RUM AROMA DESCRIPTIVE ANALYSIS

A Thesis

Submitted to the Graduate School Faculty of the

Louisiana State University and
Agricultural and Mechanical College
in partial fulfillment of the
requirements of the degree of
Master of Science

in

The Department of Food Science

by

Sabina Maza Gómez

B.S., La Salle University, Mexico City, 1998
December, 2002
 ACKNOWLEDGMENTS

I would like to thank the many people that contributed one way or another to the

realization of this work. To Dr. Witoon Prinyawiwatkul, I cannot thank him enough for

his support, guidance, and example throughout the course of my studies, and for patiently

advising me. To Dr. Willem H. Kampen and Dr. Donal F. Day, for all their valuable time

and advice.

To my friends Gabriela Rosales, Denise Pallais, Sireesha Bhattiprolu, Sirisha

Sonti, Boris Castro, Guillermo Duque, Sandeep Bhale, Noemi Pavón, Patricio Paz, María

del Pilar Paz, Manuel Rodriguez, and to Dr. Michael Saska, Dr. Arthur M. Sterling, and

Mr. Hampton Stewart for numerous ho urs spent on this project, for their creative input

and sincere help. To Dr. Micheal Moody for finding support for financing this project, to

Dr. Joan King for her willingness to review and correct this thesis.

To all the people that contributed to make my life easier during this time, Ericka

Barrientos, Maria Francisca Paz, Carmen Ochoa, Fr. Rafael Juantorena: Muchas gracias!

Very special thanks to Fernando, my beloved husband, for being a true part of

this. Thank you for never giving- up and always keeping your faith in me.

All my gratitude to Sabina Gómez Villaseñor and José Manuel Maza Alvarez; for

their great testimony, for all the many sacrifices they have had and continue to make for

me, for their unconditional love and support, and for always helping me to achieve my

goals. To Fernando and Daniel Sebastian, who have challenged and inspired me in so

many ways, life would be so different without their smiles! Thank you for helping me

prove to myself.

To God, for always showing me the path, giving me hope and enlightening me.

ii
 TABLE OF CONTENTS

ACKNOWLEDGEMENTS……………………………………………….…………. …ii

LIST OF TABLES…………………………………………………………………… ....vi

LIST OF FIGURES…………………………………………………………….…….. .viii

ABSTRACT………………………………………………………..…………..…….. …ix

CHAPTER 1. INTRODUCTION………………………………..…………….…….. ….1

CHAPTER 2. LITERATURE REVIEW………………………...……………..……..….3

2.1 Sugarcane…...……………………………………..………………………….3
2.1.1 Raw Sugar Process…..….………………..……………………...….3
2.1.1.1 Harvesting..……………………..………………...…...….4
2.1.1.2 Cane Preparation and Milling………………….……...….6
2.1.1.3 Sugar Extraction…….………….………………..…… ….6
2.1.1.4 Juice Treatment and Clarification…………......………….6
2.1.1.5 Evaporation………………..….………………..……...….7
2.1.1.6 Pan Boiling and Sugar Crystalization…….…..……….….8
2.1.1.7 Baggase Use…………..…………………….….…….. ….9
2.1.2. Sugar Refining……………………………………..…………...….9
2.1.3 By-products of Sugarcane Industry…………………….………....11
2.1.4 Blackstrap Molasses…..………………….…………….………. ...12
2.2 Rum…………………………..…………………….…………...…………...13
2.2.1 Rum History………..……………………..…………………….....13
2.2.2 Rum Manufacture………………………………………………....14
2.2.2.1 Pre-treaments……..…………………………………......16
2.2.2.1.1 Clarification…………………..…………….....16
2.2.2.1.2 Dilution with Water and Nutrient Addition…...16
2.2.2.2 Fermentation Conditions……….……………..………....18
2.2.2.2.1 Yeast……...…………………………………...19
2.2.2.2.2 Fermentation Rate and Efficiency...………......22
2.2.2.2.3 Environment…………...……………………....25
2.2.2.2.4 Bacteria……………………………………......25
2.2.2.3 Centrifugation…………………………………………...26
2.2.2.4 Distillation……...…………………………………….. ...27
2.2.2.4.1 Pot-still Distillation…………………………....28
2.2.2.4.2 Continuous Distillation……………………......28
2.2.2.5 Aging…...…………………………………………….. ...30
2.2.2.5.1 Oak Wood for Maturation………………….. ...31
2.2.2.6.2 Contribution of Oak to the Aroma of Rum…....35
2.2.3 Flavor of Rum…………………………………………………......38
2.2.3.1 Higher Alcohols……………………………..………......38

