and possible application, Nutr. Res. Rev. 7: 1-23 (1994).
10. British Nutrition Foundation, Report of the Task Force on Unsaturated Fatty Acids—Nutritional
and Physiological Significance, British Nutrition Foundation and Chapman and Hall, London,
1992.
11. I. A. Prior, F. Davidson, C. E. Salmond, and Z. Czochanska, Am. J. Clin. Nutr. 34: 1552-
1561 (1981).
12. G. O. Burr and M. M. Burr, A new deficiency disease produced by the rigid exclusion of fat from
the diet, J. Biol. Chem. 82: 345-367 (1929).
13. L. Soderhjelm, H. F. Wiese, and R. T. Holman, The role of polyunsaturated fatty acids in human
nutrition and metabolism. Prog. Chem. Fats Other Lipids 9: 555-585 (1970).
14. R. T. Holman and S. B. Johnson, Essential fatty acid deficiencies in man, in Dietary Fats and
Health (E. G. Perkins and W. J. Visek, Eds.), Am. Oil Chem. Soc., Champaign, IL, 1983.
15. M. Press, H. Kikuchi, T. Shimoyama, and G. R. Thompson, Diagnosis and treatment of essential
fatty acid deficiency in man, Br. Med. J. 2: 247-250 (1974).
16. R. T. Holman, Biological activities of and requirements for polyunsaturated fatty acids. Prog.
Chem. Fats Other Lipids 9: 607-682 (1970).
17. J. C. Allen and R. J. Hamilton (Eds.), Rancidity in Foods, 2nd ed., Appl. Sci London, 1989.
18. S. M. Grundy and M. A. Denke, Dietary influences on serum lipids and lipoproteins, J. Lipid
Res. 31: 1149-1172 (1990).
19. A. Bonanome and S. M. Grundy, Effect of dietary stearic acid on plasma cholesterol and lipopro
tein levels, N. Engl. J. Med. 318: 1244-1248 (1988).
20 . K. C. Hayes and P. Khosla, Dietary fatty acid thresholds and cholesterolemia, FASEB J. 6: 2600-
2607 (1992).
21. M. I. Gurr, N. Borlak, and S. Ganatra, Dietary fat and plasma lipids; Nutr. Res. Rev. 2: 6 3 -
8 6 (1989).
22 . R. O. Vies, Nutritional aspects of rapeseed oil, in The Role of Eats in Human Nutrition (A. J.
Vergroesen, Ed.), Academic, London, 1975.
23. K. R. Norum, E. N. Christiansen, B. O. Cristophersen, and J. Bremer, Metabolic and nutritional
aspects of long-chain fatty acids of marine origin, in The Role of Fats in Human Nutrition, 2nd
ed. (A. J. Vergroesen and M. A. Crawford, Eds.), Academic, London, 1989.
24. Department of Health, Dietary Reference Values for Food Energy and Nutrients for the United
Kingdom, Report on Health and Social Subjects, No. 41, HMSO, London, 1991.
25. A. G. Hassam, The role of evening primrose oil in nutrition and disease, in The Role of Fats in
Human Nutrition (F. B. Padley and J. Podmore, Eds.), Ellis Horwood, Chichester, 1985, pp.
84-100.
26. D. F. Horrobin, Natural does not equal safe, Pharm. Technol. Eur., December 1994, pp. 14-15.
27. R. S. Lees and M. Karel (Eds.), Omega-3 Fatty Acids in Health and Disease, Marcel Dekker,
New York, 1990.
28. M. L. Burr, A. M. Fehily, J. F. Gilbert, S. Rogers, R. M. Holliday, P. M. Sweetnam, P. C.
Elwood, and N. M. Deadman, Effects of changes in fat, fish and fibre intakes on death and
myocardial infarction: Diet and Reinfarction Trial (DART), Lancet ii: 757-761 (1989).
29. L. G. Cleland, J. K. French, W. H. Betts, G. A. Murphy, and M. J. Elliott, Clinical and biochem
ical effects of dietary fish oil supplements in rheumatoid arthritis, J. Rheumatol. 15: 1471-1475
(1988).
30. I. Tamamoto, M. Sugano, and M. Wada, Hypocholesterolaemic effect of animal and plant fats in
rats. Atherosclerosis 13: 171-184 (1971).
110 Gurr
31. R . B. McGandy, D. M. Hegsted, and M. L. Myers, Use of semisynthetic fats in determining

effects of specific dietary fatty acids on serum lipids in man, Am. J. Clin. Nutr. 23: 1288-1298
(1970).
32. A. Christophe, F. Matthys, R. Geers, and G. Verdonk, Nutritional studies with randomized butter;
cholesterolaemic effects of butter oil and randomized butter oil in man. Arch. Int. Physiol. Bio-
chem. 86: 413-415 (1978).
33. D. J. McNamara, Relationship between blood and dietary cholesterol, in Meat and Health: Ad
vances in Meat Research, Vol. 6 (A. M. Pearson and T. R. Dutson, Eds.), Elsevier, London,
1990, pp. 63-87.
34. E. Renner, Milk and Milk Products in Human Nutrition, Voikswirtschaftlicher Verlag, Munich,
1983.
35. G. Billek, Heated fats in the diet, in The Role of Fats in Human Nutrition (F. B. Padley and J.
Podmore, Eds.), Ellis Horwood, Chichester, 1985, pp. 163-172.
36. M. I. Gurr, Lipids: products of industrial hydrogenation, oxidation and heating, in Nutritional and
Toxicological Aspects of Food Processing, (R. Walker and E. Quattrucci, Eds.), Taylor and Fran
cis, London, 1988, pp. 149-155.
37. C. B. Taylor, S. K. Peng, N. T. Werthesen, P. Tham, and K. T. Lee, Spontaneously occurring
angiotoxic derivatives of cholesterol. Am. J. Clin. Nutr. 32: 40-57 (1979).
38. British Nutrition Foundation, Report of the Task Force on Trans Fatty Acids, Br. Nutr. Founda
tion, London, 1995.
39. P. W. Parodi, Composition and structure of some consumer-available edible fats, J. Am. Oil
Chem. Soc. 53: 530-534 (1976).
40. P. W. Parodi, Distribution of isomeric octadecenoic fatty acids in milk fat, J. Dairy Sci. 59:
1870-1873 (1976).
41. B. Koletsko, Trans fatty acids may impair biosynthesis of polyunsaturates and growth in man,
Acta Paediatr. 8F. 302-306 (1992).
42. R. P. Mensink and M. B. Katan, Effects of dietary trans fatty acids on high density and low
density lipoprotein levels in healthy subjects, N. Engl. J. Med. 323: 439-444 (1990).
43. J. T. Judd, B. A. Clevidence, R. A. Muesing, J. Wittes, M. E. Sunkin, and J. J. Podczasy,
Dietary trans fatty acids; effects on plasma lipids and lipoproteins of healthy men and women.
Am. J. Clin. Nutr. 59: 861-868 (1994).
44. R. P. Mensink, P. L. Zock, M. B. Katan, and G. H. Homstra, Effects of dietary cis and trans
fatty acids on serum lipoprotein(a) levels in humans, J. Lipid Res. 33: 1493-1501 (1992).
45. P. J. Nestel, M. Noakes, B. Belling, R. McArthur, P. Clifton, E. Janus, and M. Abbey, Plasma
lipoprotein lipid and Lp (a) changes with substitution of elaidic acid for oleic acid in the diet, J.
Lipid Res. 33: 1029-1036 (1992).
46. W. C. Willett, M. J. Stampfer, J. E. Manson, G. A. Colditz, F. E. Speizer, B. A. Rosner, L. A.
Sampson, and C. H. Hennekens, Intake of trans fatty acids and risk of coronary heart disease
among women. Lancet 341: 581-585 (1993).
47. A. Aro, A. F. M. Kardinaal, I. Salminen, J. D. Kark, R. A. Riemersma, M. Delgado-Rodriguez,
J. Gomez-Aracena, J. K. Huttunen, L. Kohlmeier, B. C. Martin, J. M. Martin-Moreno, V. P.
Mazaev, J. Ringstad, M. Thamm, P. van’t Veer, and F. J. Kok, Adipose tissue isomeric trans
fatty acids and risk of myocardial infarction in nine countries; the EURAMIC study. Lancet 345:
273-278 (1995).
48. T. L. Roberts, D. A. Wood, R. A. Riemersma, P. J. Gallagher, and F. C. Lampe, Trans isomers
of oleic and linoleic acids in adipose tissue and sudden cardiac death. Lancet 345: 278-282 (1995).
49. P. M. Kris-Etherton, J. Derr, D. C. Mitchell, V. A. Mustad, M. E. Russell, E. T. McDonell, D.
Salabsky, and T. A. Pearson, The role of fatty acid saturation on plasma lipids, lipoproteins and
apoproteins. I. Effects of whole food diets high in cocoa butter, olive oil, soya bean oil, dairy
butter and milk chocolate on the plasma lipids of young men, Metabolism 42: 121-129 (1993).
50. D. Kritchevsky, S. A. Tepper, D. Vesselinovitch, and R. W. Wissler, Cholesterol vehicle in
experimental atherosclerosis. 13. Randomized peanut oil. Atherosclerosis 17: 225-243 (1973).
51. K. Sickinger, Clinical aspects and therapy of fat malassimilation with particular reference to the
use of medium-chain triglycerides, in The Role of Fat in Human Nutrition (A. J. Vergroesen,
Ed.), Academic, London, 1975, pp. 115-209.
52. R. Sieber, Cholesterol removal from animal food—can it be justified? Lebensm. Wiss. Technol.
26: 375-387 (1993).
53. E. Renner and M. I. Gurr, Do we need cholesterol reduced dairy products? Dairy Ind. Int. 56:
34-35 (1991).
54. R. W. Foltin, M. W. Fischman, T. H. Moran, B. J. Rolls, and T. H. Kelly, Caloric compensation
for lunches varying in fat and carbohydrate content by humans in a residential laboratory. Am. J.
Clin. Nutr. 52: 969-980 (1990).
55. V. K. Babayan, Modification of food to control fat intake, J. Am. Oil Chem. Soc. 51: 260-
264 (1974).
56. K. W. Miller and G. S. Allgood, Nutritional assessment of Olestra, a non-calorific fat substitute,
Int. J. Food Sci. Nutr. 44(Suppl. 1): S77-S82 (1993).
57. J. C. Peters, B. N. Holcombe, L. K. Hiller, and D. R. Webb, Caprenin. 3. Absorption and caloric
value in adult humans, J. Am. Coll. Toxicol. 10: 357-367 (1991).
58. J. R. Hayes, D. H. Pence, S. Scheinbach, R. P. D ’Amelia, L. P. Klemann, N. H. Wilson, and
J. W. Finley, Review of triacylglycerol digestion, absorption and metabolism with respect to
SALATRIM triacylglycerols, J. Agrie. Food Chem. 42: 474-483 (1994).
59. L. P. Klemann, J. W. Finley, and G. A. Leveille, Estimation of the absorption coefficient of
stearic acid in SALATRIM fats, J. Agrie. Food Chem. 42: 484-488 (1994).
60. J. W. Finley, L. P. Klemann, G. A. Leveille, M. S. Otterbum, and C. G. Walchak, Caloric
availability of SALATRIM in rats and humans, J. Agrie. Food Chem. 42: 495-499 (1994).
61. N. M. Kaplan, The deadly quartet. Upper body obesity, glucose tolerance, hypertriglyceridaemia
and hypertension. Arch. Int. Med. 149: 1514-1520 (1989).
62. A. Tremblay, Human obesity: a defect in lipid oxidation or in thermogenesis? Int. J. Obesity 16:
953-957 (1992).
63. J. P. Flatt, Importance of nutrient balance in body weight regulation. Diabetes!Metab. Rev. 4:
571-581 (1988).
64. A. Tremblay, J.-P. Després, and C. Bouchard, Adipose tissue characteristics of ex-obese long
distance runners, Int. J. Obesity 8: 641-648 (1984).
65. A. Tremblay, G. Plourde, J.-P. Després, and C. Bouchard, Impact of dietary fat content and fat
oxidation on energy intake in humans. Am. J. Clin. Nutr. 49: 799-805 (1989).
66. M. I. Gurr, Dietary lipids and coronary heart disease: old evidence, new perspective. Prog. Lipid
Res. 31: 195-243 (1992).
67. R. H. Rosenman, The questionable roles of the diet and serum cholesterol in the incidence of
ischemic heart disease and its 20th century changes. Homeostasis 34: 1-44 (1993).
68 . U. Ravnskov, Cholesterol lowering trials in coronary heart disease: frequency of citation and
outcome. Br. Med. J. 305: 15-19 (1992).
69. W. E. Stehbens, The quality of epidemiological data in coronary heart disease and atherosclerosis,
J. Clin. Epidemiol. 46: 1337-1346 (1993).
70. S. Sethi, M. J. Gibney, and C. M. Williams, Postprandial lipoprotein metabolism, Nutr. Res.
Rev. 6: 161-183 (1993).
71. D. B. Zilversmit, Atherogenesis: a postprandial phenomenon. Circulation 60: 473-485 (1979).
72. B. A. Griffin and A. Zámpelas, Influence of dietary fatty acids on the atherogenic lipoprotein
phenotype, Nutr. Res. Rev. 8: 1-26 (1995).
73. S. Renaud, F. Godsey, E. Dumont, C. Thevenon, E. Ortchanian, and J. L. Martin, Influence of
long term diet modification on platelet function and composition in Moselle farmers. Am. J. Clin.
Nutr. 43: 136-150 (1986).
74. P. Marckmann, B. Sandstrom, and J. Jespersen, Effects of total fat content and fatty acid compo
sition in diet on factor VII coagulant activity and blood lipids. Atherosclerosis 80: 227-233
(1990).
75. R. Doll and R. Peto, The causes of cancer, J. Natl. Cancer Inst. 66: 1191-1308 (1981).
76. C. G. Isles, D. J. Hole, C. R. Gillis, V. M. Hawthorne, and A. F. Lever, Plasma cholesterol.
112 Gurr
coronary heart disease and cancer in the Renfrew and Paisley survey, Br. Med, J. 298: 920-
924 (1989).
77. S. B. Hulley, J. M. B. Walsh, and T. B. Newman, Health policy on blood cholesterol: time to
change direction. Circulation 86: 1026-1029 (1992).
78. P. W. Parodi, Conjugated linoleic acid: an anticarcinogenic fatty acid present in milk fat, Aus. J.
Dairy Technol. 49: 93-97 (1994).
79. M. I. Gurr, The role of lipids in the regulation of the immune system. Prog. Lipid Res. 22: 257-
287 (1983).
80. P. V. Johnson, Dietary lipids and immunity, Adv. Lipid Res. 21: 103-141 (1985).
81. D. M. Kremer and D. R. Robinson, Studies of dietary supplementation with omega-3 fatty acids
in patients with rheumatoid arthritis, in Health Effects of Omega-3 Polyunsaturated Fatty Acids in
Seafoods (World Rev. Nutr. D iet., Vol. 6 6 ) (A. P. Simopoulos, R. R. Kifer, R. E. Martin, and
S. M. Barlow, Eds.), Karger, Basel, 1991, pp. 367-382.
82. H. Engelberg, Low serum cholesterol and suicide. Lancet 339: 727-729 (1992).
83. J. R. Hibbeln and N. Salem, Dietary polyunsaturated fatty acids and depression: when cholesterol
does not satisfy. Am. J. Clin. Nutr. 62: 1-9 (1995).
84. World Health Organization Study Group, Diet, Nutrition and the Prevention of Chronic Diseases,
Tech. Rep. Ser. 797, WHO, Geneva, 1990.
85. S. B. Eaton and D. A. Nelson, Calcium in evolutionary perspective. Am. J. Clin. Nutr. 54: 281S-
287S (1991).
86 . M. I. Gurr, Fats in Human Nutrition and Dietetics (J. S. Garrow and W. P. T. James, Eds.),
Churchill Livingstone, Edinburgh, 1993, Chap. 6 .
87. M. I. Gurr, Polyunsaturated fatty acids of the n-3 family: influence on inflammatory disease. Lipid
Technol. 5: 65-68 (1993).
Extraction of Lipids from Natural Sources
Maurice A. Williams
Anderson International Corp., Cleveland, Ohio
I. INTRODUCTION ,
Lipids are removed from lipid-containing materials by two methods: mechanical crushing and
solvent extraction. Small volume processors prefer mechanical crushing to avoid having to
build a costly solvent extraction system. High volume processors prefer solvent extraction to
avoid having multiple mechanical screw presses operating in parallel. Usually volumes less
than 100 Mg/day are mechanically crushed, and volumes above 300 Mg/day are solvent ex
tracted. Volumes range from a few megagrams per day for a small crushing mill to 3000 or
more for a large solvent extraction plant.
In mechanical crushing, the oilseed is cleaned, cracked, flaked, sometimes cooked, and
conditioned to optimum moisture and temperature and then enters a mechanical screw press,
which generates enough pressure to cause oil to flow out of the oilseed. Crushing can reduce
the oil content of most oilseeds to less than 5%. In solvent extraction, the prepared oilseed
passes through a percolation vessel on a moving screen (or perforated plate or some similar
device) that transports the material through the vessel. Hot solvent, sprayed on top, percolates
through the material, extracting the oil. The solvent is then removed from the oil and the
solid residue. Solvent extraction can reduce oil levels below 1%.
IL PREPARATION
A. C leaning
Oilseeds come into the mills along with stems, leaves, stones, sand or dirt, and sometimes
weed seeds. Stems, dirt, and stones are abrasive and contain no oil. Weed seeds contain
different kinds of oil. The first preparation step is to clean the incoming oilseeds. Stems and
stones are usually larger than the oilseeds and can be screened out through revolving reels.
Weed seeds are usually smaller and can be removed along with sand and dirt through vibrating
screens. Magnets remove tramp iron.
113
114 Williams
Table 1 Preparation of Some Common Oilseeds
Cook at Dry to
Particle
Oilseed Size % H2O °C % H2O °C
Copra 3 mm — — 2 -2 .5 127
Com germ Whole germ — — 2 115
Cottonseed 0.25 mm flakes 12-14 88 2 .6-2.8 115-129
Linseed 0.25 mm flakes 11 88 2.5 115
Palm kernels 0.38 mm — — 1.5 -2 .0 115
Peanut 3 mm 9 88 2 .5 -3 .0 121
Rapeseed 3 mm flakes 12 88 3 115
Safflower 3 mm 12 88 3 71
Sesame Whole seed — — 2 .5 -3 .0 127
Soybean 1.5 mm — — 2-2.25 132
Sunflower 0.25 mm flakes 8-9 88 3 110
Tung 3 mm — — 3 121
B. Dehulling
Hulls are abrasive. They usually contain less than 1% oil, and they add to the bulk entering
the screw press. Removing hulls reduces wear on the screw press, increases oil yield (because
hulls absorb oil within the press), and increases capacity (because the screw press presses
only meats rather than whole oilseed). There are several different dehulling procedures. Bar
hullers or disk hullers are used for medium-sized seeds such as sunflower seed or cottonseed.
Hot aspiration systems are used for soybeans. Cracking machines are used for larger oilseeds.
Seeds enveloped in thick, hard shells are dehulled by various devices that crack the hulls,
then separate the hulls by screening or aspiration or by flotation and sometimes even by hand
[1]. The objective of all procedures is to remove as much hull as possible with minimal loss
of meats. A special case is cottonseed, which contains lint, the short ends of cotton fibers,
still attached to the hull.
Ce Flaking
Oilseed meats are usually flattened into thin flakes. This helps to more readily expose the oil
to solvent than would prevail with whole or cracked meats. It also reduces power consumption
of a screw press, which has to pulverize the seeds within the press to release the oil. Flaking
also ensures a more uniform cook (if the oilseed requires cooking) because all oilseed particles
are uniformly thick.
Two types of flakers are used. One type employs two heavy steel rolls in parallel, almost
touching each other and revolving in opposite directions so that the two rolls pull the meats
down between them. Mechanical jackscrews or hydraulic cylinders apply pressure against the
bearing blocks supporting the rolls so that the rolls do not spread apart when the meats pass
between them. Sometimes one roll revolves faster than the other. This helps the smearing
action of the rolls.
The second type stacks flaking rolls atop each other, usually five high. This has the advan
tage that the weight of the upper rolls provides the pressure. The meats are progressively
flattened into thinner and thinner flakes as they pass from roll to roll in their descent down
the stack.
Extraction of Lipids 115
Five-high rolls are well accepted for flaking oilseeds in mechanical crushing plants. They
are not widely used to prepare oilseeds for solvent extraction where large unbroken flakes are
desired. A five-high roll, for example, flakes the seed four times and produces more broken
flakes and fines than a single-pass flaker would.
D. Cooking and Drying

Cooking is excellent preparation for most oilseeds and is necessary for those too soft to
withstand the pressures generated within a screw press. Cooking hardens those oilseeds, simi
lar to the way cooking hardens eggs. Cooking also ruptures oil cells, thereby releasing the oil
into a more fluid state. The temperature of cooking reduces oil viscosity, allowing it to flow
more readily through the slotted walls of the screw press. Cooking inactivates troublesome
enzymes that adversely affect oil and meal quality. Cooking also sterilizes the oilseed by
destroying bacteria, molds, fungi, etc. that infest the incoming oilseed.
Cooking occurs when oilseeds are held at elevated temperature, preferably around 98°C ,
and at 1 0 -1 2 % moisture for at least 15 min without loss of moisture or drop in temperature.
When the chemical and physical reactions caused by cooking are completed, the oilseed is
dried to a lower moisture more compatible to the screw press. Sometimes separate vessels are
used for cooking and drying. Sometimes the same vessel is sectioned into different compart
ments, one or two for cooking, the rest for drying. Sometimes cooking is done by extrusion
followed by a separate drying step. Sometimes extrusion cooking is done in a high shear, low
moisture extruder, eliminating the need for a dryer.
The moisture-temperature conditions of cooking must prevail throughout the entire cook
cycle. If moisture is lost as the temperature is raised, the cooking suffers and may not occur.
A cooker vessel should have a sealed inlet and a sealed outlet to prevent moisture within it
from escaping. Feed screws pushing flakes against a hinged baffle plate as they fall into the
cooker serve as inlet seals. Similar screws or alternative devices (described below) serve as
discharge seals. Live steam (sometimes a mixture of live steam and water) is sprayed into the
cooker chamber to elevate the oilseed moisture and provide a humid atmosphere within the
cooker. Steam jackets on the shell provide heat.
Dryers, on the other hand, drive moisture out of the oilseed, reducing the moisture to
around 2-3% for more efficient pressing. Dryers are vented, and usually a draft of air is
induced to draw the water vapors away from the flakes. As flake moisture drops, the re
maining moisture is more tenaciously gripped by the solids. If temperature were not raised,
drying would stop.
Temperature, therefore, is raised. There are usually several drying vessels in series (or
separate chambers in series) each heated through a steam jacket. By the time the flakes exit
the last chamber (ready to enter the screw press) they should be at 2.5-3% moisture. They
will have to be heated to 115-126°C to reach this moisture.
Once the material reaches the desired final moisture and temperature it should immediately
enter the screw press. Long delays or, worse than that, pneumatic conveying of the dried
product can seriously upset the critical moisture-temperature conditions required for optimum
screw pressing. Preparation is summarized in Table 1.
1. Stack Cookers
Two types of jacketed cookers are used for oilseeds: stack cookers and horizontal cookers. A
stack cooker looks like a cylindrical silo standing upright. The interior is divided into several
horizontal sections, each with a steam-jacketed bottom. Down through the center is a rotating
shaft with horizontal arms that sweep across the bottom of each stack. Flakes enter the top
stack through an inlet seal (as described earlier).
116 Williams
Built into each stack bottom is a trapdoor that is pushed open every time the sweep arm
passes over it. The magnitude by which each trapdoor is opened can be adjusted by external
linkage. The flakes fall from stack to stack through these trapdoors. Each stack can be vented
to a common header to accommodate drying. The vents on the lower stacks are usually open;
those for the top stacks are closed to prevent moisture loss during cooking. Live steam (some
times mixed with water) is injected into the top stack through a hollow passage in the sweep
arm.
2. H orizontal Cookers
Horizontal cookers are arranged in clusters of separate vessels, one for cooking, three or more
for drying. Each vessel is steam-jacketed and contains a revolving shaft with protruding stir
ring paddles. Flakes enter the cooker through an inlet seal, migrate through the vessel as the
shaft tumbles them against the walls, and find a level within the cooker dependent on angle
of repose and dams that are inserted into the discharge end of the cooker.
The flakes spill over the dams into a closed chamber above a discharge screw that conveys
the flakes against a hinged plate (flake seal) as they fall away from the cooker. Live steam
and water are injected through the inlet face plate of the cooker near the bottom so that the
steam immediately contacts the flakes.
In comparison to a horizontal cooker, a stack cooker has an advantage in that it is more
compact than a cluster of horizontal vessels. A disadvantage is the time required for the
incoming material to reach cooking temperature. The incoming material falls on top of a thick
layer of material that is slowly agitated beneath its surface. It takes several minutes before
any of the incoming material contacts the steam-heated bottom. This allows oil-damaging
enzymes in the incoming material several minutes to start their action before the material
reaches inactivation temperature.
In a horizontal cooker, the incoming material immediately mixes with a large volume of
tumbling material already heated by the jackets. Therefore the incoming material reaches
enzyme inactivation temperature faster than it would in a stack cooker, and there is less
damage done by the enzymes. However, if enzymatic damage is a serious problem, as with
myrosinase in rapeseed [2], extrusion cooking (to be described below) can bring the oilseed
to inactivation temperature within a few seconds.
E. Detoxification
Several toxins are troublesome in oilseeds. Cottonseed contains gossypol. Castor seed con
tains three toxins, CB-IA being the most difficult to inactivate. Seeds like peanuts and cotton
seed that are permitted to mold can contain aflatoxin, a metabolic waste product of several
aspergini molds.
Gossypol is a deep red oil-soluble compound, unhealthy for chickens, turkeys, and baby
pigs. Gossypol also darkens cottonseed oil if allowed to dissolve into the oil. Cooking, when
done properly, can bind gossypol to the cottonseed protein [3], ensuring a light-colored oil.
Some bound gossypol in the meal is released within the animal’s digestive tract, however [4].
Most cottonseed meal is sold as an ingredient in ruminant feeds because, with their multi-
chambered stomachs, ruminants are not harmed by gossypol.
The toxin CB-IA in castor can be inactivated by procedures Rhee developed at Texas
A&M University [5,6] with chemical additives, such as Ca(OH )2 or NaOH and NaOCl and
using an extruder as a high pressure reactor vessel [7]. Aflatoxin is a metabolic waste product
of several species of molds, including ordinary bread mold, Aspergillus flavus. These molds
occur worldwide and are potential problems with any oilseed that molds before processing.
especially with peanuts, com, cottonseed, and copra. Methods have been developed for detox
ifying aflatoxin-contaminated meals with ammonia. The ammonia reacts with the aflatoxin,
breaking it into smaller nontoxic residues. Ammoniation can reduce aflatoxin levels from
300-1000 ppb to as low as 1-3 ppb, but it also reduces the meal’s nutritional value [8].
Since prevention is a less expensive procedure, there is interest in reasonable precautions
that could prevent the mold from growing in the first place [9,10]. Home et al. [9] outline
prevention methods on the farm, during harvest, and during storage that could minimize mold
growth. Hron et al. [11] and Lusas et al. [12] developed procedures for extracting aflatoxin
with alcohols. Those procedures will be discussed in more detail in Section V.B. Rhee
[5] has adapted the castor meal deallergenation process to destroy aflatoxin using NaOH and
N aO Cl.
F. Extrusion Preparation
1. Background
In 1954, Zies and Baer developed an expanding extmder using some of the technology they
used to build screw presses [13] (Fig. 1). This machine first served to cook and puff cereal
grains for the pet food industry [14] and later found use in other industries [15].
In the 1960s, realizing that the expansion and extmsion also inactivated enzymes, Baer,
Williams, and Zies developed a process for extmsion cooking of rice bran to inactivate lipase
and agglomerate the bran into spongelike collets prior to solvent extraction [16]. In the 1970s,
this expander was also used to transform flaked soybean and other oilseeds into porous col
lets [17].
An expander (or expanding extmder) consists of a rapidly rotating wormshaft within a
cylindrical barrel. Oilseed enters one end of the barrel and is forced out through a die plate
at the discharge end. Worm flighting is intermpted rather than continuous. The intermptions
mesh with stationary pins protmding from the barrel wall. Flights that rotate between station
ary pins impart a highly turbulent mastication to the oilseed. Injected steam elevates the
moisture and temperature and softens the material.
11 A I
Fig. 1 An expanding extruder. (Courtesy of Anderson International Corp.)

118 Williams
The oilseed, at 1 0 -1 3 % moisture and 110-115°C, is subjected to 1379-4137 kPa pressure

for 10-30 s before discharge from the expander. When the oilseed emerges from the ex
pander, some absorbed moisture flashes. This inflates the particles with internal pores and
surface cracks, giving the particles a porous, spongelike internal structure.
2. Cooking by Extrusion
Since expanders cook within 20 s, they counteract the activity of troublesome enzymes, such
as lipase in rice bran [18] and urease in soybean [19]. The short time between enzyme activa
tion by air and moisture and inactivation by heat destroys the enzyme before it has time to
cause damage. An expander is even more effective doing this than the horizontal atmospheric
pressure cookers described earlier.
Canola, for example, contains enzymes that release phosphorus compounds into the oil. If
flaked canola is processed through a stack cooker, the enzymes, which are triggered into
action when flaking exposes them to air, have several minutes to release phosphatides into the
oil before the slow-acting stack cooker brings the canola to a high enough temperature to
inactivate the enzymes. An expander brings the canola to full temperature in 20 s. This results
in a significant reduction of phosphorus compounds in the finished oil, 48 ppm compared to
350 ppm when using stack cookers. Reduced chlorophyll, free fatty acid, peroxide value, and
greenish color are added benefits.
Zhang et al. [20] found a similar reduction of phosphorus in soybean oil. Conventionally
prepared soybean yields degummed oil with 184 ppm phosphorus. Extrusion-prepared soy
bean yielded degummed oil with 67 ppm phosphorus.
In this application, the expander replaces the cooking tray in a stack cooker and cooks the
oilseed at 88-105°C. If the oilseed is going to a solvent extractor, it would continue to a
prepress or collet expander (same as flakes from a cooker). If the oilseed is going to a full
press, then drying to optimum crushing moisture (2-3% ) is still required.
3. Extrusion before Crushing

Expanding extruders are also used to prepare oilseeds for mechanical crushing. Nelson et al.
[21] passed soybean through a high shear extruder and converted the bean into a frothy meal
of almost fluidlike consistency. Fluidization was caused by moisture boiling through the
freshly liberated oil. The soybean, at 10% moisture and ambient temperature, was cracked
before entry into the extruder. High shear developed by the extruder pulverized the soybean
and ruptured the oil cells. Heat generated by shaft friction raised the temperature to 135°C.
This caused the moisture to flash from 10-14% down to 6-7% as the material flowed out of
the extruder.
Oilseed expanders fitted with adjustable jaw chokes to replace the die plates (Fig. 2) and
operated under low moisture, high shear conditions can transform cracked or uncracked oil
seeds, with or without preheating, into frothy, semifluid extrudates that flash to 5-7% mois
ture en route to the screw press. A rotating cone point inserted between two laterally posi
tioned stationary jaws generates the shear [22,23]. Both the proximity of the cone point to the
jaws and the opening between the jaws are adjustable.
4. Extrusion Before Solvent Extraction

The first oilseed application for extrusion was rice bran [16,18]. Rice bran comes into a
solvent extraction plant as a fine powder, which makes rice bran very difficult to extract. The
powder also contains an enzyme, lipase, that splits triglycerides into free fatty acids. Acti-
F 'iUff
Fig. 2 An oilseed extruder with an adjustable-jaw choke. (Courtesy of Anderson International Corp.)
vated when the bran is removed from the rice, lipase will raise the free fatty acid level
approximately three to seven percentage points every day [24]. The cooking conditions within
an expander inactivate lipase and convert the bran into porous, spongelike particles (collets)
that permit rapid flow of solvent. The same procedures also transform flaked soybean and
flaked cottonseed into porous collets that handle better in an extractor. Farnsworth et al. [25]
describe early extrusion research with cottonseed.
Flaked oilseeds extract well if the flakes are good quality, but if flakes are thick or crumbly
or if the feed to the flaker already contains fines, then extraction suffers. Extrusion is benefi
cial because it converts poor quality flakes into easily extracted collets. Even compared to
good flakes, extruded collets offer great advantages because collets are larger, heavier,
stronger, and more porous than flakes. This permits faster solvent flow (because of larger
particles), greater extractor capacity (because the heavier, more porous collets occupy less
space), and better drainage (because the collets are strong enough not to crumble as easily as
flakes do) [26].
120 Williams
Extrusion converts the oilseed’s protein into a gluelike condition that binds all the oilseed
particles into a densely compacted elastic mass. When the material exits the expanding ex
truder into atmospheric conditions, the mass inflates with internal pores as the superheated
moisture flashes into water vapor [27].
Some oilseeds are mechanically crushed and then solvent-extracted. The partially deoiled
“cake” from the screw press is sent to a solvent extractor. Expanding extruders can transform
the presscake into porous collets. Steam injected into the expander raises cake moisture 2 -4
percentage points. This must be done before the cake has cooled and hardened. Once cool
and hard, the denatured protein in the cake can no longer be transformed into a gelatinous,
inflatable condition. If oilseeds of high oil content are prepressed to 15-30% oil, they extrude
similarly to flaked soybean (at 18% oil) and flaked cottonseed meats (at 30-33% oil).
Slotted-Wall Expanders. Oilseeds containing more than 30-33% oil cannot be extruded
through a closed-wall expander because the oil accumulates within the expander and stops
steady-state operation. Anderson International (Cleveland, OH) developed a slotted-wall ex
pander that can make collets from high oil materials by allowing the liberated oil to escape
through the slotted-wall drainage cage [28] (Fig. 3). Field trials with canola [29] and other
oilseeds [30] showed that this Hivex expander could extrude full-fat oilseeds at high oil levels
and produce collets at 20-30% oil. Preparation for most oilseeds is to crack or flake the seeds
and heat to approximately 60-70°C while reducing the moisture to approximately 8% and
being careful not to denature the protein.
III. MECHANICAL EXTRACTION OR CRUSHING

A. Background
Most mechanical crushing of lipid-containing materials is done with screw presses. A screw
press is constructed of a screwlike wormshaft that rotates within a slotted-wall, cylindrical
A slotted-wall drainage cage on an oilseed extruder. (Courtesy of Anderson International

housing. It accepts a continuous stream of lipid-bearing material, propels it through the hous
ing, and compresses it under high pressure to squeeze out the oil. The first workable screw
press was developed by Valerius D. Anderson [31] in 1900 and improved in 1906. His im
proved screw press had a barrel made of two longitudinal halves (each an assembly of bars
and spacer shims) that were clamped together with heavy clamping bolts. Inserted between
the barrel halves were two lug bars called knife bars, 180° apart. The lugs protruded from the
barrel walls and meshed with interruptions in the wormshaft flighting. Figure 4 shows a mod
em press in cross section.
Incoming seeds are cleaned, dehulled, cracked or flaked, and sometimes cooked, as de
scribed earlier [32]. The final step is to reduce the oilseed’s moisture to a level more compati
ble for screwpressing: 2-3% for full pressing, 4-6% for prepressing. If the moisture is not
reduced, the solid material is too soft, and some of it flows out with the oil. If the moisture
is too low, the material will overheat from friction generated by the wormshaft. Moisture is
also important in the flaking and dehulling operations. Good cmshing requires proper moisture
control throughout the process.
In addition to oilseed preparation, three mechanical aspects of a screw press also influence
its performance: the wormshaft, the barrels (or drainage cages), and the choke mechanism.
The wormshaft revolves within the barrel, propelling the oilseed toward the discharge end,
where it encounters the choking mechanism. Shaft speed, screw geometry, and applied horse
power influence capacity and pressure profile along the barrel. Choke position also influences
the pressure profile, especially the final pressure before discharge. Drainage cage spacing
determines the ease with which oil rather than solids flows through the walls.
B. Full Presses Versus Prepresses

Screw presses serve two applications for oilseeds: fullpressing and prepressing. Fullpressing
requires shaft designs that generate pressure—estimated at 96,500 kPa—to squeeze out as
Fig. 4 Sectional view of a screw press showing knife bars and wormshaft flighting. (Courtesy of
Anderson International Corp.)
122 Williams
much oil as possible, preferably to 3-5% residual. Animal materials can be reduced to 7 -
10% residual.
Prepressing is used to partially deoil the seed in preparation for solvent extraction. A
prepress (Fig. 5) generates less pressure than a full press, consumes less kilowatts per kilo
gram of oilseed, and runs at higher capacity. The wormshaft revolves faster, inlet moisture is
higher, and less choking (to apply backpressure) is needed. More oil, however, is left in the
solids: 15-30%, compared to 3-5% for fullpressing.
C. The Wormshaft
The wormshaft picks up the feed entering the screw press and transports it along the barrel,
subjecting it to greater and greater pressure by means of a steadily increasing hub diameter
and/or shorter flight pitch. Ideally, wormshaft geometry should increase compaction while at
the same time compensating for loss of volume as oil escapes through the barrels [33]. This
is not easy because different oil-bearing materials generate different resistances to flow, which
influences the pressure. Even the same material with different preparation can vary in its
resistance to flow.
Hull content, for example, increases resistance to flow. Materials high in hull content do
not easily slide through the screw press’s channel (annular space between the shaft and the
barrel). Shafts designed for these materials have uniform channel depth and uniform pitch.
They rely on the material’s resistance to flow to generate pressure. An adjustable choking
device at the end of the shaft can increase the material’s resistance to flow, thereby increas
ing pressure.
Softer materials flow more easily. There is not enough drag for a uniform channel depth,
uniform pitch wormshaft to develop sufficient pressure. Shafts for soft materials rely on de
creasing channel depth and decreasing pitch to generate pressure and also on “cones” or
funnel-shaped segments inserted between the worm segments. Channel depth decreases rap
idly as the oilseed flows across the cone and can be very thin at the cone’s discharge end.
Fig. 5 A prepress for partially deoiling lipid-containing materials. (Courtesy of Anderson Interna
tional Corp.)
3.17 mm in some cases. Cones present obstacles for the propelling action of worms. Worms
preceding the cones have to generate higher pressure and therefore grind out more oil.
Some screw presses use two shafts, one that, in effect, prepresses, followed by a second
one that fullpresses (shown in Fig. 4). The first shaft is placed upright above the main (hori
zontal) shaft and pushes feed down into the main shaft. The vertical shaft often has a drain
age cage.
D. The Choking Mechanism

Most presses have a choking mechanism at the discharge end. Choking provides a means to
adjust the press’s internal pressure to compensate for differences in feed quality or preparation
and to provide a means to compensate for wormshaft and barrel bar wear. Presses designed
without chokes are not so tolerant of wear. Chokes also allow for starting up and shutting
down the press under conditions of low internal pressure (choke wide open). Once the press
is running, the choke is closed to increase pressure and bring down the residual oil. At shut
down, just before the feed flow is stopped, the choke is fully opened to ensure that all solidly
pressed material can be discharged out of the press.
There are different types of choking mechanisms. One type resembles an iris diaphragm.
A cluster of movable jaws is inserted into or retracted from the channel by means of a large
ring supporting the opposite ends of the jaws and allowing them to insert or retract depending
on which direction the ring is rotated. Rotation is accomplished by a worm gear operated
manually or through a two-directional motor.
Other mechanisms resemble tapered cones that are mounted on the discharge end of the
shaft and are slid toward the endplate of the barrel (or away from it) to restrict or clear the
channel and thereby influence pressure.
E. Cooling the Press

A screw press generates frictional heat as the wormshaft slides against the oilseed. The heat
could in time get the shaft and barrels hot enough to overheat the oil and scorch the solids. If
this is of concern, then cooling is required.
Some screw presses are cooled by recirculating product oil over the drainage cages [34,35].
Product oil, cooled in a water-cooled heat exchanger to 48.9°C, flows at approximately 150
L/min over the barrel surfaces. This dilutes the hot freshly pressed oil with a larger volume
of cool oil, flushes the barrels to rinse away any solids that linger on the barrel, absorbs heat
from the screw press, keeping temperature within acceptable limits, and also blankets the
drainage surfaces with flowing oil, thereby preventing the fresh oil from lingering on the hot
barrel, exposed to the oxidizing effects of air. Another method of cooling screw presses is to
water-cool the barrel housings. The freshly pressed hot oil then cools as it drips off the water-
cooled barrels.
Wormshafts are also water-cooled. Flowing through a rotary seal, water enters through a
tube inserted into a cylindrical opening bored longitudinally through the center of the shaft.
Water flows through the tube toward one end of the shaft, then through the bore back toward
the opposite end, and finally out through the rotary seal.
F. Slotted Barrels
The drainage cage is assembled of rectangular bars fitted into a supporting frame (Fig, 6).
Each bar is separated from the others by spacing clips, which are made in different thicknesses
124 Williams
Fig. 6 A drainage cage showing the bars. (Courtesy of Anderson International Corp.)
to allow for different gaps between the bars. High pressure full presses use narrow spacers,
0.13-0.76 mm. Lower pressure prepresses, running at higher capacities and liberating a
greater volume of oil, generally use thick spacers, 0.5-1.5 mm.
F. Removing Solids from the Oil

Freshly pressed oil contains 2-7% oil-free solids, sometimes more. The solids are usually
removed in a two-step procedure. Larger particles are removed with a vibrating screen or a
screening tank. A screening tank resembles a large reservoir for freshly pressed oil. Solids
that settle to the bottom are captured by a slow-moving drag mechanism that lifts them up a
shrouded passage on one end of the tank and slowly drags them over a drainage screen.
Excess oil drains back into the tank; the solids continue to the conveyor, taking fresh material
to the screw press. The second clarification step is a plate-and-frame filter press.
Unlike the first step, solids from a filter press do not flow continuously. They are accumu
lated within the filter press for several hours as the oil passes through the filtering cloths.
Then the press is opened, and all the accumulated solids are discharged at once. To prevent
disturbing the steady-state operation of the screw press, the discharged solids from the filter
press are dropped into a hopper equipped with a variable-speed screw that meters the solids
into the fresh material. The screw press, therefore, receives a uniform blend of fresh material,
drained solids, and filter presscake.
IV. SPECIAL CASES

A. Olives
Olive oil production is concentrated in the Mediterranean region, which is the source of 98%
of the olive oil marketed throughout the world [36]. They are processed within 3 days at low
temperature, no higher than 40°C [37, p. 345] to avoid any deterioration of oil quality.
The olives are washed, and the pulp is ground into a thick paste (called malaxer) that is
processed by one of three methods to obtain the oil. One method is pressing in batch hydraulic
presses. The pastelike malaxer is spread onto pressing bags and stacked onto the lower pres
sure plate of a vertical hydraulic plate press. When pressure is applied, an emulsion of oil
and water flows from the press. The emulsion is centrifuged, and the oil is filtered using
diatomaceous earth [36,38]. Oil yields can be increased by adding enzymes to the malaxer
[39].
The second method centrifuges the malaxer itself. A third method processes the malaxer
in an extraction vessel equipped with stainless steel blades that are inserted into the malaxer.
Because of differences in surface tension, oil rather than water clings to the blades and mi
grates along the blade’s surface to drip off and flow out of the extractor [37, p. 350].
B. Palm Fruit Versus Palm Kernels

Palm fruit oil accounted for 16% of the world’s vegetable oil production (in 1994) [40],
second only to soybean oil [41]. The fruits grow in large grapelike clusters weighing 10-20
kg and containing more than 1500 fruits [42]. The fruits arrive at the oil mill containing 20-
25% moisture [37, p. 219]. To avoid spoilage, the fruit is processed as quickly as possible.
The first step is sterilization. The fruits arrive in narrow-gauge railway wagons, which are
rolled into horizontal sterilization chambers.
After sterilization, the fruit clusters are removed from the wagons and fed into a stripper
that strips the stalks from the individual fruits. The fruits are washed to remove sand and dirt,
then pass into a digester, which breaks the structure of the skin and pulp but leaves the
kernels intact.
The digested material is then pressed in continuous, high volume, low pressure screw
presses, which press out a liquor consisting of approximately 53% oil, 40% water, and 7%
finely divided solids [42]. The liquor goes to a continuous settling tank, where it separates
into three phases: a top phase of clear oil containing small amounts of moisture and solids, a
bottom watery phase containing most of the solids, and an intermediate phase not yet sepa
rated. Continuous streams flow from the top and the bottom. The top phase (the oil) is centri
fuged to remove solids and undissolved water, then passes through an evaporative dehydrator
to remove the dissolved water.
Screw press solids (a mixture of pulp solids, skin residue, and kernels) contain about 16%
moisture. The kernels are freed of pulp in a pneumatic separator and are then conditioned
to 10-12% moisture. This shrinks the meats, loosening the hulls. The kernels then pass
through a centrifugal nutcracker [37, p. 219], which discharges a mixture of loose meats
and hulls.
Hulls are removed by slurrying the mixture with water and passing it through a liquid
cyclone. The meats pass out the bottom; the hulls pass out the top with the water. Another
method mixes the cracked kernels with a clay (or salt) aqueous slurry at 1.18-1.20 specific
gravity. The meats float; the hulls sink. The meats are dried to below 8% moisture and sent
for crushing in high pressure oilseed presses as described earlier.
126 Williams
C. Animal Materials
Lipids are obtained from both animal and vegetable sources. These source materials vary
considerably. Most vegetable lipids come from oilseeds. Most animal lipids come from fatty
tissues from animal bodies if the fat is destined for human consumption or, if the fat is not
destined for human consumption, from all lipid-bearing portions of animal bodies. Animal
materials are prepared for lipid extraction much differently than oilseeds because they contain
considerably more moisture and, on a moisture-free basis, contain more lipid, and the lipid
(called fat rather than oil) is usually solid at room temperature.
Leaf fat from hogs has the highest fat content (92-95%). Fatty tissue from cattle contains
60-80% fat. These and similar materials are usually processed under hygienic conditions to
produce edible-grade fats such as lard. All unused portions of animals from slaughterhouses,
livestock that died before slaughter, dead farm animals, fish harvested for production of fish
lipids, fish residues from fish processing plants, recycled restaurant grease— any source of
readily available fatty animal tissue—is processed under less hygienic conditions to produce
inedible-grade animal fats such as tallow, grease, and fish oil. These products find a ready
market in industries that produce soaps, lubricants, and animal feeds.
The separation of fat from animal residues (called rendering) involves various methods to
liberate the fat from the solids, remove most of the moisture, then press or solvent extract
the residue to obtain the fat [1]. To obtain edible fats where color, flavor, and keeping quali
ties are important and the amount of solids in the raw material is small, the fatty tissue is
mixed with water and agitated while being heated in order to release the fat from the tissue.
The low fat solids residue, called tankage, is dried and marketed as an ingredient for animal
feeds.
Another method of rendering (dry rendering) heats the material without addition of water,
and the material’s own moisture is boiled out. The resultant low moisture residue, containing
all the fat and solids, is drained of free fat and fullpressed or solvent-extracted as oilseeds
are. Dry rendering is the more economical of the two methods. It is done batchwise under
atmospheric conditions, batchwise under pressure, and in continuous systems operating under
atmospheric conditions, under pressure, and under vacuum [43,44].
Another method mixes the ground raw material with recycled fat to form a slurry of ap
proximately 68% fat, 10% solids, and 22% moisture. The slurry is pumped through an evapo
rator, where the water is removed under vacuum, then through a centrifuge to remove most
of the fat. The solid residue, at approximately 35% fat, is pressed to 10% fat in screw
presses [45,46].
Fish can be rendered in the same equipment used for other animal materials, but fish is
processed separately. High moisture, strong odor, and difficult processing (compared to other
materials) encourage procedures and equipment modified specifically for fish.
Bemadini [37, p. 371] describes a rendering system that continuously feeds whole fish into
horizontal cylindrical steam-heated cookers. Twenty minutes at 90-95°C sterilizes the fish
and causes them to disintegrate into an aqueous mass. The cooker discharges into a continu
ously operated, low pressure screw press (similar to those used for palm fruit), which sepa
rates the mass into an aqueous phase containing most of the oil, some suspended solids, and
most of the soluble solids, and a solid phase containing the larger solids and some oil and
having about 50% moisture content.
The liquid phase is separated into oil and water fractions. The water fraction is concen
trated by evaporation, and the solids-rich “stick water” is added to the presscake, which is
thermally dried to 10% moisture. The finished fish meal has about 10% oil.
There has also been research into wet rendering with the assistance of enzymes that pro
mote fat separation by dissolving the connective tissue. In 1982, Freeman [47] described the
use of natural enzymes to improve the rendering of catfish viscera.
V. SOLVENT EXTRACTION
A. Traditional Solvents
In the early days of solvent extraction, many solvents were investigated, but by 1947 the most
commonly used solvent was hexane. For a review of other solvents, see Johnson and Lusas
[48], Lusas et al. [49], and Hron et al. (covering biorenewable solvents) [50].
Hexane is a clear hydrocarbon that boils at 68.9°C. It is derived from petroleum, is fully
miscible with oil and immiscible with water, and does not impart an objectionable odor to the
oil or meal. As long as it is in abundant supply, it will remain the solvent of choice. But
hexane is flammable, which presents a safety hazard. Plants using hexane are advised to avoid
conditions that could generate spark-causing static electricity or could permit solvent vapors
to leak outside the vessels and accumulate in the plant. Advisory groups like The American
National Fire Protection Association (NFPA) issue manuals that contain recommended safety
precautions [51].
B. Alternative Solvents
Halogenated solvents have long been considered and sometimes have been used because of
their nonflammability. Johnson et al. [52], extracting cottonseed with dichloromethane, re
ported a significant reduction of gossypol and aflatoxin in the meal.
Acetone can extract oil, gossypol, and aflatoxin. It is, however, miscible with water and
must be protected against absorbing too much water from the incoming material. The U.S.
Department of Agriculture sponsored research into the use of an azeotropic mixture of ace
tone, hexane, and water (ratios of 53:44:3) to extract gossypol along with the oil from cotton
seed [53]. The gossypol was removed from the oil by conventional alkali refining. The resul
tant meal produced superior growth in chicks, showing that the gossypol was gone, but this
three-component solvent was never accepted by the oilseed industry because of difficulty in
evaporating an azeotropic mixture and because the acetone caused an off-odor in the meal
[54].
Another alternative solvent is liquid CO 2 at (or above) its critical temperature and pressure.
The critical point for carbon dioxide is a pressure of 7387 kPa and a temperature of 31°C.
Extraction with supercritical CO 2 usually takes place at pressures of 20,670-103,352 kPa and
temperatures of 50-80°C, depending on the product being extracted. Extractions are per
formed by circulating supercritical CO 2 through a pressurized cell filled with the oilseed. List
and Friedrich [55] reported that almost all of the oil could be extracted from 0.10 mm soy
flakes, but if the flakes are thicker (0.38 and 0.81 mm), only 8 7 % or 67%, respectively, of
the oil is extracted. The extraction vessels are operated batchwise rather than continuously.
Feidrich and coworkers [56,57] reported producing a better quality soybean oil using super
critical CO2 than is obtained with hexane. Other oilseeds studied for supercritical fluids extrac
tion are cottonseed [58], wet-milled and dry-milled com germ [59], and rice bran [60]. Such
extraction is feasible, but the required high operating pressure makes the equipment expen
sive. Supercritical fluid extraction is presently used on some high value materials such as
coffee, hops, and natural flavor extracts.
Environmental concerns and the possible short supply of petroleum-based solvents have
spurred research on extraction processes using ethanol and isopropanol. These solvents are
128 Williams
attractive because they can be produced from starchy cereal grains and they can extract some
undesirable constituents such as gossypol in cottonseed and aflatoxins.
Kamofsky and others worked with alcohol extraction [61,62] in the 1980s. Hron and co
workers [11,63-66] worked with ethanol and developed a benchtop process for extracting
cottonseed oil along with gossypol and any aflatoxin in the cottonseed using 95% ethanol
[11]. Abraham et al. [64] tempered cottonseed meats to 1 4 % H2O, flaked them to 0.229 mm,
and dried them to 2% moisture (to ensure that the ethanol would not extract moisture from the
incoming flakes). Unheated ethanol extracts gossypol in a first extraction step. The ambient
temperature mixture then passes through an absorber to remove gossypol and returns to the
first-stage extractor.
In the second extraction step, heated ethanol extracts oil and aflatoxin. The resultant oil-
alcohol mixture is chilled to 4 °C to reduce the oil’s solubility in the alcohol. Centrifuging
separates the oil and alcohol into fractions that are sent to reverse osmosis membranes and
absorbers to remove aflatoxin and any remaining gossypol. The crude oil-alcohol is then
miscella-refined. Hron et al. [65] report that aqueous 95% ethanol converts free gossypol into
bound gossypol even at ambient temperature. Acidulating the aqueous ethanol with a tribasic
acid and heating the cottonseed improves the extraction [66].
Lusas et al. [12] chose isopropanol because of its greater oil solubility, lower heat of
vaporization, and freedom from scrutiny by alcohol taxing agencies. They chose 93-96%
isopropanol to ensure a high enough oil-holding limit of the alcohol. Concentrated isopropanol
can hold 48% oil in solution with the alcohol. A constant boiling azeotrope can hold only
1 2 - 1 4 % oil.
Lusas et al. developed a pilot process that consists of adjusting cottonseed meats to 12%
moisture, flaking them cold to 0.38 mm, preconditioning them in an expanding extruder at
88-93°C, holding the extruded meats for 30 min, and then passing them through a second,
collet-forming expanding extruder at 104-110°C. The collets are then extracted with 95%
isopropanol at I T C in a simulated nine-stage countercurrent extraction. Solvent-to-meal ratio
is 1:1 by weight. Free gossypol in the meal was reduced to approximately 200 ppm. Chill
separating at 5°C removes most of the alcohol. The oil is then desolventized in a thin film
evaporator. Subsequent caustic refining and bleaching produced oils of a quality that compares
favorably with hexane-extracted oils.
The isopropanol is then reconcentrated by perevaporation (a membrane separation proce
dure in which the mixture is vaporized and the vapors are separated by the membrane). Vac
uum draws water vapor, the smaller molecule, through the membrane. The water vapor is
then condensed and discarded [67].
C. Extractors
1. Background
The extractor most commonly used today is a percolation extractor. A percolation extractor
carries the lipid-bearing material through a vaportight chamber where solvent rains down
through the material, dissolving out the oil, similar to the way a coffee percolator works. The
material is transported through the extraction chamber on a woven wire belt (or by similar
means), and the solvent makes multiple passes through the bed of material. Collection sumps
under the material and overhead spray heads separate the flow of solvent and the lipid-bearing
material into “stages.” The flow is countercurrent: fresh solvent encounters previously ex
tracted material, and fresh lipid-bearing material encounters solvent already containing some
oil. This countercurrent flow permits high levels of oil removal using as little solvent as prac
tical.
In most percolation extractors, incoming material is mixed with oil-laden solvent as the
material is conveyed into the extractor. This extracts some readily available oil but also carries
out loose particles of solids (fines). The oil-laden solvent (miscella) from this first stage is
pumped through the second-stage compartment to allow the already formed bed of material to
filter out the fines. The oil-solvent mixture then leaves the extractor as full miscella. Lean
miscella from stage 3 skips stage 2 to mix with incoming feed in stage 1 then goes on to
stage 2 as described above. The rest of the stages operate in true countercurrent fashion. Stage
3 gets miscella from stage 4, stage 4 gets miscella from stage 5, stage 5 gets miscella from
stage 6, and stage 6, usually the last stage, gets washed with fresh solvent. Then the solids
free drain for several minutes before discharge from the extraction chamber.
There are several major types of percolation extractors, classified mainly by their means of
transporting solids: perforated belt, rectangular loop, circular basket (rotary or stationary), and
sliding-bed [51].
2. P erforated B elt Extractors

Manufactured by DeSmet (Edegem, Belgium), the perforated belt extractor (Fig. 7) employs
a flexible perforated belt to carry the lipid-bearing material through the extraction chamber.
The belt, made of wire mesh or wedge-bar sections, forms a looped endless conveyor and
carries the material as one continuous bed. Bed depth is adjustable and is usually 0 .5 -2 m.
Belt speed determines residence time.
The belt travels on a slight incline through the extractor. This reinforces countercurrent
flow by causing miscella washes to seep through the solids in a direction slanted toward the
feed end. Rakes scrape ridges into the bed’s surface after each miscella spray to promote
more rapid penetration of miscella.
The miscella from each stage is pumped back onto the stage it came from. Miscella ad
vances from stage to stage by overflowing the sumps under the belt.
Fig. 7 Perforated belt extractor. (Courtesy of Extraction DeSmet SA/NV.)

130 Williams
3. Rectangular Loop Extractor

Manufactured by Crown Iron Works Co. (Minneapolis, MN) the rectangular loop extractor
employs an en masse drag conveyor that moves the solid material through a closed-loop
housing of rectangular cross section (Fig. 8). Fresh material flows into the drag conveyor just
before the conveyor starts its downward pass to the lower level and is deposited as a shallow
bed (usually less than 0.77 m deep) that travels a distance that is usually 50 times the bed
depth.
Rich miscella is pumped into the material during the downward pass. The mixture is then
pulled horizontally along a slotted vee-bar floor, through which the miscella drains into the
first-stage sump. After the solids pass through stages 3, 4, and 5, they are lifted to the upper
level and are turned upside down as the drag conveyor loops over the upper vee-bar flooring.
The solids then pass through stages 6 and 7.
Miscella from each sump can be pumped back to the stage it came from or to the next
stage. If the miscella is returned to the stage it came from, it can still move to the next stage
through overflow weirs built into the sumps.
Crown later reversed the drag’s direction of travel. Fresh material in the Model III enters
at the beginning of horizontal travel on the upper level, travels through several stages on that
level, then loops around to the lower level (turned over in transit) for the rest of the extraction.
The spent material drops out on the lower level, allowing the drag conveyor to be empty as
it rises to the upper level. Having the conveyor empty at this point reduces horsepower con
sumption of the extractor drive.
Fig. 8 Rectangular loop extractor. (Courtesy of Crown Iron Works Co.)

4. C ircular B asket Extractors

There are two varieties of circular extractors. They look similar, but the parts that move in
one do not move in the other. A “rotary” circular extractor is built around a slowly revolving
circular disk divided into pie-shaped compartments that hold the material to be extracted.
Incoming material falls into each compartment as that compartment passes under a stationary
inlet spout. The compartments (“cells”) then rotate under a circle of stationary spray heads
that spray miscella in a countercurrent fashion into the cells. The miscella flowing out of the
cells gathers in a circle of stationary sumps beneath the rotating disk. In this fashion, the
lipid-bearing material passes under a series of spray heads, then under a spray of fresh solvent,
then travels over a final drainage sump, and finally moves over a discharge hopper where the
floor of the cell swings open, and the extracted, drained, but solvent-laden solids (marc) drop
into a hopper. Continued rotation shuts the door and moves the now empty cell under the
flake inlet. Called a rotary extractor because the cells rotate, this type of extractor is supplied
by Davy International (Pittsburgh, PA) (Rotocel) (Fig. 9) and by Krupp (Hamburg, Ger
many) (Carrousel).
The other type of circular basket extractor is made of a circular disk of cells with wedge-
shaped compartments, but the disk does not rotate. An apparatus containing the flake inlet
spout and the fresh solvent and miscella spray heads revolves above the stationary circular
disk. Another apparatus containing the drainage sumps and marc hopper (connected to a com
mon drive with the top apparatus) revolves below the stationary disk. Termed a stationary
basket extractor, this variety is supplied by French Oil Mill Machinery Co. (Piqua, OH).
All versions of circular extractors handle deep beds of solids, sometimes 2 -3 m deep.
There are several mechanical variations of both versions. In one variation, the individual cells
have hinged bottoms that swing open to empty the cell. In another variation (termed the
sliding plate), the solids slide against a common one-piece bottom that has an opening through
which the solids fall when the extraction cycle is completed.
Fig. 9 A Rotocel circular basket extractor. (Courtesy of Davy International Corp.)

132 Williams
Fig. 10 Sliding-bed extractor. (Courtesy of Krupp Machienentechnik GmbH.)
5. Sliding-B ed Extractors
In a sliding-bed percolation extractor, a chain-and-sprocket assembly with drag bars slides (or
drags) the material along two stationary perforated steel plates, one above the other. The
solids enter on the upper plate, pass under spray heads that spray miscella onto them in a
countercurrent flow pattern, and then drop to the lower plate, where they form a fresh bed of
material to be slid under the rest of the miscella spray heads. The fresh hexane rinse is on the
lower plate.
Manufactured by Krupp, the sliding-bed extractor is shown in Fig. 10. How deep the bed
of material is on the plates is an easily adjusted variable. Bed depth for most materials is
0 .5-1.3 m.
D. Solvent Recovery
Solvent is removed from the miscella in a two-step process: evaporation followed by steam
stripping. Evaporation is usually done in a double-stage evaporator. In the first stage, hot
solvent vapors that were boiled out of the solids condense against the first-stage evaporator
and boil solvent from the miscella. In the evaporator’s second stage, jacket steam provides
heat. The oil leaves containing approximately 5 % solvent. It then enters a stripping column
where the last traces of solvent are stripped out by live steam injected countercurrent to the
oil flow.
Steam and solvent vapors (for this example, hexane) from the stripper are condensed, and
the liquid water and hexane go to a solvent-water separator. The lighter hexane rises, spills
over a weir at the top, and goes to the solvent work tank. The heavier water sinks, then flows
up a siphon tube (pushed upward by the combined weight of water and solvent above it), and
goes to a water stripper that strips out the last traces of solvent. The solvent-free effluent
water then leaves the system.
Solvent is removed from the solvent-laden solids in desolventizers similar to the cooker
vessels discussed earlier. The solvent is driven out thermally with heat supplied by steam
jackets and by injected live steam, which strips the solvent in the same way live steam strips
solvent from the oil. The heat and moisture (from the steam) also partially cook the solids
and are used to “toast” or inactivate urease in soymeal. The cook conditions also lower the
nitrogen solubility index (NSI) of the protein.
High NSI materials are desired for manufacturing meat analogues from vegetable proteins.
If a processor wants to make high NSI products, a flash desolventizer is used. These vessels
desolventize the marc in a circulating stream of preheated solvent vapor followed by a low
temperature, low moisture finishing step that removes the last traces of solvent without dena
turing the protein [68]. Some plants divert part of their marc to a flash desolventizer if they
have a market for the higher value, high NSI flours. The rest of their marc (soymeal, for
example) goes to a desolventizer/toaster to serve the animal feed market.
Besides oil and solids, there are two additional streams flowing out of a solvent extraction
system: effluent air and effluent water. Air enters with the incoming material and through
leaks in the vessels that operate under vacuum. Air leaves the system through a water-cooled
vent condenser followed by a refrigerated vent condenser, a charcoal absorber, or a mineral
oil scrubber and then passes through a flame arrester before release to the atmosphere.
Effluent water (in a hexane extraction system) passes through a steam-heated “wastewater
stripper,” where live steam strips out the last traces of hexane. All recovered hexane vapors
are condensed and sent to the solvent-water separator. The hexane is recycled to the extrac
tor. The water is discharged from the system.
REFERENCES
1. F. Norris, Extraction of fats and oils, in Bailey's Industrial Oil and Fat Products (Daniel Swem,
Ed.), 4th ed., Vol. 2, Wiley, New York, 1982, p. 175.
2 . J. R. Reynolds and C. G. Youngs, Effect of seed preparation on efficiency and oil quality in
filtration extraction of rapeseed, J. Am. Oil Chem. Soc. 41: 63 (1964).
J. W. Dunning and R. J. Terstage, Effect of overcooking of cottonseed meats on quality of meals,
J. Am. Oil Chem. Soc. 29: 153 (1952).
4. M. C. Calhoun, J. E. Huston, C. B. Calk, B. C. Baldwin, Jr., and S. W. Kuhlmann, Cattle
Research with Gossypol Containing Feeds, National Cottonseed Products Assoc., Memphis, TN,
1991, pp. 39-51.
5. K. C. Rhee, Detoxification and deallergenation of oilseed meals, in Proceedings of the World
Conference on Oilseed and Vegetable Oil Processing Technology (T. H. Applewhite, Ed.), AOCS
Press, Champaign, IL, 1993, pp. 346-351.
6 . The Processing of Castor Meal for Detoxification and Deallergenation, The International Castor
Oil Association, Tech. Bull. 1-1989, Ridgewood, NJ, 1989, pp. 58-62.
7. J. Horton and M. A. Williams, A cooker-extruder for deallergenation of castor bean meal, J. Am.
Oil Chem. Soc. 66: 227 (1989).
P. J. Coulty, Aflatoxin in cottonseed products— what can be done?. Oil Mill Gaz. 96: 26 (1991).
9. C. W. Home, L. L. Boleman, C. G. Coffman, J. H. Denton, D. B. Laehom, A. B. Wagner, Jr.,
M. L. Walker, R. E. Withers, Jr., and C. G. Woefel, Mycotoxins in Feed and Food-Producing
Crops, Texas Agric. Extension Service, Texas A&M Univ., College Station, TX, Bull. B-1278,
12-88 (1988) Revision, pp. 17-24.
10. L. Lee, Aflatoxin, J. Am. Oil Chem. Soc. 66: 1398 (1989).
11. R. J. Hron, Sr., M. S. Kuk, G. Abraham, and P. J. Wan, Ethanol extraction of oil, gossypol and
aflatoxin from cottonseed, J. Am. Oil. Chem. Soc. 71: 417 (1994).
12. E. W. Lusas, L. R. Watkins, and S. S. Koseoglu, Isopropyl alcohol to be tested as a solvent, Int.
News Fats, Oils Relat. Mater. {INFORM) 2: 970 1991).
13. C. W. Zies, Apparatus for the preparation of food compounds, U.S. Patent 3,108,530 (to Interna
tional Basic Economy Corp) (Oct. 29, 1963).
134 Williams
14. M. A. Williams, R. E. Horn, and R. P. Rugala, Extrusion, extrusion, extrusion. Food E ng. 49:
99 (1977) (Part 1) and 49: 87 (1977) (Part 2).
15. M. A. Williams, Preparation of oilseeds to improve extraction of fats. Extrusion Commun. 6:
12 (1993).
16. S. Baer, M. A. Williams, and C. W. Zies, Pre-treatment of oleaginous plant materials, U.S.
Patent 3,255,220 (to International Basic Economy Corp.) (June 7, 1966).
17. M. A. Williams, Extrusion preparation for oil extraction, INFORM 6: 289 (1995).
18. M. A. Williams and S. Baer, The expansion and extraction of rice bran, J. Am. Oil Chem. Soc.
42: 15 1 (1965).
19. M. A. Williams, Extruded starter pig feeds. Feed Manage. 42: 20 (1991).
20. F. Zhang, S. S. Koseoglu, and K. C. Rhee, Effect of expander process on the phospholipids in
soybean oil, J. Am. Oil Chem. Soc. 71: 1145 (1994).
21. A. I. Nelson, W. B. Wijeratne, S. W. Yeh, T. M. Wei, and L. S. Wei, Dry extrusion as an aid
to mechanical expelling of oil from soybeans, J. Am Oil. Chem. Soc. 64: 1341 (1987).
22. M. A. Williams, Adjustable-jaw expanders. Feed Manage. 44: 25 (1993).
23. M. A. Williams, Advancement of processing hardware. Oil Mill Gaz. 10: 20 (1995).
24. M. A. Williams, Extrusion of rice bran, in The Proceedings of the World Congress on Vegetable
Protein Utilization in Human Foods and Animal Feedstujfs (T. H. Applewhite, Ed.), AOCS Press,
Champaign, IL, 1989, pp. 100-102.
25. J. T. Farnsworth, L. A. Johnson, J. P. Wagner, L. R. Watkins, and E. W. Lusas, Enhancing
direct solvent extraction of oilseeds by extrusion preparation. Oil Mill Gaz. 91: 30 (November
1986).
26. E. W. Lusas and L. R. Watkins, Oilseeds: extrusion for solvent extraction, J. Am. Oil Chem.
Soc. 65: 1109 (1988).
27. M. A. Williams, Extrusion preparation for solvent extraction. Oil Mill Gaz. 91: 24 (November
1986).
28. M. A. Williams, Apparatus and method for the continuous extrusion and partial deliquefaction of
oleaginous materials, U.S. Pat. 4,901,635 (to Anderson International Corp.) (Feb. 20, 1990).
29. M. A. Williams, Expanders for high-oil seeds. Oil Mill Gaz. 95: 10 (1989).
30. M. A. Williams, Using expanders to improve extractability, INFORM 1: 959 (1990).
31. V. D. Anderson, Press, U.S. Pat. 647,354 (Apr. 10, 1900).
32. J. W. Dunning, Unit operations in a mechanical extraction mill, J. Am. Oil Chem. Soc. 33:
462 (1956).
33. M. A. Williams, Placing expanders and screw presses under automatic controls: some entry level
systems. Oil Mill Gaz. 96: 38 (June 1991).
34. R. T. Anderson, Oil settling tanks, U.S. Patent 2,365,442 (to V. D. Anderson Co.) (Dec. 19,
1942).
35. R. T. Anderson, Screening and conveying means for the solids of oil separating tanks, U.S. Patent
2,420,927 (to the V. D. Anderson Co.) (May 20, 1947).
36. S. Latta, Gourmet oils in the 1990s, INFORM 2: 99 (1991).
37. E. Bemadini, The New Oil and Fat Technology, Tecnologie s.r.i., Rome, 1973.
38. FAO, Olive Oil Technology, Food and Agriculture Organization of the United Nations, Rome,
1975.
39. F. M. Christensen, Extraction by aqueous enzymatic processes, INFORM 2: 984 (1991).
40. Y. Basiron and R. Abdullah, The palm oil situation in south east Asia, INFORM 6: 891 (1995).
41. Anon., Palm oil update, INFORM 6: 884 (1995).
42. K. G. Berger, Production of palm oil from fruit, J. Am. Oil Chem. Soc. 60: 158A (1983).
43. W. H. Prokop, Rendering systems for processing animal by-product materials, J. Am. Oil Chem.
Soc. 62 : 805 (1985).
44. S. Ottewell, Rendering meat safe?, Chem. Eng. 477: 15 (1990).
45. C. Greenfield, Process of dehydration of fatty materials, U.S. Patent 2,651,647 (Sept. 8 , 1953).
46. C. Greenfield, Dehydration of fluid fatty mixtures above normal coagulation temperature, U.S.
Patent 3,076,715 (Feb. 5, 1963).
47. D. W. Freeman, in P roceedings of the 11th Annual Meeting, Inland Commercial Fisheries Associ
ation (D. Yocum, Ed.), Arkansas Game and Fish Commission, Little Rock, AR, 1982, pp. 1-10.
48. L. A. Johnson and E. W. Lusas, Comparison of alternative solvents for oils extraction, J. Am.
O il C hem . Soc. 60: 229 (1983).
49. E. W. Lusas, L. R. Watkins, and K. C. Rhee, Separation of fats and oils by solvent extraction:
non-traditional methods, in Proceedings of the W orld Conference on Edible Fats and Oils Pro
cessing: Basic Principles and Modern Practices (D. R. Erickson, Ed.), AOCS Press, Champaign,
IL, 1990, pp. 56-78.
50. R. J. Hron, Sr., S. P. Koltun, and A. V. Grad, Jr., Biorenewable solvents for vegetable oil
extraction, J. Am Oil Chem. Soc. 59: 674A (1982).
51. NFPA, NFPA 36, Solvent Extraction Plants, 1993 ed., National Fire Protection Association,
Quincy, ME.
52. L. A. Johnson, J. T. Farnsworth, N. Z. Sadek, N. Chamkasem, E. W. Lusas, and B. L. Reid,
Pilot plant studies on extracting cottonseed with methylene chloride, J. Am. Oil Chem. Soc. 63:
647 (1986).
53. G. E. Mann, F. L. Carter, V. L. Frampton, A. B. Watts, and C. Johnson, Evaluation of cotton
seed meals prepared by extraction with acetone-water mixtures, J. Am. Oil Chem. Soc. 39: 8 6
(1962).
54. R. J. Hron, Sr. and M. S. Kuk, Acetone extracted cottonseed meals without catty odors, J. Food
Sci. 54: 1088 (1989).
55. G. R. List and J. P. Friedrich, Processing characteristics and oxidative stability of soybean oil
extracted with supercritical carbon dioxide at 50°C and 8,000 psi, J. Am. Oil Chem. Soc. 62:
82 (1985).
56. J. P. Friedrich, G. R. List, and A. J. Heakin, Petroleum-free extraction of oil from soybeans with
supercritical CO2 , J. Am. Oil Chem. Soc. 59: 288 (1982).
57. J. P. Friedrich and E. H. Pryde, Supercritical CO2 extraction of lipid-bearing materials and charac
teristics of the products, J. Am. Oil Chem. Soc. 61: 223 (1984).
58. G. R. List, J. P. Friedrich, and J. Pominski, Characterization and processing of cottonseed oil
obtained by extraction with supercritical carbon dioxide, J. Am. Oil Chem. Soc. 61: 1847 (1984).
59. G. R. List, J. P. Friedrich, and D. D. Christianson, Properties and processing of com oils obtained
by extraction with supercritical carbon dioxide, J. Am. Oil Chem. Soc. 61: 1849 (1984).
60. W. Zhao, A. Shishikura, K. Fujimoto, K. Arai, and S. Saito, Fractional extraction of rice bran
oil with supercritical carbon dioxide, Agr. Biol. Chem. 51: 1773 (1987).
61. G. B. Kamofsky and R. J. Hansotte, Process for extracting oleaginous seed materials particularly
cottonseed with aqueous alcohols, U.S. Pat. 4,359,417 (to Dravo Corp.) (Nov. 16, 1982).
62. G. Kamofsky, Ethanol and Isopropanol as solvents for full-fat cottonseed extraction. Oil Mill Gaz.
86: 34 (1981).
63. R. J. Hron, Sr. and S. P. Koltun, An aqueous ethanol extraction process for cottonseed oil, J.
Am. Oil Chem. Soc. 61: 1457 (1984).
64. G. Abraham, K. M. Decossas, R. J. Hron, Sr., and M. S. Kuk, Process engineering economic
evaluation of the ethanol extraction of cottonseed: preliminary analysis, J. Am. Oil Chem. Soc.
68: 418 (1991).
65. R. J. Hron, Sr., G. Abraham, M. S. Kuk, and G. S. Fisher, Acidic ethanol extraction of cotton
seed, J. Am. Oil Chem. Soc. 69: 951 (1992).
6 6 . G. Abraham, R. J. Hron, Sr., M. S. Kuk, and P. J. Wan, Water accumulation in the alcohol
extraction of cottonseed, J. Am. Oil Chem. Soc. 70: 207 (1993).
67. E. W. Lusas, L. R. Watkins, S. S. Koseoglu, K. C. Rhee, E. Hernandez, M. N. Riaz, W. H.
Johnson, Jr., and S. C. Doty, New isopropanol system shows promise, INFORM 5: 1245 (1994).
6 8 . E. D. Milligan and J. F. Suriano, System for production of high and low protein dispersibility
index edible extracted soybean flakes, J. Am. Oil Chem. Soc. 51: 158 (1974).
Refining
David A. Allen
Partnership Solutions, Liverpool, England
L INTRODUCTION
The large majority of oils and fats are not fully suitable for human consumption immediately
after extraction from seeds, nuts, fruits, fish, animals, etc. For markets in developed coun
tries, their tastes and odors are generally too strong and rather unpleasant. In addition, the
fats and oils may contain components that make them unsuitable for particular applications,
in terms of either quality or performance. The best known exception to this general statement
is olive oil. Cold pressed and (extra) virgin oils are much prized for their flavor and aroma.
Only the lower grades of olive oil are likely to be fully refined prior to use.
The processes that together are referred to as refining or purification are chosen to effect
removal of a large proportion of the non-triacylglycerol components from oils and fats. Ideally
the antioxidants naturally present in many crude oils and fats would not be removed by the
refining processes. Unfortunately, the process conditions necessary to remove undesirable ma
terials are also effective in removing some proportion of the natural antioxidants and vitamins.
A significant number of undesirable materials may be present in crude oils and fats and
should be either completely or almost completely removed by the refining processes. Table 1
provides a list of these non-triacylglycerol impurities. Note, however, that it is unlikely that
any one crude oil or fat would contain all of the components listed.
In spite of the large number of possible impurities in crude oils and fats, trading specifica
tions are generally focused principally, or even exclusively, on those impurities that are cer
tain to affect yield, i.e., that are commercially based rather than technically sophisticated
specifications. The principal safeguard to the refiner is found in the Good Merchantable Qual
ity (GMQ) or Fair Average Quality (FAQ) clause of the contract. Refiners will usually carry
out more extensive analysis of a crude oil shipment than is required simply to confirm that
the oil meets the commercial specification. Over the years a refiner builds up his own picture
of what is typical and acceptable in terms of non-triacylglycerol content for a range of crude
oils and fats. He may be able to relate certain analytical results to refinery yield performance
or refined oil quality parameters. Thus if, or when, he receives a parcel of oil that does not
137
138 Allen
Table 1 N o n -T ria cy lg ly cero l C om p onents o f Fats and O ils
Free fatty acids

Partial g ly c e r o ls (m on o- and d ia cy lg ly cero ls)
P h o sp h o a cy lg ly cero ls
P igm en ts
S terols
T o co p h ero ls
O xid ation products
T race elem en ts (cop p er, iron, sulfur)
Protein fragm ents
W a x es
H ydrocarbons
M oisture
Dirt
P esticid e resid ues
P o ly c y c lic arom atic hydrocarbons
meet the requirements of GMQ, he is able to recognize this quickly and provide convincing
documentary evidence to support his claim that the oil is not GMQ/FAQ.
II. PROCESSING ALTERNATIVES

As already indicated, the complete refining or purification process comprises a series of dis
crete processing steps. There are two major alternative processing sequences, which are
known in the industry as “chemical” and “physical” refining (Fig. 1). These terms are drawn
from the different methods of free fatty acid removal—either neutralization by caustic soda
(or other alkali) solution, which is a chemical reaction, or distillation, which is a physical
process.
The physical refining route can offer important advantages to the refiner. These are becom
ing ever more significant as concerns for the environment lead to increasingly stringent limits
on the quality, and quantity, of water discharged from refineries. Such limits result from the
fact that the soap produced by alkali neutralization of free fatty acids is usually acidified to
convert the soap to a mixture of free fatty acid and neutral oil (known as acid oil). As a
result of the neutralization and subsequent acidification processes, a considerable quantity of
wastewater is generated. This can be relatively high in fatty matter and suspended solids as
well as containing sodium sulfate in solution. All three parameters (fatty matter, suspended
solids, and sodium sulfate) are likely to be limited by the water authorities with heavy fines
for exceeding prescribed limits.
There is no “neutralization” (refining) step in physical refining (see Fig. 1). Thus the resul
tant need to acidify soapstock and treat wastewater is avoided. Furthermore, the yield loss on
physical refining is lower than that achieved by chemically refining the same oil. Because
neutralization and water washing steps are not required for physical refining, there is a saving
in capital investment. This is substantial because of the high cost of the sophisticated continu
ous separators or centrifuges used for soapstock and washwater separations.
In an ideal world, therefore, a refiner would choose to operate only a physical refining
process sequence. Unfortunately, unless oil quality objectives are compromised, certain oils
are not naturally suited to physical refining. This is principally because of the relatively high
contents of phospholipids and gums. To make more oils suitable for physical refining, certain
major companies have developed processes that can reduce the phosphatide content of some
Refining 139
"Physical" "Chemical"
Crude Oil Storage Crude Oil Storage

I ^ 1 4
Gum Conditioning Intensive Degumming Gum Conditioning
4
Neutralisation
4
Water Washing
4
Drying
4
Bleaching Bleaching
i 4
Filtration Filtration
4 4
Pretreated Oil Storage N&B Storage
4 4
Steam Refining Deodorisation
4 4
Polish Filtration Polish Filtration
4 4
Cooling Cooling
4 4
Edible Oil Storage Edible Oil Storage
Fig. 1 Refining process alternatives.
types of crude oils to a sufficient extent that they will physically refine satisfactorily. These
processes are summarized in Section IV.A. At present the major processes are covered by
patents, but they can be licensed from the patent holders.
III. CHEMICAL REFINING

A. Degumming
Soft or liquid oils are usually degummed with water at the completion of the extraction pro
cesses. As a result, the remaining phosphatide levels are much lower, but the majority of
these phosphatides are nonhydratable, that is, they are not affected by contact with water
alone. The use of a more aggressive medium, such as citric or phosphoric acid, will hydrate
the phosphatides, permitting separation from the oil either at a discrete degumming stage or
together with the soapstock produced at the neutralization stage.
The equipment used to bring about intimate contact between the crude oil itself and a small
quantity of concentrated acid varies according to the type of oil being processed and the level
of phosphatides present. Crude oil fed to the pretreatment process is normally heated to a
temperature in the range 60-85°C. Specialist mixers are used to disperse the small quantity
(0.05-0.2% , based on the weight or throughput of crude oil) of concentrated (85%) phos
phoric acid in the bulk of the crude oil and to provide a short residence time for the hydration
reaction to take place. Mixer types commonly employed include disk and knife mixers. Not
only do they offer different intensities of mixing but they also offer quite different residence
times, varying from a few seconds to 1-2 mins, respectively [1]. Often refineries are equipped
with both disk and knife mixers, which are selected on the basis of the type and quality of oil
being processed.
140 Allen
Hydration of phosphatides is a slow process compared with the rapid reaction of an alkali
solution with free fatty acid to produce soap. Thus additional reaction time is often provided
in the form of an in-line reaction vessel, which may be unagitated or fitted with a gentle low
shear agitator. Typical additional residence time could be 5 -9 mins [1].
If gums are to be recovered separately for use as commercial lecithin, citric acid should be
used in preference to phosphoric acid. The latter tends to char the lecithin, leading to dark-
colored product.
The above descriptions refer to continuously operating plants. Degumming pretreatment
can also be carried out in batch plants fitted with slowly rotating (approximately 30 rpm) gate
agitators. The crude oil is heated to 5 5 -8 5 °C , and the acid solution is added under agitation.
Water ( 0 .5 -2 % ) is then added. After a few hours of agitation, the hydrated gums will have
flocculated. They may then be separated from the oil using a conventional “refining” (disk-
type) centrifuge in preference to gravity settlement, which is both slow and of poorer yield.
The main reason for carrying out the pretreatment step to remove gums, both hydratable
and nonhydratable, is that certain phosphatides, most notably the calcium and magnesium
salts of phosphatidic and lysophatidic acids [2], are strong emulsifying agents. If these phos
phatides were still present in the oil during alkali neutralization and subsequent soapstock
separation, the yield of neutral oil would be adversely affected by emulsification losses.
In addition to hydrating nonhydratable phosphatides, phosphoric acid has the effect of
reducing the chlorophyll content of the oil. This effect can be very important when processing
rapeseed or canola oil produced from unripened or damaged seeds. In this case the dosage of
concentrated phosphoric acid needs to be near or above the top of the usage range quoted
above.
B. Neutralization
The primary objective of the neutralization (or refining) step is the removal of almost all the
free fatty acids present in the crude oil. As already stated, an alkali solution [usually sodium
hydroxide (caustic soda)] reacts readily with free fatty acid to produce soap and water.
Whereas the free fatty acids were fully dissolved in the oil, soap is almost completely oil-
insoluble, making subsequent separation relatively straightforward.
As soon as the oil-phosphatide-phosphoric acid mixture from the degumming stage comes
into contact with caustic soda, any excess phosphoric acid is immediately neutralized, and no
further phosphatide hydration takes place. To allow for neutralization of excess phosphoric
acid, the quantity of caustic soda dosed into the oil stream is increased to compensate for the
total addition of phosphoric acid.
From a production viewpoint, it is important to achieve the desired reduction in free fatty
acid content and the lowest practical level of yield loss. Caustic soda will react with triacyl-
glycerols as well as with free fatty acids. Thus there is inevitably some loss of neutral oil by
saponification. Neutral oil losses can also occur because of emulsification into the water phase
and occlusion of droplets into the soapstock itself.
Appropriate selection of the following process variables will enable the refiner to achieve
the desired reduction in free fatty acid content with an acceptable minimum loss of neutral oil.
1. The type of alkali used— sodium hydroxide (caustic soda), sodium carbonate (soda
ash), sodium silicate, ammonium hydroxide
2. The strength of the alkali solution
3. The excess of alkali solution above the stoichiometric quantity required to neutralize
the free fatty acid and phosphoric acid (calculated from chemical equations)
4. The temperature at which reaction is carried out
Refining 141
5. The type and degree of agitation during and after alkali addition
6. The time between alkali addition and soapstock separation
Caustic soda is the most widely used alkali for refining. It has the effect of thoroughly
cleansing the oil and effecting some decolorization. It also saponifies some neutral oil, espe
cially when used in concentrated solution or large excess. To reduce saponification losses,
sodium carbonate is sometimes used in place of or in combination with caustic soda. Sodium
carbonate is a much milder alkali, resulting in less unwanted saponification but also less
decolorization.
Pardun [3] first reported the possibility of using ammonium hydroxide for refining soybean
oil. He claimed that environmental benefits were achieved by avoiding the need for splitting
the soapstock with mineral acids. Ammonia may be boiled off the soapstock without the need
for acid treatment.
The stronger the alkali solution used, the greater the removal of gums and pigments as
well as free fatty acids, but the higher the losses due to saponification. The soapstock formed
is also thicker, making it more difficult to discharge from the centrifuge and increasing the
risk of neutral oil being lost by occlusion or entrapment in the soap itself. Lower strength
alkali solutions overcome these problems but can lead to increased oil losses due to emulsifi
cation.
The higher the excess of alkali solution used, the greater the saponification loss but the
greater the cleansing effect on the oil. Low levels of excess alkali reduce the saponification
loss but increase the risk of emulsification leading to losses. Choice of alkali type, strength,
and excess is generally based on experience with the type of oil being processed. Minor
adjustments are made depending on how well a particular parcel of oil is processing.
Combining a long contact time with high temperature and vigorous agitation leads to in
creases in saponification losses. This has the advantage of maximizing pigment removal and
improving phase separation.
Modem continuous refineries (see Figs. 2 and 3) are equipped to offer the refiner a range
of mixing time and intensity options and a range of contact times as well as ease of selection
and control of throughput rate and process temperature(s). This gives the refiner the ability to
handle a number of different oil types and a range of oil qualities using the same equipment.
The primary separator is used to separate the soapstock, plus gums and dirt, from neutral
ized oil. Almost invariably new refinery installations are fitted with self-cleaning separators
for this duty. These units are designed to permit the centrifuge bowl to be separated briefly
into two parts, leaving a small space through which viscous sludges are discharged. The very
high centrifugal forces generated at the inner surface of the centrifuge bowl are sufficient to
propel even thick sludges through the self-cleaning space during a “bowl shoot.” The refinery
operator can set the frequency and duration of these bowl shoots, depending on the quality of
the oil being processed and the time since the centrifuge was last cleaned manually.
Over time, sludge material also builds up between the conical “disks,” which are located
over the main rotating shaft in the center of the separator. This buildup restricts the flow of
oil and soapstock across the faces of the disks and vertically between disks. The process
operator will find that normal adjustment of the backpressure control valve is insufficient to
ensure satisfactory separation between the water and oil phases. There is then little alternative
but to stop production and either clean the centrifuge in place (cleaning in place, CIP) or strip
it down and clean each disk manually.
Many continuous refining plants are equipped with three separators. In this case the first
two (following the direction of oil flow) are identical, permitting use as either primary soap
stock separators or washwater separators. When one primary separator is off-line for cleaning.
142 Allen
Separator Separator
Fig. 2 Flow diagram of a modem continuous refining (neutralization) plant. (Courtesy of Westfalia
Separator Limited.)
Refining 143
the refinery can continue operating at a reduced throughput by using the second identical
separator with just one water wash separator. By this means, lost production is minimized,
but quality is maintained.
C. Washing and Drying

Neutral refined oil leaving the primary separator typically contains 500-1000 mg soap/kg oil.
This soap must be removed from the oil before further processing. Hot soft water, at a higher
temperature than that of the oil, is added at a rate equivalent to 10- 20 % of the oil flowrate.
After only brief contact with the oil, soapy water is separated from the oil in the second
centrifugal separator. In many continuous plants, there is further washwater addition followed
by another centrifugal separation. This third separator is used only for washing. It therefore
requires less sophisticated internal fittings than the so-called primary separators, making it
lower in cost. To reduce water consumption, many plants operate countercurrent washing,
whereby fresh water is used for the second wash stage. The dilute soapy water produced is
used for the first wash stage.
After two water washes, the oil should be almost completely soap-free, containing less
than 50 mg/kg. This level may be further reduced if citric or phosphoric acid is added either
to the water used for the final wash or to the oil prior to drying. Such acid treatment eliminates
the soap almost completely. Soap removal is important because during bleaching, soap will
preferentially adsorb onto the bleaching clay, effectively deactivating a portion of the surface
for color pigment adsorption.
After water washing, the refined oil is dried to a moisture content of approximately 0.1%
prior to bleaching or intermediate storage. Drying is typically carried out under a moderate
vacuum (20-50 torr) and at an oil temperature of 90-100°C.
The viscous soapstock from the primary separator is usually mixed with the washwater
from the wash separators prior to acidification with a concentrated mineral acid, most com
monly sulfuric acid. The acid reacts with the soap to convert it back to a mixture of free fatty
acid and neutral oil. This commodity is known as acid oil and is sold as a by-product either
for use in animal feed or for use in the oleo chemical industry. Acid oil is separated from the
water phase, usually by gravity settlement of the water. The water phase, comprising princi
pally sodium sulfate solution and excess sulfuric acid, is adjusted to a neutral pH, using
caustic soda solution, prior to being subjected to effluent treatment to reduce the contents of
suspended solids and separable grease and oil to permitted levels prior to discharge to, for
example, a public effluent system.
D. Refining with Dilute Alkali

An alternative to the processes outlined above that use strong sodium hydroxide solutions in
the range 16-24°Baume (3.1-5.3 Normal), is the Zenith process [4]. A schematic for this
process is shown in Fig. 4. This process sequence involves three semicontinuous steps, al
though a fully continuous version is now available. The P-unit (see Fig. 4) is a vacuum vessel
fabricated from acid-resistant steel and divided into three or four trays, depending on the
desired reaction time and the required oil throughput. Concentrated phosphoric acid is auto
matically dosed from the holding vessel on the top of the P-unit. The top tray provides steam
heating to ensure that the oil is deaerated and held at optimum reaction temperature. All trays
are provided with agitators mounted on a common drive shaft. Valves at the bottom of each
tray operate to an automatic cycle. The quantity of phosphoric acid added ranges from 0.03
to 0.1% of oil weight for low phosphatide oils such as coconut oil, palmkemel oil, tallow,
lard, and partially refined hydrogenated oils. Between 0.1 and 0.2% of phosphoric acid is
144 Allen
mÂCHING EARTH
C-ACO
(IRi p-uNrr
NEtJTRALIZER rtni
lI j 1 t; :î lxd ca xd
1 b
cX bm
1 ^
.... " i: -J
„ J
STEAM C j
a .J = = = = LJ
Fig. 4 Flow diagram of the Zenith dilute alkali refining process and associated bleaching plant. (Cour
tesy of Campro International Inc.)
required for the liquid seed oils. Up to 0.4% may be required for oils containing more than
200 mg/kg of phosphorus or high chlorophyll levels.
After the P-unit, the oil is centrifuged to separate the dirt and denatured phosphatides, in
the form of a sludge. The oil then enters the bottom of the neutralizer vessel, in which it is
transformed into a series of very fine droplets (approx 10"^ m “ ^), which rise upward through
the dilute alkali solution filling the neutralizer. During their upward passage, the free fatty
acids in the oil droplets are neutralized by the alkali and the resulting soap dissolves in the
alkali solution. Any water-soluble impurities also transfer into the aqueous solution.
Neutralized oil accumulates on the top surface of the alkali solution before passing continu
ously to the C-unit. The strength of the alkali solution is initially approximately 2°Baumé
(0.35 Normal), except for coconut oil refining, when the solution is stronger, 4.6°Baumé (0.8
Normal). As oil is refined, the alkali solution becomes weaker. When it reaches 0.1 Normal
it is replaced. The cleaning of the oil prior to neutralization coupled with the use of very
dilute alkali solutions results in relatively very low saponification of triacylglycerols. Thus,
yields of refined oil are close to theoretical.
In the C-unit, or bleacher, the refined oil is first treated with citric acid solution to split
any residual soap and to inactivate any oxidation catalysts. Bleaching clay is then added to
decolorize the oil. The quantity of bleaching clay required is likely to be approximately the
same as that required for bleaching refined, washed oil from strong alkali treatment.
Finally, bleaching clay is separated from the oil on a pressure leaf filter, followed by a
polish or guard filter to ensure that only clean oil goes into storage or for further processing.
In the Zenith process there is no need to water wash the neutralized oil because soap levels
are already low (50-100 mg/kg). Spent alkali-soapstock solution is acidulated with mineral
acid to produce an “acid oil” by-product. Because weak alkali is used, there is little decolor-
ization of the oil.
Refining 145
E. Batch Refining
Batch refining is carried out in open or, preferably, closed tanks. These are usually vertical
cylindrical vessels fitted with conical bottoms. The conical shape assists in removing only the
aqueous phase at the completion of the various process steps.
Batch refining kettles are fitted with variable-speed agitators and steam/cooling coils for
temperature adjustment. Capacities can vary greatly but are generally in the range of 5 -
25 tonnes.
At the outset of processing, oil is charged to the vessel and heated to the target temperature
for refining, 80-95°C. Caustic soda solution of the chosen strength is sprinkled slowly onto
the surface of the crude oil, which is subjected to gentle agitation. Soapstock begins to form
almost immediately. After a short reaction period following the completion of caustic soda
addition, the agitator is stopped, and the heavier aqueous phase, including the soap, is allowed
to settle under gravity. The settling rate will be affected by factors such as the size of the
soapstock particles, the size of oil and water droplets, and the viscosities of the oil and
aqueous phases. Where relatively thick soapstocks are formed, e.g., from saturated fats and
from cottonseed and sunflower oils, the use of a dilute caustic soda is preferred in order to
reduce soapstock viscosity.
When the aqueous phase has fully settled, it is run off from a valve at the lowest point of
the conical bottom of the tank. It is common practice to have a sight glass between the valve
outlet and the inlet to the discharge pipe. If this is properly illuminated, the operator can
readily determine when discharge of the aqueous phase is complete and the oil phase is start
ing to discharge.
The hot neutralized oil is washed two to four times with hot water to ensure that soapstock
has been removed insofar as possible. Note, however, that residual soapstock levels from
batch refining are significantly higher than those achieved by continuous centrifugal separation
of washwater from neutral oil. This residual soap will be adsorbed onto the clay used in the
bleaching step that usually follows refining.
It is common practice to use the same vessel for bleaching as for neutralization and water
washing. If this is done, a vacuum pressure is produced within the vessel and the oil tempera
ture is raised to dry the oil to an acceptably low moisture level prior to bleaching.
Occasionally, problems of either slow settling of soapstock or emulsion formation are en
countered in batch refining. These can often be overcome by using a dilute solution of com
mon salt for the first water wash.
Batch refining is a relatively slow process. It may be possible to refine only one batch of
oil during a production shift. This is in marked contrast with the production rates of 10-15 t/
h achieved in many continuous refining plants.
F. Bleaching
The process is known as bleaching because when first introduced its primary purpose was
seen as the removal of color pigments from the oil. More recently it has been concluded that
bleaching clays also remove or at least reduce the contents of other impurities such as soap,
trace metals, phosphatides, and sulfur compounds. Primary oxidation products, hydroperox
ides, are broken down on the surface of the bleaching clay, and secondary oxidation products,
carbonyl compounds, are adsorbed onto the surface. Decolorization still occurs during bleach
ing, but its effect is often less significant than that which takes place at the higher temperatures
of deodorization. Polycyclic aromatic hydrocarbons containing five or more benzene rings in
146 Allen
their structure are removed during bleaching only if activated carbon is used in addition to
bleaching clay.
Bleaching clays are naturally occurring bentonite and montmorillonite clays. Usually these
materials are activated using a mineral acid such as sulfuric acid. The acid removes some
minerals from the clay, resulting in larger volume micropores and smaller particle size [5].
The principal benefit of acid activation is that it permits reduced usage of clay to meet a
particular bleached oil specification. A slight negative is that acid-activated clay retains more
oil per unit weight than natural clay. The lower usage level, however, more than offsets this
effect in most cases. Mag [6] reviews the theory of adsorption bleaching, including the mecha
nisms of acid activation of bleaching clay.
As mentioned above, activated carbon can also be used for bleaching. It is generally more
expensive than activated bleaching clay. Activated carbon retains more oil than bleaching
clay. Its very fine particle size makes it difficult to handle and to filter. Certain grades of
activated carbon are used specifically to remove the so-called heavy polycyclic aromatic hy
drocarbons, which will not be stripped from the oil during subsequent deodorization. Prefera
bly, the oil is treated with bleaching clay before the activated carbon is added. This avoids
using the active sites on the carbon to adsorb the easy-to-remove components like soap and
color pigments. It has been stated that a combination of 10-20 parts of bleaching clay to one
of carbon can be more effective than bleaching clay alone [7]. Where removal of polycyclic
aromatic hydrocarbons (PAH) is a priority, a much higher proportion of activated carbon will
be required. Coconut oil, some fish oils, and some rapeseed oils may require activated carbon
treatment to reduce the content of heavy PAH [8].
Bleaching clays are expected to have good adsorption properties for the components to be
removed from the oil, not too high a retention of oil, and good filterability. The latter factor
is important in ensuring that clear, particle-free oil is produced soon after a clean filter is
brought on-line and oil flowrates do not diminish rapidly as filter cake builds up on the
filter medium.
The precise mechanism of the adsorption process is not yet fully understood [9]. Bleaching
is carried out at elevated temperature (90-120°C), almost always under a vacuum pressure to
exclude oxygen. The use of vacuum pressure tends to remove some of the residual moisture
from the oil, following the drying stage at the completion of neutralization and water washing.
It also removes some “free” water from the bleaching clay itself. This may not be desirable
because experience has shown that bleaching clay effectiveness can sometimes be enhanced
by adding small amounts of moisture to the oil prior to adding bleaching clay [10].
Temperature and time are the major process parameters affecting the bleaching achieved
with a particular dosage of an activated bleaching clay. Bleaching times of 10-30 min are
generally required to ensure that adsorption reaches or approaches equilibrium. Shorter times
may be sufficient in certain circumstances, but commercial equipment is usually sized to
provide residence times in this range at the design throughput rates for oil.
A small rise in FFA may be observed during bleaching. This is likely to be due to a
combination of mineral acid leaching out from the activated bleaching clay and some hydroly
sis of the oil.
At the completion of bleaching, the peroxide value of the oil is generally zero or very low,
unless levels of bleaching clay below about 0.5% have been used or the unbleached oil had a
peroxide value above about 10 meq/kg. Anisidine values are not reduced to the same extent
during bleaching unless uneconomic quantities of bleaching clay are employed to achieve this
particular result.
Many refineries have noted what is generally known as the press effect. This results in a
slightly lower color in the bleached oil after filtration than immediately prior to filtration. If
Refining 147
oil is truly in equilibrium with the bleaching clay prior to filtration, then there will be no press
effect [12]. The existence of a press effect in many operating situations indicates that oil and
clay are not in equilibrium at the end of the bleaching period. Some refineries are able to use
the press effect by shutting off the flow of clay to the bleacher when the filter is almost full
and then running for a period of time, during which only the press effect reduces the oil color.
1. Equipm ent fo r Bleaching

As with refining and water washing, bleaching may be carried out as a batch process. It is
more usual, however, to have a semicontinuous or fully continuous bleaching plant. Often
this takes oil directly from the outlet of the dryer, which marks the completion of the water
washing process.
Batch bleaching can be carried out in the same vessel as neutralization and washing. It can
also be carried out in a separate but similar vessel. As mentioned above, bleaching usually
takes place under vacuum. Batch bleaching has the advantage of providing easy segregation
between different oils in a multi-oil refinery. It also has the advantage of a relatively low
capital cost. Its major disadvantages are the range of bleaching times that result from a long
filtration time and the fact that the bleacher often has to be returned to atmospheric pressure
before filtration can proceed, thus exposing the oil to oxygen.
Semicontinuous and fully continuous bleaching processes are carried out in virtually identi
cal equipment. The major difference lies in the stepwise sequencing of oil through a series of
compartments in the semicontinuous process compared with continuous flow along a predeter
mined path in the continuous process. With semicontinuous bleaching, neutralized oil is
pumped through a heat exchanger to a deaeration unit and into the top tray within the main
bleacher vessel. Here the temperature is further increased to the desired bleaching tempera
ture. Bleaching clay and/or activated carbon is added directly to the agitated oil in the top
tray. All the trays within the bleacher vessel are held under the same vacuum pressure (25-
50 torr absolute pressure).
After a predetermined time interval, the oil is discharged by gravity into the tray below,
and so on until it reaches the bottom of the bleacher. This region acts as a feed tank for
the filters.
Because of the segregated tray arrangement and careful sequencing of the drop valves,
changes from one oil to another can be made in this plant with only traces of cross contami
nation.
In many plants, bleaching clay is added directly to the oil. This is generally preferred,
because Brimberg [10] showed that a bleaching effect is then achieved more rapidly than with
alternative means of clay addition. Alternatives include dispersing the clay in a small side
stream of oil before adding the resultant slurry back to the main stream and making a slurry of
bleaching clay in oil that has already been bleached, then adding that to the main oil stream.
Currently, the driving force behind the introduction of new and improved bleaching clays
and bleaching processes is the desire to reduce the amount of absorbent required for effective
bleaching. Silica hydrogels (see Section III.G) are one example of new bleaching media. One
example of a process designed to achieve a reduction in bleaching clay usage is shown in Fig.
5. This provides an alternative to the usual agitated vessel (often divided into a number of
sequential compartments) that provides the contact time between bleaching clay and oil neces
sary for effective bleaching. In the design shown, bleaching clay is dosed into neutralized
refined oil in a small agitated tank. The resultant slurry is then drawn into a tall slim vessel
within which it is initially heated to bleaching temperature while being agitated with steam.
Heated slurry overflows to fall through a vacuum space to the main residence time volume.
148 Allen
Vacuum
^ Oil Bleaching + Clay

Steam ►
Condensate
S parge Steom
SPARBLEACH
S porge Steam
▼ Oil + Bleaching Clay
Fig. 5 Diagram of the reactor for the Sparbleach process. (Courtesy of Extraction de Smet NV/SA.)
where it is again steam sparged until it flows from the vessel for filtration. As well as a saving
in bleaching clay consumption, savings in maintenance costs are also claimed because there
is no mechanical agitation.
2. Filtration o f Bleaching Clay

At the completion of bleaching, spent clay is filtered from the oil stream. A number of differ
ent filter designs are in use. These include plate-and-frame filter presses, both manual and
with fully automatic cleaning; chamber presses, which may be fitted with diaphragms to per
mit oil to be squeezed out of the filter cake; and filters that use a woven stainless steel mesh
as the filter medium. These filter “leaves” may be mounted horizontally in a vertical cylindri
cal tank and cleaned by spinning the central mounting shaft. Alternatively, they may be
mounted in a vertical or horizontal cylindrical tank and cleaned by vibration of the leaves
themselves. These filters may discharge spent filter cake either internally through a dump
valve or externally into a receiving tray or screw conveyor.
All these filters are capable of fully automatic cleaning, which may be started after a set
time period, at a preset differential pressure, or directly by a plant operator. Each filter type
requires the oil to be recycled for several minutes after a clean filter is brought into operation
to allow sufficient filter cake to build up on the filter mesh for the filtrate to be sparkling
clear. It is common practice to position a “guard” filter immediately downstream from the
main filter to trap any particles that have remained in the oil. This is often a bag filter with
relatively little holding capacity for bleaching clay. Thus it blocks quickly in the event of a
significant malfunction of the main filter, which serves to alert plant operators to the need to
take appropriate action.
During the cleaning sequence, the filter cake is blown with steam or, preferably, nitrogen
to reduce the quantity of oil remaining in the filter cake and therefore lost to the process.
Typically, the oil content of spent filter cake is around 2 0 -2 5 % , but it may be as high as
40% depending on the characteristics of the clay, the blowing medium, and the type of filter.
Refining 149
Oil recovered by blowing the filter cake is of a poorer quality than the remainder of the
production. It should be returned to crude or rebleached and not simply mixed with the fil
tered oil.
Spent bleaching clay can present a fire hazard, especially where high iodine value oils have
been processed. Spontaneous combustion usually starts in a smoldering pocket. Plant opera
tors need to be alert for early signs of trouble and be prepared to take immediate action.
3. Spent B leaching Clay

Spent clay is a waste product. Fahn and Zschau [12] present an overview of possible disposal
methods. More recently there have been attempts to regenerate bleaching clays for reuse [13],
but commercial processes seem to be several years away. Some work has been reported [14]
on regenerating silica-alumina mixtures. Early work suggests that regeneration may not be
possible if these materials are contaminated with phosphatides or soap. This is counter to a
major function of these materials. Thus any progress toward successful regenerability is likely
to be slow.
4. B leaching P lant O peration

Most bleaching processes are controlled on the basis of the Lovibond red color of the
bleached, filtered oil. For certain applications, e.g., soy oil for bottling, chlorophyll content
is also measured and bleaching clay addition is increased or decreased as appropriate. Most
plants operate with a standard time for filters to remain on-line prior to cleaning. Alternatives
are to use the differential pressure across the filter unit to indicate that the filter is ready for
cleaning or to monitor filtration rate, which decreases as the filter cake thickness increases.
Whichever method is adopted, it is crucially important to clean a filter before filter cake
thicknesses are such that there is “bridging” between adjacent filter leaves. If bridging occurs,
the filter will not discharge the filter cake correctly when subjected to the normal cleaning
sequence. It will be necessary to dismantle the filter elements and clean them manually,
entailing additional cost and downtime.
Care must be taken when withdrawing filter leaves for cleaning and when replacing clean
leaves. The woven stainless steel mesh that provides the filter medium is relatively easily
damaged. Obviously, a tear in a filter surface will result in bleaching clay passing through the
filter and rapidly filling the guard filter.
Although the filters described are cleaned automatically in place, a proportion of the
bleaching clay is driven into the apertures of the mesh and is not removed by vibration or
spinning. When a number of the filter elements are fouled in this way, they require additional
cleaning. This is often undertaken away from the filter body either by treatment with hot
caustic soda in a specialist cleaning tank or by jet washing with hot water.
Because of the need for periodic extra cleaning of filter elements, a spare set or half-set is
maintained in a clean condition for use when elements are removed for external cleaning.
This provision keeps the downtime of the filter unit to an absolute minimum.
G. Silica Refining
As mentioned in the previous section, silica hydrogel may be used for bleaching. It was
initially introduced to oils and fats processing as an alternative to bleaching clay. Silica hy
drogel has a much higher capacity than bleaching clays for soaps, phosphatides, and, in
particular, trace metals [15-19]. Unfortunately, it does not have any significant capacity for
color pigments including chlorophyll. Thus, when used as a bleaching medium, silica hy
drogel is normally used in combination with bleaching clay.
150 Allen
Preferably the silica hydrogel is added to the refined and washed oil several minutes before
the bleaching clay is added. This staged addition maximizes the benefits of silica hydrogel by
adsorbing all the soap and most of the phosphatides present, leaving the subsequently added
bleaching clay to adsorb color pigments, trace metals, and oxidation products. When using
this two-step addition, savings on total bleaching media have been achieved [20]. The use of
a combination of Trisyl (W. R. Grace) silica hydrogel with a bleaching clay was found to
improve phosphatide removal and color stability in physically refined palm oil [21].
A reduction in the amount of solid adsorbent used increases the time “bleaching” filters
remain on-line and reduces the quantity of neutral oil lost with the adsorbent. It also reduces
the amount of solid waste that must be disposed of. Financial savings are not necessarily
achieved, however, because silica hydrogel is typically around three times as expensive as
bleaching clay.
Because of the high adsorbency of soap onto silica hydrogels, a modification has been
proposed to the traditional refining and water washing process [22]. In the modified process,
residual soap in oil leaving the primary separator is adsorbed onto silica hydrogel instead of
being washed out by hot water. The main advantage of this process is the significant reduction
in the quantity of water requiring treatment in the effluent plant. It is not known to what
extent this process has been adopted in commercial practice.
Somewhat surprisingly it has been found that rather than soap competing with phosphatides
for active sites on the silica hydrogel, soap actually enhances the adsorptive capacity for
phosphatides. Combined soap and phosphatide capacities in excess of 100% on a weight basis
can be achieved [25].
A comparable process using silica hydrogel has been developed for use in physical refining
[26]. This involves adding a very small quantity of caustic soda solution with the silica hy
drogel (e.g., 0.5% of 18°Baume/3.6 Normal caustic) to produce some soap to enhance phos
pholipid removal prior to addition of bleaching clay as normal (see Section III.F.l).
A silica-based process has also been developed for acid refining high phosphorus oils such
as canola and soybean oils. The process [27,28] involves adding a degumming acid such as
citric acid. This increases the phosphatide adsorption capacity of silica hydrogel to about 65%
by weight (dry basis) at a residual phosphorus content of 5 mg/kg in the oil [29]. This capacity
is used to calculate the amount of silica hydrogel required to reduce the phosphorus content
of an oil of known phosphorus content to below 5 mg P/kg oil. For oils containing more than
200 mg/kg of phosphorus, the quantity of silica hydrogel required is impractically high. An
improved process [30] involves mixing citric acid (as a 50% solution) into heated oil, then,
after a suitable holding period, adding sufficient sodium silicate (10% solution) to neutralize
about 70% of the citric acid introduced originally. This limits the level of soap produced
while ensuring good hydration of the phosphatides. Silica hydrogel is then added at approxi
mately 80°C (100°C max) to adsorb the gums and the soaps. Thereafter, the oil may, if
desired, be bleached with clay prior to filtration to remove the silica and the bleaching clay.
Laboratory work has shown that this process can be satisfactorily applied to oils containing
levels of phosphatides up to 400 mg P/kg oil. As with the other silica refining processes
reported in the literature, it is not known whether, or to what extent, this process is being
used in commercial practice.
H. Deodorization
Almost invariably, deodorization is the final processing stage in the refining sequence. After
deodorization the oil is fully fit for human consumption and is ready for shipment in bulk to
food manufacturers or for conversion into a range of oil and fat products.
Refining 151
The purpose of deodorization is to remove from the oil those trace components that give
the oil its characteristic, and generally unpleasant, taste and odor. Because of the process
conditions that have to be applied to achieve this objective, deodorization also effects a reduc
tion in free fatty acid and decolorization of some pigments (mostly carotenoids) by thermal
degradation; removes portions of the natural tocopherol and tocotrienol antioxidants and other
unsaponifiable components such as sterols, hydrocarbons, and alcohols; and almost com
pletely removes any pesticide residues, polycyclic aromatic hydrocarbons (only those con
taining four or less benzene rings in their structure), final traces of any extraction solvent, and
any volatile sulfur compounds [31,32].
Deodorization is, in effect, a steam distillation process that is carried out at high tempera
ture (200-260°C) and low pressure (2-7 torr; 2.5-9.2 mbar). Theoretical treatments of batch
deodorization have been developed by Norris [33] (subsequently modified to include an activ
ity coefficient A [34] to reflect the deviation of fatty acid solutions from “ideal”):
FO
eK a
In
(8 ( 1)
or
/SPA
In (2 )
[ v ,/ PO ^ \P O /
where S = moles of stripping steam, O = moles of oil, P = total system pressure, =
vapor pressure of free fatty acid, Vj = initial number of moles of free fatty acid in the oil,
V2 = the final number of moles of free fatty acid in the oil, E = vaporization efficiency, A
= an activity coefficient, and K = EA, an experimental constant.
Gavin [35] developed a similar equation for continuous deodorization:
n p^s
(3)
where Vj, V2 = moles of free fatty acid in the oil entering and leaving the deodorizer,
respectively, S = moles of stripping steam per hour, O = moles of oil per hour, and K, P,
and P^ are as specified above for batch deodorization.
These equations show that the main process variables influencing free fatty acid removal
are the absolute pressure of the system, the amount of stripping steam used in relation to the
amount of oil, and the vapor pressures of the free fatty acids and other volatile components.
These are heavily influenced by the oil temperature.
The activity coefficient is determined experimentally. It is influenced by free fatty acid
type and concentration. Alternatively, it can be calculated from an equation developed by
Wilson [36]. Stage et al. [37] used this equation to produce a curve for mixtures containing
up to 5% stearic acid in tristearin and showed that the activity coefficient A lay between 0.63
and 0.65.
The vaporization efficiency, E, reflects the effectiveness of steam-oil mixing in transfer
ring free fatty acid and other volatile components into the steam. For a modem continuous or
semicontinuous plant, E is approximately 1.0. For batch deodorizers the vaporization effi
ciency is much less, probably closer to 0.5.
1. Influence o f Process Variables

From Eq. (1), it is clear that the quantity of stripping steam required to remove a specified
quantity of volatile components ( e .g ., free fatty acid) is directly proportional to the total
152 Allen
system pressure P. Thus the lower the system pressure, the smaller the quantity of stripping
steam required. There are, however, limits to the vacuum pressure that can be achieved eco
nomically. Thus commercial deodorizers usually operate in the range 2 -5 ton* (2.5-6.5 mbar)
to avoid excessive steam consumption in ejector systems.
It is the volume of the stripping steam rather than the mass that is important in stripping
out volatile components. A combination of high temperature with a low system pressure en
sures expansion of the stripping steam supplied once it enters the deodorizer itself.
Temperature is crucially important. The higher the temperature, the higher the vapor pres
sures of the volatile components that are to be removed. There are practical limits to the
temperatures used in deodorizers. Production of high temperatures is expensive, requiring the
use of high pressure steam or thermal fluids. Very high temperatures [in excess of 250°C
(485°F)] are undesirable because they increase the conversion of cis to trans isomers in highly
unsaturated liquid oils [38-41]. For this reason, deodorization temperatures for liquid oils are
being limited to 24 5°C maximum.
Deodorization effectiveness is also time-dependent, although this is not obvious from the
above equations. This is principally because it takes time to expose each “element” of the oil
to the required quantity of stripping steam for the prevailing oil temperature and system pres
sure. The move from batch to semicontinuous or fully continuous deodorization has resulted
in a reduction in deodorization times. This has been due, in part, to improved systems for
contacting the oil with the stripping steam that, among other features, have permitted the use
of higher steam velocities without problems resulting from excessive frothing or unacceptably
high oil losses because of oil droplet entrainment into vapor streams.
2. The D eodorization Process

A commercial semicontinuous or fully continuous deodorization process may be viewed as a
sequence of several steps: deaeration, preheating, high temperature heating, steam stripping
(deodorization), heat recovery, chelation, final cooling, and polish filtration. For a batch de
odorizer, the preheating and heat recovery steps do not apply. Also, all the steps, except
polish filtration, are carried out within the deodorizer vessel itself. With semicontinuous and
continuous deodorizers, several steps may be carried out externally to the deodorizer.
Deaeration. Because deodorization is carried out at relatively very high temperature, it

is essential that the oil to be deodorized be almost completely free of dissolved oxygen.
Failure to remove dissolved oxygen will lead to oxidation under the high temperature condi
tions in the process.
In batch deodorizers, where the oil is normally charged to the vessel at a moderate tempera
ture, say 40-80°C, deaeration takes place naturally as the oil is heated by steam coils within
the vessel because the oil is held under a low absolute pressure. Heating rates in batch deodor
izers are relatively low because there are practical limitations to the surface areas of the
heating coils through which high pressure steam is passed. Thus batch deodorized oil is likely
to be completely deaerated under vacuum before the oil temperature exceeds 100-120°C
(2 10 -2 5 0 T ).
With semicontinuous or fully continuous deodorizers, deaeration has to be provided as a
discrete process step. Undeodorized oil is heated to approximately 100°C and then passed into
a deaeration vessel operating at a moderate vacuum pressure. The deaerator may draw its
vacuum from the main deodorizer plant or, more likely, from an independent vacuum source
such as a liquid ring pump.
Preheating. To maximize heat recovery, deaerated oil is further heated by hot deodorized
oil either directly or indirectly and either inside or external to the deodorizer itself. Effective
Refining 153
use of heat interchange significantly reduces the heat that must be added via high tempera
ture heating.
High Temperature Heating. As mentioned already, high pressure steam (70-80 bar) or a
thermal heating fluid (such as Thermex, Dowtherm A, Santotherm, Therminol, Syntrell) is
used to raise the undeodorized oil to the desired temperature.
In recent years the European refiners trade association (FEDIOL) has implemented a volun
tary policy for members that all new installations will use only high pressure steam and that
the use of thermal oils in older installations will be phased out. These decisions were based
on concerns that thermal heating oil degraded by continuous use may pass undetected into the
food product stream. Although such leaks are unlikely and the most commonly used diphe
nyl-diphenyl oxide eutectic mixtures are easily detected in taste testing of deodorized oils, a
possibility of food contamination nevertheless exists.
Rossell [42,43] reviewed the safety of a range of thermal heating fluids and concluded that
although steam remains the preferred high temperature heating medium, the best alternatives
are Syltherm 800, which is a dimethylsiloxane polymer, and Dowtherm A and Thermex,
which are diphenyl-diphenyl oxide eutectic mixtures. Based on toxicity, ease of detection,
and chemical composition after extended use, mineral oils, synthetic organic fluids, and hy
drogenated terphenyls are not recommended.
The attraction of thermal fluid systems from an engineering viewpoint is that they operate
at low pressures relative to the high steam pressures required for high temperature heating.
Because of thermal degradation, diphenyl-diphenyl oxide mixtures should be regularly re
turned to the supplier for reprocessing.
Whichever heat transfer medium is used, high temperature heating may be carried out
inside or external to the deodorizer itself. If it is carried out within the deodorizer, it is
important that the oil be sparged with stripping steam to avoid local overheating. External
heating is usually carried out in heat exchangers, where the oil is in motion and localized
overheating is avoided.
In batch deodorizers, it is usually impractical to achieve any heat interchange between
deodorized and undeodorized oil. All the heating and cooling is carried out within the deodor
izer, using high pressure (25-30 bar) steam and then cold water.
Heat Recovery. The heat recovery step balances the preheating. Heat is interchanged with
deaerated but undeodorized oil either directly or indirectly. It is desirable to cool the deodor
ized oil as rapidly as practical to maintain the highest oil quality. If heat interchange takes
place within the deodorization vessel, the oil should be sparged with steam to provide agita
tion and to maintain the oil’s freshness.
In modem semicontinuous and continuous deodorizers, heat recovery can save between 50
and 85% of the heating costs. Interchange can be achieved in the following ways.
1. Inside the deodorizer
a. Direct interchange between undeodorized and deodorized oils
b. Indirect interchange using a heat exchange medium in a closed circuit, probably
operating on the thermosiphon principle
2. External to the deodorizer
a. Direct interchange between undeodorized and deodorized oil in an interchanger,
e.g., of the spiral design
3. Inside/extemal
a. Indirect interchange between deodorized oil in the deodorizer and undeodorized
oil in a vessel that is external to the deodorizer
154 Allen
Chelation. Once the oil is cooled below 120°C, it is almost universal practice to dose it
with a low level of citric acid (usually 50-100 mg/kg). The acid is preferably added as a 30-
50% solution in water to minimize the risk of blocking transfer lines and nozzles with crystal
line citric acid.
The function of the citric acid is to chelate trace metals present in the oil so that the
chelates may be removed by subsequent polish filtration. Even at 50 mg/kg, citric acid is
present in excess. Thus a low level of citric acid remains dispersed in the deodorized oil.
Experience has shown that this improves the oil’s resistance to oxidation, particularly if this
is measured by an accelerated method such as the one employing the Rancimat instrument.
Cooling. Deodorized oil is finally cooled to an appropriate temperature for storage, ship
ment to customers, or blending. Cold water is usually used to achieve the desired tempera
tures. It is common practice to cool liquid oils to 30-35°C and hard fats or hydrogenated oils
to approximately 10 °C above their melting points. If liquid oils are to be used for bottling it
will be necessary to cool them further to below 25°C. This may be done after deodorization
or prior to bottling. Depending on the cold water temperature and the heat exchanger design,
it may be necessary to use chilled water when ambient temperatures are high (above 30°C).
Polish Filtration. Metal chelates, particles of polymerized oil, and/or traces of bleaching
clay should be removed from the cooled, deodorized oil prior to storage and shipment or
further processing. This is typically achieved by the use of bag or cartridge filters. In a
majority of plants, duplex filter units are employed so that filter elements can be changed
without interruption to the oil flow.
The grade of filter element varies between refineries but is generally in the range 10-15
/xm. Indication that a filter element needs to be changed may be by differential pressure
measurement, reduction in oil throughput rate, or operator judgment.
Many deodorization plant laboratories run standardized filter tests to check the effectiveness
of polish filtration. Typically these involve passing a fixed quantity (say, 1 lb) of oil at a
specified fixed temperature (say, 50°C) through a specific grade of filter medium (e.g., 5-10
^m nylon sediment test disks) in a laboratory vacuum filter unit operating at a fixed vacuum
pressure. When all the oil has been filtered, the filter paper is compared with preprepared
standard papers to indicate whether the filter grade is acceptable, borderline, or unacceptable.
Similar polish filtration, with or without filter grade assessment, is usually repeated on the
oil or blend immediately prior to tank car or railcar filling and shipment from the refinery.
This represents an additional safeguard to ensure that only clear bright oil is sent to customers
or for further processing.
3. D eodorization Equipm ent

Semicontinuous deodorizers are probably the most widely used types at present. Fully contin
uous deodorizers are better suited to refineries processing a single type of oil or a small
number of similar or compatible oils. Increasingly, batch deodorizers are being reserved for
small capacity plants or for low throughput oils, which are often of high relative value and
incompatible with any high throughput oils processed at the same facility.
The term “semicontinuous” indicates that the oil progresses through the deodorization pro
cess sequence, outlined above, as a series of small discrete batches by automatic sequencing.
Thus oil flow into and out of the plant is discontinuous. “Continuous” or “fully continuous”
indicates that oil flows at a steady rate through each of the steps in the process sequence and
discharges at the same steady rate from the plant.
A minority of deodorizer designs combine semicontinuous and continuous features. Oil
may flow continuously through the deaerating, preheating, and final heating stages to be
Refining 155
treated as a batch in the steam stripping or deodorization stage and then be returned to continu
ous flow through heat recovery, chelation, final cooling, and polish filtration. Alternatively,
the above may be reversed such that only the steam stripping stage is fully continuous.
Semicontinuous Deodorizers. Semicontinuous deodorizers normally consist of a series of
trays arranged vertically in a tall cylindrical vacuum vessel. Each tray represents a process
stage in the sequence, although it is common practice to use two or three trays for the steam
stripping or deodorization stage.
Figure 6 shows a process diagram for a typical semicontinuous deodorizer. Undeodorized
oil is passed through one of two bag filters to a measuring and deaeration tank mounted above
the deodorizer vessel. This tank is used to control the quantity of oil that is passed as a small
discrete batch through the process. As the automatic sequencing valve opens, the measured
quantity is drawn into the top tray within the deodorizer shell. Here it is preheated indirectly
by hot deodorized oil lower down the deodorizer, by means of a thermosiphon arrangement.
After a predetermined interval, the heated oil passes into the second tray, where it is heated
to deodorization temperature by high pressure steam or a thermal fluid. The next two trays
down the column shown are deodorization trays, in which the oil is sparged with steam using
mammoth pumps or similar arrangements to increase stripping efficiency.
Following the deodorization tray, the hot deodorized oil first gives up heat, via the thermo
siphon, to incoming undeodorized oil and then to high pressure demineralized water, which is
subsequently flashed to produce stripping steam. Citric acid is dosed into oil in the final tray.
Cool oil leaving the bottom of the deodorizer column is water-cooled and then polish-
filtered prior to storage. It is becoming more common, as shown in the figure, to use nitrogen
injection to return the oil to atmospheric pressure, in preference to allowing it to dissolve air.
The measuring tank, every tray in the deodorizer, and the pool of oil at the base of the
deodorizer are all continuously sparged with steam to provide both agitation and some steam
stripping.
Fully Continuous Deodorizers. Figure 7 shows one design for a fully continuous deodor
izer. In this example, cool undeodorized oil is pumped through a bag filter to a deaeration
tank mounted vertically above the main deodorizer column. The oil is sprayed into the tank,
which is held under vacuum. The oil is sparged with steam and is preheated by steam coils
within the tank. Heated, deaerated oil passes to the top tray of the deodorizer, where it is
further heated by a thermosiphon arrangement from the penultimate tray lower down the
deodorizer column. Note that all the trays within the deodorizer and the pool of oil at the
base of the column are sparged with stripping steam.
The second tray from the top of the column is the main heating tray, where the oil tempera
ture is raised to deodorization temperature by high pressure steam generated locally to the
deodorizer in a specialist boiler. Below the main heating tray are two steam stripping or
deodorization trays. The design of the steam-oil contact arrangements in these trays varies
from one equipment supplier to another, each having its preferred means of enhancing strip
ping efficiency.
Hot deodorized oil interchanges heat with the incoming deaerated oil and is partially cooled
by this process. Further cooling is effected in the bottom tray, in which high pressure water
flowing through the cooling coils is partially converted to steam, which is subsequently used
as sparge or stripping steam. Citric acid solution is dosed into the oil in this bottom tray,
water being flashed off in the low vacuum pressure environment.
Deodorized, precooled oil is pumped through a cold water cooled heat exchanger and then
via polish filters to storage. Pressure is brought back to atmospheric by introducing high purity
nitrogen into the oil stream prior to polish filtration. This has the important added benefit of
01
m
3
Fig. 6 Flow diagram of a semicontinuous deodorizer. (Courtesy of Extraction de Smet NV/SA.)
Refining 157
Fig. 7 Flow diagram of a fully continuous deodorizer. (Courtesy of Extraction de Smet NV/SA.)
ensuring that the oil dissolves nitrogen and not air prior to storage. If the oil is stored in
nitrogen-blanketed tanks, then any oxidation during storage should be negligible.
Deodorizer Designs. As indicated, the vertically stacked tray design is the most common
for both semicontinuous and fully continuous deodorizers. There are a number of different
tray designs, providing alternative means of achieving good stripping steam efficiencies.
There are also various designs for stacked trays. Again their objectives are to enhance strip
ping efficiency by optimizing factors such as pressure drop, oil depth, and stripping steam
rate. This latter factor is important in that it strongly influences the size of the vacuum system
required to sustain the low pressures that are essential for good deodorization. With ever more
stringent limits being set for the quality of water discharged by refineries and with higher
costs for fresh water, the amount of motive steam required for ejectors and augmentors is
becoming critically important.
Several deodorizer designs set out to achieve a minimum oil film thickness so as to elimi
nate pressure drop and maximize stripping steam-oil contact. Such designs include falling
film deodorizers [44], packed column deodorizers [45], and climbing film deodorizers [46].
Batch Deodorization. Batch deodorizers are often used for processing small quantities of
oil that are not consistent with the relatively high throughput rates of modem semicontinuous
or continuous deodorizers. They can also be used for oils whose values and characteristics are
158 Allen
significantly different from those of the major oils being processed, because there is very little
chance of accidental commingling of oils in a batch deodorizer.
Modem batch deodorizers are likely to be fabricated from stainless steel, whereas older
units are probably of carbon steel. Heating is by high pressure steam or thermal fluid ap
plied to coils within the vessel itself. Capacities are in the range 5-30 t, most commonly 15-
25 t.
One problem with batch deodorizers has been the relatively low vaporization efficiencies
achievable because of the large depth of oil present compared with the depths encountered in
semicontinuous deodorizers, for example. These efficiencies have been improved by fitting
devices known as mammoth pumps that greatly increase the intimacy of mixing between
steam and oil.
The tops of batch deodorizers are insulated in the newer installations. This assists in ensur
ing that the top of the vessel is kept hot, at almost deodorization temperature, thereby avoiding
the problem of refluxing, in which free fatty acid vapors condense on cool surfaces and drop
back into the deodorizing oil.
Higher sparge steam rates are required to achieve satisfactory oil quality by batch deodor
ization. This can lead to significant oil losses by entrainment of oil droplets into the vapors
being drawn into the vacuum system. This effect is reduced by allowing a significant “disen
gagement” space above the level of the oil in the deodorizer and by fitting a demister either
across the deodorizer vessel or, more usually, in the vapor exit pipe.
In batch deodorizers, 2-5 h is the typical deodorization time. Because all the process
stages are carried out in the deodorizer vessel (see Section H.2), cycle times for batches are
likely to fall in the range of 6-10 h, giving an effective production rate of only approximately
3 t/h.
IV. PHYSICAL REFINING

A. Intensive Degumming
As mentioned in Section II, physical refining is an increasingly attractive process compared
with chemical refining. It is suitable only for low phosphatide oils. In order to make more
oils suitable for satisfactory physical refining, processes have been developed that remove
much greater proportions of phosphatides from liquid seed oils, in particular, to produce
feedstocks for physical refining. These contain less than 30 mg/kg of phosphorus and often
less than 10 mg/kg. Best known of these processes are the Superdegumming process, patented
by Unilever [47] and the TOP process patented by Vandemoortele [48]. The Superdegumming
process has been extended and improved subsequent to the initial patent to be capable of
producing oils containing less than 5 mg/kg phosphorus, thus matching the capability of the
TOP process. The extended Superdegumming process is often known as Unidegumming [49].
A summary of the total degumming process (TOP) is given in Fig. 8. The essential ele
ments of the process [50] are (1) virtually complete decomposition of metal-phosphatide
complexes by mixing with a dilute acid that is capable of binding metal ions very finely into
the oil being degummed, (2) complete removal of reaction products by partial neutralization
of the degumming acid, while avoiding soap formation, (3) removal of residual gums and
recycling of oil-rich gums, (4) removal of inorganic compounds and gum residues by water
washing, and (5) final removal of water by vacuum drying. The original process was equip
ment-intensive, requiring four centrifuges. Further operating experience at Vandemoortele has
led to the development of the TOP Dry process (Fig. 9), which now operates satisfactorily
Refining 159
Crude Oil
Heating
Dilute acid
Mixing
I
Holding
Dilute Alkali
Centrifugal- Low oil content

Separation gums
i
Centrifugal - High oil content Recycle
Separation gums
Water
Centrifugal - Waste water

Separation
Water
Centriftigal Waste water

Separation
Drying
I
TOP Oil
Fig. 8 Total degumming process (TOP).
with two centrifuges [51]. TOP oil typically contains less than 5-10 mg/kg phosphorus and
less than 0.2 mg/kg iron.
The Superdegumming process sequence [52,53] is shown in Fig. 10. The process appears
rather complex because of the number of stages. It offers the possibility, however, of increas
ing the amount of easily hydrated phospholipids, which facilitates the removal of phospholip
ids containing more phosphatidic acid [53]. Citric acid added to warm oil (~70°C) is dis
persed under high shear mixing. Once phospholipid hydrate particles are formed, only special
low shear mixing is used so as to prevent damaging these particles.
Unidegumming involves further processing of oil after Superdegumming. Superdegummed
oil is cooled slowly to 25°C, at which temperature an alkali solution is added to promote the
formation of larger phospholipid particles. Further residence time at 25°C allows the desired
particle agglomeration to take place. The mix is then rapidly heated to 60-65°C immediately
prior to centrifugal separation. The separator used is of the clarification type, capable of
removing 1-2% solids from the oil. Unidegummed oil contains less than 8 mg/kg phosphorus.
For some oils, levels as low as 1 mg/kg have been achieved [49].
Superdegumming, Unidegumming, and TOP degumming processes all produce oils that
are suitable for physical refining, i.e., they contain less than 30 mg/kg phosphorus and typi
cally less than 15 mg/kg when the latter two processes are employed.
160 Allen
Crude / Water Degummed Oil
Degumming Acid
Intense Mixing
Holding (2-3 minutes)

Concentrated Caustic
Soda Solution ------
Mixing
I
Separation - Gums
Wash Water
Separation/<pIarification
Separation
(optional)
Drying
I
TOP DRY Oil
Fig. 9 Total degumming process (TOP Dry).
Because Superdegumming and Unidegumming are carried out at low temperatures (25°C),
some waxes will be removed together with the phosphatides when sunflower oil is being
processed. This significantly reduces the load on subsequent dewaxing operations.
The TOP process focuses on iron removal as well as on phosphatide removal. This is
claimed to give a superior oil because of the prooxidant effect of iron.
Successful adoption of these processes on a commercial scale is determined by several
factors, including the additional capital and operating costs of the processes themselves
and the cost savings from reduced effluent treatment or disposal. As effluent disposal costs
are expected to continue to rise, at least in western countries, over the next few years as en
vironmental restrictions are tightened further, it is anticipated that intensive degumming
processes for liquid seed oils will become more attractive. This will be particularly true if
these processes are designed into new facilities, as opposed to being retrofitted into existing
refineries.
Intensively degummed oils can be caustic refined, if desired. In this case neutral oil losses
in the soapstock may be slightly reduced because of the virtually complete absence of phos
phatides in the oil and thus a much reduced tendency to emulsion formation. An overall
refining loss of 0.5% from crude to fully refined oil has been noted for caustic refining of
TOP degummed oil [50].
Refining 161
Crude / Water Degummed Oil
I
Heat to 70“C
(Modified Lecithin)
Citric Acid — -------
Mixer
1
Reaction Time (3 hours)
1
Cool below 4(TC
Water ■
\
-^ M ixer ( 2 5 0
1
Holding Time
1
Heat to 65“C
Separatgr - Gums/sludge
(M icrofiltration)--------- Very fine gum particles
1
Superdegummed Oil
Fig. 10 Superdegumming process.
B. Pretreatment
The objectives of pretreatment are similar to those of intensive degumming. Pretreatment is
applied to all oils, whether or not they require intensive degumming to be made suitable for
physical refining.
The main functions of the pretreatment process are to reduce the Totox value (Anisidine
Value + 2x Peroxide Value), the trace metal content, and the phosphorus content of the oil.
This is usually achieved by intimately mixing the oil with a low level (0.1-0.2% ) of concen
trated phosphoric acid before adding a relatively high level of bleaching clay to the oil (usually
at least 1.5 times the level used for bleaching refined or neutralized oils). The bleaching clay
is used to adsorb trace metals and oxidation products and to adsorb hydrolyzed phosphatides.
C. Physical or Steam Refining

As already indicated, the principal difference between physical or steam refining and deodor-
ization is the level of free fatty acids present in the feed material. Oils that have been neutral
ized and bleached, possibly followed by modification through hydrogenation, interesterifica
tion, or fractionation, will have free fatty acid levels below about 0.3% and probably below
162 Allen
0.1%. Oils that have been pretreated only in preparation for steam or physical refining will
have much higher levels of free fatty acids in the range of 1-5% , depending on oil type and
quality. As a result, physical refining plants have to be designed to be able to strip these high
levels of free fatty acid from the oil. In comparison with deodorizers, steam or physical
refining deodorizers have larger vacuum systems to handle the additional free fatty acid vapor
and to maintain a lower absolute pressure than is normally required for deodorization of
neutralized oils. They also achieve a longer residence time for the oil at deodorization temper
ature. They may, in addition, heat the oil to slightly higher temperatures and use greater
quantities of stripping steam. Current concerns about isomerization of cis bonds to the trans
configuration during deodorization is limiting the temperatures used for physical refining in
the same way that deodorization temperatures are being reduced by 10-15°C.
The equipment used for physical refining is very similar to that used for deodorization and
may be either semicontinuous or fully continuous. Designs are virtually identical to those used
for deodorization, with detailed differences only as discussed above. The whole range of
deodorizer designs are used for physical refining.
Palm oil is the oil that is most commonly physically refined. Care must be taken in select
ing palm oil for physical refining. Oil that has been oxidized should not be physically refined.
Satisfactory refined oil will be achieved only by chemically refining very oxidized palm oil.
Totox value is taken as the principal means of assessing whether or not a particular palm oil
consignment is suitable for physical refining. Crude palm oils with totox values in excess of
10 are not generally regarded as appropriate for physical refining. Crude palm oils with totox
values less than 10 should be suitable for physical refining provided that trace metal and
phosphorus contents are low— iron below 3 mg/kg, copper below 0.2 mg/kg, and phosphorus
below 20 mg/kg. Ideally the content of carotene should be high (over 550 mg/kg) because
this is widely regarded as a further indicator of overall good palm oil quality.
Because of its high carotene content, palm oil has a characteristic deep orange-red color.
This is little affected by pretreatment with phosphoric acid and bleaching clay. Once in the
high temperature environment of the physical refiner, the carotene is subjected to thermal
decomposition. Loncin et al. [54] showed that at 240°C the carotene content of palm oil is
reduced to only approximately 1% of its original level after a period of 35-40 min. Thus
dark-colored palm oil will be '‘heat bleached” during the physical refining process, usually
yielding a fully refined product that is lighter than 3 Red (Lovibond 5.25 in. cell) in color.
Coconut and palmkemel oils are also commonly physically refined because they are less
prone to oxidative deterioration in their crude states and because they are naturally low in
phosphorus content. Good quality, high grade tallows are also well suited to physical refining.
Batch deodorizers are used only infrequently for physical refining. This reflects the practi
cal difficulties of achieving sufficiently high temperatures and of economically providing large
enough vacuum systems to cope with the relatively high sparge steam rates. Physical refining
in batch vessels is normally limited to oils that have low free fatty acids levels, approximately
1-2%. Because heat interchange is much more difficult to achieve in batch processes than in
either semicontinuous or continuous processes, the energy costs of batch deodorization and
especially batch steam refining are much higher. This further reduces the incentive to attempt
steam refining as a batch process.
V. FUTURE TRENDS
Concerns to minimize the environmental impact of oil refining processes will become more
significant in the future. It is certain that legal requirements will impose even more stringent
limits on discharges to the atmosphere and to water. This pressure, coupled with an ongoing
Refining 163
objective of reducing processing costs and/or improving oil quality, is likely to encourage
equipment manufacturers and process material suppliers to further improve their products.
Within the limitations of trying to forecast the future, the following approaches seem likely
to be significant in contributing to the achievement of the above objectives.
Increased use o f physical refining. Physical refining does not involve the production of
soapstock with the consequent need to acidulate it to produce acid oil. This conversion
leads to the production of relatively large quantities of (usually) sodium sulfate solution
of variable pH and varying levels of oil contamination, requiring extensive treatment
prior to discharge into public sewer systems. In addition to significantly reducing the
quantity of water requiring treatment, physical refining leads to higher yields, because
of lower losses of neutral oil, and produces a higher concentration of free fatty acid in
the distillate by-product. It is likely that more widespread use will be made of current
intensive degumming processes (Section IV.A) or other innovative processes will be
developed to make possible satisfactory physical refining of a wider range of crude oil
types and qualities.
Use o f lower adsorbent levels for bleaching. The introduction of silica hydrogels has
focused attention on the relative abilities of different adsorbents to remove the range of
unwanted components from oils and fats. Since the initial approach that saw silica hy
drogel as an alternative or substitute for bleaching clay, it has been recognized that each
medium has its own strengths and weaknesses. From this has come a recognition that
the use of different adsorbents either together or in sequence can yield improved oil
qualities and lower the total level of adsorbent usage. From this have also come renewed
efforts on the part of bleaching clay manufacturers to further enhance the abilities of
bleaching clay to remove unwanted components at lower levels of clay in oil. Bleaching
clay and other solid adsorbents are expected to pose more demanding disposal problems
in the future, with high associated costs. Attempts to improve current adsorbents and
the introduction of other, better adsorbents will reduce solid waste disposal quantities.
Allied with this will probably be a reduction in the quantity of oil disposed in the spent
clay, with a resultant modest improvement in refinery yield.
Use o f regenerable bleaching media. Some work has already been reported [13,14,55]
on the possibility of economically regenerating bleaching clays for reuse through several
cycles, possibly with augmentation by fresh bleaching clay. Clearly, if this approach
can be developed into a commercial process, the quantity of absorbent to be disposed
of will be very significantly reduced compared with the current situation.
Improved deodorizer steam efficiencies. Further improvements in steam stripping efficien
cies will result in reduced steam consumption in stripping demand and vacuum genera
tion by steam ejectors. Falling film deodorizers, for example, require less stripping
steam than tray designs [1]. Mention of vacuum systems leads to consideration of odor
emissions from deodorizer distillate and/or hot wells. The present trend toward indirect
condenser systems and recycling of greasy water is expected to continue and to acceler
ate, giving further improvements to water management and odor elimination.
Integrated facilities. Where refining is carried out on the same site as crushing or extrac
tion, there is greater flexibility to optimize processing to maximize yield or to minimize
energy costs. There are opportunities also to blend certain of the waste products from
refinery processes into the meal product from the crushing plant. Examples are gums,
soapstock, and spent bleaching clay. It is important to recognize, however, that not all
these refinery wastes can be blended into the meal simultaneously because the recipient
meal can be put outside specified limits for oil content, for example, by indiscriminate
164 Allen
blending. In the future, it is possible that the practice of mixing spent bleaching clay
into oilseed meals may be restricted or even prohibited because of the presence of
antinutritive components.
A further benefit from operating a large integrated crushing and refining facility may
be the economic justification for a combined heat and power (CHP) plant, which permits
steam raising and power generation to be optimized so as to reduce the overall energy
consumption of the facility and reduce, or even remove, its dependence on externally
supplied electricity.
In addition to the specific points noted above, it is anticipated that through ongoing devel
opment work, equipment and process material suppliers will continue to make small but im
portant improvements to their products to permit achievement of high yields, improved oil
quality, greater throughputs, or less plant downtime.
VI. GENERAL
The processes described here are used successfully to produce high quality refined oils for a
wide range of applications. These may be supplied as individual oils or as blends of oils
chosen to meet particular specifications.
In addition to the refining processes described here, a refiner may need to be able to use
modification processes (described in Chapters 8-10) in order to provide the full range of oil
products required by customers. It may also be necessary to include additional processes such
as dewaxing or winterizing that are necessary to ensure that certain oils remain completely
clear even at low temperatures. This is important for bottled oils supplied for consumer use
and oils used in mayonnaise production. Where oils are being produced for bottling or for
mayonnaise production, it is important that there be no crystallization of stearins or waxes at
low temperatures. For bottling, this is an aesthetic requirement. For mayonnaise production it
is essential in order to avoid instability in the emulsion when the product is refrigerated.
Dewaxing may be carried out in a variety of ways. The simplest process involves cooling
partially refined or fully refined oil to approximately 4-5°C, holding the oil at that temperature
under very gentle agitation for a long enough period for the high melting waxes (fatty alco
hols) to crystallize, and then heating the mixture to approximately 10°C to reduce the viscosity
prior to filtration through a filter that has been precoated with a filter aid and/or using a filter
aid as a body feed with the oil.
More sophisticated processes have been developed in attempts to reduce the yield losses
associated with the basic process. It is quite common to take oil from the continuous neutral
ization plant after the primary separator, at which stage the oil may contain about 1000 mg/
kg of soap, or to dose the oil with a low level of caustic soda prior to the secondary separator.
Rapid chilling of this oil will crystallize the waxes, with the soap particles acting as nuclei.
After storage at low temperature for extended periods (12-24 h) under low shear agitation,
the mixture is heated in-line to approximately 10-15°C and then water washed as normal.
This process sequence claims to keep yield losses to a minimum. In most cases, however,
dewaxed oils produced this way do not meet the stringent chill test requirements and a second
ary chilling and filtration process is also required. Here the loading of waxes on the filter is
very low and yield losses are modest.
The term “winterizing” is often used as an alternative to dewaxing to describe the removal
of high melting point waxes. More correctly, winterizing is a very similar process to dewaxing
but its purpose is to remove traces of high melting stearins from, for example, lightly hydroge
nated oils or oils such as peanut oil that naturally contain low levels of stearins. In winterizing
Refining 165
plants, the filtration step is carried out in cold rooms because the stearins are much more
likely to melt if exposed to warm temperatures than are waxes, whose melting points are often
in excess of 65-70°C.
Oils that are to be hydrogenated are usually partially refined and occasionally fully refined
prior to hydrogenation. This is because hydrogenation catalysts are relatively easily poisoned
by some of the non-triacylglycerols found in crude oils, especially sulfur compounds. Catalyst
poisoning, or partial deactivation, leads to extended reaction times and, more important,
changes in relative reaction rates, affecting selectivity and trans isomerization. Where fish oils
and canola oils are being partially refined prior to hydrogenation, it is very important to ensure
that the bleaching process has effectively removed almost all of the sulfur compounds that
may have been present in the crude oil. For canola oil it can occasionally be crucially im
portant to reduce the chlorophyll content to low levels (0.1 mg/kg or less) prior to hydrogena
tion in order to avoid formation of a green coloration in the hydrogenated oil, which is almost
impossible to remove completely.
Overall, modem refineries have a wide choice of processing alternatives, which should
enable them to produce high quality oils and fats from a range of types and qualities of
cmde oils. Indications are that the range of alternatives available will increase even further in
the future.
REFERENCES
K. F. Carlson and J. D. Scott, Recent developments and trends: processing of oilseeds, fats and
oils, INFORM 2: 1034 (1991).
2 . A. Hvolby, Removal of non-hydratable phospholipids from soybean oil, J. Am. Oil Chem. Soc.
48: 503 (1971).
3. H. Pardun, Deacidification of vegetable oils with ammonia—an environmentally safe refining
method, Fette Seifen Anstrichm. 75 : 297 (1979).
4. F. V. K. Young, Caustic refining. Trace impurities, refining methods and edible oil quality. Pre
sented at Second ASA Symp. on Soybean Processing, Antwerp, The Netherlands, June 1981.
R. Fahn, Influence of the structure and morphology of bleaching earths on their bleaching action
on oils and fats, Fette Seifen Anstrichm. 75: 77 (1973).
6 . T. K. Mag, Bleaching— theory and practice, in Edible Fats and Oils Processing: Basic Principles
and Modern Practices (D. R. Erickson, Ed.), AOCS Press, Champaign, IL, 1990, p. 107.
7. F. V. K. Young, Processing of fats and oils: production of edible oils, in The Lipid Handbook
(F. D. Gunstone, J. L. Harwood, and F. B. Padley, Eds.), Chapman and Hall, New York, 1986,
p. 189.
8 . A. N. Sagredos, D. Sinha-Ray, and A. Thomas, Determination, occurrence and composition of
polycyclic aromatic hydrocarbons (PAH) in oils and fats. Fat Sci. Technol. 90: 76 (1988).
9. F. A. Norris, Refining and bleaching, in Bailey’s Industrial Oil and Fat Products, 4th ed., Vol.
2 (D. Swem, Ed.), Wiley, New York, 1982, p. 292.
10. U. I. Brimberg, Kinetics of bleaching vegetable oils, J. Am. Oil Chem. Soc. 59: 74 (1982).
11. J. H. Henderson, A laboratory study of the press effect in adsorptive bleaching. Presented at 83rd
AOCS Annual Meeting, Toronto, Canada, May 1992.
12. R. Fahn and W. Zschau, Possibilities for the use of spent bleaching earth. Presented at 75th AOCS
Annual Meeting, Dallas, TX, 1984.
13. R. Nebergall, D. Taylor, and B. Jaynes, Bleaching: solving the problem of disposal. Presented at
84th AOCS Annual Meeting, Anaheim, CA, May 1993.
14. G. Toeneboehn and W. A. Welsh, Regenerate adsorbent process for refining. Presented at 84th
AOCS Annual Meeting, Anaheim, CA, May 1993.
15. W. A. Welsh and Y. O. Parent, U.S. Patent 4,629,588 (W. R. Grace & Co.) (1986).
16. P. M. Parker and W. A. Welsh, U.S. Patent 4,734,226 (W. R. Grace & Co.) (1988).
166 Allen
17. J. N. Pryor, J. M. Bogdanor, and W. A. Welsh, U.S. Patent 4,781,864 (W. R. Grace & Co.)
(1988).
18. J. N. Pryor, J. M. Bogdanor, and W. A. Welsh, U.S. Patent 487765 (1989) (W. R. Grace & Co.)
19. W. A. Welsh, U.S. Patent 4880574 (1989) (W. R. Grace & Co.)
20 . A. Nock, Factors affecting the performance of silica hydrogel in edible oil refining. Presented at
8 6 th AOCS Annual Meeting, San Antonio, TX, May 1995.
21 . W.-L. Slew, Y.-A. Tan, and T.-S. Tang, Silica refining of palm oil, J. Am. Oil Chem. Soc. 71:
1013 (1994).
22 . J. M. Bogdanor, Modified caustic refining. Presented at the International Conference on Fats for
the Future, sponsored by the Royal Society of New Zealand, Auckland, February 1989.
23. J. M. Bogdanor, Modified bleaching. Presented at 80th AOCS Annual Meeting, Cincinnati, OH,
May 1989.
24. J. M. Bogdanor and G. J. Toeneboehn, Silica refining of oils with little or no chlorophyll, Pre
sented at 80th AOCS Annual Meeting, Cincinnati, OH, May 1989.
25. W. A. Welsh, J. M. Bogdanor, and G. J. Toeneboehn, Silica refining of oils and fats, in Edible
Fats and Oils Processing: Basic Principles and Modern Practices (D. R. Erickson, Ed.), AOCS
Press, Champaign, IL, 1990, p. 189.
26. J. M. Bogdanor, Modified physical refining— commercial considerations. Presented at 84th AOCS
Annual Meeting, Anaheim, CA, April 1993.
27. L. O. F. Schumutzler, U.S. Patent 5,248,799 (1993).
28. A. Nock, A novel acid refining process using silica adsorbent. Presented at 84th AOCS Annual
Meeting, Anaheim, CA, April 1993.
29. W. A. Welsh and P. M. Parker, Eur. Patent 0 234 221 (W. R. Grace) (1991).
30. A. Nock, Acid refining using silica—a process for refining high phosphorus containing oils. Pre
sented at 85th AOCS Annual Meeting, Atlanta, GA, May 1994.
31. A. Thomas, Desired quality attributes in winter and summer rapeseed, J. Am. Oil Chem. Soc. 49:
1 (1982).
32. A. Thomas, The removal of trace contaminants from oils and fats by deodorisation, Fette Seifen
Anstrichm. 84: 133 (1982).
33. F. A. Norris, Deodorization, in Bailey’s Industrial Oil and Fat Products, Vol. 3 (T. H.
Applewhite, Ed.), Wiley, New York, 1985, p. 127.
34. D. S. Szarbo Sakardi, Laboratory deodoriser with a vaporization efficiency of unity, J. Am. Oil
Chem. Soc. 35: 472 (1958).
35. A. M. Gavin, Edible oil deodorization, J. Am. Oil Chem. Soc. 55; 783 (1978).
36. G. M. Wilson, Vapour liquid equilibrium. Part XL A new expression for the excess free energy
of mixing, J. Am. Oil Chem. Soc. 86: 127 (1964).
37. H. Stage, W. Kuhns, and H. Hammer, U.K. Patent 2 046 606 (1979).
38. R. L. Wolff, Trans polyunsaturated fatty acids in French edible rapeseed and soybean oils, J. Am.
Oil Chem. Soc 69: 106 (1992).
39. L. H. Wiederman and D. R. Erickson, Soyabean oil: modem processing and utilisation, INFORM
2: 200 (1991).
40. R. G. Ackman, S. N. Hooper, and D. L. Hooper, Linolenic acid artifacts from the deodorization
of oils, J. Am. Oil Chem. Soc. 51: 42 (1974).
41. G. Devinât, L. Scamaroni, and M. Naudet, Isomerization of linolenic acid during deodorization
of rapeseed and soya oils. Rev. Fr. Corps Gras 27: 283 (1980).
42. J. B. Rossell, Heat transfer fluids in the oils and fats industry, 1. Specification, types and toxicity.
Lipid Technol. 5: 110 (1993).
43. J. B. Rossell, Heat transfer fluids in the oils and fats industry, 2. Acceptable media, characteristics
and analysis. Lipid Technol. 5: 134 (1993).
44. D. R. H. Stage, Countercurrent falling film steam distillation using externally applied temperature
field for efficient deodorization of fatty acids and edible oils, Fette Seifen Anstrichm. 78: 457
(1976).
45. W. B. Hendrix and P. Sjoberg, Evaluation of the effects of packed column deodorization on seed
oil. Presented at 8 6 th AOCS Annual Meeting, San Antonio, TX, May 1995.
Refining 167
46. R. S. Jickling, Deodorization in climbing thin films, Presented at 6 8 th AOCS Annual Meeting,
New York, 1977.
47. H. J. Ringers and J. C. Segers, U.S. Patent 4,049,686 (to Unilever) (1977).
48. A. J. Dijkstra and M. Van Opstal, U.S. Patent 4,698,185 (to Vandemoortele) (1987).
49. J. C. Segers and R. L. K. M. Van de Sande, Degumming of rapeseed oil, Presented at 83rd
AOCS Annual Meeting, Toronto, Canada, May 1992.
50. A. J. Dijkstra and M. Van Opstal, The total degumming process, in Edible Fats and Oils Pro
cessing: Basic Principles and Modern Practices (D. R. Erickson, Ed.), AOCS Press, Champaign,
IL, 1990, p. 176.
51. A. J. Dijkstra, Degumming, refining, washing and drying fats and oils. Presented at World Conf.
on Oilseed Technology and Utilisation, Budapest, Hungary, September 1992.
52. J. C. Segers, Possibilities and limitations for decreasing the environmental load from the refining
of edible oils, Fette Seifen Anstrichm. 87: 541 (1985).
53. J. C. Segers and R. L. K. M. Van de Sande, Degumming—theory and practice, in Edible Fats
and Oils Processing: Basic Principles and Modern Practices (D. R. Erickson, Ed.), AOCS Press,
Champaign, IL, 1990, p. 8 8 .
54. M. Loncin, B. Jacobsberg, and G. Evrard, Palm oil. Rev. Fr. Corps Gras 18: 69 (1971).
55. T. Jacobson and R. Nebergall, Spent bleaching earth regeneration—commercial scale results. Pre
sented at 8 6 th AOCS Annual Meeting, San Antonio, TX, May 1995.
Oil Storage, Tra nsp o rt, and Handling
Martin H. Hilder
Unilever Research Laboratory, Vlaardingen, The Netherlands
I INTRODUCTION
Oil storage and handling are no more than a necessary evil. They cost money and result in
deterioration of oil quality. At best, if the deterioration in oil quality is not kept under control,
extra costs are incurred; in extreme cases the oil can become unprocessable or unsalable.
Considering the number of times an oil may be handled or stored during processing, stor
age and handling must be the most widely used processes in the oil industry (Fig. 1). Even
this is a simplified picture, as oil will be stored by the supplier and often transported to and
stored at the port of loading, and after arrival sea cargoes often undergo further storage and
transport before they reach their destination. At each storage or transport step there is a risk
of degradation; each transfer into storage or transport involves an oil handling operation that
is equally as critical, if not more so.
The objectives of good oil storage and handling are to minimize costs and minimize oil
deterioration. Sometimes these are mutually consistent objectives, but in the end the art is to
strike an optimal balance. This requires an understanding of the source and magnitude of the
costs of oil storage and handling and the factors involved in oil degradation, which also
provide a sound basis for the proper design, operation, and management of oil storage and
handling facilities.
The following sections contain a summary of the principles involved, covering
The relative costs of capital, operation, and degradation and the factors that determine them
The principles of oil hydrolysis and oxidation, which are the main causes of degradation
The application of these principles to the practical requirements of oil storage, transport,
and handling
For more detailed information several publications are worth referring to. Basil Patterson’s
book [1], which also covers oilseeds and meal, is the most comprehensive. The chapter by
List and Erickson in the Handbook o f Soy Oil Processing and Utilisation [2] provides a useful
169
170 Hilder
Supplier
I
I
Bleaching
I
Storage
!
(Modification)
I
(Storage)
I
Deodorization
I
Storage
I
Transport
I
Customer
Fig, 1 Storage and handling in oil processing.
review of the literature and state of the art in the U.S. soybean industry up to 1980. This
review also formed the basis for List’s chapter in Volume 3 of the latest edition of Bailey [3];
previous editions of Bailey (e.g., Ref. 4) are not so informative.
A final point regarding definitions. In the United States the term “refined” oil is often used
to describe what in Europe is known as neutralized oil. In this chapter the following European
definitions are used:
Neutralized oil is referred to as “neutral” or “neutralized” oil.

“Fully refined” or “refined” oil always refers to the final deodorized oil.
II. COSTS
The costs associated with oil storage and handling can be divided into three parts:
1. Capital cost: both investment and working capital

2. Operating costs: the costs of labor, heating, pumping, cleaning, etc.
3. Oil degradation costs: any extra oil loss or processing or reprocessing costs and the
loss in final oil quality
For the present purpose it is more important to understand the relative contributions rather
than to know the actual costs; any cost data in real money are soon out of date, and every
company will have its own unit costs and financial criteria. Costs are therefore discussed here
Oil Storage, Transport, and Handling 171
in relative terms based on the “triglyceride dollar (TG$),’ with an average price of oil of
1000 TG$/ton or 1 TG$/kg.
A. Capital Costs
The capital costs of oil storage and handling are primarily determined by the quantity to be
stored. One week’s oil (buffer) stock for a 100,000 tons-per-annum (tpa) plant represents
2000 metric tons (t) of oil. On average, tanks are only half-full, and therefore the required
tank capacity is double the average stock requirement—in this example, 4000 t tank capacity
per week oil stock requirement. The actual stock requirement will depend on company policy
and supply-and-demand logistics.
Crude oil storage capacity is normally larger than refined oil storage capacity because
1. Crude oil is usually delivered in larger quantities and is more stable.

2. Refined oil is usually less stable and is usually transported in smaller amounts, and
once refined, quality (freshness) is paramount.
Ideally, the buffer requirement should be minimized to what is essential for operational pur
poses. Any extra quantities should be justified on their own merits.
L Investm ent
The capital invested in storage facilities will obviously depend on the oil storage requirements
and the scale of operation (plant size) but will ultimately be determined by the number and
size of the tanks. The type of operation, and in particular the number of crude oils and refined
products to be handled, will therefore also play a major role. One to two tanks are required
for each crude or refined oil. The choice will depend on how much segregation is required to
separate oil qualities, ensure first-in, first-out operation, and other factors.
Assuming for the sake of simplicity that 1 week’s crude oil and 1 week’s refined oil buffer
are required, there is a significant difference between a 100,000 tpa plant for
1. One crude + one refined oil: say, 4 x 2000 t tanks

2. Five crude + 20 refined oils: say, 32 x (on average) 250 t tanks
General cost engineering data for the United Kingdom and The Netherlands show that
tankyard costs increase with the 0.5-0.7 power of the tank size. For our present purposes,
we will assume total investment based on Table 1.
Table 1 Effect of Tank Size on

Tankyard Costs
Tank size Investment per tank

(t) (1000 TG$)
100 200
250 300
500 450
1000 600
2500 1000
172 Hsider
The crude and refined oil tankyard investment for our two plants is therefore about
3.5 million TG$ for 4 x 2000 t tanks
10 million TG$ for 32 x 250 t tanks
Every company will have its own criteria, but a target pretax return on investment plus depre
ciation of 20-25% per year is not unusual; say at least 5% depreciation and 15% return on
investment (ROI).
2. W orking Capital
The cost of the oil held as a buffer stock also has to be financed. If, for example, interest
rates of 10% have to be paid, the cost of financing our 2000 t of crude oil and 2000 t of
refined oil buffer stocks will be about 0.4 million (4 x 10^) TG$ per year.
B. Operating Costs
Labor costs will depend on the organization rather than on technology, so we will assume a
nominal cost.
Heating costs based on a steam requirement of 600 kg/yr per square meter of tank surface
area to maintain the oil at 50°C in tanks with 50 mm insulation would be about 15,000 TG$/
yr for a 2000 t tank (1000 m^) and 3500 TG$ for a 250 t tank (250 m^), although the single
oil plant is likely to be dealing with only liquid oils, which would not require heating.
The pumping costs based on 0.25 kWh/t for a 4 bar pressure drop amount to about 0.025
TG$/t oil, or, for pumping crude oil in and pumping refined oil out, about 0.05 TG$/t.
The agitation costs for crude oil based on an installed power of 3-5 W/t tank capacity,
depending on the tank size, will amount to 0.1-0.15 TG$/t oil processed.
The nitrogen consumption for deodorized oil sparging (0.5 m^/t), refined oil storage blan
keting (max 1 m^/m^ oil), and line blowing, etc., would amount to about 2 m^/t oil, giving a
cost of about 0.5 TG$/t oil.
The repair and maintenance (R&M) costs can be estimated as, say, 5% per year on in
vestment.
Cleaning costs and disposal of waste will depend on the local situation and possibilities.
C. Degradation Costs
The cost of any hydrolysis in crude oil storage is relatively simple to estimate. Each 0.1%
free fatty acid (FFA) increase plus the neutral oil it takes with it during processing represents
a loss of, say 1.5 kg/t, but the recovered (acid) oil can be sold at about half the price of the
original oil. With liquid oils at ambient temperature, hydrolysis should be negligible; but with
high FFA crude that has to be stored at higher temperatures (e.g., crude palm oil), some
hydrolysis is virtually unavoidable. With poor practice, examples where the FFA increase was
0.5% or more are not unknown.
The cost of oxidation is more difficult to quantify. Depending on where it occurs during
processing, it can result in extra processing costs (e.g., extra bleaching earth) or total re
processing, with all the costs and loss in effective capacity that this involves.
Based on the example given by Strecker and his colleagues [5], which is summarized in
Section V.A.2, we will assume that with good oil storage and handling practice (time and
conditions), oxidative degradation can be reduced to a minimum level, which can be compen
sated for by an additional 0.1% bleaching earth during normal processing. The costs associ
ated with this (earth, oil loss, etc.) amount to, say, 1.5 TG$/t oil more than if the crude oil
were processed directly on arrival.
A s Strecker C'emonstrated, long storage times and poor storage conditions can result in
much greater extiv:; earth requirements; with poor practice, 0 .5 % or more extra earth can be
required. We can add to this the oil loss associated with any “tank-bottoms,” the sludge which
separates out during storage. Crude oils should not normally contain more than about 0.05%
dirt. If this is allowed to settle in storage it can entrain many times its own weight of oil; as
a paste or slurry, tank-bottoms will incur oil losses of at least 1-1.5 kg/kg dirt. This oil loss
can be avoided to a great extent if tanks are agitated.
Finally, we should not forget the loss in refined oil quality due to oxidative degradation
either during or after processing. This is even more difficult to quantify, but it can be expected
to affect oil stability or shelf life and therefore may show itself in the form of reduced market
share or increased trade returns. For the present purpose, we will assume a cost equal to the
cost of oxidation during processing.
D. Total Costs
Adding up the individual items for the two 100,000 tpa examples, assuming good practice,
we arrive at the totals in the first column of Table 2. The column headed “Poor practice” has
been added to show the possible effect of long storage times and poor storage and handling
conditions. These figures are simply intended to show the relative effects of individual items
and practices and do not necessarily follow conventional accounting rules.
The first conclusion that can be drawn is that items related to capital, including R&M,
make up a large part of the total, even if we are satisfied with a lower return on investment.
Table 2 Hypothetical Costs (TG$/t oil) of Crude and Refined

Oil Storage
Item Good practice Poor practice
Average stock (weeks) 2 4

Capital (TG$/tpa) 35-100 70-200
Capital costs
Depreciation 2-5 3.5-10
ROI 5-15 10-30
Working capital 4 8
Operating costs
Labor 3 4.5
Heating 0 -1 0 -2
Pumping 0.05 0.05
Agitation 0.1-0.15 —
Nitrogen 0.5 —
R&M 2-5 3.5-10

Tank cleaning ____ a __b
Degradation costs
Hydrolysis 0-0.75 2.5
Oxidation 1.5 5
Tank bottoms — ^ 1
Final oil quality 1.5 5
Total (TG$/t) 20-37 43-78
^minimal cost,
‘’significant extra cost.
174 Hilder
The capital cost of storage is high, and this is the first good reason for minimizing storage
capacity and stocks. The other good reason is the inevitable degradation associated with any
storage. Once the capital has been invested, the capital costs are more or less fixed, but at
least half of the remaining (controllable) costs are associated with oil degradation. With good
practice, the cost of degradation can be reduced to a tolerable minimum, but with long storage
and poor conditions this cost can increase more than proportionally.
Up to now we have considered only crude and refined oil storage, but the same arguments
apply to intermediate storage. For the same reasons, any intermediate storage should be re
duced to an absolute minimum. A minimal process buffer between processing steps is the
ideal situation, so that intermediate storage as such can be eliminated.
III. HYDROLYSIS OF OILS

The reaction between water and glycerides (hydrolysis) results in the formation of free fatty
acid (FFA) and a diglyceride, monoglyceride, or glycerol. All these degradation products can
cause a financial or quality loss:
1. Any increase in FFA and, to a lesser extent, di- and monoglycerides usually results in
a direct financial (oil) loss.
2. In some products, the presence of di- or monoglycerides has an adverse effect on
product properties.
3. Glycerol in effluent will contribute to any COD or BOD (chemical or biological oxy
gen demand) costs.
Under storage conditions the FFA will normally be less than 10% and the predominant
reaction is [6]
Trigyceride + H 2O —» diglyceride + FFA
Enzymatic hydrolysis will not be considered; with normal levels of impurities in crude oil and
the proper (heat) pretreatment at an earlier stage, this can be neglected.
Fatty acids of chain length and above are essentially flavorless and do not affect the
taste of oils and fats. However, short-chain fatty acids can cause ‘‘hydrolytic rancidity” in
butter (rancid flavor) and in lauric oils and fats (soapy flavor); the flavor thresholds of lauric
and capric fatty acids are about 0.07% and 0.02%, respectively [1, p. 84; 3, p. 274].
A. Analyses
The determination of FFA and moisture are standard analyses in the oils and fats industry.
For accurate measurement of moisture contents, the Karl Fischer method is normally used.
With good laboratory practice, levels down to 0.01% H 2O can be detected, and with more
care an accuracy of ±0.001% H 2O is achievable [7], but with poor laboratory practice it can
be difficult to obtain reliable results below 0.1% H 2O.
B. Solubility and Diffusivity

Hydrolysis takes place only with water dissolved in the oil. The solubility of water in oil and
the rate of diffusion of “free” water (liquid or vapor) into the oil may therefore be relevant in
understanding the rate and extent of hydrolysis.
Values of the approximate solubility and diffusivity of water in edible oil deduced from
published data are given in Table 3. The solubility of water in oil has been measured in the
Table 3 Solubility and Diffusivity of Water in Oil
Water Water vapor

Temp. saturated at 1 0 0 mbar Viscosity Diffusivity
(°C) (%) (%) (cP) (m^/s X 10^)
0 0 .1 — — ____
25 0.15 — 62 0.25
50 0 .2 0.15 24 0.7
75 0.3 0.065 12.5 1.4
100 0.45 0.035 7.5 2 .6
125 0.65 0 .0 2 5.1 4.0
150 ~1 0.015 3.6 6 .1
temperature range 0-265°C. The saturation solubility increases with temperature, but the solu
bility of water vapor decreases with temperature [7,8].
The mole fraction solubilities are the same for all oils. The weight percent solubilities in
Table 3 (for an average oil of MW 850) are therefore dependent on oil type; for example, in
coconut oil the percent solubility is 25-30% higher. The solubility of water vapor obeys
Henry’s law, and the (mole fraction) solubility is proportional to the partial pressure of water
vapor. The solubility at 50°C and 10 mbar is therefore about 0.015%, and at 100°C and 200
mbar, about 0.07%.
The diffusivity of water in oil was measured at 24°C [9]. The other values in Table 3 have
been estimated based on the Stokes-Einstein relation; i.e., the product of the diffusivity and
viscosity divided by the absolute temperature is roughly constant.
C. Factors Affecting Hydrolysis

The rate and extent of hydrolysis under storage conditions was extensively studied by Loncin
[6,7]. He and his colleagues also made a significant contribution to the study of the solubility
of water in oil, especially on the effect of FFA [6,7,11]. The rate of hydrolysis is dependent
upon the FFA level, the concentration of (dissolved) water, the temperature, and the type
of oil.
1. F ree Fatty A cid

Hydrolysis is an autocatalytic reaction; the rate is proportional to the level of FFA. If the
hydrolysis of an oil with 1% FFA results in a final FFA of 1.5%, an oil with 5% FFA will
hydrolyze to about 7.5% FFA under the same conditions. Loncin [6,7] provides many exam
ples to illustrate this effect.
2. M oisture
It is reasonable to expect that the rate of hydrolysis is proportional to the concentration of
(dissolved) water, but direct experimental evidence for this is scarce. Sarkadi [12] showed
that the increase in FFA during deodorization is proportional to the water vapor pressure.
Based on the solubility relations, this suggests that the rate of hydrolysis is basically depen
dent on the water concentration. The situation is made more complicated by the fact that the
solubility of water in fatty acids is greater than in the corresponding triglycerides, and the
solubility of water in oil also increases with increasing FFA.
176 Hllder
Table 4 Effect of FFA on the Saturation Solubility of Water

in Palm Oil
Temperature (°C)
40 60 80
FFA (%) Saturation solubility (%)
0 .2 0.18 0.23 0.32

2.5 0 .2 1 0.27 0.36
6.5 0.25 0.32 0.42
15 — 0.42 0.52
The effect of FFA on the saturation solubility of water in palm oil [6,7,11] is shown in
Table 4. However, it is important to note that Loncin [6,7] concluded that the autocatalytic
effect is not due to the enhanced solubility of water with increasing FFA.
Moisture is also consumed in the hydrolysis reaction. Complete conversion of 0.1% H 2O
will result in the formation of about 1.2-1.6% FFA, depending on the molecular weight of
the triglycerides. If no more moisture is supplied, the dissolved water content will decrease
and the reaction will slow down. With an excess of (“ free” ) water, more water can dissolve.
In this case the rate of hydrolysis may be dependent on the rate of transfer or diffusion of the
water into the oil.
All these effects illustrate the importance of minimizing the moisture content during trans
port and storage:
1. At moisture contents at or below the saturation solubility, the hydrolysis reaction will
tend to be self-limiting.
2. With even only 1% of “free” water, the rate and extent of hydrolysis can be many
times greater.
3. Tem perature
As with any chemical reaction, the effect of temperature on the rate of hydrolysis is signifi
cant; the rate approximately doubles with each 10°C rise in temperature [6,7]. From the above
description of the effects of FFA and moisture, the rate of hydrolysis can be formulated as
d [FFA]
= k [H 20][FFA]
dt
Loncin and his colleagues worked mainly with saturated oil-FFA mixtures in the presence
of excess (“free”) water; this is not always clear from the published results [6,7] but is a
reasonable assumption. They evaluated the effect of temperature based on the equation
d [FFA]
= k ' [FFA]
dt 10
with time (tio) in units of 10 days; their results are summarized in Table 5. The rate constant
k ' is therefore the product k [H20 ]sat and is not a pure reaction rate constant; it includes the
effect of temperature (and FFA) on the saturation solubility of water, k' therefore indicates
the maximum rate of hydrolysis of an oil-FFA mixture under storage conditions.
Table 5 Effect of Temperature on

the Rate of Hydrolysis of Palm Oil
Temperature Rate constant

(°C) k' [(10 days)“ ']
37 0.025
50 0.05
60 0.125
70 0.29
80 0.51
100 1.5
4. Oil Type
Loncin [6,7,11] also investigated the effect of the type of oil and the type of fatty acid on the
rate of hydrolysis. The results show that the rate of hydrolysis increases in the order
Groundnut oil < palm oil < coconut oil
Palm acid < coconut acid < butter acid
This is also the order of decreasing molecular weight and increasing saturation solubility of
water. The influence of oil or fatty acid type is therefore at least qualitatively due to its effect
on the level of dissolved water.
The pH or addition of anionic or cationic surface-active compounds to the water phase, or
the presence of protein, pectin, carotene, cellulose, phosphatides, monoglycerides, or metal
soaps have no effect on the rate of hydrolysis [6,7].
5. O verall R ate o f H ydrolysis

The effect of FFA, moisture, and temperature on the rate of hydrolysis can therefore be
described by
d [FFA ] [H 2 O]
- = ........ [FFA ]
dt [H2O]
2 '^ J s a t
Where the time t and rate constant k' are expressed in the same units. The maximum hydroly
sis occurs when the oil-FFA mixture is saturated with water, and can be expressed as
[FFA ]
In
[FFA] 0 V [FFAJo /
where [FFA ] q is the initial FFA concentration and [FFA]i„c is the increase in FFA concentra
tion. The maximum increase in FFA over initial FFA can therefore be estimated. The effect
of storage time and temperature is illustrated in Fig. 2.
The exponential increase with time shown in Fig. 2 assumes that the oil-FFA mixture is
always saturated, i.e., that additional “free” water dissolves to make up for the increase in
saturation solubility with increasing FFA and for the moisture consumed by the reaction. In
practice this will not always happen, and the reaction will slow down as a result, giving the
approximately linear increase in FFA often observed in practical situations. Even under satu-
178 Hilder
TIME (WEEKS)
Fig. 2 The hydrolysis of palm oil saturated with water.
rated conditions the increase may appear to be linear over short periods of time within the
limits of experimental accuracy.
The rate and extent of hydrolysis at lower moisture levels or with other types of oil can be
estimated by adjusting the “rate constant” (k') for the actual moisture content and/or saturation
solubility. For example, after 4 weeks storage or transport of crude palm oil with 5% FFA at
50-60°C and 0.1-0.2% H 2O (50% saturated), one would expect an FFA increase of 0.37-
0.96%. This is probably representative for the storage and transport conditions in 1953 when
Loncin [6] quoted an increase of about 0.1-0.25% FFA per week. For neutralized or “rbd”
(refined-bleached-deodorized) oils and with modem transport and storage practices, the in
crease should be (and generally is [13]) much lower.
IV. OXIDATION OF OILS

The reaction between oxygen and glycerides is far more complex than hydrolysis, but in
general it will also lead to an increase in processing costs, oil losses, and a loss in quality. In
extreme cases it can result in an oil becoming impossible to process or impossible to sell.
The oxidation of edible oils proceeds via three steps:
1. Absorption of O2
2. Formation of peroxides
3. Degradation of peroxides to secondary oxidation products:
O2 (g) ^ 0 2 (l)—>peroxides-^ secondary products
Much has been written about the mechanisms of peroxide formation and degradation. Far
less attention has been paid to the absorption step, which is probably the most important in
industrial practice.
Oxidation can take place only if there is dissolved oxygen in the oil. If the absorption of
oxygen is prevented or minimized, oxidation cannot take place or is limited.
A. Analyses
Many methods for analyzing the extent of oxidation of edible oils have been proposed and
tested [15,16]. The two most popular and useful are those determining the peroxide value
(PV) and the anisidine value (AV).
1. P eroxide and A nisidine Values

The peroxide value is the result of a chemical analysis of the hydroperoxide content of the
oil. One millimole of hydroperoxide per kilogram of oil will give PV = 2; a PV of 1 therefore
represents the conversion of 16 mg 02/kg oil into hydroperoxides. An alternative measure is
the Lea value (LV), which is equal to the number of millimoles of hydroperoxide per kilogram
of oil and is therefore related to PV by PV = 2 LV.
The anisidine value (previously called the benzidine value) is a physical measure of the
aldehyde content based on the extinction at 350 nm. Aldehydes are the most important degra
dation products of hydroperoxides with respect to final taste and odor. The AV does not
measure all the secondary products but is accepted throughout the industry as the best stan
dard analysis.
In practice, often only the PV is measured, but this can be misleading. Referring to the
reaction scheme above, it can be seen that (1) a low PV may be the result of fast breakdown
of hydroperoxides rather than slow formation and (2) a high PV during processing is never
good, but the effect on final taste and stability may be limited by selective breakdown to
secondary products with less taste impact.
Another problem with the PV is the need for careful sampling and sample handling. It is
difficult to avoid contact with or absorption of air, but it should be remembered that every 8
mg dissolved or absorbed oxygen per kilogram of oil may give a false PV = 1 due to the so-
called oxygen error [16, p. 220]. However, dissolved oxygen is less reactive than the “active
oxygen” in hydroperoxides, and the “oxygen error” should be less than 0.5 PV unit if the
analysis is correctly executed.
2. Totox Value
A measure often used to express the total extent of oxidation of an oil is the Totox value,
which by definition is
Totox = 2 PV + AV
Although this empirical combination of chemical and physical analyses may seem strange, in
practice it has been found to be very useful for evaluating the effects of processing (including
storage and handling) on oxidation.
For pure thermal degradation of PV or physical separation of Totox, I have found that a
“Totox balance” usually works.
As an example of the use of the Totox value. Fig. 3 includes the Totox values correspond
ing to the PV and AV figures already reported by Mag [17, p. 112] to illustrate the effect of
clay (earth) dosage on the oxidation values of soybean oil. The Totox value provides a better
physical appreciation of Mag’s conclusion that “adsorption of secondary products is relatively
inefficient”; the adsorption capacity in this case was about 3.5 Totox per percent of clay. We
shall see further examples of the use of the Totox value later.
Strecker and his colleagues [5, p. 318] recognize the importance of PV and AV in com
(maize) oil processing but consider that “the significance of the AV [in the Totox value] must
be enhanced to properly express the oxidative state of an oil throughout the processing
180 Hilder
Earth Dosage (% )
Fig, 3 Oxidation values of bleached soybean oil.
scheme.” They proposed an alternative index, but their reasons for this conclusion are not
clear to me.
3. Oxygen C ontent
The dissolved oxygen content of an oil may eventually cause oxidation and therefore contri
butes to the potential oxidation of the oil. It can also be used to measure the extent of oxygen
absorption during processing. The oxygen content is not easy to measure, but various methods
have been developed in the past [15, p. 223]. For example, dissolved oxygen analysis has
been used to follow the rate of oxidation and the corresponding formation of volatile compo
nents [18]; Fedeli and Brillo [19] used an oxygen analyzer to measure the oxygen profile
during absorption in a column of olive oil.
Portable oxygen-in-oil analyzers are also available [e.g., 20] that have been successfully
used to compare and evaluate alternative oil storage and handling procedures [13,21]. These
analyzers usually measure a percentage saturation or partial pressure of dissolved oxygen,
which, if the solubility of oxygen is known, can be converted to oxygen content.
B. Solubility and Diffusivity

Oxidation takes place only with dissolved or absorbed oxygen. The solubility and rate of
diffusion of oxygen in the oil are therefore relevant in understanding the rate and extent of
oxidation. The solubility of nitrogen is also relevant because nitrogen is often used to protect
oils from oxidation.
A summary of approximate solubilities and the diffusivity of oxygen deduced from pub
lished data is given in Table 6.
The solubility of nitrogen in various oils has been determined between 25 and 150°C
[22,23]; there is no systematic effect of oil type, and other authors report similar results.
There is far more variation in the published data on the solubility of oxygen in oil, especially
with unsaturated oils and at temperatures above 60°C, where oxidation could affect the results.
The figures in Table 6 are realistic average values based on measurements with hardened lard
and olive oil in the range 25-85°C [22,23].
Oll storage, Transport, and Handling 181
Table 6 Solubility of Oxygen and Nitrogen and the Diffusivity of Oxygen in Oil
Oxygen Nitrogen Diffusivity of

Temp. at 1 bar at 1 bar Viscosity oxygen
(°C) (mg/kg) (mg/kg) (cP) (m^/s X 10^)
0 170 80 — ____
25 180 85 62 0 .8
50 185 90 24 1.3
75 190 95 12.5 1.9
100 200 105 7.5 2.5
125 — 110 — —
150 — 115 —
—
The solubility of gases usually obeys Henry’s law: the solubility is proportional to the
partial pressure of the gas. The solubility of oxygen from the air (21 v o l% oxygen) at 25°C
is therefore about 0.21 X 180 = 38 mg/kg oil. Berger [13,, p. 439] quotes 36 mg/kg for palm
olein at 60°C, and Fedeli and Brillo [19] imply a value of 39 mg/kg for olive oil at 20°C.
These solubility data will be seen to be useful when discussing the rate or extent of degra
dation.
Data on the diffusivity of oxygen in edible oils are surprisingly scarce. Crowe and cowork
ers [24] (working in the Water and Effluents Laboratory of a civil engineering department!)
reported diffusivities in olive oil of 0.87 x 10“^ m^/s at 25°C and 1.06 x 10"^ m^/s at 3TC.
They also measured the diffusion resistance of triolein and tributyrin. Schrader et al. [25]
determined the solubility and diffusivity in vegetable oil (probably soybean and/or sunflower)
to be 0.6 X 10“ ^ m^/s at 20°C. They also measured the solubility and diffusivity in butter,
mayonnaise, and two well-known brands of German margarine.
The specific diffusion resistance [ = 1/(diffusivity x solubility)] in olive oil and triolein is
about the same, as would be expected, but in tributyrin it is about 5 0 % lower, which suggests
that the solubility and/or diffusivity of oxygen in oils is dependent on the chain length of the
triglyceride, in a yet to be quantified manner.
The temperature dependence of the diffusivity of oxygen in oil also still needs be clarified.
Based on measurements of bimolecular collision frequencies of oxygen in olive oil in the
range of about 5 -6 0 °C , Subczynski and Hyde [27] concluded that the diffusivity should de
pend approximately on the square root of the value of the absolute temperature divided by the
viscosity. In the absence of any better evidence this relation has been used to estimate the
values in Table 6.
C. Absorption of Oxygen
Any oxygen dissolved or absorbed into the oil is a potential risk; it is the starting point for
any oxidation.
1. D issolved Oxygen Content

The dissolved gas content of oil saturated with air (25°C, 1 bar), calculated from the compo
nent solubilities, is shown in Table 7. The dissolved oxygen is sufficient to produce a potential
PV rise of about 2.5, and therefore the first precaution is to avoid contact with air wherever
possible. For example, according to Pardun [15, p. 223], the oxygen content of freshly filled
refined oil is indeed 36-38 mg/kg. Measurements of the dissolved oxygen content during
182 Hilder
Table 7 Gas Content of Oil Saturated with Air
Dissolved in oil
Solubility Air
Gas [mg/(kg.bar)] (voi %) mg/kg Nl/kg
Oxygen 180 21 38 0.026

Nitrogen 85 78 66 0.053
Argon 270^ 1 3 0 .0 0 2
Total O2 , N 2 , Ar 100 107 0.081
^From Ref. 23.
refining [13] show that similar values are approached during processing, in particular in stor
age; see Table 14 in Section V.
2. H eadspace Volume
The air headspace above oils, for example in closed tankcars or sample bottles, is potentially
even more damaging. The potential PV rise due to the oxygen in air in an enclosed headspace
depends on the degree of filling, as illustrated in Table 8.
3. Rate o f A bsorption
The rate of absorption is more difficult to assess. In general, the increase in average oxygen
content C in a volume V of oil will depend on (1) an absorption or mass transfer coefficient,
K, (2) the contact area between gas and oil, A, and (3) the concentration difference between
oxygen in the gas, expressed in terms of the equivalent solubility in oil (C J, and the oxygen
content in the bulk of the oil.
dC
V — =KA (C, -C)
The factors that would be expected to affect the rate of absorption, apart from the dissolved
oxygen content, are therefore (1) the surface area-to-volume ratio of the oil, (2) the oxygen
concentration (in the gas/headspace) above the oil, and (3) the effect of agitation (natural or
mechanical), especially near the oil surface, on the absorption coefficient.
4. Surface A rea-to-Volum e Ratio

If the rate of absorption is the determining factor, we would expect the increase of the oxygen
content of the oil to depend on the area-to-volume ratio (A/V). The area-to-volume ratio for
Table 8 Potential PV Rise Due to Air in an Enclosed Headspace
Container filling Headspace Potential PV rise

(vol % oil) (m^ air/m^ oil) (meq/kg)
99 0 .0 1 0 .2
90 0 .1 2
50 1 20
oil in any normal container, from sample bottle to storage tank, is also related to the filling
height H.
A_ 1
V~~H
The effect and importance of the area-to-volume ratio was quite clearly demonstrated by
Going [28], who measured the PV rise in neutral, bleached soybean oil after 5 weeks of
storage at 49°C in containers of various sizes. Similar results have been obtained in laboratory
tests [29] with fully refined soybean oil stored for 2 weeks at 23°C and 5 0% RH. Both sets
of results, with approximate contact area and filling height values, are summarized in Table
9. Similar effects have repeatedly been observed during thermogravimetric studies of oil oxi
dation [30,31].
The oxidation rate is not directly proportional to contact area. Factors other than absorption
play a role, and, as Going [28] pointed out, it is inadvisable to extrapolate these figures.
However, it is clear that contact area or filling height are very important. For example, oxida
tion rates in a sample bottle (filling height 10 cm) may be similar to those in 10 cm of oil in
a storage tank, which we would obviously not consider ideal storage conditions!
5. Oxygen Concentration in GasIHeadspace

The use of nitrogen to decrease the oxygen content in the “air” above oil will reduce both the
rate of absorption and the rate of the subsequent oxidation reaction. Going [28] clearly showed
the positive effect of nitrogen in storage. Neutralized bleached soybean oil (PV = 1 ) was
stored for 5 months, at 2 4 ± 8 ° C as 340 t lots in half-filled 750 m^ tanks. The PV of the oil
stored under air (21 vol % O2) increased by 4 units; the PV of the other half, stored under
nitrogen (1 .5 ± 0 .5 vol % O2), increased by only 0.5 unit. Both oils were subsequently lab-
deodorized, hydrogenated and lab-deodorized, and plant processed to give a hydrogenated
shortening; in all cases the flavor stability of the products made from the oil stored under
nitrogen was significantly better.
Similar results were obtained during extensive long-term storage trials with commercially
produced and packed cottonseed and soybean salad oils sponsored by the U.S. Defense De
partment [32]. Eleven different oils in bottles or cans filled under air or nitrogen were evalu
ated during 52 wk of storage in the dark at 26 and 38°C. The flavor of all the oils kept under
Table 9 Influence of Contact Area on the Rate of Oxidation
Final PV (meq/kg)
Contact Filling height
Container area (m^/m^) (mm) Ref. 28 Ref. 29
Drum (200 L) 1.2 850 28 —
Can (25 L) 3 350 39 —
Cylinder 8 120 — 11
Can (1 L) 13 75 55 —
Bottle 35 30 — 16
Petri dish 95 10 — 19
400 2.5 — 22
4000 0.25 — 44
184 Milder
nitrogen (zero O 2) was still good after 52 wk, and better than that of the oils kept under air
( 2 1 % O2) or nitrogen (with 2% O 2) for only about 5 or 15-30 wk, respectively.
The two examples given in detail show that
1. The PV of the oil kept under air increased to 6 after about 24 wk and then stayed
constant; the flavor development showed a similar trend.
2. There was no increase in PV in the oil kept under nitrogen (zero O 2), and the flavor
decreased only very gradually.
D. Hydroperoxide Formation
Hydroperoxide formation is a complex subject, and I do not intend to go into detail. Lundberg
edited a two-volume review in 1961 [33], and he presented a useful summary of the mecha
nisms [34] at about the same time.
Hydroperoxides are formed by an autocatalytic reaction between oxygen and unsaturated
oils involving free radicals. The reaction starts with the formation of free radicals (initiation).
Each free radical can combine with oxygen and an unsaturated group to produce a hydroperox
ide and another free radical, which starts a chain reaction (propagation). This continues until
the free radicals are destroyed, usually by combining to form nonreactive molecules (termina
tion). In practice this results in (1) a slow initial rate of reaction during the so-called induction
period, often followed suddenly by (2) a much faster rate of hydroperoxide formation. The
length of the induction period is often used as a measure of the basic stability of the oil.
Many factors that affect the reaction have a similar effect on both the induction period and
the subsequent, faster rate of reaction, but this is not always true.
1. Type o f Oil
The type of oil, in particular the degree of polyunsaturation, has an important effect. The
relative oxidation rates of fatty acids (Table 10) based on results at 2 0 -10 0 °C [4, p. 74] also
illustrate the effect of conjugation and isomeric form. Conjugated double bonds, which can be
detected by extinction measurements at 232 nm (dienes) and 268 nm (trienes), are particularly
susceptible to oxidation, and to make matters worse, they also tend to be formed as a result
of oxidation.
The induction periods Active Oxygen Method (AOM hours) of laboratory-extracted and
fully refined safflower and sunflower seed oils from plant breeding trials, with widely differing
18:1 and 18:2 contents [35], provide a good example. Purdy noted the dependence on 18:2
Table 10 Relative Oxidation Rates of Fatty Acids AOM- “Active

Oxygen Method”; office method of the AOCS
Fatty acid Isomer Conjugated Rate
18:0 (stearic) — — 1
18:1 (oleic) 1 X cis — 10
18:2 (linoleic) 2 X cis No 100
18:3
Linolenic 3 X cis No 200
Elaidolinolenic 3 X trans No 150
a-Eleostearic 2 X trans Yes 1000
jS-Eleostearic 3 X trans Yes 400
20:4 (arachidonic) 4 X cis No 400
Fig. 4 Effect of unsaturation on the rate of deterioration.
content, but if the reciprocal of the AOM hours is taken as a measure of the rate of deteriora
tion, the physical relationship is more obvious (Fig. 4). When the contribution of the 18:1
content is included as 7% of the rate of oxidation of 18:2 rather than the 10% indicated in
Table 10, the relationship appears to account for the total rate of oxidation.
The state of an oil also affects the potential for oxidation, as was illustrated by Going [28]
for soybean oil. After 7 wk storage of the oils at 38°C in 1 L cans, he noted a PV of about
50 for neutral bleached, 30 for neutral bleached hydrogenated (IV 25), and only 5 for crude
soybean oil.
2. Temperature
In practical terms the effect of temperature is even more important than that of the type of oil
because we can influence it. As with any chemical reaction, temperature has a very strong
effect on the rate of oxidation. Published data [4, p. 75; 36] show a doubling of the rate of
oxidation for every 10-15°C rise in temperature. This is well illustrated by figures quoted in
Ref. 37 for the influence of storage temperature on the time taken to develop off-flavor in a
fully refined soybean oil (Table 11).
Table 11 Effect of Temperature on Off-Flavor

Development in Fully Refined Soybean Oil
Storage temp. (°C) Time to off-flavor (h)
0 500
20 180
40 42
60 16
80 4
186 Milder
3. Oxygen Concentration
The concentration of oxygen in the oil also has a direct effect on the rate of hydroperoxide
formation and is again an effect we can influence. The overall, practical effect has already
been mentioned in Section IV.C.5.
In model studies the rate of oxidation was shown to be more or less proportional to the
pressure of oxygen at pressures below about 200 mbar [34], which is approximately the partial
pressure of oxygen in air. Another example was provided by Min and Wen [18], who mea
sured the decrease in dissolved oxygen in sealed oil samples. Analysis of their data shows
that the rate of decrease of dissolved oxygen is directly proportional to the dissolved oxygen
concentration.
4. L ight
Radiation in most forms, including sunlight and artificial light, accelerates oxidation. A whole
chapter could be written on this subject alone, but I will simply note that
1. Light promotes the formation of free radicals and therefore influences the initiation
of autoxidation.
2. Photo-oxidation occurs by a different mechanism involving singlet oxygen at rates
1000 or more times as fast as autoxidation; it is promoted by “sensitizers” such as
chlorophyll and inhibited by singlet oxygen “quenchers” such as carotene.
3. It is therefore advisable to store oils and fats in the dark.
4. Lea’s study on the effect of (artificial) light on the oxidation of animal fat [38] is of
interest not only for the elegant experimentation but also because it laid the foundation
for the PV analysis.
5. M etals
Some metals are pro-oxidants even at very low levels; the levels in Table 12 were found to
halve the shelf life of lard at 98°C [39]. The simplest explanation for the effect of these metals
is that they catalyze the formation of free radicals, for example via hydroperoxide breakdown
[34]. It is clear from Table 12 why copper in any form should be avoided and iron and nickel
should be removed during refining. Much lower levels are usually recommended to avoid pro
oxidant effects; for refined oils and fats, maximum levels of 0.01 ppm Cu, 0.1 ppm Fe, and
Table 12 Trace Metal Levels

at Which a Significant Effect on
Oxidation Occurs
Metal Level (ppm)
Copper 0.05
Manganese 0 .6
Iron 0 .6
Chromium 1 .2
Nickel 2 .2
Vanadium 3.0
Zinc 20
Aluminum 50
0.1 ppm Ni are typical requirements. This also explains the antioxidant effect of processing
aids such as citric and phosphoric acid, which chelate with or otherwise aid in the removal
of metals.
6. A ntioxidants
The study of antioxidants is also a major subject in its own right, which is made even more
complicated by the pro- and antioxidant effects and the synergistic effects of some of the
compounds involved. The natural antioxidants such as the tocopherols, carotene, lecithin, and
gossypol show that nature usually ensures a suitable level of antioxidant in the crude oil.
The simplest explanation for the effect of antioxidants is that they suppress or reduce the
formation of free radicals by promoting termination reactions. Most natural antioxidants are
also oxidized themselves; this loss of natural protection is one reason for the decrease in
refined oil stability often observed as a result of (feedstock) oil oxidation. Oxidation of the
pigments such as carotene and gossypol also leads to “color fixation,” which is very difficult
to remove in subsequent processing.
The best course of action is to try (where possible) to maintain the natural level of antioxi
dants by good oil processing practice. Many synthetic antioxidants have been proposed, pro
duced, and tested, and some have been accepted for use with edible oils [1, p. 61], for
example, for the protection of palm oil during transport. But this is time-consuming and
expensive. In addition to confirming their efficacy as antioxidants, they must be thoroughly
tested for product safety.
7. M oisture
The effect of moisture has always caused confusion. The consensus of opinion seems to be
that water activities of about 0 .5 ± 0 .2 (i.e., water contents of about 50% of the saturation
solubility in Table 3) are the best. At this level water tends to stabilize hydroperoxides and
facilitate free radical termination reactions. At lower levels this “hydration protection” effect
is negligible, leading to higher oxidation rates. At higher levels there is a risk of formation of
free fatty acids (which oxidize faster than glycerides), hydrolytic rancidity, enzymatic attack,
and leaching of iron. Whatever the cause, oxidation rates tend to be higher at moisture con
tents approaching or in excess of saturation.
E. Hydroperoxide Breakdown
The breakdown of hydroperoxides is, if anything, more complex than their formation. De
pending on the conditions, decomposition of hydroperoxides can produce a variety of com
pounds, including aldehydes, ketones, esters, alcohols, acids, hydrocarbons, and polymers.
Initial decomposition often involves fission at the hydroperoxide position, giving a low
molecular weight (relatively volatile) aldehyde and a high molecular weight (nonvolatile) alde-
hydic triglyceride residue. The anisidine value includes the contributions of both volatile and
nonvolatile aldehydes.
A useful summary of the main volatile compounds produced during oxidation of eight
types of oil was provided by Snijder et al. [40].
The hydroperoxides themselves are tasteless and odorless, but the short-chain aldehydes
are particularly obnoxious even at low concentrations and are organoleptically detectable at
levels on the order of 1 ppm in oil and 0.01 ppm in water. The formation of aldehydes
initially follows roughly the same trend as the formation of hydroperoxides, and they can
reach levels that are organoleptically detectable by the end of the induction period. Even
188 Hilder
Table 13 Analyses of Oxidized Soybean Oil Before and After Refining
Before refining After refining
Sample PV AV Totox PV AV Totox
0 7.5 7.5 0.1 8.1 8.3

5.0 9.5 19.5 0.1 8.9 9.1
10.7 10.4 31.8 0.1 13.8 14.0
24.6 9.2 58.4 0.1 29.9 30.1
50 9.2 109.2 1.1 49.2 51.4
105 12.2 222.2 1.0 104.2 106.2
without further formation, the decomposition of hydroperoxides will continue, and it is partic
ularly rapid at high temperature.
The aldehydes are an important cause of oxidative rancidity in oils, which explains the
interest in the edible oils industry for (1) peroxide and anisidine values, (2) induction periods,
(3) taste and flavor stability, and (4) the whole subject of oil oxidation.
1. E jfect o f Processing
It is not easy to generalize on the trend of PV, AV, and Totox during processing, especially
when storage is also involved. A decrease will occur during hydrogenation (PV and AV) and
interesterification (PV) owing to the specific reactions involved. The combined effect of other
processing steps often results in a net increase in Totox before deodorization. During deodor-
ization, all the hydroperoxides decompose (PV = 0) and the volatile AV components are
stripped out. The fully refined oil should contain only relatively stable, nonvolatile oxidation
products.
Any deterioration of the fully refined oil will therefore depend on renewed absorption of
oxygen, hydroperoxide formation, and decomposition. But the rate of deterioration is also
influenced by the oxidation history and the processing of the oil, and experience has shown
that the stability/keepability can be optimized by optimal processing (not overprocessing) but
that in the end the greater the extent of oxidation of the oil (as crude or during processing),
the worse the refined oil stability will be.
Pardun [15, p. 233], a chief chemist with a lifelong experience of oils and fats, stresses
the importance of Totox and the relation between the increase in AV during refining and the
decrease of the original PV in assessing the keepability of oils and fats. He illustrates this
with examples of the PV and AV figures of soybean oil (to which I have added Totox) given
in Table 13. These figures also illustrate the general experience that deodorization reduces
Totox by about 50%; the experience with com oil, for which a 45-50% reduction in AV
during deodorization is claimed [5, p. 139], seems to be different, unless the PV is already
low.
V. TRANSPORT AND STORAGE REQUIREMENTS

Transport and storage of oils are essential for the operations of any oil user whether he is
adjacent to a plantation, oil mill, extraction plant, or refinery or relies on these as his supplier
of cmde or refined oils. Both supplier and customer require efficient transport and storage at
an acceptable quality. Apart from normal economic considerations, the main concerns are
therefore related to contamination and degradation. By contamination I mean any form of

admixing with a foreign substance or another oil or fat during transport or storage or during
the associated handling before or after.
Transport is really no more than mobile storage. The risk of degradation is the same or
perhaps even greater due to the inherent movement and agitation. But there are two factors
that increase the chance of contamination during the transport stages. The first is that transport
is not usually under the direct control of either supplier or receiver. The second is that it
cannot be assumed that transport, especially by ship, is used for the transportation of edible
oils alone. This means that strict procedures and inspections are required at all stages to avoid
or identify any contamination.
As regards degradation, the basic principles of hydrolysis and oxidation provide an excel
lent guideline as to what and what not to do, both when deciding on the requirements and
when designing the facilities for oil storage, transport, and handling. The same principles
apply to the (crude or refined) oil before it comes under your control; how have your suppliers
handled the oil?
The type of oil will determine the degree of unsaturation or polyunsaturation and the nor
mal level of natural antioxidants, and these factors should be borne in mind when deciding
on the requirements for transport and storage. Most other factors affecting degradation are
more or less controllable:
1. Minimize contact with or absorption of air.
2. Avoid excessive moisture.
3. Avoid contact with and minimize content of pro-oxidants.
4. Keep oils in the dark.
5. Keep time and temperature to a minimum.
6. Avoid unnecessary agitation.
However, the precautions taken must also be economically justifiable, and they are usually
related to the commercial risk involved in not taking them.
The usual ranking of the oxidative stability of oils [1, p. 268] is
Crude > deodorized > neutral > bleached
but the ranking with respect to commercial risk closely follows the reverse order of pro
cessing:
Deodorized > bleached > neutral > crude
This is mainly because the further one gets, the less chance there is to rectify any degradation.
For example, the use of nitrogen to protect deodorized oil is quite common, but although the
positive effect of nitrogen in crude [5] and bleached [28] oil storage has been clearly demon
strated, it is rarely used.
Another basic factor that cannot be stressed enough is that every stage of transport or
storage involves an oil-handling step, or two oil-handling steps if the ideal alternative is
degradation-free transfer between processing stages. These oil-handling steps probably ac
count for a major part of the potential oxidative degradation. This is illustrated by the dis
solved oxygen contents during palm oleine processing reported by Berger [13], which are
summarized in Table 14. These figures are the average (and minimum-maximum) measure
ments on 12 separate days, or only 4 days for deodorized storage. As Berger,commented,
“Filling [of the bleached oil buffer and deodorized oil storage tank] involved splashing the oil
from the top of the tank and resulted in a substantial increase in oxygen level”; crude oil
storage was the next most critical stage.
190 Hilder
Table 14 Dissolved Oxygen During Processing
Process step Dissolved O2 (mg/kg) ^
Crude storage 15 (12-20)

Outlet bleacher 8 (5-18)
Bleached buffer 26 (18-33)
Outlet deaerator 12 (0.3-31)
Outlet deodorizer 8 (4-13)
Deodorized storage 31 (30-32)
^Range given in parentheses.
The requirements for transport, storage, and handling of crude and fully refined oils should
therefore be carefully considered.
In the following subsections on crude and fully refined oil, I have tried to avoid repetition,
and it should therefore be read as a whole. For example, the comments made about crude oil
transport and refined oil storage are also relevant to refined oil transport but are not repeated.
A. Crude Oil
Crude oil is relatively stable and is transported and stored in relatively large lots. The negative
effects of any poor practice will therefore tend to be diluted by the bulk of oil or take longer
to show themselves. However, all the evidence and indications we have show that any degra
dation of the crude oil results in higher processing costs and a poorer final product quality. It
is therefore worthwhile specifying crude oil transport and storage requirements properly,
rather than just waiting to see what happens.
L Transport
Crude oil is transported by ship, barge, rail, and road. Large quantities are transported by
sea. Laurie oils and palm oil products, for example, are typically transported in lots of any
where between 250 and 10,000 t in ships carrying a total of 25,000-30,000 t [41]. The
journey from Malaysia to northern Europe takes about a month.
We will therefore take marine transport as an example; the same principles apply to other
forms of transport.
The supplier or receiver (seller or buyer) will book space with a shipowner, and both will
appoint their own surveyor or superintendent to supervise the loading and/or unloading by the
crew and terminal personnel. With so many independent players, the rules of the game must
be clearly defined and agreed on to avoid mistakes being made, especially with regard to
contamination.
There are many national sets of rules, but most of the international trade in oils and fats
occurs under FOSFA (Federation of Oil Seeds and Fats Associations) contracts.
In the 1980s, requirements were reevaluated and updated. Berger [13,14] summarized
some examples of deterioration and presented the recommended practice for storage and trans
port of edible oils and fats proposed by the Palm Oil Research Institute of Malaysia (PORIM).
This document influenced FOSFA’s update of their codes of practice, which also suggested
stricter rules on adulteration and contamination [42]. Ludwiczak [41] probed the relation be
tween shipper and shipowner and gave an excellent step-by-step account of procedures from
the port of loading to the port of discharge.
It is impossible to go into details here, and I restrict myself to some examples illustrating
the principles behind good practice. A useful summary of the details is given in Fairer’s The
Shipment o f Edible Oils [43].
Contamination. Contamination from previous cargoes can be disastrous. Information on

at least the three previous cargoes should be checked. All three should preferably be on the
FOSFA “approved” list, and none should be on the FOSFA “banned” list. In intermediate
situations, the cargo should be segregated and analyzed. The minimum requirement is that the
last cargo is not on the “banned” list. If the second or third previous cargoes are on the
“banned” list, the cargo must be segregated and analyzed by a competent laboratory before
processing.
Contamination can also occur in transit or during handling (loading, unloading, etc.).
Berger [13] refers to a study by M. S. A. Kheiri of samples taken during unloading of fully
refined palm oil. During the first 25 s (about 1 t) he found about 0 .0 15 % impurities, 0.9%
moisture, and Lovibond red (5.25 in.) = 20-25, compared with the bulk of oil, which con
tained about 0.003% impurities, <0.04% moisture, and Lovibond red (5.25 in.) = 2.5.
Air Contact. Partially filled compartments should be avoided. The oil should be loaded
through a “drop line” and not allowed to free fall from deck level; the figures in Table 14 for
crude storage, bleached buffer and deodorized storage illustrate the effect of free fall into
a tank.
Temperature and Heating. The international standard is the lASC (International Associa
tion of Seed Crushers) handbook, which includes the PORIM recommendations [14], equiva
lent to
Unloading temperature: 5-15°C above the melting point or, for liquid vegetable oils, 15-
25 °C to ensure good pumpability and drainage
Storage and shipping temperature: 0 -5 °C above the melting point or, for liquid vegetable
oils, ambient temperature
This is a realistic compromise between ensuring that oils are pumpable and homogeneous for
unloading and minimizing storage temperatures and heating during transport.
The recommended maximum heating rate of 5°C per day for the last few days of the
journey is sufficient to reach unloading temperatures. However, Berger [13] suggests that
even then local overheating may cause deterioration; he gives an example of the PV of a
shipment of crude palm oil that increased by less than 1.5 units during the first 18 days at
about 40°C but then increased by more than 4 units while heating at 2 °C per day for the last
6 days. A maximum heating coil temperature would probably be a better criterion, and hot
water heating is definitely preferable to steam heating.
Metal Contact. Any copper-containing metal is totally unacceptable. Berger [13] quotes
the study by M. Pike of crude palm oil samples taken using a brass sampling device and then
stored at 50°C in glass bottles. An instantaneous sample (0.1 ppm Cu) had an AV of 6 and
had lost less than 10% of its carotene and tocopherol after 5 days. Samples left in the sampling
device for 5-20 min at 50°C showed a roughly exponential increase in copper content to 0 .2 -
1.4 ppm Cu; after 5 days these samples had an AV of 11-20 and had lost 75-90% and 9 4 -
100% of their carotene and tocopherol, respectively.
Ships’ tanks should preferably be of stainless steel or coated mild steel with stainless steel
heating coils [14,41]. Any coating should be food grade, and for effective coating the metal
surface must be properly prepared.
192 Hilder
Postscript. The above are just a few examples of transport requirements and the reasoning
behind them. A complete definition of good working practice inevitably contains a lot of
detailed instructions and requirements. For these details you should consult the cited literature
and the relevant codes of practice.
2. Storage
Crude oil is relatively stable and is stored in relatively large tanks. Both requirements and
good construction practice will therefore differ from those for final product tanks. Neverthe
less, it still makes good economic sense to bear the principles of oil degradation in mind.
Strecker et al. [5] provide a good example of the combined effect of storage time, tempera
ture, and air or nitrogen blanketing on the cost of degradation. They stored two different
crude com oils under various conditions and determined the extra earth (clay) dosage required
to achieve a final color of 2.0 red (1 in. cell) compared with the fresh oils, all on a laboratory
scale. The results are reproduced in Table 15 because they suggest some interesting relation
ships.
Apart from the usual increase in oxidation values with time, the results suggest that
1. The required earth (clay) dosage depends roughly linearly on the Totox of the cmde
oil.
2. The earth (clay) requirement for physical refining is more sensitive to oxidation of the
cmde oil than with alkali refining.
3. The decrease in tocopherol content (TOCO) is directly related to the increase in To
tox value.
We also see once again that storage under nitrogen or at low temperatures significantly re
duces the rate of deterioration and extra earth (clay) required.
Whatever the actual relationships, it is clear that oxidative degradation in storage results in
increased processing costs, and a loss of natural antioxidant that will probably be reflected in
the stability of the final product. These results suggest an extra earth (clay) requirement of
0.1-0.2% after 2 weeks of storage and a corresponding PV increase of about 0.5 unit. Al
though these were laboratory-scale results, the assumption of 0.1% extra bleaching earth con
sumption to cover the cost of oxidative degradation in storage and during processing is not un
realistic.
Some aspects of good cmde oil storage practice have already been covered directly or
indirectly in the section on transport. Some of the following points on storage are also relevant
to transport.
Table 15 Effect of Crude Oil Storage Conditions on Refinability
Physical refining Alkali refining
Storage conditions
Time, mo 0 2 2 0 3 3 6
Temp., °C — 30 30 — 10 50 50
Atm. — Air — Air Air Air
Quality
PV 0.5 1 .0 2.7 0.4 — 3.7 18.6
AV 2.7 3.0 3.7 3.5 — 3.6 4.4
Totox 3.7 5.0 9.1 4.3 — 1 1 .0 41.6
Toco (ppm) 1690 1500 1200 2000 —
1460 1160
Clay (%) 1 .0 1 .2 2 .0 0.5 0.5 1 .1 1.9
Contamination and Segregation. Crude oil tanks should be easily and completely draina
ble, not only for the removal of tank bottoms but also to minimize intermixing of different
lots of oil. Intermixing of old and new deliveries of crude oil or different types of oil is
seldom a good idea; the oils lose their identity and purity. Apart from the contamination issue,
it makes it difficult to keep track of the quality history of the oil, which can help in predicting
required or optimal processing conditions.
Where necessary, oils should be segregated according to type: vegetable, animal, lauric,
etc.
Air Contact. The size of a tank should be chosen in relation to the requirements. A tank
that is never more than 50% full obviously has an unnecessarily large surface/volume ratio
and headspace.
The method of filling will have a strong influence on the absorption and initial level of
oxygen in oil. A drop line or dip pipe to the bottom of the tank (with a siphon breaker if
necessary) can also be used for storage tanks. The use of bottom filling or even a floating
inlet that discharges just below the oil surface can also be considered [13,14].
Moisture and Impurities. The accepted levels of moisture and dirt for trading and good
merchantable quality are often very lenient. Crude oil with higher levels should be rejected.
Moisture levels lower than the saturation solubility at storage temperature should be preferred.
Any undissolved water, waxes, lecithin, and impurities will inevitably tend to settle and
form tank bottoms. Tank bottoms, especially when they are near heating coils, provide an
ideal environment for various degradation reactions and should therefore be regularly re
moved. These effects must also be allowed for in the design and operation of crude oil stor
age tanks.
The classical method for removing tank bottoms involves “sweeping” the bottom of the
tank at regular intervals. This involves labor, creates a waste disposal problem, and results in
an oil loss and a temporarily “unproductive” tank. Modem processing machines can cope with
a reasonable (preferably constant) level of moisture and impurities. Agitation of cmde oil
tanks, sufficient to maintain any potential sediment in suspension but not to promote oxida
tion, is therefore a viable alternative. This will minimize the tank bottoms problem and has
the added advantage of creating a constant quality feedstock.
Temperature and Heating. Cmde oils that are regularly used should be stored at about
10°C above their melting point. Liquid vegetable oils can usually be stored at ambient temper
ature in uninsulated tanks.
Heated tanks should always be insulated. This is not only economically sound but also
prevents the extra degradation due to excessive heating.
Hot water heating is definitely preferred for a number of reasons:
It is sufficient for the hardest fats (in insulated tanks).
It is easier to control than steam heating.
It prevents excessive coil surface temperatures.
It reduces heat losses from steam connections and appendages.
Even old steam-heated tanks can usually be easily modified for hot water heating if they
are insulated.
The whole tank should be insulated, including the roof. Uninsulated roofs will result in
condensation, corrosion, and the risk of additional iron (mst) and moisture in the oil. Even
insulation underneath the tank can be considered if feedstock fractionation is a problem.
Oils not immediately required for processing are sometimes stored without heating and
allowed to “set up” (solidify). The pros and cons of this practice should be carefully evalu
ated. The heat required for remelting and the extra degradation caused by the extreme heating
194 Hilder
conditions required will only outweigh the savings of normal heating and degradation if the
oil is to be stored cold for a long time.
Metal Contact. It is still common practice to use mild steel tanks and mild steel heating
coils. With liquid vegetable oils, a polymerized oil “coating” will gradually build up on the
tank surfaces, which will protect both the metal and the oil. It would be even better to prop
erly prepare and coat all the surfaces with vegetable oil when the tank is installed.
For oils and fats stored at higher temperatures the choice is more difficult. The higher the
temperature, the greater the chance of corrosion and iron pick-up, especially at high moisture
levels and with high FFA and high lauric oils. Under these conditions stainless steel heating
coils would seem to be a sensible choice. With palm oil deliveries, quality differences can be
detected that may be partly due to the tank material used by the supplier.
3. H andling
Some of the comments on transport and storage are related to handling; the following are
equally relevant.
Any form of air ingress via pumps, pipelines, and fittings or vortices in tanks should
be avoided.
Pipelines should be designed to be emptied and to allow gravity to help rather than hinder
drainage, whatever system of line clearing is used. Pigging systems are often recommended
[13,14], but with proper design other systems employing line blowing, drain pumps, and gas
lift-lines can be equally effective.
Steam blowing of lines can never be recommended because it only introduces unwanted
moisture. Liquid oils certainly do not require it. Transfer lines for higher melting point oils
and fats should be properly trace-heated; heated lines should always be insulated. With these
precautions and proper pipeline design, controlled air (or nitrogen) blowing combined where
necessary with gas lift-lines should be quite adequate. Trace heating should be turned on only
when it is needed, to prevent both unnecessary energy consumption and degradation of any
oil (film) left in the lines.
B. Fully Refined Oil

The requirements for the transport, storage, and handling of refined oils are of necessity more
stringent than those for crude oils. For the refiner, fully refined oil is the final product. There is
no further processing to “put things right,” but there is plenty of chance for things to go wrong.
1. Storage
With few exceptions, the principles and comments applying to crude oil also apply to refined
oil. Moisture and impurities should not be a problem.
Contamination and Segregation. Complete emptying and draining of refined oil storage
tanks is essential. The tanks should therefore be constructed with (insulated) conical or dished
bottoms. In general, mixing of old and new oil should be avoided, but there is an economic
risk vs. cost trade-off. For regular, fast-moving oils, topping up can be acceptable.
Segregation of refined oil types should be considered. Dedicated tanks may be required
for vegetable, animal, lauric, or hardened products. For example, contamination of dewaxed
sunflower oil, intended for bottling, with any hardened oil would be disastrous.
Air Contact. The results of improper handling of deodorized oil were given in Table 14.
A deodorizer is a very efficient degasser, and the oil leaving it should be oxygen-free. If the
oil contains oxygen, it has probably come from either air leaks in the discharge system or
contact with air during sampling.
Deodorized oil is like a sponge; it will “suck up” any air it comes in contact with. If the
degassed oil is top-filled into a storage tank containing air, the oil will be more or less satu
rated with air, 8 0 - 8 5 % saturated according to Table 14.
The preferred procedure is to sparge the oil with nitrogen before or during discharge and
to bottom-fill the storage tank. The nitrogen, typically 0.5-0.9 N m ^/t oil (although suppliers
may recommend more than 1.0 N m ^/t [2, p. 312]), is sparged into the oil under pressure.
According to Table 6, only 85 mg/kg (about 0.07 N m ^/t) will remain dissolved at atmospheric
pressure; the rest will come out of solution and help to flush the air out of the head-
space of the storage tank.
Considering the solubility of nitrogen, it is questionable whether such large volumes are
required for sparging. The nitrogen purity should also be carefully specified. If 0.5 Nm^/t
nitrogen of 99% purity (max. 1.0% oxygen) is sparged into oil at 4 bar and 50°C, only about
0.3 Nm^/ton (and 5 mg/kg oxygen) will dissolve. On discharge to 1 bar, the excess nitrogen
will flash off and strip out part of the oxygen, leaving, say, 2.5 mg/kg oxygen (and 0.07
N m ^/ton nitrogen) dissolved. This is equivalent to a potential PV increase of about 0.15 and
is only slightly more than the equilibrium content of 1.85 mg/kg with 99% nitrogen at 1 bar
and 50°C. For line blowing and blanketing, which affects only a small proportion of the oil,
nitrogen of 99% purity is certainly quite sufficient.
It is important to remember that sparging the oil with nitrogen does not protect it forever.
If the oil comes into contact with air again, oxygen will still be absorbed, giving a renewed
risk of oxidation.
Agitation of refined oil tanks should not be necessary and is not recommended.
Temperature and Heating. The same rules apply as for crude oil, except that a refined
oil should never be allowed to “set up” or be remelted. It is essential that the deodorized oil
be cooled to storage temperature before storage; otherwise the oil can start to deteriorate
before it has cooled down. This is particularly important for liquid vegetable oils, which
should never be stored at more than 30°C.
Metal Contact. The simplest and safest solution is to choose stainless steel or coated steel
tanks. However, economics still have to be considered. The minimum acceptable choice is
Mild steel for oils at <45°C
Stainless or coated tanks for oils and fats at >45°C
Example. The following example [44] illustrates the combined effect of some of these
parameters on the stability of deodorized sunflower oil.
Sunflower oil, deodorized under two different conditions A and B, was stored in open 1 L
cans. The results therefore represent an “accelerated oxidation test” rather than the actual
storage or shelf life of the oil.
In the first trial, the oil leaving the deodorizer was not sparged with nitrogen and was
stored at 40°C . In the second trial, the oil leaving the deodorizer was sparged with nitrogen
and was stored at 40°C and 2TC . The PV of the samples during 15 days storage is shown in
Fig. 5. The taste results were similar.
The effect of the processing conditions is less significant than the effect of the storage and
handling conditions.
1. Under the more stringent deodorization conditions, the “induction period” was about
1 day longer.
2. The influence of nitrogen sparging was negligible; this is not surprising, as the subse
quent storage was in contact with air.
3. The effect of storage temperature was the most significant. At 27°C the PV increase
and taste decrease were negligible; at 40°C they were unacceptable.
196 Hilder
24
A: o40°C A(N2)40°C +(N2)27°C
B: •4a°c A(N2)40X x (N2)27°C

20
m
16 A
UJ A •
LU
3
12 A •
> 2
LU
9
X
o
cc
UJ
CL
^ ®®
0 I—8- ■i.^ ................. ^..^ J
0 4 8 12 16
SAMPLE STORAGE (DAYS)
Fig. 5 Effect of processing and storage conditions on the stability of refined sunflower oil.
2. Transport
Transport of fully refined oil by ship, barge, road, or rail are all feasible. Road and rail
transport over up to 1000-2000 km lasting up to 2 weeks (with the necessary precautions!)
has been common practice in the United States for many years. The same principles apply as
for crude oil transport, but the actual procedures and precautions will be more stringent and
will depend on the exact application. The remarks on refined oil storage are also relevant,
since transport is also a form of storage. It is also worth remembering that problems are often
the result of a seemingly trivial oversight or mistake.
Since large amounts of refined oil travel by road, we will take road transport as an exam
ple, but the same principles apply to other forms of transport.
Contamination and Segregation. Dedicated transport is preferable. Otherwise, the tanker

should be cleaned and dried before filling each load.
Air Contact. Here again, partially filled tankers must be avoided (minimum headspace),
and bottom-filling either via a dip pipe or via the bottom outlet is essential.
Sparging the oil with nitrogen during filling is a sensible precaution, especially where long
transport distances or times or sensitive oils are involved. Nitrogen blanketing should then not
be necessary, except in extreme circumstances.
Temperature and Heating. The same temperatures apply as for storage. For short dis
tances or times, insulated and unheated tankers are adequate. If heating is required, external
coils are preferred; there is less risk of local overheating or contamination with the heating
medium, and it is easier to keep the tanker clean.
Metal Contact. Stainless steel should be considered mandatory for all forms of refined
oil transport except in exceptional cases. The requirements concerning material of construction
are in general more stringent for transport than for storage. This is understandable, consider
ing that storage is normally “dedicated” whereas transport tanks usually have to be cleaned
before each load. The combination of mild steel and cleaning increases the risk of oil degrada
tion. However, if liquid vegetable oils, for example, are to be transported in dedicated ships,
the “coating” of polymerized oil may provide sufficient protection.
3. H andling
Most aspects of handling relevant to fully refined oil have already been mentioned. As regards
line blowing, however, the clear choice is nitrogen, not air—and definitely not steam.
C. Intermediate Oil Storage

By now it will be obvious that (1) intermediate storage should be reduced to an absolute
minimum and (2) a minimum buffer between processing stages is the ideal situation. If inter
mediate storage is still required, the principles of oil degradation should be considered. The
storage (or transport) requirements will lie somewhere between those for crude oil and fully
refined oil. Although in-process buffer tanks are less critical owing to the limited residence
time of the oil, air contact, heating, etc. should still be carefully considered.
ACKNOWLEDGMENTS
My thanks go to Ir J. C. Segers, Ir A. Leniger, B. P. Backlog, Rosita van der Burg, Irma
Knieriem, Mirto Oduber, Hans Houtman, and Anne Hilder, for their help.
REFERENCES
1. H. B. W. Patterson, Handling and Storage of Oilseeds, Oils, Fats and Meal, Elsevier Science,
London, 1989.
2 . G. R. List and D. R. Erickson, Storage, handling and stabiliziation, in Handbook of Soy Oil
Processing and Utilisation (D. R. Erickson, E. H. Pryde, O. L. Brekke, T. L. Mounts, and
R. A. Falb, Eds.), American Soybean Association, St. Louis, MS, and AOCS, Champaign, IL,
1980, pp. 267-353.
3. G. R. List, Storage, handling and stabilization of edible fats and oils, in Bailey’s Industrial Oil
and Fat Products, Vol. 3, 4th ed. (T. H. Applewhite, Ed.), Wiley, New York, 1985, pp. 273-309.
4. D. Swem (Ed.), Bailey’s Industrial Oil and Fat Products, 3rd ed.. Interscience, New York, 1964.
5. L. Strecker, A. Maza, and G. Winnie, in World Conference Proceedings: Edible Fats and Oils
Processing (D. R. Erickson, Ed.), Am. Oil Chem. Soc., Champaign, IL, 1990, pp. 309-323.
6 . M. Loncin, Die spontane autokatalytische Spaltung der Triglyceride, Fette Seifen 55(1): 7 (1953).
7. M. Loncin, l’Hydrolyse spontanée autocatalytique des triglycérides. Oléagineux 7(12): 695 (1952).
8 . M. H. Hilder, The solubility of water in edible oils and fats, J. Am. Oil Chem. Soc. 45: 703
(1968).
9. M. H. Hilder, The solubility of water in edible oils and fats above 100°C, J. Am. Oil Chem. Soc.
48: 296 (1971).
10. M. H. Hilder and M. van den Tempel, Diffusivity of water in groundnut oil and paraffin oil, J.
Appl. Chem. Biotechnol. 21: 176 (1971).
11. M. Loncin, Die Löslichkeit des Wassers in Fetten und Ölen und der Dampfdruck des gelösten
Wassers, Fette, Seifen, Anstrichm. 57(6): 113 (1955).
12. D. S. Sarkadi, Hydrolysis during deodorisation of fatty oils; catalytic action of fatty acids, J. Am.
Oil Chem. Soc. 36: 143 (1953).
13. K. G. Berger, Quality control in storage and transport of edible oils, J. Am. Oil Chem Soc. 62:
438 (1985).
14. K. G. Berger, Recommended Practices for Storage and Transport of Edible Oils and Fats, Palm
Oil Res. Inst, of Malaysia, Malaysia, 1985.
198 Hilder
15. H. Pardun, Analyse der Nahrungsfette, Verlag Paul Parey, Berlin, 1976.
16. V. C. Mehlenberger, The Analysis of Fats and Oils, Gerrard, Champaign, IL, 1960.
17. T. K. Mag, Bleaching—theory and practice, in World Conference Proceedings: Edible Fats and
Oils Processing (D. R. Erickson, Ed.), AOCS, Champaign, IL, 1990, pp. 107-116.
18. D. B. Min and J. Wen, Effects of dissolved free oxygen on the volatile compounds of oil, J.
Food Sei. 48: 1429 (1983).
19. E. Fedeli and A. Brillo, Meccanismo dei fenomeni di autossidazione delle sostanze grasse, Riv.
Ital. Sostanze Grasse 52(3): 8 8 (1975).
20. Dissolved Oxygen in Oils and Fats, Brochure, Orbisphere Laboratories, A. Woerden, Holland.
21 . B. A. Bol, Loders Croklaan BV, Internal report, 1992.
22 . L. B. Parsons, Private communication, 1938, cited in Bailey's Industrial Oil and Fat Products,
3rd ed. (D. Swem, Ed.), Interscience, New York, 1964, p. 127.
23. R. Battino, F. D. Evans, and W. F. Danforth, The solubilities of seven gases in olive oil, J. Am.
Oil Chem. Soc. 45: 830 (1968).
24. M. Crowe, P. E. O’Connor, and B. Masterson, Oxygen diffusion in natural oils and pure triglycer
ides, Biochem. Soc. Trans. 75(2): 260 (1987).
25. U. Schrader, K. Becker, and R. Heiss, Der Einfluss von Diffusion und Löslichkeit auf die Reak
tion von Sauerstoff mit kompakten Lebensmittein, Verpack.-Rundsch. 31(6): 33 (1980).
26. R. Heiss, U. Schrader, and G. R. Reinelt, The influence of diffusion and solubility on the reaction
of oxygen in compact food, in Food Process Engineering, Vol. 1 (P. Linko, Y. Mälkki, and J.
Olkku, Eds.), Applied Science, London, 1980, pp. 364-370.
27. W. K. Subezynski and J. S. Hyde, Diffusion of oxygen in water and hydrocarbons using an
electron spin resonance spin-label technique, Biophys. J. 45: 743 (1984).
28. L. H. Going, Oxidative deterioration of partially processed soyabean oil, J. Am. Oil Chem. Soc.
45: 632 (1968).
29. J. Koch and J. Bittorf, Unilever Research Laboratory, Vlaardingen, Internal report, 1983.
30. H. S. Olcott and E. Einset, A weighing method for measuring the induction period of marine and
other oils, J. Am. Oil Chem. Soc. 35: 161 (1958).
31. H. Mikula and A. Khayat, Reaction conditions for measuring oxidative stability of oils by thermo-
gravimetric analysis, J. Am. Oil Chem. Soc. 62: 1694 (1985).
32. C. D. Evans, G. R. List, H. A. Moser, and J. C. Cowan, Long term storage of soyabean and
cottonseed salad oils, J. Am. Oil Chem. Soc. 50: 218 (1973).
33. W. O. Lundberg (Ed.), Autooxidation and Antioxidants, Interscience, New York, 1961.
34. W. O. Lundberg, in Symposium on Foods: Lipids and Their Oxidation (H. W. Schultz, Ed.), Avi,
Westport, CT, 1962, pp. 31-47.
35. R. H. Purdy, Oxidative stability of high oleic sunflower and safflower oils, J. Am. Oil Chem.
Soc. 62 : 523 (1985).
36. S. P. Kochhar, Influence of Temperature on Accelerated Stability Testing of Oils and Fats: A
Literature Survey. Sei. Tech. Surveys No. 161, BFMIRA, Leatherhead, UK, 1988.
37. Speciality Fat Technology: Quality I, Loders Croklaan BV, Wormerveer, The Netherlands.
38. C. H. Lea, The effect of light on the oxidation of fats, Proc. Roy. Soc. Lond. 108B: 175 (1931).
39. M. H. Gordon, The mechanism of antioxidant action in vitro, in Food Antioxidants (B. J. F.
Hudson, Ed.), Elsevier, London, 1990, pp. 1-18.
40. J. M. Snijder, E. N. Frankel, and E. Selke, Capillary gas chromatographic analyses of headspace
volatiles from vegetable oils, J. Am. Oil Chem. Soc. 62: 1675 (1985).
41. J. E. Ludwiczak, Marine transportation of edible oils and fats, in World Conference Proceedings:
Edible Fats and Oils Processing (D. R. Erickson, Ed.), AOCS, Champaign, IL, 1990, pp. 25-
27.
42. B. P. Backlog, Storage, handling and shipping practices: buyer/recipient viewpoint, in World
Conference Proceedings: Edible Fats and Oils Processing (D. R. Erickson, Ed.), AOCS, Cham
paign, IL, 1990, pp. 28-30.
43. K. T. H. Farrer, The Shipment of Edible Oils, IBC Tech. Services Ltd., London, 1990.
44. P. J. Curtis, Van Den Berghs & Jurgens Ltd., Internal report, 1991.
8________________
Fractionation
Ralph E. Timms
The Cottages, Swinderby, Lincoln, England
I. INTRODUCTION
In the industrial processing of oils and fats, the term fractionation without further qualification
usually refers to the process of fractional crystallization. In this process, the oil is first crystal
lized and then the crystals are separated from the remaining liquid by a variety of means.
Fractionation separates the various triglycerides into one or more fractions by using the differ
ent solubilities of triglycerides, which depend on their molecular weights and degree of unsat
uration.
Fractionation may also be effected by other methods such as molecular distillation, solvent
extraction, or cyclodextrin complexing. However, none of these alternative processes is eco
nomical for large-scale processing of oils and fats; nor do these processes offer the same
separation efficiency as fractionation crystallization. In recent years, much interest has also
been shown in the use of supercritical or liquid carbon dioxide for the fractionation of oils
and fats [1]. Although some interesting results have been obtained, it is unlikely that this
process could replace fractionation crystallization because of its higher costs and no real ad
vantage in separation efficiency. Where these alternative processes can have an advantage is
in the selective removal of minor, nontriglyceride components, e.g., cholesterol [2], jS-caro-
tene [3], or free fatty acids [4], or the enrichment of very unsaturated triglycerides in liquid
oils where crystallization would require very low temperatures [5].
Fractionation of fats is carried out for four reasons:
1. The removal of minor components detrimental to the application of the oil, e.g., the
dewaxing of sunflower oil
2. The enrichment of a desirable triglyceride, e.g., POP in palm oil
3. The separation into two or more fractions of wider application and hence greater value
than the parent fat, e.g., fractionation of beef tallow, palmkemel oil, or palm oil to
liquid (olein) and solid (stearin) fractions
199
200 Timms
MOLECULAR LEVEL
nucléation
growth
phase behavior
(solid solutions,
polymorphism)
PHYSICAL LEVEL
removal of heat
agitation
PARAMETERS
crystal size
crystal habit
oil viscosity
pressure
Fig. 1 Schematic diagram of fractionation process indicating factors that are important at each stage.
4. As an alternative to hydrogenation, e.g., in the fractionation of palm oil to yield

fractions that can replace hydrogenated soybean or rapeseed oils in margarine hard-
stocks
Several of these applications are considered in detail later.

The process of fractionation consists of two steps:
1. Crystallization to produce solid crystals in a liquid matrix

2. Separation of the crystals from the liquid matrix
Different factors are important at each stage as shown in Fig. 1, and these are considered in
the next section. At this point it is useful to note that the quality of the liquid fraction depends
only on the crystallization step, whereas the quality of the hard fraction depends on both the
crystallization and separation steps. By “quality” is meant the degree of concentration of the
desired triglycerides in the separated fraction. Quality is usually assessed by physical criteria
such as cloud point or solid fat content.
The crystallization step is usually crystallization of the melted fat (crystallization from the
melt), but crystallization of the fat dissolved in a solvent is also practiced industrially. These
processes are commonly referred to as dry and wet fractionation.
The economics of fractionation are not considered in this chapter. Some useful information
is given in a recent review by Hamm [6].
II. PRINCIPLES
A. Phase Behavior and Solubility
When a liquid fat is cooled, a solid phase separates whose composition and amount depend
principally on the temperature. The situation is illustrated in Fig. 2, which shows the phase
diagram of a binary mixture of triglycerides A and B that form a continuous solid solution.
Fractionation 201
Fig. 2 Schematic phase diagram of a binary mixture of triglycerides A and B showing a continuous
solid solution.
i.e., they are completely miscible in the solid state. Holding the mixture at temperature Tj
results in the formation of a solid phase (crystals) of composition c in a liquid of composition
a. The fraction of solid phase is ab/ac).
The phase behavior of fats and the principles of phase diagrams have been reviewed [7].
Where solid solutions form, as is always the case in real fats, both the composition and the
amount of crystals vary with the temperature of crystallization. This is a quite separate effect
from polymorphism, which is considered later.
Another way of looking at the same situation is to plot a solubility curve as shown in Fig.
3. Conventionally, temperature now varies along the x axis, and the concentration of the
higher melting component or solute along the y axis. A solution whose concentration at a
given temperature is exactly that given by the solubility curve is said to be a saturated so
lution.
Temperature (®C ) —►
Fig. 3 Solubility of tristearin (SSS) and tripalmitin (PPP) in triolein, trilinolein, and paraffin oil.
(Redrawn from Ref. 37.) (O) SSS in triolein; ( • ) SSS in trilinolein; (A) SSS in paraffin oil; (□ ) PPP
in triolein.
202 Timms
Fig. 4 Phase diagram of mixtures of tristearin (SSS) in triolein (OOO). (Redrawn from Lutton as
given by Rossell [38] with solidus line added.)
Consider the solubility of tripalmitin and tristearin as shown in Fig. 3. It can be seen that
the solubility is independent of the liquid oil or solvent used. The solubility depends only on
the solute, in this case, tripalmitin or tristearin. This is an example of ideal solubility, and
such solutions are called ideal solutions. Ideal solubility is observed when solute and solvent
do not interact, i.e., no solid solutions are formed. The phase diagram then looks different
from Fig. 2, as shown in Fig. 4.
Fig. 5 Solubility of tripalmitin (PPP) in 2-oleodipalmitin (POP).