iii
 2.2.3.2 Organic Acids…………………………………………...40
2.2.3.3 Esters………………………....……………………….....41
2.2.3.4 Carbonyl Compounds…………………………………...43
2.2.3.5 Acetals………………..………………………..……... ...44
2.2.3.6 Phenols……………….……………………………….....44
2.3 Rum Aroma and Taste Perception………………………………………......45
2.3.1 Odor Perception…………………….…………………......48
2.3.2 Persistence of Odors……………………………...…….....49

CHAPTER 3. DEVELOPMENT OF SENSORY DECRIPTORS FOR RUM

AROMA, FLAVOR, AND TASTE EVALUATION BY SEMI—
EXPERT JUDGES……………………………………………...…. ...52
3.1 Introduction………………..…………………………………………..…. ...52
3.2 Materials and Methods………………………………………….……..….....54
3.3 Results and Discussion...…………………………………………………....56
3.3.1 List of terms with Definitions for the Description of Rum
Aroma, Flavor, and Taste………………..……………….….….... ...56
3.3.2 Rum Evaluation Using Terms Developed by the Judges………. ...58
3.4 Significance…………………………………….………………………… ...60

CHAPTER 4. DESCRIPTIVE ANALYSIS OF THE AROMA OF RUM.…...……......61

4.1 Introduction……..…………………………………………………..……. ...61
4.2 Materials and Methods……….……………………………………...……....63
4.2.1 Panelist Screening……..…………………………………..…... ...63
4.2.2 Orientation………….………………………………………...…...66
4.2.3 Group Training Sessions……..…………………………………....67
4.2.4 Selection and Preparation of Standards and Scale Setting...….... ...76
4.2.5 Individual Training Sessions. Use of References and Scales…......89
4.2.6 Panel…………….………………………………………..…….. ...90
4.2.7 Selection of Rum Samples for Evaluation…….………………......90
4.2.8 Product Evaluation….………………………………...…………...93
4.2.9 Analysis of Data.……………………………………………….. ...95
4.3 Results and Discussion….……………………………………………..…....96
4.3.1 ANOVA Results……………………………………………...…...96
4.3.2 Overall Sample Differences………………….……………….….106
4.3.3 Principal Component Analysis….…………………………….....108
4.4 Conclusion...……………………………………………………………….110