Fractionation 203
Ideal solubility can be described and defined by the equation

1 /r^ -i/r
lnx = M ^
R
where = molar heat of melting/crystallization, = melting point of solute, and x = solu
bility at temperature T expressed as the mole fraction. This equation was used to draw the
lines in Fig. 3, and it can be seen that it describes the experimental results closely.
In practice, the triglycerides in real fats do interact to form solid solutions, and this affects
the solubility curve. Tripalmitin (PPP) and 2-oleo-dipalmitin (POP), two of the major triglyc
erides in palm oil, do form a solid solution, and the effect on the solubility curve is shown in
Fig. 5. It can be seen that at all temperatures the observed solubility is higher than the ideal
solubility predicted by the equation. The solid phase that crystallizes out is now not pure
tripalmitin but a mixture of PPP and POP. This effect of increased solubility is what is meant
when one fat or triglyceride is described as “solubilizing” another.
So far, we have considered a triglyceride as having a single melting point and a single
solubility. However, fats and triglycerides exhibit polymorphism and occur in any one of
three basic polymorphs designated a (alpha), j8' (beta prime), and p (beta), a is the least
stable, most loosely packed, lowest melting, and most soluble; ¡3 is the most stable, most
densely packed, highest melting, and least soluble. Fats and triglycerides commonly crystal
lize in the a polymorph and then quickly transform to a ¡3' or ¡3 polymorph.
B. Supersaturation
To obtain crystallization it is necessary to increase the concentration of the triglycerides to be
crystallized above the saturated solution concentration at a given temperature. In practice this
is not sufficient to cause crystallization, and solutions can exist indefinitely with concentra
tions above the saturation level without forming any crystals. Such solutions are said to be
supersaturated.
Fig. 6 Saturation-supersaturation diagram for crystallization of partially hardened soybean oil.

Effect of cooling rate on metastable/unstable boundary (dashed line). Numbers are rates of cooling
(°F/min) at an agitation speed of 120 rpm. (Adapted from Ref. 33.)
204 Timms
For any system we can draw a saturation/supersaturation diagram as shown schematically

in Fig. 6 for the crystallization of partially hardened soybean oil. The continuous line is the
normal solubility or saturation curve. Below this line, crystallization is impossible because
the solution is not saturated, and the situation is stable indefinitely. The dashed lines divide a
metastable zone from an unstable or crystallization zone. In the metastable zone, crystalliza
tion is possible but will not occur spontaneously or immediately without assistance such as
stirring or seeding. In the unstable zone, crystallization will occur spontaneously and immedi
ately. It can be seen that the position of the dashed line boundary between the metastable and
unstable zones is variable and depends on process variables such as cooling rate and agitation.
The position of the continuous line boundary between the stable and metastable zones depends
on thermodynamic factors, whereas the position of the dashed line boundary depends on
kinetic factors.
The reason for the existence of the metastable zone can be understood if crystallization is
treated as a two-step process: nucléation followed by crystal growth.
C. Nucléation
A crystal nucleus is the smallest crystal that can exist in a solution of a certain concentration
and temperature. Aggregates of molecules smaller than a nucleus are called embryos and will
redissolve if formed [8].
When molecules come together to form a crystal there are two opposing forces. First,
energy is evolved due to the heat of crystallization, which favors the process. Second, the
surface of the crystal increases as the molecules aggregate. Just as when a balloon is blown
up, increasing the surface requires energy to overcome the surface tension or pressure, a
stable crystal will form only when the energy due to the heat of crystallization exceeds that
required to overcome the surface energy. Since the surface energy is proportional to the sur
face area and hence to size to the second power (size^) and the heat of crystallization is
proportional to the volume and hence to size to the third power (size^), it is clear that the
solubility must depend on the size of the crystal. (Size is the linear dimension of the crystal,
e.g., the diameter for spherical crystals.) Using the equation derived by van den Tempel [9],
data for a typical triglyceride are shown in Table 1. Supercooling is the decrease in tempera
ture below the solubility temperature that is required to get small crystals (in a supersaturated
solution) to crystallize. The critical size is the minimum size of a crystal that is stable at the
prevailing temperature.
In practice, spontaneous or homogeneous nucléation rarely occurs in fats [8,9]. Instead,
heterogeneous nucléation takes place on solid particles such as dust, walls of the container.
Table 1 Variation of Solubility and Supercooling with Radius

of Crystals of a Triglyceride
Radius of crystal
Supercooling Increase in
/xM Â (°C) solubility
10 100000 0.004 1 .0 0 1
1 10000 0.036 1.007
0 .1 1000 0.36 1.1
0 .0 1 100 3.6 2 .1
0 .0 0 1 10 7.2 1380
Source: Ref. 10.

Fractionation 205
number of nuclei per s per
Fig. 7 Primary nucléation rate of triunsaturated triglyceride /3' and ¡3 crystals in refined palm oil as a
function of temperature. (From Ref, 11.)
or foreign molecules. Crystallization at a wall can lower the surface energy barrier to crystalli
zation. The smaller the contact angle between the crystallizing fat and the surface on which
nucléation is to take place, the easier nucléation becomes.
Once crystals have formed due to primary nucléation, secondary nucléation can occur.
Secondary nuclei form whenever small pieces of crystal are removed from the growing crystal
surface. If the pieces are smaller than the critical size, they redissolve; if larger, they act as
nuclei and grow to become crystals. Secondary nucléation is undesirable in fractionation.
Stirring is the primary cause, and stirring is usually kept to the minimum required to facilitate
heat transfer.
As we have seen, a fat may crystallize in different polymorphs that have different stabilities
and solubilities. The least stable a polymorph has the lowest surface tension as well as the
lowest heat of crystallization. A small difference in surface tension can result in a large
difference in nucléation rate in favor of the unstable phase [8], despite the fact that the stable
phase is less soluble and hence more supersaturated. The result is that the nucléation rate of
fat crystals is in the order a > P '> (3 . In Fig. 7 a large difference in nucléation rates is
observed for the crystallization of the trisaturated triglycerides from palm oil in both the ¡3'
and j8 polymorphs [11]. The difference in nucléation rates between the two polymorphs indi
cates why rapid cooling of a fat leads to unstable a or ¡3' crystals, even though the more
stable j8 polymorph is favored thermodynamically because of its lower solubility.
D. Growth
Once a crystal nucleus has formed it will start growing by the incorporation of other mole
cules. These molecules are taken from the adjacent liquid layer, which is replenished continu
ously from the surrounding, supersaturated, liquid by diffusion. The rate-determining step is
incorporation of the new molecule in the correct configuration at the correct place on the
growing crystal surface. The growth rate is proportional to the amount of supercooling and
206 Timms
growth rale(pm/h)
Fig. 8 Growth rate of triunsaturated triglyceride p' and p crystals in refined palm oil as a function of
temperature. (From Ref. 11.)
inversely proportional to the viscosity, which affects the rate of diffusion. In crystallization
from the melt, viscosity increases as the temperature falls and the growth rate can go through
a maximum and then decrease with increasing supersaturation.
If a small degree of supercooling is applied, incorporation will take place only if the new
molecule is in exactly the right configuration, because the molecule will have time to orient
itself correctly. Thus lower supercooling tends to produce crystals that are more perfect. At
higher degrees of supercooling, molecules are attached to the crystal face at a faster rate. A
new molecule may become attached when its neighbor is not in a perfect position, and disloca
tion to the regular crystal packing may occur. As a result, perfect crystals are difficult to
obtain at high degrees of supercooling, and different triglycerides can cocrystallize if they are
comparable in chain length and melting point. Mixed crystals and unstable solid solutions are
formed. This conclusion was confirmed by Dafler [12], who used X-ray diffraction to show
that when tristearin was crystallized from the melt the crystals formed at 37°C were less
perfect than those crystallized at 52°C under similar conditions.
When growth is underway, there is a substantial evolution of heat. In particular, in the
absence of stirring, local temperature rises can be substantial. As a result, the volume adjacent
to a growing crystal surface may cease to be saturated and/or existing nuclei may redissolve,
because the critical size has increased with the rise in temperature. The nucléation and growth
sequence may become erratic, leading to imperfect crystals of variable size. In the industrial
fractionation of fats it is an important aim of the process to limit this exotherm by varying the
rate of heat removal and agitation with temperature.
As with nucléation, growth rates depend on the particular polymorph crystallized. Since
growth rate is proportional to supersaturation and the more stable forms are always the more
supersaturated at any given temperature, it follows that the more stable form has the higher
growth rate. This has been observed in practice for the crystallization of the trisaturated tri
glyceride fraction from palm oil as shown in Fig. 8.
III. PRACTICE
In practice, nucléation and growth do not occur sequentially and separately. Growth of the
first nuclei will start while the temperature is still falling, thus inducing further nucléation and
growth of these new nuclei. Nevertheless, the aim should be to minimize further nucléation
Fractionation 207
and instead grow the existing crystals to a large uniform size, enabling easy separation from
the liquid.
A. Crystallization
The fat to be crystallized is first heated above its melting point to destroy all crystal embryos.
Usually, a temperature 20°C above the melting point is satisfactory. The melted fat is then
cooled until crystallization starts. Until that point the cooling medium removes only sensible
heat from the thermal mass. After crystallization starts, the cooling medium must also remove
the latent heat of crystallization, which is much greater than the sensible heat.
The rate of cooling and crystallization are determined by the commercial need to expedite
the process and the thermodynamic need to keep in step with the rate of nucleation and
growth. Because the rate of nucleation increases so much more rapidly than the rate of growth
(Figs. 7 and 8), rapid cooling leads to large numbers of nuclei and hence a lot of small,
imperfect crystals. These crystals are then difficult to separate from the liquid, leading to a
hard fraction of overall poor quality.
Crystallizers are therefore designed to provide for slow rates of cooling during crystalliza
tion; l-3°C /h is typical. For small-scale crystallizers, a simple cooling jacket surrounding the
crystallizer can provide adequate cooling capacity. The problem in the design of larger scale
industrial crystallizers (5-50 t) is the need to ensure adequate cooling surface and heat trans
fer without excessive agitation, which would induce secondary nucleation. Clearly, the vol
ume of a crystallizer vessel increases as the cube of its linear size, whereas the surface area
of the vessel increases only as the square of the linear size. There comes a point in increas
ing the size of crystallizers when additional cooling surface has to be provided [13]. This
additional surface can take the form of internal coils, plates, or a cooled stirrer as shown in
Fig. 9.
An alternative to increasing the cooling surface is to lower the temperature of the cooling
medium, thus increasing the differential between the temperature of the oil and that of the
cooling medium. The disadvantage of this is that crystals formed at the cold surface are far
from equilibrium with the bulk oil. They will contain too high a proportion of lower melting
triglycerides. For a good quality fractionation they must eventually equilibrate with the bulk
oil. Since the attainment of solid-liquid equilibrium is slow, it is unlikely to be achieved in
practice. Additionally, the crystals formed will tend to stick to the cooling surface, thus fur
ther impeding heat transfer.
A practical crystallizer of optimum design will have the following features:
Adequate cooling surface (at least 2 m^/m? vessel; 3 -4 m^/m^ is typical).

Small temperature differential between cooling medium and oil, e.g., maximum 3°C, pref
erably 1°C, although this can be higher in the initial cooling period before crystallization
starts, and it is certainly useful to be able to vary the temperature differential systemati
cally.
Gentle but effective agitation to assist heat transfer and maintenance of a uniform tempera
ture while avoiding damage to the crystals.
Slow cooling to ensure that the crystallization takes place under conditions as close as
possible to equilibrium. Crystallization times can be 10-30 h in a typical fractionation
process.
The overall aim of the crystallization step is to produce large, uniform, perfect crystals in
equilibrium with the liquid. If adequate time is allowed for the crystals to grow, not only will
208 Timms
Fig. 9 Schematic representation of the most common types of crystallizer vessels used in dry fraction
ation. (From Ref. 13.)
the molecular level factors of crystallization (Fig. 1) have been optimized, but so also will
the crystal parameters of size and habit, which have a major effect on the efficiency of the
separation process.
B. Separation
To complete the fractionation process, the desired solid triglycerides in the crystals need to
be separated from the triglycerides that are liquid at the temperature of crystallization. These
liquid triglycerides are distributed in three locations: (1) in the uncrystallized bulk oil, (2) in
the uncrystallized oil that is physically trapped or entrained in the crystals, and (3) in solid
solution with the solid triglycerides.
The extent and type of solid solutions formed depend primarily on the fundamental phase
behavior of the fat being crystallized. High degrees of supercooling tend to increase the extent
of formation of solid solutions. (At the necessarily lower temperature a greater proportion of
liquid triglycerides may take part in the solid solution.) As we saw earlier, increased solid
solution formation leads to increased solubility, so in practice a higher yield of lower quality
Fractionation 209
(a)
Fig. 10 Schematic representation of the two most common types of vacuum filters used in dry fraction
ation. (a) Rotary drum filter; (b) vacuum belt filter. (From Ref. 13.)
crystals is produced. However achieved, once the solid solution has formed in the crystals,
the separation step can do nothing to change the composition of the solid phase of the crystals.
The uncrystallized bulk oil is relatively easily removed from the crystals. The entrained oil is
more difficult to remove. The separation step must address the problem of reducing the level of
entrained oil in the final solid fraction. Three methods have been developed to achieve this.
1. Vacuum Filtration
Two types of vacuum filters are in use—the rotary drum and the belt filter—as shown in Fig.
10. Such filters operate in two stages. In the first stage the crystallized bulk oil is removed
and the crystal cake builds up. In the second stage the residual liquid is sucked out and air or
nitrogen is sucked through the cake. It is in this second stage that the entrained oil is reduced.
Although more complex and expensive, the belt filter has the advantage that more time can
be given to this second stage and the differential pressure can vary along the belt. As a result,
the belt filter yields a solid cake fraction with a lower level of entrainment than the drum
filter. Comparative results are given in Table 2.
2. Centrifugal Separation
Because of the viscosity of the uncrystallized oil at the crystallization/separation temperature
and the large amount of solids to be separated, industrial centrifugal separators cannot be used
to efficiently separate crystals from the bulk oil. A solution to this problem is to use a surface-
210 Timms
Table 2 Estimates of Entrained Oil in Stearin from Different

Fractionation Processes
Entrained oil
Yield (% of
Company Process (%) stearin)
Alfa Laval Detergent (Lipofrac) 17-23 35-52

Tirtiaux Dry + belt filter 28-33 60-67
De Smet Dry + drum filter 37-40 70-73
Bemadini Hexane 37-40 70-73
De Smet Dry + membrane filter 20-21 47-50
Source: Ref. 14.
active agent (detergent) to “wet” the crystals so they are transferred into an aqueous phase.
Separation of a light oil phase and a heavy aqueous phase containing the crystals is then easy.
The concept was patented in 1905 by Fratelli Lanza and then developed as a proprietary
process using centrifugal separation (the Lipofrac Process) by the Alfa-Laval Co. The Lanza,
Lipofrac, or detergent process enables more entrained oil to be removed than can be achieved
using vacuum filtration (Table 2). Until the development of membrane pressing, the Lipofrac
process was a popular alternative to vacuum filtration. The higher costs of the process were
outweighed by the higher yield of liquid fraction and the better quality of the solid fraction.
The process is now less favored on environmental grounds related to the disposal of the spent
detergent solution containing surface-active agent and lost oil.
3. P ressing
Vacuum filters are necessarily limited to a pressure on the crystals of less than 1 bar. If the
pressure is increased, it is possible to squeeze out more of the entrained oil. As noted in the
Introduction, improved separation improves the yield but not the quality of the liquid fraction,
whereas for the solid fraction the separation step has a decisive effect on quality.
The mechanical or hydraulic pressing process was therefore developed particularly for the
production of palmkemel stearin, where the quality of the stearin determines its utility and
value as a replacement for cocoa butter. Conventional vertical hydraulic presses capable of
100 bar pressure have been used for many years.* The process is applied in practice by
crystallizing the fat statically in blocks, e.g. in trays, in a cold room with cold air circulating
around the blocks to cool the fat. The crystallized blocks of fat are then wrapped in filter
cloths of a mesh size able to retain the crystals while allowing the entrained oil to flow out,
and hydraulic or mechanical pressing is applied. The practical limit on pressure is determined
by the point at which no further oil can be squeezed out and/or the solid crystals begin to
extrude through the cloth. Such a process is labor-intensive and, because of the need to
surround the crystals completely with the cloth, it is feasible only for solid crystalline masses
with at least —25% solid fat content.
If a crystallized slurry of the consistency produced in a stirred crystallizer is to be
pressed, a totally different design of press is required. Such presses, membrane filter presses,
were developed in the early 1980s for the fractionation of palm oil. Their main features
are that:
* Palmkemel can be used as one, or two separate words in the literature. It appears here as one term, and should not
be confused with palm oil.
Fractionation 211
FILLING SQUEEZING
Fig. 11 Operation of filtration sequence in membrane press. (From Ref. 13.)
1. A crystal slurry may be pumped into the press chambers.

2. Pressure may be applied to the cake in the press chambers.
3. Squeeze pressures of 6 bar (and more recently 15 bar) are easily achieved.
4. The filling, pressing, and emptying of the press are automated, making the process
relatively cheap to operate.
The design and operation of a membrane filter press are shown schematically in Fig. 11.
As a result of the development of the membrane filter press, entrainment levels as low as in
the Lipofrac process can be achieved at lower capital, running, and environmental costs. This
separation method is now the overwhelmingly preferred choice for new fractionation plants.
More recent developments reported by Willner and Weber [15] have shown the dramatic
effect on entrainment levels and hence on the quality of palm midfractions that can be
achieved by increasing the pressure up to 50 bar (Fig. 12). However, it does not look as if
Fig. 12 Pressing of soft palm midfraction (mid-POL) to produce palm midfraction (PMF) suitable for
CBE production. Effect of pressure on iodine value (iv) of PMF. (From Ref. 15.)
212 Timms
pressures of 100 bar as achieved in vertical hydraulic presses will prove feasible due to the
limitations of the flexible pressure membrane and filter materials.
C. Multistage Dry Fractionation

A single fractionation consists of crystallization and separation steps. In many cases more
than one fractionation is carried out to give a multistage fractionation process. Multistage
fractionation processes are used where the desired fraction is a middle-melting fraction or
where improved purity of the fraction is required—greater than can be obtained in a single-
stage fractionation. Such processes are illustrated in Section IV.
D. Solvent Fractionation
If the fat to be crystallized is dissolved in an organic solvent, the fractionation process can be
substantially improved. There are several benefits of using a solvent.
1. Nucléation and growth are faster, so higher degrees of supercooling and faster rates
of cooling can be used. This favors short crystallization times (less than 1 h) and
continuous crystallizers.
2. Lower viscosity of the liquid leads to easier filtration.
3. Dilution of the fat makes heat transfer easier and the amount of oil in the entrained
liquid smaller.
4. The ability to wash the cake with fresh solvent leads to very low levels (less than
10 %) of entrained oil.
Unfortunately, the extra costs of solvent fractionation compared to dry fractionation have
made the process uneconomical except for the production of specialized fractions to replace
cocoa butter [16].
The choice of solvent has been discussed by Hamm [16] and by Timms [17]. The selectiv
ity of triglyceride separation in the crystallization step is little affected by the choice of sol
vent, and indeed solvent fractionation has no advantage over dry fractionation here. The
choice of solvent does affect the selectivity of separation between lipid classes, particularly
the separation of triglycerides from diglycerides, free fatty acids, and other lipids more polar
than triglycerides. Of the two commonly used solvents, acetone and industrial hexane, the
more polar acetone effects a better separation of triglycerides from diglycerides and free fatty
acids than does the nonpolar hexane.
E. Countercurrent Fractionation
Although countercurrent crystallization is commonly applied in the chemical industry, it has
rarely been applied to the fractionation of oils and fats. This is attributable to the long crys
tallization times required and to the relatively inefficient solid-liquid separation methods
available. With the introduction of membrane filter presses, countercurrent fractionation be
comes feasible; the possibilities have been discussed by van den Kommer and Keule-
mans [18].
In Fig. 13 is shown a hypothetical phase diagram for the first four fractionations of a two-
stage countercurrent process. Feed of composition FI is fractionated at temperature Tj to give
an olein 01 and stearin SI. Note that 01 and SI are different from Oj and Sj, the equilibrium
compositions, due to the system not having reached equilibrium (for the olein) and the con
tamination of the stearin with entrained olein.
Olein 01 then becomes feed F2 and is refractionated at temperature T2 to give olein 02
and stearin S2. Stearin S2 is then returned to the previous stage and mixed with FI to give a
Fractionation 213
Fig. 13 Schematic hypothetical phase diagram showing the first four fractionations of a countercurrent
process. (Adapted from Ref. 18.)
new feed F3, which is fractionated at The resulting stearin S3 is enriched in component
B (compared with SI). Olein 03 is returned to the previous stage and mixed with feed F2 to
give a new feed F4, which is fractionated at T4 to give an olein 0 4 enriched in component A.
At least three cycles are required to achieve a steady state, and the process can be time-
consuming to establish, so in practice only a two-stage or perhaps three-stage countercurrent
process could be commercially practicable.
The advantage of a two-stage countercurrent process over a two-stage consecutive fraction
ation process is shown in Fig. 14 for the fractionation of palm oil. The two stearins from the
47.5/26.4 39.1/33.0 41.5/30.8
Fig. 14 Two-stage countercurrent fractionation (A) compared to a two-step process with the same
olein composition (B) or same stearin yield (C), respectively. The numbers refer to SSS/SSO concentra
tions . (From Ref. 18.)
214 Timms
consecutive fractionations are combined so that both processes result in only two output
streams, one liquid and one solid. There is little difference in the compositions of the oleins
but a large difference in the compositions of the stearins.
F, Process Economics
The commercial viability of a fractionation process depends on
The value of the raw material feedstock
The value and yield of all products
The cost of each fractionation stage
The number of fractionation stages
For simplicity assume a main product A and a by-product B, which comprises all the by
product fractions; the cost of each fractionation stage is the same. Then,
Raw m ateria l^ product A + by-product(s) B
For commercial viability,
•ÂÂ + ^ ^RM +
where are fractions of A and B, are values of A and B, ^RM is the cost of raw
material, n is the number of fractionations, and F is the cost of each fractionation.
Rearranging this equation we find that for break-even,
1 T/ 17 , , ^RM~^B
Break-even V^ ~ ^ b "•----- 1-------------
Xa Xa
In Table 3, values are given for for different values of n, x^, and V^. It can be seen
that although the break-even value increases with the number of fractionations and decreasing
yield, an even more important parameter is the value of the by-product(s) in relation to the
value of the raw material. The advantage of starting with a relatively low cost raw material
such as palm oil is clear.
Table 3 Calculation of Break-Even Value of Product A (V^) According to Equation

Given in the Text^
Break-even value of A depending on

value of by-products (Vb) in relation
Number of Fractional yield of to value of raw material (Vrm)
fractionations main fraction
n (■*a ) Vb = 1-1 V'rm Vb = Vrm Vb = 0.9 Vrm
1 0.5 1000 1100 1300

1 0.3 933 1167 1500
2 0.5 1100 1200 1400
2 0.3 1100 1333 1667
3 0.5 1200 1300 1500
3 0.3 1267 1500 1700
Âssuming value of raw material, P rm = $1000/t; cost of each fractionation stage = $50.
Fractionation 215
IV. APPLICATIONS
A. Palm Oil
Palm oil is a semisolid fat with about 20% solid fat crystals at 20°C. Fractionation to separate
these crystals as a stearin leaving an olein with a melting point of 20-24°C is carried out on
a large scale throughout the world, but mainly in Malaysia and Indonesia. In 1994, world
production of palm oil was almost 14 megatonnes, and production of palm olein is estimated
at 5.1 Mt. In Malaysia alone, 3.4 Mt of palm olein is estimated to have been produced, of
which 3.1 was exported.
Because palm oil is ideal for fractionation and because of the massive growth in palm oil
production in the last 30 years, modem fractionation processes have been developed and
optimized mainly for palm oil. Palm oil consists essentially of three groups of triglycerides:
About 8% trisaturated triglycerides, principally PPP
About 45% disaturated triglycerides, principally POP and PLinP
About 40% monosaturated triglycerides, principally POO and PLinO
The main fractionation process is carried out to produce an olein that is liquid below 2 4 °C ,
making it an ideal cooking oil in tropical countries. Essentially the process is designed to
separate PPP from the palm oil. Because of solid solution formation between PPP and POP
first stage second stage third stage

fractionation fractionation fractionation
Fig. 15 Dry multiple fractionation of palm oil. (From Ref. 21.)

216 Timms
Table 4 Properties of Palm Oleins Produced by Dry Fractionation
Cloud Percent soybean oil

Iodine point required to meet
Olein type value ro cold test at 0°C
“Traditional” 57 8 .0 100
Superolein 1 63 4.0 90
Superolein 2 65 1.3 70
“Top” olein 71 - 2 .0 0
and because of entrainment, the palm stearin produced contains substantial quantities of disa-
turated and monosaturated triglycerides. Dry fractionation processes are used, with the mem
brane filter press now the preferred method of separation. Using this process, an olein of IV
56-58 with PPP content 1% maximum can be produced in a yield of 75-80% . Deffense [20]
compared different fractionation processes for the production of this palm olein.
Palm olein from a single-stage fractionation has now become a commodity product and
one of the world’s major oils in its own right. To develop products with more added value,
multistage fractionation has been applied to produce more-liquid palm oleins as shown in Fig.
15 and Table 4.
Multistage fractionation necessarily produces middle-melting fractions, usually called mid
fractions. Palm midfractions (PMFs) concentrate the disaturated triglycerides. A major use of
PMF has been in the production of cocoa butter equivalent fats (CBEs) to replace cocoa butter
in chocolate. Essentially, POP is required, and solvent fractionation using acetone, hexane,
or 2-nitropropane has been used to get PMF of the required quality. (The use of 2-nitropro-
pane has been discontinued for toxicological reasons.)
With the improved crystallization and separation methods now applied to palm oil, it is
possible to produce a PMF by dry fractionation that is equivalent to that produced by sol
vent fractionation, albeit in somewhat lower yield. If the belt vacuum or low pressure mem
brane filter is used, then three fractionation steps are required. This process is shown schemat
ically in Fig. 15 to produce a “hard PMF.” Alternatively, if a high pressure (25-50 bar)
membrane filter press is used for separation, then only two fractionation stages are required
[15]. In Table 5 the various sorts of PMF now commonly available are compared with a
solvent-fractionated PMF. Iodine value is used as a measure of the total unsaturation, SFC as
Table 5 Properties of Palm Midfractions Produced by Dry and

Solvent Fractionation
Triglyceride
Iodine SFC at with Carbon No.
PMF type value 30°C 52 (%)
Dry “soft” 50 1 50
Dry “hard” 40 18 64
Dry “super” 34 45 74
Solvent 34 46 75
Fractionation 217
a measure of the steepness of melting (SFC at 35°C is less than 5%), and the triglyceride car
bon no. at 52 as a measure of the POP and PLinP content. It can be seen that the “super
PMF” obtained from the high pressure membrane press is similar to the solvent-fractionated
PMF.
B. Palmkernel and Coconut Oils

Palmkemel (PK) oil is fractionated in one step to concentrate the triglycerides containing
medium-chain fatty acids (lauric and myristic) into a stearin fraction. Coconut oil can also be
fractionated similarly, but the tonnages are very small in comparison with PK oil. The frac
tionation of lauric oils has been reviewed in detail by Rossell [22].
Palmkemel stearin is widely used as a cocoa butter substitute to replace cocoa butter in
compound chocolate. Palmkemel oil of IV 18 is fractionated to produce a stearin of IV 7-8
[23]. Because the stearin is the main product and olein the by-product, separation of the olein
from the stearin crystals is critical to the production of a good quality product. A dry fraction
ation process with dram or belt vacuum filtration cannot be used because the resulting stearin
contains too much entrained olein. Instead, solvent fractionation, dry fractionation with deter
gent (Lipofrac process), or dry fractionation with high pressure pressing are used as discussed
in Section III. In the detergent process, it is necessary to recycle about 30% of the olein to
the feedstock in order to lower the solid fat (crystal) content of the crystallized slurry to about
20%. If this is not done, it is necessary to accept a low yield of stearin in order to be able to
pump the slurry and not overload the centrifugal separator.
A comparison of these three fractionation processes is given in Table 6. In Fig. 16 the
melting curves of palmkemel oil, olein, and stearin are compared. The steep melting curve of
the palmkemel stearin, similar to the melting curve of cocoa butter, is clearly shown. Re
cently, the use of a high pressure membrane filter press has been described. These presses
seem to offer the advantages of the traditional hydraulic presses without their high labor costs
and difficulty of operating a clean process [24].
Table 6 Comparison of Fractionation Processes for Palmkemel Oil
Stearin
Capital Operating yield
Process cost costs (%) Olein value Remarks
Solvent High High ~50 Lowest Uses nitropropane, ac

fractionation etone, or hexane as
solvent.
Crystallization Low Medium to ~40 Intermediate A labor-intensive and
+ pressing high “messy” process as
usually practiced.
Operating costs de
pend much on labor
costs.
Detergent Medium Low to ~ 30 Highest Continuous process,
fractionation medium usually well auto
mated.
Source: Ref. 22.

218 Timms
Fig= 16 Solid fat contents (%) of palmkemel (PK) oil, olein, and stearin. (From Ref. 23.)
C. Milk Fat
Milk fat is a semisolid fat like palm oil and can be easily fractionated into a similar range of
products. The range of fatty acids and triglycerides present is much greater than in palm oil,
so the fractions produced are less distinct than palm fractions. Relatively little milk fat is
fractionated, perhaps about 250 kt/yr, due to its high price in relation to the value of the by
product fractions as discussed in the previous section. Most milk fat fractionation is carried
out not where milk fat is cheapest, e.g.. New Zealand, but in Europe, where the raw material
is expensive even with subsidies. This anomaly exists as a result of the European Common
Agricultural Policy price support scheme, whereby the value of the by-products is maintained
artificially high in relation to the raw material cost.
Although solvent and detergent fractionation processes have been investigated [25,26] and
are perfectly feasible, only dry fractionation with vacuum or membrane filtration is practiced
SO LID FAT
Fig. 17 Melting properties of milk fat fractions from three-step fractionation. (From Ref. 28.)
Fractionation 219
LIFT FACTOR
Fig. 18 Lift of puff pastry (48 layers) using butter and plasticized milk fat fractions. (From Ref. 28.)
commercially [27]. Some of the wide range of milk fat fractions that can be produced
by three-stage fractionation are shown in Fig. 17. By blending the soft (olein) and hard
(stearin) fractions, a “butter” that is spreadable at refrigeration temperatures can be produced;
this butter has been marketed by the New Zealand Dairy Board [27,28]. Milk fat stearins are
used to make pastry butter [29] with substantial improvement in performance as shown in
Fig. 18.
D. Beef Tallow
Beef tallow can be fractionated into an olein and a stearin in a similar manner to the fraction
ation of palm oil. Beef tallow has mostly been fractionated in Australia using the Lipofrac
detergent process. Some properties of commercial Australian beef tallow olein and stearin in
comparison with palm olein and stearin are shown in Table 7. Tallow stearin is used to
formulate pastry shortenings. Tallow olein is used as a frying fat or to formulate margarines.
The solvent fractionation of beef tallow has also been proposed as a method to produce
tallow midfractions for use in CBEs and other speciality fats [30,31]. Some useful fractions
can be obtained, but the process has never become commercial because of the general decline
in the use of animal fats for food, especially to replace vegetable fats, and the ready availabil
ity of palm oil, which is better suited to the production of a CBE component.
Table 7 Properties of Australian Beef Tallow Stearin and Olein Compared

with Palm Oil Stearin and Olein
Beef tallow Palm oil
Property Stearin Olein Stearin Olein
Iodine value 35 48 36 57
Melting pt (°C) 52 37 52 21
SFC at 10°C 73 43 74 40
SFC at 20°C 66 28 62 6
SFC at 30°C 49 12 41 0
SFC at 40°C 30 3 28 0
SFC at 50°C 9 0 12 0
220 Timms
E. Other Fats
Many other fats are also fractionated. Oils such as cottonseed oil and hydrogenated soybean
oil are fractionated to remove a small amount of solid so that the oils can be used as salad
oils that remain clear at refrigerator temperatures. The process is often called winterization,
because crystallization occurs naturally with the onset of winter, providing the ideal slow
crystallization conditions. Winterization is a basic dry fractionation with a relatively simple
vacuum filtration. In the winterization of cottonseed oil the stearin removed consists mainly
of PLinP, whereas for soybean oil the stearin removed contains mainly di- or triunsaturated
triglycerides containing trans acids. The process has been reviewed by List and Mounts [32].
Unlike the fractionation of natural oils, in the fractionation of hydrogenated soybean oil the
properties of the feedstock itself can be varied by the processor according to the degree of
hydrogenation. Clearly the greater the extent of hydrogenation (and the lower the IV), the
less olein of a particular IV can be produced. However, Singh [33] studied the crystallization
of hydrogenated soybean oil in detail and concluded, rather surprisingly, that there is an
optimum concentration of solid fat in the feedstock for the production of olein with the best
cold test.
In a process similar to the winterization of cottonseed and soybean oils, sunflower and
ricebran oils are fractionated to give oils that will not cloud at refrigerator temperatures.
Unlike all the other fractionation processes we have considered, the solid phase removed is
not a triglyceride but a wax, and the process is usually called dewaxing.
Sal fat, a fat extracted from the seed of Indian forest trees, is fractionated to produce sal
stearin, which is rich in SOS triglycerides for use as a component of CBEs. Although solvent
fractionation is practical and is used in Europe, sal fat is now more commonly dry fractionated
in India. The fat is crystallized in trays and pressed in a hydraulic press in a way similar to
the process described earlier for palm kernel oil. The yield of sal stearin by the dry fraction
ation process is typically 55-60% .
Shea fat, a fat extracted from the nut of trees in the dry tropical areas of West Africa, is
also fractionated to produce a shea stearin rich in SOS triglycerides. Compared with sal fat,
the content of stearin/SOS triglycerides is lower, and a solvent process is usually preferred to
give a yield of stearin of 35-40% . If acetone is used, an undesirable isoprenoid gum may
also be removed in an initial “degumming” step. Only where the product, in this case shea
stearin, has a high value can the cost of solvent fractionation be justified. With cocoa butter
at about US$4000/t, CBEs can be sold at up to US$3000/t, several times higher than the price
of most other edible oils and fats.
V. PROSPECTS
After hydrogenation, fractionation is the second most important process for modifying oils
and fats. Hydrogenation is used to modify a wide range of liquid oils, principally soybean,
rapeseed, sunflower, and fish oils, to produce semisolid plastic fats for use in margarine and
shortenings. Fractionation is used overwhelmingly to modify a single oil—palm oil—to pro
duce liquid fractions for use as frying oils and middle and higher melting fractions for use in
a wide variety of fat-based food products as well as oleochemicals. The use of fractionation
is likely to grow relative to hydrogenation for the foreseeable future because
1. Palm oil production is growing quickly, and palm oil and its associated oil, palm
kernel oil, are forecast to be the oils produced in the greatest volume in the world by
the year 2002, overtaking soybean oil [34].
Fractionation 221
2. There are nutritional and health concerns about the level of trans fatty acids in the diet
[35]. Some of these acids come from animal fats, but the major contribution in the
diet is from hydrogenated oils.
3. Fractionation is a mild process, whereas hydrogenation is a chemical process con
ducted at high temperatures using nickel and hydrogen gas. Fractionation is thus more
environmentally benign and is likely to be increasingly favored by today’s “greener”
consumers.
REFERENCES
1. W. Hamm, Trends in fractionation practice for edible oils, Presented at Soc. Chem. Ind. Symp.
Fractional Crystallization of Fats, London, 9 March 1994, SCI Lecture Paper No. 0037.
2. E. Schlimme, Removal of cholesterol from milk fat, Eur. Dairy Mag. 4: 12, 13, 16-21 (1990).
3. C. K. Ooi, Y. M. Choo, S. C. Yap, Y. Basiron, and A. S. H. Ong, Recovery of carotenoids
from palm oil, J. Am. Oil Chem. Soc. 71: 423-426 (1994).
4. W. Hamm, Liquid-liquid extraction in food processing, in Science and Practice of Liquid Extrac
tion (J. D. Thornton, Ed.), Clarendon Press, Oxford, 1992.
5. Unilever Ltd., Br. Patent 1 444 551 (1976).
6. W. Hamm, Trends in edible oil fractionation. Trends Eood Sci. Technol. 6: 121-126 (1995).
7. R. E. Timms, Phase behaviour of fats and their mixtures. Prog. Lipid Res. 23: 1-38 (1984).
8. R. Boistelle, Fundamentals of nucléation and crystal growth, in Crystallization and Polymorphism
of Eats and Eatty Acids (N. Garti and K. Sato, Eds.), Marcel Dekker, New York, 1988, pp.
189-226.
9. M. van den Tempel, Effects of emulsifiers on the crystallization of triglycerides, in Surface-Active
Lipids in Eoods, SCI Monograph No. 32, Soc. Chem. Ind., London, 1968, pp. 22-36.
10. R . E. Timms, Crystallization of fats, Chem. Ind., 20 May, 342-345 (1991).
11. K. P. A. M. van Putte and B. H. Bakker, Crystallization kinetics of palm oil, J. Am. Oil Chem.
Soc. 64: 1138-1143 (1987).
12. J. R. Dafler, Polymorphism behaviour in fully hydrogenated mono acid triglycerides, J. Am. Oil
Chem. Soc. 54: 249-254 (1977).
13. M. Kellens, Developments in fractionation technology. Presented at Soc. Chem. Ind. Symp. Frac
tional Crystallization of Fats, London, 9 March 1994, SCI Lecture Paper No. 0042.
14. R. E. Timms, Principles of fractionation. Presented at Soc. Chem. Ind. Symp. Fractional Crystalli
zation of Fats, London, 9 March 1994, SCI Lecture Paper No. 0039.
15. T. Willner and K. Weber, High-pressure dry fractionation for confectionery fat production. Lipid
Technol. May¡June: 57-60 (1994).
16. W. Hamm, Fractionation—with or without solvent?, Eette Seifen Anstrichm. 88: 533-537
(1986).
17. R. E. Timms, Choice of solvent for fractionational crystallization of palm oil, in Palm Oil Product
Technologies in the Eighties (Pushparajah and Rajadurai, Eds.), Kuala Lumpur, 1983.
18. M. van den Kommer and C. N. M. Keulemans, Developments in dry fractionation. Presented at
Soc. Chem. Ind. Symp. Fractional Crystallization of Fats, London, 9 March 1994, SCI Lecture
Paper No. 0040.
19. E. Deffense, Fractionation of palm oil, J. Am. Oil Chem. Soc. 62: 376-385 (1985).
20. E. Deffense, Superolein production and analysis of fractions. Presented at Soc. Chem. Ind.Symp.
Fractional Crystallization of Fats, London, 9 March 1994, SCI Lecture Paper No. 0036.
21. E. Deffense, Dry multiple fractionation: trends in products and applications. Lipid Technol.
March: 34-38 (1995).
22. J. B. Rossell, Fractionation of lauric oils, J. Am. Oil Chem. Soc. 62: 385-390 (1985).
23. R. E. Timms, Processing of palm kernel oil, Eette Seifen Anstrichm. 88: 294-300 (1986).
24. T. Willner, High-pressure dry fractionation. Presented at Soc. Chem. Ind. Symp. Fractional Crys
tallization of Fats, London, 9 March 1994, SCI Lecture Paper No. 0038.
222 Timms
25. P. C. Chen and J. M. de Man, Composition of milk fat fractions obtained by fractionation crystal
lization from acetone, J. Dairy Sci. 49: 612 (1966).
26. R. Norris, I. K. Gray, A. K. R. McDowell, and R. M. Dolby, The chemical composition and
physical properties of fractions of milk fat obtained by a commercial fractionation process, J.
Dairy Res. 38: 179 (1971).
27. E. Deffense, Milk fat fractionation today: a review, J. Am. Oil Chem. Soc. 70: 1193-1199 (1993).
28. C. Versteeg, L. N. Thomas, Y. L. Yep, M. Papalois, and P. S. Dimick, New fractionated milkfat
products, Aust. J. Dairy Technol. 49: 57-61 (1994).
29. W. Seibel, Use of fractionated butterfat (butter pastry fat), Zucker Susswarenwirtsch. 40: 122-
126 (1987).
30. F. E. Luddy, J. W. Hampson, S. F. Herb, and H. L. Rothbart, Development of edible tallow
fractions for specialty fat uses, J. Am. Oil Chem. Soc. 50: 240-244 (1973).
31. H. H. Taylor, F. E. Luddy, J. W. Hampson, and H. L. Rothbart, Substitutability of fractionated
beef tallow fractions for other fats and oils in the food and confectionery industries: an economic
evaluation, J. Am. Oil Chem. Soc. 53: 491-495 (1976).
32. G. R. List and T. L. Mounts, Partially hydrogenated-winterized soybean oil, in Handbook of Soy
Oil Processing and Utilization (D. R. Erickson, E. H. Pryde, O. L. Brekke, T. L. Mounts, and
R. A. Falb, Eds.), Am. Soyabean Assoc., St. Louis, MS, and Am. Oil Chem. Soc., Champaign,
IL, 1987, pp. 193-216.
33. G. Singh, Analysis of crystallization systems with applications to continuous fractional crystalliza
tion of fatty acid triglycerides, AIChE Symp. Ser. 72: 100-109 (1974).
34. S. Mielke, Trends in supply, consumption and prices, in Oils and Fats in the Nineties (V. K. S.
Shukla and F. D. Gunstone, Eds.), Int. Food Sci. Centre, Lystrup, Denmark, 1992, pp. 10-22.
35. British Nutrition Foundation, Trans Fatty Acids, London, July 1995.
36. J. Hannewijk, A. J. Haighton and P. W. Hendrikse, Dilatometry of fats, in The Analysis and
Characterization of Oils, Fats and Fat Products, Vol. 1 (H. A. Boekenoogen, Ed.), Interscience,
London, 1964, pp. 119-182.
37 . J. B. Rossell, Phase diagrams of triglyceride systems, Adv. Lipid Res. 5: 353-408 (1967).
Interesterification of O ils and Fats
A. Rozendaal
Unilever Research Laboratory, Vlaardingen, The Netherlands
A. R. Macrae
Unilever Research Laboratory, Shambrook, Bedfordshire, England
I. INTRODUCTION
Interesterification is a catalytic process involving the exchange of fatty acids between existing
esters to form new esters. When applied to a mixture of triacylglycerols (TAGs), the available
fatty acids are redistributed over all the possible triacylglycerol types. Due to this re
arrangement the physicochemical properties of the mixture, in particular the melting and crys
tallization or recrystallization properties, can be changed and steered into a desired direction.
Interesterification, next to hydrogenation and fractionation, is therefore an important tool in
modifying the physical and functional properties of oil and fat mixtures. Unlike hydrogena
tion, it leaves the overall fatty acid composition of the mixture unchanged. The wide range
of triacylglycerols present in interesterified fats usually leads to a more regular crystallization
during the manufacture of spreads and shortenings. Moreover, the rate of recrystallization
during storage of such products may be retarded, resulting in stabilization of the generally
preferred (3' modification and a reduced tendency to develop product defects. Interesterifica
tion or rearrangement processes therefore find ever wider application for converting oils and
fats into blend components that are optimally suited for incorporation into various types of
edible fat-based consumer products.
Among the catalysts nowadays available for rearrangement are alkali metal derivatives
(such as sodium methoxide and other alcoholates) and enzymes (lipases). The latter are effec
tive under mild reaction conditions and in many cases are specific in exchanging only the
fatty acids from the 1,3-positions of the triacylglycerols, thereby offering new possibilities for
the manufacture of tailor-made TAGs (structured lipids).
Interesterification is usually carried out in a single phase (liquid). With a nonselective
catalyst, rearrangement of the fatty acids will then ultimately result in their statistical or ran
dom distribution over all possible positions. However, it is also possible to carry out the
rearrangement under such conditions that part of the reaction products is continually removed
223
224 Rozendaal and Macrae
A c id o ly s is
H 2 C 0 C 0 R^ H2 COCOR4
1 1
1 1
HCOCOR2 + R4 C O O H — HCOCOR2 ■f R ^C O O H
1 I
1 1
H2 COCOR3 H2 C O C O R 3
A lc o h o ly s is
H2 CO CO R HgCO H
1 1
1 1
HCOCOR + 3 R3 OH — HCOH + 3 R3 O CO R
1 1
1 !
H2 CO CO R H2 CO H
In te re s te rific a tio n , rea rra n g e m e n t, e s te r in te rc h a n g e
H2 C O C O R 1 H2 C O C O R 4 H2 C O C O R 1
I 1
I 1 I
HCOCOR2 + HCOCOR5 — HCOCOR5 + others
1 1 I
1 1 I
H2 C O C O R 3 H2 CO C O R 6 H2COCOR2
Fig.i lypes ot mteresteritication reactions.
from the reacting mixture, usually by crystallization (directed interesterification). This process
is essentially the reverse of randomization as it tends to convert mixed triacylglycerols into
high melting, crystallizable, fully saturated TAGs on the one hand and triunsaturated TAGs
on the other hand:
SU2 + S2U ^ S3 i + U3
The term “interesterification” involves various types of reactions in which an oil, fat, or other
material composed of fatty acid esters is caused to react with fatty acids, alcohols, or esters.
It therefore includes (see Fig. 1)
1. The reaction of an ester with an acid, more specifically known as acidolysis
2. The reaction of an ester with an alcohol, known as alcoholysis (e.g., glycerolysis
or methanolysis)
3. The reaction of an ester with another ester, referred to as interesterification, ester
interchange, rearrangement, or sometimes transesterification.
Several excellent reviews on interesterification, covering both processing aspects and product
applications, are available [ 1 - 1 1 ].
II. CHEMICAL BACKGROUND AND

PRODUCT COMPOSITION
As already indicated, single-phase interesterification with a nonspecific catalyst or lipase will
eventually lead to the so-called random distribution in which the fatty acids are statistically
distributed over all possible positions (hydroxyl groups) in the system. The background to this
is that changes in free energy due to fatty acid exchange between esters are negligibly small.
The statistical nature of the fatty acid distribution implies that the equilibrium triacylglycerol
Interesterification of Oils and Fats 225
composition and thereby the physicochemical properties of interesterified fat blends are highly
predictable from the overall fatty acid composition of the mixture.
In naturally occurring oils and fats, on the other hand, the constituting fatty acids are
distributed according to specific patterns. For example, in vegetable oils, saturated and long-
chain monounsaturated fatty acids (such as 22:1 in the former high-erucic rapeseed oils) pref
erentially occupy the external positions of the triacylglycerol molecules while the 2 -position
is almost exclusively occupied by unsaturated fatty acids, especially linoleic but also oleic
and linolenic acid. Detailed stereospecific analyses [12,13] have demonstrated that the 1- and
3-positions are not identical and may show significantly different ratios of the fatty acids
attached to them. On first approximation, however, the fatty acid distribution in vegetable oils
obeys the so-called 1,3-random, 2-random distribution [14,15].
In animal fats the differences between the fatty acid compositions at the 1- and 3-positions,
are generally larger than in vegetable oils. Moreover, the 2-position may be enriched in
saturated fatty acids, notably palmitic acid. The presence of a considerable amount of StPU
triacylglycerols in lard and, as a consequence of this, a rather irregular crystallization behav
ior, has historically been one of the main driving forces for applying interesterification in the
United States (preferably directed interesterification in this case [16]).
Returning to vegetable oils, up to a saturated fatty acid content of about 30-35% , the
proportion of palmitic (P) and stearic (St) acids at the êxternal positions can be approxi
mated by
P ,3 = 1 .4 7 P o St ,,3 = 1.47 St„
in which ov refers to the overall fatty acid composition. Saturated fatty acids will thus be
present predominantly as asymmetric SUU and symmetric SUS type triglycerides, whereas fully
saturated triacylglycerols will be virtually absent. Only in the case of a high overall content of
saturated fatty acids such as in palm oil will part of the saturated fatty acids occupy the 2-posi-
tion. As a consequence, trisaturated glycerides will be present (7-10% , mainly tripalmitin) and
the SUS/SSU ratio will typically be in the range of 4 -8 .
Applying hydrogenation and fractionation to oils and fats will disturb the 1,3-random, 2-
random distribution. Fractionation will remove the high melting triacylglycerols; e.g., the
tripalmitin content will be reduced to below 1 % upon fractionation of palm oil at temperatures
of about 2T C or lower. Hydrogenation is unselective with respect to the position of the fatty
acids in the triacylglycerol molecule; the unsaturated fatty acids will thus be attacked in pro
portion to their relative concentrations at the 1,3- and 2-positions. In any case, we may con
clude from the above that natural oils and fats as well as their hydrogenated or fractionated
derivatives have a nonrandom distribution pattern, and the same is, of course, true for their
physical mixtures. Random distributions will be obtained only by interesterification (re
arrangement) at temperatures above the melting point of the triacylglycerol mixture. Due to
the randomization and thereby the formation of increased amounts of fully saturated TAGs,
the melting point of several oils, including soybean, cottonseed, palm, and cocoa butter, is
raised. On the other hand, mixtures comprising high melting components (e.g., hydrogenated
oils and fats, stearin fractions, and animal fats such as lard and tallow) show a reduction in
melting point and solid-phase content when interesterified with soft oils. Although randomiza
tion often leads to fats with flat melting curves and a long plastic range, a steepening of the
solid-phase content versus temperature profile may occur at a sufficiently wide chain length
distribution of the fatty acids present in the mixture. Tables 1 and 2 show typical examples
of changes in TAG composition as a result of interesterification.
If two monoacid triacylglycerols, AAA and BBB, are interesterified, the redistribution of
the fatty acids will result in an equilibrium mixture of six possible triacylglycerols (Fig. 2).
At equimolar amounts of A and B, the final mixture will consist of
Table 1 TAG Composition (mol %) of Palm Oil Before

and After Interesterification
After random esterification
Triacylglycerol Initial Actual Calc.
sss 8 13 14
ssu 9 13 13
sus 35 25 26
suu 36 23 24
usu 9 14 12
uuu 7 12 11
S = saturated, U = unsaturated fatty acid.
Source: Ref. 17.
12.5% of each of the parent triacylglycerols AAA and BBB

37.5% of each of the intermediate triacylglycerols A 2B and AB 2, of which two-thirds will
be asymmetric and one-third symmetric
If three different fatty acids A, B, C are present in a mixture at molar fractions a, b,

and c, respectively, in principle 27 different TAGs will be formed during interesterification.
Disregarding stereoisomers, this number falls to 18 (Table 3). The monoacid glycerides AAA,
BBB, and CCC will be present in molar amounts of b^, and c^. The molar fraction of
each of the triacid glycerides ABC, BAC, and ACB will correspond to 2abc. Diacid glycer
ides such as BBC and BCB will be present in amounts of 2P c and b^c, respectively. When
positional isomers are taken together, the amounts of each of the diacid glycerides will equal
3a^b, 3a^c, 3P c, etc., and the triacid glyceride ABC will account for a mole fraction of 6abc.
Further expressions for the theoretical composition of randomized mixtures [6] containing
n different fatty acid types are given in Table 4.
The above predictions with regard to the composition of randomly interesterified fats and
oils have been confirmed extensively. Classical studies include those of Norris and Mattil
[18] and Naudet and Desnuelle [19], who found excellent agreement between calculated and
experimental compositions for mixtures of tristearin and triolein interesterified with sodium
methoxide as catalyst (Fig. 3). Other studies involving real oils and fats have led to a similar
Table 2 TAG Composition (mol %) of Lard after Random and

Directed Interesterification
Mole % TAG
Interesterification
Triacylglycerol Initial Random Directed
SSS 2 5 14-17
ssu -f sus 26 25 15-12
suu + usu 54 44 32-37
uuu 18 26 39-34

E E
—A |- A p-A pB pB r-B
—A + \~B + h-A + +
—A L_b L-A L_b L -a L_b
Equilibrium
12.5 25 12.5 12.5 25 12.5
composition (%)
Fig. 2 Interesterification of a binary mixture (50/50) of triacylglycerols.
conclusion [20,21]. List et al. [22] recently applied interesterification between various liquid
oils (cottonseed, peanut, soybean, com, and canola) and fully hydrogenated soybean or cot
tonseed oil hardstocks for the manufacture of zero-trans margarine fats. Generally speaking,
excellent agreement was found between the experimentally determined TAG compositions and
those calculated for random distribution, from which it was therefore concluded that under
the reaction conditions employed, complete randomization of the triacylglycerol stmcture had
occurred in most cases.
Even if there are very large differences in the chain lengths of the constituting fatty acids,
randomness of the triacylglycerol composition after interesterification appears to remain
preserved. This follows from studies by Klemann et al. [23] on Salatrim fats, which are
composed of very short chain fatty acids (C2- C 4) and long-chain saturated fatty acids. These
fats are claimed to be low in caloric value, due to poor adsorption of long-chain fatty acids
on the one hand and low combustion energy of short-chain fatty acids on the other hand.
Klemann et al. applied sodium methoxide catalyzed interesterification to mixtures of C 2- C 4
triacylglycerols and completely hydrogenated canola or cottonseed oils and subsequently re
moved unreacted short-chain triacylglycerols from the reaction mixture by steam distilla
tion. The recovered compositions were examined by several analytical techniques and found
to be in good agreement with values calculated by the statistical model for random distribu
tion [23].
For specific applications one may aim at a maximum amount of a certain triacylglycerol or
Table 3 Random Triacylglycerol Composition of an Oil Mixture Containing Fatty

Acids A, B, and C in Molar Fractions a, b, and c, respectively
Triacylglycerol Molar fraction Triacylglycerol Molar fraction
AAA CAC ac^

BBB ACC 2ac^
CCC BCB b^c
ABA a^b BBC 2b^c
AAB la^b CBC bc^
ACA a^c BCC 2bc^
AAC 2a^c ABC 2abc
BAB ab^ BAC 2abc
ABB 2ab^ ACB 2abc
Table 4 Random Triacylglycerol Composition of an Oil Mixture Composed of n Fatty Acids A, B,

C, D, . . . in Molar Amounts a, b, c, d, . . .
Number with Number without Amount without

Triacylglycerol positional isomers positional isomers positional isomers
Monoacid (AAA, BBB, . . .) n n . . .

Diacid (AAB/ABA, AAC/ACA, . . .) 2n{n—1) n { n - \) 3a^b, 3a^c, . . .
Triacid (ABC, ACD, . . .) ( l/2 ) n ( n - l) (n - 2 ) {ll6)n{n—\){n —2) 6abc, 6abd, . . .
Total number in^ + n^)/2 («^ + 3«^ + 2w)/6

n= 2 6 4
n= 3 18 10
n= 4 40 20
n= 5 75 35
n= 6 126 56
class of triacylglycerols after completed interesterification. From Fig. 3 it can be seen that for
a pseudobinary system the monoacid triacylglycerols show a continuous decrease or increase,
whereas the diacid triacylglycerols show a maximum value at fatty acid molar fractions of
one-third and two-thirds, respectively. The maximum amount of these intermediate triacyl
glycerols equals
3x2/3x2/3x1/3 = 479
or 44 mol %. For a ternary system one finds a maximum for only the triacylglycerol ABC,
which is obtained at fatty acid mole fractions a = b = c= 1/3.
composition (mol %)
Fig- 3 Theoretical and experimental composition of interesterified triacylglycerols containing stearic

and oleic acids in different proportions. (From Ref. 19.)
III. REARRANGEMENT CATALYSTS

Since the finding of interesterification of oils and fats, a very large number of potentially
useful catalysts have been investigated as a means to enhance the reaction rate. Without added
catalyst, some degree of rearrangement can be obtained, but only under extreme conditions
of temperature and time, leading to undesirable effects such as isomerization, polymerization,
and decomposition. Apart from the now generally preferred basic catalysts (e.g., sodium
methoxide and other alcoholates), the following substances have been claimed to catalyze
interesterification or related reactions [2,4,8,24]:
Acids, including mineral acids, sulfonic acids, cation exchangers, and zeolites
Metals and their oxides, hydroxides, and salts, e.g., the oxides of zinc, magnesium, and
calcium, the hydroxides and chlorides of tin and zinc, the acetates and stearates of
calcium and magnesium, aluminates of alkali metals and anion exchangers
Organometallic compounds, including acetylacetonates, organotin and titanium com
pounds, metal alkyl aluminates, etc.
The above list is by no means complete. Further details can be found in the cited review
papers [1-9]. Generally speaking, none of these materials is sufficiently active or otherwise
attractive for the interesterification of edible oils and fats. In most cases reaction temperatures
of at least 200°C were required to achieve substantially complete interesterification, which
would imply a risk of undesirable side reactions and traces of catalyst being left in the oil.
On the other hand, the alkali metal alcoholates generally applied nowadays are sufficiently
active at much lower temperatures and can also easily be washed out after completion of the
reaction. Formo [25] estimated that alcoholysis catalyzed with sodium alkoxide proceeds at
least 4000 times as fast as with an equivalent amount of hydrochloric acid.
As so many substances show at least some catalytic effect at elevated temperatures, their
presence should be minimized in cases where even a slight degree of interesterification is
undesirable, e.g., during physical refining of olive oil and palm oil for the production of palm
midfraction. In both cases partial inter- or intraesterification will result in an increased amount
of saturated fatty acids at the 2 -position and a shift in the ratio of symmetric to asymmetric
TAGs. Willems and Padley [26] studied the effects of temperature and time on the SOS/SSO
ratio of palm oil; their results are reproduced in Fig. 4. A distinct drop in the above ratio
from about 8 to 6 was observed after holding times of a few hours at 240°C. Such changes,
although significant when palm oil is to be used as a feedstock for sophisticated fractionation
products, appear to be irrelevant for oils and fats to be applied in spreads and shortenings.
ratio SOS/SSO
Fig. 4 Effect of heating time and temperature on the SOS/SSO ratio of palm oil. (From Ref. 26.)
Palm oil, because of its relatively high content of diacylglycerols, is probably also more
sensitive to partial interesterification than most other oils and fats.
As indicated above, alkaline catalysts are now generally preferred for interesterification of
oils and fats. The following compounds have been proposed and/or actually used:
Alkali metals Na, K, Na-K alloys

Alkoxides (alcoholates) NaOCH3, NaOC2H5
Hydroxides NaOH, KOH
Glycerolates From NaOH + glycerol
Others E.g., NaH, NaNH^, NaR (R = alkyl or phenyl)
Most of these allow a fast rearrangement at temperatures < 10 0 °C and—provided the feed
stock is sufficiently low in moisture, free fatty acids, and oxidized compounds— at very low
catalyst doses, e.g., down to 0 .1% [2,8] or even 0.05% [11] of sodium methoxide or
ethoxide.
Among the alkali metals proposed for interesterification, metallic sodium for random re
arrangement [27] and Na-K alloy for directed interesterification [16,28] are best known. So
dium-potassium alloys are liquid at room temperature (the system is eutectic with a melting
point of —12°C at the molar composition NaK 2) and can therefore easily be dispersed at the
low reaction temperatures required for directed interesterification with crystallization as the
driving force. Metallic sodium has a melting point of 98°C and has been dosed by pressing
the soft metal through orifices into a stream of preheated oil in which the sodium wires are
melted off, followed by further reduction of droplet size in a high shear mixer to ensure
H ? C - 0- C 0- R i H2C - 0 ~ C 0 --R i
I
I
H C -O -C O -R p + NaOCHp H C -O --C O -R 2 + CH3 - O - C O - R 3
I
H 2 C -O -C O -R 3 H2C»-0 ~-Na
H 2 C -O -C O -R 1 H c -O -C O -R .
I '
I
H C -O -C O -R p + Na HC 0 “ C 0 ‘R 2 + 1/2 H p
I
I
H 2 C -O H H p C -O -N a
H g C -O -C O -R i HpC-O H H p C -O ^ C O -R . HpC-OH
1 I ' I
H C -O -C O -R p + H C -O H H C -O -C O -R p + H C -O H
I 1
1 I I
H 2 C -O -C O -R 3 HpC-ONa HpC—O—Na H2C - O - C O - R 3
|-H p O
HpC-OH
^ j
NaOH + HC“ OH
j
1
HpC-OH
Fig. 5 Formation of sodium diacylglycerolate from various precatalysts.
interesterification of Oils and Fats 231
effective reaction. Application of metallic catalysts also has a number of disadvantages, i.e.,
high reactivity, the need for extensive safety precautions during storage and in use, generation
of some hydrogen during catalyst activation (see Section IV. A and Fig. 5), and risk of
ignition of unreacted metal during catalyst inactivation with water after completion of the re
action.
Alcoholates (sodium methylate, ethylate, or glycerolate) are therefore most widely applied
in current rearrangement processes. Sodium or potassium glycerolate is preferably formed in
situ from an aqueous solution of the corresponding hydroxide and glycerol, either in a two-
stage batch process with catalyst addition and drying at low temperature followed by interes
terification at higher temperature [29] or as a one-stage continuous process in which catalyst
addition, activation, and rearrangement are all carried out at the same temperature [30]. So
dium and potassium hydroxide, without addition of glycerol, are distinctly less active, and in
general temperatures of at least 160-180°C appear to be necessary.
Sodium ethylate or methylate can in principle be dosed in several ways, e.g., as a powder,
dispersed in inert solvent or fat, or dissolved in the corresponding alcohol (e.g., as 25%
solution). For industrial use, dosing as dry powder is the most practical method. Application
of alcoholic solutions will lead to very high oil losses, because most, if not all, of the added
solvent is converted into ethyl or methyl esters. Predispersion in liquid oils should be avoided
or handled with care, as high catalyst concentrations may cause heating up and side reactions.
IV. MECHANISM OF ALKALI-CATALYZED

INTERESTERIFICATION
A. Overall Description
Several stages can be distinguished in the total interesterification process:
1. Formation of the genuine interesterification catalyst by gradual dissolution of the added

catalytic material into the oil (induction period of the reaction)
2. Interesterification by exchange of fatty acids between triacylglycerols and the genuine
catalyst (maximum reaction rate period)
3. Declining rate period due to approach of the equilibrium (random) distribution.
Much evidence is now available to prove that the genuine catalyst responsible for fatty
acid exchange in current interesterification processes of oils and fats consists of the sodium or
potassium compound of a diacylglycérol molecule, i.e..
HXONa
HCO CO R
HXOCOR
This species is structurally similar to simple alcoholates (sodium methoxide or ethoxide) but sol
uble in oil. Interesterification is thus a homogeneously catalyzed reaction, with the active cata
lyst being formed in the initial stages of the process by reaction of the added (pre-)catalyst with
Conversion (%)
Fig. 6 Kinetics of interesterification of methyl myristate and butyl palmitate. (From Ref. 31.)
tri- and/or diacylglycerols from the oil. The genuine catalyst is identical for all types of precata
lysts; only the way it is formed during the catalyst activation period is different (Fig. 5 ).
Several kinetic studies support the above view of catalyst activation as a necessary first
step in interesterification reactions. To minimize analytical complexity, most of these early
kinetic studies were carried out in model systems. Jordan de Urries and Utrilla [31], for
example, studied the interesterification of methyl myristate and butyl palmitate using sodium
metal as catalyst in the 25 - 45 °C temperature range. Their results (Fig. 6) indicate that the
degree of conversion (relative to the equilibrium composition) goes up rapidly with increasing
reaction temperature, and, especially at the lower temperatures, there is evidence that there is
an induction period. In a similar study by Coenen [7] into the interesterification of glycol
esters of short-chain fatty acids, very pronounced induction periods were observed; the actual
redistribution reaction appeared to be fast even at temperatures as low as 30-40°C (Fig. 7).
A third example refers to palm oil where the solids content at 40°C was taken as a measure
of the degree of interesterification (Fig. 8). Again it can be seen that the reaction speeds up
at higher temperatures and that at lower temperature there is an induction period.
All these data strongly suggest that the added catalysts were indeed only precursors for the
genuine interesterification catalyst, which is slowly formed during the initial stages of the
reaction. This was confirmed by the fact that no induction period is observed if the catalyst is
Conversion (%)
Fig. 7 Kinetics of interesterification of glycol esters of and Cg fatty acids. (From Ref. 7.)
Solid phase content (40°C)
Fig. 8 Kinetics of interesterification of palm oil at these different reaction temperatures (60°C, 52°C,
and 45T ). (From Ref. 7.)
predissolved in part of the oil or dosed as a molecular solution of, e.g., sodium methylate in
methanol [7].
Due to the heterogeneous nature of the catalyst activation step, the length of the induction
period will depend on numerous factors, including
1. Type of catalyst and its physical state (liquid, solid)

2. Particle size and size distribution
3. Input of mechanical energy into the system
4. Oil impurities that affect the surface condition of the catalyst particles
In the actual practice of oil and fat interesterification, the induction phenomenon may be
virtually absent, due to effective dispersion of catalyst, well refined and dried oil quality, and
small catalyst particle size. Temperature is a factor also, because the activation energy for the
formation of the genuine interesterification catalyst is higher than that of the interesterification
reaction itself.
Coenen [7] established that the interesterifications of esters of mono-, di-, and trihydric
alcohols as depicted in Figs. 6 , 7, and 8, respectively, can all be described (after the induction
period) by first-order kinetics and activation energies of 13-15 kcal/(mol • °C).
Weiss et al. [32], in their study of sodium methoxide catalyzed rearrangement of lard, in
fact distinguished three reaction phases:
1. Active catalyst formation

2. Crystal modification (associated with a reduction of S2U content)
3. Interesterification (approach to random distribution)
The activation energy for the first phase, i.e., formation of the active catalyst, was estimated
at 26 kcal/(mol • °C) against 14-17 kcal/(mol • °C) for the other two phases, which is in good
agreement with Coenen’s data.
B. Detailed Mechanism
Several investigators [2,6-8,32-36] have attempted to understand the mechanisms involved
in catalytic interesterification, and it appears now to be almost generally accepted [7,11,34,35]
H p C -O -C HpC- HpC"
O
H C -O -Q H C -- HC™
O
/^ 2 /R 2
H p C -O -Q H2C™ a.0
> 0
H2C - 0 ^ +
H2C - Q /^ 2
H2C - 0 ~ c;
,o HC“
HC“" 0 “ C HC™-
^R i
.o HpC“
H p C -O -C H2 C-
\ f
Fig=9 Mechanism of interesterification. (From Ref. 7, 11, 34, 35.)
that the fatty acid exchange occurs by interaction of triacylglycerols with the genuine catalytic
species (sodium or potassium compound of diacylglycerols) according to the mechanism sche
matically shown in Fig. 9. This is a so-called bimolecular, nucleophilic substitution reaction,
in which the slightly negatively charged diacylglycérol anion with fatty acids Rj is supposed
to attack on the slightly positive carbonyl carbon atom of one of the fatty acids R 2 of the
triacylglycerol molecule. In this way a transition complex is formed that can subsequently
decompose to the original compounds or form a new triacylglycerol molecule (R 2R 1R 1) and a
new active catalyst ion (R 2R 2O ). This anion is then available for reaction with other triacyl
glycerols, until eventually the reaction reaches an equilibrium in which fatty acid exchanges
still continue but do not lead to further net changes in composition.
o
EtÔ " N a + i E t-O -C .
\c
+ +
o
H g C -o -c : HgC-i-O-Na+i
\c
.0 o
H C -O -C HC-O-C,
\c \c
.0 o
HgC-o-c; HpC-o-c:
\ f " \ r
Fig. 10 Mechanism of catalyst activation. (From Ref. 11, 33, 34.)

HoC- O.0 H oC-O Ri H p C -O -C l

!
I
HC“ o - c: H C -d ^ H C - 0 .0
o
y ^2 /^ 2
HoC-o-c: H22C
i. - - -0 -C ^ H n C -O -C ^
Fig. 11 Mechanism of intramolecular interesterification. (From Ref. 33.)
The mechanism of catalyst activation can be described in a very similar way. For example,
with sodium ethoxide as catalyst, the ethoxide anion will attack on one of the carbonyl carbon
atoms of a triacylglycerol molecule, thus forming the catalytically active diacylglycérol anion
and an equivalent amount of fatty acid ethyl ester (Fig. 10).
Intramolecular interesterification, i.e., fatty acid exchange within the same molecule, is
assumed to proceed more quickly [37] than the intermolecular reaction, but again the mecha
nism appears to be quite similar (Fig. 11).
Weiss et al. [32] proposed a slightly different mechanism, the first step of which is the
formation of an enolate ion by hydrogen abstraction from the CH 2 group adjacent to the
carbonyl group (Fig. 12). This enolate ion would then interact with the slightly positive car
bonyl carbon atom to produce a ^-ketoester intermediate and, through a number of other
steps, an inter- or intramolecularly rearranged triacylglycerol (Fig. 13). Infrared spectral data
were in agreement with the proposed mechanism, although it was shown later [8] that the
measured IR adsorptions could also be attributed to other compounds such as soaps. Further
evidence against this mechanism was obtained independently by van der Plank [36] and Hei-
O O®
I
r O - c - C “ R. r " 0 ““ C=C"“ R
-O -C O C H p R i 1 ^ !
A H H
-O -C O C H p R p -f OCH3 4“
-O -C O C H p R p -O -CO CHpR
2^2
-O -C O C H 2 R 3
-O -C O C H p R p
©
o
1;
"0 ""C-=C-R.
1
H
- 0 -C O C H 2R2
“O -CO CH 2R3
Enolate ion
Fig. 12 Enolate ion formation as proposed by Weiss et al. [32].

236 Rozendaal and Wacrae
e o
o i; O O
i; -O -C H II H II
-0 -C--^,C-Ri rO-C-C-C-CH 2R2
/H ^ C -R , Ri
-O-COCH2R2 - 0 -C -C H 2R2 -o
I
L -O -C O C H 2 R 3 O -0-COCH2R3
-O -C O C H 2 R 3
FIg= 13 Mechanism of (intra-)molecular interesterification as proposed by Weiss et al. [32].
dal and M 0rk [35], who showed that the j8-ketoester anion is so strongly resonance-stabilized
that it is unable to catalyze interesterification reactions.
Strong support for the mechanism involving sodium diacylglycerolate as active species has
been provided by Steenhoek (cited in Refs. 7 and 11), who made a detailed study into the
kinetics of interesterification of the glycol esters of saturated Cg and unsaturated C jq fatty
acids with predissolved catalyst. The system was described by a set of six equations between
all possible triacylglycerols and all possible diacylglycérol anions (Fig. 14). Each equation
was considered as an equilibrium reaction with in principle the same rate constant 3K. Correc
tions, however, were applied to account for the fact that only part of the exchange reactions
result in a net change of composition. Therefore, effectively the rate constants of some of the
S3 + UjONa SU2 + SgONa
U3 + SpONa SpU + UpONa
SU2 + UpONa 5 = ^ U3 + SUONa

3K
SpU + SpONa S3 + SUONa

3K
SpU + UpONa SUp + SUONa

K
SUp + SpONa SpU + SUONa

K
Fig. 14 Kinetics of interesterification via diacylglycerolate as catalytic species. (From Ref. 7, 11.)
Interesterification of Oiis and Fats 237
Concentration (gmol/liter^)
1.0 I
S 2OH
_A - SUOH
U2OH
10 15 20
time (min)
Fig. 15 Calculated and experimental triacyl and diacylglycérol compositions for reaction of S 3 and U 3
with predissolved catalyst. (From Ref 7, 11.)
reactions are equal to K or IK . Results of this simulation study are given in Fig. 15, which
shows excellent agreement between calculated and experimental data not only for the triacyl-
glycerols but also for the diacylglycérol anion composition, thus clearly supporting the pro
posed mechanism.
Two concluding remarks need to be made in this section on the kinetics and mechanism of
interesterification. The first refers to the characteristic brown color that usually develops in
the early stages of the reaction and is generally seen as an indication that the reaction is likely
to proceed to completion. The nature of the coloring compound, however, does not seem to
have been unambiguously identified. The second comment refers to a certain regioselectivity
effect as demonstrated by Konishi et al. [39], who found that under specific reaction condi
tions the ester interchange rate at the 1,3-positions may be up to a factor of 2 faster than at
the 2-position. The system investigated was the sodium methoxide catalyzed interesterification
of soybean oil with methyl stearate in hexane solvent. Regioselectivity was observed at 30-
40°C but disappeared at 60°C. A kinetic factor, i.e., steric hindrance of the fatty acids
attached to the central position of the glycerol molecule, is likely to be responsible for the
observed regioselectivity.
V. DIRECTED INTERESTERIFICATION
Directed interesterification, as first described in some detail by Eckey [40], can be a valuable
extension to random rearrangement. The directed process can, for example, be applied for
1. Extending the plastic range of fats and improving their consistency properties
2. Preparing oleins from oils like palm oil for use as salad or frying oil
3. In situ preparation of hardstock components from liquid oils such as sunflower and
cottonseed oils
4. Producing low SAFA oils by directed interesterification and subsequent fractionation
The directing effect is always due to the presence of a second phase (gas or solid), into
which reaction products are transported and thus withdrawn from the liquid phase in which
fatty acid rearrangement proceeds to continually restore the disturbed equilibrium. One exam
ple of the removal of reaction products via the gas phase is described by Schmulinzon et al.
[41], who interesterified hydrogenated coconut oil with longer chain fatty acid methyl esters
in vacuum at 140°C. The low boiling esters of caprylic and capric acid were thus continuously
distilled off during the reaction. Another example is given by Barsky [42], who, by simply
heating, replaced short-chain fatty acids from a fat with free fatty acids having at least
two carbon atoms more than the fatty acids to be displaced. However, the liquid-solid system
in which high melting triacylglycerols are forced to crystallize by application of sufficiently
low temperatures is by far the most important system for practical directed interesterifica
tion.
Directed interesterification is a relatively slow process because of (1) the decreased rate of
reaction as a result of the low temperatures required for crystallization and (2) the small
driving force for crystal growth during the process. Depending on the fatty acid composition
of the system and the applied conditions of temperature and catalyst concentration, either the
interesterification or the crystallization will be the principal rate-determining step. For in
stance, an oil containing 85% unsaturated and 15% saturated fatty acids will have the follow
ing TAG composition at random distribution:
U 3 = 61%, SU2 = 33%, S2U = 6%, 83 = 0 .3%
The actual concentration of S3 and, depending on the degree of cooling, that of S2U will
exceed the solubility of these triacylglycerols when the randomized mixture is cooled down
to the directed interesterification temperature, resulting in crystallization of S3 and S2U, or
their mixed crystals. The reduced concentration of these TAGs will then be compensated for
by the reactions
SU , + SU Na i + U ,N a
S .U + U 2 N a U , + S jN a
SU - + SjN a + U^Na
In this example, temperatures as low as 0 -1 0°C will be required to effect crystallization.

With feedstocks like lard and palm oil, the S3 content after randomization will be on the order
of 5-14% , and considerably higher crystallization/reaction temperatures can be applied. In
that case directed interesterification will result in both SU2 and S2U being reduced in concen
tration [16,43].
To compensate as much as possible for the low reaction temperatures, highly active catalyst
types such as sodium-potassium alloy [16,28] or sodium ethylate or methylate must be applied
for directed interesterification. The catalyst activation step is normally carried out at somewhat
elevated temperatures, typically 40-60°C, to avoid lengthy induction periods (see Section
IV .A ).
Adding small amounts of glycerol to oils and fats appears to be beneficial for speeding up
directed interesterification [44], presumably due to the fact that saturated mono- and diacyl-
glycerols are less soluble in oil than the corresponding triacylglycerols. Various other methods
to enhance the rate or degree of completion of the process have been used also, e.g., a gradual
Fig, 16 Flow chart of directed interesterification process of lard. (From Ref. 16.)
reduction of temperature during the directed interesterification [40] and a rapid chilling from
the catalyst activation temperature to the directed interesterification temperature [16,45] to
create a large number of crystal nuclei on which further crystal growth is facilitated. Periodic
variation of temperature (cycling) between an upper and a lower temperature has also been
claimed [45,47] to enhance the overall rate of the coupled interesterification-crystallization
processes.
Determinations of triacylglycerol composition after directed interesterification of lard [48]
have shown that formation of fully saturated TAGs is favored at relatively high temperatures
(20-38°C), whereas at lower temperatures (0 -1 0°C) S2U formation is more important. In this
fraction the asymmetric triacylglycerols also dominate the symmetric ones. The rate-enhanc
ing effect of a stepwise reduction of temperature was confirmed in this study.
A flow diagram for continuous directed interesterification of lard [4,16] is reproduced in
Fig. 16. Liquid Na-K alloy catalyst is metered into a stream of dried oil and dispersed to a
droplet size below 50^m in a high shear mixer. The oil-catalyst mixture is then rapidly
cooled in a scraped surface heat exchanger to induce nucleation and crystallization in a first
crystallizer. Owing to the liberation of crystallization heat, a second cooling step is necessary,
after which the oil flows through a series of crystallizers to bring the process to the required
degree of completion. The design of the crystallizers must ensure a gentle movement through
out the crystal dispersion, preventing any stagnant zones.
VI. RANDOM INTERESTERIFICATION—

PROCESSING ASPECTS
A. Feedstock Requirements
Prior to interesterification, oils should be thoroughly pretreated and dried to remove any cata
lyst-poisoning substances such as free fatty acids, moisture, and oxidized compounds (perox
ides) as far as reasonably possible. The reactions of FFA and moisture with strongly alkaline
catalysts such as sodium methylate, ethylate, or diacylglycerolate (the genuine active catalyst,
see Section IV.B) occur quantitatively:
N adi + H ,0 N aO H + d ia c y lg ly c e ro l
C H jO N a + H ;,0 -> N aO H + CHjOH
C H jO N a + RCOOH RCOONa + C H ,O H
Consequently, each 0 .0 1 % moisture in the feedstock will inactivate about 0.03% (by
weight on oil) sodium methylate. Similarly, 0.1 wt % free fatty acids is equivalent to an
additional catalyst requirement of 0.02 wt %. For peroxides also an equimolar reaction has
been proposed [6, 8], implying an additional catalyst requirement of about 0.03 wt % for each
5 POV units, but this may be somewhat overestimated.
Based on the above, the following feedstock specifications are recommended:
Free fatty acids <0.1%, preferably <0.05%

Moisture < 0 .01%
POV < 10, preferably < 5 mgaeq/kg
The required FFA level can easily be obtained by conventional neutralization processes.
Bleaching is not generally required except when the peroxide value is too high and color
specifications of the final product are very tight.
Predrying is most effectively carried out by (1) spray drying in a vacuum chamber (continu
ous processes) or (2 ) batch drying with oil recirculation and spray drying into the evacuated
headspace. Some guidance for the required (minimum) drying conditions can be derived from
Fig. 17, which shows equilibrium moisture contents as a function of oil temperature and
pressure. Under optimal conditions, 0.05-0.06 wt % sodium ethylate or methylate, i.e., 0 .5 -
0.6 kg catalyst per tonne of oil, is sufficient to achieve complete interesterification.
water content (wt %)
Fig. 17 Equilibrium moisture contents of oils and fats as a function of temperature and pressure.
B. Process Options
In the literature, many types of random interesterification processes have been described:
1. Batch, with sodium alcoholates (CH30Na, C2H50Na) or glycerolates (in situ from
NaOH or KOH and glycerol) as catalyst
2. Continuous, with sodium [27], predispersed sodium methylate [4], aqueous sodium
hydroxide [4,49], and aqueous sodium hydroxide plus glycerol [30] as catalyst
3. Semicontinuous, again catalyzed by the sodium hydroxide and glycerol combination
[50]
1. B atch Processes
The sodium or potassium hydroxide-glycerol catalyzed batch process [29,51] consists of es
sentially two stages that are both carried out in vacuum. The first stage is carried out at about
60-80°C and comprises the steps of catalyst addition, catalyst dispersion, and removal of
most of the water added with the catalyst solution. Vigorous boiling and use of a sufficiently
dilute catalyst solution (e.g., weight ratio NaOH: glycerol: water = 1 : 2 : 7 ) are important fac
tors for obtaining good catalyst dispersion. In the second stage of the process, the oil-catalyst
mixture is heated up to first form sodium glycerolate, then sodium diacylglycerolate (Fig. 5),
the active species that catalyzes the rearrangement reaction.
Sodium ethoxide or methoxide catalyzed batch interesterification is applied more widely.
The process can be carried out in conventional neutralizing-bleaching equipment [8,9] pro
vided with adequate facilities for catalyst dosing and oil drying. As explained in Section
VI.A, the neutral oil must first be dried to a moisture content below 0.01 wt %. Care should
be taken to prevent any subsequent ingress of water; i.e., valves and coils should be absolutely
leak-free. Then the catalyst is sucked in, preferably from a nitrogen-flushed dosing funnel,
through a dip pipe that ends below the oil surface near the tip of the stirrer to ensure optimum
catalyst dispersion and fast formation of the genuine active catalyst. The generally preferred
reaction and catalyst addition temperature is between 80 and 100°C, and the allowed reaction
period is normally between 10 and 30 min, although in many cases the actual rearrangement
may be largely completed in a few minutes.
Sodium alcoholates are toxic and highly reactive materials and should therefore be handled
with care. Sodium methylate normally has a higher bulk density, a longer shelf life, and a
somewhat coarser, more uniform particle size distribution than sodium ethylate. It causes less
dust and is easier to handle. For both catalysts it is essential to prevent deterioration of activity
by avoiding contact with moisture and air. The catalysts should therefore be stored under cool
and dry conditions in closed bags and containers until the moment of use. The contents of the
catalyst bags should be such that an integral number of them are used per interesterification
batch; the presence of half-empty bags in the refinery should be avoided. When emptying
catalyst bags, safety goggles, powder masks, and gloves should be worn. Fires should never
be extinguished with water, but with sand and chemical foam.
2. Continuous Processes
The plant designs that have been proposed or actually used for continuous interesterification
are in principle all rather similar:
1. A suitable amount of catalyst is added to a purified and preheated oil stream.

2. Measures are taken to obtain a fine dispersion of catalyst particles.
P O W D E R E D SO D IU M
Fig. 18 Flow chart of continuous interesterification process with sodium methylate catalyst. (From
Ref. 5.)
3. Sufficient residence time is provided to allow for catalyst activation and fatty acid
rearrangement.
4. The catalyst is inactivated (below 100°C) by addition of water followed by separation
of the oil and aqueous phases.
The various processes differ mainly with respect to the applied catalyst and the associated
reaction temperature and regarding the relative importance of thorough drying before or after
catalyst addition. Two examples are shown in Figs. 18 and 19.
The required reaction temperatures are lowest for alkali metals [27] and predissolved metal
alcoholates [4], but, as indicated in Section III, these highly active catalysts also have some
disadvantages in terms of safety of handling and stability in use. Sodium hydroxide, generally
Fig. 19 Flow chart of continuous interesterification process with aqueous sodium hydroxide as cata
lyst. (From Ref. 5.)
sodium hydroxyde/glycerol/w ater
Fig. 20 Principle of continuous sodium hydroxide-glycerol interesterification process.
applied as a 50% aqueous solution, has a rather low activity, and reaction temperatures in the
range of 160-180°C therefore appear to be necessary. The combination of sodium hydroxide
and glycerol is intermediate in activity, and the preferred reaction temperature is therefore
about 1 2 5 -1 4 0 °C [30].
It is essential that immediately after these aqueous catalyst solutions have been dosed, the
mixture of oil and catalyst be thoroughly dried to prevent loss of catalytic activity and exces
sive oil losses due to saponification. This is most conveniently achieved by spray drying in a
vacuum chamber (Fig. 20). In this unit also the genuine interesterification catalyst is formed.
The actual redistribution reaction is very fast and can be completed in a few minutes. It is
preferably carried out in a coil reactor with a narrow residence time distribution to minimize
backmixing effects. A specific feature of this process is that the feedstock need not be care
fully neutralized. Up to a certain level of free fatty acids, neutralization and interesterification
can be combined in a single operation just by increasing the amount of sodium hydroxide
solution added to the oil.
Glycerol in the catalytic solution has been found [11] to suppress the saponification reaction
that is always in competition with the formation of the genuine active catalyst. This is due to
the fact that the equilibrium reaction
glOH + O H " :;± Glo" + H 20
lies far to the right-hand side, as a result of which the OH concentration is lower than in
systems without glycerol.
C. Posttreatment and Process Yield

All interesterification processes with alkaline catalysts have in common that on completed
reaction the catalyst is inactivated by addition of water:
Na diacylglycerolate + H 20-^ NaOH + diacylglycérol
The sodium hydroxide thus formed may cause oil losses due to saponification, especially
when the separation of oil and aqueous phase is delayed, as may be the case in batch pro
cessing. Hawley and Holman [16] showed that coaddition of some acid (citric, phosphoric)
or carbon dioxide minimizes such saponification losses. Further refining steps include conven
tional washing, bleaching, and deodorization.
As a first approximation one may assume that all added catalyst is ultimately converted
into soap. At a catalyst dose of, say, 0.06 wt % sodium methoxide, the theoretical oil loss as
soaps is therefore as low as 0 .3 % . In practice somewhat higher losses will be observed owing
to saponification and entrainment of neutral oil with the water phase (more significant in batch
settling processes than in the case of continuous centrifugal separation). The loss as ethyl or
methyl esters is also equivalent to the amount of catalyst used, again typically 0.3% at a
catalyst dose of 0.06 wt % on oil. Naturally, the actual process step in which the latter loss
occurs is the deodorization, not the washing stage. Sodium hydroxide from NaOH-glycerol
catalyst will also more or less quantitatively be converted to soaps (hence 0.1% NaOH will
yield about 0.7% soaps); the glycerol will be recovered in the rearranged oil as diacylglyc-
erols.
D. Interesterification Control Methods

A large number of analytical methods can be used [6,8,9,52] to establish whether an interes
terification reaction has proceeded to the required degree of completion. The degree of interes
terification is defined as
P -P a
P .- P a
in which P q and P^ represent the value of a characteristic property before and after interesteri
fication (equilibrium situation) and P the actual value. The property P can be either a chemical
or a physical parameter. It should be realized that not all parameters vary in the same way
during an interesterification process, hence the value of a will to some extent depend on the
parameter that is chosen for evaluation. However, these mutual differences become less im
portant when the degree of interesterification approaches 1 (virtually complete randomization,
the normally desired situation).
Chemical methods include
1. Determination of the fatty acid composition at the 2-position (pancreatic lipase
method). On full randomization the 2-position composition will be the same as the
overall composition.
2. Determination of the triacylglycerol composition by chromatographic methods (TLC,
HPLC, capillary gas chromatography, or combinations thereof). A characteristic tri
acylglycerol or a certain class of related triacylglycerols must be selected for quantifi
cation of a.
3. Determination of the free sterol content, for example, by TLC. This is a semiquantita-
tive method based on the observation that sterols take part in the fatty acid exchange
process at a rate comparable to that of triglycerides. Virtual disappearance of free
sterols is thus an indication that TAG interesterification is complete. (Note: Tocopher-
ols also participate in the fatty acid exchange process but at a considerably lower rate.
Hence, interesterification will reduce the free tocopherol content, which may have an
effect on the oxidative stability but not on the vitamin E activity.)
Physical methods for estimating the degree of interesterification are generally based on
changes in melting properties or crystallization behavior and therefore include determination
of
1. Melting point or drop point.
2. Solid-phase content as a function of temperature (NMR).
3. Differential scanning calorimetry.
4. Turbidity. In this case the fat is first melted to destroy all crystal nuclei, then cooled
under standardized conditions to a fixed, optimum temperature (dependent on the na

ture of the mixture). Crystallization will start after a certain period, and the develop
ment of turbidity is determined photometrically.
5. Cooling curve (similar principle as for turbidity, but assessment now based on deflec
tion of cooling curve due to liberation of heat of crystallization).
For all these physical methods it is essential that the measured parameter value be compared
to those of reference samples of identical composition before and after complete interesterifi
cation (e.g., prepared on laboratory scale with excess catalyst).
Several of the above methods, although suitable for a final assessment, are too complex or
time-consuming for process control and application on the factory floor. For the latter purpose
the following strategy is recommended:
1. Ensure that all criticai feedstock and process parameters are clearly specified and ad
hered to (e.g., temperature, time, and vacuum conditions for oil drying; catalyst qual
ity and dosing rate; etc.)
2. Control end point of reaction by using a quick and simple method (e.g., melting point
or crystallization-turbidity test method).
VIL ENZYMATIC INTERESTERIFICATION

A. Introduction
During the past 15 years there has been much interest in the development of enzyme-catalyzed
processes for the interesterification of oils and fats. The selectivity of enzymatic catalysis can
be exploited to generate tailored TAGs that are difficult to obtain by conventional chemical
or physical methods, and the mild conditions used for enzymatic interesterification enable
reactions with unstable materials, such as polyunsaturated fatty acids, to be run without the
generation of by-products associated with the use of more severe chemical procedures.
Extracellular microbial lipases from bacteria, yeasts, or molds are used as catalysts for the
interesterification of oils and fats. The natural function of these enzymes is to catalyze the
hydrolysis of acylglycerols and other fatty acid esters to provide nutrients for the microorgan
isms, but these reactions are easily reversed, and consequently lipases are also effective cata
lysts for various esterification reactions. The properties and industrial applications of lipases
have been described in some recent articles [53-57]. Because of the reversibility of the lipase
reaction, hydrolysis and resynthesis of ester bonds occurs when lipases are incubated with
TAGs, TAGs plus free fatty acids (FFAs), or TAGs plus alcohols, giving the interesterifica
tion, acidolysis, and alcoholysis reactions shown in Fig. 1.
Microbial lipases are produced industrially by growing yeasts, molds, or bacteria in large
fermenters. The enzymes are excreted by the microorganisms in their growth media, from
which, after removal of the cells, they can be recovered by conventional enzyme processing
techniques. In the past, because of low fermentation yields and small markets for the en
zymes, lipases were expensive in comparison to other more widely used extracellular enzymes
such as proteases and carbohydrases. However, in the last few years successful application of
recombinant DNA techniques for the production of lipases and an increasing market for the
enzymes, particularly as additives to laundry products, has enabled enzyme manufacturers to
produce lipases more efficiently. The availability of some cheap lipases is widening their
potential for application as catalysts for interesterification of oils and fats and a variety of
other industrial processes and products.
B. Lipase Catalysts
Enzymatic interesterification reaction systems consist of the lipase catalyst together with a
very small amount of water dispersed in a continuous organic phase comprising the reactants
(usually mixtures of TAGs or TAGs plus F F A) and if necessary a solvent such as hexane.
The water in these microaqueous reaction systems is dissolved in the organic phase and ab
sorbed onto the enzyme catalyst particles, and there is no distinct aqueous phase. Lipases
catalyze reactions at interfaces, and to obtain high interesterification rates a large interfacial
area between the reactants and the more hydrophilic enzyme phase is required. This can be
achieved either by production of a fine dispersion of lipase in the organic phase, e.g., by use
of a surfactant, or by immobilizing the enzyme on the internal surfaces of macroporous sup
port particles. Immobilized lipases are usually preferred for interesterification processes, be
cause immobilization generally improves lipase stability. Immobilized enzyme catalysts can
also be used in packed bed reactors and easily recovered from stirred tank reactors for reuse.
To produce effective interesterification catalysts, appropriate lipases and support materials
must be selected. The main criteria for selection of lipases for use as interesterification cata
lysts are specificity, activity, and stability in the microaqueous reaction system and acceptabil
ity for use in food processing.
Microbial lipases can show specificity with respect to both the fatty acid and alcohol parts
of their substrates [53], and for interesterification processes, lipases with appropriate specific
ity on reaction with acylglycerols must be selected. A large group of microbial lipases includ
ing those produced by Rhizopus, Humicola, Rhizomucor, and Aspergillus species catalyze
reactions regiospecifically at the outer 1- and 3-positions of glycerol, the 2-position being
unaffected (Fig. 21). This type of specificity can be exploited to generate TAGs that are
difficult to obtain by conventional chemical and physical methods. For example, high value
TAGs used in the formulation of confectionery fats can be produced from cheaper oils such
as palm midfraction or high oleate sunflower oil by an interesterification reaction with stearic
acid using a 1,3-regiospecific lipase [58-60] (Fig. 22).
A second group of lipases show no significant regioselectivity and catalyze reactions at all
three glycerol positions (Fig. 21). The lipases produced by Candida rugosa, C. antárctica,
and Geotrichum candidum are examples from this group of enzymes [53,62,63]. When these
lipases are used as interesterification catalysts, the TAG products have a random distribution
of fatty acyl groups and are similar to those produced by chemical interesterification (see
Section II) [64]. Many lipases, such as those from several Pseudomonas species, show regio
selectivity and catalyze reactions faster at the 1,3-positions than at the 2-position of glycerol
[58]. When these lipases are used as interesterification catalysts, prolonged reaction can give
TAGs with a random distribution of fatty acyl groups, but in practice products intermediate
between those formed by 1,3-regiospecific and nonselective lipases are obtained. Recent work
has shown that some microbial lipases can display stereoselectivity with TAGs, i.e., the reac
tion rate at the 1-hydroxy group is significantly different from the rate at the 3-hydroxyl group
[65]. This type of selectivity can be exploited in interesterification reactions to produce chiral
TAGs (Fig. 23) [66]. Chiral TAGs do not occur to any significant extent in most common
oils and fats and are very difficult to produce by conventional physical or chemical techniques.
In general, extracellular microbial lipases show little selectivity as regards the type of fatty
acid released from common oils and fats. Oils containing very long chain polyunsaturated
fatty acids (PUFAs) are exceptional, because these fatty acids are only slowly released from
the oils by some lipases. This selectivity has been used to enrich acylglycerol fractions in
these PUFAs [67]. A lipase excreted by the mold Geotrichum candidum is unusual, because
it shows extreme fatty acid specificity, preferentially releasing unsaturated fatty acids with a
1 ,3 -R e g io sp e cific lip a se
H2COCOR H2COCOR H2COH

: H2O I + H2O
HCOCOR = = ^ HCOCOR + HOCOR HCOCOR + 2 HOCOR
H2 COCOR H2COH H2 COH

TAG 1.2(2,3)-DAG (2-MA
N o n -re g io se le ctive lip a s e
H2 COCOR H2 COCOR H2 COCOR

1 ± 2 H2 O I I
2 HCOCOR ------------ HCOCOR'’ ^ HCOH 2 HOCOR
H2 COCOR H2 COH H2 COCOR
/
± 2H 20
/
H2 COH H2 COH H2 COH
1 ± 2H 20 I I
HCOH + 6 HOCOR ..^ HCOCOR + HCOH ■
H2 COH H2 COH H2 COCOR
Fig. 21 Hydrolysis reactions catalyzed by lipases.
double bond in the A9 position from TAGs. Saturated fatty acids and unsaturated fatty acids
without a double bond in the A9 position are only slowly released [63]. When G. candidum
lipase is used as an interesterification catalyst, only A9 fatty acyl groups are exchanged in
mixtures of TAGs or TAGs plus free fatty acid, and the enzyme has been used to enrich olive
oil in linoleoyl groups at the expense of oleoyl groups, the saturated fatty acyl content of the
oil remaining unchanged [64].
Microbial lipases vary greatly in the activity expressed in interesterification reaction sys
tems. The reactions are run under conditions of low water activity (A^), and one reason for
the variation in interesterification activity lies in the different responses of the enzymes to low
A^. Some immobilized lipases have good catalytic activity at low A^ (<0.2), whereas others
are essentially inactive under these conditions and express high catalytic activity only when
A^ is raised to >0.5 [68,69]. For interesterification catalysts, enzymes in the former category,
e.g., Rhizomucor miehei lipase, are preferred. A second reason for variation in interesterifica
tion activity probably lies in differences between lipases in the kinetic parameters of the
various component reactions of the overall interesterification process (Fig. 24). Subtle varia
tions in protein structure can alter the relative rates of hydrolysis and interesterification reac
tions.
The recovery and reuse of lipase catalysts from interesterification reactors is required for
cost reasons. Consequently lipase stability in the reactors is an important parameter. Loss of
activity during interesterification is caused by two factors: heat inactivation of the lipase and
poisoning by minor components in the reactants [70,71]. Poisoning can be largely prevented
+ 3 STEARIC ACID □
------ 0 + r
------- 0 L 3 PALMITIC ACID
L— p i ----- St I------- St
PALM COCOA BUTTER
MIDFRACTION EQUIVALENT
st r — *St
I---.0
+ 3 STEARIC ACID
fflG H OLEATE
SUNFLOWER OIL
E St
0 . U o
i— 0
COCOA BUTTER
+ 3 OLEIC ACID
EQUIVALENT
»p + 3 UNSATURATED
ACIDS ^ p . p; + 3 PALMITIC ACID
“P L -»u •— ■u
PALM
STEARINE HUMAN MILK
FAT SUBSTITUTE
Fig. 22 Interesterification reactions catalyzed by 1,3-regioselective lipases.
by careful refining of the reactants, but a lipase must be used that is resistant to heat inactiva
tion at the reaction temperature. Operation without a solvent is advantageous, and a reaction
temperature of 70°C is often necessary to ensure that the reactants and products are fully
liquid. Fortunately, the immobilized 1,3-regiospecific Rhizomucor miehei and the nonselective
Hydrolysis
H2COCOR H2COH
H2O
I
R O C O -^ C P - H R O C O -^ C ► H HOCOR
I I
H 2 CO CO R H 2 COCOR
Triacyl-sn-glycerol 2,3-diacyl-sn-glycerol
Interesterification
H2COCOR H2COCORI
I
RO CO ^ C ► H + HOCOR'l RO CO ^ C ► H -{■ HOCOR
I I
H^COCOR H2 COCOR
Chiral TAG
Fig. 23 Reactions catalyzed by a stereospecific lipase.

Candida antárctica lipases are stable at 70°C and can be used as catalysts in solvent-free
processes [62,72]. Most other immobilized lipases are less stable, and the use of a lower
reaction temperature, often with addition of a solvent, is required to maintain catalyst stability.
The most effective catalysts for use in enzymatic interesterification processes are prepared
by adsorption of enzyme onto macroporous particles by ionic or hydrophobic interactions.
Because protein is insoluble in interesterification reaction mixtures, tight irreversible associa
tion of lipase with the support materials is not required, and immobilization by covalent
bonding is unnecessary. The physical characteristics of the support particles have an important
influence on their effectiveness, and some important parameters have been investigated by
Acidolysis
H2COCOR H2COH
\ 1 + H — OCOR
^11 +
^ K — OH \
X X
+ H20 __\ OH + HOCOR

ffi—OCOR
hocor "' OCOR^ + H2 O

H — OH +
H2COH HoCOCOR -OH

OCOR ^ 1
X
Overall reaction
H2COCOR OH H2COCOR
+ HOCOR
+ HOCOR
H2”O X
Interesterification
H 2 COCOR
H2COH
-OH I -OCOR
X
H2C0C0R^ H2COH
-OH 1 -OCOR
Y Y
H2COH H2COCOR'í
OCOR I OH
X
H2COH H2C0C0R'^
-OH
I -OCOR I
Y Y
Fig. 24 Mechanisms of lipase-catalyzed interesterification reactions.

Overall
-OH
H2 COCOR ^ H2 COCOR'* H2 COCOR'i + H 2 COCOR
I I
X Y
HCOCOR HCOCOR'*
X = I ; Y = I
H2COCOR H2COCORI
H “ 0H = ENZYME
OCOR = ACYL-ENZYME COMPLEX
Fig. 24 Continued
Bosley and Clayton [73]. The macroporous particles should have sufficient internal surface
area to allow adsorption of a substantial amount of lipase, and materials with a surface area
in the range of 10-100 m^/g are normally used. The pores of the particles must be large
enough to allow enzyme molecules access to the internal surfaces of the particles. Support
materials with a mean pore diameter of >100 nM are preferred. The chemical nature of the
particle surface is also important. With ionic materials, weak anion exchangers are generally
more effective than cation or strong anion exchangers, while for hydrophobic particles satu
rated surfaces are better than aromatic or unsaturated surfaces. Co-immobilization of lipase
with an inert protein such as sodium caseinate on ion-exchange or hydrophobic particles can
give more effective interesterification catalysts [74]. The presence of another protein probably
reduces the unfolding of the lipase into an inactive form during immobilization.
The mechanical properties and size of the support particles have to be taken into account
when preparing enzymatic interesterification catalysts. In large stirred tank reactors, fragile
brittle materials can be damaged by attrition, and consequently catalysts used in those reactors
must be prepared from support materials that are strong and preferably have some elasticity.
For packed bed reactors, rigid, essentially incompressible support materials are required. Be
cause of mass transfer effects the efficiency of interesterification catalysts decreases when the
particle size is increased above a certain diameter. With large particles the reaction rate is
often limited by the rates of diffusion of the reactants and products within the pores [75]. In
stirred tank reactors, to avoid diffusion effects, small particles (0.1-0.2 mm) can be used,
but in packed bed reactors these give unacceptably high pressure drops across the beds. For
packed reactors a particle size of —0.5 mm is recommended [76]. With particles of this size,
acceptable pressure drops are obtained and any diffusional limitations are small.
Novo-Nordisk produces a highly active interesterification catalyst consisting of Rhizomucor
miehei lipase immobilized on a phenol-formalydehyde weak anion-exchange resin. This cata
lyst is particularly suited for application in packed bed reactors [72]. Very effective interesteri
fication catalysts suitable for use in both stirred tank and packed bed reactors can also be
prepared by immobilization of lipase onto hydrophobic macroporous polypropylene particles
[77].
Because of the high cost of macroporous materials, catalysts prepared by immobilization

of lipases on ion-exchange or hydrophobic particles are expensive. Cheaper catalysts are ob
tained by deposition of lipase onto flux-calcined diatomaceous earth particles. Deposition can
be achieved by solvent precipitation of enzyme in the presence of the particles or by mixing
lipase solution and diatomaceous earth to form a paste that is then vacuum dried at low
temperature [78,79]. Although they are cheaper to prepare, catalysts based on diatomaceous
earth are less active than those based on macroporous ion-exchange or hydrophobic particles.
The diatomaceous earth catalysts are also less suited to application in large-scale reactors
because of their susceptibility to attrition in stirred tanks and excessive pressure drops in
packed beds.
As mentioned previously, high interesterification reaction rates can be achieved by using
surfactants to disperse lipases in reactant mixtures. For one type of system, lipase solution is
dispersed in the continuous organic phase by formation of thermodynamically stable micro
emulsion droplets using surfactants such as AOT [sodium bis-(2-ethylhexyl) sulfosuccinate]
or lecithin [80,81]. In another method, complexes of lipase and a nonionic surfactant such as
sorbitan-monostearate are prepared. The lipase-surfactant complexes are then finely dispersed
or dissolved in an interesterification reaction mixture [82,83]. With both types of reaction
systems, rapid interesterification rates are achieved, but so far neither has been adopted for
large-scale processes. Expensive membrane reactors are required for recovery of the lipase
catalysts for reuse, and contamination of the reaction products by small amounts of surfactant
is a problem.
C. Reaction Systems
The reactants for enzymatic interesterification processes are usually mixtures of TAGs or
TAGs plus FFAs. Fatty acid methyl or ethyl esters can also be used instead of FFAs [60,84-
86]. The advantages of using esters rather than the free acids are that with some lipases faster
reaction rates are obtained and, because the esters have lower melting points, lower reaction
temperatures are possible. The main problems with using esters are their higher cost and the
inevitable occurrence of some hydrolysis during enzymatic interesterification processes. This
gives small amounts of FFA in addition to esters, TAGs, and DAGs as reaction products and
complicates product recovery and by-product recycle procedures.
Enzymatic interesterification without use of a solvent is preferable for cost reasons, but in
some cases a solvent must be added to ensure that the reactants and products are completely
liquid at the required reaction temperature. Lipase catalysts are active in the presence of a
wide range of water-immiscible solvents, but for enzymatic interesterification processes hex
ane should be selected because this solvent is already used industrially for the extraction and
processing of oils and fats. The presence of hydrocarbon solvents in enzymatic interesterifica
tion reaction mixtures has little effect on the reaction rate or product yield [85]. The applica
tion of supercritical carbon dioxide as a solvent in enzymatic interesterification reactions has
been investigated by several groups [87,88]. Good reaction rates were observed, although the
solubility of the reactants in the fluid is rather low. In the future, supercritical carbon dioxide
could be used as an environmentally friendly nontoxic solvent in interesterification processes,
particularly if its unusual properties can be exploited to improve product yields [87].
Enzymatic interesterification reaction systems must contain a small amount of water. This
water has two functions. First, it is required by the enzyme catalyst for the maintenance of
an active hydrated state. As mentioned previously, the amount and thermodynamic activity of
the water needed for this purpose can be very low. Second, water is a reactant that generates
partial glycerides and FFAs from TAGs by hydrolysis. The hydrolysis reactions reduce the
final yield of interesterified TAGs but accelerate the interesterification reactions by raising the
concentration of DAGs, which are essential intermediates (Fig. 24). In operating enzymatic
interesterification processes, the water content of the reaction mixture must be carefully con
trolled to achieve a balance between acceptable reaction rates and product yields.
Enzymatic interesterification reactions can be run using both stirred tank and packed bed
reactors. For large-scale processes, continuous packed bed reactors are preferred because they
give higher product yields and enable more oil to be processed with a given quantity of lipase
catalyst. In stirred tank reactors, catalyst particles can be damaged by attrition and not fully
recovered during recycling procedures, but with packed bed reactors these problems are
avoided and greater catalyst productivity is achieved. When regio- or stereospecific lipases
are used as interesterification catalysts, side reactions resulting in a decrease in the yield of
valuable TAGs occur if long reaction times are used. These side reactions are caused by
migration of fatty acyl groups between the 1,3- and 2-positions of partial glycerides. The acyl
migration reactions in combination with reesterification reactions give additional DAG and
unwanted TAGs (Fig. 25). Because the rate-limiting steps in the side reactions are chemical
rather than enzyme-catalyzed isomerizations of monoglycerides, increasing the lipase catalyst
concentration and reducing the reaction time essentially eliminate the unwanted side reactions
[86,90]. High catalyst concentrations and short contact times between the catalyst and re-
H2COCOR H2COH H2COH

LIPASE LIPASE
HCOCOR HCOCOR HCOCOR
I ± HOCOR / H2O I ± HOCOR / H2O
(1.2(2 ,3)-DAG) (2-MAG)
/I
CHEMICAL
H2COH H2COCOR H2COCOR

1 LIPASE 1 LIPASE I
HCOH ______
\ V 1
HCOH HCOH
I ± HOCOR''/H2O 1 ± HO COR/ H2O I
(1.3-DAG) (1(3)-MAG)
/
CHEMICAL
1/
H2COH H2COCOR H2COCOR ^
LIPASE LIPASE
HCOCOR HCOCOR HCOCOR
: HOCOR V h 2 0 1 HO CO R/ H2O
H2COH H2COH H2COCOR
Fig. 25 A cyl migrations occurring during interesterification catalyzed by 1 , 3 -regiospecific lipases.

actants are characteristic of packed bed reactors, whereas lower catalyst concentrations and
longer reaction times are used in stirred tank reactors.
For operation of packed bed reactors, the reactants are treated to remove lipase catalyst
inhibitors and poisons, and, if required, they are dissolved in hexane. A small amount of
water is dissolved in the reactant mixture, which is then pumped continuously through a bed
of catalyst particles. Operation of a typical reactor is illustrated by interesterification of a
mixture of palm oil midfraction (1 part) and stearic acid (0.4 part) dissolved in hexane. A
small amount of water (0.06%) was dissolved in the reactant mixture, which was then pumped
through a bed of immobilized Rhizomucor miehei lipase to give a mean residence time of
approximately 10 min. The temperature of the reactor was maintained at 40°C. High interes
terification activity as measured by an increase in the stearoyl content of the TAGs was ob
tained through 13 days of continuous reactor operation (Fig. 26). Analysis of the product
showed that, as expected for a 1,3-regiospecific lipase, stearoyl groups were incorporated
exclusively into the 1- and 3-positions of the TAGs with the formation of the valuable TAGs,
1,3-distearoyl-2-oleoyl glycerol (StOSt) and l(3)-stearoyl-3(l)-palmitoyl-2-oleoylglycerol
(POSt) [59] (Table 5). After an initial equilibration period, the DAG and FFA contents of the
product stream remained constant, and 5% additional DAG and 2% additional FFA were
formed as by-products in the reactor (Fig. 26). During the equilibration period, which lasted
for a few hours, water was removed from the hydrated lipase catalyst by the formation of
higher concentrations of DAG and FFA until steady-state conditions were attained. In the
steady state some of the water in the feedstream was consumed to give the additional DAG
and FFA, setting up an equilibrium between TAG, water, DAG, and FFA. Therefore, there
was variation in the water content and thermodynamic water activity (A^) of the reactant and
catalyst phases through the bed (Fig. 27). On entry to the reactor, the feedstream was almost
saturated with water, and at the top of the catalyst bed the of both phases was —0.9.
Limited but rapid hydrolysis occurred in the topmost part of the bed, leading to a reduction
in the water content and A ^ of both phases. Further down the reactor no further hydrolysis
occurred (i.e., equilibrium was reached) and the A ^ of both phases was <0.5. Interesterifica-
Catalyst activity (gTAG/hr/gcatalyst)

5
Reactants; 1part palm midfraction + 0.4parts stearic acid in hexane

Catalyst; Immobilized Rhizomucor miehei lipase.
Temperature; 40°C. Residence time; 10min.
Reactant composition; 70%TAG, 1%D A G , 2 9 % FFA.

Product composition; 63%TAG, 6 % DAG, 31%FFA .
2 4 6 8 10 12 14
Time in operation (days)
Fig. 26 Interesterification o f a mixture of palm midfraction and stearic acid in a packed bed reactor.
Table 5 TAGs Formed by Reaction of a Mixture of Palm Oil Midfraction (1 part) and Stearic Acid
(0.4 part) Using Immobilized Rhizomucor miehei Lipase as Catalyst
Amount in palm oil midfraction (%) Amount in interesterified TAG (%)
Fatty acyl group Total TAG 1,3-Position 2-Position Total TAG 1,3-Position 2-Position
16:0 57.2 8 1.1 9.5 39.4 54.4 9.4

18:0 6.0 8.6 0.7 24.4 36.3 0.7
18:1 32.3 7.8 81.4 3 1.2 5.9 81.8
18:2 3.2 0.8 7.9 3.1 0.7 7.8
Others 1.3 1.7 0.5 1.9 2.7 0.3
TAG^
SSS 2 7
POP 65 28
post 15 37
StOSt 2 15
SSO 6 2
Others 10 1
= saturated fatty acyl group; P = palmitoyl; St = stearoyl; 0 = oleoyl.
tion occurred throughout the whole of the catalyst bed, and therefore the catalyst was active
under the low conditions present in all but the top part of the reactor.
The effect of varying the water content of the feedstream to packed bed interesterification
reactors has been studied [85,91]. Typical results obtained using a mixture of olive oil and
methyl palmitate dissolved in petroleum ether as reactants are given in Table 6. As expected,
both the interesterification reaction rate and by-product (DAG) formation decreased as the
feedstream water content was lowered. With a dry reactant mixture the rate of reaction was
very low and no DAG by-products were formed. For operation of commercial reactors, the
PA LM FRA CTIO N
4- STEARIC ACID
HÔ
INTERESTER IFIED
PRO DUCTS
Fig. 27 profile in a packed bed reactor.

Table 6 Effect of Feedstream Water Content on Interesterification and By-Product

Formation in a Packed Bed Reactor^
Water content of feedstream Interesterification activity of

(% saturation) catalyst (g TAG/h/g catalyst) DAG in product
100 3.4 3.7

52 1.4 2 .0
31 1.0 1.4
10 0.5 < 1.0
0 0.1 < 1.0
^Various concentrations o f water were dissolved in a mixture of olive oil (1 part), methyl palmitate (0.96
part) and 100-120°C petroleum ether (0.85 part). The resulting solutions were pumped through beds of
immobilized Rhizomucor miehei lipase to give a mean residence time of ~ 1 h.
feedstream water content is selected to give the minimum level of by-product formation while
maintaining an acceptable reaction rate.
Immobilized lipase catalysts can be very stable under the conditions found in packed bed
reactors, and operation of reactors at temperatures as high as 70°C is possible when immobi
lized heat-stable lipases such as those produced by Rhizomucor miehei or Candida antárctica
are used [62,72]. An important cause of catalyst inactivation can be the presence of deleteri
ous minor components in the oils such as phospholipids and fat oxidation products [70,71].
These components inhibit or inactivate lipase catalysts, and to obtain good catalyst stability
and productivity, effective refining of the reactants is essential. With well-refined oils and fats
and heat-stable lipase catalysts, reactors can be run continuously at temperatures up to 70°C
for many weeks. During operation of packed bed reactors, a constant product composition is
maintained by varying the flow rate to take account of decreasing catalyst activity. When the
fiow rate becomes unacceptably low because of catalyst inactivation, the reactor is stopped
and the inactivated catalyst is replaced with fresh immobilized lipase.
The progress of interesterification reactions between TAGs and FFA or fatty acid esters
can be followed by the change in the fatty acyl compositions of the various species. The
composition of the interesterified TAGs can also be examined by HPLC or capillary gas
chromatography. For many lipases a very simple pseudo-first-order kinetic model based on
FAME analysis of the TAGs or FFA can be used to calculate interesterification reaction rates
and catalyst activity (Fig. 28). This simple model works well for 1,3-regiospecific or com
pletely nonselective lipases but cannot be used for fatty acid or regio- or stereoselective li
pases. With these types of catalysts, more complex kinetic models are required. When mix
tures of TAGs are used as reactants, the progress of the interesterification reaction can be
monitored by analysis of TAGs by HPLC and/or capillary gas chromatography. With most
mixtures of oils and fats the complexity of the TAG composition is such that simple kinetic
models for calculation of reaction rates cannot be used.
Kinetic models of varying complexity and sophistication have been developed by several
groups to describe enzyme interesterification processes [92-95]. These models can be applied
to reactions catalyzed by different types of lipase and can be used to simulate and predict the
effects of parameters such as water and DAG concentrations, reactant concentrations, mass
transfer limitations, and catalyst activity on reaction rate and product yield. These models will
be useful in the future for designing and predicting the behavior of large-scale reactors under
a variety of conditions.
Stirred tank reactor
k = 1_ \ n ( A ^
t \A - X
catalyst activity = -kV ^ In /A -B \

W Wr \A -X /
Packed bed reactor (PBR)
k = 1 In f A - B '
A-X
catalyst activity kV = V in A-B _ F In /A -B \

W Wr A 'X “ W \A - X j
k reaction rate
t reaction time
r residence time in PBR
A Calculated mole % of reactant fatty acid in TAG at completion of reaction
(t or r = o d ). For non-selective lipases calculated assuming a fully random
distribution o f fatty acyl groups amongst all three positions of TAG and FFA.
For 1,3-regiospecific lipases calculated assuming a random distribution of
fatty acyl groups amongst FFA and 1,3-positions of TAG
B mole % o f reactant fatty acid in TAG at start of reaction (t or r = 0)
X mole % of reactant fatty acid in TAG at time = t or residence time = r
V volume of reaction mixture in reactor
W weight of catalyst In reactor
F flow rate through PBR
Fig. 28 Simple kinetic model for calculation of interesterification reaction rate and catalyst activity.
D. Applications
To date, enzymatic interesterification has been applied for the production of TAGs used for
confectionery fat formulations and nutritional applications. A process for the production of
confectionery fats is outlined in Fig. 29. In this process a small amount of water is dissolved
in a 1:1 mixture of high oleate sunflower oil and stearic acid and the mixture is pumped
continuously through a bed of immobilized Rhizomucor miehei lipase at 70°C. On passage
through the catalyst bed, interesterification of the reactant mixture occurs and stearate residues
are incorporated into TAG largely at the expense of oleate residues, with the formation of
1,3-distearoyl-2-monolein (StOSt), l(3)-monostearoyl-3(l)2-diolein (StOO), and oleic acid
(Fig. 22 and Table 7). The reaction products are then passed through a distillation unit to
remove FFAs, and the resulting acylglycerols are fractionated by crystallization from acetone
to give a stearine containing a high concentration of the valuable TAG, StOSt. This stearine
can be used as a substitute for expensive fats such as shea stearine and illipe butter in the
formulation of cocoa butter equivalents. The FFA by-product of the reaction, which consists
mostly of stearic and oleic acids, can be hydrogenated and then recycled for reaction with
fresh high oleate sunflower oil, so there is no net consumption of stearic acid in the overall
process. The oleine by-product, which contains StOO, can be reacted with additional stearic
fflG H OLEATE
SUNFLO W ER OIL
STEARIC ACID
PACKED BED
REACTOR
DISTILLATION MIXED ACIDS

(STEARIC + OLEIC)
FRACTIONAL
OLEINE BYPRODUCT
CRYSTALLIZATION
(O OO + StOO)
STEARINE
(StO St)
Fig. 29 Process for the production of confectionery fats.
acid to generate more StOSt and increase the overall yield of stearine from the original high
oleate sunflower oil.
Other reactants can be used for the production of TAGs useful as components of confec
tionery fats. In particular, there are many reports on the enzymatic interesterification of mix
tures of palm oil fractions and stearic acid or stearic acid esters to produce fats containing
high concentrations of POSt and StOSt. These TAGs are the main components of cocoa
butter, and enzymatic interesterification processes can produce fats with compositions and
physical properties very similar to cocoa butter (Table 8) [59]. Olive, safflower, and high
oleate rapeseed oils have also been interesterified with stearic acid or methyl stearate to give
TAGs enriched in POSt and StOSt [96-98]. Similar products were obtained by reaction of
shea and sal fats with stearic and palmitic acids or their methyl esters [84,97]. Enzymatic
interesterification of mixtures of high oleate sunflower oil and ethyl esters obtained from
hydrogenated high-erucate rapeseed oil gave fats enriched in 1,3-dibehenoyl-2-monolein and
related very long chain fatty acid (> C 2o) TAGs. These fats are useful as cocoa butter substi
tutes, particularly because it is claimed that they inhibit bloom formation in chocolate prod
ucts [99].
In the area of nutritional fats, enzymatic interesterification is used to produce a human milk
fat substitute for use in infant formula [76]. Interesterification of a mixture of tripalmitin and
unsaturated fatty acids using a 1,3-regiospecific lipase as catalyst gives TAGs derived entirely
Table 7 TAGs Formed by Interesterification of a Mixture of High Oleate

Sunflower Oil and Stearic Acid
Amount in TAG fraction (%)
High oleate Reaction product

Component^ sunflower oil before fractionation Stearine
SSS — 1 2
SOS 2 36 79
sso — 1 3
SLnS — 3 5
so o 26 41 10
OOO + others 72 18 1
^S = Saturated fatty acyl group; O = oleoyl; Ln = linoleoyl.
from vegetable oils that are rich in 2-position palmitate with unsaturated fatty acyl groups in
the 1- and 3-positions (Fig. 22). These TAGs closely mimic the fatty acid distribution found
in human milk fat, and when they are used in infant formula instead of conventional fats the
presence of palmitate in the 2-position of the TAGs has been shown to improve digestibility
of the fat and absorption of other important nutrients such as calcium.
Enzymatic interesterification can also be used for production of oils and fats containing
nutritionally important PUFAs, such as eicosapentaenoic acid (EPA) and docosahexaenoic
acid (DHA). For example, various vegetable and fish oils have been enriched in EPA and
DHA by enzymatic interesterification reactions [100,101]. Use of the technique to produce
structured lipids with medium-chain fatty acids (Cg and Cjo) and PUFAs located specifically
in either the 2- or 1,3-positions of the TAGs has been described [89,102,103]. Lipids of this
type have been reported to have beneficial effects on a range of metabolic parameters, includ
ing immune function, nitrogen balance, and improved lipid clearance from the bloodstream
[76]. Enzyme processes are particularly suitable for the production and modification of lipids
containing PUFAs, because these unstable fatty acids are susceptible to damage under the
more severe conditions used for chemical processing.
Table 8 Compositions of Cocoa Butter and a Confectionery Fat Produced by

Enzymatic Interesterification ^
Amount in enzymatically produced

TAG Amount in cocoa butter (%) confectionery fat (%)
POP 16 16
post 41 39
StOSt 27 28
SLnS 7 8
SOO 6 4
Others 3 5
^The confectionery fat was produced by interesterification o f a mixture of palm oil midfraction (1
part) and stearic acid (0.5 part) using immobilized Rhizomucor miehei as catalyst. After removal
of the FFAs by distillation, the interesterified glycerides were fractionated from acetone to give the
confectionery fat as a midfraction.
The possible application of enzymatic interesterification for the production of lower value
nonspecialty oils and fats such as margarine hardstocks and cooking oils has been investigated
by several groups. When nonspecific lipases such as Candida rugosa and C. antárctica are
used as catalysts for interesterification of oil blends, the TAGs produced are very similar to
those obtained by chemical interesterification [64,104]. Therefore, replacement of chemical
interesterification by an enzyme process giving similar products is technically feasible, al
though it has not been adopted on a commercial scale to date largely because of the compara
tively high process and catalyst costs.
If regio- or stereospecific lipases are used to interesterify oil blends, the products formed
are different from those obtained by chemical interesterification, and they can have better
functional properties. For example, interesterification of blends of canola and palm oils using
the 1,3-regiospecific Rhizopus delemar lipase as catalyst gave oils with improved fluidity
compared with the original blends or chemically interesterified products. Because of their
higher fluidity the enzymatically interesterified oils were most effective as frying oils [105],
Interesterification of blends of palm and hyrogenated canola oils and cottonseed and hydro
genated soybean oils using 1,3-regiospecific lipases as catalysts gave fats with a low trans
acid content that were effective as margarine hardstocks [106,107]. Reaction of mixtures of
palm stearine and lauric fats using immobilized Rhizomucor miehei as catalyst also produced
fats that were functional as margarine hardstocks [72,108]. With these enzymatically interes
terified fats, margarine could be formulated without using hydrogenated fats.
Because of high catalyst and processing costs, enzymatic interesterification has not been used
to date for the large-scale commercial production of margarine hardstocks, cooking oils, or other
nonspecialty oils and fats. The technique will be adopted for the manufacture of these types of
product only if the comparatively high processing costs can be offset by addition of extra value
to the oils and fats. Higher value could come from improved functionality as indicated above or
issues such as naturalness. In the future development of novel improved processing techniques
and more effective and cheaper lipase catalysts could increase the opportunities for application
of enzymatic interesterification in the manufacture of edible oils and fats.
REFERENCES
N. O. V. Sonntag, Fat splitting, esterification and interesterification, in Bailey’s Industrial Oil and
Fat Products, 4th ed., Vol. 2, Wiley-Interscience, New York, 1979, pp. 97-173.
2. K. S. Markley, Esters and esterification, in Fatty Acids: Their Chemistry, Properties, Production
and Uses, 2nd ed., Interscience, New York, Vol 2, 1961, pp. 757-984.
K. S. Markley, Esters and esterification, in Fatty Acids: Their Chemistry, Properties, Production
and Uses, 2nd ed.. Interscience, New York, Vol. 5, 1968, pp. 3459-3560.
4. R. O. Feuge, Derivatives of fats for use as foods, J. Am. Oil. Chem. Soc. 39: 521 (1962).
5. L. H. Going, Interesterification products and processes, J. Am. Oil Chem. Soc. 44: 414A (1967).
6. DGF (German Society for Fat Research), Gemeinschaftsarbeiten der DGF, 57. Mitteilung, Die
Umesterung von Speisefetten I, II, III, Fette, Seifen, Anstrichm. 75: 467 , 587 , 663 (1973).
J. W. E. Coenen, Fractionement et interesterification des corps gras dans la perspective du marché
mondial des matières premières et des produits finis. II. Interesterification, Rev. Fr. Corps Gras
21: 403 (1974).
8. B. Sreenivasan, Interesterification of fats, J. Am. OU Chem. Soc. 55: 796 (1978).
9. S. J. Laning, Chemical interesterification of palm, palmkemel and coconut oils, J. Am. Oil Chem.
Soc. 62: 400 (1985).
10. F. V. K. Young, Interchangeability of fats and oils, J. Am. Oil Chem. Soc. 62: 372 (1985).
11. A. Rozendaal, Interesterification of oils and fats, Proc. AOCS World Conf. on Edible Fats and
Oils Processing, Champaign, IL, 1990, pp. 152-157.
12. H. Brockerhoff, A stereospecific analysis of triglycerides, J. Lipid Res. 6: 10 (1965).

13. H. Brockerhoff, Stereospecific analysis of triglycerides; an alternative method, J. Lipid Res. 8:
167 (1967).
14. R. J. van der Wal, Calculation of the distribution of the saturated and unsaturated acyl groups in
fats from pancreatic lipase hydrolysis data, J. Am. Oil Chem. Soc. 37: 18 (1960).
15. M. H. Coleman and W. C. Fulton, Enzymes of Lipid Metabolism, Pergamon, New York, 1961,
p. 127.
16. H. K. Hawley and G. W. Holman, Directed interesterification as a new processing tool for lard,
J. Am. Oil Chem. Soc. 33: 29 (1956).
17. G. Hoffmann, The Chemistry and Technology of Edible Oils and Fats and Their High Fat Prod
ucts, Academic Press, New York, 1989, pp. 264-277.
18. F. A. Norris and K. F. Mattil, Interesterification reactions of triglycerides. Oil Soap 23: 289
(1946).
19. M. Naudet and P. Desnuelle, Phenomena of transesterification between triglycerides. III. Detailed
study of the equilibria of transesterification, Bull. Soc. Chim. Fr. 14: 323 (1947).
20 . A. Kuksis, M. J. McCarthy, and J. M. R. Beveridge, Triglyceride composition of native and
rearranged butter and coconut oils, J. Am. Oil Chem. Soc. 41: 201 (1964).
21 . D. Chobanov and R. Chobanova, Alterations in glyceride composition during interesterification of
mixtures of sunflower oil with lard and tallow, J. Am. Oil Chem. Soc. 54: 47 (1977).
22 . G. R. List, T. L. Mounts, F. Orthoefer, and W. E. Neff, Preparation and properties of zero trans
soyabean oil margarines, J. Am. Oil Chem. Soc 72: 379, 383 (1995).
23. L. P. Klemann, K. Aji, M. M. Chaysam, R. P. D ’Amelia, J. M. Henderson, A. S. Huang,
M. S. Otterbum, and R. G. Yarger, Random nature of triacylglycerols produced by the catalyzed
interesterification of short- and long-chain fatty acid triglycerides, J. Agric. Food Chem. 42: 442
(1994).
24. H. P. Kaufmann, F. Grandel, and B. Grothues, Interesterification in fats. I. Theoretical back
ground and literature. Interesterification by hydrogenation, Fette, Seifen, Anstrichm. 60: 919
(1958).
25. M. W. Formo, Ester reactions of fatty materials, J. Am. Oil Chem. Soc. 31: 548 (1954).
26. M. G. A. Willems and F. B. Padley, Palm oil: quality requirements from a customer’s point of
view, J. Am. Oil Chem. Soc. 62: 454 (1985).
27. W. Herman de Groot and M. H. Hilder (Unilever Ltd), Preparation and use of alkali metal disper
sions, Br. Patent 1,236,233 (1968).
28. C. W. Placek and G. W. Holman, Directed interesterification of lard, Ind. Eng. Chem., Ind. Ed.
49: 162 (1957).
29. J. Burgers, C. W. Molt, and P. Seiden (Procter and Gamble Co.), Interesterification process, U.S.
Patent 3,170,798 (1965).
30. C. N. M. Keulemans and G. Smits (Lever Brothers Co), Interesterification process and apparatus,
U.S. Patent 4,585,593 (1986).
31. M. P. Jordan de Urries and R. M. Utrilla, Studies on transesterification. I. Kinetics and mecha
nism of the reaction between fatty acid monoesters catalysed by alkali metals, Grasas Aceites 21:
281 (1970).
32. T. J. Weiss, G. A. Jacobson, and L. H. Wiedermann, Reaction mechanisms of sodium methoxide
treatment of lard, J. Am. Oil Chem. Soc. 38: 396 (1961).
33. M. Naudet, Modifications de structure des corps gras, B. Modifications affectant leur structure
glycéridique. Rev. Fr. Corps Gras 21: 35 (1974).
34. M. Naudet, Interesterification et esterification. Mécanismes réactionnels et conséquences sur la
structure, Rev. Fr. Corps Gras 23: 387 (1976).
35. J. A. Heidal and P. C. M0 rk, On the mechanism of the base catalysed interesterification of fats,
Proc. 11th Scand. Symp. on Lipids, 1982, pp. 147-152.
36. P. van der Plank, unpublished results.
37. I. P. Freeman, Interesterification. I. Change of triglyceride composition during the course of inter
esterification, J. Am. Oil Chem. Soc. 45: 456 (1968).
38. A. Steenhoek, unpublished data cited in Refs. 7 and 11.

39. H. Konishi, W. E. Neff, and T. L. Mounts, Chemical interesterification with regioselectivity for
edible oils, J. Am. Oil Chem. Soc. 70: 411 (1993).
40. E. W. Eckey, Directed interesterification in glycerides, Ind. Eng. Chem. Ind. Ed. 40: 1183
(1948).
41. C. Schmulinzon, A. Yaron, and A. Letan, Directed transesterification of hardened coconut oil
with several saturated fatty acid methyl esters, Riv. Ital. Sostanze Grasse 48: 440 (1971).
42. G. Barsky, Restructuring glyceride oils, U.S. Patent 2,182,332 (1940).
43. R. C. A. Lago and L. Hartman, Directed interesterification of a Brazilian palm oil and analysis of
the original and interesterified oil and its fractions, J. Sci. Eood Agric. 37: 689 (1986).
44. E. W. Eckey and M. W. Formo, Directed interesterification in glycerides. Reaction in presence
of added glycerol, J. Am. Oil Chem. Soc. 26: 207 (1949).
45. R. de Lathauwer, M. van Opstal, and A. J. Dijkstra (Safinco BE), Process for the directed interes
terification of a triglyceride oil or oil mixture, U.S. Patents 4,284,578 (1981) and 4,419,291
(1983).
46. H. R. Kattenberg, Beschleunigung der gelenkten Umesterung mittels periodischer, variabeler Tem-
peraturfuehrung, Fette, Seif en, Anstrichm. 76: 19 (1974).
47. A. J. Haighton and H. R. Kattenberg (Lever Brothers Co), Interesterification process for glycer
ides, U.S. Patent 3,855,254 (1975).
48. D. G. Chobanov and M. R. Topoalova, Alterations in glyceride composition during directed inter
esterification of lard, J. Am. Oil Chem. Soc. 56: 581 (1979).
4^. G. W. Holman, T. B. Laycock, and R. L. Wille (Procter and Gamble Co), Catalytic interesterifi
cation, U.S. Patent 2,738,278 (1956).
50. C. N. M. Keulemans and A. Rozendaal (Lever Brothers Co), Process and apparatus for the interes
terification of a triglyceride oil and products therefrom, U.S. Patent 4,790,962 (1988).
51. W. E. Dominick and D. Nelson (Swift and Co), Lard crystal modification process, U.S. Patent
2,625,485 (1953).
52. R. Duterte, Méthodes analytiques applicables au contrôle de l’estérification et de l’interestérifica
tion, Rev. Fr. Corps Gras 23: 547 (1976).
53. A. R. Macrae and R. C. Hammond, Present and future applications of lipases, Biotechnol. Gen.
Eng. Rev. 3: 193-217 (1985).
54. P. Wooley and S. F. Petersen, Lipases— Their Structure, Biochemistry and Application, Cam
bridge Univ. Press, 1994.
55. S. E. Godtfredsen, Microbial lipases, in Microbial Enzymes and Biotechnology, 2nd ed., W. M.
Fogarty and C. T. Kelly, eds., Elsevier Applied Science, London, 1990, pp. 255-274.
56. M.-P. Egloff, S. Ransac, F. Marguet, E. Rogalska, H. van Tilbeurgh, G. Buono, C. Cambillau,
and R. Verger, Enzymes lipolytiques et lipolyse. Oléagineux Corps Gras Lipides, 2: 52 (1995).
57. P. Adlercreutz, Enzyme-catalysed lipid modification, Biotechnol. Gen. Eng. Rev. 12: 231-254
(1994) (see ref. 53).
58. M. Matori, T. Asahara, and Y. Ota, Positional specificity of microbial lipases, J. Ferment. Bi-
oeng. 72: 397 (1991).
59. A. R. Macrae, Interesterification of fats and oils, in Biocatalysis in Organic Syntheses, J. Tramper,
H. C. van der Plas and P. Linko, eds., Elsevier, Amsterdam, 1985, pp. 195-208.
60. S. Bloomer, P. Adlercreutz, and B. Mattiasson, Triglyceride interesterification by lipases. 1. Co
coa butter equivalents from a fraction of palm oil, J. Am. Oil Chem. Soc. 67: 519 (1990).
61. A. R. Macrae, Modifying oils— enzymic methods, in Oils and Fats in the Nineties, V. K. S.
Shukla and F. D. Gunstone, eds., Int. Food Sci. Centre, Lystrup, Denmark, 1992, pp. 199-208.
62. H. P. Heldt-Hansen, M. Ishii, S. A. Patkar, T. T. Hansen, and P. Eigtved, A new immobilised
positional non-specific lipase for fat modification and ester synthesis, in Biocatalysis in Agricul
tural Biotechnology, J. R. Whitaker and P. E. Sonnet, eds., American Chemical Society, Wash
ington, DC, 1989, pp. 158-172.
63. E. Charton and A. R. Macrae, Substrate specificities of lipases A and B from Geotrichum can-
didum CMICC 335426, Biochim. Biophys. Acta 1123: 59 (1992).
64. A. R. Macrae, Lipases catalysed interesterification of oils and fats, J. Am. Oil Chem. Soc. 60:
243A (1983).
65. E. Rogalska, C. Cudrey, F. Ferrato, and R. Verger, Stereoselective hydrolysis of triglycerides by
animal and microbial lipases. Chirality 5: 24 (1993).
66. I. C. Chandler and P. T. Quinlan (Loders Croklaan BV), Enzymic triglyceride conversion, PCX
Patent, WO 9410326 (1994).
67. K. D. Mukherjee, Fractionation of fatty acids and other lipids via lipase-catalysed reactions,
Oleag. Corps Gras Lipides 2: 365 (1995).
68 . R. H. Valivety, P. J. Hailing, A. D. Peilow, and A. R. Macrae, Lipases from different sources
vary widely in dependence of catalytic activity on water activity, Biochim. Biophys. Acta 1122:
143 (1992).
69. R. H. Valivety, P. J. Hailing, and A. R. Macrae, Rhizomucor miehei lipase remains highly active
at water activity below 0.0001, FEBS Lett. 301: 258 (1992).
70. Y. Wang and M. H. Gordon, Effect of phospholipids on enzyme-catalysed transesterification of
oils, J. Am Oil Chem. Soc. 68: 588 (1991).
71. Y. Ohta, T. Yamane, and S. Shimizu, Inhibition and inactivation of lipase by fat peroxide in the
course of batch and continuous glycerolysis of fat by lipase, Agric. Biol. Chem. 53: 1885 (1989).
72. L. H. Posorske, G. K. LeFebvre, C. A. Miller, T. T. Hansen, and B. L. Glenvig, Process
considerations of continuous fat modification with an immobilised lipase, J. Am. Oil Chem. Soc.
65 : 922 (1988).
73. J. A. Bosley and J. C. Clayton, Blueprint for lipase support: use of hydrophobic controlled-pore
glasses as a model. Biotech. Bioeng. 43: 934 (1994).
74. J. A. Bosley (Unichema Chemie BV), Supported enzyme, Eur. Patent 0424130 (1991).
75. H. J. Jung and W. Bauer, Determination of process parameters and modelling of lipase-catalysed
transesterification in a fixed bed, Chem. Eng. Technol. 15: 341 (1992).
76. P. Quinlan and S. Moore, Modification of triglycerides by lipases: process technology and its
application to the production of nutritionally improved fats, INFORM 4: 580 (1993).
77. A. R. Macrae, J. A. Bosley, and A. D. Peilow (Unilever NV), Process for the preparation of fatty
acid esters, Eur. Patent 0322213 (1989).
78. R. A. Wisdom, P. Dunnill, and M. D. Lilly, Enzymic interesterification of fats: the effect of non
lipase material on immobilised enzyme activity. Enzyme Microb. Technol. 7: 567 (1985).
79. T. Matsuo, N. Sawamura, Y. Hashimoto, and W. Hashida (Fuji Oil Co. Ltd.), Method for en
zymic interesterification of lipid and enzyme used therein, Eur. Patent 0035883 (1984).
80. K. Holmberg and E. Osterberg, Enzymatic transesterification of a triglyceride in microemulsions,
Prog. Colloid Polym. Sci. 74: 98 (1987).
81. A. G. Marangoni, R. D. McCurdy, and E. D. Brown, Enzymatic interesterification of triolein
with tripalmitin in canola lecithin-hexane reverse micelles, J. Am. Oil Chem. Soc. 70: 737 (1993).
82. S. Basheer, K.-I. Mogi, and M. Nakajima, Surfactant-modified lipase for the catalysis of the
interesterification of triglycerides and fatty acids. Biotech. Bioeng. 45: 187 (1995).
83. M. Goto, M. Goto, N. Kamiya, and F. Nakashio, Enzymatic interesterification of triglyceride
with surfactant-coated lipase in organic media. Biotech. Bioeng. 45: 21 (1995).
84. T. Matsuo, N. Sawamura, Y. Hashimoto, and W. Hashida (Fuji Oil Co. Ltd.), Producing a cacao
butter substitute by transesterification of fats and oils, U.K. Patent 2035359 (1980).
85. A. R. Macrae, Tailored triacylglycerols and esters, Biochem. Soc. Trans. 17: 1146 (1989).
86. S. Bloomer, P. Aldercreutz, and B. Mattiasson, Triglyceride interesterification by lipases: 2. Reac
tion parameters for the reduction of trisaturated impurities and diglycerides in batch reactions.
Biocatalysis 5: 145 (1991).
87. K. Nakamura, Y. M. Chi, Y. Yamada, and T. Yano, Lipase activity and stability in supercritical
carbon dioxide, Chem. Eng. Commun. 45: 207 (1986).
88 . D. A. Miller, H. W. Blanch, and Y. M. Prausnitz, Enzymatic interesterification of triglycerides
in supercritical carbon dioxide, Ann. N.Y. Acad. Sci. 613: 534 (1990).
89. A. Shishikura, K. Fujimoto, T. Suzuki, and K. Arai, Improved lipase-catalysed incorporation
of long-chain fatty acids into medium-chain triglycerides assisted by supercritical carbon dioxide
extraction, J. Am. Oil Chem. Soc. 71: 961 (1994).
90. A. R. Macrae and P. How (Unilever N.V.), Interesterification with a lipase enzyme as an interes
terification catalyst, Eur. Patent 0093602 (1983).
91. P. Forssell, P. Parovouri, P. Linko, and K. Poutanen, Enzymatic transesterification of rapeseed
oil and lauric acid in a continuous reactor, J. Am. Oil Chem. Soc. 70: 1105 (1993).
92. H. J. Jung and W. Bauer, Determination of process parameters and modelling of lipase-catalysed
transesterification in a fixed bed, Chem. Eng. Technol. 15: 341 (1992).
93. S. Basheer, K. Mogi, and M. Nakajima, Interesterification kinetics of triglycerides and fatty
acids with modified lipase in n-hexane, J. Am. Oil Chem. Soc. 72: 511 (1995).
94. H. R. Reyes and C. G. Hill, Kinetic modelling of interesterification reactions catalysed by immo
bilised lipase. Biotech. Bioeng. 43: 171 (1994).
95. S. R. Moore and J. Davies, The kinetics of lipase catalysed interesterification of fats and oils,
Inst. Chem. Eng. Res. Event 1: 526 (1994).
96. T. Tanaka, E. Ono, and K. Takinami (Ajinimoto Co. Ltd.), Method of producing improved
glyceride by lipase, U.S. Patent 4,275,011 (1981).
97. T. Matsuo, N. Sawamura, Y. Hashimoto, and W. Hashida (Fuji Oil Co. Ltd.), Method for
modification of fats and oils, U.S. Patent 4,420,560 (1983).
98. T. Gitlesen, I. Svensson, P. Adlercreutz, B. Mattiasson, and J. Nilsson, High oleic acid rapseed
oil as starting material for the production of confectionery fats via lipase-catalysed transesterifica
tion, Ind. Crops Prod. 4: 167 (1995).
99. N. Sagi, T. Izumi, H. Kida, and H. Maeda (Fuji Oil Co. Ltd.), Hard butter composition, Eur.
Patent 0227364 (1985).
100. G. G. Haraldsson and P. A. Hostuldsson, The preparation of triglycerides highly enriched with
CO-3 polyunsaturated fatty acids via lipase catalysed interesterification, Tetrahedron Lett. 30:
1671 (1989).
101. K. H. Huang and C. C. Akoh, Lipase-catalysed incorporation of n-3 polyunsaturated fatty acids
into vegetable oils, J. Am. Oil Chem. Soc. 71: 1277 (1994).
102. F. Bjorkling, S. E. Gotfredsen, and O. Kirk, The future impact of industrial lipases. Trends
Biotechnol. 9: 360 (1991).
103. C. C. Akoh, Structured lipids— enzymatic approach, INFORM 6: 1055 (1995).
104. P. Kalo, P. Parviiainen, K. Vaara, S. Ali-Yrrko, and A. Antila, Changes in the triglyceride
composition of butter fat induced by lipase and sodium methoxide catalysed interesterification
reactions, Milchwissenschoft 41: 82 (1986).
105. J. Kurashige, N. Matsuzaki, and Takahashi, Enzymatic modification of canola/palm oil mixtures:
effects on the fluidity of the mixture, J. Am. Oil Chem. Soc. 70: 849 (1993).
106. C. France and J. M. deMan, Physical properties and composition of low trans canola/palm blends
modified by continuous enzymic interesterification, Elaeis 6: 39 (1994).
107. H. M. A. Mohamed and K. Larsson, Modification of fats by lipase interesterification. 2. Effect
on crystallisation behaviour and functional properties. Fat Sci. Technol. 96: 56 (1994).
108. J. Muderhwa, M. Pina, and J. Graille, Interestérification catalysée par les lipase régiosélectives
I. 3 en milieu fondu: applications à l’huile de palme et à sa fraction concrète, Rev. Fr. Corps
Gras 36: 11 (1989).
10_________________________________
Hydrogenation of Edible Oils
Technology and Applications
Wicher T. Koetsier
Unichema International, Emmerich, Germany
I, INTRODUCTION
The hydrogenation of triglyceride oils is a heterogeneously catalyzed reaction that is induced
commercially in three-phase reactors. The three phases are oil, hydrogen gas, and solid, po
rous, Ni catalyst particles. In principle, the following reactor configurations can be used: tank
reactors (well-stirred autoclaves or loop reactors), bubble columns, or fixed bed reactors.
Patterson [1] gives an extensive overview of early and current reactor designs.
Autoclaves operated batchwise are the preferred type of reactor for edible oil hydrogenation
and are most likely used in more than 95% of all edible oil hydrogenation processes. In some
countries, for example in the former USSR, continuous fixed bed trickle flow reactors or a
continuous process in a sequence of three autoclaves in series are used. The trickle flow
process requires the use of dedicated catalyst particles, otherwise the quality of the partially
hydrogenated product will be unacceptable.
In the hydrogenation of triglyceride oils using nickel catalysts various reactions occur on
the nickel surface. These include consecutive hydrogenation reactions as depicted by the sim
ple reaction equation
+ H, + H,
B D
In edible oil hydrogenation it is important that a high quality product be obtained within a
reasonable time. In practice this means an acceptable rate of the hydrogenation reaction,
appropriate selectivity of the chemical reactions, and adequate ease of removal of the catalyst
from the oil by filtration. To obtain an acceptable hydrogenation rate one has the option of
controlling the concentration of the reacting components, the temperature, and the amount
and quality of the catalyst. The concentration of the oil molecules is a given fact, but the
mean concentration of hydrogen in the bulk of the oil phase is determined by several factors
that need to be well understood.
265
266 Koetsier
A main objective of this chapter is to discuss two correlations: the hydrogenation rate and
the hydrogenation time required to drop the initial iodine value, IV^, to a specific end iodine
value, IVend-
The overall rate of the hydrogenation process [which is usually expressed as AlV/min] is
reflected by the rate at which the hydrogen is absorbed by the oil. It can be derived from the
definition of the iodine value, that
1 IV = 36 mol H 2/mîi
and
1 IV/min = 0.6 mol H 2/(mîi • s)
which generates = 17 kcal/(m^ji • s)
= 71 kJ/(m^ji • s) or kW/mîj energy
This overall rate is determined by a number of physical steps, such as transfer of hydrogen
from the gas phase to the oil phase and diffusion of hydrogen through the intricate texture of
the catalyst particles, and the reaction rate on the nickel surface.
These aspects have been discussed in many papers and are at least qualitatively well under
stood by the hydrogenators. However, in many papers the hydrogenation rate results, depicted
as IV versus hardening time, for example, are presented in graphs containing the data not
only of relevant parameters such as temperature, hydrogen pressure, and amount of catalyst,
but also of parameters that cannot be easily interpreted for use in other systems, for example
stirring speed. This latter parameter does indeed greatly affect the hydrogenation rate, but the
stirring speed is not a meaningful parameter because its effect is also determined by the size,
type, and position of the impeller in the vessel.
It is the intention of this chapter to review the following aspects:
• In the first part the technology of hydrogenation is discussed. For example, it is ex
plained which parameter should be used as a meaningful indication of the effect of the
hydrogen transfer rate for which we use the concept of minimum hydrogenation time.
• In the second part of a number of aspects are reviewed that determine the properties of
the hydrogenated product.
The filtration rate is determined by particle size distribution, catalyst particle morphology, and
the presence of filter aid and of components in the oil that easily absorb onto the catalyst
particles and affect the characteristics of the filter cake, such as soaps and phospholipids.
These aspects are not reviewed in this chapter.
II. TECHNOLOGY OF HYDROGENATION

In the following paragraphs a simple general equation for the overall rate of the hydrogenation
process is derived that we have used successfully to study catalyst activity in a number of
different types of reactors. Moreover, we discuss whether the results obtained in small-scale
reactors can be applied to large-scale reactors.
A. Transfer of H2 from the Gas to the Catalyst Phase

In edible oil hydrogenation the solubility (maximum concentration at a given pressure and
temperature) of hydrogen in the oil is much lower than the concentration of the double bonds.
From Ostwald solubility coefficients given in Bailey [2] or from the data published by Anders-
Hydrogenation of Edible Oils 267
son et al. [3], it can be derived that at 1 bar the following hydrogen concentrations exist at
equilibrium in an edible oil.
[2 ] [3]
Temp. (mol/m^) (mol/m^)
100°C 2.60 2.76

120°C 2.76 2.91
140°C 2.92 3.05
160°C 3.07 3.2
180°C 3.20 3.3
200°C 3.36 3.4
Due to the low solubility and the low pressure that are used in these hydrogenation processes,
the concentration of hydrogen is always some orders of magnitude lower than the concentra
tion of the molecules to be hydrogenated. For example, in the hydrogenation of a fish oil with
an IV q of 180 the overall concentration of the —C = C — bonds is on the order of 7000 mol/
môii, while at 5 bar and 180°C the hydrogen solubility is only 16 mol/môji.
Such substantial differences in concentration between the two components have two im
portant consequences. First, hydrogen has to be continuously supplied to the liquid phase by
effectively dispersing a large volume of hydrogen bubbles through the liquid. Second, the
hydrogen concentration in the bulk liquid phase, will always be lower than its solubil
ity, Cjâx? because this bulk concentration, as explained in the next sections, is determined by
the equilibrium of two processes, namely, (1) the rate at which hydrogen is transferred into
the oil phase and (2) the reaction and diffusion rate within the catalyst particles—the rate at
which hydrogen is removed again from the oil phase to the catalyst phase.
The bulk liquid-phase concentration is one of the important factors determining the hydro
genation rate as well as the selectivity. Because there are no analytical techniques available
to measure this hydrogen concentration in the bulk liquid phase, it cannot be continuously
monitored. It is possible to calculate this concentration throughout the hydrogenation reaction,
but because certain assumptions have to be made there will always remain uncertainty as to
its correct value. This hydrogen concentration is not necessarily constant during hydrogena
tion. For example, during isothermal hydrogenation in which the hydrogen gas is continuously
recirculated by using a compressor, the bulk concentration will continuously increase because
the reaction rate decreases as hydrogenation proceeds.
These aspects are discussed first before coming to the general formulas expressing the
overall hydrogenation rate.
B. Rate of Hg Transfer from the Gas to the Liquid Phase

The rate J [4], expressed in moles per cubic meter of oil per second, at which the hydrogen
is transferred from the dispersed hydrogen bubbles to the bulk oil phase is directly propor
tional to the driving force (Cjâx~^buik):
J kL^(Cjjjax ^bulk) ( 1)
In this equation is the maximum concentration at the temperature and hydrogen pressure
used in the reactor. Its value can be calculated from the pressure and the solubility of hydro
gen by using the Ostwald solubility coefficients. In fact, this concentration exists at the gas/
268 Koetsier
liquid interface, but within a thin layer it drops to the hydrogen concentration Cûik, which
exists in the bulk of the liquid. The product kâ (1/s) is the characteristic rate constant. In
chemical engineering practice this parameter is often referred to as the volumetric liquid-side
mass transfer coefficient. It is the product of the liquid-phase mass transfer coefficient (m/s)
and the specific interfacial area of the hydrogen bubbles in the oil phase. For
various reasons it is important to understand which factors affect this characteristic rate con
stant. As well as for commercial reactor design, upscaling, and the choice of the lab reactor
size, for example, used to determine the catalyst performance, it is important to properly
understand what determines this characteristic rate constant. For example, important questions
are whether kâ remains constant during the entire hydrogenation process and whether it is a
function of the location within the reactor. In many reactor types this is the case. But before
going on to answer these questions, let us first review the parameters determining this charac
teristic rate constant.
One last remark: This equation is the equation to use if one wants to calculate the hydrogen
concentration within the bulk liquid phase at any time t. However, it requires that the volumet
ric liquid-side mass transfer coefficient, kija, be determined first. Then
^bulk(0 ^max J(t)lk iJ2 (2 )
L The Liquid-Phase M ass Transfer Coefficient k^^

The liquid-phase mass transfer coefficient, ^l , is usually expressed as the dimensionless Sher
wood number Sh = kidfD n^ and is determined from:
1. The diffusion coefficient of hydrogen in oil, several authors have published data
on the value of this coefficient as a function of temperature [5] . It appears that its value
is on the order of 1 x 10~^-3 x 10“ ^ m^/s, within a temperature range of 10 0 -20 0 °C .
2. The hydrodynamics of the liquid flowing around a free rising bubble having a diameter
and a velocity in a highly turbulent liquid; these hydrodynamics are determined in
particular by the conditions at the gas/liquid interface and the effect of the turbulence.
The hydrodynamics are in most cases not well understood, but in general two extreme cases
exist for free rising bubbles, one for small (d^=<3 mm) bubbles and one for bubbles larger
than 3 mm. The two cases are the result of the two different hydrodynamic situations at the
gas/liquid interface, which have been observed for free rising single bubbles and are deter
mined by surface phenomena occurring at the gas/liquid interface. Small bubbles behave like
rigid spheres, while bubbles larger then 3-5 mm do not. This effect is caused by small
differences in the surface tension of liquid elements just arriving at the interface and those
that have already been present at the interface for some time. These differences in surface
tension cause a liquid flow along the surface from low to high surface tension, hence in the
direction of the newly arriving liquid elements, which is the direction opposite that of free
rise. This surface flow causes a flow resistance of the free rising bubble that is equal to that
observed for a solid sphere.
For free rising single bubbles that behave like rigid spheres the well-known Frossling [6]
equation is applied to calculate the Sherwood numbers:
Sh = 2 + 0.6 Re«-^ Sc^-' (3)
It can be theoretically derived that the Sherwood number is 2 if there is no flow around the
bubble. This condition exists for very small (d^= <0.1 mm) bubbles, because then the effect
of the free rise velocity can be neglected. If the effect of the free rise velocity is taken into
account, Sh will be larger by the factor given by the second term, which is a function of the
Reynolds (Re = Poii^h^h^Voii) and Schmidt (Sc = Von^Poiiûi) numbers.
The free rise velocity, v^, can be calculated from the equation
Buoyancy force = drag force
(l/6)7Ti/bÂpg = C„(l/2)poiiVb^(l/4)7rJb (4)
or
1/2
= l(4 /3 )g d M (5)
in which is the drag coefficient, which is a function of the Reynolds number Re [7] and
Ap Poll Ph2 Poll •
For bubbles larger than 3 mm, the above-mentioned flow along the surface cannot, as
explained above, exhibit the effect that is observed for small bubbles. In this case the flow
pattern around the bubbles becomes quite different. No friction exists at the bubble sur
face, and the liquid flows almost frictionless along the interface. As a result, liquid ele
ments flow from the top of the bubble to the bottom and disappear again in the bulk of
the oil. The contact time of these liquid elements is the ratio of the bubble diameter to the
free rise velocity. This free rise velocity as calculated from Eq. (4) can be higher by a factor
of 2, because the drag coefficient is much smaller. The liquid-phase mass transfer coeffi
cient kj^ should now be calculated from penetration theory [4]. In this theory it has been
shown that
ki^=2iDiij7TT) 1/2 (6 )
where is the liquid-phase diffusion coefficient for hydrogen and r is the mean contact
time of the liquid elements flowing along the hydrogen/oil interface. This mean contact time
is approximately equal to
T=4/Vb (7)
Rearranging the two equations into a Sherwood correlation gives

Sh = M b /^H 2 = l-1 3 Re^’^ Sc^’^ ( 8)
From a comparison of the Frossling equation with the Sherwood equation, which stems from
penetration theory, it can be concluded that Eq. (8) results in a Sherwood number that is a
factor of 1.9 Sc^-^^ larger than calculated from Eq. (3). Schmidt numbers are on the order of
1500. Thus this calculation results in Sherwood numbers that are 1.9 x 1500®^^ = 6.6 times
larger than calculated from the Frossling equation. It reflects the change in the flow pattern
around a free rising single bubble. Usually this change occurs for bubbles that are larger than
3 mm.
Table 1 summarizes these two calculations of for six bubble sizes in an oil with a
temperature of 150 °C , a dynamic viscosity of 1.8 x 10 “ ^ N -s/m ^, a liquid density of 830 kg/
m^ (then the kinematic viscosity v = 2.2 x 10~^ m^/s), and a hydrogen diffusivity of 15 x 10~^
m^/s. Hence the Schmidt number is 1445, Sc^-^ = 38, and 11.3
The result of these calculations shows that for a wide range of bubble sizes the liquid-
phase mass transfer coefficient at 150 °C and for free rising single bubbles ranges around
2 X lO""^ m/s. It is more a function of liquid temperature than of the size of the bubbles. Only
if bubbles become smaller than 0.1 mm or larger than 3 mm does the value of start to
increase significantly. As the majority of the bubble sizes present in the hydrogenation reac-
270 Koetsier
Table 1 Liquid-Phase Mass Transfer Coefficient for Various Bubble Sizes
¿b I'b T k i , x l 0 '‘
(mm) U\v (mm/s) Sh (m/s) (S) (m/s)
0.1 20,870 0.25 2.3 4.1 0 .2 na

0.2 2,680 1 2.7 2.4 0.1 na
0.5 167 6 3.8 1.4 0.042
1.0 22 24 10 1.8 0.021 na
2 .0 3 90 25 2.3 0.011 na
4.0 1 160 41 1.8 0.025 12
na = not applicable.
tors are within the range for which ki^ has the above-mentioned constant value, it will be
understood that an improvement in cannot be easily achieved. The above estimate holds
for a free rising single bubble. For a swarm of bubbles it can be expected to give a slightly
lower ki^ value, because the free rise velocity will be smaller.
The above calculation does not take into account the effect of turbulence, presence of
catalyst particles, or the frequent collisions occurring in such dispersions.
The effect of turbulence can be estimated from the Brian-Hales [8] or Sänger-Decker [9]
theory. Brian and Hales have shown that the Sherwood number kid^lDî^ is a function of the
parameter which is a Reynolds number using the characteristic parameters describ
ing homogeneous turbulence and in which i/3 2 is the Sauter [10] mean bubble diameter, is
the mean energy dissipation per unit mass, and v is the kinematic viscosity. For example,
Sänger and Decker derived the following Sherwood correlation:
Sh - 2 + 2.309(6T^3.2^/ï^f (9)
From this equation the effect of the turbulence on the liquid phase mass transfer coefficient
can be estimated. In a well-stirred tank reactor the local energy dissipation is a strong function
of the position within the vessel. In the impeller region it can be higher than the mean (over
the whole vessel) value by a factor of 50, whereas in other parts it can be lower by a factor
of 10. Usually the mean energy dissipation is 2 kW/moy or 2.4 kW/kg. From Eq. (9) it can
be derived that a proportionality that is far from linear. In the impeller region the
effect on k^^ for 1.1 mm bubbles will be higher by a factor of 2.4 versus the value for a free
rising single bubble, and in the other regions it will be higher by a factor of only 1.7. All in
all, if all the above-mentioned effects are taken into account, the liquid-phase mass transfer
coefficient in a well-stirred reactor in which the mean energy dissipation is 2.4 kW/kg is
expected to vary between 4 x 10 “ "^ and 6 x 10 ""^ m/s.
The zone within which the hydrogen concentration drops from its maximum value to its
value in the bulk is very thin. It can be estimated from the penetration depth as given by
the equation
5 = 4(D h, t) 1/2 ( 10)
This depth of penetration is about 10“ ^ m, thus extremely small compared to the size of the
bubbles, but of a size equal to the catalyst particle size. So if a catalyst particle is within this
small zone, it will recognize the hydrogen concentration as higher than in the bulk of the oil
phase. The mixing outside the bubbles and within the bulk of the liquid is so good that it will
have no effect on the overall rate of the hydrogenation process.
2. The Specific Interfacial A rea a

The specific inteifacial area of the hydrogen bubbles in the oil, a, is given by the equation
a = 6egjd3,2 (11)
where is the gas hold-up and J 3 2 is the Sauter mean diameter (volume/surface ratio) of
the gas bubbles.
Van Dierendonck et al. [15] empirically determined scale-up correlations for gas hold-up
in well-stirred tank reactors having diameters from 0.3 to 2.6 m and a HIT = 1. The equations
have been experimentally determined for variations within the following properties:
Viscosity:
5 X 10’^'^<77oi,<5 X 10-3 N •
Density:
800<Poii<1300 kg/môii
Surface tension:
20 X lO -^ < 0 -< 7 5 X 10 ”-^ N/m
The correlation for the gas hold-up at which he arrived, if Rushton-type turbine impellers are
used, is given by the equation
, / V, y-67 o .45Vl
( 12)
where is the linear gas velocity, T the tank diameter, D the impeller diameter, H the liquid
height, the liquid height above the impeller, and a the surface tension at the interface
H 2/oil [12] (ir=^25 X lO^^ N/m), and
V^ = (N -N o)D ^/T (13)

in which N q is the minimum stirrer speed that is required to draw in bubbles from the liquid
level in the reactor,
No = 2{h^lT f\(T glp,^,^^\T ID ^) (14)
At 150°C,
o-g/poi, = ( 2 5 x 10-3)(10)/830 = 3 . 6 x lO-'^ (15a)
so
= 0 .1 3 2 (15b)
For example, for a commercial reactor that uses one Rushton-type six-blade impeller and has
a tank diameter T=2.5 m, an impeller diameter D = 0,9 m, a liquid height H = T, h ^ = l . l
m, and no recirculation of hydrogen, then N q = 0.1 s~ ^ and if the power P{ = 6 is
2 kW/moii, then V = 2.8 s~ k Hence, Vl = 0.68 m/s and 6gas ranges between 0.09 and 0.06 if
the linear gas velocity drops from 0.004 to 0 m/s, which occurs as a result of the decrease
in the hydrogenation rate. The value of 0.004 m/s is calculated from an initial hydrogenation
rate of 1 AlV/min.
Van Dierendonck gives the following correlation for the mean bubble diameter:
d s .i = (0.41a-/poi,g)°-3 ± 25% (16)
272 Koetsier
A gas hold-up of 9% (€gas = 0.09) and a Sauter diameter of 1.1 mm are the values calculated
from the above equations. The specific interfacial area a will then be 490 m^/môji.
3. The M axim um Achievable Rate C onstant kjâ

The maximum obtainable value of the characteristic rate constant kiâ depends very much on
the type and size of the reactor. In small-scale reactors it is possible to use very high stirring
speeds such that the mean energy dissipation is on the order of 25 kW/kg. Under these condi
tions one can obtain a gas dispersion of up to 30 vol %. The specific interfacial area can then
be as high as 1500 8 X 10“ "^ m/s. On the basis of these predictions a maxi
mum value for kjâ of 1.2 s “ ^ would seem to be achievable. It has been demonstrated that in
our 400 mL open lab reactors with stirring speeds of 3000 rpm and continuous introduction
of hydrogen gas below the impeller, a kua of 1 s “ ^ can indeed be achieved.
(a)
(b)
Fig. 1 Top: (a) The CD-6 and (b) Rushton impeller designs. Bottom: The kiCi enhancement factor for
the CD6 versus the Rushton impeller type as a function of the power input. (Reprinted with permission
of Chemineer Inc.)
In the example of a commercial reactor discussed above, the volumetric liquid-phase mass
transfer coefficient kiß would be between 0.16 and 0.25 s “ ^ This estimate results in a value
that is slightly higher than what is commonly seen in commercially used reactors, namely kiß
ranging from 0.10 to 0.15 s~ ^ if a stirring power of 2 kW/môy is applied. The maximum
achievable value for kiß depends on the maximum power input that can be used in such
reactors and on a large number of factors that are discussed below.
4. E ffect o f R eactor D esign on kiß

Reactor design has always been focused on getting an excellent hydrogen transfer rate, for
which it is important for the characteristic rate constant kiß to be as high as possible. This
can be achieved by ensuring
1. The presence of adequate baffles, proper position of the impellers, number, type and
size of impellers, liquid height, presence and position of heating coils, hydrogen intro
duction through spargers, etc....
2. The quality of the oil; for example, one usually obtains higher values of kiß in fish
oils than in vegetable oils.
In commercial practice there are two reactor designs in use. These operate on different
principles. One is the well-stirred tank reactor with one or two Rushton-type or Lightnin [13]
impellers and optionally a second smaller impeller at the height of the gas/liquid interface to
suck in gas bubbles, four baffles, hydrogen sparger below the bottom impeller, and sometimes
with continuous gas recirculation. Bakker et al. [14] of Chemineer, Inc., discussed a new
impeller design that should show better mass transfer performance than the conventional
Rushton-type design. The two designs are illustrated in Fig. 1. It is claimed that the kiß factor
could be higher by a factor 1.3-2 depending on the superficial gas velocity. The other is the
so-called Buss loop [15] reactor in which the liquid is recirculated and reintroduced into the
reactor through a Venturi tube into which the hydrogen is also fed into the system. In this
Venturi tube the hydrogen gets extremely well dispersed into the liquid. In this type of reactor
it seems to be possible to achieve higher values for kiß.
Even today examples have come to our attention where the hydrogen absorption rate is the
rate-determining step because the kiß value is far too low.
C. Rate of H2 Transfer from the Liquid Phase to the

Catalyst Particles
The rate J at which the hydrogen is transferred from the bulk oil phase to the catalyst particles
is directly proportional to the driving force
J ^s^catC^bulk ^ ca t) ( I "7)
In this equation the product kß^^^ (s“ ^) is again the characteristic rate constant. It is the
product of the solid-phase mass transfer coefficient k^ (m/s) and the specific contact area
between the catalyst particles and the oil aât (m\at/i^ôii)* The H 2 concentration Qat is that
existing at the outer surface of the catalyst particles.
1. The Characteristic Rate C onstant

The solid-phase mass transfer coefficient k^ is a property that, in well-stirred solid-liquid
systems, can be calculated after the Brian-Hales or Sänger-Decker theory [8,9]. For catalyst
particles that have a mean Sauter diameter of 5 ßm , the corresponding value of the solid-
phase mass transfer coefficient is 0.9 x 10“ ^ m/s. It is, in fact, mainly determined by the 2
274 Koetsier
in Eq. (9) and not by the turbulence of the liquid elements flowing around these particles.
The contribution of the second term to the Sherwood number is only 0.23 if e^ = 2.4 kW/kg
and 0.45 if €x = 50 kW/kg. The explanation is that the size of the smallest eddies, that is, the
eddies in which the energy is lost due to friction, is several times larger than the size of the
particles. Kolmogorov [16] gives the following expressions for the size and velocity of
these small eddies:
(18)
,0.25
Ve=(lêT)' (19)
For an energy dissipation of 2.4 W/kg, the Kolmogorov scale is 0.26 mm and the velocity is
85 mm/s. This scale is 50 times larger than the mean size of the particles. Therefore, the
turbulence will not have a great effect on the mass transfer rate to these particles.
The specific interfacial area between the catalyst particles and the oil is given by the
equation
^^cat^^3.2,cat (20)
in which 6cat is given by the equation
Poil mg Ni in oil
( 21)
Ppait ((% Ni in particles)(10'^)(kg oil)/
(wppm N i)(p jp p ^ )
% Ni in particles x 10"^
If 100 wppm Ni is used, then eât is 1.4 x 10“"^ and is 170 m\at/môii-
Hence, the characteristic rate constant is 1.5 s “ ^ It is one order of magnitude larger
than the characteristic rate constant ki<a for the absorption rate of hydrogen into the oil phase.
The effect of mass transfer rate from the bulk to the catalyst particles can therefore be ne
glected!
D. Reaction Rate Within the Catalyst Particles

Several reactions occur on the surface of the Ni crystallites that are randomly distributed
throughout the catalyst particles. Hydrogenation and positional and configurational isomeriza
tion of double bonds are the main reactions. As a result of the isomerization reactions, a
variety of isomers are formed as hydrogenation proceeds.
Rate equations are in general of a very complex nature. For practical purposes it is usually
sufficient to have a simple equation that describes the rate as a function of the concentration
of the reacting components and the temperature. Hydrogenation of edible oils and fatty acids
is characterized by the following facts.
1. The concentration of hydrogen, as already stated, is considerably lower than the con
centration of the double bonds. It is our experience that the hydrogenation rate can be
approximated by a process first order in hydrogen if the hydrogen concentration re
mains below 10 mol/moij.
2. The oil contains the double bonds in different types of molecules, and during hydroge
nation the number of different types of molecules increases even further due to posi
tional and configurational isomerization.
3. The hydrogenation rate is of zero order with respect to each type of component con
taining a double bond and has a zero-order hydrogenation rate constant k^ that is differ-
ent for each type of unsaturated component. Some are hydrogenated fast, while others
are hydrogenated much more slowly. Such similar systems often occur in practice.
Examples are the hydrogenation and desulfurization of a number of petrochemical
feedstocks [17]. The desulfurization rate is first-order in a pure component, but for a
mixture of many different components the '‘apparent” order in “sulfur” is second-order
if the reaction is carried out in a batch process and of 1.5 order if the reaction is
carried out in a continuous, well-stirred, tank reactor.
4. A similar situation exists for the hardening rate in our set of reactions. All these
different unsaturated oil components are each hydrogenated at their specific rate and
initially as a reaction of zero order in oil. One could describe such a system by using
a distribution of rate constants if all these rate constants were known.
A rigorous analysis of the reaction rate constants for each of these components requires an
understanding of the hydrogenation rate of each individual component. Such an analysis is
too time-consuming because it demands a detailed component analysis, and it is questionable
whether the resulting information is of much use. However, many results have indicated that
the hydrogenation of edible oils is of first order in the iodine value over a wide range of IV.
In this case the IV reflects the concentration of all components containing an unsaturated
bond. It is also an “apparent” order. It is an order of reaction that is determined by both the
type of process and the difference in reaction rates of the various components. Some of the
components react very fast because their reaction rate constants are high. The reaction rate
constants of other components may be much smaller, and as a consequence the hydrogenation
rate is much lower. And in this case also the apparent order, if the rate is expressed in terms
of IV, can be considered a first-order reaction. However, it should be understood that this
conclusion is based on the assumption that during the hydrogenation reaction the hydrogen
concentration remains constant. As already mentioned, this is usually not the case. Even
under isothermal conditions the hydrogen concentration in the bulk liquid phase will continu
ously change. Examples are given in Section II.H. From our results we have come to the
conclusion that the correlation between the hydrogenation rate at constant hydrogen concen
tration and the iodine value have to be expressed by two terms, one term reflecting the rate
for the iodine values above IV = 80 and one for the regime below 80.
All in all, we have concluded that the hydrogenation rate per unit volume of catalyst can
be approximated by the following equations.
R { t) - ^r,p[IV(0 - 75]Cbuik(0i?p(0. for IV > 8 0 (22)

^(0=/?mono = ^r,mI’V(i)Cbulk(0l?m(0, for IV<80 (23)
and
R {t)= J{t)l€ ^, (mol H 2 /(m \at' s)) (24)
in which Cbuik(0 is also the hydrogen concentration at the liquid/catalyst interface, because
we demonstrated that kâ^ » kiâ, rj(t) is the effectiveness factor, which is a function of the
Thiele modules [18], and eât is the volume fraction of catalyst particles.
Equations (22) and (23) reflect the correlation between the reaction rate R and the iodine
value. An example of these two correlations is shown in Fig. 2, in which the hydrogenation
rate is expressed as J ( = e^^^R/0.6) AlV/min, in which the factor 0.6 is the proportionality
constant between IV/min and mol H 2/(m \ifs), has been plotted as a function of the iodine
value of a fish oil hydrogenation. However, Fig. 2 does not give any information about the
mechanism. And one should remember that the reaction rate for two different oils of equal
iodine values will be different because IV can be the result of different fatty acid composi-
276 Koetsier
Hydrogenation rate in
tV/min as a function of IV
Rate in IV/min at constant bulk hydrogen concentration
Reaction rate Rpoly
Reaction rate Rmono
iO 80 100
Iodine value
Fig. 2 Hydrogenation rate in IV/min as a function of IV.
tions. Finally, it should be borne in mind that once the concentration of these unsaturated
molecules has become significantly lower, the hydrogenation rate will become first-order in
stead of zero-order in each individual component. This change will cause the “apparent” order
of the reaction to become second-order.
E. Factors Affecting the Overall Rate of the

Hydrogenation Process
In Sections I I.C and II.D, it has been explained that the overall rate of the hydrogenation
process can be described by the rate equations (1) and (22,23). Moreover, it has been esti
mated that the rate of hydrogen transfer from the bulk of the oil to the surface of the catalyst
particles is not expected to affect the overall rate of this hydrogenation process. There are,
however, three aspects that have not yet been discussed. First a general equation is derived to
describe the overall rate of the hydrogenation process.
1. Initial Saturation R ate o f Oil

In edible oil hydrogenation processes the feedstock is heated under vacuum to a specific
temperature at which the hydrogen is introduced into the reactor. From time zero onward, the
bulk hydrogen concentration is determined by the mass balance
iiCfoulkCO
ôil X “ ^L^t^max ~ ^bulkCOlôii “ ^(O^cat
dt
khîCmax - QulkCOlKil - ^r,p[IV(0 ~ 15] t] (t)C^îy,(t)V^, (25)
At t = 0 the bulk hydrogen concentration is 0.

Upon integration the following equation is obtained:
^bulk(0 K
{ l-e x p [ -( ^ -h L ) i] } (26)
^max ^ +^
in which
K = kija (27)
and
L = /:,,p[IV(0-75]i7p«6cat (28)
In Fig. 3, examples of three different solutions to Eq. (26) are given:

^ = 0.1 a n d L = 0 (29)
^ = 0.1 a n d L = 0.1 (30)
K = 0A and L = 0.1 [1 - exp ( - kt)] (31)
In which the rate constant k reflects the rate at which the catalyst is reactivated to its maximum
activity. The third solution [Eq. (31)] has to be obtained numerically because an analytical
solution cannot be derived.
Once the hydrogen concentration has reached its equilibrium value, which will occur after
a time t —Mkâ^ the derivative becomes almost zero. This derivation neglects the fact that in
the initial phase the specific gas/liquid interface a is also a function of time. It may be ex
pected that this time will be much shorter than the time constant Mkâ for the absorption
process.
2. R eactivation o f C atalyst Activity

The second process is the reactivation of the nickel that has been passivated as a result of un
avoidable contact with oxygen. Nickel is dispersed into hardened fat at the end of the manufac
turing process to prevent reoxidation of the reduced nickel surface. However, the pastilles occa
sionally come into contact with air, and then oxygen diffuses into the hardened fat and reoxidizes
a small fraction of the reduced nickel. This reoxidation is extremely slow, because the perme
ability of oxygen through hardened fat is low. The oxygen that permeates will always start to
passivate the most active part of the nickel first, which consists of the nickel sites at the outer
surface of the micrometer size particles. These nickel sites are the most active ones, because
there is no diffusional limitation for the oil molecules to reach them.
3. Poisoning o f Edible Oil H ydrogenation Catalysts

Nickel catalysts can be poisoned as a result of various reactions.
1 . Absorption of components within the porous structure of the catalyst particle. For
example, components such as phospholipids, in particular those containing the basic
C h a n g e in H 2 concentration
Cbuik / Cmax in the first two minutes
Cbulk / Cmax
1 ,2
1. no reaction
1
0,8 3. reaction;
•'— ■
— activity increases in time
0,6
0,4
0,2
0
/ 50 100
2. reaction;
no change in cat activity
150 200 250

Hydrogenation time in seconds
Fig. 3 Change in H2 concentration in the first 2 min.

278 Koetsier
NH groups, will be adsorbed onto the acid sites of the carrier. This will affect the
accessibility of the nickel sites.
2. Dissolution of the active nickel surface due to the presence of fatty acids. Free fatty
acids react with nickel sites according to the reaction
Ni + 2 FFA Ni(FFA)2 + H 2 (32)
This is an equilibrium reaction, and therefore the higher the hydrogen pressure the less
nickel will dissolve. In edible oil hydrogenation it will result in only 1-2 wppm Ni if
0.2 wt % FFA is present. In fatty acid hydrogenation it is an important contribution
to the deactivation of the catalysts during hydrogenation.
3. Poisoning of the nickel surface by reaction with sulfur components. In many oils minor
amounts of sulfur are still present in the oil after refining. These components are
desulfurized, and the nickel surface is sulfided to the extent that one S atom is present
per two atoms of surface Ni atoms [19]. The reaction with sulfur components is a
chemical reaction the rate of which depends on the type of component and the temper
ature of the reaction. If the conditions chosen are such that the desulfurization rate is
high, the nickel present within the region at the outer side of the particles will be
sulfided first, then the sulfur will penetrate further to the center of the particle. This
mechanism is responsible for the fact that small particles are poisoned much faster
than larger particles. Since small particles are the most active particles this leads to a
serious loss of activity. One would prefer all the nickel sites, irrespective of where
they are located within the particles, to be poisoned at the same rate.
The reaction rate of the sulfided nickel is much lower than the rate observed for pure
nickel. For a sulfided nickel surface, Eqs. (22) and (23) become
/?(i)=/?s,poiy = ^rs,p[IV(0-75]Cbuik(i)i?,(i), for IV >80 (33)
and
« ( 0 = /? s,m o n o = ^rs.m IV(i)Cb„,k(i)r,,(i), for IV<80 (34)

Equations (22) and (33) can be combined as
(K s.p\(vs(t)\
7?ov(0 = X+ (1 —jc) { ^ , , p [ I V ( i ) - 7 5 ] T ,p ( i ) } C b ,i k ( i ) (35)
W,p)/\T?p(i)/
in which the x is the fraction of the nickel surface that has not yet been sulfided, or
= X { k , J l Y ( t ) - 7 5 ] i7 p ( 0 n u ik (0 (36)
in which X has been substituted for [x-h(l — .
As already said, the sulfiding (and any other poisoning) of the nickel surface area is not
expected to occur homogeneously over all the catalyst particles or within a catalyst particle.
Sulfiding of the whole nickel surface area would be homogeneous only if the reaction were
slow and if all the particles were of the same size. However, the sulfiding is a fast reaction.
First all the nickel at the outer surface will be sulfided and subsequently the nickel that is
located deeper in the particle. In the case of a particle size distribution the rate of sulfiding of
the nickel will also be a function of the particle size. The sulfiding of the smallest particles is
completed much faster than that of the larger particles. To reflect these effects in the rate
equation, the factor X has to have the exponent n. Equation (36) becomes
/?„,(?) =X"{fc,,p[IV(0 - 7 5 ] t ,p ( 0 } C b , i k ( 0 (37)

In one of our fish oil experiments (see Section II.H .3) it is demonstrated that the data could
be adequately fitted by using a value of 2 for n.
F. The Overall Rate Equation of the Hydrogenation Process

Once the catalyst has reached its maximum activity after the initial induction period, the
hydrogenation rate has also reached its maximum value. Consequently, the hydrogenation rate
will slowly drop as time passes because the rate is first order in IV. As the rate drops over
time, the hydrogen concentration in the oil phase will slowly increase over time. The rate of
this increase is rather slow, and for a short time the overall hydrogenation rate can be consid
ered almost constant.
If we now eliminate the unknown concentration CûikCO from the equations (1), (24), and
(37), we arrive at the following equation for the hydrogenation rate J{t) for IV>80:
-1
1 1
m = c^ (38)
lM X”{k,„[IV(0-75]i7p(i)}e.
This equation expresses the absorption rate as a function of the driving force, which is the
maximum hydrogen concentration and the term in brackets, which expresses the total
resistance.
The first term in the large brackets in Eq. (38) represents the resistance of the physical
transfer rate of hydrogen from the gas phase to the oil phase. This term can be experimentally
determined if the hydrogenation rate is measured by using an excess of catalyst so that the
second term is no longer relevant.
The second term in the large brackets reflects the effect of the amount of catalyst and the
parameters describing the quality in terms of the chemical reaction rate. In other words this
term depends on the intrinsic rate, the extent of poisoning, the iodine value, and the extent of
diffusional limitation within the catalyst particles, but not on the hydrogen concentration, if
the above assumption of first-order kinetics in hydrogen is correct. The effect of this second
term is therefore complex. All three important parameters, namely IV, and 17, change in
time as the hydrogenation proceeds. The reaction rate constant decreases as a result of
poisoning of the nickel surface area, IV drops continuously, and increases to a maximum
value of one as the rate decreases.
Equation (38) can be rearranged as follows:
a 1 1
- = ;— + ; (39)
J(t) kâ ■ - 75]r|p(0}6eat
The left-hand term is the quotient of the maximum hydrogen concentration, which is deter
mined by the temperature and the pressure applied in the reactor and the absorption rate J{t),
a rate that can be continuously monitored.
Equation (39) is a very useful equation to describe the initial absorption rate, as a
function of the most relevant parameters:
1
(40)
X"R.p[IV(i = 0)-75]T7p(0},= (-)
or the mean absorption rate, tj, where the initial iodine value, IV q , drops to the final
value, IVg„d.
1
(41)
kâ n^,,p[IV(i)-75]7,p(f)}„ (-)
\e c J
280 Koetsier
Regimes for different type and scale

of hydrogenation reactors
Cmax Commercial stirred - and

closed lab Reactors
Jmean
poisoning
1
kL,a Loop - and Open Lab
Reactors
Fig. 4 Regimes for different types and scales of hydrogenation reactors.
If the results of of any hydrogenation are plotted as a function of 1000/wppm

Ni then a line will be obtained as illustrated in Figs. 4-6.
In Fig. 4 the difference between reactors operating under conditions of excellent hydrogen
transfer rate (low value of Hkia) and of moderate mass transfer rate (higher values for Hkija)
are reflected. The intersection with the axis indicates the value of the parameter
Hkia. On the basis of Vkija it is possible to distinguish two regimes, one in which Mkija varies
between 1 and 3 and one in which it varies between 5 and 50. This latter value is obtained in
closed reactors in which baffles are not used.
It is our experience that two different types of lines are obtained:
1. Straight lines if oils are hydrogenated that do not contain contaminants that poison the
nickel catalyst particles
2. Lines that bend upward as the parameter 1000/wppm Ni increases (which occurs if
poisons are present)
These two different lines are illustrated in Fig. 5. The curved lines are typically obtained for
the results of fish oil and rapeseed oil hydrogenations. The less catalyst is used, the higher
the value 1000/wppm Ni and the greater the effect of poisoning will be.
Three lines reflecting rate at two ka values

and effect of poisoning:
Cmax / Jmean
1 poisoning;
.................. i curved lines .........<■
—-------— T no poisoning;
1 straiaht lines
1000/w ppm Ni
Fig. 5 Typical lines obtained for vs 1000/wppm Ni when hydrogenated oils are (top two
lines) and are not (bottom line) contaminated by poisons.
Three regimes reflecting rate as a result

of different effects:
Cmax / Jmean
H2 Catalyst Effect of
absorp. rate....... activity ........ poisoning
-ST-----
.. — - -
1000 /w ppm Ni
Fig. 6 Three regimes reflecting changes in hydrogenation rate as a result of effect of hydrogenation
rate, catalyst activity, or poisoning.
Another important observation is that if two experiments are performed at the same pres
sure but under lower or higher values of kija, the lines move upward or downward parallel to
each other. The difference between the two experiments is that in the case of higher values
of kija, the actual hydrogen concentration in the liquid is higher and the reaction proceeds
faster. The fact that the lines move parallel to each other implies that the parameter determin
ing the slope remains constant. This is the parameter X^{k^^[YV(t) - 75]i7p(0}mean^ the parame
ter that describes the reaction rate. Moreover, the reaction can indeed be considered first-order
in hydrogen.
However, if the pressure is changed considerably then these lines tend to bend upward
more sharply if the pressure is higher. Apparently, if the bulk hydrogen concentration between
two experiments is different because higher pressures are used, then the slope is more signifi
cant at the higher pressures. Usually, higher pressures will result in a significant increase in
the hydrogen concentration. The stronger curvature of the slope implies that the poisoning
occurs faster as a result of which the rate becomes slower.
In Fig. 6 three regimes are shown where the effect of hydrogen absorption rate, catalyst
activity, or poisoning dominates the process. At low values of 1000/wppm Ni the predominant
effect is the hydrogen absorption rate, and at high values of this parameter the predominant
effect is poisoning.
G. General Equation for the Hydrogenation Time of the

Hydrogenation Process
In most cases the hydrogenators are interested in the time required to get the iodine value
reduced from the initial value IV^ to a particular IVend- A graph in which this hydrogenation
time is plotted on the vertical axis against the amount of catalyst, preferably expressed as
1000/wppm Ni, on the horizontal axis can be very helpful. For that reason we use the concept
of hydrogenation time. The equation expressing the hydrogenation time, HT, is
HT = 0.6(IVo-IVend)//rr (min) (42)
The proportionality constant of 0.6 is required if HT is in minutes and iniean is in moles per
cubic meter of oil per second.
282 Koetsier
By substitution of Eq. (41) for imean^ the following equation is obtained:

1 1
HT = (43)
a' '
or
1000
HT = HT^,„ + --------— B (44)
wppm N i '
in which
= 0 .6 (IV ,- IVe„d)/(^L^ Câx) (min) (45)
and
1000 B P o il/P ipart
t= 0.6(IV „-IV ,„,) W r,p [IV (0 -7 5 ]T |(0 L e a n Q
wppm N i _% Ni in particle x 10"^
(46)
The first term on the right-hand side of Eq. (44) is the minimum hydrogenation time HTjîn-
It is determined by process parameters only and not by the type or quality of the catalyst,
decreases linearly with increases in C^âx and kjâ.
In a commercial soybean oil hydrogenation process with ^La = 0.15, a pressure of 2 bar
( = 200 kPa), and A IV (= IV^ —YV^nd) = 50, the minimum hydrogenation time is already 32
min, and in fish oil hydrogenation with kjâ =0.15, a pressure of 4 bar ( = 400 kPa), and
A IV = 80, the minimum hardening time will be 25 min. If in the latter case a very high stirring
speed were used, then kâ would be 1 and the minimum hydrogenation time only 4 min.
It was explained in Section II.B.4 that the value for a specific power input can be
increased by a factor of 2 if an optimally designed impeller is used. Then the minimum
hydrogenation time is lower by a factor of 2. Moreover, if a loop-type reactor is used, the
minimum hydrogenation time can probably be reduced further.
The second term on the right in Eq. (44) is determined by IV^- IVe^d; the type, quality,
extent, and rate of poisoning; and the amount of catalyst and hydrogen pressure used in the
system, but not by other process parameters such as the stirring speed.
All in all, the minimum hydrogenation time is a meaningful parameter reflecting all factors
that affect the hydrogenation rate and are not determined by the quality or amount of the
catalyst. The performance of two reactors of different sizes can be considered almost equal if
they are operated under the same temperature and pressure and the minimum hydrogenation
times are equal. The only difference between two reactors of different size and type originates
from the difference in the distribution of the local energy dissipation, which can cause minor
differences in the concentration of hydrogen in the liquid phase. It is not expected that these
differences will have a major impact on the product qualities obtained after hydrogenation.
The following table summarizes the drop in hydrogenation time for three cases.
Reactor type Impeller type kjâ(s 0 HTî„ (min) HT (min)
Tank Rushton 0.15 32 64

Tank CD-6 0.22 22 54
Loop 0.30 16 48
In cases where the mean value of the volumetric liquid-phase mass transfer coefficient for
a loop reactor is twice as high as for a tank reactor using a conventional impeller, then the
overall hydrogenation time will be lower by only 16 min. To get the same product quality it
may be necessary to operate the loop reactor under a slightly lower pressure such that the
hydrogenation times are equal again. An advantage of the loop reactor remains that it can
be operated at a lower pressure to get the same product quality in the same hydrogenation
time.
H. Examples Obtained from Laboratory-Scale Reactors

In this section we review a number of examples of the hydrogenation of two soybean oils,
one canola oil, and three fish oils. Let us first discuss the main features of the equipment used
for our catalyst performance tests. An adequate catalyst performance test consists of two
important elements: reliable equipment and appropriate test procedures.
We selected Medimex autoclaves, which have been operating to our full satisfaction since
1990. The rate of hydrogenation is monitored and controlled by a press flow control unit as
depicted in Fig. 7. These units were originally manufactured by Peteric, but have been manu
factured for the past few years by Biichi. This system is fully automated and computer-con
trolled.
The hydrogen supply unit controls the hydrogen supply rate by keeping the hardening
reactor at a fixed pressure by continuously injecting a known small amount of hydrogen. A
computer monitors the amount of hydrogen. In this way the hydrogen absorption rate can be
monitored. Consequently, the iodine value and the rate of change in it can be calculated as a
function of hydrogenation time.
Our test reactor is operated as a dead end reactor. This has the advantage that in combina
tion with the computer control, the oil can be hydrogenated to a fixed end IV because all the
hydrogen fed into the reactor is used for hydrogenation.
Once the test is completed, the hydrogenated product can be subjected to several analyses.
The proper functioning of the equipment is controlled by measuring the change in the refrac
tive index of the oil. A second advantage of the computer control is that not only the pressure,
but also the temperature and stirring speed, are continuously monitored and adjusted. The
temperature can be kept isothermic or raised adiabatically.
I. H ydrogenation o f a Soybean Oil

Figure 8 depicts the results of hardening “soybean oil 1” at a temperature that was increased
stepwise from 150°C to 180°C at atmospheric pressure and two different stirring speeds. The
volumetric liquid-phase mass transfer coefficient, kjâ was 1 and 0.3 s ~ ^ respectively. This
difference will cause a significant difference in the bulk hydrogen concentration.
The amount of catalyst was varied between 30 and 500 wppm Ni. The two lines are almost
HYDROGEN HARDENING
SUPPLY RESERVOIR EQUIPMENT
UNIT
Fig- 7 Control scheme to monitor the rate of hydrogen supply to the Medimex autoclave.
PI = pressure indicator.
TI = temperature indicator.
284 Koetsier
Fig. 8 Characterization of soybean oil 1 at 100 kPa and 150-180°C (302-365°F).
straight and exactly parallel. There was no effect of poisoning, and it can also be concluded
that the activity of the catalyst is the same at the two different hydrogen concentration levels.
In Fig. 9 the results of an isothermal experiment of hardening of “soybean oil 2” at 180°C
and 150 kPa are depicted. In this case also the data show an almost straight line. In this case
the kiG value is much smaller, 0.14 s “ k
2. H ydrogenation o f a Canola Oil

Figure 10 depicts the results of hardening a canola oil at a constant temperature and three
different pressures, with a volumetric liquid-phase mass transfer coefficient of 0.13 s “ k
The amount of catalyst was varied between 100 and 6000 wppm Ni. In this case straight lines
are not obtained. The effect of poisoning causes the lines to bend upward. The higher the
pressure, the more the curve bends. Hence at higher pressures poisoning seems to have a
stronger effect.
In Fig. 11 the results for for two different catalysts, PRICAT 9920 and PRICAT
9910 (see Section III.B), are depicted. The higher the nickel content (the lower the value 1000/
wppm Ni), the less difference there is between the two catalysts. Only at low wppm Ni can a
Fig. 9 Characterization of soybean oil 2 at 150 kPa and 180°C (365°F).

Fig. 10 Characterization of standard canola oil hardened at 150, 300, and 500 kPa and 130°C (266°F).
difference be observed for the parameter For example, at 100 wppm Ni this differ
ence is 12 s versus 16 s. It can also be concluded from Figs. 10 and 11 that at equal stirring speed
the kija value increases if the temperature rises from 130 to 180°C, namely from 0.13 s “ ^ to
0.19 s-i-
3. H ydrogenation o f Three Fish Oils

Figures 12-17 depict the results for “fish oil 1.” The amounts of catalyst used in these experi
ments were 1760 wppm Ni (line 1), 880 wppm Ni (line 2), 660 wppm Ni (line 3), and 440
wppm Ni (line 4). The operating pressure was 900 kPa, and the temperature was adiabatically
increased from 160°C to 190°C and thereafter kept constant.
In Figs. 12 and 13, IV is plotted as a function of time, in Fig. 12 on a linear scale and in
Fig. 13 on a logarithmic scale. These results are in line with findings that have been reported
by others [20]. The plot on the logarithmic scale shows straight lines. It is wrong, however,
to conclude that this observation proves that the hydrogenation reaction is of first order in the
iodine value. Such a conclusion can be drawn only if it is certain that the bulk hydrogen
concentration had a constant value during the entire reaction time.
Fig. 11 Characterization of standard canola oil hardened at 300 kPa and 180°C (365°F).
286 Koetsier
IV Fish oil 1 as a function of

hydrogenation time
iodine Value
Fig. 12 Iodine value of fish oil 1 as a function of hydrogenation time with PRICAT 9910 and various
amounts of Ni as catalysts.
IV Fish oil 1 as a function of

hydrogenation time
Iodine Value
Fig. 13 Iodine value of fish oil 1 plotted logarithmically as a function of hydrogenation time with
PRICAT 9910 and various amounts of Ni as catalysts.
Hydrogenation rate Fish oil 1 in

IV /min as a function of time
Rate In IV / min
Fig. 14 Hydrogenation rate of fish oil 1 as a function of hydrogenation time.

In Figs. 14 and 15, hydrogenation rate is plotted as function of the hydrogenation time
(Fig. 14) and as function of the iodine value (Fig. 15). Figure 15 shows that the hydrogenation
rate increases if IV increases, but the increase is not in linear proportion to IV.
It still cannot be concluded that this simple linear correlation exists over the whole IV
range, because during this hydrogenation the bulk hydrogen concentration changes also. We
therefore now first calculate the bulk hydrogen concentration for the four cases. These results
are plotted in Fig. 16.
The hydrogen concentration in the bulk of the liquid increases as hydrogenation proceeds. In
the first 30 min the increase is rapid, but it then slows down. The change in the bulk hydrogen
concentration is, as expected, different for all three cases. The higher the amount of nickel used,
the lower the hydrogen concentration at the start of hydrogenation. In the case of the lowest
nickel content the maximum hydrogen concentration is obtained almost directly after the start of
the hydrogenation process. In the other two cases the bulk hydrogen concentration changes from
a low value to the maximum value. In the case of the highest nickel content this maximum has
not yet been achieved when hydrogenation is complete at an IV of 30 (see Fig. 16).
We can now calculate the hydrogenation rate assuming that the hydrogen concentration in
Bulk H2 cone Fish oil 1 as a function of

hydrogenation time
Fig. 16 Bulk hydrogen concentration in fish oil 1 as a function of hydrogenation time.

288 Koetsier
Hydrogenation rate in fV/min versus IV

(assuming H 2 cone would have had max value)
Fig. 17 Bulk hydrogen concentration in fish oil 1 in IV/min vs IV (assuming that the H2 concentration
would have had a maximum value).
Hardening of fish oil 2

400 kPa; 1 5 0 X - 1 8 0 X ; PRICAT 9910
Fig. 18 Hardening time for fish oil 2 vs concentration of Ni catalyst for PRICAT 9910 at 400 kPa
Fig. 19 Hardening time for fish oil 2 vs Ni concentration with PRICAT 9910 and 9920 at 400 kPa,
150-180°C, i^L^ = 0.14s-^
Fig. 20 Fitting eq. 44 to hardening time results. Model with proportionality to 1/X^ en 1/X.
the bulk of the liquid reached its maximum value right at the start of the hydrogenation
process. This calculation is, however, based on the assumption that the hydrogenation rate is
indeed first order in the hydrogen concentration. These results are plotted in Fig. 17.
From the results depicted in Fig. 17 it can be concluded that for this fish oil the hydrogenation
rate is certainly not first order in the iodine value. In all three cases there are two distinct re
gimes. For an IV of 0 to 80, the increase in the hydrogenation rate is indeed linear, but above
IV-80 the rate increases linearly with the quantity IV-75. This is the regime where most commer
cial edible oil hydrogenations are carried out. In order to reflect all these data in one picture, the
data for the lower wppm Ni had to multiplied by the factor given in parenthesis.
The next two figures depicted the results of the hardening time required to drop the IV of
fish oil 2 from 165 to 82 as a function of 1000/wppm Ni. The conditions applied were
400 kPa, 150°-180°C, with the switch point at the iodine value calculated from AIV
= 0.002(1Vq)^, and three different stirring speeds resulting in kiâ values of 0.07, 0.14, and
0.33 s “ ^ (Fig. 18). All three lines are parallel, indicating that the assumption of a reaction
being first order in hydrogen is correct for this fish oil hardening reaction as well.
In Fig. 19 the results of a similar test at kja = 0A 4 are plotted for the two catalysts PRI-
290 Koetsier
CAT 9920 and PRICAT 9910. It is clearly demonstrated that PRICAT 9910 is much more
poison-resistant than PRICAT 9920. This feature is ascribable to the lower specific nickel
surface area of this catalyst.
The next question to be answered is whether the lines in Fig. 19 can be described by Eq.
(37). This analysis requires an assumption for as given in Eq. (37). A value of
0.1 has been used, because the reaction rate constant for sulfided nickel is about one-tenth
that for pure nickel, and will be about 1. The value of x has been calculated assuming
that the surface of 200 wppm Ni of catalyst A and 275 wppm Ni of catalyst B has been
poisoned by the sulfur content of the oil. The result is depicted in Fig. 20.
In Fig, 20 the lower line reflects the correlation for « = 1 in X", and the two upper lines
(black and gray) are for n = 2. There is surprisingly good agreement with the experimental
results if « = 2, for both PRICAT 9920 (black line) and PRICAT 9910 (gray line).
Figure 21 depicts the correlation between the hardening time and the reciprocal amount
of nickel for four different cases: one for fish oil 2 and three for nb fish oil 2 , fish oil 2 that
was bleached using 0.75, 1.25, and 2.5 wt % bleaching earth. This figure clearly demon
strates the reduction of catalyst poisoning if the fish oils are bleached before they are
hardened.
III. PRODUCT PROPERTIES AFTER HYDROGENATION

A. Modification of Oils
The use of liquid oils for a variety of fat-based products requires modifications of two proper
ties: the oxidative stability of the oil and its melting characteristics. Three different modifica
tion processes can be used for this purpose. Among these three basic fat modification pro
cesses—hydrogenation, fractionation, and interesterification—hydrogenation is, even today,
more than 90 years after its commercial introduction, the most widely used. It effectively
improves the oxidative stability of the oil, promotes the interchangeability of raw materials,
and creates great diversity in the melting characteristics of the resulting fats. However, due
to several factors, such as the increased awareness of possible negative effects of trans isomers
on human health, the continuous change in the availability of the raw material mix, an in
creased preference for soft margarines, the recognition of environmental issues, and the use
of technically improved esterification processes, the picture of hydrogenation is changing.
The question of how oxidative stability and melting characteristics can be changed such
that optimal product properties will be obtained has been studied over the past 70 years by
many scientists, for example, in the early 1950s by Waterman and coworkers [21] at Delft
Institute of Technology, in the 1960s and 1970s by Coenen and colleagues [22] at the Unilever
Research Laboratory at Vlaardingen, by Albright at Purdue University, and by others. These
studies led to the conclusion that when porous Ni catalysts are used, the selectivity of the
edible oil hydrogenation reactions is the key factor determining melting characteristics.
Different selectivities have been defined to characterize the complexity of this hydrogena
tion reaction. To explain the effects of selectivity, it must be understood which types of
reactions occur during hydrogenation. It is well known that around 80% of all fatty acids in
edible oils have a carbon chain length of 18 atoms. Therefore, the consecutive hydrogenation
reactions are usually depicted by the simple reaction equation
+H2 +H.
18:3 18:2 18:1 -> 18:0
For such a set of consecutive reactions it has been observed [23] for many years that the
selectivity toward the intermediate products is strongly determined by the diffusional limita
tion of the large triglyceride molecules and by the hydrogen concentration within the pores.
Usually it has been assumed that there is no significant diffusional limitation of hydrogen
within the pores. This assumption is correct for low temperature hydrogenation, but not for
temperatures as high as 180°C, which will be discussed further in this section.
Four different selectivities have been defined:
• First, the linoleic and oleic acid selectivities, which express the preference for linoleic
acid over oleic acid and for oleic over stearic acid hydrogenation, respectively. If these
preferences are high, stearic acid is not formed until almost all polyene has been hy
drogenated. The result is a product high in linoleic and oleic acid and low in stearic
acid.
• Second, triglyceride selectivity is a measure of the extent to which the three fatty acid
chains in one fat molecule behave independently. This can be illustrated by the follow
ing three hypothetical cases.
1. The case of extreme triglyceride selectivity is given by the presence of only one
saturated fatty acid in each triglyceride molecule. This case will result in a very
low melting point.
2. In a fully random situation the same amount of saturated fatty acids will not have
been distributed equally over all the triglyceride molecules, but it will have been
distributed in accordance with a Poisson distribution. There will be a small fraction
of fat molecules containing two and three saturated fatty acids, and there will also
be a small fraction containing no saturated fatty acid molecules. This situation
yields a melting point about 10°C higher.
3. If all the saturated fatty acid molecules are concentrated in only trisaturated mole
cules mixed with unhardened oil, the result will be a mixture of high melting fat
crystals and liquid oil, usually a highly undesirable situation.
In practice, the highest possible triglyceride selectivity yields a fully random distribu
tion. If one-third of the fatty acids is saturated (S) and two-thirds is unsaturated (U), the
fraction of SSS will be almost 4%, and the fractions of UUU, SUU, and SSU will be
30%, 44%, and 22%, respectively.
• Finally, the trans selectivity describes the tendency for the formation of trans isomers,
defined as percent trans formed per unit IV reduction. Trans fatty acids greatly raise the
solid fat content below 30°C, resulting in steep melting products.
All types of selectivity, and hence melting behavior, are determined by three parameters:
temperature, bulk hydrogen concentration in the oil, and type of catalyst.
Temperature. The effect of temperature is that the rate by which each hydrogenation
reaction is increased is different, because the activation energies for each hydrogenation reac
tion are not expected to be the same. Moreover, the effect of a distribution of hydrogen
concentration within the catalyst particles will start to affect selectivity, because as tempera
ture increases, the diffusional limitation of hydrogen will start to exert its effect on selectivity.
These concentration distributions will be the result of diffusional limitation of hydrogen within
the catalyst particles. The lower the concentration, the higher the selectivity will be.
Bulk Hydrogen Concentration in the Oil. This parameter is not an independent parameter,
but it has been explained in the previous sections that the bulk hydrogen concentration is
determined by the hydrogen pressure, the volumetric liquid-phase mass transfer coeficient
kija, and the amount and type of catalyst used in the hydrogenation. The higher the pressure,
the higher the hydrogen solubility in the oil. It will increase in proportion to the pressure in
the pressure range used for hydrogenation. The higher the kua value, the faster the hydrogen
is transferred from the gas to the oil phase, which will result in a higher bulk hydrogen
292 Koetsier
concentration. It has been explained in Section II.B that higher kua values can be obtained by
a number of options, one of them being higher stirring speeds.
The more catalyst is used, the faster the hydrogen reacts away from the bulk, and the
lower its concentration will be. Equation 2 expresses the latter two effects quantitatively.
It has been experimentally observed for several systems of consecutive hydrogenation reac
tions that the selectivity for the intermediate product is better if the hydrogen concentration is
low [24]. A proper explanation for this observation has not yet been found.
Type o f Ni Catalyst. The main features of edible oil hydrogenation catalysts are (1) the
mean particle size and the particle size distribution, (2) the mean pore size and the pore size
distribution, and (3) the specific nickel surface area. The main reason for using small Ni
catalyst particles with a mean size of 4-10 jam is to minimize the phenomenon of diffusional
limitation. The molecules that are hydrogenated on the active Ni sites, which are located
throughout the whole particle volume, diffuse from the outside of the particle through the
intricate texture within the particle toward the center. At the same time there is a chance of
them reacting while passing an active site. As a result of these two simultaneously occurring
processes, a gradient will be obtained in the concentration of these two components.
The rate of diffusion of the triglyceride (TG) and hydrogen molecules through the outer
surface of the (spherical) particles (at r= R ) is given by Pick’s first law of diffusion:
dTG
^ T G ~ ~ ~~ ^pore^TG ,eff (47)
dCn2
*^H2“ ^pore^H2,eff (48)
r= R
where is the porosity of the catalyst particle, a property that also reflects the fraction of
the outer surface through which the molecules diffuse, i)TG,eff is an effective diffusion coeffi
cient of the oil molecules diffusing through the intricate structure of the catalyst particle, and
7^H2,eff is the same property for hydrogen. The concept of effective diffusion coefficient has
been defined by van Krevelen [25]. The molecules diffusing from the outer surface of the
particle in the direction of the pores have to follow a zigzag path and pass several narrow
passages, and they can be adsorbed as well. The zigzag factor r reflects the decrease in the
diffusion coefficient due to these three obstacles.
7^H2,eff —7^1H2I t (49)

The value of r has been measured for several different types of extradâtes and pellets 0.8-3
mm in size. Usually these values are on the order of 5. For highly porous catalysts one might
expect the value of r to approach 1 , because it becomes easier and easier for the molecules
to diffuse through the intricate texture of the catalyst particle.
The ratio of the rate of diffusion to the rate of reaction is decisive for the steepness of this
concentration gradient. If the diffusion rate cannot keep up with the reaction rate, then there
will be a steep concentration gradient. At the outer surface the concentration will be high, but
toward the center it can become very low, causing a decrease in the reaction rate. Such a
situation is usually described by an effectiveness factor rj, defined by the ratio
actual conversion rate within the catalyst particles

V = conversion rate in the absence of diffusional limitation (50)
In edible oil hydrogenation three regimes are observed:

1. No diffusional limitation. There is no diffusional limitation of oil molecules or hy

drogen at very low temperatures, for example at 80°C. An excessive amount of catalyst has
to be used to obtain a rate of hydrogenation that is sufficiently high, and it is therefore not of
commercial interest. In lab experiments it can be demonstrated that under these conditions the
quality of the product is not determined by the type of Ni catalyst as long as the catalyst
particles are in the micrometer size range. For example, the properties of the product obtained
by using colloidal unsupported Ni will be the same as if it were obtained by using Ni sup
ported by an inert carrier.
2. Diffusional limitation of oil molecules only. Once the temperature increases to above
100°C, the reaction rate increases to such an extent that the diffusion rate of the oil molecules
cannot keep up with the reaction rate. The quality of the hydrogenated product is now deter
mined by the type of catalyst, namely by the accessibility of the Ni sites for the large tri
glyceride molecules, the size of which is 1.5-2 nm. If the pores are long and narrow, the
diffusion rate of large triglyceride molecules cannot cope with the reaction rate, resulting in
steep concentration gradients along the length of the pores. Even at the beginning of the
reaction with polyunsaturated fatty acids still present, the deepest pore bottom only sees fully
hydrogenated products and the nickel surface area hardly contributes any more to activity.
Less deep in the pore, intermediate products may accumulate, and if the supply of the original
raw material is insufficient, through slow diffusion, intermediates cannot escape quickly from
the catalyst particle, and highly saturated products are formed. In other words, the composi
tion inside the pore runs ahead of that in the bulk of the liquid. Thus, narrowing of pores
impairs not only activity but selectivity as well. It may be necessary to increase pore radius
much further than 3 nm if high selectivity is required or if unselective conditions such as high
H 2 pressure, low temperature, or hydrogenation with low amounts of catalyst are used.
3. Diffusional limitation of oil and hydrogen molecules. At still higher temperatures,
for example at 180°C, the reaction is so rapid that the diffusional limitation of hydrogen also
starts to affect the reaction rate. A steep concentration gradient of hydrogen is then also
obtained, and the hydrogen concentration at the center of the catalyst particle will be consider
ably lower than the concentration at the outer surface. Despite the fact that a vast amount of
literature exists on the hydrogenation of edible oils, very little quantitative data have been
published on the degree of diffusional limitation of hydrogen within the nickel catalyst parti
cles. For optimization of the size of the catalyst particles, it is of interest to understand
whether any diffusional limitation of hydrogen exists.
To estimate the extent of diffusional limitation of hydrogen within the catalyst particles,
one needs to know the following properties:
Liquid phase diffusion coefficient of hydrogen in oil, see Ref. 5.

The tortuosity factor, r. One of the features of edible oil catalysts is the very high porosity.
Porosities of 70-85% are common. For such highly porous catalysts the tortuosity factor
is assumed to vary from 3 to 2, respectively.
The particle porosity, which can be calculated from the nitrogen physisorption data
and the skeletal density.
The volume of catalyst particles per wppm Ni. This can be calculated from the composition
and the particle porosity.
The Sauter mean diameter J 3 2, which is measured by the laser diffraction technique.
The mean hydrogen concentration within the bulk of the oil, which can be calculated using
Eq. (2).
The initial hydrogen absorption rate [mol/(môii • s)], which is calculated from the rate at
which IV changes over time.
294 Koetsìer
The hydrogenation rate per unit volume of catalyst is expressed by Eq. (24), which we
repeat here:
(24)
The flux /cat iri mol/(m^cat‘ s), diffusing into the particles at r = i? is
J^J,t)=R(t)d^2l6
^H 2 ^ ^H2
(51)
Ax at r = /?
or
ACH2
Ajc= — D H2 (52)
*^cat(0
And the smallest value of Ax, Ax^j„, is obtained for the highest value of the absorption
rate. Namely, for (i = 0).
'“pore ACH2
Axî = D H2
icat(?= 0)
The physical meaning of the value of Ax is the depth of hydrogen penetration if the de
crease in concentration had been linear. Of course, this decrease will not be linear, but in
any case this distance Ax can be considered a characteristic depth.
In Table 2 the value of Ax^în has been calculated for two different catalyst samples A
and B for the initial hydrogenation rate of a soybean oil. The data calculated for Ax^î^
show hardly any diffusional limitation for hydrogen within particles of a size of 5 /xm at a
temperature of 120°C, but at 180°C a severe diffusional limitation does indeed exist. Figure
22 depicts the hydrogen concentration gradient at x = R at 12 0 ° and 180°C, as can be
estimated from for catalyst A.
At 120°C a gradient at x = R can be observed, but this gradient will not result in severe
hydrogen depletion at x = 0. The gradient at 180°C is very steep, and this gradient will
result in severe hydrogen depletion at x= R . Moreover, it has been clearly demonstrated
that under these conditions hydrogenation occurs mainly within a thin layer at the outer
surface of the catalyst particle. Of course, as hydrogenation proceeds and the rate drops in
proportion to IV —75, Axîj^ will increase.
Table 2 The Depth of H2 Penetration for Two Catalysts and Two Hydrogenation Temperatures
Bulk H2
Temp. concn. J(t= 0) Gs Xat (f= 0 ) D hi
(°C) (mol/m^) [mol/(m\i,/s)] [mol/(m\3,/s)] (m^/s) (/xm)
Catalyst A, 0 .2 6
120 1.5 0.44 400 1.1 X 10-3 1 .3 x 1 0 -* 4.6

180 1.3 0.71 50 14.2 X 10-3 2.7X 10-* 0.64
Catalyst B, = 0.42
120 0.9 0.52 608 1 .9 x 1 0 -3 1.3X10-* 5.4

180 1.4 0.68 76 2 4 x 1 0 -3 2.7X 10-* 0.67
Fig. 22 The slope of the H2 concentration gradient at r = R if T = 120°d and 180°d (both of cat A).
B. Current Types of Ni Catalysts for Edible

Oil Hydrogenation
Over the past 10 years the development of new or improved versions of Ni catalyst systems
by most catalyst companies has been the result of a better understanding of the factors affect
ing activity, selectivity, catalyst poisoning, and filtration rate in combination with an under
standing of customer needs. Activity and poison resistance have in many cases been improved
to the extent that these catalysts can be used at starting temperatures as low as 120°C . New
catalysts have been introduced onto the market that exhibit considerably better selectivity and
are very useful in vegetable oil hydrogenation.
At Unichema Int. this development has led to the introduction of PRICAT 9910 and PRI-
CAT 9920. At Engelhard it has led to the introduction of Nysosel 222 and 545 and Nysosel
325 and 645; at UCI, of G 13 5 and G95D; at Mallinckrodt, of Calsicat 428 and Calsicat 472;
and at Hoechst, of 882 OF.
PRICAT 9910 is a multipurpose, sulfur-resistant, medium-pore catalyst with a high nickel
surface area suitable for all types of oils under quite selective conditions. PRICAT 9920 is
a very selective, wide-pore catalyst that has been particularly developed for vegetable oil
hydrogenation. The better poison resistance of PRICAT 9910 has been shown by the results
depicted in Figs. 11 and 19. To illustrate the difference between these two catalysts with
respect to their selectivity, their performance will be compared in a number of hydrogena
tion processes.
1. Exam ples o f C atalyst Perform ance in the M odification o f Oils

The first example below reviews the difference between the performances of the two catalysts
PRICAT 9910 and PRICAT 9920 in hydrogenating an nb-soybean oil. These data should not
be considered as characteristic of an nb-soybean oil, because the quality of an nb-soybean oil
is determined by several factors, such as phospholipid content. With other oil qualities, melt
ing points 1 - 2 ° C higher or lower may be obtained.
The data give the results of two sets of experiments, one at 150 kPa and the other at 400
kPa. In both cases the experiments were performed adiabatically at 120 and 180°C . The
volumetric liquid-phase mass transfer coefficient, kip, was 0.13 and 0.20 s~ ^ respectively.
296 Koetsier
Table 3 Characteristic Data for IV = 80 of a Hydrogenated Soybean Oil with IVq= 132.8 and a Feed
Composition of 4% C18:0, 23.1% C18:l, 53.1% C18;2 and 7.4% C18:3
120°C; 150 kPa 180°C; 150 kPA
PRICAT 9910 PRICAT 9920 PRICAT 9910 PRICAT 9920
Hydrog. time, min 75 85 74 83

Melting point, °C 34.0 29.5 31.5 30.5
Trans content, wt % 29.0 30.0 40.0 40.0
Solid fat content by NMR
A^35 1.5 0 .2 0 .2 0.1
GC analysis of fatty acids
18:0, % 9.5 7.5 6.8 7.0
18:1, % 64.0 70.5 71.5 71.8
18:2, % 13.5 10.0 9.2 9.1
18:3, % 0 .2 0.0 0.0 0.0
In Tables 3 and 4, the data for the slip melting point, A/35, and GC analysis of the Cjg
fatty acid composition, clearly indicate the higher selectivity of PRICAT 9920 versus PRICAT
9910. These differences are particularly significant when the hydrogenation is performed un
der nonselective conditions.
In Figs. 23 and 24 the slip melting points of the hydrogenated oils are plotted as a function
of the iodine value for temperatures of 120 and 180°C respectively. In Fig. 23 it is seen that
by using the wide pore catalyst PRICAT 9920 the lowest slip melting points are obtained;
whether the pressure is 150 kPa or 400 kPa; the slip melting points at I Vs ranging from 60 to
105 are not a function of pressure. This is not observed when PRICAT 9910 is used. Then
the slip melting points at a particular IV are several degrees higher. The highest values are
obtained at the highest pressure. Under the most selective conditions, namely 180°C and 150
kPa, the differences between the slip melting points at a given IV are only on the order of 1°C.
The same conclusions with respect to the difference in selectivity between these two cata
lysts can be made if one considers the other characteristic data, which are summarized in
Tables 3 and 4 for an IV of 80. For example, the solid fat content at 35°C, has its
Table 4 Characteristic Data for IV = 80 of a Hydrogenated Soybean Oil with IVq= 132.8 and a Feed
Composition of 4% C18:0, 23.1% C18:l, 53.1% C18:2 and 7.4% C18:3
120°C; 400 kPa 180°C; 400 kPa

Melting point °C 37.0 29.0 35.0 31
Trans content, wt % 28.0 27.0 39.0 40.0
3.0 0 .2 1.0 0.15
18:0, % 12.0 9.0 9.5 8 .0
18:1, % 63.0 69.0 67.5 69.0
18:2, % 13.0 10.5 10.5 9.5
18:3, % 0.1 0 .0 0.3 0.1
Fig. 23 Slip melting point versus IV; hydrogenation at 120°C.
highest value if PRICAT 9910 is used at 120°C and a pressure of 400 kPa, and the lowest
values are obtained for PRICAT 9920. Similar conclusions can be drawn from the GC analy
ses of the four Cjg fatty acids. A minimal difference is obtained if the reaction is performed
under the most selective conditions, namely at 180°C and 150 kPa, but a considerable differ
ence in selectivity is seen at 120°C and 400 kPa. Then a much higher selectivity is obtained
if the wide-pore catalyst is used. All in all, these results convincingly show the higher selec
tivity of the wide-pore catalyst PRICAT 9920.
Finally, the solid fat content curves at IV = 80 are depicted in Figs. 25 and 26. For the
experiments at 120°C it can be concluded that, using PRICAT 9920, the oil can be hydroge
nated to a lower IV to get the same curve as obtained by using PRICAT 9910. The advantage
can be that a lower 18:3 content is obtained—in other words, an oil with higher oxidative sta
bility.
Figure 26 depicts the difference in steepness of the solid fat content curve at IV = 80 for
the two cases where ( 1 ) the most selective catalyst is used under the most selective conditions
and (2) the medium-pore catalyst is used under the least selective conditions.
The advantage of a low 18:3 content while the solid fat content at 10°C, N^q, is still low
is depicted by the results given in Fig. 27. In liquid table oils a low A jq in combination with
Fig. 24 Slip melting point versus IV; hydrogenation at 180°C.

298 Koetsier
Solid Fat Content soybean oil with IV = 80

120°C, 400 kPa (points) and 150 kPa (lines)
Solid Fat Content in %

60
50
40
PRICAT 9910
30
20
10
PRICAT 9920
0
10 20 30 35
Temperature in °C
Fig. 25 Solid fat content soybean oil with IV = 80; 120°C, 400kPa (points) and 150 kPa (lines).
Solid Fat Content soybean oil with IV = 80

180°C, 150 kPa and 120°C, 400 kPa
10 20 30 35
Temperature in °C
Fig. 26 Solid fat content soybean oil with IV = 80; 180°C, kPa and 120 C, 400 kPa.
” Liquid table oils "

low solid fat and low linolenic acid content
Highly selective catalyst Medium selective catalyst
PRICAT 9920 PRICAT 9910
soHd content at 10°C{%) soUd content at 10°C (%)
linolenic acid content (%) Unolenic add content (%)

SPP24/2/94.FRS
Fig. 27 Liquid table oils: low solid fat and low linolenic acid content.
Table 5 Characteristic Data for IV = 80 of a Hydrogenated Rapeseed Oil with 113.7 and a
Feed Composition of 1.7% C18;0, 60.6% C 18 :l, 20.7% C18:2 and 8.8% C18:3
120°C; 400 kPa; 440 wppm Ni 180°C; 400 kPa; 110 wppm Ni

Melting point °C 37.0 28.0 38.5 30.5
Trans content, wt % 22.0 22.0 36.0 43.0
Nss 1.5 0.1 4.4 2.5
18:0, 1.7 12.0 11.0 12.0 10.0
18 :1, 60.6 72.0 73.0 73.2 76.0
18:2, 20.7 8.0 8.0 6.6 6.0
18:3, 8.8 0.5 0.5 0.2 0.1
a low 18:3 content is an important quality criterion. This criterion can be better achieved if
PRICAT 9920 is used.
Table 5 summarizes the same characteristic data for the product properties of a rapeseed
oil, hydrogenated to IV = 80. Again it is demonstrated that the highest selectivity is obtained
by using the wide-pore catalyst PRICAT 9920. However, in this hydrogenation the presence
of 3 wppm S in the canola oil causes, at 180°C, a significantly larger increase in the trans
content due to the fact that on the basis of the same Ni content a larger fraction of the active
nickel surface of PRICAT 9920 has been sulfided.
2. E ffect o f Phospholipids on P roduct Properties

The effect of phospholipids, added as lecithin, on the performance of these two types
of catalysts is demonstrated by the data given in Table 6 . Soybean oil was hydrogenated
Table 6 Data of a, to an IV of 80 Hydrogenated Soybean Oil with IVo= 132.8 and a Feed
Composition of 4% C18:0, 23.1% C 18 :l, 53.1% C18:2 and 7.4% C18:3
180°C; 300 kPa; 100 wppm Ni
wppm P 0 0 10 10
Melting point, °C 3 1.5 29.5 42.5 27.5
Trans content, wt % 39.5 40.0 32.0 34.5
45 38 48 32
N 20 22 17 31 13
Nso 6.5 4 14.5 2
1.5 0.0 7.5 0.0
18:0, 4.0 4.0 7.5 11.5 7.5
18 :1, 23.1 69.5 70.0 62.5 70
18:2, 53.1 10.5 11 13 10.5
18:3, 7.4 0.0 0.0 1.5 0.0
300 Koetsier
at 180°C and 300 kPa to an IV of 80, in one set of experiments without the addition of
lecithin and in one set in which lecithin was added in an amount such that 10 wppm P
was present.
Lecithin has a dramatic effect on the hydrogenated product properties if PRICAT 9910
is used but hardly any when PRICAT 9920 is used. The following effects are observed for
the two catalysts PRICAT 9910 and PRICAT 9920 if lecithin is added in amounts of 10
wppm P:
1. The slip melting point increases by 11°C and drops by 2°C, respectively.
2. Trans content decreases by 7.5% and 6%, respectively.
3. Flatter dilatation curves are obtained.
4. Considerably less selective hardening is seen for PRICAT 9910.
In other words, if PRICAT 9920 is used in such a case, then the steepness of the dilatation
curves at a particular IV will be less, but only because of the lower trans content. Hardly any
difference can be observed in the distribution of the C^g fatty acids between the two cases.
Hence, if one wants to hydrogenate a soybean oil containing only a few parts per million P
by weight, then a catalyst such as PRICAT 9920 is the preferred type.
Two other aspects should also be noted. One is the reduction in the hydrogenation rate,
and the other is the extent to which lecithin is adsorbed onto the catalyst. The hydrogenation
times (HT) have been increased by the addition of 10 wppm P from 29 min to 44 min for
PRICAT 9910 and from 32 min to 55 min for PRICAT 9920. PRICAT 9920 usually quantita
tively adsorbs all the lecithin from the oil.
C. Reduction of Trans Content

Today an important issue in hydrogenating edible oils is to reduce the trans content as much
as possible. Low temperature and high pressure are the preferred conditions to get a low trans
content. This effect is reflected in Fig. 28 for a fish oil hydrogenation. In this figure the trans
content is plotted as a function of the slip melting point. At a slip melting point of 35°C it is
possible to get a reduction of 30% in the trans content. However, it is possible to achieve a
much lower IV if PRICAT 9920 is used. This aspect is depicted by Fig. 29, in which the slip
melting point is plotted as a function of the iodine value. The same melting point is achieved
at iodine values of 72 and 83 by using PRICAT 9920 and PRICAT 9910, respectively. The
melting curve of the oil that was hydrogenated to an IV = 72 by using PRICAT 9920 is much
Fig. 28 Reduction of 30% in TRANS can be obtained at the reduced temperature.

S lip Melting Point hydrog. ftsli ©II C
— PRICAT @910
■■■■■PRICAT 9920
300 kPa
120X
0,45 wt% cat
6065 70 75 80 85 905^51<K10ai<m2a25
Iodine Value of hydrogenated fish oil
By using PRICAT 9920 the fish oil can be hydrogenated

to an IV of 10 points lower at the same SMP of 35°C
Fig. 29 By using PRICAT 9920 the fish oil can be hydrogenated to an IV of 10 points lower at the
same SMP of 35 C.
steeper than the curve of the oil that was hydrogenated to IV = 83 by using PRICAT 9910.
PRICAT 9920 produces at low temperature an oil with the same melting curve steepness as
is obtained by using PRICAT 9910 at 180°C, but with 30% less trans isomers.
IV. DISCUSSION AND CONCLUSION

The rate of hydrogenation, expressed as the hydrogenation time (Eq. (44)) required to drop
the initial IV value, IV^, to an end IV value, IVend, can be expressed by a simple equation
consisting of two terms—the minimum hydrogenation time and a term reflecting all effects
that determine the actual hydrogenation rate within the particles. The minimum hydrogenation
time reflects all factors that affect the rate at which the hydrogen is transferred from the gas
to the liquid phase, such as stirring speed and hydrogen pressure, and the drop in IV. It is a
meaningful parameter, and it provides more information than the stirring speed alone, the
parameter that has been used by many authors. To give the stirring speed as a relevant param
eter is misleading, because it does not give any valuable information. Size, type, position
of the impeller, and presence of baffles and coils are other factors affecting the minimum
hydrogenation time.
As has already been said, the second term in the hydrogenation time equation reflects all
the factors affecting the reaction rate within the catalyst particles. It has been demonstrated
that for feedstocks that do not poison the catalyst this correlation yields a straight line if the
hydrogenation time is plotted versus the reciprocal amount of catalyst. This observation was
made by Coenen [26] in 1969. If there are any components in the oil that will poison the Ni
catalyst, this correlation will no longer result in a straight line but will bend upward. The
extent to which this line deviates from a straight line is a measure of the quality of the
feedstock with respect to the amount of Ni catalyst poisons present. For example, it has been
demonstrated that for nb-soybean oils this correlation results in an almost straight line, but
for nb-canola oils it is no longer straight. The most severe poisoning occurs when fish oils
are used. However, in this case also the extent to which bleaching has a favorable effect on
the quality of the feedstock has been demonstrated.
Once the correlation between the hydrogenation time and 1000/wppm Ni has been deter
mined for one particular minimum hydrogenation time, it has been demonstrated in a number
of cases that this line can be moved upward/downward if the process is operated at other
302 Koetsier
minimum hydrogenation times. This is a very important observation, because it can be used
to predict the hydrogenation time in different types of large-scale reactors if all processing
conditions are the same. The only prerequisite is the knowledge of the minimum hydrogena
tion time.
The quality of the hydrogenation catalysts has been improved over the past decade. In
addition to medium-pore catalysts, wide-pore catalysts are now commercially available. By
using these wide-pore catalysts in vegetable oil hydrogenation, considerably better selectivity
can be obtained. All in all, we have come to the conclusion that the technique described
in this chapter is a valuable method of correlating hydrogenation time with process param
eters.
ACKNOWLEDGMENTS
I would like to express my appreciation for the valuable comments made by Mr. Senno van
der Velde, Unichema’s Business Unit Manager Catalyst; Dr. Peter Scherm, Catalyst Develop
ment Scientist; Mr. Wolfgang Geuking, Market Application Manager, and Mr. Robert Riitjes,
Catalyst Characterization Engineer. Finally I thank Mrs. Mary Stephens, Deutsche Unilever
GmbH, Hamburg, for help in correcting my English.
REFERENCES
1. H. B. W. Patterson, Hydrogenation of Fats and Oils: Theory and Practice, AOCS Press, Cham
paign, IL, 1994, pp. 52-68.
2. D. Swem (Ed.), Bailey’s Industrial Oil and Fat Products, 4th ed., Wiley, New York, 1979, Vol.
1, Chap. 3.
3. K. Andersson, M. Hell, L. Lowendahl, and N. H. Schöön. Diffusivities of hydrogen and glyceryl
trioleate in cottonseed oil, J. Am. Oil Chem. Soc. 51: 171-77 (1974).
4. J. M. Coulson and J. F. Richardson, Chemical Engineering, 3rd ed., Pergamon, New York, 1977,
Vol. 1, Chap. 8.
5. K. L. Ganguli. Measurements of H2/edible oil interfacial area in an agitated hydrogenator using a
Ziegler-Natta catalyst, Ph.D. Thesis, TH-Delft, 1975; see also Ref. 3.
6. N. Frössling, Über die Verdunstung fallender Tropfen, Beitr. Geophys. 52: 170 (1938).
7. J. M. Coulson and J. F. Richardson, Chemical Engineering, 3rd ed., Pergamon, New York, 1977,
Vol. 2, Chap. 3.
8. P. L. T. Brian and H. B. Haies, A. /. Ch. E. J. 15: 419 (1969).
9. P. Sänger and W. -D. Decker, Liquid-solid mass transfer in aerated suspensions, Chem. Eng. J.
22: 179 (1981).
10. M. Alderliesten, A nomenclature for mean particle diameters, Anal. Proc. 21: 167-172 (1984).
11. L. L. van Dierendonck, The specific contact area in gas-liquid reactors, Proc. 4th Eur. Symp.
Reaction Eng., Brussels, 1971, p. 205.
12. Bailey’s Industrial Oil and Pat Products, 4th ed. (D. Swem, Ed.), Wiley, New York, 1979, Vol.
1, Chap. 3.
13. P. Varey, Towards the flexible impeller, Chem. Eng., Sept. 24: 577-33 (1992).
14. A. Bakker, J. M. Smith, and K. J. Myers, How to disperse gases in liquids, Chem. Eng. 707(12):
98-104 (1994).
15. D. Urosevic, Buss technology for oils and fats and fatty acid derivatives, presented at the ISF-
JOCS World Congr. 1988, Tokyo.
16. A. N. Kolmogorov, The local stmcture of turbulence in incompressible viscous fluid of very large
Reynolds numbers, C. R. Acad. Sei. URSS 30: 301.
17. K. R. Westerterp, Chemical Reader Design and Operation, Wiley, New York, 1984, p. 146.
18. E. W. Thiele, Relation between catalytic activity and size of particle, Ind. Eng. Chem. 31: 916
(1939).
19. M. Perdereau, Structure, mécanisme de formation et stabilité de la ouche d’adsorption du soufre

sur le nickel, Surf. Sci. 20: 80 (1970).
20. L. F. Albright, in Hydrogenation, Proceedings of an AOCS Colloquium, Ed (R. Hastert, Ed.),
AOCS Press, Champaign, IL, 1987.
21. H. I. Waterman, Hydrogenation of Fatty Oils, Elsevier, New York, 1951.
22. J. W. E. Coenen and B. G. Linssen, in Physical and Chemical Aspects of Adsorbents and Cata
lysts (B. G. Linssen, Ed.), Academic, New York, 1970.
23. L. F. Albright, Quantitative measure of selectivity of hydrogenation of triglycerides, JAOCS 42:
350 (1965).
24. C. Niklasson, B. Andersson, and N. Schoon, Influence of hydrogen pressure on selectivity in
consecutive hydrogenation reactions, Ind. Eng. Chem. Res. 26: 1459-1463 (1987).
25. D. W. van Krevelen, Stoftransport en kinetiek bij contact katalyse, Chem. Weekbl. 48: 421 (1951).
26. J. W. E. Coenen; Hydrogenation of oils and fats, in Margarine Today, Proc. of a seminar held at
Dijon Univ., March 1969, pp. 62-91.
11
Butter, Margarine, Spreads, and Baking Fats
Eric Flack
Danisco Ingredients (UK) Limited, Bury St. Edmunds, Suffolk, England
L BUTTER
A. History
The processing of milk to produce butter has been known for thousands of years. A Sumerian
frieze from about 4000 BC shows milk being processed and, according to archaeologists, the
milk was being whipped to form butter [1]. It is also recorded in the Hindu Veda, written
over 3500 years ago, that Hindus valued their cows in relation to the amount of butter that
could be obtained from their milk [2]. Of course, the principle involved in these ancient
processes, i.e., phase inversion, still applies today although the milk used for making butter
has been substituted by more concentrated milk fat in the form of cream.
B. Early Developments
Initially, butter was produced manually by agitating and beating the milk in simple wooden
chums. Wood was long considered ideal for the constmction of chums as, apart from being
readily available and easily fashioned, it was also one of the natural materials to which butter
would not adhere during production, that would not impart off-flavors to the product, and that
would not be adversely affected by the milk or the salt used in processing. Unfortunately,
wood is difficult to sterilize, which is a severe disadvantage, and wears out with use. Thus,
the introduction of all-metal equipment to overcome these difficulties was welcomed. Other
developments in the late nineteenth century also led to the production of microbiologically
safer butter on a large scale.
The mechanical separation of cream, the pasteurization of milk products, and the introduc
tion of butter starter cultures in the 1880s enabled the establishment of dairies that could
produce butter in much greater quantities and thus heralded the beginning of modem but
termaking.
305
306 Flack
C. The Process
The process of manufacturing butter basically transforms cream into butter grains and butter
milk by agitation and by beating air into the cream [3]. The individual stages of butter grain
formation, i.e., phase reversal from oil in water to water in oil, are explained as
1. Building up foam from the skim milk

2. Adsorption of fat globules at the foam lamellae
3. Agglomeration of fat globules
4. Destabilization of the foam and mechanical clumping of the agglomerates to form
butter grains
The microstructure of the butter grains as well as the finished butter is characterized as a
dispersion of water, air, fat crystals, and fat globules in oil. The destabilization of the cream
and the subsequent phase separation involve a series of overlapping dynamic processes.
During the early stages of churning, air is whipped into the cream and the proteins (lacto-
globulin, lactalbumin, and casein) form an interfacial film around the air bubbles [4]. As a
result, an unstable foam is formed, and the mechanical stress during churning causes the fat
globules to lose their membrane so that the globule surface becomes hydrophobic [5] . Due to
their lower surface tension compared to milk proteins, they accumulate at the air/serum inter
faces of the air bubbles. Through the action of shear forces on the semisolid fat phases,
the globules become deformed and release liquid fat, which promotes agglomeration of the
fat globules.
Although large-scale batch churning is still in operation, it has been replaced to a very
large extent by the continuous buttermaking process introduced in the 1950s by Fritz.
D. Cream Preparation
Cream is prepared by the centrifugal separation of whole milk at 40-50°C to a fat content of
36-42% for continuous buttermaking and 2 5 -3 5 % for batch production. It is pasteurized at
8 5 -1 10°C for 1-30 s to destroy microbial contamination and to inactivate enzymes. The
cream is then left to rest at varying temperatures for several hours while it undergoes physical
ripening. If cultured or lactic butter is desired (as against sweet cream butter), starter culture
is added at this stage to initiate bacteriological ripening, during which lactose is partially
converted to lactic acid, thus reducing the pH from, say, 6.5 to 4 .6 -4 . 8. Simultaneously, the
culture will develop the typical lactic butter flavor through the generation of diacetyl.
E. Buttermaking
The process of buttermaking may be illustrated in Fig. 1 [6].
The main elements of the process may thus be summarized as follows:
1. Cream preparation. See Section D.
2. Churning. During the churning phase, inversion from o/w to w/o takes place and
buttermilk is separated off. Air is whipped into the cream, creating a foam at the air/serum
interface where the fat globules are concentrated. Due to mechanical forces, some of the
globules become damaged, releasing liquid fat, which promotes agglomeration of the re
maining fat globules and the fat crystals from the damaged globules. This process continues
until pea-sized butter granules are formed and the buttermilk is separated.
3. Draining and Washing. The separated buttermilk is drained off. However, where
draining off is insufficient, the butter granules may be washed with cool, clean water to
Butter, Margarine, Spreads, and Baking Fats 307
M IL K ( 3 -4 % m ilkfat)
Centrifugal separation
CREAM SKIM MILK

(3 0 -4 5 % m ilk fat) (0 .0 5 % m ilk fat)
Pasteurization
Physical ripening
Bacteriological ripening (optional)
Churning
Draining
B U T T E R G R A N U LE S B U T T E R M IL K
( 0 .2 -0 .6 % m ilk fat)
Washing (optional)
Salting (optional)
Kneading
BU TTER
(8 0 -8 2 % m ilk fat)
Packaging
F IN A L P R O D U C T
Fig.1 The process of buttermaking.
reduce the buttermilk solids content. Removal of buttermilk and wash water is controlled to
allow for the convenient adjustment of the final moisture content of the butter.
4. Salting. Invariably, butter, whether sweet cream or lactic, will be salted at levels
from 0 . 1 % to 2 % by the addition of brine prior to kneading.
5. Kneading. The butter granules, together with brine and additional water when re
quired, are kneaded either by continued rotation of the chum in batch processing or by means
of worm-screw conveyors in continuous processing. This ensures a fine water distribution and
a smooth texture in the finished butter. In some processes, kneading is carried out in two
stages and also includes vacuum treatment to remove occluded air.
F. Final Composition
The legal requirements for butter and variations containing lower fat contents are listed later
in Table 2.
308 Flack
II. MARGARINE
A. Introduction
Margarine is a fatty food resembling butter in appearance, character, and composition that is
used as a substitute for or alternative to butter. Statutory definitions may vary in different
countries (see later), but usually the term “margarine” is applied to fatty foods similar to
butter except that the fat is not, or only to a minor extent, derived from milk fat.
B. History and Early Developments

The invention of margarine in the 1860s is attributed to the French chemist Hippolyte Mége
Mouriés, in response to a competitive challenge initiated by the French government under
Napoleon III for a less expensive and less perishable substitute for butter. Mége had already
undertaken research in this field at the imperial farms, where he had observed that although
underfed or hungry cows lost weight, they still produced milk, though of lower than normal
yield, and the milk contained fat. Thus, he deduced that milk fat was derived from normal
body fat, i.e., tallow. However, as tallow itself does not possess the melting property of milk
fat, he construed that through some metabolic process a fractionation of the fat occurred, the
lower melting fractions of which were transported to the mammary glands of the udder
whereby, through the enzymatic action of pepsin, it became transformed to butterfat and
dispersed as an emulsion in the milk plasma, while the harder fractions were used by the
animal as a source of energy [7].
Having arrived at these conclusions, Mége set out to imitate the natural process by care
fully rendering fresh tallow at body heat (about 45°C) using artificial gastric juices to facilitate
separation of tissue from pure fat and then, following crystallization at 25-30°C, to extract
under pressure about 60% of a soft semifiuid fraction—oleomargarine— and about 40% of a
hard white fat—oleostearine. The softer fraction had a melting point similar to butterfat, a
pleasant flavor not unlike melted butterfat, and a pale yellow color. It could easily be plasti
cized and, although different in fatty acid composition from butterfat, it offered a useful basis
for the production of a substitute. He assumed, therefore, that this new butterfat substitute
consisted of glycerides of margaric and oleic acids (the former then considered to be a C 17
homologue of the saturated fatty acids although now known to be a eutectic mixture of pal
mitic and stearic acids) that solidified in crystals having a pearly luster. The name margarine
was derived from the Greek word margantes meaning pearl.
Mége’s process was to take a proportion of the soft fat with appropriate quantities of milk
and water, into which a small amount of udder extract was stirred. Following agitation, a
stable emulsion similar to thick butter cream formed that, on further c h u rn in g , thickened to
resemble butter. Although this process may now sound rather rudimentary, it nonetheless
comprised all the elements of margarine production as we know them today.
Mége took out patents in France and England, but the next major development took place
in Holland. In 1871, in the absence of patent law in Holland at that time, he sold his invention
to the butter merchants Jo h a n n e s and Anton Jurgens in the village of Oss. Also in the same
locality were another family of butter merchants— Simon and Henry Van den Bergh— who
commenced production shortly afterwards, later joined forces with the Jurgens, and eventually
become the largest margarine manufacturers in the world [8].
Production followed in most countries in Europe—Austria (1873/4), Italy (1874), Germany
(1874), Norway (1876) and in Denmark (1883) which was later to become the largest per
capita consumer of margarine. Mége was granted a U.S. patent in 1874.
Many variations were developed in the following years, and patents on new formulations
and processes taken out. At the same time there was strong opposition to the introduction of
margarine to the market from the farming community, and in some places anti-margarine
legislation was introduced. Opposition from such quarters continued well into the 20th century
and is not entirely unknown today.
C. Definitions and Descriptions

The basic composition of margarine, or oleomargarine, has been more or less established
since Mège’s original formulation—that is, it should be similar to butter. However, with
the new developments during more recent years involving different components and varying
combinations of vegetable oils, animal fats, and milk fat, a range of new descriptions have
come into being such as low-fat spreads, low-calorie spreads, and yellow-fat spreads. In 1993,
Moran [9] listed the variety then available as shown in Table 1.
In Europe, the adoption of Council Regulation (EC) No. 2991/94 in December 1994
brought all products, including butter, under the one description “spreadable fats” and de
scribed them as “products in the form of a solid, malleable emulsion—with a fat content of
at least 10 % but less than 90% by weight. The fat content, but excluding salt, must be at
least two-thirds of the dry matter.” These terms are further qualified as “products which will
remain solid at a temperature of 20°C and which are suitable for use as spreads.” The defini
tions and sales descriptions are detailed in Table 2.
Clause 2 of Article 3 of this regulation further states that the sales descriptions “minarine”
and “halverine” may be used in place of “half-fat margarine.” Fat spreads falling outside these
standards, for instance those with fat contents below 10% and concentrated products with fat
contents of 90% or more, may not be classed as spreadable fats.
European legislation does not specify vitamin fortification, which is left for decision at the
national level. However, in the United Kingdom the requirement will remain as previously
specified in the Margarine Regulations (SI 1867) 1967. That is, each 100 g of margarine shall
contain 800-1000 ¡xg vitamin A and 7.05-8.82 ¡xg vitamin D. This does not apply to other
Table 1 Some Current Spreads
Approx fat
Spread content (%)
High fat spreads

Butter 80
Margarine, packet 80
soft 80
Polyunsaturated fatty acids 80
(PUPA) spreads
Vegetable fat/butterfat blended spreads 80
Low fat spreads
Vegetable fat spreads 70, 60, 40
Vegetable fat/butterfat blended spreads 70, 40
Butterfat spreads 40
Very low fat spreads 20-30
Water continuous spreads 15, 9, 5
Source: Ref. 9.
310 Flack
Table 2 EC Definitions and Descriptions for Spreadable Fats
Fat Group Definitions Sales description Product categories ^
A. Milk fats. Products derived 1. Butter Milk fat content not less than 80%
exclusively from milk and/or but less than 90%. Max. water
certain milk products for which 16%; max dry nonfat milk mate-
the fat is the essential constit- rials 2 %.
uent of value.
2. Three-quarter Milk fat content not less than 60%
fat butter but not more than 62%.
3. Half fat butter Milk fat content not less than 39%
but not more than 41%.
4. Dairy spread Products with milk fat contents
x% less than 39%; more than 41%,
less than 60%; more than 62%,
less than 80%.
B. Fats. Products derived from 1. Margarine Vegetable and/or animal fats not
solid and/or liquid vegetable less than 80%; less than 90%.
and/or animal fats suitable for
human consumption, with a
milk fat content of not more
than 3% of the fat content.
2. Three-quarter fat Not less than 60%; less than 62%.
margarine.
3. Half fat margarine Not less than 39%; more than
41%.
4. Fat spreads x% Products with fat contents: less
than 39%; more than 41%, less
than 80%.
C. Fats composed of plant and/or 1. Blend Product from a mixture of vegeta-

animal products. Products de- ble and/or animal fats with a fat
rived from solid and/or liquid content of not less than 80%
vegetable and/or animal fats but not more than 90%.
suitable for human consump
tion, with a milk fat content of
between 1 0 % and 80% of the
fat content.
2. Three-quarter fat Not less than 60%; not more than
blend 62%.
3. Half fat blend Not less than 30%; more than
41%.
4. Blended spread x% Products with fat contents: less
than 80%.
Source: Official Journal of the European Communities L316, Vol. 37, Dec. 9, 1994.
Âdditional description with indication of % fat content by weight.
spreads, so vitamin fortification in these cases are matters for individual decision by the manu
facturers and/or retailers.
Although legislation in the United States and Canada is currently less detailed than in Europe,
both also require a minimum of 80% fat. In the United States FDA 166.110 describes margarine
(or oleomargarine) as “food in plastic form or liquid emulsion” containing not less than 80% fat
and describes a method for fat determination. It specifies vitamin A fortification to be not less
than 15,000 I.U. per pound but leaves vitamin D as an optional ingredient not less than 1500
I.U./lb. Although spreads are not specified in the legislation, low fat spreads constituted the fast
est growing category in 1990, with spreads containing 20-72% fat on the market [10].
In the United States, the National Association of Margarine Manufacturers describes
spreads as having fat contents ranging from 18% to 79%.
Canadian Standard B.09.016 states that margarine shall be “a plastic or fluid emulsion of
water in fat, oil or fat and oil that are not derived from milk,” and shall contain not less than
80% fat and not less than 3300 I.U. vitamin A and 530 I.U. of vitamin D. Calorie-reduced
margarine is specified in Standard B .09.017 as containing not less than 40% fat and having
50% of the calories normally present in margarine.
The conversion of international units of vitamins A and D to microgram equivalents is
complicated in the case of vitamin A as it depends on the origin of the fat. Thus, to convert
vitamin A from I.U. to ¡xg retinol in foods of animal origin, the I.U. of retinol should be
multiplied by 0.3. However, for foods of plant origin, the I.U. of ^-carotene should be
divided by 10. This arises because 1 I.U. of vitamin A is equivalent to 0.3 /xg retinol or 0.6
/xg /3-carotene. The conversion of the units for vitamin D is straightforward: 1 I.U. vitamin
D = 0.025 /xg.
D. Structure and Raw Materials

Margarine is a water-in-oil (w/o) emulsion; that is to say, the water (the disperse phase) is
distributed as droplets within the oil (the continuous phase). In margarine, the levels of each
largely simulate that found in butter and are regulated in most places by legislation: minimum
80% fat, maximum 16% water, the remainder consisting of salts, proteins, emulsifiers, vita
mins, colors, and flavors.
Whereas production initially was based upon the use of fractionated animal fats such as
tallow and lard, these were superseded later by vegetable oils and, for a period, marine oils,
both of which offer a much greater flexibility in physical characteristics and economy.
Being the main and most expensive ingredient, the fat blend is also the most important
factor in the formulation of margarine. As oil prices fluctuate, it is essential to have a fat
blend providing the desired quality at minimum cost. Thus, it is vital to be able to utilize
alternative formulations based on prevailing price conditions while still maintaining quality.
Therefore, a full understanding of the physical characteristics of the individual oils composing
the blend is critical. In this respect, factors of importance for the fat phase include the solid/
liquid fat ratio, crystallization rate, and melting properties.
Linear programming has been used for many years to formulate to optimum cost, perfor
mance, and specification. Hydrogenation has allowed greater flexibility in fat blend formula
tion. The negative aspect of trans acids as perceived by the consumer has, however, resulted
in manufacturers turning to natural sources of hardstocks (solid phase) to achieve the desired
texture and meltdown of products. The EU margarine manufacturers, for example, plan to
reduce trans acids to below 5% in margarine by the end of 1996.
In the early days of margarine production, large quantities of milk and water in almost
equal proportions were used for emulsification with the fat, with much of the excess water
312 Flack
being pressed out during kneading to arrive at a final moisture content of 16% or lower.
Nowadays, the aqueous phase is added at optimum levels and consists usually of a simple
solution of skim milk powder or whey powder in water, with the addition of salt, and the pH
is adjusted to 5 .5-6.0 by using, for instance, citric acid.
E. Fat Crystallization
The structure and physical properties of lipids have been covered in earlier chapters, but it is
worth reviewing the important issue of fat crystallization. Variations in the crystallization prop
erties of fats and differences between batches of a similar origin may cause problems in produc
tion even though processing techniques are fully controlled and highly automated [11]. The crys
tallization rates of some individual and blended fats are shown in Table 3.
Oils and fats are polymorphic and can crystallize in more than one form, which can differ in
terms of melting point, density, heat of fusion, and rate of crystallization. The three crystal
forms are commonly known as alpha (a), beta-prime (/3'), and beta (j8), with the a form having
the lowest values, ¡3' intermediate, and the most stable (3 form the highest values. Transition
from one form to another is a f 3 ' ¡3 and is not reversible without remelting the fat.
The polymorphism of crystalline fats may cause problems with the consistency of marga
rine and spreads. During production, the fats initially crystallize in the a form and normally
will rapidly transform to the j8' form. This is the desirable form for spread production as the
small needle-shaped f3' crystals (about 1 fxm long) impart good plasticity. Should the crystals
transform from the /3' form to the much larger ¡3 form (>20 /xm) they will give the spread a
grainy consistency known as sandiness, as can be seen in Fig. 2 [12].
Sandiness may also be caused by the physical interaction of triglycerides to produce higher
melting systems (compound formation). This can occur, for example, with triglycerides of the
forms POP/OPP (P = palmitic, O = oleic) when lard and palm oil are blended. Some care
must be taken, therefore, when formulating hardstocks from such blends [13]. At the same
time, the specific area of the crystal surface will reduce, allowing the liquid oil to penetrate
to the surface of the margarine, which may then lead to oiling out, especially if the margarine
comes under pressure. Some vegetable oils such as partially hardened sunflower or low-erucic
rapeseed are particularly prone to form j8 crystals and thus can cause sandiness in spreads. In
this case, sorbitan tristearate at 0.3% of the fat has been found to inhibit the transition from
¡3' to ¡3 form.
The crystal-modifying effect of a range of emulsifiers— sorbitan esters, ethoxylated sorbitan
esters, ethoxylated fatty alcohols, citric acid esters of monoglycerides (Citrem), diacetyl tar-
Table 3 Crystallization Rates of Fats
Fat Time (min.) Temp. (°C)
Coconut 3 20
Coconut/palm 1:1 4 15
Coconut/palm, interesterified 5 18
Palm hardened 5 17
Palm hardened/palm 2:1 8 13
Lard 14 10
Lard/palm 1:1 15 10
Palm 27 10
Shea fat 45 10
Source: Ref. 11.

taric acid esters of monoglycerides (Datem), sucrose monostearate, sodium stearoyl lactylate,
and polyglycerol esters— on the polymorphism of tristearin (glyceryl tristearate) has been in
vestigated [14]. In these trials it was found that sorbitan monostearate and Citrem (Cj^-Cig
fatty acids) were the most effective in preventing the recrystallization (from a to (3 form)
of tristearin.
It should be noted that when used in emulsions, surface-active materials will be absorbed
at the oil/water interface and that only lipophilic emulsifiers with high solubility in the oil
phase can perform as crystal inhibitors in emulsions.
Varying the level of saturated fat affects the degree of crystallinity of the oil phase [10].
Spreads with 40% fat were made using oil phases of varying crystallinity by blending either
liquid soybean oil or stick margarine oil with soft margarine oil. The textures of the spreads
made with these blends were visibly very different, increasing in softness in the same direc
tion as the oil itself. As a consequence of increased firmness, the stability of the resulting
spread was expected to increase with increasing crystallinity in the oil phase. A centrifugation
procedure was used to measure the stability of the emulsions (Fig. 3). The higher amount of
water and/or oil (especially water) released, the less stable the emulsion.
In summary, three main types of margarine can be distinguished according to the different
criteria for which they are formulated: table margarine (which includes the distinct categories
of tub and packet margarine), industrial or bakery (cake) margarine, and pastry margarine.
Apart from specific compositional criteria such as “all vegetable,” the main product character
istics are those of performance. This, in the case of table margarines, includes consistency at
the temperature of use, and for bakery and pastry margarines, creaming and plasticity, respec
tively [15]. The performance of the margarine in each case will relate closely to the solids
content (dilatation) of the fats used, as illustrated in Fig. 4.
F. Emulsifiers
The main function of emulsifiers in processed foods is to reduce the interfacial tension be
tween the phases of an emulsion, usually oil and water. In such two-phase systems, one phase
is dispersed as large droplets within the other. They are called oil-in-water (o/w) emulsions,
when the continuous phase is water (such as in milk or ice cream) and water-in-oil (w/o)
emulsions when the continuous phase is oil (as in butter and margarine).
The stability of two-liquid phase emulsions is kinetic stability, i.e., the system is not
thermodynamically stable [16]. Thermodynamically stable emulsions would spontaneously re
form following separation by, for instance, centrifugation, whereas experience shows that an
emulsion that has separated remains so unless mixed by some external action. In reality, the
two separated phases are in the most stable state to which all emulsions will tend. Thus, a
stable emulsion is one in which this inevitable trend has been retarded so that it is not notice
able during the normal life of the product, be it several years, even.
Margarine is a w/o emulsion in principle only because in reality it is a dispersion of water
droplets in a semisolid fat phase containing fat crystals and liquid oil [17]. Preparation of the
emulsion requires considerable energy to reduce the droplet size of the disperse phase, thereby
creating an increase in the surface area between the two immiscible phases.
The initial w/o emulsion is prepared in mixing tanks with vertical or horizontal stirrers to
ensure emulsification but without incorporating too much air. It is rather coarse in water
droplet size and is fairly unstable if not kept agitated.
In modem continuous production, the emulsion exists only for a brief period before passing
into the chilling unit where final emulsification and crystallization of the fat phase take place.
Thus, the emulsion need not be very stable against coalescence, as the water droplets become
314 Flack
(a)
(b)
Fig. 2 Fat crystals in (a) normal margarine with good consistency and (b) a margarine with large p
crystals causing a “sandy” texture (magnification 200X). (Courtesy of Danisco Ingredients, Denmark.)
316 Flack
4.5 6.3 8.6 10.4 13.0
Ctystallinity {Enthalpy - J o u le / gram)
50 Soft 75 Soft 75 50 Soft

50 liquid 25 25 50 Stick
Fig. 3 Effect of the degree of crystallinity on the emulsion stability of 40% fat spreads as measured
by ultracentrifugation. (□ ) Oil; (O) water. (From Ref. 10.)
fixed within the semisolid phase. However, the droplet size is important when considering
flavor release and microbiological spoilage. Thus, emulsifiers are used to lower the interfacial
tension between the oil and water phases, which will generally result in a smaller water
droplet size. Distribution of the water droplets in the range 2 -4 ^m will give better stability
against deterioration such as mold growth. However, it is still desirable for some of the water
droplets to be larger— 10-20 ^tm—to give a better flavor release in the mouth. For these
purposes, therefore, lipophilic emulsifiers such as monodiglycerides of long-chain fatty acids
(Ci^-Cjg) are used at levels of 0.1-0.3% , often in combination with 0.05-0.1% refined
soy lecithin.
Fig. 4 Typical solids contents of margarine fats. (From Ref. 15.)

Comparisons of the water droplet distribution in margarine emulsion and in finished marga
rine are shown in Fig. 5. The difference in water droplet size shows that the droplet size in
the emulsion is further reduced during the cooling and kneading processes in the tube chiller.
The water droplets in finished margarine are stabilized by adsorbed fat crystals as seen in
Fig. 6, which shows the microstructure of margarine by freeze-fracture transmission electron
microscopy. It is clearly seen that the water droplets are covered with fat crystals, oriented
flatly along their surface [18].
Margarine is frequently used for frying when it is especially important that it does not spatter.
Spattering is caused when the margarine melts in the frying pan, the emulsion breaks, and the
coalesced water droplets will, due to gravity, form a film of water covered with molten fat on
the base of the frying pan. When the temperature reaches the boiling point of water, the increase
in vapor pressure will cause the water phase to spatter—sometimes explosively. It is desirable,
therefore, that the margarine allow gradual evaporation of the water from the small water drop
lets and the formation of a fine, golden brown sediment that does not adhere to the frying pan.
Both formulation and processing play important roles in reducing the tendency to spattering. The
presence of salt and milk are desirable, as is a high pH, up to say 6, whereas sugars and starches
will increase the tendency to spatter. The selection and addition of the emulsifier is of consider
able importance in producing a margarine with good frying properties.
A combination of monodiglycerides and lecithin will have only a limited effect in reducing
spattering in low salt margarine, but there are other emulsifiers that, either alone or with
lecithin, will help prevent coalescence of water droplets during frying. In this respect, citric
acid esters of mono- and diglycerides (Citrem), polyglycerol esters, and thermally oxidized
soybean oil interacted with mono- and diglycerides used at 0.3-0.4% together with soy leci
thin can be very effective.
■■ » ,
Fig. 5 Water droplet size distribution in (a) margarine emulsion and (b) the finished margarine (magni
fication 200X). (Courtesy of Danisco Ingredients, Denmark.)
318 Flack
R g .6 Freeze-fracture electron micrograph showing the microstructure of margarine. Water droplets

(w) are covered with fat crystals. Bar: 1 ¡xm. (Courtesy of Dr. W. Buchheim, Keil.)
G. Processing
Modem processing methods involve weighing the major components by means of computer-
controlled load cells, which ensure accuracy, fast throughput, and constant composition. The
essential elements include
Aqueous phase preparation tanks

Fat blend preparation vats
Emulsion make-up tanks
Intermediate holding tanks
Cooling and kneading equipment
Worker units and resting tubes
The typical plant layout shown in Fig. 7 would be suitable for most types of both table and
industrial margarines, low fat spreads, and shortenings. The entire plant should be free of
copper and copper alloys to avoid the high risk of oxidation they would present.
The aqueous phase components may include salt, milk components such as whey powder
or skimmed milk powder, thickeners such as gelatin or sodium alginate, water-phase flavors
where appropriate, and preservatives such as potassium sórbate for low fat products. These
can be incorporated in the aqueous phase preparation or make-up tanks or via the emulsion
make-up tanks. The flnal oil blend will also contain the oil-soluble components such as emul
sifiers, colors, flavors, and vitamins.
The aqueous phase should be pasteurized before production of the emulsion, which requires
strong agitation, avoiding aeration, to ensure a good dispersion of the water droplets, as
discussed earlier.
Cooling and kneading can be carried out by either of two methods: (1) tube chiller or (2)
chilling drum-complector. In the former, as shown in Fig. 7, cooling and kneading take place
in a closed system and in a single process. In the latter, the cooling and kneading are carried
out separately—cooling on the chilling drum and kneading in the complector. The advantage
m
c
a
®
m
m
5"
m
©
m
(p) dp wrn^
--- m
m
Ti a.
STEAM AT
3 3 BAR m
3
— i -.p — a
f
1 T Tfit ^ I i m
^ Y ■ ; ® — ®
a>
w
fn 1 f ? ^ 3
m
-n
Q)
23 1 SET OF PIPEWORK VALVES I^Í3TRUMS^4T

22 1 ROTARY LOSE TRANSFER PUMP
21 1 OP FUl-lP
20 1 TRIPLE HOT water SET
.19 1 CONTROL PANEL
13 1 PLATE RECYCLE HEATER
17 1 REHELT TANK
16 1 resting TUSE
15 1 FirJAL WORKER UNIT
M 1 NTERMEDSATE CRYSTAUSER
13 1 K226F CHEMETATOR
12 1 EMULSION PRE-COOLER
11 1 HOLDING TUBE
10 1 M216S CHEMETATOR
9 1 TRIPLEX HIGH PRESSURE PUMP
6 1 CIP PUMP
7 1 BALANCE tank
6 1 ROTARY LOBE TRANSFER PUMP
5 1 CP PUMP
i 1 CP PUMP
3 2 EMULSION MAKE-UP TANK
2 1 ROTARY LOBE TRANSFER PLWP
1 2 WATERPHASE MAKE-UP TANK
w
Fig. 7 Flow diagram of margarine production line. (Courtesy of Crown Chemtech Limited, Reading, UK.) m
320 Flack
Table 4 Comparison of Processing Methods
Chilling drum/
Type Tube chiller complector
Block/Stick margarine XXX XXX

Soft table margarine XXX X
Low fat spread XXX X
Cake and creaming margarine XXX XXX
Puff pastry margarine XX XXX
Shortening XXX X
W/o emulsion with high water content XXX X
X = Poor; XX = acceptable; XXX = excellent.

Source: Ref. 19.
of the chilling drum-complector method is that it allows the product to rest between cooling
and kneading, which is important with formulas based on slower crystallizing fats such as
puff pastry margarine. By comparison, the tube chiller is relatively more compact considering
throughput, is easy to operate, and reduces the possibility of spoilage.
Most types of margarine can be made very satisfactorily using the tube chiller method,
although in some cases extra working units (pinning machines) may prove advantageous.
However, the chilling drum-complector was the preferred method for puff pastry margarine
and is perfectly satisfactory for normal table margarine and for cake margarine. The suitability
of both methods is summarized in Table 4. After cooling and kneading, industrial margarines
may be filled directly. Table margarines may require additional resting before going to the
filling unit.
In the case of cake margarine, the most important points are that the fat blend must be soft
and easy to incorporate into the batter and have good creaming properties. This therefore
suggests the use of lauric fats (coconut, palm kernel), which crystallize quickly and thereby
facilitate creaming. However, due to cost, lauric fats are mostly used only as components at
10- 20% of blends, which may include hardened and liquid oils such as palm, palm fractions,
and soybean to provide the particular requirements of the end product.
III. REDUCED AND LOW FAT SPREADS

As has been previously highlighted, low fat spreads have been the only growth area in a
gradually declining market. They can have fat contents of 10-79% and, except for very low
fat spreads, are w/o emulsions. However, with low fat products, it is necessary to strike a
balance between stability and mouthfeel, which can be affected by both the composition and
the method of processing.
Margarine is essentially a stable product, and the appropriate fat blend and the combination
of milk proteins, lecithins, and/or saturated mono- and diglycerides used in production ensure
both stability and the desired eating characteristics [20]. However, in reduced and low fat
spreads in which the disperse phase is in much greater proportion to the continuous phase,
problems can arise with stability, meltdown, and flavor release [9].
Although milk proteins are used in high fat products to improve mouthfeel and flavor, they
also act as o/w emulsion stabilizers, and thus their use in low fat products, together with the
considerable energy input needed, could result in phase inversion.
These problems can be overcome by a combination of formulation and processing. Im

portant factors in the formulation include the melting characteristics of the fat blend, the type
and level of emulsifier, and the addition of thickeners such as gelatin, sodium alginate, pectin,
and carrageenan to the aqueous phase. Low levels of whey protein may be used to improve
flavor release, with the further advantage that the pH of the aqueous phase can be lowered to
improve keeping properties because, unlike casein, whey proteins do not precipitate at low
pH. The speed of processing, i.e., the rate of throughput, and the emulsion temperature are
also important factors in the stability of the spread.
Examples of the effect of various types of emulsifiers on the stability of low fat spreads
with fat contents of 40% and 20% are shown in Tables 5 and 6 , respectively. The trial batches
in both cases were made on a three-tube Perfector pilot plant using a fat blend composed of
60 parts soy oil, 30 parts soy 41, and 10 parts coconut oil. Both examples indicate the
potential instability of these emulsions and the variation in the stabilizing effects of different
emulsifier blends even when used at low dosage levels.
More recently, the concept of “mixed gels” has been shown to be of interest for the stabili
zation of reduced and low fat spreads. This concept suggests that two or more gelling compo
nents that are not compatible normally may, under certain conditions, interact and thereby
increase their individual gelling capacities, thus reducing the amounts of the components in
combination compared to their use individually [21]. In particular, interactions between poly
saccharides (hydrocolloids) and proteins are of special interest.
Mixed gels are obtained when the two components, e.g., starch hydrolysates and gelatin,
of a single-phase system form continuous networks separately and can thus be a case of
homogeneous polymeric networks [22]. In this respect, one component of the mixed gel will
form a thermosetting gel while the other will be set by the action of ions. The behavior of
individual components and their use in combinations in mixed gel systems composed of gela
tin and maltodextrins have been fully investigated [23] and have also formed the basis for
many patents relating to the stabilization of water-in-oil emulsion spreads.
Table 5 Effect of Types and Blends of Emulsifiers on the Stability of a 40% Fat Spread
Water separation (%) after
Emulsifier 5 min 10 min 15 min 20 min
0 .6 % Distilled monoglyceride from 1.6 2 .6 9.5 15.8

vegetable oil, IV ~ 80.
0 .6 % Distilled monoglyceride from 0 1.0 2.6 6.3
vegetable oil, IV ~ 80
plus
0 .2 % Polyglycerol ester of fatty
acids, IV ~ 80.
0.4% Distilled monoglyceride from 0 0 0 0
vegetable oil, IV ~ 80
plus
0 .2 % Polyglycerol ester of interesterified
ricinoleic acid, IV ~ 85.
IV = Iodine value.
Formulation of spread:
Water phase (pH 4.5) 56.4% water, 0.5% whey powder, 1.5% gelatin, 1.5% salt, 0.1% potassium sórbate
Fat phase 39.2-39.4% fat blend, 0 .6 - 0 . 8 % emulsifier (as above), 4 ppm ^-carotene
322 Flack
Table 6 Effects of Type and Blend of Emulsifier on the Stability of a 20%

Fat Spread
Levels and types of emulsifiers Results
0 . 8 % distilled
monoglyceride from Emulsion split into separate phases
sunflower oil, IV ~ 105. in tube chiller.
0.5% distilled monoglyceride from Produced a fine, stable spread with
sunflower oil, IV ~ 105 good spreadability and
mouthfeel.
plus
0.5% polyglycerol ester of interester-
ified ricinoleic acid, IV ~ 85.
Formulation o f spread:
Water phase (pH 6.8) 69.9% water, 4.0% skim milk powder, 3.0% gelatin, 1.5% sodium
alginate, 1.5% salt, 0.1% potassium sórbate
Fat phase 19.0-19.2% fat blend, 0 .8-1.0% emulsifier (as above), 8 ppm ^-carotene
IV. DAIRY SPREADS AND BLENDS

The fat component of butter-containing spreads may be composed solely of milk fat as listed
in part A of Table 2 or may consist of blends of milk fat and other fats, primarily liquid
vegetable oils, as in the case of the products listed in part C of Table 2. In the case of blends,
the spread can be produced by normal buttermaking methods by adding liquid oil to the cream
prior to churning or through the normal margarine production method.
It is rather more difficult to produce low fat butter spread from butter oil due to the relative
hardness of the fat at low temperatures. Nonetheless, spread with a butterfat content of about
40% can be produced using 5% sodium caseinate and 2% sodium alginate in the water phase
and 0.5% distilled monoglyceride, IV approximately 55, in the fat phase. However, in this
case it is not possible to lower the pH of the aqueous phase without precipitating the caseinate,
which will reduce its emulsifying properties and thereby limit the keeping properties of the
spread. However, a satisfactory low fat butter spread can be produced from dairy cream,
using a distilled monoglyceride with high iodine value (80-105) in the cream and a thickener
such as sodium alginate. In this case, phase inversion from o/w to w/o can be achieved in the
tube chiller, using normal or slightly reduced cooling and operating at 40-50% of the usual
capacity. To obtain a satisfactory working effect, a high rotor speed in the tube chiller cooling
cylinder would be preferable [24].
Factors that may affect the efficiency of phase inversion in low fat emulsions are listed in
Table 7. Most of these factors concern the increase in rate of collision of oil droplets upon
which the rate of coalescence is dependent [3].
Table 7 Factors Promoting the Inversion of Oil-in-Water

Emulsions
1. Increased rotor speed

2. Controlled temperature (maintenance of optimum fat solids level)
3. Use of low HLB emulsifiers
4. Increased oil droplet size entering unit
Source: Ref. 9.
V. OIL-IN-WATER SPREADS
According to Moran [9], the advantages of o/w spreads over the more conventional w/o
spreads may be attributed to the following factors.
1. The product structure is not dependent on the type of fat used.
2. Any level of fat can be used, from 1 to 50% or more.
3. High levels of protein are possible.
4. Processing is easier and cheaper.
5. Flavor release is quicker on the palate.
Against this, however, one major drawback is that unless products are prepared at compara
tively low pH, then ultrahigh temperature processing and possibly an aseptic filling procedure
must be followed if shelf lives comparable to those of conventional spreads are required.
Oil-in-water spreads remain a relatively unexplored area of the spreadable fats market,
possibly due to the problem of microbiological deterioration. They may offer a potential for
future expansion in markets where adequate levels of preservatives are permitted.
VI. LIQUID MARGARINE

Liquid margarine is used primarily for frying and is available particularly in markets such as
the United States, Germany, and Sweden, where butter or margarine are normally used for
frying. In markets such as the United Kingdom that have traditionally used solid fats such as
lard or cooking fat or, more recently, liquid vegetable oils, there is less interest in using
liquid margarine.
A typical liquid margarine composition has a fat phase of about 82% based on soy or
sunflower oil, in which emulsifiers perform two functions. First, an emulsifier such as citric
acid ester of monoglyceride (Citrem) at, say, 0.4% plus 0.2% soy lecithin produces a stable
water dispersion with limited spattering during frying; second a selected mono- and triglycer
ide blend prevents oiling out on storage and thus gives a homogeneous product with low
viscosity. If the margarine is not intended for frying, the Citrem can be replaced with 0.2%
distilled monoglyceride. The water phase may contain proportions of skimmed milk powder
and salt as well as a preservative such as potassium sórbate. Flavoring may be added to
both phases.
The aqueous phase should be adjusted to pH 4.5-6.5 and pasteurized at 80°C. The fat
phase should be tempered to about 60°C and the emulsifiers melted into a small amount of
the liquid oil to a temperature of 65-70°C, then stirred into the main part of the fat phase.
The water phase is added to the fat phase under continuous agitation before cooling through
the tube chiller with an outlet temperature of approximately 5 -T C .
The emulsion should be rested for 2 h and then stirred vigorously for 15 min before tapping
off for packaging.
VII. BAKING FATS

Apart from margarines produced specifically for cake and pastry production, baking fats can
also take the form of shortenings or compounds.
Fat fulfills a number of roles when used in baked products. At its simplest, fat provides
lubricity in the mouth and allows products with a “short” texture to be produced. This “crum
bly” texture also associated with easy breakdown and dispersion in the mouth arises because
324 Flack
the fat interrupts the gluten network of the dough, thus preventing a hard cohesive texture
from developing. Hodge [25] listed the range of functions fat plays during the bakery process
and in the final product:
Better aeration
Emulsifying properties (greatly enhanced by emulsification)
Provision of an impervious layer (puff pastry)
Improvement in antistaling
Enhancement of flavors (particularly butter)
Texture modifier— shortening power and lubricity
Some particular applications are described below.
A. Shortenings in Cakes
The term “shortening” arises from the use of this type of fat to impart “shortness” in the
preparation of flour confectionery products such as shortbread, short pastry, and cakes [26].
Shortenings contain no moisture and consist of plastified fat or a blend of fats, often together
with emulsifiers suitable for the intended end products. The consistency and plasticity of
shortenings are therefore important to their effectiveness. For instance, high ratio shortenings
are used in cake formulas to enable a higher proportion of sugar to fat by ensuring a greater
dough strength arising from the presence of higher levels of emulsifiers, for instance 3-6%
distilled monoglycerides. In this case, it is essential that the fat blend be of a consistency that
enables good distribution of the fat and excellent creaming properties in the batter.
An important development in the use of plastified fats such as lard was the production of
pumpable shortenings to enable delivery to the factory in bulk. From bulk containers, where
the shortening is held at temperatures of 23-25°C, it can then be pumped to any part of the
production plant.
The first stage in cake making is the batter preparation. The objective is to obtain a fine
dispersion of air in the batter. This is achieved by mixing the ingredients sequentially, e.g.,
initial creaming of sugar and fat or fat and flour, or by the all-in mixing method. The first
approach has been practiced over many years whereas the all-in method is comparatively
recent and is now widely used both domestically and by large commercial bakeries.
In the initial batter phase the air cells are stabilized largely by fat crystals, although some
air will undoubtedly be in the viscous aqueous phase. At this stage the emulsion is complex,
being a mixture of domains, some water-continuous and others oil-continuous.
During the baking process the fat melts as the temperature rises and will be completely
molten by about 37°C depending on the fat. At this point the water-in-oil emulsion inverts
and the air remains trapped in the aqueous phase. With increasing temperature the starch is
increasingly hydrated and gelatinized, egg protein starts to coagulate, and the air cells form a
nucleus for further expansion by steam and carbon dioxide (e.g., from baking powder). The
baking process has been described by Shepherd and Yoell [27].
Butter and margarine are often preferred in cake making because they are able to convey
additional flavor characteristics not commonly found in shortenings (100% fat). The prime
function of fat in cake making is therefore to assist in the aeration of the batter and modify
the texture of the final product.
B. Short Pastry
As indicated earlier, fat is able to interrupt the gluten network of the flour, greatly improving the
texture of the final product. This is particularly true for short pastry where aeration is secondary.
The fat needs to be of a reasonably firm consistency such that it distributes throughout the dough
as a thin film, creating many “faults” in the dough structure. A variety of fats are used for short
pastry manufacture including lard, beef tallow olein, hardened fish oils, and hardened vegetable
oils. Margarines and butter are also used. They are generally less effective weight for weight
because they contain less fat but will generally impart an improved flavor.
C. Puff Pastry
These products rely on the barrier properties of fat, separating the individual dough layers
from one another. The expansion of gas or steam during the bakery process creates a layer
structure with a unique texture.
The key to puff pastry making is in having conditions such that the consistency of the
butter or margarine is sufficiently plastic to permit it to move coherently with the dough
layers. As these are rolled out, relayered, and rolled again to produce the multilayered dough,
the fat has to “stretch” with the dough layers, ensuring that they will remain apart during
baking. For this reason puff pastry margarine fats are normally higher melting (slip melting
point 42°C) than conventional margarine fats and have a higher solid fat content. Plasticity is
created in the margarine by postworking (work softening) of the margarine after chilling and
crystallizing the molten premix through tubular heat exchangers. Alternatively, chilled drum
followed by complector (mixing and work softening) are used [19].
In the case of butter, which has been widely used, additional functionality has been
achieved by using butter oil stearins.
D. Bakery Compounds
Bakery compounds or bread fats are water-in-oil emulsions with high water contents used in
yeast-raised white bread and producing a softening effect comparable to fat. Their composi
tion is relatively simple (see Table 8.).
Emulsifiers are used to reduce the interfacial tension between the fat and water phases, to sta
bilize the liquid emulsion, and to impart a fine, stable water dispersion to the bakery compound.
Production is carried out through emulsification of the two phases, held separately at 4 0 -
50°C, by adding the water phase slowly to the fat phase, preferably in a process involving a high
capacity circulation pump. It is essential that a water-in-oil emulsion be maintained throughout
the process. The emulsion is then cooled and kneaded via the tube chiller. Such compounds are
used at levels of 2-5% of the bread dough.
E. General
A number of more detailed reviews on the baking process and the function of fats and emulsi
fiers and lipids in baking have been published. For details on the use of butter and butter
fractions, the reader is referred to Kaylegian and Lindsay [28] and Mann [29].
Table 8 Composition of Bakery

Compound
Water Phase
Water 20-60%
Sucrose orglucose syrup 0-10%
Fat blend
Soft fat (MP 30-35 °C) 40-80%
Emulsifier 1.2%
326 Flack
Baking science and technology is described by Kamel and Stauffer [30] and Pyler [31].
Fats in baking systems are reviewed by La Guardia [32] and Podmore [33]. Lipid interactions
in breadmaking are discussed in detail by Carr et al. [34], Eliasson and Larsson [35], and
Barnes [36].
REFERENCES
E. Huber, Cloister dairies in the fourth pre-Christian milleniumn (in German), Deut. Molkerie-Z.
45: 2610 (1931).
2. L. M. Lamport, Modern Dairy Products, Chemical Pub. Company, New York, 1965, p. 282.
3. E. Frede and W. Buchheim, Buttermaking and the churning of blended fat emulsions, J. Soc.
Dairy Technol. 47(1): (1994). p. 17.
4. M. Anderson and B. E. Brooker, Dairy foams, in Advances in Food Emulsions and Foams (E.
Dickinson and G. Stainsby, Eds.), Elsevier, London, 1988, p. 221.
5. W. Buchheim, Electron microscopic investigation of the structure of whipped cream. Proceedings
19th Int. Dairy Congr., New Delhi, 1974.
C. Poot and E. Biemoth, Margarine and butter production, in The Lipid Handbook, 2nd ed.,
(F. D. Gunstone, J. L. Harwood, and F. B. Padley, Eds.), Chapman and Hall, London, 1994,
p. 293.
7. A. J. C. Andersen and P. N. Williams, Margarine, 2nd ed., Pergamon, Oxford, 1954, p. 2.
8. M. K. Schwitzer, Margarine and Other Food Fats, Leonard Hill, London, 1956, p. 60.
9. D. P. J. Moran, Reduced calorie spreads, Porim Technol. No. 15 Palm Oil Research Institute of
Malaysia, February 1993.
10. R. P. Borwanker and G. S. Buliga, Emulsion properties of margarines and low fat spread, Proc.
Food Emulsions and Foams—Theory and Practise, San Francisco, 1989, pp. 44-52.
11. J. Madsen, Product formulation and processing of margarine and yellow fat spreads, Proc. Marga
rine and Yellow Fat Seminar, Coventry, 1983.
12. J. Madsen and G. Als, Sandiness in table margarine and the influence of various blends of triglyc
erides and emulsifiers thereon, Proc. IXth Int. Soc. Fat Res. Congr., Rotterdam, 1968.
13. R. E. Timms, Phase behaviour of fats and their mixtures. Prog. Lipid Res. 23: 1-38 (1984).
14. N. Garti, E. Wellner, and S. Sarig, J. Am. Oil Chem. Soc. 59: (1981).
15. K. Berger, Palm oil products—why and how to use them. Food Technol., September 1986, p. 72.
16. S. E. Friberg, R. F. Goubron, and I. H. Kayali, Emulsion stability, in Food Emulsions (K.
Larrson and S. E. Friberg, Eds.), Marcel Dekker, New York, 1990.
17. N. Krog, T. Riisom, and K. Larsson, Applications in the food industry, in Encyclopedia of Emul
sion Technology, Vol. 2 (K. Larrson and S. E. Friburg, Eds.), Marcel Dekker, New York, 1988,
p. 321.
18. N. Krog, Interactions of emulsifiers with other components in foods, in Ingredient Interactions—
Effects on Food Quality (A. G. Gaonkar, Ed.), Marcel Dekker, New York, 1995, p. 380.
19. J. Madsen, Puff pastry margarine: a comparison of the chilling drum and tube chiller methods of
production, TP 101, Danisco Ingredients, Denmark.
20. E. A. Flack, The role of emulsifiers in reduced fat and fat-free foods, in Food Technology Interna
tional Europe (A. Turner, Ed.), Sterling, London, 1992, pp. 179-181.
21 . A. H. Clark, R. K. Richardson, S. B. Ross-Murphy, and J. M. Stubbs, Structural and mechanical
properties of agar/gelatin co-gels. Small deformation studies. Macro Molecules 16: 1367 (1983).
22 . V. B. Tolstoguzov, in Functional Properties of Food Macromolecules (J. R. Mitchell and D. A.
Ledward, Eds.), Elsevier, London, 1985, p. 385.
23. S. Kasapis, E. R. Morris, I. A. Norton, and R. T. Brown, Phase equilibria and gelation in gelatin/
maltodextrin systems. Parts I-IV, Carbohydrate Polym. 21: 243 (1993).
24. J. Madsen, Low calorie spread and melange production in Europe, Proc. World Conf. on Edible
Oils and Fats Processing, Maastricht, 1989.
25. D. G. Hodge, Fats in bakery products, Br. Nutr. Found. Nutr. Bull. 77(3): 153 (1986).
26. M. Gordon, Fats and fatty foods, in Food Industries Manual, 23rd ed. (M. D. Ranken and R. C.
Kill, Eds.), Blackie, London, 1993, p. 322.
27. I. S. Shepherd and R. W. Cake Emulsions Yoell, in Food Emulsions (S. Friberg, Ed.), Marcel
Dekker, New York, pp. 217-275. 1976.
28. K. E. Kaylegian and R. C. Lindsay, Handbook of Milk Fat Fractionation and Technology, AOCS
Press, Champaign, IL, 1995.
29. E. Mann, Modified butters and fat spreads. Dairy Ind. Int. 60: 21 (1995).
30. B. S. Kamel and C. E. Stauffer, Advances in Baking Technology, Blackie, London, 1993.
31. E. J. Pyler, Emulsions in Baking Science and Technology, Vol. I, Sosland,Chicago 1988.
32. M. K. La Guardia, High performance fat systems in baked products. Cereal Foods World 39:
147 (1994).
33. J. Podmore, Fats in bakery and kitchen products, in Fats in Food Products (D. P. J. Moran and
K. K. Rajah, Eds.), Blackie, London, 1995, p. 220.
34. N. O. Carr, N. W. R. Daniels, and P. J. Frazier, Lipid interactions in breadmaking. Rev. Sci.
Nutr. 31(3): 237 (1992).
35. A. C. Eliasson and K. Larsson, Intereactions between components, in Cereals in Breadmaking: A
Molecular Colloidal Approach (A. C. Eliasson and K. Larsson, Eds.), Marcel Dekker, New York,
1993, pp. 161-210.
36. P. J. Barnes, Lipids in Cereal Technology, Academic, London, 1983.
12__________
Ice Cream
H. Douglas Goff
University of Guelph, Guelph, Ontario, Canada
L INTRODUCTION
The importance of ice cream and related products as a food commodity is very significant,
with perhaps an annual production of 136 million hectoliters worldwide, increasing at 4-5%
per annum. Annual production figures from some of the major world producers are approxi
mately 54 million hL in the United States, 5.5 million hL in Germany, 5.2 million hL in the
United Kingdom, 4.0 million hL in Australia, 3.9 million hL in Canada, and 2.3 million hL
in Sweden. Annual per capita consumption figures are approximately 18 L in Australia, New
Zealand, and the United States; 14 L in Canada and Sweden; 9 L in the United Kingdom; 7
L in Germany; and 6 L in Norway and Finland [1]. Given also that, on average, 5-12% of
the weight of ice cream is fat [2], it becomes readily apparent that the ice cream industry is a
large consumer of fat, worldwide, approximately 800,000 tonnes per annum. Fat not only
provides flavor to ice cream and related products but is also vital in creating the smooth
eating texture that is both desired by and expected from the consumer.
The term “ice cream” is often taken to mean a family of whipped dairy products that are
manufactured by freezing and are consumed in the frozen state, including ice cream that
consists of either dairy or non-dairy fats; premium, higher fat versions; “light,” lower fat
versions; ice milk; sherbet; frozen yogurt; etc. The manufacturing processes for most of these
products are similar and involve the preparation of a liquid mix; concomitantly whipping and
freezing this mix dynamically under high shear to a soft, semifrozen slurry; incorporation of
flavoring ingredients to the partially frozen mix; packaging the product; and further quiescent
freezing (hardening) of the product in blast air. Many comprehensive reviews of ice cream
manufacture are available [2-7]. This chapter briefly summarizes the manufacture and compo
sition of ice cream and related products and focuses on the use and role of fats in these
products.
329
330 Goff
II. FORMULATIONS
Table 1 presents the range of compositional variables found in most ice cream mix formula
tions. The composition, minimum fat content, total solids content, and weight per volume of
frozen dairy products in most countries are standardized and regulated. Ice cream and ice
milk mix formulations specify the content of fat, milk solids-not-fat, sweeteners, stabilizers,
emulsifiers, and moisture that are desired. Usually one mix is used for the production of a
variety of flavors. Dairy and other ingredients used to supply these components are chosen on
the basis of availability, cost, legislation, and desired quality. Ingredients available for each
of the major components are listed in Table 2. It must be remembered, however, that in
addition to supplying the major component, most of the dairy ingredients also supply minor
amounts of the other components. Thus a detailed calculation is necessary to convert the
desired formula into a recipe, based on the ingredients chosen. Algebraic methods for the
Table 1 A Typical Compositional

Range for the Components Used in Ice
Cream Mix Formulations
Component Range (%)
Milkfat 10 -16
Milk solids-not-fat 9 -12
Sucrose 9 -12
Glucose syrup solids 4-6
Stabilizers/emulsifiers 0-0.5
Total solids 36-45
Water 55-64
Table 2 Components of an Ice Cream Mix and the Ingredients That

Supply Them
Component Ingredients ^
Milkfat Cream (msnf, water)

Butter (msnf, water)
Milk solids-not-fat Skim milk powder (water)
Condensed skim milk (water)
Condensed milk (water, fat)
Sweetened condensed skim milk (water, sugar)
Sweetened condensed milk (water, sugar, fat)
Whey powder (water)
Water Skim milk (msnf)
Milk (fat, msnf)
Water
Sweetener Sucrose
Glucose syrup solids
Liquid sugars (water)
Stabilizers/emulsifiers
msnf = milk solids-not-fat.

^Minor suppliers of the component listed in parentheses.
Ice Cream 331
solution of these problems are available [2,5], and computer programs to solve mix calcula
tion problems, based on the solution of simultaneous equations, are in widespread use in
many commercial operations.
III. INGREDIENTS
A. Fat
The fat component of the mix increases the richness of flavor in ice cream, produces a charac
teristic smooth texture by lubricating the palate, helps to give body, and aids in producing
desirable melting properties [2,4]. The fat content of a mix also aids in lubricating the freezer
barrel while the ice cream is being manufactured. Limitations on excessive use of fat in a mix
include cost, a hindered whipping ability, decreased consumption due to excessive richness,
and high caloric value. Fat contributes 9 kcal/g to the diet, regardless of its source. During
freezing of ice cream, the fat emulsion that exists in the mix will partially coalesce (destabi
lize) as a result of emulsifier action, air incorporation, ice crystallization, and high shear
forces of the blades [4,8,9]. This partial coalescence is necessary to set up the structure and
texture in ice cream, which is similar to the structure in whipped cream [10,11]. This process
is discussed in Section V.A. The fat content is an indicator of the perceived quality and/or
value of the ice cream. Ice cream must have a minimum fat content of 8-10% in most legal
jurisdictions. Premium ice creams may have fat contents as high as 14-18%. It has also
become desirable, however, to create light ice creams, < 5% fat, with similar perceived qual
ity. In addition to structure formation, fat contributes a considerable amount of flavor to ice
cream, which is difficult to reproduce in lowfat ice creams. Fat content must be altered by at
least 2-3% before any noticeable difference appears in the taste or texture [2].
1. D airy F at
Milk fat as a fat source for ice cream formulations is in widespread use in North America,
Australia, New Zealand, and parts of Europe. The triglycerides in milk fat have a wide melt
ing range, +40° to —40°C. The crystallization patterns of milk fat are also very complex, due
in part to the large variation in fatty acids and large numbers of different triglycerides present
[12-17]. Consequently, there is always a combination of liquid and crystalline fat at chilled
and subzero temperatures. Alteration of this solid/liquid ratio at freezer barrel temperatures,
through natural variation or fat fractionation, may affect the ice cream structure formed
[9,18]. The best source of butterfat in ice cream for high quality flavor is fresh sweet cream
from fresh sweet milk [2]. Other sources of butterfat include anhydrous milk fat, sweet (un
salted) butter, frozen cream, and condensed milk blends. Whey creams have also been used
but may lead to flavor or texture problems.
2. N on-D airy
Vegetable fats are used extensively as fat sources in ice cream in the United Kingdom, parts
of Europe, the Far East, and Latin America but only to a very limited extent in North
America. Five factors of great interest in selection of fat source are the crystal structure of
the fat, the rate at which the fat crystallizes during dynamic temperature conditions, the tem
perature-dependent melting profile of the fat, especially at chilled and freezer temperatures,
the content of high melting triglycerides (which can produce a waxy, greasy mouthfeel), and
the flavor and purity of the oil [4]. As will be discussed in Section V.A, it is important that the
fat droplet contain an intermediate liquid/solid fat ratio at the time of freezing. It is difficult to
332 Goff
Fig. 1 The variation of liquid fat content with temperature of fats suitable for use in ice cream.
(Courtesy of D. Finney, Unilever Research Laboratory, Colworth House, United Kingdom.)
quantify this ratio as it is dependent on a number of composition and manufacturing factors;

however, one-half to two-thirds crystalline fat at 4-5°C is a good working rule [4]. Crystalli
zation of fat occurs in three steps: undercooling to induce nucléation, heterogeneous or homo
geneous nucléation (or both), and crystal propagation. In bulk fat, nucléation is predominantly
heterogeneous, with crystals themselves acting as nucleating agents for further crystallization,
and undercooling is usually minimal. However, in an emulsion each droplet must crystallize
independently of the next. For heterogeneous nucléation to predominate, there must be a
nucleating agent available in every droplet, which is often not the case. Thus in emulsions,
homogeneous nucléation and extensive undercooling may be common [4,13].
Figure 1 illustrates the melting profiles for a range of fats that are used in ice cream
formulations. Blends of oils are often used in ice cream manufacture, selected to take into
account physical characteristics, flavor, availability, stability during storage, and cost. Hydro
genation is often necessary to achieve the appropriate melting characteristics. Palmkemel oil,
coconut oil, palm oil, and fractions thereof, plus their hydrogenated counterparts, can all be
used. Tong et al. [19] substituted up to 40% milk fat in ice cream with safflower oil, a highly
unsaturated oil, in an attempt to lower the saturated fatty acid content of the final product.
They reported that increasing the concentration of safflower oil decreased overrun (the incor
poration of air into the ice cream) but had little effect on the extent of fat destabilization.
However, Berger [4] reported that ice cream made entirely with liquid vegetable oil as the fat
source was unacceptable. The oil spread at the air surface, leading to collapse and undesirable
texture. A partly crystalline fat globular structure is thus necessary for structure development.
3. F at Substitutes
There has been great interest in the marketplace for the development of lower fat alternatives
to traditional ice cream products. As a result, a large amount of product development time
has been used in searching for a combination of ingredients that will replace the textural and
flavor characteristics of fat in ice cream. These often involve the use of fat substitutes. Such
products may be formulated with starch or other polysaccharides, proteins, or lipids, but their
main requirement is to provide fewer calories to the product than traditional fat sources in the
diet. A great deal of technical literature exists on the various properties of the products being
Ice Cream 333
marketed by a number of commercial firms. Schmidt et al. [20] studied the rheological, freez
ing, and melting properties of ice milks manufactured with protein-based or maltodextrin-
based fat alternatives. They concluded that the carbohydrate-based alternatives resulted in
greater effects on mix rheology while the protein-based alternatives were more similar to the
control, due in part to the functional contributions of proteins to food systems, especially in
the area of emulsification and air incorporation. Ice cream products are very complex systems,
both in structure and in flavor. In creating products that are meant to deliver to the consumer
the same attributes but with less fat or calories, it is imperative that the structural element be
considered to the same extent as flavor in order to deliver high quality products and develop
market share for these products.
B. Nonfat Milk Ingredients

The milk solids-not-fat (msnf) or serum solids improve the texture of ice cream, aid in giving
body and chew resistance to the finished product, are capable of allowing a higher overrun
without the characteristic snowy or flaky textures associated with high overruns, and may be
a cheap source of total solids [21]. The msnf contain the lactose, caseins, whey proteins, miner
als (ash), vitamins, acids, enzymes, and gases of the milk or milk products from which they
were derived. The content of serum solids used in a mix can vary from 10% to 14% or more.
Limitations on their use include off-flavors from certain products and an excess of lactose,
which may lead to numerous problems as discussed below [22]. Traditionally, the best sources
of serum solids for high quality products are fresh concentrated skimmed milk or spray-dried
low-heat skim milk powder. Others include sweetened condensed whole or skimmed milk,
superheated condensed skimmed milk, buttermilk powder or condensed buttermilk, and dried
or condensed whey [21]. Buttermilk powder is a by-product of the butter industry. It contains
a higher concentration of fat globule membrane phospholipids than skim milk. Thus it can be
used for its emulsifying properties to reduce the need for emulsifiers [2]. Whey is a by
product of the cheese industry and contains fat, lactose, whey proteins, and water but no
casein. Whereas skim milk powder contains 54.5% lactose and 36% protein, whey powder
contains 72-73% lactose and only about 10-12% protein [12]. Thus it can promote some of
the problems associated with high lactose as described below as well as flavor problems
associated with the high salt content. However, there are an increasing number of whey prod
ucts available that have higher protein and lower lactose and mineral contents. Many of these
can provide much higher quality than the traditional whey ingredients [23]. It is also possible
to produce concentrated protein products from the casein portion of milk proteins, which have
potential for use in ice cream [23,24]. These caseinates are formulated to give different func
tional properties, such as degrees of emulsification, foaming, gelation, or water holding [25].
Proteins contribute much to the development of structure in ice cream, including emulsifi
cation, whipping, and water-holding capacity [25-27]. Emulsification properties of proteins
in the mix arise from their adsorption to fat globules at the time of homogenization [23].
Whipping properties of proteins in ice cream contribute to the formation of the initial air
bubbles in the mix [27,28]. The water-holding capacity of proteins leads to enhanced viscosity
in the mix [25], which imparts a beneficial body to the ice cream, increases the meltdown
time of ice cream, and contributes to reduced iciness [23]. The protein content of a mix is
usually about 4%. When replacing traditional msnf sources in ice cream with newer sources
with the ratio of proteins altered from the native state, it is important to ensure that these
functional roles are being met.
Lactose is a disaccharide of glucose and galactose. Lactose does not contribute much to
sweetness; it is only one-fifth to one-sixth as sweet as sucrose [ 12 ]. Lactose content of ice
334 Goff
cream mix is about 6% if no whey powder has been used in the formulation. Excessive levels
of lactose in ice cream mix leads to reduced freezing point, causing a softening of the ice
cream and the potential for development of iciness and a greater potential for lactose crystalli
zation [29]. Lactose is relatively insoluble, and lactose crystals, particularly the a form, which
takes on a characteristic tomahawk shape [22], can produce the defect known as sandiness
when they are allowed to grow to a large size. The lactose solubility in water at room tempera
ture is about 11% [12]. During freezing, this concentration is exceeded as a result of freeze
concentration (water removal in the form of ice). When 7 5 % o f the water is frozen in a mix
consisting originally of 1 1 % m s n f (6% lactose), the lactose content in the unfrozen phase
corresponds to 39.3%. Probably much of the lactose in ice cream exists in a supersaturated,
amorphous (noncrystalline) state, however, due to extreme viscosity [30]. Stabilizers help to
hold lactose in a supersaturated state because of viscosity enhancement.
C. Sweeteners
Sweet ice cream is usually desired by the consumer. As a result, sweetening agents are added
to ice cream mix at a rate of usually 12-17% by weight. Sweeteners improve the texture and
palatability of the ice cream, enhance flavors, and are usually the cheapest source of total
solids [2]. Their ability to lower the freezing point of a solution imparts a measure of control
over the temperature-hardness relationship. In determining the proper blend of sweeteners for
an ice cream mix, the total solids required from the sweeteners, the sweetness factor of each
sugar, and the combined freezing point depression of all sugars in solution must be calculated
to achieve the proper solids content, the appropriate sweetness level, and a satisfactory degree
of hardness [3,4,31]. The most common sweetening agent used is sucrose, alone or in combi
nation with other sugars. Sucrose and lactose are most commonly present in ice cream in the
supersaturated or glassy state, with few crystals being present [4,30]. It has become common
practice in the industry to substitute sweeteners derived from starch hydrolysate syrup (glucose
syrup) for all or a portion of the sucrose [2,31]. A typical sweetener blend for an ice cream
mix usually includes 10-12% sucrose and 4-5% glucose syrup solids [2,31].
The use of primarily com, but also wheat, potato, or tapioca starch hydrolysis products in
ice cream is generally perceived to provide greater smoothness by contributing to a firm e r and
more chewy body, to provide better meltdown characteristics, to bring out and accentuate
fm it flavors, to reduce heat shock potential, which improves the shelf life of the finished
product, and to provide an economical source of solids [31,32]. The higher molecular weight
saccharides (dextrins) are effective stabilizers and provide maximum prevention against the
formation of coarse ice crystals, which is reflected in improved meltdown and heat shock
resistance. They also improve cohesive and adhesive textural properties. Smaller sugars pro
vide smoothness, sweetness, and flavor enhancement. Dextrose offers sweetness synergism
with sucrose but has the greatest freezing point depression of all the sugars, being a monosac
charide. The ratio of higher to lower molecular weight fractions can be estimated from the
dextrose equivalent (DE) of the symp. Ice cream manufacturers usually use a 28-42DE sy ra p ,
either liquid or dry [31]. When dextrose is converted to fmctose, the resultant sy m p is much
sweeter than sucrose, although it has half the molecular weight and thus contributes more to
freezing point depression than sucrose [32].
D. Stabilizers
Ice cream stabilizers are a group of ingredients (usually polysaccharides) commonly used in
ice cream formulations. The primary purposes for using stabilizers in ice cream are to produce
smoothness in body and texture; retard or reduce ice and lactose crystal growth during storage
Ice Cream 335
(or mask the effects of crystal growth), especially during periods of temperature fluctuation,
known as heat shock [33]; and to provide uniformity to the product and resistance to melting
[2,3]. They also increase mix viscosity, stabilize the mix to prevent separation of a clear
serum on meltdown, known as wheying off (e.g., carrageenan), aid in suspension of flavoring
particles, produce a stable foam with easy cut-off and stiffness at the barrel freezer for packag
ing, slow down moisture migration from the product to the package or the air, and help to
prevent shrinkage of the product volume during storage [34]. Stabilizers must also have a
clean, neutral flavor, not bind to other ice cream flavors, contribute to acceptable meltdown
of the ice cream, and provide desirable texture and mouthfeel upon consumption [34]. Limita
tions on their use include production of undesirable melting characteristics, excessive mix
viscosity, and contribution to a heavy, soggy body. Although stabilizers increase mix viscos
ity, they have little or no impact on freezing point depression.
Gelatin, a protein of animal origin, was once used almost exclusively in the ice cream
industry as a stabilizer but has gradually been replaced with polysaccharides of plant origin
due to their greater effectiveness and lower cost [2,5]. Stabilizers currently in use include ( 1 )
carboxymethyl cellulose, derived from the bulky components or soluble fiber of plant mate
rial; (2 ) locust bean gum (carob bean gum), which is derived from the beans of the tree
Ceratonia siliqua grown mostly in the Mediterranean; (3) guar gum, from the guar bush,
Cyamoposis tetragonolba, a member of the legume family grown in India and Pakistan for
centuries and now grown to a limited extent in Texas; (4) carrageenan, an extract of Chondus
crispis, Irish moss, a red algae, originally harvested from the coast of Ireland, near the village
of Carragheen; and (5) sodium alginate, an extract of another seaweed, kelp, a brown algae.
Each stabilizer has its own characteristics, and often two or more of these stabilizers are used
in combination to lend synergistic properties to each other and improve their overall effective
ness. Guar, for example, is more soluble than locust bean gum at cold temperatures; thus it
finds more application in HTST pasteurization systems. Carrageenan is a secondary colloid
used to prevent the wheying off of mix, which is usually promoted by one of the other
stabilizers [2-5].
The mechanism of action of stabilizers in enhancing frozen stability is related primarily to
their effect on the ice phase. Stabilized ice creams have been observed with electron micros
copy techniques to have smaller ice crystals than unstabilized ice cream both before and after
storage at fluctuating temperatures [33,35,36]. The control of iciness by stabilizers has been
examined by several researchers [36-44]. It was originally thought that stabilizers bound
water, rendering it unfreezable and thereby exerting their stabilizing mechanism by reducing
the quantity of ice present [2], but recent research has failed to demonstrate this effect [39-
44]. The stabilizing effect of the polysaccharides presently appears to result from an alteration
of the diffusion properties of water and solutes within the unfrozen phase following freeze-
concentration [3,36,45-47].
E. Emulsifiers
Emulsifiers have been used in ice cream mix manufacture for many years [48]. They are
sometimes integrated with the stabilizers in proprietary blends, but their function and action
are very different from those of the stabilizers. They are used for improvement of the whip
ping quality of the mix; production of a drier ice cream to facilitate molding, fancy extrusion,
and sandwich manufacture; smoother body and texture in the finished product; superior draw
ing qualities at the freezer to produce a product with good stand-up properties and melt resis
tance; and more exact control of the product during freezing and packaging operations [2,49].
Their mechanism of action can be summarized as follows: They lower the fat/water interfacial
336 Goff
tension in the mix, resulting in protein displacement from the fat globule surface, which in
turn reduces the stability of the fat globule to the partial coalescence that occurs during the
whipping and freezing process, leading to the formation of a fat structure in the frozen product
that contributes greatly to texture and meltdown properties [8]. The extent of protein displace
ment from the membrane, and hence the extent of dryness achieved, is a function of the
emulsifier concentration [4,50]. Their role in structure formation is described in detail in
Section V.A.
Egg yolk was formerly commonly used as an ice cream emulsifier. Emulsifiers used in ice
cream manufacture today are of two main types: the mono- and diglycerides and the sorbitan
esters. Mono- and diglycerides are derived from the partial hydrolysis of fats of animal or
vegetable (important for certain religious or ethnic considerations) origin. The sorbitan esters
are similar to the monoglycerides in that the sorbitan esters have a fatty acid molecule such
as stearate or oleate attached to a sorbitol molecule whereas the monoglycerides have a fatty
acid molecule attached to a glycerol molecule. To make the sorbitan esters water-soluble,
polyoxyethylene groups are attached to the glucose molecule. Thus Polysorbate 80, the most
common of these sorbitan esters, is actually polyoxyethylene sorbitan monooleate [48]. Poly
sorbate 80 is a very active drying agent in ice cream [8] and is used in many commercial
stabilizer/emulsifier blends.
IV. PROCESSING
A. Mix Manufacture
1. Blending and Pasteurization
Ice cream processing operations can be divided into two distinct stages: mix manufacture and
freezing operations. Ice cream mix manufacture consists of the following unit operations:
combination and blending of ingredients, batch or continuous pasteurization, homogenization,
and mix aging [6]. Ingredients are usually preblended prior to pasteurization, regardless of
the type of pasteurization system used. Blending of ingredients is relatively simple if all
ingredients are in the liquid form, as automated metering pumps or tanks on load cells can be
used. When dry ingredients are used, powders are added either through a pumping system
under high velocity or through a liquéfier, a large centrifugal pump with rotating knife blades
that chops all ingredients as they are mixed with the liquid [6].
Pasteurization is the biological control point in the system, designed for the destruction of
pathogenic bacteria. In addition, it serves a useful role in reducing the total bacterial load and
in solubilization of some of the components (proteins and stabilizers). Both batch and continu
ous (high temperature, short time, or HTST) systems are in common use [6]. In a batch
pasteurization system, proper ingredient amounts are blended in large jacketed vats equipped
with some means of heating, usually saturated steam or hot water. The product is then heated
in the vat to at least 69°C (155°F) and held for 30 min to satisfy legal requirements for
pasteurization, necessary for the destruction of pathogenic bacteria [51]. Various time-tem
perature combinations can be used, depending on the legal jurisdiction. The heat treatment
must be severe enough to ensure destruction of pathogens and to reduce the bacterial count to
a particular maximum (e.g., 100,000 per gram) depending on the legal jurisdiction [51].
Following pasteurization, the mix is homogenized using high pressures and then is passed
across some type of heat exchanger (plate heat exchanger or double- or triple-tube heat ex
changer) for the purpose of cooling the mix to refrigerated temperatures (4°C).
Continuous pasteurization is usually performed in an HTST heat exchanger following the
blending of ingredients in a large insulated feed tank. Some preheating, to 30-40°C, may be
Ice Cream 337
necessary for solubilization of the components. The HTST system is equipped with a heating
section, a cooling section, and a regeneration section (Fig. 2). Mix first enters the raw regen
eration section, where cold or preheated mix is heated to as warm as possible on one side of
a plate heat exchanger while the pasteurized hot mix is cooled to as low as possible running
countercurrent on the opposite sides of the plates. Following the raw regeneration section,
mix may enter the homogenizer, depending on the particular configuration. The location of
the homogenizer at this point may be preferred, as it may act as the timing pump, it makes
use of the heating effect of homogenization ( 2 -3 °C ) , and it ensures that any potential contami
nation is maintained on the raw side. Mix then flows to the heating section, where a minimum
temperature of 80°C is obtained [51]. Mix is held at this temperature for at least 25 s [51] by
flowing either through a series of holding tubes or through an additional set of plates in the
HTST unit. Holding times much longer than the minimum can be accomplished with longer
holding tubes. Holding times of 2 -3 min may produce superior mixes that retain many of the
advantages of batch pasteurization but may also leave the mix prone to cooked flavors [4,5].
After leaving the holding tube, mix flows back through the pasteurized side of the regenera
tion section and enters the cooling plates, where a chilled brine or glycol solution or chilled
water brings the mix down to around 4°C.
2. H om ogenization
Homogenization is responsible for the formation of the fat emulsion by forcing the hot mix
through a small orifice under pressures of 15.5-18.9 MPa (2000-3000 psig), depending on
the mix composition. Table 3 show the effects of homogenization on the fat emulsion [12,52],
The actual mechanism of fat disruption within the homogenizer is thought to result from
turbulence, cavitation, and velocity gradients (energy density) within the valve body [53].
VACUUM
I BREAKER
PASTEURIZED I
REGENERATOR COOLER I
CONSTANT LEVEL TANK

RAW PRODUCT
Fig. 2 A schematic illustration of the equipment used for HTST pasteurization of milk, frozen dairy
dessert mixes, and other fluid dairy products. (From Ref. 51.)
338 Goff
Table 3 Effect of Homogenization of Milk on the Size Distribution of Fat Globules
No homogenization 15 MPa (2500 psig)
Average fat globule diameter (/xm) 3.3 0.4

Maximum diameter (/xm) 10.0 2.0
Surface area (m^/mL at 3.5% fat) 0.08 0.75
Number of globules (/xm“^) 0.015 12
The 8- 10-fold increase in the surface area of the fat globules is responsible in part for the
formation of the fat globule membrane, composed of adsorbing materials attempting to lower
the interfacial free energy of the fat globules [54,55]. With single-stage homogenizers, fat
globules tend to cluster as bare fat surfaces come together or adsorbed molecules are shared.
Therefore, a second homogenizing valve is frequently placed immediately after the first with
applied backpressures of 3.4 MPa (500 psig) [6], allowing more time for surface adsorption
to occur. The net effects of homogenization are the production of smaller, more uniform fat
droplet size resulting in a greater stability of fat droplets during aging, better whipping ability,
and a smoother, more uniform final product with a greater apparent richness [2]. Homogeniza
tion also decreases the danger of churning the fat in the freezer and makes it possible to use
products that could not otherwise be used, such as butter and frozen cream.
3. A ging
An aging time of 4 h or greater at 2-4°C is recommended following mix processing prior to
freezing. This allows for hydration of milk proteins and stabilizers (some viscosity increase
occurs during the aging period), crystallization of the fat globules, and a membrane re
arrangement, to produce a smoother texture and better quality product [56]. Non-aged mix
exhibits a loose stand-up, is very wet at extrusion, and exhibits variable whipping abilities
[2,56]. The appropriate solid/liquid fat ratio must be attained at this stage, a function of
temperature and the triglyceride composition of the fat used, as a partially crystalline emulsion
is needed for partial coalescence in the whipping and freezing step [57], as discussed in
Section V.A. Emulsifiers generally displace milk proteins from the fat surface during the
aging period [8,23,58], and this is also discussed in detail in Section V.A. The whipping
qualities of the mix are usually improved with aging. Aging is performed in insulated or
refrigerated storage tanks, silos, etc., or in single-walled tanks in chilled rooms, where valves
and pipework can also be kept cold. Mix temperature should be maintained as low as possible
without freezing.
B. Ice Cream Freezing

Ice cream freezing also consists of two distinct stages: (1) passing mix through a swept-
surface heat exchanger under high shear conditions to promote extensive ice crystal nucléation
and air incorporation and (2) freezing the packaged ice cream under conditions that promote
rapid freezing and small ice crystals [3,6]. The freezing and whipping process is one of the
most important unit operations for the development of quality, palatability, and yield of fin
ished product, due to the incorporation of air creating the foam, the formation of the ice
phase, and the partial destabilization of the fat emulsion. The objective of ice cream manufac
turers is to produce ice crystals that are below, or at least not significantly above, the threshold
of sensory detection at the time of consumption. This threshold has been suggested by Ar-
Ice Cream 339
buckle [2] to be between 40 and 50 fim. Consequently the freezing steps of the manufacturing
process and the temperature profile throughout the distribution system are critical factors in
meeting this objective. Flavoring and coloring can be added as desired to the mix prior to
passing it through the barrel freezer, and particulate flavoring ingredients such as nuts, fruits,
candy pieces, or ripple sauces can be added to the semifrozen product at the exit from the
barrel freezer prior to packaging and hardening.
1. D ynam ic Freezing and Aeration

Continuous freezers dominate the ice cream industry [6]. In this type of process, mix is drawn
from the flavoring tank into a scraped surface heat exchanger, which is jacketed with a liquid,
boiling refrigerant [usually ammonia; a chlorinated fluorinated hydrocarbon (CFC) such as R-
12, R-22, or R-502; or one of the newly developed CFC substitutes]. Incorporation of air into
ice cream, termed the overrun, is necessary to produce desirable body and texture. Overrun
is calculated as [2]
volume of ice cream produced —volume of mix used

% Overrun = X 100%
volume of mix used
As a guide, maximum overrun should be 2.5-3 times the total solids content to avoid possible
defects in the finished ice cream [2].
There are three types of air incorporation systems used in continuous freezers. In older
systems, the pump configuration resulted in a vacuum either at the pump itself or on the mix
line entering the pump. Air was then incorporated through a spring-loaded, controllable needle
valve. Newer types of freezers use compressed air, which is injected into the mix. This type
of air handling system allows for ultrafiltration (0.65 /xm micropore filter) of the air prior to
its entering the mix [6]. Additionally, recent experimentation with preaeration systems, in
which the mix is whipped independently prior to entering the barrel freezer, has demonstrated
that much smaller air bubble sizes and improved body and texture can be attained [59]. Fol
lowing aeration, the water in the mix is partially frozen as the m ix-air combination passes
through the barrel of the heat exchanger. Rotating knife blades and dashers keep the product
agitated and prevent freezing on the side of the barrel. Residence time for mix through the
annulus of the freezer varies from 0.4 to 2 min, although some product may remain for much
longer times, especially with open-cage dashers; freezing rates can vary from 5 to 2T C per
min, and draw temperatures of —6°C can easily be achieved [4].
Batch freezing processes differ slightly from the continuous systems just described. The
barrel of a batch swept-surface heat exchanger is jacketed with refrigerant and contains a set
of dashers and scraper blades inside the barrel. It is filled to about one-half volume with the
liquid mix. Barrel volumes usually range from 2 to 12 L. The freezing unit and agitators are
then activated, and the product remains in the barrel for sufficient time to achieve the desired
degree of overrun and stiffness. Whipping increases with time and cannot exceed the amount
that will fill the barrel with product (i.e., 100% overrun when starting half-full). Batch freez
ers are used in smaller operations where it is desirable to run individual flavored mixes on a
small scale or to retain an element of the “homemade”-style manufacturing process. They are
also operated in a semicontinuous mode for the production of soft-serve desserts. A hopper
containing the mix feeds the barrel as product is removed.
The formation of the ice phase during this stage is responsible in part for the structure and
texture of the final product [3]. The crystallization of water to ice involves two major steps:
nucléation and crystal growth [60-62]. Rapid heat removal, which results from low tempera
tures (large AT) in the freezing medium, and rapid agitation, which are both present in the
340 Goff
barrel freezer, promote extensive nucléation and create numerous tiny ice crystals. Further
temperature reduction during hardening accounts for continued growth of the preformed crys
tals [3]. As water freezes out of solution in a relatively pure form, the formation of ice acts
to concentrate the dissolved sugars, lactose, milk proteins, salts, and hydrocolloids in an ever-
decreasing amount of water. Water and its dissolved components are referred to as the serum
or matrix of the mix [2]. Because the freezing point of the serum is a function of the concen
tration of dissolved solids, the formation of more ice concentrates the serum and results in an
ever-decreasing freezing temperature for the remaining serum. Thus at temperatures several
degrees below the initial freezing temperature, there is always an unfrozen phase present.
Bradley [63,64] presented a method of calculating initial freezing points and equilibrium
freezing curves for sweetened dairy mixes based on their composition. An equilibrium freez
ing curve for an ice cream mix, which plots the amount of water frozen as a function of
temperature, is shown in Fig. 3. This represents the maximum amount of ice that can form at
any given temperature, based on freezing point considerations. Ice cream hardness is a func
tion of temperature due to its effect on this conversion from unfrozen water to ice and further
concentration of the serum phase surrounding the ice crystals, which helps to give ice cream
its ability to be scooped and chewed at freezer temperatures.
2. Static H ardening
Ice cream following dynamic freezing, ingredient addition, and packaging is immediately
transferred to a hardening chamber (—30°C or colder, either forced convection or plate-type
conduction freezers), where the majority of the remaining water freezes [6]. As mentioned
earlier, rapid initial freezing is essential to set up as many crystal nuclei as possible so that
during the maturation or growth stage their size stays small. In the same context, rapid harden
ing is also necessary to keep ice crystals small. When hardening is slow, there is too great an
opportunity for water still remaining in the ice cream to migrate to crystal centers already
formed, resulting in large ice crystals [3]. Many factors need to be considered during the
hardening process. The main factors affecting heat transfer are the temperature difference
between the product and the freezing medium, the area of product being exposed to the
freezing medium, and the heat transfer coefficient for the particular operation [6]. Following
Fig. 3. An equilibrium freezing curve for a typical ice cream mix as calculated by the method of
Bradley [63,64] illustrating the relationship between temperature and unfrozen water as determined by
freezing point depression characteristics. This relationship illustrates the maximum amount of ice that
may form at any temperature, but due to kinetic considerations the amount of ice may be considerably
less as a result of freezing rate considerations.
Ice Cream 341
rapid hardening, ice cream storage should occur at low, constant temperatures, usually
-2 5 T .
V. ROLE OF FAT IN PHYSICAL STRUCTURE AND FLAVOR

A. Structure Formation
The development of structure and texture in ice cream is sequential, basically following the
manufacturing steps, and results from the interactions of several components. Ice, air, and fat
each occupy distinct but interrelated phases (Fig. 4), all surrounded by a temperature-depen
dent unfrozen phase [30,65]. Ice crystals and air bubbles are usually in the range of 20-50
firn and 50-100 /im, respectively [4].
To properly describe the role of fat, it is necessary to begin with the formation of the
emulsion at the time of homogenization and the role of the ingredients present at that time,
with particular reference to the fat, proteins, and emulsifiers. Following homogenization, the
emulsion is further affected by changes occurring during the aging step, that is, crystallization
of the fat and rearrangement of the fat globule membrane to the lowest free energy state. This
emulsion then undergoes both whipping and ice crystallization during the dynamic freezing
process, which contributes to the development of the three main structural components of the
frozen product: a discontinuous foam, a network of partially coalesced fat surrounding the air
bubbles, and ice crystals.
Homogenization begins the process of fat structure formation. It has been shown that a
finely divided fat emulsion with little fat globule clustering after homogenization is essential
for the manufacture of a smooth, dry ice cream [66,67]. After preheating or pasteurization,
the mix is at a temperature sufficient to have melted all the fat and emulsifier present, and the
fat passes through one or two homogenizing valves. Immediately following homogenization,
the newly formed fat globule is practically devoid of any membranous material and readily
adsorbs amphiphilic molecules from solution [68-70]. The immediate environment supplies
the surfactant molecules, which include caseins, undenatured whey proteins, phospholipids,
lipoprotein molecules, components of the original milk fat globule membrane, and any added
..... -X .
Fig. 4 The structure of frozen, aerated dairy dessert products showing ice crystals (C), air bubbles
(A), fat globules (F), and the freeze-concentrated unfrozen serum phase (S), as evidenced from cryo-
scanning electron microscopy. (From Ref. 30.) (a) Bar = 30 /xm; (b) bar= 7.5/xm.
342 Goff
FAT
Fig» 5 A schematic representation of the adsorption of casein and whey proteins to the fat globule
following homogenization. Because of the differences in conformational structure, the resulting surface
excess is much lower with the whey proteins than with the caseins.
chemical surfactants [55,68]. These all compete for space at the fat surface, shown schemati
cally in Fig. 5 [55,70].
By selecting and/or limiting the adsorbing material present at the time of homogenization,
it may be possible to predetermine the adsorbing substances and thus create a membrane with
more functional attributes, using natural proteins rather than relying on the chemical surfac
tants [23,71]. The membrane formed during homogenization continues to develop during the
aging step, and rearrangement occurs until the lowest possible energy state is reached [58,68].
The transit time through a homogenization valve is on the order of 10~^-10“ ^ s, and veloci
ties on the order of 12,000 cm/s are reached [53]. Protein adsorption or unfolding at the
interface may take minutes or even hours to be complete [52]. It is clear, therefore, that the
immediate membrane formed upon homogenization is a function of the microenvironment at
the time of its creation and that the recombined membrane of the fat globule in the aged mix
is not fully developed until well into the aging process [68].
Emulsifiers are not needed in an ice cream mix to stabilize the fat emulsion (Table 4), due
to an excess of protein and other amphiphilic molecules in solution [49,72]. If a mix is
homogenized without any emulsifier, both the whey proteins and the caseins will form this
new fat globule membrane, with the caseins contributing much more to the bulk of the ad
sorbed protein (Fig. 6a). However, if surfactants such as the added emulsifiers are present.
Table 4 Comparison of the Creaming Rates of Ice Cream Mix Emulsions Either
With No Surfactant or With 0.08% Polysorbate 80 When Both Have Sat Quiescently for
10 Days^
No emulsifier Polysorbate 80
Butterfat (%) in mix 10.00 10.00

Butterfat (%), top of 100 mL cylinder, 11.2 5 ±0.43 11.2 8 ±0.43
10 days
Butterfat (%), bottom of 100 mL cylinder, 9.47 ± 0 .1 1 9.46 ± 0 .1 1
10 days
Butterfat determined by Mojonnier ether extraction.
Source: Ref. 75.

Ice Cream 343
• . i C
“ ^ ■ 1
►: M
►; ■ 1^
■lllii
M
m
lliBii
^Bl
...
II
Fig. 6 Transmission electron micrograph of the ice cream mix emulsion in the absence (a) and pres
ence (b) of Polysorbate 80. (From Ref. 75.) F = fat globule, C = casein micelle, A = agarose matrix;
arrow points to fat crystal within the globule. Bar= 1 /xm.
they have the ability to lower the interfacial tension between the fat and water phases more
than the proteins (Table 5). Thus they become preferentially adsorbed to the surface of the fat
[8,58,73-75]. As the interfacial tension is lowered and proteins are eliminated from the sur
face of the fat, the surface excess (quantity of adsorbed material, m g/m ^) is reduced (Table
6) and the actual membrane becomes weaker to subsequent destabilization [75,76]. This is
due to the fact that the protein molecules, particularly the caseins, are considerably larger
than the emulsifier molecules so that a membrane made up entirely of emulsifier is very thin
(Fig. 6b), i.e., there is lower surface excess, although the emulsion is thermodynamically
favored due to the lowering of the interfacial tension and net free energy of the system [75].
However, Dickinson et al. [77] point out that subsequent instability to shear is not induced
by the protein displacement per se but by some change in the adsorbed layer properties caused
by the replacement of a pure protein monolayer by a mixed protein-surfactant monolayer.
Crystallization of fat also occurs during aging, creating a highly intricate structure of nee
dle-like crystals within the globule (Fig. 7). The high melting point triglycerides crystallize
first and continue to be surrounded by liquid oil of the lower melting point triglycerides. The
process of fat crystallization during aging of an ice cream mix was investigated by Iversen
and Pedersen [56]. Examining the crystallization of pure butterfat and butterfat in an emulsion
344 Goff
Table 5 Interfacial Tension Values (dyn/cm)

Between 11% Milk Solids-Not-Fat Solutions and
Anhydrous Butter Oil in the Presence of Various
Emulsifiers ^
Emulsifier Interfacial tension^ (dyn/cm)
No emulsifier 6.16
GMS 5.52 a
GMO 5.09 b
Span 60 5.64 a
Span 80 5.02 b
Tween 60 2.42
Tween 80 2.24
^n = 4. Emulsifiers were dispersed in their appropriate sol

uble phase (0.08%) at 70 °C. Interfacial tension measured
by a duNuoy ring.
Values followed by the same letter are not significant at
p < 0.05.
Source: Ref 8.
by pulsed NMR techniques, they reported that fat crystallization had completed within 1.5 h.
A partially crystalline fat is necessary for clumping to occur [52,78]. Van Boekel and Walstra
[79] found emulsion stability of a paraffin oil in water emulsion to be reduced by six orders
of magnitude when crystals were present in the dispersed phase. This has been attributed
[73,75,79] to the protrusion of crystals into the aqueous phase causing a surface distortion of
the globule (Fig. 8). The crystal protrusions can then pierce the film between two globules
upon close approach. As the crystals are preferentially wetted by the lipid phase, clumping is
thus inevitable [78]. This phenomenon may account for partial clumping of globules under a
shear force (Fig. 9). The clusters thus formed actually hold the ice cream serum in their
interstices, resulting in the observed dryness. These fat globule chains may also envelop the
air cells, thus improving overrun and adding stability to the foam structure [30,78]; however,
fat crystals have also been shown to impair overrun development in whipped cream [80]. It
is possible that in addition to crystallization, polymorphic transitions of the fat crystals are
occurring. Fat crystals exist in three polymorphic forms, a , p ' , and in order of stability
[52]. The rotation of the triglycerides from one form to another is important in many food
fats, especially those of a relatively few fatty acids, and takes time to occur [81].
Table 6 Differences in the Amount of Protein Adsorbed^ to the Fat

Globules in the Same Mixes as in Table 4
No emulsifier Polysorbate 80 (0.08%)
Mix protein (%) 4.08±0.01 4.08 ±0 .0 1

Protein in semm 3.43 ±0.07 3 .7 6 ± 0 .17
(% of mix)
Adsorbed protein 15.9 7.84
(% of total protein)
^n = 4. Determined by centrifugal separation and Kjeldahl analyses.

Source: Ref. 75.
Ice Cream 345
Fig- 7 Milk fat globule from an ice cream mix showing the intricate structure of needle-like crystals
present in the cross section of the globule, preserved by fixation at 4°C. (From Ref. 75.) F = fat globule;
arrow points to fat crystal. B ar= 1 ¡xm.
The next stage of structure development occurs during the concomitant whipping and freez
ing step. Air is incorporated either through a lengthy whipping process (batch freezers) or by
being drawn into the mix by vacuum or injected under pressure (continuous freezers) [2].
This process causes the emulsion to undergo partial coalescence or fat destabilization, during
which clumps and clusters of the fat globules form and build an internal fat structure or
network into the frozen product [2,9,78,82] in a manner very analogous to the whipping of
heavy cream [11,52,73,83]. In studying the whipping of heavy cream, Schmidt and vanHooy-
donk [10] stated that the proteinaceous membrane that envelops the air bubble is penetrated
by fat globules as the process proceeds, and this fat penetration offers foam stability to the
f e » i s .
Fig. 8 Milk fat globule from an ice cream mix. Arrow points to surface distortion of the globule
caused by the presence o f the internal crystalline structure. (From Ref. 75.) B ar= 1 /xm.
346 Goff
Fig. 9 A schematic representation of the network of partially coalesced fat globules formed during
whipping, illustrating the important role of fat crystals in cementing the coalescing globules together
(visualized as the straight lines within the globule). (Courtesy of Prof. P. Walstra, Wageningen, Nether
lands.)
whipped product. Brooker et al. [11] offered a more detailed explanation of the whipping
and foam stabilization process in heavy cream. During the initial stages of whipping,
air bubbles were stabilized primarily by ^-casein and whey proteins with little involvement
of fat. Adsorption of fat to air bubbles occurred when the fat globule membrane coalesced
with the air/water interface. Only rarely did fat spread at the air/water interface. The final
cream was stabilized by a cross-linking of fat globules surrounding each air cell to adjacent
air cells, thus building an infrastructure in the foam. In skim milk foams, the initial air/water
interface is also formed by the serum proteins and soluble ^-casein with little involvement
of micellar casein. Micelles became attached as a discontinuous layer but were not deformed
or spread [28]. It can be postulated that air cell incorporation into ice cream mix follows
a similar mechanism. Cross-linking of fat globules from one air cell to the next, thus form
ing an infrastructure, is less likely due to the reduction in dispersed phase volume from the
heavy cream system to the ice cream mix system; however, it must also be remembered that
the air bubbles, fat globules, and aqueous phase are being freeze-concentrated at the same
time.
Fig. 10 The effect of emulsifier addition on the percent of fat destabilized (determined by turbidity of
a 1:500 dilution at 5°C) during the period of time the ice cream mix was held in the barrel of the batch
freezer. Ice crystallization began after about 5 min. (From Ref. 8.)
Ice Cream 347
Relative Volume (% )
10-
10 .
Droplet Size (jim)
Fig. 11 Droplet size distribution in ice cream with butteroil as a fat source measured at room tempera
ture as a function of the stage o f the process, (a) For an aqueous solution of homogenized, aged mix. (b)
For an aqueous solution of hardened, thawed ice cream, (c) For hardened, thawed ice cream measured in
1 wt % SDS, used as a dissociating medium. (From Ref. 87.)
The fat globule clusters formed during the process of partial coalescence are responsible
for surrounding and stabilizing the air cells and creating a semicontinuous network or matrix
of fat throughout the product, resulting in the beneficial properties of dryness upon extrusion
during the manufacturing stages (aids in packaging and novelty molding, for example), a
smooth-eating texture in the frozen dessert, and resistance to meltdown or good stand-up
properties (necessary for soft serve operations) [4,9,84,85]. Kloser and Keeney [86] presented
the first evidence that fat destabilization was enhanced by some of the emulsifiers in common
use, and this has been confirmed by numerous researchers (Fig. 10) [4,8,72,78,84]. When
the emulsion is subjected to the tremendous shear forces in the barrel freezer, the thin mem
brane created by the addition of surfactant is not sufficient to prevent the fat globules from
colliding and coalescing, thus setting up the internal fat matrix. If an ice cream mix is sub
jected to excessive shearing action or contains too much emulsifier or not enough protein.
Stage of Mix Processing
Fig. 12 Changes in protein composition of aqueous phase as a function of the stage of the process in
the manufacture of ice cream with butteroil as a fat source, (a) Initial aqueous phase; (b) homogenized
mix; (c) aged mix; (d) thawed ice cream. (From Ref. 87.)
348 Goff
Fig. 13 The effect of ice crystallization as demonstrated by a reduction in draw temperature in a

continuous ice cream freezer on the fat destabilization process in ice cream. (From Ref. 8.)
the formation of objectionable butter particles can occur as the emulsion is churned beyond
the optimum level. It should be possible to predict the destabilizing power of an emulsifier
in the barrel freezer by measuring the interfacial tension between the fat and water phases in
the presence of the emulsifier to determine the extent of protein coverage that will result and
by considering the molecular size (and surface excess) of the emulsifier (Table 5) [8]. Polysor-
bate 80, having a small molecular weight and producing the lowest interfacial tension, dis
places more protein, resulting in a very thin membrane, and thus produces the maximum
amount of fat destabilization [8,75]. This can lead to the development of desired dryness or
to excessive fat clumping beyond the optimum level if not properly controlled.
Changes to the fat emulsion have recently been studied by Gelin et al. [87]. They demon
strated through light scattering measurements of fat globule size distribution and aggregation
(see Fig. 11) that the freezing step is responsible for considerable fat aggregation (Fig. 11b)
but that this aggregation is reversible through dissociation with SDS (Fig. 11c). They also
showed the changes occurring in the protein distribution between the aqueous and adsorbed
states (Fig. 12). It is obvious from this study that the homogenization step accounts for a
large amount of adsorbed protein and that casein is preferentially adsorbed over whey protein.
The aging and freezing-hardening-thawing steps each accounted for subsequent protein de
sorption, again mostly of the caseins. The emulsifier used in these mixes contained saturated
mono- and diglycerides but no Polysorbate 80. According to the results of Goff and Jordan
[8], Polysorbate 80 may have led to greater amounts of protein desorption during the aging
period. The sequential process of partial coalescence during ice cream freezing has also been
examined [8]. The incorporation of air alone, or the shearing action alone, independent of
freezing, was not sufficient to cause the same degree of fat destabilization as when ice crystal
lization occurs at the same time (Fig. 13). The freezing process causes an increase in concen
tration of the mix components in the unfrozen water phase. It is believed that the ice crystals
contribute to the shearing action on the fat globules, owing to their physical shape, and that
the concentration of components also leads to enhanced destabilization. However, to create
the desired fat destabilization, whipping and freezing must occur simultaneously [8,49].
The static freezing step is responsible for the conversion of most of the remaining unfrozen
water into ice and for the continued crystallization of the fat [3,4]. A concentrated, unfrozen
solution of sugars and salts remains, although the sugars are probably supersaturated or in the
glassy (amorphous solid) state [3,4,30,46,47]. It is likely that a glass transition temperature
is reached at temperatures less than —30 to —35°C (composition-dependent) where the re
Ice Cream 349
maining solution solidifies but does not crystallize [36,45]. Below this glass transition temper
ature, it is unlikely that any further physical or chemical phenomena occur.
B. Flavor and Structure Defects

There can be numerous flavor and textural defects associated with ice cream. This chapter dis
cusses only those associated with the fat content. However, an excellent review on ice cream
defects has been published by Body felt et al. [88]. Defects in ice cream flavor associated with
the fat phase are usually related to either autoxidation of the fat, resulting in oxidized flavors, or,
especially in the case of milk fat, lipolysis of free fatty acids from triglycerides by the action of
lipases (known in the dairy industry as rancidity). These defects tend to be present in the raw
ingredients used in ice cream manufacture rather than promoted by the manufacturing process
itself. The fat phase can also account for textural defects associated with content (too high or too
low) or degree of fat destabilization (affecting melting properties).
Oxidation of fat proceeds by the well-known autoxidation reaction [89] in three stages:
initiation, propagation, and termination. Free radicals formed during the initiation step, espe
cially from unsaturated fatty acids or phospholipids, are then able to initiate oxidation of
triglycerides, especially in the presence of copper and proteins. During propagation, antioxi
dant compounds such as tocopherols and ascorbic acid are depleted while peroxide derivatives
of fatty acids accumulate. Peroxides, which have little flavor, undergo further reactions to
form a variety of carbonyls, some of which are potent flavor compounds, especially some
ketones and aldehydes. Milk may oxidize as a result of factors either extrinsic or intrinsic to
the milk [12,90-92]. Important extrinsic factors include contamination with metals, tempera
ture of storage, oxygen tension, heat treatment, agitation, light, and acidity. Both copper and
iron may catalyze lipid oxidation, but probably only copper is significant in milk. Added
copper is much more potent than natural copper because a significant portion of added copper
goes directly to the fat globule [12]. Significant intrinsic factors affecting milk fat oxidation
include metalloproteins such as milk peroxidase and xanthine oxidase, endogenous ascorbic
acid, which acts as a cocatalyst with copper to promote oxidation, endogenous copper content,
and endogenous antioxidants, mainly tocopherols [93]. Fresh forage is well known to control
spontaneous oxidation, as indicated by obvious seasonal effects on the incidence of oxidized
flavor. This effect is probably due to increased levels of endogenous antioxidants.
Hydrolysis of fatty acid esters by the action of lipases results in the common flavor defect
known as lipolytic or hydrolytic rancidity and is distinct from oxidative rancidity [90]. Lipoly
sis in dairy fats can be extremely detrimental due to the number of highly volatile short-chain
fatty acids present, especially butyric acid. A comprehensive review of flavor impairment of
milk and milk products due to lipolysis was published by the International Dairy Federation
[94]. Lipases are unique among enzymes in that they are active at the lipid/serum interface.
In milk, lipases are ineffective unless the fat globule membrane is damaged or weakened in
some way. Lipolysis may be caused by the lipoprotein lipase (LPL) that is endogenous to
milk or by bacterial lipases. The properties of the fat globule membrane are most important
to lipolysis. The observed lactation effects may be due to reduced content of phospholipids in
the fat globule membrane in late lactation [95]. Mastitis, which alters milk composition, also
increases sensitivity of the fat globule to lipolysis [96]. Other factors that destabilize the fat
globule membrane, especially agitation and/or foaming, also promote lipolysis. Churned fat
does not appear to be a good substrate for LPL [95], but lipolysis is accelerated by the
replacement of the native membrane with surface-active material (mainly casein micelles and
whey proteins) from the plasma [97]. This effect is at least partly due to redistribution of LPL
350 Goff
from the plasma to the fat globule membrane and accounts for greatly increased lipolysis after
homogenization. In the milk from some animals, lipolysis may proceed without subsequent
thermal or mechanical activation. This effect, frequently referred to as spontaneous lipolysis,
is unlikely to occur in herd milks or in pooled milks because it is prevented by mixing affected
milk with three to five times its volume of normal milk. The major conditions that influence
spontaneous lipolysis are late lactation, insufficient fresh forage, and low-yielding cows.
Textural defects attributed to the fat component of the ice cream are associated with either
the fat content or its degree of destabilization [88]. Fat contributes smoothness to the finished
product. Lowfat mixes must therefore compensate for this lack of inherent smoothness by
altering the ratio of other components, particularly the protein, polysaccharide stabilizer, and
emulsifier components. On the other hand, mixes high in fat such as the premium products
typically have a heavier, dense texture related to both high fat and lower than normal air
content. The fat destabilization process also affects the structure and hence the texture and
meltdown of the ice cream. If too much destabilization has occurred, a defect known as “does
not melt” may occur. This results from a network of fat that gives sufficient structure to the
product to hold its shape in the absence of the ice phase after warming to temperatures suffi
cient to melt the ice [2]. Some slowness of melt is desirable and can be achieved by fat
destabilization, but it is also considered desirable for the product to melt down or collapse to
a consistency not unlike the unfrozen mix after the ice phase has melted.
VI. CONCLUSIONS
In summary, the fat content of ice cream and related products is vital both for its flavor and
for structural formation and resulting texture. While it may be possible to replace the former
attribute in creating lower fat products, it is much more difficult to replace the latter attribute.
The partial coalescence process is a complicated one, with important roles being served by
both the proteins and the emulsifiers, in addition to the fat, and occurring sequentially during
several steps: homogenization, aging, and whipping/freezing. Consumers have come to expect
a smooth-eating, creamy product with good meltdown characteristics. There is no doubt that
control of the size distribution of the ice phase is critical in producing a smooth texture [3],
but so too is the fat content. With consumer interest in lower fat products increasing, espe
cially in North America, the challenges for the ice cream manufacturer lie in developing
products that meet consumer expectations for lower fat while not compromising textural
quality.
REFERENCES
International Dairy Federation, Consum ption Statistics f o r M ilk and M ilk P rodu cts, Int. Dairy Fed.
Bull. 295, Brussels, 1994, p. 1.
2. W. S. Arbuckle, Ice C ream , 4th ed., AVI, Westport, CT, 1986, p. 1.
3. H. D. Goff and M. E. Sahagian, Freezing of dairy products, in F reezing Effects on F ood Quality
(L. E. Jeremiah, Ed.), Marcel Dekker, New York, 1996, 299-335.
4. K. G. Berger, Ice cream, in F ood Em ulsions, 2nd ed. (K. Lars son and S. Friberg, Eds.), Marcel
Dekker, New York, 1990, p. 367.
5. R. Jimenez-Flores, N. J. Klipfel, and J. Tobias, Ice cream and frozen desserts, in D airy Science
and Technology H andbook, Vol. 2, P rodu ct M anufacturing (Y. H. Hui, Ed.), VCH, New York,
1993, p. 57.
6. H. L. Mitten and J. M. Neirinckx, Developments in frozen products manufacture, in M odern
D airy T echnology, Vol. 2, A dvances in M ilk P roducts (R. K. Robinson, Ed.), Elsevier, New
York, 1986, p. 215.
Ice Cream 351
7. P. G. Keeney and M. Kroger, Frozen dairy products, in Fundamentals of Dairy Chemistry, 2nd
ed. (B. H. Webb, A. H. Johnson, and J. A. Alford, Eds.), AVI, Westport, CT, 1974, p. 873.
8. H. D. Goff and W. K. Jordan, The action of emulsifiers in promoting fat destabilization during
the manufacture of ice cream, J. Dairy Sci. 72: 18 (1989).
9. P. G. Keeney, Development of frozen emulsions, Fo oJ Technol. 36(11): 65 (1982).
10. D. G. Schmidt and A. C. M. vanHooydonk, A scanning electron microscopical investigation of
the whipping of cream. Scanning Electron Microsc. Ill: 653 (1980).
11. B. E. Brooker, M. Anderson, and A. T. Andrews, The development of structure in whipped
cream. Food Microstruct. 5: 277 (1986).
12. P. Walstra and R. Jenness, Dairy Chemistry and Physics, W iley, New York, 1984.
13. P. Walstra, Physical chemistry of milkfat globules, in Developments in Dairy Chemistry, Vol. 2,
Lipids (P. F. Fox, Ed.), Applied Science, New York, 1983, p. 119.
14. J. M. deMan, Physical properties of milkfat. II. Some factors influencing crystallization, / . Dairy
Res. 28: 117 (1961).
15. J. Vasic and J. M. deMan, Effect of temperature history on the solid fat content of milkfat, J.
Dairy Sci. 48: 1277 (1965).
16. J. E. Schaap and G. A. M. Rutten, Effect of technological factors on the crystallization of bulk
milk fat, Neth. Milk Dairy J. 30: 197 (1976).
17. I. Soderberg, L. Hemqvist, and W. Buchheim, Milk fat crystallization in natural milk fat globules,
Milchwissenschaft 44: 403 (1989).
18. E. L. Thomas, Structure and properties of ice cream emulsions. Food Technol. 35: 41 (1981).
19. P. S. Tong, W. K. Jordan, and G. Houghton, Response surface methodology to study fat destabili
zation and development of overrun in ice creams produced with polyunsaturated safflower oil and
milkfat blends, J. Dairy Sci. 67: 779 (1984).
20 . K. Schmidt, A. Lundy, J. Reynolds, and L. N. Yee, Carbohydrate or protein based fat mimicker
effects on ice milk properties, J. Food Sci. 58: 761 (1993).
21 . H. D. Goff, Examining the milk solids-not-fat in frozen dairy desserts. Mod. Dairy 77(3): 16
(1992).
22 . P. A. Morrissey, Lactose: chemical and physicochemical properties, in Developments in Dairy
Chemistry, Vol. 3, Lactose and Minor Constituents (P. F. Fox, Ed.), Elsevier, New York, 1985,
pp. 1-3 4 .
23. H. D. Goff, J. E. Kinsella, and W. K. Jordan, The influence of various milk protein isolates on
ice cream emulsion stability, J. Dairy Sci. 72: 385 (1989).
24. J. G. Parsons, S. T. Dybing, D. S. Coder, K. R. Spurgeon, and S. W. Seas, Acceptability of ice
cream made with processed whey and sodium caseinate, J. Dairy Sci. 68: 2880 (1985).
25. J. E. Kinsella, Milk proteins: physicochemical and functional properties, CRC Crit. Rev. Food
Sci. Nutr. 21(3): 197 (1984).
26. M. C. Phillips, Protein conformation at liquid interfaces and its role in stabilizing emulsions and
foams. Food Technol. 35(1): 50 (1981).
27. P. J. Hailing, Protein stabilized foams and emulsions, CRC Crit. Rev. Food Sci. Nutr. 15: 155
(1981).
28. B. E. Brooker, Observations on the air serum interface of milk foams. Food Microstruct. 4:
289 (1985).
29. D. E. Smith, A. S. Bakshi, and C. J. Lomauro, Changes in freezing point and rheological proper
ties of ice cream mix as a function of sweetener system and whey substitution, Milchwissenschaft
39: 455 (1984).
30. K. B. Caldwell, H. D. Goff, and D. W. Stanley, A low-temperature SEM study of ice cream. I.
Techniques and general microstructure. Food Struct. 11: 1 (1992).
31. H. D. Goff, R. D. McCurdy, and G. N. Fulford, Advances in com sweeteners for ice cream.
Mod. Dairy 69(3): 17 (1990).
32. H. D. Goff, R. D. McCurdy, and E. A. Gullett, Replacement of carbon-refined com symps with
ion-exchanged com symps in ice cream formulations, J. Food Sci. 55: 827 (1990).
33. H. D. Goff, Heat shock revisited. Mod. Dairy 72(3): 24 (1993).
352 Goff
34. H. D. Goff and K. B. Caldwell, Stabilizers in ice cream. How do they work?, Mod. Dairy 70(3);
14 (1991).
35. K. B. Caldwell, H. D. Goff, and D. W. Stanley, A low-temperature SEM study o f ice cream. II.
Influence of ingredients and processes, Food Struct. 11: 11 (1992).
36. H. D. Goff, K. B. Caldwell, D. W. Stanley, and T. J. Maurice, The influence of polysaccharides
on the glass transition in frozen sucrose solutions and ice cream, J. Dairy Sci. 76: 1268 (1993).
37. E. K. Harper and C. F. Shoemaker, Effect of locust bean gum and selected sweetening agents on
ice recrystallization rates, J. Food Sci. 48: 1801 (1983).
38. Y. Shirai, K. Nakanishi, R. Matsuno, and T. Kamikubo, Effects o f polymers on secondary nuclé
ation of ice crystals, J. Food Sci. 50: 401 (1985).
39. A. H. Muhr and J. M. V. Blanshard, The effect of polysaccharide stabilizers on ice crystal forma
tion, in Gums and Stabilisers for the Food Industry, Vol. 2, Applications of Hydrocolloids (G. O.
Phillips, D. J. Wedlock, and P. A. Williams, Eds.), Pergamon, New York, 1983, p. 321.
40. A. H. Muhr, J. M. V. Blanshard, and S. J. Sheard, Effects of polysaccharide stabilizers on the
nucléation of ice, J. Food Technol. 21: 587 (1986).
41. A. H. Muhr and J. M. V. Blanshard, Effect of polysaccharide stabilizers on the rate of growth of
ice, J. Food Technol. 21: 683 (1986).
42. E. R. Budiaman and O. Fennema, Linear rate of water crystallization as influenced by temperature
of hydrocolloid suspensions, J. Dairy Sci. 70: 534 (1987).
43. E. R. Budiaman and O. Fennema, Linear rate of water crystallization as influenced by viscosity
of hydrocolloid suspensions, J. Dairy Sci. 70: 547 (1987).
44. N. Buyong and O. Fennema, Amount and size of ice crystals in frozen samples as influenced by
hydrocolloids, J. Dairy Sci. 71: 2630 (1988).
45. M. E. Sahagian and H. D. Goff, Thermal, mechanical and molecular relaxation properties of
stabilized sucrose solutions at sub-zero temperatures. Food Res. Int. 28: 1 (1995).
46. H. Levine and L. Slade, Principles of cryo-stabilization technology from structure/property rela
tionships of carbohydrate/water systems. A review, Cryo-Letters 9: 21 (1988).
47. H. Levine and L. Slade, Cryostabilization technology: thermoanalytical evaluation of food ingredi
ents and systems, in Thermal Analysis of Foods, (C. Y. Ma and V. R. Harwalker, Eds.), Elsevier,
London, 1988, p. 221.
48. W. H. Knightly, The role of the liquid emulsifier in relation to recent research on ice cream
emulsification. Ice Cream Trade J. 55(6): 24 (1959).
49. H. D. Goff, Emulsifiers in ice cream: how do they work?. Mod. Dairy 67(3): 15 (1988).
50. A. Tomas, J. L. Courthadon, D. Paquet, and D. Lorient, Effect of surfactant on some physico
chemical properties of dairy oil-in-water emulsions. Food Hydrocoil. 8: 543 (1994).
51. Food and Drug Administration, Milk Pasteurization Controls and Tests, U .S. Dept, of Health and
Human Services, Cincinnati, OH, 1981, p. 25.
52. H. Mulder and P. Walstra, The Milk Fat Globule, Commonwealth Agricultural Bureaux, Famham
Royal, Bucks., England, 1974.
53. W. D. Pandolfe, Development of the new Gaulin micro-gap homogenizing valve, J. Dairy Sci.
65: 2035 (1982).
54. H. Oortwijn, P. Walstra, and H. Mulder, The membranes of recombined fat globules. 1. Electron
microscopy, Neth. Milk Dairy J. 31: 134 (1977).
55. H. Oortwijn and P. Walstra, The membranes of recombined fat globules. 2. Composition, Neth.
Milk Dairy J. 33: 134 (1979).
56. E. K. Iversen and K. S. Pedersen, Ageing of Ice Cream, Grinsted Technical Paper TP 33-le,
1982, p. 1.
57. N. M. Barfod and N. Krog, Destabilization and fat crystallization of whippable emulsions (top
pings) studied by pulsed NMR, J. Am. Oil Chem. Soc. 64: 112 (1987).
58. N. M. Barfod, N. Krog, G. Larsen, and W. Buchheim, Effect o f emulsifiers on protein-fat inter
action in ice cream mix during ageing I. Quantitative Analyses, Fat Sci. Technol. 93: 24 (1991).
59. B. W. Tharp, T. V. Gottemoller, and A. Kilara, The Role of Processing in Achieving Desirable
Properties in Health-Responsive Frozen Desserts, Germantown Manufacturing Co. Tech. Paper
100, Broomall, PA, 1992.
Ice Cream 353
60. F. Franks, Complex aqueous systems at subzero temperatures, in Properties of Water in Foods
(D. Simatos and J. L. Multon, Eds.), Martinus Nijhoff, Dordrecht, Netherlands, 1985, p. 497.
61. F. Franks, Biophysics and Biochemistry at Low Temperatures, Cambridge Univ. Press, New York,
1985, pp. 2 1, 37, 167.
62. J. M. V. Blanshard and F. Franks, Ice crystallization and its control in frozen food systems, in
Food Structure and Behavior (J. M. V. Blanshard and P. Lillford, Eds.), Academic, New York,
1987, p. 5 1.
63. R. Bradley and K. Smith, Finding the freezing point of frozen desserts. Dairy Rec. 84{6)\ 114
(1983).
64. R. Bradley, Plotting freezing curves for frozen desserts. Dairy Rec. 85(7): 86 (1984).
65. K. G. Berger, B. K. Bullimore, G. W. White, and W. B. Wright, The structure of ice cream.
Dairy Ind. 37: 419, 493 (1972).
66 . L. M. Klotzek and J. G. Feeder, Relationship of emulsifier and homogenization pressures to fat
destabilization o f soft serve ice cream, J. Dairy Sci. 46: 592 (1963).
67. L. M. Klotzek and J. G. Feeder, Effect of high second stage homogenization and added salts on
churning of soft ice cream. Ice Cream Rev. 50(12): 26 (1966).
68 . K. G. Berger and G. W. White, The fat globule membrane in ice cream. Dairy Ind. Int. 41:
199 (1976).
69. D. F. Darling and D. W. Butcher, Milkfat globule membrane in homogenized cream, J. Dairy
Res. 45: 197 (1978).
70. P. Walstra and H. Oortwijn, The membranes of recombined fat globules. 3. Mode of formation,
Neth. Milk Dairy J. 36: 103 (1982).
71. P. R. Musselwhite and D. A. Walker, The effect of the colloidal state of the emulsion on ice
cream structure, J. Text. Stud. 2: 110 (1971).
72. R. Govin and J. G. Feeder, Action of emulsifiers in ice cream utilizing the HFB concept, J. Food
Sci. 36: 718 (1971).
73. D. F. Darling, Recent advances in the destabilization of dairy emulsions, J. Dairy Res. 49: 695
(1982).
74. J. Chen, E. Dickinson, and G. Iveson, Interfacial interactions, competitive adsorption and emul
sion stability. Food Struct. 12: 135 (1993).
75. H. D. Goff, M. Fiboff, W. K. Jordan, and J. E. Kinsella, The effects of Polysorbate 80 on the
fat emulsion in ice cream mix: evidence from transmission electron microscopy studies. Food
Micro struct. 6: 193 (1987).
76. H. Oortwijn and P. Walstra, The membranes of recombined fat globules. 4. Effects on properties
of the recombined milks, Neth. Milk Dairy J. 36: 279 (1982).
77. E. Dickinson, R. K. Owusu, and A. Williams, Orthokinetic destabilization of a protein-stabilized
emulsion by a water-soluble surfactant, J. Chem Soc. Faraday Trans. 89: 865 (1993).
78. K. G. Berger and G. W. White, An electron microscopical investigation of fat destabilization in
ice cream, J. Food Technol. 6: 285 (1971).
79. M. A. J. S. van Boekel and P. Walstra, Stability of oil-in-water emulsions with crystals in the
disperse phase. Colloids Surf. 3: 109 (1981).
80. B. E. Brooker, The adsorption of crystalline fat to the air-water interface of whipped cream. Food
Struct. 9: 223 (1990).
81. S. Fee and J. M. deMan, Effect of surfactants on the polymorphic behavior of hydrogenated
canola oil, Fette, Seifen, Anstrichm. 86: 460 (1984).
82. P. G. Keeney and D. V. Josephson, A measure of fat stability in ice cream. Ice Cream Trade J.
54(5): 32 (1958).
83. M. Kalab, Microstructure of dairy foods. 2. Milk products based on fat, J. Dairy Sci. 68: 3234
(1985).
84. P. M. Fin and J. G. Feeder, Mechanism of emulsifier action in an ice cream system, J. Food Sci.
39: 108 (1974).
85. W. Buchheim, N. M. Barfod, and N. Krog, Relation between microstructure, destabilization
phenomena and rheological properties o f whippable emulsions. Food Microstruct. 4: 221
(1985).
354 Goff
86. J. J. Kloser and P. G. Keeney, A study of some variables that affect fat stability and dryness in
ice cream, Ice Cream Trade J. 55(5): 26 (1959).
87. J. L. Gelin, L. Poyen, J. L. Courthadon, M. Le Meste, and D. Lorient, Structural changes in oil-
in-water emulsions during the manufacture of ice cream. Food Hydrocolloids 8\ 299 (1994).
88 . F. W. Body felt, J. Tobias, and G. M. Trout, The Sensory Evaluation of Dairy Products, Van
Nostrand Reinhold, New York, 1988, p. 166.
89. W. N. Wassef, Lipids, in Food Chemistry (O. R. Fennema, Ed.), Marcel Dekker, New York,
1985, pp. 176-205.
90. H. D. Goff and A. R. Hill, Dairy chemistry and physics, in Dairy Science and Technology Hand
book, Vol. 1, Principles and Properties (Y. H. Hui, Ed.), VCH, New York, 1993, p. 1.
91. J. C. Bruhn, A. A. Franke, and G. S. Goble, Factors relating to development of spontaneous
oxidized flavor in raw milk, J. Dairy Sci. 59: 828 (1976).
92. W. F. Shipe, R. Basette, D. D. Deane, W. L. Dunkley, E. G. Hammond, W. J. Harper, D. H.
Kleyn, M. E. Morgan, J. H. Nelson, and R. A. Scanlan, Off flavors of milk: nomenclature,
standards and bibliography, J. Dairy Sci, 61: 855 (1978).
93. R. L. King, Oxidation of milk fat globule membrane material, J. Dairy Sci. 45: 1165 (1962).
94. W. K. Downey, Flavour Impairment of Milk and Milk Products Due to Lipolysis, Int. Dairy Fed.
Document 118, Brussels, 1980.
95. H. C. Deeth and C. H. Fitzgerald, Lipolytic enzymes and hydrolytic rancidity in milk and milk
products, in Developments in Dairy Chemistry, Vol. 2, Lipids (P. F. Fox, Ed.) Applied Science,
London, 1983, pp. 195-239.
96. S. C. Murphy, K. Cranker, G. F. Senyk, D. M. Barbano, A. I. Saeman, and D. M. Gallon,
Influence of bovine mastitis on lipolysis and proteolysis in milk, J. Dairy Sci. 72: 620 (1989).
97. M. Anderson, Milk lipase and off-flavour development, J. Soc. Dairy Technol. 56(1): 3 (1983).
13___________
Cream Alternatives
lain J. Campbell and Malcolm G. Jones

Unilever Research Laboratory, Sharnbrook, Bedfordshire, England
I. INTRODUCTION
Cream alternatives first came onto the market in Europe as short shelf life products for the
bakery and catering sectors. They were formulated such that they could be used in both
pouring and whipping applications, with the emphasis being on physical rather than organo
leptic quality. The whippable products generally had high overruns and had very good shape
retention, although the flavor and oral melt properties were poor. More recently products have
been developed for the retail market to compete directly with dairy cream, and in the United
Kingdom this growth has been particularly significant as exemplified by the Van den Berghs
Foods product Elmlea, which has achieved almost 20% of the total cream market by volume.
The dairy cream market is segmented in a number of ways. The main differentiations made
are between fresh (pasteurized), long-life (sterilized at ultrahigh temperature, UHT), and aero
sols. It can also be divided in terms of fat level, and in the United Kingdom, single cream for
pouring contains not less than 18% fat, whipping cream not less than 35% fat, and double cream
not less than 48% fat. In many other European countries whipping cream contains not less than
30% fat, and double and single creams are less common. Cream alternatives, or non-dairy
creams as they are sometimes called, have been developed to compete in all these categories,
and many are close in flavor and textural quality to the dairy equivalents. Nearly all are UHT
products, hence giving the combined benefits of product performance and extended shelf life.
Outside of Europe the market for cream alternatives to compete directly with dairy cream
is less well developed, although there is some growth of whipping creams in Japan, while in
the United States the interest is mainly for coffee creamers.
Because cream alternatives are fabricated products it is possible to tailor the physical and
organoleptic properties of the emulsion by manipulation of the fat phase, the aqueous phase,
and the emulsifier system. This chapter describes the role of the various components, how
they interact with each other and with the processing conditions, and how this knowledge can
be used to develop new product options.
355
356 Campbell and Jones
II. COMPOSITION AND PROPERTIES

The development of cream alternatives with the desired storage and subsequent in-use proper
ties presents many challenges to the product developer. There is a requirement for an exten
sive knowledge of the role of the individual ingredients that make up the complete formula
tion, and since all cream alternatives are oil-in-water (o/w) emulsions, there is also the
requirement for a complete understanding of the factors that affect the stability of such emul
sions. A complicating factor in formulating creams is that emulsion stability is always re
quired during the storage shelf life of the products. However, for the cream to function during
use, for example whipping, the emulsion must destabilize in a controlled manner. The emul
sion scientist is thus faced with the problem of controlling the instability of the cream. This
section discusses some of the important issues that need to be considered in formulating a
cream alternative.
A. Liquid Creams
Cream alternatives consist of a dispersion of small (normally < 1 fim) oil droplets in an aque
ous phase containing milk proteins. A typical formulation (Table 1) contains a blend of vege
table oils (typically palmkemel and coconut oils), a milk protein source, one or more lipid
emulsifiers, and one or more polysaccharide stabilizers. The milk proteins impart both stabil
ity against recoalescence of the oil droplets and a desirable creamy flavor to the product.
Common sources of milk proteins are skim milk powder and buttermilk powder, although in
principle any milk source can be used. Homogenization of a preemulsion with coarse (> 20
fjLm) oil droplets results in small droplets coated in milk protein. The thickness of the protein
coat controls the stability of the droplets to recoalescence by the formation of a steric barrier
preventing film drainage and film rupture between adjacent oil droplets. When two oil droplets
approach each other to close distances, the protein molecules become entangled, resulting in
an osmotic pressure that forces the droplets apart. Thus, the protein coat ensures emulsion
stability against coalescence during storage.
Oil droplets that have densities less than that of the aqueous phase are prone to creaming
during storage. Stokes’ law predicts the creaming rate of particles in a dilute system, and is
given by
Creaming rate 17 = (p^ —p^)R^lr]^
where p^ and p^ are the continuous phase and disperse phase densities, respectively, R is the
oil droplet radius, and 7]^ is the continuous phase viscosity. Although this equation is inade
quate in predicting the creaming rate in concentrated emulsions it nevertheless shows the
Table 1 Typical Cream Alternative Formulations

Ingredient Whipping cream Pouring cream
Buttermilk powder (%) 7.00 7.00

Guar gum (%) 0.10 0.15
Hardened palmkemel oil 17.50 9.00
(PK38) (%)
Coconut oil (CN) (%) 17.5 9.00
Tween 60 (%) 0.35 0.30
j8-Carotene (%) 0.25 0.25
Water (%) 57.3 74.3
Cream Alternatives 357
parameters that affect creaming. It should be appreciated that most cream alternatives at typi
cal storage temperatures (approximately 5°C) consist of partially crystallized droplets, which
means that the density difference between the oil and aqueous phases is less than might be
expected. It is known that very small (0.5 ¡jbm) oil droplets with a high protein surface cover
age can be more dense than the aqueous phase and will therefore sediment. In reality most
creams are prone to creaming, especially if the storage temperature is high, and for this reason
a polysaccharide “stabilizer” is usually added to the cream to raise the aqueous phase viscosity
and so reduce the creaming rate as predicted by Stokes’ law. A secondary function of the
polysaccharide is to impart a more viscous or creamy mouthfeel to the product. Typical poly
saccharides used in cream alternatives are locust bean gum, guar gum, alginates, and carra
geenan. Care has to be exercised in the level of usage of many polysaccharides in protein-
containing emulsions as protein and polysaccharide compete for available water, resulting in
a type of depletion flocculation [1]. While depletion flocculation leads only to weakly attracted
oil droplets, flocculated droplets in dilute emulsions (<20% fat) will cream much faster than
single droplets. At higher phase volumes of fat, a percolated network of flocculated oil drop
lets can form a weak gel that prevents creaming. It is often noticed in cream alternatives that
the emulsion becomes less viscous on stirring, particularly after a period of storage. This is
due to the action of stirring being sufficient to break the weak gel formed by oil droplet floc
culation.
A much stronger form of flocculation that can occur in cream alternatives is via a reduction
in the electrostatic charge on the interfacially adsorbed protein. Above the isoelectric point
(pH 4.7 for casein), protein carries a net negative charge. This negative charge imparts stabil
ity against flocculation by creating an electrostatic repulsion between adjacent oil droplets. If
the charge is reduced by lowering the pH, or screened by the addition of calcium ions, the
electrostatic repulsion is reduced or eliminated (at the isoelectric point) and rapid strong floc
culation occurs. This effect is always witnessed when either fresh dairy cream or a cream
alternative is added to acidic fruit juice. Flocculation due to a reduction in the electrostatic
charge is the basis for the formation of a yogurtlike texture in soured cream. In soured cream
the pH is reduced by the formation of lactic acid, which is normally produced microbiologi-
cally. Flocculation in cream alternatives can be a major problem in creams designed to be
added to coffee. Any flocculation already present in the cream combined with the heat from
the coffee can cause a “feathering” appearance. Sodium citrate is often added to creams,
especially to so-called coffee creamers, to help sequester calcium ions and reduce flocculation
during storage.
The factors affecting the stability of cream alternatives can conveniently be shown by a
simplistic energy diagram (Fig. 1) similar to those given by DLVO theory (after Deryaguin,
Landau, Verwey, and Overbeek [2,3]).
A special type of oil droplet aggregation that can be used to produce viscous or slightly
gelled creams termed “clustering,” is employed to produce cream sold as extra thick or spoon-
able. Clustering is produced by high pressure homogenization, particularly in systems con
taining milk protein. Early work by Darling and coworkers [4,5] showed that clustering is
produced when the interfacial area of the emulsion is greater than the surface-covering capac
ity of the protein, leading to protein molecules sharing two adjacent droplets. In this case the
clustering will be eliminated by increasing the concentration of protein in the emulsion. We
have found that this type of clustering is prevalent in lower fat emulsions, while the opposite
effect is found in higher fat (>30% ) emulsions, where increasing the protein level can create
more oil droplet aggregation. Figure 2 shows a plot of emulsion viscosity versus buttermilk
powder concentration in creams containing 45% fat (50% coconut oil, 50% hardened palmker-
nel oil) and either 1% or 2% sodium caseinate. In both cases it can be seen that increasing
repulsion
attraction
Fig. 1 Conceptual energy diagram showing aggregation states possible in cream alternatives.
levels of buttermilk powder lead to more oil droplet aggregation and hence higher viscosities.
The higher level of sodium caseinate led to less aggregation. Electron microscopy of the
emulsions showed that this form of aggregation was due to casein micelles adsorbing to two
adjacent oil droplets as also shown by Darling and Butcher [5] and Ogden [6]. Increasing
numbers of casein micelles led to more bonds between aggregated droplets, resulting in higher
viscosities. It is clear that the mechanism for this type of aggregation is different from that
described by Darling and Ogden and also by Mulder and Walstra [7], although the microstruc
ture of the emulsion may be similar. We propose a mechanism in which oil droplets are
disrupted by the shear in the homogenizer valve, and during the recoalescence process (the
oil droplet size produced by homogenization is a result of rapid droplet breakup and recoales
cence) casein micelles prevent film drainage by adsorbing almost simultaneously to two adja
cent oil droplets. One could describe this type of aggregation, which we refer to as high fat/
high protein clustering, as a form of resticted coalescence in which the film thickness is
determined by the thickness of the adsorbed casein micelles. We prefer to describe the aggre
gation due to insufficient protein as bridging flocculation.
Clustering such as we have described is a very strong form of aggregation and is not
broken down by stirring. High power ultrasonics are, however, effective at declustering and
can reduce an almost solid cream to a low viscosity liquid. It is interesting to note that clusters
will not re-form after disruption because there is no mechanism for casein micelles to readsorb
to two oil droplets simultaneously. We have shown in other work that a wide range of smaller
molecule emulsifiers such as monoglycerides, polysorbates, lecithin, and sucrose esters, when
used at a low level (0.3%), are very effective in preventing clustering. It is believed that the
mechanism is similar to that by which sodium caseinate reduced clustering in the examples
shown in Fig. 2. During oil droplet breakup in the homogenizer there is a large increase in
interfacial area, and the rate of recoalecence of the new oil droplets will depend on the rate
of adsorption of surface-active materials from the aqueous phase. If adsorption is slow, much
oil droplet coalescence will occur, whereas if casein micelles are present between the thinning
films, clustering will occur. Small surface-active molecules will adsorb more rapidly than
large protein molecules and will therefore prevent recoalescence and clustering by covering
Buttermilk powder level (%)
Fig. 2 Effect of buttermilk powder level on the viscosity of creams containing 45% fat and either 1%
or 2% sodium caseinate.
the interface. This use of emulsifiers has an important yet often poorly understood role in the
formulation of most cream alternatives.
Be Whipping Creams
A large proportion of creams are sold for use as whipping creams. It is the development of
these creams that presents the emulsion scientist with the greatest challenge. If one considers
the energy diagram shown in Fig. 1, it is desirable to avoid all forms of instability during
storage, yet on the application of shear and the incorporation of air the oil droplets must
overcome the steric barrier to form an oil droplet network. The oil droplet network shown
surrounding air cells in Fig. 3 is responsible for imparting a desirable firm and creamy texture
to the whipped cream. Anyone who has attempted to whip cream will bear witness to the
■»V i
Fig. 3 Electron micrograph of freeze-fractured sample of whipped cream alternative.

difficulty of obtaining exactly the required texture, not too soft yet not overwhipped. Many
cooks will have experienced the indignation of producing butter rather than a perfect whipped
cream. It is for this reason (and, of course, convenience) that the cream alternative market
has seen a huge increase in the use of aerosol creams. The following section explores the
factors affecting the stability and “whippability” of creams.
The difficulty in controlling the whipping process is well illustrated by a whipping curve
generated using a Mohr-type apparatus [8]. The cream is stirred at a fixed temperature by a
constant-speed motor. Use of a torque meter on the motor or a simpler measurement of the
motor load yields a whipping curve of cream viscosity versus time. A typical curve is shown
in Fig. 4. As air is whipped into the cream a gradual rise in cream viscosity is witnessed until
a maximum is obtained in the curve. The optimum whipped structure is normally considered
to be close to this viscosity maximum. On further whipping, a decrease in viscosity occurs,
eventually leading to a further viscosity increase and the phenomenon of buttering. This type
of whipping curve is often used to examine the influence of processing or formulation vari
ables on the time to achieve an optimum whipped cream structure. Some of the important
factors in this process are discussed below.
Before one can fully appreciate the factors that influence both the whip time and the texture
of a whipped cream, it is important to understand the mechanisms involved in the formation
of a whipped cream structure. If a solution is stirred vigorously with a whipper beater, air is
rapidly incorporated and is disrupted into smaller air cells by the action of the beater. The
creation of small stable air cells is aided by a lowering of the air/water interfacial tension by
surface-active materials in the aqueous phase. The surface-active ingredients then play a vital
role in preventing film drainage and consequent film rupture. The ability of egg white to make
a stable foam structure is shown in Fig. 5. It can be seen that as the time of whipping
increases, both the phase volume and the firmness of the foam increase. A similar plot for a
whipping cream containing 45% fat (Fig. 6) shows many significant differences compared to
the whipping behavior of egg white. In the case of a cream, a maximim overrun is achieved
and further whipping only serves to exclude air from the whipped structure, whereas for egg
white no decrease in overrun was observed. A second difference can be seen when overrun is
plotted against firmness (Fig. 7), showing a direct correlation between these two parameters
Fig. 4 A typical Mohr-type curve of torque versus whip time.

500
400
25 S '
‘E
3
300 E'
0) # firmness
^ 200
100
whip time (minutes)

Fig. 5 (■ ) Overrun and ( ^ ) firmness versus whip time for an egg white solution.
for egg white but not for cream. The firmness of cream continues to increase even after the
maximum overrun is achieved. The conclusion from these results is that the microstructures
of an egg white foam and a whipped cream are very different.
A whipped cream relies for its rigidity and also its stability against air cell coalescence on
the presence of a fat droplet network, as previously shown in Fig. 3. The network is formed
by shear-induced aggregation of oil droplets resulting from rupture of the interfacial film
between adjacent droplets during the whipping process. It is essential that the oil droplets
contain fat crystals at this stage to give them mechanical rigidity such that full droplet coales
cence is prevented. For this reason it is always necessary to cool the cream in a refrigerator
before whipping to ensure that sufficient solid fat is present. Because the whipped cream
structure depends on the formation of a fat droplet network, and because the strength of the
300 150
250
200
I overrun
>firmness
150 -
100
20 30
time (mins)
Fig. 6 (■ ) Overrun and ( ^ ) firmness versus whip time for a 45% fat cream alternative.
150
Z5 100
# Cream alternative
□ Egg white
50
0 100 200 300 400 500

overrun (%)
Fig. 7 Overrun versus firmness plots for (□ ) egg white solution and ( ^ ) a cream alternative.
network will ultimately depend on the strength of the oil neck between two partially coalesced
droplets, one would expect the nature of the fat blend to affect the whipped firmness. This is
confirmed by the plots of overrun versus firmness, shown in Fig. 8, for a series of 45% fat
creams with fat blends with different ratios of hardened palmkemel oil (PK38) and rapeseed
oil (RP). The figure shows that the maximum obtainable overrun and the firmness are directly
related to the amount of PK38 in the blend.
The importance of the fat blend composition in determining whipped cream properties
makes it imperative that full crystallization of the blend occur within the oil droplets. It is
important to appreciate that the crystallization behavior of fat in bulk and in emulsion droplets
can be very different. Figures 9 and 10 show isothermal crystallization curves at a temperature
of 5°C for PK38 and coconut oil (CN), respectively, in three different systems: bulk oil, an
emulsion with droplets similar in size to dairy cream (£>3 2 = ~3îm ), and an emulsion with
very small droplets (D3 2 = ~ l/x m ) produced by high pressure homogenization. These data
200
□
□□ Q
CO
~ 150 -
□
A. □ m 40% PK38
TO A
□ ♦ 60% PK38
1 100 A
A 80% PK38
A ^ „
♦ A □ □ 100% PK38
□
♦
50 ♦
■ ■ ^ ■.V ^
m A jrP
% 4 jg jS S ^
100 200 300

overrun (%)
Fig. 8 Overrun versus firmness for cream alternatives containing different levels o f PK38. (■ ) 40%
PK38; ( ♦ ) 60%; (A) 80%; (□ ) 100%.
* Bulk fat
^ 3 micron emulsion
-A 1 micron emulsion
0 1 2 3
Thousands
time (seconds)
Fig. 9 Isothermal crystallization curves for PK38 at 5°C. (■ ) Bulk fat; (A) 1 /xm emulsion; ( ^ )
3fjim emulsion.
show that the crystallization rate of PK38 decreases with decreasing oil droplet size as shown
earlier by Walstra and Van Beresteyn [9], although an equilibrium level of solids is achieved
in each case. With coconut oil the case is quite different in that equilibrium solids are not
achieved in either of the emulsions. In fact, no crystallization occurred in the smaller oil
droplets even after several weeks of storage. The explanation for the different crystallization
behavior between bulk fat and emulsions lies in the manner in which crystallization is nucle
ated [9]. In bulk fat and large oil droplets, nucléation is heterogeneous, which means that a
single nucleus is sufficient to promote total crystallization. In very small oil droplets the
probability of heterogeneous nucléation occurring is very low owing to the lack of nuclei
within the droplets. Nucléation therefore has to be homogeneous, which requires a signifi
cantly lower temperature. For this reason coconut oil droplets with a diameter less than 1 /xm
can remain supercooled for very long times. The equilibrium level of fat solids achieved in
the emulsions shown in Figs. 9 and 10 is controlled by the number of droplets that were
homogeneously nucleated, although it should be noted that once a droplet is nucleated, full
m Bulk fat
^ 3 micron emulsion
-A 1 micron emulsion
1 2
Thousands
time (seconds)
Fig. 10 Isothermal crystallization curves for CN at 5°C. (■ ) Bulk fat; (A) 1 /xm emulsion; ( ^ ) 3
/xm emulsion.
* Bulk fat
^ Emulsion
40 60 80 100
% PK38 (w/w) in SF oil
Fig. 11 Crystallization temperature for fat blends with different levels of PK38 in SF for bulk fat and
( ♦ ) an emulsion.
equilibrium solids for a given temperature will be achieved within that droplet. Figure 11
shows a plot of crystallization temperature (measured by differential scanning calorimetry,
scan rate —5°C/min) as a function of PK38 level in sunflower oil (SF) in the fat blend for
bulk fat and a homogenized emulsion. The emulsion had an approximate mean oil droplet
size (D3 2) of 1 ^tm. It can be seen that the emulsion requires a temperature approximately
10-12°C lower to initiate crystallization than that required for the bulk fat. This factor can
present a major problem to the product developer, particularly with healthy fat blends con
taining less saturated fat, since oil droplet crystallization may not occur at normal processing
or storage temperatures. One method to overcome this problem is to “ seed” the oil droplets
with a minor (<2% ) fraction of a high melting triglyceride (e.g., fully hardened rapeseed
oil). This fraction will be homogeneously nucleated at normal processing temperatures and
will then heterogeneously nucleate the remainder of the fat blend to ensure that full crystalliza
tion takes place.
III. PROCESSING
Although cream alternatives have been manufactured in the past using H T S T (high tempera
ture ~85°C, short time —20 s) methodology, it is only sterilized (UHT) products that are of
significant commercial interest today.
There are five main steps in the UHT processing of creams, each of which can have a
significant effect on the properties of the final product. The steps are premixing, sterilization,
homogenization, cooling, and packing, and a typical flowchart is shown in Fig. 12,
Premixing is a critical part of cream production and needs to be carried out with care. The
aqueous phase must be sufficiently agitated to avoid clumping of gums and sufficiently heated
to ensure complete hydration without going to temperatures that may lead to alteration of
protein functionality. After the initial mixing it is normal to allow sufficient time (at least 1
h) for all components to fully hydrate. The oil phase is prepared in liquid form before mixing
with the aqueous phase. Most lipid emulsifiers, e.g., monoglycerides, Tweens, lecithins, and
polyglycerol esters, should be melted into the oil before addition to the premix. The oil phase
is added to the aqueous phase, normally at 5 5 -6 5 °C , and agitated sufficiently to maintain a
homogeneous mixture until homogenization. It is essential that the elapsed time between addi-
Cream Alternatives
tion of powders and completion of processing of the premix not exceed 3 h, as growth of
thermophilic bacteria could occur, which would result in spoilage of the premix.
Sterilization of the premix normally involves a preheating stage to approximately 80°C to
bring the mix temperature to a point where the sterilizer can cope with the A T required to
ensure complete sterilization. The temperature-time regime required for sterilization is de
fined in terms of the conditions required to ensure that Clostridium botulinum spores cannot
be present in the final product. The process equipment must therefore be heated with steam
to a sufficiently high temperature for a sufficiently long time to achieve a ''botulinum cook”
at the coldest part of the line. This normally requires time-temperature combinations such as
130°C for 1 h depending on the plant design. The hydrated raw materials are given a full
"botulinum cook” in-line, and this is normally by direct steam injection to a temperature of
14 0 -15 0 °C for 2 -5 s. The in-line sterilization can also be carried out by indirect heating via
tubular heaters or plate heat exchangers. This results in very different amounts of heat being
received by the products. In the case of the direct steam process, the product temperature
rises from 80 to 150°C and back to 80°C equally quickly. In the indirect process, more time
is required to raise the product to sterilization temperature and more time is required to cool
it down again as shown in Fig. 13. It is generally believed that direct steam injection results
in better flavored products than indirect heating, and certainly there can be differences in other
product properties linked to emulsion stability.
Homogenization is normally carried out with an aseptic high pressure valve homogenizer.
The temperature of homogenization is normally in the range 5 0 -9 0 °C , where the fat phase is
liquid, and the pressure is selected to give the desired droplet size. In some instances a two-
stage homogenization is given in which the second stage is used to reduce the size of any
large droplets that escaped the first stage, hence giving a narrower size distribution. A second
homogenizer valve can sometimes be used downstream of the main homogenizer to apply a
small pressure (5-10 bar) for declustering of fat droplets that have become clustered during
the initial high pressure treatment.
Cooling of the homogenized cream emulsion is an area that is not well defined and one
that is often not well controlled in factory production, particularly if aseptic holding tanks are
used prior to packing. The primary function of cooling is to ensure that crystallization of the
fat phase takes place, because fat solids are required if the cream is to be whippable. Also, if
Fig. 13 Temperature-time profiles for direct and indirect UHT process regimes.
a cream is collected at too warm a temperature, the resultant viscosity can lead to creaming
upwards of the fat droplets. The final cooling temperature must be low enough to ensure that
fat crystal nucléation occurs in all droplets, and the subsequent storage temperature must be
low enough to ensure that crystal growth will proceed.
In most industrial processes aseptic holding tanks are used as buffers between the pro
cessing lines and packing machines. They are frequently only partially jacketed so that cooling
can be applied but not always uniformly, and they are rarely stirred. During storage of cream
in the tank after cooling, crystallization proceeds with concomitant production of heat. Lack
of stirring means that the emulsion next to the cooled tank wall will experience a different
environment from the emulsion in the center of the tank. It is also known that temperature-
induced aggregation of oil droplets can occur in the aseptic tank, hence cream coming from
the tank may often have a shorter whip time than cream that has gone directly to the packing
machine. This is a result of the tempering that it has received in the tank. A temperature rise
in-pack of as much as 10°C can be observed when a holding tank is used, whereas the in
pack temperature is normally within one degree of the final cooled temperature when product
is packed directly. A variety of packaging regimes are used for UHT products, but for cream
alternatives most products are filled into 1 L Tetrapaks, 200 m L (or 250 mL) Tetrabriks, or
142 and 284 mL Cuplets. The packaging materials are normally sterile when made and are
given a relatively mild heat treatment combined with chemicals (usually hydrogen peroxide)
to provide an adequate level of safety assurance. Shelf life of products is limited by the
physical or taste stability of the products, not by microbiological considerations. Many prod
ucts sold in the retail market are placed in the chill cabinet for image reasons, but also because
a constant storage temperature helps remove the uncertainty associated with large temperature
fluctuations and the resultant effects on product stability.
IV. CURRENT MARKET

The growth of the cream alternatives market has taken place to a greater extent in the United
Kingdom than in the other dairy cream consuming countries of the world. Some European coun
tries have followed the U.K. trend (e.g., Sweden and Italy), but countries where consumption
of dairy cream is large (e.g., Germany) have not generally followed suit. The reason for this
may be that fresh dairy cream is considered a commodity rather than a luxury product, hence the
attraction of a cheaper alternative is not so great. Alternatively, a positive product attribute per
haps related to health may be required in order to expand the market in the same way as hap
pened with low fat spreads. This will certainly be the case for the United States, where dairy
products have suffered severely from a negative health image for some time now. The overall
dairy market in the United States is forecast to fall in value terms by about 30% by the year 2000,
with the greatest losses being in those segments where the health issue is important.
About two-thirds of the U.K. dairy cream market is accounted for by fresh cream, although
its share has remained fairly static since 1989. Sales of UHT cream have expanded at a
slightly faster rate largely due to the success of aerosols, which have been the main area of
growth within the cream market and now command around 40% of the UHT sector.
After an initial impact on the market, the cream alternative sector appears to have stabilized
at around 11%. The market is dominated in the United Kingdom by Van den Bergh Foods,
which markets a full fat Elmlea range, and a half-fat Delight range. Elmlea was launched in
1984 and has grown to almost 20% of the total cream market by volume, while Delight is a
much smaller brand with around 2% of the market.
Aerosols are the main growth area within the cream market, with sales having grown by
around 20% since 1991. Anchor claims to have an 80% share of the aerosol UHT creams
market by value and is the fastest growing brand with its full fat and half-fat products, both
of which consist of real cream. Branded cream alternative aerosols are also on the market,
but they account for only a small proportion of the total. In the United States, aerosol creams
are the most widely used form of whipped cream, but this is small in comparison to the usage
of ice cream as a desert topping.
V. FUTURE TRENDS
Future products in the cream alternatives category are likely to be marketed on the basis of
added value, in order to attract higher profit margins. The added value will most probably be
in terms of health issues, such as low SAFA or high PUFA, or even fat-free. These are
areas where it is difficult for the dairies to compete, particularly with products for whipping.
Technically, low SAFA/high PUFA and low fat/fat-free are difficult to achieve for a whip
ping cream.
Cream alternatives tailored for culinary use form another area where there is scope for
technical developement of products with added value. Control of oil droplet flocculation dur
ing cooking or in acidic products where ‘'feathering” can give an unacceptable appearance is
an opportunity for technical development.
Ready whipped cream for convenience has been achieved in the past with frozen dairy
cream, but for the chill cabinet or even ambient storage the target is much more difficult.
Stabilization of air cells against disproportionation and shrinkage may require something other
than the generation of the aggregated fat particle network that is currently present in whipped
cream structures.
REFERENCES
1. E. Dickinson, An Introduction to Food Colloids, Oxford Univ. Press, New York, 1992, pp. 101-
107.
2. E. J. W. Verwey and J. Th. G. Overbeek, Theory and Stability of Lyophobic Colloids, Elsevier,
Amsterdam, 1948
3. J. Th. G. Overbeek, The interaction between colloidal particles, in Colloid Science (H. R. Kruyt
Ed.), Elsevier, New York, 1952, pp. 211-211.
4. D. F. Darling and R. J. Birkett, Food colloids in practice, in Food Emulsions and Foams (E.
Dickinson, Ed.), Royal Society of Chemistry, London, 1987, p. 31.
5. D. F. Darling and D. W. Butcher, Milk fat globule membrane in homogenised cream, J. Dairy
Res. 45: 197-208 (1978).
6. L. Ogden, Homogenisation induced clustering of fat globules in cream and model systems, Ph.D.
Thesis, University of Minnesota, 1973.
7. H. Mulder and P. Walstra, The Milk Fat Globule, Centre for Agric. Pub. D oc., Wageningen, The
Netherlands, 1974.
8. W. Mohr and E. Mohr, Mo Ik. Kaseneiz 6: 34 (1955).
9. P. Walstra and E. C. H. van Beresteyn, Neth. Milk Dairy J. 29: 3 5 -6 5 (1975).
14
Ghee, Vanaspati, and Special Fa ts in India
K. T. Achaya
CSIR Emeritus Scientist (Retired), Bangalore, India
L USAGE OF FATS IN INDIA

A. Historical
The Indus Valley civilization of north India, spanning the millenium 2500-1500 BC, has
yielded carbonized remains of three oilseeds— sesame, mustard, and linseed—and clay models
of the coconut. Beautiful steatite seals depict fine breeds of dairy cattle that still flourish in
that region and imply a knowledge of milk fat. Numerous animal and fish bones would point
to familiarity with animal body fats. Copper frying pans of thoroughly modem design testify
to the prevalence of frying, and tandoor clay ovens, to the prevalence of baking. Use of
edible fats in India has thus a hoary lineage [1].
B. Current Patterns of Fat Use

National usage today is made up of ghee 3%, hydrogenated fat or vanaspati 18%, and vegeta
ble oil 79% [2]. In recent years, the average per capita daily intake has been about 13 g, with
a statewise range of 3-20 g depending on regional culinary practices, income, and mral or
metropolitan residence. Ghee is consumed in only 7.6% of Indian households, and two-thirds
of these are confined to two large northern states [3]. Household dietary surveys reveal a per
capita daily intake of ghee that averages only 0.36 g [2], but estimates based on the total milk
that is converted into ghee give the far higher figure of 2.5 g [3], implying considerable
indirect consumption of ghee as products. Vanaspati availability calculated from production
data and population figures yields a figure of about 2.0 g per capita per day [2]. By far the
largest use is of edible vegetable oil [2], whose proportion in the total fat mix varies according
to the location of states: southern 97%, eastern 86%, western 83%, and northern, where both
vanaspati and ghee have larger shares, 56%. Most traditional oils are consumed in unrefined
form; these include, for example, groundnut, rape-mustard, safflower, sesame, coconut, and
niger. Barely 15-20% is in the form of refined oils; this includes all newer entrants such as
cottonseed, soybean, sunflower, ricebran, and palm, while some traditional oils are also being
refined [4].
369
370 Achaya
Refining practices in India are those employed everywhere and are covered elsewhere in
this volume. For larger scale manufacture of ghee and vanaspati, distinctive technologies have
been developed on the Indian subcontinent. To some extent this is also true of bakery fats,
margarines, and fat spreads, which are also briefly noted in this chapter.
II. GHEE
A. Traditional Preparation
The oldest Sanskrit text extant, the Rigveda of about 1500 BC, refers to a cake of cereal flour
fried in ghrta (ghee), and indeed the Aryan culture would countenance no other cooking fat.
To start with, its origin from the sacred cow gave it special sanctity; further, being fire-
derived from butter, golden, shining ghee was deemed to be pure in itself and to confer ritual
purity on dishes cooked with it [1]. To make ghee, boiled and cooled milk from the cow or
the buffalo was set overnight using a pat of mixed starter culture from a previous run. The
firm curd was diluted and churned in various domestic devices [5, pp. 137-138] to yield
buttermilk and butter granules, which were patted together into lumps. Very little butter was
used in the home; material pooled for a few days was boiled down over a fire while stirring
with a ladle by hand to yield ghee. It was estimated in 1951 that about 42% of all the milk
produced in India was converted into ghee [5, pp. 299-307]. At present the figure is about
28%, which would yield annually an estimated 850,000-900,000 tonnes of ghee [3].
B. Nature of Ghee
7. Ghee Attributes
The shelf life of ghee is 6 -8 months even at high ambient tropical temperatures, with cow
ghee having an edge over buffalo ghee. This high stability is attributed to the low moisture
content (0.2%) and the high content of phospholipids (—40 mg/100 g) and perhaps of free
amino acids, which are liberated from the phospholipid-protein complex into the fat phase
[6]. Flavor precursors in ghee develop in part during the souring of milk to form curd (if
omitted, the flavor is flat and oily), but in the main during the rendering step by interaction
between a protein, which may be a casein degradation product, the reducing sugar lactose,
and minerals. A range of compounds have been identified, including ketones (propanone and
butanone-2), alcohols (ethanol, n-butanol), 2-ketohydrocarbons of 3 -9 carbon chain length,
monocarbonyls, ketoglycerides, oct- and dec-2-enols, and 8- and y-lactones of various chain
lengths [6].
Aroma is highly prized, with preferences in various parts of the country for undercooked
or overcooked flavors. Also valued is a granular appearance, which generally takes the form
of solid grains floating in a liquid mass. Buffalo ghee, which is more saturated, yields larger
crystals in the form of irregular clusters, and cow ghee forms smaller spheroids [7]. After
initial seeding, slow undisturbed cooling at a temperature 1°C above the melting point, about
29-3 r c , yields larger crystals than if the cooling is too rapid.
2. Ghee Composition
Milk fat makes up 99.0-99.5% of ghee. The fatty acids range in carbon chain length from 4
to 22, and buffalo ghee carries the following types as weight percentages: saturated 46, cis-
monoene 29, trans-monotm 7, diene 13, and polyene 5 [8]. Cow ghee would have about 5%
less saturates and correspondingly more cw-monoenes. In consequence, the following glycer
Ghee, Vanaspati, and Special Fats in India 371
ide types are present in cow and b u ffa lo ghee, respectively: SSS 42, 49; SSU 42, 39; SUU
14, 11; and UUU 2, 1 [7]. Thin-layer chromatography yielded three distinct molecular weight
groups in which various fatty acids were preferentially associated. Thus in the high molecular
weight triglycerides, hardly any very short chain fatty acids were associated with the stearic
and oleic acids. In the low molecular weight group, the SSS component consisted of two
short- or medium-chain fatty acids and one long-chain fatty acid, and in the diene and polyene
triglycerides one short- or medium-chain fatty acid was associated with two long-chain unsatu
rated fatty acids [8]. Ghee carries about 4% diglycerides and 0.6% monoglycerides, probably
developed during the several steps of its isolation from milk.
The unsaponifiable matter of ghee contains, among other things, 300 mg of cholesterol, 9
mg of lanosterol, and 4, 60, 9, 28, and 6 jug of lutein, squalene, vitamin A, vitamin E, and
ubiquinones, respectively, per 100 g. Cow ghee has a distinct yellow color and carries 72 jug
of carotenoids per 100 g; the latter are totally absent in buffalo ghee, which contains biliverdin
and bilirubin and therefore has a greenish cast [6].
3. Standards fo r Ghee
Two broad categories of ghee are specified under the AGMARK rules of the Agricultural
Produce (Grading and Marking) Act of 1937. These permit maximum free fatty acid levels
of 1.4% and 2.5% (as oleic acid) and a maximum moisture level for both of 0.3%. The
marked effects of feeding cottonseed to cattle, in hardening milk fat and depressing lower
fatty acid levels, especially in summer, are reflected in the following AGMARK specifications
for ghee:
Regional (cottonseed tract)
All-India Winter Summer
Butyrorefractometer reading at 40°C 4 0 .0 -4 3 .0 4 1 .5 -4 4 .0 4 2 .5 -4 5 .0

Reichert-Meissl value, min 28 23 21
Polenske value 1 .0 -2 .0 0 .5 -1 .2 0 .5 -1 .0
Under the Prevention of Food Adulteration Act and Rules of 1976, addition of antioxidants
to ghee is not permitted.
C. Preparation of Ghee
1. Principles o f G hee-M aking
Three distinct stages are recognized in the conversion of butter or cream into ghee by the
application of heat [9]. In the first stage, the temperature of the product must be raised rather
gradually to the boiling point of water while ensuring adequate stirring to control frothing. In
the second stage of heating, free water evaporates as steam with a crackling sound. This
requires a considerable amount of heat, so the rate of heat application can be increased while
ensuring sufficient agitation to prevent local overheating and sputtering. As most of the water
evaporates, a fine stream of bubbles rises from the cast-out solids-not-fat (SNF) at the base.
The rate of heating should be reduced to prevent their charring, since that would impart bitter
flavors and a brown color to the ghee. Overheating could also drive off desirable volatile
372 Achaya
flavor materials and also appears to alter the subsequent ability of the fat to form suitable
grains on cooling. A temperature of about 103°C should be maintained throughout this stage.
In the third stage, water bound to the non-fat solids must be removed without heat damage
either to the fat or the SNF; this is done by raising the temperature to between 105 and 118°C
with constant agitation. A light amber color in the precipitated SNF and a pleasant aroma are
traditional criteria to signal the end of the operation. Regional preferences range from a but
tery, uncooked flavor in the south to a slightly burned, overcooked flavor in the west. Cream
rendering takes much longer than when butter is used. Sometimes betel or curry leaves, which
contain antioxidant principles, are added at the end of the boiling down.
The yield of ghee in domestic or village preparation is about 82% of the fat in milk; about
13% is lost in the buttermilk produced during curd churning, about 2% is held in the ghee
residue, and handling losses are about 2.5% [10]. Dairy butter yields about 92% ghee, with
a handling loss of 4.5%, and when cream is boiled down, ghee recovery is 88%, with a 9%
loss of fat in the large ghee residue.
2. Larger Scale Traditional Production

Ghee was sold in bulk in the past by wholesale merchants in towns and cities [11]. Partly
clarified ghee was assembled from small producers for final desiccation in large iron vats kept
over wood fires and stirred with iron ladles. A new development at this level of operation is
the proliferation in Indian cities and towns of minidairies owning cream separators, in which
cream of 55-60% fat content is produced from milk [12]. Such cream is ripened for a day in
vats of 50-100 kg capacity, after which it is diluted and skillfully churned using both hands
to yield butter. This in turn is melted into ghee in pans that have a simple tilting device or a
hollow central tube to facilitate ghee removal into 17 kg tins. These are then delivered to bulk
users such as traditional sweetmeat makers, hostels, hotels, hospitals, and the armed forces.
As much as 186,000 t of ghee, comprising 22% of all the ghee made in India, is produced
by the unorganized sector [3].
3. Ghee Production in the O rganized Sector

Batch Production Commonly, a steam-jacketed stainless steel vessel of 500-1000 kg
capacity is employed. This is charged with butter, and the temperature is raised to 90°C with
agitation to remove water. When almost all of the water has gone, the temperature will tend
to rise sharply, and careful control is needed. The final temperature lies between 105 and
120°C depending on the flavor desired. To remove sediment, the ghee is either centrifuged or
filtered through fine muslin or even through filter paper. The use of glass bottles and tin cans
of various sizes to pack ghee is yielding to the use of plastic jars and flexible film laminates,
which are easier to handle and comparatively impermeable to oxygen and water [6,13].
An alternative starting material for ghee manufacture is cream of 50% fat content obtained
from a standard cream separator. This contains about 5% solids-not-fat and foams copiously
when heat is applied. Washing it with an equal volume of water will reduce the SNF content
to 1-2% , not only reducing foaming but also raising the yield of ghee by reducing the quantity
of ghee residue and hence the amount of fat held up in it. After washing, the cream can be
raised to an 80% fat level using a plastic cream separator. To improve the flavor of the ghee
made, the cream can be ripened using a lactic starter to a lactic acid level of 0.2% or simply
acidified with citric acid. Rendering cream is a slower operation than when butter is used,
and precautions are necessary to achieve both an acceptable color and flavor [9,13].
A procedure involving prestratification can effect considerable economy in steam consump
tion [14,15]. Butter is held in a ghee kettle with a bottom valve at 80°C for 30 min, when
three layers are formed. The top layer is a scum of curd particles, the middle bulky layer is
essentially melted fat, and the bottom layer is serum or buttermilk, which will contain 80%
of the original water and 60% of the SNF. The bottom layer is run off, and the other two are
boiled down together to yield ghee of good flavor thanks to the scum.
Continuous Ghee Making. Many advantages are claimed for continuous preparation meth
ods that employ scraped-surface heat exchangers [16]. These are a low steam consumption,
0.35 kg/kg of ghee prepared compared with 0.48 kg in a conventional batch unit; low electric
ity consumption, 0.01 kWh/kg; and a high thermal efficiency, 71% versus 37% for batch
operation. The very high heat transfer coefficients also permit a compact design.
The equipment is shown in Fig. 1. The butter is melted in a receiving and heating vat; thereaf
ter a gravity pump moves part of it to a gravity separator and part to a balance tank [13]. In the
gravity separator the butter is separated into two portions, rich in fat and serum, respectively.
The fat-rich fluid containing a little buttermilk goes back to the same balancing tank. If cream is
the starting material, the gravity separator is bypassed. The molten butter or cream then passes
to the first scraped-surface heat exchanger. Here it is heated with steam, after which the super
heated liquid (butter or cream) is flashed to the vapor separator, where water is separated. The
fat then passes for further moisture removal to the second scraped-surface heat exchanger, where
the ghee flavor develops. If cream is the raw material, passage to a third heat exchanger may be
necessary. After centrifuging or filtering, the clear ghee is packed.
Removal of water from cream or butter using heat is expensive and not very efficient.
Mechanical means can be used to break the emulsion and free the bulk of the water, leaving
heating entirely for flavor development. In a recent development, shown in Fig. 2 [13], the
fat-in-water emulsion is broken down using centrifugal forces, after which a scraped-surface
heat exchanger and vacuum vessel serve to generate flavor. The plant is claimed to be less
energy-intensive, to result in lower losses of both fat and SNF, and to cause less environmen
tal thermal pollution. Moreover, much larger volumes can be handled.
Some 110,000 t of ghee, or 13% of total production, derives annually from the organized
sector, under a score of brand names. Five brands are each produced in quantities of 5000-
13,000 t [3]. This development is just a decade old.
III. VANASPATI
A. Growth of the Industry
1. H istorical
Edible hydrogenated fats came into India as imports from the Netherlands after World War I, for
supply initially to bulk users of ghee like traditional sweetmeat makers and restaurants [17]. The
origin of the term “vanaspati,” from pati (lord) and v a n a (forest), has been lost, but the name
was almost certainly intended to emphasize its vegetable origin in contrast to the animal product
ghee. Even the later official designation for vanaspati was Vegetable Oil Product, or V.O.P.
Between 1927 and 1932, some 20,000 t annually was imported. Production in India com
menced in 1930, and within a decade imports ceased completely as the domestic annual produc
tion climbed to 66,709 t, produced in nine factories. World War II gave a tremendous boost to
production, following the decision to include vanaspati as an approved cooking fat for the armed
forces to supplement the use of ghee, which was in short supply. This led in October 1943 to a
control on price with a formula based on ruling vegetable oil prices and conversion costs. A year
later a comprehensive V.O.P. control order was promulgated under the Defence of India Rules
to (1) promote the manufacture of vanaspati on a priority basis, (2) regulate its procurement and
distribution, (3) control its price, and (4) lay down minimum quality standards [18]. In course of
time it became the practice, as part of price regulation, for the government to import edible oils
w
-^1
YT Vat
GS Gravity separator
BT Balance tank
HE Heat exchanger
VS Vapor separator
Fig. 1 Continuous ghee-making plant. (From Ref. 13.)
>
o
3"
<m
m
Seram Serum
Fig. 2 Continuous ghee-making process using cream. (From Ref. 13.)
for supply to vanaspati manufacturers at prices lower than those ruling in the Indian market. This
practice came to an end in December 1988; with it the control on the price of vanaspati also
ceased, and the product was obliged to compete for its place in the market.
2. Usage o f B ase Oils

At the start, the only oil used as a base for hydrogenation was screw-pressed groundnut oil,
following its refining, bleaching, and deodorization in the vanaspati factory itself. In conse
quence, groundnut oil production rose dramatically from a figure of 1.5 x 10^ t in 1930 to
4 x 10^ t in 1947, and it took its place as India’s leading edible oil. Addition of 5% sesame
oil as a latent marker was enforced in 1947. The decades that followed witnessed three devel
opments. Production of vegetable oils in India grew far too slowly to keep up with population
increase and the market demand for edible oils and vanaspati. This meant in turn that vegeta
ble oils had to be imported from world supplies, and the least expensive of these, e.g.,
soybean, rapeseed, and palm, were naturally favored. Third, a number of new oil sources
came to be developed and exploited in the country, such as cottonseed, ricebran, and soy
beans. Marketing these unfamiliar oils in India posed problems, and vanaspati manufacture
offered a ready outlet in a known and standard product.
Oils permitted for vanaspati manufacture are under constant review [4]. Even rebates and
incentives have been offered from time to time to encourage the use of particular oils in
vanaspati, such as cottonseed oil in the 1960s and ricebran oil in the 1980s. Since screw-
pressed groundnut oil is greatly in demand for direct edible use, its use for vanaspati manufac
ture is not now permitted. After much hesitation, the use of three oils that are solvent-ex
tracted from expeller oilcakes, namely groundnut, rape-mustard, and sesame, is now allowed.
Thirteen other oils are also permitted. Five of these are obtained by direct solvent extraction,
namely ricebran, soybean, sunflower, maize, and sal (use of which should not exceed 10%).
The other eight permitted oils are cottonseed, mahua, nigerseed, palm, palmolein, sunflower,
safflower, and watermelon seed.
376 Achaya
Figures of actual usage reflect these shifts [4]. When oil imports were extensive, as in
1986-1987, the major oils used in vanaspati manufacture were soybean (22% ) and palm
(17%), and, a year later, palm (31%) and rapeseed (9%), besides, of course, indigenous oils.
Once indigenous oil production began to increase and imports became notional, the order of
usage of oils in making vanaspati in 1992-1993 was soybean 26%, cottonseed 23%, ricebran
16%, rape-mustard 12%, sunflower 8%, and mandatory sesame 5%, for a total of 90%. The
remaining oils included mahua, nigerseed, maize, and others [4,19].
5. Regulations and Standards

Over the years, the original Vegetable Oil Products Control Order 1942 has been subject to
frequent revision [4,19]. As it now stands, vanaspati should conform to the following specifi
cations: moisture content 0.25% (max.); capillary slip melting point 31-41°C (from an earlier
upper limit of 37°C , which is nevertheless largely observed even now); butyrorefractometer
reading at 40°C, 48.0 (max.); free fatty acids as oleic 0.25% (max.); unsaponifiable matter
2.0% (max.); and nickel content 1.5 ppm (max.). No coloring or flavoring matter, nor any
antioxidant, synergist, emulsifier, or similar substances, nor any matter injurious to health can
be added to the product. Addition of refined sesame oil is mandatory, the quantity being such
that the vanaspati should give a Baudouin color of at least 2 red units in a 1 in. Lovibond cell
when mixed in a proportion of 20:80 with refined groundnut oil and examined under specified
conditions. Also compulsory is the addition of at least 25 lU of synthetic vitamin A; addition
of vitamin D is optional, but label declaration is necessary. The product has to be manufac
tured in premises maintained under hygienic conditions, and no refined oils other than those
permitted should be stocked on the premises.
There are two provisos. Vanaspati for use by the defense forces may contain up to 4 ppm
of diacetyl. If more than 30% ricebran oil is used in manufacture, the permitted level of
unsaponifiable matter is 2.5% (max.).
4. Vanaspati Production D ata

During the last decade, vanaspati production has varied between 816 and 985, average 906,
thousand tonnes annually [4,19]. In 1993, there were 133 factories for vanaspati manufacture,
with an installed capacity of 2041 thousand tonnes, hence capacity utilization is just 44%.
The location of factories is strongly regional, so that local demands, as shown below, can be
met without excessive transport costs.
Percentage of total vanaspati production
North West East South
1966 57 19 16
1993 57 27 9
The trend is toward less consumption in eastern India and greater usage in the traditionally
ghee-consuming western region.
B. Hydrogenation Process Parameters

Chapter 9 deals with the hydrogenation of oils. Accordingly, only outlines of Indian practices
relating to vanaspati production are set down here.
1. Equipm ent and O peration

The batch hydrogenator is essentially a dish-ended autoclave of mild steel varying in capacity
from 5 to 15 t [20,21]. It is fitted with closed and open steam coils and variable turbine or
baffle stirrers. At the top is a vacuum connection and a sight glass. Oil from a storage tank
and catalyst from a doser are drawn in from the top. Compressed hydrogen gas from a holder
is admitted at the bottom through a perforated ring or a fritted ceramic disk. Temperature
indicators are provided. It is usual to have a number of hardening autoclaves so that various
oils can be separately hydrogenated.
Alkali-refined, bleached, filtered, and deodorized oil is drawn into the hydrogenator to
about two-thirds its capacity, either cold or preheated in a heat exchanger. A measured quan
tity of the nickel catalyst in the form of pellets or flakes is introduced from the catalyst
doser. Stirring is commenced, and steam is slowly introduced through the closed coils. As
the temperature rises to within 10°C of the hydrogenation temperature, vacuum is applied to
draw off dissolved moisture, which could otherwise retard the hydrogenation reaction. When
the oil is deemed dry, compressed hydrogen at a pressure of 1.1-1.3 kg/cm^ is let into the
autoclave. The hydrogenation reaction being exothermic, the temperature rises, and usually
no further heat need be supplied; in fact, if too vigorous, cold water may even be passed
through the coils. The degree of hardening is followed by noting the consumption of hydrogen
with time based on previous experience with the same oil in the same unit. When the process
is deemed to be nearly complete, a sample of the hardened fat is drawn out using vacuum
and its melting point is determined [22-25].
Cold water is now drawn through the coils until the temperature drops to about 95°C, after
which the oil passes for further cooling to a heat exchanger to warm up the next batch of oil.
This is followed by filtration in a plate-and-frame or leaf filter press at 50-60°C to remove
the suspended catalyst.
Postrefining operations include neutralization, bleaching, and deodorization. During hydro
genation, the level of free fatty acids will show an increase, and these are neutralized using dilute
lye of 5° Beaumé. Bleaching with 0.25-0.5% bleaching earth serves to absorb both coloring
matters and any nickel that may be present in the form of soap. Characteristic odors develop
during hydrogenation, which have been attributed primarily to short-chain aldehydes, ketones,
alcohols, and lactones. Deodorization is effected by intimate contact between the hydrogenated
fat held in thin layers and dry superheated steam at a high vacuum and temperature. In batch
units this can take the form of shallow trays or pans standing one below the other, with the oil
moving downward by gravity. In continuous units the stripping steam itself spreads the flowing
oil in very fine layers on tubular surfaces, yielding turbulence and high interfacial contact for
efficient and rapid deodorization. After deodorization the vanaspati is sometimes filtered again.
Hydrogenated products made from various permitted oil components are then blended in a
stirred tank, into which refined and deodorized sesame oil, appropriate quantities of vitamin
A, and vitamin D, where this is added, are also introduced. The finished vanaspati then passes
at 50-60°C to storage tanks. After being filled into containers, the latter are allowed to stand
undisturbed while cooling gradually to near ambient temperature. Fresh air is now blown into
the room to hasten the rate of cooling. When an incipient cloudiness is noticed, refrigerated
air is blown in to hasten cooling and induce a uniform and fairly firm consistency [20].
2. Types o f O peration
Two systems are in vogue for hydrogenation [21,22-25]. In the recirculation system, hydro
gen gas that passes unused through the oil and collects at the top is returned to the gas holder
after passing through a cooler, a scrubber to absorb moisture, a degreaser to trap entrained
378 Achaya
fat, and a compressor. In the dead-end system, the gas that thus accumulates, along with any
volatile matter, is simply vented away; this constitutes both a waste and a fire risk but elimi
nates the need for costly gas purification equipment. Dead-end operation is also well-suited to
monitoring the hydrogen uptake.
Individual hydrogenator units in India are frequently of 10 t capacity. Average factory
output is close to 50 tpd [4]; about half the units turn out less, and the other half more, than
this figure (Fig. 4, pp. 385).
3. Consum ption o f H ydrogen

For a unit drop in iodine value, 0.9-1.1 m^ of hydrogen is consumed by 1 t of oil. Thus the
quantities of hydrogen needed to bring about the drop in iodine value required to meet vanas-
pati specifications for any batch size of oil can be calculated. Both theoretical calculations
and practical experience show that to convert 10 t of oil to vanaspati, the requirement of
hydrogen in cubic meters at NTP for various oils is as follows: soybean 645, cottonseed 555,
ricebran 530, rapeseed 500, and palm 220. For groundnut oil, which had a long history of
use for making vanaspati in India, though at present it is not permitted, the figure is 280 m^
of hydrogen [22].
4. M onitoring o f H ydrogenation
Hydrogen uptake during the hardening of any specific oil is the main criterion by which the
endpoint is judged, and this is confirmed by determination toward the end of the run of the
capillary slip point on a sample that is withdrawn from the reactor and chilled in ice [22-25].
The filtration rate of the product after the hydrogenation run is over is also a pointer to any
major deviation from normal in the reactor. During a run, the iodine value, refractive index,
and solid fat index are not generally employed for monitoring but are determined on the
finished vanaspati product.
5. Catalysts
Up to a decade ago it was usual to generate the nickel catalyst from its salts in the vanaspati
factory [21, Chap. 10]. Dry-reduced catalysts were prepared by first precipitating nickel hy
droxide or carbonate on diatomaceous earth or other refractory support material and then
reducing the ground, dried material to the metal in a current of hydrogen. To make wet-
reduced catalysts, nickel was first converted into its formate; this was then easily reduced in
a vegetable oil medium at a relatively low temperature to give carbon dioxide, hydrogen,
water, and the metal, which again was supported on siliceous material. At present, commer
cial catalysts take the form of pastilles or free-flowing flakes that contain 18-26% reduced
nickel metal precipitated on a siliceous or silica-alumina support and suspended in almost
fully hydrogenated soybean oil. In general, oil hydrogenation nickel catalysts are designed to
exhibit a high degree of selectivity toward linoleate over oleate during hydrogenation and a
high capacity to generate trans fatty acids. Use of a high catalyst concentration also favors
selectivity as well as a high temperature and low hydrogen pressure. For 1 t of oil, 0.5-1.0
kg of catalyst pellets are commonly employed.
Oils that are solvent-extracted from screw press oilcakes after storage, in particular from
rape-mustard oilcakes, are known to carry catalyst poisons. Special catalysts have been devel
oped for such oils that have sufficient active sites still left for the absorption and hydrogena
tion of unsaturated fatty acids even after a proportion of them have been nullified by catalyst
poisons. A higher concentration of 2 kg of catalyst per tonne of oil is also generally employed.
Nickel catalyst is expensive, and nickel resources in India are scarce. It is therefore cus
tomary to use, along with fresh catalyst, 1 0 - 1 5 % of once-used catalyst. Too high a proportion
of spent catalyst will produce less saturation and more isomerization and result in products
that, even at the specified melting point, will show considerable separation of liquid glycer
ides. In practice, the catalyst from, say, two plates of a 20 plate filterpress can be set aside
and the rest used again with fresh make-up catalyst [23].
Hydrogen uptake with time during a hydrogenation run under set conditions is employed
as a sensitive measure not only of the activity of the catalyst being used but also of the
proportion of fresh catalyst that will be needed for the next run [22]. Yet another indicator of
catalyst activity during a hydrogenation run is the filtration rate of the hardened product in the
filter press. Too high a rate implies an excessively liquid oil, and too low a rate, oversatura
tion. This must be corrected when various individually hydrogenated components are finally
blended to vanaspati [22-29].
6. Oils E m ployed fo r H ydrogenation

Groundnut oil was for decades the single base stock for manufacture of vanaspati in India.
Though over a dozen vegetable oils are now permitted for conversion to vanaspati, as noted
earlier, the belief in the industry is that no single oil other than groundnut can yield a satisfac
tory vanaspati product.
Each oil presents special problems and needs an appropriate approach [22-25].
When soybean oil is to be hardened, the level of gums present through insufficient degum-
ming in the refinery is important, because they are powerful catalyst poisons. Degumming
may therefore have to be carried out using dilute acids in the vanaspati factory before the
hydrogenation of soybean oil. During the hardening operation, it is often the practice to start
the hydrogenation at a lower temperature, 140°C, so as to selectively reduce the linolenic acid
component, which averages about 5% in India-grown soybean oils, before raising the tempera
ture to the usual 180°C. Postoperation steps of alkali neutralization, bleaching, and deodoriza-
tion of hardened soybean oil also call for stronger measures than for groundnut oil.
Cottonseed oil is offered in India both as a fully refined and bleached product and as a
“washed” oil. In the latter, most of the free fatty acids present in the crude oil, but only part
of the coloring matter, have been removed by washing with dilute lyes at slightly elevated
temperatures. Frequently, washed cottonseed oils are color-fixed, and even after alkali-refin
ing, bleaching, and deodorization in the vanaspati factory, the product obtained after harden
ing will tend to be darker in color. Use of higher than usual levels of bleaching earths and
active carbons at all stages is fairly routine.
Ricebran oil also tends to have fixed red colors. If crude oils are procured, the free fatty
acid level is frequently of the order of 4 % , and such oils need degumming, chilling, alkali
refining, and bleaching before hardening. Even refined ricebran oils when procured tend to
carry gums and waxes and need to be thoroughly degummed at the vanaspati factory. Some
factories treat the oil with spent nickel catalysts to remove catalyst poisons prior to hardening.
Others recommend postbleaching of the hydrogenated product with, say, 1 % of bleaching
earth and 0 .3 % of active carbon to reduce the red color component.
Rape-mustard oil that has been solvent-extracted from screw press oilcakes, frequently
after some delay, is known to contain catalyst poisons. Some factories treat the oil with spent
catalyst prior to hardening, and others deodorize it at 140°C to reduce the content of volatile
sulfur components. Hydrogenation of solvent-extracted rape-mustard oil is recognized as be
380 Achaya
ing sluggish, with little generation of exothermic heat during hardening. Not more than 50%
of hardened rapeseed oil is employed as a vanaspati component.
Palm oil and palmolein require only slight hardening. This is performed at low tempera
tures, using mainly spent catalyst to slow down the hydrogenation and avoid excessive satu
ration.
Different oils are never hydrogenated together. To have better control of the process, each
is hydrogenated separately to a vanaspati specification end product, followed by blending.
Only if an oil component is very minor will it be incorporated in a major component before
the latter is hardened.
7. Generation o f H ydrogen
Earlier it was common practice to generate hydrogen in vanaspati factories by the steam iron
process [21, Chap. 10], but this has largely been given up. One large vanaspati plant in India
uses hydrogen from caustic chlorine cells, but a majority use hydrogen from alkaline electro
lytic cells [30, 31]. These can be worked at capacities varying from 10 to 100% merely by
varying the voltage, which is very convenient. Two types of cells are in use. In the unipolar
or tank type, each cell is independent and self-contained; electrodes are suspended in the
electrolyte in an iron tank, and alternate electrodes are connected in parallel. It requires large
currents at low voltages, in addition to massive transformers, rectifiers, and copper bus bars.
In the bipolar or filter press type, a series of flat, platelike cells are stacked together and
connected in series, with a voltage supply at either end. Operating pressures vary from 5 to
35 kg/cm^, so gas compressors can be dispensed with. The electrolyte solution in both types
of cells is 25-30% aqueous sodium hydroxide; this requires no replacement, and only the
concentration of water needs to be automatically made up as it is consumed. The purity of
electrolytic hydrogen gas is over 99.5%; the hydrogen is scrubbed to remove droplets of alkali
and compressed for storage in high pressure cylinders. The by-product oxygen, produced in
half the volume of the hydrogen, is compressed into cylinders and finds a ready outlet in
welding operations.
At 2500 A and with the electrodes at 2 V, such a cell can produce 1.142 m^ of hydrogen
and 0.571 m^ of oxygen per hour [30].
8. B leaching Earths and Active Carbons

Certain natural silicoaluminates, when suitably treated with acid, heat, and steam, become
both porous and charged and are capable of adsorbing the coloring matters present in fatty
materials [21, Chap. 9]. Special types are made in India for specific oils (depending on the
nature of the pigments present) and for vanaspati. High capacity for adsorption of pigments,
low fat absorption, easy filterability, and economy in use are the qualities looked for.
Active carbons are adsorbents prepared in India from a wide range of carbonaceous materi
als by treatment at elevated temperatures with air, steam, or metal salts. They are particularly
effective in removing the red color component of fats as well as picking up both traces of
soap and the earthy or musty flavors imparted to the oil by bleaching earth. Frequently they
are mixed in about 10% concentration with the earth, as they cause a loss of far more oil by
adsorption than do the latter [21, Chap. 9].
Postbleaching of vanaspati is carried out in cylindrical mild steel vessels. Steam coils raise
the temperature of the fat from 50 to 100°C in about 30 min in the presence of about 0.5%
of the earth-carbon mixture, with stirring at 80-120 rpm under 700 mm pressure. Filtration
thereafter is in plate-and-frame or metal leaf filters using heavy duty filter cloths, polyester
weaves, or stainless steel wire gauze [32].
Ghee, Vanaspatl, and Special Fats in India 381
9. By-Products o f Vanaspati M anufacture

A small quantity of soapstock is produced during the postneutralization of vanaspati. This is
acidified to yield acid oil, which finds a ready outlet as a hard fat component in soapmaking.
Spent nickel catalyst finally discarded from the filter presses carries between 4 and 14% nickel
held in a mass of fat (45-85%) and siliceous matter. Batches of discarded catalyst are pooled
and transported to catalyst recovery units in the country where the fat is extracted away and
the nickel is leached out with acid. Salt solutions are then reduced to the metal in a current
of hydrogen for deposition on siliceous supports suspended in a hardened fat. The fatty matter
carried over in vanaspati deodorizer distillates may be recovered and added to soapstocks.
By-product oxygen is compressed and sold.
10. Energy C onservation

Fuel and electricity consumption in Indian vanaspati units is far higher than in modem installa
tions elsewhere [33], and installation of heat exchangers at several points in the total system
is on the increase. Thus, hot hardened oil from the autoclave can be dropped into a heat
exchanger to carry the next batch of oil from a temperature of 50-60°C to even 190°C, thus
eliminating the need for heat prior to hardening [34].
There is a trend for new vanaspati units to use the Buss loop reactor [34]. This employs
two loops in which the processing occurs. In the first, the oil and suspended catalyst are put
through a heat exchanger, and in the second loop, hydrogen in very fine form is distributed
through the suspension. Reaction times are short (80-120 min, compared to up to 12 h in
batch units), and as many as eight batches can be mn in a day. Moreover, post-hardening
operations are conducted separately and do not affect plant capacity. No heat energy has to
be supplied, and hardly any cooling water, because of excellent heat efficiency. The process
is fully automated and even generates some low pressure steam for various uses [34].
11. Effluent Treatm ent

Administration of the Environment (Pollution) Act 1986 in India is the responsibility of cen
tral or state pollution boards [35]. Specifications laid down by the Bureau of Indian Standards
are followed, the chief requirements of which are shown in Table 1 along with data for
vanaspati plant effluents.
Vanaspati factory effluents are treated by employing physical, chemical, and biological meth
ods [36]. Physical methods include vibratory and rotary screening, sedimentation of solids that
are capable of settling or floating, flotation using air bubbles, and filtration by gravity or under
Table 1 Specifications for Wash Waters and for Effluents from Vanaspati Units
Wash water for discharge into
Effluent from
Inland Land for Public vanaspati
water irrigation sewers factories
pH value 5.5-9.0 5.5-9.0 5.5-9.0 2 -3

Suspended solids, mg/L 100 max 200 max 600 max 5000-25,000
Oil and grease, mg/L 10 max 10 max 20 max 500 -1 0 ,0 0 0
Biochemical oxygen demand
(5 days at 20 °C), mg/L 30 max 100 max 350 max 2000-15,000
Chemical oxygen demand, mg/L 250 — — 5000-20,000
382 Âchaya
pressure. Chemicals used are lime, alum, ferrie chloride, ferrous sulfate, and anionic or cationic
polymers. Biological treatment involves the use of living bacteria in activated sludges.
In practice, the effluent is mixed with dosed quantities of chemicals in a flash mixer, with
continuous removal of the dregs that are formed. The liquid passes into collecting tanks and
thence for dewatering to drying beds or filter presses. The clear filtrate that overflows the weir is
treated with the chemicals earlier noted, which greatly reduce its total solids content as well as
its biological oxygen demand (BOD) and chemical oxygen demand. For a further 9 0 % reduction
in BOD, the liquid is brought into contact in a mechanically aerated reactor containing nutrients
with the aerobic bacterial culture. The treated effluent is fit for discharge, while the sludge can
be dewatered in drying beds, filter presses, or centrifuges for use as a landfill [36].
Few vanaspati units now employ the full complement of treatment, which requires consid
erable capital outlay and physical space, but will eventually have to fall in line.
C. Product Characteristics and Composition

7. Selectivity C oncepts D uring H ydrogenation
Selectivity concepts have been dealt with elsewhere in this volume and are therefore only
outlined here in relation to vanaspati production.
When a cis double bond in an oil is adsorbed during the hydrogenation process at an active
center on the catalyst surface and encounters a hydrogen atom also dissociated at an active cata
lyst center, it can become saturated and then desorbed. However, if the hydrogen pressure is
low, the reaction temperature high, and the quantity of catalyst also high, it is more likely to be
desorbed with the double bond still intact but largely twisted into the trans configuration. Prefer
ential adsorption on the catalyst of linoleate, with two cis double bonds, and partial hydrogena
tion leads to both trans- and cw-monoenes in which the double bond has shifted along the chain
as a result of the formation of conjugated diene intermediates. Thus selective hydrogenation of
an oil containing linoleate will lead to an accumulation of tran^-monoenes with double bonds
scattered from carbons 6 to 16. These have melting points of around 44°C, and in effect the oil
has been partially saturated and hardened. Some dienes may also be desorbed without saturation,
and these could also carry trans double bonds. Fortuitously, a selectively hydrogenated fat high
in trans-monotms can be made to physically resemble ghee, the prized culinary fat of India, and
early promotion of vanaspati was based on this circumstance.
2. Crystallization B ehavior and G ranularity

The ghee-like grain structure desired by consumers is achieved by careful cooling of vanaspati
[20]. As the hardened product cools from, say, 50°C, the rra/i^-glycerides in the liquid phase
become supersaturated and nucléation commences; there is considerable evolution of heat,
which can even cause remelting if external cooling is insufficient. The large saturated glycer
ide crystals give the product its granular texture, and the large number of trans nuclei, its
firmness. Photomicrographs of vanaspati [37] show large granules consisting of radially ar
ranged needle-like crystals. Such a crystallization network has enough surface area to firmly
entrap the liquid phase and prevent it from separating out at ambient temperatures as an oily
layer on top.
3. N ature and Properties o f Vanaspati

Detailed studies on 20 samples of Indian vanaspati were carried out at the Palm Oil Research
Institute of Malaysia [37]. The results, shown in Table 2, reflect the nature of the products
that were being made in the early 1980s. The vanaspati being made at present uses much the
Table 2 Characteristics of an Average Indian Vanaspati
Standard
Parameter Average dev
Capillary slip melting point, °C 35.6 1.78

Iodine value 76.5 4.31
Yield value, kg/cm^
15°C 7.4 1.5
12°C 3.9 0.98
Solid fat content, %
10°C 75.1 7.63
15°C 64.6 6.86
20°C 52.6 7.51
25°C 37.1 6.98
30°C 20.7 5.80
35°C 6.8 3.79
Fatty acids, wt %
16:0 14.8 2.62
18:0 8.1 2.27
18:1 t 43.2 3.89
18:1 c 18.8 2.56
18:2 r,r 5.0 1.80
18:2 c,i 3.5 1.25
18:2 i ,c -1-20.0 1.3 1.40
18:2 c,c 3.4 1.16
Total trans acids 53.7 4.29
Total saturated acids 23.8 3.00
Total unsaturated acids 76.2 2.92
Source: Based on analysis of 20 vanaspati products, early 1980s [37].
same bulk raw materials (soybean, cottonseed, and rape-mustard oils; only ricebran oil is used
in larger proportions now) and should not be vastly different.
The capillary slip melting point averaged 35.6°C with a small standard deviation of 1.7 8 °C
and a coefficient of variation of only 5 % . The specification of a maximum of 37°C prevalent
at that time was thus being strictly observed. So firmly is this limit ingrained that it still tends
to be followed in vanaspati manufacture though the legal upper limit has been raised to 4 1°C
to permit manufacture of harder products that will not be totally liquid during the hot summer
months. The best correlation of slip melting point was with total saturated acids (0.71). The
iodine value averaged 76.5, again with only a 5 .6 % coefficient of variation.
Indian vanaspati exhibited a steep solid fat content profile, as shown in Fig. 3 [37]. This
is a well-known consequence of a high content of trans fatty acids (here averaging 53.7%)
coupled with low saturated and cw-unsaturated fatty acid levels [37]. Any increase in saturated
acids would tend to push the melting point of the product outside the permitted range, so that
linoleate selectivity is imperative during hardening. This is in contrast to vanaspati products
made in neighboring Pakistan [37], which showed higher melting points (mean 37.2°) and a
softer consistency as a result of higher contents of saturated and cw-unsaturated acids and
considerably lower levels of trans acids (mean 27.4%).
The solidification curve at 20°C determined by wide-line NMR (Fig. 4) shows that Indian
vanaspati crystallizes rapidly and attains a high final solids content of about 35% [37]. This
would make for reduced oiling out at higher ambient temperatures.
384 Achaya
Temperatiire (°C)
Fig. 3 Solid fat content profile (95% confidence level) of Indian vanaspati products. (From Ref. 37.)
4. Sesam e Oil as a L atent Color

To deter use of vanaspati as an adulterant of ghee, early legislation specified that not less than
5 % of sesame oil, whether raw or refined, be finally blended into vanaspati, with the further
proviso of a minimum color of 2 red units in the Baudouin test carried out under specified
conditions. Since raw sesame oil would give a markedly higher response than the refined
product, much smaller quantities would serve to answer the color test. At present, only the
addition of refined sesame oil is permitted; the quantity to be used is not specified, and only
a color response of 2 red units is laid down in the Baudouin test. It is usual to use a small
overage of sesame oil.
Considerable scientific effort has been expended to find other possible markers for vanas
pati [5, pp. 299-307]. These include (1) various natural pigments like ratanjot, curcumin,
kamala dye, chlorophyll, copper pheophytin, alkanet, annatto, several carotenoids, and lac-
toflavin; (2) synthetic and coal tar dyestuffs (yellow, orange, red, blue); and (3) latent colors
such as fluorescent dyes, tyrosine ethyl ester, quinine, vitamin K, alkylated dihydroxydiphe-
nylamine, acetylated methylene quinine, and butylated hydroxyanisole [5], None of these has
been adopted or stood the test of time.
•I
!2
i-
Fig. 4 NMR solidification curve of an Indian vanaspati. (From Ref. 37.)
5. A ddition o f Vitamin A
The compulsory addition of 25 lU of synthetic vitamin A to vanaspati in May 1951 was
intended to equate the product in this respect with ghee. Observations show that half or more
of the vitamin is lost on storage of vanaspati for 6 months whatever the container or tempera
ture [38]. Further, vanaspati is used mostly as a frying medium at high temperature, which
would doubtless destroy any surviving vitamin A.
6. N ickel Contents
Though once thought to be innocuous, nickel is now regarded as an essential nutrient for
humans [39]. Little of it is absorbed from the diet, and impossibly high levels of 100 mg/100
g body weight, or, say, 5 g of nickel a day, are needed to produce deleterious effects. The
daily intake of nickel from customary Indian diets has been estimated at 300-600 ¡ng [39].
Commercial vanaspati samples varied in nickel content from 0.30 to 4.54, average 1.55
ppm [39]. Since the higher contents could reflect poor manufacturing practice, the legal speci
fication was recently fixed at 1.55 ppm (max) [4].
386 Achaya
IV. BAKERY FATS

A. Definition
Even the legal definition of a bakery shortening, in the Prevention of Food Adulteration Rules
1955 as it currently stands, states that it “means vanaspati conforming to the standards pre
scribed,” with certain qualifications. These are (1) a capillary slip melting point of 4TC
(max), (2) a content of 12% (max) by volume of nitrogen, air, or any inert gas where aera
tion has been employed, and (3) optional use of mono- and diglycerides as emulsifying
agents [4].
B. Production Parameters
1. E arlier Technologies
The aim at first was simply to convert a hardened vanaspati product from the regular produc
tion unit into a material that was smooth, fine, homogeneous, and nongrainy. In the early
stages, this was achieved by chilling the hardened fat very rapidly as it emerged from the
deodorizer. When the product had become semisolid, but still pourable and pumpable, it was
filled into tins, and no further cooling was applied [20]. In a later development, the vanaspati
was cooled continuously on the surface of slowly revolving (10 rpm) hollow rolls chilled
internally with brine or ammonia. Fat was scraped from these rolls to fall into an open trough
in which a conveyor or paddle provided with blades beat air into the product as it passed in
semisolid form into the filling machine [20]. In 1968, some 38,000 t of such all-purpose
bakery fat was produced in India [21, Chap. 10].
2. C urrent Processes
Basis. The present practice in India is to manufacture three types of hardened products and
then blend and sometimes aerate them to yield bakery fats suited to specific end uses. The three
hydrogenated fats have melting points of 28, 36, and 42°C, and in blending and plasticizing
them the critical factors in achieving the desired product characteristics are (1) the components
and composition of the blend, (2) the chilling conditions, (3) whether aeration is employed, and
(4) the filling temperature [26]. It is usual in Europe to adopt a common temperature of 20°C for
the operations of filling, transport, and storage, but in the warmer Indian ambience this common
temperature is 30°C, which would in turn influence product characteristics.
Hardening. Although all the edible oils used in the manufacture of vanaspati can be
employed to make the components of bakery shortenings, experience has led to certain prefer
ences for each of the three hardened products. For the product of melting point 28°C, soybean
oil is the choice; for that of melting point 36°C , cottonseed, sunflower, and soybean oils; and
for the product of melting point 42°C, ricebran oil [23]. In preparing the product melting at
28°C, a high proportion of used catalyst and a low hydrogenation temperature of 160-170°C
appear to be preferred [23].
Blending and Plasticization. Blending the three hardened fats in different proportions
yields various types of bakery fats. Thus the preferred solids content index at 20°C for table
margarines in India lies between 4 and 20%; for cake shortenings, 20-40% ; and for puff
pastry, 60%; so that an ever greater proportion of the hardest component will be required for
each product type [40].
The right amounts of fats and the permitted additives are mixed in a heated and stirred
vessel. The next steps of converting this mixture into plastic bakery fats is described in Chap
ter 11. In brief this involves its passage through various arrangements of scraped and chilled
open tubes to achieve very fine crystallization. The chilled end products are variously pack
aged in India: in tins of 0.5, 1, 2, and 15 kg; in polybags of 1 and 5 kg; in polyjars of 1, 5,
10, and 15 kg; and in bags-in-boxes of 15 kg [26].
Products Manufactured. Nonaerated, high melting, but plastic products go into the manu
facture of puff pastry and patties and also that of a popular, highly layered, crisp and eggless
salted snack called Bombay khari. Aerated products of high creaming power are chosen for
making crisp cookies, wafers, and cream biscuits. Margarines themselves are employed in
considerable quantity in India to make cakes and cream fillings [26].
The total production of bakery fats in India is on the order of 20,000 t annually, and the
quantity is rising fast as commercial baking expands.
Related fat-based products made for the baking industry are emulsifying agents like glyc
erol monostearate, mixtures of mono- and diglycerides, and pan-releasing agents employed in
baking bread [26].
V. MARGARINES AND FAT SPREADS

Ae Background
Butter was too perishable a product to gain widespread traditional use in India, and it was
mostly rendered into ghee. Packaged table butter is now made in organized dairies only to the
extent of about 30,000 t annually. Fat spreads are in concept and practice substitutes for table
butter; they have only recently made a beginning in India, with production in 1992-1993 of
about 2000 t of margarine and 5000 t of table spreads [4,27]. The distinction is one of water
content; the maximum permissible in a margarine is 16%, and if greater than this the product
qualifies as a table fat spread [36].
B. Specifications
Three types of fat spreads are recognized by the food laws of India [4]. A milk fat spread
should have only that commodity as the fat component. In a mixed fat spread, the fat should
consist of a mixture of milk fat with one or more hydrogenated or unhydrogenated refined
edible vegetable oils or interesterified oils. A vegetable fat spread will carry only the latter.
The fat content has to be declared on the label, and in a mixed fat spread the content of milk
fat must also be declared. The word “butter” should not be associated with the labeling. The
fat content should lie between 80 and 40% by weight, and the moisture content between 16
and 56%. The fat extracted from a vegetable fat spread should have a melting point of T C
(max), an acid value of 1.5% (max), and a content of unsaponifiable matter of 1.5% (max);
the latter value is 1.0% (max) for fat extracted from a milk fat spread or a mixed fat spread.
The vegetable fat spread should contain 25 lU per gram of synthetic vitamin A when packed
and show a positive response to the Carr-Price antimony trichloride test. A vegetable fat
spread should contain sufficient raw or refined sesame oil to answer the same Baudouin color
test that is prescribed for vanaspati. The package size of any fat spread should not exceed 500
g, and loose sale is prohibited.
Certain additives are permitted but not mandatory. These include (1) common salt, 2% by
weight (max); (2) the solids-not-fat of milk; (3) starch at levels of 100-150 ppm; (4) diacetyl
at a 4 ppm (max) level; (5) BHA and TBHQ as antioxidants at a level of 0.02% (max) on the
fat content; (6) sorbic acid (as its sodium, potassium, and calcium salts) and benzoic acid (as
388 Achaya
its sodium and potassium salts), singly or in combination, at a 1000 ppm (max) level; and (7)
annatto and/or carotene as coloring agents.
One commercial milk fat spread contains 60% fat, of which one-sixth is butter. It is de
signed to be spreadable when taken out of the refrigerator and has a solids content index of
30 at 10°C, 18 at 20°C, 2.5 at 30°C, and nil at 40°C, which are all higher than for a similar
product in colder countries [27]. Two commercial table margarines are in production, and a
low calorie diet margarine is stated to be in development [29].
C. Manufacture
1. F at C om ponent
The fat component of a margarine or a vegetable fat spread is a blend of refined and hardened
edible vegetable oils [4,27-29]. A mixed fat spread will also contain milk fat. The refined
oils that can be used cover a wide range, and in India they chiefly include sunflower, soybean,
cottonseed, and sesame oils. The hydrogenated oil component could include those made from
soybean, cottonseed, ricebran, and palmolein. Soybean oil that has been hydrogenated to two
different degrees of hardness could be employed. To achieve the desired smooth texture and
physical stability, one component of the hydrogenated oil mix should have a B' crystal habit,
with considerable palmitic acid in the 2-position; these include in India cottonseed, palm, and
rape-mustard oils [27] (see Chapter 11).
Selective hydrogenation conditions are employed in hardening the vegetable oils. However,
the pressure and temperature used may be slightly lower than for a vanaspati-type hydrogena
tion, and the catalyst employed could contain a higher proportion of used catalyst.
A table margarine in India could well be a blend of three parts of liquid oil and two parts
of the solid component [40].
Fig. 5 Packing section o f an Indian mixed fat spread unit. (Courtesy o f National Dairy Development
Board, Anand.)
2. Em ulsification
The emulsion is prepared in two stages [40]. A coarse water-in-oil emulsion is first formed
by adding pasteurized salted water or milk containing the water-soluble ingredients to the fat
blend carrying the emulsifiers and fat-soluble ingredients under strong agitation at 45°C . This
is cooled and then homogenized by passing the mass under high pressure through fine orifices
to yield a fine, uniform emulsion. A part of the fatty matter crystallizes around the aqueous
droplets in the form of a fine network and keeps them from coalescing.
The next step of chilling and working the emulsion is carried out in the annular space between
a shaft and a chilled insulated outer jacket at very high speed to achieve nucléation (see Chapter
11). In the second half of the unit, the mass is worked vigorously to complete crystal formation,
followed by a resting phase to achieve phase stabilization and the stiff consistency needed for
packaging. Such units are manufactured in India [28]. The extruded margarine or fat spread is
packed in plastic tubs (Fig. 5), frequently of 100 g size for convenience of domestic use.
REFERENCES
K. T. Achaya, Oilseeds and Oilmilling in India: A Cultural and Historical Survey, Oxford and
IBH Publishing, New Delhi, 1960, p. 157.
2. S. L. Rao (Ed.), Consumer Market Demographics in India, rev. ed., National Council of Applied
Economic Research, New Delhi, 1994, Chap. 23, pp. 268-272.
Anon., Ghee in India: Production, Statewise Consumption and Product Profile, Gujarat Coopera
tive Milk Marketing Federation, Anand, 1993.
4. M. K. Kundu, Orders and production data. Directorate of Vanaspati, Vegetable Oils and Fats,
Ministry of Civil Supplies, Consumer Affairs and Public Distribution, Government of India, New
Delhi, Personal communication, February 1995.
5. K. S. Rangappa and K. T. Achaya, Indian Dairy Products, Asia Publ. House, Bombay, 1974.
6. R. P. Aneja, Dairy process and product development under Operation Flood, Indian J. Dairy Sci.
42: 53 (1991).
7. C. Arumughan and K. M. Narayanan, Grain formation in ghee (butterfat) as related to structure
of triglycerides, i . Food Sci. Technol. 16: 242 (1979).
8. C. Arumughan and K. M. Narayanan, Triacylglycerol composition of buffalo milk fat, J. Dairy
Res. 49: 81 (1982).
9. J. N. Warner, Principles of Dairy Processing, Wiley Eastern, New Delhi, 1976, Chap. 13, pp.
253-259.
10. R. C. Misra and N. S. Kushwaha, Studies in methods of preparation of ghee, Indian J. Dairy Sci.
23: 115 (1970).
11. S. De, Outlines of Dairy Technology, Oxford Univ. Press, Delhi, 1980, pp. 43 3 -4 6 6 .
12. K. L. Gaba, Recent trends in rural marketing, packing and trading of ghee, Indian Dairyman 32:
783 (1980).
13. J. S. Punjrath, Indigenous milk products of India, Indian Dairyman 42: 15 (1991).
14. B. Bahadur, M. R. Srinivasan, and S. C. Ray, Utilisation of induced stratification of butter melt
in ghee making, Indian J. Dairy Sci. 3: 94 (1950).
15. S. C. Ray and M. R. Srinivasan, Pre-Stratification Method of Ghee Making, Indian Council of
Agricultural Research, New Delhi, Res. Ser. No. 8, 1976.
16. H. Abhichandani, S. P. Agrawala, R. D. Verma, and B. S. Sector, Advances in continuous ghee
making, Indian Dairyman 30: 769 (1978).
17. K. T. Achaya, The Food Industries of British India, Oxford Univ. Press, New Delhi, 1994, Chap.
12, pp. 211-218.
18. A. C. Chhatrapati, The Vanaspati Industry: A Historical Review, Popular Prakashan, Bombay,
1985, pp. 9 -2 1 .
390 Achaya
19. Vanaspati Manufacturers Association of India, New Delhi, Statistical data, 1994.
20 . G. S. Hattiangdi, The Vanaspati Industry, Indian Central Oilseeds Committee, Hyderabad, 1958,
Chap. 3, pp. 18-32.
21 . K. S. Murti and K. T. Achaya, Cottonseed Chemistry and Technology, Publications and Informa
tion Directorate, Council of Scientific and Industrial Research, New Delhi, 1975, repr. 1984.
22 . T. N. Murthi, National Dairy Development Board, Anand, Personal communication, 1994.
23. K. Chandrasekaran, Brooke Bond Lipton India Ltd., Bangalore, Personal communication, 1994.
24. R. S. Vaidyanathan, Rasoi Ltd., Calcutta, Personal communication, 1994.
25. K. S. Holla, Tata Oil Mills Co. Ltd., Bombay, Personal communication, 1994.
26. S. Gupta and R. Somnath, Masterline Bakery Service, Brooke Bond Lipton India Ltd., Bangalore,
Personal communication, 1994.
27. A. P. Mahadevan, National Dairy Development Board, Anand, Personal communication, 1994.
28. J. K. Khaitan, Amrit Banaspati Co. Ltd., Rajpura, Personal communication, 1994.
29. M. M. Chakrabarty, Calcutta, Personal communication, 1994.
30. S. C. Jain and O. P. Narula, Hydrogen through water electrolysis, in Treatise on Fats, Fatty Acids
and Oleochemicals (O. P. Narula, Gen. Ed.), Industrial Consultants (India), New Delhi, 1994,
Chap. I-l.
3 1. A. K. Mehra and R. Viswanath, Hydrogen production processes— a review, in Treatise on Fats,
Fatty Acids and Oleochemicals (O. P. Narula, Gen. Ed.), Industrial Consultants (India), New
Delhi, 1994, Chap. 1-3.
32. O. P. Narula, Economics and practice of oil bleaching, in Treatise on Fats, Fatty Acids and
Oleochemicals (O. P. Narula, Gen. Ed.), Industrial Consultants (India), New Delhi, 1994, Chap.
F-3.
33. R. K. Khullar, Energy conservation in vanaspati industry, in Treatise on Fats, Fatty Acids and
Oleochemicals (O. P. Narula, Gen. Ed.), Industrial Consultants (India), New Delhi, 1994, Chap.
S-1.
34. 0 . P. Narula and R. K. Chopra, Heat recovery and energy savings in oils and fats plants, in
Treatise on Fats, Fatty Acids and Oleochemicals (O. P. Narula, Gen. Ed.), Industrial Consultants
(India), New Delhi, 1994, Chap. S-2.
35. R. K. Suri and A. Mehrotra, Environmental protection and pollution control in oil processing
industry, in Treatise on Fats, Fatty Acids and Oleochemicals (O. P. Narula, Gen. Ed.), Industrial
Consultants (India), New Delhi, 1994, Chap. T-3.
36. K. C. Mathur and Surinder Kumar, Effluent treatment in oil processing industry, in Treatise on
Fats, Fatty Acids and Oleochemicals (O. P. Narula, Gen. Ed.), Industrial Consultants (India),
New Delhi, 1994, Chap. T-4.
37. M. S. A. Kheiri, A Survey of Indian and Pakistani Vanaspati Products, Palm Oil Res. Inst, of
Malaysia, Kuala Lumpur, August 1982.
38. N. Nasir and K. V. Nagaraja, Loss of vitamin A in vanaspati under different storage conditions,
J. Oil TechnoL Assoc. India 19:44 (1987).
39. 1. S. Shenolikar, Nickel content of vanaspati, J. Oil TechnoL Assoc. India 23:48 (1991).
40. T. K. Roy and S. K. Banerjee, Margarine manufacture— technology and plant design, in Treatise
on Fats, Fatty Acids and Oleochemicals (O. P. Narula, Gen. Ed.), Industrial Consultants (India),
New Delhi, 1994, Chap. N-1.
15________________________________________
Chocolate and Confectionery Fa ts
Fred B. Padley
Bedford, Bedfordshire, England, Consultant, Coders Croklaan bv., Division
o f Quest International, Wormerveer, The Netherlands
I. INTRODUCTION
Fats are widely used in the confectionery industry mainly to convey specific textural attributes
and as a vehicle for fat-soluble flavors. The application that has received most scientific atten
tion has been in the area of chocolate and chocolate substitutes for ambient confectionery.
The major part of this chapter therefore focuses on the use of fats in chocolate, although
mention is made of other relevant applications including cream fillings, toffee, and ice
cream couvertures.
It is interesting to note that of all the edible oils and fats, cocoa butter has probably
received the most detailed examination, no doubt because it combines scientific interests in
relating physical behavior with chemical composition. For this reason cocoa butter, cocoa
butter equivalents, and substitute fats are discussed in some detail in this chapter, but first a
short description of the major confectionery product, chocolate.
II. CHOCOLATE
A. Chocolate Composition and Manufacture
The main component of chocolate is derived from the cocoa bean, whose continuing attraction
is due to its flavor and melting characteristics. The Aztecs of South America created a spiced
drink from the cocoa bean called chocolatl and believed it to have aphrodisiac properties, an
image continued today with the romantic associations created by advertisers of chocolate.
The extraction of cocoa butter from the cocoa bean and the formulation of chocolate was
first carried out in Holland by Van Houten in 1828. Milk chocolate was developed later by
Peters of Vevey in Switzerland in 1876. These innovations subsequently led to the develop
ment of a major industry throughout the world including the grinding of 2.5 x 10^ tonnes of
cocoa beans and the production of approximately 4.5 x 10^ tonnes of chocolate in 1995 [1].
391
392 Padfey
Chocolate, as defined within the EU [2] is “the product obtained from cocoa nib, mass,
powder and sucrose with or without added cocoa butter, having a minimum dry cocoa solids
content of 35% and at least 14% of dry non-fat cocoa solids and 18% of cocoa butter.” Other
types of chocolate, including plain chocolate and milk chocolate, are also defined. For exam
ple, milk chocolate, among other requirements, must contain
A minimum of 25% total dry cocoa solids including at least 2.5% of dry non-fat cocoa
solids
At least 14% of milk solids obtained by evaporation, including at least 3.5% of butterfat
Not more than 55% sucrose
At least 25% fat
This is, however, a simplification of what chocolate can contain, and the reader is recom
mended to read the appropriate legislation. Chocolate is essentially a mixture of finely ground
cocoa beans and sugar plus milk powder in the case of milk chocolate. However, the use of
only the cocoa beans with sugar creates a dry, powdery product because there is insufficient
fat in the mixture to create a truly fat-continuous product. To create a product with the right
texture and mouthfeel it is essential to add fat. This additional fat is obtained from additional
cocoa beans by either pressing (expelling) or extraction. The addition of extra cocoa butter at
elevated temperatures creates a fluid mass that can be used for molding or enrobing, the cocoa
butter being essentially solid at ambient temperature.
Some typical chocolate compositions are given in Table 1. The residual cocoa powder left
after fat extraction is used to formulate chocolate or cocoa drinks much lower in fat than those
prepared by the Aztecs. Cocoa powder is widely used in formulating chocolate substitutes in
which the expensive cocoa butter is replaced by cheaper sharp melting fats (see later).
Chocolate technology has been described in a number of textbooks, notably those of Mini-
Table 1 Some Chocolate Formulations
Milk chocolate
High Average Low milk crumb Dark chocolate

(%) (%) (%) (%) (%)
Cocoa mass 14.00 11.78 10 12.00 39.62

Whole milk powder"^ 23.60 19.08 18
Whole milk (roller)
Whole milk (spray)
Enzyme-treated (WMP)
Sugar 46.00 48.73 55 48.08
Added cocoa butter 16.00 19.98 16.50 18.00 11.75
Lecithin 0.40 0.35 0.38 0.35
Milk crumb^ 70.00
Salt 0.06
Vanillin 0.08 0.06 0.14
Total 100.00 100.00 100.00 100.00 100.00
Total fat 30.00 31.50 27.00 31.32 31.1
^Whole milk powder is sometimes substituted with skimmed milk powder and butter oil.
^Milk crumb FCM (Full Cream Milk) solids 37%, sugar 54%, cocoa mass solids 8%, water 1%, total 100%, total
fat 13.32%.
Source: Ref. 139.
Chocolate and Confectionery Fats 393
ROASTED BEANS
[Previously fermented, dried and cleaned]
CRACKING and REMOVAL of SHELL
GRINDING to PRODUCE COCOA MASS
PRESSING CRUMB PROCESS

[Usually of Alkalized Mass] Combine with
TRADITIONAL PROCESS - Milk
Mix with - Sugar/Syrup
-Sugar
PARTIALLY COCOA - Cocoa Butter
DEFATTED BUTTER -[CBEA/egetable Fat]
MASS - Full Fat Milk Powder DEHYDRATE
for Milk Chocolate
- Emulsifier
- other Permitted additions Add
- Cocoa Butter
FURTHER - [CBEA/egetable fat]
- Emulsifier
GRINDING/BLENDING
- Other Permitted additions
COCOA POWDER
[Fat content depends GRINDING/REFINING
on extent of pressing [Particle size reduction
or inclusion of solvent
extraction] i
CONCHING
[Additional c.b/CBE etc.]
TEMPERING, followed by
MOULDING or ENROBING and
FINAL COOLING/CRYSTALLIZATION
Fig. 1 Traditional chocolate processing. (Adapted from Beckett [4].)
fie [3] and Beckett [4]. An overview of the process is given in Fig. 1. The key features of
traditional chocolate making are
Flavor development of beans by fermentation and roasting
Separation of cocoa shell from the bean to give cocoa nib
Grinding cocoa nib to give cocoa mass
Pressing cocoa mass to obtain cocoa butter
Mixing cocoa mass, sugar, cocoa butter (and milk powder for milk chocolate), and emul
sifier
Grinding and conching to reduce particle size to less than approximately 25 /xm
Conching to remove acidic volatile flavors and to modify flavor
394 Padley
Tempering to achieve the appropriate crystal nuclei

Molding or enrobing
Packaging and long-term storage
There are, of course, variants of this, in particular the so-called crumb process, which was
developed in the United Kingdom and is now used in many other countries. In this case whole
milk is dehydrated under vacuum at an elevated temperature in the presence of either sugar to
give white crumb or sugar and cocoa mass to give chocolate crumb. Processing in this way
gives a unique partially caramelized, chocolate flavored product. The crumb is then refined
down to a given particle size. Each stage is critical in providing the customer with an accept
able product. The fat component plays a greater or lesser part at different process stages. The
key aspects of chocolate technology where fat plays a major role are in influencing the viscos
ity of molten and tempered chocolate, the nucléation and crystallization behavior during tem
pering, molding, enrobing, storage, and the dependence on fat crystal stability for adequate
shelf life. Finally the melting behavior of the stabilized fat phase and the overall flavor percep
tion depend to a large extent on the blends of fats used. The wider aspects of chocolate
technology are therefore not covered here, and the reader is referred to more comprehensive
texts as mentioned above. The reader is also referred to specialist papers on chocolate flavor
[5], quality control of chocolate [6], cocoa beans [7], chocolate manufacture [8], and cocoa
powder [9]. The areas where fat has a major role to play are described below.
B. Chocolate Rheology
Chocolate is a dispersion of finely divided sugar, protein, and cocoa particles in a continuous
fat phase. The particle size distribution is normally such that <20% of the particles are greater
than 20 ¡xm to ensure that the product texture is acceptable and does not taste grainy [10,11].
Some typical ranges of particle size found in some commercial chocolates are given in Table
2. Chocolate rheology is normally characterized using a range of instruments including vis
cometers chosen from various types, e.g., falling ball, cone-and-plate, and rotating cylinder.
Tensile strength is also measured using appropriate equipment, e.g., Instron or Bohlin. Mol
ten chocolate is a non-Newtonian fluid, that is, a finite force (yield value) is required before
the chocolate flows, and there is not a linear relationship between shear rate and shear stress.
The interpretation of shear rate-shear stress curves has been studied by a number of workers.
Steiner [12] and NCA [13] proposed that an approach devised by Casson for paint and printing
ink was also very useful for chocolate, and this has been widely adopted by the industry.
Table 2 Particle Size Distribution of a Range of Milk chocolate masse

(lower milk content)
Average residues (% by weight)

Size of sieve
(/am) USA (ref.) UK Mainland Europe
>800 0.003 0.002

>500 0.025 0.015 0.010
>400 0.050 0.030 0.020
>300 0.200 0.090 0.070
>200 0.500 0.450 0.400
>100 10.000 9.000 10.000
Source: Ref. 11.
The Casson equation relates the square roots of shear rate (D ) and shear stress (r) in the
following way:
^= k ^+ k ,V d
where t is the shear stress, D the shear rate, and K q the yield value.
In the case of rotational viscometers, corrections are made depending on the geometry of
the cylinders used [12,13], giving the equation
{ \ + a ) ^ = ~ [(\+ a )iT -2 K ^ ]
where
IWR^ 2W
radius of inner cylinder (r)

a =-
radius of outer cylinder (R)
W = angular velocity
N = revolutions per minute
From this relationship (l/^j)âp lastic viscosity and K q^ = yield value (Casson yield value).
The rheology of molten chocolate is dependent on the relative proportions of fat and dis
persed solid phase (Fig. 2) [14]. It is also very significantly influenced by the addition of
certain emulsifiers or, more correctly in this case, surface-active agents such as lecithin and
polyglycerol esters. Both infiuence viscosity; polyglycerol esters in particular greatly reduce
the yield value. Their effect is explained on the basis that the emulsifier coats the sugar
Fig. 2 The effect of fat content on chocolate rheology. Graphs depict the Casson parameters of two
milk chocolates with 0.25% lecithin. 1, Fine chocolate with 5.7% particles > 2 0 ¡xm; 2, moderately
coarse chocolate with 16% particles > 2 0 ¡xm, (According to Chevalley [14].)
396 Padley
Fig. 3 The effect of surface-active lipids on chocolate depicting plastic flow and yield value at 50°C.
A (A), with 0.3% soy lecithin; B (x), with 0.3% sucrose dipalmitate; C (o), with 0.3% polyglycerol
polyricinoleate; D (o), with 0.8% polyglycerol polyricinoleate. (From Ref. 15.)
particles, preventing flocculation and thus minimizing particle-particle interaction. The influ
ence of emulsiflers on chocolate rheology is demonstrated in Fig. 3 [15,16]. The use of
emulsifiers therefore provides the manufacturer with a method of influencing chocolate viscos
ity in an advantageous way that also allows a reduced fat level to be used to achieve a given
rheology. The reduction in yield value, especially by polyglycerol polyricinoleate, is im
portant in aiding the removal of air during the molding process and in assisting some enrobing
processes. Note that viscosity is recorded in pascal-seconds (1 Pas —10 poise) and yield value
in pascals (1 P a= 10 dyn/cm^).
Some typical viscosities of molten and tempered chocolate are given in Table 3.
C. Chocolate Tempering
The need for a chocolate tempering stage arises from the complex polymorphic behavior of
cocoa butter. Six polymorphs numbered I to VI have been described by Wille and Lutton
[17]. The most stable, forms V and VI, are (3 polymorphs, also designated ^ 2 ^nd jSj, respec
tively [ 18- 20].
Table 3 Some Typical Ranges of Viscosity^ and Yield Value

of Chocolate
Molten (40°C)
Application plastic viscosity (Pa’s) Yield value (Pa)
Molding 3 -12 3 -12

Enrobing 1.5 -2 10 -30
^Viscosity at temper will increase by a factor of up to 2 depending on the fat

content and degree of temper.
Source: Ref. 21.
Of the various confectionery fats used, only cocoa butter and cocoa butter equivalents
exhibit this complex polymorphic behavior and therefore require tempering. Other fats used
in substitute chocolate crystallize into an essentially stable j8' form and do not need tempering.
Products that have a stable melting behavior and retain good texture and surface gloss are
produced when cocoa butter crystallizes in the ^2 form or form V (see Section V). The
tempering process is used to generate a sufficient number of stable fat crystal nuclei onto
which the bulk of the fat phase of the chocolate can crystallize in a form that remains stable
over long periods of time. This is achieved by the tempering process in which the chocolate
is partially crystallized initially in an unstable form. The temperature is then raised sufficiently
to melt out all the unstable nuclei yet retain a small but sufficient level of stable form V, P 2
seed crystals. These promote the formation of form V crystals during the bulk crystallization
of the fat phase when molded or enrobed chocolate is allowed to cool, normally with the
assistance of a cooling tunnel. On the production scale, tempering is normally carried out
using a multistage continuous tempering unit with a minimum of three stages— one for cool
ing, the second to generate fat crystals, and the third to raise the temperature of the chocolate,
melting out all crystals that are not form V nucleators. Batch tempering units are also used
for both production and laboratory-scale work. A small batch tempering kettle designed by
the Leatherhead Food Research Association, England, can be used for experimental work.
The various types of tempering machines are described in detail by Nelson [21].
The amount and type of fat in crystalline form at temper is open to some debate. Some
authors [21] quote 2-4% ; however, others suggest significantly less. In a detailed article
Reade [22] discusses the solidification process of chocolate, relating the number of crystal
nuclei to crystal growth rate and solidification. As little as 0.1-0.2% fat as seed nuclei would
appear to be adequate for good temper. These apparently conflicting statements explain to a
large extent why the viscosity increase at temper is also reported to be either very small or as
much as twice that of molten chocolate.
The state of chocolate temper is normally determined using a tempermeter [6,23,24], which
essentially determines the temperature of a standard weight of chocolate as it crystallizes when
cooled in a controlled way. The Greer tempermeter is one such piece of apparatus. The
interaction of the temperature on the different stages of the tempering unit with the flow rate
of chocolate and chocolate temper has been studied in detail by Cebula et al. [25] using
a laboratory-scale three-stage continuous temper machine. The interrelationship between the
temperature of the first and third stages of a temper unit with the degree of chocolate temper
is shown in Fig. 4. Essentially this demonstrates that a standard degree of chocolate temper
can be achieved under a very wide variety of conditions. For a range of flow rates there is a
range of conditions defined in terms of the temperatures of the different stages. However, it
is only at the more elevated temperature of the final stage that the chocolate will remain in a
relatively stable state of temper, with little change in viscosity, for some time. Tempered
chocolate emerging from the unit at the lower temperatures will continue to crystallize. Infor
mation of this type can be used to relate results obtained by simple batch-scale laboratory
experiments to larger scale continuous procedures.
D. Crystallization During Enrobing and Molding

of Chocolate
An understanding of the role of crystal nuclei and subsequent crystal growth is important in the
ultimate optimization of the tempering and heat transfer/cooling tunnel processes to achieve sta
ble products with good gloss. This is still an area that has been relatively under-researched. For
further discussions on nucléation, see publications by Larsson [19] and Cebula et al. [26].
398 Padley
Cocoa Butter / Butterfat

95/5
Cocoa Butter / Butterfat

75/5
Fig. 4 The effect of fat phase on the continuous tempering performance of chocolate. Results obtained
on a small-scale continuous three-stage unit. For simplicity only the temperatures of the first and third
(last) stage are depicted. (According to Cebula et al. [25].) The “surface” indicates those conditions
giving equivalent chocolate temper.
The role of the tempering process as indicated above is to generate seed crystals that on
subsequent cooling of the semimolten mass promote the crystallization of the “stable” form V
polymorph. The fundamental questions are, How many seed crystals? and What rate of subse
quent cooling can be permitted? To a very large extent the answers are still arrived at empiri
cally by each manufacturer by controlling temperatures and flow rates, etc., such that a satis
factory product emerges from the cooling tunnel. For example, very rapid undercooling of
tempered chocolate may result in an initially very glossy product, but this will be relatively
unstable, leading to loss of gloss and bloom.
A mathematical treatment [22] shows that the final crystal size is approximately 8 to 19
times the size of the original seed crystal assuming that the seed crystal occupies 0.2 to
0.01%, respectively, of the final volume of the solidified phase. To take this analysis further
ideally requires information on the number and size of crystal nuclei and the linear growth
rate of the crystals. However, in one example Reade [22] assumes that 0.01% of the fat is
present as seed crystals with the dimensions in 10~^ fim of 10 x 1 x 1. These would grow by
approximately 19-fold to achieve complete solidification, producing final crystal sizes well
below those that would interfere with gloss and dispersion of light (wavelength of light 5 0 0 -
700 fims X 10“ ^). Again, this analysis has largely ignored the role of other particulate mate
rial in chocolate and its possible involvement with crystallization.
Numerous types of molding and enrobing equipment are available [3,21,27]. The way in
which chocolate is cooled to crystallize the bulk of the fat will depend on the type of chocolate
being used. In general, tempered chocolate is not cooled too quickly during the first stage of
the cooling tunnel. In the case of some of the nontempered substitute chocolates, the cooling
tunnel conditions have to be tailored somewhat differently; e.g., for palmkemel stéarines,
relatively cool inlet temperatures of 12°C are used. Products emerging from the tunnel should
be at a high enough temperature to prevent moisture condensation, which in extreme cases
would promote sugar bloom. Residence times can be on the order of 10-30 min.
Chocolate used in molding processes needs to exhibit good contraction prior to demolding.
This has been studied by a number of workers but is a difficult property to fully characterize
[3]. In modem processing plants, the product is wrapped immediately after it emerges from
the cooling tunnel. This demands that the chocolate be essentially solid and resistant to me
chanical damage. Solid fat content is normally associated with a degree of hardness or firm
ness of a product. For fully stabilized products formulated using similar glyceride types this
almost certainly occurs. However, during the solidification of chocolate this was observed not
to be the case [28]. Simultaneous measurements of solid fat content by NMR and of hardness
using a penetrometer were made on chocolate as it crystallized from the tempered state. This
showed that the hardness and mechanical strength of the chocolate developed at a significantly
later stage after most of the fat had been crystallized (Fig. 5). This is an intriguing observation
when one considers that in addition to the crystallized cocoa butter there is already >60%
dispersed solids in the chocolate. One explanation might be that large random agglomerates
of solid sugar or cocoa particles and fat are initially formed that then are “cemented” together
by the sintering of fat crystals. Further care must be taken during the enrobing process. The
component to be enrobed should be at a sufficiently high temperature to prevent further super
cooling of the chocolate and the crystallization of less stable polymorphic forms.
The only case where this is not adhered to is in ice cream technology where the chocolate
is invariably nontempered. In this case the fat phase crystallizes in one of the less stable (e.g.,
form IV) states, which has a lower melting point and more acceptable mouthfeel when eaten
as a frozen product. An alternative to conventional tempering and cooling to achieve the
stable crystal form of cocoa butter is to use high pressure [29]. However, even in this work.
m
“O
o
CO
Fig» 5 The relationship between solid fat content and hardness of chocolate during crystallization in a
cooling tunnel.(------ )Hardness measured by penetrometry; (----------) solid fat content measured by indi
rect NMR. (According to Cebula and Dilley, private communication, with permission of Loders
Croklaan.)
400 Padley
some stable nuclei were added to the crystallizing fat. High pressure extrusion of ‘solid"
chocolate has also been described and patented [30].
III. CONFECTIONERY FATS

A wide range of confectionery fats are used in the chocolate confectionery industry. However,
for the uninitiated not all that appears to be chocolate is chocolate. For a product to be called
chocolate it has to adhere to the definition as given under the legislation of the particular
country concerned. Other products that do not conform to the legislation will not (should not)
be labeled chocolate. These latter products are subsitute chocolates in which the main fat
phase is based on palmkemel, coconut, or partially hydrogenated fats. The flavor and color is
obtained normally from defatted cocoa powder, but some products, e.g., “supercoatings’’
based on cocoa butter equivalents (GBE) can be formulated using some cocoa mass.
As will be seen from the detail on individual fats, cocoa butter and CBE exhibit a more
complex polymorphic behavior than palmkemel, coconut, or partially hydrogenated fats. For
this reason, genuine chocolate requires a more complex method of crystallization, i.e., a
tempering stage, than the p ' stable chocolate substitute fats, which can be crystallized directly
from the melt. These latter fats therefore provide the advantage of simpler processing and can
also be tailored to have higher melting points more appropriate for warm climates despite the
fact that the product cannot be called chocolate.
Whereas in the past there was a strong commercial interest to find alternatives to cocoa
butter, there is now also, in Europe at least, a strong interest to define milk fat alternatives in
those countries where vegetable fats are legally permitted at a defined low level, normally 5%
in chocolate [2]. The main groups of fats used in chocolate confectionery are summarized in
Table 4. This grouping is made mainly on the basis of the chemistry of the different fats that
are used. The simplest in chemical terms are cocoa butter and cocoa butter equivalents; the
most complex is probably butterfat. Perhaps surprisingly, the difference in chemical composi
tion is also reflected in an inverse way in physical behavior; that is to say, the simpler the
chemical composition the more complex the physical behavior. Each of these main groups is
described in separate sections below.
Table 4 Chocolate Fat Types
Trans hardened
Symmetrical SOS type Laurie oils Butterfat
Cocoa butter Palmkemel (PK) Midfractions and non- and its fractions
fractionated trans
hardened liquid oils
Cocoa butter fractions Stearins
Cocoa butter equivalents Hydrogenated PK
stearins
Illipe Hydrogenated PKs
Sal
Kokum Coconut oil and stearins
Palm fractions
Shea and enzyme equivalents
j8-Stable -Stable /3'-Stable ^'-Stable
Require tempering Non-temper Non-temper
Note: Liquid oils are also used in confectionery and behave physically as a solvent for higher melting fats and also
as a carrier for flavor, e .g ., hazelnut oil.
Table 5 Fatty Acids (wt %) of Cocoa Butter
Quotient
(% 16:0)
Origin 16:0 18:0 18:1 18:2 (% 18:0)
Brazil 8 25.1 33.9 36.1 3.3 0.75

(2 4 .6 -2 6 .0 ) (3 0 .2 -3 5 .5 ) (3 5 .2 -3 8 .6 ) (2 .4 -4 .8 ) (0 .7 0 -0 .8 1 )
Ecuador 6 26.3 35.4 33.9 2.7 0.75
(2 6 .0 -2 7 .4 ) (3 4 .8 -3 5 .8 ) (3 3 .5 -3 4 .3 ) (2 .3 -3 .0 ) (0 .7 3 -0 .7 9 )
Venezuela 3 26.8 34.5 33.9 2.6 0.78
(2 5 .5 -3 0 .5 ) (3 2 .2 -3 6 .5 ) (3 3 .2 -3 4 .8 ) (2 .6 -2 .6 ) (0 .7 0 -0 .9 5 )
Trinidad, Grenada, Jamaica 5 26.8 34.9 34.0 3.1 0.77
(2 5 .3 -2 7 .3 ) (3 3 .2 -3 5 .1 ) (3 3 .7 -3 5 .9 ) (2 .7 -4 .0 ) (0 .7 2 -0 .8 0 )
Ghana 6 25.2 34.3 35.1 3.5 0.73
(2 3 .6 -2 5 .6 ) (3 3 .9 -3 5 .4 ) (3 4 .8 -3 7 .6 ) (3 .3 -3 .6 ) (0 .7 0 -0 .7 5 )
Nigeria 4 25.8 35.4 33.9 3.3 0.74
(2 5 .5 -2 6 .2 ) (33 .9 -3 5 .7 ) (3 3 .7 -3 5 .1 ) (2 .9 -3 .8 ) (0 .7 1 -0 .7 6 )
ivory Coast 4 25.9 34.6 34.6 3.2 0.75
(2 5 .4 -2 6 .2 ) (3 4 .2 -3 5 .2 ) (3 4 .2 -3 5 .1 ) (3 .0 -3 .4 ) (0 .7 2 -0 .7 6 )
Cameroon 3 25.9 34.3 34.8 3.4 0.75
(2 4 .0 -2 6 .3 ) (3 4 .2 -3 4 .6 ) (3 4 .8 -3 6 .0 ) (3 .2 -4 .1 ) (0 .6 9 -0 .7 6 )
Togo 3 24.9 34.3 35.2 3.3 0.72
(2 4 .6 -2 6 .0 ) (3 3 .9 -3 5 .1 ) (3 5 .1 -3 5 .3 ) (3 .2 -3 .6 ) (0 .7 2 -0 .7 7 )
Congo 1 25.0 34.2 35.6 3.4 0.73
St Thomas 1 25.3 34.2 35.2 3.4 0.74
Madagascar 1 25.7 36.1 34.0 2.4 0.71
New Guinea, Borneo 9 26.1 35.4 34.2 2.6 0.74
(2 4 .6 -2 8 .2 ) (3 3 .8 -3 6 .4 ) (3 3 .3 -3 4 .4 ) (2 .2 -2 .9 ) (0 .6 9 -0 .8 3 )
Samoa, Java 3 26.1 34.6 34.6 3.0 0.75
(2 5 .8 -2 7 .8 ) (3 3 .2 -3 4 .6 ) (3 4 .4 -3 4 .7 ) (2 .7 -3 .2 ) (0 .7 5 -0 .8 4 )
Note: Range in parentheses.

Source: Ref. 145.
In understanding the behavior of confectionery fats it is important to be able to relate phy

sical properties to chemical composition. The chemistry of cocoa butter and cocoa butter equiva
lents (CBE) is relatively simple compared to that of other oils and fats. Much of what is
described about cocoa butter also relates directly to CBE. Because cocoa butter is a natural prod
uct there is a natural variation in its composition depending on a range of factors including
climate.
A. Cocoa Butter^
The chemical composition of cocoa butter can be described in a number of ways based either
on individual glycerides, glycerides grouped according to degree of unsaturation characterized
by silver-phase chromatography, or glycerides grouped according to carbon number as deter
mined by high temperature gas-liquid chromatography [31-37]. Some typical data are given
in Tables 5 -9 .
The chirality of POSt has been determined and is seen to be essentially racemic [38]. The
major glycerides in cocoa butter are 1,3-distearoyl-2-oleoyl glycerol (StOSt), r-l-p alm ito yl-2-
^See also Chapter 2.