CHAPTER 5. GENERAL CONCLUSIONS…………………………………….…...113

REFERENCES…………………………………………………………………..…….115

APPENDIX A. BALLOT FOR THE EVALUATION OF RUM USING THE LIST

OF TERMS BY SEMI-EXPERT JUDGES…….........…………..…..123

B. BALLOT FOR PANELIST SCREENING……..…………..……… .124

iv
 C. RUM AROMA DESCRIPTIVE ANALYSIS. ORIENTATION
HANDOUT...………………………………………………………. .127

D. BALLOT FOR GROUP TRAINING SESSION NO. 1 AND 2….....132

E. BALLOT FOR GROUP TRAINING SESSION NO. 3..…………....133

F. BALLOT FOR GROUP TRAINING SESSION NO. 4…………… .134

G. BALLOT FOR GROUP TRAINING SESSION NO. 5…………… .135

H. BALLOT FOR GROUP TRAINING SESSION NO. 6…………… .136

I. BALLOT FOR GROUP TRAINING SESSION NO. 7…………… .137

J. BALLOT FOR GROUP TRAINING SESSION NO. 8, 9 AND 10...138

K. BALLOT FOR THE EVALUATION OF THE AROMA OF

RUMS………………………………………………………………..139

L. GUIDELINES FOR THE EVALUATION OF RUM AROMA…….144

VITA…………………………………………………………………………………. .145

v
 LIST OF TABLES

2.1 General Composition of Sugar Cane………….………………………..………….. …4

2.2 Harvesting and Transportation Methods for Sugar Cane….……………...……….. …5

2.3 Typical Composition of Louisiana’s Cane Molasses….…………………………… ..12

2.4 Aromatic Compounds Present in Molasses………………….…………………….. ..15

2.5 Composition of the Mash After Molasses Pre-treatments....………………………. ..17

2.6 Yeasts Isolated from Molasses or Cane Juice…….………………………………... ..20

2.7 Desirable Properties of Distilling Yeast……….……………………………….….. ..21

2.8 Compounds Identified in Oak Wood….……………………………………….…... ..36

2.9 Higher Alcohols in Rum…………...………………………………………….….... ..40

2.10 Fatty Acids Present in Rum…….…………………………………………….…... ..41

2.11 Aromatic and Other Carboxylic Acids Identified in Rum..………………………. ..41

2.12 Esters of Aliphatic Monocarboxylic Acids in Rum…………...………………….. ..42

2.13 Carbonyl Compounds in Rum…….……………………………………………… ..43

2.14 Volatile Phenols Identified in Rum………..……………………………………… ..45

4.1 Preparation of References for Caramel Aroma...…………………………………... ..77

4.2 Preparation of References for Medicinal Aroma..…………………………………. ..78

4.3 Preparation of References for Butterscotch Aroma……………………..…………. ..79

4.4 Preparation of References for Honey Aroma……………………………………..... ..79

4.5 Preparation of References for Vanilla Aroma……………………………………... ..80

4.6 Preparation of References for Woody Aroma……………………………………... ..80

4.7 Preparation of References for Fruity-artificial Aroma……………………………... ..81

4.8 Preparation of References for Almond Aroma…………………………………….. ..81

vi
4.9 Preparation of References for Plastic Aroma….…………………………………… ..82

4.10 Preparation of References for Cinnamon Aroma…………………………………. ..82

4.11 Preparation of References for Butter Aroma……………………………………... ..82

4.12 Preparation of References for Ethanol Aroma……………………………………. ..83

4.13 Preparation of References for Green Apple Aroma………………………………. ..84

4.14 Preparation of References for Smoke Aroma…………………………………….. ..84

4.15 Preparation of References for Nutty Aroma……………………………………… ..85

4.16 Preparation of References for Cardboard Aroma………………………………… ..85

4.17 Preparation of References for Isopropanol Aroma……………………………….. ..86

4.18 Preparation of References for Pineapple Aroma…………………………………. ..87

4.19 Preparation of References for Pepper Aroma…………………………………….. ..87

4.20 Preparation of References for Banana Aroma……………………………………. ..88

4.21 Preparation of References for Ocean- like Aroma………………………………… ..88

4.22 Rum Samples Selected for Evaluation………..…………………………………... ..92

4.23 Intensity of the Aroma Attributes (Part I)..…..…………………………………… ..97

4.24 Intensity of the Aroma Attributes (Part II)……………….……………………..... 101

4.25 Frequency of the Perception of Some Aroma Sensations………..…..…………… 102

4.26 Multivariate Statistics and F Approximations….……………………...…………. 106

4.27 Canonical Structure r’s Describing Group Differences Among Samples……...… 107

vii
 LIST OF FIGURES

2.1 Sugar Cane Parts and Its By-products…….……………………………………….. ..11

2.2 Embdem-Mayerhof-Parnas Pathway of Glycolysis…….………………...………... ..23

2.3 Example of a 3-Distillation Column System for Rum……………………………... ..29

2.4 Reactions of Lignin Components During Storage in 60% Ethanol Solutions.….…. ..37

2.5 Pathways of Lignin Derived Compounds Formation………..…………………….. ..37

2.6 Diagram of the Formation of Higher Alcohols in Yeast Cells………….……….… ..39

2.7 Formation of Acetal…………….……………………………………………….…. ..44

4.1 Screening…..……………………………………………………………………….. ..64

4.2 Orientation Session.……..…………………………………………………………. ..67

4.3 Group Training. Development of a List of Terms for Describing Rum Aroma…… ..75

4.4 Reference Standards Preparation…………….…………………………………….. ..77

4.5 Standard References………….…………………………………………………….. ..89

4.6 Rum Samples for Evaluation….…………………………………………………… ..91

4.7 Evaluation of Rum Samples………….…………………………………………….. ..93

4.8 Plot of Prin2*Prin3…….…………………………………………………………... 108

4.9 Plot of Prin3*Prin1……………………….………………………………………... 109

viii
 ABSTRACT

Rum contains a large number of compounds that contribute to the complex aroma,

some of which present in a very small amount. This fact represents a problem for the use

of chemical analysis as a way to ascertain the quality of rum and to determine differences

in rums products. Descriptive sensory analysis techniques can be useful in such

circumstance, not only because they are easier to achieve, but because they give valuable