402 Padlef
Table 6 Fatty Acids (wt %) of Some Cocoa Butter Equivalent Components

Origin 16:0 18:0 18:1 18:2 Others
Palm fraction 56 6 32 4 2
Shea stearin 5 57 33 3 2
Illipe butter 18 46 35 1 tr
CBE 28 34 33 3 2
tr = trace.
Table 7 Triacylglycerol Composition of Cocoa Butters from Different Countries of Origin
Triacylglycerol (%)
Country
of origin PLP POP PLSt POP StOO StLSt post StOSt StOA
Bolivia 1 1.1 3.3 3.5 2 2 .6 4.0 2.1 40.4 22.8 0.5

Brazil 6 0.9 3.9 3.7 17.9 6.7 3.2 37.1 26.0 0.04
Colombia 2 1.1 3.3 3.6 20.4 4.4 2.3 39.4 25.0 0 .6
Ecuador 3 1.2 3.0 3.2 19.2 5.4 2.3 38.4 26.9 0.4
Peru 1 1.5 4.3 3.9 18.3 7.4 3.7 35.8 24.6 0.4
Venezuela 1 0.9 1.0 3.1 20.4 2 .8 1.9 40.4 28.8 0 .8
Costa Rica 1 1.0 2.6 3.5 17.8 5.5 3.0 38.7 27.4 0.4
Dominican Republic 4 1.1 3.3 3.2 18.4 6.1 2.7 38.2 26.5 0 .6
Guatemala 1 1.0 2.3 3.4 19.3 4.9 2.2 39.0 27.5 0.4
Mexico 1 1.1 2.4 3.5 19.1 4.1 3.0 38.8 27.8 0.6
Panama 1 1.0 1.5 3.0 19.1 3.1 2.7 41.4 27.3 0.8
Cameroon 2 1.0 3.0 3.4 17.9 5.8 2.5 38.3 27.7 0.5
Gabon 1 0.9 3.7 3.5 17.5 7.3 3.0 37.1 26.5 0.4
Ghana 3 1.2 2.2 3.4 17.8 4.9 2.2 39.0 27.5 0.4
Ivory Coast 9 1.0 1.9 3.0 19.0 3.9 2.5 39.6 28.5 0.6
Nigeria 2 1.0 2.3 3.6 17.9 5.2 3.0 38.8 27.8 0.5
Indonesia 2 1.1 1.6 3.0 19.9 3.6 1.7 40.6 28.1 0.5
Malaysia 20 0.7 1.2 2.8 18.4 2.9 2.2 40.0 31.1 0.8
Solomon Islands 1 1.0 0.9 3.0 19.3 2.8 2 .0 40.7 29.5 0.7
= number of samples.
P, palmitic (16:0); St, stearic (18:0); A, arachidic (20:0); 0 , oleic (18:1); L, linoleic (18:2). S, saturated.
Source: From Ref. 146.
Table 8 Triacylglycerol Composition of CBE and Components
S3 SOS sso SLS so o Others
Palm fraction 3 80 6 6 3 2
Shea stearin 2 83 — 7 5 3
Illipe 1 94 — 1 3 1
CBE 2 83 3 7 4 1
Brazil cocoa butter 1 83 — 5 9 2
Table 9 Triglyceride Composition of Cocoa Butter 1976-1979

Triglyceride carbon number^
Origin C48 C50 ^52 C54 C56 ^58

Ghana 0.3 18.1 46.5 33.4 1.6 0.1
0.4 18.9 46.6 32.4 1.5 0.2
Ivory Coast 0.4 16.9 46.3 35.0 1.6 0.2
0.3 18.1 46.4 33.1 1.8 0 .2
0 .2 17.9 46.3 34.1 1.6 0.1
Nigeria 0.4 18 46.6 33.4 1.4 0.1
0 .2 17.8 46.2 34.1 1.5 0.2
Cameroon 0.3 17.9 46.1 33.6 1.8 0 .2
Togo 0 .2 18.3 46.3 33.4 1.7 0.1
Bahia 0.3 17.5 45.7 34.5 1.7 0 .2
0.3 15.2 44.6 38.0 1.7 0.2
0.3 17.4 46.2 34.7 1.6 0.1
Ecuador 0.4 19.4 46.2 31.5 2.1 0.3
0.3 19.8 46.7 31.4 1.8 0.1
Venezuela 0.3 22.7 47.3 27.1 2 .2 0.4
0.4 19.4 46.9 31.2 1.7 0 .2
0.5 19.6 46.4 31.9 1.5 0.2
Trinidad 0.3 2 1.2 47.0 30.0 1.3 0 .2
Grenada 0.4 18.4 46.4 32.5 1.7 0.2
Dominican Republic 0 .2 17.7 46.0 34.5 1.5 0.1
Papua New Guinea 0.4 18.9 46.5 32.3 1.6 0.2
0 .2 18.0 46.3 33.8 1.6 0.1
0.3 16.7 45.9 35.5 1.6 0.2
Malaysia 0.3 18.1 45.8 33.7 1.8 0 .2
0.3 17.0 45.8 34.8 1.9 0.2
0 .2 17.6 46.0 34.5 1.6 0.1
Sri Lanka 0 .2 17.9 46.3 34.0 1.5 0.1
Indonesia 0.2 18.2 46.2 33.6 1.6 0.2
0.3 17.8 46.0 34.1 1.6 0.2
Samoa 0.4 17.2 46.1 34.6 1.7 0 .2
^Carbon number is the number of carbon atoms in the fatty acid chains of the triglyceride, e .g ., C48 is all triglycer
ides in which the fatty acid carbon atoms add up to 48. A number of samples have been analyzed from some sources.
Source: Ref. 129.
oleoyl-3-stearoyl glycerol (POSt), and 1,3-dipalmitoyl-2-oleoyl glycerol (POP). These glycer

ides largely dictate the physical behavior of cocoa butter, but other minor glycerides, e.g.,
diglyceride and trisaturated glycerides, can influence melting and crystallization behavior
[25,39]. The polymorphism of the major cocoa butter triglycerides and of cocoa butter has
been described by Wille and Lutton [17] and more recently by Yano et al. [41]. In a recent
review, Sato [20] brought all the relevant information together. Melting point and X-ray dif
fraction data of the triglyceride polymorphs are given in Table 10. The effect of liquid oil on
the crystallization behavior of cocoa butter has also been described [42].
The chocolate processing stages of both tempering and cooling tunnel crystallization aim
to produce chocolate crystallized in form V (J82). This is stable for most practical purposes
and ensures that the chocolate retains its gloss. Under conditions of high ambient temperature
404 Padley
Table 10 Polymorphic Forms and Melting Points of StOSt, POSt, and POP
StOSt
Property a y p' P2 /3.
LS (nm) 4.83 7.05 7.00 6.50 6.50

(°C) 23.5 35.4 36.5 41.0 43.0
post
a 3 p' p
LS (nm) 4.76 7.06 6.79 6.40

(°C) 19.5 28.3 31.6 35.5
POP
a y 3 p '2 p \ P2 P,
LS (nm) 4.65 6.54 6.25 4.24 4.24 6.10 6 .1 0
CC) 15.2 27.0 29.2 30.3 33.5 35.1 36.7
Filer et al. — — — P' p’ p p
Malkin and Wilson a — p' p p
Lutton and Jackson a -2 — m h p -2 p '- 2 p '- 2 P-3 p-3
Lavery a -2 p" — l3'-2 p '- 2 p -3 p-3
Lovegren et al. 5 4.3 — 2 2 1 1
Gibon et al. a -2 LL 2 3 subj8 - 2 pseudo-/3'-2 p -3 P-3
For original references refer to Sato [20].

Source: Ref. 20.
(>25°C, say), the metastable form V transforms in a matter of weeks to form VI (jSj). This
transformation is associated with the onset of bloom. This aspect of chocolate behavior is
discussed later.
The stabilized solid content for the ternary mixture POP-PO St-StOSt is depicted in an
isosolids diagram [40] (Fig. 6). This illustrates that the melting behavior of cocoa butter will
be reduced with increasing POP contents and raised with increasing StOSt.
Recent work by Savage and Dimick [43] shows that phospholipids influence the nucléation
and crystallization behavior of cocoa butter. In particular, an increase in phosphatidylcholine
content increased nucléation and crystal growth. The converse was reported for lysophosphati-
dylcholine and phosphatidylinositol. In almost simultaneous publications from the same group
[44,45], however, the triglyceride composition is observed to have an overriding effect, fast
crystallizing cocoa butters being characterized as having high 2-oleodistearin (StOSt) contents.
This is in line with the generally acknowledged relationship between composition and crystal
lization behavior [33,46]; i.e., increased StOSt content raises the melting point and the crys
tallization rate at a given temperature as a consequence of the higher degree of supersatura
tion. Taking all these data into account, they suggest that phospholipids play a secondary
role in influencing crystallization behavior, the major influences being triglyceride and partial
glyceride composition. In relation to phospholipids one should also note that the Dimick and
coworker studies were on cocoa butter alone. The real-life situation is concerned with choco
late, which will contain added lecithin, a significant part of which will almost certainly have
adsorbed onto the sugar particles.
calculated
(a)
measured (d )
Fig. 6 Ternary phase diagram for a ternary mixture of POP, POSt, and StOSt. Calculated and mea
sured ternary isosolids diagrams of SOS-POS-POP. (a, c) Clear point diagrams; (b, d) isotherms with
25% solid fat. The PPP-POP-POO and PPP-PPO-POO ternaries also show good agreement between
measurements and calculations. The agreement near the PPP comer at the PPP-POO side of the dia
grams is somewhat less good. (From Ref. 40.) (Measured data K. Smith).
Perhaps the main point to note here is that cocoa butter is a natural product and exhibits a
significant natural variability. The sources of cocoa and cocoa butter are now extremely numer
ous, and as a consequence of different climatic and growing conditions the glyceride content
and hence the melting behavior of cocoa butter can vary quite widely [46] (Table 11). This was
demonstrated quite elegantly by Lehrian and Keeney [47]. (See also Section VII.) In addition,
the various grades of cocoa butter also depend on the quality of the beans (FFA, etc.) and on
Table 11 Solid Fat Content (%) by NMR of Three Cocoa Butters and Fractions ^
West West
West African African
Brazilian Malaysian African stearin olein
N20 66 81 76 91 48
N25 60 76 70 95 36
N30 37 55 45 73 0 .0
N35 0.0 0 .0 0.0 16 0 .0
^Samples stabilized at 26°C then 0°C. N20 = NMR solids at 20°C.

406 Padley
how the cocoa butter has been extracted from the bean—by hydraulic pressing or via solvent
extraction. There is also the potential contamination of the pure cocoa butter with cocoa shell fat
[6,48,49]. Shell and germ fats are normally considered waste fats by responsible manufacturers
and are not permitted in pressed cocoa butter as legally defined. Shell fat is more unsaturated
than cocoa butter, containing 10-12% linoleic acid (18:2) and a number of unique compounds
such as alkyl tryptamides, 5-hydroxytryptamides, and fluorescent sterol derivatives [49].
Fractions of cocoa butter (Table 11) have also been developed and offer the chocolate
manufacturer another degree of freedom in controlling chocolate properties and in optimizing
formulations [50]. The properties of blends of cocoa butter stearins and oleins with unfraction
ated cocoa butter and milk fat are described. The stearins can be used to raise the melting
behavior of cocoa butter, creating more heat-resistant products. Both stearin and olein can be
used to provide greater flexibility in using milk fat [50].
B. Cocoa Butter Equivalents

The relatively high price of cocoa butter stimulated the invention of cocoa butter equivalents
[51]. These products were developed at a time when the chemistry of fats was still poorly
understood and without the advantage of chromatographic techniques. Nevertheless, it proved
possible to mimic the chemical composition and physical properties of cocoa butter extremely
well by blending together fractions of the required triglycerides from different oils and fats.
This invention by Crossley and coworkers transformed the world of chocolate confectionery
in a number of countries and acted as a catalyst for later technical achievements in the area
(Fig. 7).
Palm oil and shea nut oil are fractionated from solvent, usually acetone, to obtain fractions
with the necessary purity and with the right melting and crystallization behavior. In the case
of palm oil fractions, increasing interest is being shown in the use of palm fractions obtained
by dry fractionation, especially for the formulation of milk fat replacers in chocolate. In all
these fractionation processes, significant amounts of by-product fractions, i.e., high melting
fraction, mainly tripalmitin and a liquid olein fraction, are produced. The viability of this
technology therefore depends as much on their economic disposal as on the value of the CBE.
The various aspects of fat fractionation are covered in Chapter 8.
The control of palm oil quality as a raw material feedstock is very important in CBE
applications [52]. The fractionation conditions and choice of solvent are also important in
determining the levels of diglyceride and tripalmitin in the midfraction rich in 2-oleodipalmitin
(POP) [33]. Alternative sources of cocoa butter triglycerides are given in Table 12. Kokum,
illipe, and to some extent sal can be used without resorting to fractionation. All the fats apart
CBE CB
Shea oil .► Shea stearin (StOSt)
Illipe ^ illipe (POSt/StOSt)
StOSt
Sal/Kokum Other sources of Post
POP
etc . ^ SOS (StOSt SUU
U
Enzymic fats etc) Others
Palm oil
Palm mid Fraction (POP)
Fig. 7 Simplified scheme illustrating CBE formulation from oils and fats and their fractions.
Table 12 The 2-Position and Overall Fatty Acid Composition of Some SUS-Rich Fats
FAC: 2-position FAC: Overall
16:0 18:0 18:1 18:2 18:3 20:0 16:0 18:0 18:1 18:2 18:3 20:0
Cocoa butter 2 2 85 11 — 0 27 33 35 3 —
2
Aceituno acid 2 4 89 4 — 0 10 28 57 3 — 1
Sal fat 0 .6 4.4 94.7 — — 0.3 7.1 44.8 42.5 — —
5.6
Mowrah tallow 3.2 3.4 57.0 35.0 1.4 0 23.1 2 2 .0 38.5 15.8 tr —
Malabar tallow 1.1 3.9 90.7 3.9 0.4 0 9.0 46.9 41.4 1.3 tr —
Kokum tallow — 4.4 94.4 1.3 — 0 1.7 61.0 36.8 0.5 — 1.4
Palm oil 11 2 65 22 — — 45 6 40 9 — —
Shea butter —
3.5 81.4 15.1 — 0 3.4 41.1 47.1 6.6 — —
Allanblackia (Dunkwa) 2 2 96 — — — 2.0 53 45 — — —
Chinese vegetable
tallow 11 0.5 84.5 4.0 0 0 65.0 2 .0 32.0 1.0 tr
Illipe butter 2.3 2.1 94 1.6 — 0 17.0 46.0 35 1.0 — 2
tr = trace.
Source: Ref. 76.
from palm oil (source of POP) are uncultivated and have to be harvested by local people
retrieving nuts from either native forest trees or natural groves. Illipé, shea, and sal are the
major sources of POSt and StOSt for the confectionery industry. Because of the diverse condi
tions associated with the harvesting and storage of the nuts, the quality and availability of
these oils is very variable [53,54]. Free fatty acid content can be very high, even up to 30%.
Diglycerides, an initial by-product of the hydrolyses, can be present at 10% or more in the
neutral fat. On the other hand, if good husbandry is followed from collection through interme
diate storage, shipment, and subsequent storage prior to extraction, then oils with low free
fatty acids and diglycerides can be obtained. Diglycerides have particular relevance to CBE
manufacture because their presence lowers melting point and retards crystallization [25,39].
In the case of sal, further complications can arise through microbial attack on the seed oil
during harvesting and storage. This introduces a range of modified glycerides in which the
oleic acid has been converted first to an epoxide, via ring opening to a dihydroxy fatty acid
and by partial esterification of the hydroxy acid to an estolide containing four fatty acids. The
triglycerides containing epoxy, dihydroxy, and estolide components are designated P, Q, and
R, respectively [55,56].
Chinese vegetable tallow (Sapium sebiferum), the outer coating of the Chinese vegetable tal
low seed, has been proposed as a rich source of pure POP. Examination of the tallow and the
seed has, however, demonstrated that there is a small but significant level of stillingia oil con
tamination. This arises from the migration of stillingia oil from the center kernel to the outer
coating [57]. The presence of toxic phorbol esters in the seed has also been described [58].
The shortage of natural sources of StOSt and POSt stimulated research to seek alternatives.
In the 1960s and 1970s synthetic routes were explored that enabled the individual contribu
tions of the major glycerides to be defined [46,59]. No commercial exploitation of this tech
nology occurred. Fermentation has also been explored in detail but again without commercial
success [60].
More recently a completely new technology has been developed, enzymatic interesterifica
tion [61-65]. This uses 1,3-specific lipases to catalyze the interesterification of, for example,
high oleic sunflower oil with stearic acid to generate 2-oleodistearin (StOSt):
408 Padley
O 1,3-specific lipase 0 p- St p- S t
O + 3S t o + - 0 + - 0 + S t + O acid s (6 0 :4 0 )
O 0 - 0 - St
16% 48% 36%
High oleic
(e.g.) sunflower oil Stearic
(O = o leic) acid (St)
The advantage of such a process is that it utilizes readily available raw materials and very mild
interesteriflcation conditions that are lipase-catalyzed. Commercial implementation relies on a
number of problems being resolved, notably the cost of the catalyst, including catalyst life and
the use of by-products. The by-products, some not shown in the above simplified equations, in
clude diglycerides generated by lipolysis. This is an undesirable but inevitable by-product that
arises because the lipase requires a finite amount of water for it to exhibit catalytic activity. Di
glyceride levels can be controlled, however, by the appropriate choice of process and process
conditions. From this new technology have emerged a range of other applications that are being
exploited commercially [64]. That technology is covered in detail in Chapter 9.
Cocoa butter equivalents can be tailored to meet a variety of requirements. Those con
taining higher amounts (>40%) of StOSt and POSt will be more expensive but much closer
in physical behavior to cocoa butter than those rich in palm midfraction (POP). Indeed, those
rich in StOSt and POSt (>60-70% ) will raise the melting range, thus improving the melting
and crystallization behavior of softer, inferior cocoa butters when used at the 15-20% level
of the fat phase. The use of high palm fraction CBE will, for example, lower the tempering
temperature and extend crystallization times to some degree. CBE formulated to closely match
cocoa butter have virtually no effect on the melting properties when blended with cocoa butter
chocolate, and tempering will also be essentially the same. The chemical and physical proper
ties of some typical CBE and of cocoa butter are given in Tables 5-10. The phase behavior
of cocoa butter-CBE mixtures has been described [66]. The isosolids ternary diagram of
StOSt, POSt, and POP given in Fig. 6 provides some guide to the effects of these major
glycerides on the physical behavior of cocoa butter-CBE mixtures.
The similar chemistry of cocoa butter and CBE has led to the wider use of CBE in so-called
supercoatings. This takes advantage of the total compatibility of the two fats and permits the
formulation of coatings containing high levels of cocoa mass, thus ensuring that the flavor
and meltdown of the product is extremely close to that of real chocolate. Some typical formu
lations are given in Table 13 [67]. CBE and vegetable fats are legally allowed at up to the
5% level in chocolate in the United Kingdom, Ireland, and Denmark and a number of other
European and non-European countries. The EU debates this issue periodically, and a wider
acceptance may occur eventually.
One of the issues that has arisen from time to time in this debate has been the inability to
determine the level of vegetable fat addition, particularly CBE, because of their close similar
ity to cocoa butter. This problem is discussed in Section VII.
C. Milk Fat2
The fatty acid composition of milk fat is given in Chapter 2. From the confectionery viewpoint
we are primarily concerned with the physical behavior of milk fat, its fractions, and blends
^See also Chapter 2.

of milk fat with cocoa butter [68,69]. Some typical solid fat content-temperature relationships
are shown in Table 14 and Fig. 8. When determining the solid fat content of cocoa butter
milk fat blends, care should be taken to adopt appropriate stabilization procedures [70]. The
important features to note are that butterfat is mainly liquid at temperatures between 20 and
35°C. The liquid oil component therefore acts as a solvent, leading to a significant melting
point depression when mixed with cocoa butter simply due to cocoa butter dissolving in the
liquid component of milk fat. The solid glyceride components, particularly those isolated in
the midfraction, also depress the melting range of cocoa butter and form a eutectic mixture
[69,71].
Milk chocolate therefore has a significantly softer texture than plain chocolate depending
on the level of milk fat added. This infiuences not only the texture but also the processing
behavior of the chocolate. Tempering performance is significantly affected as illustrated in
Fig. 4, which depicts the interrelationship between the temperatures of the first and third
stages of a three-stage pilot-scale temper unit while still retaining chocolate in temper as it
emerges from the third stage. Twenty percent milk fat in the fat phase lowers the tempering
temperature by approximately 2°C [25].
It has also been observed over many years that the addition of milk fat to chocolate inhibits
bloom formation by delaying the form V to form VI transition (see later). Milk fat is added
to chocolate in a variety of ways, directly as milk fat or full cream milk powder or even as
whole milk, which is dehydrated during the crumb process. Depending on the method of
addition and the degree to which the chocolate is refined, all or only part of the milk fat will
be released from the milk fat globule membrane and will form part of the continuous fat phase
[72]. This is relevant when trying to optimize fat content, the most expensive major ingredi
ent, and viscosity.
Table 13 Composition (%) of Coatings Using CBE and Cocoa Butter Substitutes
Cocoa
mass Cocoa Full cream Skim Coating % Fat
Coating ( 1 0 - 1 2 %) powder milk powder milk powder fat Sugar overall
Laurie ^
Dark — 14 7 31 48 32
Milk A — 5 10 10 29 46 32
MilkB — 5 0 17.5 31.5 46 32
Non-lauric ^
Dark — 20 — — 33 47 35
Milk A 10 — 6 12 28 44 35
MilkB — 5 — 17 34 44 34.5
Supercoating
Dark A 40 — — — 10 50 32
Dark B 20 10 — — 20 50 32
Milk A 10 — 20 — 22 48 32.5
MilkB 10 — — 15 27 48 32
^Palmkemel stearin based (non-temper). Cocoa powder should contain as low a level as possible of cocoa butter.
^Trans hardened fats and fractions (non-temper).
Cocoa butter equivalent (temper).
Note: Overall fat content will vary depending on the particular application in terms o f viscosity/weight control.
Source: Ref. 139.
410 Padley
Table 14 Solid Fat Content (%) of Cocoa Butter, Milk Fat, and High,
Middle, and Low Melting Fractions of Milk Fat^
Temp. Cocoa Milk

(°C) butter fat HMF MMF LMF
0 86 .0 56.8 99.7 92.1 26.1

10 81.9 42.0 99.5 90.5 11.6
15 79.1 30.2 88 .6 8.4
20 76.0 19.5 99.0 86 .2 6.1
25 68.9 13.1 80.2 0
30 31.6 8 .2 98.6 51.2
32.5 8.3 5.6 26.4
35 0 3.5 16.4
40 0 93.0 0
45 78.9
50 38.1
55 0
^For stabilization procedures used, refer to original paper.

HMF = High melting fraction; MMF = midfraction; LMF = low melting fraction.
Source: Ref. 69.
In the past the relative cost of cocoa butter and milk fat has induced some manufacturers
to develop milk chocolate containing high levels of milk fat (>20% on fat phase). In recent
years this has become less commercially attractive. In those countries where 5% vegetable
fats are legally permitted in chocolate, there has been increasing interest in partially replacing
the milk fat rather than the cocoa butter. A new group of products are therefore emerging
aimed at retaining the texture and mouthfeel of the original formulation but replacing part of
the milk fat with an appropriately tailored vegetable fat blend.
An alternative approach has been to consider using milk fat fractions. The oleins will behave
as a liquid phase, softening chocolate in a similar way to milk fat itself. The stearin will lower
the melting range, moving toward a eutectic at higher levels of addition (^50% ). Because of the
Fig. 8 Isosolids diagram for cocoa butter-milk fat mixtures after 4 weeks of storage at 13°C. (From
Ref. 69.)
presence of high melting glycerides there will also be a tendency to increase the viscosity of tem
pered chocolate. In most European countries milk fat fractions are now regarded as being legally
equivalent to milk fat. Milk fat stearin has been used, in some cases, as a bloom inhibitor. The
addition of high melting stearin will tend to increase the viscosity of the tempered chocolate.
Other derivatives of milk such as hydrogenated milk fat have also been suggested as antibloom
agents but with little practical success to our knowledge [73].
D. Paimkernel Stearins and Other Lauric-Rich

Confectionery Fats
Paimkernel oil can be easily fractionated using either solvent-based or non-solvent-based pro
cesses. Today most processes are non-solvent and rely on processes that produce a partially
crystallized fat that is amenable to separation into solid and liquid phases using hydraulic or
membrane presses [74,75]. Coconut oil can be processed in a similar way. For most confec
tionery applications paimkernel stearins, either hydrogenated or nonhydrogenated, are pre
ferred because of their sharp melting behavior. The slightly lower melting coconut stearins
find application in low melting fillings. Coconut oil is also widely used in formulating cou
vertures for ice cream. Some typical compositions and the melting behaviors of lauric stearins
are given in Tables 15 and 16. Babassu and tucum oils are also used as alternative sources of
lauric fats, but they are in limited supply. Readers are also referred to Soon [76].
Both paimkernel and coconut oils and their stearins are complex mixtures of triglycerides.
The high level of lauric acid (—55 wt %) means that most of the glycerides will contain at
least one lauric acid group, and approximately one-third will contain two lauric acids.
The complexity of the triglyceride mixture results in a fat that is -stable. The significance
of this is that the fat can be crystallized directly from the melt into its ¡3' form. The triglycer
ides of paimkernel stearin are chemically very different from those in cocoa butter. The phase
behavior of paimkernel stearin-cocoa butter mixtures [66] therefore exhibit eutectic behavior
with partial miscibility in the solid phase (Fig. 9). In practical terms this means that only
small amounts of paimkernel stearin (<2) can be mixed with cocoa butter before the melting
Table 15 Typical Lauric Fat Compositions
Fatty acids (wt %)
Fat IV 12 :0 14:0 16:0 18:0 18:1 18:2
Paimkernel oil 17.5 7 48 16 9 2 15 2

Paimkernel stearins 9 6 53 21 9 2 8 1
7 4 55 22 9 2 7 1
2 3 50 30 12 3 2 0
4 4 55 22 9 6 5 0
Paimkernel stearin (IV = 7) 0.4 4 55 22 9 9 0.5 0
hydrogenated to lower IV
Paimkernel stearin (IV = 7) 3 3 56 26 9 2 3 0.5
refractionated to lower IV
Coconut oil 8.5 15 48 18 8 3 6 2
Coconut stearin (25% yield) 4 9 55 22 9 2 3 0.5
Hydrogenated coconut stearin 1.5 9 55 22 9 5 2 0
Paimkernel olein 21 9 45 13 9 3 19 2
Coconut olein 10 17 46 17 7 3 7 3
Source: Ref. 75.

412 Padley
Table 16 Typical Laurie Fat Properties

Solid fat content (% by pulse NMR)
Slip
Fat IV N20 N25 N30 N35 N40 MP (°C)
Palmkemel oil 17.5 44 20 0 — —

28
Palmkemel stearins 9 70 48 8 0 — 30
7 82 68 29 0 —
32
2 94 84 37 3 0 34
4 87 70 24 1 —
31
Palmkemel stearin (IV = 7) 0.4 95 90 50 5 1 35
hydrogenated to lower IV
Palmkemel stearin (IV = 7) 3 94 90 60 1 0 33
refractionated to lower IV
Coconut oil 8.5 36 0 0 24
Coconut stearin 4 84 53 2 0 — 30
Hydrogenated coconut stearin 1.5 92 57 8 2 — 32
Hydrogenated palmkemel stearin 1.0 73 — 27 14 6 41
Source: Ref. 75.
point is noticeably depressed. Over a period of time phase separation will also be observed.
Conversely, only low amounts of cocoa butter (<5% ) can be used in chocolate substitutes in
which palmkemel stearin is the main fat phase. In both cases chocolate bloom will be ob
served if excess amounts of the minor component are used in the blend, i.e., palmkemel
bloom on cocoa butter chocolate, cocoa butter bloom on palmkemel chocolate 15 ~20°C or
lower (see Section V). It is the presence of these low amounts of cocoa butter in palmkemel
stearin substitute chocolate that no doubt accelerate the loss of gloss. However, even in the
absence of cocoa butter, palmkemel-based products still lose their gloss more rapidly than
cocoa butter chocolate. At elevated temperatures, however, the bloom on palmkemel choco
late is lauric-rich. Minor amounts of cocoa butter remain dissolved, but growth of lauric-rich
crystals is encouraged by holding at elevated temperatures. The interference by cocoa butter
in palmkemel stearin-based chocolate substitute has been largely removed with the develop-
Fig. 9 Phase diagram of cocoa butter-palmkemel stearin mixtures. A similar phase diagram is ob
tained with partially hydrogenated, high trans fats but with greater miscibility, i.e., the phase boundaries
move slightly toward the center. (From Ref. 6 6 .)
merit of fat-free cocoa powders [9]. Bloom in these products can also be alleviated to some
extent by the addition of surface-active agents such as sorbitan tristearate.
The processing of palmkemel stearin chocolate is greatly simplified by the elimination of the
tempering stage. Molten chocolate at, say, 45°C can be used directly for enrobing or molding.
The chocolate is then solidified by cooling rapidly at approximately 12°C in a cooling tunnel,
avoiding condensation as the product emerges by raising the temperature of the final section.
The use of palmkemel stearins has also been associated with the development of soapy
rancidity in some confectionery products processed under nonhygienic conditions. This aspect
is discussed under rancidity, but suffice to say the wide application of these fats demonstrates
that by taking reasonable precautions this problem is easily overcome.
E. Fractions of Partially Hydrogenated Oils and Fats

The ready availability of liquid oils has stimulated confectionery fat manufacturers to use
hydrogenation, with or without the additional step of fractionation, to generate suitable prod
ucts for the confectionery industry.
With the appropriate choice of catalyst and under conditions of limited hydrogen availabil
ity it is possible to minimize the level of fully saturated acids and to maximize the level of
trans fatty acids in the product (see chapter on hydrogenation). Trans fatty acids and their
triglycerides have an intermediate melting point between saturated and natural cw-unsaturated
acids. The generation of trans-rich fats therefore offers an alternative approach to obtaining
sharp melting fractions suitable for confectionery purposes. The melting behavior of unfrac
tionated, trans-harácncá fats is not very sharp, and the high melting components give the
product a waxy mouthfeel. For some confectionery applications, however, this is acceptable.
Significantly improved products are produced by fractionating the hydrogenated fat from sol
vent and isolating a midfraction, i.e., removing high and low melting components. Some
typical compositions and melting profiles are given in Tables 17 and 18 [77].
Chemically these fats are very dissimilar to palmkemel stearins. The hardened fats are based
on and Cjg fatty acids. Their glyceride composition is much more complex than that of cocoa
butter, and they can crystallize from the melt directly into the B' form, which is also the stable
polymorph for these fats. These fats therefore do not need tempering and are widely used in the
manufacture of substitute chocolate. Generally they display good gloss retention, although fats
based mainly on Cjg fatty acids have a tendency to lose gloss more rapidly. This can be counter
acted to some degree by the addition of sorbitan esters of stearic acid.
The phase behavior of trans-hardcncd fats with cocoa butter has been studied [66]. The
phase diagram is similar to that shown for palmkemel stearin-cocoa butter mixtures, but there
is a greater degree of miscibility in the solid phase. These fats are therefore more compatible
with cocoa butter than palmkemel stearins. Eutectic behavior and melting point depression
does occur, however. Another aspect relevant to these fats in particular is that of postcrystalli
zation. For a reason that is not fully understood, the solid fat content of many of these fats
Table 17 Typical Fatty Acid Compositions of Trans Hardened Fats
Fat 16:0 18:0 18:lr 18:1c S Others < 1 ^
A 28 6 44 21 1
B 12 11 59 15 3
C 19 11 34 33 3
D 23 13 39 22 3
^Sum of all others occuring at <1% individually.

414 Padley
Table 18 Typical Solid Fat Content of Traw^-Hardened Fats^
Solid Fat Content

Fat N20 N25 N30 N32.5 N35
A 86 78 41 21 3
B 82 77 62 34 4
C 70 59 38 24 12
D 62 50 28 18 6
^N20 indicates solid fat content measured by NMR at 20°C, etc. Measurement made
on a Brucker Minispec p20i on unstabilized samples.
increases significantly during storage at ambient or slightly elevated temperatures. This results
in a deterioration in melting properties. This effect is, however, tolerable, possibly beneficial,
in warmer climates for those products for which a reasonable degree of heat stability is re
quired and for those enrobed products where a slight elevation in melting point can be benefi
cial to the overall mouthfeel of the product. Also see discussion by Soon [76] and a patent
describing these types of fats and nonpostcrystallizing alternatives [78]. Because of consumer
pressure there has been a move to seek alternatives to partially hydrogenated fats, especially
as filling fats. Nutritionally this is a difficult area because replacing trans fatty acids will in
most cases result in increasing saturated acids.
F. Liquid Oils
Although not regarded strictly as confectionery fats, the liquid oils are widely used in choco
late confectionery. Typical examples are peanut and hazelnut oils present in nut-paste fillings.
The addition of up to 5% finely ground hazelnuts, peanuts, etc. is legally allowed within the
European Union without the need for declaration. A knowledge of the infiuence of liquid oils
on physical behavior is therefore important. The effect of liquid oil on the solid fat content of
cocoa butter is illustrated in Fig. 10.
Fig. 10 The effect of liquid oil on the solid fat content of cocoa butter.
The addition of liquid oil reduces the overall solid fat content, which can be advantageous
when trying to achieve softer textures. There is, however, a significantly greater risk of bloom
formation (see later). The effect of liquid oils on the polymorphic behavior of cocoa butter
has been studied [42].
IV. METHODS OF CHARACTERIZING MELTING AND

CRYSTALLIZATION BEHAVIOR
The complexity of fat crystallization and melting behavior arises because fats are polymor
phic. The most widely applied technique for determinijig the polymorphs present in a crystal
lized fat is X-ray diffraction. Various books and papers [19,20,79,80] describe the technique
and its application to fats, including the more recent application of a position-sensitive detec
tor (PSD) for more dynamic measurements [20].
Four main methods are used with differing degrees of success to characterize the melting
and crystallization behavior of fats. These are dilatometry, wide-line nuclear magnetic reso
nance (NMR), particularly pulse NMR, differential scanning calorimetry (DSC), and empiri
cal cooling curves. Dilatometry, prior to the advent of NMR, was the preferred method for
determining melting behavior [53] and relies on measuring the volumetric change arising from
the change from solid to liquid. The method has been very useful but requires trained opera
tors to obtain reproducible results. In addition, the volumetric change in going from solid to
liquid is different for fats widely different in fatty acid type. For example, lauric fats rich in
Ci 2 exhibit a smaller volume change than more conventional fats based on and C^g fatty
acids for the same amount of solid. A 100% solid fat in palmkemel stearin will give a dilation
as measured by the lUPAC method of —1800 whereas cocoa butter will give —2500 (mol/25
g fat). Solid fat index is approximately obtained by dividing the lUPAC dilation by 25.
“Pulsed NMR” [81-83] is now the preferred method, and results are given in percent solid
fat. Again, care needs to be taken in making the measurement because the signals from
different fats decay at different rates, thus influencing the way in which the instrument has to
be set up.
In the case of both dilatometry and NMR, standard methods of precrystallizing and temper
ing the fat (stabilization) have to be observed to obtain reproducible results. These stabiliza
tion conditions also depend on whether the fats are /3'- or ^-stable, different procedures being
used for palmkemel stearins (^'-stable) than for cocoa butter and cocoa butter equivalents {¡3-
stable) [53]. Care is also needed in defining appropriate stabilization regimes for cocoa butter
milk fat mixtures [70].
Differential scanning calorimetry (DSC) and differential thermal analysis (DTA) have been
widely used to determine the phase behavior of fats. These techniques have not met with wide
acceptance as methods of measuring solid fat content. This is because both polymorphic
changes and melting can occur at the same time, inducing thermal effects in different direc
tions. Even more complex is the influence of heating and cooling rates on melting and crystal
lization behavior. These effects have been amply demonstrated by Cebula and Smith [84,85],
who studied the effect of heating and cooling rates on the DSC curves for the triglycerides
POP, p o s t, and StOSt. A detailed study of the effect of composition on the crystallization
and melting behavior of a cocoa butter equivalent also demonstrates the need to understand
the chemistry of the fat in order to make rational interpretations of the data [85]. A study by
Ali and Dimick [86] on palm midfraction, cocoa butter, and milk fat also highlights some of
the difficulties associated with obtaining fully stabilized systems to obtain meaningful phase
diagrams. The technique potentially offers a useful in-house method for quality control of
routinely produced products. Careful standardization would be required for more general qual
416 Padley
ity control purposes across laboratories. The reader should also refer to other detailed publica
tions [71,87-89].
The empirical crystallization cooling curves are typified by the Jensen and Richard cooling
curves. Standard procedures have been described for both methods [3,51]. The Jensen cooling
curve is normally applied to cocoa butter and CBE. It is obtained by recording the temperature
of the fat as it crystallizes from the molten state under periodic agitation. The aim of the test
is to nucleate and crystallize the form V polymorph of cocoa butter or CBE. After super
cooling, the temperature of the fat rises to a maximum as the fat crystallizes. The characteris
tics of the curve are used by some manufacturers to anticipate the performance of the fat
during tempering and crystallization in the cooling tunnel. The Richard cooling curve has been
used for a similar purpose, but because it is obtained without agitation it is more suited to
fats that crystallize and are stable in the /3' form.
V. CHOCOLATE BLOOM
There are two basic forms of bloom that appear on chocolate, sugar bloom and fat bloom.
Sugar bloom will arise on the surface of products produced under conditions that promote
moisture condensation on the surface of the product as it emerges from the cooling tunnel.
Alternatively, it will arise during storage in conditions of high humidity and fluctuating tem
perature. The main purpose of this section, however, is to discuss fat bloom.
A. Bloom on Chocolate Substitutes

There have been numerous studies on the nature of fat bloom. We encounter bloom with
those products that include in their formulations a blend of cocoa butter with fats having a
very dissimilar chemical composition. This invariably arises because the two fat phases sepa
rate [66]. This has been briefly discussed above under palmkemel stearins (see Fig. 8), and it
is in this area that phase-separated bloom is most likely to occur. The boundaries between the
single solid solutions of either cocoa butter in j8' palmkemel stearin or palmkemel stearin
dissolved in ¡3-3 stable cocoa butter and the area of formulation containing both solid phases
defines the extremes of the safe limits for blends of both fats to ensure that bloom does not
occur. The problem will be exacerbated by temperature cycling and, perhaps surprisingly, by
storing products at low temperatures. So-called ghost bloom arises at low temperatures, for
example, in chocolate containing low but significant levels (3-5% ) of palmkemel oil or stea
rin. In this case temperatures of 10-15°C are sufficient to encourage phase separation and the
emergence of bloom rich in lauric triglycerides.
The main application of palmkemel stearins and trans hardened fat fractions is, however,
as the major fat phase in substitute chocolate. In this case the level of cocoa butter must be
carefully controlled, particularly for palmkemel stearins.
Cocoa powder is usually the only source of cocoa butter in chocolate substitute formulations.
For this reason both the fat level in the cocoa powder and the absolute level of cocoa powder
used must be controlled in palmkemel stearin-based “chocolate” such that the level of cocoa
butter in the fat phase is less than 3%. This will ensure the avoidance of rapid bloom formation
at ambient temperatures of ~20°C or less arising from the separation of a cocoa butter-rich
phase. At higher storage temperatures the loss of gloss and development of bloom on palmker-
nel-based products is almost invariably caused by the recrystallization of lauric triglycerides.
Despite taking all reasonable precautions it has been widely observed that palmkemel-based
products and some trans fats lose their gloss more rapidly than conventional chocolate. As men
tioned earlier, the development of fat-free cocoa powders does offer a solution to the problem of
cocoa butter interference in these products. Bloom can also be alleviated to some extent by the
addition of some surface-active components such as sorbitan esters.
B. Bloom Arising from Cocoa Butter

The problem of chocolate bloom for systems based primarily on cocoa butter has been widely
studied [17,73,90-94]. The starting point has been the phase diagram of the stable forms of
the major cocoa butter triglycerides StOSt, POSt, and POP (Fig. 6). Various hypotheses have
been made, including the presence of a eutectic that encourages phase separation on prolonged
storage or storage at elevated temperatures [95]. There is, however, a general consensus aris
ing from the work of Wille and Lutton [17] that loss of gloss and bloom on correctly tempered
chocolate (form V, P 2 polymorph) is directly related to a polymorphic change from form ^2
to the most stable form (form VI).
The structure of f^ 2 and (3i crystals (most stable form) of the pure symmetrical triglycerides
and of cocoa butter has been studied recently [20,41,96-99]. In the Hagemann and Rothfus
paper [96], computer modeling is used to calculate the most stable (j8j) configuration of the
triglyceride molecule StOSt. This structure shows the familiar alignment of the saturated acid
chains alongside each other. The oleic chains pack in a separate layer in a bent configuration
(Fig. 11). Sato and coworkers [20,41,97,98] have studied the J82 and polymorphs of the
pure symmetrical triglycerides using a range of spectroscopic techniques. Interestingly, a ^ 2
I3~3
Fig. 11 Speculative scheme achieved via computer modeling for the double chain-to-triple chain tran
sition of StOSt. (From Ref. 96.)
418 Padley
form of p o s t was not observed [98]. The confirmation of the oleic chain in the P 2 and Pi
polymorphs is of specific interest. The oleic acid moiety is invariably depicted as having a
“bent configuration” around the cis double bond in the /3-stable crystal lattice. However, a
number of studies have shown that a straight configuration closer to that observed for saturated
hydrocarbons can also exist. The differences in free energy for the bent and straight configu
rations is sufficiently small as to make a definitive choice difficult [100] (Fig. 12). In fact,
the oleoyl chain of crystalline cholesteryl oleate is known to be in the straight configuration.
In their model study of the jSj form of the symmetrical triglyceride SOS, de Jong et al. [100]
concluded that the oleic acid was in the bent configuration.
The “straight” configuration of the oleoyl chain is proposed both by Sato et al. [99] and by
Hagemann and Rothfus [96] for those crystal states where SOS triglycerides have adopted the
double chain length packing arrangement as found in certain y and P' states. The transformation
of SOS from jS' to ¡3 has been modeled by computer-aided techniques [96], and the realignment
of the straight oleoyl chain to the bent /3 form is clearly depicted. Could it be that the oleoyl
chain in P 2 (fo™ V) is in a straight configuration, moving to the bent configuration in the most
stable Pi form (form VI)? Recent work by Sato’s group [41] indicates that the difference be
tween the P 2 and Pi forms of SOS does in fact lie in the conformation of the oleoyl chain. How
ever, their evidence suggests that the oleic acid is “bent” in both P2 and Pi forms but transforms
from an orthorhombic to a triclinic form of packing. Any attempt to delay cocoa butter bloom
must therefore address the problem of finding a method of delaying the reorientation of the
oleoyl chain, thus preventing the ^ 2-to-j8i transition. The polymorphism and structure of the ma
jor triglycerides in cocoa butter has been recently reviewed by Sato [20].
The precise conformation of the oleoyl chain in cocoa butter crystals and the mechanism of
the jS2-to-î transformation may still be open to debate. However, based on our current knowl
edge it is possible to propose various mechanisms that would lead to a delay in the onset of
Fig. 12 Possible configurations of oleic acid. (Adapted from Ref. 100.)

Chocolate and Confectionery Fats 41 i
bloom. The development of chocolate bloom is initiated by the j82-to-î transition. This can oc
cur via either solid-solid or solid-liquid-solid transformations. In the case of solid-solid re
arrangements it is highly probable that the initial changes will occur either at the surface or at
dislocations within the crystal where there is greater room for molecular mobility. The remainder
of the crystal can then be reordered by a concerted action somewhat like a “Mexican wave” pass
ing through the crystal. Bloom would be delayed if an appropriate foreign molecule interfered
either with the initial surface transformations or with the bulk transformation. This could occur,
for example, if the foreign molecule interferes specifically with the oleic acid realignment, i.e.,
if a molecular “splint” is inserted to hold the oleic acid in the P 2 shape.
In the case of bloom being formed by solid-liquid-solid transformations, two approaches
can be envisaged. The first is to add stable P 2 seed crystals. This would encourage P 2 growth
to the detriment of jSj. The second approach is to include a foreign molecule that either
prevents nucléation or significantly delays crystal growth, i.e., a nucléation or crystal
growth inhibitor specific for (3i (form VI). The approach of ensuring that there are always
sufficient ^ 2 nuclei and seed crystals available, even at elevated temperature, to direct any
recrystallization of fat that occurs into the ^82 form has met with some success. The use of the
high melting symmetrical triglyceride 1,3-dibehenoyl-2-oleoyl glycerol (BOB) employs this
effect [101-103]. Using the P 2 form of BOB added in a finely powdered form to nontempered
chocolate at 33°C at levels of 1-5% inhibited bloom on chocolate cycled between 20 and
32°C. Interestingly, the most stable ¡3 form of StOSt (form VI) and cocoa butter was also
effective, which to some extent negates the thoughts about P 2 seeds. See also Giddey and
Clerc [92] for studies on seeding effects. This suggests that the mechanism is not simply one
of ensuring that only P2 nuclei are present. The role of tempering in delaying bloom [91]
relates to the need to ensure the presence of the right type and number of crystal nuclei for
P 2 growth. The success of extended tempering regimes such as those proposed by Kleinert is
almost certainly influencing this aspect of fat crystallization.
In the second approach the success of certain other fats in delaying bloom is explained by
the mechanisms that rely either on a “blocking” mechanism to prevent j82-to-j8j transitions or
on inhibiting Pi nucléation and growth. Milk fat, for example, has been widely used in
chocolate formulations for many years both as a component of milk chocolate and as a bloom
inhibitor. It almost certainly operates as a crystallization inhibitor or blocking agent
[3,73,94,104].
A group of triglycerides based on two long-chain saturated fatty acids (Cj^, Cjg) and one
short-chain saturated acid have proven to be eminently successful in inhibiting bloom
[105,106]. These triglycerides have been tailored to act either as crystal inhibitors for Pi or
as retainers of P 2 configuration. The intriguing observation is that these materials are effective
not only by direct addition to chocolate but also via fat migration, e.g., from a center filling.
This suggests that they are operating either by adsorption on the crystal surface, thus pre
venting the P 2 - P 1 transition, or as crystal growth or nucléation poisons for j8j.
With the continuing development of techniques such as atomic force microscopy [107] for
characterizing crystal structure, we can anticipate further advances in this area. Retention of
good gloss is clearly very important. Of equal importance are the retention of good flavor and
the prevention of rancidity.
VI. RANCIDITY IN CONFECTIONERY

A. Oxidative Rancidity
There are essentially three types of rancidity that can be encountered arising from fats in
confectionery products [108]. These are oxidative, lipolytic, and ketonic. All confectionery
420 Padlef
fats have potential for oxidative rancidity. Lipo lytic rancidity is lim ited to the lauric fats and
butterfat, and ketonic rancidity is lim ited prim arily to palmkemel and coconut oils but could
also be anticipated in butterfat-containing products. For a comprehensive view of rancidity in
foods the reader is referred to Allen and Hamilton [109]. The general topic of oxidative
rancidity has been reviewed by a number of workers, including Labuza [110] and Frankel
[111]. Factors influencing the shelf life of cocoa products have been discussed by Kattenberg
and de Muijnck [112].
Confectionery fats are relatively stable to oxidative rancidity provided that sufficient antiox
idant is present. The majority of confectionery fats contain only low levels of the oxidatively
unstable polyunsaturated acids such as linoleic or linolenic acid. The lauric fats are essentially
saturated and contain only low levels of unsaturated acids. Despite the fact that there is a low
level of risk, it is nevertheless necessary to take reasonable and sensible precautions to mini
mize autoxidation, which can occur under extreme conditions of low antioxidant, elevated
temperatures, and ready access to oxygen.
Cocoa butter is primarily based on saturated and oleic acids. It is also protected, however,
by a range of naturally occurring antioxidants, tocopherols, and flavonoids. These two factors
combine to give a fat that is extremely stable to oxidative rancidity. Related products such as
cocoa butter equivalents may need additional protection (100-200 ppm tocopherol) as these
are normally midfraction or stearins and may be depleted in natural antioxidants. This com
ment applies generally to all fractionated products apart from the olein fraction which will be
enriched in the antioxidants of the parent oil. Additional protection can also be obtained by
adding synthetic antioxidants such as butylated hydroxyanisole (BHA), tert- butylhydroqui-
none (TBHQ, in the United States), or butylated hydroxytoluene (BHT), customer and legisla
tion permitting. In the case of commercial CBE, these have virtually identical induction peri
ods as measured using the Rancimat accelerated oxidative stability test [113] to cocoa butter.
The relatively saturated lauric fats should also be handled with some care. These saturated
fats contain only low levels of natural antioxidant. Typical levels of tocopherol in palmkemel
and coconut oils are <20 ppm. Storing these fats and their fractions at elevated temperatures
M a c a d a m ia
nuts
Pecan nuts
Almoiixis
Fig. 13 Changes in peroxide value of oil in stored nuts. Mean values of 10 determinations on nuts
stored at 30°C and 55% RH. (From Ref. 114.)
in the presence of air and in the absence of any supplemented antioxidant will lead to fat
autoxidation. This will influence oil quality in terms of both flavor and physical performance,
depending on the degree of oxidation.
Lipid autoxidation also arises in a number of confectionery components, particularly nuts.
These may, of course, be a primary source of confectionery fat such as shea, illipé, or sal or
used as a key component, unmodified, of a confectionery product, e.g., as whole, small
pieces or ground nutmeats. Oxidation in nuts can occur via at least two pathways. The first is
by simple autoxidation arising from poor storage of the nuts over a period of time, i.e.,
ambient and higher temperatures in the presence of oxygen (Fig. 13) [114,115]. The relative
stability of almonds compared to pecans can be explained in part by the approximately five
times higher level of tocopherol in almonds compared to pecans.
Oxidation via the action of lipoxygenase is the second route most common in nuts and
seeds. The lipoxygenase can either be indigenous to the nuts as in the case of sunflower and
soy [116] or introduced by microbial contamination as has been observed in the case of sal
seeds [53,55,56,117]. The products of oxidation under the influence of lipoxygenase are usu
ally epoxides, which can be further modified to dihydroxy derivatives of the peanut fatty acid.
In the case of sal fat, the combined action of lipoxygenase and general hydrolytic degradation
of the oil adversely affects the physical properties of the fat, and in extreme cases the fat
becomes unsuitable for confectionery applications.
B. Lipolytic and Ketonic Rancidity

Lipolytic and ketonic rancidity are grouped together in this discussion because they both
primarily affect the lauric oils. Short-chain fatty acids have low flavor thresholds (Table 19).
Only a low level of hydrolysis, catalyzed by microbial lipases, of either palmkemel or coconut
oil is sufficient to release an organoleptically detectable level of free fatty acid. This type of
rancidity has been most prevalent in chocolate substitutes formulated with poor quality, micro-
bially contaminated cocoa powder and lauric stearins. It does not occur in products manufac
tured under good standards of hygiene as demonstrated by the wide use of lauric stearins in
substitute chocolate. Ketonic rancidity arises through microbial contamination and has been
studied by Kinderlerer and coworkers [118,119]. The microorganisms identified as being re
sponsible are xerophilic fungi of the type Eurotium amstelodami, E. chevalien, E. herbario-
rium, and P. citrinum. These fungi reduce even-chain-length fatty acids to a homologous
series of odd-chain-length aliphatic methyl ketones and secondary alcohols (C 5-C n )- Some
of the volatile off-flavor compounds that have been identified are listed in Table 20.
Table 19 Flavor Characteristics and Flavor Threshold Values of Short-Chain Free

Fatty Acids
Flavor threshold (ppm)

Flavor charactistic
Fatty acid (in milk) Milk Oil Water
Butyric Butyric 25.0 0.6 6 .8

Caproic Cowy (goaty) 14.0 2.5 5.4
Caprylic Cowy (goaty) — 350 5.8
Capric Rancid, unclean, bitter, soapy 7.0 200 3.5
Lauric Rancid, unclean, bitter, soapy 8.5 700 —
Source: Ref. 147.

422 Padley
Table 20 Volatile Off-Flavor Compounds in Rancid Coconut Oil
Rancid
Consumer coconut
Off-flavor return (mg/g) Flavor Intensity
compounds (mg/g oil) S.D. note 1:4
Pentan-2-one Trace Trace Pear drops 3

Pentan-2-ol N.D. 0 .58 ± 0 .21 Ethereal 2
Hexan-2-one 9 .3 6 ± 5.0 1 Trace Ethereal 4
Hexan-2-ol 4.88±5.45 N.D. Turpentine 3
Heptan-2-one 10 .7 2 ± 3 .5 5 5.05±0.68 Rancid almonds 4
Heptan-2-ol 9 .2 9 ± 1 .1 1 7 .0 5 ± 0 .18 Rancid coconut 4
Octan-2-one 2.57 ±0.44 N.D. Weakly ethereal 2
Nonan-2-one 19.63±0.88 2.97 ±0.07 Weakly turpentine 1
Nonan-2-ol 2.91 ±0.88 0 .3 5 ± 0 .15 Musty, stale 2
Undec-2-one 6.77 ±0.20 1.2 9 ± 0 .12 Weakly turpentine 1
Total 66.13 17.29
N .D . = n o t detected; S.D . standard deviation.

Source: Ref. 119.
VII. COCOA BUTTER ADULTERATION AND THE

DETERMINATION OF VEGETABLE FAT IN CHOCOLATE
The adulteration of food and food commodities has occurred since time immemorial. This is
very true of fats, particularly cocoa butter, over the last 100 years or more. There is no
comprehensive publication on fat adulteration, but oil and fat analysis has been well docu
mented in standard methods of analysis agreed on by international bodies such as British
Standard Methods, lUPAC, AO AC, and ISO. These have also been described in various
textbooks [53,120]. Accepted ranges of fatty acid compositions of oils and fats have been
published by FAO/WHO [ 12 1] .
A recent and notorious case of adulteration occurred in Spain when olive oil was adulter
ated with refined aniline-denatured rapeseed oil [122,123]. Olive oil is a typical case where
there is a strong commercial advantage to diluting an expensive oil with cheaper commodity
oils. Palm oil, palm fractions, and palmkemel oil have also been the subject of fraud investi
gations. At certain times, for example, there has been a commercial advantage to adulterating
palm oil with palm fractions.
In the confectionery area the adulteration of cocoa butter has been widespread. This has
become much less prevalent since the development of new analytical techniques and the inter
nal controls exercised by large responsible companies. The use of fatty acid compositions to
detect adulteration is of very limited value because of the natural variation encountered with
most oils and fats. Analyses for triglyceride and minor components, including sterols, offer
the more satisfactory approaches. Sterol analysis is useful for detecting adulteration but not
for quantitative estimation. The reason for this is that there is a wide variation in the sterol
content of individual oils and fats. This arises in part by natural variation and in part through
their partial removal during refining. Where the sterol content is widely different from that of
cocoa butter, the level of detection can be quite low, as in the case of illipé and shea, for
example [124]. The sterols and minor components of a wide range of oils and fats have been
determined [53,120].
The analysis of triglycerides, on the other hand, offers a good approach for quantifying the
composition of many fat mixtures. Triglycerides represent the bulk of the oil, and their overall
level is virtually constant, but their actual composition is subject to natural variations. Triglyc
eride analysis has been particularly useful in characterizing pure and adulterated cocoa butter
[31,32,34,125-129]. The method is based on analyzing triglycerides by high temperature gas-
liquid chromatography. This technique separates triglycerides on the basis of carbon number,
i.e., the total number of carbon atoms in the fatty acid chains (POP = 50; POSt = 52; StOSt
= 54). When the data for a wide range of genuine cocoa butters are plotted with P50 versus
P54 (C50, C52, C 54 normalized to 100), then a straight line is observed [124] (Fig. 14). This
is, in fact, a small part of the overall curve representing the C 50/C 54 relationship for all
naturally occurring vegetable fats [130,131].
The straight-line relationship for cocoa butter can be used as a vehicle to detect adulteration
by a large range of fats, in particular cocoa butter equivalents. A standard method was devel
oped for the European Union for the determination of cocoa butter equivalents in chocolate
[2]. Triglyceride analysis has also been used successfully to quantify milk fat in cocoa butter
and chocolate [132,133].
The determination of vegetable fats in chocolate presents additional problems. The total
fat has to be first determined using methods defined by the OICC, for example. A combination
of chromatographic methods including fatty acid, triglyceride, and sterol analysis will have
to be used to characterize “unknown” samples [134,135]. The determination of 2-oleo-disa-
turated (SOS) glycerides also provides a way of characterizing the type of fats present
[38,136].
The CBE method of analysis can also be used to make reasonable estimates of the level of
other fats, especially those based mainly on Cjg fatty acids, or their origin can be determined
by other means. Vegetable oils including soybean, com, rapeseed, sunflower, many ground-
Fig. 14 The determination of cocoa butter equivalents (CBE) in cocoa butter. The triglycerides with
carbon numbers 50, 52, and 54 (total number of carbon atoms in fatty acid chains) are normalized to
100%, giving P 50 , P 52 , P 54 values. (From Ref. 125.) Coberine is a commercial CBE.
424 Padley
groundnuts, and hazelnut oils fall into this category. As the C 50 content of the oil increases,
the quantitative aspect of the analysis diminishes. The CBE method in this case would deter
mine the total level of added fat but would not say specifically what was added.
Certain palm fractions can be detected in mixtures with cocoa butter using the techniques
described above. An unscrupulous operator could choose judicious blends of palm fractions
with or without other fats to try to hide the adulteration. Certain blends will be more difficult
to analyze than others. However, the combined use of silver-phase HPLC chromatography
and gas-liquid chromatographic techniques should provide a solution— at least for the more
sophisticated oil and fat laboratories.
The potential development of genetically engineered seed oils and/or enzymatically modi
fied fats with essentially the same fatty acid and triglyceride compositions as cocoa butter
would present a challenge to the analyst. These materials are not available today. Methods
of detection would rely on minor component analysis, such as sterols, hydrocarbons, and
waxes.
Cottonseed oil and other oils and fat blends that also have triglyceride carbon numbers
similar to cocoa butter would also require special treatment. In these cases the use of high
resolution capillary gas chromatography triglyceride analysis and silver ion chromatography
[77] should provide a solution. An example of high resolution capillary GC is given in Fig.
15 [35].
Many confectionery products are made up of a number of components—for example, choc
olate with inclusions and other “countlines” in which a variety of centers are enrobed in
chocolate. Many of these centers contain significant levels of other vegetable fats that rapidly
migrate into the chocolate coating [137]. Any interpretation of data must therefore attempt to
take this into account.
If a semiquantitative determination of chocolate adulteration by excessive amounts of vege
table fats is required, then a combination of modem analytical methods should provide the
answer in simple cases. The unscmpulous blending of particular fats and their fractions will
always present problems, as it no doubt does today, and the interpretation of analytical data
will depend on the analyst and the degree of complexity of the blend.
The rigorous characterization of any chocolate, e.g., cocoa solids, milk solids, and vegeta
ble fat content, can be carried out only with the aid of factory inspection, now a legal obliga
tion within the E U . In the case of factory inspection, the actual components used to fabricate
T ime (m in u te s )
Fig. 15 The determination of triacylglycerols by high resolution wide-bore capillary gas chromatogra
phy. (From Ref. 35.)
products will be available and the composition of the final product can be verified precisely
using existing analytical methods.
VIII. ADDITIONAL ASPECTS OF CONFECTIONERY FAT

APPLICATION
This chapter has been largely restricted to chocolate fat applications. Some other applications
of fats in confectionery are much less demanding as regards the physical behavior of the fat;
other applications have different physical requirements altogether. Some brief examples are
given below. The reader is referred to a number of specialized textbooks, notably those of
Manley [140] and Jackson [148].
The specialized area of low calorie fats is covered in Chapter 19.
A. Toffee Fats
One of the main attributes required from the fat in recipes such as for caramel toffee is good
flavor, in particular butterlike flavor. In other formulations, as in caramel, because aqueous
sugar solution or sugar glass is the continuous phase, the physical properties of the fat are not
critical provided that the melting behavior is not detrimental to mouthfeel, i.e., significant
solid fat at about 37°C. For this reason butterfat might be regarded as the preferred fat from
the flavor point of view but certainly not from the economic aspect. A wide range of fats are
therefore used depending on the application and processing/marketing requirements. Typi
cally, partially hardened fats with slip melting points in the range 27-40°C are used in many
products, some with added flavors or flavor precursors to provide a buttery flavor note.
B. Biscuit Cream Fillings

Biscuit cream fillings are typically flavored mixtures of aerated fat and sugar. A major require
ment is good mouthfeel, ideally cool melting, coupled with good aeration properties and
ability to adhere to the biscuit. Fats fulfilling these requirements range from palmkemel and
partially hardened palmkemel oils to trans-hardened oils and fats with appropriate slip melting
points, e.g., 27-29°C, depending on the application.
C. Ice Cream Couvertures

The most widely used couvertures for ice cream are based either on chocolate, i.e., cocoa
butter based, or on substitute chocolate in which the fat phase is almost universally coconut
oil. Some other fat blends such as mixtures of palm oil, palmkemel oil, and partially hardened
fats with slip melting points of approximately 2T C etc. are also used, however.
Whether the ice cream product is enrobed by dipping or in a conventional enrober, the
important requirement is to be able to control the weight of chocolate used for enrobing at a
time when the chocolate is crystallizing rapidly. This is done by a combination of viscosity
control (fat, emulsifier level) and process manipulation, e.g., air jets to remove excess choco
late from the products. For example, in those operations where ice cream is enrobed by
dipping, the fat content of the couverture will be between 45 and 60%.
In the case of cocoa butter it is interesting to note that for enrobed ice cream products the
chocolate is not usually tempered and crystallizes in the less stable form IV (which is stable
at ice cream temperatures). Thus the chocolate melts in the mouth at a lower and more accept
able temperature.
426 Padley
D. Center Filling Fats

Pralines are combinations of sugar, cocoa, nut paste, and confectionery fat. In this case the
selection of the confectionery fat is dependent on the product application and the mouthfeel
required. In some cases the “confectionery fat” will in fact be a liquid oil, as in hazelnut or
peanut paste. Where firmer textures are required, the melting profile of the fat can be raised
progressively, ultimately to that of cocoa butter. The fats used are therefore very diverse and
range from butterfat to fractions of coconut and palmkemel, fractions of palm oil, hardened
fats and their fractions, etc. Apart from mouthfeel and its retention during the life of the
product, another preferred requirement is that these fats can be crystallized without the need
to temper the fat as in the case of cocoa butter.
An additional consideration with many of these products in which a fat-continuous center
is enrobed in chocolate is the problem of fat migration [137]. This occurs primarily from the
softer center to the chocolate coating, although the migration occurs in both directions via the
liquid oil phase. This migration causes a significant softening of the chocolate coating and in
many cases increases the susceptibility to bloom (see Section V). The rate of fat migration is
clearly dependent on the product, solid fat content, and temperature. However, migration will
be virtually complete in most systems held at 28°C for 14 days.
To minimize any textural changes or reduction in melting point, then, wherever feasible a
center filling fat that is compatible, or most compatible, with cocoa butter should be chosen
(assuming cocoa butter is the fat phase of the coating). For example, center filling fats based
on lauric oils will, on migration, significantly reduce the hardness of the chocolate compared
to a similar fat based at least in part on SOS-type triglycerides.
The adverse publicity given to a range of fats on health grounds has encouraged the indus
try to develop unhardened, non-lauric alternatives based on palm fractions. The postcrystalli
zation and subsequent change in mouthfeel for many of these products has recently been
overcome by the judicious addition of unsymmetrical glycerides to the POP-rich fraction
[138].
E. Fats as Moisture Barriers

The subject of fats as moisture barriers is covered to some extent in Chapter 17. However,
one major aspect is mentioned here. For those confectionery products that rely on the appeal
of texture contrast created by the use of crisp wafers or similar material together with choco
late and fat fillings, it is important to prevent, or at least minimize, the uptake of water by
the wafer. This will retain product quality and extend the acceptable shelf life. This is not
always achieved, and the final result strongly depends on product design. Invariably, fat-based
coatings, particularly chocolate or chocolate couverture (substitute chocolate), are used as a
moisture barrier in confectionery products. Moisture transfer is clearly dependent on the per
meability of the coating used. This depends primarily on the solid fat content; the higher the
percent solid fat, the lower the permeability. A severe limitation to this is, however, the
potential for the coating to crack, thus allowing rapid ingress of moisture. A compromise must
be reached between overall product appeal, moisture barrier thickness, and melting profile and
cost. For further background reading there are numerous papers and reviews [141-144].
As a final comment the reader will have noted that no reference has been made to the use
of animal fats, notably beef tallow. The presence of unsymmetrical triglycerides, i.e., of the
SSO type, means that these fats form eutectics with cocoa butter. In addition there is a virtual
embargo, based on consumer response, to the use of these fats in high quality confectionery.
Those interested in gaining further insight into the interaction of glycerides, particularly
binary interactions of pure triglycerides and the phase behavior of fats, are directed to reviews
by Rossell [149] and Timms [150].
IX. CONCLUSION
The use of fats in confectionery products in becoming increasingly sophisticated with our
improved understanding of the factors that influence physical properties and the increased
demand from the marketplace. New developments will continue, driven by new technologies
and changing attitudes to nutrition.
ACKNOWLEDGMENT
I am very grateful for the support and encouragement given over many years by Loders
Croklaan.
REFERENCES
J. J. Scheu, The changing patterns of cocoa bean supply and demand, Manuf. Confect, p. 41
(1995). Also Cocoa report, Manuf. Confect. 75: 14 (September 1995).
2 . EU Council Directive 73/241/EEC of 24 July 1973 on The Approximation of the Laws of the
Member States Relating to Cocoa and Chocolate Products Intended for Human Consumption and
EU Working Paper, 1984. Also Commission of the European Communities; Directorate General
for Internal Market and Industrial Affairs. Working Group on Methods of Analysis for Foodstuffs
(Cocoa and Chocolate). Proposal by UK Delegation III/A/3.
3. B. W. Minifie, Chocolate, Cocoa and Confectionery: Science and Technology, 2nd ed., AVI,
Westport, CT, 1980.
4. S. T. Beckett, Industrial Chocolate Manufacture and Use, Blackie, London, 1994.
5. J. J. Urbanski, Chocolate flavour—origins and descriptions, Manuf. Confect. 72: 69 (November
1992).
6 . J. Kleinert, Quality control on chocolate, Zucker Siiss. Wirtsch. 44: 52 (1991).
7. W. J. Shaughnessy, Cocoa beans: planting, fermentation and effect on flavour, Manuf. Confect.
72: 51 (November 1992).
8 . E. Minson, Chocolate manufacture—beans through liquor production, Manuf. Confect. 72: 61
(Nov. 1992).
H. R. Kattenberg, The application of cocoa powders in chocolate confectionery, Manuf. Confect.
75: 73 (May 1995).
10. W. Rostagno, Particle size of non-fat solids in chocolate, Manuf. Confect. 49: 81-85 (May 1969).
11. E. A. Niediek, Particle size reduction, in Industrial Chocolate Manufacture and Use (S. T. Beck
ett, Ed.), Blackie, London, 1994, pp. 83-89.
12. E. H. Steiner, A new rheological relationship to express the flow properties of melted chocolate,
Rev. Int. Choc. 13: 290 (1958).
13. NCA (National Confectioners Association), Method for determination of Casson yield value and
plastic viscosity of chocolate, Manuf. Confect. 68: 12-75, 76-81 (January 1988).
14. J. Chevalley, Chocolate flow properties, in Industrial Chocolate Manufacture and Use (S. T.
Beckett, Ed.), Blackie, London, 1994, p. 139.
15. T. L. Harris, Surface Active Lipids in Chocolate, SCI Monograph 32, 1968, pp. 109-122.
16. H. F. Banford, K. G. Gardiner, G. R. Howat, and A. F. Thomson, The use of polyglycerol
polyricinoleate in chocolate, Confect. Prod. 359-365 (1970).
17. R. L. Wille and E. S. Lutton, Polymorphism of cocoa butter, J. Am. Oil Chem. Soc. 43: 491
(1966).
18. K. Larsson, Molecular arrangement in glycerides, Fette Seifen Ansstrichm. 74: 136 (1972).
428 Padley
19. K. Larsson, Lipids—Molecular Organisation, Physical Functions and Technical Applications,

Oily Press, Dundee, Scotland. 1994.
20 . K. Sato, Polymorphism of pure triacylglycerols and natural fats, in Advances in Applied Lipid
Research, Vol. II (F. B. Padley, Ed.), JAI Press, New York, in press, 1996, pp. 213-268.
21 . R. B. Nelson, Pumps and tempering, in Industrial Chocolate Manufacture and Use (S. T. Beckett,
Ed.), Blackie, London, 1994, p. 211.
22. M. G. Reade, The natural rate of solidification of chocolate, Manuf. Conf. 65: 59 (January 1985).
23. P. G. Allen, Chocolate temper measurement, Manuf. Confect. 85: 91 (1995).
24. M. G. Reade, Temper and temper testing. Parts I and II, Manuf. Confect. 60: January and Febru
ary (1980), pp. 29-32 and 51-54, respectively.
25. D. J. Cebula, K. M. Dilley, and K. W. Smith, Continuous tempering studies on model confection
ery systems, Manuf. Confect. 131 (1991).
26. D. J. Cebula, D. J. M. Clements, M. J. W. Povey, and P. R. Smith, Neutron diffraction studies
of liquid and crystalline trilaurin, J. Am. Oil Chem. Soc. 69: 130 (1992).
27. R. A. Martin, Chocolate, Adv. Food Res. 31: 211 (1987).
28. D. J. Cebula and K. M. Dilley, Personal communication; published by agreement with Coders
Croklaan b.v., Holland.
29. A. Yasuda and K. Mochizuki, The behaviour of triglycerides under high pressure, in High Pres
sure and Biotechnology (C. Balny, R. Hayashi, and K. Heremans, Eds.), Colloque INSERM,
John Libby Eurotext Ltd. 224, 1992, pp. 255-259.
30. M. R. Mackley, Extrusion process, Br. Patent GB 2270828A (September 1992).
31. F. B. Padley and R. E. Timms, The determination of cocoa butter equivalents in chocolate, Chem.
Ind. {Lond.) i 75: 918.
32. A. Fincke, Triglyceride distribution by carbon number in cocoa butter substitutes, Deut. Lebens.
Rundsch. 76: 187-192 (1980).
33. H. R. Kattenberg, The quality of cocoa butter, Manuf. Confect. 61: 32 (January 1991).
34. E. Geeraert and P. Sandra, Capillary GC of triglycerides in fats and oils. Part II. The analysis of
chocolate fats, J. Am. Oil Chem. Soc. 64: 100-105 (1987).
35. G. J. Sassano and B. S. J. Jeffrey, Gas chromatography of triacyl glycerols in palm oil fractions
with medium polarity wide bore columns, J. Am. Oil Chem. Soc. 70: 111 (1993).
36. S. Chaiseri and P. S. Dimick, Cocoa butter—its composition and properties, Manuf. Confect. 67:
p. 115 (1987).
37. B. S. J. Jeffrey, Silver-complexation liquid chromatography for fast high resolution separations of
triacylglycerols, J. Am. Oil Chem. Soc. 68: 289-293 (1991).
38. T. Tagaki and Y. Ando, Stereospecific analysis of mono-unsaturated triacylglycerols of cocoa
butter, J. Am. Oil Chem. Soc. 72: 1203 (1995).
39. T. Okawachi and N. Sagi, Confectionery fats from palm oil, J. Am. Oil Chem. Soc. 62 : 421
(1985).
40. L. Wesdorp, Liquid-multiple solid phase behaviour in fats, Ph.D. Thesis, Technical Univ., Delft,
Holland, 1990.
41. J. Yano, S. Ueno, and K. Sato, FTIR study of polymorphic transformations in SOS, POP, POS,
J. Phys. Chem. 97: 12967-12973 (1993).
42. N. V. Lovegren, M. S. Gray, and R. O. Feuge, The effect of liquid fat on melting point and
polymorphic behaviour of cocoa butter and cocoa butter fractions, J. Am. Oil Chem. Soc. 52:
108 (1976).
43. C. M. Savage and P. S. Dimick, Influence of phospholipids on the crystallisation of hard and soft
cocoa butters, Manuf. Confect. 75: 111 (May 1995).
44. S. Chaiseri and P. S. Dimick, Dynamic crystallisation of cocoa butter. I. Characterisation of
simple lipids in rapid and slow nucleating cocoa butters and their seed crystals, J. Am. Oil Chem.
Soc. 72: 1491 (1995).
45. S. Chaiseri and P. S. Dimick, Dynamic crystallisation of cocoa butter. II. Morphological, thermal
and chemical characteristics during crystal growth, J. Am. Oil Chem. Soc. 72: 1497 (1995).
46. F. B. Padley, C. N. Paulussen, C. J. Soeters, and D. Tresser, The improvement of chocolate
using mono-unsaturated triglycerides SOS and POS, Rev. Int. Choc. 27: 226 (1972).
47. D. W. Lehrian and P. G. Keeney, Triglyceride characteristics of cocoa butter from cacao fruit
matured in a microclimate of elevated temperature, J. Am. Oil Chem. Soc. 57: 6 6 (1980).
48. J. Kleinert, Pressed and extracted cocoa butter, Rev. Int. Choc. 19: 142 (1964).
49. U. Braceo, R. Horman, and R. Viani, Isolation and identification of new compounds in cocoa
shell fat, Riv. I tal. Sost. Grasse 53: 167 (1976).
50. M. Weyland, Cocoa butter fractions: a novel way of optimising chocolate performance, Manuf.
Confect. 72: p. 53 (1992).
51. R. L. Best, A. Crossley, S. Paul, H. Pardun, and C. J. Soeters, Improvements in or relating to
cocoa butter substitutes, Br. Patent 827,172 (1956).
52. M. G. A. Willems and F. B. Padley, Palm oil: quality requirements from a customers point of
view, J. Am. Oil Chem. Soc. 62 : 454 (1985).
53. F. D. Gunstone, J. L. Harwood, and F. B. Padley, Lipid Handbook, 2nd ed.. Chapman and Hall,
London, 1994, p. 94.
54. S. J. Kershaw and E. Hardwick, Heterogenicity within commercial contract analysis of samples
of shea nut kernels, J. Am. Oil Chem. Soc. 58: 706 (1981).
55. N. V. Bring!, Non Traditional Oilseeds and Oils of India, Oxford Univ. Press and IBH, New
Delhi, 1987.
56. N. V. Bring!, R. E. Timms, and F. B. Padley, Fatty acids and glyceride composition of sal fat,
Chemind. (Lond.) 1982: 805 (1982).
57. B. S. J. Jeffrey and F. B. Padley, Chinese vegetable tallow and its contamination with stillingia
oil, J. Am. Oil Chem. Soc. 68: 123 (1991).
58. G. Brooks, N. A. Morrice, C. Ellis, A. Aitken, A. T. Evans, and F. J. Evans, Toxic phorbol
esters from Chinese tallow. Toxicon 25(11): 1229-1233 (1987).
59. C. J. Soeters, C. N. Paulussen, F. B. Padley, and D. Tresser, Br. Patent 1,390,936 (April 1971).
60. R. J. Davies and J. E. Holds worth. Synthesis of lipids in yeast: biochemistry, physiology, produc
tion, in Advances in Applied Lipid Research, Vol. I (F. B. Padley, Ed.), JAI Press, New York,
1992, pp. 65-194.
61. A. R. Macrae, Fat process and composition, Br. Patent 1577933 (1976).
62. A. R. Macrae, Lipase catalysed interesterification of fats, J. Am. Oil Chem. Soc. 60: 291 (1983).
63. P. Eigtved, Enzymes and lipid modification, in Advances in Applied Lipid Research, Vol. I
(F. B. Padley, Ed.), JAI Press, New York, 1992, pp. 1-64.
64. S. R. Moore and P. T. Quinlan, Modification of triglycerides by lipases, INFORM 4: 580
(1993).
65. Y. Hashimoto, Production of cocoa butter like fats by enzymic transesterification Bioprocess Tech-
nol. 16: 337-352 (1993).
66 . M. H. Gordon, R. E. Timms, and F. B. Padley, Factors influencing the use of vegetable fats in
chocolate, Fette Seifen Anstrichm. 81: 116 (1979).
67. G. Talbot, Vegetable fats, in Industrial Chocolate Manufacture and Use, 2nd ed. (S. T. Beckett,
Ed.), Blackie, London, 1994, pp. 242-251.
68 . G. M. Chapman, E. E. Akehurst, and W. B. Wright, Cocoa butter and confectionery fats. Studies
using programmed temperature. X-ray diffraction and differential scanning calorimetry, J. Am. Oil
Chem. Soc. 48: 824 (1971).
69. R. E. Timms, The phase behaviour of mixtures of cocoa butter and milkfat, Lebensm. Wiss.
Technol. 13: 61 (1980).
70. B. Peterssen, Pulsed NMR method for determining solid fat content. Part II. Cocoa butters and
equivalents in blends with milkfat, Fette Seifen Anstrichm. 88: 128 (1986).
71. J. Schlichter-Aronhime, J. S. Sarig, and N. J. Garti, Reconsideration of polymorphic transforma
tions in cocoa butter using DSC, J. Am. Oil Chem. Soc. 65: 1140 (1988).
72. J. G. P. Verhey, Physical properties of dried milk in relation to chocolate manufacture, Neth. Milk
Dairy J. 40: 261-268 (1986).
73. L. B. Campbell, D. A. Anderson, and P. G. Keeney, Hydrogenated milk fat as an inhibitor of fat
bloom, J. Dairy Sci. 52: 976 (1969).
74. R. E. Timms, Processing of palm kernel oil, Fette Seifen Anstrichm. 88: 294 (1986).
75. J. B. Rossell, Fractionation of lauric oil, J. Am. Oil Chem. Soc. 62: 385 (1985).
430 Padley
76. W. Soon, Speciality Fats versus Cocoa Butter, Privately published, Malaysia, 1991. Fax: 60
373 47162.
77. K. W. Smith, J. M. Perkins, B. S. J. Jeffrey, and D. L. Phillips, Separation of cis and trans
triacylglycerols in trans hardened confectionery fats by silver ion high performance chromatrogra-
phy, /. Am. Oil Chem. Soc. 71: 1219 (1994).
78. A. D. Hughes and K. W. Smith, Non-post hardening fats, WO Patent 9424882 (1994).
79. D. Chapman, The Structure of Lipids, Methuen, London, 1965.
80. M. Garti and K. Sato, Crystallisation and Polymorphism of Fats and Fatty Acids, Marcel Dek-
ker. New York, 1988.
81. D. Waddington, Fats and Oils: Chemistry and Technology, Applied Science, London, 1981,
p. 25.
82. J. C. van der Enden, A. Haighton, K. van Putte, L. F. Vermaas, and D. Waddington, Fette
Seifen Anstrichm. 80: 180 (1978).
83. J. C. van der Enden, J. B. Rossell, L. F. Vermaas, and D. J. Waddington, Determination of the
solid fat content of hard confectionery butters, J. Am. Oil Chem. Soc. 59: 433 (1982).
84. D. J. Cebula and K. W. Smith, DSC of confectionery fat triglycerides, J. Am. Oil Chem. Soc.
68: 591 (1992).
85. D. J. Cebula and K. W. Smith, DSC of confectionery fats. Part II. Effect of blends and minor
components, J. Am. Oil Chem. Soc. 69: 992 (1992).
86 . A. R. M. Ali and P. S. Dimick, Thermal analysis of palm mid fraction, cocoa butter and milk
fat blends by differential scanning calorimetry, J. Am. Oil Chem. Soc. 71: 299 (1994).
87. D. G. Manning and P. S. Dimmick, Interpreting the thermal characteristics of cocoa butter using
differential scanning calorimetry, Manuf. Confect. 63(8): 73 (1983).
88 . V. P. Gibon, P. Blainpain, F. Durant, and C. Deroanne, Belg. J. Food Chem. Biotech. 40(5):
119 (1985).
89. E. Kaisersberger, The application of heat flux DSC for the characterisation of edible fats. Anal.
Proc. 27: 64 (1990).
90. J. Kleinert, Bloom on chocolate. Rev. Int. Choc. 8: 72 (1953).
91. J. Kleinert, Studies on the formation of bloom and methods of delaying it. Paper presented at the
15th Annual Production Conf. of the Pennsylvania Manufacturing Confectioners Assoc., Lancas
ter, PA, April 1961.
92. C. Giddy and E. T. Clerc, Polymorphism of cocoa butter and its importance to the chocolate
industry, Int. Choc. Rev. 16: 548 (1961).
93. J. S. Aronhime and N. Garti, Solidification and polymorphism in cocoa butter and bloom prob
lems, in Crystallisation and Polymorphism of Fats and Fatty Acids (N. Garti and K. Sato, Eds.),
Marcel Dekker, New York, 1988, p. 363.
94. L. B. Campbell and P. G. Keeney, Developments in fat bloom research on dark chocolate coat
ings, Manuf. Confect. 48: 77 (1968).
95. W. Andersson, Fat bloom and phase changes, Int. Choc. Rev. 18: 92 (1963).
96. J. W. Hagemann and J. A. Rothfus, Computer modelling of packing arrangements and transitions
in saturated-cis-unsaturated mixed triglycerides, J. Am. Oil Chem. Soc. 69: 429 (1992).
97. T. Arishima and K. Sato, Polymorphism of POP and SOS. Part III. Solvent crystallisation of P2
and jSj polymorphs, J. Am. Oil Chem. Soc. 66: 1014 (1989).
98. T. Arishama, N. Sagi, H. Mori, and K. Sato, Polymorphism of POS. I. Occurrence and polymor
phic transformation, J. Am. Oil Chem. Soc. 68: 710 (1991).
99. K. Sato, T. Arishima, Z. H. Wang, K. Ojima, N. Sagi, and H. Mori, Polymorphism of POP, SOS.
Part I. Occurrence and polymorphism transformation, J. Am. Oil Chem. Soc. 66: 664 (1989).
100 . S. de Jong, T. C. van Soest, and M. A. van Schaick, Crystal structure and melting points of
unsaturated triacylglycerols in the jS-phase, J. Am. Oil Chem. Soc. 68: 371 (1991).
101. T. Koyano, I. Hachiya, and K. Sato, Fat polymorphism and crystal seeding effects on fat bloom
stability of dark chocolate, Food Struct. 9: 231 (1990).
102. I. Hachiya, T. Koyano, and K. Sato, Seeding effects on the behaviour of cocoa butter and dark
chocolate solidification. Part I. Kinetics of solidification, J. Am. Oil Chem. Soc. 66: 1757
(1989).
103. I. Hachiya, T. Koyano, and K. Sato, Part II. Physical properties of dark chocolate, J. Am. Oil
Chem. Soc. 66: 1763 (1989).
104. D. J. Cebula and G. Ziegleder, Studies of bloom formation using X-ray diffraction on chocolates
after long-term storage. Fat Sci. Technol. 95: 340 (1993).
105. G. Talbot, F. Cain, and A. Hughes, Chocolate anti-bloom fat, Eur. Patents EP 530864 and EP
521549 (1991).
106. G. Talbot, Chocolate bloom, the cause and the cure, Int. Food Ing. 40-45 (1995).
107. P. J. M. W. L. Birker and J. C. G. Blonk, Alkyl chain packing in the triacyl glycerol
measured by atomic force microscopy, J. Am. Oil. Chem. Soc. 70: 319 (1993).
108. F. B. Padley, Rancidity in confectionery fats, in Rancidity in Foods, 3rd ed. (D. Allen and
R. J. Hamilton, Eds.), Blackie, London, 1994, pp. 230-255.
109. D. Allen and R. J. Hamilton, Rancidity in Foods, 3rd ed., Blackie, London, 1994.
110. T. P. Labuza, Kinetics of lipid oxidation in foods, CRC Crit. Rev. Food Technol. 2: 355
(1971).
111. E. N. Frankel, Recent advances in lipid oxidation, J. Sci. Food Agrie. 54: 495 (1991).
112. H. R. Kattenberg and L. de Muijnek, The shelf life of cocoa products as ingredients for the food
industry, Dev. Food Sci. 33: pp.311-340. (1993).
113. J. B. Rossell, Measuring resistance to oxidative rancidity. Lipid Technol. 4: 39 (1992).
114. P. C. Fourie and D. S. Basson, Predicting the occurrence of rancidity in stored nuts by means
of chemical analysis, Lebensm. Wiss. Technol. 22: 251 (1989).
115. T. Kerne, M. Messerli, J. Sheijbal, and F. Vital!, Storage of hazelnuts at room temperature under
nitrogen. Rev. Choc. Conf. Baker 8(2): 15 (1983).
116. T. Meshedani, J. Pokomy, and J. Panek, Interaction of oxidised lipids with protein, Nahrung
34: 915 (1990).
117. S. Y. Reddy and J. V. Prabhakar, Effect of triglycerides containing 9 ,10-dihydroxy stearic acid
on the solidification properties of sal fat (Shorea robusta), J. Am. Oil Chem. Soc. 62: 1126
(1985).
118. J. L. Kinderlerer and S. Johnson, Rancidity in hazelnuts due to volatile aliphatic aldehydes, J.
Sci. Food Agrie. 58: 89 (1992).
119. B. Kellard, D. M. Busfield, and J. L. Kinderlerer, Volatile off-flavour components in desiccated
coconut, J. Sci. Food Agrie. 36: 415 (1985).
120. J. B. Rossell and J. L. R. Pritchard, Analysis of Oilseeds, Fats and Fatty Foods, Elsevier,
London, 1991.
121. Codex Alimentarius Commission (1993), Report of the 14th Session of the Codex Committee on
Fats and Oils, September 1993, London. Alinorm 95/17. Issued by the Joint FAO/WHO Food
Standards Programme, Rome.
122. A. Pastana and E. Munoz, Nature 298: 608 (1982).
123. Toxic oil syndrome references archive. Fondo de Investigaciones Sanitarias, Ministero de Sani-
dady-consumo, Madrid, 1990.
124. D. Gegiou and K. Staphylakis, Detection of cocoa butter equivalents in chocolate, J. Am. Oil
Chem. Soc. 62: 1047 (1985).
125. F. B. Padley and R. E. Timms, The determination of cocoa butter equivalents in chocolate, J.
Am. Oil Chem. Soc. 57: 286-293 (1980).
126. A. Fincke, Possibilities and limits of gas chromatography of triglycerides for detecting foreign
fats in cocoa butter and chocolate fat, Deut. Lebensm. Rundsch. 78: 389-396 (1982).
127. A. Fincke, Studies on the diglyceride content of cocoa butter, Gordian 82: 246 (1982).
128. C. C. Young, New interpretation of GLC triglyceride data for the determination of cocoa butter
equivalents in chocolate, J. Am. Oil Chem. Soc. 61: 576 (1984).
129. A. Fincke, Triglyceride analysis of cocoa butter and chocolate fats according to carbon number,
Deut. Lebens. Rundsch. 76: 162-167 (1980).
130. F. D. Gunstone, The distribution of fatty acids in natural glycerides of vegetable origin, Chem.
Ind. (Lond.) 1962: 1214 (1962).
131. A. Fincke and F. B. Padley, Theoretical aspects of triglyceride composition of natural fats ex
pressed in terms of carbon number, Fette Seifen Ansstrichm. 83: 461-465 (1981).
432 Padley
132. D. O. Biltcliffe and R. Wood, The determination of milkfat in admixture with cocoa butter, J.
Assoc. Pub. Anal. 20: 69 (1980).
133. D. von. Precht, Quantitative determination of milk fat in chocolate. Fat Sci. Technol. 92: 275
(1990).
134. K. Grob, A. M. Giuffre, U. Leuzzi, and B. Mincione, Recognition of adulterated oils by direct
analysis of the minor components. Fat Sci. Technol. 96: 286-290 (1994).
135. K. Grob, H. P. Neukom, and R. Battaglia, Triglyceride analysis with glass capillary gas chroma
tography J. Am. Oil Chem. Soc. 57: 282-286 (1980).
136. H. Chaveron, M. Adenier, F. B. Padley, M. S. J. Dallas, and G. A. Jongenotter, Method of
determining mono-unsaturated symmetrical and unsymmetrical triglycerides, Ann. Falsif. Exp.
Chim. Toxicol. 74: 455-464 (1981).
137. M. Wooton, Fat migration in chocolate enrobed biscuits, Gordian 72: 95 (1972).
138. F. W. Cain, A. D. Hughes, and J. H. Pierce, Br. Patent Appl. 94 12045 (1994).
139. K. Jackson, Recipes, in Industrial Chocolate Manufacture and Use (S. T. Beckett, Ed.), Blackie,
London, 1994, pp. 258-280.
140. D. Manley, Technology of Biscuits, Crackers and Cookies, 2nd ed., Ellis Horwood, New York,
1991.
141. T. H. McHugh and J. M. Krochta, Water permeability properties of edible whey protein-lipid
emulsions, J. Am. Oil Chem. Soc. 71: 307 (1994).
142. E. Baldwin, J. M. Krochta, and M. O. Nisperos-Carriedo, Edible Film and Coatings to Improve
Food Quality, Technomic, Lancaster, PA, 1994.
143. I. Greener Donhowe and O. Fennema, The effect of relative humidity gradient on water vapour
permeance of lipid and lipid-hydrocolloid bilayer films, J. Am. Oil Chem. Soc. 69: 1081 (1992).
144. I. Greener Donhowe and O. Fennema, Water vapour and oxygen permeability of wax films, J.
Am. Oil Chem. Soc. 70: 867 (1993).
145. A. Fincke, Fatty acid composition of cocoa butter, Gordian 76: 325-327 (1976).
146. S. Chaiseri and P. S. Dimick, Lipid and hardness characteristics of cocoa butters from different
geographical regions, J. Am. Oil Chem. Soc. 66: 1772 (1989).
147. J. E. Kinsella, The flavour chemistry of milk lipids, Chem Ind. (Lond.) 1969: 36 (1969).
148. E. B. Jackson, Sugar Confectionery Manufacture, Blackie, London, 1990.
149. J. B. Rossell, The phase behaviour of glycerides, Adv. Lipid Res. 5: 353-408 (1967).
150. R. E. Timms, Phase behaviour of fats and their mixtures. Prog. Lipid Res. 23: 1 (1984).
16__________________________________
Frying Oils and Salad Oils
Timothy L. Mounts
National Center fo r Agricultural Utilization Research, Agricultural Research
Service, US. Department o f Agriculture, Peoria, Illinois
I. INTRODUCTION
The preparation of fat-containing food products uses two classes of lipid material based on
their physical state at about 25‘"C (74°F); (1) liquid oils, which include soybean, canola,
rapeseed, com, cottonseed, sunflower, safflower, peanut, and olive oils; and (2) semisolids
and solids, including lard, palm oil, coconut, and palmkemel oils.
Frying and salad oils are produced in many countries as indicated in Table 1. Soybean oil
is used in the greatest quantity in the world fats and oils industry.
II. DETERIORATION OF OILS AND FATS

All fats and oils are subject to deterioration. Hydrolysis, the reaction of fats and oils with
water, and oxidation, the chemical reaction in which oxygen combines with another substance
with the liberation of heat, are the two basic processes that result in the deterioration of
oils and fats. Oxidation is responsible for much more of the deterioration of fats and oils
than hydrolysis.
During the use of fats and oils for the frying of foods, food moisture promotes the splitting
of triglycerides to form free fatty acids (Fig. 1). Essentially, hydrolysis is the reverse of
making a fat molecule.
The deterioration of a fat or oil in the presence of oxygen is termed oxidative rancidity.
The initial step in the oxidation of an oil or fat is the addition of oxygen at or near the double
bond of a fatty acid chain to form unstable compounds generally designated as peroxides (Fig.
2) [1]. When the oxidation of an oil is followed experimentally, either by measuring the
amount of oxygen absorbed or by determining the peroxide value, the course of oxidation is
defined by the oxygen absorbed in the oil as shown in Fig. 3 [2]. During the initial or induc
tion phase, oxidation proceeds at a relatively slow and more or less uniform rate. Peroxides
are formed during this period much faster than they are destroyed so their content increases
433
434 Mounts
Table 1 Production of Salad/Frying Oils in Major Countries
Country 1989 1990 1991 1992 1993
Germany 474 524 714 710 726

Austria 65 75 75 81 84
Czechoslovakia 97 99 84 86 90
Hungary 66 74 65 95 81
Poland 112 77 70 94 110
Romania 224 222 217 220 220
Former Yugoslavia 253 208 205 190 185
East Europe 752 680 642 685 685
Egypt 362 370 380 410 430
S.Africa Rep. 162 194 220 231 229
Canada 230 241 275 293 338
United States 2777 2785 2862 2944 2829
Israel 66 69 73 75 76
Japan 165 181 174 165 170
Source: Oil World Statistics Update, Sept. 2, 1994.
more or less in parallel with the oxygen absorption. Thus, after a certain critical amount of
oxidation has occurred, the reaction enters a second phase characterized by a rapidly accelerat
ing rate of oxidation. The point at which the sample begins to smell and taste rancid coincides
with the beginning or early part of the second phase. As oil oxidation continues, with time,
the peroxides that are formed decompose to generate volatile and nonvolatile compounds that
contribute to flavor and odor deterioration of oils and fats. The extreme stages of oxidation,
polymerization, and degradation are accompanied by rapid increase in the viscosity of the oil.
There is considerable variation among fats in the manner in which their oxidation and
accompanying flavor deterioration proceeds. The amount of oxygen that must be absorbed to
produce rancidity will vary according to (1) the composition of the oil, (2) the presence or
absence of antioxidants and prooxidants, and (3) the conditions of oxidation. Generally, oxy
gen absorbed will amount to about 15-150% of the oil by volume or 0.02-0.20% by weight.
Fats high in oleic acid and low in linoleic acid will become rancid after the absorption of less
oxygen than fats in which the ratio is reversed.
The relative oxidation rates of pure fatty acid esters are shown in Table 2 [3]. This table
is based on a common factor of 1 that was arbitrarily assigned to the oxidation rate of oleic
H 0
I I!
H -c-0 -C " R H-C-O H
1I oII H -Q -H
1
1
0
II ^
O
II
H -C -O -C -R H - C -O -C -R + H O -C -R
Heat
II 0II I
I
0
II
H - C -O -C -R H -C -O -C -R "
Water
Triglyceride Diglyceride Free Fatty Acid
Heat
Fig. 1 Hydrolysis of fats and oils.

Frying Oils and Salad Oils 435
Heat
H H
Metals Polymers
1 1 Oxygen
R -C = C -R R -C -
Light 1
0 Acids
1 Alcohols
Lipid 0 Esters
(with unsaturated 1
Aldehydes
fatty acid) H
Ketones
Hydroperoxide Lactones
Aromatics
Hydrocarbons
Fig. 2 Oxidation of fats and oils.
acid. The rate at which oxygen is absorbed is markedly accelerated by heat and also by
exposure to light, particularly in the ultraviolet and near-ultraviolet range [4-6].
III. FLAVOR, ODOR, AND STABILITY

Many fats and oils develop off-flavors during storage. This has been called “reversion,” be
cause the oil flavor is reverting to the original flavor of the crude oil. However, as discussed
by Smouse [7], it is more accurately called “retrogression” because the flavor retrogrades
from a better to a worse condition. Also, aged oils do not develop off-flavors identical with
those of the crude oils, but for any given type of oil the off-flavor isâlways specific and
characteristic. Although all oils develop off-flavors, some do so at a faster rate than others.
Soybean and marine oils develop “fishy” flavors, lard and tallow develop “animal” flavors,
rapeseed and linseed oils develop “grassy/painty” flavors, and palm oil develops “painty/
rancid” flavors. All of the efforts to prevent the development of off-flavors in oils are aimed
at slowing down the rate at which they develop. Multivalent metals, such as iron or copper,
act as prooxidants when present with triglycerides. Early research [8] showed that metal inac-
Tim e
Fig. 3 Stages of progressive deterioration of fats through oxidation.

436 Mounts
Table 2 Relative Reaction Rates of Fatty Acids

with Oxygen
Approximate relative
Fatty acid oxidation rate
Oleic acid (18:1) 1

Linoleic acid (18:2) 10
Linolenic acid (18:3) 25
Source: Ref. 3.
tivators, which chelate or sequester these metals, are very effective at reducing the rate of
reversion in soybean oil. This was a major improvement in processing for soybean oil stability
and is practiced universally by the industry. Research [9] has indicated that the addition of
citric acid to canola oil is important for flavor stability in the oil.
A good discussion of the theories of soybean oil reversion is presented by Smouse [7],
who includes 121 citations to pertinent research. Each theory is discussed in detail. These
theories concern linolenic acid, isolinoleic acid, oxidative polymers, phospholipids, nonsapon-
ifiables, multivalent metals, and singlet oxygen. Each of the theories plays a role in soybean
oil flavor deterioration.
IV. ADDITIVES AND STABILITY

A. Metal Chelating Agents
Perhaps the most important additive used with edible oils to enhance stability during storage
and use is citric acid or a similar metal chelating agent [10]. Traces of metals in a reactive
form can act as catalysts that promote the autoxidation of fats and oils. Concentrations of iron
as low as 0.1 ppm or of copper as low as 0.01 ppm can accelerate the deterioration of soybean
oil [11]. If an oil has higher concentrations of metals, it is necessary to first lower the content
by treating the oil with a vigorous acid wash, e.g., with very dilute phosphoric acid, or by
using activated bleaching earth to absorb metals and metal complexes from the oil. A small
amount of citric acid (i.e., 0.05% w/w citric acid/oil) can be added to enhance the effective
ness of a bleaching earth. Other acids that have been used to chelate metals include ascorbic,
phosphoric, tartaric, and ethylenediamine tetraacetic acid. It should be noted that citric acid
does not improve the stability of olive oil [12].
B. Antioxidants
Antioxidants effectively improve the stability of fats and oils with lower levels of natural
antioxidants such as tocopherols and with lower levels of polyunsaturation. With oils that
have a high content of polyunsaturated fatty acids with significant levels of tocopherols, such
as soybean oil, antioxidants minimize the accumulation of peroxides but do not improve the
flavor stability [13]. The qualities desired in an antioxidant used to stabilize edible fats and
oils ensure that (1) its use in a product is safe; (2) it contributes no odor, flavor, or color to
the product in which it is used; (3) it is effective in the product at low concentrations; (4) it
is easily incorporated into the product; (5) it is retained after cooking processes such as baking
and frying; and (6) it is available at low cost for the application [14]. Laurie, olive, and palm
oils contain only small amounts of tocopherols compared to soybean, com, and cottonseed
oils. Synthetic antioxidants were shown to be effective in the stabilization of virgin olive oil
[12]. Butylated hydroxyanisole (BHA), butylated hydroxytoluene (BHT), and r^ rr-b u ty lh y d ro -
quinone (TBHQ) all improved the oxidative stability of olive oil during storage at 50°C
(120°F) in the dark but were not effective in preventing photooxidation of olive oil. It should
be noted that addition of antioxidants to oils that are already deteriorated will not improve the
quality or stability of the oil.
C. Nitrogen Treatment
As mentioned above, oxidation has the most detrimental effect on the quality of finished
products held in storage. Thus, exclusion of oxygen during storage is highly desirable and
practical for preventing quality deterioration [15-18]. Sparging (Fig. 4) is a practical method
for protecting finished oils from oxidative deterioration. The principle involved is saturation
of the oil with nitrogen while it is completely free of air and oxygen, i.e., after deodorization.
A sparger introduces tiny bubbles of nitrogen into the oil stream; as the gas-saturated oil falls
into the package, the effusing gas sweeps the headspace and thus removes most of the air and
oxygen from the vessel [15]. Another benefit of nitrogen sparging is the reduction of oxygen
picked up from natural leakage of air into the package. The gradual flushing action of the
sparge gas reduces the reabsorption rate of oxygen because of the pressure differential be
tween the liquid and the headspace above it [17]. A nitrogen gas sparger is illustrated in
F ig . 4.
V. OBJECTIVE METHODS FOR EVALUATING OXIDATIVE

AND FLAVOR STABILITY
A. Shelf Storage Test
The most realistic estimate of the storage life of fats and oils may be obtained by simply
storing the test oil under conditions identical to or resembling normal or anticipated storage
and handling conditions. Periodic evaluation by the available methods for detecting and mea
suring oxidation serves to pinpoint the storage time at which the oil is harmed or rendered
unacceptable by oxidation. The principal objection to this type of testing is the extensive
amount of time required to develop information.
PRO DUCT IN
438 Mounts
B. Schaal Oven Test

The accelerating effect of heat on fat and oil oxidation is a commonly used factor in speeding
up stability tests of edible oils. This is based on the estimation that for each 10°C rise in
temperature in the range of 9 7 -1 10°C, the rate of peroxide formation approximately doubles
[19]. The test is usually conducted in electrically heated convection ovens at about 6 0 -6 3°C .
Such oven testing permits estimation of fat and oil stability and antioxidant effectiveness in
days or hours rather than the weeks or even month required for normal shelf storage. Adapta
tions of the Schaal oven test have been used for the accelerated aging of edible oils prior to
sensory evaluation to assess the relative flavor stability of the oil [20].
C. Active Oxygen Method

The active oxygen (AOM) has been widely used in the fats and oils industry for measuring
the oxidative stability of fats and oils. The method measures the time (in hours) required for
a sample of fat or oil to attain a predetermined peroxide value under the specific conditions
of the test. The length of this period of time is assumed to be an index of resistance to
rancidity. The AOM is highly empirical, and close attention to details is required if reproduc
ible results are to be expected. The details of the test are included in the AOCS Official
Method Cd 12-57 [21]; however, the method has been declared surplus, effective 1996, and
is no longer recommended.
D. Oil Stability Index

The Oil Stability Index test is recommended to replace the AOM. All oils and fats have a
resistance to oxidation that depends on the degree of saturation, natural or added antioxidants,
prooxidants, and prior abuse. Oxidation is slow until this resistance is overcome, at which
point oxidation accelerates and becomes very rapid. The length of time before this rapid
acceleration of oxidation is the measure of the resistance to oxidation and is commonly re
ferred to as the induction period. In the oil stability index method for determining the induc
tion period, a stream of purified air is passed through a sample of oil or fat that is held in a
thermostated bath. The effluent air from the oil or fat sample is then bubbled through a vessel
containing deionized water. The conductivity of the water is continually monitored. The ef
fluent air contains volatile organic acids, swept from the oxidizing oil, that increase the con
ductivity of the water as oxidation proceeds. Formic acid is the predominant organic acid
formed [22]. The conductivity of the water is monitored by a computer or stripchart recorder.
The oil stability index is defined as the point of maximum change of the rate of oxidation,
and this method was adopted in 1993 as AOCS Official Method Cd 12b-92 [21].
E. Volatiles Analyses
Numerous publications have shown that volatiles analysis can correlate with flavor panel eval
uations [23-26]. AOCS Recommended Practices Cg 1-83 and Cg 2-83 (21) have been
adopted for application of these techniques. These data have identified total volatiles and/or
individual volatile compounds such as pentane, pentanal, hexanal, and 2,4-decadienal that
may serve as markers for oil stability. These volatiles are not individually responsible for the
total flavor of the oil as in a cause-and-effect relationship; however, they do increase rapidly
with increasing oil storage time and are indicative of oxidation levels. Volatile analysis by
gas chromatography is described in AOCS Recommended Practice Cg 4-94 [21]. The three
basic gas chromatographic procedures described in the recommended practice are direct injec
tion, static headspace, and dynamic (purge and trap) headspace analysis. The choice of
method will depend on available equipment and/or on the type of volatile compounds of
interest. The two headspace procedures described in the recommended practice require an
additional piece of instrumentation attached to the gas chromatograph to collect volatile com
pounds.
The static headspace method is selective toward the low molecular weight compounds such
as pentane, whereas dynamic headspace is selective toward medium and higher molecular
weight compounds such as 2,4-decadienal. The entire range of low to high molecular weight
compounds are detected by direct injection [27].
Flavor problems from nonoxidative sources are not detected by volatiles analysis. Data on
volatiles found in the literature are difficult to extrapolate to other chromatographic conditions
to determine oil stability, because the data are usually expressed in integrator counts for peaks
rather than in parts per million (ppm). Integrator counts can vary widely depending on the
chromatographic conditions used.
VI. SENSORY EVALUATION METHODS

A. Salad Oils
Salad-type oils that are liquid at room temperature are evaluated for odor and flavor to deter
mine their initial quality and/or flavor stability after storage. The American Oil Chemists’
Society (AOCS) has methods and recommended practices (Cd 8-53, Cg 1-83, and Cg 2-83)
[21], as does the American Society for Testing and Materials (ASTM) [28] for preparation
and presentation of oils for sensory testing. These practices include basic information such as
serving containers, oil temperature, and heating procedures for the oils.
The AOCS recommended practice also includes scoring scales based on a 1-10 system for
rating the quality or intensity of liquid vegetable oils. Two scales are given, one for intensity
and the other for quality. The intensity scale is based on a 1 = strong, 10 = bland (no odor/
flavor) scale. This is recommended for rating oils that have blandness as their ultimate goal.
Oils such as soy, sunflower, cottonseed, palm, low erucic acid rapeseed (canola), and saf
flower that have little or no flavor immediately after a proper deodorization are rated on this
scale. On the other hand, a quality scale could be used for oils such as com, peanut, and
olive that have natural distinctive, although desirable, flavors. Rating intensity may be less
subjective than rating an oil for quality because more exact reference standards for intensity
can be used to train judges. Flavorful oils can be rated for intensity of individual flavors;
however, overall intensity scores for these oils are not indicative of quality as the natural
flavors in addition to any off-flavors would affect the score.
The descriptors represent flavor and odors that are present from various sources ranging
from crade or partially processed oil to freshly processed oil to one that has deteriorated. For
example, crade soybean oil has beany, grassy, and hay flavors. After sufficient processing
and deodorization, soy oil has little or no odor or flavor and is usually described as bland or
as slightly nutty or buttery. If the processing or deodorizing is not adequate, then the soybean
oil may have a weak grassy or beany or hay flavor. As the oil oxidizes, odors and flavors
become detectable. In the early stages of oxidation, soybean oil may have weak to moderate
buttery odors and flavors. As oxidation continues, odors and flavors develop that are typical
of a crade or partially processed oil. Sometimes this is referred to as a reverted flavor or
retrograded flavor [7]. For soybean oils, these odors and flavors are grassy, hay, or beany.
Flavors detected in later stages of oxidation of soybean oil and canola oil include both
rancid and painty, because both linoleic and linolenic acid are present in these oils. Oils such
440 Mounts
as peanut oil that contain no linolenic acid do not develop a painty flavor, as this is a flavor
characteristic of linolenate-containing oils such as soybean and canola. Oils such as cottonseed
with high linoleic acid levels will develop a predominant rancid flavor in highly oxidized
samples. Frankel [29,30] presents reviews of fat and oil oxidation that explain both the mecha
nism of formation and the flavor significance of oxidation products.
B. Frying Oils
Methods to evaluate the room odor characteristics of heated and frying oils were developed
in the 1970s by Evans and coworkers [31,32]. The initial method included simply heating a
pan of oil in an empty laboratory. This method was upgraded by constructing small rooms
(5 X 8 X 10 ft) with controlled air flow and temperature [20,33]. Judges were instructed to rate
the overall intensity of room odor and the intensities of the individual odors detected. The
intensity scoring scale is based on a 0-10 scale (no odor to strong odor). High oleic acid
sunflower or low linolenic acid soybean oil (nonhydrogenated) heated to 190°C usually have
the weakest overall odor [34]. Low erucic acid rapeseed (canola) oil and olive oil receive
ratings of 7.0 or 8.0. Predominant odors for canola oil are fishy, acrid, and burnt. Hydroge
nated oils have a characteristic “hydrogenation” odor. Descriptions such as fruity, flowery,
and waxy are typical of hydrogenated oils.
VII. SALAD OILS, MAYONNAISE, AND SALAD DRESSINGS

Salad oil is generally a clear, light-colored vegetable oil that remains liquid at refrigerator
temperatures (40°F or 4.4°C). Most oils are refined, bleached, and deodorized to remove
flavor. Olive oil is generally not deodorized because it contributes a special flavor to salads
and foods that is considered to be desirable. Cottonseed oils are generally winterized so that
they remain liquid at refrigerator temperatures. Crystal inhibitors provide further protection
against graining at low temperatures. Oil crystallization can accelerate mayonnaise or salad
dressing separation at storage temperatures.
An excellent discussion of the preparation and use of mayonnaise and salad dressings
has recently been published [35]. Mayonnaise and mayonnaise-type dressings are emulsified
semisolid foods containing a minimum of 6 5% vegetable oil (one or a blend of two or more);
acidifying agents (vinegar mixed with less than 25% by weight of citric acid); lemon juice,
lime juice, or both; egg yolks; and, as optional ingredients, salt, sugar, com syrap, honey or
other symps, dextrose, and other seasonings such as mustard, paprika, other spices or spice
oils; and may contain ethylenediaminetetraacetate (EDTA) and antioxidants.
Stable mayonnaise emulsions can be prepared by both batch and continuous processes, and
the time and degree of agitation exert a profound influence on the stability and taste of the
final emulsion. The viscosity and stability of the product is influenced by the temperature of
the oil and other ingredients during the mixing. Because emulsion fails at temperatures above
24°C (75°F ), the initial temperatures should be 10-15°C (50-60°F) and the final product
temperature should be 1 5 -2 1 ° C (60-70°F).
Salad dressings are also emulsified semisolid foods but differ from the mayonnaise type by
using a lower level of vegetable oil (30-40%) and by employing cooked starch materials,
emulsifiers, and gums to provide stability and thickness. These food products are sometimes
confused with mayonnaise as they are similar in taste and appearance, but they are generally
cheaper.
French dressings are temporary emulsions of oil, vinegar and/or lemon juice, and season
ings. Shortly after mixing, these temporary emulsions separate, and they require shaking be
fore each use. Egg yolk and/or other emulsifying agents are used to keep the oil in suspension
to make a nonseparating French dressing.
Finished product quality is first evaluated by its appearance. The whiteness of the mayon
naise or dressing depends on the color of the eggs and oils used and the presence of some of
the optional ingredients. Injected nitrogen gas, used to control the final viscosity of the prod
uct, can help to improve the whiteness of the product. The second consideration is the flavor
and odor of the finished product, which are usually evaluated by a trained panel. Standard
microbiological procedures test the product for safety as well as quality assurance.
VIII. FRYING OILS

A. Partial Hydrogenation
Partial hydrogenation of edible oils is practiced to increase stability by the selective reduc
tion of linolenic acid. Commercially, a dual-purpose salad/cooking oil is prepared by hy
drogenation of soybean oil with nickel catalysts under selective conditions as described in
Chapter 10. The oil is hydrogenated to an iodine value of 110-115 and must be winterized
to meet the requirements of the standard American Oil Chemists’ Society cold test. Official
Method Cc 11-53 [21]. This test calls for the oil to remain clear for a minimum of 5.5 h
at 0°C. Stearine, high melting glycerides, and palmitic and stearic acid fractions are removed
by a winterization process [36]. Oil is chilled slowly to about 6°C during a 24 h period;
at this point cooling is stopped and the oil-crystal mixture is allowed to stand for 6 -8 h.
The yield of liquid oil is approximately 75-85% . By-product stearine is generally used for
shortening manufacture. Lowering of the linolenic acid content does provide increased sta
bility during the use of soybean oil as a cooking oil. It is still, however, only a partial solu
tion to the problem, and thermal oxidation will eventually produce objectionable odors and
flavors.
B. T ra n s Fatty Acids
Trans isomers of naturally occurring cw-unsaturated fatty acids are produced when liquid
vegetable or marine oils are chemically hydrogenated to produce frying fats or other hardened
fat products. An intense discussion has centered on the trans isomers formed during hydroge
nation and their potential role in coronary heart disease. The health debate has centered most
recently on work of Dutch researchers that showed that the trans fatty acids increase blood
cholesterol in humans to the same extent as saturated fatty acids [37]. In addition, researchers
at Harvard Medical School published an epidemiological study in 1993 claiming that trans
fatty acid intake was directly related to the risk of coronary heart disease (CHD) [38]. The
conclusions of the Harvard Medical School studies were criticized for biased estimates of
intake, and the scientific community was warned to be cautious about these conclusions [39].
Two subsequent studies that based their assessment of trans fatty acid intake on analysis of
the fatty acid composition of the adipose tissue of the subjects concluded that there was no
convincing evidence that trans fatty acids are important causes of CHD [40] and that there
were no correlations found between adipose tissue trans fatty acid content in men who had
died suddenly from CHD compared to living controls [41]. While there continue to be some
serious questions concerning conclusions that link trans fatty acids with CHD, a task group
of the Danish Nutrition Council recommended that the consumption of trans fatty acids be
reduced as much as possible [42]. Perhaps the most reasoned approach in this area is the
recent emphasis that the priority in dietary management of CHD should remain the reduction
of saturated fatty acids, which are the major cause of elevated cholesterol and coronary heart
442 Mounts
disease and not be diverted by the trans fatty acid factor, which fails to match the risk for
CHD that saturates pose [43].
C. Deep Frying
Deep frying is a most useful method of cooking food because it is a fast method of cooking
and most foods can be fried in under 5 min. Because of the short cooking period, it is possible
to prepare food as it is required. In this way, waste is kept to a minimum. In deep frying the
food is immersed in a very hot material that sears and seals the food, keeping in more flavor
than many other methods of cooking. The layer of fat or oil deposited on the food during
frying improves the eating quality of the food. Frying also imparts a characteristic flavor that
is different from that produced by other methods of cooking.
Multiple changes in an oil occur during its use at frying temperatures. Colors are formed ac
cording to the type of food fried and any breading used. Accelerated oxidative deterioration of
the oil occurs, with development of peroxides and polymers. Excessive foaming and an increase
in oil viscosity with formation of gummy deposits are results of oil polymerization. Differences
in oil stabilities are more apparent at high temperatures. As mentioned earlier, hydrolysis occurs
in the presence of moisture-containing foods, and this results in an increase in free fatty acids.
Polymerization can result in foaming during the frying of foods. With the development of
very high molecular weight polymers, the frying oil contains fatty acids of considerably differ
ent chain lengths. This difference in chain lengths results in foaming. Foaming is the forma
tion of small bubbles that creep slowly up the sides of the fryer. If the fat foams excessively,
it should be discarded, because foaming fat can pose a serious safety and Are hazard.
The effect of the amount of food on the foaming of a frying fat is illustrated in Fig. 5 [44].
The “zero rate” (no food frying, heating only) results in a rather high rate of oxidation,
polymerization, and foaming. Under these conditions the frying oil is unprotected by the
steam generated from moisture-containing foods or by turnover from fat absorption. More
deterioration occurs at the low rate of frying because not enough fat adsorption occurs. At the
high rate of frying, the more food fried, the more fat adsorption there is and the more rapid
the replacement of used fat with fresh fat.
As shown in Fig. 6, the amount of free fatty acid formed during frying is directly propor
tional to the amount of steam released by the food into the fat. Differences in free fatty acid
levels in frying fats do not necessarily constitute a differentiation between good and poor
Fig. 5 Foam development during deep fat frying. (Reprinted with permission from Ref. 44.)
Fig- 6 Free fatty acid development during deep fat drying. (Reprinted with permission from Ref. 44.)
quality fats in the frying system. Free fatty acid level does not correlate well with fried food
quality. Furthermore, free fatty acids are somewhat volatile at frying temperatures. Although
frying large quantities of high moisture foods tends to increase the rate of free fatty acid
development, it has a very positive effect on the life of the frying fat, because the rate at
which fresh fat is added is increased.
Methyl silicone at 0.5 -2 .0 ppm effectively retards oxidation and polymerization, thus re
ducing foaming. Frying stability increases of three- to tenfold with the antifoamer have been
confirmed in controlled laboratory frying evaluations. The degree of frying stability increase
depends on the original stability of the frying oil before the antifoamer is added; the more
stable original products show a greater increase in stability with the additive [45,46].
The mechanism by which silicone oil on the oil/air interface suppresses the thermal deterio
ration of frying oil was clarified in recent experiments [47]. The temperature of the air/oil
interface was measured by a surface thermometer while heating the frying oil and was usually
lower than that of the bulk oil in the fryer. The difference in temperatures of the surface and
bulk was greater in oil with silicone oil than in that without it. It thus appears that silicone
oil effectively slows down the convection current of frying oil.
Oil quality during frying and the quality of food produced in it are intimately related. The
degradation process has been described by Blumenthal [48] as consisting of five oil phases
(Fig. 7). The initial phase, break-in oil, is characterized by no cooked odors, no crisping of
the surface, little oil pickup by the food, and generally a white product with raw, ungelati
nized starch at the center of the fry. In the subsequent phase, fresh oil, a slight browning at
the edges of the fry is observed as well as partially cooked (gelatinized) centers, crisping of
the surface, and slightly more oil absorption. The next phase is optimum oil, in which is
obtained the golden-brown color; crisp, rigid surfaces; delicious potato and oil odors; fully
cooked centers; and optimal oil absorption. On the down side of the curve is the degrading
oil phase with darkened and/or spotty surfaces, excess oil pickup, limp product, and case-
hardened surfaces. The final phase is runaway oil, recognized by dark, case-hardened sur
faces, excessively oily products, surfaces collapsing inward, centers not fully cooked, and
off-odors and flavor (burned).
Liquid shortenings containing the methyl silicone antifoamer were introduced with claims
of the convenience of a pourable product and longer frying stabilities than the specially hydro
genated solid frying shortenings. Frying tests confirmed these claims until the antifoamer
was also incorporated into the specially hydrogenated solid frying shortenings. However, the
Mounts
Heating Tim e ------ ►
Fig. 7 Frying oil quality curve showing the five phases that a frying oil passes through during the
degradation process. (Reprinted with permission from Libra Laboratories.)
convenience of the liquid shortening, the long frying stability, and the slightly oilier food
appearance have made the liquid frying shortening type a major food service ingredient.
Before the addition of antifoamers to frying oils, foaming was the indicator used by food
service operators to decide when to replace their frying oil. Now, a frying fat that reaches the
persistent foaming stage may have gone beyond the discard point. Currently, food service
operators use several frying oil discard indicators, which include the following [49]:
Some operators rely on color, discarding the oil when it reaches a certain darkness or when
visibility is impaired at a definite distance under the oil surface.
Smoke is caused by hydrolysis and oxidation, which lower the smoke point, so some
operators replace their frying oil when smoking reaches a certain intensity.
Some operators replace their frying oil after a prescribed time has elapsed or a specified
quantity of food has been fried.
A number of test kits have been introduced to measure color, conductivity, free fatty acid,
and other factors.
Foaming is still used as an indicator by some operators and is acceptable if the amount of
foaming is not excessive.
A measure of the heat abuse taking place in an oil and an understanding of the products
formed during deep fat frying are of great interest and importance to the consumer and to the
food industry. Therefore, a review of the methods available for studying changes that take
place during deep fat frying is in order.
D. Standard Methods
Several methods to establish the quality of used frying fats have been widely used and subse
quently adopted as standard procedures.
L P olar C omponents
Total polar materials are determined by dissolving a weighed amount of fat (2.5 g) in light
petroleum ether-diethyl ether (87:13) and running it through a silica gel column that absorbs
the polar compounds [50,51]. After evaporation of the eluted solvent, the nonpolar fat is
weighed and the total polar material is estimated by difference. If desired, the polar material
remaining on the column can be eluted with diethyl ether. A level of 27% polar components
has been suggested as the limit beyond which a restaurant should discard its frying oil [52].
2. C onjugated D ienoic A cids

When polyunsaturated fatty acids are oxidized, a shift in one of the double bonds occurs,
producing a conjugated diene that can be measured by ultraviolet absorption at 232 nm [50].
The absorbance increases initially and then plateaus during frying. This has been related to
the establishment of an equilibrium between the rate of formation of conjugated dienes and
the rate of formation of polymers formed by a Diels-Alder reaction involving conjugated
dienes. This test is most useful in measuring heat abuse of polyunsaturated oils but is less
applicable to fats containing few unsaturates [53].
3. F atty Analyses and 18:2116:0 Ratio

Thompson and Aust [54] found that the total linoleic and linolenic acid content of a lightly
hydrogenated frying oil, after 100 h of frying, dropped by 50%. The relative amount of
saturated fatty acids thus increased. Augustin et al. [55] found good correlations between the
18:2/16:0 ratio in used frying oil and results from other methods.
E. Quick Tests
Several commercially available procedures have been developed to provide a rapid means of
testing for frying oil abuse. The tests are based on one chemical change taking place in the
oil and are not always accurate. However, a speedy test for estimating oil abuse may be all
that is necessary.
L D ielectric C onstant
A quick method that measures the dielectric constant has been used to estimate frying oil
degradation [56,57]. As the oil degrades, there is an increase in the dielectric constant. This
procedure is useful in some settings, but it would be difficult to adapt it to real situations
because the reading is influenced by many outside factors such as water or fat extracted from
the fried food. Nonetheless, the method is useful in defined circumstances. A portable elec
tronic instrument, the food oil sensor (FOS, Northern States Instrument Corp., Lino Lakes,
MN), measures the dielectric constant in frying fat relative to fresh oil [58]. It is suggested
that an FOS reading of 4.0 should be the limit beyond which a used fat should be discarded
[59].
2. O xifrit-Test
The Oxifrit-Test (previously the RAU test, E. Merck, Darmstadt, Germany) [60] is a colori
metric test kit that contains redox indicators that react with the total amount of oxidized com
pounds.
3. Fritest
The Fritest is a colorimetric test kit that is sensitive to carbonyl compounds (E. Merck, Darm
stadt, Germany) [61]. The mixture of sample and reagent is compared to three diagnostic
colors, with 1 representing good, 2 intermediate, and 3 bad.
446 Mounts
4. Spot Test
The spot test is a colorimetric procedure [62]. To test the oil, a drop is placed on a silica gel-
covered slide that has bromocresol green incorporated into the gel as a pH indicator. The test
monitors the free fatty acid content as an indicator of hydrolytic rancidity. The diagnostic
colors of the pH indicator are blue, green, and yellow.
5. A lkaline Contam inant M aterials, Quick Test

A colorimetric quick test produced by Libra Labs (Piscataway, N J) shows the presence and
semiquantitative concentration of alkaline contaminant materials (ACMs), such as soaps that
accumulate in used frying fats from the interaction of food materials with oil degradation
products [63]. A test kit solution containing bromophenol blue as a pH indicator, sodium
hydroxide, and hydrochloric acid is added to the oil, and the oil and solution are shaken in a
test tube. The oil quickly forms a lower layer, and the transparent upper layer assumes a color
directly proportional to the amount of soaps or ACMs extracted from the oil. Studies [64]
indicated that at 43 ppm ACM, frying oil is ready to be discarded.
In a study of nine analytical methods [65] the polymer/FOS ratio not only proved to be
adequate for monitoring the quality of the used oil but was also affected minimally by replen
ishment.
White [66] summarized recent research on comparing procedures for evaluating frying oil
abuse, i.e., polar components, conjugated dienoic acids, fatty acid analyses and 18:2/16:0
ratio, and several quick tests: dielectric constant, RAU test, Fritest, spot test, and alkaline
contaminant materials. In most studies the determination of polar components was proven to
be accurate, simple, and reproducible. White’s conclusion concerning the choosing of a
method was that the choice will depend on the accuracy of measurement desired, the money
available for establishing and running the tests, and the time allowed. Perhaps more im
portant, the choice will depend on the purpose for measuring the frying oil.
The Swedish National Food Administration (NFA) prepared a document in 1989 presenting
advice and guidelines on handling deep frying fats. An English edition was issued in June
1990 [67].
1. All the fat in the deep fat fryer must be changed before it starts smoking or foaming.
Use an indicator such as a food oil sensor or Oxifrit test to indicate when it is time
to change.
2. Strain the fat and clean the fryer once a day. Rinse carefully after cleaning. Solid
material in the fat and detergent residues accelerate breakdown of the fat. Store
strained fats at room temperature or at lower temperatures in a covered stainless steel
vessel. If iron pots are used, they should be rinsed only with hot water. Detergents
remove the protective film of polymerized fat that builds up during use.
The frying temperature should be 160-180°C (320-356T). At lower temperatures,
the product absorbs more fat. At higher temperatures the fat deteriorates more quickly.
4. Use fat that is specially intended for frying.
5. Avoid salting or seasoning the fried food over the fryer. Salt or seasoning can acceler
ate breakdown of the fat.
6. Lower the temperature when not frying, and protect the fat from light.
7. The fryer should have no iron, copper, or brass parts that come into contact with the
heated fat.
Keep a constant level of fat in the fryer. Fry a little at a time to keep the temperature
as even as possible. Prefry when large amounts are to be prepared.
9. Use a separate fryer, if possible, for frying potatoes. The fat deteriorates more rapidly
when meat or fish is fried than when only potatoes are fried.
Caution: Do not overheat. If the fat temperature rises above 300°C (5 7 2 °F ), the fat may start
to bum.
The purpose of the guidelines was to encourage employees in food establishments to pre
pare high quality fried foods. In Sweden, antifoam agents, such as silicones, are not permitted
in frying oils because they mask the natural foaming in deteriorated oils [68].
IX. SAFETY ASPECTS OF FRYING FATS AND OILS

Clark and Serbia [69] present an excellent review of the scientific literature concerning the
safety of heated fats and oils. It is well established that the heating of fats can result in the
formation of compounds with antinutritional properties. Also, studies involving exaggerated
stress of fat or oil, well in excess of the conditions actually encountered in home or commer
cial frying, have produced toxic reaction products. Cooking at such conditions yields totally
unpalatable foods. The conclusion was that although excessive cooking conditions can pro
duce toxic compounds in fats and oils, the fats and oils become unacceptable from a sensory
standpoint much earlier.
Recent research has indicated mutagenic activity of polar fractions isolated from repeatedly
used deep frying fats [70]. In these studies, samples were collected from restaurants and snack
bars after sensory indication of abuse. Of the 20 deep frying fat samples obtained, levels of
diglycerides and polymeric triglycerides in 13 of the samples exceeded the threshold level of
10% above which fats are rejected for use. The significance of these findings remains to
be determined.
X. NEW OILSEED VARIETIES

Soybean and canola varieties that yield oils with low linolenate content have been produced
by hybridization. At cooking and frying temperatures, these modified oils lack the objection
able fishy odor generated by normal soybean oil and the hydrogenation odor of partially
hydrogenated cooking oil.
The development of these genetically modified varieties has generated much interest in the
food oil industry. Currently, soybean and canola oils must be chemically altered by catalytic
hydrogenation for improved stability and extended use at high cooking and frying tempera
tures in restaurants, institutions, and elsewhere in the food industry. Theoretically, soybean
and canola oils that are naturally low in linolenate should not require hydrogenation for most
applications. In addition, foods produced from such oils would not contain the positional and
geometrical isomers of poly- and monounsaturated fatty acids produced by catalytic hydroge
nation. Studies [71-75] have reported the oxidative stability and flavor quality of modified
soybean and canola oils having less than 3% linolenate for use as cooking and frying oils.
Other efforts have focused on raising the oleate content. High oleic sunflower production
has experienced rapid growth since 1988, with high oleic sunflower seed currently accounting
for 1 0 -1 5 % of the total U.S. oil seed crop. Industry researchers believe that the potential for
specially tailored sunflower oil, particularly high oleic oil, may overshadow that of conven
tional sunflower oil [76].
Recent research [77] evaluated blends of commercial cottonseed oil and high oleic sun
flower oil to achieve ranges of linoleic acid of 9-54% and of oleic acid, 20-80% . Results
indicated that potato chips and fried potatoes prepared in 100% cottonseed oil had the highest
448 Mounts
intensities of fried food flavor; however, this positive flavor characteristic decreased with
decreasing levels of cottonseed oil in the blends. This work suggested that there may be a
limitation on the level of the oleic acid content of sunflower oil consistent with obtaining
flavor quality of the foods prepared therein.
REFERENCES
1. D. R. Erickson and G. R. List, Storage, handling and stabilization of edible fats and oils, in
Bailey's Industrial Oil and Fat Products, Voi. 3 (T. H. Applewhite, Ed.), Wiley, New York,
1985, p. 275.
2. E. G. Perkins, Formation of non-volatile decomposition products in heated fats and oils, Food
TechnoL 21: 611 (1967).
3. F. D. Gunstone and T. P. Hilditch, The union of gaseous oxygen with methyl oleate, linoleate
and linolenate, J. Chem. Soc. 1945: 836.
4. G. R. Greenbank and G. E. Holm, Photochemical oxidation of cottonseed oil, Ind. Eng. Chem.
25: 167 (1933).
5. G. R. Greenbank and G. E. Holm, Promoting the oxidation of fats and oils. Relative effectiveness
of different bands of the visible spectrum, Ind. Eng. Chem. 33: 1058 (1941).
6 . S. W. Thompson, Proceedings of the Committee on Food Research, Conference on Deterioration
of Fats and Oils, Quartermaster Corps Manual 17-7, 1945.
7. T. H. Smouse, Flavor reversion of soybean oil, in Flavor Chemistry and Fats and Oils (D. B.
Min and T. H. Smouse, Eds.), Am. Oil Chem. Soc. Champaign, IL, 1985, p. 85.
8. H. J. Dutton, A. W. Schwab, H. A. Moser, and J. C. Cowan, The flavor problem of soybean oil.
IV. Structure of compounds counteracting the effect of prooxidant metals, J. Am. Oil Chem. Soc.
25: 385 (1948).
9. K. Warner, E. N. Frankel, and T. L. Mounts, Flavor and oxidative stability of soybean, sunflower
and low erucic acid rapeseed oils, J. Am. Oil Chem. Soc. 66: 558-564 (1989).
10. K. S. Law, A review on the use of citric acid in the processing of oils and fats, Oleagineaux 39:
89 (1984).
11. J. C. Cowan, Advances in research on the flavor stability of edible soybean oil— a review. Food
TechnoL 19: 1413 (1965).
12. A. K. Kiritsakis, C. M. Stine, and L. R. Dugan, Effect of selected antioxidants on the stability
of virgin olive oil, J. Am. Oil Chem. Soc. 60: 1286 (1983).
13. T. L. Mounts, K. A. Warner, G. R. List, J. P. Friedrich, and S. Koritala, Flavor and oxidative
stability of hydrogenated and unhydrogenated soybean oils; effects of antioxidants, J. Am. Oil
Chem. Soc. 55: 345 (1978).
14. P. P. Coppen, in Rancidity in Foods (J. C. Allen and R. J. Hamilton, Eds.), Elsevier, London,
1983, p. 67.
15. L. M. Wright, Finished oil handling and storage, J. Am. Oil Chem. Soc. 53: 408 (1976).
16. T. W. Eselgroth, Inert gas: safeguard of quality. Food Eng. 23: 72 (1951).
17. Anonymous, Treatment and Protection of Edible Oils to Preserve Quality and Increase Storage
Life, Tech. Data Proc. Bull., Air Reduction Sales Co., Madison, WI, 1961.
18. L. H. Going, Oxidative deterioration of partially processed soybean oil, J. Am. Oil Chem. Soc.
45: 632 (1968).
19. V. C. Mellenbacher, The Analysis of Fats and Oils, Garrard, Champaign, IL, 1960.
20. T. L. Mounts and K. Warner, Evaluation of finished oil quality, in Handbook of Soy Oil Pro
cessing and Utilization (D. R. Erickson, E. H. Pryde, O. L. Brekke, T. L. Mounts, and R. A.
Falb, Eds.), Am. Soybean Assoc, and Am. Oil Chem. Soc., Champaign, IL, 1980, p. 262.
21. Official Methods and Recommended Practices of the American Oil Chemists' Society, 4th ed.,
AOCS, Champaign, IL, 1990.
22. J. M. deMan, F. Tie, and L. deMan, Formation of short chain volatile organic acids in the
automated AOM method, J. Am. Oil Chem. Soc. 64: 993 (1987).
23. C. D. Evans, G. R. List, R. L. Hoffman, and H. Moser, Edible oil quality as measured by
thermal release of pentane, J. Am. Oil Chem. Soc. 46: 501 (1969).
24. K. Warner, C. D. Evans, G. R. List, H. P. Dupuy, J. I. Wadsworth, and G. E. Goheen, Flavor
score correlation with pentanal and hexanal contents of vegetable oil, J. Am. Oil Chem. Soc. 55:
252 (1978).
25. H. W. Jackson, G. K. Mirmira, and W. M. Wagner, Oxidative oil stability using volatiles method
ology, J. Am. Oil Chem. Soc. 63: 117 (1986).
26. K. Warner, E. N. Frankel, and K. J. Moulton, Flavor evaluation of crude oil to predict the quality
of soybean oil, J. Am. Oil Chem. Soc. 65: 386 (1988).
27. J. Snyder, E. N. Frankel, E. Selke, and K. Warner, Comparison of gas chromatographic meth
ods for volatile lipid oxidation compounds in soybean oil, J. Am. Oil Chem. Soc. 65: 1617
(1988).
28. American Society for Testing and Materials, STP 434, Manual on Sensory Testing Methods,
ASTM, Philadelphia, PA, 1968.
29. E. N. Frankel, Chemistry of autoxidation: mechanism, products and flavor significance, in Flavor
Chemistry of Fats and Oils (D. Min and T. H. Smouse, Eds.), Am. Oil Chem. Soc., Champaign,
IL, 1985, p. 1.
30. E. N. Frankel, Volatile lipid oxidation products. Prog. Lipid Res. 22: 1 (1982).
31. C. D. Evans, H. A. Moser, G. R. List, H. J. Dutton, and J. C. Cowan, Edible oil evaluation by
room odor tests, J. Am. Oil Chem. Soc. 48: 711 (1971).
32. C. D. Evans, K. Warner, G. R. List, and J. C. Cowan, Room odor evaluation of oils and cooking
fats, J. Am. Oil Chem. Soc. 49: 578 (1972).
33. T. L. Mounts, Odor considerations in the use of frying oils, J. Am. Oil Chem. Soc. 56: 659
(1979).
34. K. Warner, Sensory evaluation of flavor quality of oils, in Flavor Chemistry of Fats and Oils (D.
Min and T. H. Smouse, Eds.), Am. Oil Chem. Soc. Champaign, IL, 1985, p. 207.
35. A. Moustafa, Salad oil, mayonnaise, and salad dressing, in Practical Handbook of Soybean Pro
cessing and Utilization (D. R. Erickson, Ed.), Am. Oil Chem. Soc., Champaign, IL, 1995, p.
314.
36. G. M. Numenz, Old and new in winterizing. J. Am. Oil Chem. Soc. 55: 396A (1978).
37. R. P. Mensink and M. B. Katan, Effect of dietary trans fatty acids on levels of high-density and
low-density lipoprotein cholesterol in healthy subjects, N. Engl. J. Med. 323: 439 (1990).
38. W. C. Willett, M. J. Stampfer, J. E. Manson, G. A. Colditz, F. E. Speizer, B. A. Rosner, L. A.
Sampson, and C. H. Hennekens, Intake of trans fatty acids and risk of coronary heart disease
among women. Lancet 34F. 581 (1993).
39. P. McKeigue, Trans fatty acids and coronary heart disease: weighing the evidence against hard
ened fat. Lancet 345: 269 (1995).
40. A. Aro, A. F. M. Kardinaal, I. Salminen, J. D. Kark, R. A. Riemersma, M. Delgado-Rodriguez,
J. Gomez-Aracena, J. K. Huttunen, L. Kohlmeier, B. C. Martin, J. M. Martin-Moreno, V. P.
Mazaev, J. Ringstad, M. Thamm, P. van’t Veer, and F. J. Kok, Adipose tissue isomeric trans
fatty acids and risk of myocardial infarction in 9 countries: the EURAMIC study. Lancet 345:
273 (1995).
41. T. L. Roberts, D. A. Wood, R. A. Riemersma, P. J. Gallagher, and F. C. Lampe, Trans isomers
of oleic and linoleic acid in adipose tissue and sudden cardiac death. Lancet 345: 278 (1995).
42. Transfedtsyrers betydning for sudhenen—Report of a task force of the Danish Nutrition Council,
Copenhagen, Denmark.
43. P. J. Nestel, Comment on trans fatty acids and coronary heart disease risk. Am. J. Clin. Nutr. 62:
522 (1995).
44. H. Lawson, Food Oils and Fats—Technology, Utilization, and Nutrition, Chapman & Hall, New
York, 1994, p. 6 6 .
45. I. P. Freeman, F. B. Padley, and W. L. Sheppard, Use of silicones in frying oils, J. Am. Oil
Chem. Soc. 50: 101 (1973).
46. S. Ohta, Effect of silicone oil on the thermal degradation of frying oil, Yushi 42: 19 (1989).
45G Mounts
47. S. Ohta, Y. Shimizu, M. Kinugawa, K. Tosaki, and Y. Tottori. Functions of silicone oil in frying
oil. X. Reconsideration of previous assumptions, Yukagaku 39: 23 (1990).
48. M. M. Blumenthal, A new look at the chemistry and physics of deep-fat frying, Food TechnoL
45: 68 (1991).
49. R. D. O’Brien, Food service use of fats and oils, INFORM 4: 913 (1993).
50. Anonymous, Evidence of purity and determination from ultraviolet spectrophotometry, Method
2.505 and Determination of polar compounds in frying fats. Method 2.507, in Standard Methods
for Analysis of Oils, Fats and Derivatives, 7th ed. (C. Paquot and A. Hautfenne, Eds.), Blackwell,
Oxford, England, 1987.
51. Anonymous, Polar components of frying fats. Method 18.074, in Official Methods of Analysis,
14th ed., AOAC, Washington, DC, 1984.
52. A. J. Paradis and W. W. Nawar, Evaluation of new methods for the assessment of used frying
oils, i. Food Sci. 46: 449 (1981).
53. M. Peled, T. Gutfinger, and A. Letan, Effect of water and BHT on stability of cottonseed oil
during frying, J. Sci. Food Agric. 26: 1655 (1975).
54. L. U. Thompson and R. Aust, Lipid changes in french fries and heated oils during commercial
deep frying and their nutritional and toxicological implications. Can. Inst. Food Sci. TechnoL J.
16: 246 (1983).
55. M. A. Augustin, A. Telingai, and L. K. Heng, Relationships between measurements of fat deterio
ration during heating and frying in RBD olein, J. Am. Oil Chem. Soc. 64: 1670 (1987).
56. V. J. Graziano, Portable instrument rapidly measures quality of frying fat in food service opera
tions, Food TechnoL 33: 50 (1979).
57. C. W. Fritsch, D. C. Egberg, and J. S. Mangnuson, Changes in dielectric constant as a measure
of frying oil deterioration, J. Am. Oil Chem. Soc. 56: 746 (1979).
58. C. W. Fritsch, Measurements of frying fat deterioration; a brief review, J. Am. Oil Chem. Soc.
58: 272 (1981).
59. L. M. Smith, A. J. Clifford, C. L. Hamblin, and R. K. Creveling, Changes in physical and
chemical properties of shortenings used for commercial deep-fat frying, J. Am. Oil Chem. Soc.
63: 1017 (1986).
60. L. B. Croon, A. Rogstad, T. Leath, and T. Kiutamo, A comparative study of analytical methods
for quality evaluation of frying fat, Fette Seifen Anstrichm. 88: 87 (1986).
61. H. V. Von Zeddelmann, Probleme bei der Beurteilung von Fritierfetten, Lebensm.-Rundsch. 69:
81 (1973).
62. H. Robem and L. Gray, A colorimetric spot test for heated oils. Can. Inst. Food Sci. TechnoL J.
14: 150 (1981).
63. M. M. Blumenthal, J. R. Stockier, and P. J. Summers, Alkaline contaminant materials in used
frying oils: a new quick test, J. Am. Oil Chem. Soc. 62: 1373 (1985).
64. M. M. Blumenthal and J. R. Stockier, Isolation and detection of alkaline contaminant materials
(ACM) in used frying oils, J. Am. Oil Chem. Soc. 62: 1373 (1985).
65. P.-F. Wu and W. W. Nawar, A technique for monitoring the quality of used frying oils, J. Am.
Oil Chem. Soc. 63: 1363 (1986).
66. P. J. White, Methods for measuring changes in deep-fat frying oils. Food TechnoL 45: 75 (1991).
67. Anon., General advice on handling frying fats (SLV FS 1990:2), National Food Administration,
Uppsala, Sweden, 1990.
68. D. Firestone, Worldwide regulation of frying fats and oils, INFORM 4: 1366 (1993).
69. W. L. Clark and G. W. Serbia, Safety aspects of frying fats and oils. Food TechnoL 45: 84
(1991).
70. G. Hageman, R. Kikken, F. Ten Hoor, and J. Kleinjans, Assessment of mutagenic activity of
repeatedly used deep-frying fats. Mutagen Res. 204: 593 (1988).
71. T. L. Mounts, K. Warner, G. R. List, R. Kleiman, W. R. Fehr, E. G. Hammond, and J. R.
Wilcox, J. Am. Oil Chem. Soc. 65: 624 (1988).
72. H. Liu and P. J. White, High-temperature stability of soybean oils with altered fatty acid composi
tion, J. Am. Oil Chem. Soc. 69: 533-538 (1992).
73. K. Warner and T. L. Mounts, Effect on frying stability of canola oils with modified fatty acid
compositions, J. Am. Oil Chem. Soc. 70: 983 (1993).
74. T. L. Mounts, K. Warner, and G. R. List, End-use performance evaluation of genetically modified
soybean oils, J. Am. Oil Chem. Soc. 71: 157 (1994).
75. T. L. Mounts, K. Warner, G. R. List, W. E. Neff, and R. W. Wilson, Low-linolenic acid soybean
oils— alternatives to hydrogenated cooking oils?, J. Am. Oil Chem. Soc. 71: 495 (1994).
76. B. Fitch Haumann, Modified oil may be key to sunflower’s future, INFORM 5: 1198 (1994).
77. K. Warner, P. Orr, and M. Glynn, Effect of fatty acid composition on fried-food flavor quality
and stability, INFORM 6: 507 (1995).
17________________________________________
Edible Coatings and Film Barriers
Thomas H. Shellhammer and John M. Krochta
University of California-Davis, Davis, California
I. INTRODUCTION
An edible coating can be defined as a thin film of edible material that is formed on a food
surface. With some materials, a stand-alone film can be formed for placement between food
components. The purpose of an edible coating or film is to control mass transfer, improve
mechanical handling, and/or modify external appearance. Controlling mass transfer involves
inhibiting the movement of moisture, permanent gases, aromas, and lipids among heteroge
neous food components or between a food and its external environment. Release of food
ingredients such as flavors, antioxidants, or antimicrobial agents is a possibility from within
the film or coating itself. Edible coating and film materials comprise three general categories
of materials: polysaccharides, proteins, and lipids. The first two categories are hydrophilic
and are therefore effective at minimizing the transport of gases (at low relative humidity) but
are poor moisture barriers. Lipids, on the other hand, are very effective moisture barriers and
somewhat ineffective gas barriers.
Lipid coatings are used primarily to inhibit moisture loss from foods and also to improve
consumer appeal by adding a glossy finish to the coated product. The application of lipid film
coatings to retard moisture loss, induce flavor changes, or alter product appearance is not a
new practice. As early as the twelfth or thirteenth centuries, the Chinese used wax to ferment
oranges and lemons [1]. Enrobing food in fat, a process called larding, was used in England
during the sixteenth century to minimize moisture loss [2]. For many years, waxes have been
used as protection against contamination and moisture loss [3]. More current roles for lipid
coatings include carrying ingredients to a food surface, protecting encapsulated materials from
moisture uptake, and reducing gas transport for the purpose of decreasing respiration in fresh
produce. Lipids are used as coating agents for a broad and diverse set of products. They are
used on fruits and vegetables, confectioneries, nuts, dried fruit, refrigerated and frozen meat,
starch-based baked products, and moisture-sensitive encapsulated ingredients. While conven
tional packaging can offer protection from mass transfer for bulk materials and individually
packaged products, edible lipid coatings offer continued protection for the product after the
453
454 Shellhammer and Krochta
conventional package has been opened. Yet the incomparable advantage of lipid coatings and
films comes from their ability to retard moisture transport within heterogeneous food materi
als. It is not feasible to include synthetic, inedible packaging materials within food products.
However, edible lipid coatings and films can conveniently perform a moisture barrier function
within a heterogeneous food without otherwise affecting food quality.
Many material-specific properties determine the appropriateness of a particular lipid for use
as a coating or film. These include blocking (self-adhering), color, compatibility with other
ingredients, crystalline structure, ductility, edibility, emulsifiability, flavor, flexibility, gas
permeability (oxygen, carbon dioxide, nitrogen, and ethylene), hardness, melting point, odor,
plasticity, resistance to bacteria and fungi, sheen and reflectivity, stability to light and oxi
dation, surface tension, and water vapor permeability [4]. Additionally, other non-material-
specific factors such as availability, price, ease of handling (i.e., bulk, slab, flakes, or pow
der), and legal restrictions (i.e., health, safety, or patents) must be considered.
H. FACTORS AFFECTING MASS TRANSPORT

IN LIPID FILMS
Barrier properties of film and coating materials are generally described through a permeability
coefficient. The permeability of a penetrant, such as water or oxygen, through a film is a
function of three factors: (1) the nature of the film material, (2) the nature of the penetrant,
and (3) interactions between the penetrant and the film [5]. The transport process is a combi
nation of sorption of the penetrant on the high concentration side of the film, diffusion of the
penetrant through the film, and desorption of the penetrant from the film on the low concentra
tion side (Fig. 1).
Sorption is a general term used to describe dispersal of a penetrant into a matrix to form a
mixture [6]. The equilibrium condition existing between the concentration of the penetrant
outside a film matrix and its concentration in the matrix (assuming a dilute concentration) can
be described through Henry’s law [7]:
Pa ( 1)
The molecular concentration outside the film material is expressed in terms of its partial
pressure, pp, (e.g., kilopascals), and the concentration of the penetrant has units of moles per
unit volume (mol/cm^). The constant describing the equilibrium relationship is a solubility
coefficient, or Henry’s constant. Figure 1 displays this relationship graphically. The concen-
>Flux of penetrant A
1 2
Fig.1 The transport of a penetrant through a film. = partial pressure of A; = molar concentration
of A.
Edible Coatings and Film Barriers 455
tration in the film at point 1 or point 2 is merely the partial pressure of A outside the film
divided by the material-specific solubility coefficient.
Diffusion is the process by which a penetrant molecule travels from one position in a
matrix to another position as a result of random molecular motions and because of an imposed
concentration gradient. This process can be described as a Fickian diffusion [8], which can
be represented by the following equation in one dimension:
J=-i ( 2)
dx
The flux J of species A is proportional to the concentration driving force (dcp^/dx) via the
diffusion coefficient or diffusivity D. Assuming both D and S are constant, Eqs. (1) and (2)
can be combined and integrated to yield a third equation that describes the permeability of the
penetrant through the film solely in terms of the partial pressure driving force,
^P a
J =P (3)
Ax
where P = permeability [g •mm/(m^ •day -kPa)], = partial pressure driving force (kPa),
Ax = film thickness (mm), and J = flux of molecule A [g/(m^-day)].
The permeability coefficient is the product of the diffusivity and solubility coefficients
(P=DS). Thus, the transport through the film can be characterized by controlling only the
concentration of the penetrant outside the film rather than having to determine its concentra
tion within the film. Determining the concentration of the penetrant in the film matrix is
achieved through sorption and desorption isotherms. The process of measuring the isotherms
is time-consuming and tedious. Therefore, researchers often focus their efforts on measuring
the permeability of the penetrant through the film material, since the partial pressures of the
components outside the film are somewhat easier to control.
Because the transport of a molecule through a film depends on both its solubility in and
diffusivity through the film, factors that affect either of these two properties influence the
barrier properties of the film. With regard to solubility, the state of the film material structure
influences the penetrant’s compatibility in the matrix. Relative polarities of the penetrant and
the matrix material strongly influence solubility, as the general principle of “like dissolves
like” is applicable [9]. Nonpolar species such as permanent gases and organic vapors are
readily soluble in hydrophobic films, whereas water, which is strongly polar, is sparingly
soluble. Consequently, hydrophobic films are permeable to hydrophobic penetrants but are
effective barriers to moisture. Such is the case with lipid films. Of course, the converse is
true, as in the case of hydrophilic films such as ethyl vinyl alcohol and edible protein films.
In these films, water is very soluble, whereas hydrophobic species are not [10,11].
Anything that affects the “tightness” of the molecular structure of the film material network
will have a significant effect on the diffusivity, and therefore the permeability, of the penetrant
molecule. Factors affecting the permeability, especially the diffusivity, are the film material
density, its free volume, coefficient of expansion, cohesive energy density, and crystallinity
(morphology and orientation) [12]. Increasing temperature increases molecular diffusivity
[13]. The increased thermal motion of lipid molecules in their liquid state, along with changes
in their solubility characteristics, increases the movement of the penetrant through the matrix
[5]. The barrier properties of lipid films are highly temperature-dependent for these reasons.
Many lipids crystallize at slightly elevated temperatures and can be a mixture of crystalline
and liquid phases at room temperature. Thermal motion of the lipid molecules is much higher
in the liquid phase than in the crystalline phase; hence the transport of penetrants in these
mixtures is much higher than in completely crystalline systems.
III. LIPID MATERIALS AVAILABLE FOR EDIBLE COATINGS

AND FILMS
Many types of lipids have been examined for use as coating or film-forming materials. These
include waxes, fats, oils, and lipid-containing materials such as chocolate liquor and choco
late. Table 1 lists the lipid materials approved by the U.S. Food and Drug Administration
(FDA) for use as edible coatings, their applications, and their federally regulated limits as
listed in the Code o f Federal Regulations (CFR). This table was produced from lists of food
ingredients in CFR title 21, chapter I, subchapter B, parts 172 and 184 [14,15]. CFR 21.172
lists food additives that are permitted for direct addition to food for human consumption,
while CFR 21.184 refers to food substances that have been reaffirmed as generally recognized
as safe (GRAS). Table 1 is a representative list of lipid coating materials rather than a compre
hensive list. Lipid materials, such as hydrogenated soybean oil, that are not listed in Table 1
can be used as coatings but are not explicitly stated as coating materials in CFR 21.172 or
21.184. Table 2 lists the physical properties of common lipid coating materials.
Waxes are naturally occurring esters of long-chain carboxylic acids (Ci6 or greater) [16].
Waxes can be classified on the basis of their source into five general groups: mineral (fossil
and petroleum), vegetable, animal, synthetic, or compound [4,17]. The mineral wax category
is dominated by paraffin wax, to a lesser degree by microcrystalline wax, but also comprises
petroleum, montan, ozocerite, and peat waxes. Of these materials, only paraffin and petro
leum wax are approved for use as a food coatings, although microcrystalline and montan
wax are approved for use on food contact surfaces as in packaging materials [18]. Paraffin
wax is a purified mixture of hydrocarbons obtained from petroleum that has a crystalline
structure and no color, odor, or taste [4]. It is sold in various grades, with the main difference
being melting point. Refiners produce paraffin waxes covering a wide range of melting points
from 27 to 69°C. Higher melting waxes, which are sold as microcrystalline, amorphous,
and petroleum waxes, are derived from heavier lubricating oil fractions of petroleum distil
lates.
Of the vegetable waxes, candelilla, camauba, and ricebran wax are approved for food use
at a level defined as “good manufacturing practices.” Cotton wax, esparto (or reed) wax, fir
wax, flax wax, Japan wax, palm waxes, and sugar cane wax can be classified as vegetable
waxes but do not have U.S. federal approval for use in or on foods. Candelilla wax is pro
duced by reedlike plants {Euphorbia antisiphilitica, Euphorbia cerifera, and Pedilanthus pa-
vonis) that grow wild in northwestern Mexico and southern Texas [4]. It is composed in large
part of hydrocarbons with some wax acid esters and alcohols. Camauba wax is a plant exudate
from the Brazilian “tree of life” {Copernica cerifera). It is the hardest, highest melting, natu
ral commercial wax [4] and is composed almost entirely of wax acid esters of C 24 and C28
carboxylic acids and the C 32 and C34 straight-chain primary alcohols [16]. Rice bran wax is
refined wax from rice bran.
The animal waxes comprise beeswax, insect (or Chinese) wax, shellac wax, spermaceti
wax, and wool wax. Beeswax is produced by honeybees for building honeycomb cells. It is
a mixture of wax esters, wax acids, and hydrocarbons. Hydrolysis of these esters yields
mainly C 26 and C28 carboxylic acids and the C30 and C 32 straight-chain primary alcohols [16].
Although not considered a lipid material, shellac is often used as a protective and decora
tive coating for foods because it forms a barrier to moisture and gas and because of its
physical toughness. “Shellac” is the word used today referring to all forms of lac. Lac is the
hardened resinous secretion of the scale insect, Laccifer lacca, and is the only commercial
resin of animal origin [19]. It is composed of more than 90% resin and less than 5% wax
[20]. Shellac has a molecular weight of approximately 1000 and forms films because of exten
sive hydrogen bonding [21]. Shellac was reaffirmed as GRAS for use as food coatings in
1990 b y th e U.S. F D A [15].
Synthetic waxes are waxes only in the sense that they possess physical properties similar
to those of natural waxes. They are made of long-chain polymers of ethylene or ethylene
oxide, halogenated hydrocarbons, montan wax derivatives, hydrogenated waxes, and many
reaction products involving fatty acids [17]. Some of these materials are accepted for use in
or on food. The primary material used from this category is oxidized polyethylene or polyeth
ylene wax. Oxidized polyethylene is the basic resin produced by the mild air oxidation of
polyethylene. The polyethylene used in the oxidation process must conform to the density,
maximum w -hex an e-ex tractab le fraction, and maximum xylene-soluble fraction specifications
prescribed by section 177.1520(c) in title 21 of the CFR in order to be permitted on selected
foods. The oxidized polyethylene has a minimum number average molecular weight of 1200,
as determined by high temperature vapor pressure osmometry, contains a maximum 5 % w/w
total oxygen, and has an acid value of 9-19 (CFR 21.172.260).
Wax materials containing a combination of various waxes or containing different sub
stances are considered “compounded” waxes [22].
The other major group of lipid coatings other than waxes consists of fatty acids and their
glyceride and sucrose esters. Of particular interest as coatings are the acetylated glyceride
monoesters of long-chain fatty acids, otherwise known as acetylated monoglycerides or com
mercially as Myvacet. Distilled acetylated monoglycerides form flexible films at room temper
ature and retain their flexibility at low temperatures. They are fairly effective barriers to
moisture and moderate barriers to permanent gases. The oxidative stability of some of these
products has seldom, if ever, been approached by any other fat [4]. Also contained in this
general category of lipids are triglycerides and mixtures thereof. These are generally termed
“fats” and “oils,” with the only difference being their melting point [23]. Fats are solid at
room temperature; oils are not. Full hydrogenation of cottonseed, coconut, peanut, soybean,
or other vegetable oils or animal oils yields a waxlike substance [4]. Increasing the melting
point of these materials through hydrogenation produces a lipid that has less liquid-like and
more viscoelastic mechanical properties, greater chemical stability, and increased moisture
barrier properties.
IV, BARRIER PROPERTIES OF LIPID FILMS

The primary use for lipid coatings is due to their moisture barrier properties. Features neces
sary for a lipid film to have a low water permeability include a regular chemical structure,
with a high degree of saturation such that crystallization or close packing is encouraged, and
the absence of highly polar groups [24]. Purified triglycerides and waxes have low water
vapor barrier properties for these very reasons [25]. Waxes, in general, are the best moisture
barriers because of their high melting point and high degree of hydrophobicity. Compared to
fatty acids, fatty alcohols, and hydrogenated or crystalline triglycerides, waxes are superior
moisture barriers [26-28]. The best barrier to moisture of all naturally occurring lipids is
candelilla wax [29,30]. Because of its hydrophobic nature and highly crystalline character,
the moisture barrier properties of candelilla wax are better than those of low density polyethyl
ene, the industry standard for efficient, inexpensive moisture barrier polymers (Table 3).
However, with regard to oxygen permeability, waxes and high melting point lipids have bar
rier properties 100-1000 times poorer than polyvinylidene chloride (Saran) or ethyl vinyl
alcohol [29,31] (Table 4). These latter two polymers represent the forefront of gas barriers
for synthetic polymers. This relationship between moisture and gas barrier properties indicates
that lipid materials are well suited as coating materials for fresh fruit and vegetables. Shellac,
Table 1 U.S. Food and Drug Administration Approvals of Lipid Materials Used for Edible Coatings
Material (CFR citation) Uses and limitations
Acetylated monoglyerides (21.172.828) The food additive is used at a level not in

excess of the amount reasonably required
to produce its intended effect in food, or in
food processing, food packing, or food
storage equipment.
Beeswax (21.184.1973) GRAS
Current good manufacturing practice results
in a maximum level, as served, of
0.065%— chewing gum
0.005%— confections and frostings
0.04%—hard candy
0 . 1 %— soft candy
0 .0 0 2 % max.— all other food categories.
Carnauba wax (21.184.1978) GRAS
No limitations other than good manufacturing
practice.
Candelilla wax (21.184.1976) GRAS
No limitations other than good manufacturing
practice.
Castor oil (21.172.876) As an antisticking agent in hard candy
production at levels not to exceed 500
ppm.
As a component of protective coating for
vitamin and mineral tablets.
Cocoa butter substitute from coconut oil, palmkemel GRAS
oil, or both oils (21.172.861) As coating material for sugar, table salt, vita
mins, citric acid, succinic acid, and spices.
In compound coatings, cocoa creams,
cocoa-based sweets, toffees, caramel
masses, and chewing sweets.
Fatty acids, including salts of fatty acids (21.172.860) As a component in the manufacture of food
grade additive at levels not to exceed good
manufacturing practice.
Methyl esters of fatty acids produced from edible fats The additive is used or intended for use at a
and oils (21.172.225) level not to exceed 3% by weight in an
aqueous emulsion in dehydrating grapes to
produce raisins, whereby the residue of the
additive on the raisins does not exceed 2 0 0
ppm.
Mineral oil (21.172.878) On raw fruits and vegetables at levels not to
exceed good manufacturing practice. In
frozen meat, as a component of hot-melt
coating at levels not to exceed 0.095% of
meat. As release agent and as sealing and
polishing agent in the manufacture of con
fectionery at levels not to exceed 0 .2 % of
confectionery.
Mono- and diglycerides (21.184.1505) GRAS
Morpholine (21.172.235) Used as the salt(s) of one or more of fatty
acids, as a component of protective coat-
Table 1 Continued
Material (CFR citation) Uses and limitations
ings applied to fresh fruits and vegetables

at a level not in excess of that reasonably
required to produce its intended effect.
Odorless light petroleum hydrocarbons (21.172.884) As a coating on shell eggs at levels not
exceeding good manufacturing practice.
Paraffin (21.172.274) Used or intended for use as a protective
coating or component of protective coat
ings for fresh grapefruit, lemons, limes,
muskmelons, oranges, sweet potatoes, and
tangerines.
Petrolatum (21.172.880) In confectionery; as release agent and as
sealing and polishing agent at levels not to
exceed 0.2% of confectionery.
Petroleum wax (21.172.886) On cheese and raw fruits and vegetables as a
protective coating and as a component of
microcapsules for spice-flavoring sub
stances at levels not to exceed good
manufacturing practice.
Petroleum wax, synthetic (21.172.888) On cheese and raw fruits and vegetables as a
protective coating at levels not to exceed
good manufacturing practice.
Polyethylene, oxidized or polyethylene wax Used or intended for use as a protective
(21.172.260) coating or component of protective
coatings for fresh avocados, bananas,
beets, coconuts, eggplant, garlic, grape
fruit, lemons, limes, mango, muskmelons,
onions, oranges, papaya, peas (in pods),
pineapple, plantain, pumpkin, rutabaga,
squash (acorn), sweet potatoes, tangerines,
turnips, watermelon, Brazil nuts, chest
nuts, filberts, hazelnuts, pecans, and
walnuts (all nuts in shells).
Ricebran wax (21.172.890) Candy, 50 ppm
Fresh fruits and vegetables, 50 ppm
Sperm oil (21.172.210) As an adjuvant for citrus fruit coatings.
Stearic acid (21.184.1090) GRAS
Sucrose fatty acid esters (21.172.859) As components of protective coatings applied
to fresh apples, avocados, bananas, banana
plantains, limes, melons (honeydew and
cantaloupe), papaya, peaches, pears,
pineapples, and plums to retard ripening
and spoiling. This material is to be used in
accordance with current good manufactur
ing practice and in an amount not to
exceed that reasonably required to
accomplish the intended effect.
Synthetic isoparaffinic petroleum hydrocarbons As a component of coatings on fruits and
(21.172.882) vegetables and as a coating on shell eggs
at levels not to exceed good manufacturing
practice.
Table 2 Physical Properties of Common Lipid Coating Materials
Melting point Specific gravity Refractive Iodine

Material (°C) (g/mL) index number
Cocoa butter 30-35 0.858-0.864 1.454-1.458 35-40

Acetylated monoglycerides 32-43 0.96-1.00 1.447-1.44 2 1-14
Paraffin 50 -51 0.88-0.915 1.4 10 -1.4 5 0 0
Stearyl alcohol 54-57 0.85 —
<2
Beeswax 6 1-65 0.950-0.960 1.434-1.445 8 -11
Candelilla wax 66-71 0.982-0.9856 1.456 19-44
Stearic acid 69 0.847 1.430 0
Camauba wax 82-86 0.996-0.998 1.463 7 - 14
Polyethylene wax 12 4 -12 7 0.96 — —
Source: Ref. 4.
which has an oxygen permeability similar to the high barrier waxes, delivers a high gloss but
can easily cause coated fruit to respire anaerobically [32]. Shellac coatings tend to block pores
in the citrus peel where gas exchange occurs [33]. The result is off-flavor development rather
than an extension of storage life. Due to its low moisture permeability and very high gas
permeability, oxidized polyethylene appears to be the best material for coating high respiration
rate produce [33-36].
With regard to lipids other than wax, many factors determine their ability to retard the
transport of moisture, including crystal morphology, polarity, and carbon chain length. In
general, the higher the melting point of the lipid, the greater is its barrier to moisture transport
[26]. Tempering lipid films can decrease oxygen and moisture permeabilities due to healing
of crystal imperfections and the development of a more extensively linked arrangement of
crystalline platelets [37]. Lipid crystal polymorphology and, in turn, molecular density might
affect moisture transport in lipid films as it does in polymeric films [24,38]. The a form is
the least stable polymorph and has hydrocarbon chain packing characterized by a hexagonal
subcell [39]. A more stable form is the ¡3' polymorph, followed by the /3 polymorph. The ¡3'
has a more dense hydrocarbon packing characterized by an orthorhombic subcell. Thus, the
order of thermodynamic stability is P > ¡3' > a, which is also the order of decreasing molecu
lar density [23]. Although one might expect the ¡3 form to have a lower moisture permeability
because of its denser, more restricted structure, Kester and Fennema [40] conversely found
that the moisture permeability increased following transition from a to ¡3' in a blend of fully
hydrogenated soybean and rapeseed oil. They hypothesized that the polymorph may have
had a higher hydration capacity (moisture sorption) and hence the higher permeability. The
oxygen permeability, however, was lowest for the form.
Of the non-wax lipids, acetylated monoglycerides have been used in a broader range of
applications than any other lipid material. Lovegren and Feuge [41,42] were the first to report
the oxygen and moisture barrier properties of acetylated monoglycerides as well as their po
tential use as protective coatings. Compared to other lipid materials, particularly waxes, it is
clear that acetylated monoglycerides are rather poor barriers to both moisture and permanent
gases (see Tables 3 and 4). Nevertheless, many foods have been coated with them with
positive results. One possible reason for these results may be the mechanical properties of this
material. Acetylated monoglycerides produce flexible films even at low temperature, whereas
wax films crack under similar conditions.
Table 3 Water Vapor Permeabilities of Lipid Coating Materials

Thickness WVP RH Temp.
Material (mm) [g • mm/(m^ *day • kPa)] difference C O
Lipid films
Candelilla wax [30] 0.14 0.0120 0- 100% 24.9
Paraffin [42] 0.66 0.0194 0- 100% 25
Beeswax [30] 0.14 0.0888 0- 100% 25.9
Camauba wax [25] 0.130 0.098 0- 100% 27.5
Tripalmitin [25] 0.130 0.194 0- 100% 27.5
OPE-oleic acid emulsion [36] 0.0127 0.61 0-75% 30
Anhydrous milk fat fraction [25] 0.130 0.886 0- 100% 24.9
Chocolate [135] 0.593 2.313 0-80% 26
Hydrogenated peanut oil [42] 3.39 3.33 0- 100% 25
Acetylated monoglyceride [42] 1.3 1-2 .7 3 1.9 1- 12 .8 0- 100% 25
Synthetic polymers
PVDC [25] 0.019 0.024 0- 100% 27.60
LDPE [25] 0.010 0.031 0- 100% 27.6
Polyester [25] 0.025 0.168 0- 100% 25
PVC [25] 0.012 0.617 0- 100% 27.6
Multicomponent emulsion films containing lipids
45% SA -55% HPMC [50] 0.019 0.0267 0-85% 27
BW on gluten-DATEM [192] 0.090 0.0360 0-32% 30
SA, PA, HPMC, PEG [26] 0.041 0.0480 0-85% 25
BW on SA/PA/MC/HPMC/PEG [45] 0.056 0.0572 0-85% 25
BW on MC/PEG [44] 0.051 0.0953 0- 100% 25
44% Paraffin-44% M C -12% PEG [52] 0.087 0.2646 22-84% 25
SA-PA -H PM C-PEG [43] 0.041 0.3429 0-97% 25
40% PA-60% com zein [58] 0.081 1.64 50-100% 25
37.5% BW -62.5% NaCas [54] 0.075 3.60 0-97% 25
2% A M - 68% gluten-30% Gly [193] 0.066 4.84 0- 1 1 % 23
28% BW-56% W PI-16% Sorb [56] 0.140 5.28 0-97% 23
40% BW -56.3% W PI-3.7% Gly [30] 0.140 10.82 0-98% 25
Chitosan-lauric acid [61] 0.0254-0.0380 154.28 0- 100% 25
AM = Acetylated monoglyceride; BW = beeswax; DATEM = diacetyl tartaric acid esters of monoglycerides; Gly =
glycerol; HPMC = hydroxypropyl methylcellulose; MC = methylcellulose; NaCas = sodium caseinate; OPE = oxidized
polyethylene; PA = palmitic acid; PEG = polyethylene glycol; SA = stearic acid; Sorb = sorbitol; WPI = whey protein
isolate.
V. MULTICOMPONENT FILMS CONTAINING LIPIDS

Since lipid materials generally have poor mechanical properties and are often brittle, incorpo
ration of other film-forming materials can be used to aid the integrity and performance of the
film. Many studies have examined the barrier properties of multicomponent films composed
of chemically modified cellulose and lipids [26,43-52] or protein and lipids [30,53-58].
These efforts have been focused more on improving the moisture barrier properties of the
various hydrophilic film-forming materials than on improving the mechanical integrity of the
lipid films. From the standpoint of the lipid film, unless the multicomponent film is produced
as a bilayer, added hydrophilic materials act as pinholes in the lipid substrate, thereby reduc
ing its moisture barrier properties [51,52].
Table 4 Oxygen Permeabilities of Lipid Coating Materials
Thickness O2 perm. Relative Temp.

Material (mm) [cm^ • ¡imJ{m^ •day •kPa] humidity (°C)
Lipid films
Camauba wax [194] 0.04-0.05 157.2 0% 25
Candelilla wax [194] 0.04-0.05 175.4 0% 25
Beeswax [194] 0.04-0.05 931.4 0% 25
Acetylated monoglyceride [42] 1.74 1360.9 0% 26
Microcrystalline wax [194] 0.04-0.05 1536.2 0% 25
OPE:oleic acid emulsion [36] 0.0127 8523.1 50% 30
Synthetic Polymers
EVOH (70% VOH) [10] 0.1 0% 23
EVOH (70% VOH) [10] 12.0 50% 23
Polyvinylidene chloride (Saran) [195] 3.2 0% 25
Polyester [195] 19.7 0% 25
High density polyethylene [195] 493.5 0% 25
Low density polyethylene [195] 2960.8 0% 25
Multicomponent emulsion films containing lipids
Chitosan-lauric acid [61] 0.025-0.038 6.28-9.40 0% 25
29% BW:55% WPI:16% Sorb [11] 0.110 11.6 50% 23
Delamination of the lipid layer from the hydrophilic support can be a problem because of
the high surface energy existing between these two phases. Alternatively, the lipid component
can be dispersed throughout the hydrophilic phase in the form of an emulsion. Proteins,
especially casein and whey proteins, are effective emulsifiers because of their amphiphilic
nature and therefore lend themselves to forming protein-lipid emulsions. When cellulose
ethers are used, a separate emulsifier must be incorporated to stabilize a lipid emulsion. The
stability of film-forming emulsions plays an important role in their final moisture barrier prop
erties. Unstable emulsions lead to phase separation during drying of the film. Consequently,
the orientation of the film during moisture permeability testing results in differences in the
measured values [54-56].
Lipid content in multicomponent films plays a significant role in determining the final water
vapor permeability (WVP) of the film [26,30,59]. Lipid type also has a significant effect on
the WVP of emulsion films. Film-forming solutions that contain wax generally offer the best
resistance to moisture transport [54]. However, in some cases medium-chain fatty acids (myr-
istic, palmitic, or stearic) or triglycerides perform better than waxes in composite systems
[30,47,55].
Both the barrier properties and mechanical properties of the dispersed phase contribute
significantly to the barrier properties of an emulsion film. For whey protein emulsion films
made of four separate lipids (candelilla wax, camauba wax, beeswax, and a high melting
industrial milk fat fraction) and with a wide range of lipid concentrations (0-80%), the trend
in moisture barrier properties of the emulsion films was contrary to those of the bulk lipid
material [30]. The low moisture-transmitting lipids (i.e., candelilla wax and camauba wax)
yielded emulsion films with the largest moisture permeabilities, while the higher moisture-
transmitting lipids (beeswax and the milk fat fraction) provided emulsion films that were better
barriers. These results correlated well with the degree of viscoelasticity of the bulk lipids.
The beeswax and milk fat fractions were more viscous, less elastic, and more easily de
formable lipids that either the candelilla or camauba wax [60]. As the emulsions dried, the
softer lipid particles may have come together and yielded to the internal forces of the shrink
ing, collapsing protein structure. As a result, these films may have contained an intact internal
network of lipid that provided the better moisture barrier. The significance of an internal lipid
network within a hydrophilic support with respect to the moisture barrier properties of edible
films has also been suggested by Koelsch and Labuza [47].
Chitosan, like many carbohydrate polymer films, has poor water vapor barrier properties.
The moisture permeability of chitosan-lipid films was most effectively reduced by forming
an emulsion with lauric acid. Other fatty acids and esters did not have the same effect of
significantly reducing moisture permeability [61]. These differences may have been due more
to morphological differences in the film microstructure of the various chitosan-lipid films
than to the barrier properties of the emulsified lipids.
VI. APPLICATIONS OF LIPID FILMS AND COATINGS

A. Fruit Coatings
Waxing fruit is one of the largest uses of lipid coatings. The goal of using these coatings is
not only to reduce moisture loss but also to extend storage life by reducing the fruits’ respira
tion and, in some cases, by carrying fungicides to the fruits’ surface. Additionally, the appear
ance of the fruit is improved in the eyes of the consumer by the high gloss of some lipid
coatings. The waxing of citrus fruit is so widespread that it is difficult for American consum
ers to find oranges and lemons that have not been coated. In some instances, the goal of using
these coatings may only be to impart a glossy shine to the fruit [62]. Although a broad range
of lipid-coated fruits are reviewed here, we go into depth in discussing citrus coating because
it deals with all of the issues found in waxing nearly any fruit.
Waxing of citrus fruits is not a process new to the twentieth century. As early as the
twelfth or thirteenth century, oranges and lemons were coated with a thick layer of molten
wax to ferment the fruit [1]. In the 1920s and 1930s, citrus growers began waxing fruits and
vegetables to reduce moisture loss during storage [63-65]. More recently, the emphasis in
waxing fresh produce has been less on extending shelf life and more on aesthetics. Consumers
desire produce that has a bright shine. Kaplan [66] summarizes four reasons why fruit is
waxed:
1. To reduce shrinkage due to water loss

2. To provide a barrier to free gas exchange
3. To improve appearance and marketability by application of a shiny, sweat-resistant
film
4. To provide a carrier for decay control fungicides, growth regulators, and coloring
agents
From the viewpoint of extending the product’s shelf life, waxing citrus fruits is performed,
in part, because the natural waxes are removed during washing. The fruits’ natural wax, a
mixture of fatty alcohols and fatty acids [67], picks up dust, dirt, fungal and bacterial spores,
and preharvest sprays before and during harvesting. These contaminants, along with the wax,
are removed by washing; after a brief drying step, the fruit is waxed with either a water-based
or solvent-based wax to replace the original waxy coating.
Wax coatings are applied to citrus fruit by using either a solvent wax system or, more
commonly now, a water wax emulsion. A solvent wax system consists of the coating material
dissolved in a highly volatile petroleum solvent typically composed of 70-80% aliphatic hy
drocarbons with up to 25% aromatic hydrocarbons and possibly including solvents such as
acetone or ethyl acetate [68]. A water wax system is an oil-in-water emulsion of the hydropho
bic coating material. An additional difference between these two systems is that the solvent
wax requires that the fruit be dried prior to coating but may not require drying after coating
because of the volatility of the solvent. This is not the case with the water wax emulsion,
since the fruit can be wet prior to coating but must be dried following application of the
emulsion [66]. However, total drying times using water waxes may not necessarily be longer
than with solvent waxes, since the fruit in the latter case does not need to be dried prior to
coating with the emulsion [69]. Lipid and hydrophobic materials used for both waxing sys
tems on citrus fruit are listed in Table 5. Section 21.172.210 of the U.S. CFR lists all coatings
(including those not specifically lipid-based) that may be applied to fresh citrus fruit.
Citrus fruit storage life is extended because of a reduction in moisture loss and a decrease
in respiration rate [70-72]. Lipid-coated citrus fruit displays an increase in carbon dioxide
and a decrease in oxygen compared to noncoated fruit [33,73-76]. One of the problems
associated with waxing citrus fruits is the application of too thick a coating. Providing too
effective a gas barrier will certainly reduce moisture loss, but it can cause internal oxygen
concentrations to drop below a critical level such that ethanol and off-flavors are produced
[32,35,75,77,78]. There is some evidence that lipid coatings may enhance the flavor of or
anges by causing a buildup of components important to orange flavor (acetaldehyde, ethyl
acetate, ethyl butyrate, and methyl butyrate) [76]. However, the same study indicated that
the oranges coated with partially hydrogenated vegetable oil or propylene glycol esters of
fats and fatty acids displayed a significant loss in limonene, a twofold increase in meth
anol, and a sixfold increase in ethanol. A minimal coating sufficient to provide a glossy
appearance and reduce weight loss has been suggested, with a maximum level of 0.2-0.3
mg/cm^ [35].
Table 5 U.S. Food and Drug Administration

Approvals of Principal Film Formers Used in Citrus
Fruit-Coating Formulations
Material 21 CFR citation^
Beeswax 184.1973, 184.1975

Camauba wax*’ 184.1978
Candelilla wax 184.1976
Cottonseed oil 172.894
Fatty acids 172.860
Mineral oiU 172.878
Montan wax 178.3770
Oleic acid 172.862
Paraffin"^ 172.275
Petrolatum‘s 172.880
Petroleum wax^^ 172.886, 172.888
Polyethylene, oxidized‘s 172.260
Rice bran wax 172.890
Sperm oil 172.210
PVDC 172.210
^Code of Federal Regulations, Title 21.

‘’Not permitted on citrus exported to the United Kingdom.
^Not permitted on citrus exported to Japan.
^Not permitted for use in Italy.
Waxing may partially or completely plug stomatal pores in citrus fruit, restricting the trans
port of ethylene, oxygen, and carbon dioxide while not restricting the movement of moisture
[79]. Kalmar [80] suggests that the wax coating be matched to the respiratory quotient (ratio
of carbon dioxide production to oxygen consumption) of the particular fruit or vegetable.
Those products having a low respiratory quotient can be more completely “sealed” than those
that have a high quotient. By spraying a solvent wax on citrus fruit, a glossy appearance is
achieved, but the film coating is discontinuous and porous, thereby allowing sufficient gas
exchange to prevent the fruit from respiring anaerobically. Commercial fruit waxing materials
have a wide range of gas and moisture permeabilities [75]. It is important that a producer of
waxed fruit understand the differences in these materials so that the desired effect is achieved
(gloss and extended storage life) while the undesirable effects of anaerobic respiration are pre
vented.
Compared to wax coatings, polyethylene-based coatings have high gas permeabilities and
therefore do not cause a buildup of ethanol and off-flavors in the coated citrus fruit, but they
still manage to impart a glossy appearance and reduce water loss in these products [33-36].
This makes sense, since polyethylene is a very good moisture barrier but a rather ineffective
gas barrier [36].
Lipid coatings have also been used in conjunction with fungicides to extend the storage
life even further than with lipid coatings alone. A wide range of fungicides have been tested
in wax emulsions on citrus products, including 2-aminopyridine [81], thiabendazole [82],
benomyl [83], sodium orthophenyIphenate, and hexamine [69,84,85]. Treating oranges with
either of the two hormones 2,4-D (2,4-dichlorophenoxyacetic acid) or 2,4,5-T (2,4,5-trichlor-
ophenoxyacetic acid) at levels of 1000-2000 ppm followed by waxing with a paraffin-car-
nauba wax emulsion [86] resulted in a delay of ripening and a reduction in fungal attack,
which increased the storage life by 900% over a control [87,88].
An additional issue that must be considered when selecting wax materials for coating is
the degree of gloss. Natural waxes alone are not necessarily effective at producing a high
gloss finish [66,68]. Hydrocarbon wax provides low gloss, and polyethylene wax provides
intermediate gloss, while shellac and resin esters can give a high gloss appearance on citrus
fruit [32]. The glossy finish of some lipid coatings can be affected by exposure to moisture.
When fruit has been removed from cold storage, moisture will condense on the exterior of
the fruit; this “sweating” may irreversibly dull the finish of some coatings. For example,
camauba wax and polyethylene wax coatings containing palmitic or stearic acids produced a
whitish blush upon wetting with water, while the same waxes emulsified using oleic acid did
not produce the whitish blush [74]. Further issues include requirements for buffing of the
coating to achieve a glossy appearance and the temperature of the fruit during coating. Brush
ing of coated produce can result in damage to the tissue [89]. Furthermore, problems occur
when trying to coat refrigerated fruit, because moisture condensation on the exterior of the
coating surface can stall or reverse the drying process. A patented process claims to overcome
this problem [90]. The surface of refrigerated fruit is preheated with hot (100-160T) water,
the fruit is coated and then dried, and the core temperature of the fruit does not rise above
15T.
B. Deciduous Fruit Coatings

Apple coating represents another large area of lipid-coating use. The minimization of moisture
loss in apples has been achieved using various lipid coatings. Several factors affect the effec
tiveness of the coating in extending the storage life of apples: (1) maturity of the fruit (the
greater the maturity of the fruit, the greater the gas resistance of the skin); (2) type of coating
(waxes are better than oils); (3) coating thickness; (4) temperature of storage (affecting the
apples’ rate of respiration); and (5) variety of apple [91]. The order of decreasing moisture
loss prevention was found to be paraffin wax > carnauba wax > beeswax; peanut oil, com
oil, and light mineral oils enhanced moisture loss [91]. Commercial coatings are often not a
single lipid system. A typical apple wax may contain any of the following ingredients: water
(solvent), isopropanol (cosolvent), shellac, carnauba wax, propylene glycol (plasticizer), mor
pholine or ammonium oleates, casein, soy protein (binders), and methyl siloxanes (antifoam
agents) [92].
Red Delicious apples coated with wax and stored at room temperature have displayed
reduced respiration rate, increased sugar and soluble solids, and slower development of solu
ble pectic substances over noncoated controls [93]. As with citms fruit, too thick a wax
coating on apples and pears reduces respiration to such an extent that alcoholic fiavors develop
[94]. Methyl esters of fatty acids, fatty acids, and triglyceride-type oils applied to Jonathan
apples after harvest reduced levels of hexanal in the fruit and the incidence of “soft scald”
that developed during cold storage [95]. However, the use of pure tripalmitin had little effect
on the occurrence of soft scald. This may have been due to tripalmitin’s limited solubility in
ethanol and hence an incomplete coverage with a solvent system, or it may have been due to
crack development in this very brittle substance during storage. Waxing Golden Delicious
apples increases their color and reduces their weight loss during storage, yet firmness was not
affected. Increasing the drying temperature to 60°C improved firmness and color but increased
weight loss [96]. Consumer acceptability scores were linearly related to the amount of wax
applied to apples, with good scores for appearance directly related to high level of wax appli
cation [97].
Bananas ripen quickly at high temperature and therefore have a short shelf life. Lipid
coatings have been used on bananas chiefly as a means of controlling gas transport and to a
lesser degree for preventing moisture transport. Coating bananas with sucrose esters of fatty
acids, glycerides, or paraffin-sugarcane wax emulsions had a significant effect on reducing
moisture loss and delaying the ripening process [98-100], while polyethylene wax coatings
had no significant effect on speed of ripening [100].
In addition to citrus fruit, apples, and bananas, many different types of fruit have re
sponded positively to lipid coatings. Emulsion coatings of carnauba wax, paraffin wax, poly
ethylene wax, and mixtures thereof had a significant effect on reducing moisture loss, altering
metabolic activities, extending postharvest storage life, and improving the appearance of
apples [97], apricots [101], avocados [102], bananas [103], grapes [101], peaches [101], pears
[104], pineapples [105], plums [101], mangoes [106-108], and nectarines [101]. Sucrose
fatty acid esters have been used effectively on many of the previously mentioned fruit plus
durian fruit [109], guavas [110], and melons (honeydew and cantaloupe). The addition of
fungicides to the wax coatings has significantly increased storage life by reducing fungus-
associated decay in addition to moisture loss in apples [71,111,112], bananas [113], guavas
[114,115], mangoes [116], melons [117], pears [118], and rambutan [119]. Baldwin and
coworkers [92,120] have reviewed the application of coatings to a wide range of fresh and
lightly processed fruits and vegetables with many references to commercial coatings. A simi
lar review exists on coatings for processed foods [121].
C. Vegetable Coatings
Vegetable coating, like fruit coating, is directed toward reducing moisture loss, reducing the
respiration rate (in some cases), and improving appearance. Vegetables that have low respira
tion rates are better candidates for lipid coatings than those with high respiration rates. Al
though the coating may have a significant effect on the moisture loss, the gas barrier properties
of the coating may have varying effects on the respiration rate. Coating a highly respiring
vegetable with a lipid film will tend to make it respire anaerobically, and this leads to off-
flavor development. Commonly coated vegetables are cucumbers, rutabagas, and mature
green tomatoes [92]. Bell peppers, cucumbers, eggplant, squash, and tomatoes are typically
coated with a mineral oil-based coating, while paraffin wax emulsions are frequently used to
coat root crops such as rutabaga, turnip, and yucca [122]. Other crops that are occasionally
given wax coatings include asparagus [123], avocados [75], beets, carrots, celery [123], egg
plant [75], green beans [124], kohlrabi, okra, parsnips, potatoes, radishes, summer squash,
turnips, and winter squash [123]. Waxing leafy vegetables does not extend their storage life
[125].
The application of surfactants to potato tubers is associated with a resistance to carbon
dioxide transport in the surfactant films, increased carbon dioxide levels in the tuber, and
prevention of greening during cold storage [126]. Addition of plant acids, such as tartaric,
ascorbic, or citic, reduced the carbon dioxide permeability of the surfactant layer, thereby
allowing a thinner coating to achieve the same antigreening results [127].
The whitish blush that develops on the surface of peeled carrot pieces (manufactured “baby
carrots”) is a defect noticeable by the consumer. The white appearance comes from a combi
nation of irreversible dehydration and lignification as a result of wounding [128]. An opti
mized sodium caseinate-stearic acid emulsion eliminated the formation of the white blush
[129]. The optimal formulas consisted of 91% sodium caseinate and 9% stearic acid (dry
basis). Although the emulsions contained stearic acid, the authors hypothesized that the hydro
philic protein phase contributed to moisturizing the carrot surface, thereby reducing surface
whiteness.
D. Candy and Confectionery Coatings

The largest use for lipid coatings in the candy and confectionery industry is for polishing
panned products. A panned confection is produced by repeated application and drying of a
sugar syrup or chocolate coating on candy pieces. These pieces are contained in a rotating
drum, called a pan, such that they tumble against each other, which ensures an even distribu
tion of the coating material. Application of a wax coating in the final stage of panning candies
provides numerous benefits such as preventing them from sticking to surfaces and other can
dies; improving their appearance by providing a glossy finish; preventing cracking, splitting,
or crumbling; and improving shelf life through the reduction in moisture loss or pickup
[130,131]. For the polishing stage of panning, the pan is first coated with a layer of molten
beeswax; depending on the coating thickness and the volume of product being polished, this
coating will last from 2 weeks to 6 months [132]. When hard sugar-coated candies, such as
Jordan almonds, are to be polished, they are placed in the coated polishing pan and tumbled
with added carnauba wax or beeswax [132,133]. During the polishing stage it is important to
maintain a steady flow of cold, dry air (less than 60°F and 50% RH) over the candy pieces in
the pan [134]. Wax usage is low, approximately 0.5 oz per 300 lb of candies.
Since the polishing process can take some considerable time, a faster technique involves
melting beeswax and carnauba wax and then adding them to odorless volatile mineral spirits
[130]. The polishing liquid is added to the finished pieces in the polishing pans, and the
pieces are tumbled until they are uniformly coated. The evaporated solvent is exhausted, and
the pieces are then polished.
Soft panned candies, such as jelly beans or marshmallow eggs, must be first coated with a
sugar syrup that is allowed to dry. This final sugar coating provides a thin shell over the
candy on which the polishing material can deposit [132]. It also prevents dusting of the soft
finish underneath. The candies are then polished as if they were hard candies.
Wax is not the exclusive material for polishing candies. Shellac, dissolved in ethanol, is
another common polishing agent, which is referred to as confectioners glaze. Relative to
shellac, wax polishes impart a less glossy matte finish [13 3 ]. A combination of shellac and
wax may be used to improve the glossy appearance. Other waxes sometimes used for pol
ishing are spermaceti, ouricouri, and candelilla wax [133]. Mineral oil can be used to prevent
sticking in hard gum candies made from gelatin or starch, such as Ju Ju Bees or Gummi
candies. The drawback from using mineral oil is that it is easily transferred from one surface
to another; thus the inner surface of the package containing these coated candies will end up
covered with mineral oil [130].
Chocolate coatings, although not solely lipid, are used throughout the candy and confec
tionery industry to provide a moisture barrier on nuts and to minimize the transport of mois
ture among candy components of differing water activity. The WVP of chocolate lies between
that of hydrogenated oils and those of high melting waxes and is a function of the amounts
of lipid and carbohydrates present in the mixture [135]. Incorporation of alkali derivatives of
soy protein [136], calcium stearate [137], or hydroxy lecithin [138] can reduce the moisture
transmission rate in chocolate by 30-60%. Additionally, the type of tempering and, in turn,
the nature of the fat crystals influence the moisture barrier properties of chocolate coatings
[139]. Rapidly cooled cocoa butter has a moisture permeability higher than that of the tem
pered fat [140]. Unlike the other coatings used on candies, chocolate imparts it own flavor or
additional flavors such as milk or mint.
In addition to chocolate, similar materials called compound coatings can be used. Com
pound coatings are typically white and made from refined vegetable butters that are often
based on imported lauric oils such as palmkemel and coconut oil [130]. There are a number
of drawbacks to the use of lauric-based fats. Hydrolysis into lauric acid can produce an unde
sirable “soapy” flavor, and fats containing laurin do not provide a good moisture barrier [130].
Furthermore, most fats used for compound coatings are not compatible with cocoa butter in
that they produce rapid bloom formation in the mixture. This can be alleviated by using an
anhydrous milk fat fraction [133] or modified fats [141]. Use of water-in-oil emulsions for
chocolate and other fat-containing confectionery coatings has been patented as a means of
incorporating water-soluble flavoring and coloring in the fat-based coating [142].
A coating of com zein and acetoglycerides is able to resist moisture transfer and oil migra
tion in candies [130], This coating can provide sufficient gloss that it can be used as a replace
ment for confectioner’s glaze or mineral oil coatings. The main drawback to using com zein
is that the coating dissolves slowly and unevenly during the eating of hard candies [130].
E. Nut and Raisin Coatings

The two main defects in stored nuts are oxidative rancidity and moisture uptake. Lipid coat
ings have been effective at minimizing or delaying the onset of these defects. Mixtures of
various antioxidants (TBHQ, BHA, BHT) and citric acid in hydrogenated vegetable oil or
acetylated monoglycerides have been effective at minimizing oxidative rancidity in peanuts
[143]. When the nuts are granulated for use in food items, the increased surface-to-volume
ratio decreases the time it takes for the product to become unacceptable because of oxidation
and moisture uptake. Coating granulated peanuts just after roasting has proven to delay these
deteriorative processes [144]. Coating nuts with acetylated monoglycerides helped them retain
crispness when used in ice cream and also reduced rancidity [145]. A similar study examining
the effects of coating pecans with acetylated monoglycerides also showed that the coating
reduced oxidative rancidity and extended the shelf life of the product, with or without antioxi
dants [146]. Film-forming solutions of com zein and acetoglycerides can retard moisture sorp
tion and rancidity in nuts and other items [130,147,148].
Mineral oil has traditionally been used to prevent the clumping of raisins as well as to
improve their appearance [149]. However, the major deteriorative process for raisins is exces
sive moisture loss. Raisins can be stabilized against moisture loss by coating with beeswax.
Mixtures of beeswax and cottonseed oil or other vegetable oils have increased water vapor
permeability and are not as effective as beeswax coating alone [150,151]. Precoating raisins
with polysaccharides that have a molecular weight above 5000 followed by wax coating im
proved the integrity of coating and reduced the moisture loss of the product [150]. Coating
dehydrated fruit pieces with acetylated monoglycerides protected them from moisture loss or
uptake in cake mixes [152]. Foods containing dehydrated fruits or other intermediate moisture
products have been coated with a patented film-forming dispersion of protein, polysaccha
rides, a plasticizer, and an immiscible liquid such as an edible oil [153,154].
F. Meat, Fish, and Cheese Coatings

Postmortem changes in the meat of warm-blooded animals results in considerable dehydration
of meat protein. Rapid freezing to —40°F and low temperature storage can extend the storabil-
ity of the product. However, a considerable amount of moisture is lost during frozen storage;
this is known as freezer bum. Coating frozen meat with an oil-in-water emulsion composed
of plant oils, waxes, and methyl cellulose retained moisture during frozen storage [155,156].
Ethyl cellulose and oil [157], mono- and triglycerides [158], fatty acids and cetyl alcohol in
a glycerol-water base [159], and acetylated monoglycerides [145] have been patented as coat
ing materials for the prevention of moisture loss in frozen or refrigerated meat.
Dehydration and oxidation are the primary sources of degradation in frozen fish. Ice glaz
ing can slow both of these processes [160], but the loss of an ice barrier due to sublimation
or cracking and chipping requires that the fish be periodically reglazed [161]. Whey protein
isolate and acetylated monoglycerides, alone and in combination, have been examined as a
means of reducing these types of degradation [162]. Spraying the frozen fish with a low
melting point (—13°C) acetylated monoglyceride had a significant effect on reducing moisture
loss and a noticeable, though not significant, effect in reducing peroxide values of the frozen
stored fish.
Coating meats with lipids does not come without its own set of problems. Acetylated
monoglycerides have been tested on beef and lamb cuts [163,164]. When used on meats that
were vacuum-packaged, the coated cuts purged less moisture and had lower microbial counts;
however, the authors of the study felt that the uncharacteristic aroma and oily appearance
might detract from the benefits of improved shelf life [165]. Addition of chlortetracycline to
the coating further decreased microbial loads and improved appearance during storage [166].
Yet the consumption of antibiotics with coated product remains a controversial issue. Coating
sub-scalded chicken broiler parts with acetylated monoglycerides prevented dehydration and
skin darkening in storage at 40°¥ [167]. Yet chicken parts coated with the same material had
unacceptable flavor scores when the coated meat was stored in the presence of other foods
[168] . The addition of cellulose acetate butyrates to acetylated monoglyceride coatings for
meat and fish improved the performance of the coating by making it a more mechanically
resilient package. This film must be peeled from the meat prior to cooking and consumption
[1 6 9 ] .
Cheese has historically been coated with wax for the purpose of minimizing moisture loss
and protecting the product from contamination. Single or multiple dips in paraffin wax or
mixtures containing paraffin and petrolatum have been used on cheddar and gouda. The use
of microcrystalline wax helps to improve the moisture barrier properties of the wax coating,
while the incorporation of elastomers (such as polystyrene or polyethylene) in the wax mixture
aids in peeling the coating from the wax [17]. The addition of antimicrobial agents, such as
sorbic acid, to edible hydrophilic films has proven to reduce the surface concentration of
micoorganisms on a cheese analog [170]. The permeability of composite films of cellulose
esters and various lipids to sorbic acid and potassium sórbate has been studied with a similar
intention [49,171].
G. Coating Starch-Based Products

Moisture uptake in dry cereal-based foods is the leading mode of deterioration for these prod
ucts. Above a water activity of 0.40-0.45, enough water is present to plasticize dry cereal-
based snacks to the extent that they lose their crispness [172]. Minimizing moisture uptake
through the application of lipid coatings has proven effective in a number of instances.
Applying edible coconut oils to cereal flakes resulted in a number of advantages including
improved storage stability, enhanced sweetness when sugar-coated, and improved texture and
flavor [173]. Yeast-raised doughnuts suffer such rapid moisture loss and starch rétrogradation
that their shelf life is only a matter of hours. Coating this type of doughnut with an oil-
in-water emulsion of acetylated monoglycerides, triglyceride shortening, and ricebran wax
minimized moisture loss and extended its shelf life [173]. Coating the interior surface of ice
cream cones with two layers of a lipid mixture containing high melting triglyceride shorten
ing, tristearin, and cocoa improved the moisture repellency and extended the crispness of
these products [174]. In practice the interior of cones is often sprayed with chocolate or
couverture to provide an effective moisture barrier. In addition to ice cream cones, ice cream
sandwich wafers, pie shells, and some types of candy bars have been successfully coated with
acetylated monoglycerides [145].
H. Fat Encapsulation of Moisture-Sensitive Ingredients

Moisture-sensitive food ingredients can be protected by encapsulation in a lipid material.
Coating food grade acids with hydrogenated vegetable oils prevents the acid from accelerating
food component deterioration, such as starch hydrolysis, color degradation, and flavor loss,
until the food is heated. The coating helps reduce the acid’s hygroscopicity, reduces its dust
ing, and improves its flowability by minimizing clumping [175]. Sodium acid pyrophosphate,
glucono-6-lactone, and lactic acid have been encapsulated with lipids to improve their use in
foods such as restructured pork meat [176,177]. Ascorbic acid has been protected from water
and oxygen by lipid encapsulation for use as a dough conditioner [178].
Sweeteners represent another large group of lipid-encapsulated ingredients. Similar to coat
ing acids, encapsulating sweeteners reduces their hygroscopicity, improves their flowability,
and prolongs their perceived sweetness. Encapsulated sugar can be used in chewing gum to
extend the perception of sweetness. The coated sugar requires a higher temperature and shear
ing to release its sweetness than the uncoated sugar [179]. Several techniques for encapsulat
ing aspartame (L-aspartyl-L-phenylalanine) to improve its stability against moisture and heat
in chewing gum have been patented [180-183].
In addition to acids and sweeteners, a wide range of flavorings and vitamins [184,185]
have been encapsulated in fat for protection. Nucleoside-5'-phosphate is used in flavoring a
wide range of foods and beverages, but it is susceptible to decomposition by phosphatase. In
cases where the enzyme phosphatase has not been inhibited and the flavoring agent is desired
(such as with miso), or when it is not feasible to add the flavoring agent after heat (ham.
sausage), the flavoring agent must be protected. Coating the flavoring agent with a high melt
ing fat protects it until phosphatase has been inactivated or until the food has been heated and
is ready to be consumed (instant soups) [186].
Salt in some meat products can act a prooxidant and result in rancidity development during
storage. Such is the case with pork sausage. This problem is confounded during frozen storage
by the concentration of unfrozen salt-containing liquid. The potential for rancidity can be
reduced by coating the salt with a lipid substance that itself has excellent flavor stability.
When the meat is cooked later, the fat melts and releases the salt to flavor the product. The
lipid source can be hydrogenated triglyceride oils or mono- and diglycerides [187]. It must
have a sufficiently high melting point to protect the salt during storage, but not so high a
melting temperature that some of it remains crystalline even during cooking. Approximately
20-60% by weight of fat is used in practice. A similar objective was achieved for use with
salt or sugar using a multicomponent emulsifier mixture consisting of polyoxyethylene (20)
sorbitan monostearate (Polysorbate 60), polyoxyethylene (20) sorbitan monooleate (Polysor-
bate 80), and a blend of mono- and diglycerides of fat-forming fatty acids [188].
I. Coating Fat with Fat

In some cases it is desirable to coat bars of cooking fat so that they may be stored in contact
with each other without sticking inseparably. The patent describing this procedure is suffi
ciently vague to state that any suitable fat material may be used provided it has been “hard
ened” to a melting point greater than 43°C. Only 0.25-2 mm thickness is required, and the
patent suggests coextruding the two fats using an attachment on the end of a Votator scraped
surface heat exchanger [189].
VII. CONCLUSION
The development and use of edible coatings is motivated by quality, aesthetic, and economic
factors [3]. Minimizing moisture loss, enhancing appearance, and reducing the overall cost
and complexity of packaging systems are goals that can be achieved by using edible lipid
coatings. Lipid coatings have proven themselves as moisture barriers for encapsulating mois
ture-sensitive ingredients and as a means of extending the postharvest life of fresh produce.
The use of edible lipid coatings on and in foods has, without question, improved the salability
of many products by extending their shelf life and, in some cases, improving their appearance.
Yet consumers remain somewhat unaware of the extent to which these materials are used.
The FDA’s ruling in January 1993 on postharvest coatings used for fresh produce requires
shippers to list the origin of the coating, rather than its subingredients, on the shipping con
tainer and on prominently and conspicuously displayed signs at retail outlets [190]. These
regulations became effective on May 8, 1994. The consequences of these regulations are far-
reaching, since most consumers have been unaware of the materials used for coating fresh
produce. Should consumers come to believe that their food is coated with unsafe “chemicals,”
any benefit these coatings may have provided will be of little consequence because the coated
product will not be purchased. Currently there are certain countries that will not accept fruit
coated with specific waxes. Norway will not import fruit coated with any wax [191]. In this
light, the future for coated fruit is somewhat uncertain.
The future of edible lipids will be affected by their use in multicomponent edible coatings
and films. This area has not yet been fully explored. Considerable research is being focused
on improving the moisture barrier properties of polysaccharide and protein edible films by
incorporating lipid materials. With the proper materials and film-forming techniques, an edible
film will be produced that can control gas, aroma, and lipid transport but also remain an
effective moisture barrier.
REFERENCES
1. Anonymous, Invisible package—long life to fruits and vegetables. Mod. Pack. 18: 75-80 (1944).'
2. T. P. Labuza and R. Contreras-Medellin, Prediction of moisture protection requirements from
foods. Cereal Foods World 26: 335 (1981).
3. S. Guilbert, Technology and application of edible protective films, in Food Packaging and Preser
vation: Theory and Practice (M. Mathlouthi, Ed.), Elsevier Applied Science, London, 1986.
4. H. Bennett, Industrial Waxes, Chemical Publishing, New York, 1975.
5. C. E. Rogers, J. A. Meyer, V. Stannett, and M. Szwarc, Some factors affecting the permeability
constant. Part II, in Permeability of Plastic Films and Coated Paper to Gases and Vapors (TAPPI
Monograph Ser. No. 23) (V. Stannett, M. Szwarc, R. L. Bhargava, J. A. Meyer, A. W. Myers,
and C. E. Rogers, Eds.), Technical Association of the Pulp and Paper Industry, New York, 1962.
6. C. E. Rogers, Permeation of gases and vapours in polymers, in Polymer Permeability (J. Comyn,
Ed.), Elsevier Applied Science, London, 1985, Chap. 2.
7. W. J. Moore, Physical Chemistry, Prentice-Hall, Englewood Cliffs, NJ, 1972.
8. J. Crank, The Mathematics of Diffusion, Oxford Univ. Press, New York, 1975.
9. C. E. Rogers, Solubility and diffusivity, in Physics and Chemistry of the Organic Solid State,
Vol. II (D. Fox, M. M. Labes, and A. Weissberger, Eds.), Interscience, New York, 1965. pp.
510-635.
10. M. Salame, Barrier polymers, in The Wiley Encyclopedia of Packaging Technology (M. Bakker,
Ed.), Wiley, New York, 1986. pp. 48-54.
11. T. H. McHugh and J. M. Krochta, Sorbitol- versus glycerol-plasticized whey protein edible films:
integrated oxygen permeability and tensile property evaluation, J. Agric. Food Chem. 42: 841-
845 (1994).
12. V. T. Stannett, The permeability of plastic films and coated papers to gases and vapors. Convert.
P ack. 22-26 (1985).
13. R. B. Bird, W. E. Stewart, and E. N. Lightfoot, Transport Phenomena, Wiley,New York, 1960.
14. Code of Federal Regulations, Title 21, Part 172, 1977 ed., amended 1989.
15. Code of Federal Regulations, Title 21, Part 184, 1977 ed., amended 1983.
16. A. Streitwieser and C. H. Heathcock, Introduction to Organic Chemistry, Macmillan, New
York, 1985.
17. A. H. Warth, The Chemistry and Technology of Waxes, Reinhold, New York, 1956.
18. Code of Federal Regulations, Title 21, Part 175, 1977 ed.
19. J. W. Martin, Shellac, in Treatise on Coatings, Vol. 1, Part 3, Film-Forming Compositions
(R. R. Myers and J. S. Long, Eds.), Marcel Dekker, New York, 1972. pp. 442-478.
20. P. K. Bose, Y. Sankaranarayanan, and S. C. S. Gupta, Chemistry of Lac, Indian Lac Research
Institute, Namkum, India, 1963.
21. M. Weiss and S. Beinfest, Systematic analysis of shellac. Soap and Chemical Specialities 38:
135, 137, 139, 14 1, 149 (1962).
22. H. Bennett, Industrial Waxes, Chemical Publishing, New York, 1975.
23. A. E. Bailey, Melting and Solidification of Fats, Interscience, New York, 1950.
24. J. A. Barrie, Water in polymers, in Diffusion in Polymers (J. Crank and G. S. Park, Eds.),
Academic, New York, 1968. pp. 259-314.
25. T. H. Shellhammer and J. M. Krochta, Water vapor barrier and rheological properties of simulated
and industrial milkfat fractions, Trans. ASAE (In press) (1997).
26. S. L. Kamper and O. Fennema, Water vapor permeability of edible bilayer films, J. Food Sci.
49: 14 7 8 -14 8 1, 1485 (1984).
27. J. J. Kester and O. Fennema, Resistance of lipid films to water vapor transmission, J. Am. Oil
C hem . Soc. 66: 113 9 - 114 6 (1989).
28. O. Fennema and J. J. Kester, Resistance of lipid films to transmission of water vapor and oxygen,
in Water Relationships in Food (H. Levine and L. Slade, Eds.), Plenum, New York, 1991. pp.
703-719.
29. I. G. Donhowe, Water Vapor and Oxygen Permeability of Wax Films, The American Chemical
Society Edible Film Symposium, St. Louis, MO. August, 1992.
30. T. H. Shellhammer and J. M. Krochta, The performance of whey protein emulsion films as af
fected by the viscoelasticity and water vapor permeability of the dispersed phase, J. Food E ng.:
(1996).
31. J. J. Kester and O. Fennema, Resistance of lipid films to oxygen transmission, J. Am. Oil Chem.
Soc. 66: 112 9 - 113 8 (1989).
32. R. D. Hagenmeier and R. A. Baker, Layered coatings to control weight loss and preserve gloss
of citrus, HortScience 30: 296-298 (1995).
33. R. D. Hagenmeier and R. A. Baker, Reduction in gas exchange of citrus fruit by wax coatings,
J. Agrie. Food Chem. 41: 283-287 (1993).
34. S. Ben-Yehoshua, Some physiological effects of various skin coatings on range fruit, Israel J.
A grie. R es. 17: 17 -2 7 (1967).
35. P. L. Davis and R. C. Hofman, Effects of coatings on weight loss and ethanol buildup in juice of
oranges, J. Agrie. Food Chem. 21: 455-458 (1973).
36. R. D. Hagenmeier and P. E. Shaw, Permeability of coatings made with emulsified polyethylene
wax, J. Agrie. Food Chem. 39: 1705-1708 (1991).
37. J. J. Kester and O. Fennema, Tempering influence on oxygen and water vapor transmission
through a stearyl alcohol film, J. Am. Oil Chem. Soe. 66: 1154-1157 (1989).
38. B. Pascal, Study of some factors affecting permeability, in Food Paekaging and Preservation (M.
Mathlouthi, Ed.), Elsevier Applied Science, London, 1986. pp. 7-24.
39. K. Larsson and P. Dejmek, Crystal and liquid structures of lipids, in Food Emulsions (K. Larsson
and S. E. Friberg, Eds.), Marcel Dekker, New York, 1990. pp. 97-125.
40. J. J. Kester and O. Fennema, The influence of polymorphic form on oxygen and water vapor
transmission through lipid films, J. Am. Oil Chem. Soe. 66: 1147-1153 (1989).
41. N. V. Lovegren and R. O. Feuge, Permeability of acetostearin products to carbon dioxide, oxy
gen, and nitrogen, J. Agrie. Food Chem. 4: 634-638 (1956).
42. N. V. Lovegren and R. O. Feuge, Permeability of acetostearin products to water vapor, J. Agrie.
Food Chem. 2: 558-563 (1954).
43. S. L. Kamper and O. Fennema, Water vapor permeability of an edible, fatty acid, bilayer film, J.
F o o d S e i. 49: 1482-1485 (1984).
44. I. K. Greener and O. Fennema, Barrier properties and surface characteristics of edible, bilayer
films, J. F o o d S e i. 54: 139 3-139 9 (1989).
45. J. J. Kester and O. Fennema, An edible film of lipids and cellulose ethers: barrier properties to
moisture vapor transmission and structural evaluation, J. Food Sei. 54: 1383-1392 (1989).
46. V. Sapru and T. P. Labuza, Dispersed phase concentration effect on water vapor permeability in
composite methyl cellulose-stearic acid edible films, J. Food Proeess. Preserv. 18: 359-368
(1994).
47. C. M. Koelsch and T. P. Labuza, Functional, physical and morphological properties of methyl
cellulose and fatty acid-based edible films, Lebensm. Wiss. Teehnol. 25: 404-411 (1992).
48. D. C. Rico-Peña and J. A. Torres, Oxygen transmission rate of an edible methylcellulose-palmitic
acid film, J. Food Proeess. Preserv. 13: 125-133 (1990).
49. D. C. Rico-Peña and J. A. Torres, Sorbic acid and potassium sórbate permeability of an edible
methyl cellulose-palmitic acid film: water activity and pH effects, J. Food Sei. 56: 497-499
(1991).
50. R. D. Hagenmeier and P. E. Shaw, Moisture permeability of edible films made with fatty acids
and (hydroxypropyl) methylcellulose, J. Agrie. Food Chem. 38: 1799-1803 (1990).
51. M. Martin-Polo, C. Mauguin, and A. Voilley, Hydrophobic films and their efficiency against
moisture transfer. 1. Influence of the film preparation technique, J. Agrie. Food Chem. 40: 407-
412 (1992).
474 Sheilhammer and Krochta
52. F. Debeaufort, M. Martin-Polo, and A. Voilley, Polarity homogeneity and structure affect water
vapor permeability of model edible films, J. Food Sci. 58: 426 (1993).
53. J. M. Krochta, A. E. Pavlath, and N. Goodman, Edible films from casein-lipid emulsions for
lightly-processed fruit and vegetables, in Preservation Processes and Related Techniques (W.
Speiss and H. Schubert, Eds.), Elsevier Applied Science, London, 1990. pp. 329-340.
54. R. J. Avena-Bustillos and J. M. Krochta, Water vapor permeability of caseinate-based edible films
as affected by pH, calcium cross-linking and lipid content, J. Food Sci. 58: 904-907 (1993).
55. T. H. McHugh and J. M. Krochta, Water vapor permeability properties of edible whey protein-
lipid emulsion films, J. Am. Oil Chem. Soc. 71: 307-312 (1994).
56. T. H. McHugh and J. M. Krochta, Dispersed phase particle size effects on water vapor permeabil
ity of edible emulsion films, J. Food Process. Preserv. 18: 173-188 (1994).
57. N. Gontard, C. Duchez, J. L. Cuq, and S. Guilbert, Edible composite films of wheat gluten and
lipids: water vapour permeability and other physical properties, Int. J. Food Sci. Technol. 29: 39-
50 (1994).
58. J. W. Park, R. F. Testin, H. J. Park, P. J. Vergano, and C. L. Weller, Fatty acid concentration
effect on tensile strength, elongation, and water vapor permeability of laminated edible films, J.
Food Sci. 59: 916-919 (1994).
59. C. M. Koelsch, The functional, physical and morphological properties of edible fatty acid and
methyl cellulose based barriers, Ph.D. Thesis, Univ. Minnesota, 1992.
60. T. H. Sheilhammer, T. R. Rumsey, and J. M. Krochta, The viscoelastic properties of lipid film
forming materials, J. Food Eng. (in press).
61. D. W. S. Wong, F. A. Gastineau, K. S. Gregorski, S. J. Tillin, and A. E. Pavlath, Chitosan-
lipid films: microstructure and surface energy, J. Agric. Food Chem. 40: 540-544 (1992).
62. C. Lecos, How to shine an apple, FDA Conserv. 16: 8-11 (1982).
63. F. Ahulberg, Preserving oranges and other fruit by coating with scale wax, U.S. Patent
1,618,159 (1927).
64. E. M. Brogden, Preserving fresh fruit, U.S. Patent 1,585,370 (1926).
65. E. M. Brogden, Preserving fresh fruit, U.S. Patent 1,585,371 (1926).
66. H. J. Kaplan, Washing, waxing, and color-adding, in Fresh Citrus Fruits (W. F. Wardowski, S.
Nagy, and W. Grierson, Eds.), AVI, Westport, CT, 1986. pp. 379-398.
67. T. Robinson, The Organic Constituents of Higher Plants, Cordus, Amherst, MA, 1975.
68 . D. J. Hall, Innovations in citrus waxing— an overview, Proc. Fla. State Hort. Soc. 94: 258-
263 (1981).
69. W. F. Newhall and W. Grierson, A low-cost, self-polishing, fungicidal water wax for citrus fruits,
Proc. Am. Soc. Hort. Sci. 66: 146-154 (1955).
70. W. A. Farooqi, M. S. Ahmad, and Zain-ul-Adbin, Effect of wax coatings on the physiological
and bio-chemical aspects of Kinow fruit, Pak. J. Sci. Ind. Res. 31: 142-145 (1988).
71. M. Meheriuk and S. W. Porrit, Effects of waxing on respiration, ethylene production, and other
physical and chemical changes in selected apple cultivars. Can. J. Plant. Sci. 52: 257-259
(1972).
72. O. Paredes-Lopez, E. Camargo-Rubio, and Y. Gallardo-Navarro, Use of coatings of candelilla
wax for the preservation of limes, J. Sci. Food Agric. 25: 1207-1210 (1974).
73. I. L. Eaks and W. A. Ludi, Effects of temperature, washing, and waxing on the composition of
the internal atmosphere of orange fruits, Am. Soc. Hort. Sci. 76: 220-228 (1960).
74. R. D. Hagenmaier and R. A. Baker, Wax microemulsions and emulsions as citrus coatings, J.
Agric. Food Chem. 42: 899-902 (1994).
75. R. D. Hagenmeier and P. E. Shaw, Gas permeability of fruit coating waxes, J. Am. Soc. Hort.
Sci. 117: 105-109 (1992).
76. M. Nisperos-Carriedo, P. E. Shaw, and E. A. Baldwin, Changes in volatile flavor components of
pineapple orange juice as influenced by the application of lipid and composite films, J. Agric.
Food Chem. 38: 1382-1387 (1990).
77. P. L. Davis and J. J. Smoot, Effect of polyethylene and wax coatings, with and without fungi
cides, on rind breakdown and decay of citrus. Citrus Ind. 41: 6 -7 ,9 (1960).
78. S. M. Norman, The role of volatiles in storage of citrus fruits, Proc. Int. Soc. Citriculture 1:
238-242 (1977).
79. S. Ben-Yehoshua, S. P. Burg, and R. Young, Resistance of citrus fruit to mass transport of
water vapor and other gases. Plant Physiol. 79: 1048-1053 (1985).
80. A. F. Kalmar, Wax coating preserves citrus for the market in field-fresh condition. Citrus Grower
316: 7-11 (1960).
81. J. R. Winston, 2-Aminopyridine, a promising inhibitor of decay in oranges. Citrus Ind. 29: 5 -
7,16-17 (1948).
82. P. T. Pesler, Decay control in Washington Navel and Valencia oranges by application of TBZ of
benomyl suspended in water or incorporated into waxes. Citrus. Subtrop. Fruit J. 463: 12-
13 (1972).
83. D. J. Hall, Comparative fungicidal activity of benomyl and its breakdown product 2-benzimidaz
ole carbamate (MBC) on citrus, Proc. Fla. State Hort. Soc. 94: 341-344 (1980).
84. E. F. Hopkins and K. W. Loucks, Prevention of the phytotoxic action of sodium orthophenylphe-
nate on citrus fruits by hexamine. Science 112: 720-721 (1950).
85. V. B. Dalai, S. D ’Souza, H. Subramanyam, and H. C. Srivastava, Wax emulsions for extending
the storage life of Sathgudi oranges. Food Sci. 11: 232-235 (1962).
86. S. A. Trout, The preparation of emulsions for coating fruits and vegetables., J. Council Sci. Ind.
Res. 15: 26-32 (1942).
87. W. S. Stewart, Effect of 2 ,4-dichlorophenoxy acetic acid and 2,4,5-trichlorophenoxy acetic acid
on citrus fruit in storage, Proc. Am. Soc. Hort. Sci. 54: (1949).
88. S. B. Lodh, S. De, S. K. Mukherjee, and A. N. Bose, Storage of mandarin oranges, II. Effects
of hormones and wax coatings, J. Food Sci. 28: 519-524 (1963).
89. K. W. Loucks and E. F. Hopkins, Commercial packinghouse operations in relation to decay in
citrus fruits. Citrus Ind. 30: 5 -7 , 9 (1949).
90. G. U. Glascow and A. J. Wright, U.S. Patent 3,488,200, cited in Gloss coating for fruits. Edible
Coatings and Soluble Packaging (R. Daniel), Noyes Data Corporation, Park Ridge, NJ, 1973,
pp. 85-92.
91. S. A. Trout, E. G. Hall, and S. M. Sykes, Effects of skin coatings on the behavior of apples in
storage. I. Physiological and general investigations, Aust. J. Agric. Res. 4: 57-81 (1952).
92. E. A. Baldwin, Edible coatings for fresh fruits and vegetables: past, present and future, in Edible
Coatings and Films to Improve Food Quality (J. M. Krochta, E. A. Baldwin, and M. O.
Nisperos-Carriedo, Eds.), Technomic, Lancaster, PA, 1994. pp. 25-64.
93. J. D. Tewari, N. Pandey, R. M. Rai, and C. B. Ram, Effect of wax emulsion on physiological
factors influencing the storage behaviour of Red Delicious apples, Ind. J. Plant Phys. 23: 257-
265 (1980).
94. R. M. Smock, Certain effects of wax treatments on various varieties of apples and pears, Proc.
Am. Soc. Hort. Sci. 33: 284-289 (1936).
95. R. B. H. Wills, K. J. Scott, and G. Hopkirk, Reduction of soft scald in apples with methyl
linoleate, HortSci. 12: 72-73 (1977).
96. S. R. Drake and J. W. Nelson, Storage quality of waxed and nonwaxed ‘Delicious’ and ‘Golden
Delicious’ apples, J. Food Qual. 13: 331-341 (1990).
97. J. K. Fellman, S. R. Drake, and J. M. Mattheis, Shellac and camauba waxing of Delicious
apples held in refrigerated storage studied. Good Fruit Grower 43: 16-18,20-21 (1992).
98. S. Krishnamurthy and C. G. Kushalappa, Studies on the shelf life and quality of Robusta bananas
as affected by post-harvest treatments, J. Hort. Sci. 60: 549-556 (1985).
99. S. Goburdhun, Chemical ripening of Dwarf Cavendish bananas (cv. Naine). I. Effects of ethrel
and ethylene on ripening. II. Extension of shelf-life of ripened fruits.. Rev. Agric. Sucr.re-de-
rile-Maurice 73: 36-43 (1994).
100. V. B. Dalai, P. Thomas, N. Nagaraja, G. R. Shah, and B. L. Amla, Effect of wax coating on
bananas of varying maturity, Ind. Food Pack. 24: 36-39 (1970).
101. L. L. Claypool, The waxing of deciduous fruit, Proc. Am. Soc. Hort. Sci. 37: 443-447
(1939).
102. Anonymous, Facelift for avocados, Food Ind. So. Africa, March, 1975, pp. 31, 33, 47.
103. S. Muthuswamy, R. Sadasivam, J. S. Sundararaj, and V. Vasudevan, Storage studies on Dwarf
Cavendish banana, Ind. J. Agric. Sci. 41: 479-484 (1971).
104. G. Suemnue and L. Bayindirli, Effects of Semperfresh® and Johnfresh® fruit coating on post
storage quality of Ankara pears, J. Food Process. Preserv. 18: 189-199 (1994).
105. R. E. Pauli and K. G. Rohrbach, Juice characteristics and internal atmosphere of waxed ‘Smooth
Cayenne’ pineapple fruit, J. Am. Soc. Hort. Sci. 107: 448-452 (1982).
106. R. B. Patel and N. P. Goswami, Post harvest diseases of mango fruits {Mangifera indica) and
their control, Ind. J. Micro. 24: 140-141 (1984).
107. M. Ullah, M. A. Haq, and M. S. Bhatti, A new technique to boost export of grafted mango fruit
by expanding its marketing season from June to November by the use of water soluble wax and
bioregulants at room temperature, J. Agric. Res. Pakistan 15: 311-314 (1977).
108. L. Gomez-Brito and M. Peleg, The effect of coating with TAG on postharvest ripening of man
goes, Arch. Latin Nut 24: 513-525 (1974).
109. S. Sriyook, S. Siriatiwat, and J. Siriphanich, Durina fruit dehiscence—water status and ethylene,
HortScience 29: 1195-1198 (1994).
110. R. G. McGuire and G. J. Hallman, Coating guavas with cellulose- or camauba-based emulsions
interferes with postharvest ripening, HortScience 30: 294-295 (1995).
111. R. B. Patel, Post harvest diseases of apple (Pyrus malus) fruits and their control, Ind. J. Micro.
24: 14 2 -14 3 (1984).
112. G. Bombiex and F. Morgat, Anti-scald fungicidal waxes and storage of apples. Fruits 32: 189-
196 (1977).
113. C. A. Shillingford, Effect of plastic coating and polyethylene wraps on banana fruit rot and
quality. Tropic. Agric. 55: 173-179 (1978).
114. R. V. Singh, M. C. Joshi, H. B. Ram, and N. S. Bisht, Effect of wax coating and prepackaging
on the storage behaviour of guava cv. Allahabad Safeda, Ind. Food Pack. 38: 80-85 (1984).
115. R. N. Agnihotri and H. B. Ram, Storage behaviour of guava variety Allahabad Safeda (Psidium
guajava L.) treated with fungicidal wax-emulsion. Food Ind. J. 5: 13-16 (1972).
116. H. B. Ram and W. B. Date, Studies on the storage behaviour of two late varieties of mangoes
treated with fungicidal wax emulsion at room temperature, Ind. J. Hort. 28: 206-209 (1971).
117. J. P. Singh and D. K. Bhatnagar, Controlling post-harvest decay of muskmelon fruits by coating
wax-emulsion and fungicides, Haryana Agric. Univ. J. Res. 12: 295-297 (1982).
118. M. Meheriuk and O. L. Lau, Effect of two polymeric coatings on fruit quality of ‘Bartlett’ and
‘d’Anjou’ pears, J. Am. Soc. Hort. Sci. 113: 222-226 (1988).
119. D. B. Mendoza, E. B. Pantastico, and F. B. Javier, Storage and handling of rambutan (Nephe-
lium lappaceum L.), Phil. Agric. 55: 322-332 (1972).
120. E. A. Baldwin, M. O. Nisperos-Carriedo, and R. A. Baker, Edible coatings for lightly processed
fruits and vegetables, HortScience 30: 35-38 (1995).
121. R. A. Baker, E. A. Baldwin, and M. O. Nisperos-Carriedo, Edible coatings and films for pro
cessed foods, in Edible Coatings and Films to Improve Food Quality (J. M. Krochta, E. A.
Baldwin, and M. O. Nisperos-Carriedo, Eds.), Technomic, Lancaster, PA, 1994. pp. 65-88.
122. L. A. Grant and J. Bums, Application of coatings, in Edible Coatings and Films to Improve
Food Quality (J. M. Krochta, E. A. Baldwin, and M. O. Nisperos-Carriedo, Eds.), Technomic,
Lancaster, PA, 1994. pp. 189-200.
123. R. E. Hardenburg, Wax and Related Coatings for Horticultural Products—A Bibliography, U.S.
Dept, of Agric., Agric. Res. Service, Washington, DC, 1967.
124. R. P. Singh, F. H. Buelow, and D. B. Lund, Storage behaviour of artificially waxed green snap
beans, J. Food Sci. 38: 542-543 (1973).
125. H. Platenius, Wax Emulsions for Vegetables, Cornell Agricultural Experiment Station Bulletin
No. 723, 43 pgs. (1939).
126. P. A. Poapst and F. R. Forsyth, The role of internally produced carbon dioxide in the prevention
of greening in potato tubers, Acta Hort. 38: 900-903 (1974).
127. P. A. Poapst, I. Price, and F. R. Forsyth, Prevention of post storage greening in table stock
potato tubers by application of surfactants and adjuvants, J. Food Sci. 43: 900-903 (1978).
128. L. Cisneros-Zevallos, M. E. Saltveit, and J. M. Krochta, Mechanism of surface white discolor

ation of peeled (minimally processed) carrots during storage, J. Food Sci. 60: 320-323, 333
(1995).
129. R. J. Avena-Bustillos, J. M. Krochta, and M. E. Saltveit, Optimization of edible coating formu
lations on zucchini to reduce water loss, J. Food Eng. 21: 197-214 (1994).
130. J. J. Alikonas, Candy Technology, AVI, Westport, CT, 1979.
13 1. D. H. Wing, Enrobing of bakery products, Proc. Am. Soc. Bakery Eng. 51: 136-141 (1975).
132. C. D. Barnett, Candy Making—As a Science and Art, Gussow, New York, 1960.
133. B. W. Minifie, Chocolate, Cocoa, and Confectionery Science and Technology, AVI, Westport,
CT, 1980.
134. M. J. Lynch, Panning: an overview, Manuf. Confect. 72: 59-64 (1992).
135. B. Biquet and T. P. Labuza, Evaluation of the moisture permeability characteristics of chocolate
films as an edible moisture barrier, J. Food Sci. 35: 989-997 (1988).
136. F. P. Colten and M. Kaplan, U.S. Patent 3,784,713, cited in Chocolate-coated confections with
improved moisture barrier. Candies and Other Confections (M. T. Gillies), Noyes Data Corp,
Park Ridge, NJ, 1974, pp. 2-4.
137. J. L. McReynolds and H. H. Topalain, Chocolate-coated confections with improved moisture
barrier, U.S. Patent 3,784,714, cited in Candies and Other Confections (M. T. Gillies), Noyes
Data Corp, Park Ridge, NJ, 1974.
138. F. P. Colten and M. Kaplan, U.S. Patent 3,895,104, cited in Chocolate-coated confections with
improved moisture barrier. Candies and Other Confections (M. T. Gillies), Noyes Data Corp,
Park Ridge, NJ, 1979, pp. 4.
139. N. W. Kempf, The conging process—tests on the moisture barrier properties of coatings, Manu-
fac. Confect. 38-40 (1967).
140. W. Landman, N. V. Lovegren, and R. Feuge, Permeability of some fat products to moisture, J.
Am. Oil Chem. Soc. 37: 1- 4 (1960).
14 1. S. A. Matz, Cookie and Cracker Technology, AVI, Westport, CT, 1992.
142. H. A. Pinkalla, G. L. Neuser, L. R. Cook, and R. F. Korfhage, U.S. Patent 3,223,532, cited in
Coating emulsions for confectionery, Edible Coatings and Soluble Packaging (R. Daniel), Noyes
Data Corporation, Park Ridge, NJ, 1973. 140-145.
143. R. Shea, New edible protective coating keeps nut fresh. Food Eng. 26: 148 (1965).
144. M. W. Hoover and P. J. Nathan, Influence of tertiary butylhydroquinone and certain other surface
coatings on the formation of carbonyl compounds in granulated roasted peanuts, J. Food Sci. 47:
246-248 (1981).
145. G. T. Luce, Acetylated monoglycerides as coatings for selected foods. Food Technol. 21: 1462-
1468 (1967).
146. S. D. Senter and W. R. Forbus, Effects of acetylated monoglyceride coatings on pecan kernel
shelf-life, J. Food Sci. 44: 175 2 -17 5 5 (1979).
147. H. B. Cosier, Stroud Jordan medal winner extends nut candy shelf life with a coating. Candy
Ind. Confect. J. 3: 17,18,22,35,36 (1958).
148. H. B. Cosier, Prevention of staleness, rancidity in nut meats and peanuts. Peanut J. Nut World
37: 10 ,11 ,15 (1958).
149. S. P. Kochar and J. B. Rossell, A vegetable oiling agent for dried fruits, J. Food Technol. 17:
661 (1982).
150. G. G. Watters and J. E. Brekke, U.S. Patent 2,909,435, cited in Raisins for cereal products.
Edible Coatings and Soluble Packaging (R. Daniel), Noyes Data Corporation, Park Ridge, NJ,
1973, pp. 109-112.
15 1. G. G. Watters and J. E. Brekke, Stabilized raisins for dry cereal products. Pood Technol. 15:
236-238 (1961).
152. R. A. Shea, U.S. Patent 3,516,836, cited in Fat coated fruits for cake mixes. Edible Coatings
and Soluble Packaging (R. Daniel), Noyes Data Corporation, Park Ridge, NJ, 1973, pp. 127-
133.
153. J. R. Durst, U.S. Patent 3,323,922, cited in Film-forming dispersions. Edible Coatings and
Soluble Packaging (R. Daniel), Noyes Data Corporation, Park Ridge, NJ, 1973, pp. 250-255.
154. J. R. Durst, U.S. Patent 3,434,843, cited in Coating of food bars. Edible Coatings and Soluble
Packaging (R. Daniel), Noyes Data Corporation, Park Ridge, NJ, 1973, pp. 255-259.
155. C. D. Bauer, G. L. Neuser, and H. A. Pinkalla, U.S. Patent 3,406,081, cited in Simple applica
tion and solidification of water-in-oil emulsion. Edible Coatings and Soluble Packaging (R. Dan
iel), Noyes Data Corporation, Park Ridge, NJ, 1973, pp. 196-200.
156. C. D. Bauer, G. L. Neuser, and H. A. Pinkalla, U.S. Patent 3,483,004, cited in Simple applica
tion and solidification of water-in-oil emulsion, Edible Coatings and Soluble Packaging (R. Dan
iel), Noyes Data Corporation, Park Ridge, NJ, 1973, pp. 200-204.
157. L. I. Berkowitz, U.S. Patent 2,957,772, cited in Simple application and solidification of water-
in-oil emulsion containing ethycellulose. Edible Coatings and Soluble Packaging (R. Daniel),
Noyes Data Corporation, Park Ridge, NJ, 1973, pp. 212-214.
158. J. Scheide, U.S. Patent 3,667,970, cited in Meat coating containing mono and triglycerides.
Edible Coatings and Soluble Packaging (R. Daniel), Noyes Data Corporation, Park Ridge, NJ,
1973, pp. 208-212.
159. T. R. Anderson, U.S. Patent 2,989,402, cited in Meat coating containing fatty acids and cetyl
alcohol. Edible Coatings and Soluble Packaging (R. Daniel), Noyes Data Corporation, Park
Ridge, NJ, 1973, pp. 204-206.
160. A. Khayat and D. Schwall, Lipid oxidation in seafood. Food Technol. 37: 130-140 (1983).
161. F. W. Wheaton and T. B. Lawson, Processing Aquatic Food Products, Wiley, New York, 1985.
162. Y. M. Stuchell and J. M. Krochta, Edible coatings on frozen king salmon: effect of whey protein
isolate and acetylated monoglycerides on moisture loss and lipid oxidation, J. Food Sci. 60: 2 8 -
31 (1995).
163. M. Stemmier and H. Stemmier, Composition for the preparation of coatings on meats and sau
sage goods, U.S. Patent 3,936,312 (1976).
164. M. Stemmier and H. Stemmier, Method for preserving freshly slaughtered meat, U.S. Patent
3,851,077 (1974).
165. J. T. Keeton, R. Leu, C. Vanderzant, J. J. Bohac, D. B. Giffin, J. W. Saveli, and H. R. Cross,
Evaluation of fresh vacuum-packaged beef steaks coated with an acetylated monoglyceride, J.
Food Sci. 53: 701-704 (1988).
166. J. C. Ayres, Use of coating materials of film impregnated with chlortetracycline to enhance color
and storage life of fresh beef. Food Technol. 13: 512-515 (1959).
167. C. W. Woodmansee and O. J. Abbott, Coating sub-scalded broiler parts in order to afford protec
tion against dehydration and skin darkening in fresh storage. Poultry Sci. 37: 1367-1373 (1958).
168. M. E. Zabik and L. E. Dawson, The acceptability of cooked poultry protected by an edible
acetylated monoglyceride coating during fresh and frozen storage. Food Technol. 23: 87-91
(1963).
169. W. L. Clark and R. J. Shirk, A hot-melt transparent peelable coating for food. Food Technol.
19: 15 6 1-15 6 7 (1965).
170. A. J. Torres and M. Karel, Microbial stabilization of intermediate moisture food surfaces. III.
Effects of surface preservative concentration and surface pH control on microbial stability of an
intermediate moisture cheese analog, J. Food Process. Preserv. 9: 107-119 (1985).
17 1. F. Vojdani and J. A. Torres, Potassium sórbate permeability of methylcellulose and hydroxypro-
pyl methylcellulose multi-layer films, J. Food. Process. Preserv. 13: 417-430 (1989).
172. E. E. Katz and T. P. Labuza, Effect of water activity on the sensory crispness and mechanical
deformation of snack food products, J. Food Sci. 46: 403 (1981).
173. R. K. Scharschmidt and L. Murphy, Flake cereal coated with edible oil, U.S. Patent 4,211,800,
cited in Edible Oils and Fats, Development Since 1978 (S. Torrey), Noyes Data Corp, Park
Ridge, NJ, 1983, 364-366.
174. S. Werbin, I. H. Rubenstein, and D. Weinstein, U.S. Patent 3,526,515, cited in Fat coatings for
ice cream cones. Edible Coatings and Soluble Packaging (R. Daniel), Noyes Data Corporation,
Park Ridge, NJ, 1970. pp. 280-282.
175. L. E. Werner, Encapsulated food acids. Cereal Foods World 25: 102 (1980).
176. J. C. Cordray and D. L. Huffman, Restructured pork from hot processed sow meat: effect of
encapsulated food acids, J. Food Prot. 48: 965 (1985).
Edible C oatings and Film B arriers 479
177. P. J. Parcel and D. W. Perkins, Process of preparing a particulate food acidulant, U.S. Patent
4,537,784, August 27, 1985.
178. Anonymous, Specially coated natural ingredient improves bread. Food Process. 34: 10 (1973).
179. F. Shahidi and X.-Q. Han, Encapsulation of food engredients, CRC Rev. Food Sci. Nutr. 33:
501-547 (1993).
180. T. Cea, J. D. Posta, and M. Glass, Encapsulated APM and method of preparation, U.S. Patent
4,384,004, May 17, 1983.
181. A. M. Schoebel and R. K. Yang, Encapsulated sweeteners composition for use with chewing
gum and edible products, U.S. Patent 4,824,681, April 25, 1989.
182. R. K. Yang and A. M. Schoebel, Chewing gum composition with encapsulated sweetener having
extended flavor release, U.S. Patent 4,911,934, March 27, 1990.
183. S. R. Cherukuri and G. Mansukhani, Sweetener delivery systems containing polyvinyl acetate,
U . S. Patent 4,816,265, March 28, 1989.
184. A. Koff, U.S. Patent 3,080,293, cited in Encapsulating with fatty acids. Edible Coatings and
Soluble Packaging (R. Daniel), Noyes Data Corporation, Park Ridge, NJ, 1973. pp. 329-331.
185. Anonymous, Fat matrix encapsulation controls ingredients release—reactions are temperature-
specific, Food Process. 37: 72 (1976).
186. E. Fujita, K. Ishii, H. Nakatani, Y. Nakai, and K. Katagiri, U.S. Patent 3,389,000, cited in Fat
encapsulation. Edible Coatings and Soluble Packaging (R. Daniel), Noyes Data Corporation,
Park Ridge, NJ, 1973. pp. 306-308.
187. S. S. Chang and P. X. Hoynak, U.S. Patent 3,261,692, cited in Encapsulating fat, Edible Coat
ings and Soluble Packaging (R. Daniel), Noyes Data Corporation, Park Ridge, NJ, 1973, pp.
314-319.
188. L. J. Henthom, U.S. Patent 3,507,665, cited in Encapsulating emulsion coating. Edible Coatings
327.
189. H. W. L. Westenberg, U.S. Patent 1,238,336, cited in Hardened fat coating. Edible Coatings
345.
190. Federal Register, Vol. 58, No. 3, January 6 , 1993.
191. Anonymous, General news: waxed fruit banned in Norway, Postharvest News Info. 1: 433
(1990).
192. N. Gontard, S. Marchesseau, J. L. Cuq, and S. Guilbert, Water vapor permeability of edible
bilayer films of wheat gluten and lipids, Int. J. Food Sci. Technol. 30: 49-56 (1995).
193. A. Gennadios, C. L. Weller, and R. F. Testin, Property modification of edible wheat, gluten-
based films, Trans. ASAE 36: 465-470 (1993).
194. I. K. Greener, Physical properties of edible films and their components, Ph.D. Thesis, Univ. of
Wisconsin—Madison, 1992.
195. M. Karel, Protective packaging of foods, in Principles of Food Science, Part II, Physical Princi
ples of Food Preservation (O. R. Fennema, Ed.), Marcel Dekker, New York, 1975. pp. 399-
466.
18________________________________________________________
Spray Processing of Fat-Containing Foodstuffs
Spray Drying and Cooling
Keith Masters
NIRO A/S, Soeborg, Denmark
L INTRODUCTION
Spray drying is one of the few continuous industrial drying techniques that converts heat-
sensitive liquid feedstocks into powders without the product quality degradation often associ
ated with dryers that use hot air to promote moisture evaporation. This important characteristic
of spray drying is further enhanced by the ability of a spray dryer not only to produce pow
ders, but to produce powders having specific desired properties defined by their particulate
structure (powders, agglomerates, granules), particle size distribution, bulk density, moisture
content, and other properties that define overall powder quality.
It is no surprise, therefore, that the food industry has long recognized the potential of
applying spray drying whenever liquid food concentrates are required in powder form. The
ability to handle production capacities of any size in one dryer and to adopt the latest process
automation ideas to continuous operation while achieving reproducibility of powder quality
during production are also factors that have played a part in firmly establishing spray drying
in food processing.
Spray drying [1] involves the atomization or spraying of a pumpable fluid food formulation
into hot air entering a drying chamber through an air disperser. Various configurations of air
disperser and drying chamber designs can be used, but whatever the design it is essential that
a controlled air flow is created in the drying chamber and that the droplets in the spray created
by the atomization stage are contained in this air flow. In this way, the best conditions are
created for rapid completion of moisture evaporation from each droplet and the formation of
a stable, dried particulate form.
The process description of spray drying indicates a sense of simplicity of process, but in
reality spray drying is a highly complex operation. Spray formation involves the creation and
handling of billions and billions of droplets, all of different sizes and moving in the confined
volume of a drying chamber. The movement of these droplets must then be so controlled that
all are subjected to similar heat and humidity environments. This will give the best possible
481
482 M asters
G. E. GRAY.
PROCESS o r DESICCATIHO.
ir r u iO A T io B riL S B i « i . le . i® u .
1,107 .784 . Patented Aug. 18.1S14.
Fig. 1 The first commercially successful spray dryer for food formulations: The Gray Jensen design,
patented 1914.
circumstances to obtain uniformity of the physical properties and overall quality of the
powder.
Every spray dryer is characterized by the flow of droplets, particles, and air within the dryer
layout. To achieve a continuous operation, extreme care has to be taken in selecting the design
of the air disperser, atomizer, and drying chamber and in specifying the operating conditions in
relation to the product being dried.
Different products have different drying histories, but any product that is hygroscopic or has
a sticky phase during its drying history is particularly problematic if it has to be spray dried. Full
control of the air flow pattern in the spray dryer is therefore essential to prevent droplets from
reaching the walls of the drying chamber too quickly. There must be sufficient airborne drying
time available to carry out the physical conversion of wet droplets into dried particles before the
resulting particulates reach the vicinity of the wall. If the surfaces of the particulates, on contact
with the wall, are too soft due to excessive moisture or heat, then wall deposits will occur. This
is an unwanted phenomenon of spray drying, as deposit formation leads to product quality degra
dation (misshapen, discolored, hard particles) and poor dryer performance (too frequent plant
shutdown for wet or dry cleaning). Furthermore, a safety hazard can be created as overheated
Spray Processing 483
deposits can be a potential ignition source for a powder explosion or fire. All powdered food
stuffs, by their very organic chemical structure, and especially those containing fats, exhibit ex
plosive and inflammable properties.
Many foodstuffs or food ingredients required in powder form contain constituents that give
rise to stickiness, and into this category falls the range of fat-containing products. The drying of
such products provides some of the greatest challenges to the spray dryer designer and plant op
erator. Successful spray drying is not as easy as it looks.
Spray drying was first used in industrial production at the beginning of the twentieth century,
and the earliest products handled included milk and eggs. Although the spray drying concept
was first patented in the 1860s, the early designs failed due to the lack of appreciation of the
importance of how droplets are contacted with hot air and the requirements of atomization and
air distribution to achieve this. In fact, it took nearly 50 years to establish the technology as in
dustrially feasible. The Grey Jensen design (Fig. 1) of 1914 was the first to show the way and
started a development that has accelerated up to the present day . With the range of designs now
available,^ it is possible to spray dry the most difficult products, and these include products hav
ing a high fat content.
II. PRINCIPLES OF SPRAY DRYING