information on the actual perceived aroma of the products. The first part of this work

consisted of the development of lexicon for describing rums. A group of 5 semi-expert

judges evaluated a variety of 15 commercial rums from different origins, raw materials

and processing conditions, and created consens ually a list of 33 terms with description

useful for describing the aroma, flavor, and taste of rums. The second part of the work

was focused on the creation of a method used to describe the aroma of different rum

products, to discriminate among different rums, and to relate the perceived aroma of the

samples prepared from different processing protocols. Using a group of 12 trained

panelists, a modified descriptive analysis technique for the evaluation of rum aroma, and

the adequate descriptors and references that were generated, it was possible to describe

and discriminate rum samples. The results obtained from the evaluation of the 9 different

commercial rums and one experimental sample can be used to relate the perceived aroma

of the different rum samples to their processing protocol. Among the many possible

applications of this study is the use as a tool for quality control, new product

development, and brand identification. The information obtained from the description of

the products can also be a useful tool for marketing purposes.

ix
 CHAPTER 1. INTRODUCTION

Rum is defined as “any alcoholic distillate from the fermented juice of cane syrup,

sugarcane molasses or other sugarcane derivative, distilled at less than 190º proof

(whether or not such proof is further reduced prior to bottling to no less than 80º proof) in

such a manner that the distillate possesses the taste, aroma and characteristics generally

attributed to rum and includes mixtures solely of such distillates” (27 CFR 5:21).

Rum production is centered in the West Indies. Traditional rum making countries

are Jamaica, Martinique, Puerto Rico, Cuba, Barbados, Trinidad, Haiti, Guadeloupe, the

Virgin Islands, the Dominican Republic, and Guyana. Other rum producing countries

include Brazil, Peru, Mexico and parts of Asia and Africa (Lehtonen and Suomalainen,

1975).

Rum can be used for direct consumption in a pure form or as mixed drinks (Ruter,

1975). It can also be used as a flavoring agent in chocolates, liquors and in tobacco

(cigars) and bakery products. Rum contains numerous compounds that contribute to the

aroma. Some aroma compounds present in rum are higher alcohols, fatty acids, fatty acid

esters, carbonyl compounds, phenolic compounds and lactone. Esters are the most

important group of compounds in the aroma of rums. Many types of rums can be

produced from using different processing protocols. For instance, heavy aromatic rums

are generally produced in long-duration fermentations, and are distilled in pot-stills. The

presence of certain bacteria also increases the amount of congeners in rum yielding a

distillate, which upon aging yields a heavy aroma. Light rums are generally produced in

short-duration fermentations and distilled in a series of continuous columns. Rums can

1
also be classified by their ester number, which is the concentration of esters (mg) in 100

ml of ethanol (100%) (Lehtonen and Suomailanen, 1977).

The types of raw material used, the method of clarification, fermentation

conditions (rum yeast), distillation conditions, and aging conditions will affect the aroma

of the final rum product. The objectives of the present study were:

1. To develop a lexicon to describe the aroma, flavor, and taste of rum;

2. To develop a methodology suitable for the aroma descriptive analysis

of rum;

3. To describe the aroma of different rum products;

4. To relate the perceived rum aroma to the processing conditions of the

products.

2
 CHAPTER 2. LITERATURE REVIEW
2.1 Sugarcane

Sugarcane is a hybrid of several species of the genus Saccharum. The word

Saccharum seems to originate form the Sanskrit word Sarkara. Saccharum spontaneum

describes the wild cane varieties and S. officinarum L. the developed varieties used for

commercial sugar production. Saccharum officinarum L. has long been considered the

“noble” cane, as for centuries it has been the major source of commercial sugarcane. It

originates in New Guinea, from where it spread due to migration (Blackburn, 1984).

Different breeding and selection techniques have been applied to sugarcane to improve

sugar yield. Selections have been made based upon cane yield, sugar content, fiber

content, habit, ratooning, disease resistance, insect resistance, and other characteristics

such as flowering, spines, brittleness, and herbicide tolerance (Walker and Simmonds,

1984).

Sugarcane is a perennial tropical grass that produces unbranched stalks. It

consists of root, stalk (with nodes), flower, and leaves (Jones and Scard, 1921). Its stalks

can be 3-4 m tall and 5 cm in diameter. Sugar is extracted from these stalks of cane, and

it is contained in the fibro- vascular bundles. The composition of sugarcane (Saccharum

officinarum L.) varies depending on such factors as variety, soil condition, climate, and

use of fertilizers (Ruter, 1975). The general composition of sugarcane is listed in Table

2.1.