Spray drying involves the process stages of atomization and hot air introduction, which com
bine to give the conditions of spray-air contact necessary to accomplish droplet drying and
particle shape formation during the residence time made available in the drying chamber.
Atomization can be conducted by a rotating wheel (rotary) atomizer or nozzles.
A. Atomizers
For drying high fat products, pressure nozzle atomizers (Fig. 2a) are widely used. In such a
nozzle atomizer, the liquid feed is fed to the nozzle under pressure (50-300 bar depending on
the feed properties, nozzle capacity, and droplet size distribution required). Pressure energy,
imparted to the feed by the high pressure pump, is converted to kinetic energy in the form of
a high speed thin conical film of liquid issuing from the nozzle orifice. A swirl insert within
the nozzle body creates the necessary liquid rotation to form this film. The film is very unsta
ble and disintegrates into droplets under the influence of internal forces and external turbu
lence effects. However, the atomized liquid has a distinctive spray (cone) angle on leaving
the nozzle, and this angle varies according to the product application, the required nozzle
conditions of operation, and the sizes of orifice and swirl inserts selected. Spray angles are
normally in the range of 50-110°. The size of the droplets in the spray varies inversely with
the pressure (exponential 0.6) and directly with feed rate and viscosity (exponential 0.2).
However, these exponentials are valid only over the somewhat limited capacity range of the
nozzle. As the turndown ratios of these nozzles are low, multinozzle assemblies must be used
in spray dryers designed for high feed capacities. However, within the capacity range of a
given nozzle, the feed rate varies with the square root of pressure.
Rotary (wheel) atomizers (Fig. 2b), on the other hand, utilize centrifugal energy. The
liquid feed is fed onto the center of the wheel, which is rotating with a high peripheral speed
(60-160 m/s). The liquid moves to the edge of the wheel under the influence of centrifugal
force, passing along vanes that prevent liquid slippage over the wheel surface. The liquid thus
achieves the peripheral velocity of the wheel and is able to use the available centrifugal energy
*Major companies supplying spray dryers to the food industry include APV Anhydro A/S Copenhagen, Denmark;
NIRO A/S Copenhagen, Denmark; and Stork Friesland B V , Gorredijk, The Netherlands.
484 Masters
(a)
0 m
(b)
Fig. 2 (a) Pressure nozzle in operation, showing its distinctive spray angle, (b) Rotary atomizer in
operation, showing the horizontal trajectory of spray droplets from the wheel. (Courtesy of the NIRO
Group.)
to the full. The liquid is ejected from the wheel edge as a thin film that readily disintegrates
into droplets involving a mechanism similar to film break-up from nozzle atomizers. Air
turbulence caused by the wheel rotation is also a major factor in the atomization mechanism
(film break-up). The size of the droplets created depends on the wheel speed (peripheral
velocity). The higher the speed, the smaller the droplet size (exponential 0.8). Rotary atomiz
ers have infinite turndown potential, and there is very little effect of feed rate on droplet size
within the feed rate variations occurring in a given spray drying operation. Viscosity becomes
an influential factor only at high viscosity values.
B. Spray Dryer Chamber Designs

The design of the drying chamber in which the atomizer and air disperser are located de
pends on
1. The droplet size distribution required from the atomizer that will form the basis for
the end product powder specification (size and structure) (Fig. 3).
2. The time droplets and particles must remain in the chamber (residence time) to com
plete drying down to the specified value of residual moisture content in the end
powder.
3. The maximum temperature to which dried and semidried particles can be subjected
and reach before heat degradation occurs.
4. The air flow pattern required in the drying chamber to minimize wall deposit for
mation.
For the drying of fat-containing products, two categories of design are applicable:
1. Cocurrent, in which droplets, particles, and air pass through the drying chamber in the
same direction.
2. Mixed flow, in which droplets and particles pass downward through the drying cham
ber. The air initially flows downward only for the flow to be reversed so as to enable
a top exhaust.
Drying chambers that conform to these categories are shown in Fig. 4.
Figure 4a. The standard cocurrent, conical base chamber operates with a two-point dis
charge system. Powder is discharged from the chamber base, whereas air and entrained fines
(particles from the lower end of the size distribution) are exhausted from the side of the cone into
cyclones for fines collection and reprocessing. The air disperser located centrally in the roof of
the drying chamber creates a cyclonic air flow around the atomizer and down into the cone of the
drying chamber. Both nozzle and rotary atomizers can be utilized. Pneumatic powder-conveying
systems can be installed at the base of the chamber to give the so-called one-stage drying layout,
i . e . , powder leaves the drying chamber at the residual moisture content required in the end prod
uct. However, vibrating fluidized beds can also be installed at the base of the chamber for
achieving two-stage drying. This lowers the product temperature during drying by operating the
spray dryer at a lower exhaust air temperature. This in turn results in powder leaving the drying
chamber at a higher residual moisture content than that specified for the end product. Therefore,
drying is completed in the second-stage vibrating fluidized bed, which also includes powder
cooling prior to discharge. Two-stage drying produces agglomerated powders of improved
flowability and wettability and creates a spray drying operation of higher thermal efficiency.
Drying chambers of this design are suited for only low fat content products and products having
virtually no sticky and hygroscopic characteristics.
The air temperature profile in this design is given in Fig. 5a. The temperature of powders
486 Masters
(a)
1mm:
m (c)
Fig. 3 Spray-dried, fat-containing food ingredients, (a) Typical agglomerated free-flowing powders.
(b) High fat powder ingredients: Butter (sample shown), fish oil, lard, lecithin, tallow, vegetable oil.
(c) Dairy foods: Baby formula, butter, buttermilk, casein, caseinate, cheese, cream, lactose hydrolyzed
whey (sample shown), permeate, skim milk, sweetened condensed milk, whey, whole milk, yogurt.
(Courtesy of the NIRO Group.)
discharged from the chamber is approximately 20-25°C lower than the exhaust air tempera
ture required to achieve the required residual moisture content in the powder.
Figure 4b. The tower (tail-form) cocurrent, conical base chambers are used exclusively
with nozzle atomizers. The chamber can be supplied with an oversize lower cone section
(bustle) to maximize powder discharge and minimize powder carryover to the cyclones. The
air disperser creates a streamlined air flow with a minimum of turbulence down the chamber.
This maintains both droplets and particles airborne, keeping them away from the surrounding
walls until well into the lower cone section. A typical air temperature distribution through the
chamber is shown in Fig. 5b.
(C)
Fig. 4 Chamber designs used in the spray drying of foodstuffs, (a) Cocurrent, conical base, wide
body; (b) cocurrent, conical base, tower body; (c) cocurrent, integrated belt, Filtermat body; (d) mixed
flow, integrated fluidized bed, MSD body. A = air, F = feed, P = dried product.
This design is used for low to medium fat content products, baby foods, and milk and
coffee powders, especially when large individual particles are required to maximize bulk den
sity and flowability. The drying chamber design can be operated with a vibrating fluidized bed
attachment acting as powder cooler/conditioner for fat-containing powders and as an afterdryer
agglomerator/cooler for instant, agglomerated powders.
Figure 4c. Integrated belt, cocurrent chambers feature nozzles spraying feed down into
hot air and onto a moving belt. The first section of the drying chamber resembles a short
nozzle tower with an air disperser creating a streamline, nonturbulent air flow pattern. Semi-
dried powder accumulates on the belt and moves with the belt into further sections of the
drying chamber where completion of drying and cooling is achieved with secondary warm
and cool air flows. All incoming air passes through the powder layer and belt fabric prior to
exhausting. In this way, the powder acts as an effective particulate filter and hence negligible
amounts of fines are collected in the cyclone. In many cases, further air cleaning equipment
such as scrubbers or bag filters is not required after the cyclone, as environmental emission
standards can be met with just a cyclone.
488 Masters
= A ir ooo {> = P roduct ^3^ = F lu id feed
Fig. 5 Air temperature profiles in spray dryer chambers. Numbers are temperatures in degrees Celsius.
Note: Profiles are for operating conditions with inlet hot air temperatures much higher than those used
in spray drying fat-containing formulations. This is to illustrate that even with these higher inlet levels,
air temperatures in the drying chamber are maintained low and the corresponding product temperatures
even lower.
With the previously described cocurrent designs, air and powder residence times in the
drying chambers are comparable. In the integrated belt design, the residence time available
for the powder as it moves slowly on the belt through the drying chamber is much longer
than the time required for the air to flow from the air disperser to the cyclone. This makes
this design very suitable for high fat content products and products that slowly crystallize,
i.e., sugar-containing foodstuffs.
The fact that all particles during their most soft-surface (stickiest) state are lying on the
belt and not swirling in air flows makes the design ideal for products that cause the greatest
problems of wall deposits in more conventional chamber designs. Powder adhesion on the
belt is prevented by the selection of the belt material. A further important feature of operation
is the very low exhaust drying air temperature required to achieve completion of drying.
Powder temperatures are maintained very low, which is very important for fat-containing
powders. Powders produced on this type of design have an agglomerate form.
Figure 4d. The mixed flow integrated fluidized bed drying chambers represent a design
frequently used to handle the more difficult drying operations, i.e., with fat-containing prod
ucts or foodstuffs exhibiting hygroscopic and sticky characteristics. However, for the highest
of fat contents and for sugar-containing products, the integrated belt design is preferred.
Nozzle atomizers are normally used, although in large-diameter drying chambers, rotary
atomizers are equally successful. The air disperser creates a strong downward flow of air so
that the spray is projected deep down into the lower cone of the chamber toward the fluidized
bed that is installed in the chamber base. Operation of a drying chamber with an integrated
fluidized bed enables lower outlet temperatures to be used than with the standard cocurrent
and tail-form tower designs and thus leads to lower powder temperatures and higher thermal
efficiencies. A typical temperature distribution through the chamber is shown in Fig. 5c. The
drying air from the roof-mounted air disperser together with the air supplied to the fluidized
bed becomes mixed in the central part of the drying chamber and then flows back up the
chamber for exhaust via ducts mounted at the top of the chamber walls or through the roof,
away, however, from the air disperser. The fluidized bed operates as a totally mixed fluidizing
system as this allows powder that cannot be directly fluidized at the moisture content it pos
sesses on entering the fluidized layer to be successfully and continuously handled at the base
of the chamber, thereby achieving completion of drying and agglomeration. This type of
drying chamber is used for heat-sensitive, sticky products where agglomerated free-flowing
powders are required.
All four designs described above have full cleaning-in-place (CIP) capability.
C. Drying Mechanisms
Irrespective of the design of the drying chamber, the mechanism of droplet drying follows the
same principles.
The moisture in a spray droplet can be present in two forms: bound and unbound. The
nature of the solid and accompanying moisture determines the drying characteristics. The
bound moisture in the solid exerts an equilibrium vapor pressure lower than that of the pure
water at the same temperature. Water retained in small capillaries in the solid absorbed at
solid surfaces as solutions in cells or fibers falls into the category of bound moisture. The
unbound moisture in a hygroscopic material is that moisture in excess of the bound moisture.
All water in a nonhygroscopic material is unbound water that exerts an equilibrium vapor
pressure equal to that of pure water at the same temperature. The equilibrium moisture is the
moisture content of a product when it is at equilibrium with the partial pressure of the water
vapor of the surroundings. The free moisture is the moisture in excess of the equilibrium
moisture and consists of unbound and some bound moisture. Free moisture is evaporated
during spray drying.
The mechanism of moisture flow through a droplet during spray drying is diffusion supple
mented by capillary flow. The drying characteristics of the droplet depend upon whether
bound or unbound moisture is evaporated, as each has distinct features. As long as unbound
moisture exists, drying proceeds at a constant rate and will continue while the rate of moisture
diffusion within the spray droplet is fast enough to maintain saturated surface conditions.
However, for food-based products, constant-rate periods are very short. When diffusional and
capillary flows can no longer maintain these conditions, the critical point is reached, and the
drying rate will decline until an equilibrium moisture content condition is attained. However,
in practice, the residence time for particles in the spray dryer is too short to allow equilibrium
conditions to be established with the exhaust air.
The residual moisture content for a given dryer operation is thus related to the temperature
and humidity levels of the exhaust air. In the majority of industrial spray dryer operations,
saturation conditions at the powder discharge area are never approached, and thus maintaining
the exhaust air temperature at a set point is control enough for the residual powder moisture
content. This outlet temperature is maintained constant either by (1) adjustment to the feed
rate to the atomizer at fixed heat input to the air heater supplying air to the air disperser or
(2) adjustment of the heat input at the air heater at constant feed rate to the atomizer. To
compensate for humidity changes (from day to night, or from winter to summer), the outlet
temperature set point may have to be readjusted up and down as required to maintain a
constant residual moisture content value.
490 Masters
Control systems that account for those temperature and humidity variations are available if
such a degree of control is required for continuity of operation and reproducibility of powder
quality. Systems for continuous measurement of moisture content are also available, but it is
seldom that these measurements are used to control the spray dryer. It is the outlet temperature
measurement that is the important control variable in all spray dryers for foodstuffs.
As the resistance to moisture transfer from within each spray droplet to the surface differs
from product to product, the shape of the particle also differs, even under the same atomiza
tion and air dispersion conditions. Furthermore, droplets of the same product are always sub
jected to different local temperature, humidity, and velocity environments, and therefore
spray-dried particles are never homogeneous in shape and size. Powder viewed under a micro
scope displays a wide range of structures including individual and agglomerated particles and
misshapen and broken particulates. The inability to produce homogeneous, uniform particles
in the chamber is no longer an important criterion, because with today’s development of spray
drying systems, the original atomized droplet and particle size is used as the starting material
for aftertreatment techniques in integrated fluidized beds and belts or externally mounted vi
brating fluidized beds to create the uniformity of the final agglomerate size and structure
required by today’s quality standards.
III. TYPICAL HIGH FAT SPRAY-DRIED POWDERS

One of the most important groups of fat-containing foodstuffs is based on milk powders
produced by spray drying. Such products are intended for both human consumption and ani
mal feeding and often have a fat content of more than 35%. Products for human consumption
can either serve the needs in direct household use or act as raw materials for further industrial
processing. Few high fat milk powders employ only milk fat as the fat constituent. To this
group belong high fat, full cream powders, dried cream, and dried butter powder. A common
feature of all other products is that the cream has been removed and the milk fat has been
replaced by vegetable or animal fat. Regardless of the origin or kind of fat used, the composi
tion of these products can vary greatly.
Historically, the production of fat-filled milk powders started in the 1960s. These were
used as milk replacers for feeding calves. The original milk replacers were produced with a
fat content of 18-22%, Such levels of fat enabled conventional one-stage cocurrent spray
dryers with pneumatic transport systems to be used (see later, Fig. 6). Typical installations
had evaporative capacities of 500 kg/h and used moderate inlet drying air temperatures. It
was soon recognized that the installed spray dryer capacity could be better utilized by pro
cessing skim milk in the peak season and filled milk powder of high fat content during the
off-peak season. This production trend created the need to better understand the drying mecha
nisms of high fat milk formulations and stimulated development in design to achieve improved
continuous and high quality production. This involved dry mixing and recombination tech
niques. High fat powder was blended with skim milk powder to obtain a product of the same
composition as that produced by drying the original raw material directly. High fat content
concentrates prepared for spray drying contained initially about 35% fat, a percentage that has
increased over the years with the development of the technology to handle higher fat levels.
Today, fat levels of 60% are by no means uncommon, and the spray dryers suited for handling
such formulations incorporate fluidization or belt drying technology for the necessary opera
tional stages of drying, cooling, and powder handling (Figs. 4c and 4d).
Milk replacers for calves can be considered the pioneer product that opened the way to
more sophisticated fat-containing products for both household and industrial use. Many
branches of the food industry have always used dairy-based ingredients in their classical form.
i.e., skim milk, full cream milk, cream, butter, and whey, as components in their recipes.
However, automation of these industries has demanded modified raw materials that are better
suited for dosing and transportation and have a longer shelf life and improved functional prop
erties.
The driving force of the milk powder industry to develop new products was the demand
and pressure from food manufacturers—i.e., bakeries and manufacturers of soups and sauces
in liquid and dried form, ice cream, confectioneries, sweetened condensed milk, cream-based
liquors, chocolate, and meat products—in fact, all industries who were seeking new, innova
tive, and functional food ingredients that could assist in extending their range of products of
improved quality, all of course at reduced manufacturing costs.
IV. SPRAY DRYING HIGH FAT FORMULATIONS

Spray-dried powders of milk-based formulations containing fat include
Full cream milk powder with fat content greater than 35%
Full milk powder of high free fat content
Fat-filled milk powder with fat content of 40-60%
Fat-filled milk powder with glycerine monostearate content of 30%
Fat-filled whey powder with fat content of 40-60%
Butter powder with fat content of up to 80%
Soft-serve ice cream mix powders
Coffee whiteners
Cream liquor powders
Each of these products appears in numerous compositional variations with functional proper
ties tailormade to the needs of a particular customer. The raw materials available for use in
clude
Caseinates Whey (sweet/acid) Glucose

Cream Whey protein concentrate Lard
Emulsifiers Whole milk Sequestering agents
Free-flowing agents Animal fat Soybeans
Gelatin Antioxidants Sugar
Skim milk Coconut Sunflower seed
Stabilizers Com Tallow
Starches Cottonseed Vegetable oil
Vitamins and minerals
A. Fat-Filled Powders
Irrespective of the formulation, the mix preparation is very important to ensure a trouble-free
drying operation and production of a quality powder. Most formulations contain skim milk
or whey solids as the non-fat component. A typical processing line based on skim milk is
as follows.
Skim milk is concentrated in a multiple-effect, falling film evaporator to achieve 50-55%
total solids level after mixing in the fat. The fat components, together with emulsifiers and
possibly vitamins, are stored at 50-55°C, but not for longer than 24 h. The fats and concen
trated skim milk are then mixed in the ratio necessary to achieve the desired final fat content.
The resulting mix is stored in agitated tanks prior to homogenizing and spray drying.
492 Masters
Homogenization is conducted in two stages. For mixes containing approximately 50% fat
in the total solids, homogenizing pressures in the first stage are in the range of 70-80 bar and
in the second, 20-30 bar. At higher fat contents, pressures should be proportionally lowered.
Following homogenization, the mix concentrate is spray dried. Moderate inlet drying air
temperature levels of 180°C are used for formulations having fat contents of less than 50%.
Should the fat contents be higher, lower inlet drying air temperatures are recommended. Skim
milk powder containing glycerine monostearate as the fat component is processed in a simi
lar way.
For the processing of formulations using whey powder as the non-fat constituent, similar
procedures are used, but it is advantageous to precrystallize the whey concentrate and reheat
to ~50°C before mixing with the fat. The process has been subjected to much fundamental
study including the stability of emulsion when fat-filled whey powders of about 50% are
applied for producing milk replacers. Two problems have been highlighted during reconstitu
tion: foaming and excessive creaming of the emulsion. The explanation of these phenomena
relates to cluster formation during homogenization when using high pressures and at low
protein/fat ratios. This problem is reduced with the addition of a calcium sequestering agent
prior to homogenization.
B. Butter Powder
With its 70-80% milk fat (approximately the same level as in normal table butter), butter
powder is most difficult to dry and handle. Mix preparation involves the addition of sodium
caseinate, non-fat milk solids, emulsifiers, and stabilizing salts. Spray dryers with belt or
fluidized bed coolers are essential, as is also a final moisture content of the powder, which
must be below 1%. If the powder is to be used in further processing, a free-flowing agent is
added. The product is used exclusively in bakeries as a source of milk fat in dry form for
making croissants.
C. Whole Milk Powder with High Free Fat Content

High free fat content powder having a total fat content of 28% does not by definition belong
to the category of high fat milk powders. However, the characteristics of this product and the
difficulties experienced in drying it justify its mention in the coverage of high fat powder pro
duction.
The free fat content of such powder should be more than 90% of the total fat, and this is
achieved by precrystallizing the lactose in the whole milk concentrate. Lactose in conventional
whole milk powder is present in its amorphous form and provides the continuous phase of the
milk powder particle encapsulating the fat globule. Higher degrees of crystallization and even
higher free fat content are achievable in the spray dryer designs that feature an integrated
fluidized bed or belt where the extended residence time of the moist powder in the fluidized
bed and on the belt enables continuing aftercrystallization.
D. Coffee Whiteners and Coffee Creamers

Coffee whiteners and coffee creamers are available to the consumer with varying recipes, but
most of them do not contain any milk solids. These products were developed to avoid the
flocculation that appears when whole milk powder is used in hot coffee or tea. This floccula
tion was attributed to the whey protein. The pH of coffee and tea is quite low, sometimes well
below 5, which together with the temperature of the beverage creates favorable conditions for
denaturing of whey proteins on the surface of particles before acceptable dispersion and dis-
solving could take place. Typical constituents that make up coffee whiteners include glucose
(com symp solids), sodium caseinate, vegetable fats, emulsifiers, stabilizers, stabilizing salts,
and possibly sucrose and some non-fat milk solids. The formulation is spray dried in dryers
with or without integrated fluidized beds to produce a fairly agglomerated product that will
give good functional properties to the powder. Compared with high fat formulations, coffee
whiteners and creamers are classified as among the easier products to spray dry.
E. Ice Cream Powder

An ice cream powder can be considered a whole milk powder with a fat content of up to
45%. However, part of this fat can be vegetable fat. Besides milk solids, powders also consist
of sucrose and emulsifiers. Ice cream mix powders can be flavored and colored prior to the
preparation of the ice cream or left neutral. Owing to the fat and sugar content, ice cream
powder is produced on spray dryers featuring integrated fluidized beds or belts.
F. Cream Liquor Powder

Cream liquor powder is produced by spray drying a feed formulation consisting of sodium
caseinate, cream (48% fat), sugar, starch, glucose, coloring and flavoring ingredients, emulsi
fiers, and milk or vegetable fat, forming a mixture containing approximately 60% fat on a dry
weight basis. This is homogenized prior to spray drying to produce a stable emulsion. The
high fat content requires drying in spray dryers with integrated fluidized beds or belts.
V. SUITABILITY OF SPRAY DRYER DESIGN: GUIDE

TO SELECTION
The suitability of a given spray dryer design depends on the feed formulation and its total fat
content. The higher the fat content, the more difficult the spray-drying operation becomes, and
the more sophisticated the design of the plant required. For fat contents up to 35%, conventional
layouts suffice (Fig. 6), but for higher fat contents, fluidized bed cooling systems become essen
tial (Figs. 7 and 8), and the problem of handling powder fines in the cyclones must be addressed.
Cyclones have to be sized for low pressure drop operation so as to prevent excessive deposit
formation and bridging of the powder at the outlet. All mechanical handling of such fine particles
Fig. 6 Spray dryer layout with pneumatic conveying of powder from drying chamber and cyclone.
494 Masters
Fig. 7 Spray dryer layout featuring external vibrating fluidized beds as powder cooler/afterdryer-
cooler.
must be substantially reduced or eliminated, and therefore the plant layout must allow for gravity
flow of these fine particles from the cyclones, etc., to the cooling bed. Formulations with up to
60% fat can be handled on spray dryers with rotary atomizers; however, there is the added risk
of increased wall deposits at increasing fat content. This is because in spray dryers that use rotary
atomizers and air dispersers creating a cyclonic air flow in the drying chamber, particles cannot
be prevented from moving radially toward the drying wall under the influence of the cyclonic
Fig. 8 Spray dryer layout featuring tower (tail-form) drying chamber 1, external vibrating fluidized
bed, and powder fines recycle from cyclones. Final exhaust air cleaning in bag filter optional.
Fig. 9 Spray dryer layout (MSD) featuring drying chamber with integrated fluidized bed, external
vibrating fluidized bed, and powder fines recycle from cyclones.
turbulent air flow. If the particles move to the wall too fast and contact is made while the parti
cles are soft and in a sticky state, then excessive deposits will occur. For this very reason, when
powders of fat contents higher than 60% are required, dryers with alternative spray and air flow
patterns are necessary, i.e., nozzles must be used and a more streamlined air flow established
■down the center of the dryer as in the tower-type design or in the spray dryer with an integrated
fluidized bed or belt (see Figs. 8-10).
Fig. 10 Spray dryer layout (Filtermat) featuring drying chamber with integrated belt and powder fines
collection in cyclone. No fines recycle required.
496 Masters
Table 1 Suitability of Design
spray dryer design^
A B c D E
Product Fat content (%) (Fig. 6) (Fig. 7) (Fig. 8) (Fig. 9) (Fig. 10)
Milk replacer ~35 1 2 3 3 4

-4 0 -5 0 0 2 3 3 4
-5 0 -6 0 0 1 3 3 4
Fat-filled whey -3 5 1 2 3 3 4
-4 0 -5 0 0 2 3 3 4
-5 0 -6 0 0 1 3 3 4
Coffee whitener -4 0 - 4 5 1 3 4 4 4
Butter >70 0 0 1 2 4
^Classification: 0, Not recommended. 1, Possible, especially as a multipurpose plant with non-fat or low fat powders
as main products. 2, Suited. 3, More suited. 4, Most suited.
As a general rule, the higher the fat content, the more difficult the drying operation due
to the increased tendency for wall deposits to build up. This tendency can be reduced by
using lower inlet drying air temperatures, lower homogenizing pressures in the feed sys
tem, and higher feed concentrations and cooling the chamber walls. Furthermore, powder
flowability can be improved by incorporating the recycling of the very fine particles collecting
in the cyclone back to the spray drying chamber to coat the spray droplets with powder as
they leave the atomizer. The fines return system can also be used to add other solid ingredients
a' :-o
Fig. 11 Spray dryer with installed safety suppressant system. (Courtesy of the NIRO Group.)
such as skim milk powder in a planned powdering technique. The use of such techniques
means that the feed must have a higher fat content than is required in the final product. The
suitability of spray dryer designs available to the food processing industry for the handling of
fat-containing feedstocks is summarized in Table 1. The spray dryer with integrated belt is
considered the most suitable for fat-containing products. An industrial-size dryer is shown in
Fig. 11.
VI. SAFETY ASPECTS

Spray dryers operating on fat-containing products have a fine safety record, but there are
examples where explosion and fire have caused structural damage. Fat-containing powders
can form explosive mixtures in air, and it is impossible to prevent the likelihood of dangerous
powder-air concentrations existing in some part of the spray dryer chamber and auxiliary
powder handling equipment. The fat content itself presents a fire hazard under any circum
stances. Therefore, during the spray drying of fat-containing formulations every precaution
must be taken to eliminate ignition sources. Experience has shown that the usual source is
overheated deposits falling through the swirling powder cloud in the spray drying chamber
and into the fluidized bed integrated within or attached to the chamber. Therefore close atten
tion must be given to the prevention of deposits on the roof air disperser and atomizer assem
bly, as these are located in the hottest zones of the dryer. Furthermore, the use of systems
il
ai
'^1
I
Fig. 12 Spray dryer chamber with removable air insulation panel. (Courtesy of the NIRO Group.)
498 Masters
lia
^ -'-"R
^ M l.
;î''.'t4-r--'‘.
i|H |
Fig. 13 Filtermat spray dryer designed for food ingredients including fat-based formulations. (Cour
tesy of the NIRO Group.)
that return fines via the atomizer must not permit interference of the overall air flow pattern
in the drying chamber and create a turbulence that can direct the fines, which consist of
powder of small particles, back into the air disperser and hot air zone. Deposits on the lower
wall areas are a lesser danger as they are at much lower temperatures.
All suppliers of spray dryers for high fat formulations have to address the safety issue by se
lecting and providing the most suitable design. For these types of products, designs have to be
quite sophisticated; therefore they represent a fairly expensive investment. As a minimum the
design must be equipped with an automatic fire extinguishing system for protection against the
spreading of fire, and overpressure venting for protection against a sudden pressure rise in an
explosion situation. Installation of overpressure vents inevitably influences plant performance
by increasing the possibilities of air leakage and liquid seepage during CIP cleaning. There is
always an advantage, therefore, to minimizing the venting areas, and this is done without com
promising safety by strengthening the chamber structure to a higher pressure shock-resistant rat
ing so that a smaller vent area can be used. An alternative to pressure shock-resistant designs is
the use a suppressant system, but this approach is only beginning to be considered as a practical
alternative within food drying. Pressure systems involve canisters of a fire extinguishing chemi
cal that is dispersed throughout the spray dryer as soon as an explosion hazard situation arises.
Figure 12 shows a spray dryer equipped with such a system. The canisters are on the cone and
can be clearly seen. However, venting and strengthened structures still remain the most widely
used safety features on spray dryers handling fat-containing products. Whenever venting is used,
it is most important that all venting areas be ducted to safe areas, preferably outside the building,
so as to prevent the possibility of a secondary explosion and fire occurring just outside the spray
dryer structure. Ducting also protects the operator in the working areas and protects the building
against partial or total collapse.
Aspects of safety always focus on degrees of deposit formation that can be expected with
a given fat-containing formulation, since the nature of the products gives a high possibility of
deposits forming at the walls. Deposit formation is usually minimized through the correct
adjustment of the air disperser and location of the atomizer assembly, but the quality of the
wall insulation and effective overall cooling also has an effect on the positioning of deposits.
In recent years, the whole situation regarding deposit formation has been improved by the use
of air insulation panels that are also removable. This gives not only the advantage of reducing
deposit tendencies but also the ability to inspect the wall easily for cracks or pinholes in the
stainless steel plate that can cause safety problems associated with bacterial contamination. A
dryer chamber equipped with such removable insulation panels is shown in Fig. 13.
ACKNOWLEDGMENTS
I thank the management of the NIRO Group of Companies for permission to use NIRO photo
graphs and flow diagrams in illustrating the chapter. I am also grateful to Dr. J. Pisecky for
his advice and technical input.
REFERENCE
1. K. Masters, Spray Drying Handbook, 5th ed., Longmans, London, 1991.
BIBLIOGRAPHY
Bucheim, W., Electron microscopic localization of solvent-extractable fat in agglomerated spray-dried
whole milk powder particles. Food Microstruct. 1: 233-238 (1982).
Buma, T. J., Free fat in spray dried whole milk. I. General introduction and brief review of literature,
Neth. Milk Dairy J. 25: 33-41 (1971).
Buma, T. J., Free fat in spray dried whole milk. II. An evaluation of methods for the determination of
free fat content, Neth. Milk Dairy J. 25: 42-52 (1971).
Buma, T. J., Free fat in spray dried whole milk. III. Particle size. Its estimation, influence of processing
parameters and its relation to free fat content, Neth. Milk Dairy J. 25: 53-72 (1971).
Buma, T. J., Free fat in spray dried whole milk. IV. Significance of free fat for other properties of
practical importance, Neth. Milk Dairy J. 25: 88-106 (1971).
Buma, T. J., Free fat in spray dried whole milk. V. Cohesion; determination; influence of particle size,
moisture content and free fat content, Neth. Milk Dairy J. 25: 107-122 (1971).
Buma, T. J., Free fat in spray dried whole milk. VIII. The relation between free fat content and particle
porosity of spray dried whole milk, Neth. Milk Dairy J. 25: 123-140 (1971).
Buma, T. J., Free fat in spray dried whole milk. IX. The size distribution of fat globules in concentrated
milk and spray dried milk, Neth. Milk Dairy J. 25: 151-158 (1971).
Buma, T. J., Free fat in spray dried whole milk. X. A final report with a physical model for free fat in
spray dried milk, Neth. Milk Dairy J. 25: 159-174 (1971).
Faeldt, P., and B. Bergenstahl, Fat encapsulation in spray-dried food powders, J. Am. Oil Chem. Soc.
72: 171-176 (1995).
Faeldt, P., B. Bergenstahl, and G. Carlsson, The surface coverage of fat on food powders analyzed by
ESCA (electron spectroscopy for chemical analysis). Food Struct. 12: 225-234 (1993).
500 Masters
Hansen, P. S., Production of agglomerated fat-filled milk powder, J. Soc. Dairy Technol. 33: 19-
23 (1980).
Hilaire, J. R., Spray-drying: principle and recent developments, Via-Lactea 8: 13-15,17-20 (1976).
Horn, J. D., Spray dried fats, Int. Food Flavor. Food Addit. 7: 64-65, 68 (1976).
Hussmann, P., Process for producing emulsified, edible fat preparations of good keeping quality, Ger
man Patent Appl. 1 417 445 (1970).
Kautz, K., Method for manufacture of free flowing powders from fat and flour, German Patent Appl. 1
417 553 (1968).
Kharitonov, V. D., Distribution of free fat on the surfaces of dried milk particles, Molochn.-Prom. 7:
12-15 (1975).
Kubicki, M., and S. Misicki, Spray drying of emulsion in powdered fat manufacture, Tluszcze-Jadalne
18: 225-238 (1974).
Lipatov, N. N., G. B. Dvoretskii, L. A. Kabanov, and N. N. Mizeretsikii, Method for production of
dried rapidly soluble fat-containing milk products, USSR Patent 428 734 (1974).
Loo, C. C., Process for the production of rapidly dispersible dried milk, or of a milk based fat con
taining dried product, German Patent Appl. 1 492 773 (1969).
McMahon, T., Powdered lipid composition and its production and use, UK Patent Appli. 2 031 937A
(1980).
Mol, J. J., The milk fat globule membrane and the solubility of whole milk powder, Neth. Milk Dairy
J. 29: 212-224 (1975).
Nava, L. J., J. T. Hutton, J. Shields, and C. Kempf, Process for manufacturing instant fat containing
milk, German Patent Appl. 1 492 784 (1969).
Onwulata, C., P. W. Smith, J. C. Craig, Jr., and V. H. Holsinger, Physical properties of encapsulated
spray-dried milk fat, J. Food Sci. 59: 316-320 (1994).
Onwulata, C. L, and V. H. Holsinger, Thermal properties and moisture sorption isotherms of spray-
dried encapsulated milk fat, J. Food Process. Preserv. 19: 33-51 (1995).
Pisecky, J., The new generation of spray-driers for milk products, Deut. Molkerei Z. 104: 886, 888-
891 (1983).
Pyle, J. R., Instant milk powders and the principles of high-fat powder manufacture, in Winter School
on Spray Drying, see FSTA (1976) 8 2P275, 1975.
Rayner, P. B., Spray dried fats—their nature and applications, Flavour Ind. 4: 379-380 (1973).
Sadini, V., Modification of organoleptic properties and presentation of butter, Int. Dairy Congr. E:
891 (1978).
Tamsma, A., F. E. Kurtz, A. Kontson, and C. Sutton, Reconstitutability of spray dried milk products
containing 26 or 13% fat, J. Dairy Sci. 58: 1433-1436 (1975).
Tamsma, A., and A. Kontson, Preparation of a foam spray dried whole milk type product with good
sinkability, J. Dairy Sci. 57: 1149-1151 (1974).
19__________________
Low Calorie Fats
John W. Finley,* A. McDonald,f and L. P. Klemann
Nabisco Foods, East Hanover, New Jersey
I. INDUSTRIAL SIGNIFICANCE
The development of food products with flavor and quality attributes similar to their full fat and
calorie counterparts but reduced in both fat and calories has presented a major technological
challenge for food manufacturers. Reducing fat content requires replacement of part or all of the
fat by an ingredient or system of ingredients that can effectively duplicate the diverse functional
roles of the fat it replaces in a wide variety of product applications. Reducing fat content is the
most efficient means for reducing caloric content because, with a caloric contribution of more
than twice that of carbohydrate or protein, less than half as much fat has to be removed to
achieve an equivalent caloric reduction. Consumer demand for high quality products low in both
fat and calories has stimulated the development of a number of new technologies and innovative
ingredients aimed at replacing fat in food products without compromising its functional contri
butions to flavor delivery, textural characteristics, and heat stability.
II. APPROACHES TO REDUCTION OF CALORIES

FROM FAT
Caloric contribution from fat can be reduced by one or more approaches including fat reduc
tion by emulsion technology, fat mimetics, low calorie fats, and fat substitutes. Emulsifiers
are also lipid-based, but some may include carbohydrate moieties as well. Fat mimetics do
not have any lipid components, being derived entirely from either protein or carbohydrate
sources. Low calorie fats are fats that exhibit only partial absorption. Fat substitutes are fatlike
materials that are not digested but function like fat in food applications. In addition to use as
food ingredients, some fat replacers may also function by forming films that serve as barriers
^ C u r r e n t a ffilia tio n : Monsanto Company, Chesterfield, Missouri.
'\C u r r e n t a ffilia tio n : Consultant, Chicago, Illinois.

501
502 Finley et al.
to fat absorption. These types of fat replacers are useful in frying applications to reduce the
amount of fat retained by batter-coated or breaded products.
Fat replacement and caloric reduction are readily achieved by fat mimetics because they
are not lipid compounds. Because they are not fats, their applications are sometimes limited
by their lack of heat stability and true fat functionality. Emulsifiers generally serve as fat
extenders, enabling reduced levels of fat to perform as effectively as larger amounts by max
imizing functionality. Most emulsifiers have the caloric value of fat on a weight basis but
contribute fewer calories to finished food because only small amounts are used. Low calorie
fats and fat substitutes, on the other hand, must accomplish fat reduction through structural
modifications that reduce digestibility. If less fat is available for absorption, its metabolic
caloric value is reduced, as is the possibility of other metabolic effects typically associated
with high fat intakes.
A. Low Calorie Fats and Fat Substitutes

Low calorie fats are genuine triacylglycerol mixtures that have been chemically and physically
modified to provide fewer than the usual 9 cal/g normally contributed by lipids. This reduction
in caloric value is accomplished through designating a specific fatty acid compositional mix
and the positional arrangement of these fatty acids on the triacylglycerol backbone to reduce
availability. In contrast, fat substitutes are engineered lipids that bear only a resemblance to
the natural triacylglycerols of food lipids, although edible fats and oils are frequently used as
starting materials in the synthesis of these compounds. Fat substitutes are actually structural
analogues of natural lipids that either completely or almost completely resist hydrolysis by
digestive enzymes and thus contribute minimal or no caloric value. Although fat substitutes
can technically be considered low calorie fats because they have lower caloric values than
regular fats, they are distinguished from low calorie fats in this chapter on the basis of their
origins. As such, use of the term “fat” is applied only to those compounds that are true fats
in the sense of their structural similarity to the triacylglycerol moieties of edible fats and oils.
B. Emulsifiers
Emulsifiers have been in wide use in fat replacement systems for some time. Incorporating
emulsifiers such as mono- and diacylglycerols into the formulation of fat-reduced products is
appealing because these compounds are familiar ingredients with an established record of
safety. Emulsifiers perform as fat replacers most optimally when used in combination with
other emulsifiers and as part of a fat replacement system with other ingredients such as water
or flavorings. Classification of the structural, functional, and metabolic properties of emulsifi
ers place these compounds somewhere between carbohydrates and lipids. The extent to which
they exhibit properties of lipids is a function of the degree of esterification present [ 1 ].
Emulsifiers may be used as fat extenders in partial replacement of fat or to completely
replace fat in a product. Emulsifiers most commonly used in fat replacement systems at the
present time are those surface-active molecules that have had a previous history of successful
use as stabilizers in a variety of food products. Examples of such emulsion stabilizers are
polyglycerol esters, poly sórbate, and propylene glycol mono- and diesters [2]. When emulsi
fiers are used in fat replacement, properties not considered of consequence in conventional
applications may become increasingly more significant as the amount of fat is reduced. Ad
justments to formulations may need to be made when emulsifying agents are added to replace
fat or extend fat functionality to accommodate the emergence of properties not normally of
concern in full fat products. These properties include interactions with proteins, complexing
of starch, and modification of the crystallization characteristics of other fats in the system [2].
Low Calorie Fats 503
C. Fat Mimetics
Fat mimetics function in fat replacement by simulating the physical attributes of fat. These
types of fat replacers are structurally engineered to form spheroidal particles that will slide
over one another upon hydration. The desirable diameter for reproducing the mouthfeel of fat
is between 0.1 and 3.0 ¡im. Larger particles are too powdery or grainy, and smaller particles
are imperceivable on the tongue [3]. Fat mimetics are effective at calorie reduction because
their protein or carbohydrate natures render them intrinsically lower in calories on a dry
weight basis than the fat they replace. The caloric value of these compounds is reduced further
by dilution when reconstituted with the water required for functional activity. The hydrated
state of fat mimetics lowers the caloric value by one-fourth to one-third of the value of the
dry powder, to 1-1.33 cal/g. Carbohydrate-based mimetics may be used to replace 50-100%
of fat in a product, while protein-based mimetics may replace 75-100% , but neither of these
substitutions can replace fat in a gram-equivalent exchange.
The protein-based fat mimetics currently available have been derived from egg, skim milk,
whey concentrate, or grain proteins. These proteins are subjected to microparticulation or
dénaturation processes to shape spherical particles similar in diameter to natural lipid particles.
The prolific carbohydrate-based fat mimetics include hydrocolloid compounds such as
starches, dextrins, maltodextrins, fibers, and gums. These compounds may also be processed
to further enhance their intrinsic water-holding capacity. It is the hydrating property of these
hydrocolloid particles that slows their clearance from the mouth, allowing the lubricant and
flow characteristics of fat to be effectively duplicated.
Like the emulsifiers in current use, fat mimetics are widely used as fat replacers because
they too are derived from substances with prior histories of safe use in food systems for
purposes other than fat replacement. The carbohydrate-based fat mimetics are the most fre
quently used fat replacers presently on the market. The utility of protein-based fat mimetics
for fat replacement is more limited by the instability of these compounds to high temperatures
and the potential of the protein core components to provoke allergic responses. As with emul
sifiers, fat mimetics must frequently be used in a fat replacer system with other ingredients
before all of the functions of fat can be effectively replaced. Because large amounts of water
are integral to the functional properties of fat mimetics, flavor release may be altered when
these compounds are used, making it also necessary to adjust or add new flavor delivery
systems. The effectiveness of fat mimetics in achieving functional properties of fat may vary
with the particular application, making it unworkable to use the same fat mimetic satisfactorily
in every category of fat-reduced products.
III. STRUCTURAL AND TECHNICAL ASPECTS OF LOW

CALORIE FATS AND FAT SUBSTITUTES
The structural modifications of lipid-based compounds enabling them to function as fat re
placers with lower bioavailability and caloric value than natural lipid compounds take one of
two general approaches. In the first approach, the basic triacylglycerol structure is preserved
but the composition and positional sequence of the fatty acid esters is dictated by requirements
for decreased bioavailability or low metabolic energy value. Alternatively, the basic structures
are synthesized from chemicals unrelated to triacylglycerols but possess the physical and
chemical attributes of natural lipids that allow them to function as lipids in food systems while
resisting the hydrolytic digestive enzymes that make them available metabolically. With either
approach, the end products are lipid-based compounds that possess all of the organoleptic
504 Finley et al.
properties of conventional lipids but deliver fewer calories per gram. Most low calorie fats
and fat substitutes are more accurately described as “families” of compounds related by their
core structural elements. Minor adjustments in the physical or chemical natures of substituents
on this core can expand the number of applications for a particular low calorie fat or fat
substitute by creating different physical forms with a range of temperature or other physico
chemical characteristics.
Low calorie fats are manufactured using the first approach, which modifies the identity and
position of fatty acid substituents on the glycerol backbone of triacylglycerols obtained from
edible fats and oils. Also referred to as structured triacylglycerols, these compounds retain the
conventional properties of natural food lipids but with a bioavailability reduced by about 45%.
During the synthesis of low calorie fats, selected fatty acids are esterified in a positional
sequence on the glycerol backbone of the triacylglycerol defined according to random interes
terification theory and designed to create a lower caloric profile [4]. These fatty acids may be
of short- or medium-chain lengths (Cg-Cio), which are handled metabolically in a manner
more similar to glucose than to fatty acids, thus yielding a lower fuel value and a decreased
propensity for storage as body fat. Saturated fatty acids of longer chain lengths ( > € 15) may
also be used in the synthesis of low calorie fats. These long-chain saturated fatty acids, which
are relatively more poorly absorbed in their free forms, are esterified preferentially at the first
and third hydroxyl groups on the glycerol backbone, where they will be more readily hy
drolyzed by digestive enzymes to their poorly absorbed free forms.
Fat substitutes are synthesized using the second approach, where structures that are unlike
natural triacylglycerols are forged from a variety of chemical compounds and fatty acids,
which are often obtained from edible fats and oils [5]. These synthetic structural analogues of
triacylglycerols may include, but are not exclusive to, compounds where the glycerol back
bone has been replaced with alternative alcohols such as neopentyl alcohol or sucrose polyol
to sterically protect the ester bonds from enzymatic hydrolysis [6,7]; compounds such as
branched triacylglycerols where the fatty acids have been replaced with alternative acids that
also sterically protect the ester bonds [8]; compounds such as retrofats where the ester linkage
has been reversed so that a polyfunctional acid serves as the backbone esterified with long-
chain alcohols [9]; or compounds such as glyceryl ethers where the ester linkage has been
reduced to an ether linkage [10,11]. Although they are more accurately labeled lipid ana
logues, fat substitutes may be loosely considered low calorie fats because they are lipid-
based structures that deliver fewer calories per gram than conventional lipids. Because of their
similarities to low calorie fats, fat substitutes are included here in the discussion of this type
of fat replacer.
The advantages that the use of low calorie fats has over the use of other fat replacers stem
from the authentic triacylglycerol nature of these compounds. Fat substitutes offer the same
technical advantages as low calorie fats, but the synthetic basis of these lipid analogues raises
questions about safety. Both low calorie fats and fat substitutes can perform identically to
conventional fats in food systems and thus are better able to preserve the textural and flavor
attributes imparted by fat. Both compounds also provide thermal stability at the high tempera
tures needed for baking and frying. Low calorie fats and fat substitutes are well suited for fat
replacement in low moisture baking applications such as cookies and crackers. Either type of
lipid-based fat replacer may also be used with equally good results in high moisture food
systems such as ice cream and cheese. Fat replacement in low moisture systems has been
among the most troublesome technical hurdles for food manufacturers to overcome because
the common fat replacers presently used depend on hydration for functional effectiveness.
The fact that low calorie fats are partially digested and absorbed does not permit the degree
of reduction in fat calories that fat substitutes may achieve by their indigestibility and that
nonlipid-based fat replacers may achieve by completely replacing fat calories with calories
from protein or carbohydrate. Fat substitutes are unique among fat replacers in their ability to
completely reduce total fat calories in food systems on a weight-equivalent basis with little or
no additional caloric contributions associated with their use. By being virtually noncaloric, fat
substitutes are able to exceed the fat calorie reduction utility of low calorie fats, emulsifiers,
and fat mimetics. As lipids, they can substitute for equivalent amounts of fat without requiring
adjustments in formulations. The unique ability of fat substitutes to completely reduce the
total amount of fat and its corresponding caloric contribution in a product without compromis
ing product attributes is functionally important, but it raises a number of safety issues. The
metabolic impact from ingesting the lipid analogue structures of fat substitutes is less predict
able and less well understood than that of low calorie fats and other fat replacers that consist
of or have been derived from conventional food lipids or common food ingredients with a
prior history of safe use.
A. Low Calorie Fats

L Caprenin
Developed jointly by the Procter and Gamble Company (Cincinnati, OH) and Grinstead Prod
ucts, Inc. (Kansas City, MO) as a substitute for cocoa butter, caprenin (caprocaprylobehenin)
is a structured triacylglycerol formed by esterification of glycerol with the medium-chain satu
rated fatty acids caprylic acid (Cg) and capric acid (C^q) and the very long chain saturated
fatty acid behenic acid (022)- All of these fatty acids are derived from natural food sources.
Caprylic and capric acids are obtained by fractionation of palmkemel and coconut oils. Be
henic acid is produced from complete hydrogenation of (high erucic) rapeseed oil and is also
found in peanuts and marine oils. Behenic acid is only partially absorbed and metabolized,
whereas caprylic and capric acids are metabolized rapidly for energy, yielding a caloric value
similar to that of carbohydrate. The total caloric value of caprenin is 56% of the caloric value
of conventional fats at approximately 5 cal/g.
Caprenin is limited by its particular fatty acid compositional mix to confectionery applica
tions as a replacement for cocoa butter. There have been few early applications for this fat
replacer because of high initial production costs. Caprenin has been used in combination with
the fat mimetic polydextrose in several commercially available reduced fat chocolate candy
bars. A GRAS self-affirmation petition has been submitted for caprenin for use as a fat re
placer in soft candy and confectionery coatings [12]. Further information is available in Refs.
1 3 -1 6 .
2. M edium -C hain Triacylglycerols

Medium-chain triacylglycerols (MCTs) were originally developed for therapeutic purposes to
provide a source of energy for individuals with compromised gastrointestinal systems. MCTs
have been used extensively in parenteral nutrition formulations. MCT are structured triacyl
glycerols composed of fatty acids with chain lengths between 8 and 10 carbons obtained from
splitting, distilling, and fractionating coconut oil. Caprylic and capric acids comprise more
than 96% of the fatty acids in MCT. The presence of these medium-chain saturated fatty acid
substituents create a triacylglycerol with a low melting point that is liquid at room temperature
and stable to oxidative rancidity. MCTs are readily hydrolyzed by digestive enzymes, and the
fatty acid end products are rapidly absorbed directly into the bloodstream following a post
prandial pattern similar to that observed for glucose [17]. Medium-chain fatty acids are used
as immediate sources of energy by the liver, yielding fewer calories per gram than longer
506 Finley et al.
chain fatty acids. The caloric value of MCT preparations is 7 kcal/g, or 22% lower than that
of the long-chain triacylglycerols present in vegetable oils [18].
3. Salatrim
Salatrim comprises a family of structured triacylglycerols consisting of short- and long-chain
saturated fatty acids randomly distributed on the glycerol backbone. Salatrim is synthesized
using a sodium methoxide-catalyzed interesterification process developed by Nabisco Brands,
Inc. (East Hanover, NJ) that reacts triacetin, tributyrin, and tripropionin with triacylglycerols
of long-chain fatty acids from hydrogenated canola, soybean, or cottonseed oils [19,20]. Un
reacted short-chain fatty acid triacylglycerols are removed from the reaction mix by vacuum
steam deodorization [19]. Each newly formed triacylglycerol contains one short-chain and one
long-chain fatty acid and either a short- or long-chain fatty acid at the third site, in proportions
that can be theoretically predicted from the compositional profile of the starting materials
[21]. Nuclear magnetic resonance imaging is consistent with the chromatographic character
izations of the compounds, confirming that the interesterification reaction is random and that
fatty acids are distributed during the process at attachment sites on the glycerol backbone
predictable by random reaction statistics [19].
A maximum of five variants of triacylglycerols can be produced from this interesterification
reaction encompassing all possible combinations of short- and long-chain fatty acid arrange
ments to be derived from the starting materials [4,21]. Short-chain fatty acids typically make
up 33-66 mol % of the mix of triacylglycerol end products, with the balance made up of
long-chain saturated fatty acids [22]. Stearic and palmitic acids are the dominant long-chain
saturated fatty acids in the mix. The ratio of stearate to palmitate in the final product mix
reflects the proportions of these fatty acids in the original source of hydrogenated oil. Canola
oil has the highest ratio of stearate to palmitate at 17.2 followed by soybean at 5.5 and
cottonseed at 2.8 [22]. Stearic acid is absorbed much less efficiently than palmitic acid, and
thus a higher proportion of stearate would contribute to a lower caloric value for Salatrim.
The ratio of stearic acid to palmitic acid is one of two determinants of the metabolic caloric
value assessed for Salatrim. The other determinant is the intrinsically lower metabolic fuel
value of the short-chain fatty acids, which, at 4 kcal/g, is more consistent with values for
carbohydrate than with the 9 kcal/g value typical for fat. The absorption efficiency for Sala
trim is a function of the molar ratio of short- to long-chain fatty acids, which may from vary
from 0.51 to 1.99. As values approach the upper end of the range, absorptive efficiency
increases from 15% to 70% and the caloric value derived per gram increases correspondingly
from 2.56 to 6.39 kcal. According to both human clinical studies and rat bioassays, the
metabolic caloric value observed for Salatrim ranges from 4.5 to 5.5 kcal/g. Based on these
data, an average caloric value of 5 kcal/g has been assigned to Salatrim for labeling pur
poses [23].
The potential for preparing various combinations of fatty acid compositional mixes and
positional arrangements in Salatrim increases the versatility of this fat replacer, expanding its
uses to a wide range of applications including confectionery coatings, bakery cream fillings,
and milk fat and shortening replacements. The ability to modify the ratio of short-chain to
long-chain fatty acids in the Salatrim mix permits the physical properties of this fat replacer
to be tailored to match those of a number of different conventional food fats [20]. Because
Salatrim is completely saturated, it is highly stable to the effects of oxidation. The initial
focus for Salatrim has been the development of a preparation that closely matches the melting
profile of cocoa butter to create the cooling mouthfeel and smooth melt associated with choco
late. The preparation currently available will facilitate a 45% reduction in calories from fat
and a 25% reduction in calories in chocolate products [24]. Pfizer Food Science Group (New
York, NY) has licensed from Nabisco the exclusive rights for the manufacture and commer
cialization of Salatrim internationally. Nabisco has retained certain rights in North America
associated with their core businesses.
B. Fat Substitutes
1. Carboxyl/C arboxylate Esters
Hydroxycarboxylic acids such as malic or tartaric acids form the backbone of the carboxyl/
carboxylate ester group of fat substitutes. Carboxy, methyl carboxy, carboxylate, or methyl
carboxylate groups extend from the carbon core. Fatty acid esters are formed with the acid
groups, and fatty alcohol esters with the hydroxyl groups. These fat replacers are partially
digested, yielding about 4% of the calories potentially available depending on the particular
fatty alcohol and fatty acid substituents. Physical forms may also vary from oil-like to hard
fats. This compound may be particularly useful as a coconut oil mimetic. Possible applications
for carboxyl/carboxylate esters include butter cream, ice cream, vanilla wafers, and crackers.
The process patent for producing these compounds belongs to Nabisco Brands, Inc. (East
H a n o v e r, N J) [2 5 ,2 6 ].
2. D ialkyl D ihexadecylm alonate

Dialkyl dihexadecylmalonate (DDM) is a fatty alcohol ester of malonic and alkylmalonic acids
patented by Frito-Lay, Inc. for use as a fat substitute in high temperature frying media [27].
This compound is a dimer consisting of a dicarboxylic acid carbon backbone substituted with
hydrogen or C 1- C 20 alkyls and esterified with alkyl, alkenyl, or dienyl groups [28]. Lower
molecular weight forms may be synthesized by reacting a malonyl dihalide with a fatty alco
hol such as oleyl. To produce higher molecular weight forms, an alkyl halide is added in
basic solution. Impurities are removed by vacuum distillation or silica gel chromatography.
The unique structure of DDM is almost completely resistant to digestive enzymes as a result
of steric protection provided for the ester linkages. About 0.1% of the compound consumed
is absorbed. A blend of DDM and soybean oil was used to prepare potato chips and tortilla
chips to be tested by panelists. The products were found to be as crisp as those fried in
conventional vegetable oils but without being as oily. DDM itself contributes no calorie value,
but in a soybean oil blend it yields a fried product reduced in fat by 6 0% and reduced in
calories by 33% [29].
3. Esterified P ropoxylated Glycerol

The propoxylated glycerols are analogues of triacylglycerols in which a propoxyl group has
been inserted between the glycerol backbone and the fatty acid side chain to replace the ester
linkage with an ether linkage. Glycerol is reacted with propylene oxide to form a polyether
glycol and then esterified with fatty acids obtained from any edible fat or oil. Lard, tallow,
soybean, cottonseed, com, and canola oils have all been used in preparing esterified propoxy
lated glycerol (EPG). Digestibility of the molecule decreases with the number of ether link
ages formed. One to three polyoxypropylene groups of various lengths may be added to the
glycerine backbone prior to esterification. Variations in the length of the propoxyl units, in
types of fatty acids, and in the vegetable oil blends in which it may be contained contribute
to the adaptability of this family of compounds to a wide range of applications. EPG is
thermally stable and suitable for use in baking and frying applications as well as in formulated
products. It is currently under development as a noncaloric fat and oil substitute for general
508 Finley et al.
u se in all applications in a joint effort by ARCO Chemical Company and CPC International/
Best Foods [30].
4. P olyorganosiloxanes
Polyorganosiloxanes are chemically inert, linear polymeric structures of silica. They are stable
compounds that maintain their viscosity over a wide range of temperatures. They are also
resistant to oxidation, hydrolysis, and degradation. Polysiloxanes exhibit solubility character
istics similar to those of nonpolar lipids. These compounds are not absorbed and thus do not
contribute any caloric value. Phenyl and methyl substitution of the polysiloxane backbone
forms an oil having a viscosity similar to soybean oil (Dow Coming 550 Fluid, Contour
Chemical, North Reading, MA). The preferred stmcture for this molecule takes the general
formula (CH3)3SiO[(CH 3)2SiO]^Si(CH3)3, where x has an average value between 25 and 500
[31]. Phenylmethylpolysiloxane has been tested as a noncaloric fat substitute in animals [32].
Although safety and toxicity have been established by Dow Coming Corporation (Midland,
MI), anal leakage appears to be a problem when the product is fed to rats at level of 6.5% by
weight [33]. This problem may be corrected by use of antianal leakage agents such as long-
chain saturated fatty acids or cellulose [34].
5. Sorbestrin (Sorbitol Polyester)

Sorbestrin is one of the family of carbohydrate-based fatty acid polyesters that also includes
sucrose polyesters [35]. Sorbitol and sorbitol anhydrides serve as the backbone of this com
pound, which is esterified with fatty acids of varying chain length and degree of saturation.
Sorbitol hexaoleate can be synthesized by heating sorbitol at 180°C with a fivefold excess of
ethyl oleate in dimethylacetamide for 5 h with sodium methoxide as a catalyst [7]. The final
stmcture must have more than four fatty acid esters to provide the steric hindrance needed to
inhibit digestion. Thus polyesters with less than this number of sorbitol units are not useful
fat replacers but function well as emulsifiers. Sorbitol polyesters with less than four fatty acid
esters can be synthesized by enzymatic as well as nonenzymatic methods, but the higher
number of substitutions that have utility as fat substitutes cannot. The functional uses for
sorbitol polyesters as fat substitutes are determined by the composition of the fatty acid esters
on the sugar-alcohol backbone. Sorbestrin is a thermally stable liquid fat substitute suitable
for use in all vegetable oil applications including salad dressing, mayonnaise, baked goods,
and fried foods. It is currently under investigation by Pfizer Food Science Group (New
Y o rk , N Y ).
6. Sucrose P olyester (Olestra)

The process for chemically linking lipophilic fatty acids to hydrophilic sugars was originally
patented in 1952 for preparation of a nonionic surfactant to function as a detergent [35]. The
development of sucrose polyesters currently being tested for use as fat substitutes has evolved
from this technology. These sucrose polyesters may consist of glycosides with between 4 and
14 acetate or hydroxyl groups available for substitution with fatty acids. Olestra is the generic
name assigned to sucrose polyesters with six to eight fatty acid esters, reflecting the identity
of the fatty acid methyl ester of oleic acid used in its synthesis. Soybean oil is currently the
most common source of fatty acids used in preparation of olestra.
The resistance of sucrose polyesters to lipolytic enzymes is increased by the steric hin
drance resulting when the degree of substitution, or the number of fatty acids esterified at the
acetal and hydroxyl sites on the glycoside unit, exceeds three. The types of fatty acids esteri-
fied may also add to an even greater steric hindrance and contribute further to decreased
digestibility of the molecule. The extent to which lipophilic attributes are expressed by sucrose
polyesters is also a function of the degree of substitution or esterification of the molecule.
These functional properties may be more finely tuned by altering the chain length and degree
of saturation of the fatty acid substituents. Olestra is a completely indigestible, noncaloric fat
substitute that exhibits thermal stability at the high temperatures used in baking and frying.
Its physical properties are parallel to those of the partially hydrogenated soybean oil used in
its preparation. Liquids are generated from lightly hydrogenated soy oil, and solids are gener
ated from the more heavily hydrogenated forms of this oil [18].
A number of methods for preparing sucrose polyesters have been reported [35]. Solvent-
free interesterification processes involve admixing long-chain fatty acid methyl esters synthe
sized from edible oils with molten sucrose at temperatures between 110 and 180°C using 1 -
2% sodium metal as a catalyst [36,37]. This method eliminates the need for using toxic
solvents such as dimethylacetamide, dimethylformamide, and dimethylsulfoxide polyesters
that promote formation of the homogeneous melts that increase product yields but that also
make this process unsuitable for preparing food grade polyesters [38].
Olestra has been under investigation by the Procter and Gamble Company (Cincinnati, OH)
for use as a fat substitute since the original patent was filed in 1971 [7]. The first food additive
clearance petition was filed with the FDA in April 1987 for replacement of 35% of fat in
shortening and oils for home use and up to 75% of fats used industrially for deep fat frying
and manufacture of snack foods [39]. The petition has since been modified a number of times.
In August 1990, the focus of the petition was narrowed to restrict the intended use of olestra
to savory snacks for the purpose of accelerating the approval process. In late 1995, an advi
sory panel to the FDA recommended approval for the petition, and an official ruling by the
agency is expected in early 1996.
7. Trialkoxytricarballyate
Trialkoxytricarballyate (TATCA) is a retrofat or a structural analogue of a triacylglycerol in
which the ester linkages have been reversed. In these molecules, the glycerol moiety has been
replaced by thermally stable polycarboxylic acids such as tricarballylic, malonic, or citric
acids, which are esterified with saturated or unsaturated long-chain alcohols instead of fatty
acids [9]. Trialkoxycitrate (TAG) displays viscosity and surface tension properties similar to
those of com oil [40]. Polymorphic behavior was displayed upon melting for both compounds.
Both TATCA and TAG are virtually indigestible oil-like compounds and thus contribute no
caloric value. The potential for thermal decomposition of citrate esters in TAG prevent it from
being used in frying oils. The primary interest in TATCA has been used as a fat replacer for
vegetable oils in formulating fat-free and fat-reduced margarine, mayonnaise, and similar
products. TATCA has been successfully used in the formulation of mayonnaise-type products
and of a fat-reduced margarine that was softer and melted more quickly than regular marga
rine. The patent for this group of compounds is held by CPC International, Division of Best
Foods (Englewood Cliffs, NJ) [9].
IV. UTILITY OF LOW CALORIE FATS IN FAT-REDUCED

FOOD PRODUCTS
The ideal fat replacer has been described as one that is safe, retains all of the organoleptic
and functional properties of fat, substitutes for fat on a one-to-one basis, and contributes
significantly fewer calories than the fat it replaces [2]. It should also be readily available at a
510 Finley et al.
reasonable cost. The quality, and thus consumer acceptance, of food products reduced in fat
is dependent on the degree to which each of the many functions of fat is successfully dupli
cated by the specific fat replacer used. Among the contributions fats make to foods are a
pleasing mouthfeel, flavor and aroma, moisture and lubricity, smoothness and spreadability,
elasticity, thickness, body, heat and structural stability, emulsification, dispersion, aeration,
processing ease, color, and gloss [41,42]. The difficulty of accomplishing all of these tasks
without fat is enormous.
Low calorie fats and fat substitutes are the simplest fat replacement systems because they
allow for substitution of fat in weight-equivalent amounts. Both compounds exhibit the chemi
cal and physical properties of edible fats and oils and express the same functional properties
of conventional lipids in food systems. When low calorie fats and fat substitutes are used for
fat replacement, product formulations do not need to be adjusted to account for the loss of fat
functionality. Fat replacement systems must be used with emulsifiers and fat mimetics because
these nonlipid compounds cannot replace all of the functions of fat if used alone. To satisfy
the need for a desired set of fat functional attributes, fat mimetics and emulsifiers must fre
quently be accompanied by additional ingredients such as water, additional emulsifiers, bulk
ing agents, and flavor enhancers in amounts and proportions that depend on the particular
application. Consequently, weight-equivalent replacement of fat with fat mimetics or emulsi
fiers is not possible.
Use of fat mimetics provides a direct approach to reducing fat and calories because it
involves substitution of fat calories with nonlipid components intrinsically lower in calories.
Structural manipulations of proteins and carbohydrates enable the modified forms of these
compounds to simulate many of the properties of fat but with fewer calories. Mouthfeel can
be reproduced by dispersing spherical particles with diameters of 0.1 -3 .0 ^tm in water as
achieved during microparticulation of proteins [3 ] . This process creates the sensory perception
of an oil-in-water suspension. Fat mouthfeel may also be duplicated by the films formed
when interfacial tension between water and other phases is relieved by compounds such as
methylcellulose [2]. Gelatin-based stabilizers or other thickeners can provide the viscous stim
uli perceived as the creamy or smooth mouthfeel that is usually associated with fat [41]. The
effectiveness of protein- and carbohydrate-based mimetics as fat replacers depends on individ
ual structural features that determine how each will perform in a particular application. Mouth
feel and texture have been successfully duplicated by most of the currently used emulsifiers
and fat mimetics in many applications, but flavor delivery, heat stability, and shelf life are
often compromised [ 1 ].
Preserving the flavor attributes of products reduced in fat is paramount to ensuring a high
degree of quality and acceptability of these products. The large amount of water integral to
the structures of fat mimetics needed to create the mouthfeel of fat can upset the flavor bal
ance, potency, and release in fat-reduced products [43]. The presence of fat in a food system
solubilizes and lowers the vapor pressures of flavor components. Consequently, when the
amount of fat in the system is reduced, the lack of a solvent medium makes the lipid-soluble
flavors stand out in the profile. The increased vapor pressures that also result intensify the
sensory perceptions of these flavors relative to water-soluble flavor components and change
the order in which the flavors are perceived. Flavor imbalances create the off-flavors fre
quently experienced with fat-reduced products compared to their full fat counterparts.
The extent to which the functionality of fat is successfully imitated by a particular fat
replacer may vary considerably from product to product, even among those in the same cate
gory, depending on the set of attributes preferred for the product and the degree of fat replace
ment desired. As with the conventional fats they replace, some fat replacers are preferred
over others in a particular application on the basis of a particular property desired such as
melting point, smoke point, or viscosity. Minor alterations in the core molecule or the substi
tution of fatty acids with different chain lengths and degrees of saturation can be employed to
prepare a family of compounds for low calorie fats and fat substitutes. Oils, semisolids, and
hard fats can then be made available for a particular low calorie fat or fat substitute to support
different fat preferences, thus covering a broader range of applications. With fat mimetics and
emulsifiers, entirely different compounds must often be chosen for different applications, and
the adjustments in formulations needed to accommodate each one will have to vary accord
ingly.
Low calorie fats and fat substitutes may be used for partial or total fat replacement in all
applications, giving them the maximum utility for fat replacement. Of value technically is the
ability of these compounds to be used in low moisture baking systems, which has presented
the most difficult challenge in fat replacement for the food industry. Fat mimetics and emulsi
fiers are prevented from being used in low moisture systems by the hydration required to
attain their functional properties. Of value nutritionally is the ability of low calorie fats and
fat substitutes to replace all of the fat in cooking oils and much of the fat in salad oils, which
account for a large proportion of total fat in the American diet. Use of fat replacers in cooking
oils in particular could contribute considerably to the extent of fat reduction possible in a
number of applications. The flavor attributes and instability at high temperatures exhibited by
many fat mimetics makes these fat replacers inappropriate for either partial or total fat replace
ment in cooking oils.
V» NUTRITIONAL IMPLICATIONS OF LOW CALORIE FATS

Fat replacers can make a significant contribution to improving the healthfulness of the food
supply if their use increases consumer acceptance of fat-reduced products. A number of health
policy groups have recommended that dietary fat intake be decreased to reduce risk for coro
nary heart disease, adult-onset diabetes, and breast, colon, and prostate cancers [44,45].
Moreover, a high fat intake has also been linked to weight gain and obesity, and a reduction
in fat intake has been suggested as a means for facilitating weight control [46,47]. The number
of adults who are considered to be overweight has been increasing over the past 20 years, and
the rise in obesity among children over this same time period has also been considerable
[48,49].
The wider the range of applications that can utilize a particular fat replacer, and the more
closely it is able to duplicate the functional and organoleptic properties of fats, the more
successful this compound will be in helping to improve nutritional health. Low calorie fats
and fat substitutes have more versatility in fat replacement applications than other fat replacers
because they are heat-stable and can be tailored molecularly to yield different physical forms
offering different sets of defined functional characteristics appropriate for a wide range of ap
plications.
A. Blood Cholesterol Levels

One nutritional goal that most fat replacers may achieve equally well is the contribution to a
reduction in total dietary fat intake in amounts that could significantly influence blood choles
terol levels and reduce the associated risk of coronary heart disease [50-53]. One dietary
assessment study has estimated that substitution of fat with a fat replacer in frozen desserts,
processed cheese, and commercial baked goods could reduce total fat intake by more than 8 %
for people who regularly consumed these products [52]. When dairy foods and salad dressings
incorporating the same fat replacer were also included in the assessment, the potential reduc
512 Finley et al.
tion in total fat intake for these people was d o u b le d to approximately 17%. Either of these
magnitudes of reduction in total fat intake could have a considerable impact on closing the
gap between current fat intake and the recommendations made by most health policy groups
for a total fat intake of no more 30% of total calories [44].
Displacement of saturated fatty acids contributed by conventional fats and oils with fatty
acids that do not adversely affect cholesterol levels is one means by which low calorie fats
may exert a positive influence on blood cholesterol levels. In this regard, the impact of low
calorie fats on blood cholesterol would be greatest when used to replace the more highly
saturated fats in food products. The short- and medium-chain fatty acids typically contained
in low calorie fats have not been found to elevate blood cholesterol levels [5 4 ], and the
dominant long-chain saturated fatty acid in Salatrim, stearic acid, is considered to be neutral
in its effects on blood cholesterol levels [55]. The poor absorptive e fficien cy for stearic acid
also lowers its bioavailability to participate in this or any potentially adverse metabolic effect.
Fat substitutes may contribute to reducing dietary fat intake not only through displacement
of fat calories but also through reducing amounts of fat absorbed. Poorly digestible lipid
compounds cause a lipid phase to persist in the small intestines, providing a solvent medium
for lipophiles that transports these compounds through the gastrointestinal tract unabsorbed
[56]. Solubilization of endogenously secreted bile acids does not occur to the extent that bile
acid pool size or bile composition, saturation, and reabsorption are adversely affected [57].
However, the amount of dietary cholesterol absorbed intestinally and the amount of endoge
nously secreted cholesterol reabsorbed enterohepatically in bile are reduced when olestra is
consumed [57]. Ingestion of olestra in amounts consistent with its use as a fat replacer was
associated with lower levels of triacylglycerols and lower levels of both total and LDL choles
terol in human subjects than what would have been expected from caloric reduction alone
[58].
B. Weight Control
Regardless of whether low calorie fats and fat substitutes are able to achieve a physiologically
significant reduction in saturated fat to affect blood cholesterol levels, these compounds could
alternatively produce favorable blood lipid profiles by assisting in weight loss and facilitating
weight control. Caloric reduction is a primary objective in the development of fat-reduced
products and is accomplished by low calorie fats and fat substitutes through decreased or lack
of digestibility. Use of these fat replacers will often achieve a greater reduction in calories
from fat than in total calories per se depending on the contribution originally made by fat to
the total calories in the product. For example, incorporation of Salatrim into a chocolate
coating formulation reduces total calories from fat in the finished product by 45% while reduc
ing total calories by just 25% [24]. Nevertheless, a growing body of evidence suggests that
fat calories may make a greater contribution to body fat weight gain than calories from other
sources [46].
The success with which weight can be controlled by replacing full fat products with their
fat-reduced counterparts will depend on whether compensation for caloric dilution will occur
as a consequence of ingesting these products. It will also depend on whether the use of fat
replacers in fat-reduced products results in a caloric dilution of a significant enough magnitude
to trigger compensatory responses. If a reduction in total energy intake does occur when foods
reduced in fat are ingested, individuals may compensate for this energy loss to different de
grees [59-62]. Ingestion of foods containing fat replacers by healthy adults over a 12 week
period produced a weight loss similar to that observed with ingestion of foods naturally low
in fat [63]. The fact that the control group in this study gained a small but not statistically
significant amount of weight suggests that caloric compensation did not occur among individu
als who reduced their fat intake by either method. Only minimal caloric compensation was
observed among adults consuming fat-reduced products containing olestra over a 20 day pe
riod [64]. Inconsistencies found in these studies between the weight loss observed and that
predicted on the basis of the expected caloric reduction may have resulted as much from
the inaccuracies of measuring caloric intake using current dietary methods as from caloric
compensation and thus should not be interpreted to suggest unequivocally that caloric compen
sation had occurred [47].
The probability of compensation for caloric dilution from ingesting foods containing fat
replacers may depend on age, sex, degree of obesity, and motivational factors. Children and
lean adult men may be more likely to compensate for calories displaced by fat replacers than
obese adult men or women [65,66]. When compensation for energy dilution occurs, it appears
to be specific for energy and is not selective for a particular macronutrient such as fat [67,68].
Energy is the primary factor driving appetite among children to order to ensure an adequate
caloric intake to support growth [65]. This essential drive for caloric density may be mani
fested by children as a preference for fat [66]. For some adults, psychological factors may also
be a powerful motivation to compensate for caloric dilution [52,67,69]. Studies examining the
effects of fat replacers on caloric intake have not satisfactorily taken into account whether the
knowledge that fat-reduced foods were being consumed may have influenced the decision to
eat more of these foods or may have provided the discipline needed to consume less.
C. other Nutrient Intakes

Consumption of fat-reduced products will assist consumers in attaining a healthier dietary
intake only if the use of these products promotes a lower fat intake, and the corresponding
decrease in calories, without compromising nutritional health in other ways. Fat is important
nutritionally as a source of linoleic acid and other long-chain unsaturated fatty acids. It also
serves as a solvent medium that assists in the conveyance of fat-soluble vitamins to intestinal
absorptive surfaces and facilitates their uptake by mucosal cells. Long-chain unsaturated fatty
acids have important physiological roles as structural constituents of cell membranes and as
precursors of functional compounds regulating cellular reactions.
At the present time, the common uses of fat replacers result most frequently in displace
ment of vegetable oils [70]. These oils are the primary sources of unsaturated fatty acids in
the diet, including both the essential o)-6 fatty acid linoleic acid and the important oj-3 fatty
acid linolenic acid. The requirement for linoleic acid can be met with intakes as low as 1 -
3% of total calories (2-7 g/day) [71]. This level of linoleic acid can usually be ingested if
vegetable oils account for a minimum of 10% of total caloric intake. The applications most
frequently using low calorie fats and fat substitutes will involve food categories such as baked
goods, confectioneries, and salty snacks, which are not considered to be important contribu
tors of essential amounts of fat to the diet. Use of fat substitutes and low calorie fats to
replace all or most of the fat in salad and cooking oils could make it more difficult to achieve
the required level of linoleic acid intake. Replacement of soybean oil in particular would
remove a principal source of linolenic acid from the diet.
Fat-soluble vitamins depend on lipids in foods to assist in their transport and absorption.
The presence of fat also stimulates the secretion of bile acids required for the uptake of these
vitamins by mucosal cells. Displacement of fat by low calorie fats is unlikely to interfere with
this absorptive process because only small amounts of fat are required to satisfy this need. A
reduction in total fat intake to below the critical amount of 10 % of calories is not likely to
occur with consumption of low calorie fats in all their potential applications. As partially
absorbed lipids, low calorie fats should function in the transport and absorption of fat-soluble
vitamins in the same manner as conventional food lipids. Feeding Salatrim to minipigs for 4
514 Finley et al.
weeks at levels equivalent to 3, 6, and 10 % of the diet did not significantly affect serum
levels of vitamin A or vitamin E compared with controls fed 10 % com oil [72]. Prothrombin
time was also unaffected in these animals, indicating that vitamin K status was unimpaired.
In addition, concentrations of calcium and phosphoms in bone were similar between Salatrim-
fed minipigs and controls, indicating that vitamin D status was also unimpaired.
Use of fat substitutes that are virtually indigestible raises other issues regarding the absorp
tion of fat-soluble vitamins. Theoretically, poorly absorbed or completely unabsorbed fat sub
stitutes could solubilize fat-soluble vitamins, keeping some fraction of these vitamins in solu
tion where they would be eliminated in the feces along with the undigested fat substitute [57].
Ingestion of olestra by human subjects in amounts of 17-27 g/day over a 16 week period was
not found to adversely affect semm levels of vitamins A or D or prothrombin time, but semm
vitamin E levels were notably decreased [58]. Concern about the possible depletion of vitamin
E reserves over time has generated a recommendation that olestra preparations be supple
mented with this vitamin. Because vegetable oils are the primary contributor of vitamin E in
the diet, providing approximately one-third of the total intake of this vitamin, use of olestra
as a fat replacer in salad oils may not be appropriate [73].
D. Adjustment to Low Fat Diets

One of the most significant contributions that low calorie fats and fat substitutes may make
toward improving health is in assisting consumers to become accustomed to low fat diets.
The preference for fat is a powerful response regulated by both sensory and postingestive
factors [70,74]. It may be possible to dissociate the sensory aspects controlling fat intake from
the metabolic properties and to develop a preference for low fat foods if the organoleptic
qualities of fats are successfully duplicated by fat replacers [69,74]. Low calorie fats and fat
substitutes may make an additional contribution to helping consumers achieve better adher
ence to low fat diets over the long term by enabling them to maintain their usual food selection
patterns [70,75]. Products containing fat replacers were consumed by adult women participat
ing in the Women’s Health Trial who were able to maintain a 20% fat intake over a period
of several years [76]. Much more remains to be learned about the relative contributions made
by fat taste and fat-energy interactions in acquiring a preference for fat before the contribution
that fat replacers may make in helping consumers maintain a low fat intake can be fully
understood [69].
VI. REGULATORY ASPECTS OF LOW CALORIE FATS

As food ingredients, low calorie fats and fat substitutes are classified as food additives and
therefore must receive regulatory approval before they may be used in fat-reduced foods. The
approach to obtaining this approval will vary depending upon whether the fat replacer was
developed as a result of creating a new molecule, applying new technology to make an old
molecule serve new purposes, or expanding old technologies to apply to old molecules for the
first time [77]. Fat replacers derived from traditional food sources are also handled differently
from fat replacers produced through chemical synthesis or derived from novel food sources
such as jojoba oil or the membrane lipid phytanyl phosphatidylglycerol phosphate produced
by Halibacterium halohium [31,70,78].
A. Safety
The creation of a new molecule by chemical synthesis was the method used to develop olestra
and the other fat substitutes. Regulatory approval for these new molecules requires the submis
Low Calorìe Fats 515
sion of a food additive clearance petition using guidelines for safety testing as outlined in the
FDA monograph commonly referred to as the “Red Book” [79]. Fat replacers developed by
either old or new technologies involving traditional molecules with prior histories of safe use,
or by chemical synthesis from these traditional molecules, need only be shown to be safe in
their new forms and/or applications. These methods are generally used to develop low calorie
fats as well as emulsifiers and fat mimetics. For these types of fat replacers, a GRAS (gener
ally regarded as safe) affirmation petition must be submitted for approval. Whether regulatory
approval is sought through a food additive clearance petition or a GRAS affirmation petition,
the burden of proof of safety is on the manufacturer [77,78].
The procedures for safety testing outlined in the first version of the “Red Book” did not
address the additional considerations raised in testing additives such as fat replacers that are
added to foods in substantially larger amounts than has been typical for food ingredients in
the past. The standards established for determining the amount of preclinical and clinical
testing needed for microadditives such as preservatives, colorants, and flavorings upon which
the original guidelines were based cannot be interpreted in a practical manner for macroaddi
tive fat replacers. Consequently, traditional safety concerns and standard toxic endpoints are
difficult to apply [78].
The amount of clinical testing required to assess food additive safety is based on guidelines
found in the most recent version of the Red Book in the section “Concern Levels as Related
to Chemical Structure and Exposure” [80]. If these numbers are used, macroadditives would
be placed at the highest concern level for safety at almost any appreciable exposure, which
would almost always require extensive human testing to establish the safety of fat replacers
[81]. The extensive testing triggered by the relatively high average exposure levels typically
found for these ingredients compared with microadditives unfairly imposes a greater burden
on manufacturers of macroadditives [81]. Although traditional evaluation of food additive
safety does not allow for risk/benefit assessment, this approach may be more appropriate for
macroadditives [78].
B. Labeling
1. Ingredients Listing and N utrition L abel Inform ation
Food products formulated with low calorie fats or fat substitutes must carry information on
ingredients and nutritional content on the package label in compliance with current labeling
regulations [82]. Ingredients must appear in the list of ingredients according to their contribu
tions by weight to the final product. If the fat replacer also contributes caloric value, it must
be included in the calculations of total fat, fat calories, and total calories used to provide
nutrient information for the label using values supported by experimental data provided by the
manufacturer. The preferred method for estimating caloric value of a new ingredient is to
establish an Atwater factor using data on weight gain obtained from animal bioassays.
2. N utrient C ontent Claims and H ealth Claims

The presence of low calorie fats and fat substitutes in a packaged food product may qualify
the product for a nutrient content claim. Many of these products may qualify for a health
claim as well [78]. Nutrient content claims allow use of descriptors such as “fat-free,” “low
fat,” “reduced fat,” and “light” or “lite” to appear on the label, but the product must meet a
specified level of fat reduction per reference amount and label serving size relative to the fat
content of the same type of product or that of a standard value for the product. The “reduced
fat” descriptor allows claims on products with at least a 25% reduction in fat, which can be
516 Finley et at.
readily achieved with low calorie fats and fat substitutes. To qualify for “light” claims, the
food must be reduced in fat by 50% if the comparative product is high in fat (more than 50%
of calories) or it must be reduced in calories by 33% or in fat by at least 50% if the compara
tive product is not high in fat.
The “low fat” definition can be met if the product contains no more than 3 g of fat per
reference amount. “Low saturated fat” requires that no more than 1 g of saturated fat be
present in the reference amount and no more than 15% total calories come from saturated fat.
The rigorous definition of “fat free,” i.e., trivial or negligible amounts of fat and no added
ingredient that is fat, may allow only foods containing nonlipid fat mimetics to qualify. Health
claims regarding cancer and coronary heart disease may also be made for products formulated
with low calorie fats and fat substitutes as long as these products meet the definition for low
fat. Additional qualifiers for saturated fat content or fiber content may also be needed de
pending upon the particular health claim being made.
VII. FUTURE TRENDS IN DEVELOPMENT OF LOW

CALORIE FATS
Continued investment in the development of new low calorie fats and fat substitutes in the
future will depend on whether these new ingredients are embraced or dismissed by consumers
and health policymakers. In spite of how these ingredients may be received by consumers and
health policymakers, incentives for developing new fat replacers will still be limited by the
willingness of the regulators to modify traditional food ingredient safety testing to accommo
date macroadditives. The basis of consumer acceptance will evolve from hedonic appeal of
fat-reduced products and attainment of concrete health benefits such as controlling weight by
consuming these products. Acceptance by health policymakers will be forthcoming if results
of studies convincingly and consistently demonstrate the effectiveness of these products in
promoting specific health outcomes without compromising nutritional health. Changes in regu
latory barriers are the most difficult to predict, but proactive pressure applied to regulators by
consumer groups and health policymakers can assist manufacturers in promoting a more favor
able regulatory climate for obtaining approval for macroadditives.
Future innovations used in the synthesis of low calorie fats and fat substitutes must be
managed in ways that will minimize toxicological concerns. Fat replacers derived from novel
food sources may receive more attention in the future. Biotechnology may also play an in
creasingly important role in generating low calorie fats and fat substitutes from natural
sources. For example, genetic engineering may allow bacterial synthesis of sucrose polyesters
with degrees of substitution greater than 4, which is currently impossible because steric hin
drance interferes with this ability [35]. Fat substitutes produced through biotechnology would
not present the toxicological concerns associated with the presence of chemical impurities that
are frequently introduced during industrial synthesis.
The demand for fat-reduced products should continue to grow well into the next decade as
the aging population begins to increase disproportionately in numbers. Health concerns will
remain or increase in importance as a dominant motivating factor in selection of foods.
Weight control will in all likelihood be the overriding health concern. The availability of
foods reduced in fat and calories but hedonically comparable to full fat foods will allow the
need for indulgence to be satisfied while fulfilling the drive toward a more healthful diet. The
use of low calorie fats and fat substitutes that successfully duplicate the organoleptic and
functional attributes of fat with fewer calories should give fat-reduced food products an even
stronger competitive edge in the marketplace.
REFERENCES
1. J. D. Gussow, Are we really fixing up the food supply?, J. A m . Diet. Assoc. 93: 1300-1304
(1993).
2. Calorie Control Council, Fat Reduction in Foods, 1994.
3. G. E. Gaull, Role of microparticulated protein fat substitutes in food and nutrition, Ann. N .Y .
A cad. Sci. 623: 350-355 (1991).
4. L. P. Klemann, J. W. Finley, and G. A. Leveille, Estimation of the absorption coefficient of
stearic acid in SALATRIM fats, J. Agric. Food Chem. 42: 484-488 (1994).
5. D. J. Hamm, Preparation and evaluation of trialkoxytricarballylate, trialkoxycitrate, trialkoxyglyc-
eryl ether, jojoba oil, and sucrose polyester as low calories replacements of edible fats and oils,
J. F o o d Sci. 49: 419-428 (1984).
6 . A. Minich, Dietetic compositions, U.S. Patent 2,962,419 (1960).
7. F. H. Mattson, M. Healthy, and R. A. Volpenhein, Low calorie fat-containing compositions, U.S.
Patent 3,600,186 (1971).
8 . D. D. Whyte, Triglyceride esters of branched carboxylic acids, U.S. Patent 3,579,548 (1971).
9. D. J. Hamm, Low calorie edible oil substitutes, U.S. Patent 4,508,746 (1985).
10. V. W. Trost, Low calorie fat substitutes, Can. Patent 1,106,681 (1981).
11. H. K. Mangold, Biological effects and biomedical applications of alkoxylipids, in Ether Lipids,
Chemistry, and Biology, 1972, p. 154.
12. A. M. Thayer, Food additives, Chem. Eng. News 70: 26 (1992).
13. The Proctor and Gamble Co., Cincinnati, Ohio, GRAS Affirmation Petition: Caprenin, to Foods
and Drug Administration, Washington, DC (4.1.1991).
14. K. D. Fisher (Life Science Research Office, Federation of American Societies for Experimental
Biology, Bethesda, Maryland): Evaluation of the Health Aspects of Caprenin; prepared for P&G
(Jan. 1991).
15. D. R. Webb and R. A. Sanders. Caprenin I. Digestion and rearrangement in thoracic diet cannu-
lated rats, J. Am. Coll. Toxicol. 10: 325 (1991).
16. D. R. Webb, J. C. Peters, R. J. Jandacek, and N. E. Fortier. Caprenin 2. Short-term safety and
metabolism in rats and hamsters, J. Am. Coll. Toxicol. 10: 341 (1991).
17. V. K. Babayan, Modification of food to control fat intake, J. Am. Oil Chem. Soc. 51: 260-
264 (1974).
18. C. E. Stauffer, Fats that cut calories. Baking Snack, May 1995, pp. 44-49.
19. L. P. Klemann, K. Aji, M. M. Chrysam, R. P. D ’Amelia, J. M. Henderson, A. S. Huang,
M. S. Otterbum, R. G. Yarger, Random nature of triacylglycerols produced by the catalyzed
interesterification of short- and long-chain fatty acid triglycerides, J. Agric. Food Chem. 42: 442-
446 (1994).
20. R. E. Smith, J. W. Finley, G. Leveille, Overview of SALATRIM, a family of low-calorie fats,
J. Agric. Food Chem 42: 432-434 (1994).
21. B. J. Softly, A. S. Huang, J. W. Finley, M., Petersheim, R. G. Yarger, M. M. Chrysam,
R. L. Wieczorek, M. S. Otterbum, A. Manz, G. J. Templeman, J. Agric. Food Chem. 42: 461-
467 (1994).
22. J. M. Talbot, S. A. Anderson, and K. D. Fisher, The evaluation of the health aspects of using
certain triacylglycerols as food ingredients. Life Sciences Research Office, Federation of American
Societies for Experimental Biology, Rockville, MD, 1993.
23. J. W. Finley, L. P. Klemann, M. S. Oterbum, and C. G. Walchak, Caloric availability of SALA
TRIM in rats and humans, J. Agric. Food Chem. 42: 495-499 (1994).
24. J. Stowell, Salatrim—a versatile family of reduced-calorie fats. Paper prepared for Die Ernahrungs
Industrie, 1995.
25. L. P. Klemann and J. W. Finley, Low calorie fat mimetics comprising carboxy/carboxylate esters,
Eur. Patent 303,503 (1989).
26. L. P. Klemann and J. W. Finley, Low calorie fat mimetics comprising carboxy/carboxylate esters,
Int. Patent 89/01293 (1989).
518 Finley et al.
27. M. E. Spearman and J. G. Fulcher, Malonate Esters: New Synthetic Fat Substitutes for Food Use.
Annual Meeting of the Oil Chemists Society, Cincinnati, Ohio, May 3 -7 , (1989).
28. J. Fulcher, Synthetic cooking oils containing dicarboxylic acid esters, U.S. Patent 4,582,927
(1986).
29. J. D. Dziezak, Fats, oils, and fat substitutes. Food Technol. 43(7): 66-74 (1989).
30. Charles F. Cooper, Reduced calorie fat substitute, U.S. Patent 528888 4A.
31. W. E. Artz, and S. L. Hansen, Other fat substitutes, in Carbohydrate Polyesters as Fat Substitutes
(C. C. Akoh and B. G. Swanson, Eds.), Marcel Dekker, New York, 1994, pp. 197-236.
32. E. F. Bracco, N. Baba, and S. A. Hashim, Polysiloxane: potential noncaloric fat substitute; effects
on body composition of obese Zucker rats. Am. J. Clin. Nutr. 46: 784-789 (1987).
33. J. W. Ryan, Prevention of anal leakage of polyorganosiloxane fluids used as fat substitutes in
foods, Eur. Patent 368,534 (1989).
34. J. W. Ryan, Food composition containing a siloxane polymer and a particulate silica, U.S. Patent
4,925,692 (1990).
35. C. C. Akoh, Synthesis of carbohydrate fatty acid polyesters, in Carbohydrate Polyesters as
Fat Substitutes (C. C. Akoh and B. G. Swanson, Eds.), Marcel Dekker, New York, 1994, pp.
9-35.
36. G. P. Rizzi and H. M. Taylor, A solvent-free synthesis of sucrose polyesters, J. Am. Oil Chem.
Soc. 55: 398-401 (1978).
37. C. C. Akoh and B. G. Swanson, Optimization of sucrose polyester synthesis: comparison of
sucrose polyesters, raffinanose polyesters and salad oils, J. Food Sci. 55: 236-243 (1990).
38. R. O. Fuege, H. J. Zeringue, Jr., T. J. Weiss, and M. Brown, Preparation of sucrose esters by
interesterification, / . Am. Oil Chem. Soc. 47: 56-60 (1970).
39. Procter and Gamble, Olestra as a calorie free substitute for fats and oils. Petition submitted to the
FDA, 1987.
40. S. D. Huhn, P. S. Given, Jr., and L. P. Klemann, Ether bridged polyesters and food compositions
containing ether bridged polyesters, U.S. Patent 4,888,195 (1989).
41. A. Drewnowski, Sensory properties of fats and fat replacements, Nutr. Rev. 50: 17-20 (1992).
42. L. Schinhofen, Fat substitutes. Fat and Cholesterol Issues Monitor, March 1995, pp. 16-26.
43. J. P. Schirle-Keller, G. A. Reineccius, and L. C. Hatchwell, Flavor interactions with fat replacers:
effect of oil level, J. Food Sci. 59: 813-875 (1994).
44. National Research Committee (U.S.), Committee on Diet and Health, Fats and other lipids, in
Diet and Health: Implications for Reducing Chronic Disease Risk/Committee on Diet and Health,
Food and Nutrition Board, Commission on Life Sciences, National Research Council, National
Academy Press, Washington, DC, 1989, pp. 205-231.
45. J. A. Marshall, R. F. Hamman, and J. Baxter, High-fat, low-carbohydrate diet and the etiology
of non-insulin-dependent diabetes mellitus. Am. J. Epidemiol. 134: 590-603 (1991).
46. J. P. Flatt, Use and storage of carbohydrate and fat. Am. J. Clin. Nutr. 51: 952S-959S (1995).
47. A. Kendall, D. A. Levitsky, B. J. Strupp, and L. Lissner. Weight loss on a low-fat diet: conse
quence of the imprecision of the control of food intake in humans. Am. J. Clin. Nutr. 53: 1124-
1129 (1991).
48. S. L. Gortmaker, W. H. Dietz, A. M. Sobol, and C. A. Wehler, Increasing pediatric obesity in
the United States, Am. J. Dis. Child. 141: 535-540 (1987).
49. National Institutes of Health, Consensus Development Panel on the Implications of Obesity, Health
implications of obesity, Ann. Intern. Med. 103: 1073-1077 (1985).
50. D. M. Hegsted, R. B. McGandy, M. L. Meyers, and F. J. Stare, Quantitative effects of dietary
fat on serum cholesterol in man. Am J. Clin. Nutr. 17: 281-295 (1965).
5 1. A. Keyes, J. T. Anderson, and F. Grande, Serum cholesterol response to changes in the diet. IV.
Particular saturated fatty acids in the diet. Metabolism 14: 776-787 (1965).
52. B. J. Lyle, K. E. McMahon, and P. A. Kruetler, Assessing the potential dietary impact of replac
ing dietary fat with other macronutrients, J. Nutr. 122: 211-216 (1992).
53 . National Heart, Lung, and Blood Institute, Consensus Development Conference, Lowering blood
cholesterol to prevent heart disease, JAMA 253: 2080-2086 (1985).
54. S. M. Grundy and M. A. Denke, Dietary influences on serum lipids and lipoproteins, J. Lipid
R es. 31: 114 9 - 117 2 (1990).
55. S. M. Grundy, Influence of stearic acid on cholesterol metabolism relative to other long-chain
fatty acids. Am. J. Clin. Nutr. 60 (Suppl.): 986S-990S (1994).
56. C. M. Bergholz, Olestra and the potential role of a nonabsorbable lipid in the diets of children,
A nn. N .Y . A cad. Sci. 623: 356-368 (1991).
57. R. J. Jandacek, The effect of nonabsorbable lipids on the intestinal absorption of lipophiles. D rug
M etab. R ev. 13: 695-714 (1982).
58. M. J. Mellies, C. Vitale, R. J. Jandacek, G. E. Lampkin, and C. J. Glueck, The substitution of
sucrose polyester for dietary fat in obese, hypercholesterolemic outpatients. Am. J. Clin. Nutr. 41:
1 - 1 2 (1985).
59. L. Lissner, D. A. Levitsky, B. J. Strupp, H. J. Kalkwarf, and D. A. Roe, Dietary fat and the
regulation of energy intake in human subjects. Am. J. Clin. Nutr. 46: 886-892 (1987).
60. K. P. Porikos, G. Booth, and T. B. Van Italic, Effect of covert nutritive dilution on the spontane
ous food intake of obese individuals: a pilot study. Am. J. Clin. Nutr. 30: 1638-1644 (1977).
61. K. P. Porikos, M. F. Hesser, and T. B. Van Italic, Caloric regulation in normal-weight men
maintained on a palatable diet of conventional foods. Physiol. Behav. 29: 293-300 (1982).
62. R. D. Mattes, C. B. Pierce, and M. I. Friedman, Daily caloric intake of normal-weight adults:
response to changes in energy density of a luncheon meal. Am. J. Clin. Nutr. 48: 214-219 (1988).
63. R. D. Mattes, Fat preference and adherence to a reduced-fat diet. Am. J. Clin. Nutr. 57: 373-
381 (1993).
64. C. J. Glueck, M. M. Hastings, C. Allen, E. Hogg, L. Baehler, P. S. Gartside, D. Phillips, M.
Jones, E. J. Hollenbach, A. S. Braun, and J. V. Anastasia, Sucrose polyester and overt caloric
dilution. Am. J. Clin. Nutr. 35: 1352-1359 (1982).
65. L. L. Birch, S. L. Johnson, M. B. Jones, and J. C. Peters, Effects of a nonenergy fat substitute
on children’s energy and macronutrient intake. Am. J. Clin. Nutr. 58: 326-333 (1993).
66. L. L. Birch, Children’s preferences for high fat foods, Nutr. R ev. 50: 283-290 (1992).
66 . R. G. Campbell, S. A. Hashim, and T. B. Van Italic, Responses to variation in nutritive density
in lean and obese subjects, N. Engl. J. Med. 285: 1402-1407 (1971).
67. R. W. Foltin, M. W. Fischman, T. H. Moran, B. J. Rolls, and T. H. Kelly, Caloric compensation
for lunches varying in fat and carbohydrate content by humans in a residential laboratory. Am J.
Clin. N utr. 52: 969-980 (1990).
68 . B. Rolls, P. A. Pirraglia, M. B. Jones, and J. C. Peters, Effects of olestra, a noncaloric fat
substitute, on daily energy and fat intake in lean men. Am. J. Clin. Nutr. 56: 84-92 (1992).
69. B. Rolls, Changing the preference for fat in foods, Nutr. Rev. 52: 21-23 (1993).
70. D. J. Mela, Nutritional implications of fat substitutes, J. Am. Diet. Assoc. 92: 472-476 (1992).
71. Committee on Dietary Allowances, Food and Nutrition Board, National Research Council, Recom
mended Dietary Allowances, 10th ed.. National Academy Press, Washington, DC, 1989, pp.
4 4 -51.
72. J. R. Hayes, N. H. Wilson, M. C. Roblin, P. C. Mann, and A. L. Kiorpes, 28-day continuous
dosing study in minipigs with a SALATRIM structured triacylglycerol composed of stearate, ace
tate, and propionate, J. Agric. Food Chem. 42: 563-571 (1994).
73. S. P. Murphy, A. F. Subar, and G. Block, Vitamin E intakes and sources in the United States.
A m . J. C lin. N utr. 52: 361-367 (1990).
74. B. J. Rolls, and D. J. Shide, The influence of dietary fat on food intake and body weight, Nutr.
R ev. 50: 283-290 (1992).
75. American Dietetic Association, Position of the American Dietetic Association: fat replacements,
J. Am. Diet. Assoc. 91: 1285-1288 (1991).
76. A. R. Kristal, E. White, A. L. Shattuck, S. Curry, G. Anderson, A. Fowler, and N. Urban,
Long-term maintenance of a low-fat diet: durability of fat-related dietary habits in the Women’s
Health Trial, J. A m . D iet. A ssoc. 92: 553-559 (1992).
77 . C. J. Glueck, P. A. Streicher, E. K. Illig, K. D. Weber, Dietary fat substitutes, Nutr. Res. 14:
16 0 5-16 19 (1994).
520 Finley et a l
78. W. H. Glinsmann and J. E. Vanderveen, Fat substitutes: a regulatory perspective, Annu. Rev.
N utr. 12: 473-487 (1992).
79. Food and Drug Administration, Toxicological principles for the safety assessment of direct food
additives and color additives used in foods. Bureau of Foods, Washington, DC, 1982.
80. Food and Drug Administration, Toxicological Principles for the Safety Assessment of Direct Food
Additives and Color Additives Used in Foods, 2nd ed.. Bureau of Foods, Washington, DC, 1995.
81. W. H. Glinsmann, Commentary— focus on substitutes that alter nutrition and gastrointestinal phys
iology, ILSI Workshop on Safety and Regulatory Aspects of Macronutrient Substitutes, 1994.
82. Food and Drug Administration, Nutrition labeling of food. Code of Federal Regulations 21.101.9,
Office of the Federal Register, National Archives and Records Administration, Washington, DC,
1990.
20__________________
Food E m u lsifie rs
Niels Krog
Danisco Ingredients, Brabrand, Denmark
I. INTRODUCTION
Modem industrial food production requires surface-active lipids (emulsifiers, surfactants) as
processing aids to facilitate uniform quality and ensure long shelf life of the finished products.
Foods are very complex colloidal systems that may undergo changes during storage, resulting
in deterioration of appearance and texture and loss of flavor characteristics. Food formulations
and composition in terms of calories, fat content, vitamins, or minerals change all the time
due to new insights into the effect of eating patterns or health aspects. This leads to a constant
demand for optimization of product formulations, including optimal selection of raw materi
als. A range of food grade emulsifiers based on edible fats and oils are available. They possess
many specific applications in foods.
Food emulsifiers are polar lipids needed to increase colloidal stability and provide interfa
cial interactions between food components such as lipids, proteins, and carbohydrates. Such
interactions are important factors in obtaining emulsion stability, foam formation of whipped
products, and increased shelf life in many foods.
Food emulsifiers are esters of edible fatty acids and polyvalent alcohols such as glycerol,
propylene glycol, sorbitol (sorbitan), and sucrose. These esters may be further esterified with
organic acids such as acetic acid, lactic acid, tartaric acid, succinic acid, or citric acid to
modify their hydrophilic properties and functional effects.
The chemistry of food emulsifiers is described in detail by Lauridsen [1], Schuster [2],
Benson [3], and Krog [4].
The total world production of food emulsifiers is not known exactly due to lack of statisti
cal information from many countries. However, it is estimated that the amount of food grade
emulsifiers produced is about 250,000 metric tons. Mono- and diglycerides, including distilled
monoglycerides and their organic acid esters, amount to approximately 75% of the total pro
duction [5], making monoglycerides the most important product from both an economic and
a functional point of view.
521
522 Krog
All food emulsifiers are regulated by the health authorities in each country. Emulsifiers for
foods are evaluated by the Joint Expert Committee on Food Additives (JECFA) of the Food
and Agriculture Organization/World Health Organization (FAO/WHO). In Europe the legisla
tion is administered by the Commission of the European Communities’ Scientific Committee
for Food, and in the United States by a department under the Food and Drug Administration
(FDA). Data from toxicological studies are used by FAO/WHO’s Codex Alimentarius Com
mission to issue recommendations to national health authorities concerning the use of emulsi
fiers in foods. A compendium of food additive specifications was issued by JECFA in 1992
[6], and it is currently being updated.
II. MONO-DIGLYCERIDES, EU 471, US/FDA/CFR §184.1505

Industrial production of monoglycerides began in the 1930s, when their use in margarine
production started [7]. Mono-diglycerides is a common term describing a mixture of mono-
and diglycerides, which is produced by glycerolysis [8] of fats as shown schematically in Fig.
1 or by esterification of glycerol with fatty acids [9]. Depending on the glycerol/fat ratio in
the reaction blend, the monoglyceride content in the equilibrium mixture obtained after glycer
olysis may vary from 10% to 60%. Commercial mono-diglycerides usually contain 45-55%
monoglyceride, 38-45% diglyceride, and 8-12% triglyceride.
The main applications of mono-diglycerides in foods are typically in fat-based products,
such as margarine, spreads and bakery fats (shortenings), and cake mixes. Mono-diglycerides
are added by way of the fat phase, often in combination with other more hydrophilic emulsi
fiers. In dairy emulsions, mono-diglycerides are used in ice cream and recombined milk in
combination with hydrocolloids (see Chapters 11-13).
A. Distilled Monoglycerides
Monoglycerides are separated from the di- and triglycerides by a process referred to as molec
ular distillation, a thin-film, high vacuum technique [1]. The total monoglyceride content of
C H /O C O R ’
I
CHO CO R'
I
CHÔCOR"
C K O -C O R ’ CHÔCOR'
1
CHOH CHOCOR'
I I
CHpOCOR" CH,OH
1,3 - Diglycerides 1,2 - Diglycerides

CH^-OCOR’ CH,OH
1
CHOCO-R" CHOH CH^-OCOR^ CHÔH
I
CHÔCO-R" CH,OH CHOH CHOCOR’
!
CH,OH CHjOH
Triglycerides + Glycerol
1 - Monoglycerides 2 - Monoglycerides
CHpOH
I
CHOH
t
CH,OH
V
R\ R^ and R^ = Fatty acids
Fig. 1 Reaction scheme for production of mono-diglycerides by glycerolosis.

Food Emulsifiers 523
distilled products is 93-97% , with the 1-monoester content being at least 90%. An alternative
monoglyceride production method is to esterify fatty acids with glycerol, followed by molecu
lar distillation. High purity distilled monoglycerides with a single fatty acid composition, for
example glycerol monopalmitate or glycerol monooleate, can be manufactured using this
process.
B. Physical Properties
Monoglycerides always have higher melting points than their corresponding triglycerides [10].
With distilled monoglycerides the increase in melting point can be as much as 10°C compared
to the triglycerides used in the glycerolysis process.
Monoglycerides are polymorphic like triglycerides [11] and solidify from melt in a metasta
ble a crystal form, which transforms to a sub-o: form on further cooling. Both a forms are
metastable and transform into a stable high melting jS crystal form when stored. Typical
melting points of some distilled monoglycerides with different fatty acid compositions are
shown in Table 1.
Distilled monoglycerides have better dispersibility in water than mono-diglycerides, be
cause distilled monoglycerides form liquid crystalline mesomorphic phases [4], whereas
mono-diglycerides in most cases form emulsions due to a relatively high content of triglycer
ides. The mesomorphic behavior of distilled monoglycerides in water is caused by their purity
and well-defined molecular structure. This allows distilled monoglycerides to form ordered
structures in water with bilayers of the fatty acid chains separated by water layers associated
with the polar groups [4]. A schematic structure of the lamellar phase is shown in Fig. 2. The
thickness of the lipid bilayers and that of the water layers (J^), as well as the specific
surface area of the polar group {S) can be measured by X-ray diffraction analysis, and the
water layers can be up to several hundred angstroms in thickness [4].
Table 1 Melting and Crystallization Points and a -» sub-o: Transition Temperature of Distilled
Monoglycerides
Melting point, Crystallization point, Transition point,

j8 form a form a sub-a form
Monoglyceride (°C) (°C) (°C)
Monocaprin 52 16 —
(90% Cjo fatty acid)
Monolaurin 58 38 19
(92% Ci2 fatty acid)
Monomyristin 64 50 26
(92% Ci4 fatty acid)
Monopalmitin 68 62 34
(92% Cj6 fatty acid)
Monostearin 74 68 42
(85% C l8 fatty acid)
Monoolein 35 2 —
(80% 18:1 fatty acid)

Monolinolein 12 -6 —
(65% 18:2 fatty acid)

Monobehenin 84 78 54
(90% C22 fatty acid)
524 Krog
Lipid bilayers
Water
Fig- 2 Schematic structure of the lamellar mesophase. d, d^, d^, and S are structural parameters
measured by low-angle X-ray diffraction analysis. (Courtesy of Danisco Ingredients, Denmark.)
Such ordered structures have a long-range order but no short-range order similar to the
crystalline state, and such structures are referred to as liquid crystals or mesomorphic phases.
The fatty acid chains are in a liquid-like state with rotating hydrocarbon chains. The associa
tion of distilled monoglycerides with water results in formation of various mesomorphic
phases with lamellar, cubic, or reversed hexagonal structures depending on the fatty acid
composition of the monoglyceride, temperature, and concentration [4]. A detailed description
of lyotropic, liquid crystalline phases of monoglycerides, phospholipids, and glycolipids was
recently given by Larsson [12].
Formation of lamellar phases of saturated Ci^-Cjg distilled monoglycerides in water takes
place at temperatures above the Krafft point (T^) as shown in Fig. 3a. This figure shows a
binary phase diagram of saturated, distilled monoglycerides in water. The Krafft point for a
monoglyceride is 20-25°C lower than its bulk melting point. At high water concentrations,
“dispersions” of lamellar units are formed. These lamellar units have unique properties in low
fat foods, where monoglyceride dispersions can be added by way of the water phase. This is
used to facilitate interactions with starch components in cereal-based products or processed
potato products, etc. The use of mesomorphic phases of monoglycerides to improve the tex
ture of fat-free foods such as spreads, dressings, and bakery products was recently described
in the patent literature [13].
When cooling a mesophase below the Krafft temperature, an a gel phase is formed due to
crystallization of the fatty acid chains. The water layers remain between the polar head groups
as in the lamellar phase, so the only difference is in the crystallinity of the fatty acids.
The gel phase is an important physical state for the function of monoglycerides or other
emulsifiers used in whippable emulsions [14] or as texturizing agents in fat-free foods. The
gel phase has water-binding capacity, enhances destabilization of whippable emulsions, or
increases the relative viscosity of low fat products.
The use of monoglyceride “hydrates” as antistaling agents or crumb-softening agents in
wheat bread has been known for decades [15], and this application accounts for a large pro
portion of the world production of distilled monoglycerides. The so-called hydrates are pro
duced by hydrating 20-45% distilled monoglycerides in water above the Krafft point where
(a) Distilled, saturated monoglycerides
% water
Fig. 3 Binary phase diagrams of (a) saturated, distilled monoglycerides (DIMODAN P, Danisco Ingre
dients, Denmark) and (b) unsaturated, distilled monoglycerides (DIMODAN LS, Danisco Ingredients,
Denmark) in water. = Krafft temperature; = temperature where the a gel is formed. (Courtesy of
Danisco Ingredients, Denmark).
they form a lamellar phase. This is cooled under agitation to a temperature at which the a gel
phase is formed. The mix is then acidified to a pH of about 3.5 under continuous agitation,
resulting in a transformation from the gel phase to a suspension of jS crystals in water. The
final product is a white, pastelike product that is easy to incorporate into bread dough. The
advantage of hydrating the monoglycerides before use is that the specific surface area of the
crystals formed by recrystallization in water is much greater than what can be made by me
chanical processes (spray-cooling or milling). The hydrated aqueous preparations of mono
glycerides are more effective as starch complexing agents than any powdered products [16].
Unsaturated distilled monoglycerides (mainly monooleates or monolinoleates) form cubic
mesophases in water at ambient temperature as shown in Fig. 3b. The interaction of unsatu
rated monoglycerides with water has a specific function, particularly in whippable emulsions,
where the affinity of unsaturated monoglycerides to water provides partial destabilization of
the fat globules [17]. This is desired to improve the aeration properties and foam stability of
whipped cream or ice cream products. The mesomorphic behavior of polar lipids in water can
be an important factor in many different applications of emulsifiers in foods (e.g., bakery
products).
526 Krog
CH2-O-CO-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH3
CHOH
I
CH2OH
(a)
CHj-O-CO-CHj-CHj-CHj-CHj-CHj-CHj-CHj-CHj-CHj-CHj-CHj-CHj-CHj-CHj-CHj-CHj-CHj
CHOH
I
CHj-O-COCHj
(b)
CHP-0-CO-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH3
CHp-O-CO
CHp
(C)
CH2-O-CO-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH2-CH3
CHOH
HC-O-COCH3
(d)
Fig. 4 Chemical structure of organic acid esters of monoglycerides, (a) Glycerol 1-monostearate; (b)
acetic acid ester of glycerol monostearate; (c) lactic acid ester of glycerol monostearate; (d) diacetyltar-
taric acid ester of glycerol monostearate. (From Ref. 21.)
The uses of distilled monoglycerides in foods are numerous, and replacement of mono
diglycerides with distilled monoglycerides has occurred in many products. This is the case in
dairy products, where distilled monoglycerides are used instead of mono-diglycerides in ice
cream, coffee whiteners, and whippable creams. Combinations with lecithin or other more
polar emulsifiers are used in many emulsions.
III. ORGANIC ACID ESTERS OF MONOGLYCERIDES

Derivatives of monoglycerides containing organic acids such as acetic acid, lactic acid, diace-
tyltartaric acid, or citric acid exhibit properties very different from those of monoglycerides
in terms of crystalline behavior and surface activity (polarity). Emulsifiers with different func
tional properties in foods can therefore be made by reacting monoglycerides with selected
organic acids.
Organic acid esters are in principle made by reacting mono-diglycerides or distilled mono
glycerides with anhydrides of the organic acid in question on a mole/mole basis, yielding a
mixture of components where the organic acid is esterified to one of the free hydroxyl groups
on the monoglyceride molecule [4]. Chemical structures and molecular models of organic acid
esters compared to distilled monoglycerides are shown in Fig. 4.
A. Acetic Acid Esters of Monoglycerides—EU 472a, US/

FDA/CFR § 172.828
Acetylated monoglycerides (ACETEMs) are based on distilled monoglycerides, and products
with 50, 70, and 90% acetylation of free hydroxyl groups are available. Acetylation of mono
glycerides lowers their melting point to about 25-30°C below that of the corresponding mono
glycerides. Furthermore, ACETEMs are stable in the a crystal form and show no polymor
phism. The chemical structure of an ACETEM is shown in Fig. 4b.
Fully acetylated ACETEMs are not surface-active but behave like plastic fats. Such prod
ucts are used as low calorie fats in emulsions and some bakery products (see Chapter 19).
Fully acetylated ACETEMs form flexible films with a low degree of permeability of oxygen
and water vapor [18], and they are therefore used as coating agents for frozen meat or fruit.
Partially acetylated ACETEMs have low surface activity depending on their degree of
acetylation, which is usually between 50% and 70%. Such products can form a gel structures
with water, and this is the basis for their application in specialty fats for toppings, whippable
emulsions, or cake mixes.
B. Lactic Acid Esters of Monoglycerides—EU 472b, US/

FDA/CFR § 172.852
Mono-diglycerides or distilled monoglycerides can be esterified with 15-35% lactic acid,
forming lactic acid esters (LACTEMs). The chemical structures vary, but a main component
of a LACTEM is shown in Fig. 4c. Together with ACETEMs, LACTEMs are often referred
to as “a-tending” emulsifiers due to their monomorphic properties and stability in the a crystal
form. The melting point of LACTEMs is ~45°C, and they form a gel structures with water
[19]. LACTEMs are mainly used in specialty fats for dessert products (non-dairy creams,
toppings, cake mixes).
C. Diacetyltartaric Acid Esters of Monoglycerides—EU

472e, US/FDA/CFR § 182.4101
Diacetyltartaric acid esters of monoglycerides (DATEMs) are made by reacting mono-diglyc
erides or distilled monoglycerides with diacetyltartaric acid anhydride. Many variations in
composition exist, depending on the ratio of monoglyceride to tartaric acid anhydride used
during esterification, and the final products may contain more or less nonesterified monoglyc
erides. A main component of DATEMs is shown in Fig. 4d. The physical form may vary
from a viscous liquid to a solid, depending on the fatty acid composition of the monoglycer
ides used. DATEMs based on Ci^-Cjg monoglycerides melt at ~45°C and crystallize from
melt to a stable a crystal form. DATEMs are anionic-active, very hydrophilic emulsifiers and
are dispersible in water. A DATEM has a free carboxyl group, and its acidity in water lowers
the pH to below 2. If the pH is adjusted to above 4.5, the DATEM is partially neutralized, and
under such conditions it forms stable lamellar mesophases in water. DATEMs have multiple
528 Krog
applications in foods, mainly as dough strengtheners in yeast-raised bakery products (bread,

rolls, etc.), but they are also effective emulsifiers in many emulsions.
D. Citric Acid Esters of Monoglycerides—EU 472c, § US/

FDA/CFR 172.832
Citric acid esters of monoglycerides (CITREMs) are produced by reacting citric acid in an
amount of 12-20% by weight of the final product with mono-diglycerides or distilled mono
glycerides. CITREMs are anionic-active, hydrophilic emulsifiers forming a milky dispersion
in water and are only slightly soluble in oils and fats.
The typical use of CITREMs is in emulsions. They are used in frying margarine as anti
spattering agents and in meat emulsions to prevent fat separation.
IV. FATTY ACID ESTERS

Several emulsifiers are produced by esterification of polyvalent alcohols with fatty acids. The
main fatty acids are palmitic, stearic, and oleic acids, but products with other fatty acids
are available.
A. Polyglycerol Esters of Fatty Acids—EU 475, US/FDA/

CFR § 172.854
Esterification of condensed glycerol, mainly tri- and tetraglycerol, with fatty acids, usually
palmitic, stearic, or oleic acids, yields products with 20-40% mono-fatty acid esters. The
degree of polymerization of the glycerol and the degree of esterification can vary considerably
depending on application and legal aspects.
Polyglycerol esters of fatty acids (PGEs) are stable in the a crystal form, are water-dispers
ible, and form reversed hexagonal mesomorphic phases above their melting point (55-60°C).
PGEs are slightly more polar than monoglycerides, and they are mostly used in emulsions,
often in bakery products such as cakes.
B. Propylene Glycol Fatty Acid Esters—EU 477, US/FDA/

CFR § 172.856
Propylene glycol can be esterified with palmitic and stearic acids, forming a blend containing
55% monoester and 45% diesters. This can be purified by molecular distillation, yielding a
product containing a minimum of 90% monoesters. Propylene glycol monostearates (PGMSs)
have a melting point of about 40°C. They are capable of forming a gel phases in water below
the melting point, making them effective emulsifiers in whippable emulsions (toppings) [20],
bakery shortenings, cake mixes, etc. A PGMS is often used in combination with monoglycer
ides to make stable a crystalline blends of emulsifiers. Such combinations are mostly used in
non-fat cake systems.
C. Sorbitan Monostearate—EU 491, US/FDA/CFR § 172.842

Sorbitan is derived from sorbitol by dehydration and by esterification with palmitic and stearic
acid blends, yielding products with different hydrophilic and lipophilic properties depending
on the degree of esterification. Sorbitan mono-fatty acid esters, usually referred to as sorbitan
monostearates (SMSs), are hydrophilic, water-dispersible emulsifiers. SMSs can be reacted
with ethylene oxide to form polyoxyethylene sorbitan esters (POEMS), which are among the
most hydrophilic, water-soluble emulsifiers available. SMSs are used in many types of emul
sions (dressings, sauces, etc.) often in combination with POEMS.
Blends of monoglycerides and POEMS are used in dairy products (ice cream), but the
use of POEMS in foods is limited due to restrictive legislation regarding these products in
many countries.
D. Sorbitati Tristearate (STS)—EU 492

By using a high proportion of fatty acids relative to the amount of sorbitan during esterifica
tion, sorbitan tristearate (STS) is produced. STS is a fat-soluble emulsifier with a molecular
structure close to that of triglycerides [21], as shown in Fig. 5. It is stable in the a crystal
form and can cocrystallize with fats and form solid solutions. This makes STS an effective
crystal modifier in fat products. STS is specifically used as a crystal modifier in margarine
blends, stabilizing the ¡5' crystal form, or as a bloom inhibitor in chocolate and confectionery
fats. STS has no surface-active properties and is not used as an emulsifier for stabilizing emul
sions.
CH 3-(CH 2)i 6-C - 0 '
Rg=5 Chemical structure of sorbitan tristearate. (From Ref. 2 1 .)
E. Sucrose Esters of Fatty Acids—EU 473, US/FDA/CFR

§ 172.859
Sucrose esters are a new generation of food emulsifiers made by esterification of sucrose with
fatty acids. By varying the ratio between fatty acids and sucrose, a number of esters with
different hydrophilic/lipophilic properties can be made. Sucrose esters containing up to 70%
monoesters are water-dispersible, and sucrose esters with a low monoester content (10-30%)
are oil-soluble.
Sucrose esters can be used in many types of foods from emulsions to starch- or fat-based
products. However, sucrose esters are relatively new to the food industry. The widest scope
of application is found in Japan, where these products are generally permitted for use in
foods. In Europe and the United States, the use of sucrose esters is still somewhat limited.
A mixture of sucrose esters and mono-diglycerides, so-called sucroglycerides (EU 474),
made by transinteresterification of triglycerides and sucrose, is also available but not widely
used.
F. Sodium Stearoyl-Lactylate—EU 482, US/FDA/CFR

§ 172.844
Reaction of stearic acid with lactic acid in the presence of sodium hydroxide yields sodium
stearoyl-lactylates (SSL), which is a water-dispersible, anionic active emulsifier with a melt
ing point of ~45°C. SSL is a very versatile emulsifier with many different applications. Due
to its anionic properties it is an effective emulsifier in various types of emulsions, and it also
works as a dough strengthener and starch-complexing agent in bakery products.
530 Krog
V. SPECIFIC EMULSIFIER INTERACTIONS IN FOODS

The function of emulsifiers in foods is very complex and can be only briefly described here.
In principle, functions of emulsifiers can be divided into interactions taking place during food
processing (e.g., emulsification, aeration, dough strengthening) and interactions with food
components leading to extended shelf life and improved texture of finished products (e.g.,
crumb softening of bread, texture of ice cream, emulsion stability). Very often it is not possi
ble to correlate physical properties of emulsifiers with their functions in foods. It is also
difficult to make reliable predictions about the functionality of a specific type of emulsifier in
a given type of food. Although more and more basic knowledge about functions and interac
tions of emulsifiers is available [22], selection and optimization of an emulsifier system are
still based on experimental tests.
A. Emulsion Stability and Destabilization

The effect of emulsifiers on emulsion stability is related to their ability to form viscoelastic
interfacial films, preventing droplet coalescence [23]. Most food emulsions contain proteins
that are responsible for the emulsion’s stability in, for example, milk or cream, due to adsorp
tion of proteins to fat globule surfaces. In combination with emulsifiers, mixed interfacial
films are formed [24], thus increasing the stability.
When anionic emulsifiers (DATEM, SSL) are used, their affinity to proteins may further
bind proteins to the interface [25] and consequently provide increased emulsion stability.
In whippable emulsions (ice cream, cream), partial desorption of interfacially bound pro
tein is desired to enhance aeration properties. This is provided by monoglycerides, which have
higher surface activity than proteins at low temperatures and therefore replace the proteins at
the surface of the fat globules [26]. The competition between emulsifiers and proteins for
adsorption at the surface of fat globules can be studied by interfacial tension measurements as
demonstrated in Fig. 6. From this it appears that the interfacial tension, in the presence of
monoglycerides, is reduced when the temperature is lowered, whereas a system in which only
Fig. 6 Measurements of interfacial tension versus temperature of purified sunflower oil-water systems
containing (a) oil with 0.2% monoglycerides (GMS) against pure water, (b) pure oil against water with
0.01% milk proteins, and (c) oil with GMS against water with protein. Temperature program: Cool
from 40°C to 5°C at 0.3°C/min, hold 60 min at 5°C, heat to 40°C at 2°C/min, and hold for 90 min.
(Courtesy of Danisco Ingredients, Denmark.)
proteins are present is not affected by a change in temperature. This means that the interfacial
concentration of the monoglycerides is increased at low temperatures. Consequently, when
both monoglycerides and proteins are present, the monoglycerides will displace the proteins
from the interface. When the protein concentration at the surface is reduced, the fat globules
tend to flocculate and form clusters, and this enhances the whipping properties and improves
the texture and stability of the finished whipped cream. It is well known that agglomerated
fat particles (clusters) stabilize the air cells in whipped emulsions such as ice cream or dairy/
non-dairy creams [27,28]. (See Chapters 12 and 13.)
B. Interactions with Starch Components

Distilled monoglycerides have specific functions in starch-based foods that cannot be obtained
with mono-diglycerides. Due to their water dispersibility, distilled monoglycerides are more ef
fective in interactions with water-soluble food ingredients. Consequently, they work better than
mono-diglycerides as starch-complexing agents [29] in potato products or bakery products.
E
o
Ü
<E
Fig. 7 Effects of monoglycerides on the amylose-lipid complex formation (A-LC), the rétrogradation
of amylopectin (A), and the crumb firmness (F) of bread after 3 days of storage at 22°C. (From Ref. 22.)
The shelf life of wheat bread (toasting bread, high fiber bread, etc.) is very important for the
baking industry and has great economic aspects. The complex formation between the water-sol
uble amylose and saturated monoglycerides delays the staling process and increases the crumb
softness of bread [30]. The interaction of monoglycerides with the starch components amylose
and amylopectin is a function of the concentration of the monoglycerides added to the flour part
of the dough. Their influence on the crumb firmness of the bread is shown in Fig. 7. It appears
that at low concentrations the monoglycerides interact primarily with the amylose fraction, and
532 Krog
this complex formation decreases the crumb firmness in fresh bread. At higher monoglyceride
concentrations, complex formation with the amylopectin fraction takes place, thus reducing the
rate of staling during storage. Consequently, the optimal concentration of monoglycerides is be
tween 1% and 2% based on flour weight. The complex formation with starch components also
improves the texture of extruded cereals, processed potato products, pasta, etc.
C. Dough-Strengthening Effects
Anionic-active emulsifiers, such as DATEM or SSL, are used as processing aids in dough to
increase the strength of the gas cell membranes formed during fermentation and to provide
improved dough elasticity to prevent the dough structure from collapsing due to mechanical
abuse during production in automated plants. The resistance of dough to mechanical shock is
important when producing bread or rolls with uniform volume and texture.
The relative baking effect of flour is related to the content of the polar lipids in it [31]. The
polar lipids in flour consist of phospholipids and glycolipids, and it has been demonstrated that
emulsifiers such as lecithin and DATEM have the same effect in bread dough as the polar
flour lipids [32]. This can be seen from Fig. 8, which shows the relative effect of flour lipids,
lecithin, and DATEM on loaf volume of bread made from defatted wheat flour. The function
of both the polar flour lipids and emulsifiers such as DATEM, SSL, or lecithin in dough is
related to the formation of lamellar, liquid crystalline phases that are part of the gas cell
membrane [33].
850
550
Normal flour Defatted flour
Fig. 8 Influence of flour lipids and emulsifiers (lecithin, DATEM) on bread loaf volume. The follow
ing amounts of lipids were added based on flour weight; 0.7% total wheat lipid (TL), 0.15% nonpolar
wheat lipid (N-PL), 0.3% lecithin (Lee), and 0.3% DATEM. The wheat flour (12.3% protein) was
used before and after partial extraction of lipids. No emulsifiers were added to the control test. (From
Ref. 22.)
IV. CONCLUDING REMARKS

The variety of emulsifiers and their properties provides selective choices for various food
products, with applications ranging from film coatings that protect foods from oxidation or
loss of humidity, through stabilizing or destabilizing emulsions, dough strengthening and
crumb softening of bakery products, texturizing of starch-based foods, and modification of fat
Food Em ulsifiers 533
crystallization. Progress in research on the interfacial and functional properties of em u lsifie rs

has greatly increased the understanding of how emulsifiers work in various food systems. The
need for emulsifiers is evident in most industrially manufactured foods, particularly to ensure
uniform quality and prolonged shelf life during storage. Low fat products certainly need emul
sifiers as well as other texturizing ingredients to provide textural characteristics similar to
those of full fat products.
The development of new chemical types of emulsifiers is not likely to happen due to the
high cost of toxicological tests necessary to ensure their safety for use in foods. Instead,
developments can be expected to deal with improving the purity of the existing range of
emulsifiers and making better defined products in terms of chemical composition.
New developments in biochemical synthesis of emulsifiers, such as monoglycerides, using
lipolytic enzymes are in progress. Different enzymatic methods for glycerolysis, selective
hydrolysis of fats and oils, or esterification of fatty acids with glycerol or other polyols are
possible [34].
ABBREVIATIONS
ACETEM Acetic acid ester of monoglyeeride
LACTEM Lactie acid ester of monoglyceride
DATEM Diacetyltartaric acid ester of monoglyceride
CITREM Citric acid ester of monoglyceride
PGE Polyglycerol ester of fatty acid
PGMS Propylene glycol monostearate
POEMS Polyoxyethylene sorbitan monostearate
SMS Sorbitan monostearate
STS Sorbitan tristearate
SSL Sodium stearoyl-lactylate
REFERENCES
J. B. Lauridsen, Food emulsifiers: surface activity, edibility, manufacture, composition and appli
cation, J. Am. Oil Chem. Soc. 53: 400 (1976).
2. G. Schuster (Ed.), Emulgatoren für Lebensmittel, Springer Verlag, Berlin, 1985.
3. F. R. Benson, in Nonionic Surfactants (M. J. Schick, Ed.), Marcel Dekker, New York, 1987,
p. 247.
4. N. J. Krog, Food emulsifiers and their chemical and physical properties, in Food Emulsions (K.
Larsson and S. Friberg, Eds.), Marcel Dekker, New York, 1990, p. 127.
5. G. Als and N. Krog, Emulsifiers as processing aids, in Proceedings. World Conference on
Oleochemicals into the 21st Century (T. H. Applewhite, Ed.), Am. Oil Chem. Soc. Press, Cham
paign, IL, 1990, p. 67.
6. Compendium of Food Additive Specifications, FAO Food and Nutrition Paper 52, Food and Agri
culture Organisation of the United Nations, Rome, 1992.
7. W. Clayton, Margarine, in The Theory of Emulsions and Their Technical Treatment (W. Clayton,
Ed.), J. & A. Churchill, London, 1943, p. 313.
R. O. Feuge and A. E. Bailey, Modification of vegetable oils. VI. The practical preparation of
mono- and diglycerides. Oil Soap 23: 259 (1946).
9. N. O. V. Sonntag, Fat splitting, esterification, and interesterification, in Bailey's Industrial Oil
and Fat Products, Vol. 2 (D. Swem, Ed.), Wiley, New York, 1982, p. 97.
10. M. W. Formo, Physical properties of fats and fatty acids, in Bailey's Industrial Oil and Fat
Products, Vol. 1 (D. Swem, Ed.), Wiley, New York, 1979, p. 192.
534 Krog
11. K. Larsson and P. J. Quin, Physical properties; structural characteristics, in The Lipid Handbook,
2nd ed. (F. D. Gunstone, J. L. Harwood, and F. B. Padley, Eds.), Chapman & Hall, London,
1994, p. 401.
12. K. Larsson, Liquid-crystalline lipid-water phases, in Lipids—Molecular Organization, Physical
Functions and Technical Applications (K. Larsson, Ed.), The Oily Press, Dundee, Scotland, 1994,
p. 47.
13. I. Heertje, Eur. Patent Specification No. 0 558 523 B1 (1994), to Unilever N.V ., Rodderdam (NL).
14. N. Krog, The role of Low-polar emulsifiers in protein-stabilized food emulsions, in Emulsions—A
Fundamental and Practical Approach (J. Sjoblom, Ed.), Kluwer, Boston, 1992, p. 61.
15. G. H. Doerfert, Improved hydrates of hard monoglycerides. Bakers Dig. 42: 56 (1968).
16. N. Krog and B. Nybo Jensen, Interaction of monoglycerides in different physical states with
amylose and their anti-firming effects in bread, J. Food Technol. 5: 11 (1970).
17. N. M. Barfod, N. Krog, G. Larsen, and W. Buchheim, Effects of emulsifiers on protein-fat
interaction in ice cream mix during ageing. I: Quantitative analysis. Fat Sci. Technol. 93: 24
(1991).
18. O. Fennema and J. J. Kester, Resistance of lipid films to transmission of water and oxygen, in
Water Relationships in Food (H. Levine and L. Slade, Eds.), Plenum, New York, 1991, p. 703.
19. J. M. M. Westerbeek and A. Prins, Function of alpha-tending emulsifiers and proteins in whipp-
able emulsions, in Microemulsions and Emulsions in Foods (M. El-Nokaly and D. Cornell, Eds.),
Am. Chem. Soc., Washington, DC, 1991, p. 146.
20 . W. Buchheim, N. M. Barfod, and N. Krog, Relation between microstructure, destabilization phe
nomena and rheological properties of whippable emulsions. Food Microstruct. 4: 221 (1985).
21 . N. Krog and D. Last, Palm-oil based food emulsifiers, Malaysian Oil Sci. Technol. 4: 93 (1995).
22 . N. Krog, Interactions of emulsifiers with other components in foods, in Ingredient Interactions:
Effects on Food Quality (A. G. Gaonkar, Ed.), Marcel Dekker, New York, 1995, p. 377.
23. E. H. Lucassen-Reynders, Interfacial viscoelasticity in emulsions and foams, Food Struct. 12:
1 (1993).
24. G. Doxastakis and P. Sherman, The interaction of sodium caseinate with monoglyceride and di
glyceride at the oil-water interface and its effect on interfacial rheological properties. Colloid
Polym. Sci. 264: 254 (1986).
25. E. Dickinson and S.-T. Hong, Surface coverage of beta-lactoglobulin at the oil-water interface:
influence of protein treatment and various emulsifiers, J. Agric. Food Chem. 42: 1602 (1994).
26. N. Krog, Thermodynamics of interfacial films in food emulsions, in Microemulsions and Emul
sions in Foods (M. El-Nokaly and D. Cornell, Eds.), Am. Chem. Soc., Washington, DC, 1991,
p. 138.
27. W. Buchheim and P. Dejmek, Milk and dairy-type emulsions, in Food Emulsions, 2nd. ed. (K.
Larsson and S. Friberg, Eds.), Marcel Dekker, New York, 1990, p. 203.
28. B. E. Brooker, M. Anderson, and A. T. Andrews, The development of structure in whipped
cream. Food Microstruct. 5: 211 (1986).
29. N. Krog, Amylose complexing effect of food grade emulsifiers, StarkelStarch 23: 206 (1971).
30. N. Krog, S. K. Olesen, H. Tômæs, and T. Jonsson, Rétrogradation of the starch fraction in wheat
bread. Cereal Foods World 34: 281 (1989).
31. F. MacRitchie, Flour lipids and their effects in baking, J. Sci. Food Agric. 28: 52 (1977).
32. Y. Pomeranz, El-Baya, W. Seibel, and H. Stephan, Toast bread from defatted wheat flour. Cereal
Chem. 61: 136 (1984).
33. A.-C. Eliasson and K. Larsson, Physicochemical behaviour of components, in Cereals in Bread
making, (A.-C. Eliasson and K. Larsson, eds.), Marcel Dekker, New York, 1993, p. 42.
34. U. Bomscheuer, Lipase-katalysierte Synthèse von Monoglyceriden, Fat Sci. Technol. 97: 241
(1995).
21
Lipid Emulsions for Intravenous Nutrition
and Drug Delivery
Robert T. Lyons and Eric G. Carter
P harm acia & Upjohn In c., C layton, N orth Carolina
L INTRODUCTION: THE STRUCTURE AND PROPERTIES

OF OIL-IN-WATER EMULSIONS
A liquid emulsion is a dispersion of one immiscible phase within another, with a surfactant
used to stabilize the dispersed droplets. Water-in-oil emulsions for vaccine delivery are given
by intramuscular injection and are designed for sustained release. In contrast, fat emulsions
for parenteral nutrition need to be compatible with blood and therefore are of the oil-in-water
type [1]. Those currently in clinical use are composed of either soybean or safflower oil, and
some products contain oil blends with medium-chain triglycerides. The latter are composed
mainly of saturated Cg and Cio fatty acids. Suitable vegetable oils are emulsifled with either
egg yolk or soybean phospholipids. The emulsified lipid droplets typically range in size from
about 0.2 to 0.6 /xm and are metabolized in a manner similar to that of natural chylomicrons
[1,2]. Commercial examples in the United States include Intralipid (Pharmacia) and Liposyn
(Abbott Laboratories). In Europe, several others are also available, including Lipofundin (B.
Braun), and in Japan, Soyacal (Green Cross). These products are described in more detail in
Section III. Their relatively low toxicity and excellent safety record, exemplified by extensive
use in admixtures for total parenteral nutrition, have made this type of fat emulsion an attrac
tive choice as a carrier system or vehicle for lipophilic drugs and vitamins [3-7]. These
applications are described in Section II.
Purified egg yolk phospholipids, the most commonly used emulsifier, stabilize the dis
persed submicron oil droplets by two important mechanisms, both of which act at the oil/
water interface to inhibit aggregation and coalescence. Trace quantities of ionized molecular
species, such as nonesterified (free) fatty acids and phosphatidic acid, impart a net negative
surface charge (zeta potential) that stabilizes the dispersion by charge-charge repulsion, even
at elevated temperatures. Simultaneously, zwitterionic species such as phosphatidylcholine
and phosphatidylethanolamine and neutral lipids such as cholesterol establish a monomolecu-
lar interfacial film that binds water molecules at the surface and serves as a physical barrier
535
536 Lyons and Carter
Charge
< ---------------- >

Repulsion
B o u n d W ater F a tty A c id P h o s p h o lip id
Fig. 1 Stabilization of oil-in-water emulsions. (A) Nonionic emulsifiers at the oil/water interface stabi
lize dispersed droplets by forming a physical barrier consisting of a monomolecular film and bound
water molecules. (B) Ionic emulsifiers provide charge-charge repulsive forces. (C) Intralipid is made
with egg yolk phospholipids, which combine both nonionic and ionic stabilizing principles.
against coalescence [8-10]. These stabilizing principles are summarized in Fig. 1. Fat emul
sions designed for parenteral nutrition are chemically and physically stable enough to be termi
nally heat sterilized in steam autoclaves at 121°C.
During prolonged storage, e.g., over 18 months, certain physical changes may be ob
served. These events, including flocculation, coalescence, and creaming [6,8-10], are sum
marized in Fig. 2. Creaming is the result of lower density lipid droplets floating upward in
the aqueous phase and is readily reversible by gentle shaking. Flocculation is the formation
of droplet clusters or aggregates. These are not readily redispersed by agitation. Coalescence
is also irreversible, and the resulting large oil droplets can make an emulsion product unsuit
able for injection [11].
Possible chemical changes that may occur during long-term storage of these emulsions
include hydrolysis and/or peroxidation of the unsaturated fatty acid side chains on the phos
pholipid emulsifier as well as in the triglyceride oil phase [12,13]. However, these reactions
Lipids Emulsions for IV Nutrition and Drug Delivery 537
Fig. 2 Physical changes possible in a lipid emulsion. (A) Freshly prepared emulsion. (B) Creaming is
easily reversible. (C) Flocculation and (D) coalescence are irreversible. (E) Rapid creaming after coales
cence. (F) Rapid creaming after flocculation. (G) Total separation of oil and water phases (“broken”
emulsion) is irreversible.
may be minimized by careful pH adjustment during homogenization and by stringent protec

tion from oxygen. For example, continuous nitrogen blanketing during all production steps is
standard practice to minimize oxidation of unsaturated lipids and of the drug component,
when present.
II. INTRAVENOUS LIPIDS: AN ALTERNATIVE DELIVERY

SYSTEM FOR DRUGS AND VITAMINS
A. Potential Advantages of Using Lipid Emulsion Vehicles
for Lipophilic Drugs
The most obvious reason to consider using a lipid emulsion as a vehicle is to solubilize
lipophilic drugs. Although chemists can usually derivatize any compound to enhance water
solubility, lipophilic forms are often more effectively taken up by target cells. Amphipathic
molecules, which contain both strongly nonpolar and strongly polar groups, are only partially
lipid-soluble. Yet amphipathic compounds may also be attractive candidates because they tend
to localize at the oil/water interface and intercalate in the phospholipid monolayers [5,14].
In contrast, total immersion in an oil phase may stabilize certain otherwise labile drugs
against hydrolysis and/or oxidation during processing, sterilization, and storage [7]. In certain
cases, inclusion of pH buffering agents such as amino acids, antioxidants such as «-tocoph
erol, and metal chelators such as EOT A should be helpful. Although most nutritional fat
emulsions are stable for up to 18 months at room temperature, lipid-drug emulsions may
require controlled storage at lower temperatures, often 4-8°C. In practice, the shelf life of a
lipid-drug emulsion is more often limited by degradation of the incorporated drug than by
deterioration of the emulsion vehicle. Loss of drug potency is usually monitored by HPLC-
based analysis of dégradants; hydrolytic and/or oxidative reactions are often involved. There
fore, a perennial challenge for formulation chemists and process engineers is to optimize
combinations of excipients, pH, drug concentration, and process conditions in order to mini-
538 L yons and Carter
Fig. 3 Propofol (2, 6 -diisopropylphenol). This intravenous anesthetic agent is highly lipophilic. It has
been formulated in a 1 0 % soybean oil emulsion that is terminally heat sterilized and is successfully
marketed as Diprivan.
mize rates of degradation while retaining those properties that keep the emulsion physically
stable.
Reduction of drug toxicity by means of association with lipids has been well documented;
examples include the antifungal agent amphotericin B [15,16] and the antineoplastic drug
penclomedine [17]. Utilization of a lipid emulsion vehicle also eliminates the need for organic
solvents such as propylene glycol, ieri-butanol, or dimethyformamide. These solvents are
often associated with pain on injection, phlebitis, and other acute toxicities. Propofol (Fig.
3), an injectable anesthetic formerly in a solvent vehicle, was reformulated in a 10% soybean
oil emulsion about 15 years ago to prevent vein irritation [18,19]. Now being marketed as
Diprivan, this product represents a classical example of the benefits of lipid-based drug de
livery.
Another potential advantage of using lipid emulsions as a vehicle for certain lipophilic
agents is to reduce adsorption onto plastic infusion sets [7]. Documented examples of such
losses include Perilla ketone, an investigational cytotoxic drug [20], and the fat-soluble vita
mins A, D, E, and Kj (see Table 2).
Possibilities for improved pharmacokinetics, such as some form of sustained release and/
or directed drug delivery to various organs, is a complex and still evolving field. Controlled
release systems require the preparation of emulsified lipid droplets with long circulation times
in the blood and slow drug transport rates from the dispersed to the continuous phases. In
general, the clearance of lipid droplets from circulation depends on size (smaller particles
clear more slowly), surface charge (highly charged particles clear more rapidly), and surfac
tant composition [1]. For example, Jeppsson and Rossner [21] observed a fractional clearance
rate of about 7%/min following intravenous administration of an egg yolk phospholipid-stabi
lized emulsion in rabbits. The addition of an amphipathic polymeric surfactant, specifically
poloxamer 338 (Pluronic F108), resulted in a sixfold prolongation of circulation time.
B. Potential Disadvantages of Using Lipid Emulsion

Vehicles for Lipophilic Drugs
Potential problems with the use of lipid drug emulsions include the risk of embolism. Particles
greater than about 7/xm in diameter may lodge in capillary beds, especially those in the lungs
[22]. Therefore, manufacturing processes for such parenteral products must be under stringent
quality control. Intravenous fat emulsions stabilized by egg phospholipids have an excellent
safety record and are an integral part of total nutrient admixtures, along with dextrose and
amino acids [23]. However, inclusion of diverse surfactants or the use of alternative oils may
increase vehicle toxicity. For example, although medium-chain triglycerides are an excellent
solvent for many lipophilic drugs, too rapid infusion of these emulsions causes metabolic
acidosis and central nervous system toxicity in dogs [24].
Another important concern is the interaction of emulsified oil droplets with plasma pro
teins. For example, a strong association with fibronectin could lead to opsinization and rapid
uptake by phagocytic cells in the reticuloendothelial system (RES) [25]. Alternatively, a rapid
dissociation of surfactant from the surface of these droplets could induce flocculation in vivo,
which in turn would increase the risks of RES involvement or even of lung embolism. Poten
tial biological effects such as these must be monitored carefully during controlled animal
studies in at least two species before human trials can begin.
Yet another consideration is possible drug crystallization during prolonged storage and
shipping. Emulsions with crystals in the oil droplets tend to be destabilized if polar domains
pierce and weaken the interfacial film, leading to unsuitable flocculation or coalescence [26].
Also, crystal formation in any parenteral formulation will result in unacceptable particulate
counts [27].
C. The Design, Testing, and Manufacture of Lipid

Emuisions for Drug Deiivery
There are basically two ways to prepare a lipid drug emulsion. One is referred to as the
extraneous ^ addition of a lipophilic compound to a preformed, sterile fat emulsion, usually
with the assistance of an organic solvent such as ethanol or dimethyl sulfoxide. This method
is not recommended, for several reasons. First, even small amounts of solvent can damage
emulsions and produce large, poorly emulsified oil droplets. Second, undissolved crystals of
drug are not easily detected in an opaque white matrix. Third, it is difficult to ensure sterility
after entering a previously sterilized container. The preferred method of preparation is an ab
initio^ emulsification of a drug-containing oil phase, either aseptically or preferably in con
junction with terminal heat sterilization. In this manner, an emulsion system is custom-de
signed to accommodate the unique requirements of a particular therapeutic agent.
The major steps required for the production of a drug-containing fat emulsion are summa
rized in Fig. 4. All manipulations are performed under a protective blanket of nitrogen gas.
Typically, the oil phase is first heated together with drug, surfactant, and cosurfactant (if
any). Then the aqueous phase, containing a tonicity agent such as glycerin, is added slowly
while running a high shear overhead mixer. Next, this preemulsion is transferred to a homoge-
nizer in which the liquid is pumped through a spring-loaded valve system under very high
pressure. The resulting shear and cavitation break large oil drops into submicron droplets.
Multipassing through these valves creates a more uniform and less polydisperse size distribu
tion. For every formulation, we find that there is a set of optimum process conditions that
includes pH, temperature, pressure, and even the number of homogenizer passes. Finished
emulsion is then filtered, vials are filled under nitrogen, and rubber stoppers with overseals
are applied. Finally, vials are sterilized in a steam autoclave [28,29].
A sequence of characterizations is normally performed on each batch of lipid drug emulsion
to assess stability and biocompatibility in the case of new formulations under development
and to maintain quality and ensure safety for products already approved for marketing. Some
examples of analytical testing commonly performed on lipid drug emulsions are summarized
in Table 1 [29].
D. First Generation Injectable Lipid Drug Emulsions

“First generation” emulsions are those already approved and being marketed in Europe and/or
the United States. As shown in Table 2, only five lipid drug emulsions are in this category.*
*“Extraneous” here means “coming from the outside.”
^“Ab initio” here means “from the beginning.”
Fig, 4 Major production steps for a lipid drug emulsion. The preparation of a commercial lipid drug
emulsion is technically demanding because of the requirement that droplet mean diameter remain in the
submicrometer range. Repeated passing through a high pressure piston homogenizer reduces polydisper-
sity in the drop size distribution. At all steps, processing is performed under a nitrogen blanket to
minimize oxidative degradation of unsaturated components.
Alphabetically, the first is Diazemuls, which contains the sedative diazepam (Fig. 5) at 5 mg/
mL in a soybean oil base. Drug solubility is enhanced with diacetylated monoglycerides, and
the emulsifier consists of egg yolk phospholipids. There are no quantitative or qualitative
differences in the pharmacological properties of diazepam, whether administered in an etha
nol-propylene glycol solution or in this emulsion. However, clinical studies demonstrate that
the lipid formulation reduces the incidence of local pain and thrombophlebitis seen with the
solution. Diazemuls is used preoperatively to relieve anxiety and provide sedation and light
anesthesia. It is also used as an anticonvulsant [30,31].
The second is Diprivan, which contains the anesthetic propofol (Fig. 3) at 10 mg/mL,
using Intralipid 10% as vehicle. Following injection, propofol exhibits a high metabolic clear
ance rate and a large steady-state distribution volume, suggesting that the drug rapidly dissoci
ates from the oil droplets in vivo. Diprivan is a short-acting sedative hypnotic agent, suitable
for induction and maintenance of anesthesia or sedation. There is swift recovery with little
postoperative nausea and vomiting [18,19].
Table 1 Common Analytical Procedures for Lipid Drug Emulsions
Analytical procedure Purpose of evaluation Possible defects
Visual appearance Uniformity Presence of nonemulsified surface

Oil
Phase-contrast microscopy Uniformity Large droplets or aggregates; drug
crystals
Laser light scattering Drop size distribution Large droplets or aggregates
Single-particle optical counting Concentration of particles > 1 /xm High counts due to crystals, oil
droplets
Hydrogen ion electrode Aqueous phase pH Out of specified physiological range
Osmolality Compatibility with blood Out of specified physiological range
HPLC Drug potency and impurities High levels of degradation products
Free fatty acids Hydrolysis of oil and phospholipid High levels cause blood cell hemo
lysis
Peroxides Oxidation of oil and phospholipid High levels toxic and may degrade
d ru g
Zeta potential Electrostatic surface charge Low charge predicts poor stability
Limulus amebocyte lysate Endotoxin concentration High levels pyrogenic (cause fever)
Table 2 First Generation Injectable Lipid Drug Emulsions
Trade name Manufacturer Marketer Drug Activity Comments
Diazemuls Pharmacia Dumex in Europe Diazepam Sedative Reduced venous irri

(Dizac in U.S.) & Upjohn Ohmeda in U.S. tation
Diprivan Pharmacia Zeneca in Europe; Propofol Anesthetic Induction and main
& Upjohn Stuart in U.S. tenance
Fluosol-DA Green Cross Green Cross in Perfluorodecalin Oxygen de O2 delivery during
Japan livery balloon angio
plasty
Vitalipid Pharmacia Pharmacia & Upjohn Vitamins A, D, Parenteral Infused as admixture
& Upjohn in Europe E, Kj nutrition
Lipo-PGE, Green Cross Green Cross in Lipo-PGE, Vasodilator, Stable to heat steril
Japan Prostaglan platelet ization
din Ej inhibitor
Fluosol-DA is the most extensively tested example of a perfluorocarbon emulsion. This

product is composed of perfluorodecalin (14% w/v) and perfluorotripropylamine (6% w/v),
emulsified with a mixture of Pluronic F-68, egg yolk phospholipids, and potassium oleate.
The Fluosol emulsion is supplied in a 500 mL flexible plastic container but is unstable and
must be stored frozen and then thawed and mixed with electrolytes just before use. It is
approved only for intracoronary administration, to provide oxygen to heart muscle during
balloon angioplasty [32].
Vitalipid is also a 10% soybean oil emulsion, essentially Intralipid with added fat-soluble
vitamins A, D, E, and Kj. This emulsion is first used to reconstitute a second vial of lyophi-
lized water-soluble vitamins. The mixture is then diluted in regular Intralipid prior to infusion
CH.
Fig. 5 Diazepam. This sedative drug of the benzodiazepine class occurs as a white crystalline powder
and is only slightly soluble in water (3 mg/mL).
for adult and pediatric patients. Vitalipid is terminally heat sterilized and has a 2 year shelf
life, yet it contains no added solubilizers, antioxidants, or antimicrobial preservatives [33].
Prostaglandin Ei (PGEj) is a potent vasodilator as well as an inhibitor of platelet aggrega
tion. Incorporation of PGEj into a 10% soybean oil emulsion is reported to increase biological
activity and reduce undesirable side effects [34].
E. Second Generation Injectable Lipid Drug Emulsions

It should be apparent by now that first generation products are, with the exception of Fluosol-
DA, essentially modified versions of Intralipid. However, these products have important mar
ket niches, and they have validated the concept of lipid-based drug delivery. In contrast,
“second generation” projects demonstrate increasing divergence from the original Intralipid
concept. By “second generation” we mean formulations that are currently being evaluated for
tolerance and efficacy in human clinical trials. Some examples are summarized in Table 3.
Trade names should be considered tentative and for identification only, as these are not ap
proved products.
Eltanolone (Fig. 6) is a steroid with hypnotic properties intended for induction and mainte
nance of intravenous general anesthesia. This endogenous metabolite of progesterone has no
hormonal action. The emulsion product contains 4 mg/mL pregnanolone in a 20% soybean
oil base with diacetylated monoglycerides as cosolubilizer and egg yolk phospholipids as
emulsifier. The clinical hypothesis being tested is that potential advantages over existing prod-
Table 3 Second Generation Injectable Lipid Drug Emulsions
Trade name Company Drug Activity Comments
Eltanolone Pharmacia & Upjohn Pregnanolone Anesthetic Similar to Diprivan

Imagent Alliance Pharm. Perflubron Imaging agent Liver and spleen
Intraiodol Pharmacia & Upjohn lodinated oil Imaging agent Liver and spleen
Kynacyte Sphinx/Eli Lilly Safingol PKC inhibitor Chemopotentiator
Oncosol HemaGen/PFC Perfluorocarbon Oxygen delivery Radiation sensitizer
Supercytes HemaGen/PFC Perfluorocarbon Oxygen delivery Heart surgery
None NCI Penclomedine Antineoplastic Cytotoxic drug
None PDT Pharmaceuticals Porphyrin derivative Antineoplastic Photodynamic therapy
Freedox Pharmacia & Upjohn Tirilazad Neuroprotective Stroke, CNS trauma
Fig. 6 Eltanolone. Also known as pregnanolone, this lipophilic anesthetic agent is intended for intrave
nous administration. Eltanolone is a nonhormonal, water-insoluble steroid derivative.
ucts will include absence of pain on injection, wide therapeutic margin, favorable cardiovascu
lar profile, minimal respiratory depression, and rapid recovery [35].
Imagent is an emulsion containing perflubron (perfluorooctylbromide), a perfluorocarbon-
type contrast agent developed for CT imaging of either lymph nodes (LN version) or blood
pool (BP version). The latter is designed to enhance the detection of small (<1 cm) tumors
in the liver and spleen [36]. Intraiodol is another emulsified contrast agent for liver imaging
that is now in clinical trials. It is composed of iodinated ethyl esters of fatty acids derived
from poppyseed oil, emulsified with egg yolk phospholipids plus 0.1% phenylalanine for
improved stability [37,38].
Kynacyte emulsion contains as its active ingredient safingol, a stereoisomer of dihydro-
sphingosine. This sphingolipid is an inhibitor of protein kinase C and a potential adjunctive
treatment for cancer. Preclinical studies have demonstrated that Kynacyte can enhance the
effectiveness of approved anticancer drugs such as doxorubicin or cisplatin without increasing
their toxicity [39].
Oncosol and Supercytes are emulsions that exploit the ability of perfluorocarbons to dis
solve 20 times more oxygen than water. Malignant tumors often have a subpopulation of
hypoxic cells that are radiation-resistant. By delivering oxygen to these tissues, Oncosol
should enhance the therapeutic effect of ionizing radiation. Supercytes will be used to promote
gas exchange during open heart surgery.
Penclomedine, a cytotoxic drug being developed by the National Cancer Institute, shows
significant activity against leukemia and mammary tumors in preclinical studies. Penclomed
ine is very lipophilic but is also heat-sensitive. It has been formulated into an Intralipid-like
emulsion that is prepared aseptically [7,17].
Another example of a novel emulsion for oncology is being developed by PDT Inc. A
proprietary lipophilic porphyrin derivative serves as a photosensitizer for photodynamic ther
apy and has been formulated in a lipid emulsion. Following intravenous administration, this
green porphyrin drug is selectively retained by tumor tissue. Irradiation by red laser light then
induces the formation of intracellular oxygen radicals, which destroy the host cells [40].
Pharmacia & Upjohn Company is developing a second generation emulsion delivery system
for tirilazad. This 21-aminosteroid antioxidant is being evaluated as an intravenous injectable
drug for the treatment of central nervous system trauma and ischemia [41,42].
F. Conclusions
Although the concept of lipid-based drug delivery has been discussed and evaluated for over
20 years, relatively few such products have actually been approved and marketed. However,
product development pipelines currently contain a large assortment of novel lipid emulsions
for drug administration as well as for the delivery of vaccines and diagnostic agents. It is
now apparent that the unique requirements of each drug and therapeutic regimen will require
customized emulsion technology if these new lipid-based carrier systems are to achieve their
true potential: clinically significant improvements in drug efficacy, preferably accompanied by
reductions in toxicity.
III. LIPID EMULSIONS FOR INTRAVENOUS NUTRITION

A. Historical Overview
Intravenous or parenteral nutrition (PN), the provision of a sterile solution of nutrients by
intravenous administration, has evolved in step with the acquisition of knowledge of chemis
try, physiology, and microbiology. Although the injection of substances into the bloodstream
dates back to the seventeenth century, it was not until the middle of the nineteenth century
that physicians recognized the therapeutic value of infusing large volumes of saline to counter
fluid losses incurred from cholera infections [43]. Subsequently, attempts to supplement nutri
ents intravenously were unsuccessful until the concepts of microbial infection became widely
accepted. Early this century, as principles of nutrition and metabolism became recognized,
attempts were made to replenish protein deficits parenterally. Initially this was hampered by
allergic complications, but in one of the first successful experiments of parenteral nutrition,
in 1915, Henriques and Anderson [44] maintained nitrogen equilibrium in goats for over 2
weeks by the infusion of a mixture of hydrolyzed protein together with glucose and salts.
Protein hydrolysates were subsequently shown to be a safe and effective source of nitrogen
for human beings.
In due course, numerous complicating and limiting metabolic and technical issues became
apparent when clinicians sought improvements in the parenteral delivery of nutrients. The
provision of basal energy and nitrogen was not always sufficient to address catabolic stress.
On the other hand, excess volume would result in pulmonary edema, and concentrating the
glucose solution for caloric support was not the answer, as this resulted in venous thrombo
phlebitis.
The infusion of lipids was an obvious answer to the question of how to effectively meet
energy needs. The high caloric density and low thrombogenic potential of lipid emulsions
made these substrates obvious candidates for PN. Murlin and Riche infused a lipid emulsion
into animals in 1915, and in 1920 Yamakawa was able to provide a lecithin-stabilized lipid
emulsion for human parenteral use [43]. In 1945, McKibbin et al. [45] advocated the provi
sion of calories from lipid emulsions, and shortly thereafter the first commercial preparations
based on castor and cottonseed oils became available.
The tendency of these early emulsions to form fat globules and the high incidence of
fevers, back pain, and liver dysfunction led to their disuse and to their eventual withdrawal
in the United States in 1964 [46]. Meanwhile, in Europe, Wretlind and colleagues were devel
oping an emulsion based on soybean oil stabilized with egg yolk phospholipids that was
shown in clinical trials to be safe as well as effective as a source of calories and essential
fatty acids. Supporting clinical results from the United States [47] led to FDA approval of
Intralipid and Liposyn in 1981. To date in excess of 100 million units of lipid emulsion have
been administered to patients as an integral part of PN.
B. Uses in Parenteral Nutrition

1. Composition
Lipid emulsions have three main components:
L Triacylglycerols (TAG), which make up the bulk of the emulsion particle, provide the
substrate for energy and the fatty acids that are essential for health. TAG are character
ized by their chain length:
a. Long-chain TAG (LCT) are provided by soybean or safflower seed oils, which
are rich in Ci6-Cig polyunsaturated fatty acids of the n-6 (or co-6) series.
b. LCT may also be of the n-3 (or o)-3) series, derived from fish oil.
c. Medium-chain TAG (MCT) are provided by coconut or palmkemel oil and consist
of fatty acids with chain length Cg-Cio-
2. Phospholipids (PL) derived from egg yolk act as the emulsifying agents.
3. Glycerol renders the solution isotonic.
0
II ,
CH, OC -R’
TriacylglycerQl O
I! 'I
C O CH
I
CH, OC -R'
li
O
R’, and R^ are saturated and unsaturated fatty acid residues
Linoleic Acid 18;2n-6)
o
H H H H II
C C= C C = C C C C - - OH
H3C c C
O
II
^ CH,OCR’
R^COCH
I
CKR'
R^ and R^ are saturated fatty acid residues

R^ is primarily either choline or the ethanolamine ester of phosphatidic acid
- 0^ P - 0CH,CH,N(CH3
)3 O - P -OCH3CH3NH3
I
0“ O'
Phosphatidylcholine Phosphatidylethanolamine
Fig. 7 Chemical and structural formulas of the main components of lipid emulsions.
Table 4 Com position o f Selected Commercial Lipid Emulsions
Oil (%) Fatty acid (%)
Name Soy Coconut/ lipid Glycerin Linoleic Oleic Palmitic Linolenic Steari Capric
(Manufacturer) bean Safflower palm (%) (%) (18:2, n-6) (18 :1, n-9) (16;0) (18:3, n-3) (18;0: (8:0) ( 10:0)
Intralipid 10% 10 — — 1.2 2.25 50 26 10 9 3.5

(Pharmacia
& Upjohn)
Intralipid 20% 20 — — 1.2 2.25 50 26 10 9 3.5
— — 2.25 50 26 9 3.5
Intralipid 30% 30 1.2 10
Liposyn II 10% 5 5 — 1.2 2.5 65.8 17.7 8.8 4.2 3.4
(Abbott)
Liposyn II 20% 10 10 — 1.2 2.5 65.8 17.7 8.8 4.2 3.4
Liposyn III 10% 10 — — 1.2 2.5 54.5 22.4 10.5 8.3 4.2
Liposyn III 20% 20 — — 1.2 2.5 54.5 22.4 10.5 8.3 4.2
— — 49-60 21-26 9-13 6-9 3-5
Nutrilipid 20% 20 1.2 2.21
(Kendall McGaw)
Lipofundin 20% 10 — 10 1.2 2.5 ~ 25 ~ 12 ~5 ~4 ~3 -35 45
(Braun)
The chemical and structural formulas of the main components of lipid emulsions are shown
in Fig. 7.
Commercially available lipid emulsions vary in their composition depending primarily on
the source and concentration of the TAG and PL. Likewise, the exact composition of the PL
will depend on the origin of the egg yolk source. The compositions of selected commercial
lipid emulsions are shown on Table 4.
Lipid emulsions were designed such that the particles resembled chylomicrons, which
could then be transported in the circulation and provide TAG fatty acids to different tissues.
Chylomicrons belong to a family of lipoprotein particles that are involved in the transport of
lipid in the circulation. The relative concentrations of constituents of a typical lipid emulsion
particle compared with that of native chylomicrons and other lipoproteins are shown in Fig.
8 [48].
The emulsion particle, like the chylomicron, has a diameter ranging from about 0.2 to 0.6
ptm and consists of a TAG core stabilized by a PL surface layer. In contrast, there are no choles
terol esters or apoproteins on the surface of lipid emulsion particles, and the overall concentra
tion of PL is typically greater. Differences in surface properties may reflect subtle changes in
the distribution, uptake, and metabolism of emulsion particles and native chylomicrons [49].
500 nm
Lipid Emulsion Particle
100 nm 20 nm 10 nm
5
33
40 22
VLDL HDL
Protein
Phospholipid
Cholesterol & Cholesterol Ester
TAG
Fig. 8 Relative concentrations of constituents of a typical lipid emulsion particle compared with that
of native chylomicrons and other lipoproteins.
Due to the relatively elevated PL concentration, lipid emulsions contain two particle popu
lations that can be separated by centrifugation. The low density TAG-rich population resem
bles chylomicrons. A higher density liposome-like population is rich in PL. The proportion
of TAG-rich particles to liposome-like particles depends on the proportion of TAG to PL in
the emulsion [49].
2. L ip id M etabolism
Metabolism o f Lipid Emulsion Particles. The traditional view is that, owing to their struc
tural similarity, lipid emulsions are metabolized like chylomicrons [49]. Acquisition of surface
C and E apoproteins allows the particle to interact with and activate lipoprotein lipase at the
capillary endothelial surface. In turn, TAG hydrolysis is initiated with the release of fatty
acids (FA) at the 1 and 3 positions and subsequent uptake of the resultant monoglyceride and
some of the FA into the adjacent tissues. The remaining FA remain in the circulation bound
to albumin.
As more TAG is hydrolyzed from the particle by peripheral lipoprotein lipase, the surface
apoproteins change, and as the particle shrinks in size it becomes enriched by cholesterol
esters. The remnant, now TAG-poor but cholesterol-rich, is further hydrolyzed by hepatic
lipase, preferentially taken up by hepatocytes and degraded. Other types of cells, notably
tissue macrophages, are capable of taking up emulsion particles, particularly when in the
remnant form.
The composition of the emulsion particle will influence the rate of hydrolysis and uptake,
and this is determined not only by the FA chain length but also by its relative position on the
glycerol backbone, with FA at position 2 having the most influence. LCT emulsions are
cleared at the same rate as chylomicrons [50, 51], whereas the clearance of MCT emulsions
is more rapid [52].
Recent animal research with labeled emulsion particles suggests that (1) the traditional
view may be incomplete in that, in contrast to chylomicrons, emulsion particles may be taken
up before substantial lipolysis has taken place and (2) a large proportion, more than 50%, of
this uptake occurs in extrahepatic tissues [53].
The amount of PL relative to TAG can influence the hydrolysis of lipid emulsions. Thus
20% emulsions, or 10% emulsions with a lower PL content than regular 10% emulsions, are
able to clear TAG more quickly [54].
Phospholipid-rich particles may interfere with the binding of TAG-rich particles to lipopro
tein lipase, thereby decreasing the rate of TAG hydrolysis as well as the binding of other
lipoprotein species to their receptors [54,55]. In this way and by their affinity to cholesterol,
these particles may influence intravascular lipid metabolism in ways that remain to be fully
elucidated.
Phosphatidylcholine, a component of PL in emulsion particles, has recently been promoted
as a source of choline. Choline is an essential component of membrane PL. It is a precursor
of the neurotransmitter acetylcholine and a source of methyl groups. Choline deficiency leads
to fatty liver and, in some animal models, to cancer [56].
Oxidation o f Fatty Acids. Free fatty acids are transported to tissues in the circulation
bound to albumin. After cellular uptake, the long-chain FA are taken up by mitochondria in
a carnitine-dependent manner and undergo /3 oxidation to acetyl CoA and hence to carbon
dioxide and water, via the citric acid cycle with the concomitant formation of high energy
ATP. Acetyl CoA formed in excess will be converted to ketone bodies (acetoacetate, P~
hydroxybutyrate), primarily in the liver and these may be exported to other organs to be used
as an alternative source of fuel. Medium-chain FA are metabolized in a different manner.
MCT are cleared faster than LCT from plasma. Medium-chain FA have a lower carnitine
requirement than long-chain FA and so are oxidized more rapidly with increased ketone body
formation. Reesterification of medium-chain FA is negligible, and as a result, administration
of MCT is less likely than LCT to result in tissue deposition of lipid [57]. High levels of
plasma medium-chain FA have been associated with neurological toxicity, and this occurs in
a dose-dependent manner [58]. In part to minimize this risk, MCT has been combined with
LCT in certain lipid emulsions (see Section III.C).
3. Intravenous N utrition
Lipid emulsions form an integral part of intravenous nutrition as a source of energy and fatty
acids essential to life. Energy from lipids is delivered to tissues in the form of free FA, which
are released from adipose tissue by lipolysis and distributed in the circulation bound to albu
min [59]. Fatty acids are the primary fuel source for resting skeletal muscle, heart, and liver.
In contrast, brain, skin, and elements of the blood are dependent on carbohydrates for energy
as they cannot readily oxidize FA.
Certain FA are termed “essential” in that they cannot be synthesized endogenously to meet
demand. Linoleic and linolenic acids are classic examples because a well-characterized essen
tial fatty acid deficiency (EFAD) syndrome results when they are omitted from the diet. This
occurs primarily as a result of changes in the structure of biological membranes of which
essential FA are important constituents. Eicosanoids are active metabolites of membrane-
bound EFA, and they play a significant role in orchestrating inflammatory responses. The
minimum requirement of EFA is 1-3% of dietary caloric intake [60].
Principles o f Administration, p a t i e n t s e l e c t i o n . Malnutrition leads to wasting and

functional impairment, and increasing severity of malnutrition is associated with increasing
morbidity and mortality. Adequate nutritional status can be achieved only by timely and ap
propriate nutritional support. Parenteral nutritional support is required to maintain nutritional
status or to correct status deficiencies when there is preexisting or anticipated malnutrition in
the setting of gut failure and/or the inability to receive adequate nutrients orally or enterally.
Typical conditions include short bowel syndrome, prolonged postoperative ileus, acute in
flammatory bowel disease, fistula, and chronic wasting diarrhea of AIDS.
Parenteral preparations consist of a mixture of solutions or

PAR EN TER A L FO R M U L A TIO N .
emulsions of amino acids, glucose, lipids, electrolytes, and other minerals, vitamins, trace
elements, and water. Certain drugs such as insulin or H2 blockers, compatible with the nutri
tion admixture, may also be added. These nutritional admixtures are tailored and formulated
to meet the precise nutritional needs of individual patients as well as some or all of their fluid
and electrolyte requirements.
Protein needs are determined on the basis of body weight and degree of metabolic stress
experienced by a patient. This can be further refined by ascertaining nitrogen balance, the
difference between intake and losses and assessing the needs accordingly. Fluid and electro
lyte needs are likewise determined on the basis of close monitoring of the clinical status.
Caloric requirements and protein needs during disease and catabolic stress are interrelated.
Inadequate caloric support from glucose and/or fat will result in the utilization of the amino
acids as a source of energy. Caloric needs are based on calculated or measured estimates of
caloric expenditure. In critical illness the aim is usually to provide sufficient calories to meet
expenditures. During recovery, extra calories may be administered to correct remaining
deficits.
Calories are typically provided from carbohydrate and lipid sources. This mixture is analo
gous to a mixed diet and is felt to be more physiological than the provision of a large excess
of one of the components. Dextrose solutions are available in a wide range of concentrations
and provide about 3.5 kcal/g dextrose. Lipid is available as 10, 20, and 30% emulsions
providing 1.1-3.0 kcal/mL. There is no clear recommendation as to the proportion of calories
that comes from carbohydrate and from lipid. Typically, 20-50% of the non-protein calories
are provided as lipid. (It is traditional to exclude parenteral protein from caloric content calcu
lations because protein is supplemented to meet protein rather than caloric needs. Ultimately,
though, the breakdown of exogenous protein will contribute to the energy pool.) Approxi
mately 100 g of glucose is required to meet the needs of the glucose-dependent central nervous
system, so some carbohydrate has to be provided. Excess glucose administration, on the other
hand, results in hyperglycemia, which is associated with osmotic diuresis, protein glycosyla
tion leading to immune dysfunction, and other impairments. Fatty deposition, particularly in
the liver, and excessive generation of carbon dioxide requiring greater respiratory effort will
also occur.
Parenteral nutrition admixtures are typically administered via a canula inserted into a cen
tral vein. The high blood flow and quick dispersion allow a concentrated admixture to be
delivered over 12-24 h each day for prolonged periods of time, enabling nutritional goals to
be met with a low risk of fluid overload.
ADMIXTURES. Total parenteral nutrition (TPN) is the provision of a complete nutritional ad
mixture adequate to support life. This may be formulated and infused as one admixture, which
typically contains protein in the form of amino acids, carbohydrate in the form of glucose,
and lipid in the form of soybean or safflower oil emulsion as well as micronutrients (electro
lytes and minerals, vitamins, and trace elements), in which case it is referred to as a 3-in-l
admixture. The stability of these complex admixtures is governed by the physicochemical
interactions of the constituents. The presence of a lipid emulsion can either stabilize or desta
bilize the admixture, depending on a number of factors. Because of this, some practitioners
prefer to administer the lipid independently of the rest of the admixture constituents. The lipid
emulsion is then either piggybacked onto the main TPN line via a Y-connector at the time of
infusion or administered independently via a peripheral vein. This is referred to as a 2-in-
1 admixture.
Admixtures have to be prepared aseptically using compatible types and amounts of ingredi
ents, and strict mixing rules must be followed to ensure that the mixture remains stable and
free of large particulate matter. Automatic compounders are available that can be prepro
grammed to deliver the correct proportion of ingredients in an appropriate sequence into a
plastic bag to ensure that a stable and sterile admixture is prepared. In the United States, the
compounding of admixtures is typically done immediately prior to their administration to
avoid changes in stability that may occur with storage. European regulations allow for a
storage period prior to administration. Increasingly, an in-line filter is used in the infusion of
parenteral solutions as added safety against particulate matter that may have formed during
the preparation stage as a result of incorrect mixing or storage or incompatible ingredients.
PERIPHERAL ADMINISTRATION. Parenteral nutrition may be delivered by infusion via a pe

ripheral vein. Peripheral parenteral nutrition (PPN) is indicated for patients with a nonfunc
tioning gut who are not in catabolic stress and who require nutritional support for less than 2
weeks. The composition of the PPN admixture differs from that of TPN administered though
a central line. The main complication of PPN is phlebitis, which can be minimized by infusing
a solution with a low osmolarity. To achieve this, the proportion of lipid emulsion is increased
to approximately 70% of non-protein calories, the relative proportions of glucose and amino
acids are decreased, and the mixture is further diluted with water for injection.
Complications Associated with Lipid Emulsion Administration. Intravenous administra

tion of lipid emulsions is well tolerated, and the incidence of adverse events is less than 0.1%.
Most complications arise when the rate of administration exceeds the rate of TAG clearance
and hypertriglyceridemia ensues. For this reason lipid emulsions are contraindicated in the
rare individuals with genetic or acquired conditions that predispose them to hyperlipemia.
Abnormality in liver function is one of the more common adverse events reported following
TPN administration, and lipid emulsions have been implicated as part of this process, particu
larly in patients requiring prolonged support. Typically, fatty infiltration of the liver occurs
when lipids are absent or are provided in excess of need [61]. Deficiencies of essential fatty
acids [62] and of choline [56] as well as an excess of phospholipid-rich particles have been
associated with liver disease.
Respiratory difficulties have been reported in some patients infused with lipid emulsions.
This has been blamed on eicosanoids and other metabolites of FA acting via a vasoactive
effect on pulmonary vasculature [63]. Lipid accumulation in lung tissue macrophages and
changes in oxygen-diffusing capacity of red blood cells have also been implicated in the
process [64]. Whether or not lipids or other factors are involved in pulmonary compromise
remains controversial. Patients with pulmonary insufficiency may actually benefit from lipids
as a source of energy because fat oxidation is associated with a lower rate of carbon dioxide
production (and so a decreased need for elimination) than carbohydrate oxidation. If there is
a negative effect of lipid administration on the lungs, it most likely only occurs with excessive
dosing [65].
By accumulating in macrophages and other cells involved in the immune response and by
modulating the inflammatory process through the elaboration of active metabolites such as
eicosanoids, lipid emulsions consisting predominantly of LCT have also been associated with
impairment of the immune system. These conclusions were drawn primarily from animal and
early human studies. A review of more recent human trials proposes that other factors oc
curring concomitantly could play a more important role in immune dysfunction in patients
sick enough to require TPN and that there is actually no evidence of a detrimental effect of
lipids on the immune system [66].
Conflicting results have been published on the effect of lipid emulsions on the ability of
blood to clot. However, it appears that in the absence of significant hyperlipemia there is no
effect on platelet function or on the clotting cascade [67].
Polyunsaturated fatty acids (PUFA) are susceptible to damage by reactive oxygen species
and may participate in a chain reaction process of lipid peroxidation [68]. Very long chain
FA are particularly susceptible. When this process occurs in cell membrane phospholipids,
cellular damage may result. To minimize this, micronutrients (vitamins A, C, and E and
selenium) and enzymes (catalase, superoxide dismutase) act as antioxidants. Whether or not
the administration of lipid emulsions containing PUFA contributes to the overall oxidative
stress of sick patients with additional detrimental health effects is not known.
C. Future Trends
For over a decade, attempts have been made to manipulate the composition of lipid emulsions
to improve disease outcomes without compromising the ability of the emulsions to provide a
concentrated source of energy and adequate essential fatty acids to meet needs.
An early approach was to combine medium-chain and long-chain triacylglycerols. The aim
(as detailed above) was to benefit from the specific properties of MCT, such as faster clear
ance from the bloodstream, more rapid oxidation, a lesser tendency to tissue incorporation,
and improved protein sparing compared to LCT provided isocalorically [69,70]. The outcome
was an emulsion composed of a 1:1 physical mixture of MCT and LCT (Table 4). This
product has been marketed in Europe but is not currently available in the United States. A
second mixture with a higher concentration of MCT than LCT (3:1) has been submitted for
U.S. approval. It is claimed that the higher proportion of MCT makes the emulsion better
suited for critically ill patients with an enhanced need for more rapid lipid fuel utilization
[71]. Whether or not this is the case has yet to be confirmed. Nevertheless, it has been
demonstrated not only that the rate of release of total free FA by emulsion hydrolysis is
greater for MCT than for LCT but that it also depends on the amount of MCT in the mix
ture [72].
Structured triglycerides are an alternative to the mixture of MCT and LCT. These triacyl-
glycerols are the products of MCT and LCT oils that have been hydrolyzed and reesterified
with random redistribution such that any one glycerol backbone contains a mixture of both
medium-chain and long-chain FA [73]. Studies in animals and human beings have indicated
that structured triglycerides may have clinical advantages over the MCT-LCT mixes [74].
The rate of hydrolysis of structured triglycerides is moderated with a less pronounced rise in
plasma medium-chain FA and therefore in the subsequent risk of ketoacidosis and adverse
neurological effects [24]. Overall, it would appear, for reasons that are not completely under
stood, that the administration of structured triglycerides, compared with that of LCT or a
mixture of LCT and MCT, is associated with an improved protein metabolic profile [75,76].
It must be noted, however, that large, well-controlled comparative trials have yet to be pub
lished, so definitive conclusions on the relative merit of one type of emulsion versus another
cannot be drawn at this time.
Attempts at defining the optimal proportion of essential FA to nonessential FA, of monoun-
saturated FA to polyunsaturated FA, and of n-6 series FA to n-3 series FA is ongoing, as is
the role of long-chain (> C 2o) PUFA and the benefits of supplementation of antioxidant rela
tive to the content of PUFA. The physiological rationales for these approaches vary. The most
notable is the attempt to modify the eicosanoid profile by changing the ratio of n -3 FA to n -6
FA in the lipid emulsion. Eicosanoids are important in moderating communication between
cells and modulating the inflammatory response that coexists with disease. Metabolites of the
n-6 FA series have been more closely associated with inflammation and immunosuppression
than those of n-3 series FA, and this has been the rationale for inclusion of fish oil rich in n-
3 series FA into lipid emulsions [77,78]. Unfortunately, convincing clinical trials have yet to
confirm the early promising animal and human studies [79].
For a new emulsion to become widely accepted into medical practice, it will have to
demonstrate, in a cost-effective manner, that its use results in a positive clinical outcome.
Adequate and well-controlled trials will have to be designed and carried out in patients with
a broad spectrum of diseases sharing a need for lipids. This is a formidable task given the
heterogeneity of the patient population, the large sample size needed, the difficulties in defin
ing accepted clinically relevant and economically appropriate endpoints, and the wide variety
of possible types and concentrations of FA and TAG that are available to be tested.
ABBREVIATIONS
CNS central nervous system
CT computed tomography
EDTA ethylenediaminetetraacetic acid
EFA essential fatty acid
EFAD essential fatty acid deficiency
FA fatty acid
LCT long-chain triacylglycerols
MCT medium-chain triacylglycerols
PKC protein kinase C
PL phospholipids
PN parenteral nutrition
PPN peripheral parenteral nutrition
PUFA polyunsaturated fatty acid
RES reticuloendothelial system
TPN total parenteral nutrition
REFERENCES
1. S. S. Davis, J. Hadgraft, and K. Palin, Medical and pharmaceutical applications of emulsions, in
Encyclopedia of Emulsion Technology (P. Becher, Ed.), Marcel Dekker, New York, 1985, VoL
2, pp. 159-238.
R. L. Juliano, Factors affecting the clearance kinetics and tissue distribution of liposomes, micro
spheres, and emulsions, Adv. Drug Deliv. Rev. 2: 31-54 (1988).
P. K. Hansrani, S. S. Davis, and M. J. Groves, The preparation and properties of sterile intrave
nous emulsions, J. Parent Sci. Technol. 37: 145-150 (1983).
M. Singh and L. J. Ravin, Parenteral emulsions as drug carrier systems, J. Parenter. Sci. Technol.
40: 34-41 (1986).
S. S. Davis, C. Washington, P. West, L. Ilium, G. Liversidge, L. Stemson, and R. Kirsh, Lipid
emulsions as drug delivery systems, Ann. N.Y. Acad. Sci. 507: 75-88 (1987).
L. C. Gold, R. T. Lyons, and L. C. Bartholow, Parenteral emulsions for drug delivery, Adv.
Drug Deliv. Rev. 5: 189-208 (1990).
R. J. Prankerd and V. J. Stella, The use of oil-in-water emulsions for parenteral drug administra
tion, J. Parenteral Sci. Technol. 44: 139-149 (1990).
J. T. Rubino, The influence of charged lipids on the flocculation and coalescence of oil-in-water
emulsions, J. Parenter. Sci. Technol. 44: 210-215, 247-252 (1990).
G. Zografi, H. Schott, and J. Swarbrick, Disperse systems, in Remington’s Pharmaceutical Sci
ences (A. Gennaro, Ed.), Mack, Easton, PA, 1990, Chap. 19, pp. 298-309.
10. B. J. Carroll, The stability of emulsions and mechanisms of emulsion breakdown, in Surface and
Colloid Science (E. Matijevic, Ed.), Wiley, New York, 1976, pp. 1-67.
11. S. S. Davis, L. Ilium, P. West, and M. Galloway, Studies on the fate of fat emulsions following
intravenous administration to rabbits and the effect of added electrolyte, Clin. Nutri. 6: 13-19
(1987).
12. I. Hakansson, Physico-chemical changes in artificial fat emulsions during storage, Acta Chem.
Scand. 20: 2267-2281 (1966).
13. I. M. A. Kemps and D. J. A. Crommelin, Chemical stability of phospholipids in pharmaceutical
preparations. I. Hydrolysis of phospholipids in aqueous solutions; II. Peroxidation of phospholipids
in aqueous solutions, Pharm. Weekbl. 123: 355-363, 457-469 (1988).
14. S. S. Davis, Eur. Patent EP 456106A (1991).
15. W. R. Perkins, S. R. Minchey, L. T. Boni, C. E. Swenson, M. C. Popescu, R. F. Pasternack,
and A. S. Janoff, Amphotericin B-phospholipid interactions responsible for reduced mammalian
cell toxicity, Biochim. Biophys. Acta 1107: 271-282 (1992).
16. M. R. Lance, C. Washington, and S. S. Davis, Structure and toxicity of amphotericin B/triglycer-
ide emulsion formulations, J. Antimicrob. Chemother. 36: 119-128 (1995).
17. R. J. Prankerd, S. G. Frank, and V. J. Stella, Preliminary development and evaluation of a
parenteral emulsion formulation of penclomedine: A novel, practically water insoluble cytotoxic
agent, J. Parenter. Sci. Technol. 42: 76-81 (1988).
18. N. Mackenzie and I. S. Grant, Comparison of the new emulsion formulation of propofol with
methohexitone and thiopentone for induction of anaesthesia in day cases, Br. J. Anaesth. 57: 725-
731 (1985).
19. J. Dundee and R. Clarke, Propofol, Eur. J. Anaesth. 6: 5-22 (1989).
20. M. Paborji, C. Riley, and V. Stella, A novel use of intralipid for the parenteral delivery of perilla
ketone, an investigational cytotoxic drug with a high affinity for plastic, Int. J. Pharm. 42: 243-
249 (1988).
21. R. Jeppsson and S. Rossner, The influence of emulsifying agents and of lipid soluble drugs on the
fractional removal rate of lipid emulsions from the blood stream of the rabbit, Acta Pharmacol.
Toxicol. 37: 134-144 (1975).
22. W. R. Burnham, P. K. Hansrani, C. E. Knott, J. A. Cook, and S. S. Davis, Stability of a fat
emulsion-based intravenous feeding mixture, Int. J. Pharm. 13: 9 (1983).
23. K. Y. Warshawsky, Intravenous fat emulsions in clinical practice, Nutr. Clin. Pract. 7: 187-
196 (1992).
24. J. M. Miles, M. Cattalini, F. W. Sharbrough, L. E. Wold, et al.. Metabolic and neurologic
effects of an intravenous medium-chain triglyceride emulsion, J. Parenter. Enteral Nutr. 15: 37-
41 (1991).
25. D. L. Seidner, E. A. Mascioli, N. W. Istfan, et al.. Effects of long-chain triglyceride emulsions
on reticuloendothelial system function in humans, J. Parenter. Enteral Nutr. 13: 614-619 (1989).
26. D. Johansson, B. Bergenstahl, and E. Lundgren, Water-in-triglyceride emulsions. Effect of fat
crystals on stability, J. Am. Oil Chem. Soc. 72: 939-949 (1995).
27. M. Groves, Parenteral Technology Manual, 2nd ed., Interpharm, Prairie View, IL, 1988, pp.
159-179.
28. E. Dickinson, Emulsions and droplet size control, in Controlled Particle, Droplet, and B ubble
Formation (D. J. Wedlock, Ed.), Butterworth-Heinemann, Oxford, UK, 1994, pp. 190-216.
29. S. Benita and M. Y. Levy, Submicron emulsions as colloidal drug carriers for intravenous admin
istration: comprehensive physicochemical characterization, J. Pharm. Sei. 82: 1069-1079 (1993).
30. R. Jeppsson and S. Ljungberg, Anticonvulsant activity in mice of diazepam in an emulsion formu
lation for intravenous administration, Acta Pharmacol. Toxicol. 36: 312-320 (1975).
31. R. Jeppsson, Plasma levels of diazepam in the dog and the rabbit after two different injection
formulations, emulsion and solution, J. Clin. Pharm. 1: 181-187 (1976).
32. K. Tremper, Perfluorochemical Oxygen Transport, International Anesthesiology Clinics, Vol 23,
Little, Brown, Boston, MA, 1985.
33. H. H. Wu, G. W. Dombi, and R. L. Poland, Parenteral delivery of vitamins with lipid emulsion,
Pediatr. Res. 27 (Part 2): 293A (1990).
34. Y. Mizushimna, A. Yanagawa, and K. Hoshi, Prostaglandin Ej is more effective when incorpo
rated in lipid microspheres for treatment of peripheral vascular diseases in man, J. Pharm. Phar
macol. 35: 666-667 (1983).
35. S. Hogskilde, J. W. Nielsen, P. Carl, and M. B. Sorensen, Pregnanolone emulsion. Anaesthesia
42: 586-590, 1045-1050 (1987).
36. Alliance Pharmaceutical Corp., San Diego, CA, Annual Report, 1993.
37. K. Ivancev, A. Lunderquist, R. McCuskey, P. McCuskey, and A. Wretland, Experimental investi
gation of a new iodinated lipid emulsion for computed tomography of the liver, Acta Radiol. 30:
1-7 (1989).
38. K. Ivancev, A. Lunderquist, R. McCuskey, P. McCuskey, and A. Wretland, Effect of intrave
nously injected iodinated lipid emulsion on the liver, Acta Radiol. 30: 291-298 (1989).
39. Sphinx Pharmaceuticals Corp., Durham, NC, Annual Report, 1992.
40. PDT Inc., Santa Barbara, CA, Photodynamic Therapy Report, 1993.
41. J. C. Fleishaker, G. R. Peters, K. S. Cathcart, and R. C. Steenwyk, Evaluation of the pharmacoki
netics and tolerability of tirilazad mesylate, a 21-aminosteroid free radical scavenger, J. Clin.
Pharm. 33: 182-190 (1993).
42. E. D. Hall, The role of oxygen radicals in traumatic injury: clinical implications, J. Emerg. Med.
11: 31-36 (1993).
43. A. W. Wilkinson, Historical background of intravenous feeding, Nutr. Dieta 5: 295-297 (1963).
44. V. Henriques and A. C. Andersen, Über parenterale Ernährung durch intravenöse Injektion,
Hoppe-Seylers Z. Physiol. Chem. 88: 357-369 (1913).
45. J. M. McKibbin, A. Pope, S. Thayer, R. M. Ferry, Jr., and F. J. Stare, Parenteral nutrition:
studies on fat emulsions for intravenous alimentation, J. Lab. Clin. Med. 30: 488-497 (1945).
46. H. L. Lehr, O. Rosenthal, H. M. Rawnsley, J. E. Rhoads, and M. B. Sen, Clinical experience

with intravenous fat emulsions. Metabolism 6: 666-672 (1957).
47. L. M. Hansen, W. R. Hardie, and J. Hidalgo, Fat emulsion for intravenous administration: clinical
experience with intralipid 10%, Ann. Surg. 184: 80-88 (1976).
48. E. B. Feldman, Nutrition and diet in the management of hyperlipidemia and atherosclerosis, in
Modern Nutrition in Health and Disease (M. E. Shils, J. A. Olson, and M. Shike, Eds.), Lea and
Febiger, (Philadelphia, 1994).
49. Y. A. Carpentier, Intravascular metabolism of fat emulsions: the Arvid Wretlind Lecture, ESPEN,
Clin. Nutr. 8: 115-125 (1988).
50. D. Hallberg, Elimination of exogenous lipids from the blood stream. An experimental methodolog
ical and clinical study in dog and man, Acta Physiol. Scand. 65 (Suppl. 259): 1 (1965).
51. D. Hallberg and J. Wersall, Electron-microscopic evaluation of chylomicrons and fat emulsions
for intravenous use, Acta Chir. Scand. 325 (Suppl): 23 (1964).
52. M. J. Ball and K. White, Metabolic effects of intravenous medium- and long-chain triacylglycerols
in critically ill patients, Clin. Sci. 76: 165-170 (1989).
53. M. Hultin, C. Cameheim, K. Rosenqvist, and T. Olivecrona, Intravenous lipid emulsions, re
moval mechanisms as compared to chylomicrons, Clin. Nutr. 14: 2 (1995).
54. Y. A. Carpentier, M. Richelle, and J. Bury, Phospholipid excess of fat emulsion slows triglyceride
removal and increases lipoprotein remodelling. Arteriosclerosis 7: 541a (1987).
55. C. Rojas, T. Olivecrona, and G. Bengtsson-Olivecrona, Comparison of the action of lipoprotein
lipase on triacylglycerols and phospholipids when presented in mixed liposomes or in emulsion
droplets, Eur. J. Biochem. 197: 315-321 (1991).
56. S. H. Zeisel, Phosphatidylcholine: the unappreciated nutrient in total parenteral nutrition, in Organ
Metabolism and Nutrition (J. M. Kinney and H. N. Tucker, Eds.), Raven Press, New York
City, 1994.
57. N. Sato, R. J. Deckelbaum, G. Neeser, Y. A. Carpentier, and J. M. Kinney, Hydrolysis of mixed
lipid emulsions containing medium-chain and long-chain triacylglycerol with lipoprotein lipase in
plasma-like medium, JPEN 18: 112-118 (1994).
58. J. M. Miles, M. Cattalini, F. W. Sharbrough, L. E. Wold, R. E. Wharen, J. E. Gerich, and M.
W. Haymond, Metabolic and neurologic effects of an intravenous medium-chain triglyceride emul
sion, JPEN 15: 37-41 (1991).
59. S. W. Coppack, M. D. Jensen, and J. M. Miles, The in vivo regulation of lipolysis in humans,
/. Lipid Res. 35: 177-193 (1994).
60. R. T. Holman and S. B. Johnson, Essential fatty acid deficiency in man, in Dietary Fats and
Health (E. G. Perkins and W. J. Visek, Eds.), Am. Oil Chem. Soc., Champaign, IL, 1983.
61. I. Tulikoura and K. Huikuri, Morphological fatty acid changes and function of the liver, serum
free fatty acids and triglycerides during parenteral nutrition, Scand. J. Gastroenterol. 17: 177-
185 (1982).
62. N. L. Keim and J. A. Mares-Perlman, Development of hepatic steatosis and essential fatty acid
deficiency in rats with hypercaloric, fat free parenteral nutrition, J. Nutr. 114: 1807-1815 (1984).
63. T.-L. Hwang, S.-L. Huanng, and M.-F. Chen, Effects of intravenous fat emulsion on respiratory
failure. Chest 97: 935-938 (1990).
64. D. B. Allardyce, The postmortem interval as a factor in fat embolism. Arch. Pathol. 92: 248-
253 (1971).
65. D. Haumont, C. Rossle, and A. Cercx, Modifications of surfactant phospholipid pattern in prema
ture infants treated with Curosurf: clinical and nutritional correlations, Biol. Neonate 61 (Suppl.):
37-43 (1992).
6 6 . J. Palmblad, Intravenous lipid emulsions and host defense— a critical review, Clin. Nutr. 10: 303-
308 (1991).
67. V. C. Herson, C. Block, and L. Eisenfeld, Effect of intravenous fat emulsion on neonatal neutro
phil and platelet function, JPEN 13: 620-622 (1989).
6 8 . M. Logani and R. Davies, Lipid oxidation. Biologic effects and antioxidants: a review. Lipids 15:
485-495 (1980).
69. A. C. Bach, D. Storck, and Z. Meraihi, Medium-chain triglyceride based fat emulsions: an alter
native energy supply in stress and sepsis, JPEN 12: 835-845 (1988).
70. Z. Jiang, S. Zhang, and X. Wang, A comparison of medium-chain and long-chain triglycerides in
surgical patients, Ann. Surg. 217: 175-184 (1993).
71. M. Jeevanandam, N. J. Holaday, T. Voss, R. Buier, and S. R. Petersen, Efficacy of a mixture of
medium-chain triglyceride (75%) and long-chain triglyceride (25%) fat emulsion in the nutritional
management of multiple-trauma patients. Nutrition 11: 275-284 (1995).
72. N. Sato, R. J. Deckelbaum, G. Neeser, Y. A. Carpentier, and J. M. Kinney, Hydrolysis of mixed
lipid emulsions containing medium-chain and long-chain triacylglycerol with lipoprotein lipase in
plasma-like medium, JPEN 18: 112-118 (1994).
73. R. Sandstrom, A. Hyltander, U. Komer, and K. Lundholm, Structured triglycerides to postopera
tive patients: a safety and tolerance study, JPEN 17: 153-157 (1993).
74. A. Hyltander, R. Sandstrom, and K. Lundholm, Metabolic effects of structured triglycerides,
Nutrition in Clinical Practice 10: 91-97 (1995).
75. K. Magnusson, G. Sandberg, and H. Ronnholm, Effect of structured triglycerides on protein
synthesis in liver and muscle tissue during TPN in an injured animal model, Clin. Nutr. 5 (SuppL):
141 (1986).
76. K. T. Mok, A. Maiz, and K. Yamazaki, Structured medium-chain and long-chain triglyceride
emulsions are superior to physical mixtures in sparing body protein in the burned rat. Metabolism
33: 910-915 (1984).
77. M. M. Gottschlich, Selection of optimal lipid sources in enteral and parenteral nutrition. Nutrition
in Clinical Practice 7: 152-165 (1992).
78. C. H. Simoens, M. Richelle, C. Rossle, M. Derluyn, R. J. Deckelbaum, and Y. A. Carpentier,
Manipulation of tissue fatty acid profile by intravenous lipids in dogs, Clin. Nutr. 14: 177-185
(1995).
79. M. Katan, Fish and heart disease (editorial), N. Engl. J. Med. 332: 1024-1025 (1995).
22________________________
Th e Role of L ip id s in Animal Feeds
Julian Wiseman and Phil C. Garnsworthy

University o f Nottingham, Sutton Bonington Campus, Loughborough, England
I. INTRODUCTION
Modem high performing genotypes of farm livestock (particularly pigs, poultry, and dairy
cattle) have greater nutrient and dietary energy requirements than former strains. Fats and oils
have approximately twice the dietary energy-yielding capacity of carbohydrates. They may
contain essential fatty acids and fat-soluble vitamins, their physical texture reduces dust in
feed mills, and they promote palatability of diets (which is of particular importance in the pet
food industry, where diet acceptance is of major importance in marketing and the value of
fats and oils as energy-yielding ingredients is perhaps of less concern). These materials, ac
cordingly, have assumed considerable importance as ingredients in compound animal feeds in
recent years although it is perhaps pertinent to mention that, with the exception of essential
fatty acids, there is no obligatory requirement to include fats and oils in diets. Rather, it is
their ability to provide dietary energy that dictates whether they are employed.
It is the purpose of this review to consider the role of fats and oils in the context of their
nutritional value and their influence on the quality of animal products (for example, carcass
and milk quality). Decisions on whether fats and oils will be incorporated into animal diets
will also be based on other factors including their cost and the presence of appropriate milling
technology, which often limits the actual amounts that may be used. Thus fats and oils employed
in animal feeding are frequently blends of a number of individual ingredients providing a final
mixture with a melting point in the region of 40-50°C. Storage of blends is usually within this
temperature range, requiring specialized facilities (which should be stainless steel or polymer-
based) and they are usually added in the liquid state to other dietary ingredients. Most compound
feeds are pelleted subsequent to mixing, and this process is not effective if added fat levels are
excessive (beyond approximately 40 g fat/kg diet). Addition of further amounts of fat, which is
frequently cost-effective, is through sophisticated liquid fat spraying equipment after pelleting.
To overcome these technological problems, so-called dry fat products are becoming in
creasingly available; these include fat and oil blends absorbed onto a solid carrier, spray-dried
557
558 Wiseman and Garnsworthy
products, and calcium soaps of fatty acids (of value only to ruminants). Finally, the use of
oilseeds prior to oil extraction (e.g., soybeans) is a well-established means of adding oil to
animal diets, a process that has the additional advantage of simultaneously providing other
nutrients (e.g., amino acids).
The nutritional value of any dietary raw material is governed primarily by its chemical
composition (i.e., its ability to provide energy-yielding compounds and specific nutrients) and
the degree to which it is digested by the animal. In terms of the latter, nonruminants and
ruminants are very different and accordingly are considered separately in terms of both nutri
tional value and product quality.
II. UTILIZATION OF FATS AND OILS BY NONRUMINANTS

The detailed chemical structure of fats and oils together with their major sources have been
considered earlier in this book (Chapters 1 and 2). In addition, the various processes to which
fats and oils are subjected have been described (Chapters 4-9). The relevance of this to the
nutrition of nonruminants is that fats and oils available for incorporation into animal diets are
of diverse origin and chemical composition—the animal feed industry is frequently the recipi
ent of by-products from other processes. A wide range of commodities are thus available,
including crude vegetable oils, soapstocks, hydrogenated materials, rendered animal tallows,
recovered vegetable oils from human food production, and even products from the refining of
oils such as bleaching earths, which are sometimes available as dry fat products.
Although all these materials (with the exception of the last category) probably have similar
total fat contents approaching 1000 g/kg, there is considerable variability in chemical compo
sition, which has a pronounced influence on their digestibility. This topic has been the subject
of a number of previous reviews, which have considered both the physiological basis for the
digestion and absorption of fats and the factors that are responsible for the large differences
in their subsequent dietary energy values [e.g., 1-4].
A. Influence of Chemical Structure of Fats and Oils on

Digestibility and Dietary Energy Value
It is customary in most countries to express dietary energy values of compound diets or raw
materials for poultry and pigs in terms of their apparent metabolizable energy (AME) and
digestible energy (DE) value, respectively. The SI units for AME and DE are MJ/kg, although
the dietary calorie (kcal or Cal) is still found, as is the pound (lb) as the unit of mass.
Descriptions of the detailed methodologies employed in deriving AME and DE values are
outside the scope of this chapter but are considered elsewhere [5].
There is considerable confusion relating to the AME and DE values of fats, primarily
because descriptions of the products evaluated are confined to names and origins with no
accompanying chemical characterization of a more precise nature. Systematic studies on the
influence of chemical composition of fats and oils on dietary energy values for nonruminants,
in terms of the quantitative contribution of the variables involved, are limited. Thus tables of
nutritional value of fats and oils are limited to descriptions such as “beef tallow” and “vegeta
ble oil.” The need for greater precision in defining source materials in a chemical rather than
descriptive manner is of particular importance to animal feeding as fats and oils are rarely fed
as a single commodity but rather as blends of a variety of materials. The two variables of
most importance are the degree of saturation and content of free fatty acids, with chain length
of the constituent fatty acids being of secondary concern.
Role of Lipids in Animal Feeds 559
1. D egree o f Saturation and Free F atty A cid Content

Physiologically, the mechanisms of fat digestion and absorption in nonruminants are well
documented [2]. The major site of fat digestion in pigs and poultry is the duodenum, and
basically it consists of emulsification of dietary fat by conjugated bile salts, followed by
hydrolysis of triglycerides by pancreatic lipase into mixtures consisting essentially of 2-mono-
glycerides and free fatty acids. The subsequent absorbability of these products is dependent
on their solubility in bile salt micelles. Polar solutes are more readily incorporated into mi
celles, and this explains the relatively higher absorbabilities of unsaturated fatty acids com
pared to saturated fatty acids and the well-established observation that unsaturated fatty acids
are more digestible than those that are saturated [e.g., 6]. Accordingly, oils have a higher
dietary energy value than the more saturated fats; this also explains why hydrogenation of oils
(even partial) is associated with a reduction in dietary energy value.
However, this is an overly qualitative description, and to be of any value in assigning
dietary energy values to the wide range of commodities available, a quantitative measurement
of the degree of saturation is essential. Although reference has been made to the digestion of
individual fatty acids, some measure of the overall fat or oil is more important. What has
been used [5,7] is the ratio of unsaturated (U) to saturated (S) fatty acids (giving U/S). In
creasing the degree of unsaturation of a fat by mixing a saturated fat with an unsaturated oil,
which is associated with higher U/S ratios, appears to be associated with a nonlinear improve
ment in dietary energy value [7].
The relative superiority of an intact triglyceride compared to hydrolyzed fat in terms of
dietary energy value is also well known [e.g., 8,9]. Increasing the proportion of FFA would
appear to be associated with a linear reduction in fat digestibility [1]. The FFA content is
expressed in terms of grams per kilogram of fat.
Systematic studies on the influence of U/S ratio and FFA content of fats and oils on AME
and DE values have recently been concluded [10-14]. The approach adopted was to select
raw materials of known U/S and FFA content that covered a range sufficiently wide that they
represented commodities that would be employed in the feeding of nonruminants, and all
blends were evaluated for AME and DE content. Chain length of constituent fatty acids was
predominantly in the range of 16-20 carbon atoms. Data generated confirmed that the re
sponse of AME and DE to the U/S ratio was curvilinear (Fig. 1), with the greatest improve-
Fig. 1 Response of dietary energy of fats and oils to increasing degree of unsaturation, as measured
by the ratio of unsaturated (U) to saturated (S) fatty acids (U/S).
FFA (g/kg fat)
Fig. 2 Response of dietary energy of fats and oils to increasing free fatty acid content.
ment in dietary energy value occurring over the lower range of increase in U/S. The response
of dietary energy to FFA was linear (Fig. 2).
Nutritional evaluation through biological experimentation is, of necessity, a lengthy and
costly procedure. Accordingly there have been many studies attempting to predict dietary
energy value from chemical composition. Having identified the two major chemical compo
nents that have nutritional relevance and generated AME and DE values, regression analysis
Table 1 Prediction Equations Relating the Apparent Metabolizable

Energy (AME, MJ/kg—Poultry) and Digestible Energy (DE, MJ/kg—
Pigs) to Ratio of Unsaturated to Saturated Fatty Acids (U/S) and Free
Fatty Acid Content (FFA, g/kg Fat) of Fats and Oils^’*"
AME or DE (MJ/kg fat) =A + B'X FFA + C x e',(Z) X U /S )
a. Poultry (AME)
Age 1 Age 2
Constant (PV = 0.816) (PV = 0.925)
A 3 8 .1 12 ± 1.4 18 39.025 ± 0.557

B -0.009 ±0.002 -0.006 ±0.001
C -15 .3 3 7 ± 2 .6 3 6 -8 .50 5 ±0.746
D -0 .5 0 6 ± 0 .18 6 -0.403 ±0.088
b. Pigs (DE)
Age 1 Age 2
Constant (PV = 0.802) (PV = 0.768)
A 37.890 ± 1.69 0 36.898 ± 0 .50 1

B -0.005 ±0.002 -0.005 ± 0.001
C -8 .2 0 0 ± 1.7 5 0 -7 .3 3 0 ± 2 .7 0 0
D -0 .5 15 ± 0 .3 7 6 -0.906 ±0.452
"Âge 1 and 2 in poultry refer to 1.5 and 7.5 weeks of age, respectively, and to
1 0 -2 0 kg and 3 5 -8 5 kg live weight in pigs, respectively.
^PV = proportion in variance of dependent variables accounted for by function.
Fig. 3 Prediction of the apparent metabolizable energy (AME) of fats for poultry as influenced by
age, ratio of unsaturated (U) to saturated (S) fatty acids (U/S), and FFA content (employing two levels—
100 and 400 g/kg fat). Young and old birds aged 1.5 and 7.5 weeks.
Table 2 Mean Chemical Analyses of Coconut-

Palmkemel Oil Blend (CP) and Its Acid Oil
(CPAO)—All Data Expressed as g/kg Oil
Fatty acid
profile^ CP CPAO
8:0 95 42
10:0 71 41
12:0 494 462
14:0 171 170
16:0 78 101
18:0 24 26
16:1 0 0
18:1 53 129
18:2 12 21
18:3 0 0
20+ 1 7
U/S^ 0.07" 0.19

2 .66^ 2.36
8.78" 6.86
Free fatty acids
(g/kg fat) 13.8 839.0
^Notation indicates length of carbon chain followed by

number of double bonds.
^Ratio of unsaturated (U) to saturated (S) fatty acids.
Calculation of U/S ratio:
Âll saturated fatty acids included in “saturated”
fraction.
Ônly 14:0, 16:0, and 18:0 appearing in “saturated”
fraction.
Ônly 16:0 and 18:0 appearing in “saturated” fraction.
Rg. 4 Prediction of the apparent metabolizable energy (AME) of fats for poultry as influenced by
ratio of unsaturated (U) to saturated (S) fatty acids (U/S) using the linear responses ratios described in
Table 2 and FFA content compared with data (as indicated by the *) generated for mixtures of coconut
or palmkemel oil and the respective acid oil.
was undertaken to relate dietary energy value (dependent variable) to both U/S and FFA
content (independent variables).
Those functions combining both independent variables are presented in Table 1. Separate
functions were derived for both “young” and “old” animals, it having been established that
the ability to digest fats improves with age [15-17]. As an example of the influence of both
U/S and FFA together with age, solutions to functions, for poultry, are presented in Fig. 3.
These functions do represent a considerable improvement, in terms of accuracy of prediction
of AME and DE, over those based on rather more empirical approaches. An example of the
latter would be prediction equations employing iodine value as the independent variable. Thus
the iodine values for soybean oil and rapeseed oil, for example, would be different (the former
would be higher), whereas in terms of dietary energy value the two commodities would be
very similar.
The functions described in Table 1 are, however, not without their problems. Thus it may
not be appropriate to consider applying the functions derived to those fats containing saturated
fatty acids of chain length shorter than those of the current study (i.e., below C^^), as these
may be associated with improved digestibility [6]. This problem has been studied [18] where
a combination of coconut and palmkemel oils (together with the respective acid oil and with
mixtures of the two to give blends of intermediary free fatty acid content) was evaluated.
Both commodities consist predominantly of saturated fatty acids, but of chain lengths shorter
than Cj4 (Table 2).
The AME data generated are presented in Fig. 4, which also illustrates the responses as
predicted employing the functions in Table 1 for the three U/S ratios. It is evident that content
of saturated fatty acids used to calculate the U/S ratio should be based on the sum of myristic
(14:0), palmitic (16:0), and stearic (18:0) acids, but that lauric acid (12:0) behaves like an
unsaturated fatty acid.
2. H eat-D am aged Fats and Oils

As described above, fats and oils destined for use in animal feeding are usually by-products
of other industrial processes. As such it is likely that they have undergone some form of heat
treatment, often in the presence of oxygen. Fats and oils are generally relatively unstable (the
more so the greater the degree of unsaturation) and are therefore prone to some form of
degradation under a variety of conditions (which explains why protection of fats and oils
through the use of antioxidants, added as early as possible in the manufacturing process be
cause oxidative degeneration is irreversible, is crucial for animal feeding). This has prompted
numerous studies on the chemical commodities produced during heating and the nutritional
implications (see, e.g., the reviews in Refs. 19-21).
A large number of modifications to the chemical structure of fats and oils following heating
have been identified, ranging from simple oxidation products through to dimerization and
polymerization (linear and cyclic) of both fatty acids and triglycerides depending on the sub
strate in question and the conditions operating.
The biological effects of feeding these modified structures are also extremely varied in
terms of both the actual response in the animal and its severity. It should be noted that
even minor adverse biological consequences would have serious repercussions for output from
intensive livestock enterprises. Initially, digestibility and hence dietary energy value would be
reduced. The consequences would be that the animals would not perform to expectations and
also that an increasing amount of dietary fat would pass through the gastrointestinal tract
and be excreted. This may have serious implications for the conditions under which animals
and birds are maintained. Thus, for example, the litter in housed broiler chickens, if allowed
to deteriorate through becoming “greasy,” will give rise to a poorer environment and problems
of both bird welfare and product quality. It is therefore possible that the presence of modified
fat structures within the gastrointestinal tract may interfere with the overall digestive process
such that general nutrient uptake is impaired. Thus there may be overall nutrient deficiency.
Furthermore, an actively oxidizing fat or oil will destroy other nutrients present, including
some vitamins.
Perhaps more concern has been expressed over whether any toxic products are generated
following fat and oil heating or oxidation. It does appear, however, that the majority of these
products are only sparingly absorbed and would thus not present a threat to the metabolic
process. However, in the case of oxidized fats and oils, defense mechanisms in the gut muco
sae to prevent absorption may be stretched such that overall nutrient absorption might be
reduced. It is certainly the case that death in laboratory animals fed heated fats and oils has
been recorded [22], but these are extreme cases.
Because of the potential adverse effects of feeding heat-damaged/oxidized fats and oils,
there has been considerable interest in developing chemical methods for the detection of such
damage. It is important to note that any method adopted has to be one that measures all
products collectively if it is to have any practical application. Peroxide value (PV) has been
employed widely for this purpose, but it is an unsound method. In tracing the change in PV
over time [23], an increase followed by a reduction was observed. Thus a low PV value may
indicate on the one hand a commodity that had not undergone any degradation or, on the
other hand, one that had been seriously denatured. Measurement of “oxidized fat” has been
employed, but the evaluation is solvent-dependent and would not measure complexes that are
insoluble in polar solvents.
Free fatty acid content has been employed frequently to assess damage to oils used in the
human food industry but is inappropriate for fats and oils for animal feeding. This is because
soapstocks, for example, are perfectly acceptable ingredients for blends (although of lower
energy value than the original triglyceride) even though the free fatty acid content is invariably
high. This also explains why assessments of molecular weights or sizes are inappropriate.
Thus a fatty acid trimer (of no dietary energy-yielding value) would generate data similar to
those of a triglyceride (of high value).
A technique that has found favor is one based on estimating the total nonelutable material
(NEM) of a fat or oil through quantitative gas-liquid chromatography [21,24]. Although this
method merely measures collectively most degraded structures within fat or oil, it does at
Table 3 Mean Chemical Analyses of Refined (RS) and

Sunflower Acid Oil (SAO)—All Data Expressed as g/kg
Oil
Fatty acid
profile^ RS SAO
14:0 0.7 1.4

16:0 69.3 82.4
17:0 2 .0 2.7
18:0 44.7 47.9
16:1 1.1 2 .0
18:1 196.1 190.5
18:2 651.1 619.6
18:3 1.3 2.0
20+ 16.4 25.3
Other 17.3 26.2
u/s” 7.32 6.25
Nonelutable material 41.0 136.6

Free fatty acids 0 .0 388.0
Water 2 .0 8.2
Unsaponifiable material 0 .6 82.0
Oxidized fatty acids 0.4 14.8
Impurities 0.01 0.1
^Notation indicates length o f carbon chain followed by number

of double bonds.
‘’Ratio of unsaturated (16:1 + 18:1 + 18:2+ 18:3) to saturated
(16:0+ 18:0) fatty acids.
least provide guidance as to whether the commodity has been damaged and has proved useful
in identifying those commodities (e.g., some recovered vegetable oils from frying operations)
that are liable to have been excessively heated.
In a trial designed to examine the reduction in dietary energy value likely to result from
damage [25], a refined sunflower oil was extensively heat damaged (Table 3). Chemical analy
sis revealed an increase in FFA and NEM content following such treatment but little differ
ence in the proportion of individual fatty acids. The two commodities (together with mixtures
of the two) were evaluated for AME, and data generated were compared with predicted values
based on the regression equations described above and in Table 1, which employed FFA and
U/S. From analysis of the data it was evident that the dietary energy value of the NEM
fraction, in this material, was on the order of zero. This indicates the problems identified with
heat-damaged fats, although no account was taken of associated issues of the presence of the
NEM fraction (e.g., reduction in general nutrient uptake).
3. Contam inants W ithin Fats and Oils and N aturally Occurring Fatty
Acids with N onnutritional Activity
In addition to the products arising from heat damage and oxidation, fats and oils may contain
contaminants that are fat-soluble. Perhaps the most important group of contaminants are pesti
cides. Not only are they potentially damaging to the animal itself, residues within the product
are also of concern to the human consumer. Furthermore, many of these commodities are
rendered even more dangerous following the action of heat [e.g., 26]. Levels of many such
contaminants within commodities destined for use in animal feeds are strictly controlled.
Other contaminants found in fats and oils include unsaponifiable matter (plant waxes) that
acts as a diluent, water (also acting as a diluent but also having a role in the oxidation
process), and polythene (which would compromise fat-spraying equipment). Effective quality
control procedures should evaluate fats and oils for these products. Finally it is of interest to
note that, in any process designed to purify a fat or oil, the materials employed in such
purification will themselves become contaminated—this is of particular relevance to the prod
ucts employed in the animal feed industry.
There are numerous examples of naturally occurring fatty acids that have been associated
with antinutritional activity. Erucic acid (22:1) is an important component of some varieties
of rapeseed, and it is alleged that it might be responsible for fatty acid infiltration of heart
muscle [e.g., 27]. While this is a contentious issue (a fatty acid imbalance in rapeseed oil
rather than erucic acid per se was thought responsible [27], it has resulted in a major plant
breeding program designed to reduce the levels of erucic acid in rapeseed cultivars.
Finally the cyclopropenoid fatty acids found in, for example, cottonseed oil have been
implicated in the inhibition of fatty acid metabolism [28], which led, ultimately, to discolor
ation of egg yolks [29].
B. Effect of Dietary Fats and Oils on Meat Quality

An important feature of modem livestock production is increasing interest in quality of prod
uct, and this section proposes to consider the role of dietary lipids in the meat quality of
nonmminants. Carcass fatty acids in pigs and poultry can arise from two discrete sources—
de novo synthesis or direct deposition from dietary sources. The latter is the route by which
polyunsaturated fatty acids (principally linoleic acid, which is nutritionally essential for nonm
minants) appear in the carcass.
The degree to which dietary factors influence the fatty acid profile of carcass fat is con
trolled by a number of factors, including the actual fatty acid profile of the dietary fat and the
energy status of the diet (direct deposition of dietary fat is more likely the higher the overall
dietary energy level). The basic effects of dietary fat have been recognized for some consider
able time, and it was demonstrated [30,31] that diets based on com and soybean oil have a
greater softening effect than peanut and rice oils. Carcass linoleic acid levels, specifically,
have been identified as perhaps the major determinant of the degree of softness of carcass fat.
Investigations into the role of linoleic acid have led to recommendations that concentrations
above 150 g/kg fatty acid in body fat are to be avoided if excessive softness of backfat is to
be prevented [32].
7. E ffect on Physical Texture and on K eeping and Eating Quality

The increased use of diets with high AME and DE levels (particularly as it has been associated
with including significant amounts of relatively unsaturated dietary fats) has been implicated
in promoting softer carcass fat. There is an additional problem with pigs because, with the
progressive move to leaner carcasses, what adipose tissue is left has a higher concentration of
unsaturated fatty acids. The consequences of this are significant for a number of reasons.
Meat from pig carcasses with softer fat does not “set” (it is floppy), and the tissues separate
easily, making slicing and packaging for retail sales difficult [33].
In addition to softness, unsaturated carcass fat in both pig and poultry meat is accompanied
by a risk of a negative effect on the keeping quality of the carcass because of the risk of
oxidative breakdown of unsaturated fatty acids resulting in development of peroxides and
rancidity [34]. This is because oxidation products, being volatile, give rise to off-odors that
will reduce the shelf life of meat. However, the evidence that carcass fatty acid profiles
influence the eating quality of meat is less conclusive. In a trial conducted at the University
of Nottingham (unpublished), a range of 67.8-203.6 g linoleic acid per kilogram of carcass
fatty acids in the backfat (subcutaneous fat) was produced, with the highest level therefore
extending well beyond that at which it is suggested that problems of meat quality would
become apparent. However, influences on the subsequent eating quality of the meat were not
evident although tests were conducted on fresh samples, and it is possible that storage charac
teristics might alter depending on linoleic acid levels in backfat. A subsequent trial (unpub
lished) examining linolenic acid levels in backfat did establish a negative correlation between
levels in the adipose tissue of pigs and eating quality. It would appear that the more unstable
the fatty acid, the greater the risks. Certainly, feeding oils with high levels of long-chain
polyunsaturated fatty acids (e.g., fish oil) is not advisable.
Although current recommendations appear to suggest that the polyunsaturated fatty acid
content of pig and poultry adipose tissue should be limited (on both textural and eating quality
grounds), an opposing viewpoint is occasionally put forward by human nutritionists. Thus a
British government report [35], in recommending that total dietary fat should represent no
more than 33% of total energy intake of the national diet, also proposed that the proportion
of this total fat that was saturated should be no more than one-third. Recommendations were
also established in connection with unsaturated fatty acid intake, which represents an increas
ing awareness of the importance of different classes of unsaturated fatty acids in human nutri
tion. The proposed cis-monounsaturated/cis-polyunsaturated fatty acid ratio was 2:1, and
furthermore, the ratio of cis-a-linoltic acid (of the n-6 or co-6 essential fatty acid family) to
cw-a-linolenic acid (n-3 or co-3 essential fatty acid family) was 5:1.
These recommendations apply to the total diet and should not necessarily be used to define
optimum fatty acid ratios for individual food items in the human diet. Nevertheless, there is
growing interest in modifying the fatty acid profiles of animal products, which is compara
tively easy in pigs and poultry, because of the somewhat negative reputation that these prod
ucts have of being high in saturated fatty acids, which has been implicated in consumer
avoidance of these food items. It is possible that production of animal commodities that are
aligned better to the “optimum” would have a significant impact on consumer acceptance of
them. Nevertheless, such a development should not proceed unless there is clear evidence that
such modifications to carcass adipose tissue levels are not accompanied by deterioration in
meat quality as defined by technological and organoleptic criteria. In this context the protec
tive function of dietary a-tocopherol acetate has recently been studied [36], and it was consid
ered that although it would not improve performance, it had an important effect on meat
stability. Such investigations are proceeding at the University of Nottingham.
3. R ate o f C hange o f Carcass Adipose Tissue F atty A cids

Problems of unsaturated fatty acids within adipose tissue need to be placed alongside the
higher dietary energy values associated with more unsaturated fat blends. Such seemingly
mutually exclusive objectives might be more easily reconciled if the speed with which carcass
fatty acid profiles respond to changes in dietary levels could be established. The more rapid
these changes, the later the introduction of the more saturated lower dietary energy fat. The
responsiveness of carcass fat to dietary fat has been reported [37], where, feeding pigs a diet
based on 100 g tallow/kg following 5 weeks on a diet containing 100 g safflower oil/kg, it
was possible to restore the fatty acid pattern of safflower oil-fed pigs to that of the control
animals. Data indicated that the major changes in fatty acid composition due to diet occurred
Fig. 5 Rate of change of pig adipose tissue fatty acids (for linoleic, C l8:2; palmitic, C l6:0; and
linolenic, C l8:3 acids) following a dietary change from rapeseed oil to tallow at 55 kg live weight.
within 4 -5 weeks. It seems reasonable, therefore, that carcass quality of pigs can be altered
through dietary fat change during the growing and finishing period, although the response
remains to be adequately quantified.
A further report [38] presented data from a trial where pigs had been fed diets based on
soybean oil (S, 60 g/kg), rapeseed oil (R, 60 g/kg), or tallow (T, 65.5 g/kg) while their
weight increased from 25 to 55 kg. Thereafter the animals were either maintained on the same
diet or transferred to either of the other two diets, giving nine treatment combinations in total.
The pattern of change of fatty acid concentration in adipose tissue for one of the experimental
treatments is given in Fig. 5, from which it is evident that the bulk of the change in carcass
fatty acid profile takes place within approximately 25 days of a dietary change. Similar studies
with poultry (University of Nottingham, unpublished) employing diets based on tallow and
safflower oil indicate responses comparable to those of pigs, although somewhat more rapid
(Fig. 6).
Fig. 6 Rate of change of poultry abdominal fat pad tissue fatty acids (for oleic, C18:l; palmitic, C16:0;
and linoleic. C l 8 : 2 acids) following a dietary change from safflower oil to tallow at 2 1 days of age.
III. UTILIZATION OF FATS AND OILS BY RUMINANTS

Ruminants have evolved an extremely efficient digestive system, principally for using the
poor quality forages found in their natural diet. The rumen is a large fermentation chamber
containing billions of microorganisms that break down plant material into volatile fatty acids
(principally acetate, propionate, and butyrate) that can be used by the host animal as energy
sources. During this process, microorganisms grow and multiply, converting nitrogenous sub
stances into high quality microbial protein. This protein is made available to the animal when
the microorganisms pass out of the rumen for digestion in the small intestine. Although mod
erate levels of performance can be achieved from diets containing only grass or forage, high
producing animals require additional energy and protein from supplementary sources. This is
particularly true for dairy cows in early lactation, whose appetite is usually reduced to such an
extent that the cow cannot eat enough to satisfy her nutrient requirements for milk production.
Traditional sources of supplementary energy include starchy materials such as cereals, or
fibrous materials such as sugar beet pulp, and other food processing by-products. The propor
tion of starchy materials in the diet must be limited because they are digested by different
sorts of bacteria than forages. High starch diets can result in rapid fermentation in the rumen,
and the low pH induced may inhibit forage-digesting bacteria, which are less tolerant of low
pH conditions than amylolytic microorganisms. This results in lower feed intake and reduc
tions in the butterfat content of the milk. For diets with very high energy concentrations it is
necessary to use fats.
A. Rumen Fermentation, Digestion, and Absorption of Fats

Fats have a gross energy content about twice that of grass and cereals. In ruminants they are
absorbed from the intestine and used very efficiently (Table 4), and their digestibility is little
affected by chain length and degree of unsaturation.
The fat content of natural diets for ruminants is less than 50 g/kg. However, if added fat
increases the total fat content of the diet to more than about 100 g/kg, digestive problems can
occur. Rumen microorganisms cannot use large quantities of fats, and long-chain unsaturated
fatty acids have a detergent effect on bacterial cell walls [43]. Under normal circumstances,
the ester linkages of triacylglycerides are rapidly hydrolyzed by bacterial lipases in the rumen.
Once released from ester combination, unsaturated fatty acids are subsequently hydrogenated
to detoxify them [44]. Fats also have a physical effect on fiber in the rumen; fiber particles
become coated with fat, rendering them inaccessible to microbial attack [45]. The magnitude
of these effects depends on level, source, and type of fat; dietary carbohydrate source; and
feed intake. A further problem is that long-chain free fatty acids can form soaps with calcium
Table 4 Gross Energy and Metabolizable Energy of Selected Feeds for Ruminants
(MJ/Kg Dm)
Metabolizable Net energy for

Feed Gross energy energy milk production
Grass 18.7 11.2 6.9

Grass silage 20.9 11.0 6 .8
Barley grain 18.5 13.3 8.1
Fat prills 39.0 33.0 26.4
Tallow 39.3 30.5 24.4
Sources: Refs. 3 9 -4 2 .
and magnesium in the rumen. This will detoxify the fatty acids, but it can also reduce the
availability of the minerals.
Unlike the situation for nonruminants, it is not usually possible to alter the fatty acid
composition of carcass fats by altering the fatty acid profile of the diet. This is because
hydrogenation in the rumen results in the predominance of stearic acid. Rumen microorgan
isms do synthesize shorter chain fatty acids, so there is some scope for altering carcass or
milk fatty acids by dietary manipulation, but for unsaturated fatty acids to be directly incorpo
rated into milk and meat it is necessary to protect them against rumen modification.
1. Protected Fats
Various products have been developed that are referred to as “protected fats.” This is really a
misnomer, as the main requirement is to protect the rumen from the detrimental effects of fat,
rather than to protect the fat from rumen degradation. However, protected fats do offer the
potential for increasing the absorption of unsaturated fatty acids from the small intestine.
The three methods of protection that have received the most attention are formaldehyde
protection, fat prills, and calcium soaps. Formaldehyde protection involves spray drying an
emulsion of fat and casein that has been treated with formaldehyde to render it undegradable
in the rumen. Fat prills rely on selected fatty acids that have a high melting point, thus
remaining solid at rumen temperatures, and small particle size, thus passing through the ru
men quickly. Calcium soaps are produced by attaching calcium to the carboxyl groups of fatty
acids, rendering them inert in the rumen; the calcium is detached when the fat reaches the
low pH environment of the abomasum. The cost of formaldehyde-treated casein makes the
first method of protection very expensive. The majority of protected fats used are in the form
of calcium soaps or prills.
In addition to manufactured protected fats, there are certain naturally occurring feed materi
als that have a high oil content but do not cause digestive problems, even when the quantity
of added fat is up to 0.5 kg/day for dairy cows. These feeds are oilseeds, such as soybeans,
rapeseed, sunflower seed, and cottonseed. They must be fed in the full fat form, although
they can be heat-treated and rolled or ground. Extracted oil from these products can be used
in only small quantities before rumen disturbances occur. In the unextracted form, rumen
bacteria digest the fiber and protein of the oilseeds slowly and therefore release the oil at a
rate that is compatible with normal rumen function. Cottonseed is often fed with the lint still
attached to the seed. This provides plenty of attachment points for rumen bacteria and in
creases digestibility. Soybeans contain trypsin inhibitors that may interfere with intestinal
digestion of protein. Therefore, if large quantities of whole soybeans are being fed, it is better
if they are heat-treated to denature the trypsin inhibitors. Soybeans should not be used in
conjunction with urea, because the urease in whole soybeans leads to rapid release of ammo
nia from the urea, which can cause urea toxicity and feed refusal.
A study was conducted on the effects of free fatty acids (as palm acid oil) and protected
fats (in the form of a calcium soap or fat prills) on the digestibility of fiber and fat in sheep,
where fats replaced barley and tapioca to provide 15% of digestible energy intake [46] (Table
5). At the maintenance level of feeding, fiber digestibility was not affected by the inclusion
of fat in the diet. Fiber digestibility was lower at the twice-maintenance level of feeding and
was significantly higher for the diet containing calcium soap than for the control diet and diets
containing free fatty acids or fat prills. The digestibility of added fat was also higher at the
maintenance level of feeding. At the twice-maintenance level, the fat added as a calcium soap
was more digestible than free fatty acids or fat prills. In a further study comparing fat prills
and calcium soaps [47], it was found that with pelleted diets the fat digestibility of the two
Table 5 Apparent Digestibilities of Fiber and Added Fat in Sheep Fed at Two Levels of Intake
Maintenance Twice maintenance
Free Calcium Free Calcium Fat

Control fatty acids soap Control fatty acids soap prills
Fibre 0.71 0.71 0.71 0.55 0.55 0.61 0.53

Added fat 0.82 0.76 0.56 0.63 0.48
Source: Ref. 42.
forms was similar (prills 0.902, soap 0.965), but when diets were not pelleted the digestibility
of fat in the form of prills was considerably lower than that of calcium soaps (prills 0.506,
soap 0.932). It is possible that heat generated during the pelleting process altered the nature
of the fat prills.
B. Use of Fats in Commercial Situations

L Sheep
The use of fat supplements for sheep under commercial conditions is limited because the
animal production level is much lower than that of dairy cows. If supplementary feeding is
required to complement grazing, the most common form is barley with a protein balancer.
There is some scope for the use of fats with ewes in late pregnancy, when food intake is
limited. It is important that high intakes of energy be maintained during the last month of
pregnancy, particularly with ewes in poor body condition, but caution is needed because it is
undesirable to make the ewes too fat at lambing. Under intensive conditions there may be a
case for supplementing housed ewes, or those on restricted pasture, in early lactation. It has
been shown [48] that a protected fat supplement in the form of a calcium soap fed at levels
of 75, 150, 200, and 250 g/day significantly increased milk yield and milk fat concentration
in Finn Dorset ewes, leading to increased growth rates of lambs, which were on average 1.0
kg heavier at 5 weeks of age.
2. B e e f Cattle
Fats are not used extensively in diets for beef cattle in the United Kingdom. The normal
objective is to achieve high carcass weights with a minimum amount of fat cover [49]. Al
though the efficiency of utilization of metabolizable energy for live-weight gain increases with
the energy content of the diet, the rate of carcass fat deposition also increases. Extensive use
of added dietary fat would result in overfat carcasses or carcasses with the desired fat content
but at a lighter weight. In systems where heavier and fatter carcasses are preferred, e.g., in
northern France, supplementary fats are used to achieve faster live-weight gains and heavier
weights at slaughter.
Fats are used for preruminant calves in milk replacers, which are based on skim milk or
artificial milk with added fat in the form of vegetable oil or tallow to increase the fat content
to that of whole milk [50]. This is essential because the calf’s capacity for carbohydrate
digestion is limited [51]. It is important that the fat be broken down into small particles (3-4
fxm in diameter) by homogenization to ensure efficient reconstitution of the milk and utiliza
tion by the calf. The fat source influences calf performance through a combination of factors
[52]. Digestibility decreases with increasing chain length and increases with the degree of
unsaturation at any given chain length. Short-chain fatty acids are absorbed faster than medi
um- or long-chain fatty acids. Palmitic acid is used more efficiently when present in the
middle position of the triglyceride molecule, as in lard, than when occupying positions 1 or
3. With fats of low digestibility, there is an increase in digestibility with age.
3. D airy Cows
By far the greatest value of fats for ruminants is in the feeding of dairy cows. Estimates of
the upper limit to the absorption of fatty acids have ranged from 900 to 2180 g/day [53]. It
would be very difficult to achieve these large quantities without the use of protected fats.
The response of dairy cows to supplementary fat is complex and not always predictable.
Responses that have been reported include increased milk yields, increased or decreased milk
fat content, decreased milk protein content, increased live-weight gain, and decreased live-
weight loss. The observed response will depend on the quantity of fat, its fatty acid profile
and degree of protection, the other components of the diet, and the overall feeding level,
stage of lactation, and genetic merit of the cow.
Milk yield responses can normally be explained by the increase in total energy intake when
fats are given and the increased efficiency of utilization of energy from fats. The response
will not be as great as expected if rumen fermentation is disrupted, because digestion of the
non-fat components of the diet, and subsequent energy release, will be reduced. Stage of
lactation and genetic merit affect both the milk yield response and the live-weight response.
Dairy cows in early lactation and those of higher genetic merit partition energy toward milk
production at the expense of body fat reserves. Cows normally lose 0.5 -1 .0 kg of body
weight each day for the first 8 weeks of lactation, and this is mostly body fat. Therefore
increased energy intake at this stage of lactation could result in further increases in milk yield
if the cow’s genetic potential has not been reached and/or a reduction in the daily amount of
body fat mobilized [54,55]. In later lactation, when appetite is greater and milk yield less.
Total dietary fat (% )

Fig„ 7 Milk protein concentrations of cows fed added fat relative to controls (control = 100).
Y = 101.1 - 0.6381X4-0.0141X^, r^ = 0.24. Added fat sources: animal-vegetable fat blend (4-); calcium
salts of fatty acids (A); prilled fat ( ♦ ) ; protected tallow (o); tallow (V); oilseeds (•); yellow grease
(■ ). (From Ref. 51.)
partition of energy switches more toward body reserves, so the increase in the energy supply
results in greater fat deposition in body reserves. Cows of low genetic merit have a greater
propensity for fat deposition and will partition a greater proportion of surplus energy in this di
rection.
Milk fat is derived from intramammary synthesis using acetate and /3-hydroxybutyrate and
from triglycerides or fatty acids absorbed directly from the bloodstream. Short-chain fatty
acids, up to 10 carbon atoms in length, are virtually all synthesized. Acids with 18 carbon
atoms are absorbed, and intermediate chain length acids originate from both sources. If added
dietary fats interfere with normal fiber digestion in the rumen, acetate production will be
reduced and the shortage of precursors in the mammary gland may lead to reduced de novo
synthesis of milk fat. On the other hand, added fat may increase the quantity of fatty acids
available for absorption and secretion in milk. In both cases, the proportion of C4 _i^ fatty
acids in milk fat will decrease and the proportion of longer chain acids will increase. Total
milk fat output will depend on the relative magnitude of the effects on synthesis and absorp
tion, which are affected by protection and degree of unsaturation.
The biggest problem in using supplementary fats is the depression in milk protein concen
tration. In a review of 47 published experiments [56] involving 1396 cows, it was found that
milk protein concentration decreased by 1-5% of control values when dietary fat concentra
tion was increased within the range 3-15% (Fig. 7). Factors affecting milk protein secretion
have received a lot of attention in recent years because this component has a greater effect on
milk price than milk fat. The prediction of response is complicated by the fact that dietary
protein supply, except in deficiency cases, has less effect on milk protein secretion than di
etary energy supply or, more precisely, glucose and its precursors. The system is illustrated
in Fig. 8, which explains the priorities for nutrient use in the dairy cow [57]. The three main
nutrients of interest are amino acids, glucose, and energy substrates including acetate, buty
rate, and long-chain fatty acids. The first priority for amino acids is to satisfy the nitrogen
requirements for maintenance of the animal. The next priority is to provide glucose if it is in
short supply, followed by the requirements of the growing fetus. Milk protein comes fairly
low on the list of priorities for amino acids and also requires energy-yielding substrates to
support protein synthesis. Alterations in the diet that affect glucose supply (which is mostly
from propionate absorbed across the rumen wall) or requirements will alter the extent of
utilization of amino acids for gluconeogenesis. There are two possible scenarios in which
dietary fat can reduce glucose supply. First, direct effects on rumen fermentation may reduce
propionate production. This is not likely to be a major effect, as fats tend to reduce the
amount of fiber digested which would be expected to cause a reduction in the production of
acetate. Propionate is produced mainly by the digestion of starch in the rumen, which is not
greatly affected by fats. The second possibility is that since fats normally replace starch in the
ration, the total quantity of glucose precursors may be reduced (both propionate in the rumen
and starch digested in the small intestine). Where fat supplements have replaced grains on an
isoenergetic basis [58], milk yield was similar for both treatments but milk protein content
was reduced. However, it has been reported [59] that abomasal infusion of glucose increased
the milk yield of cows but reduced the milk protein content. Another possible mechanism that
has been proposed to explain the depression of milk protein content with high fat diets is
insulin resistance. It has been noted that plasma insulin is elevated in high fat diets and that
the rate of glucose removal is negatively related to insulin concentration [60]. Therefore, high
fat diets may cause insulin resistance, which would retard the uptake of amino acids by
mammary tissue for protein synthesis [56]; however, changes in plasma insulin concentrations
in response to fat supplementation have been inconsistent [61-63].
Fig. 8 Order of priority for use of major classes of nutrients by dairy cows. Glucose includes absorbed
glucose and absorbed propionate. Energy substrates include absorbed acetic and propionic acids plus
absorbed long-chain fatty acids. Solid lines indicate pathways for use of substances in meeting defined
animal functions. Dashed lines indicate use of resources in support of synthetic processes. (From Ref.
52.)
There is one further way in which dietary fat may affect milk protein: through miscalcula
tions of protein supply during ration formulation. The Metabolizable Protein system for pro
tein evaluation recently proposed by AFRC [39] calculates the supply of metabolizable protein
to the animal from digestible undegraded dietary protein (DUP) and microbial protein pro
duced in the rumen. Microbial protein production, where it is limited by energy intake, is
calculated from fermentable metabolizable energy. Although dietary fats contribute to the
metabolizable energy supply of the animal, they are not fermented in the rumen, so they do
not provide energy for microbial growth. Diets with added fat often need extra DUP or fer
mentable carbohydrates to avoid induced metabolizable protein deficiencies. In an experiment
conducted at Nottingham University [64], it was found that increasing the fat content of the
diet from 25 to 76 g/kg DM using a calcium soap increased milk yield and protein yield but
Table 6 Milk Yield and Composition o f Four Groups of Cows Over Weeks 4 -1 5 o f Lactation
Treatment group
Protected
Protected fat and
Control fat Lactose lactose SED
Milk yield, kg/day 24.6^ 21J^ 25.6"’^^ 26.5"’'’ 1.20

Milk fat, g/kg 46.3 46.9 46.8 48.0 2.99
kg/day 1.13 1.30 1.20 1.27 0.09
Milk protein, g/kg 32.3^’^ 30.7" 32.7'’ 31.9"’'’ 0.79
kg/day 0.79 0.85 0.83 0.85 0.05
SED = Standard error of the difference.

Means within a row with different superscripts are significantly ( P < 0 .0 5 ) different.
decreased milk protein concentration. When lactose, a readily fermentable carbohydrate, was
included in the diet, the milk protein concentration was the same as control values (Table 6).
In a further experiment [65] it was found that protected protein also partially alleviated the
negative effect of dietary fat on milk protein concentration. However, when a combination of
protected fat, protected protein, and lactose was used in the diet, the results were complemen
tary. A significant increase in milk yield was found without any depression in milk protein
concentration (Table 7).
Table 7 Milk Production, Milk Composition, and Dry Matter Intake of Dairy Cows
Treatment group
Protected fat
Protected fat & protein +
Control & protein Lactose lactose SED
Milk yield, kg/day 21.3" 23.4"’'’ 22.7"’'’ 27.8'’ 2.18

Milk fat, g/kg 42.6 41.7 42.5 41.8 2.68
Milk protein, g/kg 33.7 33.6 33.7 33.8 1.38
Dry matter intake, kg/day 17.3 17.6 18.9 17.9 1.3
SED = Standard error of the difference.

^ Means with different superscripts are significantly ( P < 0 .0 5 ) different.
REFERENCES
1. C. P. Freeman, Digestion and absorption of fat, in Digestion and Absorption in the Fowl (K. N.
Boorman and B. M. Freeman, Eds.), British Poultry Science, Edinburgh, UK, 1976, p. 117.
2. C. P. Freeman, The digestion, absorption and transport of fat— non-ruminants, in Fats in Animal
Nutrition (J. Wiseman, Ed.), Butterworths, London, 1984, p. 105.
3. J. Wiseman, Assessment o f the digestible and metabolisable energy content of fats for non-rumi
nants, in Fats in Animal Nutrition (J. Wiseman, Ed.), Butterworths, London, 1984, p. 211.
4. A. Krogdahl, Digestion and absorption of lipids in poultry, J. Nutr. 115: 675 (1985).
5. J. Wiseman, Variability in the nutritive value of fats for non-ruminants, in Feedstuff Evaluation
(J. Wiseman and D. J. A. Cole, Eds.), Butterworths, London, 1990, p. 215.
6. R. Renner and F. W. Hill, Utilisation of fatty acids by the chicken, J. Nutr. 74: 259 (1961).
7. T. S. Stahly, Use of fats in diets for growing pigs, in Fats in Animal Nutrition (J. Wiseman, Ed.),
Butterworths, London, 1984, p. 313.
8. R. J. Young, The energy value of fats and fatty acids for chicks. 1. Metabolisable energy. Poult.
Sci. 40: 1225 (1 9 6 1 ).
9. D. Sklan, Digestion and absorption o f lipids in chicks fed triglycerides or free fatty acids. Synthe
sis of monoglycerides in the intestine. Poult. Sci. 58: 885 (1979).
10. J. Wiseman and F. Salvador, The influence of free fatty acid content and degree of saturation on
the apparent metabolisable energy value of fats fed to broilers. Poult. Sci. 70: 573 (1991).
11. J. Wiseman, F. Salvador, and J. Craigon, Prediction o f the apparent metabolisable energy content
of fats fed to broiler chickens. Poult. Sci. 70: 573 (1991).
12. J. Powles, J. Wiseman, D. J. A. Cole, and B. Hardy, Effect of chemical structure of fats upon
their apparent digestible energy value when fed to growing/finishing pigs, Anim. Prod. 57: 137
(1 9 9 3 ).
13. J. Powles, J. Wiseman, D. J. A. Cole, and B. Hardy, Effect of chemical structure of fats
upon their apparent digestible energy value when given to young pigs, Anim. Prod. 58: 411
(1 9 9 3 ).
14. J. Powles, J. Wiseman, D. J. A. Cole, and S. Jagger, Prediction of the apparent digestible energy
value of fats given to pigs, Anim. Sci. 61: 149 (1995).
15. L. B. Carew, R. H. Machemer, R. W. Sharp, and D. C. Foss, Fat absorption by the very young
chick. Poult. Sci. 52: 738 (1972).
16. M. R. Fedde, P. E. Waibel, and R. E. Burger, Factors affecting the absorbability o f certain
dietary fats in the chick, J. Nutr. 70: 447 (1960).
17. J. Wiseman and F. Salvador, The influence of age, chemical composition and rate of inclusion on
the apparent metabolisable energy o f fats fed to broiler chicks, Br. Poult. Sci. 30: 653 (1989).
18. J. Wiseman and A. Blanch, The effect of free fatty acid content on the apparent metabolisable
energy content of coconut/palm kernel oil for broiler chickens aged 12 and 52 days, Anim. Feed
Sci. Technol. 47: 22 5 (1 9 9 4 ).
19. N. R. Artman, The chemical and biological properties of heated and oxidised fats, Adv. Lipid Res.
7: 2 4 5 (1 9 6 9 ).
20 . J. Wiseman, Antinutritional factors associated with dietary fats and oils, in Recent Advances in
Animal Nutrition—1986 (W. Haresign and D. J. A. Cole, Eds.), Butterworths, London, 1986,
p. 47.
21 . B. K. Edmunds, Chemical analysis of lipid fractions, in Feedstuff Evaluation (J. Wiseman and
D. J. A. Cole, Eds.), Nottingham Univ. Press, Nottingham, UK, 1990, p. 197.
22 . J. S. Andrews, J. F. Mead, and W. H. Griffith, Toxicity of lipid peroxides in the rat, J. Nutr.
70: 199 (1 9 6 0 ).
23 . C. E. Poling, W. D. Warner, P. E. Mone, and E. E. Rice, The influence of temperature, heating
time and aeration upon the nutritive value o f fats, J. Am. Oil Chem. Soc. 39: 315 (1962).
24. A. E. Waltking, W. E. Seery, and G. W. Bleffert, Chemical analysis o f polymerisation products
in abused fats and oils, J. Am. Oil Chem. Soc. 52: 96 (1975).
25 . J. Wiseman, B. K. Edmunds, and N. Shepperson, The apparent metabolisable energy of sunflower
oil and sunflower acid oil for broiler chicks, Anim. Feed Sci. Technol. 36: 41 (1992).
26. L. D. Metcalfe, Proposed source of chick edema factor, J. Assoc. Off. Agrie. Chem. 55: 542
(1 9 7 2 ).
27. J. K. G. Kramer, H. W. Huían, H. L. Trenholm, and A. H. Comer, Growth, lipid metabolism
and pathology of two strains o f rats fed high fat diets, J. Nutr. 109: 202 (1979).
28. R. Jeffcoat and M. R. Pollard, Studies on the inhibition of the desaturases by cyclopropenoid fatty
acids. Lipids 12: 480 (1977).
29. F. S. Shenstone, J. R. Vickery, and A. R. Johnson, Studies on the chemistry and biological
effects of cyclopropenoid compounds, J. Agrie. Food Chem. 13: 410 (1965).
30. N. R. Ellis and H. S. Isbeli, Soft pork studies. II. The influence of the character of the ration
upon the body composition of body fat of hogs, J. Biol. Chem. 69: 219 (1926).
31. N. R. Ellis and H. S. Isbeli, Soft pork studies. III. The effect of food fat upon body fat as shown
by the separation of the individual fatty acids of the body fat, J. Biol. Chem. 69: 239 (1926).
32. F. M. Whittington, N. J. Prescott, J. D. Wood, and M. Enser, The effect of dietary linoleic acid
on the firmness o f backfat in pigs of 85kg live weight, J. Sci. Food Agrie. 37: 753 (1986).
33. A. J. Kempster, A. W. Dilworth, D. G. Evans, and K. D. Fisher, The effect of fat thickness and
sex on pig meat quality with special reference to the problems associated with overleanness, Anim.
Prod. 43: 517 (1986).
34. M. Enser, in Proceedings of Meat Research Institute Symposium No. 3 (C. L. Cutting, Ed.), Meat
Research Institute, Bristol, 1974.
35. COMA, Dietary reference values for food and nutrients in the United Kingdom, Department of
Health and Social Services, Report on Health and Social Subjects 41, HMSO, London, 1991.
36. I. Bartov and M. Frigg, Effect of high concentrations of dietary vitamin E during various age
periods on performance, plasma vitamin E and meat stability of broiler chicks at 7 weeks of age,
Br. Poult. Sci. 33: 393 (1992).
37. D. E. Koch, A. M. Pearson, N. T. Magee, J. A. Hoefer, and B. S. Schweigert, Effect of diet on
the fatty acid composition of pork fat, J. Anim. Sci. 27: 360 (1968).
38. J. Wiseman, J. Agunbiade, and D. J. A. Cole, The influence of changes in dietary oil content on
the fatty acid profile of backfat in pigs, Anim. Prod. 54: 497 (1992).
39. AFRC, Energy and Protein Requirements of Ruminants, An advisory manual prepared by the
AFRC Technical Committee on Responses to Nutrients, CAB International, Wallingford, 1993.
40. MAFF, UK Tables of Nutritive Value and Chemical Composition of Feedingstuffs, Rowett Re
search Services, Aberdeen, Scotland, 1990.
41. P. McDonald, R. A. Edwards, J. F. D. Greenhalgh, and C. A. Morgan, Animal Nutrition, Long
man, Harlow, UK, 1995.
42. NRC, Nutrient Requirements of Dairy Cattle, National Academy Press, Washington, DC, 1989.
43. H. Galbraith and T. B. Miller, Effect of metal cations and pH on the antibacterial activity and
uptake of long chain fatty acids, J. Appl. Bacterial. 36: 635 (1973).
44. C. G. Harfoot, Lipid metabolism in the rumen, in Lipid Metabolism in Ruminant Animals (W. W.
Christie, Ed.), Pergamon, Oxford, 1981, pp. 2 1 -5 5 .
45. C. Devendrá and D. Lewis, The interaction between dietary lipids and fibre in sheep. 2. Digestibil
ity studies, Anim. Prod. 19: 67.
46. H. Ainsworth and E. L. Miller, The effects of protected and unprotected fats in rations fed
to sheep at two intakes on the apparent digestibilities of fat and fibre, Anim. Prod. 40: 534
(1985).
47. J. C. Mathers, N. Sadler, and E. L. Miller, Some effects of dietary fats on rumen metabolism and
digestion in sheep, Proc. Nutr. Soc. 42: A64 (1983).
48. M. Perez Hernandez, J. J. Robinson, R. P. Aitken, and C. Fraser, The effect of dietary supple
ments of protected fat on the yield and fat concentration of ew e’s milk and on lamb growth rate,
Anim. Prod. 42: 455 (1986).
49. D. Allen, Planned Beef Production and Marketing, BSP Professional Books, Oxford, UK, 1990.
50. W. Thickett, D. Mitchell, and B. Hollows, Calf Rearing, Farming Press, Ipswich, UK, 1986.
51. J. B. Owen, Cattle Feeding, Farming Press, Ipswich, UK, 1991.
52. J. H. B. Roy, The Calf, Butterworths, London, 1980.
53. J. E. Storry, The effect of dietary fat on milk composition, in Recent Advances in Animal Nutri
tion—1981 (W. Haresign, Ed.), Butterworths, London, 1981, pp. 3 -3 3 .
54. G. P. Jones and P. C. Garnsworthy, The effects of dietary energy content on the response by
dairy cows to body condition at calving, Anim. Prod. 49: 183 (1989).
55. P. C. Garnsworthy and C. D. Huggett, The influence of the fat concentration of the diet on the
response by dairy cows to body condition at calving, Anim. Prod. 54: 1 (1992).
56. Z. Wu and J. T. Huber, Relationship between dietary fat supplementation and milk protein concen
tration in lactating cows: a review. Livestock Prod. Sci. 39: 141 (1994).
57. J. D. Oldham and G. C. Emmans, Prediction of responses to protein and energy yielding nutrients,
in Nutrition and Lactation in the Dairy Cow (P. C. Gams worthy, Ed.), Butterworths, London,
1988, pp. 76-96.
58. N. E. Smith, W. L. Dunkley, and A. A. Franke, Effects of feeding protected tallow to dairy cows
in early lactation, i. Dairy Sci. 61: 747 (1978).
59. R. A. Frobish and C. L. Davis, Effects of abomasal infusions of glucose and propionate on milk
yield and composition, J. Dairy Sci. 60: 204 (1977).
60. D. L. Palmquist and E. A. Moser, Dietary fat effects on blood insulin, glucose utilisation and
milk protein content of lactating cows, J. Dairy Sci. 64: 1664 (1981).
61. K. A. Cummins and J. L. Sartin, Response of insulin, glucagon and growth hormone to intrave
nous glucose challenge in cows fed high fat diets, J. Dairy Sci. 70: 277 (1987).
62. A. D. Kennedy, F. R. Tekpetey, J. R. Ingals, and W. M. Palmer, Effect of stage of lactation and
diet on semm insulin level and mononuclear leukocyte insulin receptor characteristics in dairy
cows. Can. J. Anim. Sci. 67: 721 (1987).
63. G. Gagliostro, Y. Chilliard, and M. J. Davicco, Duodenal rapeseed oil infusion in early and mid
lactation cows. 3. Plasma hormones and mammary apparent uptake of metabolites, J. Dairy Sci.
74: 1893 (1991).
64. P. C. Gams worthy. The effects on milk yield and composition of incorporating lactose into the
diet of dairy cows given protected fat, Anim. Sci. 62: 1 (1996).
65. P. C. Gams worthy. Effects of feeding lactose with protected fat and protein on milk production
of cows in mid lactation, Anim. Sci. 63: in press (1996).
23______________________________________________
Anionic Detergents
Maurice R. Porter
Maurice R. Porter and Associates, Sully, Vale of Glamorgan, South Wales
I. INTRODUCTION
Anionic surfactants are manufactured and used in greater volume than all other types of sur
factants. Soap, the alkali metal salt of a carboxylic acid, derived from natural oils and fats,
has been used as a detergent for thousands of years. It is produced by reacting an alkali with
an animal fat or vegetable oil that contains the glyceryl esters of the long-chain fatty acids,
the process known as saponification. Soap was used by the Phoenicians and the Romans, and
a soap-making process was recorded by Plinius in A.D. 70.
The very early soaps were made from animal fats and wood ash. Saponification is shown
in Fig. 1. Improved products were introduced in the Middle Ages by the Arabs, who used
vegetable oils. By the twelfth century, soap was being made commercially in Bristol, En
gland.
Saponification requires alkali. All the early processes used wood ash and needed consider
able quantities of timber to produce the ash. It was not until Leblanc invented his process for
the manufacture of caustic soda in 1778 that large-scale soap manufacture became possible.
Crosfields was founded in 1814, Pears in 1879, and Lever in 1884, so that the latter half of
the nineteenth century was the main period when the production of soap from lipids was
carried out on an industrial scale.
Up until the nineteenth century, soapmaking was an empirical science, but by the 1800s
inorganic chemistry was well developed so that the industrial production of alkalis and acids
was on a scientific basis. The organic chemistry of lipids was understood, the variations of
lipid raw materials were being investigated, and the sulfonation of castor oil and olive oil was
being carried out industrially in the period 1850-1870. The cheaper sulfonated castor oil was
employed by Crum to assist in dyeing with alizarin dyestuffs, the so-called Turkey red style.
Since that time sulfonated castor oil has been known as Turkey red oil. Later (about 1900),
the sulfonation of oleic acid was commercialized and was used in fat splitting during soap
manufacture.
579
580 Porter
C„H2n+lCOOÇH2 CH2OH
I
C„H2„+iCOOCH + NaOH 3C„H2„-HCOONa + CHOH
I I
C„H2n+lCOOCH2 CH2OH
Fig. 1 Saponification of fats with caustic soda.
Soap is an excellent general household detergent and was and still is widely used for
personal washing (toilet soaps), clothes washing, and wall and floor washing, but only in soft
water. In hard water, soap is not soluble; it precipitates and thereby loses its detergent proper
ties. Soap will also precipitate in the presence of acids. At the beginning of the twentieth
century the reason for this was found to be the length of the hydrocarbon chain of the lipid
that was used in the manufacture of the soap. The calcium or magnesium salts are insoluble
in water if n in Fig. 1 is greater than 16. As animal fats, the cheapest lipids used in soap
manufacture have principally alkyl chains of 16 and greater, soaps made from tallow give
precipitates in hard water. If n= 10-14, the solubility improves considerably. However, at
the turn of the century the availability of lipids containing alkyl chains of 10-14 was limited.
Such lipids are the vegetable oils such as coconut and palmkemel oil. Soap made with alkyl
chains of 10-14 gave superior hard water detergency but at a considerably higher price and
could only be used for toilet soaps. Even so, fatty acids of alkyl chain length 10-14 are still
insoluble in water and therefore can only be used in alkaline conditions. It was found that the
sulfonated vegetable oils (e.g., Turkey red oil) were stable to hard water and limited acidic
conditions. The availability of suitable vegetable oils was limited for the mass market to
laundry soaps, which had been developed in the early 1900s. It was the discovery of lime
soap dispersing agents, builders, and bleaches that could be added to soap to overcome the
hard water problem that resulted in the widespread use of soap for laundry. These additives,
generally inorganic, in some way overcame the solubility properties of sodium and potassium
salts of long-chain fatty acids. The discoveries came about entirely by trial and error, as the
reason for the action of surfactants lies in their physical chemistry, which has only recently
begun to be understood.
By the 1930s, laundry products for domestic use were formulated mixtures, with soap
(generally from tallow) as the major surfactant in the formulation. Research still continued on
chemical methods of modification to overcome the hard water problem, and a very large
number of new synthetic chemicals were synthesized in this period. By 1940 probably all the
presently known surfactants had been made on a laboratory scale. As an example, nonionic
surfactants based on ethylene oxide were found to give good detergency and excellent solubil
ity in both hard and soft water over a wide pH range, but ethylene oxide was not available
on a large scale.
It was the development of the petrochemical industry in the period 1940-1970 that brought
about a revolution in the type of surfactants available on the market. The reason was twofold.
First, the hydrophobic group (usually a saturated paraffinic chain Ci 2_i8 chain length) was
now available in large quantities at much cheaper prices than the equivalent hydrophobic
group from lipids (animal fats and vegetable oils). Second, hydrophobic groups containing
the benzene ring were now available in large quantities that allowed far more variety in
attaching hydrophilic groups to the hydrophobic group, e.g., sulfonation using sulfur trioxide.
As a result, a large number of “synthetic” surfactants appeared on the market from 1940
Anionic Detergents 581
Table 1 Soap Production in

Western Europe
Year Soap production (1000 t)
1972 866
1982 776
1993 580
Source: Ref. 1.
(depending on the country) onward, and the products that were successful were those whose
properties were superior to those of soap in hard water. Soap was replaced in most household
and domestic detergents with the exception of toilet soap, and this decline in soap production
still continues in Western Europe (see Table 1).
The principal anionic surfactants that have replaced soap in detergents are listed in Table
2. (Note the different alkyl chain lengths that are necessary to obtain a detergency equivalent
to that of soap.)
It must be emphasized that the replacement of soap by a synthetic surfactant is not a simple
matter of comparing the relative prices and surfactant properties of the two surfactants but
concerns the price/performance ratio of a fully formulated detergent. A detergent is the prod
uct that does the washing. The surfactant is the component that has the major influence on
the properties of an aqueous solution in relation to wetting, foaming, dispersing solids, emul
sifying oils, and removing dirt from a fabric. A modem heavy duty laundry detergent for
horizontal dmm machines will have at least two surfactants, a builder, a bleaching system,
an enzyme, antiredeposition agents, foam stabilizer and control additive, fluorescent whiten
ing agents or optical brighteners, corrosion inhibitor, perfume, dyestuff, and fillers. There
will be a considerable number of interactions between the various components and the surfac
tants, and therefore the testing and comparison of competitive surfactants is a complex pro
cedure.
Due to the different washing habits in various countries, different formulations will gain
market dominance. Therefore the major anionic products used in both domestic and industrial
detergents can vary from one country to another. Nevertheless, the surfactant or the mixture
of surfactants plays the major role. Linear alkylbenzene sulfonates (LABS) are undoubtedly
the major anionic component used in detergents worldwide to replace soap, a-olefin sulfonates
are used in the United States but not in Europe. Paraffin sulfonates originated in Europe and
are only now being used in the United States. Nonionic surfactants, such as fatty alcohol
Table 2 Principal Anionic Surfactants That Have Replaced Soap in

Synthetic Detergents
Chemical name Abbreviation
Linear alkyl (Cjo- m) benzene sulfonate, sodium salt LABS

a-Olefin (Ci4_i6) sulfonate, sodium salt AOS
Fatty alcohol (C12- 14) sulfate, sodium salt AS
Fatty alcohol ( € 12- 14) ether sulfate, sodium salt AES
Secondary alkane (Ci4_i7) sulfonates (paraffin sulfonates) SAS
Alpha-sulfonated methyl esters of fatty acids (Ci6_i8), sodium salt FES
582 Porter
ethoxylates, are playing a much more important role as washing temperatures decrease.
Lipids from animal fats and vegetable oils, once the dominant source of raw materials for
detergents (i.e., soap) have had a declining share of the domestic and industrial detergent
market. This situation could well change significantly in the future due to possible shortage
in the supply of crude petroleum, the demand for “mild” detergents, the demand for “natural”
products, and the environmental problems of synthetic surfactants. The overall result of these
factors will probably be a slow decline in the consumption of anionic synthetic detergents
based on petrochemicals, which will be replaced by new semisynthetic products based on
natural products rather than soap. For instance, the major new surfactant type in the last
decade is the alkyl poly glycosides, which use the polyhydroxy groups in carbohydrates as the
hydrophilic entity with a long-chain fatty alcohol (from fats, oils, or petrochemicals) as the
hydrophobic entity.
II. RAW MATERIAL ROUTES TO ANIONIC SURFACTANTS

Tallow soap has a hydrophobic group of C 17H 35 with a molecular weight of 239 and a hydro
philic group COONa with a molecular weight of 67. Sodium lauryl sulfate has a hydrophobic
group of C 12H 25 with a molecular weight of 169 and a hydrophilic group of S04Na with a
molecular weight of 119. These are typical surfactants used in detergents, and, as shown by
the relative molecular weights of the hydrophobic and hydrophilic groups, the major portion
of the surfactant molecule is the hydrophobic group. This has economic significance because
the major cost component of detergents is the surfactant and the major cost component of
the surfactant is the hydrophobic group; therefore the main cost component of detergents is
the hydrophobic group of the surfactant used in the detergent. The cost of manufacture of the
surfactant and other raw materials will rarely equal the cost of the raw materials that supply
the hydrophobic group.
The detergent industry requires hydrophobic groups consisting of paraffinic chains, gener
ally saturated to give chemical stability and linear to give good biodegradability. The sources
of these paraffinic chains are either petroleum or animal fats [2] or vegetable oils [2]. Petro
leum itself is, of course, the result of biological action on plant material. Figure 2 gives a
general overview of the origin and processing of surfactant raw materials for anionic surfac
tants.
Figure 2 shows that there are some anionics, namely alkylbenzene sulfonates, alkane sulfo
nates, and olefin sulfonates, that are derived solely from petroleum. These are not discussed
in this chapter, but it should be borne in mind that these three groups represent the largest
production of anionic surfactants today, due mainly to their low cost and excellent detergent
properties. The other two groups of anionic surfactants—group I in Fig. 2, derived from fatty
acids, and group II, derived from alcohols—can be manufactured from lipids (fats and oil)
or petroleum.
A. Anionic Surfactants— Fatty Acids

In practice, very little of the fatty acids made from petroleum is used in the production of
surfactants and fatty acids for detergent use. In Europe, the United States, and Japan, fatty
acids are made from natural fats and oils consisting of glyceryl esters of linear, saturated, or
unsaturated fatty acids with an even number of carbon atoms. There are three principal groups
of fats and oils:
1. Vegetable oils with a high proportion of lauric (C 12) and myristic (C 14) acids, e.g.,
coconut oil
2. Animal fats and oils with a high content of palmitic (Ci^) and oleic (Cjs) acids
3. Vegetable oils containing high quantities of mono-, di-, and triunsaturated acids [oleic
(Cjg) acid, linoleic (Cjg) acid, linolenic (Cjg) acid] of varying ratios.
The fats or oils will contain many other natural impurities such as sterols, seed particles,
and dirt. The direct reaction of triglycerides with alkali gives a soap and glycerine; this is
described more fully in Section III. The fatty acids can be separated from soap by the addition
of inorganic acids, but the industrial manufacture of fatty acids is carried out by splitting the
triglycerides with water. The major problem of all large-scale processes is the insolubility of
water in the fat and the stepwise hydrolysis of the triglyceride to diglyceride to monoglyceride
to fatty acid. At high temperatures (260°C) and pressures (60 bar), water becomes more
soluble in the fat phase; therefore, fat splitting is carried out under such conditions, usually
in a continuous process, for large-scale manufacture where the unsaturated content of the fatty
acid is low. Less harsh conditions of hydrolysis must be used for fats and oils with polyunsat
urated oils to avoid the production of polymers. Category 3 products are also obtained directly
as fatty acids (known as tall oil) as a by-product in the Kraft process for paper.
Depending on the end use of the fatty acid derivative, the crude fatty acid can be used in
the manufacture of a surfactant, e.g., crude tall oil soaps. The great majority of fatty acid
based surfactants are made from purified fatty acids. Fatty acids are distilled to effect the
removal of impurities (e.g., sterols, phosphatides), and this is usually carried out under low
pressure to avoid oxidation of unsaturated acids. The fractional distillation process also re
moves short-chain fatty acids (e.g., Cg) and Cjo) that can give disagreeable odors. The fatty
acids can then be hydrogenated to remove any unsaturation; such a process is called harden
ing. For specific surfactants (e.g., sulfonated oleic acid), specific high purity fatty acids are
required that cannot be obtained by fractional distillation. Fractional crystallization [2,3] with
584 Porter
the aid of a solvent or a surfactant (known as hydrophilization [2]) processes are used in
this case.
The resulting fatty acids can then be used in the manufacture of various surfactants, but
fatty alcohols, many specialty surfactants, and soap are made from the methyl esters of the
fatty acids. The methyl esters can be made either by esterification of the fatty acids or by
transesterification of triglycerides with methanol [2]. Transesterification tends to be the domi
nant process as it will proceed under extremely mild conditions in the presence of an alkaline
catalyst at 50-70°C and atmospheric pressure, and therefore oxidation of unsaturated groups
is avoided. Higher temperature processes are often employed as these can not only transester-
ify but also esterify any free fatty acid mixed in with the triglyceride. The methyl esters are
also easier to distill than the corresponding fatty acids. In 1982, western Europe produced
190,000 tonnes of methyl esters of fatty acids, the majority being further converted to fatty al
cohols.
B. Anionic Surfactants— Fatty Alcohols

Fatty alcohols are produced from methyl esters of fatty acids by catalytic high pressure hydro
genation [2]. In principle, hydrogenation could start with the triglycerides, but the harsh con
ditions result in poor glycerine recovery and greater use of hydrogen and catalyst. The same
objection is found with the hydrogenation of fatty acids, which requires higher temperatures
and gives greater corrosion than using the methyl ester. There are various complex processes
for the production of fatty alcohols from methyl esters [3].
As Fig. 2 shows, fatty alcohols are also obtained from petrochemical sources, and here
there are two main routes, starting with either « -p araffin s or ethylene.
1. Via n-Parafftns and the Oxo Process

By chlorination of « -p araffin s and subsequent elimination of hydrogen chloride, a mixture of
linear olefins is obtained with olefin bonds distributed over the entire chain length. The subse
quent mixture of olefins is then subjected to the oxo process, also known as hydroformylation.
This proceeds as shown in Fig. 3.
Figure 3 shows the addition of an aldehyde group to the double bond in a long-chain olefin
(an “internal” olefin). The olefin product is usually a mixture of olefins in the range C jq- C i^
and contains both odd and even-numbered carbon chains. The aldehyde group is then hydro
genated to an alcohol group. The resulting fatty alcohol is a mixture of alcohols randomly
distributed along the chain. Some of this mixture will have the alcohol group at the end of
the chain. Therefore the fatty alcohol produced by this method will differ significantly from
the fatty alcohol produced from lipids via the methyl ester and hydrogenation in two ways:
R -C H = C H -R + CO + Ì-Ì2
R -C H 2-C H -R + R -C H -C H 2-R
CHO CHO
Fig. 3 The 0 x 0 process.

1. The alcohol group will be randomly distributed along the fatty alcohol chain as distinct
from the fatty alcohol from lipids, which have the alcohol group at the end of the
chain.
2. The paraffinic chain will have both odd and even numbers of carbon atoms, as distinct
from the fatty alcohol from lipids, which has only an even number of carbon atoms.
The first difference is important to the surfactant properties, as the shape of the subsequent
surfactant will differ depending on whether the hydrophilic group is at the end of the chain or
within it, and this will affect surfactant properties [4, pp. 39-42]. The second difference has
no significant effect on surfactant properties, but these petrochemical-derived alcohols are
chemically different from “natural” products and can therefore be easily identified as such
by analysis.
2. Via Ethylene and the A lfol Process

The differences between fatty alcohols from lipids and those from internal olefins were nearly
entirely overcome by the Ziegler process discovered in 1953 where the polymerization of
ethylene with triethylaluminum at high temperatures gave long paraffinic chains with even-
numbered C atom chains terminating in an aluminum atom. These compounds can be oxidized
and hydrolyzed (the Alfol process) to give long-chain primary alcohols very similar to those
obtained from lipids. Chemically the primary alcohols from both sources are identical except
for the distribution of the alkyl chain. The distribution of the fatty chain in the alcohol from
lipids will depend on the original distribution in the vegetable oil(s) used as feedstock and the
fractionation in the distillation stages of fatty acid, methyl ester, and fatty alcohol. In alcohols
from ethylene it will depend on the distribution in the Ziegler process and any fractionation
in the purification stages of the alcohol. With some degree of effort, a “synthetic” fatty alcohol
from ethylene can be made with properties identical to those of a fatty alcohol derived from
lipids.
The choice of which alcohol to use in the manufacture of surfactants and detergents will
depend on their relative cost and availability, but the concept of “mildness,” “natural materi
als,” and “green-like” can sometimes influence the choice.
RCOONa(soap)
RCON(CH:0CIbCOONa
(sarcosinates)
RC0 N(CH3)CH2CH2S03Na
(taurates)
RCOOCH2CH2S03Na(isethionates)
RCH(SO 3 Na)COOCH 3 (fatty ester

sulfonate)
Sulfated glycerides
Fig. 4 Surfactants from fatty acids.

586 Porter
There has been no attempt here to describe the very important roles that a-olefins, ethylene
oxide, and alkylbenzenes play as raw materials for the detergent industry as these are all
derived from petrochemicals.
III. SURFACTANTS DERIVED FROM FATTY ACIDS AND/OR

FATTY ACID ESTERS
A considerable number of surfactants can be made from fatty acids, and the main products
are shown in Fig. 4. The four main groups are (1) soaps; (2) fatty ester sulfonates or alpha-
sulfonated acids or esters; (3) sarcosinates, taurates, and isethionates; and (4) sulfated glyc
erides.
A. Soaps
Soap is defined as the product formed when glycerides (fats or oils) or fatty acids are heated
with alkalis or organic amines. Personal (toilet) soaps are made by reacting animal fats and/
or vegetable oils with caustic soda solutions. Glycerol is a by-product. A soap may be looked
upon as a salt of a fatty acid,
RCOOM
where R is a saturated or unsaturated paraffinic chain with 8-20 carbon atoms and M is an
inorganic or organic base.
The degree of ionization and hence the water solubility of the soap will depend on three
factors: ( 1 ) the length of the paraffinic chain, (2 ) the degree of unsaturation of the paraffinic
chain, and (3) the nature of the base (M).
Increasing temperature will increase the solubility of soaps in water. Table 3 shows the
effect of paraffinic chain length on water solubility. It shows that the longer the paraffinic
chain, the less water-soluble the soap. The incorporation of one unsaturated group (oleic) into
the soap molecule increases the water solubility of the soap significantly. Substituting potas
sium or an organic amine (e.g., ethanolamine) for the sodium will also improve water solubil
ity. The water solubility will determine the surfactant properties: the less soluble the soap, the
better the surfactant properties such as foaming, wetting, solubilization, dispersing of solids,
and detergency [5]. However, the amount of surfactant present in a solution will diminish as
the solubility decreases; therefore in practice there will be an optimum solubility for a particu
lar practical application. For a more detailed discussion, see Porter [4, pp. 26-89]. Different
practical applications will need different chain lengths of fatty acids or different bases to
achieve the optimum conditions. A toilet soap used in warm or even cold water will need
Table 3 The Effect of Paraffinic Chain Length and Unsaturation on the Water Solubility of Soaps
Temperature needed
Length of Unsaturation in to obtain a 10%
Product carbon chain carbon chain saturated solution (°C)
Sodium stearate 18 None 71

Sodium palmitate 16 None 63
Sodium myristate 14 None 51
Sodium tallowate Mixed 14/16/18 40% oleic 48
Sodium laurate 12 None 36
Sodium oleate 18 One double bond 23
shorter chain lengths or more unsaturation than soaps used at high temperatures for clothes
washing.
Soap needs to be stable on storage and not affected by exposure to air (autoxidation),
which could lead to rancidity or discoloration. Autoxidation is accelerated by catalysts such
as copper and iron. The raw materials used in the preparation of the soap must be pretreated
before saponification to remove color and insoluble impurities. The principal methods are
water washing, adsorption bleaching with activated earth, and alkali refining.
The direct saponification of oils is carried out in large open pans or kettles using live steam
(as a heat source) and agitation. In continuous methods the raw materials are mixed with strong
caustic solution at 75°C in a sealed reactor under pressure. Saponification is complete within 2 -
3 min; the reaction is exothermic, and recirculation takes place in the presence of preformed
soap, which acts as an autocatalyst. The soap is recovered by washing with hot brine (to remove
the glycerol) and possibly other purification processes depending on the raw materials. The final
product is neat soap, which is a viscous liquid at 90°C and is 30-33% water. This needs drying
(usually spray drying) and appropriate additives in order to become a usable soap. The recovery
and purification of the glycerol is very important for economic reasons.
The neutralization of fatty acids is a simpler process and is usually carried out by mixing
with caustic soda solutions at 110-120°C in a turbodisperser with recycle. No further purifi
cation is needed.
Soap made from the methyl esters of fatty acids is made in a manner similar to the neutral
ization of fatty acids except that the methanol that is formed has to be separated out.
Soap manufacture was originated by empirical methods, and it was found that the solid
soap produced had different properties depending on its water content and the heat treatment
given to the soap. The reason for this is that soap exists in a polymorphic crystalline state at
normal room temperature. At least four crystalline states have been identified by X-ray diffrac
tion methods and designated a, 8, co by Ferguson et al. [5]. Toilet soap contains predomi
nantly the jS phase and is formed in the soapmaking process by vacuum drying and rapid
cooling. The (3 phase gives a high degree of foam and swells easily in water, whereas the
other phases produce less foam and the soap tablet (bar) cracks easily.
B. Fatty Ester Sulfonates or Alpha-Sulfonated Acids

and Esters
Sulfonation at the alpha position of the fatty esters does not occur readily, and sulfur trioxide
must be used, leading to problems of poor color, particularly if the starting esters have unsatu
rated groups. The sulfonation reaction is complex (see Fig. 5), which summarizes the overall
reaction and indicates the complexity of the reaction and the formation of “disulfonates.” At
the end of the reaction there is a mixture of alpha-sulfonated ester acid and disulfonate.
Neutralization with sodium hydroxide gives the results shown in Fig. 6 .
The neutralization of a sulfonated fatty acid ester will give
1. The monosodium sulfonate salt of the ester, —80% of the active weight of constituents.
2. The disodium salt of the alpha-sulfonated carboxylate, —20% of the active weight
of constituents.
3. The sodium salt of the fatty acid produced by hydrolysis of the starting ester that was
not sulfonated, that is, soap, —3% of the active weight of constituents.
Thus commercial products will be a mixture of these three products.

Sulfonation is usually carried out in a continuous plant very similar to the one described in
the discussion of the preparation of alkyl sulfates (see Section IV). Even under optimum
588 Porter
RCH2COOCH3 ) 2SO3 ------- ^ RÇHCOOSO2OCH3
SO3H
disulfonate
RCHCOOSOîOC,^^ RCH2COOCH3 ------ ► 2RCHCOOCH,

I I
SO,H SO3H
alpha sulfonated methyl ester
Fig. 5 Sulfonation of fatty esters.
sulfonation, dark-colored products are reported in the literature, and special bleaching of the
acid and salts is often needed.
The solubility of the salts where R is a C 12- C 14 alkyl chain is excellent, but the solubility
of the salts where R is a Cj^-Cig alkyl chain is poor compared to other detergent sulfonates.
Salts of the disulfonate will tend to crystallize out of solution. The commercial products can
be solubilized with hydrotropes.
The lowest surface tensions obtainable are for the C 14, 34 m N /m ; for the Cj^, 33 m N /m ;
and for the Cjg, 33 mN/m. The foaming properties of the C 12- C 14 are better than those of the
C 16- C 18 sulfonated fatty esters; thus alpha-sulfomethyl tallowate gives low foam. Ci^-C^g
based detergents showed superior detergency to LABS in the absence of polyphosphates at
low concentration and at high water hardness.
Co Sarcosinates, Taurates, and Isethionates

The sarcosinates, taurates, and isethionates are produced from fatty acids by first making the
acid chloride and then reacting this with the appropriate active group as shown in Fig. 7.
More details of the preparation of these surfactants can be found in Porter [4, pp. 99-168].
The need to make the acid chloride intermediate has severely restricted the use of these prod
ucts compared to many other surfactants, mainly on account of the extra cost. In addition, the
products will contain considerable quantities of sodium chloride unless they are purified,
which again adds to the cost.
alpha sulfonated ester

NaOH
RCHCOOCH3 RCHCOOCH3 RCHCOONa
i I 1
SO3H SOjNa S03Na
monosodium salt disodium salt
disulfonate
NaOH
RCHCOOSO2OCH3 RCHCOONa
I
SO3H S03Na
disodium salt
Fig. 6 Neutralization of sulfonated esters.

R CO N CH 2CO O N a
I sarcosinates
C\h
RC0NCH2CH2S03Na
RCOCl (taurates)
CH,
RC00CH2CH2S03Na
(isethionates)
Fig. 7 Manufacture of sarcosinates, taurates, and isethionates.
The sarcosinates and taurates can be looked upon as modified soaps with the alkaline salt
of the carboxy group still present but modified with an additional polar group— in the case of
the sarcosinate with the amide group, and in the case of the taurates with a sulfonate group.
Either of these additional groups improves water solubility and reduces the effect of hard
water. In the case of the isethionate, the carboxyl group has been replaced with the sulfate
group, again improving water solubility and reducing the adverse effects of hard water.
Before the introduction of synthetic surfactants from petroleum, these three products were
used to overcome the disadvantages of soap in hard water. The introduction of cheaper petro
chemical-based products led to the decline in their use. However, there is now considerable
interest in this group of products, particularly the isethionates, due to their non-petroleum
origins, and they have been found to cause little eye and skin irritation. Excellent shampoos
can be formulated due to their high foaming properties.
D. Sulfated Glycerides
Reaction of sulfuric acid (usually with oleum) with the glyceride of a fatty acid, i.e., an
animal fat or vegetable oil, will usually give a complex mixture of sulfates and sulfonates.
The type of products formed will depend on the nature of the glycerides and the conditions
of sulfation. The most common reactions are the following.
1. If a hydroxyl group (e.g., castor oil) is present, then sulfation of the hydroxyl group
proceeds preferentially.
R—OH + H 2S04^ R -O S O 3H
2. If a carbon-carbon double bond in the hydrocarbon chain of the fatty acid (e.g., oleic
acid) is present, this will sulfonate:
— CH = CH- - + H 2S04^ — CH(S03H)CH2
3. If a fatty monoglyceride or diglyceride is present, then sulfation of the free —OH
group proceeds as in 1, above. Hydrolysis of the ester group in triglycerides will
occur in acid solution followed by sulfation of the hydroxyl group. Alternatively,
transesterification of triglycerides in acid solution can take place if excess glycerol
is present.
590 Porter
Even with one specific oil, a large number of products with different properties can be
made depending on the conditions of sulfation and the addition of fatty acid, soap glycerol,
glycols, or alcohols. Neutralization will give mixtures of sulfated products, soap, and free
oil, very often in the form of emulsions and microemulsions.
The most common sulfated oils are Turkey red oil (ricinoleic acid triglyceride) made from
castor oil; sulfated methyl and ethyl ricinoleates; sulfated methyl esters such as butyl oleate;
and fish oil, lard oil, tallow, palmkemel oil, tall oil, and rapeseed oil sulfates.
Although the products are complex mixtures, they are all sulfate esters and subject to acid
and alkaline hydrolysis. In addition, any ester groups from the glyceride portion of the mole
cule will also be susceptible to hydrolysis. The sulfated monoglycerides are excellent deter
gents. Partial glycerides containing unsaturated acids or hydroxy acids have more than one
site available for sulfation and possess specific wetting, emulsification, and other properties.
Sulfated castor oils (e.g., Turkey red oil) will give good wetting, penetrating, and emulsi
fying properties if sulfated to a high degree. Sulfated methyl esters have good wetting proper
ties but usually with low foam, which is rare in anionic surfactants.
The sulfated monoglycerides of coconut fatty acids were once made on a very large scale
for use in household detergents but have been superseded by synthetic surfactants from petro
chemicals. Sulfated castor oil was used as emulsifier for pine oil and creosote. The ammonium
salt of cocoacid monoglyceride sulfate was used as the basis for a popular shampoo in the
United States. The product is very similar to the equivalent lauryl sulfate, but due to the
additional hydroxyl groups it has slightly better water solubility.
IV. SURFACTANTS DERIVED FROM FATTY ALCOHOLS

A. Fatty Alcohol Modification
The production of fatty alcohols has already been described in Section II, and many types of
anionic, nonionic, and amphoteric surfactants can be made from these alcohols. However, the
manufacture of anionic surfactants from fatty alcohols, e.g., by sulfation, is often preceded
by modification of the fatty alcohol. Fatty alcohols will readily react with ethylene oxide,
propylene oxide, butylene oxide, and mixtures of these oxides (see Fig. 8).
Figure 8 shows n molecules of ethylene oxide reacting with one molecule of a fatty alco
hol. The overall result is the addition to the alkyl chain of the fatty alcohol (of a polyethoxy
chain) of 2n carbon atoms joined by oxygen but with a primary alcohol group at the end of
the product molecule. The resulting compound, a fatty alcohol ethoxy late, will react with
chemical reagents in the same manner as the starting fatty alcohol as long as the chemical
reagent does not attack the C— O— C group that is present in the polyethoxy chain. The
polyethoxy chain confers increased water solubility on the fatty alcohol, and with a fatty alkyl
group of C i2_i4 and n = 6 - l0 the products are excellent wetting agents and detergents. By
varying the amount of ethylene oxide, the water solubility of the starting fatty alcohol can be
increased by increasing the value of n [4, p. 170]. The addition of propylene oxide or butylene
oxide to fatty alcohols proceeds in exactly the same manner, i.e., the addition of a polypro-
poxy or polybutoxy chain to the fatty alcohol with a hydroxyl group at the end of the chain.
R-OH + nCH2=CH2 R-(OCH2CH2)n OH

\ /
O
Fig. 8 Reaction of a fatty alcohol with ethylene oxide.
However, in the case of the addition of propylene oxide or butylene oxide the solubility of
the fatty alcohol will decrease. The practical result of the addition of alkylene oxide is that
starting with one fatty alcohol the water solubility or hydrophobic/hydrophilic properties of
the raw material can be varied in a controlled manner.
B. Sulfation
In the period 1940-1960 sulfates from both lipids and petrochemicals were the largest class
of surfactants, but during 1970-1990, they were replaced by the sulfonates from petrochemi
cals in terms of volume consumption. However, the alkyl sulfates are now being increasingly
used in domestic detergents and consumption is increasing again.
Organic sulfates are the esters of sulfuric acid,
ROH + H 2S04^ ROSO3H + H2O
The sulfur atom is joined to the carbon atom of the hydrophobic chain via an oxygen atom.
The acid ester is unstable and can revert back readily to the alcohol and sulfuric acid (particu
larly in acid conditions), whereas the neutralized salts are stable at neutral pH. The acid
hydrolysis releases sulfuric acid, which catalyzes the hydrolysis.
Acidic conditions:
ROSO3H + H2O ^ ROH + H2SO4
Alkaline conditions:
R 0 S03Na + NaOH ^ ROH + Na2S04
The alcohol sulfates are therefore more stable in alkali than in acid. The stability depends on
a number of factors, but for aqueous formulations, to give long-term shelf stability at room
temperature, a pH range 5-9.5 is preferred. At elevated temperatures (>50°C); sulfates are
unstable. It is for this reason that sulfation is generally carried out in continuous plants and
the acid that is formed is immediately neutralized. Sulfation can be carried out with oleum,
chlorsulfonic acid, or sulfur trioxide,
R 0H + S03 - ^ R 0 S03H
The most efficient plants in terms of cost and economics use sulfur trioxide. See Fig. 9. The
alcohol flows downward under gravity as a thin film on the inside walls of the tubular reactor,
which is cooled. The sulfur trioxide and air pass downward cocurrently with the liquid phase.
In the lower sections of the reactor, some separation of the liquid and gaseous phases takes
place, but often the separation of the gases is carried out in a separate section (as shown in
Fig. 9). The reaction is very fast in the tubular reactor; the air dilutes the reaction mixture
and helps in cooling. The big advantage of this type of equipment is that very little degrada
tion takes place, so very light colored materials can be produced. The limitations are in the
molecular weight of the alcohol; below about a C jq chain length the fatty alcohol is Volatile
in the air stream, and at long chain lengths (>Cig) it is not liquid enough to form a thin film
and flow downward. There are a number of different types of reactors, but most depend on
the principles illustrated in Fig. 9.
The properties of the sulfates depend on the properties of the hydrocarbon chain and those
of the sulfate group. The alkali metal salts show good solubility in water, and most surfactant
sulfates show what is known as the salt effect. The addition of inorganic salts (sodium chlo
ride, sodium sulfate) to dilute (below about 30%) solutions increases the viscosity, whereas
addition to concentrated solutions decreases the viscosity. Any salts formed during the sul
fation and/or neutralization can lead to viscosity changes.
592 Porter
Aqueous solution of
Na salt of the alcohol
sulphate
Fig. 9 Continuous sulfation plant using SO3. (Courtesy of M. R. Porter.)
A sulfate group can replace a hydroxyl group at the end of the alkyl chain (from fatty
alcohols) or the end of the polyethoxy chain (from the fatty alcohol ethoxylates). See Fig. 10.
The addition of the sulfate group gives a marked increase in water solubility of the surfactant
and generally gives high foam, good wetting, and excellent detergency if the alkyl chain is in
the region of C 12- C 16.
The sulfated alcohols and sulfated ethoxy alcohols are by far the most important surfactants
made from alcohols derived from lipids, but there are other speciality surfactants, which are
summarized in Fig. 11. The R group in Fig. 11 can be either a paraffinic alkyl chain or a
R-(OCH2CH2)nOH + SO3 ----- ► R-(0CH2CH2)n0S03H
Fig. 10 Sulfation of an ethoxylated alcohol.
R-OPO3 H Phosphate esters
r-oco |: h 2
R-OH
NaOCOCHSOsNa sulfosuccinate
R0(CH2CH20)nCH2 COONa
Ethoxy carboxylates
ROCO(R')COONa
Ester carboxylates
Fig. 11 Other specialty anionic surfactants.

modified chain with ethylene oxide, propylene oxide, or butylene oxide, and therefore count
less variations are possible. All these specialty surfactants are manufactured and used in deter
gent formulations. They are very rarely used alone; they are nearly always used in conjunction
with other anionic compounds as described here or with nonionic and/or amphoteric surfac
tants. Where mildness to skin and eyes is required, the ethoxy carboxylates have been shown
to be particularly good and are also stable in electrolyte solutions [6].
V, HOUSEHOLD DETERGENTS
Approximately 50% of all synthetic detergents in western Europe are used in household prod
ucts, and anionic surfactants are the major type used in detergent formulation. About 90% of
soap is used in personal washing.
Within a household, many different types of detergents will be used, but there are three
major groups of products, those used for
1. Personal washing, e.g., toilet soap, shampoos
2. Washing clothes
3. Washing hard surfaces, e.g., dishwashing products
Within each of these groups there are many individual products. This is in marked contrast to
the situation 50 years ago, when often only one or two soap-based products would suffice. In
washing clothes the development of the automatic washing machine has generated the need
for quite different properties, e.g., low foam in a horizontal drum machine. Automatic dish
washers are becoming more common in the home and the modem tendency to shampoo every
day has led to the need for mild shampoos. In addition, detergent manufacturers need to
market different products for hard water areas than for soft water areas. Environmental de
mands, real or perceived, also generate different needs depending on the local legislation or
public awareness. Finally, the activities of marketing departments will continually generate a
need to make products “different” from those of their competitors. The result Is that a typical
1990s household may have 20-30 individual detergents for household cleaning and personal
use.
To satisfy all these demands, technical, economic, and environmental, requires a wide
variety of surfactants possessing different properties and characteristics, but it is uneconomical
to make and use a different surfactant in each different detergent. The answer has been found
in formulating mixtures of chemicals that will give these different properties. The surfactant,
or mixture of surfactants, is the main functional chemical in any detergent formulation. The
properties of the surfactant can be modified considerably by adding other chemicals. In addi
tion it is well known that mixtures of surfactants can often produce synergistic effects, i.e.,
that a mixture of surfactants can produce a particular property that cannot be obtained by any
of the individual surfactants in the mixture. Thus the skill of the detergent manufacturer is
not in the organic chemistry of surfactant manufacture but in the physical chemistry of mixing
surfactants and other chemicals to produce the required physical properties at the appropriate
cost. The formulation of detergents is still very much an empirical science, but in recent years
there has developed a much better understanding of the theory of detergency, the theory of
surfactant behavior, and the theory of synergism, which helps in understanding the mode of
action of components in a formulation.
A. Personal Washing
Solid soap tablets (bars) are the preferred form of detergent for most personal uses with the
major exception of shampoos (see Section VI). In 1990 the U.K. toilet soap market was
594 Porter
50,000 tonnes, and this has been slowly decreasing for some years. The reason for the reduc
tion is the increasing popularity of non-soap-based products such as bathing and shower prepa
rations, facial cleaners, and scrubs. Unlike the United States, there are no major brands of
non-soap or ‘‘syndet” toilet bars based on synthetic detergents.
The traditional bar of soap was mainly tallow soap (80% tallow-20% coconut oil); the
modem bar was originally formulated as a feminine cosmetic soap but has achieved more
general use. It was originally made from a blend 50% beef tallow and 50% coconut oil with
the addition of 8-9% fatty acid as the superfatting ingredient. Modem superfatted soaps are
made from a blend of 60-70% tallow and 30-40% coconut oil with 5-7% coconut oil fatty
acid added to the liquid soap prior to drying. Superfatted soaps give a smooth feel to the skin
during and after washing due to the adsorption of the free fatty acid. They have a shorter in-
use life than the traditional soap (80% tall o il- 20% coconut oil) as they tend to form a mush
in the presence of water. Soaps can also be superfatted with different emollient systems, e.g.,
light mineral oil, natural oils, fatty alcohols, polyethylene glycols. Other ingredients in soap
can be dyes, perfumes, and opacifiers.
Transparent soaps are made from carefully selected blends of tallow, palm oil, coconut oil,
palmkemel oil, oleic acid, and castor oil, and the glycerol is not removed but stays in the
product. Translucent soaps are prepared from toilet soap formulations to which crystallization
inhibitors have been added before drying and by increased mechanical working. The translu-
cency is produced by a soap stmcture that has both solid and liquid phases, the two having
different crystalline stmctures and refractive indexes. Working the soap and/or adding crystal
inhibitors brings about a new equilibrium between the two phases in which the large crystals
of the solid phase are reduced in size and some of the solid phase is changed into a liquid
phase. Both of these increase the soap’s translucency.
The most important characteristic of a toilet soap is the foam or lather. The volume and
type will depend on the chain length of the fatty acids in the soap. Maximum foam production
occurs with soap made from C 12- C 14 saturated fatty acids, which is why coconut oil is a
necessary component in toilet soaps, with the longer chain (Ci^-Cjg) fatty acids in the tallow
soap being more responsible for the physical form of the soap. The increased level of C 12-
Cj4 fatty acids in the superfatted soaps gives a much richer and denser foam than the old
conventional soaps.
Baby soaps, those containing a minimum of nonfunctional additives, are made from the
traditional (80% tallow - 20% coconut) blend with the possible addition of glycerol as plasti
cizer. These are claimed to be “pure” soaps.
Medicated soaps are essentially toilet soaps that contain mild medicaments to assist in the
treatment of minor skin problems associated with spots and pimples. Deodorant soaps are
usually based on the traditional soap with a bactericide and a fragrance that is “fresh” smell
ing. Shaving soaps must produce a large volume of fine, stable lather. The lather must wet
the beard and not dry out during use (be stable), and it must be not irritating to freshly shaven
skin. Shaving soap is made from coconut oil and commercial stearin (80% stearic acid-20%
palmitic acid) saponified with a mixture of sodium and potassium hydroxides with the addition
of 5-6% glycerol and then adjusted to pH 7 with free fatty acid.
Liquid soaps are a small market, and most of the products found in domestic kitchens do
not contain soap at all but are surfactant blends that are mild in nature (alkanolamides, ether
carboxylates, betaines, etc.). True liquid soaps are made by the saponification of liquid fatty
acids such as coconut or palmkemel oils and oleins with potassium hydroxide or blends of
monoethanolamine and triethanolamine.
Non-soap bars are known as “syndet” (synthetic detergents) or “combo” (combinations of
soap and synthetic detergents). Both types are designed to be superior to ordinary toilet soap
in relation to their performance in hard water, in mildness, and in skin care (perceived or
real). In Europe these products have not achieved any significant success compared to the
United States, possibly due to the increased interest in specialty liquid products for bathing
and showering. Most European products are based on the use of combinations of fatty alcohol
sulfates (the fatty alcohol from lipids), isethionates (from fatty acids), and sulfosuccinates
(from fatty alcohols, probably from lipids). Products used in the United States are based on
sodium cocomonoglyceride sulfate, sodium cocoisethionate, and sodium cocoglyceryl ether
sulfonate.
In recent years, showering has become the most popular form of personal hygiene. It is
quick, efficient, and energy-saving and uses far less water than tub bathing. Liquid and gel
shower preparations are tending to replace soap bars.
Special hand cleaners for removing oil and grease are also found in the home. These are
described in Section V.D.
B. Washing Clothes
The nature of present-day textiles is extremely diverse due to the availability of various fibers,
finishes, dyes, and manufacturing processes. Many fabrics cannot be easily washed in water
but need to be dry-cleaned, a process that involves the use of specialty anionic surfactants
(sulfosuccinates). However, the tendency is for more and more garments to be washed in
automatic washing machines. Nevertheless there is still a wide variety in washable fibers and
finishes. For instance, the proportion of colored materials in an average household wash is
increasing. In the Federal Republic of Germany in 1970, 46% of clothes were white, but by
1980 the proportion of whites had fallen to 2 1 % . Some textile dyes will bleed at high tempera
ture, and this is one reason for the significant fall in average washing temperatures in the last
20 years.
Detergents for use in washing machines are known as heavy duty detergents to differentiate
them from light duty detergents, which are used for hand washing temperature-sensitive gar
ments such as woollens. Two typical formulations for a heavy duty detergent in powder form
are shown in Table 4. These two formulations are typical, but actual formulations can vary
widely. The chain length of the alkylbenzene, the fatty alcohol, and the soap; the amount of
ethylene oxide on the fatty alcohol; and the alkali metal in the sulfonate and soap are not
specified, so many variations are possible. Formulations in the United States, for instance,
may well have a fatty alcohol ether sulfate instead of the fatty alcohol sulfate and probably
much higher levels of the nonionic component. The two formulations shown represent mix
tures containing phosphorus derivatives and those not containing phosphorus. The demand for
non-phosphorus-containing formulations is entirely due to environmental needs in preventing
algal growth in lakes and slow-moving rivers. The soap would be derived from lipids. The
fatty alcohol for the ethoxylate and the fatty alcohol sulfate could be derived from lipids or
petrochemicals; the choice would depend on their relative cost.
More recent trends have been the development of compact powders and liquid products.
Liquid products have the advantage of cheaper production costs as the large spray driers used
in powder manufacture are not needed and energy costs are reduced, but the packaging costs
are higher for equivalent detergent performance. Liquid products are widely used in the
United States but only to a small extent in Europe. The difference is due to the prevalence of
a separate bleach step in the United States whereas the bleach is incorporated in the detergent
in Europe. If the bleach is present in a liquid detergent it becomes unstable on storage,
although there has been considerable research and development in liquid stable bleaches and
enzyme systems. The major problem with liquid products is that the inorganic builders (e.g..
596 Porter
Table 4 Heavy Duty Detergent Powder Formulations in Western Europe
Content (%)
Ingredient With P No P Function
Linear alkylbenzene sulfonate 8 10 Main anionic surfactant; removes

polar dirt
Fatty alcohol sulfate 2 —
Subsidiary anionic surfactant;
synergistic effects
Fatty alcohol ethoxylate 7 5 Nonionic surfactant; removes
oily dirt
Soap 3 3 Foam-controlling agent
Fatty acid monoethanolamide 1 — Foam booster
Sodium triphosphate 30 — Chelater and builder
Zeolite A 5 25 Ion exchange for hard water
Sodium carbonate — 7 Alkali acts as builder
Sodium citrate 2 — Co-builder
Sodium perborate 15 20 Bleach
Tetraacetyl ethylenediamine 1 1.5 Bleach activator
Ethylenediamine tetraacetate 0.3 0.3 Bleach stabilizer
Cellulose ether 1 1 Antiredeposition agent
Stilbene-disulfonic acid 0.2 0.2 Optical brightener
Sodium silicate 3 3 Corrosion inhibitor
Fragrance q.s. q.s. Odor
Dyestuffs q.s. q.s. Color
Sodium sulfate Balance Balance Filler
q.s. = as necessary
phosphates) present solubility problems for the surfactants because an increase in electrolyte
content will decrease the solubility of an anionic surfactant. This led to the development of
liquid detergents without phosphate builders and a high concentration of surfactants. A typical
formulation would have 30-40% active surfactant material of composition one-third linear
alkylbenzene sulfonate (LABS), one-third soap, and one-third fatty alcohol ethoxylate; the
soap would be derived from lipids, but the fatty alcohol could be derived from petrochemicals
or lipids, probably the former. In addition, products based on mixtures of fatty alcohol sul
fates, soap, fatty alcohol ethoxylates, and amphoteric compounds (alkyl polyaminocarboxyl-
ates) that do not contain LABS or builders have been put forward.
“Compact detergents” is the name given to heavy duty powder detergents that have a much
lower filler level than the ones shown in Table 4. Such products can be based on the use of
fatty alcohol sulfates to wholly or partly replace the LABS of the conventional detergents.
The fatty alcohol can be derived from lipids or petrochemicals.
There are a number of specialty detergents used for washing clothes, particularly in Eu
rope. The major ones are intended for washing wool by hand and for washing curtains. Those
based on synthetic surfactants are free of anionics and are based on nonionics with added
cationics to give a soft handle. There are some products based entirely on flaked toilet soaps
that are used to wash very delicate fabrics; the flake form is necessary to obtain rapid solubil
ity in warm water.
The major “workhorse” in modem heavy duty detergents has been linear alkylbenzene
sulfonate (LABS), and from a cost-effective point of view it is probably still the most efficient
surfactant on which a heavy duty detergent can be based. However, there have been a number
of criticisms of LABS, particularly with respect to biodegradation, leading to extensive re
search for an environmentally friendly substitute (see Section VIII). The major candidates
have been the alkyl polyglycosides (APG), the sulfonated fatty acid esters (FES), and the
alkyl sulfates (AS). The APGs are apparently nonionic in constitution but behave very simi
larly to anionics in aqueous solution [7] and are made from glucose and fatty alcohols [10],
which can be derived from lipids. The FES are derived from the glycerides via the fatty acid
or ester (see Section III). The alkyl sulfates are derived from the corresponding fatty alcohols,
which can be sourced from either lipids or petrochemicals (see Section IV). With alkyl sul
fates, the effect of chain length and addition of nonionics is important to obtaining an efficient
heavy duty detergent [8].
C. Hard Surface Cleaners

There are numerous hard surfaces inside and outside a home— walls, floors, windows, ceil
ings, worktops, dishes, furniture, toilets, etc. The most common are dishes, which have to
be washed several times a day. The primary purpose of all dishwashing agents is the removal
of soil from hard surfaces, the soil consisting of food material residues. In the cleaning of
other hard surfaces, the soil composition varies much more widely. The need for different
detergents for the different types of soils often means that the most efficient cleaning agents
are not compatible and often cannot be combined in one universal hard surface cleaner. If the
article, e.g., a dish, can be completely immersed in the detergent, then the temperature can
be raised and the cleaning will be much more efficient. For example, the use of a dishwashing
machine gives a degree of cleanliness not possible using manual dishwashing. However, the
majority of hard surfaces in the home cannot be immersed in the detergent. The surfactants
used in formulations for hard surface cleaners are basically similar to heavy duty detergents,
i.e., mixtures of anionics and nonionics, but they have much lower levels of builders or none
at all.
1. D ishw ashing
The oldest agent for washing dishes was the running stream of water. It was found that an
alkaline aqueous solution (e.g., soda ash) was much more efficient than plain water combined
with agitation but had the serious shortcoming of adverse skin reactions with manual dish
washing. These disadvantages are not present in machine dishwashing, and therefore the basic
detergent in machines is an alkaline solution with phosphates and silicates with a low foam
nonionic wetting surfactant.
In manual dishwashing the most difficult soil to remove consists of the fats, colorants, and
tannins, which are all insoluble in water. However, there is usually vigorous agitation, and it
has been found that a surfactant solution that will wet the greasy plates, disperse the solid
particles, and emulsify the liquid insoluble oils and greases is perfectly adequate without the
need for alkali or builders. The surfactant must operate at relatively low temperatures and in
hard water areas; this excludes most soaps. Surfactants with the sulfate or sulfonate group are
very cost-effective dishwashing detergents; linear alkylbenzene sulfonates, a-olefin sulfonates,
paraffin sulfonates, fatty alcohol sulfates, fatty alcohol ether sulfates, and fatty ester sulfonates
(FES) can all be used to make efficient dishwashing detergents either alone or in mixtures.
Probably the best known mixture is the 3:1 mixture of a linear sodium salt of a Ci2_i5 alkyl
benzene sulfonate and the sodium salt of a Ci 2_i4 fatty alcohol plus 3 mol of ethylene oxide
sulfate. This mixture has been very widely used as it shows a synergistic detergent effect.
Normally small quantities of a hydrotrope (sodium xylene sulfonate) and a foam stabilizing
598 Porter
Table 5 Typical Manual Washing-Up Formulation
Component Percent (w/w)
LABS 12
C i2_ i6 alcohol, (E0 )3S0 4 Na 4
Cl2-14 diethanolamide 2
Sodium toluene sulfonate 2
Salt As needed to thicken
Perfume/coloring q.s.
Water To 100
q.s. = as necessary
agent (coconut diethanolamide) are present plus perfume, coloring, and salt (which thickens
anionic surfactant solutions). The surfactants are generally derived from petroleum as such
surfactants are usually cheaper, but the fatty alcohol ether sulfate could originate from lipids,
and FES formulations (based on fatty acids) are very efficient. A typical formulation is shown
in Table 5.
In recent years there have been some changes in formulations, mainly to make milder (non
skin-irritating) products, as the surfactant mixtures described above readily defat the human
skin, which can lead to skin irritation problems. The major changes are the replacement of
LABS by fatty alcohol sulfates or fatty alcohol ethoxy sulfates and/or the addition of ampho-
terics, particularly the alkyl polyaminocarboxylates. The replacement of LABS by fatty alco
hol sulfates can reduce skin irritation, but the addition of amphoterics to anionics has a pro
found effect on reducing skin and eye irritation [9]. There are also claims that paraffin
sulfonates-amphoteric and fatty ester sulfonate-amphoteric mixtures are equally cost-effec
tive, and it is not clear which will be the most cost effective formulations. If these new
dishwashing detergents are claimed to be made from natural raw materials, then it is likely
that the fatty alcohols used in the fatty alcohol sulfates will originate from lipids. Fatty acid
derivatives such as the fatty ester sulfonates and the alkyl polyaminocarboxylates would also
seem to have a bright future. There is also interest in formulations with products that are non-
petroleum-based. A recent German patent [10] specifies alkyl polyglycosides as the nonionic
component with anionic components from non-petroleum sources.
2. G eneral-Purpose H ard Surface Cleaners

Hard surface cleaners for walls, floors, bathtubs, washbasins, worktops, etc. are liquids of a
composition similar to that of the washing-up liquids but with the addition of a builder and a
solvent for the oily material. A wide variety of surfactants can be used as long as they are
soluble and the chain length is the optimum for detergency. Alcohol sulfates, fatty alcohol
ether sulfates, paraffin sulfonates, and fatty ester sulfonates can all be used. A typical formula
tion using ether sulfates is shown in Table 6.
3. H ousehold D isinfectants and D etergent Sanitizers

Household disinfectants are commonly used in many situations. A concentrated solution is
poured into an excess of water to produce a cloudy appearance with the separation of insoluble
germicidal products that will adhere to solid surfaces. At one time the surfactant used was a
soluble soap produced by saponification of castor oil with caustic soda. The major constituent
of castor oil is 12 -hydroxy oleic acid, and the hydroxyl group gives increased water solubility
Table 6 General Hard Surface Cleaner
Cl2-14 fatty alcohol ether (2EO) sulfate, sodium

salt 6
Tetrapotassium pyrophosphate 5
Sodium carbonate 2
Butyl cellusolve 2
Perfume/preservative q.s.
Water To 100
q.s. = as necessary
to the soap. The disinfectant would be a 10% active solution of the castor oil soap with 2 -
5% chlorinated xylenols as bactericide but would also contain a small quantity of pine oil or
terpene derivatives to give a “fresh pine” smell. In recent years synthetic surfactants (ether
sulfates) have replaced the castor oil soap because of the price volatility of the castor oil.
Detergent sanitizers are normal detergents that have added bactericides that will leave sur
faces both clean and sterile, e.g., for toilets and bathroom tiles. The disinfectant described
above will show detergent properties if the surfactant content is increased. Sodium hypochlo
rite is a cheap but very effective bactericide and is used in household products, but the addi
tion of many surfactants will destabilize and deactivate the bactericide, so the choice of surfac
tant is very important.
¥1. SHAMPOOS
A. Shampoo Formulations
Hair naturally has an oily layer that tends to collect dirt. A shampoo should remove the dirt
and some or most of the oil, which has to be replaced or the hair will appear dry and be harsh
to the touch. Hence the need for conditioners that will add “body” to the hair and replace the
oil until the natural oil is renewed. Anionic detergents have been found to be very efficient
for removing the soil and the natural oil. The most useful conditioners are the cationics,
which are normally incompatible with anionics and are best applied separately. Alkanola-
mides, although considered nonionic surfactants, are compatible with anionic surfactants yet
have enough substantivity to hair to act as conditioners. “Conditioning” is a phrase that covers
prevention of dry hair static, increased body, mending of split ends, improved combability,
and improved gloss. The alkanolamides also act as foam-stabilizing agents in conjunction
with anionics. Many formulations therefore contain mixtures of anionics (fatty alcohol sulfates
or fatty alcohol ether sulfates) and alkanolamides. A typical simple formulation for a shampoo
in the form of a clear gel is given in Table 7. The addition of sodium chloride will thicken
the aqueous solution to a thick gel. Most anionics, particularly fatty alcohol ether sulfates,
will increase in viscosity when an electrolyte is added. This is known as the salt effect and
can be very dramatic, the viscosity increase changing a low viscosity liquid into a thick gel
by the addition of only 1-3% of sodium chloride. The exact amount of sodium chloride will
depend on impurities in the ether sulfate, the amount and type of alkanolamide, and the
amount and type of betaine present.
Shampoos are generally used straight from the bottle, and a thin liquid shampoo is not
easy to use, particularly in a shower. Therefore control of the rheology of the shampoo is
600 Porter
Table 7 Shampoo Formulations
Percent (w/w)
Component Clear gel Frequent wash
Fatty alcohol ether (2EO) sulfate, sodium salt 15 7

Coconut diethanolamide 4 1.5
Cocoamidopropylbetaine — 3
Sodium chloride q.s. q.s.
Preservative/coloring/perfume q.s. q.s.
Water To 100 To 100
q.s. = as necessary
very important. Shampoos are made in the form of clear liquids, lotion, cremes, pearlized
liquids, and gels. The rheology and appearance is controlled by the type of anionic detergent,
the type of alkanolamide, and the addition of electrolyte or other additives. The properties of
the foam generated during shampooing will depend on these same factors, and therefore the
exact formulation will depend on the assessment and testing of three factors—the rheology of
the shampoo, the type and stability of the foam produced during shampooing, and the cleanli
ness and conditioning of the hair.
The ether sulfate can be replaced by fatty alcohol sulfate, but the ammonium or triethano
lamine salt is preferred to give better solubility. The fatty alcohol used in the manufacture of
these surfactants can originate from either lipids or petrochemicals, which function equally
efficiently, but most large shampoo manufacturers will specify a lipid origin or a composition
that is indistinguishable from one derived from lipids. The optimum chain length of the fatty
alcohol has been found to be in the C jo- h range, with Cj 2 being predominant. The ether
sulfates will give large copious foams. The fatty alcohol sulfates give denser, creamier foams
depending on the distribution of the alkyl chain lengths.
The gel formulation shown in Table 7 is very efficient and will clean both dirt and natural
oil from hair, particularly if used frequently. Reduction in the concentration of the ether
sulfate and addition of a cocoamidopropylbetaine will modify the detergent properties and
conditioning properties. The frequent wash formulation shown in Table 7 helps to prevent
damage to the hair and improves the conditioning.
B. Frequent Wash Formulations and “Mild” Surfactants

In recent years the shampoo market has changed to meet the demands of a more discerning
consumer with a busy lifestyle. There has been a trend toward milder formulations because
of the tendency to wash the hair every day and to wash it while showering. Daily shampooing
can cause damage to the hair (e.g., split ends) but also prevents the formation of the natural
oils that lubricate and help manage the hair. In addition, shampooing during a shower can
facilitate the entry of the shampoo into the eyes, and hence very low eye irritation becomes
a requirement. Low irritation is generally accompanied by low skin irritation, which in it
self is highly desirable. Shampoos for babies are probably those needing the mildest condi
tions.
There are various routes to a mild shampoo. The easiest and cheapest is to supply a much
diluted normal shampoo, but these usually suffer from inadequate performance. A second
Eye
Irritation
H)0
50
100 SEES GAP AC 100 SEES 0 APAC

0 APAC 100 SEES 0 APAC 100 SEES
Fig. 12 Detoxification of ethyl sulfates with alkyl polyaminobetaines. (From Ref. 9.)
route is replacement of the major eye irritant in the standard shampoos. This was found to be
the anionic fatty alcohol sulfate or fatty alcohol ether sulfate. Variations in the alkyl chain
length, the cation (magnesium is less irritant than sodium), and the degree of ethoxylation
could reduce the irritancy but not to a low enough level without adversely affecting perfor
mance. Complete replacement with imidazoline-based amphoteric surfactants, sulfosuccinates,
sarcosinates, and especially ethoxycarboxylates, all of which are derived from lipids, is tech
nically possible to give excellent shampoos but is very expensive. Nevertheless, the amphoter-
ics have been used in the United States for many years in Johnson and Johnson baby shampoo.
Finally, a mild shampoo can be achieved by partial replacement of the anionics, particularly
of fatty alcohol ether sulfates by other surfactants. This was found to give a cost-effective
solution, as it was found that the addition of small quantities of some surfactants to anionics
reduced the eye irritation to a very large degree. Addition of betaines, sulfosuccinates, sarcosi
nates, ethoxycarboxylates, amidopropylbetaines, or alkyl polyaminobetaines can give this ef
fect, known as detoxification. Figure 12 illustrates detoxification with alkyl polyaminobe
taines.
SUES is the sodium salt of a Ci 2_i4 fatty alcohol ethoxylated with 2 or 3 mol of ethylene
oxide and sulfated with sulfur trioxide. APAC is the tallow (Ci 6_ig) alkyl polyaminocarboxyl-
ate. The curves show remarkable similarity and show that eye irritation is closely related to
the critical micelle concentration (CMC). Lomax [9] describes this effect as due to the anionic
SLES forming a mixed micelle, which gives a lower CMC. The effect of a lower CMC is
that there are fewer free monomeric surfactant molecules outside the micelles (as long as the
concentration is above the CMC). Eye irritation can be due to the adsorption of surfactant
monomer at the eye protein/phospholipid interface, and hence the lower the CMC the lower
the eye irritation. This is a simplistic view, but the experimental evidence clearly shows that
eye irritation is lower at lower CMC, at least for similar chemical species. The practical effect
is that anionic surfactants such as fatty alcohol ether sulfates that have excellent detergency
on hair but show eye irritation may have the latter reduced by the addition of a small quantity
of the tallow (Ci6_ig) alkyl polyaminocarboxylate.
C. Conditioning Shampoos
The separate application of cationic surfactants or cationic polymers is undoubtedly the most
efficient way of imparting softness and conditioning to hair. Cationics are normally incompati
ble with anionics in solution and will precipitate out as a anionic-cationic complex. A single
602 Porter
shampoo-conditioner mixture made with anionic or cationic surfactants is difficult to formu

late. Careful choice of the cationic is possible, but a more practical solution is the discovery
that some anionic surfactants are compatible with the cationic polymers used as softeners.
The ethoxy carboxylates, which can be used to partly (2 0 -4 0 % ) replace the fatty alcohol
sulfates or fatty alcohol ethoxy sulfates in these formulations, will give some compatibility
with cationic surfactants.
D. Bath Additives
The principal bath additive is for foam (bubble) baths. Foam bath formulations look basically
very similar to shampoo formulations—there is the same need for rheology control, low skin
and eye irritation, and copious foam—but there are some important differences. Bubble baths
are often used with bath oils that are insoluble in water, and therefore emulsifying properties
are required. Shampoos are applied directly from the bottle onto the hair, whereas the foam
bath is highly diluted. This feature is important to the requirements to avoid skin and eye
irritation, and therefore formulations suitable for foam baths may be quite unsuitable for mild
shampoos. Toilet soap is usually used in a bath, and the foam in a bubble bath must be stable
to the addition of soap. It should be stable in warm water and to both hard and soft water.
However, the foam must not be too stable, as it must vanish on draining the bath and/or
collapse with the addition of cold water. The typical surfactants used are the fatty alcohol
ethoxy sulfates, which may be derived from either lipids or petrochemicals.
E. Shower Gels
Normal shampoos can be and often are used for washing the body, but they generally generate
too much foam to be efficient. The shower gel cleanser formulations are based on the shampoo
formulations but need greater control of the rheology and also less foam. The major need is
to prevent an overly rapid loss of product under the shower spray. Consequently the products
are usually in the form of high viscosity liquids or gels. The active concentration of surfactant
is generally higher than in shampoos.
VII. INDUSTRIAL AND INSTITUTIONAL DETERGENTS

A. Definition
Industrial and institutional detergents are very often considered as one group in statistics on
surfactant consumption, and there is confusion on definition of these end uses. Industrial
detergents are used by industry as part of the process of manufacturing some other product,
e.g., scouring raw cotton with detergent in the spinning of cotton. Such detergents have been
formulated for a specific purpose, and there can be many specialized formulations depending
on the particular industry. On the other hand, institutional detergents used by hotels, catering
establishments, large offices, etc. are normally very similar in formulation to domestic deter
gents. However, very large catering establishments will have large automated equipment,
e.g., dishwashing equipment that automatically dispenses the detergent, and they will need a
specially formulated detergent that is not used in the home. Hospitals may insist on a higher
standard of hygiene and use germicidal detergents. It is difficult to distinguish between the
two market areas of institutional and industrial cleaning. Many small hotels, small restaurants,
pubs, etc. use ordinary domestic detergents bought in supermarkets.
B. Disinfectants and Biocidal Cleaners

Many detergent products are disinfectant and/or biocidal cleaners. Biocides are products that
either kill bacteria or control the level of bacteria and prevent them multiplying. Disinfectants
usually have biocidal action but do not give any detergent action, i.e., they do not clean.
They are usually used on relatively clean surfaces that have previously been cleaned with a
detergent. Biocidal detergents are products that clean and have some biocidal action. Nearly
all detergents will remove bacteria from surfaces, but very few detergents have a biocidal
action as well. It can be very difficult to distinguish between detergent use and disinfectant
use in the technical sense.
The technology ranges from the very simple to extremely complex products. In use, the
concentration of the active ingredients can vary from very low (just hot, very dilute alkali
solution with zero surfactant is used in some bottle-washing applications) to up to 90% in the
case of dry cleaning detergents. Some hospital applications (e.g., surgical scrubs) have great
demands on performance and cost is secondary. This is not general, as most users are very
concerned about the cost of the cleaning operation and will assess performance more objec
tively than the housewife. This results in products being sold on a price/performance basis
rather than by advertising.
C. Major Industrial and Institutional Areas for Detergents

1. Textile Cleaning
Natural fibers such as wool and cotton contain dirt that must be removed before processing.
During processing both natural and synthetic fibers are treated with processing aids (e.g.,
spinning oils and lubricants) that are generally removed prior to dyeing. The process of dyeing
often requires removal of excess and loose dyestuffs. Therefore many specialty detergents and
surfactants are used. Foam is normally a nuisance, so formulations are seldom based on
anionic surfactants and more often based on synthetic nonionics from petrochemicals.
2. Com m ercial Laundry

Depending on the size of the laundry, three different types of machine systems are used. The
first type is simply a standard washing machine and dryer similar to the domestic versions
that have been modified to take larger loads. This type of machine would use the same deter
gents as domestic machines. The second type is the conventional “batch” machine, and the
third type is the counterflow or “continuous” machine, which has become more common in
the last 20 years. The latter is much more efficient but also capital-intensive.
In hard water areas most commercial laundries have water softening equipment, and there
fore soap-based products can give excellent results. The relative use of soap and synthetic
surfactants does not depend only on the relative price. Soap is more difficult to rinse off
clothes than linear alkylbenzene sulfonates and the fatty alcohol and fatty alcohol ethoxy
sulfates. In addition, soap is an excellent detergent for cotton but not so good for synthetic
textiles such as polyester. The synthetic fabrics, particularly polyester-cotton blends, are be
coming a much larger fraction of commercial laundry loads, so the use of soap would be
expected to decrease. Lower washing temperatures, which save energy, will also lead to a
decline in the use of soap. Large commercial laundries can formulate their own products from
surfactants, builders, bleach, etc. The use of soap versus synthetic detergent varies widely
from country to country.
604 Porter
3. D ry Cleaning
Dry cleaning solvents contain a small quantity of water and an anionic surfactant to aid in the
removal of water-soluble stains. The surfactant is usually a petrochemical-based anionic
product.
4. H ard Surface Cleaners

A very common application of detergents in industrial and institutional settings is the clean
ing of hard surfaces. Special detergents are needed when the soil is of a special nature, e.g.,
rust on metal, burned carbon on cooking utensils, dried blood in abattoirs, or where the soil
is difficult to get at as in beer pumps and dairy equipment. With mechanical application,
spray equipment, foaming must be avoided and anionics are not used; nonionics and ampho-
terics are more appropriate. For manual application as in offices and hospitals, formula
tions very similar to household products are normally used. Cream cleaners (containing anion
ics) are used undiluted for application on vitreous enamel, stainless steel, and ceramic sur
faces.
There are many special hard surface cleaners, particularly for dairies, food processing
plants, and breweries. In these applications a disinfectant is used after cleaning with the
detergent, the main biocide being hypochlorite. Other disinfectants include iodophores, qua
ternaries (e.g., benzalkonium chloride), and chlorphenolics. Formulated biocidal detergents
are blends of nonionics with quaternaries, hypochlorite, or chlorinated isocyanurates.
The choice of one type of product over another depends on a number of factors, the main
ones being cost, government approval, and freedom from taste or taint. Anionics are rarely
used for hard surface cleaning because of high foam.
If the soiling is of such a nature that foaming is not a problem, then standard anionics can
be used. A typical powder detergent formulation is given in Table 8, and a typical liquid
detergent formulation in Table 9.
Table 8 Powder Detergent for Hard Surface Cleaning
Trisodium phosphate 70
Soda ash 25
Sodium dodecylbenzene sulfonate flake 5
Table 9 Liquid Detergent for Hard Surface Cleaning
Sodium metasilicate 10
Sodium dodecylbenzene sulfonate 3
Phosphate ester 4
Water To 100
5. M etal Cleaners
Metal cleaning in a manufacturing plant involves degreasing and derusting. This is usually
done with the use of high pressure sprays, which precludes anionic surfactants except for
special phosphate esters (based on fatty alcohols), which have excellent electrolyte compatibil
ity and relatively low foam.
6. M achine D ishw ashing

Automatic dishwashing detergents depend on the detergency and alkalinity of the detergent to
remove proteinaceous soils. The hot alkali can cause saponification, which gives a foam that
is stabilized by the protein present. It is essential to prevent the formation of foam, and
therefore an antifoam agent (not a defoamer) must be present. In addition, the antifoam agent
must be stable to alkali. EO-PO copolymers and alcohol EO-PO condensates are good anti
foam agents, but if there is a high concentration of alkali, then the primary hydroxyl group
must be blocked because of its instability to alkali.
7. F ood Processing Cleaners

Of the many types of food processing equipment that require cleaning the simplest is a tank
containing residues that can be filled with a detergent solution, which is heated and stirred,
and then drained off. In this situation an anionic alkaline solution gives excellent service, and
a typical formulation for cleaning beer tanks is shown in Table 10. However, in the majority
of cases, foaming must be controlled, so anionics are seldom used.
8. Transport Cleaners
There are many detergent products used in transport, in car manufacture (e.g., preunderseal
cleaning applied in conjunction with high pressure steam), in servicing (e.g., engine degreas
ing, which uses anionics), car shampoos, and traffic dirt cleaners for buses, trucks, and trains.
Wherever possible the anionics (normal linear alkylbenzene sulfonates) are used, but where
spray is used (e.g., for cleaning motortrucks), then amphoterics tend to be used.
De H and C le a n e rs an d Gels
Hand cleaners and gels for industrial use must be capable of removing grease and oil from
hands. Liquid soaps for washing hands are based on fatty alcohol sulfates or ethoxy sulfates
rather than soap. A typical formulation is shown in Table 11. The fatty alcohol can be based
on lipids or petrochemicals. Usually cost is the major consideration.
Industrial hand cleaners are also sold as a clear gel with a marketing gimmick of a “ring
ing” gel. The principal components are surfactants, a solvent (usually deodorized kerosene).
Table 10 Beer Tank Cleaner Formulation
Sodium dodecylbenzene sulfonate 7

Sodium xylene sulfonate 7
Sodium metasilicate 10
Water To 100
606 Porter
Table 11 Liquid Soap

Monoethanolamine lauryl sulfate (33%) 50
Cocomonoethanolamide 3
Sodium chloride 0.5
Formaldehyde 0.2
Perfume and color q.s.
Water To 100
and an alcohol or glycol. The product is often in the form of a microemulsion. The formula
tion of a typical soap-based product is given in Table 12. Soap is used in this formulation,
but synthetic surfactants are also used instead.
Anionics make very good detergents but with a few exceptions give high foaming products.
Foam is usually a nuisance in industry, and therefore anionics tend to be used in a minor way
where detergency is required.
There is a very wide and diverse use of anionics in industry in nondetergent uses involving
wetting, dispersing, emulsifying, and foaming properties. The main industries are agricultural
chemicals, cement manufacture, corrosion inhibitors, food, leather, paint, emulsion polymer
ization, paper, textiles, and polishes. Petrochemical-based anionics are mainly used. Lipid-
based alcohols are perfectly satisfactory raw material bases for fatty alcohol sulfates and
ethoxy sulfates, and the only determining factor is the price and availability relative to the
petrochemical-based materials.
VIII. FUTURE TRENDS AND ENVIRONMENTAL PROBLEMS

In the oil crisis of 1974, the cost of crude oil escalated suddenly and increased the cost of
petrochemicals relative to those of to natural fats and oils. The synthetic surfactant manufac
turers suddenly found the price of their raw materials dramatically higher, and they also
realized that these prices were controlled by countries who may not always be friendly to the
developing countries. At the same time it was widely believed that crude oil supplies would
peter out in the not-so-far future. These two factors led to increased research and the commer
cial development of new surfactants using raw materials not based on crude oil and with
properties superior to those of soap. Some of those products have now reached large-scale
use and production; an example is the sulfonated methyl esters of fatty acids (see Table 2).
Table 12 “Ringing” Hand Gel (A hand gel

which “rings” when tapped or shook.)
Deodorized kerosene 20
Oleic acid 10
Laurie diethanolamide 4
Hexylene glycol 5
Triethanolamine 3.5
Potassium hydroxide 0.2
Water To 100
Although the price of crude oil has stabilized (at the time of writing), there will undoubtedly
be severe fluctuations sometime in the future, which will ensure that surfactant manufacturers
will need alternative raw materials.
The early synthetic anionic surfactants based on branched long-chain paraffinic hy
drophobes were nonbiodegradable in natural streams and in sewage works and produced visi
ble stable foam in bodies of water into which they were discharged. Although this problem
was soon overcome by changing to linear paraffinic chains that biodegrade easily, the bio
degradation problem will not go away, as new and more stringent methods of assessing
biodegradability are being examined [11]. The present linear alkylbenzene sulfonates will
biodegrade satisfactorily under aerobic conditions [ 12 ] but not under anaerobic conditions.
This is not necessarily a practical problem in terms of environmental acceptability [13].
Early detergents based on petrochemical-based surfactants were yellow or brown in color,
had a bad odor, and were harsh to the skin. No commercial surfactant is a pure chemical
compound but a mixture, and many synthetic surfactants contain by-products such as tetralins
and sulfones in linear alkylbenzene sulfonates [14]. Some of these products, e.g., 1,4-dioxane
in fatty alcohol ether sulfates, although present in very small amounts, are known to be toxic
in the pure form. In this latter case the 1,4-dioxane occurs as a result of the sulfation process
on the polyethylene oxide chain [15] and will be present whether the starting material is lipid-
based or petrochemical-based. However, the surfactants now used in household products are
of extremely good color, have no odor, and have been thoroughly tested for impurities, and
the toxicity of any such impurities to humans or the environment is negligible.
Price, toxicity of impurities, and biodegradability can all be quantitatively measured, but
there are other factors operating to replace petrochemical-based surfactants with lipid-based ma
terials that are not easily measured. At present there is a definite market requirement for more
“natural” products. This has often been interpreted as a demand for non-petroleum-based prod
ucts, because even soap uses synthetic chemicals in its manufacture. There is a also a demand
for “mild” detergents. Although soap can adversely affect the skin and eyes, the average con
sumer has grown up with soap and is well aware of the problems. Some of the early synthetic
products caused more severe skin irritation than soap and, in a few cases, severe skin sensitiza
tion to a small percentage of the population. However, considerable research has now developed
surfactants, particularly modified soaps and amphoterics, that irritate the skin and eyes signifi
cantly less than soap. Such products are usually manufactured from “natural” raw material, i.e .,
fats and oils, so they can be marketed not only as “mild” but also as “green” and “natural.” Mild
surfactants can be defined as those producing low toxicity, zero irritation to the eyes, zero skin
irritation, and zero skin sensitization, with fast biodegradation, no toxic impurities, and no prob
lems with the environment. Such a product probably does not exist, but some new surfactants
can get nearer to meeting these requirements than soap can.
The overall results of these factors will probably be a slow decline in the consumption of
anionic synthetic detergents based on petrochemicals and their replacement by semisynthetic
products based on natural products rather than soap. For instance, the major new surfactant
type in the last decade is the alkyl poly glycosides, which use the polyhydroxy groups in
carbohydrates as the hydrophilic entity, with a long-chain fatty alcohol (from fats, oils, or
petrochemicals) as the hydrophobic entity. Lipids will probably be the major source for the
hydrophobic group of such surfactants.
REFERENCES
D. Postlethwaite, A life-cycle inventory for the production of soap in Europe, Tenside Surfact.
D e te r g . 3 2 : 157-169 (1995).
608 Porter
2. D. Sudati, Overview of processing steps in the production of intermediates for detergents from
natural raw materials, Chimaoggi July-August: 53 (1989).
3. M. Biermann, F. Lange, R. Piorr, U. Ploog, H. Rutzen. J. Schindler, and R. Schmid, Synthesis
of surfactants, in Surfactants in Consumer Products, Theory, Technology and Application (J.
Falbe, Ed.), Springer-Verlag, Berlin, 1987, pp. 3 3 -3 5 .
4. M. R. Porter, Handbook of Surfactants, 2nd ed.. Chapman and Hall, London, 1994.
5. R. H. Ferguson et al., Ind. Eng. Chem. 35: 1005-1012 (1943).
6. E. Stroink, Ethercarboxylates for industrial and institutional applications, in Industrial Applications
of Surfactants, 2nd ed. (D. R. Karsa, Ed.), Royal Soc. Chemistry, London, 1990, pp. 6 2 -7 5 .
7. D. Balzer, Alkylpolyglycosides, their physico-chemical properties and their uses, Tenside Surfact.
Deterg. 28: 419-427 (1991).
S. Tagata, A. Ishikawa, and H. Hayashi, Effect of washing temperature and added nonionic sur
factants on the detergency o f alkyl sulfates. Común. J. Com. Esp. Deterg. 25: 251, 253-266
(1994).
9. E. Lomax, The mechanism of surfactant detoxification. Spec. Chem., November 1992, pp. 3 7 -3 9 .
10. U. Hees, R. Jeschke and M. Weuthen, Alkyl oligoglycoside-containing liquid cleaners for hard
surfaces, Ger. Patent DE 4, 311,159 (1994).
IL H. A. Painter, Testing strategy and legal requirements in Biodegradability of Surfactants (D. R.
Karsa and M. R. Porter, Eds.), Blackie, London, 1995, p. 118.
12. H. A. Painter, The question of the anaerobic biodegradability of linear alkylbenzene sulfonates,
Proc. 3rd CESIO Int. Surfactants Congress and Exhibition, Sect. E, (1 -5 June 1992), pp. 3 4 -4 1 .
13. R. R. Birch, W. E. Gledhill, R. J. Larson, and A. M. Nielson, Role of anaerobic biodegradability
in the environmental acceptability of detergent materials, Proc. 3rd CESIO Int. Surfactants Con
gress and Exhibition, Sect. E, (1 -5 June 1992), pp. 2 6 -3 3 .
14. J. P. Hughes, The determination o f trace components in surfactants. Crit. Rep. Appl. Chem. 32:
90, 99 (1991).
15. K. L. Matheson, The effect of ethylene oxide adduct distribution on the formation o f 1,4-dioxane
during sulfation of alcohol ethoxylates, Proc. 3rd CESIO Int. Surfactants Congress and Exhibition,
Sect. C, (1-5 June 1992), pp. 211-221.
24______________________
Cationic Surfactants
Alan D. James
Akzo Nobel Chemicals Ltd., Littleborough, Lancashire,
England
I INTRODUCTION AND MARKETS

In this review of the market, manufacturing methods, properties, and applications of cationic
surfactants, only commercially important products are considered in detail. Amphoteric sur
factants are excluded from the discussion.
All commercially available cationic surfactants are nitrogen derivatives, and in most cases
the alkyl portion of the surfactant is derived from natural fats and oils such as tallow or
rapeseed, palm, fish, tall, or coconut oils. Cationic surfactants can be classified into four
product groups, the manufacture and chemistry of which is described below: fatty amines,
amidoamines and imidazolines, esteramines, and etheramines. Whereas in fatty amines the
hydrophobic alkyl chain is directly attached to the cationic nitrogen, in the other product
groups it is separated by an amido, imidazoline, ester, or ether group, respectively (Fig. 1).
The members of each product group can be derivatized to give a wide range of cationic
and amphoteric surfactants, which can be subclassified according to their head group chemis
try into, for example, quaternary ammonium salts (“quats”); /V-alkylpropylene diamines (“di
amines”); A,A-dim ethylalkylam ines (tertiary amines), ethoxylated amines, and amine oxides.
Some of these classifications may not be mutually exclusive in a strict chemical sense.
Worldwide production of cationic surfactants can be estimated at 350,000-500,000 tonnes
per annum (tpa). European production of fatty amines has been estimated at 120,000-150,000
tpa [1-3], amidoamines at ~ 30,000 [2], and esteramines at 61,000 tpa [4]. Estimates of
North American production of fatty amines reach 200,000-250,000 tpa; amidoamines, 50,000
tpa [2]. Worldwide production of amine ethoxylates reaches 50,000 tpa [5], amine oxides
30,000 tpa [2], and diamines 18,000 tpa [2]. But a significant portion of the fatty amines and
amidoamines produced are used as intermediates in the manufacture of amphoteric surfactants,
oil additives, and other industrial products in which the cationic character of the amine is lost.
Table 1 lists major manufacturers of cationic surfactants. Akzo Nobel is the largest producer
[2,6] and Witco the second largest.
609
610 James
R-NK R-CONHCH 2CH2 CH N(CH3^2

)
Fatty amine Amidoamine
R -O C H ^ C H p H ^ N H ^
»N- C K Etheramine
R-C R -C 0 0 C H 2C H 2N (C H 3)3
'N 'C H 2 Esteramine
C H 2C H 2N H 2
Imidazoline . ( C H ^ C H ^ O )^
R-N
' (C H p H p )^ H
R -N H C H 2C H 2CH 2 N H 2
Amine ethoxylate
Diamine
R -N (C H 3)2 R-N(CH3)3 Cl R -N (C H 3)20

Tertiary amine Quaternary amine Amine Oxide
Fig, 1 Examples of cationic surfactants.
The most important market for cationic surfactants is the detergent and personal care sec
tor, and the single most important application is as softening agents (fabric conditioners) [ 1 ]
with a worldwide market of 200,000 tpa [4], 66,000 t in Europe. Industrial applications such
as mineral processing, bituminous road construction, cleaning agents and biocides, agrochemi
cals, organoclay manufacture, and textile treatment each account for 5 -1 0 % of total cationic
surfactant consumption.
Table 1 Some International Producers of Cationic Surfactants

Main manufacturing Chemical types
Company^ locations^ produced^
Akzo Nobel Sweden, U.K., U .S.A ., Belgium, FA, AA, EA, EST
Canada, Brazil, Japan
Witco U .S.A ., Germany, Spain FA, AA, EA, EST
Kao Japan, Spain, Mexico, Malaysia FA, AA, EST
Hoechst Germany, U.K., Brazil FA, AA, EA, EST
Elf France FA, AA, EST
Fina Belgium FA, AA, EST
Henkel Germany, Spain FA, EST
Stepan U .S.A ., France EST, AA
Ûltimate owner.
Încluding joint ventures.
^FA = fatty Amines; AA = amidoamines and imidazolines; EA = etheramines; EST = esteramines.
Cationic Surfactants 611
II. MANUFACTURE
A. Fatty Amines
1. The N itrile Process
In the most important route to cationic surfactants, fatty acids or their methyl esters are reacted
with ammonia to form nitriles, which are hydrogenated to primary and secondary amines [7 ].
Conversion of fatty acid to amide and then nitrile is carried out in a single batch or continuous
process using alumina or zinc oxide as catalyst. Amide is not isolated.
RCOOH + NH 3 (RCONH2) -> RCN + 2H 2O ( 1)
In the batch process, nitrile is produced at 280-360°C in the liquid phase; then it is usually
distilled. In the continuous process, nitrile is separated in vapor form from liquid fatty acid
and amide, and further distillation is not necessary. Production of nitriles directly from fats
and oils has also been investigated on a pilot scale [8].
Fatty nitriles are not cationic surfactants and have few industrial applications. They are
hydrogenated to fatty amines in batch or continuous liquid-phase processes at 100-150°C
using nickel or cobalt catalysts.
RCN + 2H 2 RCH 2NH 2 (2)
Depending on the reaction conditions, dialkyl- or even trialkylamines may be formed.
2RCN + 4H 2 -> (RCH2)2NH + NH 3 (3)
Formation of the higher alkylated products is suppressed by the addition of ammonia. Primary
amines produced by the nitrile process contain some residual amide and dialkyl secondary
amine (3-5%). Dialkylamines produced by this process typically contain monoalkylamine
(~ 12%) and trialkylamines (~ 3%) [9].
The primary and secondary amines can be further derivatized (see below).
2. P roduction o f F atty Am ines via F atty Alcohols

Fatty alcohols produced by the hydrogenation of fatty acid methyl esters or fatty acids [10],
RCOOCH 3 + 2H 2 RCH 2OH + CH 3OH (4)
or from petroleum feedstocks can be used to manufacture fatty amines.

Alcohols can react directly with an excess of ammonia and hydrogen to form dialkylamines
[ 1 1 ], but in today’s most important process, dimethyl tertiary amines are prepared from alco
hol and dimethylamine.
R 0H + HN(CH3)2 R—N(CH 3)2 + H20 (5)
The reaction can be carried out under hydrogenation conditions. The dimethyl tertiary amines
produced are important intermediates for quaternary amines, betaines, and amine oxides. A
similar reaction can be used with high molecular weight dialkylamines to form trialkylamines
in good yield.
612 James
In commercially less important processes, the fatty alcohol can first be converted to alkyl
sulfate or alkyl halide, and this intermediate can be reacted with the secondary amine. Alkyl
halides can also be used to prepare surfactants by quatemization:
R Cl + Py* R Py^ C f
(6 )
B. Amidoamines and Imidazolines

The second major route from fats and oils to cationic surfactants is reaction with polyamines
[12,13]. In a typical reaction, fatty acid is condensed with a polyamine such as diethylenetria-
mine (DETA) at a temperature of 150-170°C without catalyst in a batch process.
RCOOH + NH 2CH 2CH 2NHCH 2CH2NH 2 RCONHCH2CH2NHCH 2CH 2NH 2 + H 2O (7)
At higher temperatures (190-250°C) and under vacuum or in the presence of solvents such as
xylene or higher alkylbenzenes, which help water removal by azeotropic distillation, further
condensation to imidazoline occurs.
N—CH,
RCONHCH,CH,NHCH,CH,NH, RC
N—CH,
CH2CH2NH2 (8)
The fat itself or fatty acid methyl ester can be used instead of fatty acid, in which case the
leaving group is glycerol or methanol. A wide range of polyamines is used commercially; the
most important of these are diethylenetriamine, aminoethylethanolamine, aminoethylpipera-
zine, triethylenetetramine, tetraethylenepentamine, dimethylaminopropylamine, and ethylene-
diamine. Fatty acid pitches or acid oils reacted with polyamine residues such as those obtained
as by-products in the manufacture of hexamethylenediamine or polyethylenepolyamines pro
vide cationic surfactants suitable for many industrial applications.
When the polyamine contains more than one primary or secondary amine group, polyalky-
lated products will result. Depending on the ratio of fatty acid to poly amine, there may also
be significant amounts of unreacted polyamine in the product. The relative amounts of ami-
doamine components and imidazolines can affect the physical properties and performance of
the products and their derivatives.
C. Esteramines
Fatty acids or esters react with alkanolamines to form esteramines.
2R CO O H + (H 0CH 2CH 2)2N CH 3 (R C O O C H 2 C H 2 )2 N C H 3 + 2 H 2 O (9)
When the alkanolamines also contain primary or secondary amine groups, amido products are
generally favored, with esteramines present only as by-products [14].
R C O O C H 3 + (H 0 C H 2 C H 2 )2 N H R C 0 N (C H 2 C H 2 0 H ) 2 + C H 3 OH (10)
Fatty diethanolamides are nonionic surfactants and are not discussed further in this chapter.
Reaction of fatty acids or esters at 15 0 -2 5 0 °C with triethanolamines gives a mixture of
mono-, di-, and trialkylesteramines, depending on the ratio of acid and amine.
*where Py = pyridine.
Alkanolam ines important com m ercially for the manufacture of esteramines include trietha
nolamine, methyldiethanolamine, diethylethanolamine, aminoethylethanolamine, and dimeth-
yldihydroxypropylamine [15].
2 R C 0 0 H + H 0C H 2CH (0 H )C H 2N (C H 3)2
RC00CH2CH(00CR)CH2N(CH3)2 + 2H 2O ( 11)
N -CH ,
2RCOOH -f NH 2 CH 2 CH 2 NHCH 2 CH 2 OH - R -C
\ N -CH ,
CH2CH2OOCR (12)
In principle, fatty alcohols can also be used to prepare esteramines by reaction with
amino acids.
2R 0 H + (H 0 0 C CH 2)2N C H 3 (R 00 C C H 2)2N C H 3 + 2H 20 (13)
D. Etheramines
Another class of alcohol-based cationic surfactants are the so-called etheramines produced by
reaction with acrylonitrile and hydrogenation:
R0H + C H = C H 2C N - ROCH 2CH 2CN (14)
ROCH2CH2CN + 2H 2 ' R O CH oCH oCH .N H , (15)
Hydrogenation proceeds like reaction with fatty nitriles [Eqs. (2) and (3)], and ammonia is
used to suppress the formation of dialkyletheramines. A lower temperature may be used to
reduce decyanoethylation, but nevertheless commercial etheramines may contain 5 -1 5 % free
alcohol. Most etheramines are produced from synthetic alcohols derived from petroleum feed
stocks.
E. Derivatization
1. Neutralization
Partial or full neutralization of amines with inorganic or low molecular weight organic acids
gives water-soluble or dispersible products, and the amine acetates in particular are commer
cially supplied products. Neutralization with long-chain carboxylic acids or petroleum sulfo
nates gives oil-soluble products.
2. Alkoxylation
The uncatalyzed reaction of secondary or primary amines with ethylene oxide at ~ 10 0 °C
gives the A-ethanol derivatives.
O ( 16)
/ \
RNH 2 + 2 CH 2CH 2 RN(CH 2CH 20H )2
614 James
O
/ \
RNH2 -f 3CH2CH2 + CH3COOH - RN^CH2CH20H)3 CH oCOO' (17)
Under acid conditions, quaternary amine salts are obtained [16].

Alkaline catalysis gives polyalkoxylates. To ensure complete conversion to tertiary amine,
polyalkoxylates are usually prepared via the A-ethanol derivatives:
O
/ \ ^ (C H 2CH 20),H
RN(CH2CH20H)2 -f (x+y-2)CH2CH2 - RN
M C H 2CH 20)yH (18)
A typical alkaline catalyst is potassium hydroxide at a reaction temperature of 120-170°C.

Polyalkoxylates contain a range of components that differ in their degree of alkoxylation (Fig.
2) and often include some poly alky lenegly col.
3. Alkylation
Primary and secondary amine groups can be methylated with formaldehyde under reducing
conditions.
R 2NH + HCH 0 ->R2NCH 3 + H 2O (19)

The reaction is usually carried out at ~ 180°C with hydrogen over a nickel catalyst. Amines
may also be alkylated with higher aldehydes under similar conditions.
RNH 2 + C7H i5CHO-^RNHC 8H i7 + H 2O (20 )

or with fatty alcohols [Eq. (5)].
2 3 4 5 6 7 8 9
moles ethylene oxide
Fig. 2. Distribution of ethoxylated cocoamines in nominal 5 mol ethoxylate. (From Ref. 17.)
4. Q uaternization
Alkylation with ethyl, benzyl, or methyl chloride, diethyl or dimethyl sulfate, or chloracetic
acid leads to quaternization.
RN(CH3)2 + CH3C1 RN(CH3)3C1 (21)
The reactions are normally carried out at 5 0 -1 2 0 ° C in water or an alcohol solvent (typically
isopropanol, ethanol, or a glycol). Alkylation of primary or secondary amines leads to the
formation of an acid.
R 2N H + 2C H 3CI ^ R 2N (C H 3) 2C 1 + H 0 ( 22)
which must be neutralized if the reaction is to go to completion, and the resulting salts must
be removed. For this reason, most quaternary amines are prepared from tertiary amines.
Quaternization of esteramines and some amidoamines and imidazolines presents particular
problems because these amines may hydrolyze or react with alcoholic solvents during the
quaternization process [13]. Quaternization at high activity under anhydrous conditions with
careful pH control with very active quatemizing agents (e.g., dimethyl sulfate) may be fa
vored.
Quaternization of propoxylated amines and higher ethoxylated amines may also necessitate
the use of dimethyl sulfate. Quaternization with ethylene oxide [Eq. (17)] or alkyl halide [Eq.
(6)] are described above. Commercially available quaternary amines often contain 1 - 2 % of
unquatemized material, inorganic salts from the neutralization of acid by-products, and traces
of alcohols or ethers formed from the reaction of alkylating agent with the solvent.
5. C yanoethylation
Cyanoethylation is used to add amino groups to fatty amines and etheramines.
R N H 2 + C H 2= C H C N - RNHCH 2CH 2CN (23)
RNHCH2CH 2CN + 2H 2 ' > RNHCH2CH 2CH 2NH 2 (24)
The addition, Eq. (23), is acid- or base-catalyzed. One or two moles of acrylonitrile can be
added to a primary amine to give a diamine or a “Y” triamine product. The nitrile intermediate
can be reduced as in Eq. (24), with ammonia if necessary to prevent the formation of dialkyl
products. Commercially available diamines generally contain 2-10% monoamine due to decy-
anoethylation during reduction. The addition and reduction processes can be continued to
form higher polypropylenepolyamines, and these diamines and polyamines can be derivatized
by alkoxylation, neutralization, and quaternization.
6. Oxidation
Oxidation of tertiary amines with hydrogen peroxide gives amine oxides [18,19]:
R N (C H 3)2 + H 2O 2 R N (C H 3 )2 0 + H 2O (25)
The reaction is generally carried out at 50-90°C in aqueous or alcoholic solution to give the
product activity of 30-50%. Chelating agents like EDTA may be added to prevent decomposi
tion by metal ions [20] .
616 James
III. PROPERTIES
A. Chemical
1. Therm al and H ydrolytic Stability
In the absence of oxygen, primary, secondary, and tertiary fatty amines and etheramines are
generally stable to high temperatures (>180°C) even in aqueous solution, whereas quaternary
amines degrade above 100°C to give tertiary amines; in alkaline conditions olefins may be
formed [21]. Amine oxides can also degrade to olefins at elevated temperatures.
Amidoamines may disproportionate to dialkyl products on heating, and similar “reshuf
fling” of alkyl chains occurs in many amidoamine and esteramine formulations when heated.
r \ r \ AA
2RCONHCH2CH2N NH RCONHCH2CH2N NCOR -f NH2CH2CH2N NH (26)
v_y w uy
Esteramines, amidoamines, and their quaternary amines are liable to undergo hydrolysis,
R C00 CH 2C H 2N (C H 3)2 + H 2 O R C O O H + (CH3)2NCH2CH20H (27)
which is most rapid in alkaline solution. Formulations based on esteramines and ester quater
naries are generally adjusted to pH 2 -4 , where hydrolysis is slowest [22] (Fig. 3). Imidazoline
rings open in alkaline conditions to give secondary and tertiary amidoamines. The alkaline
nature of the imidazolines themselves means that they are liable to undergo hydrolysis during
storage if not kept water-free, but quatemized imidazolines are stable to hydrolysis.
2. A c id -B a se Behavior
With the exception of fully quatemized amines, cationic surfactants show acid-base behavior
in aqueous solution.
RNH, RNH 2 + H^ (28)
100
storage Time
[ weeks]
3 2
Acidity [ pH I
Fig. 3 Effect of Storage time and pH on the hydrolysis of esterquatemary. (From Ref. 22.)
Table 2 Values of Cationic Surfactants
Base Structure Ref.
Cg-Cjs alkylamines RNH2 10.3-- 1 0 . 6 17

C5- C 18 dialkylamines R2NH 1 1 .0 -- 1 1 . 2 23
C5-C 8 dimethylalkylamines RN(CH3)2 9 .3 - 1 0 . 0 17
A-Alkylmorpholines RN(CH;2CH2 ) 2 0 7.5--8.5 24
Dodecylamine oxide RNO 4,,8 25
^ Of acid form .
Aqueous solutions of unquatemized cationic surfactants contain both the neutral amine form
and the protonated acid form, their ratio depending on pH. The acid and base forms of the
cationic surfactants have quite different surfactant properties, and it is impossible to under
stand the properties of cationic surfactants in aqueous solution without considering their acid-
base behavior. Some typical values for the protonated amines are shown in Table 2. Most
amines are strong bases with values over 9; the exceptions are amine oxides and A-alkyl
morpholines with much lower values. The tertiary amine in substituted piperazines is also
only weakly basic.
It is not possible, however, to conclude from the values that in aqueous systems the
amine groups are most usually protonated, since at the surface of cationic micelles or other
cationic surfactant phases or emulsion droplets, the local pH may be up to two pH units
higher than the bulk pH, owing to the influence of the positive surface charge. For a similar
reason the counterion at octadecylammonium phosphate monolayers was found to be HP 04^~,
not H 2PO4“ [26]. In the presence of anionic surfactants, on the other hand, amine oxides are
protonated well above their Zeta potential measurements on oil droplets stabilized with
cationic surfactants show that as the pH is increased, their surfaces become less charged [27-
30], as expected, as the head groups become deprotonated, whereas quaternary amines are
less affected by pH.
B. Physical and Surface Chemical Properties

1. B oiling and M elting Ranges
The boiling ranges of simple fatty amines lie in the range 180-300°C and can be related to
the boiling points of the corresponding alkanes [31]. The melting ranges of the neutral (base)
forms of cationic surfactants are related to the melting ranges of the originating fatty acid
sources (Table 3). Amidoamines have melting points above those of the corresponding fatty
amines [33], whereas etheramines and esteramines have lower melting points. Tertiary amines
have significantly lower melting points.
The anhydrous forms of quaternary amines, amine oxides, and the salts of amines with
inorganic acids have much higher melting points, in the range 100-200°C, but also often
show degradation at these temperatures. Dialkyl fatty amine and esteramine derivatives in
which some of the alkyl chains are branched show significantly lower melting points. Products
containing 2-ethylhexyl groups are commercially available [34], and esterquats based on mix
tures of tallow and 2-ethylhexanoic acid are claimed to provide liquid fabric conditioners at
high activity [35].
618 James
Table 3 Melting Points of Cationic Surfactants Derived from Fats and Oils
Fatty acid source of alkyl chain ^
Formula ^10 C 12 Coco C,4 C .6 Oleic Tallow HT ^18

RNH 2 15*^ 26 15 38 44 12 35 45 53
RN(CH3)2 -2 0 -1 0 -7 10 -10 3 20 23
RN(CH2CH20H)2 -1 0 -9 23 40 42
R 2 NH 42 47 b 43 61*^ 67*’ 62 72»
R2N(CH3) -5 15" -2 24= 35" 33 40"
RNHCH.CH^CH.NH^ 27 15 50 60
RC0NH(CH2)3N(CH3)2 29'' 5 57 63
RCOOH 30 43 33" 54 62 6 42 59 65
^Carbons in fatty acid precursor.

‘^From Ref. 7.
^From Ref. 32.
^From Ref. 12.
®Hardened grade.
Source: Refs. 7, 12, 32 as noted. All other data from the product datasheets of Akzo Nobel (midpoint of ranges
given).
2. P hase Behavior in Aqueous System s

The base forms of cationic surfactants, with the exception of amine oxides and highly ethoxy-
lated amines, are generally insoluble in water. Monoalkyl cationic surfactants formed by the
acidification or quatemization of the base forms show phase behavior similar to that of anionic
surfactants. There have been few studies of commercial surfactants that contain a mixture of
several different alkyl chain length homologues, but studies of two-component mixtures sug-
Table 4 Critical Micelle Concentrations of Cationic Surfactants

Derived from Dodecanoic Acid
CMC
Head group (mM) Ref.
15 36
— N(CH3)2H Cl 23 36
—N(CH3)3 Cl 13, 21 37, 38
—N(CH2CH20H)2CH3 Cl 4.6 37
— (CH3)i5(CH2CH2)20-C1 21 38
—N(CH3)2CH2Ph Cl 4.3 38
—N(CH2CH20H)2CH2Ph Cl 1.9 37
—I5H2CH2CH2CH2NH3 2C1 33 39
— n (CH3)P 0.76 41
—N(CH3)20H Cl 3.2 41
— C0NHCH2CH2CH2N(CH3)2H Cl 0.5 42
— 00CCH2N(CH3)3 Cl 5.5 40
gest relatively small influence. Molecular and micellar solutions dominate at concentrations
up to 30% in water. Some representative critical micelle concentration (CMC) values are
given in Table 4. Except in the case of the nonionic amine oxide, the CMC values of all the
monoalkyl cationic surfactants are similar. At higher concentrations a variety of liquid crystal
line phases are seen.
Monoalkyl cationic surfactants with certain counterions such as salicylate form highly vis
cous and viscoelastic phases that have been extensively studied [43,44]. The behavior of these
surfactants is apparently due to the formation of long (1000 nm) rodlike micelles. The prod
ucts have been suggested as drag-reducing agents and viscosifiers. The formation of rodlike
micelles is also implicated in the use of amine oxides or ethoxylated amines as viscosifiers
for acid cleaners or hypochlorite bleaches [45].
In contrast to the monoalkyl surfactants, the phase diagrams of dialkyl surfactants are
dominated by lamellar phases. With the exception of products with chains shorter than 10
carbons, the surfactants form dispersions of bilayers at even low concentrations in water. This
dispersion contains single- or multiwalled vesicles similar to those of phospholipids [46,47].
The vesicles can shrink or swell under the influence of osmotic forces and become leaky to
solutes at clearly defined temperatures. They have been studied because of their importance
in the physical properties of fabric softener formulations [48] and their potential applications
in drug delivery and magnetic fluids [49].
3. H ydrophilicity o f Cationic Surfactants

The relative importance of the hydrophobic and hydrophilic portions of a surfactant can
clearly influence its properties, and there have been various attempts either to measure or to
calculate an overall “HLB value” for a surfactant or to separate the contributions of the two
portions. Partition measurements fall into the second approach. The contribution of the indi
vidual functional groups in a molecule to its partition between two solvents can be separated,
and this is a tool used, for example, in drug design. By combining data for the head group
and the methylene groups, the partition behavior of untested surfactants can be predicted. The
water/heptane partition data for dodecyl derivatives calculated from measurements on shorter
chain length analogues are shown in Table 5. Low values of the partition coefficient corre
spond to more hydrophilic surfactants. So, for example, the diethanolamine is considerably
Table 5 Partition Data for Cationic Surfactants Between Water and Heptane
Log P increment for Calculated log P for

Head group head group dodecyl derivative
— NH2 - 2 .9 4.8
— N(CH3)2 - 1.3 6.4
— NH3 C 1 - 1.8 -1 1
— N(CH2CH20H)2 - 4 .7 3.0
— NHCH2CH2CH2NH2 - 5 .0 2.7
— N(CH2CH20)5H2 - 6.1 1.5
— N(CH2CH20)ioH2 -8 .5 - 0.8
— N(CH2CH2)20 - 2.0 5.7
(— CH2— ) (+0.64) —
(— CH2CH2 0 — ) (-0.47) —
— COOH - 3 .4 3.2
Source: Ref. 50 for C11H23COOH; all others from or calculated from data in Ref. 17.
620 James
Table 6 HLB Values of Cationic Surfactants (Davies Extended Scale 0-40)
Calculated HLB of
Head group HLB increment dodecyl derivative^
2 0 .0 21.3
8.5 9.8
—-NHCH2 CH2 CH2 NH 2 17.4 18.7
— NH 2 CH2 CH2 CH2 NH 3 38.6 39.9
-^(C H 3)3 2 2.0 23.3
—N(CH3)2 22.5 19.1'’
— N(CH2CH20H)2 11.3 12.6
-^5(CH3)p 2 1 .0 22.3"
^The HLB for the dodecyl derivative was calculated from
HLB = (HLB increment for the head group) - (1 2 x 0.475) + 7
‘’Didodecyl derivative.
"pH 3.
more hydrophilic than the dimethylamine. The technique is not sensitive for highly hydro
philic surfactants such as the charged amine salts and quaternaries.
An electrokinetic method has been used to estimate the HLB values of the more hydrophilic
surfactants, and some results are shown in Table 6 [51]. A high HLB value corresponds to a
more hydrophilic material. The results show that the cationic form of the surfactant is very
much more hydrophilic than the neutral base forms. Ethoxylation or additional amine groups
also make a significant contribution to the hydrophilic nature.
4. Adsorption o f C ationic Surfactants at Interfaces

The adsorption of cationic surfactants at interfaces has been reviewed [52] and is not consid
ered in depth here. Surface and interfacial tensions of aqueous solutions containing unquater-
nized cationic surfactants vary with pH. Mixed films of the neutral (base) form and the cat
ionic form give the lowest interfacial tensions at a particular concentration below the CMC.
Cationic surfactants are adsorbed on negatively charged surfaces, and this is related to their
use in mineral flotation, corrosion inhibition, and textile softening. Initial adsorption is by
neutralization of the negative sites. As the surfactant concentration increases, further adsorp
tion can occur, leading to reversal of the surface charge. Adsorption generally levels out at
the CMC, but further adsorption to give bi- or multilayer structures may occur in some sys
tems, especially those containing dialkyl surfactants. Acidity and basicity can have a large
effect on adsorption because in general the charge on both the surface and the surfactant will
vary with (local) pH. Differences between the adsorption isotherms of different cationic sur
factants in the low concentration region often disappear when plotted against reduced concen
tration (i.e., concentration/CMC) [17], suggesting that hydrophobic interactions are important
in the adsorption process and head group structure is relatively unimportant.
IV, APPLICATIONS
A. Fabric Conditioners
Fabric conditioners (softeners) account for ~ 40% of cationic surfactant use. Adsorption of a
cationic surfactant onto the surface of fabric improves handle, eases ironing, imparts antistatic
CH3
R C 0 N H C H 2C H 2C H 2\ I
/ NH(CHJ.CI
RCOOCH2CH2 //N'c h
C H 3S O ,
R= lallow CH
C H 2C H 2N H C O R
(R'C0NHCH2CH2)2N(CH3)R" .X R= tallow
R'=tallow, H'lallow
R"=GH3, CH2CH2OH
X= Cl. melhosulphate R 2N (C H 3)2.X
R= tallow, oleyl, H-tailow
(R'C00CH2CH2)2N(CH3)R" .X X= Cl, methosulphate
R'=tallow, H-tallow
R"=CH3, CH2CH2OH
Cl. melhosulphate RCOOCH.
C H C H 2N (C H 3)2CI
RCOO^
R= tallow
Fig. 4 Examples of fabric softener bases.
properties, and helps perfume retention [53]. Usually the softener’s active ingredient is added
as an aqueous dispersion to the laundry during rinsing at a level of 0 . 1 % fabric weight.
Alternatively, it may be impregnated into a fabric sheet that is added to the laundry during
tumble drying, or formulated into the detergent itself and added in the wash cycle (so-called
softergents).
Fabric softener bases are usually dialkylamines or quaternaries (Fig. 4). A typical rinse
cycle formulation contains 5-20% active ingredients plus nonionic components, perfume,
color, pH control additives, and electrolyte. The nonionic components may have a softening
function or help to solubilize perfumes and control viscosity; pH control additives may help
ensure the hydrolytic stability of the components; electrolytes such as sodium or calcium
chloride help control the viscosity of the formulation.
There has been a trend toward high activity (>15%) products to reduce packaging require
ments. The traditional softener base dihydrogenated tallow dimethylammonium chloride
(DHTDMAC) is difficult to formulate into highly active products and has been partly replaced
by amidoamines and imidazolines in these products [54].
Although the balance of evidence is that DHTDMAC does not seriously damage the envi
ronment, it has been largely replaced in Europe and to some extent in the United States and
Japan by esterquatemary amines and amidoester products, which biodegrade faster and are
less toxic to aquatic organisms [1,4,22,55]. Fabric softeners are today by far the major use for
esteramine derivatives. Whereas DHTDMAC accounted for 85% of softener base in Europe as
recently as 1990, it reached less than 40% in 1993 [56]. The resulting overcapacity in fatty
amine manufacturing led to a major restructuring by the prpducers [6].
A new range of esteramine softener actives based on polyols such as glycerol and amino
acids promises even lower aquatic toxicity than conventional products, especially when the
aquatic toxicity of the biodegradation intermediates is considered [57].
B. Organoclays
Certain clay minerals have a layered structure in which negatively charged metal oxide plate
lets are neutralized by alkaline earth or alkali metal cation in between the layers. These cations
622 James
can be replaced by cationic surfactants to form organoclays, which are used as thixotropic
additives for oil-based drilling fluids, coatings, cosmetics, greases, and inks. Cationic surfac
tants are used in this application at a rate of 13,000-20,000 tpa [58]. In a typical process, a
slurry of Wyoming bentonite or heetorite is centrifuged to remove impurities and optionally
ion-exchanged to convert it into the sodium form. It is then mixed with quaternary amine at
60-70°C in aqueous dispersion. The relatively hydrophobic organoclay product is filtered off,
dried, and milled. In an alternative process the reaction is carried out with almost dry clay.
The final products may contain 25-50% organic material representing 90-120% exchange of
the inorganic cations. Anionic surfactants and nonionic materials may be deliberately coad
sorbed to provide special products with improved dispersibility.
The most used products in this application are dihydrogenated tallow dimethylammonium
chloride, stearyldimethylbenzylammonium chloride, and dihydrogenated tallow methylben-
zylammonium chloride, although other quaternary amines or fatty amine acetates are used in
special products. Esteramine quaternary amines have been found useful, especially for or
ganoclays used in oil drilling fluids based on low toxicity vegetable oils [59].
C. Mineral Flotation
Cationic surfactants are used as collectors in the separation of minerals by froth flotation
[60,61], and mineral flotation accounts for perhaps 5-15% of total cationic surfactant usage
[ 1 ]. The technique involves passing air through a slurry of finely ground ores that has been
pretreated with typically 0.1-1 kg/t surfactant (collector). Selective adsorption of surfactant
on the target minerals makes their surface hydrophobic, leading to their adhesion to the air
bubbles and concentration in the froth. In direct flotation the valuable mineral is collected in
the froth; in reverse or indirect flotation the impurities are collected. Selective adsorption of
Table 7 Cationic Mineral Flotation Collectors
Valuable mineral Gangue mineral ^ Typical collector Ref.
Potassium chloride Sodium chloride Tallowamine 62, 63

Kainite Sodium chloride Cocoamine
Camallite Sodium chloride V-Cocomorpholine 63, 64
Calcite Silicates Tallowdiamine, dialkylquat- 65
emary
Magnesite Silicates DialkyIquatemary, tallow 66
diamine
Hematite/magnetite Silicates Alkyletheramine 67
Silicates Alkyletherdiamine 67
Phosphate rock Silicates Amidoamines 68
Quaternary amine, tallow
amine 69
Oxidized Zn minerals Various Cocoamine, tallowamine 70
Feldspar Quartz Tallowamine, tallow- 71
diaminedioleate
Mica Quartz Cocoamine, alkyletheramine 72
Kaolin Feldspar, quartz Alkyletheramine 73
Pyrochlore Various Tallowdiamine 74
Alkyl imidazoline 75
^Waste product.
surfactant is ensured by adjusting the conditions, especially the pH, until only the target
mineral surface is negatively charged. In a typical process a slurry of ground ore is first mixed
with acid or alkali to adjust pH, then mixed (“conditioned”) with a solution or dispersion of
cationic surfactant before air is introduced in flotation cells.
The choice of surfactant depends on the minerals to be separated. Cationic flotation is used
primarily for the upgrading of industrial minerals, but the reverse flotation of iron ore and the
flotation of oxidized zinc minerals are also significant. See Table 7.
Flotation performance is expressed in the grade (purity) of the mineral concentrate and its
yield or recovery. As the chain length of surfactant increases, there is usually an increase in
yield but a decrease in grade. Quaternary amines generally give higher grade products than
primary amines, but yields may be lower.
The use of ethoxylated amines or ether amines together with quaternary amines has been
claimed to improve yields [75,76], and combinations of cationic and anionic surfactants are
important in specific applications, especially feldspar flotation [77,78].
D. Use in Road Construction

Bitumen (known as asphalt in the United States) is a residue from the refining of crude oil
that can be used as a binder in road construction materials. Because of bitumen’s high viscos
ity, construction materials are usually prepared and used hot. Alternative cold techniques
involve the use of bitumen emulsions.
Cationic surfactants are used as adhesion agents in hot bitumen and as emulsifiers for
bitumen [79], and together these applications account for —10% of total cationic surfactant
consumption [ 1 ].
Adhesion agents (also called antistripping agents or wetting agents) are added to bitumen
at a level of 0.2-1% . Treated bitumen can displace water from the surface of wet aggregates
(active adhesion), and road materials prepared from treated bitumen are resistant to water
damage (“stripping”). Traditional adhesion agents were based on hydrogenated tallow diamine
supplied in pellet form for easy addition to the bitumen, but these have been largely replaced
by liquid products based on amidoamines and imidazolines. Bituminous binders are stored hot
(120-200°C), and at these temperatures adhesion agents degrade either by reaction with acid
components in the bitumen to form amides or through oxidation [80]. So-called heat-stable
products based on less reactive tertiary amines or polyamines have been developed for situa
tions where storage of treated binder cannot be avoided.
Bitumen emulsions typically contain 40-70% bitumen dispersed in water, as droplets 1 -
20 /xm in diameter, stabilized with 0.15-2% surfactant. Of the approximately 7 million tonnes
[81] of emulsion produced worldwide, probably more than 85% is stabilized by cationic sur
factants. Contact of the cationic emulsion with the negatively charged surface of aggregates
leads to its destabilization and coalescence of the bitumen droplets, a process known as break
ing or setting. Emulsions are classified according to their stability in contact with aggregate
into rapid-, medium-, and slow-setting grades, which contain different concentrations and
types of emulsifiers and are suitable for various road construction techniques (Table 8). The
actual mechanism of the setting process may involve the electrophoresis of bitumen droplets
to the negatively charged aggregate surface, depletion of cationic emulsifier from the droplet
surface onto the negative mineral surface leading to démulsification, and pH changes that lead
to neutralization of droplet charge [82,83].
Trends in the market are toward cold emulsion processes instead of hot processes (favoring
emulsifiers against adhesion agents) and toward techniques such as cold recycling, slurry
surfacing, and cold mix, which use slow-setting emulsions.
624 James
Table 8 Cationic Bitumen Emulsifiers
Type of
emulsion Typical emulsifier Use level (%) Application
Rapid-setting Tallo wdiamine 0 .2 Chip sealing

Medium-setting Tallo wdiamine 0.3-0.5 Open mixes
Tallo wquat 0.3-0.5 Tack coat
Slow-setting Tallo wdiquat 0 .4 -1 .0 Slurry surfacing
Amidoamines 0 .8 - 2 . 0 Slurry surfacing
Ethoxylated amines 0 .8 - 2 . 0 Dense mixes
Tallowpolypropylenepolyamines 0 .6 - 1 . 0 Dense mixes
E. Biocides
Some cationic surfactants are toxic to microorganisms such as bacteria, fungi, and algae and
as a result find use in disinfectants and preservatives. The multifunctionality of cationic surfac
tants allows biocidal properties to be combined with wetting and detergency or with corrosion
inhibition and substantivity. Applications include disinfectants and sanitizing cleaners for hos
pitals, the food industry, animal husbandry, and around the home; antiseptic skin care prepara
tions; control of algae and bacteria in water systems; wood preservation; and the control of
bacterial corrosion in oil production and storage.
Quaternary ammonium compounds are the most used products (Fig. 5). The optimum alkyl
chain length depends on the application but ranges from 12-16 carbons for monoalkyl quater
nary compounds to 8-12 carbons for dialkyl products. Some data for the biostatic effect of
cationic surfactants are given in Table 9. The minimum inhibitory concentration is the concen
tration that prevents microbial growth; higher concentrations are required to kill the organ
isms. A typical disinfectant may contain 1-5% active biocide and be diluted up to 100-fold
before use [84]. The effectiveness of cationic disinfectants is decreased by the presence of
hardness salts or soils [85]. Disinfectant and sanitizing cleaners may also contain nonionic
R -N H C H 2C H 2C H 2 N H 2.2 C H 3C O O H
Fally diamine diacetate
R= coco-
R-N(CH3)3 X R-N(CH3)3(CH2C,H3) Cl
Quaternary amine Benzalkonium chloride
R= coco-, X= Cl R= COCO
R= hexadecyl-, X= Br RA tetradecyl-
R= oleyl-, X= Cl
R= decylbenzyl-. X= Cl
R 3N (C H 3)2.CI
R -N (C H 3)20
Dialkyl quat
Amine Oxide R=coco-
R= coco- R=decyl-
Fig. 5 Examples of cationic surfactant biocides.

C ationic Surfactants 625
Table 9 Biostatic Performance of Cationic Surfactants— Minimum Inhibitory Concentrations (ppm)
Cationic Bacteria
surfactants—
common name Gram neg. Gram pos. SRB Yeasts Fungi Algae
Cocoamine acetate 50-3200 10-25 400+ 6-25 100-400 ____
Didecyl quat 5-200 3 -6 100-200 3 3-12 —
Dicoco quat 400-800 3-25 3200 400-800 6-12 —
Benzalkonium chloride 100-200 6-25 50-200 12-50 12-200 <1

Cocodiamine diacetate 25-800 12-25 50-100 6-25 50-200 —
Cocoquat 30-150 1-15 50-100 15 5 —
Cocoamine oxide 3200 50 50 50 —
SRB = Sulphate-reducing bacteria.

Source: Averaged data taken from Refs. 45, 83, 84, 86.
surfactants or amphoteric surfactants and chelating agents to improve detergency and hard
water tolerance. In oil production and storage, quaternary amines or diamines can be used to
control sulfate-reducing bacteria [86] that produce hazardous hydrogen sulfide gas and cause
corrosion. The diamines in particular also provide direct corrosion inhibition.
In the field of wood preservation, cationic surfactants can replace the environmentally
hazardous chemicals pentachlorophenol and copper chrome arsenate in some applications.
Quaternary amines can be used to control sapstain (blue coloration on freshly cut wood) and
provide above-ground rot protection [87].
A study of a range of quaternary amines concluded that ethoxylation had a negative effect
on wood preservation [88]. Combinations of cationic surfactants and other organic or inor
ganic fungicides such as copper and borate may be required to totally match the performance
of traditional wood preservatives [89].
F. Miscellaneous Applications
In addition to the major applications outlined above, cationic surfactants are used in a wide
variety of industrial and household applications (see Table 10), which together may account
for 20-30% of total consumption.
Most cationic surfactants are incompatible with anionic surfactants, and only amine oxides
have widespread use in cleaning compositions such as dishwashing liquids, bath foams, and
shampoos [45]. Quaternary amines and ethoxylated amines find use in consumer goods such
as acid cleaners and disinfectants, hair conditioners, and vehicle cleaners. The agrochemical
industry is a significant consumer of cationic surfactants, especially alkoxylates, which can
improve the penetration and rainfastness of herbicides [90]. New agrochemical uses include
“blossom thinning,” in which an application of cationic surfactant to fruit trees in bloom stops
fertilization and thus reduces the number of fruits produced. Fewer fruits means larger, more
valuable fruits, and the surfactant treatment avoids hand thinning of young fruits. In another
new application, “bud breaking,” cationic surfactants promote early blooming in fruit trees
[91].
The anticaking treatment of granular and crystalline fertilizers such as potash and ammo
nium nitrate typically involves applying 0.05-0.2% stearylamine to the hot fertilizer either as
a solution in oil or on a talc or clay carrier [92].
Cationic surfactants such as imidazoline salts are used as emulsifiers or corrosion inhibitors
in oil-based drilling fluids. Quaternary amines and diamines are used in corrosion inhibitor
626 James
Table 10 Miscellaneous Applications of Cationic Surfactants
Industry Application Chemical type
Cleaning and personal care Dishwashing Amine oxide

Shampoo Amine oxide
Bath foam Amine oxide
Hair conditioner Quat
Toilet cleaners Quat, amine oxide
Car rinses Dialkylquats
Laundry Amine oxides
Agrochemicals Fertilizer anticaking Stearylamine
Herbicide adjuvant Ethoxylated amine
Blossom thinning Alkoxylated amine
Bud breaking Alkoxylated amine
Petroleum Oil-based drilling fluid Alkylimidazoline
Demulsifier for wet crude Alkoxylated amines
Biocides Quats
Corrosion inhibitors, gasoline additives Alkylimidazolines
Textile Viscose spin bath additives Ethoxylated amines
Dye leveling Quats
Softening/handle Quats
Polymers Antistatic agents Ethoxylates, quats
PVC adhesion promotors Amidoamines
PU foam stabilizer Alkylmorpholines
Cationic SBR latex Ethoxylates, quats
formulations [93] and biocides for oil production. Ethoxylated amines are used in demulsifiers
for crude oil [94]. Cationic surfactants are also used as diesel and gasoline additives for
corrosion inhibition, and as detergents for carburetors, injectors, or valves.
In the manufacture of rayon, the cellulose xanthate solution (viscose) is spun into an acid
salt bath to regenerate the cellulose in fiber form. Ethoxylated amines and amidoamines are
added to both the cellulose xanthate itself and the spin bath to improve fiber quality. Other
textile applications include additives to improve dyeing and softener treatments.
Ethoxylated amines and quaternary amines are used in antistatic treatments for polyethyl
ene, propylene, styrene, and urethanes [95,96]. Ethoxylates may be incorporated in the poly
mer; quaternary amines are most often used in surface treatments. Amidoamines improve the
adhesion of protective coatings based on PVC plastisols. Other small applications include
the charge reversal of anionic rubber latex for use in the construction industry [97] and in
polymer foams.
In addition to having a variety of applications in industry and in consumer products, amines
of all types are important intermediates in the manufacture of other types of surfactants such
as amphoterics and sulfosuccinimates as well as nonsurfactants such as epoxy resins and
waxes.
G. Future Developments in Cationic Surfactants

Cationic surfactants provide the functions required for a wide variety of domestic and indus
trial applications. As a group their main disadvantages lie in the area of aquatic toxicity, their
irritating or corrosive effects on skin and eyes, and their incompatibility with anionic surfac
tants and other anionic dispersants used in cleaning compositions, pigments, and coatings.
Because of their rapid degradation, esteramines provide a solution to the problem of aquatic
toxicity, and with the growth in esteramine manufacturing capacity, substitution of other cat
ionic surfactants in applications outside fabric conditioning is to be expected. Esteramine
chemistry is flexible in terms of both the hydrophobic and hydrophilic parts of the surfactant
molecule, which means that a wide range of product requirements can be satisfied, although
their inherent hydrolytic instability may prevent their use in some cleaning and industrial
applications. Cationic sugar derivatives [96] may also offer improved ecotoxicological proper
ties, but these products are not yet commercialized.
Amphoteric surfactants can provide many of the same functions of cationic surfactants
with the advantages of lower ecotoxicity, less irritant nature, and compatibility with anionic
surfactants. Amphoterics may therefore also offer an alternative to cationic surfactants where
the pH of the system can be adjusted to ensure expression of their cationic properties. How
ever, regardless of any trend toward esteramines or amphoteric surfactants, the hydrophobic
portion of surfactants is likely to continue to be derived mainly from natural fats and oils
rather than petroleum.
REFERENCES
1. B. F. Haumann, European firms face new challenge, I N F O R M 2(5): 43 8 -4 4 7 (1991).
2. Hewin International Inc., Industrial cationic, amphoteric and amide surfactants and related prod
ucts, quoted in INFORM i ( l l ) : 984 (1990).
3. Frost and Sullivan International Report, quoted in I N F O R M 4 ( 2 ) : 179 (1993).
4. R. B. McConnell, Worldwide trends in fabric softeners, I N F O R M 5(1): 7 7 -8 5 (1994).
5. J. Baker, Surfactant prices head skywards, E u r o p e a n C h e m i c a l N e w s , Sept. 19, 1994, pp. 14-15.
6. M. Ward, Akzo Nobel marriage poised for new heights, E . C . N . , Sept. 19, 1994, pp. 23, 26.
7. R. A. Reck, Cationic surfactants derived from nitriles, in C a t i o n i c S u r f a c t a n t s — O r g a n i c C h e m i s
t r y , Surfact. Sci. Ser. 34 (J. M. Richmond, Ed.), Marcel Dekker, New York, 1990, pp. 163-186.
8. S. Billenstein and G. Blaschke, Industrial production of fatty amines and their derivatives, J. A m .
O i l C h e m . S o c . 6 1 ( 2 ) : 3 5 3-356 (1984).
9. R. G. Laughin, C a t i o n i c S u r f a c t a n t s — P h y s i c a l C h e m i s t r y , Surfact. Sci. Ser. 37 (D. D. N. Ru-
bingh and P. M. Holland, Eds.), Marcel Dekker, New York, 1990, p. 455.
10. U. R. Kreutzer, Manufacture of fatty alcohols based on natural fats and oils, J. A m . O i l C h e m .
S o c . 6 1 ( 2 ) : 343-348 (1984).
11. Hoechst, German Patent 2255701.
12. F. E. Friedli, Amidoamine surfactants, in C a t i o n i c S u r f a c t a n t s — O r g a n i c C h e m i s t r y , Surfact. Sci.
Ser. 34 (J. M. Richmond, Ed.), Marcel Dekker, New York, 1990, pp. 51 -1 0 0 .
13. G. W. Earl, Imidazoline surfactants, in C a t i o n i c S u r f a c t a n t s — O r g a n i c C h e m i s t r y , Surfact. Sci.
Ser. 34 (J. M. Richmond, Ed.), Marcel Dekker, New York, 1990, pp. 101-144.
14. K. E. Visek, Fatty amide surfactants, in C a t i o n i c S u r f a c t a n t s — O r g a n i c C h e m i s t r y , Surfact. Sci.
Ser. 34 (J. M. Richmond, Ed.), Marcel Dekker, New York, 1990, pp. 1 -5 0 .
15. R. Puchta, Cationic surfactants in laundry detergents and laundry after treatment aids, J . A m . O i l
C h e m . S o c . 6 1 ( 2 ) : 367 -3 7 6 (1984).
16. Keith D. Stanley, Akzona Inc, U .S. Patent 4376736, 1983.
17. J. Wates, Solution behaviour of cationic surfactants relevant to industrial applications, Ph.D The
sis, University of Salford, UK, 1990.
18. J. D. Sauer, Amine oxides, in C a t i o n i c S u r f a c t a n t s — O r g a n i c C h e m i s t r y , S u r f a c t . S c i . S e r . 34
(J. M. Richmond, Ed.), Marcel Dekker, New York, 1990, pp. 2 7 5-296.
19. C. J. Turney, F. E. Friedli, and P. J. Frank, Kinetics and preparation o f amine oxides, J . A m .
O i l C h e m . S o c . 7 1 ( 1 ) : 7 9 3 -7 9 4 (1994).
628 James
20 . P. A. Vening, W. E. Adam, H. Bruil, and H. Rorig, Amines and amine derivatives, especially
amine oxides, XIII Jomadas del Comité Español de la Detergencia tensioactives y afines Barce
lona, Feb. 2 4 -2 6 , 1982.
21 . E. H. White and D. J. Woodcock, in T h e C h e m i s t r y o f t h e A m i n o G r o u p (S. Patai, Ed.), Intersci
ence, New York, 1968, pp. 4 0 9-427.
22 . U. Sebold, Fabric softeners worldwide, 3rd World Conference and Exhibition on Detergents, Mon
treux, Switzerland, Sept. 2 6 -3 0 , 1993.
23. J. W. Smith, Basicity and complex formation, in T h e C h e m i s t r y o f th e A m i n o G r o u p , S. Patai,
Ed.), Interscience, New York, 1968.
24. A. D. James, J. M. Wales, and E. Wyn-Jones, Determination of the hydrophilicities of nitrogen-
based surfactants by measurement of partition coefficients between heptane and water, i . C o l l o i d
I n t e r f a c e S c i . 1 6 0 : 158-165 (1993).
25. S. Ikeda, M. A. Tsunada, and H. Maeda, The application of the Gibbs adsorption isotherm to
aqueous solutions of a nonionic-cationic surfactant, J . C o l l o i d I n t e r f a c e S c i . 67(2): 336-348
(1978).
26. D. Vollhardt, M. Wittig, G. Petrov, and G. Malewski, Infrared spectroscopic study of phosphate
counter ions bonded in skimmed monolayers and transferred multilayers of octadecylamine, J .
C o l l o i d I n t e r f a c e S c i . 706(1): 2 8 -3 2 (1985).
27. J. S. Laskowski and Qi Dai, Collector-extender flotation in the processing of coarse fractions
of potash ores, Proc. XVIII Int. Mineral Processing Congress, Sydney, May 2 3 -2 8 , 1993, pp.
1045-1050.
28. A. Samadijavan, Elektrophoretische Untersuchung des Brechverhaltens von Bitumenemulsionen,
B i t u m e n 4 : 151-156 (1994).
29. J. M. Wales and A. D. James, Zeta potential measurements of bitumen emulsions and road aggre
gates, Proc. First World Congress on Emulsion, Paris, Oct. 1 9 -22, 1993, Vol. 1, 1-40/089.
30. J. M. W. Mackensie, Electrokinetic properties of Nujol-flotation collector emulsion drops, T r a n s .
S o c . M i n i n g E n g . A I M E , 2 4 4 : 3 9 3-400 (1969).
31. A. A. Woolf, Estimation of boiling points in homologous series with a new linear relationship,
C h e m . I n d . , Feb. 4, 1991, pp. 9 6 -9 7 .
32. K. R. Smith, Cationic surfactants from petroleum feedstocks, in C a t i o n i c S u r f a c t a n t s : O r g a n i c
C h e m i s t r y (J. M. Richmond, Ed.), Marcel Dekker, New York, 1990.
33. W. M. Linfield, Fatty oxazolines and imidazolines, J . A m . O i l C h e m . S o c . 67(2): 437-441 (1984).
34. Akzo Chemicals Inc., Industrial Surfactants Nitrogen Derivatives General Catalog, 1992.
35. Pulcra SA, German Patent 4334365-A l.
36. N. Ikeda, N. Sanefujo, K.-K. Lou, M. Aratona, and K. Moromura, Surface adsorption and mi
celle formation of the mixture of A-methylated dodecylammonium chloride and dodecylammonium
chloride, J . C o l l o i d I n t e r f a c e S c i . 1 6 4 : 439 -4 43 (1994).
37. M. te Nijenhuis, Akzo Nobel Research Centre, Deventer, Holland, unpublished work.
38. B. Rozyika-Roszak, Micellisation of some amphiphillic quaternary ammonium chlorides, J . C o l
l o i d I n t e r f a c e S c i . 1 4 0 ( 2 ) : 5 3 8 -5 4 0 (1990).
39. H. Hagslatt and K. Fontel, Divalent surfactants— the influence o f headgroup charge on the phase
behaviour of alkyldiamine surfactant/water system, J . C o l l o i d I n t e r f a c e S c i . 1 6 5 : 4 3 1 -4 4 4 (1994).
40. B. Rozyika-Roszak, S. Przestalskis, and S. Witek, Calorimetric studies of the micellization of
some amphiphilic betaine ester derivatives, J . C o l l o i d I n t e r f a c e S c i . 1 2 5 : 80 (1988).
41. M. Tuncay and S. D. Christian, A study of the binding of dimethyldodecylamine oxide by
j8-cycodextrin using surface tension measurements, J . C o l l o i d I n t e r f a c e S c i . 1 6 7 : 181-185 (1994).
42. T. M. Muzyczko, S. Shore, and J. A. Loboda, Fatty Amidoamine derivatives: N , N-dimethyl-N-
(3 alkylamido-propyl) amines and their salts, J . A m . O i l C h e m . S o c . 4 5 : 720 -7 2 5 (1968).
43. N. Nemoto and M. Kuwahara, Self diffusion and viscoelasticity of elongated micelles from cetyl-
trimethylammonium bromide in aqueous sodium salicylate solution. II. Temperature effect. C o l
l o i d P o l y m . S c i . 2 7 2 : 846-854 (1994).
44. J. Myska and P. Stem, Properties of a drag reducing micelle system. C o l l o i d P o l y m . S c i . 2 7 2 :
542-547 (1994).
45. H. Rorig and R. Stephan, Amine oxides and their applications, in I n d u s t r i a l A p p l i c a t i o n s o f S u r

(D. R. Karsa, Ed.), Spec. Publ. No. 77, Royal Society of Chemistry, London, 1990,
fa c ta n ts II
pp. 211-234.
46. J. Matsumoto, Internal and interfacial structure of small vesicle in aqueous colloid of didodecyldi-
methylammonium bromide. C o l l o i d P o l y m . S e i . 2 7 0 : 49 2 -4 9 7 (1992).
47. R. G. Laughlin, The role of swelling methods in surfactants phase science; past, present and
future, A d v . C o l l o i d I n t e r f a c e S e i . 4 1 : 57-59 (1992).
48. A. D. James and P. H. Ogden, Effect of manufacturing variables on the viscosity of aqueous
di(hydrogenated tallowalkyl)dimethyl ammonium chloride, J . A m . O i l C h e m . S o c . 56(4): 5 4 2 -
547 (1979).
49. C. Monager and V. Cabuil, Synthesis of magnetic DDAB vesicles, C o llo id P o ly m . S e i. 272:
1295-1299 (1994).
50. R. Smith and C. Tanford, Hydrophobicity of long chain /i-alkylcarboxylic acids, as measured by
their distribution between heptane and aqueous solutions, P r o c . N a t l . A c a d . S e i . U . S . A . 7 0 ( 2 ) :
289-293 (1973).
51. O. Boen, Akzo Nobel Chemicals, Deventer, Holland, Unpublished work.
52. B. T. Ingram and R. H. Ottewill, Adsorption of cationic surfactants at interfaces, in C a t i o n i c
S u r f a c t a n t s — P h y s i c a l C h e m i s t r y , Surfact. Sei. Ser. Vol. 37 (D. N. Rubingh and P. M. Holland,
Eds.), Marcel Dekker, New York, 1990, pp. 8 7-140.
53. S. D. Escher and E. Oliveros, Quantitative study of factors that influence the substantivity o f
fragrance chemicals on laundered and dried fabrics, J . A m . O i l C h e m . S o c . 77(1): 3 1 -3 9 (1994).
54. R. O. Keys, Formulation and production of concentrated rinse cycle fabric softeners, H a p p i ,
March 1995, pp. 95-97.
55. G. R. Whalley, Fabric conditioning agents, H a p p i , February 1995, pp. 5 5 -5 8 .
56. H. Berenbold, Aktuelle Situation der Wascheweichspuler, S O F W - J . 1 2 0 ( 1 1 ) : 6 7 8 -6 8 4 (1993).
57. Akzo Nobel Surface Chemistry, A N e w , F l e x i b l e S o f t e n e r G e n e r a t i o n , Product data leaflet.
58. J. P. Tatum, Organophilic clays, in I n d u s t r i a l A p p l i c a t i o n s o f S u r f a c t a n t s (D. Karsa, Ed.), Royal
Society of Chemistry, London, 1987, pp. 289-306.
59. O. Ponsati, A. Truis, C-P Herold, H. Müller, C. Nitsch, S. von Tapavicza, Patent applicant;
Henkel Kommanditgesellschaft auf Aktien, Use of selected oleophilic compounds with quaternary
nitrogen to improve the oil wettability of finely divided clay and their use as viscosity improvers.
World Patent Appl. W092/19693, 1992.
60. D. W. Fuerstenau and R. Herrera-Urbina, Adsorption of cationic surfactants and the flotation of
minerals, in C a t i o n i c S u r f a c t a n t s — P h y s i c a l C h e m i s t r y , Surfact. Sei. Ser., Vol. 37 (D. N. Rubingh
and P. M. Holland, Eds.), Marcel Dekker, New York, 1990, pp. 4 0 7 -4 4 7 .
61. H. L. Shergold, Cationic surfactants in mineral processing, in I n d u s t r i a l A p p l i c a t i o n s o f S u r f a c
t a n t s (D. Karsa, Ed.), Spec. Publ. No. 59, Royal Society o f Chemistry, London, 1987, pp.
269-288.
62. R. J. Roman, M. C. Fuerstenau, and D. S. Seidel, Mechanisms of soluble salt flotation. Part 1,
T r a n s . S M E , March 1968, pp. 5 6 -7 0 .
63. H. Schubert, What goes on during potash flotation, E n g . M i n i n g J . , March 1967, pp. 94-97.
64. D. Mörstedt, M. Winzer, R. Köhler, and F. Püschel, Verfahren zur Flotation von NaCl aus Salz
gemischen, E. German Patent 64443, 1968.
65. W. J. Trahar, The concentration by flotation of cement rock from Fyansford Victoria, Report No.
612, CSIRO, Melbourne, 1961.
66. Financial Mining Industrial and Shipping Corporation, The separation o f magnesite from its con
taminants, U .S. Patent 3979251, 1975.
67. O. Burghardt and H. D. Wasmuth, Über die Anwendung der indirekten kationaktiven Flotation
bei der Augbereitung von Eisenerzen, A u ß e r e i t u n g s . T e c h n i k . 3 : 103-117, 1975.
68. Dow Chemical Co., V-Aminopiperazine condensates for the benefication of phosphate ore, U .S.
Patent 4301004, 1981.
69. International Minerals and Chemical Corporation, Flotation process for the recovery of phosphate
values from ore, U .S. Patent 4737273, 1988.
630 James
70. C. Caproni, R. Ciccu, M. Ghiani, and I. Trudi, The processing of oxidised lead and zinc ores in
the Campo Pisano and San Giovanni plants (Sardinia), in P r o c e s s i n g o f O x i d i s e d a n d M i x e d O x
i d e - S u l p h i d e L e a d Z i n c O r e s S e m i n a r , Proc. X lllth Int. Mineral Processing Congress, Warsaw,
1979, (J. Laskowski, Ed.), Polish Scientific Publishers, pp. 7 1 -9 1 .
71. W. H. Eddy, E. W. Collins, J. S. Browning, and G. V. Sullivan, Recovery of feldspar and glass
sand from South Carolina waste granite fines, Report o f Investigations 7651, U .S. Dept, of the
Interior, Bureau of Mines, 1972.
72. K. Hanumantha Rao, B.-M . Antti, and K. S. E. Forssberg, Flotation of mica minerals and selec
tivity between muscovite and biotite while using mixed anionic/cationic collectors. M i n e r a l s M e t -
a l l . P r o c e s s . , August 1990, pp. 127-132.
73. N. O. Clark, Pochin and Co. Ltd., Improvements in or relating to the treatment of clay, Br. Patent
1224877, 1967.
74. J. A. R. Faucher, Concentration o f pyrochlore, T r a n s . S M E , September 1964, pp. 2 55-258.
75. S. R. Rao, R. Espinosa-Gomez, J. A. Finch, and R. Biss, Effects o f water chemistry on the
flotation of pyrochlore and silicate minerals. M i n e r a l s E n g . 7(3): 189-202 (1988).
76. Kenobel AB, Material and method for upgrading of carbonate mineral, Swedish Patent 458747.
77. T. Jiaxing, S. Baoqi, C. Zhengbing, C. Xianguo, Z. Huiling, and Z. Shaozhong, Theoretical
studies and production practice of acidless and fluoless flotation of silica sand, Proc. XVIIIth Int.
Mineral Processing Congress, Sydney, May 2 3 -2 8 , 1993, pp. 851-855.
78. K. Hanumantha Rao and K. S. E. Forssberg, Solution chemistry of mixed cationic/anionic collec
tors and flotation separation of feldspar and quartz, Proc. XVIIIth Int. Mineral Processing Con
gress, Sydney, May 2 2 -2 8 , 1993, pp. 8 37-843.
79. A. D. James and D. Stewart, Cationic surfactants in road construction and repair, in I n d u s t r i a l
A p p l i c a t i o n s o f S u r f a c t a n t s , Vol. 2 (D. Karsa, Ed.), Spec. Publ. 77, Royal Society of Chemistry,
London, 1990, pp. 338-355.
80. A. D. James, P. D. McCool, D. Stewart, G. E. Varcoe, and A. R. Woodside, A gas chromatogra
phy method for the determination of antistripping agents in hot binders and the significance of this
technique to surface dressing, Proc. Int. Symp., C h e m i s t r y o f B i t u m e n s , Rome, Italy, June 5 -8 ,
1991, pp. 627-653.
81. Syndicat des Fabricants D ’Emulsions Routières de Bitume, A few figures about emulsions con
sumption around the world, Proc. 1st World Bitumen Emulsion Manufacturers Congress, Paris,
Oct. 18, 1993, p. 61.
82. A. R. Lane and R. H. Ottewill, The preparation and properties o f emulsions stabilised by cationic
surface-active agents. T h e o r y a n d P r a c t i c e o f E m u l s i o n T e c h n o l o g y , SCI, 1974.
83. Akzo Chemicals, Technical Data Sheet B-10, Amersfoort, Holland, 1992.
84. B. Davis and P. Jordan, An overview o f the biocidal activity of cationics and ampholytes, in
I n d u s t r i a l A p p l i c a t i o n s o f S u r f a c t a n t s , Vol. 2 (D. Karsa, Ed.), RSC Spec. Publ. No. 77, Royal
85. R. A. Stephenson, Optimisation of the performance of quaternary ammonium compounds, in I n
d u s t r i a l A p p l i c a t i o n s o f S u r f a c t a n t s , Vol. 2 (D. Karsa, Ed.), RSC Spec. Publ. No. 77, Royal
86 . E. Bessems and A. F. Clemmit, Quaternary ammonium compounds: evaluation and application in
the control of sulphate-reducing bacteria, in C h e m i c a l s in th e O i l I n d u s t r y (P. H. Ogden, Ed.),
RSC Spec. Publ. No. 45, Roy. Soc. Chem., London, 1983, pp. 159-170.
87. A. F. Preston, P. J. Walcheski, P. A. McKaig, and D. D. Nicholas, Recent research on alkylam-
monium compounds in the U .S ., A m . W o o d P r e s e r v e r s P r o c . I I : 3 3 1 -3 4 6 (1987).
88 . H. Tang and J. N. R. Ruddick, Evaluating the potential of amine chemicals for use as wood
protecting agents. Part 1. Investigation of cationic component of quaternary ammonium com
pounds, Proc. 25th Annu. Meeting Int. Res. Group on Wood Preservation, Bali, Indonesia, May
29-June 3, 1994.
89. R. J. Tillot and C. R. Coggins, Non-arsenical water-borne preservatives— a review o f performance
and properties. R e c o r d o f th e 1 9 8 1 A n n u a l C o n v e n t i o n o f th e B r i t i s h W o o d P r e s e r v i n g A s s o c i a t i o n ,
pp. 32 -4 6 .
90. D. J. Turner and P. Tabbush, Studies with alternative glyphosate formulations, in A p p l i c a t i o n a n d

B io lo g y (E. S. E. Southcombe, Ed.), Monograph No. 28, British Crop Protection Council, 1985,
pp. 135-145.
91. The Fruits of Research, P r o g r e s s w i t h A k z o N o b e l S u r f a c e C h e m i s t r y , Akzo Nobel Surface Chem
istry AB, Stenunsund, Sweden, Autumn 1994, pp. 4 -5 .
92. J. R. Parks and J. Granok, Fertilizer conditioning problems solved with those versatile amines,
F a r m C h e m . , October 1967, pp. 5 1 -6 2 .
93. A. Nestle, Corrosion inhibitors in petroleum production primary recovery, in C o r r o s i o n I n h i b i t o r s
(C. C. Nathan, Ed.), NACE = National Association of Corrosion Engineers, Houston, 1973, pp.
6 1 -75.
94. R. Bohm and R. Kupfer, Improved demulsifier chemistry: a novel approach in the dehydration o f
crude oil, Friedrich Straiss, Society of Petroleum Engineers of AIME, August 1991, pp. 3 34-338.
95. W. Jerzykiewicz and J. Wypych, Studies on the influence of antistatics based on tertiary and
quaternary amines on the behaviour of plasticized polyvinyl chloride, T e n s i d e D e t e r g . 1 3 ( 2 ) : 7 4 -
76 (1976).
96. B. Davis, Antistatic/conductivity additives in paint and plastics, in I n d u s t r i a l A p p l i c a t i o n s o f S u r
f a c t a n t s (D. Karsa, Ed.), RSC Spec. Publ. No. 59, Roy. Soc. Chem., London, 1987, pp. 3 0 7 -1 7 .
97. Dr. Hermann Perrey, Dr. Matner Martin, and Heinz-Gerd Nonnenbruch, Bayer AG, W. German
Patent AppL, DE 3628902-A l, 1986.
98. T. Kida, K. Yurugi, A. Masuyama, Y. Nakatsuji, D. Ono, and T. Takeda, Preparation and prop
erties of new surfactants containing D-glucosamine as the building block, J . A m . O i l C h e m . S o c .
72(7): 1 1 3 - l S O (1995).
25_______________________
Nonionic Surfactants
Guido Bognolo
I C I P e r fo r m a n c e C h e m ic a ls , E v e r t e r g , B e lg iu m
I. INTRODUCTION
N o n io n ic su rfacta n ts are o n e o f th e m o st im p o rta n t cla sse s o f a m p h ip a th ic m a te ria ls, se c o n d
only to the anionic surfactants in terms of volume consumed and certainly more widespread
in terms of end uses and applications. Their key feature is the absence of discrete electric
charges on the molecules, which confers different properties from those of the anionic, cat
ionic, or amphoteric surfactants (Fig. 1), and this often results in unique interfacial effects.
Chemically the nonionic surfactants are esters, ethers, amides, amines, or amine oxides.
Each of these classes generally includes several subclasses, depending on the nature of the
hydrophilic and hydrophobic moiety. These in turn include many individual components dif
fering in the chain length of the hydrophobe, the degree of polymerization of the hydrophile,
and the relative proportion of hydrophile and hydrophobe. It follows that the number of com
mercial products available to the formulator is astonishingly high, probably in the region of
5000 different chemicals or more. Table 1 gives an overview of the nonionic surfactants of
commercial interest.
Products of natural origin with structure and properties typical of nonionic surfactants have
been used specifically or (most often) accidentally in various preparations for hundreds of
years. Examples include glycerides and their hydrolysis products, which find use in cosmetics,
food, and colors; and saponins from plant roots, which are used in fabric cleaning. However,
the real development of nonionic surfactants as a stand-alone class of products with identified
chemistry and performance properties was prompted by two factors:
On the one hand, the emergence of fat substitutes required emulsifiers that enabled the
incorporation of water into oil phases, which favored the development of the chemistry
of the glycerol esters and ensured a market for the new products. This paved the way
for the use of polyhydric hydrophiles other than glycerol, e.g., sorbitol, which was
becoming available in large quantities and at low cost as a by-product in the manufac
ture of mannitol, which was used in the explosives industry for the production of deto
nators based on mannitol hexanitrate.
633
634 Bognolo
(A)
nonionic surface-active agent
amphoteric surface-active agent
Fig. 1 Schematic representation o f the four main categories of surface-active agents. (A) Nonionic
surface-active agent; (B) anionic surface-active agent; (C) cationic surface-active agent; (D) amphoteric
surface-active agent.
On the other hand, the industrial availability of ethylene oxide (EO) and the growing
understanding of its reactivity with molecules containing active hydrogen atoms enabled
the production of more water-soluble surface-active agents and emulsifiers. This broad
ened the range of industrial applications, and at the same time created new chemicals,
providing solutions to emerging technological demands, initially from the textile and
food industry, but expanding rapidly to other industrial sectors such as cosmetics, crop
protection, plastics, and paints, to name just a few.
Fundamental in this respect is the work from Schoeller, Wittwer, Steindorff, Balle, Horst, .
and Michel, who in the early 1930s developed and patented the synthesis of ethoxylated
nonionics from fatty acids, alcohols, amines, and alkylphenols [1,2]. Early in the next decade,
the ethoxylation of low polarity glycerol and sorbitol esters enabled the production of more
hydrophilic compounds. These could then be blended, in various proportions, with the origi
nal esters to emulsify a range of oils of different chemical structures and physical properties,
both in the form of water-in-oil (i.e., with oil being the continuous external phase, abbreviated
w/o) and oil-in-water (i.e., where water is the continuous external phase, abbreviated o/w).
The introduction of the hydrophilic-lipophilic balance (HLB) concept [3] and the extension
of its subsequent evolutions and modifications [4-17] gave a tremendous impetus to the appli
cation technology of surfactants in general and of the nonionics in particular, creating new
applications for existing products as well as meeting the demand for new chemical species
capable of solving challenging wetting, dispersion, and emulsification problems. In recent
years the development of new chemicals has slowed down, partly because of the exhaustion
of the synthetics possibilities, partly because of the reduced funds allocated to basic research,
and partly because of the ever-growing constraints from regulations and legislation. As a
result, there has been a renewal of interest in chemistry and products that initially received
Nonionic S urfactan ts 635
Table 1 Nonionic Surfactants of Commercial Interest
Common
Chemical class Hydrophobe Hydrophile denomination
Ester derivatives Fatty acid EG, POE^ Glycol and polyethylene gly
col esters
Gly, PGL’^ Glycerol and polyglycerol
esters
Sor^’ Sorbitol and sorbitan esters
Suc*" Sucroesters
Suc + Gly Sucroglycerides
Glu^ Glucoside esters
Ether derivatives Fatty alcohol POE" Ethoxylated alcohols

(natural or
synthetic)
Gly, PGL’’ Glycidyl ethers
Glu Alkylpolyglucosides
Alkyl phenol ^ POE" Ethoxylated alkylphenols
Propylene oxide‘s POE Ethylene oxide/propylene
oxide copolymers
Glu^’ Glucosides ethers
Amides Fatty acid amide POE" Ethoxylated amides
Fatty acid Gla Glucamides
Amines Fatty acid amines POE*’ Ethoxylated amines
Amine oxides Fatty acid amines Oxy Amine oxides
Note\ See list of abbreviations preceding Reference at end of chapter.

^The hydrophobic character of the final molecule can be enhanced by adding to this moiety other hydrophobic groups
(usually POP).
‘’The hydrophilic character of the final molecule can be reduced by adding to this moiety other hydrophobic groups
(usually POP) or enhanced by adding other hydrophilic groups (usually POE).
Încludes higher aromatics, e .g ., j8-naphthol.
Încludes all block, block-etheric, and etheric POE/POP copolymers with different initiators.
only limited attention. This process is stimulated by the increasing demand for properties that
were disregarded or were of low relevance in the past, prompted by changes in public opinion
or by raw materials availability or by economic and ecotoxicological considerations.
For example, alkylpolyglucosides, which were used in small volumes— essentially ex
ploiting their properties as hydrotropes in high alkalinity bottle-washing formulations—are
now experiencing a rapid expansion into sectors such as household detergents and cosmetics,
driven by their mildness to the skin, biodegradability, and natural origin and renewability of
the constituent raw materials. Similarly, sugar esters, which were almost confined to the role
of laboratory curiosities because of synthesis difficulties and the consequent high manufactur
ing costs, are now being paid increased attention, particularly in food and personal care appli
cations as high HLB, ethylene oxide-free emulsifiers.
A new stimulus to the use of nonionic surfactants arose in recent years from their superior
stabilization power for both emulsions and suspensions. The development of the theories of
steric stabilization and the appreciation of the performance of these products as steric stabiliz
ers enhanced that application in various fields.
636 Bognolo
II. WORLD PRODUCTION

Since the initial development of modem nonionic surfactant chemistry, this industry has expe
rienced constant growth. It is estimated that in 1994 about 2.6 million metric tons of nonionic
surfactants was produced on a worldwide basis for a value in excess of 1.8 billion pounds
sterling. This represents about one-third of the total estimated surfactants market.
Because of the complex nature of the product range (often related to other surfactants or
chemicals), of the routes to market, and of the industry stmcture and evolution, it is impossi
ble to produce a precise list of producers of nonionic surfactants. Table 2 attempts to give a
snapshot of the worldwide situation at the end of 1994.
Table 2 Major Producers o f Nonionic Surfactants (1994, World Situation)
Europe
Austria Th. Boehme
Belgium Oleofina
Bulgaria Bulgarska Roza AD
Velila
Denmark Palsgaard Ind.
Danisco (Grindsted)
France Ceca (Atochem)
Gattefosse
ICI
Marchon (Albright & Wilson)
PPG Ouvrie
Stearineries Dubois
Seppie
Sidobre-Sinnova (Henkel)
Stepan
FSU (former Soviet Union) Barva
Biok
Fine Organic Synthesis
Ivhymprom
Kaprolactam
Kaustik
Organic Synthesis
Orgsteklo
Salavatnefteorgsinthez
Sintez
Volzhski Kauchuck
Germany BASF
Bayer
Th. Boehme
Buna
Chem-Y (Kao)
Grunau (Henkel)
Henkel
Hoechst
Huels
ICI
Peter Greven
Nonìonic Surfactants 637
Table 2 Continued
Re wo (Witco)
Stockhausen (Huels)
Th. Goldschmidt
Unichema
Zschimmer & Schwarz
Holland Dow
Dr. Kolb
Quest
Servo (Huels)
Shell
Sybron Chemie
Unichema
Italy Akzo Nobel
DAC (Condea)
Enichem Augusta (RWE)
Fratelli Lamberti
Geronazzo (Rhone-Poulenc)
Henkel
Hoechst
Zschimmer & Schwarz
Norway Pronova
Unger
Poland Rokita
Portugal Companhia Petroquimica Do
Barreiro
Spain ICI
Kao
Pulcra (Henkel)
Unichema
Union Derivan
Slovak Republic Novaky
Sweden Berol (Akzo Nobel)
Switzerland Dr. Kolb
United Kingdom Baker Performance Chemicals
Croda
Harcros
Hickson Manro
ICI
Lanstar
Shell
Unichema
Africa
South Africa Croda
Hoechst
Manro
Americas
Canada ICI
United States ADM
638 Bognolo
Table 2 Continued
Air Products
Amerchol (Union Carbide)
American Ingredients
Baker
BASF
Calgene
Croda
Du Pont
Eastman
Ethox
Ecolab
Grinsted
Harcros
Hodag
Hoechst Celanese
Henkel
Heterene
High Point (Kao)
ICI
Karlshamns
Lipo
Lonza
McIntyre
Mona
Milliken
Nalco
Olin
Piedmont
PPG/Mazer
Rhone-Poulenc
Sandoz
Scher
Sip
Stepan
Shell
Sherex (Witco)
Texaco
Union Carbide
Vista
Van Dyck
Van den Bergh
Witco
Mexico Arancia Tensioactivos
Canamex
Christianson
Henkel
Hoechst
Poliquimia Apizaco
Poliquimia de Tlaxcala
Polyurechemia
Brazil Aquatec
Nonionic Surfactants 639
Table 2 Continued
Cliffs
Henkel
Hoechst
Oxiteno
Ultraquimica/Oxiteno
Colombia Incolgasos
Protechnio
Venezuela Ethoxyl
Argentina Dispersol San Luis
Asia
India Albright & Wilson
Bayer India
Colour Chemicals (Hoechst)
Dai Ichi
Globe Industries
Henkel India
Hico Products
ICI
India Glycols
Laffans Petro
Manali Petro
Silvolical Industries
S M Dye Chemicals
Sunshield Chemicals
Swastik
United Pestichem
Unitop Chemicals
Vashuda Chemicals
Venus Ethox
Japan Asahi Denka
Chuo Kasei
DKS
Henkel Hakusui
Kao Corporation
Kawaken Fine Chemicals
Lion Corporation
Marubishi Oil Chemicals
Matsumoto
Miyoshi Oil & Fat
Mitsubishi Kasei
New Japan Chemicals
Nicca Chemicals
Nikko Chemical Co.
Nippon Nyukazai
Nippon Oils & Fats
Nippon Shokubai
Nissan Chemical Ind.
Nisso Yuka
Riken Vitamin
Sakamoto Yakuhin
Sanyo Chemical Industries
640 Bognolo
Table 2 Continued
Taiyo Chemical Industries

Takeda Chemical
Takemoto Oils and Fats
Toho Chemical Industries
Tokai Seiyu Ind.
Yokkaici Gosei
Yoshimura Oil Chemicals
Taiwan Jintex
Pan Asia Chemical
Sino-Japan Chemical
TWSC
People’s Republic of China Beijing Synthetic Chemicals Plant
Jilin
Nanying Zhongshou Chemical Plant
Shanghai Auxiliary Plant
Shanghai Gao Qiao Petrochemical Plant
Shanghai Henkel Oleochemical Company
Tianjin Auxiliary Plant
Korea Dong Nam
Dong Sung
Han Nong
II Chil
KOA
Korea Polyol
Singapore Croda
EMPL
Salim/A&W
Shell/EGS
Thailand Kao
Australia Albright & Wilson
Henkel/Nopco
Harcros
ICI
III. RAW MATERIALS

Nonionic surfactants are manufactured using raw materials derived from renewable resources
as well as from fossil feedstock. Table 3 gives a summary of the natural/renewable component
versus the petrochemical component in the major nonionic classes. It should be made clear
that even those surfactants produced from completely natural and renewable raw materials
require chemical processing and products, e.g., catalysts, for their manufacture and therefore
have an environmental impact to a greater or lesser extent. Section VII deals in detail with
the toxicological and environmental aspects of nonionic surfactants.
Figure 2 illustrates schematically the raw materials “road map” for commercial nonionic
surfactants. The following subsections deal specifically with the individual raw materials.
Table 3 Natural/Renewable versus Petrochemical Component in the Major Nonionic Classes
Surfactants class Hydrophile Hydrophobe Percent “renewable”
Fatty alcohols POE P (POE) P (SFA) 0

P (POE) N (NFA) ----a
Fatty alcohols POE/POP P (POE/POP) P (SFA) 0
P (POE/POP) N (NFA) — ^
Alkyl phenols POE P (POE) P (AP) 0
Alkylpolyglucosides N (Glu) N (N¥A)^ 100
Glycerol esters N (gly/PGL)^ N (NFAc) 100
(inch polyglycerol)
Sorbitan esters N (Sor)*^ N (NFAc) 100
Sorbitan esters POE P (POE) N (NFAc) __ a
POE/POP copolymers P (POE) P (POP) 0
Sugar esters N (Suc) N (NFAc) 100
Polyglycerol ethers N (PGL)b N (NFA)^ 100
PEG esters P (POE) N (NFAc) __ a
Amines POE and POE/POP P (POE/POP) N (NFAm)^ a
P (POE/POP) P (SFAm)" 0
Amides POE P (POE) N (NFAm) __ a
Amine oxides P (Oxy) N (NFAm) __ d
Note: See Abbreviations at end of chapter.

^Depends on the HLB of the surfactant.
Âlthough the origin of the raw material is renewable, significant chemical processing is required.
Ûncommon.
^Depends on structure and molecular weight of the amine.
A. Ethylene Oxide
Ethylene oxide is one of the major building blocks for nonionic surfactants. It is a highly
reactive three-member cyclic ether that gives nucleophilic addition reactions of the SN2 or
SNl type [18,19] with molecules containing active hydrogen atoms through cleavage of the
carbon-oxygen bond. It was first prepared in 1859 by Wurtz [20] by dehydrohalogenation of
2 -chloroethanol, and the initial industrial processes were based on this chemistry [2 1 ].
The direct oxidation of ethylene dates from 1931 [22,23]. Nowadays, the vast majority of
ethylene oxide is produced by silver-catalyzed oxidation in air-based or oxygen-based pro
cesses. Other than for the manufacture of surfactants, it is used for the production of a variety
of chemical intermediates, e.g., polyether polyols for polyurethanes, or polyethylene glycol
for polyesters or products such as antifreeze glycols, glycol ether solvents, fire-resistant hy
draulic fluids, quenchers, and lubricants.
It is estimated that the worldwide production capacity for EO in 1994 was 11.1 million
metric tons split approximately 30% in Europe, 40% in the Americas, 20% in Asia and
Australia, and 10% in Japan.
Much is known about the physicochemical properties of EO [19], but the nature of the
chemical bond is not completely established. Figure 3 gives the structure, bond length, and
bond angles as determined experimentally [24,25]. The carbon-oxygen-carbon angle is about
50° less than that in the dimethyl ether [26], and the length of the carbon-carbon bond (1.470
A) is intermediate between that of a single bond and that of a double bond (1.55 and 1.35 A,
respectively). The CH 2 groups lie in a plane perpendicular to the plane of the ring [27].
Petroleum Starch m
Natural m Oils & Fats lO
m f T" r T” -------- T " -------- T p --------- y
fO Benzene Wax Kerosene
“Ethylene r -f-
Propylene Fatty Fa% Glycerol Glucose Sucrose
Gas Oil Acids
T isobutylene
T t Esters
Linear Ethylene Propylene
Paraffins Oxide Oxide
ID.
-L L t
p Alkyl Polyglycerol
Phenol ^ Phenol •Fatty— — I
Alcohols
0 Polyisobutylene
CL __ 1__ Sorbitol
Fatty
1 Amines
___
Oxide .
1 f Propylen BOxide
Fai ly Acid CoPolyn- ers
EtI oxylates Amine
Ethoxylates , L—Ì
S3
O Fatty
P Alcohol
Ethoxylates
Polyglycerol
Esters
c Amine
Oxides
Alkyl Phenol
Ethoxylates
Alkyl Sorbitan
Glucosides Esters
Glycerol
Esters
Sorbitan
Polyethylene Glycerol Ester
Glycols Ester Ethoxylates
Ethoxylates
m
o
m
3
O
o
Fig. 3 Structure, bond lengths and bond angles for ethylene oxide. (From Ref. 199, p. 10.)
Because of its physicochemical properties and reactivity, EO must be handled with extreme
care. It presents both physiological and explosive hazards, and its manufacturers continue to
make considerable effort to improve the procedures and standards for its safe transport and
handling [28].
B. Glycerol
Glycerol was discovered in 1779 by Scheele and was recognized as the ubiquitous component
of fats and oils by Chevreul in 1813. It can be obtained either as a by-product in the conver
sion of fats and oils to fatty acid or fatty acid methyl esters (natural glycerol) or from syn
thetic processes.
Fats and oils are the main sources of natural glycerol, which is obtained from them by
means of [29]
1. High pressure splitting. Water and fat are fed into a splitting column in a countercur
rent fashion at 5 -6 bar and 250-260°C.
2. Continuous methanolysis.
3. Saponification of natural oils with caustic soda.
Synthetic glycerol is produced from propene through any of the following intermediates:
1. Allyl chloride (IG Farben, 1943) [30-32]
2. Acrolein (Shell, 1958) [33]
3. Propene oxide [34]
The major uses of glycerol are in pharmaceuticals and cosmetics, carboxylic acid esters,
alkyd resins, polyols, food, cellulose, and organic nitrates.
Polyglycerols are produced industrially by the base-catalyzed condensation of glycerol and
as by-products of epichlorohydrin hydrolysis [35]. The worldwide consumption of glycerol is
on the order of 600,000 tpa, with an approximate ratio of natural to synthetic of 3:1.
644 Bognoio
C. Sorbitol
Sorbitol (D-glucitol) occurs in nature in the fruits, leaves, and bark of many vegetable species.
It is produced commercially by high pressure hydrogenation of glucose in the presence of
heterogeneous catalysts, e.g., nickel phosphate [36], or homogeneous catalysts, e.g., ruthe
nium dichlorotriphenyl phosphine [37].
D. D-Glucose
D-Glucose (dextrose) is the most abundant sugar in nature, either as free monosaccharide or
linked with other sugars (such as in sucrose) or as a polymer (as in starch or cellulose).
Commercially, D-glucose is produced by the hydrolysis of a variety of starches. The high
temperature acid-catalyzed process [38] has been replaced by the acid-enzyme or enzyme-
enzyme process, in which the initial starch hydrolysis is carried out with acid [39] or enzyme
(bacterial a-amylase [40,41]), respectively, whereas the subsequent hydrolysis is achieved
with an enzyme (fungal glucoamylase [42]). Commercial products include the crystalline or
monohydrate form as well as liquid solutions.
E. Sucrose
Sucrose (a-D-glucopyranosyl-^-D-fructofuranoside) occurs in practically every existing sper-
matophyte. It is extracted commercially from sugarcane, sugar beets, sorghum, and sugar
maple. Hydrolysis of sucrose under mild acidic conditions releases the two component sugars,
glucose and fructose. Hydrogenation with Raney nickel as catalyst produces a mixture of
sorbitol and mannitol. Sucrose is used widely as a sweetener and in food processing as a
preservative, bulking agent, flavor enhancer, and texturizer.
F. Propylene Oxide
Propylene oxide was also first prepared in Wurtz laboratory, by converting dibromopropane
to propylene glycol, subsequently converting it to 1 -chloro-2 -propanol followed by dehydro
chlorination with diluted potassium hydroxide. In the first industrial process (chlorohydrin
process), the chlorohydrin is prepared by the reaction of propylene and chlorine in excess
water followed by dehydrohalogenation. Because of the large volume of effluent produced in
the process, the direct oxidation of propylene with hydroperoxides in the presence of a catalyst
(oxirane process) is preferred [43-46].
The ring strain in propylene oxide is 29.57 kcal/mol [47], compared with 27.28 kcal/mol
for ethylene oxide [48], and their reactions and chemistry are similar. In addition to its use in
the manufacture of surfactants, propylene oxide is extensively used in polyether polyols, lubri
cants, and quenching fluids.
G. Fatty Acids
The fatty acids used in the manufacture of nonionic surface-active agents are virtually all of
natural origin, an even-numbered series containing 8-22 carbon atoms [49,50]. The principal
raw materials from which they are derived are tallow, coconut, palmkemel, soybean, rape-
seed, and sunflower oils, marine oils, and crude tall oil. Except for those derived from tall
oil, the acids are obtained from the triglycerides by hydrolysis with water or methanol:
CH2OCOR CH2OH
I I
CHOCOR ■f 3 H2O -►3 RCOOH + CHOH
I I
CH2OCOR CH2OH
CH2OCOR CH2OH
I
CHOCOR + 3 CH3OH 3 RCOOCH3 + CHOH
! I
CH2OCOR CH2OH
High temperature and high pressure hydrolysis can be performed without a catalyst or with
alkaline catalysts, e.g., lime or zinc and magnesium oxides. Hydrolysis at atmospheric pres
sure uses, preferentially, sulfuric or sulfonic acid catalysts.
Enzymes are used for hydrolytic processes operating at atmospheric pressure and ambient
temperature [51]. The fatty acids obtained from hydrolysis are then purified by distillation or
separated into their components by fractional distillation, pressing, or crystallization using
solvents or, as in the Henkel SOT process, water-surfactant solutions.
Methanolysis gives methyl esters that are separated by fractional distillation. The grades
suitable for detergent production are hydrogenated to alcohols; the others (usually the shorter
chain lengths) are hydrolyzed to acids. Tall oil fatty acids are obtained by fractional distilla
tion of crude tall oil.
The entire range of chain lengths, from Cg to C2 2 , linear or branched, saturated or unsatu
rated, and substituted fatty acids are the starting raw materials to produce POE esters, glyc
erol, sorbitan, and sucrose esters and their alkoxylated derivatives. Fatty acids of vegetable
origin are preferred by certain segments of the industry in response to the movement in public
opinion toward the protection of animals. The Cjg chains are most commonly used, because
the surfactants thus obtained generally have good detergency and emulsification properties.
Because of the ease of handling in its liquid form, oleic acid is preferred in many industrial
processes. The C 12- C 14 “cuts” give surfactants that are milder and therefore preferred in
applications involving extensive skin contact, e.g., hair and body shampoos, creams and lo
tions for skin care, and washing-up liquid detergents.
Tall oil fatty acids from the sulfate or Kraft paper process are sometimes used to produce
low cost esters. Isostearic acid, despite its high cost and limited availability, is used for
specialty surfactants requiring a combination of the properties from the Cjg chain length typi
cal of stearic acid with the liquidity conferred by oleic acid but without the disadvantage of
the unsaturation (which could lead to adverse effects, e.g., in polymerization reactions).
A building block that is most useful for the production of polymeric emulsifiers, described
later, is 12 -hydroxy stearic acid.
It is estimated that, in 1990, about 1.8 million metric tons of fatty acids (including tall oil
fatty acids, but excluding the fatty acid salts from continuous soapmaking processes) were
produced in the United States, Europe, and Japan, in the approximate proportions 1:1:0.3.
H. Fatty Alcohols
Fatty alcohols are obtained either from natural sources (i.e., by hydrogenolysis of fatty acids,
fats, or fatty esters) or synthetically by hydroformylation (0x0 process) or by the Ziegler
process. Hydrogenolysis is performed at high temperature (200-350°C) and high pressure
(100-200 bar) [52] if a copper chromite catalyst is used. This process yields saturated alco
646 Bognolo
hols. Unsaturated alcohols can be produced by using selective mixed oxide catalysts (AI2O 3/
Cd0 /Cr203)
RCOOH + 2H 2 -> RCH 2OH + H 2O
The hydroformylation process involves the addition of carbon monoxide and hydrogen to
an a-olefin in the presence of a catalyst to give an aldehyde that is then reduced to an alcohol
containing one carbon atom more than the starting a-olefin [53]:
RCH2CH2CHO — ^ .->RCH2CH2CH2QH
CO z ’
r CH=CH,
H2 \
RCH(CH3)CH0- ->RCH(CH3)CH20H
The alcohols are mixtures of linear and branched molecules, the composition being dependent
on the a-olefin feedstock [54,55], the reaction conditions, and the catalyst [56,57]. The
branching confers some degree of differentiation compared with the natural, or Ziegler, alco
hols in terms of reactivity and physical properties.
The Ziegler process allows the production of straight-chain synthetic alcohols by reacting
triethylaluminium and ethylene (58) (see Scheme 1). The alcohols synthesized by this route
have carbon chains with an even number of carbon atoms, and their chain length composition
follows a Poisson distribution. Branched chains are formed to a minor extent (below 5%)
owing to side reactions, so the alcohols are predominantly linear.
, CH2CH3 ,(CH2CH2)x CH2CH3
A i ------ CH2CH3 + n CH2 = CH2 • A l" (CH 2CH 2)y CH2CH3
CH2CH3 (CH 2 C H 2)z C H 2C H 3
i
+3/2 O 2
i
P (CH2CH2)x CH2CH3
A l- O (CH 2GH 2)y CH2CH3
0 (CH 2CH 2 )z CH2CH3
i
+3 HsO^
i
HO(CH2CH2) x CH2CH3
HO(CH 2C H 2)y CH2CH3
HO(CH2CH2) z CH2CH3
Scheme 1 A 1(0 H )3
A type of alcohol that is generating a growing interest is that in which a CH 2OH group is
substituted at, or near, the middle of a hydrocarbon chain. Such molecules were first prepared
by Guerbet in 1899 by heating shorter chain alcohols under alkaline conditions and are named
after him. A key feature of the Guerbet alcohols is the much lower melting point compared
with primary or secondary alcohols with a similar number of carbon atoms.
The oxidation of linear paraffins in the presence of boric acid yields secondary alcohols of
the same chain length. The process is used on a larger scale in the former Soviet Union (FSU)
and has gained significance in Japan. The oxidation is performed in the liquid phase at 150-
170°C with a mixture of 3-4% oxygen in nitrogen. The alcohols produced have the secondary
group distributed statistically along the whole carbon chain.
C„H2„^ 2+>/202 (C„H2„+,)0H
Secondary alcohols can also be produced by hydration of a-olefins from the thermal cracking
of petroleum waxes.
The total world capacity for detergent alcohols was in the region of 1.5 million tonnes in
1990, the “synthetics” being predominantly used in the United States and the “naturals” in
Asia, with Europe having a balance between the two. Since then the expansion has drifted
in favor of the natural materials, driven by environmental pressure, resource renewability
considerations, production incentives, and the essential interchangeability of the two types in
the large majority of detergents and other industrial applications. Economic considerations
have not been of major relevance, as the prices of natural versus synthetic alcohols are fluctu
ating. The prices depend on the balance of supply and demand of the key raw materials
(ethylene/a-olefins and oils and fats) more than the difference in the conversion costs for the
various production routes.
J. Fatty Amines
Fatty amines were first introduced commercially in the mid-1930s by the Armour Industrial
Chemical Company. They are produced by way of the hydrogenation of nitriles [59,60]:
NH,
R C O O H ------> R C 0 0 - N H 4 R C = N ------^RCHoNHo
Nitriles can be produced by using the reaction of fatty acids with anhydrous ammonia in the
liquid phase at about 300°C with catalysts such as alumina, titanium and zirconium alcohol-
ates, and zinc and manganese oxides. Continuous processes can use fatty acids or triglycerides
at temperatures between 300 and 400°C with aluminum or thorium oxide catalysts.
Nitrile reduction is performed at 1 0 0 -1 5 0 ° C [61,62]; higher temperatures produce higher
yields of secondary amines [63]. Catalysts of the Raney nickel type are preferred; metal
oxides can be used, but they yield mixtures of primary, secondary, and tertiary amines.
A-Alkyl-1,3-propanediamines are prepared by the addition of acrylonitrile to A-alkyla-
mines, followed by subsequent reduction of the resultant jS-cyanoethylalky lamine :
R N H 2 + C H 2= C H C N -----> R N H (C H 2) 2C N ------> R N H (C H 2) 3N H 2
K. Alkylphenols
Alkylphenols are produced industrially by catalytic alkylation of phenol with olefins [64]:
Diisobutene and di-^-butene yield octylphenols.
Propene trimer yields nonylphenol.
Propene tetramer and tri-n-butene yield dodecylphenol.
The preferred catalysts are ion-exchange resins, boron trifluoride, acid-activated clays, and
synthetic aluminosilicates. The reaction temperature and the phenol/olefin ratio are dependent
on the type of catalyst used and the end product specifications. They can be on the order of
648 Bognolo
50-85°C and 1.2:1 to 1.4:1, respectively, for the production of octylphenol with BF3 catalyst
[64]. Boron trifluoride catalyst has the advantage of having a high activity in relatively mild
conditions and of yielding essentially para-substituted monoalkyl derivatives.
Higher alkylphenols are mainly used as building blocks for technical surfactants through
alkoxylation processes. They are intermediates for the synthesis of antioxidants and light stabi
lizers. Dodecylphenols are the starting materials for a number of lubricating oil additives. It
is estimated that the worldwide production of higher alkylphenols is on the order of 550,000
tpa, with western and eastern Europe accounting for about 50%, the United States about 40%,
and Japan about 5%, the balance of 5% being supplied by the other regions.
IV. BASIC CHEMISTRY AND INDUSTRIAL PRODUCTION

The synthesis of nonionic surfactants is essentially based on well-known and long-exploited
chemistry and centers on two fundamental types of reactions:
1. Condensation (i.e., water elimination through esterification, etherification, or amide
formation)
2. Alkylene oxide addition
In the majority of cases only one type of reaction is involved. However, specialty surfactants
may require both types, e.g., an esterification followed by etherification or alkylene oxide ad
dition.
It follows that the industrial processes and equipment needed for production are compara
tively simple and inexpensive. The major issues originate from the human safety and handling
issues of ethylene oxide (and, to a lesser extent, propylene oxide), from the high temperatures
associated with some esterification and etherification reactions, and from the finishing steps
for the isolation, purification, and conditioning (if required) of the commercial products.
A. Esterification
The compounds described in this section are manufactured by way of esterification reactions.
1. G lycerol Esters
Glycerol esters can be produced by the direct esterification of fatty acids with glycerol, which
yields mixtures of mono-, di-, and triglycerides depending on the temperature and time of
reaction, the catalyst, and the relative proportions of the reagents [65]. The immiscibility of
glycerol and fatty acids is a major disadvantage that can be overcome only by operating at
high temperatures, typically in excess of 260°C.
The most important industrial method for the manufacturing of partial esters of fatty acids
is the glycerolysis of fats. This is carried out at 180-250°C in the presence of alkaline cata
lysts, e.g., sodium hydroxide, sodium methylate, sodium and potassium soaps, potassium
carbonate, trisodium phosphate, calcium oxide, or metallic zinc and aluminum. The resulting
products are mixed mono-, di-, and triglycerides. The solubility of glycerol in fats (22% at
180°C and 40% at 250°C) limits the extent of glycerolysis [66,67]. Products with more than
90% monoglyceride can be obtained by molecular distillation of glyceride mixtures [68,69].
However, since the reaction between triglycerides and glycerol is reversible, monoglycerides
can be subject to internal alcoholysis on heating or prolonged storage and disproportionate
into glycerol and di- and triglycerides, especially if residual catalyst is present [67,70].
Transesterification (the interchange of acyl radicals between glycerides in the presence of
appropriate catalysts) can be used to rearrange partial ester mixtures. By controlling the reac
tion conditions and the relative proportions of reagents, nearly pure mono- and diglycerides
can be obtained [71].
Partial glycerol esters are hydrophobic surfactants that are suitable essentially for w/o types
of emulsion. They can be converted to more hydrophilic species by the addition of ethylene
oxide.
2. P olyglycerol Esters
Heating glycerol in the presence of alkaline or acid catalysts at 2 5 0 -2 7 5 ° C results in conden
sation and the formation of polyglycerols [72]. Esterification of polyglycerol with fatty acids
can be carried out at 190-220°C using the same catalysts as for the etherification reaction
[73,74]. Polyglycerol esters can also be produced by the hydrolysis of triglycerides with
poly glycerol.
3. Tetritol and P entitol Esters

The synthesis of esters of erythritol and xylitol has been reported [75,76], but these products
are of limited commercial significance. Pentaerythritol reacts readily to form tetraesters
[75,77] that find use as lubricants and oil additives.
4, H exitol Esters
The direct esterification of hexitol alcohols with fatty acids under industrial manufacturing
conditions leads to the formation of internal ether linkages in the polyol moiety, so that the
commercial products are, in reality, anhydrohexitol esters. The anhydrization is favored by
heat and acid catalysts, e.g., sulfuric or p-toluene sulfonic or phosphoric acid [78], whereas
alkaline catalysts such as sodium or potassium hydroxide or the alkaline salts of fatty acids
promote the esterification reaction [79]. By an appropriate choice of catalyst and reaction
conditions, particularly temperature, it is possible to produce esters of significantly different
compositions, ranging from low polarity, heavily anhydrized diesters to the more hydrophilic,
nonanhydrized monoesters. These processes, however, are hardly suitable for large-scale
manufacture, which requires reasonably fast reaction kinetics, maximum yield, and product
consistency combined with a minimal need for purification and postreaction treatments,
e.g., to remove unreacted material or reduce color. Thus, statistically, the commercial prod
ucts are heterogeneous mixtures containing free polyols and fatty acids and various sorbitan
(1,4-, 1,5-,2,5-sorbitan) and isosorbide mono-, di-, and triesters. Figure 4 illustrates some of
the species present in a commercial sorbitan ester.
Hexitol esters can be prepared using the entire range of fatty acids [77-79], although the
esters with lauric, palmitic, stearic, and oleic acid are those of major commercial significance.
Sorbitol is the most common polyhydric alcohol; mannitol esters are equally available but are
confined to niche applications.
Sorbitan esters are versatile o/w emulsifiers but are also the building blocks for the more
hydrophilic ethoxylated species commonly referred to as polysorbates [80,81]. The products
most successful from the commercial point of view are obtained by the addition of 4 -5 mol
of ethylene oxide to one theoretical mole of sorbitan monoester, or of 20 mol of ethylene
oxide to one theoretical mole of mono- or triester.
Sorbitan esters and polysorbates have been in use for nearly half a century (they were
originally introduced in the mid-1940s by Atlas Powder Co. under the trademarks SPAN and
TWEEN, respectively) and have a remarkable record of safety in use, supported by extensive
toxicological data, to the point that they are often used in toxicological studies as the bench-
650 Bognolo
(A)
H^COH — HXOH
I V
HCOH
I
HO^H
HCOH '
HOCH
1 1
I
HOCH
HCOH
1
HCOH
'
HC
HO CH
1
HCOH
n Ç
HO
CH
HCOH
I 1 11 1
HCOH HCOH HC ----- HC -----i
1 1 1
H,ioH H3COH H3COH H fiO H
SO RBITO L 1-4 SORBITAN 1 -5 SORBITAN 2-5 SO RBITAN iSO SO R B ID E
(B)
C,7C„“ C00H
(C)
h,coocc,,H33
/ \ ?»
HOCH H.C'^ CHCHCH300CC,,H33 c,,H3,COO
HOCjîH
HCOH HOCH----------CH
I
HCOH
¿H
H,ioH
H3COH
HOCH
I H3C
/ \ ?»
ÇHCHCH 3 OH C,,H33C00
I
h o (| h
HC00CC,,H33 HOCH----------CH OH
HCOH ¿0CC„H,3
hL :OH
SORBITOL ESTERS SORBITAN ESTERS ISOSORBIDE ESTERS
Fig. 4 Most common species present in a commercial sorbitan ester. (A) Free polyols; (B) free fatty
acids; (C) sorbitol/sorbitan/isosorbide esters.
in
HOC
HOCH
I
I
HCOH
I
HCOH
H,ic
HOCH
HOCH
I
I
HCOH
h2 iOH
H,COH
H,c/ \ T 300CC,,H33
Hocin
I CHCHCH
HCOH
I
HCOH
I HOCH-------CH
HC00CC„H33 ¿0CC,,H33
hI öh
H3COH
HOCH
I
I
HCOH
I
HCOH
I
HCOH
I
H3C00CC,,H33
SORBITAN E ST ER S
Fig. 4 Continued
H,COH
I
C,,H,,COOCH
I
HOCH
I
HCOH
I
HCOH
H,COH
I
HOCW
I
C,,H,,COOCH
I
HCOW
I
WCOW
I
H,COW
H,COOCC,,H,,
I
MOqH
I
HOCH
I
HCOOCC,,H,,
I
HCQH
I
H,GOOCC,,H,,
SOWBITOL ESTERS SORBITAN ESTERS
Fig. 4 Continued
mark against which other surfactants are tested. They are reported in many pharmacopoeias
[82-84] and are included in the positive lists of many regulatory bodies, e.g., FDA [85],
EGA [86], and EU directives [87], as additives for polymers, adhesives, and paper products
that come into contact with food and as direct additives for human food and animal feed.
Sorbitan esters and polysorbates also form a homogeneous series of molecules of different
polarity and, as such, are particularly easy to formulate by applying, for example, the princi
ples of the HLB (hydrophile-lipophile balance system) concept described in Section V. F.
Because of the safety characteristics, the extensive regulatory approval, the good emulsifica
tion and dispersion properties, and the ease of formulation, these products are used in an
amazing variety of applications, ranging from cosmetic creams and lotions through emulsion
and suspension polymerization, pigment dispersion, industrial fluids, and food, to agricultural
formulations, to name just a few.
A special category of hexitol esters is that in which the polyol is ethoxylated first and the
resulting product is further esterified with fatty acids. The products with the best performance
contain 6 -7 mol of Cjg fatty acids per theoretical mole of ethoxylated hexitol. Because of
their structure, these molecules can function as steric stabilizers and are particularly useful in
the formulation of crop protection chemicals and metalworking fluids.
5. Sugar E sters
The economic manufacture of fatty acid esters using mono- and oligosaccharides has been the
subject of considerable research. This was— and still is—motivated by the drive to produce
high HLB nonionic surfactants without using ethylene oxide. This is a target that is particu
larly attractive to the personal care and food/feed industries. It has already been shown that
sorbitol esters are only a theoretical option, whereas polyglycerol esters are both expensive
and, comparatively speaking, still too hydrophobic.
The difficulties associated with the industrial manufacture of sugar esters are the insolubil
ity of the saccharides in the fatty acids and their derivatives, the charring of the sugar at the
temperatures of esterification, the poor conversion rates, and the need for extensive purifica
tion of the finished product. Direct transesterification of sucrose with triglycerides, without
further purification, yields complex mixtures that have found limited use as emulsifiers for
animal feed, but they are too ill-defined and contain too many impurities to be suitable for
other applications.
Manufacture of products with a broader range of applications involves the use of solvents
such as dimethylformamide [88], dimethyl sulfoxide [89], or hydrocarbons [90] for the reac
tion or the subsequent purification. This adds significantly to the manufacturing costs and
detracts from the “naturaFV^green” image that these products could otherwise have. Recently,
esters of a-m eth y l glucoside produced through enzymatic esterification have been proposed,
and it is possible that, in the long term, enzymes will provide an economic route to the
manufacture of other saccharide esters.
So far, apart from the sucroglycerides for animal feed, the few sucroesters commercially
available have found only those niche applications that can tolerate their high cost or are used
in specific foods in which the use of ethylene oxide derivatives is not allowed from a regula
tory point of view. They continue, however, to receive considerable attention from the formu
lation chemists because of their “natural” connotation.
The ethoxylation of sucrose esters has been reported [91], but the products do not seem to
have generated a commercial response. The derivatives obtained by propoxylation of sucrose,
followed by esterification or transesterification, have some limited use as antifoam agents in
the food and fermentation industries.
654 Bognolo
6. Polyoxyalkylene Esters o f Fatty Acids

Polyoxyalkylene esters of fatty acids can be prepared by a variety of methods [92]. However,
only the direct reaction of a fatty acid with an alkylene oxide olefin (usually ethylene oxide)
and the esterification of a fatty acid with polyalkylene glycols are of industrial relevance.
Ethylene oxide reacts with fatty acids to give monoesters that, during the reaction, undergo
a further transesterification, so that the final product is a mixture of polyoxyethylene glycol
monoester, diester, and free polyoxyethylene glycol [93].
RC00H + ^ z(CH2CH20) ------

> RC00(CH2CH20)„H
2RC00(CH2CH20)„H-----> RC00(CH2CH20)„0CR + H0(CH2CH20)„H
Similar mixtures are obtained by the direct esterification of fatty acids with polyoxyethylene
glycols [94] if a 1:1 molar ratio of the reagents is used. A higher ratio of fatty acids favors
the formation of diesters, whereas more monoester is produced with an excess of polyoxyethy
lene glycol. The esterification reaction is performed at temperatures of 100-250°C, with or
without catalyst. The use of a catalyst (usually sulfuric or organic sulfonic acids) reduces the
reaction temperature and shortens the reaction time. The reaction is reversible and is therefore
never complete. In industrial processes the equilibrium is shifted either by removing the water
of reaction as an azeotrope or by vacuum stripping or inert gas sparging.
B. Etherification
Apart from the production of ethers by ethylene oxide addition (discussed in detail later), the
most important reactions of etherification by far are those leading to alkyl polyglycosides, in
particular the glucose-derived products. Chemical glycosylation processes include acid-cata
lyzed reactions leading to complex mixtures of oligomers [95,96] and kinetically controlled,
irreversible, mostly stereospecific substitution reactions with suitably activated carbohydrate
substrates [97-100]. Enzymatic synthesis of alkyl glycosides [101] is not yet suitable for
commercial scale application, and the industrial processes operated today are based on the
Fisher process through different technical adaptations, depending on the molecular weight of
the alcohol and the dextrose/alcohol ratio used to obtain the desired degree of polymerization.
In addition, depending on the physical form and purity of the raw material, a direct acetaliza-
tion or transacetalization may prove economically more attractive. These routes are illustrated
schematically in Fig. 5.
The direct glucose conversion with fatty alcohols is a two-phase process, a disadvantage
that can be minimized by using catalysts that do not accumulate in the polar phase, e.g.,
alkylsulfonic acids [102]. The molar ratio of glucose to fatty alcohol is typically 1:5, and the
excess alcohol is removed (e.g., by using thin-film evaporators), purified, and recycled. Fol
lowing dilution with water (fatty alcohol polyglucosides are highly hygroscopic and cannot be
handled as such), a final purification may be required to remove colored species, by-products,
or unreacted carbohydrate. During the synthesis, care must be exerted to prevent overreaction,
which leads to excess polyglucosides with no surface activity, as well as underreaction, which
leads to poor yield.
Industrial fatty alcohol polyglucosides are complex mixtures of isomeric mono-, di-, and
triglucosides, with each higher oligomer formed in decreasing amounts and with an average
degree of polymerization ranging from 1.3 to 1.8 units. The fatty alcohols more commonly
used have chains of 8 -10 carbon atoms if the alkylpolyglucoside is intended for wetting
applications, 12-14 for cleaning/detergency, and 16-18 for emulsification. (In the latter case.
(CH^3—CH3
,CCH3)„-ch3
Fig. 5 Schematic route for the synthesis of alkylpolyglucosides.
h o w e v e r, larg e q u a n titie s o f u n re a c te d a lc o h o l are p re se n t b e c a u se o f th e im p o ssib ility o f

achieving its complete removal.)
The Williamson ether synthesis [103] is used to block the terminal OH of ethoxylated
nonionics with alkyl groups such as methyl, benzyl, or allyl. Alkyl-terminated nonionics not
only have reduced foaming and reduced water solubility but also are essential in the formula
tion of hydrolytically unstable actives, e.g., diketenes or alkenyl succinic anhydrides in paper
sizing or crop protection chemicals, that would otherwise be attacked by the free hydroxyl
groups.
Etherification reactions are also involved in the synthesis of small-volume specialty surfac
tants, e.g., polyglycerol ethers, that have a higher chemical stability than the corresponding
esters. Synthetic processes based on alcoholates, epoxides, or chloroparaffin might be used.
C. Amide Formation
There are several routes for the synthesis of fatty acid amides [104], including reaction of
fatty acid chlorides and anhydrides with ammonia, hydrolysis of nitriles, and aminolysis of
glycerol esters. Industrially, amides are prepared by reacting fatty acids and ammonia, with
the subsequent dehydration of the ammonium salt [60,105],
RCOOH 4-NH3 RC0 0 'NH4'" -► RCONH2 + H 2O
Alkanoyl-A-methyl glucamides are a new class of nonionic surfactants that appear to be

particularly suitable for the formulation of liquid dishwashing detergents because of their
favorable performance and ecotoxicological properties. The synthesis involves reacting a
tryglyceride [Eq. (1)] or a fatty acid ester of a lower alcohol [Eq. (2)] with A-methyl gluca-
mme:
656 Bognolo
CH2OOCR
I
CHOOCR + 2 CH3NHCH2(CH0H)4CH20H '
I
CH2OOCR
CH2OH
I
2 RC0N(CH3)CH2(CH0H)4CH20H + CHOH
I
CH2OOCR ( 1)
RCOOCH3 + C H 3 N H C H 2(C H 0H )4C H 20H
R C 0 N (C H 3)C H 2(C H 0 H ) 4C H 20H 4- C H 3O H ( 2)

The reaction temperature is kept within the range of 130-150°C, and there is no need for a
catalyst. For practical applications, coconut fatty acids are preferred.
De Alkylene Oxide Addition

The general reaction of alkylene oxide addition to compounds containing active hydrogen can
be described as
RAH + /2 C2H4O -> RA(C 2H40)„H
The reaction can be viewed as a sequence of discrete steps that each add one alkylene oxide
unit. This leads to a number of species with different alkylene oxide content, whose distribu
tion is determined by the nature of the hydrophobe, the reaction rate, the catalyst, the equip
ment, and the reaction conditions:
RAH + /1 C2H4O -> RA(C 2H40)H + - 1 )C 2H40 ■
RA(C 2H40)2H + (/z- 2)C 2H40
RA(C 2H40),-H + (/z- z)C 2H40 - RA(C 2 H4 0 ),-+1 H + (zz- z- 1 )C2 H4 0
RA(C2H40)„_iH-hC2H4 0 ---- > RA(C2H40)^H

The ring-opening reactions of ethylene oxide are nucleophilic substitutions [106,107]. With
basic catalysts, there is first an attack of the nucleophilic at a ring carbon atom:
RA~ + C,H 40 RAC2H40
which is then followed by proton exchange or further reaction with another alkylene oxide
molecule as seen above:
RAH + RAC2H4O R A - + RAC.H4OH
or
RAC 2H40 " + C2H40 - RAC 2H4OC 2H4O"
The rate-determining step is the first addition, which is dependent on both the concentration
of the nucleophile and that of the alkylene oxide; the substitution is therefore of the SN2 type.
Acid catalysts can also be used but are less favored because of the formation of larger
quantities of by-products such as polyoxyethylene glycols, dioxane, and 2 -methyIdioxolane.
In the acid-catalyzed reaction, the slow stage is the formation of a carbocation following
protonation and ring opening of the oxirane ring:
C2H40 -hH-^ C2H 4OH
C2H4OH C H 0C H 2 OH
R A H + C H .C H .O H - R A C H o C H .O H + H''
In ideal conditions the reaction can be considered to be of the SNl type. It has been suggested
that in practice the reaction is intermediate between an SNl and SN2 [107].
1. Polyoxyethylene Alkylphenols
Nonylphenol and ethylene oxide are, respectively, the most common hydrophobe and hydro
phile used. A typical industrial manufacturing process involves oxyethylation in an autoclave
at temperatures within the range 140-200°C, ethylene oxide pressure of 2-5 bar, with 0 .1 -
0.5% of sodium or potassium hydroxide as catalyst. Other basic catalysts such as sodium
acetate, potassium carbonate, and sodium methylate have also been described [108,109]. The
products are statistical molecules, the composition of which approximates to a Poisson distri
bution with basic catalysts and a relatively low degree of ethoxylation (usually up to 10-15
mol of EO per mole of hydrophobe initiator [110]). Different distributions are possible, but
the essential feature of a spread of molecules with different degrees of ethoxylation around a
nominal average remains.
Octylphenol has a similar reaction pattern, but despite the nominal difference of only one
carbon atom in the alkyl chains, the interfacial properties of its alkoxylated adducts differ
markedly from those of the nonylphenol homologues owing to the structure of the alkyl chain.
As such, octylphenol adducts are used in specialty niche applications, e.g., as emulsifiers in
crop protection formulations. Dodecylphenol is a starting material for the synthesis of lubricat
ing oil additives and antioxidants.
The polarity of the alkylbenzene ring confers to alkylphenols unmatched interfacial proper
ties, but the limited industrial availability of hydrophobes restricts the range of possible HLB
modifications. When a more pronounced hydrophobic character is needed, the water-insoluble
moiety is extended by the addition of propylene oxide.
2. Polyoxyalkylene F atty Alcohols

The oxyethylation of an alcohol was first reported in 1926, but it was the work of Wittmer
and Schoeller in the early 1930s that produced the first derivatives of fatty alcohols. Industrial
processes in current use favor base-catalyzed reactions with sodium or potassium hydroxide
or sodium alcoholates at temperatures ranging from 120 to 200°C and pressures of 2 -8 bar
[109,111]. Mixing by impeller or loop circulation [112] must ensure adequate heat transfer
throughout the reaction (which is highly exothermic) as well as maintaining intimate contact
between the liquid phase and the ethylene oxide vapor phase. Primary alcohols give faster
reaction rates than secondary or tertiary alcohols, while the rate of ethoxylation decreases as
the carbon chain length of the alcohol increases. After the initial ethylene oxide addition, the
differences in the rate of reaction diminish as the environment around the reacting center
becomes similar following the growth of the polyoxyethylene oxide chains.
The product from the reaction is a mixture of alcohol ethers, with a distribution that is
much wider for the secondary and tertiary alcohols. This distribution depends on the process
parameters, in particular the catalyst and the mixing system used [113,114]. On completion
658 Bùgnolo
of the reaction, the unreacted ethylene oxide is removed by vacuum stripping, and the catalyst
is neutralized with organic or inorganic acid [115,116] and removed by filtration. The product
may also be bleached to reduce color.
The propoxylation of fatty alcohols follows, in principle, the same procedure as ethoxyla-
tion [109].
3. Polyoxyethylene Alkylam ines

The preparation of surface-active amines by the ethoxylation of fatty amines was first de
scribed by Schoeller and Wittwer. A large variety of addition products from primary, second
ary, and tertiary amines as well as polyamines have been reported, although only the deriva
tives of the primary amines and of the A-alkyl-1,3-propane diamine are commercially
important. In the reaction of ethylene oxide with primary amines, both reactive hydrogens are
substituted prior to any further ethoxylation [ 110 ]:
C2H4OH
I
R -N H 2 + 2(C2H40) «► R -N
I
C2H4OH
C2H4OH (C2 H4 0 )xH

I I
R -N + n (C2H4O) •R"»N
I I
C2H4OH (C2H40)yH
The reaction is initiated through ethylene oxide addition at about 100°C and is furthered by
raising the temperature to 150°C or higher in the presence of base catalysts.
4. P olyoxyethylene Alkylam ides

The polyoxyethylene alkylamides include the mono- and diethanolamide derivatives, which
are the most important commercial products, together with the polyoxyethylene adducts.
Monoethanolamides can be prepared from fatty acids [117], from fatty esters [118], and
by transamidification [119]. Diethanolamides from fatty acids have been prepared by reacting
1 mol of fatty acid with 2 mol of diethanolamine at 150-170°C [120] or by heating a fatty
acid methyl ester with diethanolamine at about 115°C in the presence of sodium methoxide
catalyst and distilling off the methanol released [121]. The reactions are schematically
RCOOH + NH.CH.CHoOH-----> RCONHCHXH.OH + H ,0
and
R C 00H + NH(CH2CH20H)2-----> RC0N(CH2CH20H)2 + H20
In the reaction of fatty acids with alkanolamides, ester amides are formed, which usually
detracts from the foaming properties of the products. The formation of ester amides can be
minimized by balancing the reaction parameters, in particular temperature and reaction time,
and by using alkali metal catalysts [ 122 ].
Polyoxyethylene amides are prepared by reaction of amide or of a preformed alkanolamide
with ethylene oxide in the presence of basic catalysts:
RCONH2 + n(C 2H40) - > RC 0 NH(C 2H 40)„H

R C O N H C H 2C H 2O H + n (C 2H 40) RCONH(C 2H40)„+i H
RC0N(CH2CH20H)2 + n(C2H40) RC0N(C2H40)„+2H
5. P olyalkylene Oxide Copolym ers

The polyalkylene oxide copolymers probably constitute the richest class of nonionic com
pounds. Several parameters—apart from the usual ones of catalyst type and reaction condi
tions— can be varied to achieve an amazing range of products and effects. In particular, the
following variables can be taken into consideration.
1. Type of initiator. This can be mono- or polyfunctional (i.e., it may contain from one
to several active hydrogen atoms). There can be further diversification depending on
the nature of the carbon chain (a monofunctional alcohol initiator may have carbon
chains containing from 1 to 18 atoms or more, a tetrafunctional diamine may have
from 2 to 6 carbon atoms in the central bridge) and of the atoms or groups carrying the
active hydrogen (usually oxygen or nitrogen, but possibly sulfur, amido, or carboxy).
2. Sequence of addition of alkylene oxide to the initiator, in particular
a. A sequence of ethylene oxide units followed by one of propylene oxide units or
by a mixture of ethylene oxide and propylene oxide.
b. A sequence of propylene oxide units followed by one of ethylene oxide units or
by a mixture of ethylene oxide and propylene oxide.
c. A mixture of ethylene oxide and propylene oxide.
3. The total molecular weight of the adduct as well as the molecular weight of each
specific sequence and, within a mixed sequence of ethylene oxide/propylene oxide,
their relative proportions.
4. The type of alkylene oxide having the hydrophobe function (i.e., in a number of cases,
butylene oxide has replaced propylene oxide units).
Polyalkylene oxide copolymers are peculiar in that, with a few exceptions, the amphipathic
character of the molecule is attributable more to the balance of alkylene oxides in the molecule
than to the type of initiator. It is impossible in this chapter to give a detailed discussion of
the chemistry and manufacture of such a broad range of products. The most relevant commer
cially, apart from the glycerol, sorbitol, and sucrose polyols used as building blocks in poly
urethane polymers, include derivatives from
1. Monofunctional alcohols. Alcohols of various hydrocarbon chain lengths are reacted
in sequence with propylene oxide, followed by ethylene oxide addition [123]. Sodium
hydroxide can be used as catalyst, the temperature for the propoxylation stage being
about 100°C and for the ethoxylation stage about 140°C. The order of addition of the
alkylene oxides can be reversed, yielding products with significantly different proper
ties [124]. Tergitol XD, XH, and XJ from Union Carbide are produced starting from
a Cj-Cg alcohol, which is reacted first with a mixture of propylene oxide and ethylene
oxide and subsequently with pure ethylene oxide [125].
2. Bifunctional alcohols. The Pluronic products from BASF are made by the sequential
addition of propylene oxide and ethylene oxide to a propylene glycol initiator [126]
using sodium hydroxide as catalyst and a reaction temperature of about 120°C . Block
copolymers prepared by sequential condensation of butylene oxide and ethylene oxide
have been reported [127].
3. Tetrafunctional poly amines. The Tetronic products from BASF have the structure
660 Bognolo
H(C2H40)x (C3H60)y ^ ( C 3H60)y(C2H40)xH

NCH2CH2NV
H(C2H40)x (C3H60)y '(C 3H60)y(C 2H40)xH
with each polyoxypropylene group containing about 4-110 propylene oxide units and
ethylene ©side making up 20-90% of the total weight of the product.
E. Amine Oxides
Amine oxides are produced from the reaction of hydrogen peroxide on tertiary amines
[128,129]:
:H3
Î
RN - H2O2 O -f H2O
I I
CH 3 CH3
The tertiary amine (usually dimethyidodecylamine) is added to aqueous hydrogen peroxide
with vigorous agitation, keeping the temperature in the region of 6 0 -6 5°C . Initially, water is
added in small amounts to prevent gel formation and to finally reach an active concentration
of 3 0 -3 5 % while the temperature is raised to 70-75°C. Agitation is continued for another 2 -
3 h, and the excess peroxide is removed by addition of sodium sulfite.
F. Miscellaneous Nonionic Surfactants

Various nonionic compounds have been described, including sulfoxides [130], phosphine and
arsine oxides [131,132], polyoxyethylene phosphonates and phosphonites [133,134], and acet
ylenic glycols and their ethoxylates. Of these, only the acetylenic derivatives, e.g., Surfynols
from Air Products, have achieved commercial relevance as wetters, defoamers, pigment dis
persants, and hydrotropes.
V. PHYSICOCHEMICAL PROPERTIES OF
NONIONIC SURFACTANTS
The behavior of nonionic surfactants is determined by
1. The chemical class. This is at the root of many bulk and surface properties, e.g.,
physical form, surface and interfacial adsorption, micellar structure, and physical and
chemical stability.
2. The distribution of molecular species. All the commercial products are statistical
compounds, either polydispersed or heterodispersed [135]. In a polydispersed material,
the variation of structure and properties of the components is gradual, and clusters
form around a mean, as for example in ethoxylated alcohols or nonylphenols, alkoxy-
lated amines or amides, and PEO/POP copolymers. In a heterodispersed material the
variation is discontinuous and the individual constituent molecules have significantly
different structures and properties as, for example, in the case of the sorbitan or glyc
erol esters.
These parameters impart to the different species distinct physicochemical properties that are
the key to their performance in practical applications:
Critical micelle concentration
Foaming
Solubilization
Surface and interfacial adsorption
Wetting
Emulsification
Dispersion
Detergency
A. Critical Micelle Concentration

The critical micelle concentration (CMC) is the concentration at which the surface adsorption
is complete to form an interfacial monomolecular layer. The surfactant then ceases to exist as
an individual molecule in the bulk of the solution and begins to aggregate in clusters (mi
celles) of various shapes. The physicochemical properties of aqueous solutions of surfactants
change abruptly at the CMC, and it is their value at concentrations higher than the CMC that
determines the behavior and performance of a given surfactant.
Given the large variety of chemical classes and their significantly different properties, it is
not possible to give more than broad generalizations here. Ethoxylated nonionic surfactants
have been more widely investigated for the following reasons:
1. They are of major commercial interest.

2. They can be synthesized as well-characterized and pure molecules, using William
son’s condensation.
3. The polydispersed nature of the commercial products allows their use (instead of the
pure products) for accurate and representative physicochemical studies.
Nonionic ethoxylates have CMC values that are significantly lower than those of ionic
materials of comparable chain length and surface activity (see Table 4). In general, the CMC
of nonionic surfactants
1. Decreases as the number of carbon atoms in the hydrophobic chain increases

2. Decreases with the temperature
3. Is hardly affected by electrolytes (except for hydrogen ions at very low pH)
4. Decreases for ethoxylated products as the number of ethylene oxide units decreases
Becher [136] produced an extensive review of the values of the CMC in aqueous solutions
of homogeneous and heterogeneous ethylene oxide derivatives and of homogeneous surfac-
Table 4 Typical CMC Values of Ionic and Nonionic Surfactants
Temp. CMC Percent

(°C) (mol/L) (w/w)
Anionic
C,2H25SO-4Na+ 40 8 .6 x 10-3 0 .2
Ci2H25C6H4SOlNa+ 60 1 .2 x 1 0 -3 0.04
Cationic
C.2H25N+ (CH 3 ) 3 Br- 25 2 .0 x 1 0 -2 0.46
Amphoteric
C 1 2 H2 5 N+H 2 CH2 C0 0 - 27 1 .3 x 1 0 -3 0.033
Nonionic
C,2H25 0 (CH2 CH2 0 )4 H 25 4 .0x10-3 0.0014
C . 2 H2 5 0 (CH 2 CH2 0 ) 2 H 25 5.0 X 10-3 0.0025
662 Bognolo
tants with a hydrophile other than ethylene oxide. A more recent review, including CMC
values in nonaqueus media, has been reported by Rosen [137].
B. Foaming
In aqueous media, nonionic surfactants generally produce less, and less stable, foam than
ionic surfactants. For a given hydrocarbon chain length, the foam volume and stability reach
a maximum at a particular POE chain length and then decrease.
In many industrial processes it is useful to have surfactants that exhibit surface or interfa
cial activity without producing foam. This can be achieved by structural changes that
1. Affect the diffusion characteristics, thus destroying the elasticity of the surface film.
For this effect, large, straight-chain hydrophobes can be substituted with branched
isomers, or the hydrophilic group can be positioned centrally in the hydrophobe chain.
2. Increase the area per molecule, thus forming loosely packed films. Suitable modifica
tions involve putting two hydrophilic or hydrophobic groups onto the same molecule,
using highly branched or unsaturated hydrophobes, or attaching two bulky hy
drophobes on the same carbon atom. Of these the “capping” of terminal OH groups in
polyoxyethylene nonionics with alkyl groups or propylene oxide is the most effective
and preferred.
The structures of very low foaming nonionic surfactants are given in Table 5.
Table 5 Structures of Low Foaming Nonionic Surfactants
CH, CH, CH, CH,

x+y<4
I I
CH3CH CH2C— C = C — CCH 2CHCH3
HOiCjHÔ), (0 C2H4)3,0 H
(0 C2H4),0 H
/
RCH R < C i,; x = y:S5
(0C2H4)3,0H
(0 C2H4),0 H
/
RN R = Cio; x = y < 3
\
(0 X2H4)^0 H
HO(C2H40UCH2)n(OC2H4),OH x + y < 12
H0(C2H40),(CH2CH2CH2CH20)/C2H40),H y <27; x + z<82
I
CHj
H0(C2H40),(CHCH20)/C2H40),H y = 35; x + z = 45
I
CH3
C8H,2(0CHCH2),(0C2H4)3,0 H x = y = lO
Source: Ref. 137.

C. Solubilization
Solubilization is the spontaneous incorporation of a substance by reversible interaction with
the micelles of a surfactant in a solvent, to form a thermodynamically stable isotropic solu
tion. Solubilization into aqueous media is of major importance for processes such as cleaning
and soil removal, micellar catalysis of organic reactions, emulsion polymerization, and en
hanced oil recovery. Solubilization in nonaqueous media is the basis of dry-cleaning pro
cesses.
The solubilization by nonionic surfactants of polar molecules containing large hydrophobes
and small hydrophiles, e.g., alcohols or amines, is believed to occur through comicellization
with the surfactant’s molecule (Fig. 6). For less linear molecules, e.g., aromatic and substi
tuted aromatic hydrocarbons or branched esters, the solubilization may occur through inclu
sion in the interior of the micelles or through adsorption into their outer hydrophilic layer,
depending on the polarity of the solubilisate (Fig. 7). In the case of polyoxyethylene nonionics
it has been suggested that polar molecules may be solubilized into the outer region of the
polyoxyethylene tails (Fig. 8). Because of the low critical micelle concentration, nonionic
surfactants are better solubilizing agents than cationics and anionics of the same hydrophobic
chain length.
The solubilization power seemingly attainable from the surfactant structure is influenced
by the nature of the solubilisate, which determines the solubilization mechanism. In aqueous
solutions, the solubility of aliphatic hydrocarbons increases with the chain length of the hy
drophobic moiety and with a decrease in the number of ethylene oxide units. This reflects the
increase in the aggregation number of the micelles produced by these changes and the proba
ble dissolved state of the solute in the micellar core [138]. Ethylbenzene, which is more polar
and probably solubilized both in the ethylene oxide chains and in the inner micellar core,
shows an increase in solubility with an increase in the POE chain length [139].
Solubilization also depends on temperature and cloud point. In POE nonionic surfactants,
the micellar weight increases and the CMC decreases with an increase in temperature [140-
148]. Thus nonpolar materials that are solubilized in the inner core of the micelles show
increased solubility as the temperature rises. However, as the increase in temperature causes
Fig, 6 Solubilization of polar molecules.