2.1.1 Raw Sugar Process

The initial steps in sugar manufacture are: (1) handling of cane including

harvesting, weighing, dispatching, and washing (if required); (2) cane preparation and

3
milling; (3) clarification of the juice; (4) evaporation and concentration; (5) sugar

crystallization; (6) drying, storage, and packaging, and (8) steam and power production

from bagasse.

Table 2.1 General Composition of Sugarcane

Millable cane:
Water 73-76%
solids 24-27%
• fiber 11-16%
• soluble solids 11-16%
Juice composition:
Sugars 75-92%
• Sucrose 70-88%
• Glucose 2-4%
• Fructose 2-4%
Salts 3.0-4.5%
§ Inorganic acids 1.5-4.5%
§ Organic acids 1-3.0%
o Carboxylic acids 1.1-3.0%
o Amino acids 0.5-2.5%
Other organic nonsugars
§ Protein 0.5-0.6%
§ Starch 0.001-0. 1%
§ Gums 0.30-0.60%
§ Waxes, fats, phosphatides 0.05-0.15%
Other 3.0-5.0%
Source: Chen and Chu (1993).

2.1.1.1 Harvesting

Sugarcane cannot be stored without deterioration, thus manufacture is done

immediately after harvesting. Sugarcane is commonly harvested during the cooler months

in each hemisphere and before harvesting cane is commonly burnt to remove unwanted

leaves (trash). Cane can be cut by hand using knifes, cutlass or machete. The advantage

of cutting cane by hand is a reduction in trash entering sugar mills. Mechanical cut of

4
cane reduces labor, therefore is a common practice in developed countries. There are

different machines employed to cut cane. Whole stalk harvesters cut the stalks at the

bottom and the top. The cane stalk is thickest at the bottom and thus contains the most

sugar, hence one wants to cut it close to the soil. The topping of the cane is to remove top

leaves (trash). The tops contain the most reducing sugars. Chop harvesters gather the

stalks as the whole stalk method, but cut them into pieces of 6 to 9 inches. Extraction fans

blow off most of the leaves. Different methods and machines have been developed for

specific regions according to their needs (such as climate) (Blackburn, 1984).

Table 2.2 Harvesting and Transportation Methods for Sugarcane

Harvest Loading into Transport Transport System

Vehicles
Manual
1. Green cane Manual Bullock-drawn carts
2. Burnt cane Mechanical Tractor –drawn trailers
Trucks
Rail trolleys

Mechanical
1. Whole sticks Mechanical Tractor-drawn trailers
2. Chopped billets Trucks
Rail trolleys
Source: Hunsigi (1993).

After harvesting, sugarcane is transported to the sugar factory by various means

(Table 2.2). The delivery vehicles are commonly weighed by electronically operated

scales. A common system is usually sampling all trucks or railcars. These samples are

analyzed for sucrose, fiber and dirt. Payment formulas include the net weight of the

delivered cane, sugar content, fiber and trash. The most common methods for unloading

cane are (1) unload of cane using an overhead moving platform, (2) tilting of trucks and

5
trailers using hydraulic platforms, and (3) chain net unloaders. Chopper harvested cane is

usually transported into bin-type containers and unloaded by tipper or dumper trucks.

2.1.1.2 Cane Preparation and Milling

Harvesting of burned cane allows for harvesting and yields of cane with much

less trash (dead leaves, weeds, etc). Cane should always be processed as soon as possible

after harvesting to avoid quality deterioration. Cane may be further washed with water to

remove clay and foreign materials. The high amounts of water used make it necessary to

reuse the water. Washing in Louisiana cut chop cane typically causes 10-16 lbs of

sucrose losses per ton. Cane preparation is required for good sugar extraction. This is

commonly accomplished using rotary knifes and/or shredders, and crushers. The goal is

to achieve a very high open crushers number in order to obtain better sugar extraction.

2.1.1.3 Sugar Extraction

Sugar extraction is achieved after crushing the mat of sugarcane in a mill tandem

of multiples stages. Addition of water during milling improves the efficiency of the sugar

extraction. The extraction efficiency of modern mills is about 92-96%. Higher extractions

will result in the extraction of more non-sugar components that will have to be removed

later on.