664 Bognolo
I SURFACTANT
SOLUBILISATE
Fig. 7 Solubilization of low polarity molecules.
a partial dehydration of the ether oxygen atoms in the POE groups together with their tighter
coiling, the solubilization of more polar species that may occur through surface adsorption or
inclusion in the interfacial micellar layer is decreased.
D. Surface and Interfacial Adsorption

The key feature of surface-active agents is their adsorption at interfaces in an oriented fashion.
The surface concentration and molecule orientation are critical in determining the performance
of a surfactant with respect to many properties such as wetting, dispersion, emulsification,
foaming, and detergency.
Adsorption may occur at liquid/gas, liquid/liquid, and liquid/solid interfaces. The adsorp
tion at liquid/gas and liquid/liquid interfaces can be described by the Gibbs adsorption equa
tion [149],
I SURFACTANT
• SOLUBILISATE
Fig. 8 Solubilization of polar molecules by polyoxyethylene nonionics.

d y= - Y jid iX i
i
where d y is the change in surface or interfacial tension of the solvent, the surface excess
concentration of the zth component of the system, and the change in chemical potential
of the zth component of the system. The surface excess concentration is defined as the excess
per unit area of interface of any component over that present in a reference system of the
same volume in which the bulk concentration in the two phases remains uniform up to a
hypothetical dividing surface.
For dilute solutions of nondissociated surfactants, the above equation can be simplified to
d y = -R T T d ln C
where C is the molar concentration of the surfactant in the bulk of the solution. For surface-
active agents, the surface excess concentration can be equated to the actual surface concentra
tion; hence for solutions of nonionic surfactants containing a single species.
1 dy
r= —
RT d In C
A typical plot of F versus In C is given in Fig. 9. At low surfactant concentration, any

addition of surfactant results in a decrease in y, i.e., the surface is becoming progressively
saturated with surface-active molecules. At high surfactant concentration, y is practically con
stant, indicating that the surface is fully saturated. The break point at which the slope of the
curve significantly changes corresponds to the critical micelle concentration (see Section
V.A).
The surface concentration at surface saturation is a measure of the maximum value that
saturation can attain and is consequently a useful measure of the effectiveness of adsorption.
Table 6 gives the values for F and surface area per molecule for some common nonionic
surfactants in comparison with ionic, cationic, and amphoteric reference compounds. Addi
tional values are reported by Rosen [150].
The adsorption at solid/liquid interfaces can be calculated from the equation [158]
zzq A xj
m
where is the total number of moles in solution before adsorption; Axj is the
mole fraction of component 1 before adsorption; Xj, X2 are the mole fractions of components
1 and 2 at adsorption equilibrium; m is the mass of adsorbent in grams; and rf^ and zz^2 are
Fig. 9 Schematic representation of the variation of y in function of surfactant concentration.

666 Bognolo
Table 6 Surface Excess Concentration (F) and Surface Area per Molecule for Some
Common Nonionic Surfactants in Comparison with Anionic, Cationic, and Amphoteric
References
Temp. (mol/cm^ a^m

Surfactant (°C) X lO'O) (A") Ref.
CgHivOCH (CHOH)^ 25 4.0 41 151
C i2H25(OC2H4)40H 25 3.6 46 152
C i2H25(OC2H4)60H 20 3.6 45 153
C i 2 H 2 5 (O C 2 H 4 )sO H 25 2.5 66 152
«-C i(,H33(OC2H4)60H 25 4.4 38 154
/1-Ci6H33(OC2H4)90H 25 3.1 53 154
C8H12OCH2CH2OH 25 5.2 32 151
C12H25N+ (CH3)2CH2C0 0 - 23 3.5 46 155
C,2H25N+ (CH3)3CH3S0-4 25 2.7 61 156
Ci2H25SO^Na-" 25 2.9 57 157
• D-Octylglucoside.
Source: Ref. 137.
the numbers of moles of components 1 and 2 adsorbed per gram of adsorbent at adsorption
equilibrium.
In the case of dilute solutions of surfactants the above equation can be simplified to
ACi y
n\ =
m
where AC i = C i ^q—C i , C i q is the molar concentration of component 1 in the liquid phase

before adsorption, Ci is the molar concentration in the liquid phase of component 1 at adsorp
tion equilibrium, and V is the volume (in liters) of the liquid phase.
The surface concentration (C^j, mol/cm^) can be calculated from the surface area per unit
mass of solid adsorbent (a^, cm^/g):
ACi V
m
and the surface area per adsorbate molecule (in A^) is

10^6
NC\
where N is Avogadro’s number.
An adsorption pattern commonly observed in dilute solutions of surfactants is the Langmuir
type isotherm [159], which can be expressed as
Cl
Ci~\- a
where is the surface concentration of the surfactant (mol/m^) at monolayer adsorption, Ci
is the concentration (mol/L) of the surfactant in the liquid phase at adsorption equilibrium,
and a is a constant related to the absolute temperature and the free energy of adsorption at
infinite dilution.
Nonionic Surfactants 6S7
There are relatively few data on the adsorption of nonionic surfactants on different sub
strates. A review has been given by Von Rybinski and Schwuger [160]. Polyoxyethylene
(POE) derivatives adsorb on silica surfaces following Langmuir type isotherms, possibly
through hydrogen bonding between the SiOH groups on the surface and the oxygen atoms of
the POE chains [161]. The interest in coal-water slurries in the mid 1980s prompted studies
on the adsorption of alkoxylated alkylphenols on coal substrates [162]. The adsorption values
decrease as the size of the POE chain increases, and the most likely orientation of the surfac
tant is with the hydrophobic chain adsorbing on the hydrophobic regions on the coal surface,
leaving the POE chains to extend into the water phase.
E. Wetting
In a broad sense, wetting is the displacement from a surface of one fluid by another, although
in practice it tends to be related to the displacement of air by water or aqueous solutions.
Wetting is the basis of many natural and industrial processes. It is often the first necessary
step for the further development of interfacial effects. The cleaning of oily, or particulate soil
from garments, the dispersion of a pigment in an aqueous or a nonaqueous medium, the
spreading of a crop protection formulation over the leaves of plants, and the transport of
material across biological membranes are all dependent on wetting phenomena.
Wetting can be of three types [163]:
1. Spreading wetting: A liquid in contact with a substrate spreads over it and displaces
another fluid, e.g., air. The work of spreading is then
w ^s= (rsL + yL v)-ysv

where yî^ and jj^y are the interfacial tensions between substrate and vapor, sub
strate and liquid, and liquid and vapor, respectively. W§ is also referred to as the
spreading coefficient. If it is negative, spreading can occur spontaneously.
2. Adhesional wetting: A liquid not originally in contact with a substrate makes contact
with that substrate and adheres to it. The work of adhesion is
” 7 sl “ ( T sv + T l v )
3. Immersional wetting: A substrate not previously in contact with a liquid is immersed

completely in the liquid. The work of immersion is
Wi = ysL-Tsv
Introducing the contact angle 6 from the Young equation,
T s v ~ T sl + Tl v cos 6
the three equations above become

Ws= - T lv ( cos e-l)
Wa = " T lv ( cos e+l)
and
IVj — 4yLY cos 0
and
Spreading wetting is spontaneous when 6 = 0.
Adhesional wetting is spontaneous when 6< 180°.
Immersional wetting is spontaneous when 6<90°.
668 Bognolo
Table 7 Examples of Contact Angles and Surface Tension for Nonionic Surfactants
Contact angle ^ (°)

Surface tension ^
Surfactant (dyn/cm) Glass PET Steel
POE (5) C9CH alcohol 27.9 10 6 32

POE (8 ) C9C 11 alcohol 33.3 8 17 34
POE (5) C 13C 15 alcohol 31.5 11 11 43
POE (9) C 13C 15 alcohol 33.2 9 16 41
POE (4) sorbitan monolaurate 37.0 23 48 63
POE (5) sorbitan monooleate 36.3 20 52 55
POE (20) sorbitan trioleate 40.1 15 59 70
POE (10) octyl phenol 37.7 12 22 40
POE (8.75) nonyl phenol 36.4 12 21 43
Alkyl polyglucoside^^
C9- C 10- C 11, D.P. 1.4 28.7 13 29 39
C9-C io-C n , D.P. 1.8 35.4 22 64 64
Âverage of three measurements. 0.1% w/w, 18°C.

‘’Average o f 1 2 measurements.
^D.P. = degree o f polymerization.
The above considerations hold when the substrate to be wetted is a hard solid, so that the
thermodynamics of the wetting process is the major controlling factor. In this situation, the
effectiveness of a surfactant in modifying the wetting properties of a liquid can be assessed
by means of surface tension and contact angle determinations. Zisman [164] has demonstrated
that complete wetting can occur only when the surface tension of the wetting liquid falls
below a critical value (critical wetting tension) characteristic of each substrate. The critical
wetting tension can be determined by plotting cos 0 versus 7 lv for liquids of different surface
tensions and extrapolating for 6 = 0, i.e., for a zero spreading coefficient.
Table 8 Examples of Wetting Times for Nonionic Surfactants in Nonequilibrium

Conditions
Wetting time in
Cone. distilled water
Surfactant (%) (s) Ref.
C8H17CHOHCH2OH 0.047 7 165

C8Hi7(0C2H4)20H 0.1 5 166
C ioH2 i (OC2H4)2 0 H 0.1 10 166
C ioH2 i (OC2H4)4 0 H 0.1 5 166
C i2H25(OC2H4)4 0 H 0.1 19 167
P-Í-C8Hi7C6H4(0C2H4)40H 0.05 50 168
C8Hi7C0(0C2H4)20H 0.1 72 166
C h H23C0(0C2H4)40H 0.1 23 166
C ioH i9(A1 0 - 1 1)C0 (0 C2H4)4 0 H 0.1 6 166
Source: Ref. 137.

The contact angles and surface tensions for a range of nonionic surfactants are given in
Table 7.
In substrates with large surface areas, the kinetic effects are predominant over the thermo
dynamic equilibria. For nonionic surfactants at concentrations considerably above the CMC,
the kinetics of wetting, as determined by the wetting time, depend on the diffusion constant,
the molecular area of the surfactant at the air/solution interface, and the molecular weight of
the surfactant. Under these conditions, high diffusion constants, large surface area, and low
molecular weight will produce shorter wetting times. Table 8 gives some examples of wetting
times for selected nonionic surfactants in nonequilibrium conditions.
F. Emulsification
Emulsification is the process of dispersing one immiscible liquid into another in the form of
droplets with a particle size that is characteristic of the emulsion type and with a stability that
is determined by the emulsion type and the intended use. Emulsion types are distinguished by
particle size:
Emulsions (sometimes referred to as macroemulsions) have a particle size that is typically

above 400 nm; they are opaque and thermodynamically unstable.
In contrast, microemulsions have a particle size below 100 nm; they are transparent and
thermodynamically stable.
Miniemulsions can be described as systems with intermediate particle size and properties.
Emulsions can be of the water-in-oil (w/o) or oil-in-water (o/w) type depending on whether
the water is the internal (disperse) or external (continuous) phase. Two less common catego
ries are the multiple emulsions, in which the dispersed particles are themselves emulsions,
and the bicontinuous microemulsions, which are of neither the w/o nor the o/w type, but in
which the water and oil phases form an uninterrupted continuous phase.
Similarly to ionic surfactants, nonionics aid the emulsification process by reducing the
interfacial tension between immiscible liquids and counter the increase in interfacial free en
ergy from the increased surface area. In microemulsions the ultralow interfacial tension en
sures that the emulsification occurs spontaneously. The nonionics, however, differ from the
other surfactants in that the stabilization against coalescence of the dispersed droplets is the
consequence of steric effects. In addition, nonionic surfactants are free of electrostatic repul
sions between charged heads and, if properly formulated from a combination of different
molecular species, can produce closely packed interfacial films of high interfacial elasticity.
This further enhances the resistance to coalescence as the emulsion droplets collide, e.g.,
under the effect of Brownian motion. It is also an essential feature in the production and
stabilization of w/o macroemulsions, since the water droplets in such emulsions have no sig
nificant electrical barrier against coalescence. Nonionic surfactants offer advantages over the
ionic species in stabilizing emulsions in a colloidal sense but are not as effective as some
anionics in giving small particle size and fast emulsion development (the “blooming” effect).
For this reason synergistic combinations of anionic and nonionic surfactants are used, for
example, for the formulation of crop protection actives, and in emulsion polymerization.
Both macro- and microemulsions are preferentially prepared by using combinations of sur
factants rather than single ones. In macroemulsions, a mixed surfactant system gives a more
dense interfacial packing and greater mechanical strength to the interfacial film. In microemul
sions a mixture of surfactants and (usually) a cosurfactant are needed to give the ultralow
interfacial tension and the proper balance between hydrophilic and lipophilic properties for
the oil and water phases.
670 Bognoio
Table 9 Hydrophile-Lipophile Balance for Some

Common Oil Phases
Oil-in-water emulsion HLB (± 1 )
Acetophenone 14
Acetylated lanolin 14
Acid, dimer 14
Acid, isostearic 15 - 16
Acid, lauric 16
Acid, linoleic 16
Acid, oleic 17
Acid, ricinoleic 16
Alcohol, cetyl 15 - 16
Alcohol, decyl 15
Alcohol, hexadecyl 11-12
Alcohol, isodecyl 14
Alcohol, isohexadecyl 11-12
Alcohol, lauryl 14
Alcohol, oleyl 13 - 14
Alcohol, stearyl 15 - 16
Alcohol, tridecyl 14
Alcohols, C 12- C 15 13
Arachidyl propionate 7
Arlamol E 7
Arlamol F 11
Beeswax 9
Benzene 15
Benzonitrile 14
Benzophenone-3 7
Bromobenzene 13
Butyl stearate 11
Caprylic/capric triglyceride 5
Carbon tetrachloride 16
Carnauba wax 15
Castor oil 14
Castor oil (hydrogenated) 8
Ceresine wax 8
Cetyl palmitate 10
Chlorinated paraffin 12 - 14
Chlorobenzene 13
Cocoa butter 6
Coconut oil 5
Com oil 6
Cottonseed oil 5-6
Cyclohexane 15
Decahydronaphthalene 15
Decyl acetate 11
Diethylaniline 14
Diisooctyl phthalate 13
Diisopropyl adipate 9
Diisopropylbenzene 15
Dimethyl silicone 9
J-Limonene (varies widely) 6-7
Table 9 Continued
Oil-in-water emulsion HLB (±
Ethylaniline 13
Ethyl benzoate 13
2-Ethylhexyl oxystearate 10
2-Ethylhexyl-/?-dimethylamino benzoate 10
2-Ethylhexyl stearate 7.5
Fenchone 12
Hydrogenated peanut oil 6 -7
Hydrogenated polybutenes 11
Isodecyl isononanoate 9
Isoparaffinic hydrocarbon (M W ~ 190) 12.5
Isopropyl myristate 11-12
Isopropyl lanolate 14
Isopropyl palmitate 11-12
Isosorbide monolaurate 10
(Arlamol ISML)
Isotridecyl isononanoate 8
Jojoba oil 6 -7
Kerosene 12
Lanolin, anhydrous 9
Lanolin, liquid 9
Lard 5
Lauryl amine 12
Menhaden oil 12
Methyl phenyl silicone 7
Methyl salicylate 16
Methyl silicone 11
Mineral oil (light), naphthenic 11-12
Mineral oil, paraffinic 10
Mineral oil (light), paraffinic 1 0-11
Mineral oil (medium), paraffinic 9
Mineral spirits 14
Mink oil 5
Myristyl myristate 8.5
Nitrobenzene 13
N,N-Diethyl-m-toluamide 7 -8
Nonyl phenol 14
Octocrylene 6 -8
Octyl methoxycinnamate 8
Orthodichlorobenzene 13
Palm oil 10
Paraffin wax 10
Petrolatum 7 -8
Petroleum naphtha 14
Pine oil 16
Polyethylene wax 15
Polyoxypropylene 11 stearyl ether 11
Polyoxypropylene 15 stearyl ether 7
Polyoxypropylene 30 cetyl ether 10-11
Propene, tetramer 14
Rapeseed oil 6
672 Bognolo
Table 9 Continued
Oil-in-water emulsion HLB (± 1 )
Silicone oil 5
Silicone oil (volatile) 7 -8
Soybean oil 6
Squalene 10
Styrene 15
Synthetic spermaceti 10
(cetyl/stearyl palmitate)
Toluene 15
Trichlorotrifluoroethane 14
Tricresyl phosphate 17
Tridecyl stearate 6 -9
Trimethylolpropane tri Cg-Cjo ester 8
Vitamin A palmitate 6
Vitamin E 6
Xylene 14
Water-in-oil emulsions HLB (± 1 )
Beeswax 5
Gasoline 7
Kerosene 6
Mineral oil, heavy 4
Mineral oil, light 4
Mineral oils, paraffinic and naphthenic 4 -6
Paraffin 4
Petrolatum 4
Source: Ref. 169.
Em pirical rales for the choice of em ulsifying systems have been developed:
1. Surfactants with a broad molecular distribution or blends of surfactants are preferred
over single species
2. There must be some solubility in the water and oil phase, but it must not be excessive,
to allow migration to the interface.
3. Surfactants that are oil-soluble tend to form w/o emulsions, whereas o/w emulsions
are usually produced with water-soluble surfactants.
To have a degree of consistency in the classification of nonionic surfactants and to save
time in the formulation process, a systematic scheme for identifying the relatively few emulsi
fiers suitable for any given application was introduced in the late 1940s by Griffin [3]. This
is called the hydrophile-lipophile balance, or HLB, system.
In the HLB system, each nonionic surfactant is assigned a number related to the proportion
of hydrophilic and lipophilic moieties in the molecule. These numbers are calculated from the
structure of the molecule [4] or determined experimentally. A similar range of numbers are
assigned to oil phases. These are derived from experimental data. The HLB theory postulates
that the em ulsifying agents that w ill give the best performance for any specific o il are those
whose HLB number more closely approaches that of the oil. To achieve this objective, usually
a combination of surfactants has to be used.
Once the optimum HLB has been identified, there is still a need for further optimization to
select the emulsifier type. This can be done by testing blends of emulsifiers of the same HLB
but of different chemical structure.
Table 9 gives examples of the HLB value for some common oil phases. Table 10 gives
the calculated HLB values for a number of nonionic emulsifiers.
Although the HLB system can assist in the process of the initial screening, it has numerous
disadvantages, notably that it is not indicative of either the efficiency or the effectiveness of
an emulsifier or of emulsifiers being used in combination. More recently Shinoda and Arai
[5,7] proposed a selection method for nonionic emulsifiers based on the determination of the
temperature at which an emulsion changes its nature (i.e., inverting from o/w to w/o or vice
versa). At this phase inversion temperature (PIT) the lipophilic and hydrophilic properties of
the surfactant are in balance for the particular system of oil and water phases, the interfacial
tension is at a minimum, and the emulsion should have the finest particle size.
According to the PIT selection method, an emulsion is made with equal weights of oil and
water phases and a given amount of emulsifier. The temperature at which the emulsion inverts
is then determined. Suitable emulsifiers for o/w or w/o emulsions should give a PIT 20-60°C
and 10-40°C, respectively, lower than the storage temperature of the emulsion. The PIT is
affected and can be changed by a number of parameters [7,171]. These are summarized in
Table 1 1 .
The maximum solubilization method is based on the empirical observation [10,11] that a
surfactant, or a blend of surfactants, producing the smallest particle size for a given water and
oil combination are the ones that solubilize the largest amount of the water phase into the oil
phase before inversion occurs.
G. Dispersion
The stabilization of dispersions in a colloidal sense can be achieved through electrostatic or
steric mechanisms or a combination of the two. In the first instance, stabilization results from
the repulsion of the electrical double layers surrounding the particles. This repulsion prevents
them from approaching that critical distance below which the attractive forces become effec
tive. Electrostatic stabilization is described by the DLVO theory [172,173] and relies on mo
nomeric or oligomeric anionic surfactants, e.g., alkylbenzene sulfonates or naphthalene sulfo
nate-formaldehyde condensates.
In steric stabilization (reviewed in Ref. 174), amphipathic molecules are adsorbed onto the
surface of the particles, which are prevented from reagglomerating by osmotic and entropie
repulsions that originate as the layers of surfactants surrounding the particles overlap on short-
range approach. This is the mechanism of stabilization of nonionic surfactants in aqueous
media and generally for nonaqueous media, where the low dielectric constant of the liquid
does not allow extensive electrostatic interactions to take place.
The steric/electrostatic stabilization relies on high molecular weight poly electrolytes, e.g.,
lignin sulfonates, where the electrostatic repulsion from the ionic groups is superimposed on
the steric hindrance created by the bulky molecules.
The stabilization of dispersions can be explained by considering that the attractive forces
between particles (van der Waals forces) are balanced by electrostatic or steric repulsions.
This allows the development of equations relating the attractive/repulsive potential to the elec-
674 Bognolo
Table 10 Calculated HLB Values

for Nonionic Emulsifiers
Chemical composition HLB
Ethoxylated sorbitol esters

POE (40) sorbitol septaisostearate 9.0
POE (40) sorbitol hexaoleate 10.2
POE (40) sorbitol septoleate 9.0
POE (50) sorbitol hexaoleate 11.4
POE (40) sorbitol tall oil ester 9.7
Ethoxylated amines
POE (18) fatty amine 15.5
POE (5) fatty amine 10.6
Mono- and diglycerides
Glycerol mono- and distearate 3.2
Ethoxylated alcohols
POE (4) lauryl alcohol 9.7
POE (23) lauryl alcohol 16.9
POE (2) cetyl alcohol 5.3
POE (2) stearyl alcohol 4.9
POE (2) oleyl alcohol 4.9
POE (2) oleyl alcohol, spec, quality 4.9
Ethoxylated fatty acids
POE ( 8 ) stearate 11.1
POE (20) stearate 15.0
Sorbitan esters
Sorbitan monolaurate 8 .6
Sorbitan monopalmitate 6.7
Sorbitan monostearate 4.7
Sorbitan tristearate 2.1
Sorbitan monooleate 4.3
Sorbitan trioleate 1.8
Synthetic C 13/C 15 alcohol ethoxylates
POE (1) synthetic primary C 13/C 15 alcohol 2 .8
POE (2) synthetic primary C 13/C 15 alcohol 5.9
Table 10 Continued
Chemical composition HLB

Nonyl phenol ethoxylates
POE (1) nonylphenol 3.3
POE (5.5) nonylphenol 10.7
POE (6 ) nonylphenol 10.9
POE ( 8 ) nonylphenol 12.3
Ethoxylated sorbitan esters
POE (4) sorbitan monolaurate 13.3
POE (20) sorbitan monolaurate 16.7
POE (20) sorbitan monopalmitate 15.6
POE (4) sorbitan monostearate 9.6
POE (20) sorbitan monostearate 14.9
POE (20) sorbitan tristearate 10.5
POE (5) sorbitan monooleate 10.0
POE (20) sorbitan monooleate 15.0
POE (20) sorbitan trioleate 11.0
Source: Ref. 170.
trical double layer, or adsorbed layer thickness, as a function of the interparticle distance,
yielding curves of the type schematically illustrated in Figs. 10a and 10b.
Of the two energy potential minima in the curve of Fig. 10a, the shallow one at large
interparticle distance corresponds to a weak, reversible flocculation (i.e., particles cluster to
gether but can be easily redispersed with a minimal input of mechanical energy). The energy
maximum represents a barrier against particles approaching at close distances. However, if
this maximum is overcome, e.g., because of Brownian motion, a large attractive energy de
velops and irreversible agglomeration occurs.
In Fig. 10b, apart from a possible shallow minimum at large distance, the interaction
energy is always positive (i.e., there is a continuous interparticle repulsion, even at close
distances). Because of this, dispersions produced with nonionic surfactants are more stable
676 Bognolo
Table 11 Parameters Affecting the Phase Inversion Temperature (PIT)
Parameter PIT
Surfactant HLB Increases with HLB.

Surfactant concentration Decreases with concentration, more sharply at a low degree of
ethoxylation. Reaches a constant value at a surfactant con
centration of 3-5%.
POE distribution Increases with broader POE distribution. Broader POE distribu
tion gives greater emulsion stability.
Polarity of oil phase Increases as polarity decreases.
Oil/water ratio Increases as the oil/water ratio increases.
Mixed oils Weighted average (by volume) of the PITs of the emulsions
made from individual oils using the same emulsifying agent.
Source: Ref. 7, 171.
than those made with ionic molecules, even though they might not always allow the achieve
ment of high dispersed phase ratios. An additional reason for the excellent stabilizing proper
ties of nonionic surfactants is the significant extension of the steric barrier surrounding the
particles, which can be adjusted by changing the degree of ethoxylation of the molecule and
is further enhanced by the hydration of the POE chains. The minimum layer thickness for
effective stabilization is on the order of 25 A, which corresponds to approximately 20 POE
units. In practice, optimum performance is obtained with 50-100 POE units, and the adsorp
tion on the surface of the particles can be enhanced by modifying the chain length and the
polarity of the hydrophobic moiety through the addition of POP units [162]. In addition, the
layer of hydrated oxyethylene groups mimics the aqueous phase and significantly reduces the
van der Waals attraction between particles.
A summary of suitable nonionic dispersing agents as a function of the dispersion medium,
and of the polarity of the dispersed phase, is given in Table 12.
Fig. 10 Attractive/repulsive potential in function of interparticle distance for electrostatic (a) and steric
(b) stabilization mechanisms.
Table 12 Suitable Nonionic Dispersing Agents as a Function of the Dispersion Medium and Polarity
of the Dispersed Phase
Dispersion medium
Polarity of
solid Aqueous Nonaqueous
Low Long-chain alcohols, POE, and POE/POP Low-polarity amphipathic molecules,

e.g., alkylbenzenes
High Alkylphenols, POE, and POP/POE, Polyamines, High molecular weight
Unsaturated alcohols, POE, and POP/POE, polyesters, Low polarity POE alcohols
POE/POP block copolymers, Di- and and alkylphenols
polyamines
H. Detergency
Detergency and cleaning represent by far the largest outlet for surface-active agents. There is
extensive literature coverage of the physicochemical, formulation, and application aspects
[175-177]; therefore these are not discussed here.
Polyoxyethylene nonionic surfactants are particularly efficient in removing oily type soil
from hydrophobic substrates (probably because of the low CMC) and in preventing its redepo
sition (probably because of better particle coverage and steric stabilization). Oily type soil
removal appears to be at a maximum when the cloud point of the surfactant is 15-30°C below
the wash temperature; thus for a homologous series of POE alcohols maximum detergency
occurs at increasing POE chain length as the washing temperature increases [178]. In addition
to this, an increase in washing temperature demands an increase in the chain length of the
hydrophobe [179], and the higher the polarity of the hydrophobic chain, the better the removal
of oily type soil from nonpolar surfaces [180].
On cellulosic substrates there is evidence that POE nonionics adsorb through hydrogen
bonding between the hydroxyl groups of the cellulose and the POE chains [181]. This reduces
oily type soil removal and enhances redeposition.
Particulate soil is better removed with anionic surfactants; however, nonionics will help to
reduce redeposition. Nonionics also solubilize those anionics that would otherwise be precipi
tated by multivalent ions in hard water, so they are powerful synergists in enhancing the
performance of anionic surfactants.
VI. APPLICATIONS
Given the broad range of chemicals available, it is not surprising that nonionic surfactants are
used in a very large number of formulations and manufacturing processes. The richness of
product availability is the obvious consequence of the versatility of the alkylene oxide chemis
try that allows the making of systematic, controlled changes in the polarity of the emulsifiers
and consequently the production of tailor-made molecules fitting precisely the performance
requirements in different, and often highly specific, applications. This capability is not
matched by the chemistry of anionic, cationic, or amphoteric surfactants.
Another advantage of nonionic surfactants is their electrical neutrality. This ensures full
compatibility with ionic species and the retention (if not the enhancement) of interfacial prop
erties. But perhaps the most valuable consequence of electrical neutrality is that the interfacial
678 Bognolo
Table 13 Major Applications for Nonionic Surfactants
Application Nonionic surfactant Effect
Detergents and cleaners

Heavy duty powder detergents POE fatty alcohols Soil removal
(POE alkylphenols) ^ Foam suppression
Liquid laundry detergents POE fatty alcohols Soil removal; wetting
Automatic dishwashing Capped fatty alcohols Low foam wetting
POE/POP copolymers
Hand dishwashing Alkanolamides Soil removal
Amine oxides Foam booster
Alkyl polyglucosides Foam booster; anti-irritant
Hard surface cleaners POE fatty alcohols Soil removal and wetting
Amine oxides Soil removal and wetting
Floor polishes POE fatty alcohols Wetting; Leveling
POE alkylphenols
Dry cleaning Sorbitan esters; POE fatty alco Water solubilisation and
hols; (POE nonylphenols) ^ soil removal
Industrial detergents POE fatty alcohols; POE alkyl Soil removal and wetting
phenols
Industrial dishwashers Capped fatty alcohols; POE/POP Low foam wetting
copolymers
Metal cleaners and degreasers POE fatty alcohols; POE nonyl- Cleaning and emulsifi
phenols cation
Proprietary blends Electrolytic cleaning
Personal care
Water-in-oil creams Glycerol esters; sorbitan esters; Emulsification
specialty esters; proprietary
blends
Oil-in-water creams Polysorbates; POE fatty acid es Emulsification
ters; sugar esters; polyglycerol
esters; POE fatty alcohols; pro
prietary blends
Perfume solubilization Polysorbates; POE fatty alcohols Solubilization
Hair and body shampoos Fatty acid alkanolamides Foam booster; thickening
Amine oxides Foam booster
POE fatty acid diesters Thickening
Polysorbates Anti-irritant; solubili
zation
Ethlene glycol/propylene glycol Opacifiers
mono- and diesters
Undecylenic acid esters and Antidandruff
amides
Sorbitan esters Oiling and solubilization
Pharmaceutical
Creams/gels Sorbitan esters; polysorbates; Emulsification
glycerol esters
Vitamin/actives Mannide esters; polysorbates; Solubilization
isosorbide ethers
Table 13 Continued
Animal care
Vitamin solubilization Polysorbates; POE castor oils Solubilization
Calves’ milk replacers Glycerol esters; sucroglycerides; Emulsification
polysorbates; POE fatty acid
esters
Pigments, paints, inks
Printing pastes POE fatty alcohols; POE nonyl- Dispersion and emulsifi
phenols; sorbitan esters; poly cation
sorbates; proprietary blends
POE fatty acid diesters; proprie Thickening
tary blends
Paints POE fatty alcohols Emulsification
POE nonylphenols Dispersion
POE/POP copolymers Antifoam
Proprietary blends Wetting
Printing inks Sorbitan esters Emulsification
Polysorbates Dispersion
Dyes/pigment dispersion Proprietary copolymers Stabilization
POE fatty alcohols Dispersion
POE alkylphenols Wetting
Proprietary blends; POE alkyl Dust suppression
phenols
Metalworking
Rust prevention Sorbitan, glycerol, pentaerithritol Spreading/adhesion of
esters; alkanolamides; POE protective oil; water
amines scavengers
Cutting and grinding oils POE alkylphenols; sorbitan es Emulsification
ters; polysorbates; POE sorbi
tol fatty acid esters
Rolling oils POE fatty alcohols; POE fatty Emulsification
acid esters; POE alkylphenols;
sorbitan esters; polysorbates
Hydraulic fluids Polymeric surfactants; POE Emulsification
nonylphenols
Construction
Cement asphalt mortar Proprietary blends Emulsification
Concrete water reducer and POE nonylphenols Foaming; rheology modi
air entrainer fication
Energy
Coal-water slurries POE/POP high molecular polya Dispersion and rheology
mine and nonylphenol; poly modification
meric surfactants
Oil-based drilling muds Polymeric surfactants Dispersion; emulsification
Crude oil demulsifiers POE and POE/POP polymeric Emulsion breaking
derivatives
Fuel dehazing agents POE/POP copolymers Water scavengers
Fertilizers
Anticaking agents POE fatty acid esters; POE Anticaking
nonylphenols
680 Bognolo
Table 13 Continued
Phosphate production Sorbitan esters Antifoam

Crop protection
Emulsifiable concentrates POE sorbitol fatty acid esters^; Emulsification
POE/POP copolymers'"; POE
alkylphenols POE castor
oils'"; sorbitan esters'"; polysor-
bates
Wettable powders POE alkylphenols'"; POE fatty Wetting, dispersion, and
alcohols'" suspension stabilization
Flowables POE alkylphenols; POE/POP Dispersion
copolymers
Adjuvants Proprietary surfactants and Increase the biological ac
blends; POE fatty amines; tivity of crop protec
POE fatty alcohols; POE alkyl tion chemicals
phenols; sorbitan esters; poly-
sorbates
Pulp and paper
Pulp deresination POE nonylphenols Wetting; dispersion; emul
sification
Paper deinking POE nonylphenols^; POE fatty Dispersion, ink removal,
alcohols ^ ink collection, and
emulsification
POE/POP copolymers'" Low foam wetting; ink
removal
Felt cleaning POE fatty alcohols; POE nonyl Cleaning
phenols
Paper coating Sorbitan esters; polysorbates; Emulsification of silicone
POE castor oils; POE fatty antifoams and wax/
alcohols; POE alkylphenols resin coatings
Textiles
Staple spin finishing POE fatty alcohols; POE nonyl Emulsification, lubrica
phenols; POE/POP copoly tion and antistatic
mers; POE fatty acid esters;
POE fatty amines; POE fatty
amides
Scouring agents POE fatty alcohols''; POE nonyl Wetting and cleaning
phenols ''
Dyeing auxiliaries POE fatty alcohols Dye leveling; wetting
POE fatty amine Dye leveling and re
tardant
Proprietary blends Dye dispersion; leveling
Textile printing POE fatty alcohols Wetting
POE fatty acid diesters Thickening
Proprietary blends Leveling
Resin finishing Proprietary blends Softening
Polymer compounding and fabri
cation
Static electricity control Glycerol esters; POE fatty Static charge dissipation
amines; POE/POP copolymers
Table 13 Continued
Anticondensation (agricultural Sorbitan esters; glycerol esters Water vapor coalescence

films)
Anticondensation (food Glycerol esters; polysorbates Water vapor coalescence
wrapping)
Lubricants Glycerol esters; polyol esters; Reduce internal and exter
fatty acid esters; proprietary nal friction during poly
blends mer processing
Dye dispersants (liquid col Sorbitan esters; polysorbates; Dispersion
orants) POE alkylphenols; glycerol
esters
Plastisols Polymeric surfactants Rheology modifiers
Polymerization
Suspension PVC Sorbitan esters Control of particle mor
phology
Emulsion PVC POE nonylphenols; polysorbates Rheology modification for
finished polymer
Acrylic and vinyl polymers POE fatty alcohols Emulsification; emulsion
and copolymers stabilization
Thermoset resins Alkoxylated bisphenol A Polymer building block
Acrylamide Sorbitan esters; polymeric surfac Emulsification
tants
POE fatty alcohols; POE nonyl Emulsion inversion
phenols; POE/POP copoly
mers
Mining
Emulsion explosives Sorbitan esters; polymeric surfac Emulsification
tants
Food
Margarine Mono- and diglycerides Emulsion stabilization;
crystallization and rhe
ology modification
Polyglycerol esters Emulsification; antispat
tering
Sorbitan esters Crystallization modifiers
Mold release Polyglycerol esters; thermally ox Emulsification
idized soybean oils
Coffee whiteners Monoglycerides; sorbitan esters; Emulsification and dis
polysorbates persion
Whipped toppings Monoglycerides; sorbitan esters; Emulsification and dis
polysorbates; polyglycerol es persion
ters; propylene glycol mono
stearate
Ice creams Mono- and diglycerides; sorbitan Freeze stabilization; over
esters; polysorbates run; taste development,
and melting profile
Bread Mono- and diglycerides; poly Starch complexation and
glycerol esters; sucroglycer- volume build-up
ides; polysorbates; acetyl tar
taric esters; mono- and
diglycerides
682 Bùgnolo
Table 13 Continued
Cakes Mono- and diglycerides; propy Structure stabilization

lene glycol monostearate; sor-
bitan esters; polysorbates
Chocolate and chocolate Polyglycerol polyricinoleate; Rheology modification
toppings sucroglycerides
Monoglycerides Dispersion
Sorbitan esters Antiblooming
Polysorbates Tempering aids
Synthetic creams Mono- and diglycerides; polysor- Emulsification and air
bates; polyglycerol esters entrainment
Soft drinks Sorbitan esters; polysorbates Essential oils and colors
solubilization
În only a limited number o f countries and diminishing use.

•"Used often in combination with anionic emulsifiers in proprietary blends.
Mostly used in proprietary formulations.
Ûsed often in combination with anionic and amphoteric surfactants.
effects are the result of steric rather than electrostatic phenomena. This has advantages in
terms of thermodynamic stability, while the molecular structure makes the molecules resilient
to chemical destabilizers such as water hardness and multivalent ions. In this respect, nonionic
surfactants have the advantageous feature of solubilizing the salts from sulfates and sulfonated
anionics with multivalent ions. This allows the performance of these products to be enhanced
in hard water conditions.
Nonionic surfactants are also much milder than their ionic antagonists, an advantage that
is more and more valued in the current safety-conscious environment.
A. Industrial End Uses

Nonionic surfactants are used in such a broad spectrum of products and processes that it is
impossible to give a full account of their industrial applications. Table 13 gives a summary
(albeit incomplete) of the major outlets. It must be remembered that significant volumes of
nonionic surfactants are consumed for the manufacture of intermediates (discussed later) or of
other products used in further processes. Examples are silicone or wax emulsions for antifoam
agents or paper coating, resin emulsions for leather finishing or paints, and emulsion polymers
for adhesives. Many of these uses are not listed in the table, which essentially addresses those
applications in which nonionic surfactants have a major identifiable role.
B. Nonionic Surfactants as Intermediates

Significant volumes of nonionics are consumed every year for the production of other ionic
surfactants. These are discussed in the approximate order of decreasing commercial impor
tance.
1. Sulfated Polyoxyethylene Aliphatic Alcohols

The major applications of sulfated polyoxyethylene aliphatic alcohols are in liquid hand dish
washing formulations and hair and body shampoos. In these, the three (or less commonly six)
POE adducts of Ci 2“ Ci4 (natural or synthetic) or 0X0 alcohols are used.
2. Sulfated Polyoxyethylene Alkylphenols

Sulfamic acid is the common sulfating agent because it avoids ring sulfonation and the forma
tion of inorganic salts. The original application in textile auxiliaries and hand dishwashing
formulations used the low ethoxylated species. Nowadays, the domestic detergent applications
have been replaced almost completely by other uses, e.g., emulsion polymerization where the
20-50 POE derivatives are used as steric/electrostatic stabilizers.
3. Sulfosuccinates
Sulfosuccinates are among the synthetic surfactants that emerged in commercial applications
after World War II. They are produced by the esterification of maleic anhydride with 1 or 2
mol of a hydroxyl compound, e.g., on Cg-Cig fatty alcohol, an alkylphenol, or their ethoxy
lated derivatives, followed by sulfonation of the double bond with sodium metabisulfite.
The hemiesters are mild surfactants with good foaming and emulsification properties. The
diesters are harsher, are poor foamers, and have low water solubility, but they are powerful
wetters as low concentrations.
Sulfosuccinates find use in emulsion and suspension polymerization, crop protection formu
lations, textile dyeing formulations, oil spill dispersants, hair and body shampoo, dry-cleaning
processes, concrete, and latex foamers.
4. Phosphate Esters
A large variety of phosphate esters are available, almost all directed toward specialized appli
cations. They are stable in highly alkaline and high ionic strength environments. They are
also effective as low temperature emulsifiers and have excellent metal-binding properties. All
these attributes lead to their use as compatibilizing agents and hydrotropes, as emulsifiers in
water-based hydraulic fluids and in personal care formulations as rust inhibitors in metal
working fluids, and as antiwear agents in lubricants.
5. E ther Carboxylates
Ether carboxylates were originally developed with a view to upgrading the performance of
soaps by improving water solubility in general and in hard water in particular. Their interfacial
properties are related to the of the carboxyl group, which is in the region of 5- 6 , and
change significantly above and below this value. In alkaline medium, ether carboxylates func
tion as excellent emulsion or dispersion stabilizers, whereas at lower pH they behave as emul
sifiers. In this respect, ether carboxylates may be viewed as “fugitive” surfactants.
The products are comparatively mild and can be further processed to give a series of
amides that can replace alkanolamides, for example, in shampoo formulations. The other
major application of ether carboxylates is in metalworking fluids.
6. Sulfonated M ethyl Esters

The major applications of sulfonated methyl esters are as dispersants, leveling agents, and
retardants in textile auxiliaries formulations. They are also used in specific emulsion processes
for polyvinyl chloride polymerization.
7. E ther Sulfonates
Ether sulfonates are an emerging class of anionic surfactants, first developed in the United
States. They are meant to provide the features of a strongly ionized anionic group (hence
684 Bognoio
surface and interfacial properties essentially independent of pH) without the hydrolytic insta
bility inherent in the sulfate moiety. Manufacture is complex, and these surfactants are sig
nificantly more expensive than ether sulfates or carboxylates and are in more limited supply.
Nevertheless, they are gaining increasing acceptance in high added value applications such as
special emulsion polymerization processes.
8. Sulfated M onoglycerides and Alkanolam ides

Nowadays, sulfated monoglycerides and alkanolamides have been almost completely dis
placed by other, more cost-effective surfactants.
C. Polymeric Surfactants
Polymeric surfactants are an emerging class of amphipathic molecules that exploit interfacial
rather than surface effects [i.e., the molecules have strong interfacial activity with low (or
nil) surface activity]. The implications are much broader than it might appear; by targeting a
specific solid/liquid or liquid/liquid interface it is possible to develop the real chemicophysical
effects wanted and avoid undesirable consequences from surface phenomena, e.g., foam, or
from the ecotoxicological characteristics of monomeric species.
The chemistry of polymeric surfactants is not new. It involves mainly etherification, esteri
fication, sulfonation, and addition reactions. Some products have been in the trade for a
considerable number of years. What is new is the recognition of the unique opportunities that
this technology can offer and its growing use in many industrial applications. Examples of
anionic and nonionic macromolecules having strong interfacial effects include ( 1 ) polyacry
lates, (2) naphthalene sulfonate/formaldehyde condensates, (3) lignin sulfonates, (4) polyvinyl
alcohol and polyvinyl alcohol-acetate copolymers, (5) high molecular weight POE and block
and random POE/POP glycols, (6) high molecular weight polypeptides and polysaccharides,
and (7) POE sorbitol polyesters.
More recently, a number of high molecular weight polyesters and polyethers [182] have
been produced and commercialized under the Hypermer trademark by ICI Surfactants. The
versatility of the alkylene oxide chemistry allows the balancing of a number of parameters in
the polymers, from relative hydrophilicity through solubility to molecular geometry, and the
development of ad hoc structures for maximum interfacial effects.
Polymeric surfactants can be viewed as containing a number of anchoring groups (ensuring
adsorption on the dispersed phase particles) and stabilizing chains that extend into the continu
ous phase. With copolymers, the anchoring is provided by the fraction of the molecule that
has higher affinity for the dispersed phase. In the case of homopolymers the anchoring is
driven by the free energy of binding polymer segments to the surface through van der Waals
forces, hydrogen bonding, or solvent rejection.
Because of the many anchoring points, polymeric surfactants can adsorb strongly (some
times almost irreversibly) at solid/liquid or liquid/liquid interfaces. Therefore, they provide a
much more resistant interfacial coverage than simpler monomeric molecules that are easily
displaced on particle collision. Furthermore, as the stabilization is the consequence of steric
effects (entropic and osmotic), there is no real minimum in the potential energy curves (see
Fig. 10b). Thus the emulsions, or dispersions, should be thermodynamically stable (i.e., the
particles retain indefinitely their individuality, without any coagulation or coalescence). If
phase separation occurs, it does so because of weak flocculation. The original properties of
the system can be easily restored by a small input of mechanical energy.
Polymeric surfactants have proved their value in the realization of disperse systems for
extreme conditions (i.e., systems characterized by unusually high conditions of dispersed/
continuous phase ratios, ionic strength, operational temperature, and pressure). These are
encountered, for example, in [182] oil-based drilling muds, emulsion explosives, hydraulic
fluids, rolling oils, nonaqueous dispersions, inverse-phase polymerization processes, and mul
tiple emulsions. Under these conditions conventional surfactants would just not function, be
cause of displacement from the surface, or salting-out, or precipitation.
VII. ECOTOXICOLOGICAL ASPECTS

The human safety and environmental impact of nonionic surfactants depends on the chemical
class and structure.
Alcohol ethoxylates are the nonionic surfactants produced in the largest volume, with
ethoxylates based on linear, or essentially linear, alcohols being the major type. These find
use in household detergent products, institutional and industrial cleaning, cosmetics, agricul
tural formulations, and many industrial products and processes. The toxicity of alcohol ethox
ylates has undergone extensive assessment in animal studies in view of the potential human
exposure in consumer products. Alcohol ethoxylates exhibit a low order of acute toxicity by
oral, dermal, and inhalation routes of exposure. The acute oral toxicity is low, with reported
oral median lethal dose values in the range from 544 to greater than 25,000 mg/kg bodyweight
[183]. Acute dermal toxicity in animal studies is equally low, with dermal median lethal dose
levels typically exceeding 2000 mg/kg bodyweight [184]. Although undiluted and aqueous
dilutions of certain alcohol ethoxylates can be skin irritants under exaggerated exposure condi
tions associated with animal studies, irritation is rarely encountered in humans under normal
conditions of use and exposure [185]. Alcohol ethoxylates are reported to give negative re
sponses in a variety of in vitro genotoxicity assays [186].
Polyethoxylated sorbitan esters (TWEEN, polysorbates) in acute and chronic toxicity stud
ies have shown very low orders of effects and minimal potential for skin and eye irritation.
Extensive clinical testing of TWEEN products has revealed little potential for human skin
irritation or evidence of skin sensitization. This supports their application as safe-for-use in
cosmetic products [187].
Polyethylene glycol esters of stearic acid (MYRJ, PEG stearates) are used extensively in
cosmetics-related applications. The acute oral toxicity of PEG stearates is very low, with oral
median lethal dose levels reported as exceeding 2 g/kg bodyweight. Skin and eye irritation in
animal studies is reported to be minimal with mild to moderate ocular effects and mild irrita
tion to the skin. The responses from clinical studies indicate no evidence of irritation or
sensitization in preparations containing up to 25% PEG stearates [188].
The sorbitan fatty acid esters (e.g., SPAN) are generally recognized as having minimal
skin and eye irritation, and nonsensitizing effects in animals. The results of oral toxicity
studies on sorbitan fatty acid esters indicate very low orders of toxicity by ingestion, with
typically the median lethal dose exceeding 30 g/kg bodyweight. Considerable longer term test
data addressing general systemic toxicity, reproductive toxicology, and carcinogenicity have
revealed no significant evidence of effects, supporting their extensive use in food packaging
and cosmetics applications [189].
Nonylphenol ethoxylates have undergone a substantial degree of toxicity testing to establish
their safety in industrial and consumer applications. The overall conclusion is that nonylphe
nol ethoxylates exhibit a low order of toxicity to mammalian species by oral, dermal, or
inhalation exposure. Nonylphenol ethoxylates have not been found to cause reproductive,
genotoxic, or carcinogenic effects. Skin irritation studies indicate responses ranging from non
irritating through to moderately irritating. The degree of ocular irritation appears to depend
686 •Bùgnolo
on the degree of ethoxylation, with generally the shorter chain variants showing more poten
tial for causing eye irritation [183].
Fatty acid diethanolamides based on coco, lauric, and oleic acids are recognized as having
a low order of oral toxicity, with the median lethal dose typically greater than 2 g/kg body-
weight. There is some variation in eye and skin irritation characteristics, with cocoamide DBA
giving minimal eye irritation and moderate skin irritation in the rabbit, while lauramide DBA
and linoleamide DBA were mild to moderate eye irritants and mild to severe skin irritants.
Oleamide DBA is reported to be nonirritant to the rabbit eye, but a moderate skin irritant
following single and cumulative application [190].
The environmental fate and effect of nonionic surfactants is increasingly recognized as a
key determinant in their selection and application. The growing impact of legislative develop
ments and the corresponding recognition of these developments has a major bearing on this
process of selecting nonionic surfactants that match either regulatory driven or customer-
led requirements.
Nonionic surfactants used in detergent applications in the Buropean Union are required to
comply with specific requirements set out in Council Directive 82/242/BBC, which prohibits
the placing on the market, and the use of, a detergent if the level of biodegradability of the
contained nonionic surfactant is less than 80% in accordance with specific test methods. The
most common alkoxylated nonionic surfactants fully meet the conditions of this directive.
Linear and branched alcohol ethoxylates, nonylphenol ethoxylates, and many alcohol alkox-
ylates all meet the minimum 80% primary biodegradation level in either the static or confir
matory test methods. The linear block copolymers of ethylene and propylene oxide, in con
trast, do not meet the minimum primary biodegradation level, primarily through a combina
tion of high molecular weight and the influence of the pendant methyl groupings on the
oxidative-based biodegradation pathway [191].
The ultimate biodegradation (effectively the measure of whether complete mineralization
of the organic compound to carbon dioxide, water, and biomass occurs) for nonionic surfac
tants has been extensively investigated. Linear and essentially linear fatty alcohol ethoxylates
Table 14 Overview of Environmental Releases from the Production of Specific

Nonionic Surfactants ^
AE3- AE3- AE3- AE7- AE7- ABB AEn- APG- APG-

PC PKO CNO PC PKO ONO PO PKO CNO
Atmospheric emis
sions
Particulates 4.31 5.41 3.87 4.14 4.95 3.88 5.67 13 11.9
Nitrogen oxides 17.9 13.6 11.1 16.6 13.6 11.8 13.6 15.1 13.2
Hydrocarbons 36.2 25.2 24.2 36.8 29.8 29.1 30.5 16.4 15.7
Sulfur oxides 13.2 10.5 8.68 11.7 10 8.77 9.23 15.5 14.1
Carbon monoxide 1.41 1.48 0.89 1.3 1.36 0.95 1.47 2.22 1.78
Methane 0.012 14.5 0.015 0.015 10.1 0.018 15.7 10.9 0.0056
Waterborne emissions
BOD 1.34 1.55 9.51 1.45 1.67 7.23 1.72 1.02 7.02
COD 2.7 3.72 10.7 2.73 3.51 8.41 3.85 2.81 8.08
Dissolved solids 1.94 4.38 25.5 1.51 3.18 17.9 4.41 5.37 21.3
Suspended solids 0.17 0.82 0.11 0.16 0.62 0.12 0.62 12.8 12.2
Solid waste (total) 66.97 74.1 57.9 64.04 69.9 58.3 62.9 147.8 135.4
^Values are expressed as k g/1000 kg surfactant.

Source: Ref. 196.
Table 15* Overview of the Resource

Requirements for the Production of
Specific Nonionic Surfactants^’*"
Raw material (kg)
Total AE3-PC 2078.6

Total AE3-PKO 3175
Total AE3-CNO 3031
Total AEv-PC 2143.1
Total AE7-PKO 2939
Total AE7-CNO 2859
Total AEn-PO 2695
Total APG-PKO 2905.8
Total APG-CNO 2831.9
^See Abbreviations preceding References at

end of chapter.
^Values are expressed as k g/1000 kg surfac
tant (1.3)
Source: Ref. 197.
are recognized as having a significant level of ultimate biodegradability, showing a BOD28:

COD ratio greater than 60%. Alkylphenol ethoxylates, in contrast, although showing a high
degree of primary biodegradability, reveal a lower extent of ultimate biodegradation. This
lower degree of biodegradation is considered to reflect the influence of the highly branched
alkyl chain and the steric effect of the branched phenyl entity on cleavage of the hydrophile
from the hydrophobe, which is the accepted mechanism for biodegradation of essentially lin
ear alcohol ethoxylates [192].
The effect of exposure of aquatic organisms to nonionic surfactants is generally well de
fined for individual types, but there is no one general trend relating toxicity with molecular
structure. For alcohol ethoxylates, alteration of the hydrophilic moiety due to an increase in
the degree of ethoxylation is responsible for a significant reduction in acute toxicity, for
example, by at least two orders of magnitude by extending the EO number from 10 to 30
(LC50 increasing from 1 to 100 mg/L) [193].
With linear ethylene oxide/propylene oxide copolymers, typically the acute aquatic toxicity
is greater than 100 mg/L, with reported 96 LC 50 values for rainbow trout in the range of
200-1000 mg/L. By contrast, the alcohol alkoxylates introduced as alternatives to the block
copolymers, effectively on the basis of improved biodegradation profile, have a higher order
of toxicity in the range of 0.2-5 mg/L [194].
Ethoxylated fatty amines are used extensively as agrochemical surfactants, wetting agents,
and emulsifiers. Their environmental characteristics are comparable with those of fatty alcohol
ethoxylates, with acute toxicity to aquatic species generally decreasing with increasing ethoxy
lation, with reported EC 50 values to Daphnia magna ranging from 1.3 to 30 mg/L for 2-25
ethylene oxide units [191]. The biodegradation of ethoxylated fatty amines is considered to
be substantial and generally consistent with the criteria developed to distinguish substances as
readily biodegradable [195].
Currently, the environmental aspects are receiving growing attention, prompted by the
awareness of the environmental consequences of manufacturing and using chemicals in gen
eral, and surfactants in particular, and the need for a better understanding of the impact on
the environment on a global basis. Thus life-cycle assessments are becoming common practice
to help to understand the environmental implications of a product over its entire life span,
from the extraction of raw materials through manufacture of the intermediate, to product use
688 Bognolo
Table 16 Overview of the Energy Requirements for the Production of Specific Nonionic Surfactants^
Energy summary (GJ)
Process Transport EMR Total
30.3 1.35 51.5 83.2

22.2 1.79 46.4 70.4
AE3-CNO 19.1 2.01 55 76.1
AE7-PC 27.7 1.32 49.9 78.9
AE7-PKO 21.8 1.71 46.6 70.1
AE7-CNO 19.6 1.86 52.6 74.1
AEh-PO 19.5 2.22 60.1 81.8
APG-PKO 26.6 1.99 32.1 60.7
APG-CNO 24.3 2.15 38.6 65.1
Energy profile (GJ)
Natural gas Crude oil Coal Hydropower Nuclear Biomass Other Total
AE3-PC 43 32.2 4.93 0.33 2.8 0 0.0022 83.2

AE3-PKO 18.1 17.9 4.1 0.74 1.69 27.9 0 70.4
17.9 15.5 4.1 0.35 1.69 36.5 0 76.1
AE7-PC 39.8 31 4.82 0.34 2.86 0 0.0016 78.9
AE7-PKO 22.5 21.2 4.2 0.62 2.09 19.5 0 70.1
AE7-CNO 22.4 19.6 4.2 0.34 2.09 25.5 0 74.1
AEii-PO 23.1 22 4.1 0.23 2.22 30.2 0 81.8
APG-PKO 12.2 8.59 6.51 0.6 1.69 31.1 2.21E-05 60.7
APG-CNO 12.2 6.82 6.51 0.3 1.69 37.6 2.21E-05 65
^Values are expressed as GJ/1000 kg surfactant.

Note: EMR (energy o f material resource) represents the energy equivalent that is diverted from other uses (e.g ., fuel)
to manufacture the specific surfactant.
Source: Ref. 196.
and final disposal. The first step in the assessment, the life-cycle inventory (LCI), has been
recently completed for a range of large-volume products by the European Surfactants Study
Group under the administrative support of the group ECOSOL of the European Chemical
Industry council CEFIC. The group included 13 major manufacturers of surfactants and for
mulation detergents. It assessed the resources requirements and environmental releases during
the production of detergent surfactants. Tables 14-16 summarize some of the key parameters
for the nonionic surfactants included in the study [196-198].
As yet, there is no agreed-upon impact assessment methodology to enable a comparison
on a qualitative basis of the global environmental effects arising from the production and use
of products. This would involve aspects such as global warming, acid rain, surface water
quality, and landfill implications. The data produced are, however, a powerful tool in under
standing process issues on a much more complete basis than hitherto possible, and conse
quently for identifying and implementing improvements.
ACKNOWLEDGMENTS
I am indebted to Mr. L. Hughes and Dr. H. Thomas of ICI Surfactants for their contribution
to the section on the ecotoxicological aspects of nonionic surfactants. I am also indebted to
Nonìonlc Surfactants 689
the works of M. J. Schick {Nonionic Surfactants (M. J. Schick, Ed.), Marcel Dekker, New
York, 1966] and M. J. Rosen {Surface and Interfacial Phenomena, 2nd ed., Wiley, New
York, 1989), which are milestones in the bibliography of nonionic surfactants and have pro
vided ideas and material for the compilation of this chapter.
ABBREVIATIONS
AE3; AE7; AEji fatty alcohol (natural or synthetic origin) reacted with 3, 7, or 11 ethylene
oxide units
AP Alkylphenol
APG Alkylpolyglucoside
BOD Biochemical oxygen demand
CN O Coconut oil
COD Chemical oxygen demand
DEA Diethanolamine
EG ethylene glycol
EM R Energy of material resource
G la Glucamine
G lu Glucose
G ly Glycerol
N Natural/renewable origin
NFA Natural fatty alcohol
N FA c Natural fatty acid
N FA m Natural fatty amine
Oxy Oxygen
P (or PC) Petrochemical origin
PET Polyethylene terephthalate
PGL Polyglycerol
PK O Palmkemel oil
PO Palm oil
PO E Polyoxyethylene
POP Polyoxypropylene
SFA Synthetic fatty alcohol
SFA m Synthetic fatty amine
Sor Sorbitol
Sue Sucrose
REFERENCES
1 . C. Schoeller and M. Wittwer, U .S. Patent 1, 970, 578 (Nov. 24, 1931).
2. A. Steindorff, G. Balle, K. Horst, and R. Michel, U .S. Patent 2, 213, 477 (Nov. 18, 1936).
3. W. C. Griffin, / . S o c . C o s m e t . C h e m . 1 : 311 (1949).
4. W. C. Griffin, / . S o c . C o s m e t . C h e m . 5 : 249 (1954).
5. K. Shinoda and H. Arai, J . P h y s . C h e m . 6 8 : 3585 (1964).
6 . K. Shinoda and H. Arai, J . C o l l o i d S c i . 2 0 : 9 3 (1965).
7. K. Shinoda, Proc. 5th Int. Congr. Detergency, September 1968, Barcelona, Voi. 2, p. 275.
8 . K. Shinoda and H. Saito, J . C o l l o i d I n t e r f a c e S c i . 3 0 : 258 (1969).
9. H. Arai and K. Shinoda, J . C o l l o i d I n t e r f a c e S c i . 2 5 : 396 (1967).
10. T. J. Lin, H. Kurihara, and H. Ohta, J . S o c . C o s m e t . C h e m . 2 8 : 457 (1977).
11. T. J. Lin, J . S o c . C o s m e t . C h e m . 3 0 : 167 (1979).
12. J. T. Davies, I n t e r f a c i a l P h e n o m e n a , Academic, New York, 1963, pp. 3 71-383.
69Q Bognolo
13. C. D. Moore and M. Bell, S o a p P e r f u m e r y C o s m e t . 2 9 : 893 (1956).

14. H. L. Greenwald, G. L. Brown, and M. H. Fineman, A n a l . C h e m . 2 8 : 1963 (1956).
15. L. Marszall, A c t a P o l y m . P h a r m . 3 2 : 397 (1975).
16. L. Marszall, C o l l o i d P o l y m . S c i . 2 5 4 : 674 (1976).
17. L. Marszall, F e t t e S e i f e n A n s t r i c h m . 8 0 : 289 (1978).
18. C. K. Ingold, S t r u c t u r e a n d M e c h a n i s m in O r g a n i c C h e m i s t r y , Cornell Univ. Press, Ithaca, NY,
1953, p. 341.
19. R. E. Parker and N. S. Isaacs, C h e m . R e v . 5 9 : 737 (1959).
20 . H. M. Stanley, in E t h y l e n e a n d I t s I n d u s t r i a l D e r i v a t i v e s (S. A. Miller, Ed.), Ernest Benn, Lon
don, 1969, pp. 16, 17, 527, 528.
21 . J. N. Cawse, J. P. Henry, M. W. Swartzlander, and P. H. Wadia, in K i r k - O t h m e r E n c y c l o p e d i a
o f C h e m i c a l T e c h n o l o g y , 3rd ed., Vol. 9 (M. Grayson, Ed.), Wiley-Interscience, New York,
1980, p. 432.
22 . S. Lenher, J . A m . C h e m . S o c . 5 3 : 3737 (1931).
23. S. Lenher, J . A m . C h e m . S o c . 5 3 : 3752 (1931).
24. G. L. Cunningham, A. W. Boyd, R. J. Myers, W. D. Gwinn, and W. I. Levan, J . C h e m . P h y s .
1 9 : 676 (1951).
25. T. E. Turner and J. A. Howe, J . C h e m . P h y s . 2 4 : 924 (1956).
26. L. Pauling, T h e N a t u r e o f t h e C h e m i c a l B o n d , 3rd ed., Cornell Univ. Press, Ithaca, N Y, 1960,
p. 1 1 2 .
27. N. Schachat and H. Green wald. Nonionic surfactants, in S u r f a c t a n t s S c i e n c e , Vol. 1 (M. J.
Schick, Ed.), Marcel Dekker, New York, 1966, p. 10.
28. CEFIC, G u i d e l i n e s f o r th e D i s t r i b u t i o n o f E t h y l e n e O x i d e , Revision 1, Brussels, 1992.
29. G. J. Jackobson, F. W. Kathagen, M. Klatt, and U. Steinbemer, U l l m a n n E n c y c l . C h e m . T e c h n o l .
A 1 2 : 480 (1989).
30. Shell Development C o., U .S. Patent 2,605,293 (1952).
31. Shell Development Co., U .S. Patent 2,810,768 (1957).
32. Pittsburgh Plate Glass C o., GB Patent 926,804 (1963).
33. Shell Development Co., U .S. Patent 2,779,801 (1957),
34. G. Jackobson, F. W. Kathagen, M. Klatt, and U. Steinbemer, U ll m a n n E n c y c l . C h e m . T e c h n o l .
A I 2 : 485 (1989).
35. H. Dillenburg and H. Schuffems (to Deutsche Solvay Werke), DE-OS 3, 410,520 A1 (1984).
36. R. J. Capik and L. Wright (to Atlas Chemical Industries), U .S. Patent 3,588,019 (1970).
37. G. G. Boyers (to Engelhard Industries, Inc.), U .S. Patent 2,868,847 (1959).
38. R. W. Kerr (Ed.), C h e m i s t r y a n d I n d u s t r y o f S t a r c h , 2nd ed.. Academic, New York, 1950,
Chapter 14.
39. A. D. Woolhouse, S t a r c h a s S o u r c e o f C a r b o h y d r a t e S w e e t e n e r s , Report CD 2237, Dept, of
Scientific and Industrial Research, Aukland, New Zealand, 1976, p. 20.
40. E. R. Kool and F. C. Armbmster, in S t a r c h C h e m i s t r y a n d T e c h n o l o g y , Vol. 7 (R. L. Whistler
and E. F. Paschall, Eds.), Academic, New York, p. 562.
41. R. V. MacAllister, in A d v a n c e s in C a r b o h y d r a t e C h e m i s t r y a n d B i o c h e m i s t r y , Vol. 36 (R. S.
Tipson and D. Horton, Eds.), Academic, New York, 1979, p. 40.
42. Novo Industry A/S, D e x t r o s e a n d S t a r c h S u g a r , Brochure B-007 g GB, Bagsvaerd, Denmark,
1977.
43. W. Kuhn, H y d r o c a r b o n P r o c e s s 59(10): 123 (1979).
44. R. A. Scheldon and J. A. Van Doom, J . C a t a l . 3 4 : 242 (1974).
45. R. A. Scheldon and J. A. Van Doom, J . C a t a l . 3 1 : A l l (1973).
46. K. H. Simmrock, H y d r o c a r b o n P r o c e s s 57(11): 105 (1978).
47. G. C. Sinke and D. L. Hildebrand, J . C h e m . E n g . D a t a 7(1): 74 (1962).
48. A. S. Pell and G. Pilcher, T r a n s . F a r a d a y S o c . 6 1 : 71 (1965).
49. E. S. Pattison (Ed.), F a t t y A c i d s a n d T h e i r I n d u s t r i a l A p p l i c a t i o n s , Marcel Dekker, New York,
1968.
50. D. Swem, B a i l e y ’s I n d u s t r i a l O i l a n d F a t P r o d u c t s , 4th ed., W iley, New York, 1979.
51. Anon., Novel process for enzymatic fat splitting, C hem . W eek, March 23, 1988, p. 28.
52. E. F. Hill, G. R. Wilson, and E. C. Steinle, Ind. E ng, C hem . 46(9): 1917 (1954).
53. A. Landgraf and O. Roelen, German Patent 931,405 (1955).
54. August 1964, p. 19.
C h e m ic a l W e e k ,
55. Continental Oil, Tech. Bull. AL-11, 1963.
56. I. Wender, S. Metun, S. Ergun, H. W. Sternberg, and H. Greenfield, J. A m . C hem . Soc. 78:
5401 (1956).
57. L. F. Hatch, H ig h e r O x o A lc o h o ls , Wiley, New York, 1957.
58. K. Ziegler, Br. Patent 763,828 (1956).
59. A. W. Ralston, W. O. Pool, and H. J. Harwood (to Armour and C o.), U .S. Patent 2,061,314
(1936).
60. R. H. Potts (to Armour and Co.), U.S. Patent 2,546,521 (1950).
61. IG Farbenindustrie A. G ., Fr. Patent 781,960 (1935).
62. A. W. Ralston, C. Hoerr, W. Pool, and H. Harwood, J . O r g . C h e m . 9 : 102 (1944).
63. W. O. Pool and R. Potts (to Armour and Co.), U .S. Patent 2,358,030 (1944).
64. H. Fiege and H. W. Voges, U l l m a n n E n c y c l . C h e m . T e c h n o l . A 1 9 : 314-347 (1991).
65. J. G. Baldinus (to Colgate Palmolive), U .S. Patent 2,910,491 (1959).
6 6 . R. B. R. Choudhury, J . A m . O i l C h e m . S o c . 3 7 : 483 (1960).
67. R. O. Feuge and A. E. Bailey, O i l S o a p , 23, 259 (1946).
68. N. H. Kuhrt, E. A. Welch, and F. J. Kovarik, /. A m . O il C hem . Soc. 27: 310 (1950).
69. N. H. Kuhrt (to Eastman Kodak Co.), U.S. Patent 2,634,278 (1953).
70. A. Gran, C hem . U m schau. 32: 225 (1925).
7 1. E. W. Eckey and M. W. Formo, J. A m . O il C hem . Soc. 26: 207 (1949).
72. C. S. Mimer and N. N. Dalton, G ly c e r o l, Reinhold, New York, 1953, p. 366.
73. B. R. Harris, U.S. Patent 1,917,257 (1933).
74. B. R. Harris, U.S. Patent 2,023,388 (1935).
75. H. Burrell, Ind. E ng. C hem . 37: 86 (1945).
76. K. G. Mizuch, Z. A. Gorbunova, J. A. Roetermel, and V. A. Blinov, USSR Patent 125,572
(1960).
77. E. R. Mueller, P. L. Eness, and E. E. McSweeney, I n d . E n g . C h e m . 4 2 : 1537 (1950).
78. K. R. Brown (to Atlas Powder C o.), U .S. Patent 2,322,820 (1943).
79. W. L. Kubie, J. L. O ’Donnel, H. M. Teeter, and J. C. Cowan, J . A m . O i l C h e m . Soc. 40:
105 (1963).
80. W. C. Griffin (to Atlas Powder Co.), U.S. Patent 2,374, 931 (1945).
81. W. C. Griffin (to Atlas Powder C o.), U .S. Patent 2,380,166 (1945).
82. U .S . P h a r m a c o p o e i a 2 3 : NF 18 (1995).
83. B r i t i s h P h a r m a c o p o e i a , London, Her Majesty’s Stationery Office (1993).
84. E u r o p e a n P h a r m a c o p o e i a published under the direction of the Council of Europe, European Treaty
Series No. 50, Maisonneave S.A. Saint Ruffine, France. (July 1995).
85. 2 1 s t C o d e o f E e d e r a l R e g u l a t i o n s (FDA 21 CFR).
8 6 . E m p f e h l u n g d e r K u n s t o f f - K o m m i s i o n d e s B u n d e s g e s u n d h e i t s a m t e s (BGA), Kunstoffe im Lebens
mittel verkehc, Karl Heymanns Verlag, Köln, 1995.
87. European Parliament and Council Directive No 95/2/EC of Feb. 20, 1995; O fficial J o u r n a l o f th e
E u r o p e a n C o m m u n i t i e s L 6 1 1 1 , March 18, 1995.
8 8 . H. B. Hass, F. D. Snell, S. C. York, and L. I. Osipow (to Sugar Research Foundation, Inc.),
U.S. Patent 2,893,990 (1959).
89. W. F. Huber and N. B. Tucker (to Procter and Gamble Co.), U .S. Patent 2,812,324 (1958).
90. L. Osipow, F. D. Snell, and A. Finchler, J . A m . O i l C h e m . S o c . 3 4 : 185 (1957).
91. S. Komori, M. Okahara, and K. Okamoto, J . A m . O i l C h e m . S o c . 3 3 : 44 (1956).
92. W. B. Satkowki, S. K. Huang, and R. L. Liss, in N o n i o n i c Surfactants (M. J. Schick, Ed.),
Marcel Dekker, New York, 1966, pp. 142-174.
93. K. Nagase and K. Sakaguchi, K o g y o K a g a k u Z a s s h i 6 4 : 1035 (1961).
94. J. D. Malkemus and J. I. Swan, J . A m . O i l C h e m . S o c . 3 4 : 342 (1957).
692 Bognolo
95. E. Fisher, B er. 26: 2400 (1893).

96. J. Defaye and C. Peddersen, Z u c k e r in d u s tr ie 1 1 6 {A ): 271 (1991).
97. W. Konings and E. Knorr, B er. 34: 957 (1901).
98. G. Zemplen, B er. 62 : 985 (1929).
99. J. Bamoub and D. R. Bundle, C a n . J . C h e m . 5 7 : 2085 (1978).
100. R. J. Ferrier and N. Prasad, J . C h e m . S o c . ( 1 9 6 9 ) : 571.
101. K. G. I. Nilsson, T r e n d s B i o t e c h n o l . 6 : 256 (1988).
102. J. E. Davis and J. C. Letton, Eur. Patent 0,132,043 (1987).
103. W. D. Niederhauser and E. J. Smialkowski (to Rohm and Haas Co.), U .S. Patent 2,856,434
(1958).
104. N. O. V. Sonntag, in F a t t y A c i d s : T h e i r C h e m i s t r y , P r o p e r t i e s , P r o d u c t i o n a n d U s e s , Part 3,
2nd ed. (K. S. Markley, Ed.), Wiley-Interscience, New York, 1964, Chapter XVI.
105. R. H. Potts (to Armour and C o.), U .S. Patent 2,555,606 (1951).
106. C. K. Ingold, S t r u c t u r e a n d M e c h a n i s m in O r g a n i c C h e m i s t r y , Cornell Univ. Press, Ithaca, NY,
1953, p. 341.
107. R. E. Parker and N. S. Isaacs, C h e m . R e v . 5 9 : 737 (1959).
108. Jefferson Chemical Co., E t h y l e n e O x i d e , Tech. Brochure, Houston, TX, 1962.
109. J. D. Malkemus, J . A m . O i l C h e m . S o c . 3 3 : 571 (1956).
110. H. G. Nadean, D. M. Oaks, W. A. Nichols, and N. P. Carr, A n a l . C h e m . 3 6 : 1914 (1964).
111. E. J. Mills and D. G. Leis, U.S. Patent 2,988,572 (1961).
112. R. W. Wilson (to Monsanto Co.), U.S. Patent 2,586,767 (1952).
113. G. J. Stockburger and J. D. Brandner, J . A m . O i l C h e m . S o c . 4 0 : 590 (1963).
114. A. N. Wrigley, A. S. Stirton, and E. Howard, Jr., J . O r g . C h e m . 2 5 : 439 (1960).
115. W. P. Krause (to Jefferson Chemical Co.), U.S. Patent 2,983,763 (1961).
116. W. D. Beauchamp, R. E. Booth, and E. R. Degginger (to Allied Chemical Corp.), U .S. Patent
3,016,404 (1962).
117. R. Ernst (to Textilana Corp.), U .S. Patent 3,024,260 (1962).
118. G. C. Tessoro (to Onyx Oil and Chemical Co.), U .S. Patent 2,844,609 (1958).
119. Aktiebolaget Ferrosan, Br. Patent 803,769 (1958).
120. J. Kritchevsky, J. A m . O il C hem . Soc. 34: 178 (1957).
121. V. Lamberti and G. J. McCrimslisk (to Unilever N .V .), Fr. Patent 1,321,790 (1963).
122. J. V. Schurman (to Colgate Palmolive C o.), U .S. Patent 2,877,246 (1959).
123. D. R. Jackson and L. G. Lundsted (to Wyandotte Chemicals Corp.), U .S. Patent 2,677,700
(1954).
124. D. R. Jackson and L. G. Lundsted (to Wyandotte Chemicals Corp.), U .S. Patent 3,036,130
(1962).
125. A. B. Steele, Jr., E. J. Mills, Jr., and D. G. Leis (to Union Carbide Corp.), U .S. Patent
3,078,315 (1963).
126. L. G. Lundsted (to Wyandotte Chemicals Corp.), U .S. Patent 2,674,619 (1954).
127. L. G. Lundsted (to Wyandotte Chemicals Corp.), Br. Patent 722,746 (1955).
128. D. B. Lake and G. L. K. Hoh, J . A m . O i l C h e m . S o c . 4 0 : 628 (1963).
129. W. Pilcher and S. L. Eton (to Procter and Gamble Co.), U .S. Patent 2,999,068 (1961).
130. I. Webb (to Union Oil Co.), U.S. Patent 2,787,595 (1957).
131. J. H. Wilson and J. Rudy (to Unilever Ltd.), Belg. Patent 604,156 (1951).
132. Procter and Gamble Co., Belg. Patent 641,766 (1964).
133. A. H. Lewis and R. D. Steiner (to California Research Co.), U .S. Patent 2,670,367 (1964).
134. W. H. Woodstock (to Victor Chemicals Works), U.S. Patent 2,586,897 (1952).
135. R. A. Gibbson, N a t u r e 2 0 0 : 665 (1963).
136. P. Becher, in N o n io n ic Surfactants, (M. J. Schick, Ed.), Marcel Dekker, New York, 1966, pp.
480-486.
137. M. J. Rosen, S u r fa c ta n ts a n d I n te r f a c ia l P h e n o m e n a , W iley, New York, 1989, pp. 7 0 -8 0 , 128-
132, 149, 264-268, 296.
138. H. Saito and K. Shinoda, J . C o l l o i d I n te r f a c e S c i. 2 4 : 10 (1967).
139. J. Xia and Z. Hu, in S u r f a c t a n t s in S o l u t i o n , Vol. 5 (K. L. Mittal and P. Botherel, Eds.),
Plenum, New York, 1986, pp. 1055-1065.
140. T. Nakagawa, H. Inue, K. Tori, and K. Kuryiama, N i p p o n K a g a k u Z a s s h i 7 9 : 1194 (1958).
141. K. Kuryiama, K o l l o i d Z. 1 8 0 : 55 (1962).
142. K. Kuryiama, K o l l o i d Z. 1 8 1 : 444 (1962).
143. K. Kuryiama, H. Inue, and T. Nakagawa, K o l l o i d Z. 1 8 3 : 6 8 (1962).
144. R. R. Balmbra, J. S. Clunie, J. M. Corkill, and J. F. Goodman, T r a n s . F a r a d a y S o c . 5 8 :
1661 (1964).
145. P. H. Elworthy and C. B. Macfarlane, J . C h e m . S o c . 1 9 6 3 : 907.
146. C. W. Dwiggins, Jr. and R. J. Bolen, J . P h y s . C h e m . 6 5 : 1787 (1961).
147. M. J. Schick, J . P h y s . C h e m . 6 7 : 1796 (1963).
148. J. M. Corkill, J. F. Goodman, and S. P. Harrold, T r a n s . F a r a d a y S o c . 6 0 : 202 (1964).
149. J. W. Gibbs, T h e C o l l e c t e d W o r k s o f J . W. G i b b s , Vol. 1, Longmans and Company, New York,
1928, p. 119.
150. M. J. Rosen, S u r f a c e a n d I n t e r f a c i a l P h e n o m e n a , 2nd ed., Wiley, New York, 1989, pp. 7 0 -
81.
151. K. Shinoda, T. Yamanaka, and K. Kinoshita, J . P h y s . C h e m . 6 3 : 648 (1959).
152. M. J. Rosen, A. W. Cohen, M. Dahanayake, and X. Y. Hua, J . P h y s . C h e m . 8 6 : 541 (1982).
153. J. E. Carless, R. A. Challis, and B. A. Mulley, /. C o l l o i d S c i . 1 9 : 201 (1964).
154. P. H. Elworthy and C. B. Macfarlane, J . P h a r m . P h a r m a c o l . 1 4 : 100 (1962).
155. A. H. Beckett and R. J. Woodward, J . P h a r m . P h a r m a c o l . 1 5 : 422 (1963).
156. H. Lange and M. J. Schwuger, K o l l o i d Z . P o l y m . 2 4 3 : 120 (1971).
157. M. Dahanayake, A. W. Cohen, and M. J. Rosen, J . P h y s . C h e m . 9 0 : 2413 (1986).
158. R. Aveyard and D. A. Hoy don. A n I n t r o d u c t i o n t o th e P r i n c i p l e s o f S u r f a c e C h e m i s t r y , Cam
bridge Univ. Press, Cambridge, 1973 pp. 201, 212.
159. I. Langmuir, J . A m . C h e m . S o c . 4 0 : 1361 (1918).
160. W. Von Rybinski and M. J. Schwuger, in N o n i o n i c S u r f a c t a n t s , P h y s i c a l C h e m i s t r y , (M. J.
Schick, Ed.), Marcel Dekker, New York, 1987, pp. 4 5 -1 0 7 .
161. J. R. Aston, D. N. Furlong, F. Grieser, P. J. Scales, and G. G. Warr, in A d s o r p t i o n a t th e G a s /
S o l i d a n d L i q u i d / S o l i d I n t e r f a c e (J. Rouquerol and K. S. W. Sing, Eds.), Elsevier, Amsterdam,
1982, pp. 97-102.
162. L. Trochet Mignard, P. Taylor, G. Bognolo, and Th. F. Tadros, C o l l o i d s S u r f a c e s A , P h y s i c o -
c h e m . E n g . A s p e c t s 9 5 : 3 7 -4 2 (1995).
163. H. J. Osterhof and F. E. Bartell, J . P h y s . C h e m . 3 4 : 1399 (1930).
164. W. A. Zisman, A d v a n c e s in C h e m i s t r y , No. 43, American Chemical Society, Washington, DC,
1964.
165. M. J. Rosen and C. C. Kwan, Novel Surfactants From d-Olefins, in S u r fa c e -A c tiv e A g e n ts , Soc.
Chem. Ind., London, 1979, pp. 9 9 -1 0 5 .
166. J. K. Weil, R. E. Koos, W. M. Linfield, and N. Parris, J . A m . O i l C h e m . S o c . 5 6 : 873 (1979).
167. A. W. Cohen and M. J. Rosen, J . A m . O i l C h e m . S o c . 5 8 : 1062 (1981).
168. E. H. Crook, D. B. Fordyce, and G. F. Trebbi, J . A m . O i l C h e m . S o c . 4 1 : 231 (1964).
169. ICI Surfactants, T h e HLB S y s t e m , A T i m e - S a v i n g G u i d e t o E m u l s i f i e r S e l e c t i o n , Wilmington,
DE, 1992.
170. ICI Surfactants, P r o d u c t C a t a l o g u e , Brussels, 1993.
171. T. J. Mitsui, Y. Machida, and F. Harusawa, B u l l C h e m . S o c . J p n . 4 3 : 3044 (1970).
172. B. Derjaguin and L. Landau, A c t a P h y s i c o c h i m . 1 4 : 633 (1941).
173. E. Vervey and J. T. G. Overbeek, T h e o r y o f th e S t a b i l i t y o f L y o p h o b i c C o l l o i d s , Elsevier, Am
sterdam, 1948.
174. G. Bognolo, Proc. World Surfactants Congress, III C-3, pp. 185-199.
175. W. G. Cutler and E. Kissa (Eds.), D e t e r g e n c y : T h e o r y a n d T e c h n o l o g y , Marcel Dekker, New
York, 1986.
176. A. M. Schwartz, in S u rfa c e a n d C o llo id S c ie n c e (E. Matijevic, Ed.), W iley, New York, 1972,
Vol. 5, pp. 195-244.
694 Bognolo
177. H. Lange, in Solvent Properties of Surfactants Solutions (K. Shinoda, Ed.), Marcel Dekker, New
York, 1967, Chap. 4.
178. H. L. Benson, presented at 77th Annu. Am. Oil Chem. Soc. Meeting, Honolulu, 1986.
179. H. L. Benson, presented at 73rd Annu. Am. Oil Chem. Soc. M eet., Toronto, 1982.
180. K. W. Dillan, /. Am. Oil Chem. Soc. 61: 1278 (1984).
181. A. Waag, Chim. Phys. Appl. Prat. Ag. Surface C.R. Congr. Int. Deterg. 5th, Barcelona, 1968,
III, p. 143.
182. G. Bognolo, Polymeric surfactants, paper presented at the Colloid and Surfaces Chemistry Group
Meeting, Soc. Chem. Ind., Imperial College of Science and Technology, London, 12 Dec. 1990.
183. S. Talmage, Environmental and Human Safety of Major Surfactants, Soap and Detergent Associ
ation, Lewis Publishers, 1994, pp. 9 5 -1 0 2 .
184. G. M. Benke, M. N. Brown, M. J. Walsh, and R. B. Drotman, Food Cosmet. Toxicol. 15:
309-318 (1977).
185. R. Gingell and C. Lu, J. Am. College Toxicol. i0(4): 477-486 (1991).
186. J. Yam, K. A. Boorman, W. Broddle, J. Geiger, J. E. Heinze, Y. L. Lin, K. McCarthy, S.
Reiss, V. Sawin, R. I. Sedlack, R. S. Slesinski, and G. A. Wright, Surfactants: a survey of
short term genotoxicity testing. Food Chem. Toxicol. 22(9):761-769 (1984).
187. J. Am. Coll. Toxicol. 5(5) 1-82 (1984).
188. J. Am. Coll. Toxicol. 7(6) 881-910, (1988).
189. J. Am. Coll. Toxicol. 4(3) 65-121 (1985).
190. J. Am. Coll. Toxicol. 5(5) 415-454, (1986).
191. K. J. Bock, L. Huber, and P. Schorbel, Tenside 25 Surfact. Deterg. 25(2): 8 6 -9 8 (1988).
192. L. Kravetz, J. P. Salanitro, P. B. Dorn, and K. F. Guin., J. Am. Oil Chem. Soc. 68: 610 (1991).
193. W. Guhl and P. Gode, Tenside 26(282): 3-8 (1989).
194. T. G. Blease, J. G. Evans, L. Hughes, and P. Loll, in Defoaming (P. Garnett, Ed.), Marcel
Dekker, New York, 1993, p. 305.
195. C. G. Van Ginkel, C. A. Stroo, and A. G. M. Kroon, Tenside 30(3): 213-216 (1988).
196. M. Stalmans, H. Berenbold, J. L. Bema, L. Cavalli, A. Dillarstone, M. Franke, F. Hirsinger,
D. Janzen, K. Kosswig, D. Postlethwaite, T. Rappert, C. Renta, D. Scharer, K. P. Schick, W.
Schul, H. Thomas, and R. Van Sloten, European life-cycle inventory for detergent surfactants
production, Tenside 32(2): 8 4 -1 0 9 (1993).
197. W. Schul, F. Hir singer, and K. P. Schick, Tenside 32(2): 171-192 (1995).
198. F. Hirsinger and K. P. Schick, Tenside 32(2): 193-199 (1995).
199. M. J. Schick (Ed.), Nonionic Surfactants, Marcel Dekker, New York, 1968.
26_____________________
Lipids
Their Use in Personal Care Products
Keith Coupland and Julie A. Nichols

C r o d a U n iv e r s a l L t d ., H u l l , E n g la n d
I. INTRODUCTION
The aim of this chapter is to outline the use of lipids in cosmetics and personal care products.
It covers the types of lipids used, the reason for their use, and how they are formulated.
Cosmetics are defined as “something intended to adorn or beautify the body, especially the
face” [1]. The common perception of the word is decorative, and it has an implied female
bias. Since lipids are used in many products other than color cosmetics, the term “personal
care product” is preferred.
The European Cosmetics Directive [2] provides a general description of personal care prod
ucts that is the accepted European standard:
Any substance or preparation intended to be placed in contact with the various external parts of the
body (epidermis, hair system, nails, lips and external genital organs) or with the teeth and the mucous
membranes of the oral cavity with a view exclusively or mainly to cleansing them, perfuming them,
changing their appearance and for correcting body odors and/or protecting them or keeping them in
good condition.
Increasingly, new products are appearing on the market that extend beyond these defined
boundaries for personal care products into the verging area of pharmaceuticals. These products
have been termed “cosmoceuticals.”
Lipids or lipid derivatives are used in most, if not all, personal care products sold today.
They are found in emulsions such as creams and lotions as carriers for other lipophilic materi
als and as emollients.
II. LIPIDS USED IN PERSONAL CARE PRODUCTS

The emollient most commonly used by the personal care industry is hydrocarbon or mineral
oil. Its consumption is estimated to be around 700,000 tonnes per year (tpa) and is increasing
by 2% per year [3]. Despite a long history of safe use, concern about trace contents of
polycyclic aromatic hydrocarbons and a move toward renewable resources have resulted in
695
696 Coupland and Nichols
the increasing use of natural lipids. There is a distinct preference for vegetable-derived oils,
fats, and waxes.
Many personal care products contain ingredients with no real function other than to attract
the customer to buy the product. However, if some benefit is claimed such as “antiwrinkle”
or “moisturizer,” it must be substantiated in accordance with the European Directive 6th
Amendment [2] [Article 7al(0)].
A. Triacylglycerols
Most plant lipids used in personal care products are triacylglycerols, i.e., esters of glycerol
and carboxylic acids. The fatty acids are generally linear, even-numbered monocarboxylic
acids in the range They can be saturated, monounsaturated, or polyunsaturated. In
recent years it has become common to use a shorthand description of fatty acids indicating
the number of carbon atoms, the degree of unsaturation, and the position of any double bond
relative to the terminal methyl group. Using this scheme, y-linolenic acid (all-cA-octadeca-
6,9,12-trienoic acid) becomes 18:3 (n-6).
Table 1 lists five fatty acids contained in plant lipids and their plant sources. There are two
other sources for triacylglycerol lipids used in personal care products: animal fats and marine
oils. These are not considered as important as the plant oils both for reasons outlined earlier
and because the polyunsaturation of many marine oils gives rise to oxidative instability.
The formulator has many good reasons for incorporating lipids into personal care products.
1. Lipids are emollients, or superfatting agents.
2. They have market appeal and are a renewable source.
3. They are mineral oil or hydrocarbon free.
4. Secondary benefits include ultraviolet absorbance.
5. Lipids serve as solubilizers or vehicles for lipophilic ingredients, e.g ., botanical ex
tracts.
Emolliency describes the smooth lubricant effect of applying an oil to the skin. This is
particularly useful because dry skin is a common complaint caused by the environment, deter
gents, and disease. Removal of skin lipids by washing in detergent or soap is easily remedied
by applying an emulsion containing a lipid, which has a “superfatting” action on the skin.
Market appeal is an obvious benefit, and a label claim stating, for example, “Contains
Evening Primrose Oil,” is often sufficient to attract the customer and sell the product.
The use of lipids as a replacement or alternative to mineral oil is of interest because it was
oils, fats, and waxes that were used historically as cosmetics [4].
Several triacylglycerol lipids confer secondary benefits to the product such as ultraviolet
absorbance. Unsaturated oils have a weak absorbance at 210 nm, but certain lipids such as
shea butter contain ultraviolet absorbers in the unsaponifiable fraction of the oil. Shea butter
absorbs strongly at 271 nm.
Table 1 Fatty Acids Contained in Plant Lipids
Fatty acid Designation Source
Myristic 14:0 Coconut oil, palm kernel oil

Oleic 18:1 (n-9) Olive oil, hybrid sunflower seed oil
Linoleic 18:2 (n-6) Safflower seed oil
y-Linolenic 18:3 (^-6 ) Evening primrose oil, borage oil
Stearidonic 18:4 (/z-3) Blackcurrant seed oil
Lipids in Personal Care Products 697
Lipids may be used as solubilizers for other lipophilic components. Castor oil is a good
solvent for bromoacid dyes in lipsticks. Vegetable oils are also employed as the extractant
and carrier for plant or botanical extracts such as Calendula, Saint-John’s-wort, Carrot, and
Walnut.
In recent years it has become apparent that some triacylglycerol lipids are not simply
passive lipophilic vehicles providing emolliency. Several oils containing polyunsaturated fatty
acids of the n-3 and n-6 series are readily incorporated into skin lipids and are metabolized.
Dry skin is the result of a high rate of transepidermal water loss (TEWL). There are many
causes for this symptom, including environment, diet, and essential fatty acid (EFA) defi
ciency [5]. EFA deficiency caused by poor diet is rare in developed countries, but it can be
induced in disease conditions. A series of papers from the Unilever Laboratories demonstrated
that the reduction of TEWL by topical application of EFA-containing lipids was not simply
an occlusive effect but resulted from the repair of the impaired natural moisture barrier in the
skin. EFA deficiency causes severely dry skin in patients with chronic fat malabsorbance. The
cutaneous symptoms were corrected by applying an oil rich in linoleic acid [6]. This study
was extended [7] to demonstrate that TEWL caused by detergent insult could also be corrected
by cutaneous application of linoleic acid.
The repair of skin barrier function was also achieved by application of y-linolenate [18:3
(n-6)] but not dihomo-y-linolenate [20:3 (n-6)] [8]. Brod et al. [9] reported the use of an
emulsion containing blackcurrant seed oil, which is rich in y-linolenic acid containing triglyc
erides, to correct human dry skin deficiency.
B. Wax Esters
The second widely used group of lipids are known collectively as wax esters. They differ
from the triacylglycerols by having fatty acid esterified to fatty alcohol. There are several
sources in nature for both animal and vegetable wax esters. The best known of the wax esters
Table 2 Plant Lipids Used in Personal Care Formulations
Lipid Plant source Key fatty acids
Apricot kernel oil Prunus armeiaea 1 8 :1 /1 8 :2

Avocado oil Persea americana 18:1
Babassu oil Oribgnya speciosa 1 2 :0 /1 4 :0
Blackcurrant seed oil Ribes niger 1 8 :2 /1 8 :3 (n-6)^
Borage oil Borago officianalis 1 8 :2 /1 8 :3 (n-6)
Castor oil Ricinus communis 18:1 (OH)
Cocoa butter Theabroma cacao 1 8 :0 /1 8 :1
Coconut oil Cocus nucifera 1 2 :0 /1 4 :0
Evening primrose oil Oenothera biennis 18:2/18:3 (n-6)
Macadamia nut oil Macadamia tetraphylla 1 8 :1 /1 8 :2 /1 6 :1
Meadowfoam oil Limnanthes alba 20:1 (n-l5)
Olive oil Olea europa 18:1
Sesame oil Sesamum indica 1 8 :1 /1 8 :2
Shea butter Butyrospermum parkii 1 8 :0 /1 8 :U
Sunflower oil Helianthus annus 18 :2 /1 8 :1
Sweet almond oil Prunus amygdalus 1 8 :1 /1 8 :2
Wheatgerm oil Triticum speciosa 1 8 :2 /1 8 :P
Âlso contains stearidonic acid 18:4 (n -3 ).

’’ Contains cinnamate ester.
‘'High vitamin E content.
are lanolin (purified wool wax derived from sheep’s wool), jojoba oil (extracted from the
jojoba plant), and orange roughy oil (from the deep sea fish Hoplostethus atlanticus).
The composition of lanolin is surprisingly complex; it has a range of around 200 fatty acids
including linear acids (Cg-C38), branched-chain acids (C7-C41), and hydroxy acids (C1Q-C36).
These fatty acids are esterified or interesterified to a range of some 70 alcohols and sterols.
Wool wax alcohols are characterized by their high cholesterol content ( — 3 0 % ) .
The analysis and structural determination of lanolin are detailed in a series of papers by
Fawaz and coworkers [10-14].
Lanolin is an excellent water-in-oil emulsifier, skin conditioner, and emollient. These ef
fects have been quantified using skin image analysis, profilometry, dermal torque determina
tion, and cohesography [15].
C. Minor Lipids
Several other classes of lipids are used in personal care products in relatively small amounts
[16]. These include glycerol ethers, which are found in shark liver oil; phosphoglycerides;
and sphingolipids, particularly the ceramides. This latter class has excited interest among
cosmetic scientists following the discovery of the vital role played by ceramide in the perme
ability barrier of the mammalian epidermis [17]. Extraction of ceramide from tissue is expen
sive and gives variable results. The use of synthetic and semisynthetic products is expected
to increase.
III. REFINING LIPIDS FOR PERSONAL CARE PRODUCTS

As with any raw material for personal care products, quality is of extreme importance. Trace
impurities such as pesticide residues, oxidation products, thermal degradation products, pig
ments, and alkaloids may lead to problems with taste, odor, color, and irritation. The pro
cesses used to refine oils can themselves give rise to impurities. Although the use of chemical
bleaching gives very pale colored lipids, it leaves behind autodecomposition products.
For very high purity lipids, process-scale absorption chromatography is used [18,19]. This
has the advantage of removing polar lipids and pigments while leaving the substrate chemi
cally unchanged. Peroxide values of less than 1 and complete removal of autoxidation prod
ucts are possible. The resulting lipids are odorless, colorless, and tasteless.
There is a trend in the marketplace toward colorless oils and emulsions that are brilliant
white. The use of oils that are pale yellow gives off-colored emulsions, particularly water-in
oil emulsions, though the slight coloration may be tolerated because it is deemed a natural ar
tifact.
The expected shelf life of personal care products is usually 3 years. This includes the
physical stability of emulsions, clouding, precipitation, and color and odor development. This
entails that lipids with unsaturation must be adequately protected with antioxidants. Typical
antioxidants include vitamin E, ascorbyl palmitate, butylated hydroxytoluene (BHT), butyl-
ated hydroxyanisole (BHA), and octyl gallate.
IV. FORMULATING PERSONAL CARE PRODUCTS

There are many considerations in formulating personal care products, and despite increased
knowledge of the chemistry and physics of the systems formulation still remains a mixture of
art and science. Modem formulations can contain as many as 20 components, and optimizing
these systems is not trivial. Several examples of formulating with lipids are included in Ap
pendix 1.
Lipids Use in Personal Care Products 699
A. Emulsions
The majority of skincare products are in the form of emulsions. Formation of emulsions is
facilitated by the incorporation of surface-active emulsifiers. Emulsions may be oil-in-water
(o/w), in which the continuous phase is water, or water-in-oil, in which the continuous phase
is oil. This allows a wide variety of emulsion characteristics to be obtained by varying water/
oil ratios.
There are associated difficulties with the emulsification of vegetable oils, and formulation
optimization is not always easy. One guideline technique often used is the HLB scale.
B. HLB Scale
The required HLB (hydrophilic-lipophilic balance) of a vegetable oil or emollient can help in
identifying a suitable emulsifier, typically a blend of emulsifiers. However, the required HLB
of the lipid does not mean that any surfactant of that HLB will produce the required emulsion.
The chemical type or class of the surfactant used is at least equally as important, and a range
around the required HLB will probably be needed. The lipophile of the surfactant needs to
exhibit the greatest affinity for the oil-phase droplets. Similarly, the “like attracts like” concept
may aid in surfactant choice; unsaturated lipophiles are attracted to unsaturated oils, and satu
rated lipophiles to saturated oils [20]. The surfactant type can also affect the texture, feel,
viscosity, stability, and so on of an emulsion.
It should be noted that the HLB scale is limited to nonionic emulsions, and the use of
anionic and cationic emulsifiers should not be ignored. A selection of required HLBs for o/w
emulsions is shown in Table 3. In most cases other more complex factors have to be taken
into consideration, but the HLB scaling provides a good starting point.
Other important points to consider are (as with other emulsions)
1. Ratio of oil phase to emulsifier
2. Ratio of oil phase to aqueous phase
3. Composition of the oil phase
4. Composition of the aqueous phase
5. Emulsifying method
6. Emulsifying equipment
7. Emulsifying temperature
All of these factors can affect the production and ultimately the stability of an emulsion.
Table 3 Required HLB
Required HLB
Material (o/w)
Mineral oil 1 0 - 1 2
Cottonseed oil 7 -5
Castor oil 14
Cocoa butter 6
Com oil 10
Jojoba oil 6 -7
Palm oil 10
Rapeseed oil 6
Soybean oil 6
Source: Ref. 20.

C. Polarity Index
Another interesting consideration is the so-called polarity index [21], a measure of the interfa
cial tension of an oil against water. The more polar vegetable oils often afford better emol-
liency than nonpolar oils such as hydrocarbon or mineral oils. Polar oils offer improved skin
surface distribution and solvency but are not the easiest materials to emulsify. The interfacial
tension is affected by the emulsifiers used, which influence emulsification by their mobility,
diffusion, and interfacial orientation.
The polarity index values for some common oils are listed in Table 4.
D. Mono-, DI-5 and Triglycerides

Mixtures of mono-, di-, and triglycerides esterified with a variety of fatty acids allow esters
to be obtained that function as emollient, stabilizer (viscosity modifier), and coemulsifier [22].
Varying the fatty acid composition and percentage of the three glyceride types can produce
the desired end emulsion with the minimum number of components used in a formulation.
This is the subject of a recent patent by Frank Johannsen [23].
E. Water-in-Oil Emulsions
Typically w/o emulsions are more difficult to stabilize than o/w emulsions. The stabilization
of w/o emulsions is difficult with conventional mineral oil, and the same is true of water-
in-triglyceride oil emulsions. In addition to optimization of the emulsifier in such w/o triglyc
eride emulsions, the quantity of fat crystals (solid triglycerides in the vegetable oil) is im
portant. The lipid crystals can thicken the external oil phase and absorb at the water/oil in
terface, and they may form bridges between the discrete water-phase droplets. By varying the
quantity and type of lipid crystals, stability or instability can be induced [24]. Other classi
cal thickeners or rheology modifiers, e.g., clays/resins and particulate matter, may also be
used.
F. Formulations
Natural lipids have a long history of use as emollients in personal care, originally for the
simple oiling of skin and hair. Advances in technology allow their incorporation into a whole
Table 4 Polarity Indices ^
Common oils (INCI) Polarity index (mN/m)
Mineral oil 38.3

Cetearyl octanoate^ 28.6
Caprylic/capric triglyceride 21.3
Peanut oil 20.5
Almond oil 20.3
Sunflower seed oil 19.3
Castor oil 16.9
^Substances below 30 mN/m are classified as polar; those above 30

mN/m are nonpolar.
^Cetearyl octanoate is a mixture of cetyl (16:0) and stearyl (18:0)
octanoates.
host of personal care formulations for a multitude of reasons as previously discussed. A selec
tion of such formulations supplied by Croda Chemicals Limited is given in Appendix 1.
Further details and information can be obtained from the company. Details of other suppliers
are given in Appendix 2.
V. CONCLUSION
The use of lipids in personal care products is supported by many of the special or unique
features they can offer the formulating chemist and consumer. They may be employed for
functional benefits and provide an attractive marketing concept. Advances in technology and
identification of novel lipid materials will ensure future interest in these materials.
APPENDIX 1 - FORMULATIONS^
1. Moisturizing Evening Primrose Facial Oil C1610. This simple light facial moisturiz
ing oil employs Evening Primrose oil, a material rich in y-linoleic acid (GLA), to provide
moisturizing effects and an attractive marketing angle. (See following data.)
A reconstituted vegetable oil, caprylic/capric triglyceride, is the base emollient vehicle and
facilitates spreading the preparation over the skin.
Moisturizing Evening Primrose Facial Oil C1610
Ingredient Supplier Percent (w/w)
Crossential EPO (super-refined evening primrose oil) Croda 1.00

Crodamol GTCC (caprylic/capric triglyceride) Croda to 100
Tocopheryl acetate Roche 6.0
BHT — 0.05
Perfume, preservatives, color — qs^
^qs = quantum sujficit (as much as will suffice to produce the desired effect).
METHOD OF MANUFACTURE. Incorporate BHT and preservative into Crodamol GTCC with
gentle heating. Stir to cool, continue stirring, and add remaining ingredients.
2. Avocadin Cream Cl776M2. This preparation employs Avocado oil unsaponifiables
(a material rich in plant sterols; sitosterol, campesterol, and stigmasterol), to confer superior
moisturization to the skin. This emollient o/w preparation uses nonionic ethoxylated fatty
alcohols (Ceteareth 20) and sorbitan esters (Polysorbate 60), a small amount of anionic Tri
ethanolamine Stearate stabilized with Glyceryl Stearate, and Cetearyl alcohol. (See Table 5.)
METHOD OF MANUFACTURE. Heat oil and water phases separately to 6 5 -7 0 °C . Add water
phase to oil phase while stirring. Stir to cool, perfume at 40°C.
3. Carrot Oil Cream C l812. This light moisture cream employs Carrot Oil Extra to aid
U V absorption and condition the skin. Peanut oil is used to extract the carrot. The flexible
emulsification system of Sorbitan Stearate and Polysorbate 60 is used. (See Table 6.)
METHOD OF MANUFACTURE. Heat oil phase and water phase separately to 65-70°C. Add
water phase to oil phase while stirring. Stir to cool, perfuming at 40-45°C. Fill off at 30°C.
Âvailable from Croda Chemicals Limited.

Table 5 Avocadin Cream C1776M2
Avocadin (avocado oil unsaponifiables) Novarom 5.00

Crodamol IPP (isopropyl palmitate) Croda 7.50
Crodamol CAP (cetearyl octanoate) Croda 2.00
Silicone 200/100 cS (dimethicone) Dow Coming 0.50
Crodacol CS90EP (cetearyl alcohol) Croda 2.50
Crosterene SA4310 (stearic acid) Croda 0.50
Mineral oil (25 cS at 25°C) — 5.00
Cithrol GMS N/E GE0803 (glyceryl stearate) ^ Croda 1.00
Cosmowax D (cetearyl alcohol and ceteareth 20) Croda 2.10
Crillet 3 (Polysorbate 60) Croda 2.50
Deionized water —
to 100
Propylene glycol — 2.00
Triethanolamine 99% — 0.20
Perfume, preservative, color — qs
^GMS = glycerol monostearate.
4. Highly Emollient Lanolin Cream C1440. Ideal for rough, dry skin, this enriched
moisturizing cream employs a combination of lanolin and hydrogenated lanolin (see Table 7).
This cream affords a degree of occlusivity. Lanolin acts as an emulsifier and stabilizer.
METHOD OF MANUFACTURE. Heat oil and water phases separately to 70°C. Add water phase
to oil phase with stirring. Stir to 30-35°C , then pass over triple roll mill. Fill off.
5. Neck Cream C1267. Formulation C l267 is a highly emollient-rich cream designed
for use on the neck. The apricot kernel oil, squalene, and super sterol ester all contribute to
its performance. (See Table 8.)
METHOD OF MANUFACTURE. For formula 1, predisperse Collasol in water (10% of total).
Hydrate Carbopol in warm water, 60-65°C. Add remaining water-phase ingredients except
triethanolamine, and heat to 65-70°C . Heat oil phase to 65-70°C and add to water phase
while stirring. Allow to cool, adding predispersed Collasol phase for formula 1 at 30-35°C.
Fill off.
Table 6 Carrot Oil Cream C1812
Crodamol GTCC (caprylic/capric triglyceride) Croda 13.00

Crodamol OC (octyl cocoate) Croda 3.00
Crodamol SS (cetyl esters) Croda 3.00
Crodamol CS (cetearyl alcohol) Croda 3.00
Carrot Oil Extra (peanut oil, carrot extract, and isopropyl myristate) Novarom 4.00
Crill 3 (sorbitan stearate) Croda 2.00
Deionized water — to 100
Croderol GA7000 (glycerine) Croda 3.00
Table 7 Lanolin Cream C l440
White microcrystalline wax (No. 3520) — 3.40

Satulan (hydrogenated lanolin) Croda 2.50
Corona pure new lanolin (lanolin) Croda 9.50
Light mineral oil (25 cS at 25°C) — 16.00
White petroleum jelly, continental grade no. 3 — 2.50
BHT — 0.01
Magnesium sulfate — 0.70
6. Night Cream C l520. Rich skincare properties are given by the inclusion of orange
roughy oil (wax ester). The emulsification system for this o/w cream is PEG-24 stearate, an
ethoxylated fatty acid, stabilized with glyceryl stearate and cetearyl alcohol. (See Table 9.)
METHOD OF MANUFACTURE. Predissolve Collasol in water (10%). Heat oil and water phases
separately to 65-70°C. Add water phase to oil phase while stirring. Stir to cool, perfuming
at 45°C and adding Collasol solution at 30-35°C.
7. Physical Block Sunscreen Cream C l643. A highly protective water-resistant sun

screen that uses microfine titanium dioxide as the physical UV blocker. Caprylic/capric tri
glyceride, a reconstituted vegetable oil derived in combination with shea butter, provides
Table 8 Neck Cream C l267
Percent (w/w)
Formula 1
(with Formula
Ingredient Supplier Squalene) 2
Polawax GP200 (nonionic emulsifying wax) Croda 4.00 4.00

Super-refined apricot kernel oil Croda 5.00 15.00
Squalene — 10.00 —
Crodamol MM (myristyl myristate) Croda 2.00 2.00

Super sterol ester (C l0 -3 0 cholesterol/lanosterol esters) Croda 2.00 2.00
Crosterene SA4310 (stearic acid) Croda 3.00 3.00
BHA — 0.05 0.05
Dow Coming 580 Wax (stearoxytrimethylsilane and
stearyl alcohol) Dow Coming 2.00 2.00
Deionized water — to 100 to 100
Croderol GA7000 (glycerine) Croda 5.00 5.00
Triethanolamine — 1.50 1.50
Carbopol 941 (Carbomer 941) BF Goodrich 0.30 0.30
Deionized water — 10.00 —
Collasol (1%) Croda 1.00 —
Perfume, preservative, color — qs qs

Table 9 Night Cream C1520

Super-refined orange roughy oil (orange roughy oil) Croda 8.00

Silicone fluid 200/300cS (dimethicone) —
1.00
Crodalan LA (cetyl acetate and acetylated lanolin
alcohol) Croda 1.00
Crodamol MM (myristyl myristate) Croda 3.00
Crodet S24 (PEG-24 stearate) Croda 3.00
Crodacol CS90 EP (cetearyl alcohol) Croda 2.00
Cithrol GMS N/E GE0803 (glyceryl stearate) Croda 7.00
Propylene glycol — 3.00
Collasol 3% (soluble collagen and water) Croda 1.00
Deionized water (for collasol) — 10.00
Perfume, preservative — qs
emollience. This o/w system uses an emulsification system of a nonionic emulsifying wax and
triethanolamine stearate (soap) stabilized with glyceryl stearate. (See Table 10.)
METHOD OF MANUFACTURE. Heat oil phase (including Titanium Dioxide MT-IOOT) and wa
ter phase separately to 80-85°C. Add water phase to oil phase with vigorous agitation. Add
triethanolamine. Continue stirring until temperature reaches approximately 55°C, then pass
through suitable homogenizing equipment.
8. W/O Cream with Zinc Oxide C545. This protective w/o baby cream uses lanolin
alcohols as the low HLB emulsifier. Lanolin acts as a stabilizer as well as a skincare emol
lient. (See following data.)
W/O Cream with Zinc Oxide C545
Super Hartolan (lanolin alcohols) Croda 2.00

Corona Pure New Lanolin (lanolin) Croda 4.50
Mineral oil (70 cS at 25°C) — 17.00
White petroleum jelly, continental grade No. 3 — 13.30
BHT — 0.01
Zinc oxide —
7.00
Perfume, preservative, color —
qs
METHOD OF MANUFACTURE. Heat Water and oil phases to 70°C. Add water phase to oil
phase under continual stirring. A planetary mixer with side scraping arms is preferred. Add
the zinc oxide at 60°C and continue stirring. Add perfume at 45°C. Homogenize at 30-35''C
using suitable equipment, e.g., colloid or roller mill.
9. Avocadin Facial Scrub Cream C l817. This traditional detergent cleanser contains a
combination of almond oil and Avocadin to aid moisturization and improve skin feel (Table
Table 10 Sunscreen Cream C l643
Crodamol MM (myristyl myristate) Croda 1.5

Silicone 200/200 cS (dimethicone) Dow Coming 7.0
Cithrol GMS N/E (glyceryl stearate) Croda 2.5
Crosterene SA4310 (stearic acid) Croda 3.0
Antaron V220 (PVP eicosene copolymer) ISP 0.5
Polawax GP200 (nonionic emulsifying wax) Croda 2.0
Crodamol OHS (octyl hydroxystearate) Croda 5.0
Super-refined shea butter Croda 5.0
Crodamol GTCC (caprylic/capric triglycerides) Croda 3.0
Triethanolamine 99% — 0.9
Titanium Dioxide MT-IOOT (titanium dioxide) Mitsubishi 10.0
Perfume, preservatives, color — qs
11.) It may be modified into a facial scrub by addition of exfoliating agents such as polymer
granules or crushed seeds or stones.
METHOD OF MANUFACTURE. Hydrate the Veegum in water. When fully hydrated, add the
remaining water components (except Hydromond) and heat to 65-70°C. Heat oil phase to
65-70°C. Add water phase to oil phase while stirring. Stir to cool, adding perfume and
Hydromond at ~40°C.
10. Two-Layer Conditioning Shampoo with Jojoba Oil C1569M1. This unusual sham
poo (Table 12) employs jojoba oil in the upper opaque/cream layer for extra conditioning and
market appeal.
Table 11 Avocadin Facial Scrub Cream C1817
Cithrol GMS A/S ES0743 (glyceryl stearate and

PEG-100 stearate) Croda 6.00
Dehydag Wax E (sodium cetearyl sulfate) Henkel 7.50
Polawax GP200 (nonionic emulsifying wax) Croda 1.00
Crovol M40 (PEG-20 com glycerides) Croda 2.00
Super-refined sweet almond oil Croda 1.00
Volpo L3 Special (C 1 2 - C 1 3 Pareth-3) Croda 2.50
Avocadin (avocado oil unsaponifiables) Novarom 2.00
Deionized water — to 10 0
Veegum Regular (magnesium aluminum silicate) Vanderbilt 1.00

Crodasinic LS35 (sodium lauroyl sarcosinate) Croda 5.00
Empicol LZ (sodium lauryl sulfate) Albright & Wilson 7.50
Hydromond (hydrolyzed almond protein) Croda 0.50
Table 12 Two-Layer Conditioning Shampoo C1569M1
Empicol ESB3 (sodium laureth sulfate) Albright & Wilson 30.00

Crodasinic LS35 (sodium lauroyl sarcosinate
and water) Croda 8 .0 0
Crodafos SG (PPG-5-ceteth-lO phosphate) Croda 1 .0 0
Crotein Q (hydroxypropyl trimonium hydrolyzed

collagen) Croda 0.50
Incromine Oxide C (cocamidopropylamine oxide
and water) Croda 0.50
Incronam 30 (cocamidopropyl betaine) Croda 5.00
Solan X (PEG-150 lanolin) Croda 2 .0 0
Empilan CDE (cocamide DEA) — 3.00

Deionized water — to 1 0 0
Jojoba oil — 1 .0 0
Crodamol OSU (dioctyl succinate) Croda 4.00

METHOD OF MANUFACTURE. Dissolvc Crotein in water at 60°C. Add Crodafos while stirring;
warm if necessary. Add remaining water-soluble ingredients, stirring until dissolved; then add
emollient oils. Stir to disperse. Fill off while stirring.
11. Wheat Germ Shampoo C l535. Incorporation of wheat germ oil into this clear sham
poo system provides a suitable treatment for dry hair and an attractive marketing concept.
Crillet 1 and Crodafos SG are used in combination as the solubilizing agents for the oil, as
the detergent base alone is not sufficient when used at a level o f 1%. (See Table 13.)
METHOD OF MANUFACTURE. Solubilize w h ea t g e rm oil and perfume in Crillet and Crodafos.

Combine the rest of the ingredients, and when homogeneous add Crillet-Crodafos premix.
Stir until clear and fill off.
Table 13 Wheat Germ Shampoo C1535
Empicol ESB3 (sodium laureth sulfate) Albright & Wilson 50.00

Incronam 30 (cocoamidopropyl betaine) Croda 5.00
Super-refined wheat germ oil Croda 1 .0 0

Crodafos SG (PPG-5-ceteth-lO phosphate) Croda 3.00
Triethanolamine 99% —
to pH 6 .5 -7 .0 0
Deionized water — to 1 0 0
Sodium chloride — a
^To required viscosity.

Table 14 Lipstick C l496
Fluilan (lanolin oil) Croda 2 0 .0 0
Novol (oleyl alcohol) Croda 15.00

Candelilla wax — 1 0 .0 0
Carnauba wax — 1 0 .0 0
BHT — 0.05
Polytrap 603 (acrylates copolymer) Witco 0 .2 5 -0 .5 0
Castor oil — 44.60
Pigment
Ariabel Ruby 300503 Warner Jenkinson 1.75
Ariabel Saffron 300508 Warner Jenkinson 0.75
Ariabel Magenta 300522 Warner Jenkinson 2.25
Ariabel Pearl White 300330 Warner Jenkinson 1 0 .0 0
Titanium Dioxide 300309 Warner Jenkinson 1.25
12. Lipstick C1496. Lipsticks contain a wide variety of oils, fats, and waxes that impart
properties such as gloss, moisturization, and ease of application. This formulation (Table 14)
uses castor oil as the primary emollient, blended with lanolin oil and oleyl alcohol to reduce
drag and improve application. Wax combinations impart structure to the stick, preventing
bleeding, sweating, and blooming if correctly formulated.
METHOD OF MANUFACTURE. Mill pigments in castor oil, add waxes, and combine with stir
ring over heat at 80°C. When homogeneous, stir slowly to deaerate. Stir to cool, perfuming
and filling into molds at 55°C.
APPENDIX 2. OTHER SUPPLIERS

1. F Hoffman - La Roche, Novarom c/o Croda
CH-4002, Basle, Switzerland. Chemicals Ltd, Cowick Hall,
(Tel 00 4161 688 n i l ; Snaith, Goole, North Humberside,
Fax 00 4161 688 9391) D N 14 8NF U K .
(Tel 01405 860551;
Fax 01405 86025)
3. Dow Coming Europe, BF Goodrich Chemical Europe,
62 Rue General de Gaulle, 742 Rue de Verdun,
1310 La Hulpe, Brussels, Belgium. B-1130 Bmssels, Belgium.
(Tel 00 322 655 2 1 1 1 ; (Tel 00 32 2 247 19 11;
Fax 00 322 655 2001) Fax 00 32 2 247 1991)
5. ISP Europe Ltd, Mitsubishi Corp (UK) Ltd,
40 Alan Turing Road, Bow Bells House, Bread Street,
Surrey Research Park, London E C 4M 9BQ, U K .
Guildford, Surrey G U 2 5 Y F , U K . (Tel 0 171 822 0192;
(Tel 01483 301757; Fax 0 171 822 0082)
Fax 01483 302175)
7. Henkel KG A, 8. R. T. Vanderbilt Company Inc,

Postfach 1100, 30 Winfield Street (P.O. Box 5150),
D-4000 Dusseldorf 1, Germany. Norwalk, CT 06856-5150.
(Tel 00 49 211 7970; (Tel 203 853 1400;
Fax 00 49 211 7987696) Fax 203 853 1452)
9. Albright & Wilson Ltd 10. Witco Corporation,
(Detergents Group), 10 Rue Cambaceres,
Whitehaven Works, 75008 Paris, France.
P.O. Box 15, Cumbria CA28 9QQ (Tel 00 33 142 65 9903;
UK. Fax 00 33 142 65 6761)
(Tel 01946 693131;
Fax 01946 68181)
11. Warner Jenkinson Europe Ltd,
Old Meadow Road,
Hardwick Industrial Estate,
Kings Lynn,
Norfolk PE30 4JJ, UK.
(Tel 01553 770550;
Fax 01553 770707)
REFERENCES
1. Oxford Dictionary.
2. EEC Cosmetics Directive, updated version, August 1994, Dir. 7617681EEC.
3. Anon., Natural selection. Soap, Perfumery, and Cosmetics, November 1994, pp. 4 5 -4 7 .
4. G. De Navarre, Oils and fats the historical cosmetics, J. Am. Oil. Chem. Soc. 55: 435 (1978).
5. G. O. Burr and M. M. Burr, On the nature and role o f the fatty acids essential in nutrition, J.
Biol. Chem. 82: 345 (1929).
C. Prottey, P. J. Hartop, and M. Press, Correction of the cutaneous manifestations of essential
fatty acid deficiency in man by application of sunflower seed oil to the skin, J. Invest. Dermatol.
64: 228 (1975).
7. C. Prottey, P. J. Hartop, J. G. Black, and J. I. McCormack, The repair of impaired epidermal
barrier function in rats by the cutaneous application of linoleic acid, Br. J. Dermatol. 94: 13
(1976).
8. P. J. Hartop and C. Prottey, Changes in transepidermal water loss and the composition of epider
mal lecithin after applications of pure fatty acid triglycerides to the skin of essential fatty acid-
deficient rats, Br. J. Dermatol. 95: 255 (1976).
9. J. Brod, H. Traitler, A. Studer, and O. De Lacharriere, Evolution of lipid composition in skin
treated with blackcurrant seed oil, Int. J. Cosmet. Sci. 10: 149 (1988).
10. F. Fawaz, C. Miet, and F. Puisieux, Analyse des pommades, des huiles et des cires. XII. Etude
de la composition chimique de la lanoline. 1. Examen préliminaire et fractionnement de la l ’en-
chantillon étudié, Ann. P harm. Fr. 31: 63 (1973).
11. F. Fawaz, M. Chaigneau, and F. Puisieux, Analyse des pommades, des huiles et des cires. XIII.
Composition de la lanoline. 2. Etude des acides non hydroxyles de la lanoline totale et des differ
entes fractions, Ann. P harm. Fr. 31: 217 (1973).
12. F. Fawaz, C. Miet, and F. Puisieux, Analyse des pommades, des huiles et des ceres. XIV. Com
position de la lanoline. 3. Etude des acides hydroxyles de la lanoline totale des differentes frac
tions, Ann. P harm. Fr. 32: 59 (1974).
13. F. Fawaz, M. Chaigneau, and F. Puisieux, Analyse des pommades, des huiles et des cires. XV.
Composition chimique de la lanoline. 4. Etude des alcools aliphathiques de la lanoline totale et
des ses differentes fractions, Ann. P harm. Fr. 32: 215 (1974).
14. F. Fawaz, M. Chaigneau, and F. Puisieux, Analyse des pommades, des huiles et des cires. XVI.
Composition chimiques de la lanoline. 5. Etude des sterols et des alcools triterpeniques de la
lanoline totale et des differentes fractions, Ann. P harm. Fr. 32: 301 (1974).
15. J. Thewlis, The emollience of lanolin, Fragr. J. Jpn. 9: 108 (1992).
16. K. Coupland, Natural lipids— valuable raw materials in cosmetics, Croda, internal report.
17. D. T. Downing, Lipid and protein structures in the permeability barrier o f mammalian epidermis,
J. Lipid. R es. 33: 301 (1992).
18. K. Coupland, Super-refining natural lipids for pharmaceutical adjuvants, Pharm. Manuf. Rev . ,
March 1990.
19. K. Coupland and N. A. Langley, Modifying natural oils by adsorption chromatography, in Oils
and Fats in the Nineties (V. K. S. Shukla and F. D. Gunstone, Eds.), Denmark, 1992, p. 176.
20 . ICI Americas, Inc. The HLB System, ICI publication.
21 . N. Lawson and G. Dahms, Formulating better cosmetics, Manuf. Chem., April 1985.
22 . H. Traitler and H. Winter, UK Patent GB 2 118 567 (1984).
23. F. Johannsen, Fat compositions and their use in cosmetic and pharmaceutical emulsion products,
Int. Pub. No. WO 92/16184, October 1992.
24. D. Johansson, B. Bergenstahl, and E. Lundgren, Water-in-triglyceride oil emulsions. Effect of fat
crystals on stability, J. Am. Oil Chem. Soc., 72(8): 939 (1995).
27
The Use of Oils and Fatty Acids in Paints and
Surface Coatings
John Bentley*
Consultant, Buckinghamshire, England
I. INTRODUCTION
This chapter describes the use of oils and fatty acids in what are generally referred to as
surface coatings, and in this context, because they use essentially the same range of materials,
we include inks. Surface coatings are generally seen as those materials applied to substrates
ranging from wood and paper to a variety of metals, plastic, and many composite assemblies.
They generally have a dual role, which is to protect and to decorate, the latter including an
ability to disguise. The protective role is that of shielding the substrate from such environmen
tal agents as radiation, moisture, and oxygen and possibility from more aggressive attackers
such as atmospheric SO2 in industrial environments. In addition to its role in decoration, paint
may disguise inferior construction materials or even prevent recognition of the object, as with
camouflage and IR reflective paints. Inside this spectrum of uses, inks may be seen as a subset
of produets, eharacterized as being more highly pigmented and used as thin coatings laid
down in a precise manner, where properties of color and hiding power are paramount.
Paints consist of three principal components: binder, pigment, and solvent [1]. The first
two are the permanent constituents, with the binder providing the adhesion and cohesion,
keeping the pigment within the coating and ensuring that the paint remains attached to the
substrate. Pigments provide color and opacity. Solvents are present to aid in manufacture and
application. Varnishes are nonpigmented coatings. Binders have evolved through the ages
from the initial range of naturally available gums and resins such as casein, tree, and fossil
resins, through to a vast range of carefully designed synthetically based polymeric materials.
Of all the natural substances used, oils have played a dominant role over at least the last 200
years, with various blending and enhancing modifications. Latterly, they have been chemi
cally combined into a range of polymer compositions, as will be described. Although the
overall use of oils has decreased in recent years, a number of reasons make prediction of their
imminent demise premature, as will be discussed later.
Previously with ICI Paints, Slough, England.

711
712 Bentley
This chapter gives an overview of the size and breakdown of the paint industry worldwide
and its use of oils and fatty acids, as far as can be deduced. After an overview, the use of
oils in modem surface coatings and the chemistries involved are discussed in some depth.
This includes the impact of legislation and environmental concerns as referred to above and
attempts a brief look into the future. It is not possible here to cover the full chemistry or
composition of all surface coatings, and for that the reader is referred elsewhere [2,3]. Older
compositions are also not covered in as much depth where their use has significantly dimin
ished or ceased, but again references give more information for those interested. Hence, the
reactions, processes, and products discussed are principally those connected with the use of
oils or fatty acids in coatings or resins in current or recent manufacture [4].
II. OVERVIEW AND ECONOMICS OF THE

COATINGS INDUSTRY
The world coatings industry is split by volume on a geographical basis [5] as follows:
North America 22%

Western Europe 24%
Eastern Europe 20%
Asia 13%
Japan 10%
South America 8%
Rest of world 3%
Separately sourced figures give the total western European coatings market for 1993 as
5,854,000 tonnes, with a value of E C U 15 billion and a global market share of 31%, em
ploying 110,000 people [6]. European printing ink production for 1993 is reported to have
been 679,000 t with a value of ECU 2,547,000. The total U.S. paint and coatings market for
1993 has been put at 1100 million gallons US (—4,200,000 t) [7]. The Japanese market for
that year was 1,956,000 t including thinners [8]. The total world surface coatings output can
hence be estimated at around 20 million tpa (tonnes per annum).
Of the western European market above, the western part of Germany contributed just over
one-third of the total, followed by Italy, France, and the United Kingdom with 18.5%,
16.6%, and 13%, respectively. To add perspective to these figures, in the United Kingdom,
an industrial size comparison, taking the size of the total paint, varnish, and printing ink
industry at a level of 100, has plastic packaging at 115 and glass and glassware at 110 [9].
The stationery market follows at 95, and perfume, cosmetics, and toiletries at 90.
The markets for paint are variously subdivided, with major sectors being “architectural,”
meaning the interior and exterior of land-based structures; product or original equipment man
ufacturer (OEM), embracing the largest industrial sectors, including automotive, coil, can,
and appliance manufacturers; and special-purpose (depending on which statistics are used),
which includes car refinishing, powder coatings, road marking paints, and inorganic coatings.
Since the industry still uses, and also blends for resale, large quantities of organic solvents,
solvent sales are significant as both visible and hidden parts of the industry totals, not always
separable from the figures given.
In the United States the market split is 50% architectural, 30% product (OEM), and 20%
special-purpose [7]. The European decorative (architectural) market has similarly been esti
mated to be 50% of the total. It should be noted that some areas, notably automotive and can.
Oils and Fatty Acids in Paints and Surface Coatings 713
are fast becoming world markets, with a few manufacturers taking large quantities of material
to a global specification, also from a limited range of suppliers. By contrast, the large archi-
tectural/decorative market, though increasingly dominated by large suppliers, sells in small
pack sizes to users who may often be individual homeowners, with product tailored both to
local conditions and to fashions that may be shaped by advertising.
The polymer or resin systems now used are wide and varied, and the relative importance
of the different types continues to shift. Aqueous addition polymer dispersions made by emul
sion or dispersion polymerization dominate in volume terms and find application in all mar
kets. Nevertheless, oil-modified alkyd resins, discussed in detail in this chapter, are the next
highest in volume. These two resin types have this position from their use in architectural/
decorative paints. Other mixed aqueous polymer dispersions, partly or principally epoxy-
based, have achieved significance in automotive primer and can coatings. More traditional
industrial coatings now include a range of solvent-borne polymers including oil-modified
alkyds, polyesters, acrylics, epoxies (including epoxy esters), and phenolics, with additional
cross-linking resin including melamine formaldehyde and urethane polymers. Non-oil-con
taining resins are not discussed further in this chapter.
Oil use figures for the ink and coatings industries are not directly available but may be
estimated from partial breakdowns of resin use by type and from general knowledge of these
coatings. While trends in decorative paints continue toward water-borne emulsion paint, alkyd
paint still accounts for 25-35% of this sector, varying with country. Thus 16% of total Euro
pean paint and ink output is estimated still to be based on alkyd resins, of which the predomi
nating decorative alkyd typically has an oil content of around 70% based on the solid resin;
in Japan, 18% of resin production is oil-based. A recent survey in the United Kingdom [10]
estimated that 80,000 t of alkyd are produced annually, requiring 48,000 t of oil. This would
imply a European use of around 400,000 t for coatings. The European demand for fatty acids
for the manufacture of alkyds and derivatives is estimated at around 120,000-160,000 tonnes
as part of the above total. In Europe, consumption of vegetable oils for printing inks is
currently 12,000-15,000 tpa, predicted to rise to possibly as much as 25,000 tpa by 1997
[ 11].
III. OIL USE IN PAINTS

Table 1 shows the principal oils now used in surface coatings, in probable order of their use,
and also their fatty acid content. Of those oils or fatty acids listed, linseed, soybean, castor,
and coconut oils are all derived from seeds or nuts; tall oil (in fact available only as fatty acid
and not combined with glycerol) is a by-product of paper manufacture and derives its name
from the Swedish name for pine. Dehydrated castor oil, included in order to give its fatty
acid content here for comparison, is derived from castor oil and is described in a later section.
Oils used in the coatings industry will always have been refined, by processes described
elsewhere in this book, to remove waxes, free fatty acids, lecithins, and other natural contami
nants, because the presence of those substances would in almost all cases have deleterious
effects on the derived coating resins. Tall oil in particular is now used almost exclusively in
distilled form, in order to free it from any rosin content [12,13]. While all except tall oil
occur naturally as a triglyceride, free fatty acids are available from almost all oils as alterna
tive reactants to extend formulating flexibility. The use of both oils and fatty acids in coating
resins is described later.
Oils are classified as drying, semidrying, or nondrying, depending on whether the oil on
its own has sufficient unsaturation in its total fatty acid chains to autoxidize fully to a dry
film. Linseed oil, with over 60% of its fatty acids doubly or triply unsaturated, is classed as
714 Bentley
Table 1 Oils Used in Coatings and Their Fatty Acid Composition
Component, wt %
Saturated Oleic 9,12- 9,12,15- Other

Oil acid acid Linoleic acid Linolenic acid acid
Soybean oil 14 2 2 -2 8 5 2 -5 5 5 -9 0
Tall fatty acid 3 3 0 -3 5 3 5 -4 0 2 -5 10-15^^
Linseed oil 10 2 0 -2 4 14-19 4 8 -5 4 0
Castor oil 2 -4 6-8 3 -6 0 85-8 7 '’
Castor oil, dehydrated 2 -4 6-8 4 8 -5 0 0 4 0 -4 2
Coconut oil 8 9 -9 4 6-8 0-2 0 0
'‘Conjugated and pinolenic (octadeca-5,9 ,12-trienoic) acids, amounts dependent on source and method of refinement.
'’ Principally ricinoleic (octadeca-12-hydroxy-9-enoic) acid.
‘'Conjugated 9,11-linoleic acid.
a drying oil, whereas soybean, tall, and dehydrated castor oil are generally classified as semi
drying. Castor oil and coconut oil are nondrying. The concepts of drying and autoxidation are
discussed later in this chapter. In all of these oils, with one exception, the saturated fatty
acids are mainly palmitic and stearic acids; the exception is coconut oil, which contains very
high amounts of lauric acid.
Many other oils have been, and in various parts of the world still are, used in coatings.
Historically tung oil (or China wood oil) was used extensively in oleoresinous materials.
However, it has not been generally available in good supply for a number of years; in any
case it is less favored because of its highly conjugated unsaturation and hence tendency to
yellow in derived resins. Sunflower and safflower oils have significant use and are similar to
soybean oil in application. Other oils are mentioned in the following sections.
IV, TERMS USED IN THE COATINGS INDUSTRY AND

FORMULATING PRINCIPLES
Since this chapter uses terms specific to the industry, it is essential to define some of these
now to avoid confusion. The term “surface coating” could be very widely embracing, but in
our context it refers to essentially organic paints or finishes. Varnishes are unpigmented coat
ings, so the substrate will at least partly show through. They are used on wood and over
pigmented coatings to provide additional protection. The term “lacquer” may confuse, but to
the technologist it normally refers to those coatings that form a film only by solvent evapora
tion; lacquers may be clear or pigmented. “Drying” covers all mechanisms of film formation,
which is the transformation of applied liquid paint into a solid film, and may involve physical
change with polymer particle coalescence and chemical change, not just the evaporation of
solvent. “Curing” most usually applies to processes in which applied agents such as heat and
light are used to effect drying. Drying and curing are generally polymerization processes.
The terms “resin,” “vehicle,” and “binder” are all used for the film-forming part of the
paint, which is nowadays a polymer (wholly or partly synthetic), though in early paints it
may have been simply oil or gum. Resins are made in kettles or reactors and thinned in
chums. The heat polymerization processing of oils is frequently referred to as “bodying.”
Both paints and resins are characterized by their nonvolatile or solids content, their viscosity,
and frequently their acidity, expressed as acid value (mg KOH/g resin). For resins containing
oil, the terms “long oil,” “medium oil,” and “short oil” are used to categorize the oil or fatty
acid content of the total resin. This concept guides both the flexibility of the film and, if the
film contains drying oil, the rate and extent of oxidative drying to be expected. Oleoresinous
vehicles containing less than 65% oil are termed “short oil”; those with more than 65% oil,
“long oil.” For alkyds, “long oil” contains more than 60% oil, “medium oil” 40-60% , and
“short oil” less than 40%. The terms “drying,” “semidrying,” and “nondrying” have already
been referred to.
The age-old balance required with paint formulation and hence directed to the resin prepa
ration is that of adequately achieving final properties, particularly mechanical strength, getting
surface coverage and build, and being able to apply the paint evenly, all at acceptable cost.
In molecular terms, good mechanical properties require solid polymeric binders of high or
infinite molecular weight; most application techniques demand low viscosities, and hence
paint compositions always contain some solvent or diluent to make the product fluid for appli
cation. If the binder is of high molecular weight, a proportionately higher quantity of solvent
will be required for application than for a low molecular weight binder; the latter, however,
has inferior mechanical properties. The effect of this can be seen from spirit varnishes and
nitrocellulose paints (lacquers), which may contain 80-90% solvent as they are applied. The
well-recognized solution to this is to employ a curing or drying reaction so that the binder
will increase in molecular weight and its properties improve following application.
Paint systems can be applied, for example, by brush, spray, roller coating, and dipping.
Reaction can be assisted by component or catalyst addition just prior to application and by
the use of heat or other radiation after application. Utilizing the drying of oils through autoxi-
dative cross-linking has provided one of the most enduring examples of this type of system.
Solvent or diluent, as already stated, is an essential component. Solvents of choice have
moved on from turpentines to white spirit; short oil alkyds and acrylic and epoxy resins
require the use of the full range of organic solvents to achieve full solubility. Organic solvent
content is expressed as the volatile organics content (VOC) of the coating, and users expect
VOC to fall within certain limits for different coating types. These limits are being enforced
by legislation. Water is now the preferred diluent, with polymers being used in dispersions
rather than solutions.
V. CHEMICAL PROCESSES IN OIL AND FATTY

ACID USAGE
A. Chemistry of Oil and Fatty Acid Reactions
Apart from hydrolysis and esterification reactions involving oils and fatty acids and their
esters, a range of other reactions are possible when one or more double bonds, or attached
groups such as hydroxyl, are present on the fatty acid chain. Reactions carried out at the
unsaturation site include dimerization, maleinization, epoxidation, vinylation, dehydropoly
merization, and hydrogenation. Oxidation reactions are described in the next section.
Epoxidized oils find little use in coatings, and their manufacture and use are discussed in
another chapter in this book. However, naturally occuring vemonia oil has recently been
advocated as a reactive diluent for high solids alkyds [14].
Heat bodying dimerization of unsaturated oils is an integral part of the process for produc
ing oleoresinous and alkyd resins (see later). This heat bodying may also be carried out as a
separate process to form what are referred to as stand oils. This process, carried out at temper
atures around 300°C (lower with catalysis), consists predominantly of bimolecular additions
or combinations resulting in so-called dimer fatty acid structures in the oil. The strongest
contributor is probably the Diels-Alder reaction, as shown in Fig. 1, which involves conju-
716 Bentief
CHs -(CH2 )4 -CH = CH -CH2 -CH = CH- (CH2 >7 -C O O R
CHs - (CH2 )4 >CH2 - CH = CH - CH = CH - (CH2)? -C O O R
CHs -(CH2 )4 CH2 -CH = CH- (CH2 )7 -C O O R
\ /
CH-CH
/ \
CH3 -(CH2 >5 -CH CH- (CH2 >7 -C O O R
\ /
CH = CH
Fig. 1 Thermal dimerization of fatty acids.
gated pairs of double bonds on one of the fatty acid moieties (formed in situ if not already
present) [15]. The separate dimer fatty acids are used both in alkyd resin manufacture and in
the preparation of polyamide resins, which are used as thixotropic modifiers for alkyd resins.
Maleinization of oils and fatty acids is carried out both separately and in situ in alkyd resin
preparation to provide a flexible, higher functionality acid component of formulations. In the
stand-alone process, typically linseed oil fatty acid (LOFA) is reacted with maleic anhydride
at 200°C, and addition occurs principally by the ene reaction (Fig. 2a), because the bonds
involved are nonconjugated [16]. Where conjugated double bonds are present, a Diels-Alder
reaction is possible (Fig. 2b), which proceeds rapidly and exothermically. At higher maleini
zation levels of LOFA, since the ene reaction moves the double bond from the nonconjugated
-CH2-CH = CH- -CH=CH-CH
CH -C CH - c ; '
CH -C >
1
CH2-C.
(A) o
CH = CH - CH = CH - CH = CH
/ \
-CH JCH -
\ /
+ CH = CH —— > CH
j - CH
1
1 1
c c c c
/i' \ / '^ // \ / ^
O 0 o O 0 0
(B)
Fig. 2 (a) Ene reaction with maleic anhydride, (b) D iels-A lder reaction with maleic anhydride.
to the conjugated configuration, it is possible that a sluggish ene reaction is followed by a

faster Diels-Alder reaction on the same fatty acid.
Unsurprisingly to the polymer chemist, with the potential availability of the double bond
for addition reaction, vinylation with styrene or vinylbenzene, either as a preliminary or final
stage of manufacture of an alkyd is practiced to obtain polymers that combine polyester and
chain addition polymer structures in order to exploit the properties of both [17]. The process
is designed to graft the two structures together to overcome the consequences of incompatibil
ity, which would otherwise be very apparent. In practice, the reactivity of double bonds to
addition reaction is not high, and some grafting is also achieved by the use of peroxidic
initiators such as di-tert-butyl peroxide, which almost certainly adds grafting sites by abstrac
tion of hydrogen atoms from positions adjacent to double bonds. Graft sites to the main
polymer in alkyds have also been provided by the reaction of small amounts of methacrylic
acid into the polymer prior to vinylation.
A novel reaction demonstrating the susceptibility of unsaturated fatty acids to hydrogen
abstraction is that of '‘dehydropolymerization” [18]. With this, treatment of a range of oils,
including soybean oil and tall oil, with di-tert-butyl peroxide at 145°C was found to generate
dimer structures superficially similar to those described for heat bodying but believed to have
a single -C-C- link between chains.
Dehydration of castor oil, composed mainly of glyceryl triricinoleate, is carried out to
generate a second double bond, and hence semidrying oil status, by removing the hydroxyl
group on the carbon in the 12-position along with the hydrogen atom on one of the adjacent
carbons [19]. This process is carried out at temperatures of 280-300°C in the presence of
strong acids such as sulfuric acid and is illustrated in Fig. 3. Since oil bodying as described
earlier can then occur, skill in temperature control, choice of catalyst, and use of high vacuum
is necessary to produce low viscosity, good colored products. As will be evident from the
structure, both conjugated and nonconjugated linoleic acids may be produced (see Table 1).
As a semidrying oil, both dehydrated castor oil (DCO) and the fatty acids (DCOFA) find
application in alkyds and epoxy esters. However, owing to a residual characteristic surface
“nip” when DCO/DCOFA is used, its use is restricted to industrial and mainly stoving applica
tions.
Castor oil fatty acids may be hydrogenated to produce 12-hydroxy stearic acid, which can
be self-condensed to a limiting molecular weight by virtue of its residual stearic acid content
[20]. Poly-12-hydroxy stearic acid is the essential aliphatic-soluble oligomeric component of a
range of graft copolymers (see later). Castor oil may also be sulfonated (sulfated), and sulfated
castor oil, or Turkey red oil, is used as a pigment dispersant.
R R
i I
H - c - H H - C - H H - C
I II
H OH H - C H - C
I I
H - C - H - H2O H - C H - C -
I I I
H - C H - C H - C
II II II
H - C H - C H »C
I 1 !
R’ R' R'
Fig. 3 Dehydration of ricinoleic acid to form conjugated and unconjugated forms of linoleic acids.
[R ^ C sH h and RHCH2 ) 7 C0 0 H]
718 Bentley
With the different reactivities shown by conjugated and nonconjugated fatty acids and also
some differences in reactivity of their isomeric forms, isomerization is a useful process to
enhance properties, principally drying. As an example of the possibilities, linoleic acid can
be isomerized to the conjugated form and further modified from the cis,trans to the trans,trans
configuration [21]. A number of patented processes exist, and isomerized fatty acids are used
particularly in medium oil air-drying alkyds. They improve the initial drying, color, and resis
tance properties of the alkyds.
B. Drying Mechanisms
Drying or polymerization of fatty acid moieties is the most widely used of all hardening
mechanisms for surface coatings. It is an oxidative process, and early evidence (though not
then necessarily fully understood) includes Priestley’s use of linseed oil to remove the oxygen
that had been sustaining rats enclosed in a bell jar. The process, generally referred to as
autoxidation, has drawbacks that encourage continued examination of its mechanisms to fur
ther control and overcome them. Foremost, it is a cross-linking mechanism that does not stop
at a convenient point when mechanical properties are achieved but continues, ultimately caus
ing embrittlement and degradation. Second, control of drying speed, through-drying, and dry
ing in adverse conditions depends on the use of combinations of metal driers, whose cost,
balance, availability, and toxicity may all cause concern. Furthermore, side reactions cause
production of unpleasant volatile materials, and yellowing occurs to some degree in most
environments; both of these latter are features that paint chemists would like to control.
In the past, commercial pressures have triggered spurts in investigation. More recently,
moves to water-based paints and legislation requiring lead removal have led to changes in how
driers are used. However, it is notable that the availability of new investigative techniques has
over time enabled step changes in the quality of reported investigations, so that the literature
shows application of tracer techniques [22], much IR investigation [23], including time lapse
[24] and recently FTIR spectroscopy, ^^C-NMR spectroscopy, and most recently SIMS [25],
E X A F S [24], and S A X S [26].
Although parts of the process are incompletely understood, the overall process clearly
involves initial formation of hydroperoxides followed by their breakdown, induced by cobalt
in modem systems, and subsequently by free radical linking and scission processes to form
cross-links. Practical complications in investigation arise through the range of fatty acids pres
ent in typical oils and the number of reactive positions, the practice of adding a blend of
driers to a typical system, the fact that only the surface is available for oxygen ingress and
the escape of volatile products, and the presence of other components in the paint that may
affect drying.
For a preliminary overview, the reader is urged to consult the excellent review by Wexler
[27], which gives much background on oxidation and radical reactions and affords insight
into formulation practice in the context of drier use.
The first stage in drying, often preceded by an induction period, is the formation of hydro
peroxides, with attack occurring predominantly at allylic hydrogens, whose position is acti
vated by double bonds. The reaction is enhanced when the double bonds are multiple and
conjugated. Khan [28], using a variety of techniques, showed with nonconjugated linoleic
acid the clear-cut formation of 9- and l3-cis.trans-hydroperoxides with a shift both to conju
gation and to the trans configuration at the position adjacent to the hydroperoxide. Figure 4
illustrates this. A free radical mechanism is now thought most unlikely.
Much investigation has been carried out on purified esters, and this has led to the prepara
tion of high purity crystalline linoleate hydroperoxide [29].
Hx / H - - 0
CH2 O2
■CH = C H ' 'C H = C H - ---------> -C H = CH C = C-
I I
c/s H H
HOO^ ,
/C=C^ "h
-C H = CH _ H
trans
Fig. 4 Hydroperoxide formation and rearrangement. (From Ref. 28.)
Part of the subsequent reaction involves not just useful polymerization but also scission
and breakdown products. These are responsible for the characteristic odor of paint drying,
which can be acrid and lachrymatory in conditions of poor ventilation. A number of examina
tions have been published, of which some of note are those by Loury and Forney [30,31],
Frankel [32], and the recent studies by Leeves [33,34]. All studies identified a multiplicity of
alcohols, aldehydes, ketones, and esters. Of these, the aldehydes propanal and hexanal and
acids such as formic are major contributors to the unpleasantness of drying alkyd fumes.
Thus, for example, Loury and Forney’s examination of methyl linolenate oxidation under
mild conditions at 20°C found 29 volatile compounds [30]. They also produced a complete
mechanistic scheme showing how by oxygen fixation at all possible positions followed by
scission, almost all of these compounds could be formed. Other esters examined were methyl
linoleate and methyl oleate [31]. Frankel identified 27 compounds using similar conditions
[32].
In addition to examining the same esters as Loury and Forney, Leeves extended his investi
gation to alkyds using typical drying oils (soybean, linseed, DCO, and tall oils), an iso-
merized acid, and purified linoleic acids [33,34]. He used two typical drier combinations
and investigated less common catalyst systems. Differences were found, but interpretation
was difficult.
Cobalt, as has been stated, is now the major drier used, with supplementary driers includ
ing calcium, zirconium, and aluminum; lead is no longer in general use because of environ
mental and toxicity concerns. These driers are now used as octoates (octanoates), except for
aluminum, which is used as the ethyl acetoacetate substituted alkoxide. Manganese and bar
ium have been used, and zinc may also be considered partially active as a drier in paints,
since zinc oxide when used in pigmentation may partly dissolve as soap formed through
reaction with any residual acid.
The cobalt redox system is shown in Fig. 5, which illustrates why only a low concentration
of cobalt is required and the free radical nature of decomposition. The nature of the cobalt
drier complex [35] and the kinetics and mechanism of the complete process have been studied
by Waters [36]. Film weight studies show that linseed oil initially takes up about 40% by
ROOH + Co 2 ^ C o3^ 4- RO« -f OH«
ROOH -i- Co 3^ Co 2-^ 4- ROO* 4- H^
Fig. 5 Cobalt-catalyzed hydroperoxide decomposition.

720 Bentley
ROOH RO • + « OH Scission
RH RO « ROH + R* Abstraction
Ro 4- R« R - R
Oxidation
R« 4- RO « R - O - R
and
R. + O2 ROO*
Dimérisation
R. + ROO• R - 00 - R
Fig. 6 Dimerization reactions occurring with hydroperoxide decomposition.
weight of oxygen, around half of which is ultimately lost in the form of volatile scission
products. It has been concluded that polymerization, which forms both -C-0- and -C-C- cross
links, primarily involves the addition of peroxy and alkoxy radicals to a conjugated double
bond, with ether and peroxy links predominating. Other recent work using NMR and SIMS
confirms this conclusion [25], showing the links to result essentially from dimerization reac
tions. Figure 6 shows many of the possible reactions, where R indicates attachments to fatty
acid chains. This and other studies indicate that oleates can have at least some participation
in the later stages of cross-linking. The final film polymer structure will be a cross-linked
three-dimensional network, swellable but not soluble unless bond breakage occurs.
Various views have been expressed on the roles of the ancillary driers, but although their
effectiveness has been fully demonstrated, the mechanisms can only be postulated at present.
Nevertheless, formulators will be aware of the way different combinations can be chosen to
be effective with different pigmentation systems.
The auxiliary metal driers are known to enhance the catalytic activity of cobalt, thus im
proving through-drying and low temperature curing. It is also probable that driers such as lead
react preferentially with some of the autoxidation by products, which would otherwise build
up in the film, interfering with the catalytic effect of the cobalt. Aluminum and zirconium, in
particular, are capable of reacting with functional groups in the medium, thus yielding a cross-
linked structure; both function by forming coordination and covalent eross-links [37,38].
Investigation continues into the drying mechanism in new situations. For example, a recent
paper [39] suggests that in high solids systems, behavior is different to that found in conven
tional alkyds, with singlet oxygen believed to be involved in the reaction.
C. other Drying Processes

Other processes are available for the euring of finishes containing drying oil. The processes
described above occur under ambient conditions in air. However, under stoving conditions
with industrial paints, clearly the thermal polymerization reactions described earlier also play
a part. In addition, paints that contain a strong acid catalyst, primarily to promote cross-
linking reactions with added urea formaldehyde resin, for example, may also have autoxida
tion promoted by these acids.
Strong acid catalysts such as alkyl- and arylsulfonic acids have been shown to accelerate
alkyd resin drying at room temperature [40]. To be most effective, semidrying oil fatty acid
must be present in an alkyd structure formulated with a significant excess hydroxyl content.
With 1-6% of eatalyst, drying is comparable to that with conventional driers, and the mecha
nism postulated is the acid-catalyzed decomposition of peroxy groups formed in the initial
stage of other autoxidative mechanisms.
Photoinitiated autoxidation has also been demonstrated, with autoxidation initiated either
directly by UV radiation or with the aid of photoinitiators. In other applications, photoinitia
tion of unsaturated polyester reaction by straightforward radical mechanisms is an effective
process, using the combination of a ketone such as benzil and a reducing agent capable of
reducing the photoexcited state of the ketone. In translating this to autoxidative systems, it
has been shown that the drying of alkyd resin can be initiated by the use of benzil alone in
daylight [41,42]. It is believed that benzil in its excited state abstracts a methylenic hydrogen
from unsaturated fatty acid, which produces radicals and results in cross-linking. In this case,
although hydroperoxide is present during drying, it is doubtful that these peroxides contribute
significantly to the drying process.
Enzymatically catalyzed autoxidation has been reported, with cytochromes, heme, and sim
ilar metal-containing enzymes cited as being active: Whether this is principally metal-cata
lyzed autoxidation or involves other mechanisms is unclear, and much of this area has yet to
be explored.
D. Yellowing
Yellowing is an intrinsic problem with autoxidative drying, its effect being more noticeable
with films kept in the dark and in the atmospheres found in kitchens and bathrooms. The effect
of ammoniacal compounds has long been recognized, and accelerated testing of yellowing is
frequently carried out in an ammonia cabinet. It should be noted, however, that bleaching
processes also occur, and yellowing-bleaching cycles can be demonstrated, so that the pro
cess is to some extent reversible.
Yellowing is a process linked with residual unsaturation, as is drying ability, and drying
and yellowing proceed together. In general, chromophoric groups include unsaturated and
aromatic ring structures, -N = C - and > N = 0 , and those always present in oils, including
-C = C - and > C = 0 , and already involved in the drying process. Mechanisms causing the
formation of -NR2, -OH, etc. as auxochromes can readily be postulated.
The process of yellowing can be seen as proceeding stepwise, involving the production of
colorless precursors followed by further reaction to colored compounds [43]. Yellowing of
aldehydes, themselves produced in the process, can be demonstrated, with unsaturated alde
hydes yellowing most severely. Davison [44] showed that hexadienal yellows the worst of
those unsaturated aldehydes she tested in the dark at VO'^C. Other authors have identified
benzoquinone structures in films [45].
E. Degradation and Weathering Mechanisms

In many respects, degradation mechanisms can be seen as extensions or reversals of the poly
merization and drying processes. It has long been recognized, particularly with respect to the
autoxidation mechanism, that to get an initial cure that allows handleability and hardness to
develop in reasonable time for the coating to withstand normal wear and tear, the system
requires high concentrations of unsaturation. Hence a continuation of this process is likely,
which leads to a progressive increase in cross-link density and the loss of the pendant groups
plasticizing the structure, producing embrittlement. This may be followed by degradative pro
cesses with bond breakage causing polymer structure failure. Some very detailed studies of the
processes that occur have been published [46]. Unsaturation is both reacted and regenerated in
the drying process, and residual unsaturation provides target sites for a host of degradative
mechanisms. Depending on the use of the painted article, oxygen, UV, and moisture are all
likely to be aggressors in various ways.
722 Bentley
Many of the polymer structures used are ester-based, and fatty acid moieties are also linked
into the structure by ester groups. Hydrolysis is therefore a factor in film breakdown when
films are in contact with moisture, especially if soluble salts are present. A factor in facilitat
ing this process is the number and location of the more polar hydroxyl and carboxyl groups
on the polymer, as their presence allows water to enter and swell the film, thus enabling
attack. The drying and scission processes result in increases in the concentrations of both of
these groups, thus accelerating the process. In damp conditions, fungal and bacterial agents
contribute to breakdown.
Pigmentation and additives are well recognized as having significant effects. Different
grades of titanium dioxide can, for example, either shield the resin from radiation or, alterna
tively, photocatalyze breakdown. Inhibitors and UV absorbers are frequent additives, though
they must be carefully chosen to avoid interference with basic curing.
It should be said, however, that the breakdown processes are relatively slow, so a properly
applied decorative alkyd paint system, with the use of primer, undercoat, and the optimum
thickness topcoat, can give 3-5 years of full protection before some partial failure at, for
example, edge joints necessitates retreatment. Normally, the first apparent failure mode is that
of surface dulling and loss of gloss, and possibly some chalking (powdery loss of the surface
layer); in early stages this will be seen as an aesthetic rather than mechanical failure. Alkyd-
based industrial paints, for which reactions other than autoxidation provide the major curing
following application, generally retain a satisfactory appearance and other properties for the
service life of the product (car, appliance, etc.).
VI. SURFACE COATING RESIN SYSTEMS

A. Oleoresinous Systems Including Treated Oils
In the United Kingdom, linseed oil probably came into use in the 1400s. Certainly by the
1600s, white and red lead were used in what we would now identify as oleoresinous binders.
These comprised linseed oil and resins such as rosin and shellac. Thus we can trace over 400
years of use of indigenous linseed oil, which was used with imported tung oil along with
rosin and turpentine as naval stores, and with white and red lead included as protective pig
ments, also providing catalysis for autoxidation.
Oleoresinous media are made by heating together oils and chosen solid resins until compat
ibility and a predetermined viscosity are reached [47]. Solvent and driers are added to give
an unpigmented varnish composition. For paints, the solid components, which are now classi
fied as pigments if they are colored or scatter light significantly and otherwise are classified
as extenders, must be dispersed in the resin. Only drying oils (with possibly a limited amount
of semidrying oil) can be used in this class of paint vehicle. This type of vehicle has use now
only in varnishes, in some undercoats and primers, and in printing inks.
The major oils used in oleoresinous systems have always been linseed oil and tung oil,
with the former dominant, now overwhelmingly so as there are very limited supplies of tung
oil. Oiticica oil has been a satisfactory replacement for the latter, although a range of oils
blended with linseed oil have been used. These include fish oils, dehydrated castor oil, soy
bean oil, and tall oil esters with glycerol or pentaerythritol. The choice of oil blend will
always be such that in order to keep up the drying potential, proportions of highly unsaturated
drying oil, drying oil, and semidrying oil need careful control, with the amounts of soybean
oil or tall oil esters strictly limited. Clearly, in various parts of the world, indigenous oils
have been used, and a long list of those useable includes safflower, sunflower seed, rubber
seed, and tobacco seed oils.
Natural resins used have been the copal fossil resins found in a number of locations and
rosins obtained either as gum rosin from the live tree or wood rosin from stumps. As required
with oils, preheat treatments for these resins evolved. Thus the copals were normally “run”
to effect partial decomposition and improve oil solubility. Rosins were distilled, maleinized,
and partly polymerized by a variety of techniques.
Synthetic resins have been used since they became available; these include phenol formal
dehyde (PF), coumarone indene, and most recently epoxy resins. PF resins may be modified
with rosin to aid oil solubility and reduce cost. Unmodified and modified resole type PFs [3]
have become the most used resins in oleoresinous vehicles. Epoxy resins, which are used to
impart hydrolysis and chemical resistance, are more expensive.
Oleoresinous vehicles are manufactured by heating oils and solid resin together at tempera
tures up to 260°C. Since the heat polymerization of oils, particularly tung oil, can lead to
gelation, skilled techniques are required to ensure sufficient reaction that the resin is “solubi
lized” while avoiding gelation. Methods to stop reaction involve cooling back either with oil
held out of the formulation or with thinning solvent. Monitoring techniques have often been
primitive, involving judging incipient gelation by the tendency of samples taken from the
process to “string” and judging the attainment of compatibility by the clarity of a “pill” of
material cooled on a metal plate. The potential for fire during processing is great but can be
reduced by an inert gas sparge and by heating reaction vessels indirectly. These processes are
undesirable in that considerable fumes are produced from decomposition of the oils and resins
used, and this requires containment for environmental reasons.
While tung oil heat-bodies and gels rapidly and hence can be used only directly in vehicles,
it is advantageous to use linseed and other oils prebodied as “stand oils”; this is still true of
alkyd resins. This heat treatment, previously done in directly heated open pots, is now more
safely carried out in indirectly heated stirred vessels under inert gas. All drying and semidry
ing oils may be treated in this way, the process improving the drying, elasticity, and durability
of the derived coatings, possibly because polymerization is through direct -C-C- links, reduc
ing residual unsaturation. The unsaturation may also have undergone some rearrangement (to
conjugated form or cis-trans transformation), and there will be some early attendant loss of
volatiles, mimicking the effects of partial drying.
A simple wood treatment involves the use of so-called boiled linseed oil, which was once
prepared by dissolving metal oxides in the “boiling” oil but is nowadays made by blending
soluble driers with linseed stand oil.
Blown oils have also been manufactured; the heated oil is preoxidized by an air sparge,
the temperature being kept to around 130 °C . Blown castor oil was an ingredient used in the
manufacture of wrinkle finishes.
B. Alkyd Resins
1. Alkyd Overview
Alkyd resins, sometimes known as glyptals or oil-modified alkyds, are polyester resins whose
structures incorporate a significant amount of combined oil-derived fatty acids. This may be
achieved through processing via the parent oil, and in these cases glycerol will constitute at
least part of the alkyd’s polyol content. Alkyd resins may be made equally as well by using
preseparated fatty acid and the fatty acid process, as described later. The name “alkyd” resins
possibly derives from their manufacture from alcohol and an aciJ, with the “cid” changed to
“kyd.” Alkyds have been subclassified as oil-modified and oil-free. “Alkyd resin” is now
generally used as the generic name for the former type; the term “oil-free” is less usually
applied to polyester resins, normally formulated from polyhydric alcohols and di- and tribasic
724 Bentley
acids but sometimes containing short non-oil-derived monobasic acids as minor components.
Alkyd resins were one of the first large-scale applications of a synthetic polymer in surface
coatings and have endured because of the variety of ingredients available, which make it
possible to use alkyds in virtually all areas of coatings application [48]. As a resin class, they
are cost-effective and can be made with relatively simple equipment.
Alkyd resins have a polyester backbone structure, and the acid most commonly used in the
backbone is orthophthalic acid, always as the anhydride (PA). A range of di- or polyhydric
alcohols are used. The polymer-forming reaction is hence esterification, with water the reac
tion by-product. The process can be single-stage with all ingredients charged together if fatty
acid is used (the so-called fatty acid process), but it is more economical, and also more
sensible if glycerol is to be part of the formulation, to incorporate oil directly. However,
this does not occur readily, since oils are triglycerides, and a preliminary stage known as a
monoglyceride stage is normally carried out to make the oil reactive. As discussed below,
two functional groups are required to allow molecules to react to form a polymer structure,
and hence each oil molecule must shed two fatty acid groups to make available two of the
three hydroxyl groups of the glycerol for subsequent reaction.
Although any polyol (or polyol blend) may be used, the monoglyceride process is illus
trated in Fig. 7 with glycerol, which is used in simpler alkyds at all levels of oil content. This
process is typically carried out at around 240°C in the presence of a catalyst such as sodium
hydroxide and requires around 30 min. The process does not go to the completion implied
above but approaches an equilibrium mixture consisting of some unreacted oil and glycerol,
and also the diglycerides, in addition to the expected and predominant monoglycerides. It is
essential to ensure that sufficient randomization has occurred, and, depending on the polyol,
this may be tested by a simple alcohol tolerance test. In some circumstances, where a polyol
such as PE is used, it will be necessary to perform a small-scale trial alkyd preparation, taking
the monoglyceride sample and additional polyol and phthalic anhydride (PA) as appropriate,
heated in a wide-necked tube or small flask. The penalty for incomplete equilibration can be
the preparation of an alkyd that is cloudy or separates or is even prone to gelation. Equally,
where the oil is prone to heat bodying, it is essential not to prolong this stage, and the best
practice is to standardize both the heat-up rate and the time of reaction for this stage so that
subsequent stages will be predictable and reproducible.
The résinification or polyesterification stage is carried out by heating the reaction mixture
at temperatures between 180 and 260°C. Catalysis is not essential in conventional alkyd prepa
ration. High temperatures are used for long oil alkyds, where oil bodying during the process
is wanted to increase viscosity. The lower temperatures may be required with short oil alkyds
formulated to higher functionality with a strong gelling potential, so that they are more con
trollable. Water of reaction is removed at this stage, though this is not practically a reliable
indicator of extent of reaction. If a monoglyceride stage has been carried out, the charge will
have been cooled so that additional polyol and dibasic acid can be added safely before reheat
ing. This stage, using monoglyceride, is shown schematically in Fig. 8 .
CH2-OOC-R CH 2 -OH CH 2 - O O C - R CH 2 -OH

I ! Catalyst | 1
CH - O O C - R 2CH-OH — > CH - OH + C H - OOC- R
i 1 1
1
1 1 1
CH2-OOC-R CH 2 -OH CH2-OH CH2-OH
Oil Glycerol a Monoglyceride P Monoglyceride
Fig, 7 Monoglyceride formation. (— OOC— R is a fatty acid moiety.)

CH2-00C-R
1
11 CH - OH + n [ HOOC COOH ]
I
CH2-OH
CH2-OOC-R CH2-OOC-R CH2-OOC-R

I
HOOC COO - CH CH -OOC - 4 coo- CH
I
CH2-OOC —--C O O - CH2 n-2 CH2-OH
+ n H2O
Fig. 8 Alkyd formation with monoglyceride and dibasic acid. (— OOC— R is a fatty acid moiety;
HOOC— COOH is a dibasic acid.)
The progress of reaction will be followed by observing the increase in viscosity, which is
related to an increase in molecular weight, and also the decrease in the acidity of the mixture,
measured as acid value. The rate of viscosity increase generally increases as reaction pro
ceeds, and sampling intervals decrease during preparation. As stated above, some preparations
are prone to gelation, and in this case operators need be ready to stop the reaction by cooling
and, after the initial cooling by adding solvent. When a preparation runs out of control and
gelation occurs, this may often be reversed and the reactor contents recovered if prompt action
is taken in adding fatty acid or glycerol; this is further indication that a process of randomiza
tion (by transesterification or ester interchange) is always occurring to some degree during
processing.
Where an oil such as tung oil is concerned, it is not generally possible to carry out a
monoglyceride process because of the very high rates of oil bodying. Nevertheless, a propor
tion of any oil, including tung oil, may be incorporated directly by using the fatty acid-oil
process. Here, fatty acid and oil are charged together, reacting directly with polyols and
phthalic anhydride in a single stage during which sufficient transesterification can take place
to make a satisfactory product. Another case where oil can be esterified directly is when
unmodified castor oil is used, which in formulation terms is considered a triol because all
three hydroxyl groups are reactive.
It will be clear from the foregoing that normal processing is either single-stage or, if a
monoglyceride stage is carried out, two-stage. However, stepwise addition of components has
been claimed to give narrower molecular weight and functional group distributions and lower
viscosity. The original claims concerned better drying from tall oil alkyds [49-51], and the
concept is potentially capable of being of benefit with high solids alkyds (see later).
An alternative process for incorporating oil into alkyd resins is the acidolysis process, in
which oil is equilibrated with phthalic anhydride and the polyol is added during a second
stage. This process is not widely practiced.
It has been claimed that some of the properties of alkyds are derived from the presence of
“microgel” in the structure [52]. A physical explanation of gelation is that it occurs through
separation of microgel particles, which at high concentrations coalesce to form a “macrogel.”
Otherwise, gelation is thought of as occurring when molecular weight becomes infinite, and
continued reaction results in the development of a three-dimensional structure in the system.
Practically, a gel is characterized by a significant presence of insoluble material.
726 Bentley
2. Alkyd Formulation and Manufacture

As has already been implied, the basic requirement for the formation of an alkyd or polyester
polymer structure is for a blend of molecules with two hydroxyl groups and a dibasic acid to
be reacted. Clearly in a system with equimolar amounts of glycol and difunctional acid pres
ent, complete reaction would be capable of producing a polymer of very high molecular
weight, limited in extent only by system viscosity and the ability to remove the water of
reaction in manufacture.
Practically, mixtures of components are used, and very high molecular weight is not re
quired. Inspection reveals that, taking all molecules into account, there will be a finite pres
ence of polymer ends, and hence if we average the reacted number of all groups per molecule
charged we will find that the value always falls short of 2 by a small amount. This concept
of mean functionality of the system is the basis of formulating with mixtures of components
[53,54 ], where in practical systems monoreactive acids such as benzoic acid, in particular,
may be used satisfactorily, as may polyols of functionality 3 or 4. Since we have also implied
that reaction overall may not be complete, a further consideration in both formulating and
actual preparation is the extent of reaction. In typical formulations, there may be an excess of
hydroxyl groups present, such that while the extent of acid group reaction may approach
unity, that of the hydroxyls may be much less. Again, if reactants with more than two func
tional groups are present, and if the average functionality of all groups present is greater than
2 , then if the extent of reaction is taken to a high level, gelation can very possibly occur.
Practical formulating will involve consideration of both stoichiometry and oil length and
calculation of the extent of reaction at gelation using equations derived from Flory-Stock-
meyer theory, of the type
(^gBf
^ (S /2 A -2 /A )(V fi-S g B )
where p is the degree of reaction of acid groups at gelation; / , g are functionalities of carboxy
and hydroxy moieties, respectively; and A, B are the number of moles of carboxy and hydroxy
moieties present.
The effectiveness of these equations has been tested [55]. It should be noted that it will
normally be necessary to consider phthalic anhydride as having a functionality of less than 2
because of its nonideal behavior. This includes a tendency to form oligomeric ring compounds
and also to exhibit some reversibility in its reaction, and theory has developed to take this
into account.
Those formulating alkyds are likely to use computer programs to carry out the necessary
calculations. Programs are available for purchase [56], and from raw material suppliers [57].
Alkyd resin manufacture requires high temperatures, and while esterification processes are
only mildly endothermic, more significant heat input is required to remove water of reaction
in the polymerization stage. Also, heat-up rates need to be reasonably fast for consistency in
manufacture, and for this reason heating systems need to be quite powerful. Alkyd resins
have been made in primitive equipment, venting water and fumes to containment systems; in
this case, formulations need to allow for a loss of around 5% of such volatile ingredients as
phthalic anhydride.
Better control is achieved using equipment with provision for using xylene as azeotroping
solvent, where xylene boiled from the batch is condensed, water is separated, and cold xylene
is returned to the batch. Good design is necessary to prevent overheads from being blocked
by phthalic anhydride sublimate, and plants variously have fractionating columns or bubble
cap scrubbers to overcome this.
Polymerization stages need control with frequent sampling, testing for viscosity rise and
acidity fall (measured by acid value). Endpoint prediction may be employed by graphing
results, possibly using guide lines from previous batches. Shorter oil alkyds, which are partic
ularly prone to gelation, need the most frequent sampling.
At the endpoint, resins require cooling to stop further reaction and will be thinned to
maintain handleability, using the appropriate solvent.
3. M odified Alkyds
Although some resin additives have already been discussed, some further alkyd modifications
merit mention. These are the ones recognized as significant in terms of imparting special or
unique properties.
The use of a thixotropic additive is a useful enhancement to brush-applied paints to give
nondrip and improved film build properties, and alkyds may be made thixotropic by modifica
tion with polyamide resin. The resins used are simple linear structures with diamine and
dimerized fatty acids. It is normal practice to prepare a long oil thixotropic alkyd with a
higher level of polyamide modification (5%) for blending with other resins to give the required
properties. In practice, great care is required in selecting the polyamide resin as well as skill
in formulation and preparation [58,59]. It may be difficult to characterize the resin in rheological
terms, as much subjective judgment is required in properties such as brush loading and brush
mark flow-out. Modification is carried out after alkyd condensation, and strict time, tempera
ture, and endpoint control are necessary to gain optimum clarity, viscosity, and gel strength,
with the most critical thixotropic properties peaking and then falling off during preparation.
Urethanation involves the replacement of some of the difunctional acid with a diisocyanate,
with toluene diisocyanate (TDI) the most common. In resin terminology, a urethane oil is a
resin formulated with complete replacement of PA by TDI, whereas a urethane alkyd is made
with partial replacement. The latter is the more usual. In this case, a two-stage (or strictly
three-stage if monoglyceride) process is used in which an alkyd is first prepared to low molec
ular weight and acid value but with a significant excess of hydroxyl groups. Reaction with
diisocyanate is then a further stage, carried out at lower temperature. As it is essential to react
isocyanate groups to completion, the reaction is principally controlled at the alkyd-making
stage. If a monoglyceride stage is included, care is needed to avoid using a catalyst that will
cause side reactions of isocyanate groups in the urethanation stage. These resins are harder
and tougher than unmodified alkyds, and for this reason they are blended with other alkyds
to improve these properties.
Styrenation (vinylation) of oils has already been described. It is used to improve “lacquer
dry” and hardness in medium oil alkyds for maintenance paints and primers. The process is
as described earlier. Although drying is improved, durability may be adversely affected. This
modification of principally DCO alkyds is carried out on the thinned resin after the alkyd
preparation stage is complete.
Silicone resin modification may be used to improve heat resistance and weatherability in
both alkyds and other resin types such as polyesters. Although it is an expensive ingredient,
silicone resin will give improvement in direct proportion to the amount used [60]. These
modified resins are either hydroxy or methoxy functional and are generally partly reacted into
the alkyd after its own condensation is complete.
4. Use o f A lkyds in P aint

Long oil drying oil alkyds remain the major vehicles used in glossy exterior paints and var
nishes, and in volume terms they are by far the largest volume oil-containing resins now
728 Bentley
made. These alkyds were developed with around 70% linseed oil, with PE and PA as the
other components. With the use of white paints increasing, partly due to the development of
nonchalking grades of rutile titanium dioxide, formulation has changed to the use of lower
yellowing semidrying oils, notably soybean oil. Soybean oil alkyds are generally formulated
with a significant proportion of soybean stand oil in the composition. The use of linseed oil
alkyds is now reserved for undercoats and varnishes. Other semidrying oils such as sunflower
and safflower may also be used.
Soybean oil is widely available, with its major use in food, but naturally suffers some
seasonal variation in yield and therefore in price. Because of this variation, distilled tall oil is
now used as an alternative to soybean oil. For applications in decorative alkyds particularly,
where tight cost control is necessary, manufacturers may have formulations with tall oil
(TOFA) available as alternatives to soybean oil resins so they can switch as price varies.
Since tall oil does not contribute any glycerol to the formulation, long oil tall alkyds may be
made with all PE and by the fatty acid process. Higher temperatures are then used.
Long oil alkyds of this type are of relatively low molecular weight and viscosity, used at
high solids, and diluted with an aliphatic solvent such as white spirit. Decorative paint is
pigmented with titanium dioxide for white paints, with inorganic pigments rather than the
more expensive organic pigments used where possible to provide color. The most expensive
high durability organic pigments are generally too costly for use in decorative paints.
Decorative paints are generally formulated with a blend of alkyds and include urethane
alkyd to toughen, thixotropic alkyd to improve application rheology, and possibly another
modified alkyd to provide, for example, easy brush cleaning. These have been described pre
viously.
These paints are generally applied by brushing, mainly to wood substrates (doors, win
dows, trim) in interior and exterior building decoration, but also to metal where appropriate.
Long oil alkyd base paints of this type are also used in commercial transport paints and in
maintenance of marine structures.
Medium and short oil alkyds have for many years been the workhorses of almost all indus
trial coatings markets, and they remain so despite being displaced by polyesters and acrylics
for spray-applied top coats and by epoxies for more corrosion resistant primer and undercoat
vehicles. The range of oils used is extremely broad, all available polyols and acids are to be
found, and the range of modifiers is extensive. To illustrate this, some typical applications
are described below.
Medium oil semidrying oil alkyds are still used for spray-applied vehicle refinishing as an
alternative to lacquers where two-pack finishes cannot be used. Since rapid drying to a good
hardness is required, the formulations may include isomerized fatty acids and chain-stopping
acids such as benzoic or para-tert-huiylhtnzoic acids (ptBBA). These will be partly aliphatic-
soluble; both the use of ptBBA and trimethylol propane (TMP) may be used to increase this
solubility by making the polyester part of the structure more aliphatic in character. The overall
alkyd properties are greater hardness and durability and faster drying speed than long oil
alkyds.
Medium oil alkyds made from a range of semidrying oils have been used in a wide range of
industrial stoving paints. For example, light fittings and the full range of domestic appliances
(refrigerators, etc.) may use stoving paints with soybean oil, dehydrated castor oil (DCO),
and tall oil alkyd, along with urea formaldehyde (UF) resin or melamine formaldehyde (MF)
for higher performance. Alkyd resins of this type can be made less expensive by incorporating
rosin, which will increase drying speed and hardness but decrease durability. Performance
can be enhanced by using isophthalic acid in place of PA and by careful choice of polyol.
The range of polyols includes glycerol, PE, TMP, neopentyl glycol, and other glycols. Nowa
days this kind of composition competes with thermosetting acrylics, which (admittedly with
the use of more expensive and specialized monomers) have better performance regarding
detergent resistance, for example.
Blending or partial reaction with phenolic or epoxy resins can be used to improve chemical
resistance, though these will reduce both yellowing and weathering performance.
Short oil coconut oil alkyds were used with MF resins to provide factory-applied stoving
automotive top coats and for some appliances. For automotive use, these are now fully sup
planted by acrylic resins, which can give the rheology control necessary in application for
metallic finishes, but other industrial uses remain.
Castor oil alkyds have been used in industrial stoving appliance finishes and, formulated
to higher hydroxyl content, still have particular use as plasticizing resins in nitrocellulose lac
quers.
5. W aterborne Alkyds
Although the ability to modify alkyds to confer partial or total water solubility has been well
used over many years, this has been explored further with recent trends to overall solvent
reduction, where the change to water as diluent achieves this aim. Alkyds can be made water-
dispersable by the incorporation of the hydrophilic polyethylene glycol, as a polyol, into the
alkyd structure [61]. This has been used to produce alkyds that confer water brush washability
in decorative paints and to produce alkyds with surfactant properties for a variety of industrial
cleaning purposes [62].
To attain higher levels of dispersability or complete water solubility, highly polar groups,
which will in practice need to be ionizable, will have to be incorporated in the resin, and
these are typically acid groups. Hence the alkyd is formulated to a much higher acid value
than normal, using, for example, maleinized fatty acid or using trimellitic anhydride in place
of phthalic anhydride. In the latter case, a technique to get high acidity involves a two-stage
process whereby terminal hydroxyl groups are reacted with trimellitic anhydride in a final
“ring-opening” stage, where the anhydride ring is reacted with one hydroxyl to give a single
ester group and two acid groups.
These resins require preparation in the absence of water and will be dissolved or dispersed
subsequent to polymerization. This is of advantage in circumstances where another resin such
as a cross-linking resin needs be coemulsified with the solubilized resin.
The presence of ionizable groups is not in itself sufficient to confer water solubility, and
in the case of acid groups, partial or full neutralization is necessary to confer full hydrophilic
properties. Ammonia, amines [63], or alkalis may be used, depending on the application. The
former clearly will be released and hence help to lower water sensitivity of the final film but
may present toxicity problems. In the case of alkali neutralization, since they will remain in
the film, cross-linking mechanisms need to be sufficiently effective to make the film fully
insoluble. Their use reduces the resin’s VOC and avoids problems caused by amine interfering
with autoxidative drying.
Dimethylol propionic acid (DMPA) is one material specially exploited in making this type
of resin, where the hydroxyl groups react into the resin structure but the acid groups, normally
reluctant to react, remain for subsequent salt formation. Recent use has been made of the
alkali metal salts of sulfo-isophthalic acid, claimed to be very effective at imparting solubility
and dispersability [64]. These two structures are shown in Fig. 9.
While acid groups have been described here as the groups to confer solubility, with other
resin types amine groups are an alternative to give cationic rather than anionic resins. These
are not, however, easy to obtain in an alkyd structure, and their use has been with acrylic and
epoxy resins.
730 Bentley
CH2 OH NaSOs ^ COOH

I
CH3 - C - COOH
I
CH2 OH COOH
Fig. 9 Structures of dimethylol propionic acid and a sodium salt of sulfo-isophthalic acid.
A drawback to using alkyds as waterborne vehicles is that the alkyd backbone is prone to
hydrolysis, and for this reason, to aid storage stability, water dispersability rather than in
creased water solubility is often more practical. Hence alkyd emulsions are being explored as
solvent-free paints. While these may be prepared by blending surfactant with the alkyd prior
to emulsification, the use of reactive surfactant reacted into the alkyd structure is more satis
factory [65]; the resin may include ionizable groups as described above. The phase inversion
temperature method is the most efficient for emulsification [66].
Although there are problems in the use of these water-soluble resins, with hydrolyzability
and with the inability to get the highest gloss finishes, it should be noted that alkyd emulsions
have been used successfully in a number of European countries. Novel techniques have
evolved for their preparation, and microemulsion techniques have been examined [67]. Cer
tain industrial products have been made with water-based technology, and for many years
alkyd-based waterborne primers, especially dip primers, were used on larger structures such
as car bodies.
6. H igh Solids Alkyds

As an alternative to full removal of solvent and the use of water to attain approval for certain
environmental labeling requirements (e.g.. Blue Angel), some current formulations are sup
plied at a higher solids content with smaller amounts of solvent required to achieve application
properties. “High solids” is generally taken to mean >85% solids in a brush-applied paint.
To get high solids in alkyds, it is necessary to formulate to higher oil length, since viscosity
constraints require lower molecular weight, which in turn requires a greater extent of reaction
to achieve cross-linking. To this end, compositions may contain higher functionality synthetic
oils, for example, full fatty acid esters of polyols such as di- and tripentaerythritol [68,69].
Considerable attention must be given to technique as well as design, since a narrower molecu
lar weight distribution is very desirable; it is claimed that this is achieved by replacing phthalic
anhydride by isophthalic acid [70].
An additional technique for formulating high solids alkyds is to add “involatile monomer,”
exploiting the ability of the free radical nature of cross-linking to copolymerize other unsatu
rated compounds into the structure [71]. Although this is not a recent discovery, a number of
materials have been used in this modem application. In addition to vinyl monomers, allyl
ether compounds have been used as diluents. The allyl ether group has autoxidative ability
and has been extensively investigated as a synthetic replacement for the drying oil moiety
[72]. As a reactive diluent an oil-like stmcture, pentaerythritol diallyl ether di-TOFA ester,
has been used.
Finally, with considerable effort being made to control polymer architecture in other poly
mer systems, such as group transfer polymerization (Du Pont), and in making dendrimers and
hyperbranched stmctures [73], dendritic stmctures have been synthesized as high solids alkyds
[74] based on a PE/DMPA core, with the many hydroxyl groups of the final product esterified
with fatty acid.
C. Epoxy Esters
The third major class of coatings, which have been made in significant volume in one particu
lar application, is epoxy esters, formed by the reaction of fatty acids with epoxy resins [75].
Epoxy resins are made by the reaction of epichlorohydrin with diphenylol propane (also
known as Bisphenol A) and have a structure with both in-chain hydroxyl groups and terminal
epoxy groups. Although their preparation is difficult and hazardous owing to the toxicity of
epichlorohydrin, a range of these resins are available from the “monomeric” form with « = 0.5
to grades with « = 1 2 or more. A generalized form of this structure is shown in Fig. 10. It is
common practice now for larger users to “chain extend” liquid grades with diphenylol propane
to higher molecular weight.
Both the terminal epoxy groups and the secondary hydroxyl groups can be reacted with
fatty acid, and the resin with « = 4 is typically used. The reaction is carried out at 240-260°C
under refiux to an acid value and viscosity endpoint in a manner similar to that used with
an alkyd.
It is possible to use the full range of fatty acids, though linseed and DCO are most com
mon. Although a nomenclature similar to that used for other oil-containing resins may be used
(high, medium, short oil length), there is another specific code in which, for example, “L 8
resin” refers to a 4-type epoxy resin modified with 0.8 equivalent of linseed oil fatty acids to
1.0 epoxy. These “straight” epoxy esters have been used for primer formulations and a range
of industrial finishes in which the higher chemical resistance of the epoxy backbone is of
benefit.
Medium and long oil esters of drying oil fatty acids have been used in air-drying finishes,
while short oil drying and nondrying fatty acid esters are used in industrial stoving primers
and finishes. Stoving finishes frequently include formaldehyde resins and phenolic resins to
provide cross-linking and further modify properties.
Water-thinnable epoxy esters have been prepared by a number of routes. An epoxy resin
ester emulsion for use in water-thinned primers can be made by an inversion emulsification
process in which water is added with stirring to a D4 (DCO-containing) epoxy ester that has
been preblended with amine and linseed oil fatty acids. An alternative method is to prepare
an epoxy ester using maleinized linseed oil fatty acid, and this kind of resin has been made
in high volume for use in dip primers, particularly automotive electrocoat primers [76]. In
this latter process the anionic resin, which will have been partly neutralized again with alkali
or amine, is deposited onto a metallic object by passing a current through a bath containing
the paint. In this case the object (car body) is made the anode of the electrical system and
resin is deposited onto it and becomes insolubilized; as film resistance increases deposition
slows and film thickness is relatively even. This resistance increase also promotes “throw,”
whereby paint is deposited inside box sections of chassis (where spray or brush processes
cannot reach). These resins are stoved to harden them completely and are pigmented with
anticorrosive pigments. In the main, this process has been superseded by the cathodic process
using non-fatty acid-containing resins.
o CHs OH CHa O
/ \ I I / \
CHs - CH CH2 O C O CH2 CH CH2 O O CH2 CH - CH2
1
CHs CHa
Fig. 10 Bisphenol epoxy resin structure.

732 Bentley
0. Other Oil Applications in Coatings

Castor oil, by virtue of its hydroxyl content, has also been used as a polyol. Hence it has
served as a component of polyurethane resin, where it is reacted with an isocyanate such as
toluene diisocyanate.
As mentioned earlier, 12-hydroxy stearic acid can be self-condensed to a limiting molecular
weight by virtue of its residual stearic acid content, and this polymer is the essential aliphatic-
soluble oligomeric component of a range of graft copolymers [20]. These have been the key
to the manufacture of nonaqueous dispersion (NAD) polymers [77]. This versatile, though
now lesser exploited, technique has been employed to prepare both addition and condensation
polymers, used as rheology modifiers in automotive paints and currently in woodstain formu
lations [78]. The graft copolymers have also found separate application in industrial detergents
and cleaners [60].
A number of attempts have been made to incorporate drying oil fatty acids into addition
polymer acrylic resin backbones to prepare an analogue of the drying oil alkyd resin, but the
results have not had significant commercial success. Saturated fatty acid derived monomers
are, however, used in significant quantities in acrylics, notably lauryl acrylate and methacry
late and to a much lesser extent stearyl methacrylate.
VIL CURRENT STATUS AND FUTURE TRENDS

Current environmental concerns and legislation, principally targeted to reduce and ultimately
eliminate solvent emissions, expressed as VOC, have enabled oil-based polymers to continue
to find use through some reformulation to higher solids and in water-emulsified form.
As already described, drying oils remain nearly unique in their ability to autoxidize to
“convert” a liquid polymer to a solid cross-linked durable film at room temperature, despite
some adverse features of the process. Only the allyl grouping, incorporated as either the allyl
ether or the vinyl cyclic acetal, has shown useful autoxidative ability [72], and research into
other synthetic alternatives may still be warranted. Although there are side effects from the
use of coatings containing drying oil in their contribution to “true” VOC via emissions from
their autoxidative curing, they are currently tolerated because of their history of consumer use
over many years. Again, there is currently no alternative to the use of drying oil alkyds in the
formulation of high gloss trim paints, so that while consumer demand for these continues,
their use seems ensured.
With environmental awareness now a permanent force in shaping both consumer accep
tance and industrial practice, and with developments being steadily enforced by legislation,
the movement to high solids and particularly waterborne coatings [79] looks set to continue.
There is now consumer pressure to use renewable rather than petroleum-based resources. On
cost grounds, oils compete well with the petrochemically derived ingredients normally used
to form polymers. In certain cases, pressures are triggering some new applications where
there is demand for ecologically responsible products; one is the replacement of mineral oil
in inks with rapeseed oil; in the United Kingdom, 1000-2000 t is already used annually [10].
This type of substitution will undoubtedly continue.
One further factor in materials use is the demand for recycling of products, where “cradle-
to-grave” life cycle studies are being carried out [80, 81]. These studies can include energy
requirement calculations for all components used in manufacture. Some goods are being rede
signed so they can be fully recycled, and legislation is beginning to enforce this. It might be
added that the role of paints is often “green” in greatly prolonging the life of products, pre
venting their premature scrapping and hence the energy and environmental burden of provid
ing new structures and scrapping the old ones. The use of a few micrometers of paint properly
applied externally plus appropriate inner surface treatments enormously extend the lifetimes
of motor bodies, for example. It is difficult to imagine this paint coating itself being recycla
ble, but certainly changes in ingredient use, already shown by restricting use of lead and
chromates, will continue. Recycling does, however, affect paint waste and used containers.
“Cradle-to-grave” studies on paint itself are inspiring ecolabeling that will embrace perfor
mance requirements, toxicity of ingredients, and limitations on the release of volatiles as
voc.
Where oils are concerned, with crop yields already improved by selection and by farming
methods, genetic engineering looks set to change fatty acid compositions and produce new
varieties. [82,83] The effects of this will become evident in future years.
ABBREVIATIONS
DCO Dehydrated castor oil
DCOFA Dehydrated castor oil fatty acid
DMPA Dimethylol propionic acid
EXAFS Extended X-ray fine structure analysis
FTIR Fourier transform infrared spectroscopy
LOFA Linseed oil fatty acid
MF Melamine formaldehyde
NAD Nonaqueous dispersion
NMR Nuclear magnetic resonance
OEM Original equipment manufacture
PA Phthalic anhydride
PE Pentaerythritol
PF Phenol formaldehyde
ptBBA para-tertiary-EuiylhQnzoic acid
SAXS Small-angle X-ray scattering
SIMS Secondary ion mass spectroscopy
TDI Toluene diisocyanate
TMP Trimethylol propane
TOFA Tall oil fatty acid
UF Urea formaldehyde
VOC Volatile organic content
REFERENCES
1. R. Lamboume (Ed.), Paint and Surface Coatings, Ellis Horwood, London, 1987.
2. D. H. Solomon, Chemistry of Film Forming Polymers, Krieger, 1977.
3. J. Bentley, Organic film formers, in Paint and Surface Coatings (R. Lamboume, Ed.), Ellis Hor
wood, Chichester, England, 1987, Chap. 2.
4. Surface Coatings Association o f Australia, Surface Coatings, Vol. 1, Raw Materials and Their
Usage, 3rd ed.. Chapman and Hall, London, 1993, Chaps. 4 and 5.
5. A. Kimber, Recovery hampered by raw material hikes. Fur. Chem. News, 14 Nov. 1994, pp. 2 0 -
21; data from Akzo Nobel.
6 . CEPE Annual Statistical Review 1994, presented at CEPE Annual Conference 1994.
7. M. S. Reisch, Paints and coatings, Chem. Eng. News, 18 Oct. 1993, p. 34.
8 . Japanese Paint Manufacturers Association. (Data from Coatings COMET, No. 3, 1994, p. 65).
9. British Coatings Federation, U.K. Coatings Industry Facts & Figures 1994.
734 Bentley
10. Industrial Markets for U.K. Produced Oilseeds, Home Grown Cereals Authority, U .K ., Research
Review No. OS9, January 1995.
11. J. Mispreuve, Oil Fats Int. 10: 3 (1994).
12. K. S. Ennar and J. Oxley, Composition and fractionation of tall oil, J. Oil Colour Chem. Assoc,
50: 557 (1967).
13. A. Hase and S. Pajakkale, Tall oil as a fatty acid source, Lipid TechnoL 6: 110 (1994).
14. P. Muturi, Epoxidised vegetable oils as reactive diluents 1. Comparison of vemonia, epoxdised
soya and linseed oils. Prog. Org. Coatings 25(1): 85 (1994).
15. L. F. Byrne, The chemistry, manufacture, and uses of dimerised fatty acids. Paint TechnoL 26(3):
28 (1962).
16. H. R. H. Hoffmann, The ene reaction, Angew. Chem. Int. Ed. 8: 556 (1969).
17. H. W. Chatfield, Varnish Constituents, Leonard Hill, London 1953, p. 76.
18. R. Lamboume, The dehydropolymerisation and copolymerisation of drying oils and other reactive
compounds exhibiting a-methylenic activity, J. Oil Colour Chem. Assoc. 51: 310 (1968).
19. H. W. Chatfield, Varnish Constituents, Leonard Hill, London, 1953, p. 47.
20 . F. A. Waite, The development of graft copolymers as macromolecular surfactants for use in non-
aqueous media, J. Oil Colour Chem. Assoc. 54: 342 (1971).
21 . P. L. Nichols, S. F. Herb, and R. W. Reimenschneider, Isomers of conjugated fatty acids. I.
Alkali-isomerised ricinoleic acid, J. Am. Chem. Soc. 73: 247 (1951).
22 . N. A. Khan, Application of tracer techniques to studies on autoxidation reactions, J. Am. Oil
Chem. Soc. 30: 273 (1953).
23. T. L. Helmon and M. Möller, Förändringar i Klarlackers Kemiska och Fysikaliska Egenskaper
vid Utomhus-och Laboratorieprovning 5b. IR-spektroskopiska studier av härdnings-och nedbryt-
ningsförlopp, Färg Lack 13: 155 (1967) (in Swedish).
24. J. H. Hartshorn, Time-lapse infrared spectroscopic investigation of alkyd and linseed oil cure, J.
Paint TechnoL 54 (687): 53 (1982).
25. W. Z. Muizebelt, J. C. Hubert, and R. A. M. Venderbosch, Mechanistic study of drying of alkyd
resins using ethyl linoleate as a model substance. Prog. Org. Coatings 24: 263 (1994).
26. C. Carr, I. S. Dring, and N. A. R. Falla, Studies of the curing of soya oil/alkyd films containing
metal driers, XVI Int. Conf. Org. Coat. Sei. TechnoL, Athens, July 1990, p. 9 9 -1 0 9 .
27. H. Wexler, Polymerisation of drying oils, Chem. Rev. 64(6): 591 (1964).
28. N. A. Khan, Some specific features in autoxidation of methyl linoleate. Can. J. Chem. 37: 1029
(1959).
29. W. J. Bailey and G. L. Barlow, Preparation of crystalline linoleate hydroperoxide for drying oil
research, J. Paint TechnoL 42(544): 287 (1970).
30. M. Loury and M. Forney, Etude de F autoxidation de l ’acide linolénique. Rev. Corps Gras 16:
167 (1969).
31. M. Loury and M. Fomey, Étude de T autoxidation de l’acide linoléique, Rev. Corps Gras 15:
663 (1968).
32. E. N. Frankel, Lipids 16 (5): 279 (1981).
33. R. A. Hancock, N. J. Leeves, and P. F. Nicks, Studies of autoxidation. Part I. The volatile by
products resulting from the autoxidation of unsaturated fatty acid methyl esters. Prog. Org. Coat.
17: 321 (1989).
34. N. J. Leeves, Autoxidative degradation of unsaturated fatty acid esters. Thesis, Royal Holloway
College, Univ. London, 1985.
35. L. A. O ’Neill and N. R. Falla, Structure of the cobalt drier complex, Br. Polym. J. 5: 115 (1973).
36. W. A. Waters, The kinetics and mechanism of metal-catalysed autoxidation, J. Am. Oil Chem.
Soc. 48: 421 (1971).
37. J. H. W. Turner and P. Womersley, Aluminum compounds as major components of surface coat
ings, J. Oil Colour Chem. Assoc. 62: 419 (1979).
38. J. H. W. Turner and P. Womersley, Coordination drying: an alternative to catalysed autoxidative
polymerisation for paints, Chem. Ind., March 1975, p. 203.
39. K. H. Zabel, J. C. Hubert, W. J. Muizebelt, and R. A. M. Venderbosch, Singlet oxygen cross-
linking in high solids alkyd paints, ACS Polym. Mater. Sei. Eng. 70: 64 (1994).
40. IC I, Eurp. Patent Spec. 0 009 356 (1980).

4L IC I, Germ. Off. 2,350,646 (1974).
42. R. A. Hancock, N. J. Leeves, and P. F. Nicks, Studies in autoxidation. Part II. An evaluation o f
the by-products formed by the autoxidation of fatty acid modified alkyd resins under the influence
of different promoters. Prog. Org. Coat. 17:331 (1989).
43. 0 . S. Privett, M. L. Blank, J. B. Covell, and W. O. Lundberg, Yellowing o f oil films, J. Am.
Oil Chem. Soc. 38: 22 (1961).
44. S. E. L. K. Davison, Chemical processes accompanying the autoxidation o f paint films. Thesis,
Royal Holloway and Bedford College, Univ. London, 1989.
45. H. G. McAdie and R. V. L. Nicolls, in Film Formation and Film Degradation (C. R. Bray don,
Ed.), Interscience, New York, 1958.
46. T. Helmen and M. Möller, series of 15 papers entitled Förändringar i Klarlackers Kemiska och
Fysikaliska Egenskaper vid Utomhus-och Laboratorieprovning, with the last in Färg Lack 15: 155
(1969) (in Swedish).
47. H. W. Chatfield, Varnish Constituents, Leonard Hill, London 1953, Chap. 4.
48. J. F. Rooney, Practical application of alkyds. Off. D ig. , August 1964, Part 2, p. 32.
49. W. M. Kraft, High polymer alkyd technique. Am. Paint J. 47(28): 6 (1957).
50. H. Tobita and Y. Ohtani, Control of molecular weight distribution in step growth polymerisation:
the intermediate monomer feed method. Polymer 55:801 (1992).
51. H. Tobita and Y. Ohtani, Control of molecular weight distribution in step growth polymerisation
by an intermediate monomer feed method: effect of interchange reactions. Polymer 33: 2194
(1992).
52. E. G. Bobalek, E. R. Moore, S. S. Levy, and C. C. Lee, Some implications of the gel point
concept to the chemistry of alkyd resins, J. Appl. Polym. Sei. 8: 625 (1964).
53. L. A. Tysall, Calculating Techniques in the Formulation of Alkyd and Related Resins, Paint Re
search Association, Teddington, 1982.
54. T. C. Patton, Alkyd Resin Technology: Formulating Techniques and Alkyd Calculations, Intersci
ence, New York, 1962.
55. J. J. Bernardo and P. Bruins, Gelation prediction and related concepts, J. Paint Technol. 40:
558 (1968).
56. P. J. C. Nelen, Recon 36X, a computer program for the design of polycondensation resins, 17th
FATIPEC Congress 1984, v.4, p. 283-298.
57. AMOCA Chemicals Company, “Polycalc II,” polyester characterisation computer program. Tech.
Rep. No. GTSR48; Perstorp Polyols, Guidelines to Perstorp Alkyd Calculations—Percale 3.0,
Tech. Info. Leaflet No. 0005.
58. W. E. Shackleford and D. W. Glaser, Thixotropic modification of alkyd resin paints with polyam
ides, J. Paint Technol. 38: 293 (1966).
59. 1. Garbett, Long life air drying paints, P.R.A . 2nd Middle East Conference, Bahrein 1995, Paper 3.
60. J. C. Martin, Silicone copolymers. Paint Polym. Colour J. 777(4182): 20 (1987).
61. IC I, U .K . Patent 1,370,914 (1972).
62. A. S. Baker, Applications of polymeric surfactants. Paint Resin Manuf., Mar/Apr. 1982, p. 37.
63. C. A. Hare, Anatomy of paint— alkyd resins, J. Prot. Coatings Linings, February 1994, p. 97.
64. Eastman Chemicals, U .S. Patent 5,378,757 (1995).
65. K. Holmberg, Polymerisable surfactants. Prog. Org. Coatings 20: 325 (1992).
66 . D. Fenstel, Emulsifier systems for modem surface coatings. Surface Coatings Australia, January/
Febmary 1994, p. 6 .
67. G. Österberg, M. Hulden, B. Bergenstähl, and K. Holmberg, Alkyd emulsions. Prog. Org. Coat
ings 24: 281 (1994).
68 . BASF, Eur. Patent 87/00589 (1987).
69. IC I, Ger. Offen. 2,818, 102.
70. L. W. Hill and Z. W. Wicks, Design considerations for high solids coatings. Prog. Org. Coatings
10: 55 (1982).
71. K. E. J. Barrett and R. Lamboume, Air drying alkyd paints, J. Oil Colour Chem. Assoc. 49:
443 (1966).
736 Bentley
72. K. B. Gilbes and T. Hunt, New oils and fatty acids for surface coating resin manufacture, J. Oil
Colour Chem. Assoc. 51: 399 (1968).
73. R. Engel, Cascade molecules, Polym. News 17: 301 (1992).
74. B. Pettersson and K. Sorenson, High-solid alkyds based on hyperbranched (dendritic) polymers—
a new concept with new opportunities, Proc. Waterborne, High Solids and Powder Coatings Sym
posium, New Orleans, 1994, p. 1-1 3 .
75. Shell Chemicals, Epikote Resins for Paints, publication Res: 64:4, Shell Chemicals, London.
76. E. I. DuPont de Nemours, U .S. Patent 4,042,478 (1977).
77. K. E. J. Barrett (Ed.), Dispersion Polymerisation in Organic Media W iley, New York, 1975.
78. C. W. A. Bromley, Non-aqueous dispersion polymer microgels as film formers in architectural
paints, J. Coatings Technol. 67(768): 39 (1989).
79. P. K. Nielsen and J. H. Hanson, Paint and pollution— a question of solids, Fdrg Lack Scand.
6:113 (1992).
80. I. Sarvimaki, The implications of life cycle assessment for R&D in the paint industry. Surf. Coat
ings Int. 1994(S): 339.
81. P. J. A. Geurink and E. L. J. Bancken, Life cycle assessments of decorative paints, PRA 12th
Int. Conf, Waterborne Coatings, Milan, 1992, Paper 12.
82. E. M. S. van Hamersueld, F. P. Cuperus and J. T. P. Derksen, Innovation in paint technology
using new vegetable oils and novel concepts, PRA 15th Int. Conf, The Future o f Industrial Coat
ings, Brussels, 1995, Paper 7a.
83 . J. T. P. Derksen, F. P. Cuperus, and P. Kolster, Renewable resources in coatings technology: a
review. Prog. Org. Coatings 27: 45 (1996).
28___________
Lubricants
Theo Mang
Fuchs Petrolub AG Oel + Chemie, Mannheim, Germany
I. INTRODUCTION
Environmental awareness, the desire to protect resources, and the aim of agricultural produc
ers to use available capacities for the cultivation of technical raw materials—harvestable raw
materials—have led to a new generation of lubricants and functional fluids. However, it is
increasing environmental awareness that has been most instrumental in accelerating the devel
opment of rapidly biodegradable lubricants and greases.
Although lubricants are not especially hazardous from an environmental point of view,
their innumerable applications make them omnipresent. Mineral oil, representing the largest
single component in lubricants (>90% ), and some chemical additives are environmentally
undesirable because of their poor biodegradability. This is particularly true from the point of
view of protecting soil and groundwater.
Mineral oil products were not always used as the base substance during the long history of
lubrication technology, or tribology, as the scientific study of friction, wear, and lubrication
is called today. Until the second half of the 19th century, nearly all lubricants were based on
natural vegetable or animal oils and fats. Only after its industrial production and refining was
mineral oil used as a base oil for lubricants. Refined mineral oils displayed much better
thermal and oxidative stability than natural fatty oils. With the development of modem engi
neering practices and the help of chemical additives, it became possible for these “inert”
mineral oils, which alone are not particularly good lubricants, to deliver the desired proper
ties. The additive technology developed since the early 1950s has now gained a thoroughly
“high tech” character. It appeared that mineral oil was seen as a solvent for the chemical
additives and that the disadvantages of mineral oil compared to natural fatty oils could be
more than balanced out.
Fortunately, lubricant developers never forgot the benefits of natural fats and their deriva
tives; some are still used in the form of refined oils and chemically modified substances as
additives because to their good lubricity. In the area of metalworking oils in particular, the
use of natural raw materials never went out of favor. As metalworking, wire drawing, and
737
738 Mang
press oils form a particular focal point in some specialized laboratories, the use of rapeseed
oil, oxidized rapeseed oil, rapeseed oil esters, and other natural additives was never aban
doned.
As long ago as the late 1970s, the lube industry began developing environmentally harm
less and non-workshop-polluting metalworking oils. This work was successfully continued in
the 1980s within the scope of large-scale projects sponsored by the Federal Ministry for Re
search and Technology in Germany, for example.
At the beginning of the 1970s, at the time of the oil and energy crisis, scientists throughout
the world began investigating the possibilities of using harvestable “alternative” raw materials
in order to conserve resources. Much of the work done at this time was put aside when oil
prices fell. Natural fatty base oils for lubricant development regained favor in the early 1980s,
and this f i t well into development programs for environmentally harmless and non-workshop-
polluting products.
II. THE SEARCH FOR ENVIRONMENTALLY FRIENDLY

LUBRICANTS—THE DRIVING FORCE BEHIND T H E USE
OF VEGETABLE OILS AND THEIR DERIVATIVES
Apart from the efforts made in agriculture to find new areas of application in non-foodstuff
sectors, the search for environmentally harmless lubricants is the main reason vegetable oils
and their derivatives are used as base oils. To understand the development of lubricants based
on natural fatty oils and their derivatives, one must examine the criteria used to define the
environmental compatibility of lubricants. It then becomes clear that vegetable oils and their
oleochemical derivatives have significantly better chances than petrochemical alternatives even
though industrially manufactured synthetic carboxylic acids have produced good results in
lubricants. In terms of cost also, natural oils and their derivatives are thoroughly com
petitive.
More than one-third of lubricants sold end up polluting the environment either via total-
loss applications, spillages, evaporation, or in other ways (Fig. 1). In western Europe, 1.5-
Re-used or disposed“
of by customers
11%
B urnt in engines
6%
Collected as waste
oil
47 %
Lost in circulation
systems
28 %
Total-Loss
lubrication
8%
Fig.1 Lubricant consumption and lubricants in the environment in Western Europe.

Lubricants 739
2.0 million tonnes per annum (tpa) pollute the environment. This fact and ever-increasing
environmental awareness led to the development of environmentally harmless lubricants at the
beginning of the 1980s. “Environmentally harmless” means in particular, rapidly biodegrad
able and ecologically nontoxic. A significant increase of such lubricants is expected in the
next few years, helped to some extent by the agricultural/political situation in Europe. Agri
culture will get the opportunity to produce technical raw materials.
Rapidly biodegradable two-stroke outboard marine engine oils were developed as long ago
as the late 1970s for ecological reasons. But the use of rapidly biodegradable base oils for
other lubricants only began in the early 1980s. Predominant among these were total-loss lubri
cants based on vegetable oils, especially rapeseed oil. Since then, the additive industry has
supported these developments with environmentally harmless lube additive packages.
Lubricant manufacturers and users are now in a continuing process of incorporating current
know-how into further product improvements and adapting machine designs to this new gener
ation of lubricants [1-4].
A. Biodegradability and Additional

Environmental Requirements
The meaning of “environmentally friendly” or “ecologically compatible” has been subject to
much debate over the last few years, with existing conventional lubricants used as the yard
stick. With regard to ecology, lubricants are not a particularly critical group of substances,
but they are omnipresent, and mineral oil as well as some chemical additives are not desired
in the environment. “Bio-oils” are also a good idea if they are used in biologically inactive
areas, thus providing additional groundwater protection.
B. Biodegradability
Rapid biodegradability is a generally desirable feature for substances that may eventually enter
the soil and waters and also for biological purification plants.
Evaluations are based on laboratory procedures because the numerous systems presently
used have greatly differing microbiological degradation conditions. An important aspect of
natural degradation is the distribution of the substance to be degraded. Biodegradation is more
rapid if the substance is finely distributed in the soil or water and there is sufficient oxygen.
If a substance reaches the biologically inactive (deeper) reaches of the ground because it
degrades too slowly or because the ground is too porous, it can become a long-term source
of water contamination.
The following laboratory biodegradability tests are generally used in Germany and in other
European countries (test duration in parentheses): closed bottle test OECD 301 D (28 days);
AFNOR test OECD 301 A (28 days); MITI test OECD 301 C (14 days); OECD screening
test OECD 301 E (19 days); Sturm test EC 79/831 (28 days); CEC test CEC-L-33 (21 days);
Zahn-Wellens test OECD 302 B (28 days). These international tests were developed mainly
for water-soluble substances. Only the CEC test was developed for lubricating oils (two-stroke
outboard engines).
In general, the tests use oxygen consumption, carbon dioxide production, the drop in DOC
(dissolved organic carbon), or the reduction in CH valency fluctuations in the infrared spec
trum (CEC test) to measure biodegradation. The duration of the tests is 14-28 days; de
pending on the test, rapid biodegradability means degradation of (at least) 60-80% . The
Zahn-Wellens test defines a potential degradation of >20% and is of significance for the
evaluation of some additives. Even though considerable variations can occur between different
740 Mang
ÖVO 6 4 7 4 -
RAL UZ 64 4— | 1
(RAL UZ 48)4-^ |
ICOMIA 3 8 -8 8 ^ — | j i
~ r
Mineral oil
1
M
’ M
• 1
Polyethylene 1
1
g ly co ls 1
White oil
Hydrocracked 1
base oil
PAO 1
1
Diesters
Polyolesters
1
Vegetable oils
1
1
_L
10 20 30 40 50 60 70 so W 1CK)
Biodegradability, %
Fig. 2 Biodegradability of base oils. Ovo, Austrian law, May 1, 1992; RAL, German authority for
labeling (ecolabel for lubricants, 1988, 1991); ICOMIA, International Council of Marine Industry Asso
ciation (two-stroke oils), 1988.
test procedures on some substances, the basic statement on a substance’s degradation remains
valid, and as a rule the relative order stays the same in a comparison of tests of a given series
of substances.
The biodegradability of used lubricants can be altered by contamination. Measurements
have shown that a deterioration of up to 15% can occur. This means that a lubricant that is
90% degradable when fresh may only be 75% degradable when used.
Figure 2 shows the CEC biodegradability of various base oils and the threshold values
contained in European specifications for rapidly biodegradable lubricants.
C- Toxicity Values
To develop environmentally compatible lubricants, both biodegradability and ecological/toxi-
cological criteria must be taken into consideration. The aim is to protect life in various areas,
especially in waters (aquatic areas) and in nonaquatic areas (terrestial areas). The following
ecological/toxicological test procedures are of significance:
Lubricants 741
Bacteria toxicity according to DIN 38 412, part 8 ; this determines cell multiplication inhibi
tion (EC 10 and EC 50 values). The Pseudomonas type used for the test is found in waste-
water and in the soil.
The bacteria toxicity test according to ISO 8192 determines acute toxicity by the inhibition
of oxygen consumption; results of this test are EC 50 values.
The algae toxicity test according to DIN 38 412, part 9 is yet another test for aquatic
systems (measurement of chlorophyll fluorescence and determination of ECjo and EC50
values).
One of the most important test procedures in German legislation concerning aquatic areas
is a test on small living organisms {Daphnia magna Straus, water flea, small crustacean)
called the Daphnia test, according to DIN 38 412, part 11 or OECD guideline 202. In the
aquatic area, fish toxicity according to DIN 38 412, part 15 performed on a special fish
{Leuciscus idus) is of importance. Test results are the LC q, LC50, and LCjoo values. The
German environmental seal also incorporates fish toxicity according to OECD guideline 203.
Possible pollution of the nonaquatic terrestial area, i.e., soil and plants, is evaluated by the
plant growth test according to OECD guideline 208 (e.g., testing of wheat, cress, and rape
seeds).
Naturally, toxicity testing for environmental protection purposes must also include mamma
lian and human toxicity, all the more so if safety-at-work and environmental protection con
cerns are combined. Toxicity terminology and toxicity classes according to national laws
and EC guidelines need to be considered. The lethal dose LD 50 is an important measure
of toxicity.
D. Mobility
If one includes groundwater in environmental protection, the mobility of lubricants in the
ground is a further measure of environmental compatibility. Rapidly degradable but easily
soluble substances can drain through the soil and reach underground waterways. This fact has
led to criticism of the use of water-soluble, rapidly degradable glycols in lubricants, especially
hydraulic fluids.
III. ENVIRONMENTAL LEGISLATION

German legislation does not prohibit the use of mineral oil based lubricants for any applica
tion. However, a series of environmental protection provisions and orders encourage the use
of rapidly biodegradable lubricants.
A. Water Protection
German water legislation defines the water pollution potential of substances. The arithmetic
average of three toxicity values (mammalian toxicity, fish toxicity, and bacterial toxicity)
forms the basis of a water pollution number (WGZ). Water pollution numbers 0 -6 form the
basis for four water pollution categories: WGK 0 (WPN 0-1.9), generally non-water-pollut
ing; WGK 1 (WPN 2.0-3.9), slightly water polluting; WGK 2 (WPN 4.0-5.9), water pollut
ing; and WGK 3 (WPN 5.9-6.0), highly water polluting. Merit ratings, e.g., for biodegrad
ability and bioaccumulations, are included in water pollution categories along with water
pollution numbers. Ratings within a list are decided by an Environment Ministry commission.
At present, the following ratings apply for lubricants:
742 Mang
Volume in WGKO W GKl WGK 2 WGK 3
<0.1 stage A stage A stage A stage A
0.1 -1 stage A stage A stage A stage C
1.0» 10 stage A stage A stage B stage D
10 » 100 stage A stage A stage B stage D
100 - 1000 stage A stage B stage D stage D
> 1000 stage A stage C stage D stage D
Fig. 3 Definition of water pollution potential according to four hazard stages.
WGK 0 Some rapidly biodegradable lubricants, most of the natural fatty oils
WGK 1 Mineral oils without additives
WGK 2 Mineral oils with additives
WGK 3 Water-miscible mineral oils with additives
EC guideline 91/325/EWG, 1991 specifies the criteria contributing to the description “haz
ardous to the environment” for aquatic and terrestial areas. Step by step this is being adopted
into national environmental legislation and will lead to the classification of numerous lubricants.
Water pollution categories for water-polluting substances have a major influence on the
storage and handling of these substances. Considerably greater care must be taken with highly
water polluting substances than for less polluting substances. The hazard potential based on
water pollution categories and quantity is divided into four levels when manufacturing, pro
cessing, and application plants for water-polluting substances are planned. The highest hazard
level is achieved when only 1 m^ of cutting fluid (WGK 3) is stored (Fig. 3).
Rapidly biodegradable lubricants can attain WGK 0 if suitable additives are used, and this
has become a general aim of development work.
The German wastewater levy provides a further motivation to develop rapidly biodegrad
able lubricants. Industrial customers have to pay additional charges if nondegraded organic
substances (CSB) are disposed of into waters. This logically raises the level of interest in
easily degradable substances.
Special requirements apply to certain protected water zones. In some areas, prohibitions
exist for the transport of water-polluting substances, the use of grass mowers with internal
combustion engines, or DIY engine oil changing.
B. Soil Protection
A certain degree of soil protection is offered by water protection legislation. A specific law
for the protection of soil is currently being drafted in Germany and will be influenced by the
previously mentioned EC guideline.
Lubricants 743
In Europe, the “Holland list” applies for the clean-up of soil contaminated with mineral
oils. This states that a clean-up is necessary if the contamination is greater than 500 mg/kg.
At present clean-up costs in Germany are about DM 1500/m^. This can be reduced to DM
150m^ if the lubricant is rapidly biodegradable.
Austria was the first country to impose a ban on mineral oil based chain saw oils (effective
May 1, 1992). This ban requires this type of total-loss lubricant to be at least 90% biodegrad
able (CEC test) along with fulfilling the other technical specifications. In addition, the prod
ucts must not be water-soluble, which eliminates glycols as base fluids and promotes the use
of natural fatty oils and some derivatives.
C. Safety at Work, Law on Chemicals

The development of rapidly biodegradable lubricants must also take into account, as pre
viously mentioned, other ecological criteria, particularly the selection of nontoxic raw materi
als. Such products are especially interesting in terms of safety at work and thus their effect
on workshop personnel. The German Law on Chemicals, which contains numerous European
guidelines, categorizes substances according to their hazard potential and establishes classifi
cation regulations. The German environmental seal for rapidly biodegradable lubricants se
verely limits the use of hazardous substances as defined by the Law on Chemicals.
D. Keeping the Atmosphere Clean

The evaporation losses of rapidly biodegradable lubricants are generally much lower than
those of conventional oils. This can assist in meeting emission threshold requirements. An
important example is the significant reduction of diesel particulate emissions by engine oils
that display low evaporation losses (Fig. 4).
IV. ENVIRONMENTAL AWARDS FOR LUBRICANTS

More and more countries are awarding environmental seals for products that are environmen
tally friendly. In Germany, the first environmental seal (Blue Angel) was awarded in 1988
Evaporation loss, %
A lk^- Solvent VHVI hydro- PAO Polyol- Rapeseed

benzenes neutral cracked esters oil
Viscosity ISO VG 32
Fig. 4 Evaporation loss of lubricant base oils (Noack evaporation, 250°C, 1 h).
744 Mang
for rapidly biodegradable chain saw oils. In 1991, the scope of this award was extended to
cover rapidly biodegradable lubricants and mold release oils. A third award was being pre
pared for hydraulic oils in 1995.
The German “Blue Angel” is a quasi-official seal with a considerable public profile. It is
awarded by a panel of experts that includes members of the Federal Environment Ministry.
If the term ‘‘environmentally friendly” is defined in line with the two German environmen
tal seals for lubricants, the following requirements result:
1. The products must not contain substances that need to be classified according to the
law on hazardous substances (either under the threshold concentration that leads to
classification or absolutely none).
2. None of the components must be in water pollution categories 2 or 3.
3. The products must not contain any organic chlorine bonds or nitrite.
4. The products must not be ecologically or toxicologically problematic. They must pass
OECD guidelines 202 (Daphnia test), 203 (acute fish toxicity test), and 208 (growth
on higher forms of plants test).
5. Each of the fluids must be at least 70% biodegradable according to OECD guideline
301 A-E and E G C 3-7. Biodegradability according to CEC-L-33-T-82 is also a possi
ble test procedure (for which the RAL environmental seal UZ 64 requires at least
80% degradation).
The additive content must not exceed 5%. The additives must be potentially degrad
able (Zahn-Wellens test).
7. Ecological and toxicological threshold values apply to the additives.
In December 1991, the EC environment ministers decided to introduce a European environ

mental seal. Lubricants are not yet included in the first 12 product groups.
V. BASE OILS FOR RAPIDLY

BIODEGRADABLE LUBRICANTS
In line with the previously described ecological/toxicological requirements, the selection of
base oils concentrates mainly on ester substances. A wealth of know-how has been gathered
over the years for this group of substances in the lubricant area even though their use as base
oils has been limited to a few specialized applications.
A. Natural Fatty Oils

Natural fatty oils such as castor oil, palm oil, rapeseed oil, neatsfoot oil, lard, and degras or
sperm oil have been used in lubricants for years. Over the past decade, these oils have been
used mainly as additives. Since the early 1980s, rapeseed oil has gained increasing acceptance
as a base fluid because it represents a good compromise in terms of availability, price, low
temperature properties, viscosity, and thermal/oxidative behavior when suitable additives are
used for some applications in the temperature range between —25 and +80°C.
Overall, one can say that as the concentration of saturated long-chain fatty acids in the
triglyceride molecule (e.g., palmitic acid, stearic acid) increases; the low temperature behav
ior worsens. As the polyunsaturated fatty acids (e.g., linolenic acid, linoleic acid) increase,
the thermal/oxidative resistance worsens. Even very long monounsaturated fatty acids (e.g.,
erucic acid, 22:1) worsen the low temperature behavior. Today’s low erucic acid rapeseed oil
has considerably better low temperature characteristics than the rapeseed oil of the 1970s,
which contained erucic acid [8,9].
Lubricants 745
High oleic sunflower oil (HOSO)
HOSO 4- Antioxidant
h
Mineral Oil
Mineral Oil + Antioxidant

F
I---- !----- 1-----1-----1-----1----1------1-----1-----1-----1----1------ 1-----1-----1-----1---- 1------1-----i-----1-----1----1------1
0 50 100 150 200
--------------------► time (minutes)
Fig. 5 Improvement of oxidation sensitivity of high oleic sunflower oil (rotary bomb test). (From
Ref. 10).
B. Cultural Improvements to Natural Fatty Oils for Use

in Lubricants
From an application point of view, the objectives of the cultivation of “technical rapeseed” is
to achieve a different fatty acid pattern with a higher content of monounsaturated oleic acid
and less double and triple unsaturated (linoleic and linolenic) acids. A high oleic acid sun
flower oil (~ 90% oleic acid) is presently undergoing technical trials. High oleic acid rapeseed
oil (70% oleic acid) is already available in limited quantities. Apart from these objectives of
obtaining high oleic acid products by cultural improvements, efforts are also being made to
produce linoleic acid-free rapeseed oil.
Figure 5 shows how the sensitivity of a new high oleic acid sunflower oil to oxidation can
be improved by addition of an antioxidant [ 10].
The weakness of vegetable oil based lubricants can best be described by the structural
elements of the double bond and the /3-CH group (they cause thermal and oxidative instabil
ity). The ester group can hydrolyze, i.e., it can be split by water [11].
Lubricant applications generally employ fully refined rapeseed oil, i.e., foodstuff quality
oil that has undergone all refining stages such as dewatering, desliming, neutralization,
bleaching, and steaming (water vapor). These, however, still do not guarantee optimum lubri
cant suitability.
C. Chemical Modification of Natural Fatty Oils

The agricultural improvement of natural fatty oils, in particular rapeseed and sunflower oils,
is a slow process. Moreover, their potential for use in lubricants is not sufficient to optimize
the objectives for lubricants when new plants are cultivated or genetic alterations are made.
The question, therefore, is Can fundamental processes such as chemical modification lead to
more rapid success? An interesting example from lubrication technology is the manufacture
of 12-hydroxy stearic acid from castor oil. 12-Hydroxy stearic acid as an alkaline or earth
alkaline soap (mainly lithium soap) is the most important thickener for greases and is obtained
by hydrogenation.
746 Wang
Selective hydrogenation is a process to remove polyunsaturated fatty acids and thus im
prove the oxidative stability of vegetable oils. In this regard, the work of Behr et al. [12]
should be consulted. In this work, polyunsaturated fatty acids in rapeseed oil and rapeseed
fatty acid methyl esters were reduced to 1% without significantly increasing the stearic acid
content. The latter would negatively affect cold flow properties. Also important is the fact
that the cis-ltrans 18:1 ratio remains very high because large amounts of 18:1 in the trans
form also detrimentally affect cold flow properties.
Further modifications to the multiple double bonds to improve thermal oxidative stability
are possible, and it is conceivable that more cost effective processes will be developed to
increase the use of natural fatty oils that have been chemically modified.
D. Synthetic Esters
The term “synthetic esters” covers a broad range of chemicals with differing qualities and
prices. Their base components (alcohols and carboxylic acids) may have a natural or synthetic
origin (0-90% content of naturally grown components). They can be more or less biodegrad
able, depending on their structure; their thermal/oxidative stability and cold flow properties
can be much better than those of triglycerides. The toxicological properties are in many cases
comparable to those of naturally grown fatty oils.
Fig. 6 Hydrolysis and oxidation of various base oils according to H. G. Schmidt. (From Ref. 11.)
Lubricants 747
Table 1 Time Line for the Development of Synthetic Esters as

Lubricant Base Fluids
1937-1944 Production and research on 3500 ester-based oils.

In Germany, basic acids from the coal chemistry
(from phenols and cresols).
1940- Esters from neopentyl polyols (neopentyl glycols,
trimethylolpropane, pentaerythritol). Produced by
IG Farben.
1960- Great importance of aviation jet engines with me
dium-chain carboxylic acids.
1970- Importance of other synthetic lubricants.
1985- Importance of biodegradable lubricants.
2000- Increasing importance as lube base oils.
Chemistry offers a wide variety of possibilities in the area of synthetic esters. A major
development objective is to find inexpensive esters among all the possible substances that
meet the specific requirements of a particular application.
Before ecological aspects became part of lubricant development, ester oils were used in
special lubricants for technical reasons, e.g., as base fluids for aviation turbine oils and com
ponents for fuel economy oils.
The most important groups of esters include monoesters, diesters, polyol esters, and com
plex esters. At present, polyol esters such as trimethylolpropane esters (TMP esters) or penta
erythritol esters dominate. The basis of these are mainly alcohols from the petroleum and
chemical industries and fatty acids derived from natural oils.
As for hydrolytic stability, “normal” polyol esters (trimethylolpropane or glycerol trioleate)
differ only slightly from rapeseed oil; the difference in oxidative resistance is much greater.
Both characteristics are significantly improved with complex esters. Normally an improvement
in hydrolytic stability worsens the substance’s biodegradability. However, there are complex
esters [medium chain length saturated fatty acids on trimethylolpropane (TMP) or glycerol]
that combine excellent thermal/oxidative characteristics with good hydrolytic resistance and
good biodegradability. Figure 6 compares the oxidative stability and hydrolytic stability of
rapeseed oil to those of glycerol trioleate and TMP trioleate and a complex ester.
With regard to the important application criterion of oil viscosity, the combination of dicar-
boxylic acids (e.g., azelaic acid and sebacic acid) with fatty acids is an excellent instrument
for adjusting viscosity.
Table 1 shows the historical development of esters as lubricant base fluids.
Table 2 lists esters that are presently offered by the oleochemical industry for use in lubri
cants. The table is arranged by product viscosity at 40°C. This is one of the most important
values used to define the possible applications of a product. Low temperature behavior is
described as the pour point.
E. Polyalkylene Glycols
As with esters, polyalkylene glycols (aliphatic polyethers) are highly diverse and display
greatly differing characteristics (e.g., water-soluble and non-water-soluble products). Water-
miscible ethylene glycols are well known as lubricant base fluids and have achieved consider
able significance in fire-resistant HFC hydraulic fluids. They are also easily biodegradable.
Compared to esters, they suffer from the disadvantage that they are highly mobile in soil and
748 Mang
Table 2 Oleochemical Ester Base Fluids for Lubricants
Pour point
Viscosity cold flow
at 40°C Viscosity behavior
(mm^/s) index (°C) Chemical name
4,7 313 -15 Methyl oleate

4,7 207 16 Methyl tallowate
5,2 173 12 Isopropyl palmitate
5,4 205 -12 Methyl colzate
5,8 237 -18 Methyl erucate
7,7 148 < -5 3 Di-(2-ethylhexyl) adipate
8,0 196 -30 2-Ethylhexyl colzate
8,6 196 -4 2-Ethylhexyl palmitate
8,6 193 -21 2-Ethylhexyl oleate
8,9 179 -2 1 2-Ethylhexyl oleate
8,9 179 -6 2-Ethylhexyl tallowate
9,2 187 3 Isotridecyl stearate
10 246 -10 w-Decyl oleate
14 146 < -6 0 Diisodecyl adipate
16 15 1 -27 Capryl-caprinic acid ester
27 148 < -6 0 Diisotridecyl adipate
30 192 -33 Neopentylglycol dioleate
45 198 -4 5 TMP trioleate
53 163 -21 TMP trioleate
60 186 -4 5 TMP ester with long-chain fatty acids
86 157 -9 TMP ester with long-chain fatty acids
88 146 -2 7 TMP ester with long-chain fatty acids
102 146 -3 TMP ester with long-chain fatty acids
148 140 -24 PE tetraisostearate
thus have a greater water pollution potential. In applications where water solubility is an
advantage (e.g., in the foodstuff and beverage industries), ethylene glycols remain an im
portant group of base fluids.
The chemical combination of the above-mentioned esters and ethers can lead to suitable
base oils from both technical and environmental points of view.
VL ENVIRONMENTALLY FRIENDLY ADDITIVES

In applications that use natural and synthetic esters as base fluids, a series of mineral oil
additives would be suitable. However, a selection should be made according to ecological
criteria, especially toxicological values, water pollution categories, biodegradability, and
waste management considerations. The German environmental seal defines the selection crite
ria more precisely. Apart from such selections, there is still a need to develop new additives
whose technical performance and ecological characteristics are optimized.
The German environmental seal for rapidly biodegradable lubricants only approves of addi
tives in water pollution category 1. This raises the question of whether, for example, 0 .5 % of
a water pollution category 1 oxidation inhibitor is a greater hazard to water than 0 . 1% of a
Lubricants 749
water pollution category 2 substance. Such points need clarification. On the other hand, some
additives are unsuitable for rapidly biodegradable, environmentally friendly lubricants; this
applies especially to substances containing heavy metals.
A. Improving Low Temperature Properties

Many synthetic biodegradable esters have excellent cold flow characteristics. The use of envi
ronmentally friendly pour point depressants can give vegetable oils such as rapeseed oil ade
quate long-term cold flow characteristics (rapeseed oil, —30°C). Mixing in easily biodegrad
able synthetic esters can improve the pour point, long-term low temperature behavior, and
low temperature viscosity. Figure 7 illustrates this for a high oleic acid sunflower oil [10].
New laboratory procedures have been developed for vegetable oils that define cold flow char
acteristics over time.
B. Extreme Pressure and Antiwear Additives

Most vegetable oils and synthetic esters used in lubricants display better lubricity in boundary
lubrication conditions than mineral oils because of their polarity. This has been proved by a
series of tests [13,14]. Recently a twin-disk test apparatus showed that the coefficient of
friction of esters and vegetable oils is half that of mineral oils.
Table 3 shows Michaelis and Höhn’s results [14] for rapeseed oil, various synthetic esters,
poly glycols, and mineral oils. The table lists products with viscosities between 32 and 220
mm^/s. The tests were performed on base fluids without additives. The generally used solvent
raffinates were used for the mineral oil tests [14].
Sulfurized esters have established themselves as environmentally friendly extreme pressure
(EP) additives, and the additive sulfur in vegetable oils and esters also has an antiwear (AW)
effect. Sulfur carriers with 15% total sulfur and 5% additive sulfur have proven themselves
particularly effective in rapidly biodegradable esters and vegetable oils [15].
750 Mang
Table 3 Coefficient of Friction of Rapidly Biodegradable

Lube Base Oils
Type of lubricant ISO

base oil viscosity Coefficient of friction ¡i
Rapeseed oil VG 32 0,023

VG 46 0 ,0 2 2
VG 220 0,078
Synthetic esters VG 46 0 ,0 2 2
VG 46 0 ,0 2 2
VG 46 0,045
VG 6 8 0 ,0 2 0
VG 6 8 0,079
VG 220 0 ,0 2 2
VG 220 0,078
Polyglycols VG 32 0 ,0 2 0
VG 6 8 0 ,0 2 0
Refined mineral oils VG 32 0,050

VG 46 0,050
VG 6 8 0,057
VG 220 0,040
Source: Ref. 14.
Table 4 shows a comparison of antiwear values and friction co efficien ts such as are used
in numerous lube laboratories, for example, for a rapeseed oil raffinate and for an equiviscous
VG 32 mineral oil raffinate. Here again, the values show that rapeseed oil has much better
lubricity than mineral oils without additives. One can assume that the favorable friction and
wear behavior of natural fatty acids and most esters leads to a corresponding reduction in
additive content. From an environmental point of view, this reduction in chemical additives
can play an important role in the move toward rapidly biodegradable products.
Table 4 Wear Protection and Friction Behavior for Rapeseed Oil and
Solvent-Refined Mineral Oil (Solvent Neutral VG-32)
Solvent neutral
Test Rapeseed oil VG 32
VKA test according to DIN 51 350 T2 1300-1400 1 0 0 0 -1 1 0 0
welding load (N)

FZG test according to DIN 51 354 T 2
Breakdown load stage 9 6
Spec, wear (mg/kWh) 0.16 0.28

Timken test according to IP 240
OK load (lb) 1 2 10
Tannert test 2070 160

according to DIN 51 387
slide index (N)
Lubricants 751
C. Antioxidants
The focus is on phenolic products more than amine products as antioxidants. The best known
product, BHT (2, 6-di-t^ri-buty 1-1 -4-methyIphenol), is effective at lower temperatures in esters
and vegetable oils. It has the advantage of being in water pollution category 1 and having
FDA foodstuff approval. Another phenolic product with these advantages and high tempera
ture suitability is named [15 ]. Mixtures based on amine components have been used success
fully in biodegradable greases.
D. Viscosity Index Improver

The viscosity index of lubricants shows how much viscosity drops when the temperature
increases. The paraffinic mineral oil raffinates generally used in lubricants today have a vis
cosity index of about 100. For a number of lubricant applications, polymer viscosity index
improvers are used to improve viscosity-temperature behavior. The disadvantage of these
polymer compounds is that they can be mechanically destroyed when subjected to high shear
ing forces. This in turn lessens or eliminates the desired effect, and chemical reactions with
the molecular fragments can negatively influence the lubricant (e.g., sludge formation, depos
its on machine elements).
Natural fatty oils have much better viscosity indices than mineral oils, mostly over 180, so
that no viscosity index improvers are necessary. As no polymer compounds are present in the
lubricant, it is not affected when shearing forces increase. Many synthetic esters have viscos
ity indices over 150, with some over 200.
VII. PLANT OIL AND ESTER-BASED LUBRICANTS

IN THE MARKET
Biodegradable two-stroke oils have been available since 1976. In Germany, this development
was helped by inland boating regulations. The CEC-L-33-T-82 biodegradability test developed
for two-stroke oils was agreed on in 1982. Biodegradable chain saw oils have been available
since 1984, and biodegradable hydraulic oils for almost as long. In 1996, about 35,000 t of
rapidly biodegradable lubricants were sold in Germany, about 4% of total lubricant sales [4].
A. Total-Loss Lubricants
Total-loss lubrication is a technology in which the lubricant performs its function for only a
relatively short period of time before it is lost into the general environment. This type of
lubricant does not create any used products that have to be disposed of because it is designed
to enter the environment. Table 5 shows some typical total-loss applications of rapidly biode
gradable lubricants. These make up about 7-8% of total lubricant sales. In terms of volume,
chain saw oils (0 .3-0.6 L/m^ wood) and some greases are of particular significance [16].
As long-term characteristics of total-loss lubricants, in particular their thermal/oxidative
behavior, are of limited significance, rapidly biodegradable products are relatively simple to
formulate. The first German environmental lubricant award was dedicated to chain saw oils
in 1988 (RAL UZ 48); in 1991 a second environmental award was issued for total-loss lubri
cants and mold release oils (RAL UZ 64).
752 Mang
Table 5 Total-Loss Lubricants and Their Applications
Lubricant Application
Two-stroke oil Air and water cooled two-stroke engines

Chain saw oils Chain saws
Tacky oils Exposed lubrication points on industrial
machinery
Greases Exposed and central lubrication systems
in commercial vehicles and industrial
machinery
Railroad switch greases Railroad switches
Wheel flange lubricants Railroad wheel flanges
Spray lubricants Lubrication and corrosion protection of
machinery
Mold release oils Construction
B. Hydraulic Oils
The development of rapidly biodegradable hydraulic oils attracted attention early on because
of their volume potential (11-14% of lubricant sales) and their ecological relevance in auto
motive hydraulic applications. Lubricant manufacturers, additive producers, and universities
have performed basic research, which is still in progress [17-19]. Of the oils sold in Ger
many, considerably more than half are based on rapeseed oil (~ 40% are based on synthetic
esters). The problems that arose with vegetable oils and esters, especially with respect to
elastomer compatibility [20], nonferrous metal corrosion caused by hydrolysis [21], and vis
cosity increase, could all be clearly eased.
The draft of the German norm for hydraulic oils based on natural triglycerides with HETG
or synthetic esters with H EES or polyglycols are characterized with H EPG [22]. An ISO
standard of international importance is being drafted. New threshold values need to be estab
lished for used oil evaluation; e.g., the common measurement of the neutralization number,
as with mineral oils, is no longer relevant. Figure 8 details the results of a three-year long
term test in two stationary hydraulic systems of deep-drawing presses at a car maker. In spite
of a neutralization number (NZ) of 7 and a viscosity increase of about 40-70, no problems
occurred (Figure 8a).
C. Hydraulic Oils for Mobile Equipment

Good results have been achieved with biodegradable hydraulic fluids in agricultural forestry
and off-road construction. A broad range of applications under mild conditions may be cov
ered using products based on vegetable oils (in Europe, mainly rapeseed oil). More severe
demands concerning high temperature stability or cold flow properties, or both, may be satis
fied by the use of synthetic ester base fluids. Similar aspects should be considered for gear
oils or oils for combined gear and hydraulic systems. They are often used in farm tractors
with mounted agricultural equipment [28].
Special problems are involved with integrated wet brakes, because their function is depen
dent on the friction coefficient. This is different for mineral oils and natural or synthetic esters
or other biodegradable base fluids. Apart from this, we believe that mineral oils can be re
placed by biodegradable products less harmful to the environment in a considerable number
of hydraulic fluids and gear oils in mobile equipment.
Lubricants 753
Viscosity, 40 *C Neulrolizotion number (mgKOH/g)
10
0 2000 4000 6000 8000 10000 12000

Operoting time, h
Fig. 8 Long-term trial of a rapeseed oil based hydraulic fluid in two stationary hydraulic systems.
D. Ester-Based Machine Tool Fluid Family

The development of a new family of fluids based on rapidly biodegradable esters in water
pollution category 0 offers not only considerable ecological but also economic advantages.
This involves the replacement of water-miscible cutting fluids by low viscosity esters. Hydrau
lic oils, gear oils, and other machine lubricants are also formulated on biodegradable esters
that are compatible with the cutting fluids. Even though such products are more expensive
than conventional mineral oil based lubricants and functional fluids, the total process cost
analysis still shows advantages. Along with the general ecological considerations regarding
the handling of water-polluting fluids and disposal costs, the improved workshop conditions
offered by “natural” raw materials play an important role. This model has been successfully
realized in a large diesel engine manufacturing plant for a couple of years now [23,24,27].
Figure 9 illustrates the complexity of the “biofluid” family circuit based on oleochemical
esters. This completely new concept closes the material circle with losses due to only evapora
tion and residual oil on chips.
754 Mang
TO METAL RECYCLING WORKPIECES

WITH RESIDUAL OIL TO SUBSEQUENT
OPERATIONS
Fig. 9 Closed fluid circuit in chip-forming machining for an ester-based machine tool fluid family.
(From Refs. 23 and 24.)
Table 6 shows the viscosities of the most important products in the machine tool fluid
family, and Table 7 comprehensively lists the characteristics of the special biohydraulic oil.
E. Engine Oils
Premixed rapidly biodegradable two-stroke engine oils based on esters have been available
since 1976 and are of particular relevance to ecological demands and endangered waters. In
Germany, Mercedes-Benz was the first company to perform ecological evaluations of engine
oils [25]. These put a certain emphasis on biodegradability. A project initiated by the German
Federal Ministry for Agriculture [26] foresees the development of biodegradable engine oils
based on vegetable oils and their derivatives. At the end of 1993, a leading European manu-
Table 6 Biofluid Family for Machine Tools

Biofluid Viscosity
Cutting oil 8 mm^/s

Hydraulic oil 32 mm^/s
Gear oil 220 and 320 mm^/s
Way lubricant 6 8 and 2 2 0 mm^/s
Other machine tool oils 6 8 mm^/s
Grease N LG I 000
Lubricants 755
Table 7 Specification for Ester-Based Machine Tool Hydraulic Fluids
Characteristic Value Test method
Density at 15°C 914 kg/m^ DIN 51 757

Viscosity at 40°C 32 mm^/s DIN 51 562
Viscosity at 100°C 7.3 mm^/s DIN 51 562
Viscosity index 198 DIN ISO 2909
FZG mechanical > stage 12 DIN ISO 51 354-2
gear rig test
Rotary vane pump test < 1 2 0 mg DIN 51 389-1
Ring vanes < 3 0 mg DIN 51 389-1
Steel corrosion degree 0 -A DIN 51 585
Copper corrosion degree 1-1000 A 3 DIN 51 759
Air release < 5 min DIN 51 381
Démulsification 2 0 min DIN 51 599
Biodegradability >95% CEC-L-33-T-82
Water pollution category WGK 0
facturer of diesel engines issued a first approval for an ester-based oil whose biodegradability
is greater than 80% (Fuchs Plantomot 5W-40).
At present, the use of chemically unchanged vegetable oils with traditional fatty acid pat
terns as base oils for normally lubricated four-stroke internal combustion engines does not
appear to be possible. The previously mentioned research project also foresees a development
in which the continuous burning of some of the engine oil should eliminate aging products.
This would make less thermally stable oils suitable for engine lubrication. A rapeseed oil
based engine oil has been run for about 15,000 miles in the diesel engine of a Volkswagen
passenger car. Figure 10 shows the lubrication model flowchart. A prerequisite is that a ‘Tow
Fig. 10 Vegetable oil based engine oil with partial “total-loss lubrication” in a diesel engine flowchart
for environmentally compatible diesel engine lubrication. 1, Engine; 2 oil sump; 3, tank for fuel-used
oil mixture; 4, fuel-used oil mixture feed to injection; 5, fuel filter; 6 , fresh oil tank; 7, fresh oil dosing
pump; 8 , fresh oil feed; 9, coupled pumps for used oil and fuel; 10, connecting pipes; 11, control device
for oil volume/level; 12, control device for fuel-used oil mixture; 13, pump for internal motor oil
circuit; 14, motor oil; 15, fuel injection pump; 16, 17, control lines; 18, fuel feed from main tank.
756 Mang
chemical” oil is burned with few emissions. In spite of high oil consumption, the injection of
2% used oil into the fuel is an economically acceptable solution.
VIII. SUMMARY AND OUTLOOK

Increasing environmental awareness has intensified the development of lubricants and has
generated new evaluation criteria. The new generation of rapidly biodegradable lubricants was
in its infancy in the 1980s but continues to gain importance in the 1990s. Considerably more
development work needs to be done, and consumers must be informed about the benefits of
lower environmental risks because the products are usually significantly more expensive than
conventional mineral oil based products.
Some of the applications of rapidly biodegradable lubricants and functional fluids men
tioned in this chapter have shown that this area of work can best be promoted by not just
focusing on lubricant developments but also by machine tool manufacturers and their suppliers
demanding new technical solutions for environmentally friendlier lubrication technologies.
Figure 11 shows the possibility of substituting 90% of mineral oil based lubricants with
Factor 1
10%
Factor 2
20 %
Cost factor 1
I Vegetable oils Incl. HO-types and other adapted
cultivation products
Cost factor 2
II IMP oleates, other low-price polyol esters,
chemically modified plant oils, some
dicarbonic acid esters
Cost factor 4
III IMP or glycerol complex esters with saturated fatty acids,
some dicarbonic acid esters
Cost factor > 6

IV Complex esters of various structures,
other saturated special esters
V Not biodegradable
Fig. 11 Technical realization o f rapid biodegradability for 90% of total lubricant volume (hypothesis)
Lubricants 757
rapidly biodegradable products. The rapidly biodegradable base oils are divided into four
groups (I-IV) according to cost factors. Group V is the 10% nonbiodegradable that would
remain in use. By the year 2000, it is forecast that 10% of German lubricant demand could
be met by such products. Base oils will be mainly regular vegetable oils (rapeseed), new
vegetable oils, and synthetic esters of an oleochemical nature. Economically interesting alter
natives for used oils could be treatment and incineration or recycling if larger quantities are in
volved.
REFERENCES
1. T. Mang, S c h m i e r s t o f f e und F u n k t i o n s f l ü s s i g k e i t e n a u f P f l a n z e n ö l b a s i s , E r f a h r u n g e n e i n e s H e r s
t e l l e r s , Schriftenreihe des Bundesministers für Ernährung, Landwirtschaft und Forsten, Reihe A,
Heft 391, 1990.
2. T. Mang, Umweltschonende und arbeitsplatzfreundliche Schmierstoffe, T r i b o L S c h m i e r u n g s t e c h .
3 8 (4); 23 1-2 36 (1991).
3. H. Ihrig, Umweltverträgliche Schmierstoffe in den 90er Jahren, T r i b o L S c h m i e r u n g s t e c h . 3 9 (3):
12 1- 12 5 (1992).
4. T. Mang, Schmierstoffe und Umwelt—die Schmierstoffentwicklung im Einfluss der Um
weltgesetzgebung , 9th Int. Kolloq. Tribologie, Esslingen, Jan. 11-13, 1994, pp. 4.1.1-4.1.10.
5. T. Mang, Lubricants and legislation in the Federal Republic of Germany, E r d ö l K o h l e 42(10):
400-407 (1989).
6. T. Mang, Legislative influences on the development, manufacture, sale and application of lubri
cants in the Federal Republic of Germany, CEC Symp., Paris, Apr. 19-21, 1990.
7. T. Mang, Environmentally harmless lubricants, current status and relevant German environmental
legislation, N L G I S p o k e s m a n 5 7 (6 ) 9-233-15-239 (1993).
8. U. J. Möller, with T. Mang and P. Studt, Gesellschaft für Tribologie, GfT-Arbeitsblatt Pflanzenöle
als Schmierstoffe, Moers, 1993.
9. A. Hubmann, Additivierung pflanzlicher Schmierstoffe, in B i o l o g i s c h s c h n e l l a b b a u b a r e S c h m i e r s
to f f e u n d A r b e i t s f l ü s s i g k e i t e n (W. J. Bartz, Ed.) Expert-Verlag, Ehingen, 1993.
10. K. Lai and V. Carrick, Performance testing of lubricants based on high oleic vegetable oils. 9,
Int. Kolloq. Tribologie, Esslingen, Jan. 11-13, 1994, pp. 2.9.1-2.9.14.
11. H.-G. Schmidt, Komplexester aus pflanzlichen Ölen, 9th Int. Kolloq. Tribologie, Esslingen, Jan.
1 1 - 1 3 , 1994, pp. 2.2 .1-2 .2 .9 .
12. A. Behr, N. Döring, S. Durowicz-Heil, B. Ellenberg, C. Kozik, C. Lohr, and H. Schmidke,
Selektive Härtung mehrfach ungesättigter Fettsäuren in der Flüssigphase, Henkel-Referate 30/
1994, pp. 2 1-26 .
13. S. Odi-Owei, Tribological properties of some vegetable oils and fats, L u b . E n g . 4 5 (11): 685-
690 (1988).
14. K. Michaelis and B.-R. Höhn, Reibungsverhalten biologisch leicht abbaubarer Schmierstoffe, 9th
Int. Kolloq. Tribologie, Esslingen, Jan. 11-13, 1994, pp. 8 .1.1-8.1.8.
15. A. Fessenbecker and J. Korff, Additive für ökologisch unbedenkliche Schmierstoffe, 9th Int. Kol
loq. Tribologie, Esslingen, Jan. 11-13, 1994, pp. 11.12.1-11.14.14.
16. W. Dresel, Schmierfette auf pflanzlicher Basis, in B i o l o g i s c h s c h n e l l a b b a u b a r e S c h m i e r s t o f f e u n d
A r b e i t s f l ü s s i g k e i t e n (W. J. Bartz, Ed.), Expert-Verlag, Ehingen, 1993.
17. W. Backe, Bestandsaufnahme und Trends der Fluidtechnik, Ö l h y d r a u l . P n e u m . 3 6 (1): 16-19
(1992).
18. C. Busch and W. Backe, Biologisch schnell abbaubare Hydraulikflüssigkeiten, T r i b o L S c h m i e r
u n g s t e c h . 4 1 (1): 17 -2 3 (1994).
19. T. Mang, Umweltbedingte Alternativen zu Mineralölen als Grundkomponenten in Hydraulikflüs
sigkeiten, lOth Aachener Fluidtech. Kolloq., Mar. 17-19, 1992, pp. 11-36.
20. K. Nagdi, Dichtungswerkstoffe für umweltfreundliche Flüssigkeiten, Ö l h y d r a u l . P n e u m . 3 4 (1):
42-50 (1990).
21. H. F. Eichenberger, Biodegradable hydraulic lubricant— an overview of current development in
758 Wang
central Europe, SAE Tech. Paper Ser., 42nd Earthmoving Ind. Conf., Peoria, IL, Apr. 9-10,
1991.
22 . E. Pelzer, Normung und Vergaberichtlinien des Umweltzeichens von umweltschonenden Hydrau
likflüssigkeiten, in B i o l o g i s c h s c h n e l l a b b a u b a r e S c h m i e r s t o f f e u n d A r b e i t s f l ü s s i g k e i t e n (W. J.
Bartz, Ed.), Expert-Verlag, Ehingen, 1993, pp. 117-135.
23. T. Mang, C. Freiler, H. Hanagarth, and E. Strobel, Niedrigviskose Esteröle als Kühlschmiers
toff—Grundlagen und Anwendung in der Zerspanung eines Nutzfahrzeugmotorenwerkes, IWT-
Seminar, Bremen, October 1995.
24. T. Mang, C. Freiler, M. Spilker, W. Göttert, H. Hanagarth, and E. Strobel, Hydrauliköle für die
Werkzeugmaschine—Bestandteil einer neuen Fluidfamilie für die spanende Fertigung, Ö l h y d r a u L
P n e u m . 1 0 : 759-762 (1995).
25. Fuchs Mannheim and Motoren-Werke Mannheim AG, R&D project sponsored by the German
Ministry of Agriculture and Forestry, Umweltfreundliche Lösungen zur Schmierung von Dieselmo
toren, 1993.
26. Fuchs Mannheim and Mercedes-Benz AG, Stuttgart, R&D project sponsored by the German Min
istry of Research and Technology, Entwicklung emissionsarmer Schmierstoffe unter Verwendung
nachwachsender Rohstoffe, Published by Bundesumweltamt in UFOKAT ’92, p. 276.
27. Mercedes-Benz AG, Stuttgart, Tech. Paper 1992.
28. K. Höhn, Bio-Hy-Gard in John Deere Traktoren, Mannheimer Morgen, July 30, 1993.
29________________
Epoxidized Oils
Frank D. Gunstone
Mylnefield Research Services Ltd., Scottish Crop Research Institute,
I. INTRODUCTION
Epoxidised oils are made from unsaturated acids or from their glycerol or alkyl esters such as
soybean oil, linseed oil, rapeseed oil, tall oil (fatty acids), oleic acid, or alkyl oleates. For the
most part, they are produced by reaction with hydrogen peroxide in some appropriate form;
most often this is a peroxy acid (RCO 3H).
In conducting epoxidation reactions it should always be remembered that the reaction is
exothermic and that high concentrations of peroxy acid should be avoided. Whatever proce
dure is employed, it is advisable to add the epoxidizing agent slowly or at an appropriate rate
to the olefin. The reverse procedure can be hazardous. Further, it should be noted that the
epoxides themselves are reactive compounds, especially in acidic solution, and efforts must
be made to ensure that the epoxides are not converted to other products following the process
of epoxidation.
Other reviews of this topic are available elsewhere [1,2].
II. NATURAL EPOXY ACIDS

Since the discovery of vemolic (cis-12,13-epoxyoleic) acid in Vernonia anthelmintica seed oil
some 40 years ago [3], several other natural epoxy acids have been recognized (Table 1).
Epoxy acids are among the monomers of cutin and the phytoaxelins (toxic antimicrobials in
plants) [5].
Vemolic acid is still the most readily available natural epoxy acid and continues to be the
most widely studied. It attains high levels in several seed oils (Table 2), and interest has been
expressed in developing some of these as new crops of potential industrial value. This applies
particularly to Vernonia galamensis (in the United States) and Euphorbia lagascae (in
Europe).
These vemolic oils can be epoxidized to produce material rich in diepoxystearate such as
epoxidized soybean oil, but it seems to me that this is not the best approach. Such material
759
760 Gunstone
Table 1 Natural Epoxy Acids
Acid'^ Chain length and unsaturation
9,10-Epoxy 18:0 18:1 ( 1 2 c)'’ 18:1 (12a) 18:2(12cl5c) 18:2 (3il2c)

12,13-Epoxy 18:1 (Çc)"’“ 18:2 (6c9c) 18:2 (6t9c)
15,16-Epoxy 18:2 (9cl2c)
Âll these acids are c is epoxides except for the 18:0 acid, which occurs in both c is and tr a n s forms.
^Coronaric acid.
‘^Vemolic acid.
^ € 2 0 homologue— alchomoic acid.
would probably be more expensive than epoxidized soybean oil, and, more significantly, it is
wasteful of the multifunctionality present in vemolic acid. It would be better to exploit the
fact that vemolic acid contains one epoxide group and one unsaturated center with differing
reactivities and to use the acid as a source of specialty chemicals (i.e., low volume, high
value products). This is discussed further in Section VII.
Useful studies have already been carried out on Vernonia galamensis seed oil. There were
reports in 1990 that this plant was being cultivated in Kenya (—200,000 hectares) and in
Zimbabwe (20,000-30,000 ha) [6], and procedures for extraction of the oil (40-42%) [7] and
its refining [8] have been described. Typical fatty acid and triacylglycerol analyses are given
in Table 3. It is clear that with around 80% vemolic acid this oil contains substantial propor
tions of trivemolin (50-60%) and divemolins (20-30%). It is readily biodegradable [12] by
Acinetobacter and Pseudomonas species.
In Europe attempts are being made to develop Euphorbia lagascae as a new commercial
oilseed crop. The seed contains about 45-50% oil, and the oil yield in the Netherlands is at
present 500-1000 kg/ha. This is expected to rise to about 1200 kg/ha or to even higher values
[13]. The fatty acid composition of this vemolic-rich oil is given in Table 4. The oil is
reported to be about 41% trivemolin and 18% divemolic triacylglycerols.
The extracted oil is bright yellow and has a low level of phosphoms (<1 ppm) and a high
specific gravity. Though more viscous than soybean or rapeseed oil, it is less viscous than
epoxidized soybean oil or epoxidized linseed oil. It can be used, like these epoxidized oils,
as a stabilizer for PVC. But it can also be used as a diluent in paints, avoiding the necessity
for volatile organic solvents. It has been estimated that the addition of a vemolic oil to paint
at around the 10% level would reduce the amount of volatile organic compounds by 160
million pounds in the United States alone [15].
Table 2 Natural Occurrence of Vemolic and Coronarie Acids in Seed Oils
Vemolic acid (12,13-epoxyoleic), wt %

Vernonia anthelmintica 72 Euphorbia lagascae 57-62
V. galamensis 73-78 Erlangea tomentosa 52
Stokes aster 65-79 Crépis aurea 52-54
Cephalocroton cordofanus 62 Crépis biennis 68
Coronarie acid (9,10-epoxy-12-octadecenoic), wt %
Chrysanthemum coronarium 14 Helichrysum bracteatum 14
Xeranthemum annuum 8
S o u rce: Ref. 4.
Epoxidized Oils 761
Table 3 Composition of Vernonia galamensis Seed Oil
Fatty acids, wt %
Vemolic 79-81 Stearic 2 -3
Linoleic 11-12 Palmitic 2 -4
Oleic 4 -6
Triacylglycerols,^ wt %
V. v.x vx. X. Ref.
59.2 28.1 9.5 3.2 9b

50 21 20 9 10"
52 11^*
= vem olic acid; X = acids other than vem olic.
‘"Gas chromatography.
"Reverse phase HPLC.
^Mass spectrometry.
III. THE QUALITATIVE AND QUANTITATIVE ANALYSIS OF

EPOXY OILS
Total epoxide content can be measured by hydrogen bromide in acetic acid (Durbetaki reac
tion [16]), by gas chromatography, and by a colorimetric procedure based on reaction with
picric acid. This last has also been used as a specific spray reagent for epoxides in thin layer
chromatography. Epoxides show characteristic absorption bands in the infrared spectrum at
848 and 826 cm“ ^
The position of an epoxide function in a long chain is easily determined by mass spectrom
etry or by chain fission with periodic acid [3]. Epoxide configuration can be derived from the
NMR spectrum (Section VI) or from the thin layer chromatographic behavior of epoxide-
derived diols on silica impregnated with boric acid [17].
The stereochemical configuration of epoxy acids has been determined by gas chromatogra
phy of appropriate derivatives [18, 19]. If unsaturated, the epoxide is first subjected to mild
catalytic hydrogenation, and the resulting saturated epoxide is converted to an allylic hydroxy
compound by reaction with selenophenoxide anion and hydrogen peroxide. This is converted
to its menthoxycarbonyl derivative, oxidized by ozonolysis, methylated with diazomethane,
and analyzed by gas chromatography using appropriate reference compounds.
Hammond [20] has suggested HPLC examination of unsaturated oils after epoxidation as
an alternative to silver ion chromatography. The oil is epoxidized with m-chloroperoxybenzoic
acid, and the separation is effected on a silica column. Each double bond is replaced by an
Table 4 Composition of Euphorbia lagascaea Seed Oil
Fatty acid
14:0 16:0 18:0 18:1 18:2 Vemolic Acid Other
Whole oil 1.3 3.8 1.5 19.1 8.6 64.3 1.4

2-M A G " - — — 10.2 13.8 75.0 1.0
2-Monoacylglycerol remaining after lipolysis.

762 Gunstone
O OH OMC OMC
/\ I
-C H C H CH .- ■ -CH CH CH , -CH CH =CH "~- - CHCOOMe
I
SePh
epoxide function, with the possibility of stereoisomerism when there is more than one such
group. Though developed as a way of examining unsaturated oils, it can be applied directly
to epoxy oils. The method is less satisfactory at high epoxide levels because of the complica
tions arising from stereochemistry. It would therefore be more suitable for epoxidized palm
oil than for epoxidized soybean or linseed oil. See also Reference 21.
Epoxy esters have long been examined by mass spectrometry, either as epoxides or as
derivatives based on the epoxide function [22]. Marx and Classen [23] describe the mass
spectra of the dimethyloxazoline derivatives prepared by reaction of the epoxy acid with
amino isobutanol [Me2C(NH 2)CH 20 H] alone at 180° for 1 h or in the presence of dicyclohex-
ylcarbodimide at ambient temperatures for 4 h. Many fragments are observed in the mass
spectrum, but the most significant result from cleavage a or to this epoxide group. This is
applied to compounds with up to three epoxide groups, which may be saturated or unsatu
rated.
A
CH3(CH2), CHn CH2CHCH(CH2)„C^
N*
0~J
P a
Gunstone [24] has used NMR to indicate the composition of epoxidized oils. Because
of the presence of saturated acyl chains and chains with one to three epoxide groups, some of
which exist in stereoisomeric forms, the spectra are fairly complex. Nevertheless it is possible
to distinguish between and to determine the levels of triepoxides from a-linolenic acid, diep
oxides from linoleic acid, monoepoxides from oleic acid, and nonepoxy compounds from
saturated acids.
IV. THE EPOXIDATION PROCESS

Before considering examples of epoxidation in Section V, some general comments about the
epoxidation process are in order.
• When carrying out epoxidation reactions it must always be remembered that the reaction
is exothermic, that high concentrations of peroxy acid should be avoided, and that
epoxides are acid-labile.
• The epoxidation reaction is second-order, and reaction rate depends on the concentra
tions of alkene and peroxy acid. The situation is more complex if the reaction is bi-
phasic.
• The rate of epoxidation increases with the degree of substitution as shown in the relative
rates for the following alkenes;
CH2=CH2 CH3CH=CH2 CH3CH=CHCH3 (CH3)2C=CHCH3

25 600 6000
Oleic and linoleic acid are disubstituted alkenes and therefore react quite readily, cis-
Alkenes are more reactive than their trans isomers.
^Mixtures of two compounds, only one structure shown; ^mc = menthoxy carbonyl.
Epoxidized Oils 763
• The reactivity of peroxy acids is related to their values, and the order of reactivity
of the more common peroxy acids is given by the listing
Peroxytrifluoroacetic > monoperoxymaleic > monoperoxyphthalic > p-nitroperoxyben-
zoic > m-chloroperoxybenzoic > peroxyformic > peroxybenzoic > peroxyacetic
These are not pure compounds, and the level of peroxy acid should always be deter
mined before use.
• Reaction rates are influenced by solvent, and epoxidation is facilitated particularly by
halogenated solvents. Many laboratory-scale epoxidations are carried out in dichloro-
methane solution.
• Epoxidation is a cis addition process, so cis alkenes give cis epoxides and trans alkenes
give trans epoxides. A cis, cis diene, such as linoleic acid, will give two racemic
diastereoisomeric bis epoxides. Bartlett’s proposal for the mechanism of this reaction
has been criticized but remains the most widely accepted
/C — H H - C \ H—C /C — H
o(l I / O O o (|
^C—H H C ^C —H
I I
H— C — H H— C— H H—C—H H— C— H
I 1
/ C — H H” C\ C— H H— C n
o( ' 0 '
V -H C— H H— C
I I
\ /
C/% H. ’*0 \ /
11 0 H( ^
/ \ ^ R
/ \
0 R
The epoxides resulting from reaction between alkene and peroxy acid iire reactive com
pounds, especially in acidic solution (Section VII). They are readily converted to mo-
noacylated diols, and this is the major cause for reduced yields. With the more reactive
carboxylic acids it may be necessary to buffer the reaction mixture to inhibit this further
reaction.
O
RCO3H / \ RCO2H
- CH=CH-------------- > -C H C H -------------- > - CH(OH)CH(OCOR)-
The kinetics of oxirane cleavage in epoxidized soybean oil by acetic acid at tempera
tures between 60 and 90° have been reported [25].
The most important alkenes used as substrates for the purpose of this chapter include
1. Glycerol esters containing olefinic acids, especially soybean oil and linseed oil, but
also palm oil and some others
2. Alkyl esters of olefinic acids such as oleic acid or tall oil fatty acids
3. a-Olefins (R C H = C H 2) produced by oligomerization of ethene.
The reagents most used on a commercial scale for the epoxidation of the materials listed
above are peroxyacetic acid in various forms or peroxyformic acid. Other peroxy acids
or mixtures of hydrogen peroxide and catalyst are used mainly for smaller scale or
laboratory operations.
764 Gunstone
* Peroxyacetic acid can be made from acetic acid and hydrogen peroxide or by atmo
spheric oxidation of acetaldehyde (ethanal). It is available from a major supplier in a
number of forms:
1. Peroxyacetic acid in excess of acetic acid prepared in situ, i.e., in the presence of
the alkene to be epoxidized.
alkene
C H 3C O 2H + H 2O 2 ^... ^ C H 3C O 3H + H 2O ---------> epoxide
2. Preformed peroxyacetic acid. One such preparation (Proxitane 4001) contains per
oxyacetic acid (36-40%), acetic acid (44-46%), hydrogen peroxide (3-4% ), water
(13-17%), and sulfuric acid (0.5-1.0% ). This is often used along with sodium
acetate as a buffer.
3. Distilled aqueous peroxyacetic acid (Proxilate) is made by collecting the peroxya
cetic acid-water azeotrope from the distillation of a dilute aqueous peroxyacetic
acid equilibrium mixture. It contains peroxyacetic acid (up to 35%) and water (60-
65%) with only a little acetic acid (less than 5%). Its advantages are that it contains
no mineral acid, has a low level of acetic acid, and has a high content of water. Its
main disadvantage is its relatively high reversion rate to hydrogen peroxide and
acetic acid. This makes it less suitable for transportation, and it is more desirable
to produce it on site. The epoxidation reaction will usually be biphasic.
4. Higher grades of peroxyacetic acid can be obtained from Proxilate by extraction
into organic esters such as ethyl acetate (Proxisolv I) or isopropyl acetate (Proxisolv
II). These contain 25-30% peroxy acid in ester solution. Epoxidation occurs in a
one-phase system.
V. EXAMPLES OF EPOXIDATION
A. Epoxidation on a Technical Scale
Soybean oil is the most extensively epoxidized oil, with annual world production of around
100,000 t. This is generally produced with peroxyformic or peroxyacetic acid generated in
situ from hydrogen peroxide and the appropriate alkanoic acid (Table 5). The final oxirane
level at 6.5-7.1% is lower than might have been expected (Table 6). It depends on the level
of unsaturation in the starting oil and the final product. The major side reaction is epoxide
ring-opening, which leads to a lower oxirane value. Sometimes the reaction is carried out in
the presence of a solvent such as hexane, heptane, or toluene.
Table 5 Typical Epoxidation Processes for Soybean Oil with (A)

Peroxyformic Acid or (B) Peroxyacetic Acid Prepared in situ
A B
Soybean oil (units) 1 1

Formic acid (units) 0.14 —
Acetic acid (units) — 0 .2

Hydrogen peroxide (70% w/w) (units) 0.3 0.3
Sulfuric acid (units) — 0 .0 1
Temperature (°C) 60 60
Reaction time (h) 1 0 -1 2 10-15
S o u rc e : Epoxidation (Interox, publishers).

Epoxidized Oils 765
Table 6 Oils Used for Epoxidation

Theoretical
Iodine epoxide
Oil value value ^
Soybean 130 8.2

Linseed 180 11.3
Safflower 145 9.1
Sunflower 125 7.8
Rapeseed 99 6.2
Palm 54 3.4
^Based on the iodine value; the values actually at

tained are slightly lower than these.
Other vegetable oils can be epoxidized in a similar way, but only soybean and linseed oil
are reacted on a commercial scale. Epoxidized palm oil is being exploited in Malaysia.
Epoxidized soybean oil, epoxidized linseed oil, and epoxidized octyl oleate are used mainly
as plasticizers and stabilizers in PVC. They add flexibility to the polymer and scavenge any
hydrogen chloride that is liberated by heat or light. Epoxidized palm oil with its lower oxirane
value is less effective in this respect, but it is a satisfactory plasticizer.
B. Epoxidation on a Laboratory Scale

The same procedures can be applied on a laboratory scale though it is then more convenient to
use a preformed peroxy acid such as m-chloroperoxybenzoic acid in dichloromethane solution.
Hammond [20] suggests that mchloroperoxybenzoic acid be used at the level of 1.3 mol equiv
(when adjusted for purity) for 1.25 h at 25-30°C. Some examples are listed in Table 7. The
reaction is reported to be quicker if conducted in a microwave oven [35].
VI. SPECTROSCOPIC PROPERTIES

Reference has already been made to the use of infrared, NMR, and mass spectrometry in the
recognition and identification of epoxy esters (Section III). Some additional information on
NMR shifts (^H and ^^C) is collected in Table 8. With more than one epoxide function the
Table 7 Epoxides Obtained from

Unsaturated Oils, Acids, and Methyl
Esters
Epoxides from Ref.
18:1 isomers 26
18:2 isomers 27,28
Linoleate 29,30
Linolenate 31
Eicosatetraenoate 32
Eicosapentaenoate 33
Lesquerella oil 34
Meadowfoam oil 34
766 Gunstone
Table 8 Chemical Shifts (^H and in cis Epoxy Esters
and shifts
Epoxy ester a b c d e
O
/\
- C H 2C H CH CH 2 - 1.43 2.70* —
a b b a
0 0
/\ /\
-CH 2CHCHCH2CHCHCH2 - 1.45 2.79 2.91 1.60
a b e d c h a
0
/\
- CH2CHCHCHCHCHCH2 - 1.45 2.73 2.1-2.3 5.41 2.03
a b b c d d e
0
/\
- CH2CH2CH2CHCHCHCH2CH2 - 56.67 27.48 26.25 28.85 (cis)
d c b a a b c d
58.84 31.89 25.80 28.98 (trans)
^2.45 in tr a n s isomer.
S o u rc e : Refs. 24, 2 6 -2 8 .
NMR spectra become rather complex, though, as already indicated, some useful informa
tion can be obtained [24].
VII. CHEMICAL REACTIONS

The epoxide function is very reactive, especially in acid solution, and undergoes a wide range
of reactions.
O OH
/ \ / \
-C H C H - -C H C H -
reactive very reactive
Addition of molecules designated HZ, usually in the presence of acid, will give a mixture
of two products if the epoxide is unsymmetrical:
O
/ \ HZ
-C H C H --------- -> -CH(OH)CH(Z) - and -C H (Z )C H (O H )-
The reaction with hydrogen chloride to give a chlorohydrin is the basis of the use of
epoxides as stabilizer for PVC, which can liberate hydrogen chloride under the action of light
or heat. Other examples, with the product type indicated in parentheses are water (diol),
alcohol (hydroxy ethers), carboxylic acid (hydroxy esters that are readily hydrolyzed to diols),
hydrogen sulfide (hydroxy thiols), sulfites (hydroxy sulfonates), thiourea or thiocyanates (thi-
oepoxides), hydrogen cyanide (hydroxy nitriles), alkyl carboxylates (lactones), ammonia, or
primary or secondary amines (hydroxyamines), carbon dioxide (carbonate), and others.
The acid-catalyzed alcoholysis of methyl vemolate with methanol, ethanol, propanol, and
isopropanol at reflux temperature takes 4, 48, 72, and 264 h, respectively. The product is
Epoxidized Oils 767
mainly the 12-hydroxy-13-alkoxy derivative rather than its isomer (80:20) [36]. These are
mainly derivatives of ricinoleic acid, and it would be interesting to study them in the typical
reactions of this acid. Reaction with diols gives polymeric material with free hydroxyl groups
that can be converted to polyurethanes [37].
Reaction between epoxidized palm oil and acrylic acid gives a vinyl derivative that can be
cured by radiation:
O
/\
- C H C H - +C H 2=C H C 00H — - CH(0H)CH(0C0CH=CH2) -
Rearrangement of epoxides to ketones occurs with sodium iodide in polyethylene glycol,

with 9 , 10 -epoxy stearate being converted to the 9- and 10-oxostearates [38].
Reaction with glycine (in the presence of aluminum trichloride and dimethylformamide)
gives a fatty morpholide that shows activity against bacterial and fungal groups [39].
R R'
O
H 2NCH 2CO 2H > -<
/\
RCHCHR' -> N O (and isomer)
^ -< 0
Other reactions of fatty epoxides are related to the acid/ester function and (when present)
to the double bond. For example, epoxy oils can be converted to amides with C4- C 6 amines
at room temperature (reaction is 80% complete after several days), at reflux temperatures ( 1 -
5 h), or in boiling hexane solution (2-5 days) [40]. Compounds of this type have been used
as antirust derivatives in water-based cutting fluids [41].
Several dibasic acids have been prepared from vemonia oil. Oxidation with nitric acid (18
h, 90-100°C) gives a mixture of C ^-C ^ dibasic acids from which 95% pure suberic (Cg) and
azelaic (C 9) acids have been isolated in yields of 15 and 11%, respectively [42]. After saponi
fication and hydrogenation, oxidation of the 12,13-epoxystearic acid with chromic acid gives
dodecanedioic acid (C 12) in 4 -7 days at room temperature [43].
The oil reacts with dibasic acids such as suberic to give a polyester that produces a tough
ened elastomer with styrene and with divinylbenzene [44].
O CH(OH)R
RCH(OH)^
/\
RCHCHR' + H02C(CH2)6C02H CH0C0(CH2)6C00^
R’ (and other isom ers) R’
In attempts to produce high value products from new oilseed crops, Derkson and Cuperus
[45] converted vemolic acid from Euphorbia lagascae to compounds such as bombykol, bom-
bykal, traumatic acid, and some lactones.
Bombykol (Cj6) CH3(CH2)2CH=CHCH=CH(CH2)8CH20H

Traumatic acid (C12) H 0 0 C C H = C H (C H 2 )8 C 0 0 H
REFERENCES
1. D. Swem, Organic peroxy acids as oxidizing agents. Epoxidation, in Organic Peroxides (D.
Swem, Ed.), Wiley Interscience, New York, 1971, p. 355.
768 Gynstone
2. F. D. Gunstone, The formation and reactions of long-chain epoxy acids, in Fatty Acids (E. H.
Pryde, Ed.), AOGS, Champaign, IL, 1979, p. 379.
3. F. D. Gunstone, The nature of the oxygenated acid present in Vernonia anthelmintica seed oil, J.
Chem. Soc. (Lond.): 1611 (1954).
4. F. D. Gunstone, J. T. Harwood, and F. B. Padley (Eds.), Lipid Handbook, 2nd, Chapman and
Hall, London, 1994.
5. E. Blée, Oxygenated fatty acids and plant defenses, INFORM 6: 852 (1995).
6. Editor, Natural plant material. Oils Fats Int. (2): 9 (1990).
7. F. O. Ayorinde, K. D. Carlson, R. P. Paulik, and J. McVety, Pilot plant extraction of oil from
Vernonia galamensis seed, J. Am. Oil. Chem. Soc. 67: 512 (1990).
8. F. O. Ayorinde, B. D. Butler, and M. T. Clayton, Vernonia galamensis: a rich source of epoxy
acid, J. Am. Oil. Chem. Soc. 67: 844 (1990).
9. K. D. Carlson and S. P. Chang, Chemical epoxidation of a natural unsaturated epoxy seed oil from
Vernonia galamensis and a look at epoxy oil markets, J. Am. Oil. Chem. Soc. 62: 934 (1985).
10. W. E. Neff, R. O. Adlof, H. Konishi, and D. Weisleder, HPLC of the triacylglycerols of Ver
nonia galamensis and Crépis alpina seed oils, J. Am. Oil. Chem. Soc. 70: 449 (1993).
11. M. A. Anderson, L. Collier, R. Dilliplane, and F. D. Ayorinde, Mass spectrometric characterisa
tion of Vernonia galamensis oil, J. Am. Oil Chem. Soc. 70: 905 (1993).
12. E.-S. Yalley, F. D. Ayorinde, and B. E. Eribo, Biodegradation of Vernonia galamensis seed oil
by Acinetobacter and Pseudomonas sp., J. Am. Oil. Chem. Soc. 69: 1046 (1992).
13. L. J. M. van Soest and F. Mulder, Potential new oilseed crops for industrial use. Lipid Technol.
5: 60 (1993).
14. B. G. Muuse, F. D. Cuperus, and J. T. P. Derksen, Ind. Crops Products 1: 57 (1992).
15. J. T. P. Derksen, F. P. Cuperus, and P. Kolster, Paints and coatings from renewable resources,
Ind. Crops and Products 3: 225 (1995).
16. American Oil Chemists’ Society, Oxirane oxygen, AOCS method Cd 9-57 (89), AOCS, Cham
paign, IL.
17. L. J. Morris, Separation of long-chain polyhydroxy acids by thin layer chromatography, J. Chro-
matog. 12: 321 (1963).
18. M. Hamberg, A method for determination of the absolute stereochemistry of ajS-epoxy alcohols
derived from fatty acid hydroperoxides. Lipids 27: 1042 (1992).
19. P. Fahlstadius and M. Hamberg, A gas-liquid chromatographic method for steric analysis of epoxy
acids, Chem. Phys. Lipids 51: 15 (1989).
20. E. W. Hammond, Chromatography for the Analysis of Lipids, CRC Press, Boca Raton, FL, 1993,
pp. 147-152.
21. H. C. Gérard, R. A. Moreau, W. F. Fett, and S. F. Osman, Separation and quantitation of
hydroxy and epoxy fatty acids by HPLC with an evaporative light-scattering detector, J. Am. Oil
Chem. Soc. 69: 301 (1992).
22. F. D. Gunstone, Fatty Acid and Lipid Chemistry, Blackie, London, 1996, p. 150.
23. F. Marx and E. Classen, Analysis of epoxy fatty acids by GC-MS of their 4 ,4-dimethyloxazoline
derivatives. Fat Sei. Technol. 96: 207 (1994).
24. F. D. Gunstone, The study of natural epoxy oils and epoxidised vegetable oils by ^^C NMR
spectroscopy, J. Am. Oil Chem. Soc. 70: 1139 (1993).
25. F. A. Zaher, M. H. El-Mallah, and M. M. El-Hefnawy, Kinetics of oxirane cleavage in epoxidised
soybean oil, J. Am. Oil Chem. Soc. 66: 698 (1989).
26. F. D. Gunstone and F. R. Jacobsberg, The preparation and properties of the complete series of
methyl epoxyoctadecanoates, Chem. Phys. Lipids 9: 26 (1972).
27. F. D. Gunstone and H. R. Schuler, Epoxyoctadecenoates, dihydroxyoctadecenoates, and diepoxy-
octadecanoates, Chem. Phys. Lipids 15: 174 (1975).
28. F. D. Gunstone and H. R. Schuler, PMR spectra of several epoxyoctadecenoic, epoxyoctadecy-
noic, and diepoxyoctadecanoic esters, Chem. Phys. Lipids 15: 189 (1975).
29. G. Maerker, E. T. Haeberer, and S. F. Herb, Epoxidation of methyl linoleate. II. The two isomers
of methyl 9,10; 12,13-diepoxy stearate, J. Am. Oil Chem. Soc. 43: 505 (1966).
Epoxidized Oils 769
30. A. Debal, G. Rafaralahitsimba, A. Bonfand, and E. Ucciano, Catalytic epoxidation of methyl

linoleate— cyclisation products of the epoxy acid esters, Fat Sci. TechnoL 97\ 269 (1995).
31. M. Alaiz, M. P. Maza, R. Zamora, F. J. Hidalgo, F. Millan, and E. Vioque, Epoxidation of
ethyl (Z)-9-(Z)-12-(Z-15)-octadecatrienoate with m-chloroperbenzoic acid, Chem. Phys. Lipids 49:
221 (1989).
32. E. J. Corey, H. Niwa, and J. R. Falck, Selective epoxidation of eicosa-5,8,11,14-tetraenoic (ara-
chidonic) acid and eicosa-8,11,14-trienoic acid, J. Am. Chem. Soc. 101: 1586 (1979).
33. M. Van Rollins, Synthesis and characterisation of cytochome P-450 epoxygenase metabolites of
eicosapentaenoic acid. Lipids 25: 481 (1990).
34. K. D. Carlson, R. Kleiman, and M. O. Bagby, Epoxidation of Lesquerella and Linanathes (mead-
owfoam) oils, J. Am. Oil Chem. Soc. 71: 175 (1994).
35. M. S. Lie Ken lie and C. Yan-Kit, The use of a microwave oven in the chemical transformation
of long-chain fatty acid esters. Lipids 23: 367 (1988).
36. K. A. A. Bryant, C. P. Nwaonicha, M. A. Anderson, and F. O. Ayorinde, Acid-catalyzed alco
holysis of Vernonia galamensis oil, J. Am. Oil Chem. Soc. 69: 1023 (1992).
37. B. Dahlke, S. Helbart, M. Paetow, and W. H. Zech, Polyhydroxy fatty acids and their derivatives
from plant oils, J. Am. Oil Chem. Soc. 72: 349 (1995).
38. G. Stoll and K. Worschech, Keto fatty acids— a simple procedure for their preparation. Fat Sci.
TechnoL 94: 332 (1992).
39. R. Agarwal, M. H. Ansari, M. W. Y. Khan, M. Ahmad, and K. D. Sharma, Synthesis and
antimicrobial activity of fatty 2-morpholinones prepared from epoxy fatty acid methyl esters, J.
Am. Oil. Chem. Soc. 6: 825 (1989).
40. K. A. A. Bryant, C. P. Nwaoicha, M. Hassan, M. A. Anderson, and F. O. Ayorinde, Synthesis
and isolation of epoxy secondary amides via direct amidation of Vernonia galamensis seed oil, J.
Am. Oil Chem. Soc. 70: 457 (1993).
41. S. Watanabe, T. Fujita, and M. Sakamoto, Epoxides of unsaturated fatty acids as anti-rust addi
tives in water-based cutting fluids, J. Am. Oil Chem. Soc. 65: 1311 (1988).
42. F. O. Ayorinde, G. Osman, R. L. Sheppard, and F. T. Powers, Synthesis of azelaic acid and
suberic acid from Vernonia galamensis seed oil, J. Am. Oil Chem. Soc. 65: 1774 (1985).
43. F. O. Ayorinde, F. T. Powers, L. D. Streete, R. L. Shepard, and D. N. Tabi, Synthesis of
dodecanedioic acid from Vernonia galamensis oil, J. Am. Oil Chem. Soc. 66: 690 (1989).
44. O. A. Ofolabi, M. E. Aluko, G. C. Wang, W. A. Anderson, and F. O. Ayorinde, Synthesis of
toughened elastomer from Vernonia galamensis seed oil, J. Am. Oil Chem. Soc. 66: 983 (1989).
45. J. T. P. Derkson and F. P. Cuperus, High value-added applications from unusual fatty acids,
Lecture presented to ISF Congress, The Hague, October 1995.
30_______________________________________________
Bio fu e ls
Y. M. Choo and A. N. Ma
Palm Oil Research Institute of Malaysia (PORIM), Kajang, Selangor Darul
Ehsan, Malaysia
A. S. H. Ong
Malaysian Palm Oil Promotion Council, Kuala Lumpur, Malaysia
I INTRODUCTION
The energy crisis in the mid-1970s coupled with the fast diminishing energy reserves, greater
environmental awareness, and increasing energy consumption have led to an intensified global
search for viable alternative sources of energy. In recent years a great deal of attention has
been directed to plant-based sources for fuels, particularly fuels for diesel engines. Some
plants produce oil directly. Vegetable oils that are renewable have the greatest potential as
alternative fuel sources. The choice of vegetable oil depends on the cost of production and
reliability of supply. At present, annual worldwide vegetable oil production is around 77
million tonnes, with 14.3 million tonnes of soybean oil, 10 million tonnes of palm oil, and
7.7 million tonnes of rapeseed and sunflower oil [1]. Of all the vegetable oils available, palm
oil, being the highest oil yielding crop would be the preferred choice (Table 1). Since factors
other than yield must be considered, biofuels are made from palm (Malaysia), coconut (Philip
pines), rapeseed (Europe), and soy and tallow (United States), as well as from used frying oil
(Japan and other countries), i.e., in general from oils in greatest supply.
Many researchers have investigated the possibility of using vegetable oils straight or
blended as diesel substitutes. In fact, the use of vegetable oil as fuel for compression-ignition
engines is not new. The inventor of the diesel engine, Rudolf Diesel, displayed an engine
fueled with peanut oil at the Paris Exposition of 1900 [3]. The use of palm oil as a diesel fuel
was reported as early as 1920 [4]. Since then many studies have been conducted in various
parts of the world, particularly in the period 1920-1940 [3,5-8]. The 1973 oil embargo and
the ensuing energy crisis reactivated the search for alternative sources of fuel. A good account
of these attempts is given by Goering et al. [9] and in the 1983 AOCS symposium on vegeta
ble oils as diesel fuels [10-13]. The AOCS symposium revealed that vegetable oils have good
potential as alternative fuels if the following problems could be overcome satisfactorily: high
viscosity, low volatility, and the reactivity (polymerization) of the unsaturated hydrocarbon
chains when the oils are highly unsaturated. These give rise to cooking on the injectors.
771
772 Choo et ai.
Table 1 Average Yield of

Oil-Producing Crops
Oil Yield [kg/(ha-yr)]

Palm 3636
Coconut 1272
Peanut 909
Soybean 432
Rapeseed 999
Sunflower 801
carbon deposits, oil ring sticking, and thickening and gelling of the lubricating oil as a result
of contamination with vegetable oils.
It is possible to reduce the viscosity of the vegetable oils by incorporating a heating device
in the diesel engine as has been successfully demonstrated by the Elsbett engine manufacturer
[14]. Other factors that may have long-term effects on the engine are free fatty acids and
gummy substances found in the crude vegetable oils. The incomplete combustion residues
may contribute to undesirable deposits on the engine components. The gummy substances
cause filter plugging. This calls for more regular and frequent servicing and maintenance of
the engine. The long-term effects of the oil on the engine have to be monitored.
“Biofuel” is the term usually used to describe biodiesel. According to Reed [15], the term
“biodiesel” refers to methyl or ethyl esters derived from vegetable oils. It can also refer to
pyrolysis products, diesel-vegetable oil blends, microemulsions of alcohols and water in veg
etable oils, and fermentation butanol. This chapter, however, deals exclusively with alkyl
esters derived from vegetable oils. Esters are used because they have lower viscosity and
better fuel properties than vegetable oils. Vegetable oil esters are used on a commercial scale
in many cities of the world where there are air pollution problems. Engines run on vegetable
oil esters are known to emit less dark smoke and CO and no SO2. Though vegetable oil esters
cannot meet total diesel demand in the transport sector, they can help reduce air pollution if
they are used in heavy vehicles such as buses and trucks.
IL PRODUCTION TECHNOLOGY OF ALKYL ESTERS FROM

VEGETABLE OILS AND FATS
Commercially, methyl esters of fatty acids can be manufactured either by esterification of
fatty acids or by transesterification of fat triglycerides. The predominant process for the manu
facture of methyl esters is the transesterification of fats and oils with methanol. The ester
interchange, i.e., the replacement of the alcohol component glycerol by methanol, takes place
quite easily at low temperatures, 50-70°C, and under atmospheric pressure with an excess of
methanol and in the presence of an alkaline catalyst such as sodium hydroxide [16-19].
These mild reaction conditions, however, require an oil neutralized by means of alkaline
refining, steam distillation, or preesterification of free fatty acids.
Esterification may be carried out batchwise at 200-250°C under pressure. In order to obtain
a high yield, the water produced during the reaction has to be removed continuously. Esterifi
cation can also be carried out continuously in a countercurrent reaction column using super
heated methanol [17]. Esterification is the preferred method for ester preparation from specific
fatty acid cuts such as palm stearin and olein.
Biofuels 773
Vegetable oil + Alcohol ------Esters -f Glycerol

(Triglyceride) 0.1 tonne 1 tonne 0.1 tonne
1 tonne
CHÔCOR' R'C00CH3 CHÔH
R?OCOCH +3ROH — -> R^COOCH3 4- CHOH

I 1
CHjOCOR' R^œOCHs CHÔH
Some of the available ester production technology for biodiesel application is shown in
Table 2. It is also possible to convert vegetable oils such as crude palm oil to esters with
varying amounts of free fatty acids in a continuous process by combining the esterification
and transesterification processes as developed by PORIM. This has been successfully demon
strated in a 3000 tpa pilot plant [21].
The novel aspect of this patented process is the use of solid acid catalysts for the esterifica
tion [21-23]. The reaction is carried out below 100°C at atmospheric pressure in a column of
solid catalysts. The resultant reaction mixture, which is neutral, is then transesterified in the
presence of an alkali catalyst. The conventional washing stage or neutralization step after the
esterification process is obviated, and this is an economic advantage.
III. EVALUATION OF VEGETABLE OIL ALKYL ESTERS

AS BIOFUEL
The use of vegetable oils in the form of esters as diesel substitute has been quite extensively
reviewed [1,24,25]. Vegetable oils such as soybean, rapeseed, sunflower, canola, coconut.
Table 2 Available Biodiesel Production Technologies
Reaction conditions
P T Mode of
Company (atm) CC) Catalyst operation
Transesterification
Comprimo/Vogel and Nott 1 Ambient KOH Batch
University of Idaho 1 Ambient KOH Batch
Novamont/Technimont 1 > Ambient Organic Batch
Conneman/Feld and Hahn 1 60-70 NaOH Continuous
Lurgi 1 60-70 Alkaline Continuous
IFP/sofiproteal 1 50 -130 Alkaline/acidic Batch
Gratech 3.5 95 ? Continuous
Desmet 50 200 Nonalkaline Continuous
PORIM 1 < 10 0 Acidic/alkaline Continuous
Others: Oleofina,
Procter & Gamble, Me KFT 360-380
Hydrogenation 40-100 Cobalt/molybdenum Continuous
Cammit/Arbokem
S o u rc e : Refs. 1 and 2 .
774 Choo et al.
Table 3 Fatty Acid Composition of Vegetable Oils and Tallow
Cotton Lin
Fatty Coconut ' Babassu Castor seed Karanja seed Maize Palm Palmkemel
acid oil oil oil oil oil oil oil oil oil
6 :0 0 .0 - 0 . 6 nd — nd — — nd ns 0 .0 - 0 . 8
8 :0 4.6 -9 .4 2.6-7.3 — nd — — nd ns 2.4 -6 .2
1 0 :0 5 .5 -7 .8 1.2-7.6 — nd — — nd ns 2 .6 -5 .0
1 2 :0 45.1-50.3 40.0-55.0 — 0 .0 - 0 . 2 — — 0.0-0.3 0 .0 -0 .4 41.0-55.0
14:0 16.8-20.6 11.0-27.0 — 0 . 6 - 1 .0 — — 0.0-0.3 0 .5 -2 .0 14.0-18.0
16:0 7.7-10.2 5.2-11.0 1 21.4-26.4 10 8 8.6-16.5 40.1-47.5 6.5-10.0
16:1 nd nd — 0 .0 - 1 .2 — — 0 .0 -0 .4 0 .0 - 0 . 6 ns
17:0 nd nd — nd — — nd ns ns
17:1 nd nd — nd — — nd ns ns
18:0 2.3-3.5 1.8-7.4 la 2.1-3.3 7 5 1.0-3.3 3.5 -6 .0 1.3-3.0
18:1 5.4-8.1 9.0-20.0 3 14.7-21.7 49 21 20.0-42.2 36.0-44.0 12.0-19.0
18:2 1 .0 - 2 .1 1 .4-6.6 4 46.7-58.2 19 64 39.4-62.5 6.5-12.0 1.0-3.5
18:3 0 .0 - 0 . 2 nd tr 0.0-0.4 — 2 0.5-1.5 0.0-0.5 —
2 0 :0 0 .0 - 0 . 2 nd — 0.2-0.5 4 — 0 .3 -0 .6 O .O -l.O —
2 0 :1 0 .0 - 0 . 2 nd — 0 .0 - 0 .1 — — 0 .2 -0 .4 ns }
2 0 :2 nd nd — 0 .0 - 0 .1 — — 0 .0 - 0 .1 ns }
2 2 :0 nd nd — 0 .0 - 0 . 6 5 — 0.0-0.5 ns }
2 2 :1 nd nd — 0.0-0.3 } — 0 .0 - 0 .1 ns } 0.0-1.0%
2 2 :2 nd nd — 0 .0 - 0 .1 }6 — nd ns }
24:0 nd nd — 0 .0 - 0 .1 } — 0 .0 -0 .4 ns }
24:1 nd nd nd } nd ns
90»
com, safflower, cottonseed, peanut [26], palm [2], buffalo gourd, and Chinese vegetable
tallow [27] have been studied and reported to be potential feedstocks for biodiesel production.
Animal fats such as tallow [28] and used cooking oils of all types [29] have also been consid
ered for use as fuels.
Table 3 shows the fatty acid composition of the vegetable oils used for biodiesel produc
tions [30]. Table 4 shows the fuel characteristics of the methyl esters of these oils [25], and
Table 5 summarizes engine testing of alkyl esters from various vegetable oils and tallow
[31-50].
A. Fuel Characteristics
The fuel properties of vegetable oil esters (Table 4) clearly show that they can be attractive
as diesel substitutes. The specific gravities of all the vegetable oil esters are slightly higher
than that of diesel. Though their viscosities are somewhat higher, they flow under warm
weather conditions. The fluidity of the esters depends on the fatty acid composition (FAC) or
the iodine values (IV) (unsaturation) of the vegetable oils. Obviously, oils with high IV will
produce esters with better fluidity and will not solidify at low temperature. For example, palm
oil has an IV of 52, and its esters, which have a pour point of 15°C, will solidify at that
temperature. Rapeseed oil, which has an IV of >90, will solidify at a much lower tempera
tures, about 4°C. However, the high pour point can be lowered by adding pour point de
pressing agents or by blending with solvents [25].
Biofuels 775
Rapeseed
Palm Rapeseed (low emcic Ricebran Soybean Sunflower Premier and
stearin Peanut oil oil acid) oil oil oil edible tallow
nd nd nd nd
nd 0.1 nd nd nd
nd nd nd nd < 2 .5
0 .1-0 .4 0- 0.1 nd 0 . 0 - 0.1 0 .0- 0.1
1. 1- 1.8 0- 0.1 0.2 0 .0- 0.2 0.4 0 .0 - 0.2 0 . 0- 0.2
48.4-73.8 8 .3-14 .0 1.5-6 .0 3 .3 - 6.0 16.4 8 .0 -13.3 5 .6 - 7.6 1.4-78
0.05-0.2 0- 0.2 0.0-3.0 0 . 1- 0.6 0.1 0 .0- 0.2 0.0-0.3 17 -3 7
nd nd nd 0.0-0.3 nd nd 0.7-8.8
nd nd nd 0.0-0.3 nd nd 0.5-2.0
3.9-5.6 1.9 -4.4 0 .5 -3 .1 1.1 - 2 .5 1.8 2 .4 - 5.4 2 .7 - 6.5 < 1.0
15.6-36.0 36.4-67.1 8-60 52 .0 - 66.9 42.3 17 .7 - 26.114.0-39.4 6.0-40
3.2-9.8 14.0-43.0 11- 2 3 1 6 . 1 - 24.8 37.1 49.8- 57.148.3-74.0 26-50
0 . 1- 0.6 0- 0.1 5 - 13 6 .4 - 14 .1 1.3 5 .5 - 9.5 0 . 0- 0.2 0.5-5.0
0.3-0.6 1.1- 1.7 0.0-3.0 0 . 2- 0.8 0.4 0 . 1- 0.6 0.2-0.4 < 2 .5
nd 0 .7 -1.7 3 - 15 0 .1-3 .4 0.2 0.0-0.3 0 . 0 - 0.2 < 0 .5
nd O .O -l.O 0 . 0- 0.1 0 .0 - 0.1 nd < 0 .5
nd 2 .1- 4.4 0 . 0- 2.0 0.0-0.5 0.3-0.7 0 .5 -1.3
nd 0-0.3 5-60 0 . 0- 2.0 0.0-0.3 0 .0- 0.2
nd 0 . 0- 2.0 0 . 0- 0.1 nd 0.0-0.3
nd 1. 1- 2.2 0 . 0- 2.0 0 . 0- 0.2 0.0-0.4 0.2-0.3
nd 0-0.3 0.0-3.0 0.0-0.4 nd nd
nd = not detected; ns = not specified.

^Dihydroxy stearic acid.
^Ricinoleic acid 18:1 (OH).
All vegetable oil esters have higher cetane numbers and flash points than petroleum diesel.
The higher cetane numbers indicate shorter ignition delay characteristics, and the higher flash
points represent greater safety in storage and transportation.
B. Engine Testing
Engine testing of alternative fuels usually comprises two phases: (1) bench endurance tests
(single-cylinder or multicylinder engine tests) and (2) vehicle field tests.
1. B ench E ndurance Tests

Many researchers have carried out bench endurance tests with biodiesel from various vegeta
ble oils as feedstocks [31-54]. Table 5 summarizes the results reported by them. As the tests
were carried out under different conditions with different engine makes, it is expected that the
results will vary accordingly.
However, in general, biodiesel shows improved exhaust emission compared to diesel in
terms of smoke opacity, unbumed hydrocarbons (HC), and carbon monoxide (CO). Though
sulfur dioxide (SO2) was not monitored except for palm oil methyl esters, it is expected to be
much lower as all the vegetable oils (with the possible exception of poorly refined rapeseed
Table 4 Fuel Characteristic o f Methyl Esters of Various Vegetable Oils
Vegetable oil
Rapeseed Typical
Cotton Rapeseed (non- Sun diesel
Test conducted Coconut Palmkemel Palm Peanut seed (erucic) erucic) flower Soybean Linseed fuel
Iodine number 10.0 15.0 52.0 94.0 105.0 98.0 114 .0 129.0 13 1.0 183.0 _
Density at 15°C (g/L) 872.0 872.0-877.0 882.0 885.0 891.0 830.0-840.0
Lower heating value
MJ/kg 35.3 5.5 37.0 37.2 37.0 37.7 37.2 37.1 37.1 37.0 42.7
MJ/L 30.8 32.4 32.8 32.8 33.0 35.5
Viscosity at 40°C 2.7 4.3-4.5 4.3-4.5 4.2 4.0 3.7 2.0-3.5
(mmVs)
Cetane number 62.7 64.3-70.0 51.0 -59 .7 61.2 56.0 52.5 51.0
Biofuels 777
oil) contain negligible amounts of sulfur. There are conflicting results on the nitrogen oxides
(NOJ in the exhaust emissions. Most reported an increase in NO^ with biodiesel fuel.
All the tests experienced a drop in engine power output and increased fuel consumption
with biodiesel fuel compared to diesel fuel. This is expected to be marginal as the energy
content of biodiesel is about 10% less than that of diesel. Most of the differences reported,
especially in the combustion process, were mainly attributed to differences in the ignition
delay and load conditions between the two fuels.
Some material compatibility problems with biodiesel have been observed and reported
[2,35,38,55,56]. Paints, various plastic and rubber construction materials, and various adhe
sives experience degradation through hardening and/or deterioration. These observations sug
gest that fuel lines and other components should be replaced with a more compatible material.
2. Vehicle F ield Tests

There is limited published information available on the exhaustive field testing of biodiesel
[1,2,25,55-58]. In France, 5-100% blends of biodiesel (from rapeseed oil) and diesel are
used experimentally in buses and trucks in about 30 cities [1]. In Italy, it was reported that
19 cities used Iveco engines in 72 buses running on biodiesel starting in the latter part of
1991 [1] in a field test involving 520,000 km of operation. Results showed that the power
output and fuel costs were comparable to those of diesel fuel. There was marked reduction in
smoke emissions, unbumed hydrocarbons, and particulate matter and little dilution of lubricat
ing. In Germany, following the favorable results obtained with both single-cylinder and multi
cylinder engine bench tests, Mercedes-Benz AG conducted exhaustive field tests on the rape-
seed oil methyl esters (RME) in trucks, buses, and taxis from January 1991 to October 1992
[25]. Very encouraging results were obtained: Fuel consumption and lubricating oil consump
tion were comparable to those obtained during diesel fuel operation and the drivers were
satisfied with the vehicles’ behavior; the slightly lower rated output due to the lower caloric
value of the RME only rarely led to complaints. However, the unpleasant odor of the exhaust
gas (smell like French fries) was noted! This is very subjective. Also there were some prob
lems with the auxiliary heaters installed in the vehicle, which must be operated with RME.
Similarly, in the United States, blends of biodiesel from soybean oil and diesel in a 20 : 80
or 30 : 70 ratio were used to run 1200 buses, trailer trucks, and vans in 400 cities in 1992-
1993 in operational tests over a total of 4.8 million km [1]. Observations similar to those in
tests conducted in Europe were obtained. It was also found that biodiesel involves no technical
problems and has many energy and environmental advantages, particularly in urban areas.
Field trial results have so far confirmed that biodiesel and diesel behave in a similar manner
and that the use of biodiesel requires no specific engine modification.
Perhaps the most systematic and thorough evaluation of vegetable oil methyl esters as
diesel substitutes was conducted in Malaysia. Over the past ten years, palm oil methyl esters
have been extensively evaluated as diesel substitutes by evaluating their fuel characteristics,
followed by bench endurance tests and finally exhaustive vehicle field trials involving a wide
range of diesel engines such as stationary engines, passenger cars, lorries, trucks, taxis, trac
tors, generators, land cruisers, portable generators, water pumps, and vans running in Malay
sia [2]. There were two fleets of a total of 30 buses mounted with Mercedes engines operating
in Kuala Lumpur, the capital city of Malaysia using crude palm oil methyl esters as fuel from
1992 to 1995. Direct injection engines were used in these fleet tests. Each bus traveled
300,000 km, the exhaustive field trial distance recommended by the engine manufacturer. Ten
buses were run on 100% palm oil methyl esters, 10 on 50 : 50 blends of palm oil methyl
esters and petroleum diesel, and 10 on 100% petroleum diesel as control. Another fleet of six
-
.
r:
0
.o - d
+
I +-;
v;
a,
I
II
3- g I
$5
a s
a
s
a
A
C a ,z . g m m m
.0" .s 2
-
". m
E
barn
m
.s
0
.c "m
,
s
+d
V: V:
.
"i.52
8 -
2s
a
g z e E 2 2
G . 2 G
-3m -- g ~ . a Y~ g2
O'f??Ss
0 rr. r r . 0 0 0
3 f f
YS 3 n
+a
s . 5 &
~a 5
a &
g ag 8e ' 5a
. v ; a,
9
4 A
.$$-go b ap s Q 9-
0 >
a,
iE I e?s 0
I S 0 8178 C1 (Tests include 4C, DI, TC NOx increased as concn.
blending with petroleum 6C, DI, TC Power output, smoke, C 0 ,
diesel) 6C, DI, TC and decreased as concn. of
Ie
4C, DI, NA
4C, DI, TC
On-road vehicle test. 4C, ID1 Fuel consumption and volumetric speci
Durability tcst, 1418 h. sumption increased.
Power test and emission test Volumetric calorific value,
(Tests include blending with and CO decreased.
petroleum oil). Power and torque differ
No lube oil dilution.
Bench test 100 h and mixed DI Lube oil dilution increased with increas
driving test (Incorporation 4C, ID1 double bonds of ME in
of methyl esters of babassu No lube oil dilution with
oil). Lube oil normal at 7500
Lube oil dilution tcsts.
2. Rapeseed oil methyl es- Various load conditions; 20- lC, DI Smoke increased.
ters (RME) 100% rate output. Thermal efficiency and ignition delay
creased.
EMA cycle 200 (h), advance 3C, DI, NA Smoke, exhaust gas, and
injection timing 2 deg. creased.
Advance injection timing 2 Smoke, peak pressure, and
deg . creased.
Power output increased.
Short-term and engine durabil- 4C, DI, TC Fuel consumption and injection coking
ity tests 300 (h) (Tests in- lC, PC increased.
cludes ethyl esters of rape- Output power (EE lowest),
seed oil). est), smoke, and exhaust
On-road vehicle test 364 (h) 6C, DI, TC, IC Power output, fuel consumption,
(Test includes blending with decreased.
petroleum diesel).
On-road vehicle test 364 (h) 8C, PC, NA Power output and smoke decreased.
(Test includes blending with
petroleum diesel).
Perkins 4.236 engines on a cy- Brake thermal efficiency
clic load: endurance test. Fuel consumption slight increased.
Smoke decreased.
Power output: comparable.
6. Peanut oil methyl esters Endurance test 100 h Lube oil viscosity decreased.
(PEME)
Perkins 4.236 engines on a cy- Brake thennal efficiency increased.
clic load; endurance test Fuel consumption slight increased.
Smoke decreased.
7. Ricebran oil methyl esters Various load conditions; 20- 1C, 191 Smoke increased.
(RBME) 100% rated output Ignition delay decreased.
8. Karanja oil methyl esters Various load conditions; 20- le,DI Smoke increased.
(KAME) 100% rated output Thermal efficiency and ignition delay d
creased.
9. Tallow methyl esters SAE J 1003 6C,DI, TC, AC NOx and fuel consumption increased.
(TAME) Hydrocarbons, CO, and particulates de
creased.
10. Maize oil methyl esters Perkins 4.236 engines on a cy- Brake thermal efficiency increased.
(MME) and ethyl esters clic load. Fuel consumption slight
WEE) Endurance test. Smoke decreased.
11. Castor oil methyl esters Perkins 4.236 engines on a cy- Fuel consumption, brake thermal
(CME) and castor oil clic load. and ignition delay increased.
ethyl ester (CEE) Endurance test. Smoke value decreased but
speed, low load.
AC, air-cooled; BMEB, brake mean effective pressure; DI, direct ~njection;EGR, exhaust gas recirculation; IC, intercooled; ID1, indirect injection;
der; WA, naturally aspirated; PC, precombustion chamber; TC, turbocharged; nC, n cylinders.
782 Choo et at.
buses, three running on 100% palm oil methyl esters and three on 100% petroleum diesel, were
tested at about the same time. A different lubricating oil was tested with these six buses. Each
bus traveled between 200,000 and 250,000 km. The field trial has been completed, and detailed
analyses of the results are being carried out. However, throughout the field trial period, no tech
nical problem was reported as long as the engines were maintained according to engine service
manuals. The operators were not able to detect any difference in terms of general engine perfor
mance. The conclusions derived from the above evaluation are as follows:
1. No modification of the diesel engines is required.
2. The performance of the engines is generally good. The engines are easy to start, with
no knocking and with smooth running. There is a shorter ignition lag.
3. The exhaust gas emission of the engines is much cleaner with reduction of hydrocar
bon, NO^, CO, and SO2 content; therefore palm oil methyl esters are more environ
mentally friendly. These results are based on steady-state emission.
4. The engine oil is still usable at the recommended mileage as no engine oil dilution
was observed either by infrared spectroscopy or by gas chromatography.
Palm oil methyl esters do not produce explosive air-fuel vapor. They also offer en
hanced safety characteristics with a higher flash point (174°C compared to 96°C of
petroleum diesel).
The cetane number of palm oil methyl esters is 64, which is higher than that of petro
leum diesel, 30-40. Thus palm oil methyl esters can act as a cetane number improver.
7. Carbon build-up on the injector nozzles was normal except that the nature of the
carbon was different.
8 . The fuel consumption of the engines on palm oil methyl esters is comparable to that
with petroleum diesel (e.g., 3 -4 km/L for the buses under trial).
9. Palm oil methyl esters attack low grade plastic and rubber products such as hoses and
seals and also react with the binding material in cement floors.
IV, GENERAL SUMMARY OF VEGETABLE OIL ALKYL

ESTERS AS BIOFUEL
The highlights of the findings are as follows.
1. Methyl esters of vegetable oils are better diesel substitutes than the original oils.
2. Simple alkyl esters, in contrast to triglycerides, improve viscosity, but volatility re
mains poor. This may cause a problem in direct injection engines.
3. Introduction of a double bond resulted in increased efficiency; thus ethyl oleate had
the highest thermal efficiency.
4. Among the methyl esters of saturated acids, thermal efficiency is inversely related to
the chain length of the fatty acid.
5. A study of the efficiency of various esters of fatty acid as diesel fuel suggests that a
vegetable oil having a high content of oleic acid or short-chain saturated acid transes-
terified to produce ethyl esters should produce a good test material as an alternative
diesel fuel. Hence, the above observations indicate that alkyl esters of palm oil have
the potential of being good diesel fuel substitutes.
V, CONCLUSION
Various technologies have been developed for the production of alkyl esters from vegetable
oils containing varying amounts of free fatty acids. Many studies have demonstrated that
Biofuels 783
vegetable oil esters (or biodiesel) can be used as fuel to run standard diesel engines without
modification. The diesel engines are able to run on pure biodiesel or a blend with petroleum diesel.
Biodiesels have higher cetane numbers than diesel, and they can be used as diesel improvers.
All the studies from bench engine tests to exhaustive field trials have demonstrated that there
was no discernible difference in terms of engine performance and fuel consumption in field
trials. However, the exhaust emission was found to be cleaner. It emits no dark smoke and
less CO and SO2. Thus it is a more environmentally benign fuel. Nevertheless, the high
cost of biodiesel compared to petroleum diesel at the present time is a major obstacle in the
commercialization of biodiesel. It is believed that biodiesel has a strategic role to play in the
near future.
ACKNOWLEDGMENT
We thank the Director General of PORIM for his permission to publish this chapter.
REFERENCES
1. T. Murayama. Evaluating vegetable oils as a diesel fuel, I N F O R M 5: 1138 (1994).
2. Y. M. Choo, A. N. Ma, and Y. Basiron, Palm oil methyl esters as diesel substitute. Presented at
An Alternative Energy Conference—Liquid Fuels, Lubricants and Addition from Biomass, June
15-18, 1994, Kansas City, M O , 1994.
3. R. W. Nitske and C. M. Wilson, R u d o l f D i e s e l , P i o n e e r o f th e Power A g e , Univ. Oklahoma
Press, Norman, OK, 1965.
4. E. G. Shay, Diesel fuel from vegetable oils: status and opportunities. B i o m a s s B i o e n e r g y 4 : 227
(1993).
5. J. Walton, The fuel possibilities of vegetable oils. G a s O i l Power, July 1938, p. 167.
6. R. H. Seddon, Vegetable oils in commercial vehicles. G a s O i l P o w e r , August 1942, p. 136.
7. A. W. Baker and R. L. Sweiget, A comparison of various vegetable oils as fuels for compression-
ignition engines. P r o c e e d i n g s o f th e 1 9 t h O i l a n d G a s P o w e r C o n f e r e n c e , American Society of
Mechanical Engineers, Cleveland, OH, 1947, pp. 40-48.
8. J. S. Aggarwal, H. D. Chowdury, S. N. Mukerji, and L. C. Vermoh, Indian vegetable oils as
fuels for diesel engines. B u l l . I n d i a n I n d . R e s . 1 9 (1952).
9. C. E. Goering, M. D. Schrock, K. R. Kaufman, M. A. Hanna, F. D. Harris, and S. J. Marley,
Evaluation of vegetable oil fuels in engines, ASAE Paper No. 87-1586, ASAE, St. Joseph, MI,
1987.
10. C. Adams, J. F. Peters, R. J. Rand, B. J. Schroer, and M. C. Ziemke, Investigation of soybean
oil as a diesel fuel extender: endurance tests II, J . A m . O i l C h e m . S o c . 6 0 : 1574 (1983).
1 1 . W. K. Klopfenstein and S. Walker, Efficiencies of various esters of fatty acids as diesel fuels, J .
Am. O i l Chem. Soc. 60: 1596 (1983).
12. E. H. Pryde. Vegetable oils as diesel fuels: overview, J . A m . O i l C h e m . S o c . 6 0 : 1557 (1983).
13. R. C. Strayer, J. A. Blake, and W. K. Craig, Canola and high erucic rapeseed oil as substitutes
for diesel fuel: preliminary tests, J . A m . O i l C h e m . S o c . 6 0 : 1587 (1983).
14. Y. Basiron and A. Hitam, C o s t E f f e c t i v e n e s s o f th e C P O F u e l in t h e M e r c e d e s E l s b e t t E n g i n e
Car, PO RIM Inf. Ser. No. 4 (Published by PO R IM ), July 1992.
15. T. B. Reed, An overview of the current status of biodiesel. Proceedings of the First Biomass
Conference of the Americas, Voi. II, 1993, p. 797.
16. N. O. V. Sonntag, Fat splitting, esterification and interesterification, in B a i l e y ’s I n d u s t r i a l O i l a n d
F a t P r o d u c t s , 4th ed.. Voi. 2 (D. Swem, Ed.), Wiley, New York, 1982, pp. 97-173.
17. U. R. Kreutzer, Manufacture of fatty alcohols based on natural fats and oils, J . A m . O i l C h e m .
Soc. 61: 343 (1984).
18. B. Freedman, E. H. Pryde, and T. L. Mounts, Variables affecting the yields of fatty esters from
transesterified vegetable oils, J . A m . O i l C h e m . S o c . 6 1 : 1638 (1984).
784 Choo et al.
19. B. Freedman, R. O. Butterfield, and E. H. Pryde, Transestérification kinetics of soybean oil, J.

Am. O i l C h e m . S o c . 6 3 : 1375-1380 (1986).
20 . R. D. Parries, Methyl esters in the fatty acid industry, J . A m . O i l C h e m . S o c . 5 6 : 1 1 0 A (1979).
21 . Y. M. Choo, A. S. H. Ong, K. Y. Cheah, and A. Bakar, Production of alkyl esters from oils and
fats, Australian Patent 43519/89 (1992).
22 . Y. M. Choo and S. H. Goh, Esterification of carboxylic acids/glyceride mixtures, U.K. Patent
214 8897 (1987).
23. Y. M. Choo and A. S. H. Ong, Carboxylic acid esterification, U.K. Patent 216 1809 (1989).
24. S. C. Borgelt, T. S. Kolb, and L. G. Schumacher, Biodiesel: world status, in P r o c e e d i n g s o f a n
A l t e r n a t i v e E n e r g y C o n f e r e n c e (B. E. Dale, Ed.), ASAE, St. Joseph, MI, 1994, pp. 67-76.
25. A. Schafer, The use of biofuel in modem diesel engines. Presented at 1995 PORIM International
Biofuel Conference, Jan. 17-18, 1995, Langkawi, Malaysia.
26. C. L. Peterson, M. E. Casada, L. M. Saflley, Jr., J. D. Broder, L. Perkins, and D. L. Auld,
Agriculturally produced fuels, Paper No. 905532, Am. Soc, Agrie. Engineers, St. Joseph, MI,
1990.
27. L. J. Icerman and E. B. Shultz, Buffalo gourd as a source of diesel fuel and ethanol on arid lands:
preliminary economic analysis. P r o c e e d i n g s o f S e m i n a r I I I o n V e g e t a b l e O i l A s D i e s e l Fuel,
Agrie. Rev. and Manuals, USD A Agricultural Research and Service, Peoria, IL, 1983, pp. 21-22.
28. W. F. Marshall, Effects of methyl esters of tallow and grease on exhaust emissions and perfor
mance of a Cummins LIO engine, IIT Research Institute, Nat. Inst, for Petroleum and Energy
Research, Bartlesville, OK. (Prepared for Fats and Proteins Research Foundation, Inc., Fort Myers
Beach, FL), 1993.)
29. M. Mittelbach and P. Tritthart, Diesel fuel derived from vegetable oils. III. Emission tests using
methyl esters of used frying oil, J . A m . O i l C h e m . S o c . 6 5 : 1185 (1988).
30. F. D. Gunstone, J. L. Harwood, and F. Padley (Eds.), T h e L i p i d H a n d b o o k , 2nd ed.. Chapman
and Hall, London, 1994, pp. 72, 77, 92, 116, 118-119.
31. L. M. Ventura, A. C. Nascimento, and W. Bändel, First results with Mercedes-Benz Di diesel
engines, Int. Conf. on plant and Vegetable Oils as Fuels, Fargo, ND, August 1982.
32. L. E. Wagner, S. J. Clark, M. D. Schrock, and P. O. Pienaar, Soybean oil esters as alternative
diesel fuels, SAE Paper No. 841385, SAE, Warrendale, PA, 1984.
33. J. Einfalt and C. E. Goering, Methyl soyate as a fuel in a diesel tractor, T r a n s . A S E 2 8 : 70 (1985).
34. P. Srinivasa Rao and K. V. Gopalakrishnan, Vegetable oils and their methyl esters as fuels for
diesel engines, I n d i a n J . T e c h n o l . 2 9 : 292 (1991).
35. L. G. Schumacher, W. G. Hires, and S. C. Nörgelt, Fueling a diesel engine with methyl-ester
soybean oil, ASAE Proceedings of an Alternative Energy Conference, Nashville, TN, Dec. 14-
15, 1992.
36. D. Fosseen and W. Goetz, Evaluation of methyl soyate/diesel blend in a DDC 6V-92TA engine:
optimization of NOx emissions, Fosseen Manufacturing and Development, Radcliffe, lA. (Report
prepared for National Soy Diesel Development Board, Jefferson City, MO), 1993.
37. K. W. Scholl and S. C. Sorenson, Combustion of soybean oil methyl ester in a direct injection
diesel engine, SAE Paper No. 930934, SAE, Warrendale, PA, 1993.
38. L. G. Schumacher, S. C. Borgelt, W. G. Hires, and J. K. Humphrey, Biodiesel on the road— a
report from Missouri, ASAE Paper No. 93-5017, ASAE, St. Joseph, MI, 1993.
39. L. G. Schumacher, S. C. Borgelt, and W. G. Hires, Soydiesel/petroleum blend research, ASAE
Paper No. 93-6523, ASAE, St. Joseph, M I, 1993.
40. G. H. Pischinger, A. M. Falcon, R. W. Siekmann, and F. R. Femades, Methyl esters of plant
oils as diesel fuels, either straight or in blends. V e g e t a b l e O i l F u e l s (Proc. Int. Conf. on Plant and
Vegetable Oil as Fuels), Am. Soc. Agrie. Eng., St. Joseph, MI, 1982, pp. 198-208.
41. R. W. Siekmann, G. H. Pischinger, D. Blackman, and L. D. Carvalho, The influence of lubricant
contamination by methyl esters of plant oils on oxidation stability and life. V e g e t a b l e O i l F u e l s
(Proc. Int. Conf. on Plant and Vegetable Oil as Fuels), ASAE, St. Joseph, MI, 1982, pp. 209-
217.
42 . M. E. Feldman and C. L. Peterson, Fuel injector timing and pressure optimization on a DI diesel
Biofuels 785
engine for operation on biodiesel, in L i q u i d F u e l s f r o m Renewable R e s o u r c e s — P r o c e e d i n g s o f a n

Nashville, TN, 1992.
A lte r n a tiv e E n e r g y C o n fe r e n c e ,
43. C. L. Peterson, D. L. Reece, R. Cruz, and J. Thompson, A comparison of ethyl and methyl esters
of vegetable oil as diesel fuel substitutes, ASAE Alternative Energy Conference, Nashville, TN,
Dec. 14-15, 1992.
44. D. L. Reece and C. L. Peterson, A report on the Idaho on-road vehicle test with RME and
neat rapeseed oil as an alternative to diesel fuel, ASAE Paper No. 93-5018, ASAE, St. Joseph,
MI, 1993.
45. M. Mittelbach, P. Tritthart, and H. Junek, Diesel fuel derived from vegetable oils. I. Emission
tests using rape oil methyl ester. E n e r g y A g r i c . 4 : 208 (1985).
46. G. Zhang, M. Feldman, and C. L. Peterson, Diesel engine durability when fueled with methyl
ester of winter rapeseed oil, ASAE Paper No. 88-1562, ASAE, St. Joseph, MI, 1988.
47. L. A. Perkins, C. L. Peterson, and D. L. Auld, Durability testing of transesterified winter rape
oil ( B r a s s i c a n a p u s 1) as fuel in small bore, multi-cylinder, Di, Ci engines, SAE Paper No.
911764, SAE, Warrendale, PA, 1991.
48. S. M. Geyer, M. J. Jacobus, and S. S. Lestz, Comparison of diesel engine performance and
emissions from neat and transesterified vegetable oils, T r a n s . A S A E 2 7 : 375 (1984).
49. C. S. Hawkins and J. Fuls, Comparative combustion studies on various plant oil esters and the
long term ignition engine. V e g e t a b l e O i l F u e l s (Proc. Int. Conf. on Plant and Vegetable Oil as
Fuels), ASAE, St. Joseph, M I, 1982, pp. 184-197.
50. H. Masjuki and M. Sohif, Performance evaluation of palm oil diesel blends on small engine, J.
125 (1991).
E n e rg y H e a t M a ss T ra n sfe r 13:
51. D. Fosseen and W. Goetz, Performance of a 6V-92 engine using methyl ester soybean oil, Fosseen
Manufacturing and Development, Radcliffe, lA. (Report prepared for National SoyDiesel Develop
ment Board, Jefferson City, MO, 1993.)
52. W. F. Marshall, Status of biodiesel work at the National Institute of Petroleum Energy Research
(NIPER), Presented to the research committee of the National SoyDiesel Development Board,
Dec. 10, 1993.
53. J. R. Needham and D. M. Doyle, The combustion and ignition quality of alternative fuels in light
duty diesels, SAE Paper 852101, SAE, Warrendale, PA, 1985.
54. C. L. Peterson, Vegetable oil as a diesel fuel: status and research priorities, T r a n s . A S A E 2 9 :
1413 (1986).
55. J. Fuls, C. S. Hawkins, and F. J. C. Hugo, Tractor engine performance on sunflower oil fuel, J.
A g r ic . E n g . R e s . 3 0 : 2 9 -35 (1986).
56. C. L. Peterson, Vegetable oil as a diesel fuel: status and research priorities, T ran s. A S A E 2 9 :
1413 (1984).
57. J. J. Bruwer, B. Van D. Boshoff, F. J. C. Hugo, L. M. du Plesis, J. Fuls, C. Hawkins, A. N.
van der Walt, and A. Engelbrecht, Sunflower seed oil as an extender for diesel fuel in agriculture
tractors. Symposium of the South African Institute of Agricultural Engineers, June 11, 1980.
58. C. R. Engler, L. A. Johnson, W. A. LePori, and C. M. Yarbrough, B i o m a s s E n e r g y , A Mono
graph, Texas A & M University, 1985, pp. 174-212.
31_____________________________
Products from Castor Oil
Past, Present, and Future
Henri-Jean Caupin
Elf Atochem S-A, Paris, France
I INTRODUCTION
Castor oil has a long history and special status relative to other vegetable oils. Castor oil
seeds have been found in Egyptian graves of 4000 BC. The seed was used in medicine (strong
purgative action), and the hollow stalks of the plant were used to store cosmetics (especially
kohl). Known as huile de carapathes, it was widely used in Africa for care of the skin and
hair. In India, oral tradition described 59 illnesses cured by internal and external use of castor
oil. The oil is clearly different from other vegetable oils, and it is soon apparent that it can
only be ingested in a controlled fashion.
II. COMPOSITION
Castor oil is characterized by the presence of ricinoleic acid (12-hydroxyoleic acid) to the
extent of almost 90%. Its components are listed in Table 1.
Ricinoleic acid has the molecular formula C H 3 (C H 2 )5 C H (O H )C H 2 C H = C H ( C H 2 )7 C O O H .
The presence of a midchain hydroxyl group influences both the physical and chemical proper
ties of this acid and its esters. These substances show many interesting chemical reactions,
some of which are exploited commercially. For this reason the oil should be expressed in a
clear and neutral form, as bleaching and other refining processes may result in its deteriora
tion. Many of the important uses of castor oil depend on the presence of the /3-hydroxyalkene
unit — C H ( 0 H ) C H 2 C H = C H — .
The more sophisticated downstream products are more valuable than the long-established
and local uses of the oil. There are reports of castor oil being used as a fuel when other fuels
are in short supply.
787
788 Caupin
Table 1 Fatty Acid Composition of

Castor Oil (wt %)
Acid Wt %
Ricinoleic 89.5
9,10-Dihydroxystearic 0.7
Palmitic 1.0
Stearic 1.0
Oleic 3.0
Linoleic 4.2
Linolenic 0.3
Arachidic 0.3
Source: Frank N aughton o f Cambrex.
III. MAJOR USES OF CASTOR OIL AND RICINOLEIC

ACID ESTERS
Castor oil has a very wide range of uses [1,2]. Some are simple and traditional and exist only
in the producer countries; others are more advanced and involve chemical modifications car
ried out only in the developed world.
The oil has well-recognized medicinal and cosmetic effects, and significant amounts are
applied to the skin and hair. Almost 60 such uses are recognized in the oral traditions of
India, and some of these uses are reflected in European cosmetic products.
A. Sulfation (Turkey Red Oil)

The first industrial modification of castor oil came in 1874 with the production of Turkey red
oil. Reaction with sulfuric acid and sulfur trioxide results in sulfation of the hydroxyl group
and further modification of the double bond:
— CH(O H)--------> - C H ( 0 S 0 2 0 H ) —
This yields a product with improved surfactant properties. Apart from the soaps known from
ancient times, Turkey red oil was the first anionic surfactant. It is used in textile processing,
inks, industrial detergents, and leather treatment and in lubricant additives for cutting oils and
hydraulic fluids. The sulfated hydrogenated oil has the consistency of an ointment and gives
adjustable viscosity to water-based formulations with excellent skin compatibility.
B. Atmospheric Oxidation
All highly unsaturated oils such as linseed undergo atmospheric oxidation when blown, i.e.,
heated in air. This change occurs much more readily with castor oil, which has free hydroxyl
groups. Oils of different viscosities result depending of the condition of oxidation, and they
have different uses. The products are highly compatible with other resins, polymers, and
waxes. They have low volatility, and there is no migration. They are used extensively in
nitrocellulose lacquers.
C. Dehydration
Dehydration of castor oil (DCO) and of castor acids (DCOFA) converts castor oil to a mixture
of diene esters and acids that are partly conjugated:
Products from Castor Oil 789
12— OH 18:1 (9c) —> 18:2 (9c 1H, 9c 12?, and other isomers)
They have been used as a valuable alternative to drying oils such as tung oil (rich in eleos-
tearic acid), which are not always available in the quantity required. DCO is used in paints
and coatings in traditional nonaqueous systems and in more recently introduced aqueous
systems.
Nonaqueous coatings require drying oils, and DCO has been much used for this purpose.
Other drying oils present problems with smell, availability, and price stability. DCO and
DCOFA are used in the now-preferred aqueous systems and can be incorporated into acrylic
paints and into polyacrylate and epoxy ester systems.
D. Hydrogenation
Hydrogenated castor oil and hydrogenated castor acids have higher melting points than the
nonhydrogenated materials. The products still contain the hydroxyl group, and they are used
in cosmetics as well as in paints, coatings, and greases.
Greases prepared from tallow had many shortcomings, and their quality was dramatically
improved when derivatives of hydrogenated castor oil were incorporated. Lithium salts in
crease the hardness and raise the melting point of the products, permitting their use under
more severe operating conditions and higher temperatures.
Hydrogenated castor products are also used in melamine resins, where they increase flexi
bility; in acrylics; and in paints, where their thixotropic properties permit control of product
viscosity.
E. Reactions with Ethylene Oxide and Propylene Oxide

Reaction of the hydroxyl group in castor oil with ethylene or propylene oxide gives rise to a
nonionic surfactant. By suitable adjustment of the reactants, products of diverse water solubil
ity and of oil/water compatibility can be obtained. Although the products are of high quality,
they are in competition with cheaper products, and supplies of castor oil have been too uncer
tain to warrant the large investment necessary for surfactant markets.
F. Reactions with Isocyanates

Like other hydroxyl-containing molecules, castor oil reacts with appropriate isocyanates to
produce polyurethanes, which find many uses in the developed world. In America polyure
thane coatings for houses made of wood are in great demand, but this is not true in Europe,
where wooden buildings are less common. Polyurethanes based on castor oil have also been
developed for use as encapsulating materials, especially in the electrical industry. Develop
ments in this area are now concerned more with formulations than with new scientific issues
(see Ref. 5) [3].
G. Splitting with Caustic Soda (Cg and Products)

Sodium hydroxide splits the Cjg ricinoleic acid molecule into Cg and C jq products, the exact
nature of which depends on the reaction conditions. At 180-200°C with a caustic/castor ratio
of 1:1, the major products are 2-octanol and 10-hydroxydecanoic acid. At 250-275°C, with a
caustic/castor ratio of 2 : 1 , the products are mainly 2 -octanol and sebacic (decanedioic) acid.
Of these the most valuable product is the dibasic acid (sebacic). This has been used along
with the diamine to produce polyamide 6, 10 , which has better solvent and water resistance
than the shorter chain polyamides. The dibasic acid has also been combined with a range
of alcohols to give diesters that serve as efficient lubricants at both low and high temperatures.
790 Caupin
H. Splitting with Steam (C7 and Products)

The historical procedure for splitting the Cjg ricinoleic acid chain into C7 and Cn fragments
suffered from low yields and high environmental pollution. It was carried out mainly to give
C7 and Cji aldehydes, which are used as perfumes and fragrances. Heptanal is still condensed
with benzaldehyde to give amyl cinnamic aldehyde (P hC H =C (C 5C 1i)CHO). Undecenal, a
key component of high quality perfumes and a quality enhancer of large volume products, is
still made from undecenoic acid. The older uses are only a small part of the current applica
tions of these materials, to be discussed in detail later.
The antifungal properties of the zinc and calcium salts of undecenoic acid were recognized
in the early 1940s, and these compounds are still used extensively for this purpose, particu
larly for the care of adult feet and baby bottoms.
The C7/C 11 splitting process has been much improved by the development of a continuous
steam cracking process. The production streams are managed on the basis of the sales profile,
and in recent years there have been many new developments in the use of these C7 and
compounds. These are discussed in the following section.
IV. NEW USES OF CASTOR OIL DERIVATIVES WITH

SPECIAL REFERENCE TO C7 AND C^ COMPOUNDS
A. C7 Compounds
The C..7 compounds are available as aldehyde, acid, and alcohol. These compounds are of
interest in that they have borderline properties between short- and long-chain compounds.
1. C orrosion Control
Heptanoate salts act as stable anticorrosion agents even in hard water. The sodium salt gives
efficient corrosion control at concentrations between 0.5 and 4.0 g/L. The pK of calcium
heptanoate at these levels is such that there is no precipitation of the metallic soap. This is
also true for the acid but not for derivatives of the Cg, C9, and Cjo acids.
The protective coatings of heptanoic acid obtained on copper, iron, aluminum, and zinc
are probably molecular layers of organized crystals. Further investigation is needed to find the
optimum levels and formulation conditions for corrosion control. It is known that the acid
is too short to provide protection and that whereas the Cg, C9, and Cio acids are protective
their efficiency is handicapped by hard water precipitation at the concentration required for
corrosion control. It appears that the C7 acid is unique in combining corrosion control with
hard water compatibility [4].
2. Low Tem perature Properties

The boiling and freezing points of heptanoic acid derivatives are generally such that these
compounds are not lost in normal high temperature use, and they remain liquid at low temper
atures. For example, they make good plasticizers for polymers to be used at very low tempera
tures.
Among the many esters examined as potential lubricants for jet engines, automobiles, and
stationary engines, those based on heptanoic acid continue to find a place because of their
outstanding properties.
B. Compounds
In addition to its antifungal properties, undecenoic acid finds other uses both actual and poten
tial based on its differing functional groups (double bond and carbonyl) at either end of the
molecule. Some recent studies are described in patents [2, 5-10].
1. P olyam ide 11
Currently the major use of undecenoic acid is the production of 11-aminoundecanoic acid
and its polymerization to give polyamide 11. This has superior water resistance and solvent
compatibility compared with polyamides 6 and 66. Under dry conditions its mechanical prop
erties are weaker than those of the shorter chain polyamides, but they are much better main
tained under long-term use and wet conditions than those of the polyamides.
The physical mixture of polyamide 11 and silk has been known for some time, since
polyamide 11 is a better quality improver for silk than the shorter chain polyamides. This is
due to a very meaningful and highly surprising technical feature: Silk takes in about 10%
water, but the dimensional changes are sideways—the length of the filament remains constant,
as does that of polyamide 1 1 .
2. N ew Uses fo r Compounds
Other uses now developed for compounds depend on the previously established antifun
gal, antismell, and insecticidal properties of the various derivatives, of undecenoic acid.
Biostability and Basic Antifungal Properties. Formulators of performance products are
much concerned with the preparation of water-oil emulsions that maintain physical stability
over the life of the product. This holds for a wide range of water-based products such as
paints, cosmetics, cutting oils, and other types of emulsified compounds. Once the necessary
physical stability has been secured, biological stability must also be guaranteed from produc
tion through to consumption.
Currently, the necessary biocides are added late in the production process, and they must
be selected to have no deleterious effect on emulsion stability. Appropriate rules and regula
tions have to be applied to comply with EPA and FIERA documents. A compound can be
described as a biocide only if it decreases the existing germs in a water-based formulation by
a factor of 100-1000 in a given very short time.
Some compounds combine natural and limited bioresistance as well as surfactant activity.
Fats present in hair that exercise a protective function include significant levels of acids,
no or very little Cj 2 acid, and a mix of C 14, and C^g acids. This is important when the
high fermentation situation around the hair is recognized. With a constant mild temperature
and permanent humidity at root level, it is necessary to have an effective bioresistant and
renewed mixture. This must allow for constant and healthy control of microorganisms and for
permanent monitoring of the scalp lice population [ 1 1 , 12 ].
The situation for many organic-water emulsions is similar to that of the hair and scalp,
and the problem is how to introduce an approved compound without destroying the physi
cal stability of the emulsion. Many trials over the last 10 years have failed because the Cn
compounds differ from conventional surfactants and tend to destroy the emulsions. However,
a number of specialized derivatives are now available that can be used in the preparation
of emulsions and are introduced into the formulation early rather than at a later stage [13].
The Cji derivatives mentioned are the counterparts of the even-chain fatty acid derivatives,
already long known by any cosmetics formulator.
792 Caupin
Economic and regulatory constraints encourage formulations with controlled levels of ap
proved biocides. This can be achieved through fine-tuning of combinations of naturally de
rived bioresistant materials. compounds at 10-20% levels are already allowed in OTC
formulations, and the addition of such compounds at levels of 0.1-5.0% could assist in reduc
ing biocide levels.
Animal Waste Smells. esters decrease the malodorousness of pig and other animal
wastes to acceptable levels. Efficient addition rates are below 100 g/t (i.e., <100 ppm).
Smells associated with water treatment stations and with industrial effluents can be reduced
by Cii esters in the same way.
Antilouse Activity o f Selected Derivatives. Compounds similar to those just described
have also been used for the control of lice in human situations. Details are given in Ref. 7.
Boron-Free Cutting Oils Technology. C7 and Cn monoethanolamides are used in cutting
oils where they replace boron compounds and reduce the environmental impacts of the efflu
ent. The combined efficiencies allow optimization of the corrosion control as well as of the
natural bioresistance of the emulsions.
Evidence o f Specific Properties Linked to Acetylenic Bifunctionality in the Linear
Chain. The products are unique in that they contain different functional groups (one
carboxyl group and one unsaturated center) at the ends of the molecule. The common unsatu
rated acids have midchain unsaturation. The a,co difunctional aliphatic compounds are either
short-chain (C3, C4) or neutral (a-olefins) or contain two identical functional groups (dibasic
acids, diols), so differentiation becomes difficult.
There is increasing information about the surface effects of compounds that appear to
be spontaneously organized.
Typical “self-assembled biolipid tubular templates” are shown in Ref. 14, and new meso-
genic diacetylenic monomers and polymers in Ref. 15. Perfluoroalkylated dienes can be ob
tained from perfluoroalkylated alkynes, and these lead to fluorinated analogues of Lepidoptera
pheromones [16]. An important pheromone from Leucoptere scitella can be obtained from
conjugated diyne diols [17, 18].
Molecular assemblies of functionalized diynes are described in Refs. 19 and 20 along with
polymerization of supramolecular assemblies requiring a monoyne and conjugated C22
diynes. Rare dienoic acids of marine origin can be obtained starting from undecynoic acid
[21]:
CH2=CH(CH2)8C00H-^ CH ^CH 2)8C00H -> H00C(CH2)8C^C-C^C(CH2)8C00H

Receptor-ligand interactions have been identified using diyne compounds as part of the
lipid system. Direct colorimetric detection of the poly-ynes can be exploited for diagnostic
applications in various pharmaceutical fields [22]. Glycerides combined with C22 diyne struc
tures derived from the acid inhibit the influenza virus in vitro [23].
In addition there are many abstracts making reference to diynes based on the medium-chain
Cii material. A partial list of applications over the period 1980-1994 is given in Table 2.
It is recognized that the cost of undecynoic acid and its derivatives is very high and there
is no industrial source of these at reasonable and reliable cost. Laboratory-prepared material
is reported to be available at $100 per gram for 99% pure grade and around $20 per gram for
a 95% pure grade.
Undecynoic acid, its alcohol, and other derivatives have been known and used for experi
mental purposes since the beginning of the century. It is clear from the list in Table 2 that
there is a potential demand for acetylenic compounds of this chain length and the materials
Table 2 Applications of Diynes Based on Undecynoic Acid—A Time Line
1980 Photographic element

1985 Nonlinear optical elements
1986 Electrically conductive polydiacetylenes
1986 Semiconductor device comprising an organic material
1987 Recording medium for optical memory
1987 Urethane-substituted diacetylene monomers
1987 Thermochromie polydiacetylenes as temperature indicators
1987 Polym erized diacetylene Lan g m uir-B lo dg ett films as optical modulators
1988 Laser recording
1989 Monosubstituted diacetylene compounds for electrically conductive polymers and resist materials
1989 Polarized inner-shell absorption spectroscopy of organic molecular and polym eric solids
1989 Preparation of polymerizable fatty acids
1990 Reaction of diynoic acid chlorides with para-substituted phenols
1990 Coating for increasing sensitivity o f a radiation imageable polyacetylenic film
1990 Polydiacetylene derivative-containing thin film for electric conductor
1990 Process for sensitizing polyacetylenic im aging dispersion
19 9 1 Organic film materials
19 9 1 Lan g m uir-B lo dg ett organic films and preparation thereof
19 9 1 Recording medium of improved stability
19 9 1 Immunolesioning : selective destruction of neurons using immunotoxin to rat N F G receptor
19 9 1 Optical recording medium and method
19 9 1 Preparation of heteroarylaminoalkanols as drugs
1992 Preparation of P L A 2 inhibitors as anti-inflammatories
199 2 Process for the preparation of 2 -a lk y l-4 ,6-fór/-butylphenol compounds
1992 Manufacture of liquid crystal orienting film
199 2 V isu a lly imageable polyacetylene salt dyes
1992 Recovery of imageable polyacetylene compounds from binder containing aqueous dispersions
199 2 Gas-encapsulated or gas precursor encapsulated microbubbles for ultrasound contrast agents,
A D N preparation, and polymerization of amphiphiles therefor
199 2 Process for acylation of 2-alkyl-6-fórí-butylphenol
1993 Preparation of 2,3-epo xy derivatives as anti retroviral pests
1993 Contrast agents consisting of carbohydrate particles and amphiphilic carboxylic acids
1993 Spicam ycin derivatives and the use thereof
1993 Phospholipids for modulation of microstructure morphology
1993 Preparation of cis olefins
1994 Photochemical labeling of nucleic acids with europium-chelating reagents and their use in gene
probe systems
1994 M olecular composites
derived from them. This eould be a significant and valuable development of castor oil, be
cause there is no convenient source of these undecynoic acid derivatives other than the steam
cracking of this oil.
V. FUTURE TRENDS FOR CASTOR OIL AND

ITS DERIVATIVES
Castor oil is the most convenient source of C7 and compounds and materials made from
them. This arises from the unique structure of ricinoleic acid (12-hydroxyoleic acid) and from
the fact that these products are easily available only from castor oil. If the demand for C7/C 11
794 Caupin
chemicals increases in the next 5-10 years, then the pressure to find a bioengineered solution
will be strengthened. Attempts are already being made to transfer the required gene(s) to the
rape plant so that it generates ricinoleic acid. Such a modified rape could be grown in nontrop-
ical areas in North America and northern Europe.
The most recent Berkeley publications report on delicately organized “nanochemical” lay
ers for many applications. The basic requirement appears to be a difunctional chain with
terminal ethylenic or acetylenic unsaturation that would provide layers of 1 nm (1 x 10 “^ m)
thick. We are still a long way from that, with layers of 1 fxm thickness, i.e., 1000 times
thicker. There is a need for creative thinking by new generations of chemists and technologists
to develop this nanochemistry, and ricinoleic acid, as the most convenient source of
compounds, will have a significant role in this.
The existing uses of castor oil will continue to be significant, and the oil of Ricinus commu
nis will continue to be the biggest source of ricinoleic acid for many years. New developments
can only add to the value of this material.
ACKNOWLEDGMENTS
This review is based on information in the public domain and on information coming from
Elf Atochem relating to that company’s products and to its castor oil cracking process. I also
thank Frank Naughton and Frank Vignolo, both of whom played a significant role in ICO A,
itself part of the American Oil Chemists’ Society. Frank Gunstone of St Andrews, a joint
editor of this book, made useful suggestions based on his knowledge of oleochemistry. He
also assisted me in the production of the final version of this chapter. I also thank many others
who have published in this field and whose work is cited.
Five key companies making extensive use of castor oil issue booklets and data sheets on
their range of products: Alberdingk Boley GmbH, Germany; Caschem Inc, Elizabeth, NJ
(former Baker Castor Oil Co); Elf Atochem, Paris, France; Itoh Oil Manufacturing Co, Yok-
kaichi, Japan; and Thai Castor Oil Industries Co, Bangkok, Thailand.
REFERENCES
1. K irk-O thm er Encyclopedia, Interscience, New Y o rk , 1994, pp. 3 0 1 -3 2 0 , includes 12 0 references.
2. F . C . Naughton and Robert L . Vig n olo , T he C h e m istry o f C a s to r O il D e r iv a tiv e s a n d T h eir A p p li
c a tio n s , Int. Castor O il A sso c., W estfield, N J; includes 39 1 references.
3 . E lf Atochem S -A , Undecylenate deodorants for animal manures, U .S . Patent 006050 17 3 (Dec.
19 , 1990).
4. E lf Atochem S -A , Corrosion inhibitory formulation based on carboxylic acid and corrosion inhibi
tion activity for ferrous and nonferrous metals, Fr. Patent 92 14 2 33 (Mar. 24, 199 5).
5. E lf Atochem S -A , Deodorizing the muds and sludges from water treatment stations using C n
ester, Eur. Patent 0434 5 2 2 B 1 (Mar. 1, 199 5); U .S . Patent 5 .3 3 8 .5 1 1 (Aug. 16 , 1994).
6. E lf Atochem S -A , A nim al nutrients deodorizing agents using undecylenic acid ester, Eur. Patent
0 434 524 B 1 (June 15 , 1994).
7. E lf Atochem S -A , L ic e killing formulation using undecylenic acid derivatives, Fr. Patent 92 06
149 (M ay 20, 1992).
E lf Atochem S -A , Deodorizing paper m ill effluents with undecylenic acid esters, Fr. Patent 92 09
457 (July 30, 1992).
9. E lf Atochem S -A , New polymer resin loaded with undecylenic ester, Fr. Patent 92 13 369 (Nov.
6, 1992).
10. E lf Atochem S -A , A n im al dejecta deodorizing formulation and full process treatment, Fr. Patent
93 07 6 3 1 (June 2 3 , 1993).
11. L. Leon, L. D. Gershbein, and B. S. Metcalfe, J. Invest. Dermatol. 46(5): 477-479, (1966).
12. W. Stodt Meister, H. P. Nissen, C. Schirren, and H. W. Kreysel, Fettsaüreverteilung der Haarlip-
ide bel der Psoriasis vulgaris. Arch. Dermatol. Res. 264: 399-343, (1979).
13. A. Domsch, Chemistry and application of undecylenic acid and derivatives thereof for cosmetics,
SÔFWJ. 6: 120 (1994).
14. S. L. Burkett and S. Mann, Spatial organization and patterning of gold nanoparticles on self-
assembled biolipid tubular templates, Chem. Commun. 1996: 321.
15. P. T. Hammond and M. F. Rubner, Synthesis and characterization of new mesogenic diacetylene
monomers and their polymers. Macromolecules, 28(4) (1995).
16. F. Saremi, E. Maasen, and B. Ticke, Self-assembled alternating multilayers built up from diacety
lene and bolaamphiphiles and poly(allylamine hydrochloride): polymerization properties, structure,
and morphology, Langmuir 11: 1068-1071 (1995).
17. Z. Wang and X. Lu, Stereoselective synthesis of perfluoroalkylated (E, E) dienes from perfluoro-
alkylated alkynes: The synthesis of fluorinated analogs of Lepidoptera pheromones. Tetrahedron
51: 11765-11774 (1995).
18. B. Crousse, O. Pichón, and J. Soulié, Synthèse totale du (lOZ, 13Z, 16Z)-nonadeca-10,13,16,18-
tetraen-l-ol, Elsevier, Paris, 1993.
19. W. Spevak, J. O. Nagy, and D. H. Charych. Molecular assemblies of functionalized polydiacety
lenes, in VCH, Weinheim, 1995.
20 . D. F. O’Brien, C. S. Marvel Laboratories, Polymerization of supramolecular assemblies. Trip
2(6): June (1994).
21 . B. A. Kulkami, A. Chattopadhyay, and V. R. Mammdapur, A facile synthesis of (lOZ, 15Z)
cicosa-10,15-dienoic acid, a marine natural product, J. Nat. Prod. 57(4): 537-538 (1994).
22 . D. H. Charych, J. O. Nagy, W. Spevak, and M. D. Bednarski, Direct colorimetric detection of a
recepto-ligand interaction by a polymerized bilayer assembly. Center for Advanced Materials,
Lawrence Berkeley Laboratory, Berkeley, CA.
23. J. C. Paulson, Polymerized liposomes containing C-glycosides of sialic acid: potent inhibitors of
influenza virus in vitro infectivity, in The Receptors (M. Conn, Ed.), Academic, New York.
24. Elf Atochem S-A, Filtration process equipment for a mud and specific equipment, Fr. Patent 94
03 814 (March 31, 1994).
Index
Numbers in italic refer to figures; “t” refers to tables.
Absorption bands, epoxides, 761 [Adulteration]

Aceituno oil, fatty acid composition, 25 cocoa butter equivalents, determination of,
Acetic acid esters, food emulsifiers, 526, 527 423
Acetone deoiling, phospholipid processing, 55-56, Aeration, ice cream freezing and, 339-340,
56t 340
Acetylation, phospholipids, 60 Aging, ice cream mix manufacture, 338
Acetylenic acids, 7t, 7-8 Alcohol, dilute, refining process, 142, 143-144,
Acidolysis, interesterification, 224 144
Acyl derivatives, alcohols, 16 Alcoholates
Active oxygen method, 438 interesterification, 229
Acylglycerols, 11-13, 12-13t rearrangement catalysts, 231
diacylglycerols, 12 Alcohol fractionation, phospholipid processing,
diglycerides, 12 56-57, 57t
monoacylglycerols, 12 Alcoholysis, interesterification, 224
monoglycerides, 12 Aleurites moluccana {see Candlenut oil)
triacylglycerols, 11, 12, 13t Aleurites triloba {see Candlenut oil)
triglycerides, 11 Aleurites trisperma {see Candlenut oil)
with palmitic, oleic acids, alone, 12t Algae toxicity test, lubricants, 741
Additives Alkali-catalyzed interesterification, 231-
frying, salad oils, 436-437 237
antioxidants, 436-437 Alkali metals, interesterification, 230-231
metal chelating agents, 436 Alkaline contaminant materials, in frying oil,
nitrogen treatment, 437, 437 446-447
lubricant, environmentally sound, 747-751 Alkanoic acids
Adulteration names, 4t
chocolate, determination of, 423 physical properties, 4t
cottonseed oil, 424 Alkenes, as substrates, epoxidized oils, 763
palm fractions, 424 Alkoxides, interesterification, 231
soybean oil, 423-424 Alkyd resins, 723-730
797
798 Subject Index
Alkylene oxide, nonionic surfactants, 656-660 [Animal feeds]

polyalkylene oxide copolymers, 659-660 protected fats, 569-570, 570t
polyoxyalkylene fatty alcohols, 657-658 sheep, 570, 570t
polyoxyethylene alkylamides, 658-659 Anionic detergents {see Detergents)
polyoxyethylene alkylamines, 658 Anisidine value, 179
polyoxyethylene alkylphenols, 657 Anteiso acid, 8
Alkyl esters, biofuels, from vegetable oils, fats, Antifungal properties, castor oil, 791-792
782 Antilouse activity, castor oil, 792
production technology, 772-773, 773t Antimicrobials, epoxy acid, 759
Alkyl oleates, epoxidized oils, 759 Antioxidants
Alkylphenols, nonionic surfactants, 647-648 as additives, to salad, frying oils, 436-437
Allenic acid, 7t, 7-8 lubricants, 751
Almond oil, 5 phospholipids, 68
botanical name, 20, 21 Antistatic treatments, cationic surfactants, 626
fatty acid composition, 25 Antiwear additives, lubricants, 749-750, 750t
Aluminum, oxidation, effects on, level, 186 Apricot oil
Alzheimer's disease, phospholipids and, 71 botanical name, 20
Amide formation, nonionic surfactants, 655-656 fatty acid composition, 25
Amidoamines, cationic surfactant, 612 Arachidic acid, 5
Amine oxides, nonionic surfactants, 660 physical properties of, 4
Amphoterics, manufacture of, 626 usage of term, 1
Anacardium occidentale {see Cashew) Arachidonic acid, 2, 3, 7
Animal feeds, lipids in, 557-577 structure, double bond position, 6
metabolizable, digestible energy, prediction Arachis hypogea {see Arachis seed oil)
equations, 560t Arachis seed oil, botanical name, 20
nonruminants, 558-567 Argane, fatty acid composition, 25
adipose tissue fatty acids, rate of change, Argania spinosa {see Argane)
566-567, 567 Armeniaca vulgaris {see Apricot oil)
coconut-palmkemel oil blend, chemical Artist's oil, botanical name, 20
analyses, 56It Asphalt {see Bitumen)
contaminants, 564-565 Atomizers, spray drying, 483-485, 484
cyclopropenoid fatty acids, 565 Attalea speciosa {see Babassu oil)
digestibility, chemical structure and, 558-565 Auricolic acid, 10
erucic acid, 565 Autoxidation
fat, oil utilization enzymatically catalyzed, surface coatings, 721
dietary energy value, chemical structure photoinitiated, surface coatings, 721
and, 558-565 Avenasterol, sterol composition, 29
free fatty acid content, 559-562, 559-562, Avocadin cream, formulation, 701, 702t
560-56U Avocado oil
heat damage, 560t, 562-564, 564t botanical name, 20
meat quality, effect on, 565-567 fatty acid composition, 25
pesticides, 564-565 Aztecs, use of chocolate, 391
physical texture, effect on, 565-566
polytene, 565 Babassu oil, 411
saturation, 559-562, 559-562, 560-56It botanical name, 20, 21
sunflower acid oil, chemical analyses, 564t fatty acid composition, 25
ruminants, 568-574 Bacteria toxicity test, lubricants, 741
absorption, 568t, 568-570 Baking fats, 323-326, 325t, 386
beef cattle, 570-571 bakery compounds, 325, 325t
commercial setting, use of fats in, 570-574 butter, 325
cows, 571, 571-574, 573 fatty acids, 323-326, 386-387
milk, 574t India, 386
digestion of fats, 568t, 568-570 basis, 386
feeds, energy, 568t blending, 386-387
fermentation, 568t, 568-570 defined, 386
Subject Index 799
[Baking fats] Bitumen, 623

hardening, 386 cationic surfactant, 623,624t
production parameters, 386-387 Blackcurrant oil, 34
technologies botanical name, 20
current, 386-387 Black mustardseed oil, botanical name, 20
early, 386 Bleaching, refining process, 145-149
phospholipids, 65-66 bleaching clay
plasticization, 386-387 filtration, 148-149
puff pastiy, 325 spent, 149
shortening in cakes, 324 equipment for, 147-148,148
short pastiy, 324-325 plant operation, 149
Barrier properties, coatings, films, 457-461, Sparbleach process, reactor for, 148
461-462t Bleaching earths, vanaspati production, 380
Bassia longifolia {see Illipe butter, Nowrah butter) Blood cholesterol level, low calorie fats and,
Bath additives, anionic, 602 511-512
Beechnut oil, botanical name, 20 Bloom
Beef cattle, lipids in feed, 570-571 chocolate substitutes, 47ft 416-417
Beef tallow, fractionation, 219, 219t cocoabutter, 405, 417-418, 4\1A\9
Beeswax, 16 double chain-to-triple chain transition, StOSt,
as coating, 460 417
Behenic acid, 5 oleic acid configurations, 417
physical properties of, 4 polymorphic change, 417
Bench endurance tests, biofuels, 775-777, triglycerides, in inhibiting bloom, 419
778-78U fat, chocolate, 416
Berthrolletia myrtaceae {see Brazil nut) 'Blossom thinning," cationic surfactant, 625
Binder, defined, 714-715 Bodying, defined, 714-715
Biocidal cleaners, anionic, 603 Borage seed oil, 34
Biocides, cationic surfactant, 624, 624t, 624-625 botanical name, 20
Biodegradability, lubricants, 739-740, 740 Borago officinalis {see Borage seed oil)
Biodiesel biofuels, production technologies, 773t Borneo oil, 5
Biofuels, 771-785, 772t, 776t botanical name, 20
alkyl esters, from vegetable oils, fats, 782 Boron-free cutting oils technology, castor oil, 792
production technology, 772-773, 773t Brain
bench endurance tests, 775-777, 778-78It fatty acid composition, 53
engine testing, 775-782 phospholipids, 52
fuel characteristics, 774-775, 776t Branched-chain acids, 8
methyl esters anteiso acid, 8
castor oil, 781 iso acid, 8
cottonseed oil, 780-781 phytanic acid, 8
karanjaoil, 781 phytol-based acid, 8
maize oil, 781 pristanic acid, 8
palm oil, 780 Brassicajuncea {see Brown mustardseed oil;
peanut, 781 Indian mustardseed oil)
rapeseed oil, 779-780 Brassica napus {see Colza oil)
ricebran oil, 781 Brassica nigra {see Black mustardseed oil)
soybean oil, 778 Brassica rapa {see Turnip seed oil)
sunflower oil, 780 Brassicasterol, sterol composition, 29
tallow, 781 Brazil nut oil, fatty acid composition, 25
vegetable oils, fuel characteristic, 776t Brown mustardseed oil, botanical name, 20
production technologies, biodiesel, 773t Buffalo gourd oil, fatty acid composition, 25
vegetable oils Bureau of Indian Standards, vanaspati production,
alkyl esters, 773-782, 774-775t requirements, 381
fatty acid composition of, 774t-775t Butanoic acid, 4 {see Butyric acid)
yield, oil-producing crops, 772t Butter, 44-45, 45t, 305-307,400 {see Milk fat)
Biscuit cream fillings, 425 in baking, 325
800 Subject Index
[Butter] Castor oil, 1, 9, 31

churning, 306 antifungal properties, 791-792
composition of, 307, 310t antilouse activity, 792
cream preparation, 306 atmospheric oxidation, 788
draining, 306-307 as biofuel, 781
drying, 492 biostability, 791-792
early developments, 305 boron-free cutting oils technology, 792
fat content, 309 botanical name, 20
fatty acids, 23,45t C7 products, 790-793
historical overview, 305 Cg products, 789
kneading, 307 C,o products, 789
manufacture, 306-307 Cji products, 790-793
production of, 22 caustic soda, splitting with, 789
salting, 307 composition, 787, 788t
spreads, fat content, 309 corrosion control, 790
triacylglycerols, 45t dehydration, 788-789
washing, 306-307 dehydration of, surface coatings, 717
Butter powder, spray drying, 492 diynes, applications of, Undecynoic acid—time
Butyric acid, physical properties of, 4 line, 793t
Butyrospermum paradoxum {see Shea butter) ethylene oxide, reactions with, 789
Butyrospermum parkii (see Shea butter) fatty acid composition, 23, 788
future products, 793-794
Cakes shortening, 324 hydrogenation, 789
Calendic acid, 7 isocyanates, reactions with, 789
Calendula officinalis (see Calendic acid) low temperature properties, 790
Calorie reduction, from fat (see Low calorie fats) polyamide 11, 791
Camelina saliva (see Cameline seed oil. Gold of production of, 22
Pleasure) products from, 787-795
Camelina sinensis (see Teaseed oil) propylene oxide, reactions with, 789
Cameline seed oil, botanical name, 20 ricinoleic acid, molecular formula, 787
Campesterol, sterol composition, 29 saponification, iodine values, 27
Cancer, lipids and, 105-106 steam, splitting with, 790
Candelilla wax, as coating, 460 sulfation, 788
Candlenut oil uses of, 788-790
botanical name, 20 Calalpa ovata (see Catalpic acid)
fatty acid composition, 25 Catalpic acid, 7
Candy coatings, lipid, 467-468 Catalyst (see Rearrangement catalyst)
Cannabis oil, botanical name, 20 Caustic soda, castor oil, splitting with, 789
Cannabis saliva (see Cannabis oil) Ceiba penlandra (see Kapok seed oil)
Canola oil (see Rapeseed oil) Center filling fats, 426
Caprenin, 505 Centrifugal separation, fractionation, 209-210,
Capric acid, 4 210t
physical properties of, 4 Cephalin, usage of term, 51
Caprylic acid, 4 Cephalocrolon cordofanus, vemolic acid, 760
physical properties of, 4 Ceramide, 16, 17
Carbons, vanaspati production, 380 Cerebroside, 17
Carboxyl/carboxylate esters, fat substitute, 507 Chamber designs, spray dryer, 485-489, 486-488
Carboxylic acid, 1 Cheese coatings, lipid, 469-470
Cardiovascular disease, lipids and, 103-105, Chemical refining, 139-158
104t batch refining, 145
Camauba wax, as coating, 460 bleaching, 145-149
Carrot oil cream, formulation, 701, 702t bleaching clay
Carlhamus linclorus (see Safflower oil) filtration, 148-149
Carya illipensis (see Pecan) spent, 149
Cashew oil, fatty acid composition, 25 equipment for, 147-148,148
Subject Index 801
[Chemical refining] [Chocolate]

plant operation, 149 crystallization, 415-416
Sparbleach process, reactor for, 148 crystal size, 398-399
degumming, 139-140 defined, 392
deodorization, 150-158 dilatometry, 415
batch deodorizer, 157-158 differential scanning calorimetry, 415
chelation, 154 enrobing
cooling, 154 crystallization during, 397-400, 399
deaeration, 152 equipment, 399
designs, for deodorizer, 157 fat, types, 400t
equipment, 154-158 formulations, 392t
fully continuous deodorizer, 155-157,157 ice cream technology, nontempered chocolate,
flow diagram, 157 399
heat recovery, 153 manufacture of, 391-394, 392t, 393
high temperature heating, 153 melting, 415-416
polish filtration, 154 milk chocolate
preheating, 152-153 contents of, 392
process variables, 151-152 development of, 391
semicontinuous deodorizer, 155,156 formulation of, 392
flow diagram, 156 molding
theoretical treatments, 151 crystallization during, 397-400, 399
with dilute alkali, 142, 143-144,144 equipment, 399
Zenith, 144 nuclear magnetic imaging, particularly pulse,
drying, 143 415
mixer combinations, multipurpose refining nontempered, 399
plant, 142 nuclear magnetic resonance, wide-line, 415
neutralization refining, 140-143,142 particle size distributions, 394t
pretreatment, 139-140 processing, 392-394
silica refining, 149-150 rheology, 394t, 394-396,395-396, 396t
washing, 143 shear rate-shear stress curves, 394-395
Chemical Shifts, 766t, 766 tempering, 396-397, 398
Cherry oil, fatty acid composition, 25 role of, 398
China wood oil {see Tung oil) tempermeter, for chocolate tempering, 397
Chinese vegetable tallow, 4,407 vegetable fat content, determination of, 423,
botanical name, 20 423-424, 423-425
fatty acid composition, 25, 407 com oil, 423-424
Chocolate, 391-400, 415-425 rapeseed oil, 423-424
Aztecs, South American, 391 sunflower oil, 423-424
bloom, 416 viscosity
fat bloom, 416,417-419 phospholipids, 65
sugar bloom, 416 ranges, 396t
Casson equation, shear rate-shear stress X-ray diffraction, 415
curves, 394-395 yield values, ranges, 396t
cocoa bean, 391 Chocolate substitutes, bloom, 410, 416-417
extraction of cocoa butter, 391 Cholesterol
cocoa butter, 392 levels, low calorie fats, 511-512
adulteration, 422-425, 423-424 sterols, 29
sterol analysis, 422 composition, 29
triglyceride analysis, 422-423 Chromatographic purification, phospholipid
polymorphic behavior, 396-397 processing, 57-58, 58t
composition, 391-394, 392t, 393 Chromium, oxidation, effects on, level, 186
contraction, prior to demolding, 399 Chrysanthemum coronarium, coronaric acid, 760
cooling curves, empirical ciystallization, Churning, in buttermaking, 306
415-416 cis addition process, epoxidized oil, 763
crumb process, 394 cis epoxy esters, epoxidized oils, 766t
802 Subject Index
Citric acid esters, food emulsifiers, 528 [Cocoa butter]

Cleaning, natural sources, before extraction, 113 botanical name, 20
Clothes washing detergents, anionic, 595-597, chemical composition, 401
596t in chocolate, 392
Coatings (see also Surface coatings) crystallization, phospholipids, 404
food, 453-479 enzyme interestification by, 258t
applications, 4C3-471 equivalents, 401-404t, 406, 406-408,407t, 409t
barrier properties, 457-461, 461-462t aceituno acid, fatty acid composition, 407
candy, 467-468 adulteration, determination of, 423
cheese, 469-470 Chinese vegetable tallow, 407
Code of Federal Regulations, 456 fatty acid composition, 407
confectionery, 467-468 Dunkwa allanblackia, fatty acid
diglycerides, 458 composition, 407
of fat with fat, 471 enzymatic interesterification, 407-408
Fickian diffusion, 455 fatty acid composition, 407t
fish, 469-470 film barrier, 460
fruit, 463-465, 464t illipe butter, fatty acid composition, 407
deciduous fruit, 465-466 kokum tallow, fatty acid composition, 407
mass transport in, 454, 454-455 malabar tallow, fatty acid composition, 407
materials, 456-457,458-460t mowrah tallow, fatty acid composition, 407
meat, 469-470 palm oil
moisture-sensitive ingredients, fat fatty acid composition, 407
encapsulation, 410-47 \ fractionated from solvent, 406
nut, 468-469 quality control, 406-407
oxygen permeabilities, 462t sal
physical properties of, 460t fatty acid composition, 407
raisins, 468-469 microbial attack on, 407
starch-based product coatings, 470 shea oil
starch-based products, 470 fatty acid composition, 407
U.S. Food and Drug Administration, fractionated from solvent, 406
approvals, 458t-459t, 464t in supercoatings, 408
vegetable coatings, 466-467 tailoring of, 408
water vapor permeabilities, 46It triacylglycerol composition, 402t
oils, fatty acids in extraction from cocoa bean, 391
autoxidation, en2ymatically catalyzed, fatty acids, 40It, 402t
721 composition, 24,407
chemical processes, 715-722 form V, chocolate crystallized in, 403-404
cobalt redox system, 719-720 fractions of, 405-406
degradation, 721-722 glycerides in, 401,403
diying, 720-721 grades of, 405-406
dryers, 720 nucléation
scission, 719 ciystallization, 404
weathering mechanisms, 721-722 phospholipids and, 404
yellowing, 721 polymorphic behavior, 396-397
Cobalt redox system, surface coating, 719-720 saponification, iodine values, 27
Cocoa bean, 391 stabilized solid content, ternary mixture
extraction of cocoa butter from, 391 POP-POSt-StOSt, 404,405
Cocoa butter, 5, 31-32, 401-406 symmetrical disaturated oleo glycerol esters,
adulteration, 422-425, 423-424 32t
sterol analysis, 422 triacylglycerol composition, country of origin,
triglyceride analysis, 422-423 402t
bloom triglyceride composition, country-by-country
double chain-to-triple chain transition, StOSt, listing, 403t
417 variability of, 404-405
oleic acid configurations, 417 Coconut oil, 32-33, 33t, 411 561
Subject Index 803
[Coconut oil] [Confectionery fats]

acid, 4 shea oil
botanical name, 20 fatty acid composition, 407
component acids, 33 fractionated from solvent, 406
crystallization, 312 in supercoatings, 408
crystallization rate, 312 tailoring of, 408
fatty acid composition, 23 triacylglycerol composition, 402t
fractionation, 217, 217t, 275 fatty acids, 40It, 402t
production of, 22 composition, 407
rancidity, off-flavor compounds, 422t form V, chocolate crystallized in, 403-404
saponification, iodine values, 27 fractions of, 405-406
tocopherols, vitamin E, 30 glycerides in, 401, 403
triacylglycerols, 33t grades of, 405-406
triglycerides, 411 nucléation, phospholipids and, 404
Coconut/palm oil, interesterifled, crystallization stabilized solid content, ternary mixture
rate, 312 POP-POSt-StOSt, 404, 405
Cocos nucífera (see Coconut seed oil) triacylglycerol composition, country of
Cod liver oil, tocopherols, vitamin E, 30 origin, 402t
Coffee whiteners, creamers, spray drying, 492-493 triglyceride composition, country-by-country
Colza oil, botanical name, 20 listing, 403t
Compressed gas deoiling, phospholipid variability of, 404-405
processing, 56 coconut oil, 411
Conditioning shampoo, formulation, 705-706, triglycerides, 411
706t cooling curves, empirical crystallization,
Confectionery fats, 391, 400t, 400-416,419-422, 415-416
425-427 dilatometry, 415
Babassu oil, 411 differential scanning calorimetry, 415
biscuit cream fillings, 425 enzyme interesteriflcation by, 258t
butterfat, 400 ice cream couvertures, 425
center filling fats, 426 lauric stearins, 41 It, 411-413
chocolate fat types, 400t properties of, 412t
coatings, 467-468 liquid oils, 414, 414-415
cocoa butter, 401-406 hazelnut oil, 414
chemical composition, 401 peanut oil, 414
crystallization, phospholipids, 404 melting, 415-416
equivalents, 401-404t, 406, 406-408, 407t, milk fat, 398, 408-411, 410, 410t
409t bloom formation, 409
aceituno acid, fatty acid composition, 407 coatings using, 409t
Chinese vegetable tallow, 407 cost of, 410
fatty acid coruposition, 407 fat content-temperature relationships, 409
Dunkwa allannlackia, fatty acid fractions, 410-411
composition, 407 melting fractions, 410t
enzymatic interesterification, 407-408 texture, 409
fatty acid composition, 407t milk fat alternatives, 400
illipe butter, fatty acid composition, 407 moisture barriers, fats as, 426-427
kokum tallow, fatty acid composition, 407 nuclear magnetic imaging, particularly pulse,
malabar tallow, fatty acid composition, 407 415
mowrah tallow, fatty acid composition, 407 nuclear magnetic resonance, wide-line, 415
palm oil Palm oil, 215, 406-411 (see Fractionation)
fatty acid composition, 407 palmkemel oil, 41 l-412t, 411-413, 412
fractionated from solvent, 406 triglycerides, 411
quality control, 406-407 partially hydrogenated oils, fats, fractions of,
sal 413-414,413-414t
fatty acid composition, 407 trans fatty acids, 413
microbial attack on, 407 phospholipids, 64-65, 65, 66t
804 Subject Index
[Confectionery fats] Cows, lipids in feed, 571, 571-574, 573

polymorphic behavior, 400 Crambe abyssinica (see Sea kale oil)
rancidity, 419-421 Crambe oil, 20t, 24, 38-39, 39t
coconut oil, off-flavor compounds, 422t Cream
ketonic, 421-422, 421-422t pasteurization, 355
lipolytic, 421-422,421-422t preparation, in buttermaking, 306
oxidative, 419-421, 420 sterilization, 355
short-chain free fatty acids, flavor and, 42It Cream alternatives, 355-368
toffee fats, 425 composition of, 356-364
trans hardened fats creaming rate, dilute system particles, 356
with cocoa butter, phase behavior, 413-414 current market, 367-368
fatty acid composition, 413t formulations, 356t
solid fat content, 414t future products, 368
tucum oil, 411 liquids, 356t, 356-359, 358-359
X-ray diffraction, 415 processing, 364-367, 365-366
Conjugated dienoic acids, frying oil quality properties, 356-364
measurement, 445 sterilized products, 364
Conjugated unsaturated acids, 7t whipping creams, 358, 359-364, 359-364
Cooking, before extraction, 115-116 Creaming rate, dilute system particles, 356
horizontal cookers, 116 Cream liquor powder, spray drying, 493
stack cookers, 115-116 Crépis aurea, vemolic acid, 760
Copper, oxidation, effects on, level, 186 Crépis biennis, vemolic acid, 760
Copra oil Critical micelle concentration, nonionic
geographic area of production, 22 surfactants, 66It, 661-662
preparation of, 114 Crude oil
Coronario acid, 760t handling of, 194
Com germ, preparation of, 114 transport, 190-192
Com oil, 21, 33 air contact, 191
botanical name, 20 contamination, 191
in chocolate, content determination, 423-424 heating, 191
fatty acid composition, 23, 53 metal contact, 191
phospholipids, 52 temperature, 191
production of, 22 Cmmb process, chocolate formulation, 394
saponification, iodine values, 27 Cmshing, extraction by, 120-124
tocopherols, vitamin E, 30 Crystallization, 415-416
vitamin E, 29 chocolate, 415-416
Coronario acids, in seed oils, epoxidized oils, 760t cocoa butter, 404
Corylus aveliana (see Hazelnut oil) phospholipids, 404
Cosmetics, phospholipids, 66-67 cooling curves, 415-416
liposomes, 67 empirical crystallization, 415-416
pro-liposome principle, 67 dilatometry, 415
Costs (see Economics) differential scanning calorimetry, 415
Cottonseed oil, 4, 22t, 33-34 during chocolate enrobing, 397-400, 399
as biofuel, 780-781 emulsifers, 523t
botanical name, 20 fractionation and, 200, 205-206, 207-208, 208
fatty acid composition, 23 crystallizer vessels, dry fractionation, 208
geographic area of production, 22 margarine, 312t, 312-313, 314-316
phospholipid composition of, 28 nuclear magnetic imaging, particularly pulse,
preparation of, 114 415
production of, 22 phospholipids, cocoa butter, 404
saponification, iodine values, 27 soybean oil, 203
tocopherols, vitamin E, 30 vanaspati, 382
vitamin E, 29 Crystallizer vessels, dry fractionation, 208
Council of Europe Regulations, 309 Crystal size, chocolate, 398-399
Countercurrent fractionation, 212-214,213 Cucurbitafoetidissima (see Buffalo gourd)
Subject Index 805
Cucurbita maxima {see Pumpkin seed oil) [Detergents]

Cuphea seed oil, 4 beer tank cleaner, 605t
botanical name, 20 biocidal cleaners, 603
Cuphea species {see Cuphea seed oil) clothes washing, 595-597, 596t
Curing, defined, 714-715 commercial laundry, 603
Cutins, 16 dishwashing, 597-598, 598t
epoxy acid, 759 machine, 605
Cyclic acids disinfectants, 603
cottonseed oil, 9, 565 dry cleaning, 604
cyclopentene acid, 9 environmental problems, 58It, 606-608
cyclopropane acid, 9 fatty acids, 583-584, 585, 586-590
cyclopropene acid, 9 esters, 585, 586-590
kapok seed oil, 9 fatty alcohols, 584-586, 590-593
malvalic acid, 9 modification, 590, 590-591
sterculia oil, 9 fatty esters, sulfonates, 587-588, 588
sterculic acid, 9 food processing cleaners, 605, 605t
cyclopropenoid fatty acids, 565 frequent wash formulations, 600-601, 601
future trends, 58It, 606-608
Dairy fat, in ice cream, 331 general-purpose hard surface cleaners, 598,
Decanoic acid {see Capric acid) 599t
Deep frying oils, 442-444, 442-444 hand cleaners, 605-606, 606t
Degumming hard surface cleaners, 597-599, 599t, 604,
intensive, 158-161,159-161 604t
refining process, 139-140 household detergents, 593-599
superdegumming process, 161 household disinfectants, 598-599
total degumming process, 159, 160 industrial, 602-606
Dehulling, before extraction, 114 institutional, 602-606
Dehydropolymerization, reaction of, surface isethionates, 588-589, 589
coatings, 717 liquid soap, 606t
Densipolic acid, 10 metal cleaners, 605
Deodorization, refining process, 150-158 "mild" surfactants, 600-601, 601
batch deodorizer, 157-158 personal washing, 593-595
chelation, 154 raw materials, 582, 582-586
cooling, 154 sanitizers, 598-599
deaeration, 152 sarcosinates, 588-589, 589
designs, for deodorizer, 157 shampoos, 599-602
equipment, 154-158 conditioning, 601-602
fully continuous deodorizer, 155-157, formulations, 599-600, 600t
157 shower gels, 602
flow diagram, 157 soaps, 586t, 586-587
heat recovery, 153 sulfated glycerides, 589-590
high temperature heating, 153 sulfation, 591-593, 592
polish filtration, 154 taurates, 588-589,589
preheating, 152-153 textile cleaning, 603
process variables, 151-152 transport cleaners, 605
semicontinuous deodorizer, 155,156 western Europe, soap production in, 58It
flow diagram, 156 Deterioration, oils {see Stability)
theoretical treatments, 151 Detoxification, extraction and, 116-117
de-oiling, 85-86, 86t D-glucose, nonionic surfactants, 644
Depression, lipids and, 107 DHA {see Docosahexaenoic acid)
Detergency, nonionic surfactants, 677 Diacetyltartaric acid esters, food emulsifiers, 526,
Detergents {see also Surfactants) 527-528
anionic, 579-608, 580, 58It, 584, 586t, 596t Dialkyl/dihexadecylmalonate, fat substitute, 507
a-sulfonated acids, esters, 587-588, 588 Dilatometry, melting, crystallization of fats, 415
bath additives, 602 Diazemuls, lipid emulsion, 540
806 Subject Index
Dielectric constant, frying oil quality [Drying]

measurement, 445 coffee whiteners, creamers, 492-493
Diglycerides, 12 cream liquor powder, 493
Diglycerides, lipid coatings, 458 fat-filled powders, 491-492
Dihomo-gamma-linolenic acid, structure, double ice cream powder, 493
bond position, 6 whole milk powder, high free fat content,
Dilute alcohol refining process, 142, 143-144,144 492
Dimethyloxazoline derivatives, mass spectra of, principles, 483-490
epoxidized oil, 762 drying mechanisms, 489-490
Diprivan, lipid emulsion, 540 spray dryer chamber designs, 485-489,
Directed interesterification, 237-239, 239 486-488
Dishwashing detergents, anionic, 597-598, 598t safety issues, 497-498, 497-499
machine, 605 selection, 493-496, 493-497, 496t
Disinfectants, anionic, 603 surface coatings, 720-721
household, 598-599 dryers, 720
Dispersion, nonionic surfactants, 673-677, 676, fatty acid moieties, 718
677t mechanisms, 718-720, 719
Diynes, applications of, undecynoic acid—time Dry skin, transepidermal water loss, 697
line, castor oil, 793t Durbetaki reaction, 761
Docosahexaenoic acid, 2, 3
structure, double bond position, 6 "Ecologically compatible," usage of term,
Docosanoic acid {see Behenic acid) lubricants, 739
Docosapentaenoic acid, structure, double bond Economics of fractionation, 214, 214t
position, 6 Economics of storage, handling of oil, 170-174
Docosenoic acid, 3, 5 capital costs, 171-172
Dodecanoic acid {see Laurie acid) investment, 171t, 171-172
Dososemoic acid {see Erucic acid) working capital, 172
Crambe abyssinica, 6 degradation costs, 172-173
mustard seed oil, 6 operating costs, 172
Dough-strengthening effects, emulsifiers, 532, 532 total costs, 173t, 173-174
Draining, in buttermaking, 306-307 Effluent treatment, vanaspati production, 38It,
Driers, auxiliary, metal, surface coatings, 720 381-382
Drilling fluids, oil-based, cationic surfactant, 625 Egg
Drug deliveiy fatty acid composition, 53
intravenous lipids, 537-544 phosphatidylcholines, molecular species, 58t
lipid emulsions for, 535-556 phospholipids, 28, 52
design, 539, 540, 54It processing, 54
first generation injectable, 538, 539-542, Eicosanoic acid {see Arachidic acid)
541t, 542 Eicosapentaenoic acid, 2, 3
manufacture, 539, 540, 541t structure, double bond position, 6
second generation injectable, 542t, 542-543, Eicosenoic acid, 3, 5
543 erucic acid, 5
testing, manufacture, 539, 540, 54It Elaidic acid, 5
Dry cleaning detergents, anionic, 604 Elaies guinensis {see Palm oil)
Dry fractionation, multistage, 212 Eleostearic acid, 7
Drying Eltanolone, lipid emulsion, 542
defined, 714-715 Emulsification, nonionic surfactants, 669-673,
before extraction, 115-116 670-672t, 674-676t
refining process, 143 Emulsifiers
spray, fat-containing foodstuffs, 481-500, 482 food, 521-522, 521-534
spray-dried powders, high fat, 487, 490-491, destabilization, 530, 530-531
493 dough-strengthening effects, 532, 532
spray drying fatty acid esters, 528-529
high fat formulations, 491-493 polyglycerol esters, 528
butter powder, 492 propylene glycol, 528
Subject Index 807
[Emulsifers] Epoxidized oils, 759-769

sodium stearoyl-lactylate, 529 absorption bands, 761
sorbitan monostearate, 528-529 alkenes, used as substrates, 763
sorbitan tristearate, 529, 529 alkyl oleates, 759
sucrose esters, 529 chemical reactions, 766-767
interactions, 520-532 cis epoxy esters, chemical shifts, 766t
intravenous nutrition, 585-586 commercial, 763
mono-diglycerides, 522, 522-526 composition of, NMR, 762
distilled monoglycerides, 522-523 coronarie acids, in seed oils, natural occurrence
crystallization point, 523t of, 760t
melting point, 523t cutin, epoxy acid, 759
transition temperature, 523t degree of substitution, 762
physical properties, 523t, 523-526, 524-525 dimethyloxazoline derivatives, mass spectra of,
monoglycerides, organic acid esters of, 526, 762
526-528 Durbetaki reaction, 761
acetic acid esters, 526, 527 epoxidation
citric acid esters, 528 cis addition process, 763
diacetyltartaric acid esters, 526, 527-528 oils used for, 765t
lactic acid esters, 526, 527 process, 762-764
stability, 530, 530-531 on technical scale, 764-765, 764-765t
starch components, interactions with, 531, epoxide function, wide range of reactions,
531-532 766-767
ice cream, 335-336 epoxides, as reactive compounds, 759
margarine, 313-318, 317-318 examples, epoxidation, 764-765
Emulsion, microemulsions, distinguished, laboratory scale, epoxidation, 765, 765t
for drug deliveiy, 585-586 linseed oil, 759
nonionic surfactants, 669 natural epoxy acids, 759-761, 760-76It
Encapsulation, of moisture-sensitive ingredients, oleic acid, 759
lipid coatings, 470-471 peroxyacetic acid
Engine oils from acetic acid, hydrogen peroxide, 764
lubricants, 754-756, 755 distilled aqueous, 764
marine, outboard, two-stroke, lubricants, rapidly preformed, 764
biodegradable, 739 from Proxilate, higher grades, 764
Engine testing, biofuels, 775-782 peroxy acid, 759
Enrobing, chocolate, crystallization during, high concentrations of, avoiding, 762
397-400, 399 phytoaxel ins, epoxy acid, 759
"Environmentally friendly," usage of term, Proxilate, 764
lubricants, 739 qualitative, quantitative analysis, 761-762
"Environmentally harmless," usage of term, rapeseed oil, 759
lubricants, 739 from reaction between alkene, peroxy acid, 763
Enzymatic interesterification, 245-259 reaction rate, 762
applications of, 248, 256-259,257, 258t solvent, influence, 763
cocoa butter, compositions, produced by, reactivity, peroxy acids, 763
258t source of, 759
confectionary fat, compositions, produced soybean oil, 759
by, 258t with peroxyformic acid, peroxyacetic acid
feedstream water content, effect of, prepared in situ, 764t
255t spectroscopic properties, 165-166, 766t
lipase catalysts, 246-251, 247-250 stereochemical configuration, epoxy acids, 761
overview, 224, 245 tall oil, 759
reaction systems, 249-250, 251-256, 252-254, toxic antimicrobials, epoxy acid, 759
254-255t, 256 trivemolin, 760
triacylglycerols, from high oleate sunflower unsaturated oils, acids, methyl esters, epoxides
oil, stearic acid and, 258t obtained from, 765t
EPA {see Eicosapenth noic acid) vemolic acid, 759
808 Subject Index
[Epoxidized oils] [Extraction]

in seed oils, natural occurrence of, 760t cooking by extmsion, 118
in Vernonia anthelmintica, 759 before cmshing, 118, 779
Vernonia galamensis oil, trivemolin in, 760 oilseed extmder, adjustable-jaw choke, 779
Epoxy acids, 10, 759-761, 760-76It slotted-wall expanders, 120,120
Epoxy resins, manufacture of, cationic surfactant, drainage cage, 120
626 before solvent extraction, 118-120
Erlangea tomentosa, vemolic acid, 760 mechanical, 120-124
Erucic acid, 3, 5, 565 background, 120-121, 727
Essential fatty acids, 86-87 (see Lipid metabolism) choking mechanism, 123
Esteramines, cationic surfactant, 612-613 cooling press, 123
Ester-based lubricants, 751-756 full presses, vs. prepresses, 121- 122,122
Ester-based machine tool fluids, 753-754, 754, screw press, 121
754-755t slotted barrels, 123-124,124
Esterified propoxylated glycerol, fat substitute, drainage cage, 124
507-508 solids, removing from oil, 124
Ester interchange, interesterification, 224 wormshaft, 121, 122-123
Esters, 16, 528-529 natural sources, 113-135
polyglycerol esters, 528 olives, 125
propylene glycol, 528 palm fruit, vs. palm kernels, 125
sodium stearoyl-lactylate, 529 preparation, 113-120, 114t
sorbitan monostearate, 528-529 cleaning, 113
sorbitan tristearate, 529, 529 cooking, 115-116
sucrose esters, 529 horizontal cookers, 116
synthetic, lubricants, 746, 746-747, 747-748t stack cookers, 115-116
Ester waxes, 16 dehulling, 114
Etanolone, lipid emulsion, 542 detoxification, 116-117
Ethanolamine derivatives, 16 drying, 115-116
Etheramines, cationic surfactant, 613 flaking, 114-115
Ether carboxylates, nonionic surfactants, 683 solvent extraction, 127-133
Ether lipids, 15-16 alternative solvents, 127-128
ethanolamine derivatives, 16 traditional solvents, 127
platelet-activating factor, 16 solvent recovery, 132-133
Ethylene oxide Extractors, 128-132
castor oil, reactions, 789 background, 128-129
nonionic surfactants, 641-643, 643 circular basket extractor, 131, 737
Euphorbia lagascae (see Vemolic acid) Rotocel, 131
seed oil, composition of, 76It perforated belt extractor, 129, 729
vemolic acid, 760 rectangular loop extractor, 130, 730
Euphorbia lathyris (see Spurge seed oil) sliding-bed extractor, 132, 732
Europe, Council of Regulations, definitions, 309 Extmsion process, 117-120
Evening primrose oil, 34-35, 34-35t background, 777, 117-118
botanical name, 20 cooking by extmsion, 118
Extraction, 113-135 before crushing, 118, 779
animal materials, 126-127 oilseed extruder, adjustable-jaw choke, 779
crushing, 120-124 slotted-wall expanders, 120,120
extractors, 128-132 drainage cage, 720
background, 128-129 before solvent extraction, 118-120
circular basket extractor, 131,131
Rotocel, 131 Fabric conditioners, cationic surfactant, 620-621,
perforated belt extractor, 129,129 621
rectangular loop extractor, 130,130 Fabry’s disease, sphingolipids and, 17
sliding-bed extractor, 132,132 Facial oil, moisturizing, formulation, 701
extrusion, 117-120 Fagus sylvatica (see Beechnut oil)
background, 777. 117-118 Farber’s disease, sphingolipids and, 17
Subject Index 809
Fat bloom, chocolate, 416-419 [Fatty acids]

Fat-filled powders, spray drying, 491-492 linoleic acid, 6
Fat mimetics, 503 molecular structure, 3
Fat-soluble vitamins, 94-95 monoene acids, 5-6
vitamin A, 94 nomenclature, 1
vitamin D, 94-95 nonionic surfactants, 644-645
vitamin E, 95 number of, 3
Fat substitutes, 502, 507-509 numbers, use of to designate, 2
carboxyl/carboxyiate esters, 507 nutrition, 79-112
dialkyl/dihexadecylmalonate, 507 octadecatetraenoic acids, with conjugated
esterified propoxylated glycerol, 507-508 unsaturation, 7t
olestra, 508-509 octadecatrienoic acids, with conjugated
polyorganosiloxanes, 508 unsaturation, 7t
sorbestrin, 508 olefrnic compounds, 3
trialkoxytricarballyate, 509 oxygenated acids, 9-10
Fatty acids epoxy acids, 10
acetylenic acids, 7t, 7-8 furanoid acids, 10
allenic acids, 7t, 7-8 hydroxy acids, 9-10
animal feeds, lipids in, 557-577 paints, 711-736
anionic detergents, 579-608, 583-585 palmitic acid, 4-5
baking fats, 323-326, 386-387 personal care products, 695-709
biofuels, 771-785 phospholipids, 28t, 51-77
branched-chain acids, 8 polyene acids, 7t, 7-8
butter, 305-307 polyunsaturated acids, 3
castor oil products, 787-795 reaction rates, with oxygen, 436t
chocolate, 391-400, 415-425 refining process, 137-167
classes, in food natural fats, 82 salad oils, 433-441, 447-448
coatings, edible, 453-479 saturated acids, 4t, 4-5
composition, 28t, 53t sources of lipids, 19-50
oils, fats, 23t, 23-24, 24t, 25t spray processing, fat-containing foodstuffs,
confectionery fats, 391,400-416,419-422, 481-500
425-427 spreads
cream alternatives, 355-368 baking fats, 323-326
cyclic acids, 9 dairy, 322, 387-389
epoxidized oils, 759-769 liquid margarine, 323
extraction of oil-in-water, 323
from natural sources, 113-135 reduced fat, 320-322
natural sources, 113-135 stearic acid, 4-5
film barriers, 453-479 cocoa butter, 5
food emulsifiers, 521-534 storage, 169-198
fractionation, 199-222 of oil, 188-197
frying oils, 433-440,441-448 structure, 1- 10,5
ghee, 370-373 surface coatings, 711-736
handling, 169-198 surfactants
of oil, 169-188 cationic, 609-631
hydrogenation, triglyceride oils, 265-303 nonionic, 633-694
ice cream, 329-354 with trans olefrnic unsaturation, 8
interesterification, 223-263 transport, 169-198
lipid emulsions, intravenous nutrition, drug of oil, 188-197
delivery, 535-556 vanaspati, 373-385
low calorie fats, 501-520 Fatty alcohols, nonionic surfactants,
lubricants, 737-758 645-647
margarine, 308-320, 387 anionic detergents, 584-586, 590-593
methylene-interrupted polyene acids, 6t, 6-7 Feeds, animal, 557-577
gamma-1inolenic acid, 6 nonruminants, 558-567
810 Subject Index
[Feeds] Food and Drug Administration, film, coating

adipose tissue fatty acids, rate of change, approvals, 458t-459t, 464t
566-567, 567 Food emulsions, phospholipids, 64
coconut-palmkemel oil blend, chemical Fractionation, 199-222, 225
analyses, 56It applications, 215-220
contaminants, 564-565 beef tallow, 219, 219t
cyclopropenoid fatty acids, 565 coconut oil, 217, 217t, 278
digestibility, chemical structure and, 558-565 confectionery fat, 406
erucic acid, 565 countercurrent fractionation, 212-214,213
fat, oil utilization ciystallization, 200, 205-206, 207-208,208
dietaiy energy value, chemical structure ciystallizer vessels, dry fractionation, 208
and, 558-565 economics of, 214, 214t
free fatty acid content, 559-562, 559-562, fatty acids, 199-222
560-56U growth, 205-206,206
heat damage, 560t, 562-564, 564t interesterification, 225
meat quality, effect on, 565-567 milk fat, 218-219,278-279
pesticides, 564-565 three-step fractionation, melting properties,
physical texture, effect on, 565-566 218
polythene, 565 multistage diy fractionation, 212
saturation, 559-562, 559-562, 560-56It nucléation, 204t, 204-205,205
sunflower acid oil, chemical analyses, 564t olein, solid fat contents of, 278
phospholipids, 62-63 overview, 199-200,200
ruminants, 568-574 palm oil, 215, 215-217, 216t, 217t, 278
absorption, 568t, 568-570 dry multiple fractionation, 275
beef cattle, 570-571 midfractions, from diy, solvent fractionation,
commercial setting, use of fats in, 570-574 216t
cows, 571, 571-574, 573 oleins, from dry fractionation, properties of,
milk, 574t 216t
digestion of fats, 568t, 568-570 processes compared, 217t
energy, 568t solid fat contents of, 278
fermentation, 568t, 568-570 phase behavior, solubility and, 200-203,
protected fats, 569-570, 570t 201-202
sheep, 570, 570t centrifugal separation, 209-210, 210t
Feedstream water content, effect of, enzymatic pressing, 210-212,277
interesterification, 255t triglycerides, binary mixture of, 201
Fickian diffusion, lipid coatings, 455 tripalmitin
Film barriers in 2-oleodipalmitin, solubility of,
acetylated monoglycerides, 460 202
barrier properties of, 457-461, 461-462t solubility of, 201
cocoa butter, 460 tristearin
Code of Federal Regulations, 456 solubility of, 201
Fickian diffusion, 455 in triolein, phase diagram, 202
lipid, 453-479 vacuum filtration, 209,209, 210t
materials, 456-457, 458-460t principles of, 200-206
mass transport in, 454, 454-455 sal fat, 220, 407
Fish coatings, lipid, 469-470 separation, 208-212
Fish oils, 45-47, 47t vacuum filtration
component acids, 47t rotary drum filter, 209
fatty acid composition, 23 vacuum belt filter, 209
hydrogenation, 285-290,286-289 shea fat, 220
production of, 22 solvent fractionation, 212, 406
Flaking, before extraction, 114-115 stages, 200
Flavor, of ice cream, fat and, 341-350 stearin, solid fat contents of, 278
Fluosol-DA, lipid emulsion, 541 supersaturation, 203, 203-204
Foaming, nonionic surfactants, 662, 662t crystallization, soybean oil, 203
Subject Index 811
[Fractionation] Ghee, 370-373

triglyceride crystals, supercooling, solubility, attributes of, 370
variations, 204t batch production, 372-373
vacuum filtration, vacuum belt filter, 209 composition, 370-371
winterization, 220 continuous ghee making, 373, 374-375
Freezing, ice cream, 338-341 nature of, 370
aeration, 339-340, 340 organized sector, preparation in, 372-373
dynamic, 339-340,340 preparation, 371-373
overrun principles of, 371-372
calculation of, 339 standards for, 371
defined, 339 traditional preparation, 370
static hardening, 340-341 large scale, 372
Friedreich's ataxia, phospholipids and, 71 GLA {see Gamma-linoleinic acid)
Fritest, frying oil quality measurement, 445 Glycerides, sulfated, anionic detergents, 589-590
Fruit coatings, lipids, 463-465,464t Glycerol, 11
Flying oils, 433-448 Glycerolates, interesterification, 231
additives, 436-437 Glycerol esters, nonionic surfactants, 648-649
antioxidants, 436-437 3-glycerophosphoric acid, 14
metal chelating agents, 436 Glycine max (see Soybean oil)
nitrogen treatment, 437, 437 Glycolipids, 93
deep flying, 442-444, 442-444 Glycosyldiacylglycerols, 13
deterioration of, 433-435, 434-435, 436t Gold of pleasure (see also Cameline seed oil)
flavor, 435-436 botanical name, 20
new oilseed varieties, 447-448 fatty acid composition, 25
partial hydrogenation, 441 Gossypium species (see Cottonseed oil)
production, country-by-country listing, 434t Granularity, vanaspati production, 382
quality measurement, 444-445 Grapeseed oil
18:2/16:0 ratio, 445 botanical name, 20
alkaline contaminant materials, 446-447 fatty acid composition, 25
conjugated dienoic acids, 445 Groundnut oil, 1, 5, 35, 35t (see also Peanut oil)
dielectric constant, 445 botanical name, 20
fatty analyses, 445 fatty acid composition, 23
Fritest, 445 geographic area of production, 22
Oxifrit-Test, 445 phospholipid composition of, 28
polar components, 444-445 production of, 22
quick tests, 445 447 saponification, iodine values, 27
RAU test, 445 tocopherols, vitamin E, 30
spot test, 446 triacylglycerol composition, 35t
reaction rates, with oxygen, 436t Guizotia abyssinica (see Nigerseed oil)
safety issues, 447
stability, evaluation of, 437-439 "Half-fat margarine," regulation of, 309
active oxygen method, 438 "Halverine," regulation of, 309
oil stability index, 438 Hand cleaners, anionic, 605-606, 606t
Schaal oven test, 438 Handling of oil
sensory methods, 439-440, 440 costs, 170-174
salad oils, 439-440 capital costs, 171-172
shelf storage test, 437 investment, 171t, 171-172
volatiles analyses, 438-439 working capital, 172
trans fatty acids, 441-442 degradation costs, 172-173
Furanoid acids, 10 operating costs, 172
urofuranic acids, 10 total costs, 173t, 173-174
crude oil, 194
Ganglioside, 17 hydrolysis, 174
Gaucher disease, sphingolipids and, analyses, 174
17 diffusivity, 174-175, 175t
812 Subject Index
[Handling of oil] Hempseed oil

water in oil, 175t botanical name, 20
factors affecting, 175-178 fatty acid composition, 25
free fatty acid, 175 Heptane, water, partition data, cationic surfactants,
solubility, 176t 619t
moisture, 175-176, 176t Herring oil, tocopherols, vitamin E, 30
oil type, 177 Hevea bmsiliensis {see Rubberseed oil)
overall rate, 177-178, 778 Hexadecanoic acid {see Palmitic acid)
palm oil, saturated with water, 178 Hexanoic acid {see Caproic acid)
temperature, 174-175, 177t Hexitol esters, nonionic surfactants, 649-653,
solubility, 174-175, 175t 650-652
water in oil, 175t Hibiscus esculentus {see Okra seed oil)
hydroperoxide breakdown, 187-188 Homogenization, ice cream mix manufacture,
hydroperoxide formation, 184-187 337-338, 338t
antioxidants, 187 fat globule size, 338t
light, 186 Honesty seed oil, 6
metals, 186t, 186-187 Household detergents, anionic, 593-599
effect on oxidation, 186t Household disinfectants, anionic, 598-599
moisture, 187 Huile de carapathes {see Castor oil)
oxygen concentration, 186 Huntington's disease, phospholipids and, 71
temperature, 185, 185t Hydraulic oil lubricants, 752, 752
type of oil, 184t, 184-185,185 for mobile equipment, 752
oxidation, 178-188 Hydrogen, generation of, vanaspati production,
absorption of oxygen, 181-184 380
dissolved oxygen content, 181-182, 182t, Hydrogenation
190t canola oil, 284,285
gas/headspace, oxygen concentration in, castor oil, 789
183-184 catalyst
headspace volume, 182, 182t nickel, 285, 288, 295-300
oil saturated with air, gas content, 182t particles, reaction rate within, 214-216, 276
rate, 182 performance, examples, 295-299, 296t,
surface area-to-volume ratio, 182-183, 183t 297-298, 299t
analyses, 179-180 examples, 283, 283
bleached soybean oil values, 180 fish oils, 285-290,286-289
oxygen content, 180 H2 transfer
peroxide value, 179 from gas to catalyst phase, 266-267
totox value, 179-180,180 from gas to liquid phase, rate of, 267-273
anisidine value, 179 effect of reactor design, 272, 273
diffusivity, 180-181, 181t interfacial area, 271-272
oxygen in oil, 181t mass transfer coefficient, 268-270,270t
solubility, 180-181, 181t bubble sizes, 270t
nitrogen, 18 It maximum achievable rate constant, 272-273
oxygen, 18 It from liquid phase to catalyst particles
refined oil, 197 characteristic rate constant, 273-274
storage requirements, 188-197, 190t rate, 273-274
transport requirements, 188-197, 190t interesterification, 225
Hard surface cleaners, anionic, 597-599, 599t, 604, iodine value and, 265-303
604t overview, 265-266, 300-302
Hazelnut oil, 414 phospholipids, 60-61, 6 '
botanical name, 20 effect on product propv rties, 299t, 299-300
fatty acid composition, 25 product properties after, 290-300
Heat damage, animal feed, 560t, 562-564, modification of oils, 290-295, 295
564t rapeseed oil, 299t
Helianthus annus {see Sunflower seed oil) rate of
Helichrysum bracteatum, coronarie acid, 760 equation, 279-283,280-281
Subject Index 813
[Hydrogenation] [Ice cream]

factors affecting, 276-279 formulations, 330t, 330-331
initial saturation rate of oil, 276-277, 277 ingredients, 331-336
poisoning, hydrogenation catalysts, emulsifiers, 335-336
277-279 fat, 331-333
reactivation of catalyst activity, 277 dairy fat, 331
soybean oil, 283-284,284, 296t non-dairy, 331-332, 332
technology, 266-290 substitutes, 332-333
tmns content, reduction of, 300,300-301 nonfat milk ingredients, 333-334
triglyceride oil, 265-303 stabilizers, 334-335
vanaspati production, 376-382 sweeteners, 334
Hydrogen bromide, in acetic acid measure, nontempered chocolate, 399
Durbetaki reaction, epoxidized oil, 761 powder, 493
Hydrolysis processing, 336-341
analyses, 174 freezing, 338-341
diffusivity, 174-175, 175t aeration, 339-340,340
water in oil, 175t dynamic, 339-340,340
factors affecting, 175-178 overrun
free fatty acid, 175 calculation of, 339
moisture, 175-176, 176t defined, 339
oil type, 177 static hardening, 340-341
overall rate, 177-178,178 mix manufacture, 336-338
temperature, 174-175, 177t aging, 338
palm oil, 177t blending, 336-337, 337
free fatty acid, saturation solubility and, 176t homogenization, 337-338, 338t
palm oil, saturated with water, 178 fat globule size and, 338t
solubility, 174-175, 175t pasteurization, 336-337, 337
water in oil, 175t structure
storage, transport of oil, 174 defects, 349-350
analyses, 174 fat and, 341-343, 341-349, 342t, 344t,
diffusivity, 174-175, 175t 345-348
water in oil, 175t Ice cream couvertures, 425
Hydroperoxides, 184-188 Ice cream powder, spray drying, 493
factors affecting, 175-178 Illipe butter, 5
free fatty acid, 175 botanical name, 20
moisture, 175-176, 176t fatty acid composition, 25
oil type, 177 Imagent, lipid emulsion, 542, 543
overall rate, 177-178,178 Imidazolines, cationic surfactant, 612
temperature, 174-175, 177t Impatiens balsamina {see Parinaric acid)
palm oil, 177l India, usage of fats in, 369-370
free fatty acid, saturation solubility and, 176t bakery fats, 386
palm oil, saturated with water, 178 fat spreads, 387-389
solubility, 174-175, 175t fat use, current patterns of, 369-370
water in oil, 175t ghee, 370-373
Hydroperoxides, formation of, surface coatings, historical development, 369
718 margarines, 387-389
Hydrophilic-lipophilic balance, in personal care patterns of, 369-370
products, 699, 699t vanaspati, 373-385
Hydroxides, interesterification, 231 Indian mustardseed oil, botanical name,
Hydroxy acids, 9, 10 20
Hydroxylation, phospholipids, 60 Industrial detergents, anionic, 602-606
Inflammatory disease, lipids and, 106-107
Ice cream, 329-354, 330t, 344t Instant food products, phospholipids, 66
emulsifers, 325-336 Institutional detergents, anionic, 602-606
flavor, fat and, 341-350 Interesterification, 223-263,224
814 Subject Index
[Interesterification] Intravenous nutrition, 544-552

alkali-catalyzed interesterification, 231-237 lipid emulsions for, 535, 544-552
catalyst activation, 234 lodinated oil, lipid emulsion, 542
described, 230, 231-233,252-255 Iodine values, 26-27, 27t
intramolecular interesterification, 235 hydrogenation, 265-303
mechanism, 233-237,234, 234-237 linseed oil, 760, 761
in animal fats, 225 triglyceride oil hydrogenation and, 265-303
chain lengths, fatty acids, variation, 227 vegetable oils, unsaponifiable material in, 27t
chemical background, 224-228, 226-228t, Iron, oxidation, effects on, level, 186
227-228 Isethionates, anionic detergents, 588-589,589
defined, 224 Iso acids, 8
directed, 237-239,239 Isocyanates, castor oil, reactions, 789
enzymatic, 245-259 Isoricinoleic acid, 10
applications of, 248, 256-259,257, 258t
cocoa butter, compositions, produced by, Jacaranda mimosifolia (see Jacaric acid)
258t Jacaric acid, 7
confectionaiy fat, compositions, produced Jatropha curcas (see Jatropha seed oil)
by, 258t Jatropha seed oil, botanical name, 20
feedstream water content, effect of, 255t Jojoba oil, 6
lipase catalysts, 246-251,247-250 meadowfoam oil, 6
overview, 224, 245 Juglans regia (see Walnut oil)
reaction systems, 249-250, 251-256,
252-254, 254-255t, 256 Kapok seed oil
triacylglycerols, from high oleate sunflower botanical name, 20
oil and stearic acid, 258t fatty acid composition, 25
fatty acids, 223-263 Karanja oil, as biofuel, 781
fractionation, 225 Kneading, in buttermaking, 307
hydrogenation, 225 Kokum butter, 407
lard, after random, directed interesterification, Krabbe disease, sphingolipids and, 17
triacylglycerol composition, 226t Kusum seed, 5
in naturally occurring oils, fats, 225 Kynacyte, lipid emulsion, 542, 543
palm oil, before, after, triacylglycerol
composition, 226 Labeling, low calorie fats, 515-516
product composition, 224-228, 226-228t, Lacquer, defined, 714-715
227-228 Lactic acid esters, food emulsifiers, 526, 527
random, 239-245 Lanolin cream, formulation, 702, 703t
control methods, 244-245 Lard, 4,45,46t
feedstock requirements, 239-240, 240 after random, directed interesterification,
posttreatment, 243-244 triacylglycerol composition, 226t
process options, 241 crystallization rate, 312
batch processes, 230, 241 fatty acids, 23, 46t
continuous processes, 241-243,242-243 tocopherols, vitamin E, 30
process yield, 243-244 triacylglycerols, 46t
random distribution, 224 Laundry detergents, anionic, 603
rearrangement catalysts, 223, 229-230, Laurie acid, 3,4
229-231 Laurie stearins, 41 It, 411-413
triacylglycerols, 223, 227t properties of, 412t
types of reactions Lecithin, 27-28, 6J
acidolysis, 224 alcohol fractionation, 57t
alcoholysis, 224 cholesterol-lowering effect, 70t
interesterification, rearrangement, ester composition of, 54t
interchange, 224 deoiled, composition, 56i
in vegetable oils, 225 granular, 28
Intraiodol, lipid emulsion, 542, 543 liquid, 28
Intramolecular interesterification, 235 soybean, composition of, 54t
Subject Index 815
[Lecithin] [Lipids]
types, phospholipids, 66t emulsions, intravenous nutrition, drug delivery,
usage of term, 51 535-556
Lesquerella densipila {see Densipolic acid) epoxidized oils, 759-769 {see Epoxidized oil)
Lesquerolic acid, 9, 10 evening primrose oil, 34-35, 34-35t
Licania rigida {see Oiticica oil) extraction
Lignoceric acid, 5 from fatty acids, natural sources, 113-135
physical properties of, 4 from natural sources, 113-135
Limnanthes alba {see Meadowfoam oil) fat-soluble vitamins, 94-95
Linola, 36t vitamin A, 94
Linoleic acid, 2, 3, 6, 7 vitamin D, 94-95
structure, double bond position, 6 vitamin E, 95
usage of term, 1 fatty acid classes, in food natural fats, 82
Linolenic acid, 3, 5, 7 film barriers, 453-479
usage of term, 1 fish oils, 45-47, 47t, 285 289
Linseed oil, 1, 36, 36i component acids, 47i
botanical name, 20 component acids of, 47t
epoxidization, 759 food emulsifiers, 521-534
fatty acid composition, 23, 36t fractionation, 199-222 {see Fractionation)
iodine value, 765 flying oils, 433-448
preparation of, 114 gamma linolenic acid, fatty acid compositions,
production of, 22 34t
saponification, iodine values, 27 ghee, 370-373
triacylglycerol composition, 36t glycolipids, 93
Linum usitatissimum {see Linseed oil) groundnut oil, 35, 35t
Lipase catalysts, enzymatic interesterification, triacylglycerol composition, 35t
246-251,247-250 handling, of oil, 169-188
Lipids hydrogenation, triglyceride oils, 265-303
in animal feeds, 557-577 ice cream, 329-354
anionic detergents, 579-608 inflammatory disease and, 106-107
baking fats, 323-326, 386-387 interesterification, 223-263
biofuels, 771-785 iodine value, 26-27
blackcurrant seed oil, 34-35, 34-35t lard, 45, 46t
borage oil, 34 fatty acid content, 46t
butter, 44-45, 45t, 305-307 triacylglycerols, 46t
fatty acids in, 45t interesterification, 226t
triacylglycerols, 45t linola, fatty acid composition, 36t
cancer and, 105-106 linseed oil, 36, 36t
cardiovascular disease and, 103-105, 104t fatty acid composition, 36t
castor oil, 31, 787-795 triacylglycerol composition, 36t
chocolate, 391-400, 415-425 low calorie fats, 501-520
coatings, edible, 453-479 lubricants, 737-758
cocoa butter, 31-32, 32t {see Cocoa butter) margarine, 308-320, 387
oleo glycerol esters, symmetrical metabolism, 69-70, 70t, 83-87
disaturated, 32t biosynthesis, 85
coconut oil, 32-33, 33t {see Coconut oil) breakdown, tissue lipids, 87
component acids, 33 fatty acid oxidation, for cellular energy, 87
triacylglycerols, 33t lipolysis, 87
confectionery fats, 391,400-416, 419-422, peroxidation, 87
425-427 {see Confectionery fats) digestion, 83, 84
com oil, 33 {see Com oil) essential fatty acids, 86-87
cottonseed oil, 22t, 33-34 {see Cottonseed oil) deficiency, 86
Crambe oil, 20, 24, 38-39, 39t metabolism, 86, 86-87
cream alternatives, 355-368 lipoproteins, 84
depression and, 107 transport, 83-85, 84
816 Subject Index
[Lipids] [Lipids]
unsaturated fatty acids, metabolism, 86 liquid margarine, 323
mustard oil, 6, 24, 38-39, 39t margarine, 308-320
nutrition, 79-112 oil-in-water, 323
obesity and, 101-103 reduced fat, 320-322
body fat distribution, 101-102 steroids, 93-94
fat calorie, vs. carbohydrate calorie, weight cholesterol, 93, 511-512
gain and, 102-103 plant steroids, 93-94
olive oil, 36-37 (see Olive oil) sterol composition, 29
paints, 711-736 (see Paints) storage, 80
palmkemel oil, 33t, 37-38, 38t of oil, 188-197
component acids, 33 structure, 10-11, lit, 80
triacylglycerols, 33t cholesterol, 80
palm oil, 33t, 37-38, 38t glycerol, 11
triacylglycerol composition, 38t glycolipids, 80
personal care products, 695-709 phospholipids, 80
phospholipids, 27-28, 51-77, 93 sunflower seed oil
processing, 95-101 stereospecific analysis of, 42t
digested reduced energy fats, 100-101 triacylglycerols, 42t
fractionation, 99 surface coatings, 711-736
medium-chain triacylglycerols, 99 surfactants
reduced cholesterol foods, 99 cationic, 609-631
heating, 96 nonionic, 633-694
hydrogenation, 97-98 sunflower seed oil, 41-42, 42t (see Sunflower
interesterification, 98-99 seed oil)
irradiation, 96 tall oil, fatty acids, 42-43,43t
mixing, 95 tallow, 4, 45, 46t, 78It
reduced fat products, 99-101 fatty acid content, 46t
spreads, 99-100 stereospecific analysis, 46t
undigested fat analogues, 100 triacylglycerols, 46t
randomized, 226t thrombosis and, 105
rapeseed oil, 38-39, 39t (see Rapeseed oil) tocopherols, 29-30
stereospecific analysis, 39t in transit, 80
refining process, 137-167 transport, of oil, 188-197
rice bran oil, 39-40 triacylglycerols, 24-26, 87-93
safflower seed oil, 40 distribution of fatty acids in, 93
stereospecific analysis, 40t metabolic energy and, 88
salad oils, 433-441,447-448 monounsaturated acids, 89
saponification value, 26-27 palatability, 87
sesame oil, 40 polyunsaturated fatty acids, 89-93
sources, 19-50 eicosanoids, production of, 90
botanical name, plant, 20t essential fatty acids
geographical areas of production, 22t, 22-23 requirements for, 91
names, 19-21, 20-2It roles, 89-91, 90
oil, name of, 20t n-3 family, 82, 92-95
production of, 21-22, 22t n-6 family, 86, 91-92
annual, 21-22, 22t saturated fatty acids, 88
soybean oil, 40-41, 41t (see Soybean oil) structural lipids, supply, 88
triacylglycerols, 4It tung oil, 43 (see Tung oil)
stereospecific analysis, 41t unsaponifiable material content, 26-27
spray processing, fat-containing foodstuffs, vanaspati, 373-385
481-500 vemonia oil, 43t
spreads vitamin C, 29-30
baking fats, 323-326 waxes, 30, 44,44t
dairy, 322, 387-389 composition of, 44t
Subject Index 817
[Lipids] [Lubricants]
wheatgerm oil, 30t, 43 biodegradability, 739-740, 740
Lipolytic rancidity, confectionery fats, 421-422, "ecologically compatible," usage of term, 739
421-422t engine oils, 754-756, 755
Lipophilic drugs, 535-544 environmental awards, 743-744
lipid emulsion vehicles for, 537-539, 538, 54It environmental concerns, 738, 738-741
Lipoproteins, 84 legislation, 741-743
Liposomes, in cosmetics, 67 atmosphere, 743, 743
Lipostatic theory, weight gain, 102-103 soil protection, 742-743
Lipstick, formulation, 706-707, 707t water protection, 742
Liquid cream alternatives, 356t, 356-359, 358-359 "environmentally friendly," usage of term, 739
Liquid margarine, 323 "environmentally harmless," usage of term, 739
Liquid oils, 414, 414-415 {see Frying and Salad ester-based, 751-756
oils) ester-based machine tool fluids, 753-754, 754,
hazelnut oil, 414 754-755t
peanut oil, 414 hydraulic oils, 752, 753
Liquid soap, anionic, 606t for mobile equipment, 752
Liver therapy, phospholipids and, 70 low temperature properties, 748-749, 749
Long-chain saturated acids, 5 mobility, 741
arachidic acid, 5 natural fatty oils, 744
behenic acid, 5 chemical modification, 745-746
groundnut oil, 5 improvements, 745, 745
kusum, 5 outboard marine engine oil, two-stroke, rapidly
lignoceric acid, 5 biodegradable, 739
Lophira procera, 5 plant oil, 751-756
Lophira alata, 5 polyalkylene glycols, 747
rambutan tallow, 5 pressure, 749-750, 750t
Long oil, defined, 714-715 rapidly biodegradable lubricants, base oils,
Lophiea procera, 5 744-747
Lophira alata, 5 synthetic esters, 746, 146-141, 747-748t
Low calorie fats, 501-520 {see also Fat substitutes) total-loss lubricants, 751-752, 752t
adjustment to, 514 toxicity testing values, 740-741
blood cholesterol level, 511-512 viscosity index, 751
calorie reduction, from fat, 501-503 Lumbang oil, botanical name, 20
caprenin, 505 Lunaria biennis {see Honesty seed oil)
emulsifiers, 502 Lurie acid, physical properties of, 4
fat mimetics, 503
future products, 516 Macadamia oil, fatty acid composition, 25
industrial significance, 501 Macadamia tetraphylla {see Macadamia)
nutrient intakes, 513-514 Madhuca longifolia {see Illipe butter)
nutrition and, 511-514 Maize oil, 21 {see also Com oil)
regulation of, 514-516 as biofuel, 781
labeling, 515-516 botanical name, 20
safety, 514-515 Malabar kallow, 407
salatrim, 506-507 Manganese, effects on oxidation, 186
triacylglycerols, medium-chain, 505-506 Mangifera indica {see Mango)
utility of, 509-511 Mango oil, fatty acid composition, 25
weight control, 512-513 Margarine, 308-320
Low fat spreads, 320-322, 321-322t crystallization, 312t, 312-313, 314-316
Lubricants, 737-758 rates, 312t
additives, environmentally sound, 747-751 definitions, 309-3 lOt, 309-311
algae toxicity test, 741 descriptions, 309-3 lOt, 309-311
antioxidants, 751 early developments, 308-309
antiwear additives, 749-750, 750t emulsifiers, 313-318 , 317-318
bacteria toxicity test, 741 Europe, Council Regulation, 309
818 Subject Index
[Margarine] [Metabolism of lipids]

fat content, 309t transport, 83-85,84
"half-fat margarine," regulation, 309 unsaturated fatty acids, metabolism, 86
"halverine," regulation, 309 Metal cleaners, anionic, 605
historical overview, 308-309 Metals, interesterification, 229
India, production in, 387-389 Metal chelating agents, 436
emulsification, 388, 389 Methylene-interrupted polyene acids, 6t, 6-7
fat component, 388 arachidonic acid, 7
manufacture, 388-389 gamma-linolenic acid, 6, 7
specifications, 387-388 linoleic acid, 6
liquid, 323 a-linolenic acid, 7
"minarine," regulation, 309 Mead's acid, 7
packet, fat content, 309 Methyl esters
processing, 318-320, 319, 320t biofuels
methods compared, 320t castor oil, 781
raw materials, 311-312 cottonseed oil, 780-781
soft, fat content, 309 karanjaoil, 781
spreadable fats, defined, 310t maize oil, 781
"spreadable fats," regulation, 309 palm oil, 780
structure, 311-312 peanut, 781
Marine engine oil, outboard, two-stroke, rapeseed oil, 779-780
lubricants, rapidly biodegradable, 739 ricebran oil, 781
Mass spectrometry, 962 soybean oil, 778
Mass transport, in lipid coatings, 454, 454-455 sunflower oil, 780
Mayonnaise, 440-441 tallow, 781
Meadowfoam seed oil, 6 vegetable oils, fuel characteristic, 776t
botanical name, 20 epoxides obtained from, 765t
fatty acid composition, 25 sulfonated, nonionic surfactants, 683
Mead's acid, 7 Microemulsions, emulsion, distinguished, nonionic
structure, double bond position, 6 surfactants, 669
Meat coatings, lipid, 469-470 Milk, phospholipids, 52
Medium-chain saturated acid, 4 Milk chocolate {see also Chocolate)
Medium oil, defined, 714-715 contents of, 392
Melting of fats, 415-416 development of, 391
dialtometry, 415 formulation of, 392
differential scanning calorimetry, 415 Milk fat, 398, 408-411, 410, 410t {see Butter)
emulsifers, 523t alternatives to, confectioneiy fats, 400
nuclear magnetic resonance bloom formation, 409
pulse, 415 coatings using, 409t
wide-line, 415 cost of, 410
X-ray diffraction, 415 cow milk, 574t
Melting points, cationic surfactants, 618t fat content-temperature relationships, 409
Memory, phospholipids and, 71 fractionation, 218-219,218-219
Menhaden oil, tocopherols, vitamin E, 30 three-step fractionation, melting properties,
Metabolism of lipids, 83-87 218
biosynthesis, 85 fractions, 410-411
breakdown, tissue lipids, 87 melting fractions, 410t
fatty acid oxidation, for cellular energy, 87 texture, 409
lipolysis, 87 powder, 492
peroxidation, 87 "Minarine," regulation of, 309
digestion, 83,84 Mineral flotation, cationic surfactant, 622t,
essential fatty acids, 86-87 622-623
deficiency, 86 Mixer combinations, multipurpose refining plant,
metabolism, 86, 86-87 142
lipoproteins, 84 Modification of oils, hydrogenation, 290-295, 295
Subject Index 819
Moisture barrier, fats as, 426-427 Night cream, formulation, 703, 704t
Moisture-sensitive ingredients, fat encapsulation, Nitrogen treatment, oils, 437, 437
lipid coatings, 470-471 Niemann-Pick disease, sphingolipids and, 17
Molding, chocolate, equipment, 399 NMR, 762
Molecular structure, fatty acids, 3 Nomenclature, 1
Monoacylglycerol, 12 Nondiying, defined, 714-715
Mono-diglycerides, food emulsifiers, 522, 522-526 Nonionic surfactants
distilled monoglycerides, 522-523 esterification, 648-654
crystallization point, 523t ether sulfonates, 683-684
melting point, 523t industrial end uses, 678-682t, 682
transition temperature, 523t interfacial adsorption, 664-667, 665, 666t
physical properties, 523t, 523-526, 524-525 as intermediates, 682-684
Monoene acids, 5-6 mircormulsions, defined, 669
docosenoic acid, 5 natural/renewable, V5. petrochemical component
eicosenoic acid, 5 in, 64It
hexadecenoic acid, 5 physicochemical properties, 660-677
palmitoleic acid, 5 polymeric surfactants, 676, 684-685
zoomaric acid, 5 production, 648-660
honesty seed oil, 6 country-by-countiy listing, 636t-640t
jojoba oil, 6 raw materials, 640-648, 64It, 642
nervonic acid, 6 solubilization, 663-664, 663-664
octadecenoic acid, 5 surface adsorption, 664-667, 665, 666t
oleic acid, 5 world production, 636-640,636-640t
Monoglycerides {see Mono-Diglycerides) Nonruminants, lipids in feed, 558-567
aa'-diglyceride, 12 adipose tissue fatty acids, rate of change,
a-monoglyceride, 12 566-567, 567
P-monoglyceride, 12 coconut-palmkemel oil blend, chemical
organic acid esters of, food emulsifiers, 526, analyses, 56It
526-528 contaminants, 564-565
acetic acid esters, 526, 527 cyclopropenoid fatty acids, 565
citric acid esters, 528 dietary energy value, chemical structure and,
diacetyltartaric acid esters, 526, 527-528 558-565
lactic acid esters, 526, 527 digestibility, chemical structure and, 558-565
sulfated, nonionic surfactants, 684 erucic acid, 565
Mowrah, fatty acid composition, 25,407 free fatty acid content, 559-562, 559-562,
Mowrah butter, botanical name, 20 560-56 It
Mustardseed oil, 6, 24, 38-39. 391 heat damage, 560t, 562-564, 564t
fatty acid composition, 24 meat quality, effect on, 565-567
Myristic acid, 3, 4 pesticides, 564-565
physical properties of, 4 physical texture, effect on, 565-566
polytene, 565
Neck cream, formulation, 702, 703t saturation, 559-562, 559-562, 560-56It
Nephelium lappaceum {see Rambutan tallow) sunflower acid oil, chemical analyses, 564t
Nervonic acid, 6 Non-triacylglycerol components, of fats, oils, 138t
Neurology, phospholipids and, 71 Nowrah butter, botanical name, 20
Neutralization refining, 140-143, ¡42 Nuclear magnetic resonance
Nickel pulse, melting, crystallization of fats, 415
catalysts, for triglyceride oil hydrogenation, wide-line, melting, crystallization of fats, 415
285, 288, 295-300 Nucléation
hydrogenation catalyst, 285, 288, 295-300 cocoa butter, crystallization, 404
oxidation, effects on, level, 186 fractionation and, 204t, 204-205,205
rearrangement catalyst, use in India, 378 Nut coatings, lipid, 468-469
in vanaspati production, 385 Nutrition, 79-112
Nicotiana tabacum {see Tobacco) dietary lipids, 81-83
Nigerseed oil, botanical name, 20 composition of, 81-83,82
820 Subject Index
[Nutrition] [Nutrition]
sources of, 81, 8It interesterification, 98-99
fat-soluble vitamins, 94-95 irradiation, 96
carotenoids, 94 mixing, 95
retinol, 94 reduced fat products, 99-101
vitamin A, 94 spreads, 99-100
vitamin D, 94-95 structured fats, 99-101
vitamin E, 95 undigested fat analogues, 100
fatty acid classes, in food natural fats, 82 steroids, 93-94
glycolipids, 93 cholesterol, 93
health issues, 101-108 plant steroids, 93-94
cancer, 105-106 storage lipids, 80
cardiovascular disease, 103-105, 104t structural lipids, 80
depression, 107 cholesterol, 80
inflammatory disease, 106-107 glycolipids, 80
lipostatic theory, weight gain, 102-103 phospholipids, 80
obesity, 101-103 in transit, 80
body fat distribution, 101-102 triacylglycerols, 87-93
fat calorie, vs. carbohydrate calorie, weight aroma, 87
gain and, 102-103 distribution of fatty acids in, 93
thrombosis, 105 flavor, 87
human body metabolic energy and, 88
functions of, 80t monounsaturated acids, 89
lipids in, 80t, 80-81 palatability, 87
intravenous, lipid emulsions for, 544-552 polyunsaturated fatty acids, 89-93
low calorie fats, 511-514 eicosanoids, production of, 90
metabolism of lipids, 83-87 essential fatty acids
absorption, 83, 84 requirements for, 91
biosynthesis, 85 roles, 89-91, 90
breakdown, tissue lipids, 87 n-3 family, 82, 92-95
fatty acid oxidation, for cellular energy, n~6 family, 86, 91-92
87 saturated fatty acids, 88
lipolysis, 87 structural lipids, supply, 88
peroxidation, 87 texture, 87
digestion, 83,84
essential fatty acids, 86-87 Obesity, 512-513 {see also Weight control)
deficiency, 86 lipids and, 101-103
metabolism, 86, 86-87 body fat distribution, 101-102
lipoproteins, 84 fat calorie, vs. carbohydrate calorie,
transport, 83-85, 84 102-103
unsaturated fatty acids, metabolism, weight gain and, 102-103
86 Octadecanoic acid {see Stearic acid)
phospholipids, 71-74, 93 Octadecatetraenoic acids, with conjugated
crustaceans, 72-73, 73, 73t unsaturation, 7t
fish, 73-74, 74t Octadecatrienoic acids, with conjugated
piglets, 72, 72t unsaturation, 7t
processing, effect of, 95-101 Octadecenoic acid, 5
digested reduced energy fats, 100-101 Octanoic acid {see Caprylic acid)
fat substitutes, 99-101 Oenothera biennis {see Evening primrose oil)
fractionation, 99 Oenothera lamarkiana {see Evening primrose oil)
medium-chain triacylglycerols, 99 Oil-in-water emulsions {see Cream Alternatives)
reduced cholesterol foods, 99 phospholipids, 64
heating, 96 properties of, 535-537, 536-537
homogenization, 95 structure, 535-537, 536-537
hydrogenation, 97-98 Oil-in-water spreads, 323
Subject Index 821
Oil stability index, 438 [Oxidation]

Oiticica oil antioxidants, 187
botanical name, 20 light, 186
fatty acid composition, 25 metals, 186t, 186-187
Okra seed oil, botanical name, 20 moisture, 187
Olea europea {see Olive oil) oxidation rates, fatty acids, 184t
Olefinic compounds. 3 oxygen concentration, 186
Oleic acid, I, 2, 3, 5 temperature, 185, 185t
cocoa butter bloom, 417 effect of, on off-flavor development, 185t
epoxidized oil, 759 trace metal levels, 186t
usage of term, 1 type of oil, 184t, 184-185,185
Olein, solid fat contents of, 218 unsaturation, effect on deterioration rate, 185
Olestra, 508-509 solubility, 180-181, 181t
Olive oil, 1,5, 36-37 storage, transport of oil, 178-188
botanical name, 20 effect of, on off-flavor development, 185t
extraction, 125 solubility, 180-181,181t
fatty acid composition, 23 trace metal levels, 186t
production of, 22 type of oil, 184t, 184-185,185
saponification, iodine values, 27 unsaturation, effect on deterioration rate, 185
tocopherols, vitamin E, 30 Oxidative rancidity, confectionery fats, 419-421,
Olives, lipid extraction, 125 420
Oncosol, lipid emulsion, 542, 543 Oxifrit-Test, frying oil quality measurement,
Orange roughy oil, 16 445
Orbignya species (see Babassu oil) Oxygenated acids, 9-10
Organoclays, cationic surfactant, 621-622 epoxy acids, 10
Organometallic compounds, interesterification, vemolic acid, 10
229 furanoid acids, 10
Oryza sativa (see Rice bran oil) urofuranic acids, 10
Outboard marine engine oil, two-stroke, hydroxy acids, 9-10
lubricants, rapidly biodegradable, 739 auricolic acid, 10
Overrun, ice cream production castor oil, 9
calculation of, 339 densipolic acid, 10
defined, 339 isoricinoleic acid, 10
Oxidation, 178-188 lesquerolic acid, 9, 10
absorption of oxygen, 181-184 mid-chain, without conjugated
dissolved oxygen content, 181-182, 182t, unsaturation, lOt
190t ricinoleic acid, 9, 10
gas/headspace, oxygen concentration in, Oxygen permeabilities, lipid coatings, 462t
183-184
headspace volume, 182, 182t PAF (see Platelet-activating factor)
oil saturated with air, gas content, 182t Paints, oils, fatty acids in, 711-736
rate of absorption, 182 autoxidation
surface area-to-volume ratio, 182-183, 183t enzymatically catalyzed, 721
influence of contact area, 183t photoinitiated, 721
analyses, 179-180 castor oil, dehydration of, 717
anisidine value, 179 chemical processes, 715-722
oxygen content, 180 chemistry, oil, fatty acid reactions, 714t,
peroxide value, 179 715-718, 716-717
soybean oil, 180 cobalt redox system, 719-720
totox value, 179-180,180 degradation, 721 -722
diffusivity, 180-181, 18 It dehydropolymerization, reaction of, 717
effect of processing, 188, 188t drying, 720-721
hydroperoxide breakdown, 187-188 dryers, 720
soybean oil, before, after refining, 188t metal, 720
hydroperoxide formation, 184-187 fatty acid moieties, 718
822 Subject Index
[Paints] Palmitoleic acid, 3, 5

mechanisms, 718-720, 719 Palmkemel oil, 33t, 37-38, 38t, 114, 125,411-
hydrogen abstraction, unsaturated fatty acids, 412t, ^72,413 {see Confectionery fats)
717 Papaver somniferum {see Poppyseed oil)
hydroperoxides, formation of, 718 Paraffin, as coating, 460
maleinization, oils, fatty acids, 716 Parenteral nutrition, 549-551
phospholipids, 62 administration, 549-550
polymerization, 719 admixtures, 550
fatty acid moieties, 718 complications, 550-551
scission, 719 emusions, lipid, 545-551
unsaturated oils, heat bodying dimerization formulation, 549-550
of, 715 lipid metabolism, 548-549
weathering mechanisms, 721-722 emulsion particles, 548
yellowing, 721 oxidation, fatty acids, 548-549
Palm fruit, v^. palm kernels, lipid extraction, peripheral administration, 550
125 a-Parinaric acid, 7
Palm hardened fats, crystallization rate, 312 Parkinson's disease, phospholipids and, 71
Palm oil, 1, 4, 5, 33t, 37-38, 38t, 407t, Peroxide value, 179
411-413 Partial hydrogenation {see also Hydrogenation)
as biofuel, 780 fractions of, 413-414, 413-414t
botanical name, 20 flying oils, 441
component acids, 33 Passijlora edulis {see Passionflower)
crystallization rate, 312 Passionflower oil, fatty acid composition, 25
fatty acid composition, 23 Pasteurization, ice cream mix manufacture,
fractionation, 275, 215-217, 216t, 217t, 336-337, 337
218 Pastiy
diy multiple fractionation, 215 puff, 325
processes compared, 217t short, 324-325
fractions, 424 Peach kernel oil
geographic area of production, 22 botanical name, 20
handling, 178 fatty acid composition, 25
hydrolysis, 177t, 178 Peanut oil {see also Groundnut oil) 414
interesterification, 226 as biofuel, 781
kernel preparation, 114 botanical name, 20
olein {see Fractionation) preparation of, 114
phospholipid composition of, 28 Pecan oil, fatty acid composition, 25
production of, 22 Penclomedine, lipid emulsion, 542
saponification, iodine values, 27 Penitol esters, nonionic surfactants, 649
solid fat contents of, 218 Perflubron, lipid emulsion, 542
stearin {see Fractionation) Perfluorocarbon, lipid emulsion, 542
tocopherols, vitamin E, 30 Perilla frutescens {see Perilla seed oil)
triacylglycerol composition, 38t, 226 Perilla seed oil, botanical name, 20
triglycerides, 411 Peroxide value, 179
vitamin E, 29, 30 Peroxyacetic acid
Palmitic acid, 4-5 from acetic acid, hydrogen peroxide, epoxidized
Chinese vegetable tallow, 4 oil, 764
cottonseed oil, 4 distilled aqueous, epoxidized oil, 764
lard, 4 preformed, epoxidized oil, 764
palmitoleic acid, 5 Peroxy acids, 764
palm oil, 4 epoxidized oil, 759
physical properties of, 4 high concentrations of, avoiding, 762
tallow, from sheep, cattle, 4 reaction between alkene, epoxidized oil, 763
usage of term, 1 reactivity, epoxidized oil, 763
zoomaric acid, 5 Peroxyformic acid, epoxidized oil, 764t
Palm midfraction, 216t Persea americana {see Avocado oil)
Subject Index 823
Personal care products, 695-709 Phospholipids, 27-28, 28t, 51-77, 62t, 65, 93
avocadin cream, formulation, 701, 702t Alzheimer's disease, 71
carrot oil cream, formulation, 701, 702t animal phospholipids, 54, 54
conditioning shampoo, formulation, 705-706, antioxidants, 68
706t applications, 61-74
diglycerides, 700 cephalin, usage of term, 51
dry skin, transepidermal water loss, 697 classes of, 52t
emulsions, 699 cocoa butter, nucléation, crystallization, 404
facial oil, moisturizing, formulation, 701 egg, 28, 52, 54, 58t
facial scrub cream, formulation, 704-705, 705t fatty acid composition of, 28t, 52t, 52-53,
formulations, 698-707 52-53t
hydrophilic-lipophilic balance, 699, 699t brain, 52
lanolin cream, formulation, 702, 703t com, 52
lipids, 695-698 egg, 52
lipstick, formulation, 706-707, 707t milk, 52
minor lipids, 698 rapeseed, 52
monoglycerides, 700 soy, 52
neck cream, formulation, 702, 703t sunflower seed, 52
night cream, formulation, 703, 704t cosmetics, 66-67
polarity index, 700, 700t liposomes, 67
refining lipids, 698 pro-liposome principle, 67
sunscreen cream, formulation, 703-704, ciystallization, cocoa butter, 404
705t fatty acid composition, 28t, 53t
triacylglycerols, 696t, 696-697 brain, 53
triglycerides, 700 com, 53
water-in-oil emulsions, 700 egg, 53
wax esters, 697-698 rapeseed, 53
wheat germ shampoo, formulation, 706, 706t soy, 53
Pesticides, in animal feed, 564-565 sunflower seed, 53
Petroselinic acid, 5 feed products, 62-63
Pharmaceuticals, phospholipids, 67-68 food applications, 63-66
Phase behavior, solubility and, fractionation, baking, 65-66
200-203,201-202 chocolate, viscosity, 65
centrifugal separation, 209-210, 210t confectionery, 64-65, 65, 66t
confectionery fats, 400-427 food emulsions, 64
POP, post, StOSt, 404-405 instant food products, 66
pressing, 210-212,277 lecithin types, 66t
triglycerides, binary mixture of, 201 oil-in-water emulsions, 64
tripalmitin separation agents, 64
in 2-oleodipalmitin, solubility of, 202 water-in-oil emulsions, 64
solubility of, 201 Friedreich's ataxia, 71
tristearin Huntington's disease, 71
solubility of, 201 hydrogenation, effect on, 299t, 299-300
in triolein, phase diagram, 202 lecithin, 27-28
vacuum filtration, 209,209, 210t cholesterol-lowering effect, 70t
Phosphate esters, nonionic surfactants, 683 granular, 28
Phosphatidic acid, 14 usage of term, 51
Phosphatidylcholine lipid metabolism, 69-70, 70t
enzymatic hydrolysis of, 15t liquid, 28
lipid-lowering effect, 70t liver therapy, 70
Phosphoglycerides, 13-15, 14-15t, 51-52 memory, 71
3-glycerophosphoric acid, 14 neurology and, 71
phosphatidic acid, 14 nutrition, 71-74
phosphatidylcholines, enzymatic hydrolysis of, crustaceans, 72-73, 75 73t
15t fish, 73-74, 74t
824 Subject Index
[Phosholipids] Platelet-activating factor, 16

piglets, 72, 72t Plum oil
occurrence, 51-52, 52t botanical name, 20
paints, 62 fatty acid composition, 25
Parkinson's disease, 71 Polyalkylene glycols, lubricants, 747
pharmaceuticals, 67-68 Polyamide 11, castor oil, 791
phosphatidylcholine, lipid-lowering effect, Polyene acid, 7t, 7-8
70t methylene-interrupted, 6t, 6-7
phosphoglycerides, 51-52 Polyethylene, antistatic treatment, 626
phosphosphingolipids, 51-52 Polyethylene wax, as coating, 460
Physical properties, 415-416,460t emulsifers, 528
physicochemical action principles of, 63 Polyglycerol esters, nonionic surfactants, 649
physiology of, 68-74 Polymerization, surface coatings, 719
processing, 54-61 Polymorphic behavior {see Phase behavior.
acetone deoiling, 55-56, 56t Bloom)
alcohol fractiopation, 56-57, 57t cocoa butter, 396-397
soybean lecithin, 57t confectionery fats, 400
chromatographic purification, 57-58, 58t Polyorganosiloxanes, fat substitute, 508
chromatographic systems, 58t Polyoxyalkylene esters, fatty acids, nonionic
commercial phospholipid products, uses, surfactants, 654
55t Polyoxyethylene aliphatic alcohols, sulfated,
compressed gas deoiling, 56 nonionic surfactants, 682
egg yolks, 54 Polyoxyethylene alkylphenols, sulfated, nonionic
phosphatidylcholines, molecular species, surfactants, 683
58t Polyunsaturated fatty acids, 3, 82-95, 309,445
modification, 59-61 spreads, fat content, 309
acetylation, 60 Poppyseed oil
chemical, 60-61 botanical name, 20
enzymatic, 59 fatty acid composition, 25
hydrogenation, 60-61, 61 Porphyrin derivative, lipid emulsion, 542
hydroxylation, 60 Pregnanolone, lipid emulsion, 542
solvent fractionation, 55, 55t Pressing, fractionation, 210-212,211
soybean lecithin, 61 Pretreatment
deoiled, composition, 56t in physical refining, 161
products, obtaining, 53-54 in refining process, 139-140
tardive dyskinesia, 71 Pristanic acid, 8
technical industry, 62 Processing, 54-61, 95-101
terminology, 51 acetone deoiling, 55-56, 56t
Tourette's disease, 71 alcohol fractionation, 56-57, 57t
varnishes, 62 soybean lecithin, 57t
vegetable, 53-54, .-)4t chromatographic, 57-58, 58t
soybean lecithin, composition of, 54t commercial phospholipid products, uses, 55t
Phosphosphingolipids, 51-52 compressed gas deoiling, 56
Physical refining, 158-162 digested reduced energy fats, 100-101
intensive degumming, 158-161,159-161 egg yolk, 54
pretreatment, 161 phosphatidylcholines, molecular species, 58t
steam refining, 161-162 fractionation, 99, 199-222
superdegumming process, 161 medium-chain triacylglycerols, 99
total degumming process, 159, 160 reduced cholesterol foods, 99
Physicnut oil, botanical name, 20 heating, 96
Phytanic acid, 8 hydrogenation, 97-98, 265-303
Phytoaxelins, epoxy acid, 759 interesterification, 98-99, 223-263
Phytol-based acid, 8 irradiation, 96
Phytosphingosine, 16 mixing, 95
Pistachio oil, fatty acid composition, 25 modification, 59-61
Subject Index 825
[Processing] Rapeseed oil, 4, 5, 6, 38-39, 39t

chemical, 60-61 as biofuel, 779-780
acetylation, 60 botanical name, 20
hydrogenation, 60-61, 61 in chocolate, content determination, 423-424
hydroxylation, 60 epoxidation, 763
enzymatic, 59, 245-259 fatty acid composition, 23, 24, 53
reduced fat products, 99-101 geographic area of production, 22
spreads, 99-100 {see Spreads) hydrogenation, 284, 285, 299t
solvent fractionation, 55, 55t, 199-222) iodine value, 765
soybean lecithin, 61 phospholipids, 28, 52
deoiled, composition, 56t preparation of, 114
undigested fat analogues, 100 production of, 22
Pro-liposome principle, 67 saponification, iodine values, 27
Propylene, antistatic treatment, 626 stereospecific analysis, 39t
Propylene glycol, 528 sterols, 29
Propylene oxide tocopherols, vitamin E, 30
castor oil, reactions, 789 Rapidly biodegradable lubricants, base oils,
nonionic surfactants, 644 744-747
Protected fats, 569-570, 570t RAU test, frying oil quality measurement, 445
Proxilate Rayon manufacture, cationic surfactant, 626
epoxidized oil, 764 Reaction systems, enzymatic interesterification,
peroxyacetic acid from, epoxidized oil, 764 249-250, 251-256, 252-254, 254-255t,
Prunus amygdalus {see Almond oil) 256
Prunus armaniaca {see Apricot oil) Reactive compounds, epoxides as, 759
Prunus domestica (see Plum) Reactivity, peroxy acids, epoxidized oil, 763
Prunus dulcís {see Almond oil) Rearrangement catalysts, 229-230, 229-231
Prunus pérsica {see Peach oil) acids, 229
Psitachio vera {see Pistachio) activation, interesterification, 234
PUFA spreads, 309 alcoholates, 229
Puff pastry, 325 alkali metals, 230-231
Pumpkin seed oil, 25 alkoxides, 231
botanical name, 20 enzymes, 245-259
Púnica granatum {see Punicic acid) glycerolates, 231
Punicic acid, 7 hydrogenation, performance, examples,
295-299, 296t, 297-298, 299t
Quality measurement, frying oils, 444-445 hydroxides, 231
quick tests, 445-447 of metal, 229
spot test, 446 interesterification, 223
metals, 229
Raisin coatings, lipid, 468-469 nickel, use in India, 378
Rambutan tallow, 5 organometallic compounds, 229
Rancidity {see also Stability) oxides, of metal, 229
confectionery fats, 419-421 salts, of metal, 229
lipolytic, 421-422, 421-422t sodium methoxide, 229
oxidative, 419-421, 420 vanaspati production, 378-379
short-chain free fatty acids, flavor and, Reduced fat spreads, 320-322, 321-322t
421t Refined oil
Random interesterification, 239-245 handling, 197
control methods, 244-245 transport, 196-197
feedstock requirements, 239-240,240 air contact, 196
posttreatment, 243-244 contamination, 196
process options, 241 heating, 196
batch processes, 230, 241 metal contact, 196-197
continuous processes, 241-243,242-243 segregation, 196
process yield, 243-244 temperature, 196
826 Subject Index
Refining plant, multipurpose, mixer combinations Rheology, chocolate, 394t, 394-396, 395-396, 396t
in, 142 particle size distributions, 394t
Refining process, 137-167 shear rate-shear stress curves, 394-395
alternatives, 138-139,139 viscosity, ranges, 396t
chemical refining, 139-158 Rihes nigrum {see Blackcurrant pip oil)
batch refining, 145 Rice bran oil, 39-40
bleaching, 145-149 as biofuel, 781
bleaching clay botanical name, 20
filtration, 148-149 fatty acid composition, 24
spent, 149 saponification, iodine values, 27
equipment for, 147-148,148 Ricinoleic acid, 3, 9, 10
plant operation, 149 esters, uses of, 788-790
Sparbleach process, reactor for, 148 molecular formula, 787
degumming, 139-140 usage of term, 1
deodorization, 150-158 Ricinus communis {see Castor oil; Ricinoleic acid)
batch deodorizer, 157-158 Road construction, cationic surfactants in, 623
chelation, 154 Rotaiy drum filter, vacuum filtration, fractionation,
cooling, 154 209
deaeration, 152 Rubberseed oil, botanical name, 20
designs, for deodorizer, 157 Ruminants, lipids in feed, 568-574
equipment, 154-158 absorption, 568t, 568-570
fully continuous deodorizer, 155-157,157 beef cattle, 570-571
flow diagram, 157 commercial setting, use of fats in, 570-574
heat recovery, 153 cows, 571, 571-574, 573
high temperature heating, 153 milk, 574t
polish filtration, 154 digestion of fats, 568t, 568-570
preheating, 152-153 feeds, energy, 568t
process variables, 151-152 fermentation, 568t, 568-570
semicontinuous deodorizer, 155,156 protected fats, 569-570, 570t
flow diagram, 156 sheep, 570, 570t
theoretical treatments, 151
with dilute alkali, 142, 143-144,144 Safflower seed oil, 5, 40
Zenith, 144 botanical name, 20
drying, 143 fatty acid composition, 24, 25
mixer combinations, multipurpose refining iodine value, 765
plant, 142 preparation of, 114
neutralization refining, 140-143,142 saponification, iodine values, 27
pretreatment, 139-140 stereospecific analysis, 40t
silica refining, 149-150 tocopherols, vitamin E, 30
washing, 143 vitamin E, 29
fatty acids, 137-167 Safingol, lipid emulsion, 542
future of, 162-164 Salad dressing, 440-441
non-triacylglycerol component, fats, oils, 138t Salad oils, 433-441, 447-448 {see also Salad
overview, 137-138, 138t dressing)
physical refining, 158-162 additives, 436-437
intensive degumming, 158-161,159-161 antioxidants, 436-437
pretreatment, 161 metal chelating agents, 436
steam refining, 161-162 nitrogen treatment, 437, 437
superdegumming process, 161 deterioration of, 433-435, 434-435, 436t
total degumming process, 159, 160 flavor, 435-436
Resin, defined, 714-715 odor, 435-436
Resin systems, surface coatings, 722-732 production, countiy-by-country listing, 434t
alkyd resins, 723-730 reaction rates, with oxygen, 436t
epoxy esters, 731, 731 stability, 435-437
oleoreinous systems, 722-723 evaluation of, 437-439
Subject Index 827
[Salad oils] [Saturated acids]

active oxygen method, 438 oleic acid, 5
oil stability index, 438 shea butter, 5
Schaal oven test, 438 Schaal oven test, oils stability evaluation, 438
shelf storage test, 437 Schleichera trifuga {see Kusum seed)
volatiles analyses, 438-439 Scrub cream, facial, formulation, 704-705,
sensory evaluation methods, 439-440 705t
salad oils, 439-440 Sea kale oil, botanical name, 20
Salatrim, 506-507 Semidrying, defined, 714-715
Sal fat Sensory evaluation methods, oil stability,
botanical name, 20 439-440
fatty acid composition, 25 Separation, fractionation, 208-212
fractionation, 220 vacuum filtration
confectionery fat, 407 rotary drum filter, 209
Salting, in buttermaking, 307 vacuum belt filter, 209
Sanitizers, detergents, anionic, 598-599 Separation agents, phospholipids, 64
Sapium sebiferum {see Chinese vegetable tallow) Sesame seed oil, 21,40
Saponification values, 26-27, 27t botanical name, 20, 21
Sarcosinates, anionic detergents, 588-589, 589 fatty acid composition, 23
Saturated acids, 4t, 4-5 preparation of, 114
alkanoic acids production of, 22
names, 4t saponification, iodine values, 27
physical properties, 4t tocopherols, vitamin E, 30
butanoic acid, 4 vanaspati production, 384
capric acid, 4 Sesamum indicum {see Sesame seed oil)
caprylic acid, 4 Shampoos
coconut oil acid, 4 anionic, 599-602, 600t
cuphea oil, 4 conditioning
lauric acid, 4 anionic, 601-602
long-chain, 5 formulation, 705-706, 706t
arachidic acid, 5 wheat germ, formulation, 706, 706t
behenic acid, 5 Shea fat, 5
groundnut oil, 5 botanical name, 20
kusum, 5 crystallization rate, 312
lignoceric acid, 5 fatty acid composition, 25,407
Lophira procera, 5 fractionation, 220, 406
Lophira alata, 5 confectionery fat, 402, 406-408
rambutan tallow, 5 Sheep, lipids in feed, 570, 570t
medium-chain, 4 Shorea macrophylla {see Illipe butter)
myristic acid, 4 Shorea stempierà {see Illipe butter)
palmitic acid, 4-5 Short-chain saturated acid, 4
Chinese vegetable tallow, 4 Short oil, defined, 714-715
cottonseed oil, 4 Shortening, 324
lard, 4 Short pastiy, 324-325
palm oil, 4 Shower gels, anionic, 602
tallow, from sheep, cattle, 4 Silica refining, 149-150
palmkemel oil, 4 Simarauba glauca {see Aceituno)
rapeseed oil, 4 Simmondsia chinensis {see Sal fat)
short-chain, 4 Sinapsis alba {see White mustardseed oil)
stearic acid Sitosterol, sterol composition, 29
Borneo tallow, 5 Soaps {see also Detergents, Surfactants)
cocoa butter, 5 anionic, 586t, 586-587
illipe tallow, 5 Sodium methoxide, interesterification, 229
linoleic acid, 5 Sodium stearoly-lactylate, 529
linolenic acid, 5 Soil protection, lubricants, legislation, 742-743
828 Subject Index
Solubility, phase behavior and, fractionation, [Spingolipids]

200-203,201-202 Gaucher’s disease, 17
centrifugal separation, 209-210,210t Krabbe’s disease, 17
pressing, 210-212,211 Niemann-Pick’s disease, 17
triglycerides, binary mixture of, 201 occurrence of, 17t
tripalmitin phytosphingosine, 16
in 2-oleodipalmitin, solubility of, 202 sphingenine, 17
solubility of, 201 sphingomyelin, 16, 17
tristearin sphingosine, 16
solubility of, 201 structure, 17t
in triolein, phase diagram, 202 Tay-Sachs disease, 17
vacuum filtration, 209,209, 210t Sphingomyelin, 16, 17
Solvent extraction, 127-133 Sphingosine, 16
alternative solvents, 127-128 Spray processing, foodstuffs, 481-500, 482
recovery, 132-133 spray drying, 483-490
traditional solvents, 127 atomizers, 483-485, 484
Solvent fractionary, 212, 406 drying mechanisms, 489-490
Sorbestrin, 508 high fat formulations, 491-493
Sorbitan stearates, 528-529, 529 butter powder, 492
Sorbitol coffee whiteners, creamers, 492-493
nonionic surfactants, 644 cream liquor powder, 493
polyester (see Sorbestrin) fat-filled powders, 491-492
Sources of lipids, 19-50 ice cream powder, 493
South America, Aztecs, use of chocolate, 391 whole milk powder, high free fat content,
Soybean oil, 40-41,411 492
as biofuel, 778 spray-dried powders, high fat, 487, 490-491,
botanical name, 20 493
crystallization, 203 spray dryer
epoxidized oil, 759, 764t chamber designs, 485-489, 486-488
fatty acid composition, 23, 53 safety issues, 497-498, 497-499
geographic area of production, 22 selection, 493-496, 493-497, 496t
hydrogenation, 283-284,284, 296t "Spreadable fats" (see also Spreads)
iodine value, 765 defined, 310t
lecithin, 61 regulation of, 309
alcohol fractionation, 57t Spreads
composition of, 54t baking fats, 323-326
deoiled, composition, 56t butterfat, 309
phospholipids, 28, 52 dairy, 322, 387-389
preparation of, 114 fat content, 309
production of, 22 fatty acids, 323-326
saponification, iodine values, 27 India, production in, 387-389
tocopherols, vitamin E, 30 emulsification, 388, 389
triacylglycerols, 411 fat component, 388
stereospecific analysis, 41t manufacture, 388-389
vitamin E, 29 specifications, 387-388
Solvent fractionation, 212 liquid margarine, 323
Spectoscopic properties, 765-766t low fat, 320-322, 321-322t
Sperm whale oil, 16 oil-in-water, 323
Sphingenine, 17 polyunsaturated fatty acid, fat content, 309
Sphingolipids, 16-17, 17t PUFA, 309
ceramide, 16, 17 reduced fat, 99-100, 320-322, 321-322t
cerebroside, 17 vegetable fat, fat content, 309
Fabry’s disease, 17 vegetable fat/butterfat blended, fat content, 309
Färber’s disease, 17 water continuous, fat content, 309
ganglioside, 17 Spurge seed oil, botanical name, 20
Subject Index 829
Squash seed oil, botanical name, 20 [Storage of oil]

Stability moisture, 193
food emulsifiers, 530, 530-531 segregation, 193
flying, salad oils, evaluation of, 437-439 temperature, 193-194
active oxygen method, 438 hydrolysis, 174
oil stability index, 438 analyses, 174
Schaal oven test, 438 factors affecting, 175-178
sensory, 439-440 free fatty acid, 175
shelf storage test, 437 moisture, 175-176, 176t
volatiles analyses, 438-439 oil type, 177
ice cream, 334-335 overall rate, 177-178,778
Starch-based product coatings, lipid, 470 temperature, 174-175, 177t
Static hardening, in ice cream freezing, 340-341 palm oil, 177t
Steam, castor oil, splitting with, 790 free fatty acid, saturation solubility and, 176t
Steam refining, 161-162 palm oil, saturated with water, 178
Stearic acid, 2, 3, 4-5 solubility, 174-175, 175t
Borneo tallow, 5 water in oil, 175t
as coating, 460 intermediate oil storage, 197
cocoa butter, 5 oxidation, 178-188
illipe tallow, 5 absorption of oxygen, 181-184
linolenic acid, 5 dissolved oxygen content, 181-182, 182t,
oleic acid, 3, 5 190t
physical properties of, 4 gas/headspace, oxygen concentration in,
shea butter, 5 183-184
Stearidonic acid, structure, double bond position, headspace volume, 182, 182t
6 oil saturated with air, gas content, 182t
Stearin, solid fat contents of, 218 rate of absorption, 182
Steaiyl alcohol, as coating, 460 surface area-to-volume ratio, 182-183, 183t
Sterilized cream alternatives, 364 influence of contact area, 183t
Steroids, 93-94 analyses, 179-180
cholesterol, 93 oxygen content, 180
cocoa butter, 422 peroxide value, 179
plant steroids, 93-94 soybean oil, 180
Sterol composition, 29, 29t totox value, 179-180,180
vegetable oils, 29t effect of processing, 188, 188t
Stigmasterol, sterol composition, 29 hydroperoxide breakdown, 187-188
Stillingia oil soybean oil, before, after refining, 188t
botanical name, 20 hydroperoxide formation, 184-187
fatty acid composition, 25 antioxidants, 187
Stillingia sebiferum {see Stillingia) light, 186
Stokes aster, vemolic acid, 760 metals, 186t, 186-187
Storage of oil, 80, 169-198 moisture, 187
costs, 170-174 oxidation rates, fatty acids, 184t
capital costs, 171-172 oxygen concentration, 186
investment, 17 It, 171-172 temperature, 185, 185t
working capital, 172 effect of, on off-flavor development, 185t
degradation costs, 172-173 trace metal levels, 186t
operating costs, 172 type of oil, 184t, 184-185,185
total costs, 173t, 173-174 unsaturation, effect on deterioration rate,
crude oil, 173t, 192t, 192-194 185
air contact, 193 solubility, 180-181, 18 It
contamination, 1 3 refined oil, 173t, 194-196
heating, 193-194 air contact, 190t, 194-195
impurities, 193 contamination, 194
metal contact, 194 examples, 195-196,196
830 Subject Index
[Storage of oil] [Surface coatings]

heating, 195 chemistry, 714t, 715-718, 716-717
metal contact, 195 cobalt redox system, 719-720
segregation, 194 defined, 714-715
temperature, 195 degradation, 721-722
Strophanthus species {see Isoricinoleic acid) dehydropolymerization, reaction of, 717
Structural lipids, 80 drying, 720-721
cholesterol, 80 dryers, 720
glycolipids, 80 fatty acid moieties, 718
phospholipids, 80 mechanisms, 718-720, 719
Structure of fatty acids, 1-10, i hydrogen abstraction, unsaturated fatty acids,
Styrene, antistatic treatment, 626 717
Suberins, 16 hydroperoxides, formation of, 718
Sucrose esters, 529 polymerization, 719
Sucrose, nonionic surfactants, 644 resin systems, 722-732
Sucrose polyester {see Olestra) alkyd resins, 723-730
Sugar bloom, chocolate, 416 epoxy esters, 731, 757
Sugar esters, nonionic surfactants, 653 oleoreinous systems, 722-723
Sulfation scission, 719
castor oil, 788 terminology, 714-715
detergents, anionic, 589-590, 591-593, 592 unsaturated oils, heat bodying dimerization of,
a-sulfonated acids, detergents, anionic, 587-588, 715
588 weathering mechanisms, 721-722
Sulfonates, 587-588, 588 yellowing, 721
Sulfosuccinates, nonionic surfactants, 683 Surface coatings, 711-736
Sunflower seed oil, 5, 41-42,42t, 564t Surfactants
as biofuel, 780 cationic, 609-631, 625t, 626t
botanical name, 20 amidoamines, 612
in chocolate, content determination, 423-424 amphoterics, manufacture of, 626
enzymatic modification, 258 applications, 620-627
fatty acid composition, 23, 25, 53 biocides, 624, 624t, 624-625
geographic area of production, 22 bitumen, 623,624t
high-oleic, 258t "blossom thinning," 625
iodine value, 765 chemical properties, 616-617
phospholipids, 28, 52 surface, 617-620
preparation of, 114 derivatization, 613-615
production of, 22 epoxy resins, manufacture of, 626
saponification, iodine values, 27 esteramines, 612-613
stereospecific analysis of, 42t etheramines, 613
tocopherols, vitamin E, 30 fabric conditioners, 620-621, 621
triacylglycerols, 42t fatty amines, 611
vitamin E, 29 future products, 626-627
Sunscreen cream, formulation, 703-704, 705t imidazolines, 612
Supercooling, triglyceride crystals, solubility, manufacture, 611-615
variations, 204t markets, 609-610, 670. 61 Ot
Supercytes, lipid emulsion, 542, 543 melting points, 618t
Superdegumming process, 161 mineral flotation, 622t, 622-623
Supersaturation, fractionation, 203, 203-204 oil-based drilling fluids, 625
crystallization, soybean oil, 203 organoclays, 621-622
Surface coatings, 711-736 physical properties, 617-620
autoxidation polyethylene, antistatic treatment, 626
enzymatically catalyzed, 721 producers of, international, 610t
photoinitiated, 721 properties, 616-620
castor oil, dehydration of, 717 propylene, 626
chemical processes, 715-722 rayon manufacture, 626
Subject Index 831
[Surfactants] [Surfactants]
road construction, 623 raw materials, 640-648, 64It, 642
styrene, 626 alkylphenols, 647-6tfb
urethanes, antistatic treatment, 626 D-glucose, 644
water, heptane, partition data, 619t ethylene oxide, 641-643, 643
nonionic, 633-694, 634, 635t fatty acids, 644-645
alkylene oxide addition, 656-660 fatty alcohols, 645-647
polyalkylene oxide copolymers, 659-660 fatty amines, 647
polyoxyalkylene fatty alcohols, 657-658 glycerol, 643
polyoxyethylene alkylamides, 658-659 propylene oxide, 644
polyoxyethylene alkylamines, 658 sorbitol, 644
polyoxyethylene alkylphenols, 657 sucrose, 644
amide, formation of, 655-656 solubilization, 663-664, 663-664
amine, oxides, 660 surface adsorption, 664-667, 665, 666t
applications, 677-685, 678t wetting, 667-669, 668t
chemistry, 648-660 world production, 636-640, 636-640t
commercial, listing, 635t Sweeteners, ice cream, 334
critical micelle concentration, 66It, 661-662 Sunflower seed oil, 41-42, 42t
detergency, 677 Synthetic lubricants, 746, 746-747, 747-748t
dispersion, 673-677, 676, 677t Systematic names, fatty acids, conversion rules,
ecotoxicological issues, 685-688, 686-688t 1-2
emulsification, 669-673, 670-672t, 674-676t
emulsion Tall oil
defined, 669 epoxidized oils, 759
microemulsions, distinguished, 669 fatty acids, 24, 42-43,43t
esterification, 648-654 Tallow
glycerol esters, 648-649 as biofuel, 781
hexitol esters, 649-653, 650-652 fatty acid composition, 23
penitol esters, 649 production of, 22
polyglycerol esters, 649 from sheep, cattle, 4
polyoxyalkylene esters, fatty acids, 654 stereospecific analysis, 46t
sugar esters, 653 tocopherols, vitamin E, 30
tetritol esters, 649 triacylglycerols, 46t
etherification, 654-655, 655 Tardive dyskinesia, phospholipids and, 71
foaming, 662, 662t Taurates, anionic detergents, 588-589, 589
industrial end uses, 678-682t, 682 Tay-Sachs disease, sphingolipids and, 17
interfacial adsorption, 664-667, 665, 666t Teaseed oil, botanical name, 20
as intermediates, 682-684 Tempering of chocolate, 396-397,398
ether carboxylates, 683 Tempermeter, for chocolate tempering, 397
ether sulfonates, 683-684 Tetracosanoic acid (see Lignoceric acid)
phosphate esters, 683 Tetradecanoic acid (see Myristic acid)
sulfated monoglycerides, 684 Tetraenes, trivial names, 7
sulfated polyoxyethylene aliphatic alcohols, Tetritol esters, nonionic surfactants, 649
682 Textile cleaners, anionic, 603
sulfated polyoxyethylene alkylphenols, 683 Thea sansanqua (see Tea)
sulfonated methyl esters, 683 Theobroma cacao (see Cocoa butter)
sulfosuccinates, 683 Theobroma oil, botanical name, 20
mircoemulsions, defined, 669 Thrombosis, lipids and, 105
natural/renewable, v^. petrochemical Tirilazad, lipid emulsion, 542
component in, 64It Tobacco oil, fatty acid composition, 25
physicochemical properties, 660-677 Tocopherols, 29-30, 30t
polymeric surfactants, 676, 684-685 Tocotrienol, 29
producers, country-by-countiy listing, Toffe fats, 425
636t-640t Total degumming process, 159, 160
production, 648-660 Total-loss lubricants, 751-752, 752t
832 Subject Index
Totox value, 179-180,180 [Transport of oil]

Tourette's disease, phospholipids and, 71 hydroperoxide breakdown, 187-188
Transepidermal water loss, diy skin, 697 soybean oil, before, after refining, 188t
trans fatty acids, 8, 300, 300-301 hydroperoxide formation, 184-187
fractions of, partially hydrogenated oils, fats, antioxidants, 187
413 light, 186
flying oils, 441-442 metals, 186t, 186-187
trans hardened fats moisture, 187
with cocoa butter, phase behavior, 413-414 oxidation rates, fatty acids, 184t
fatty acid composition, 413t oxygen concentration, 186
solid fat content, 414t temperature, 185, 185t
Transit, lipids in, 80 effect of, on off-flavor development, 185t
Transition temperature, 5231 trace metal levels, 186t
Transport cleaners, anionic, 605 type of oil, 184t, 184-185,185
Transport of oil, 169-170,170 unsaturation, effect on deterioration rate,
crude oil, 190-192 185
air contact, 191 solubility, 180-181, 18 It
contamination, 191 refined oil, 196-197
heating, 191 air contact, 196
metal contact, 191 contamination, 196
temperature, 191 heating, 196
hydrolysis, 174 metal contact, 196-197
analyses, 174 segregation, 196
diffusivity, 174-175, 175t temperature, 196
water in oil, 175t requirements, 188-197, 190t
factors affecting, 175-178 Triacylglycerols, 11,12, 24-26, 87-93, 227t
free fatty acid, 175 cocoa butter, 402, 422-423
moisture, 175-176, 176t coconut oil, 33t, 411
oil type, 177 composition, 36t
overall rate, 177-178,178 distribution of fatty acids in, 93
temperature, 174-175, 177t enzymatic interesterification, 258t
palm oil, 177t groundnut oil, 35t
free fatty acid, saturation solubility and, 176t interesterification, 223
palm oil, saturated with water, 178 medium-chain, low calorie fats, 505-506
solubility, 174-175, 175t metabolic energy and, 88
water in oil, 175t monounsaturated acids, 89
oxidation, 178-188 number of, number of acids, relationship, 13t
absorption of oxygen, 181-184 palatability, 87
dissolved oxygen content, 181-182, 182t, with palmitic, oleic acids, alone, 12t
190t palmkemel oil, 411
gas/headspace, oxygen concentration in, polyunsaturated fatty acids, 89-93, 309, 445
183-184 eicosanoids, production of, 90
headspace volume, 182, 182t essential fatty acids
oil saturated with air, gas content, 182t requirements for, 91
rate of absorption, 182 roles, 89-91, 90
surface area-to-volume ratio, 182-183, 183t n-3 family, 82, 92-95
influence of contact area, 183t n-6 family, 86, 91-92
analyses, 179-180 saturated fatty acids, 88
anisidine value, 179 structural lipids, supply, 88
oxygen content, 180 Trialkoxytricarballyate, fat substitute, 509
peroxide value, 179 Trienes
soybean oil, 180 source, 7
totox value, 179-180,180 trivial name, 7
diffusivity, 180-181, 181t Triglyceride crystals, supercooling, solubility,
effect of processing, 188, 188t variations, 204t
Subject Index 833
Triglyceride oil hydrogenation, 265-303 (see [Vanaspati]

Hydrogenation) process parameters, 376-382
Trilinolenate, saponification equivalent, iodine selectivity, 382
value, 27 nickel contents, 385
Trioleate, saponification equivalent, iodine value, operation, 377
27 types of, 377-378
Tripalmitate, saponification equivalent, iodine production, 373-382
value, 27 properties of, 382-383, 383t, 384-385
Tristearate, saponification equivalent, iodine value, rearrangement catalysts, 378-379
27 regulations, 376
Triticum aestivum (see Wheatgerm oil) sesame oil, color and, 384
Trilaurate, saponification equivalent, iodine value, standards, 376
27 vitamin A, addition of, 385
Trivemolin, in epoxidized oils, 760 Varnishes
Tucum oil, 411 defined, 714-715
Tung oil, 43 phospholipids, 62
botanical name, 20 Vegetable coatings, lipid, 466-467
fatty acid composition, 24 Vegetable fat/butterfat blended spreads, fat
preparation of, 114 content, 309
saponification, iodine values, 27 Vegetable fat spreads, fat content, 309
Turnip seed oil, botanical name, 20 Vegetable marrow seed oil, botanical name, 20
Vegetable oil alkyl esters, biofuels, 773-782,
Ucuhuba oil, fatty acid composition, 25 774-775t
Unsaponifiable material content, 26-27, 27t Vegetable oils, biofuels, fatty acid composition of,
Urethanes, antistatic treatment, 626 774t-775t
Urofuranic acids, 10 Vegetable phospholipids, 53-54, 54t
U.S. Food and Drug Administration, film, coating Vemolic acid, 3, 10, 759, 760
approvals, 458t-459t, 464t epoxidized oils, 759
in seed oils, natural occurrence of, 760t
Vaccenic acid, 5 Vernonia anthelmintica, vemolic acid, 760
Vacuum filtration, fractionation, 209,209, 210t Vemonia seed oil, 43t
rotary drum filter, 209 botanical name, 20
vacuum belt filter, 209 composition of, 76It
Vanadium, oxidation, effects on, level, 186 divemolic in, 760
Vanaspati, 373-385 fatty acid composition, 24
active carbons, 380 saponification, iodine values, 27
base oils, usage of, 375-376 trivemolin in, 760
bleaching earths, 380 Virola sufinam (see Ucuhuba)
Bureau of Indian Standards, requirements, 381 Viscosity, of chocolate, phospholipids and, 65
by-products, 381 Vitalipid, lipid emulsion, 541
characteristics, 382, 383t Vitamin A, in vanaspati production, 385
composition, 382 Vitamin C, 29-30
cream, usage of, 375 Vitamin E, 29-30, 30t
crystallization, 382 Vitellaria paradoxa (see Shea butter)
effluent treatment, 38It, 381-382 Vitis vinifera (see Grapeseed oil)
energy conservation, 381 Volatiles analyses, oils, 438-439
equipment, 377
granularity, 382 Walnut oil
historical evolution, 373-375 botanical name, 20
hydrogen fatty acid composition, 25
consumption of, 378 vitamin E, 29
generation of, 380 Washing, in refining process, 143
hydrogenation Water continuous spreads, fat content,
monitoring of, 378 309
oils for, 379-380 Water-in-oil emulsions, phospholipids, 64
834 Subject Index
Water protection, lubricants, legislation, 741-742,

742
Water vapor permeabilities, lipid coatings, 46It
Waxes, 30, 44, 44t
composition of, 44t
Wax esters, personal care products, 697-698
Weathering mechanisms, surface coatings,
721-722
Weight control, 512-513 {see also Low calorie
fats. Obesity)
Wetting, nonionic surfactants, 667-669, 668t
Wheatgerm oil, 30t, 43
botanical name, 20
fatty acid composition, 24
saponification, iodine values, 27
shampoo, formulation, 706, 706t
tocopherols, vitamin E, 30
vitamin E, 29
Whipping cream, 358, 359-364, 359-364
White mustardseed oil, botanical name, 20
Whole milk powder, high free fat content, spray
drying, 492
Winterization, 220
Wool wax, 16
Xeranthemyum annuum, coronaric acid, 760

X-ray diffraction, melting, crystallization of fats,
415
Yellowing, surface coatings, 721

Yields, oil-producing crops, l i l t
Zea mays (See Maize oil. Com oil)

Zenith dilute alcohol refining process, 144
Zinc, oxidation, effects on, level, 186
Zoomaric acid, 5
Zoosterol, 29 (See also Cholesterol)

Frank D Gunstone - Lipid Technologi PDF

Uploaded by

Copyright:

Available Formats

Frank D Gunstone - Lipid Technologi PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Frank D Gunstone - Lipid Technologi PDF

Uploaded by

Copyright:

Available Formats

What is the book about?

What is the book about?

What are some of the major uses of oils and fats mentioned?

What are some of the major uses of oils and fats mentioned?

Lipid Technologies

Marcel Dekker, Inc . N ew York • Basel •H ong Kong

Lipid technologies and applications / edited by Frank D. Gunstone,

This book is printed on acid-free paper.

Copyright © 1997 by MARCEL DEKKER, INC. All Rights Reserved.

MARCEL DEKKER, INC.

Current printing (last digit):

PRINTED IN THE UNITED STATES OF AMERICA

10. Hydrogenation of Edible Oils: Technology and Applications 265

Part III Food Emulsions

Part IV Nonaqueous Foods

Part V Special Food Applications

Part VI Nonfood Uses

K. T. Achaya CSIR Emeritus Scientist (retired), Bangalore, India

Alan D. James Akzo Nobel Chemicals Ltd., Littleborough, Lancashire, England

John M . Krochta Department of Food Science and Technology, University of California-

Theo M ang Fuchs Petrolub AG Oil + Chemie, Mannheim, Germany

Thomas H. Shellham m er Department o f Food Science and Technology, University of

* Current affiliation: Consultant, Chicago, Illinois

Maurice A. Williams Anderson International Corp., Cleveland, Ohio

Julian Wiseman Department of Agriculture and Horticulture, The University of Nottingham,

II. FATTY ACID NOMENCLATURE

which may also be represented by the line drawing

Gamma-linolenic acid GLA

III. FATTY ACIDS— MAIN STRUCTURAL FEATURES

Table 1 Names and Selected Physical Properties of Some Alkanoic Acids

Acid Methyl ester

Chain Systematic Trivial mp bp mp bp Mol

4 Butanoic Butyric - 5 .3 164 — 103 8 8 .1

^bp at 760 mm or at 1 or 2 mm as indicated by superscript.

IV. SATURATED ACIDS

A. Short- and Medium-Chain Saturated Acids (C4-C14)

B. Palmitic and Stearic Acid

VI. METHYLENE-INTERRUPTED POLYENE ACIDS

Table 2 Methylene-Interrupted Polyene Acids

Trivial name Structure Double bond position^

Linoleic 18:2 {n-6) 9,12

®A11 double bonds have cis configuration.

VII. OTHER POLYENE, ACETYLENIC, AND ALLENIC ACIDS

Table 3 Natural Octadecatrienoic and Octadecatetraenoic Acids with

^Also kamlolenic, 18-hydroxy-a-eleostearic (Mallotus philippinensis).

VIII. ACIDS WITH TRANS OLEFINIC UNSATURATION

IX. BRANCHED-CHAIN ACIDS

COOH phytanic acid (G 2 0 )

COOH pristanic acid (C 19 )

C H 3(C H 2)5C H C H (C H 2)„C 00H

XI. OXYGENATED ACIDS

Table 4 Mid-Chain Hydroxy Acids Without Conjugated Unsaturation

Chain length and Position of Source

Source: Adapted from Ref. 1, p. 15.

CH3(CH2)m " (CH2)„C00H

R = CH3 or H R = CH3(CH 2)4 or CH3(CH 2)2

/w = 2 or 4 n = 6,8, or 10 or CH3CH2CHOH or CH3CH 2CO

furanoid acids urofuranic acids

XII. LIPID STRUCTURE

The composition of some natural phospholipids is discussed in Chapter 2. Many phospho

Palm- Cotton- Soy- Ground Sun