2.1.1.4. Juice Treatment and Clarification

The purpose of clarification is to remove the soluble and insoluble impurities

from the cane juice. Lime and heat are commonly used for this purpose. Lime is used to

neutralize the acidity in the juice by forming insoluble lime salts, mainly calcium

phosphate. The heating above boiling temperature is used to will be coagulate albumina

and other fats, waxes, and gums. The precipitate that is formed will entrap other particles

6
suspended in the juice. Once the precipitate settles and forms a mud, the juice is then

filtered, and the remaining mud can be compressed and used as fertilizers.

The expressed juice obtained after milling contains many impurities that have to

be removed. The impurities include floating solids, colloidal matter, phenols, coloring

compounds, starch, glucose, fructose, minerals, and amino acids. Then the pH of the juice

is raised from about 5.6 to 7 to prevent sugar inversion. The steps involved in the juice

treatment are (1) initial heating to 70-75°C, (2) addition of Ca(OH)2 , (3) addition of

dissolved phosphates, (4) bubbling of sulfur dioxide or calcium dioxide through juice, (5)

second-stage heating to >100°C, (6) addition of polyelectrolites to aid coagulation of

precipitates, (7) sedimentation and decantation of clear juice, and (8) sediment filtration

to separate solids and reprocessing of juice. At the end of this process clear juice will be

produced, it is a transparent, light golden liquid.

2.1.1.5 Evaporation

The clarified juice obtained has a soluble solid content (brix) of 13-15%, and it

needs to be concentrated to a syrup with about 65 % solids. This can be accomplished by

using a multiple effect evaporator. The number of effects used are normally four or five.

The larger the number of effects, the higher the steam economy. The lesser the number of

effects, the higher the evaporation capacity of each. The first body typically maintains a

steam pressure of 1-1.5 kg cm-2 and a temperature of 115-125°C, while the last body has

a vacuum of 650 mm Hg and a temperature of 55°C (Hunsigi, 1993). A feed forward

process is used so the sugar solution, upon concentration boils at a progressively lower

temperature. This avoids caramelization and other degradation reactions. Evaporators can

be of (1) short-tube construction (Roberts type), with tubes of 2000 mm long.

7
Evaporators used in the raw sugar industry are typically of the short-tube type. The

Robert or standard evaporator has short tubes and a large central down take. The external

steam heats the juice in tubes, as a result its density decreases and it starts rising. Out of

the tube flows liquid plus vapor. The latter moves overhead to the steam chest of the

following vessel, which operates at a lower pressure. The juice returns through the central

down take to the bottom of the evaporator. A portion, at the correct brix, flows forward

the next evaporator where the process repeats itself. This way, one pound of steam can

evaporate theoretically four pounds of water (3.5 practically). This technology was

invented by Robert Rillieux around 1846 in New Orleans.

Long-tube rising or falling film evaporators are also available, just like rising or

falling film plate evaporators. These have higher overall heat transfer coefficients and

evaporate more water per square foot per hour than Robert type evaporator. Additionally,

the residence time is significantly reduced, and thus there are color formation and

degradation reactions.

2.1.1.6 Pan Boiling and Sugar Crystallization

Pan boiling and sugar crystallization is an important operation in a sugar factory.

The syrup is concentrated to a super saturated stage where sugar crystallizes. This is done

in single-effect vacuum-pans (CC) and at reduced boiling temperatures. The first pure or

A- solution is usually seeded with small sugar nuclei, while less pure solutions or B- and

C- strikes (massecuite) are usually seeded with magma (pure juice + C- crystals). The A-

strikes are centrifuged and separated into A-sugar and A- molasses. The latter is used to

make a B- strike, which is allowed to undergo additional crystallization in open horizontal

crystallizers at lower temperatures, or in continuous vertical crystallizers. Next it is a

8
centrifugation step yielding B- sugar and B- molasses. The latter is used to make a C-

strike, which upon crystallization and centrifugation yields C- sugar and C- or blackstrap

molasses. A- and B- sugar form the raw sugar, which is refined into white sugar at

refineries. The final molasses is the chief raw material for rum production. During

crystallization, it is important to obtain crystals of uniform size, shape and color. The

presence of colloidal or coloring impurities trapped in the crystals should be minimized.

Crystallization is carried-out in several stages because it is impossible to crystallize all

sugar in syrup in one stage.

Commercial sugar is a disaccharide (sucrose) of two monosaccharides fructose

and glucose. The empirical formula for sugar is C12 H22O11 . It is soluble in water and

ethanol.

2.1.1.7 Bagasse Use

As mentioned before, the steam and electricity demands of a sugar factory are

generally covered by the use of bagasse as fuel. Bagasse has a calorific value of 2,200

kcal/kg on wet basis. Factories use baggase to produce high-pressure steam.

2.1.2 Sugar Refining

Raw sugar is priced according to the pol value, moisture content, ash, grain size,

dextran content and color. It is normally delivered to the refiners in bulk in truck loads, or

by rail, or by barge. The refining process of raw sugar is divided into several steps.

The first step is called affination. This process consists on the removal of a

molasses film that is occluded in the raw sugar crystals. The molasses film contains most

of the impurities in the sugar. The separation is carried out by using undersaturated syrup

9
at 75 Brix. The mixture is centrifuged and then washed with cold water. Then the washed

raw sugar is mixed with water and melted.

The next step is the defecation or clarification of washed raw liquor in which all

insoluble and colloidal matters are removed. The two types of clarification systems used

in refineries are phosphatation and carbonation. Phosphatation is accomplished with the

use of phosphoric acid and lime. It produces calcium phosphate floc difficult to filtrate.

The common way of separating calcium phosphate is by using air flotation. Carbona tion

process consists of the addition of lime and bubbling carbon dioxide to produce a calcium

carbonate precipitate in the washed raw melt liquor. The calcium carbonate crystals

entrap and adsorb other impurities. Clarification can be aided by the additio n of

quaternary ammonium compounds.

The removal of impurities that imparts color in the raw sugar is called

decolorization. Such impurities include phenolic compounds, melanoidins, caramels, and

invert degradation products. Decolorization is accomplished with the use of adsorbents in

which the aromatic structures with extensive conjugated double bond systems are

adsorbed through hydrophobic bonding. The most commonly used adsorbents are

carbonaceous adsorbents and synthetic resins. Decolorization can also be done with the

use of cationic color precipitants.

The following step is the concentration of dark sweet-waters, wash waters and

other materials with low sugar concentration not suitable for melting wash raw.

Evaporation of multiple effects is commonly used for the concentration of such materials.

After evaporation, further concentration is done in boiling pans. The next step is to

centrifuge and wash white sugar.

10
 The final step for white sugar production is the drying and conditioning. The

obtained white sugar requires drying to remove moisture. The most commonly used dryer

is the granulator. It consists of two drums in series, one for drying and the other one for

cooling.

Sugarcane Filter muds fertilizer, animal feed, cane wax

Fuel gases

Cane top and leaves

Furnace ash

Protein from mud

Bagasse As fuel Electricity, charcoal, gas production

Fibrous products Bleached pulp, fiberboard,

particleboard, writing paper.
Miscellaneous Furfural, alpha cellulose, xylitol,
plastics, poultry litter, animal feed,
bagasse concrete, soil amendment

Raw sugar Refined sugar tech. Dextran, surface-active

agents, sorbitol.
High-test molasses Protein molasses (SCP)

Molasses Direct utilization Exportation, fertilizer, animal feed

Distillery industry Rum, ethyl alcohol, rectified spirits,

anhydrous alcohol, alcohol
derivatives.
Other fermentation Vinegar and acetic acid, butanol
industries acetone, citric acid, lactic acid,
glycerol, yeast (bakers and feed).
Miscellaneous Aconitic acid, monosodium
glutamate, Dextran, L- lysine,
Xanthan gum, Itaconic acid.

Figure 2.1 Sugarcane Parts and Its By-products. Sources: Manohar Rao (1997); and
Hunsigi (1993).

11
2.1.3 By-products of Sugarcane Industry

A number of value-added products can be obtained from sugarcane in addition to

sugar. Figure 2.1 shows some of the possible by-products from the sugarcane industry.

The sugar industry can diversify to produce such product as glycerol, inositol, lactic acid,

lysine, alcohol, baker’s yeast, proteins, etc.

2.1.4 Blackstrap Molasses

Molasses is the main byproduct of the raw sugar process. It is a heavy viscous

liquid from which no further sugar can be obtained by simple means. Per ton of cane, up

to 8 gallons of molasses may be produced. Molasses contains about 50 to 60% total

sugars, of which 16-17% is sucrose, the rest being glucose and fructose. The general

composition of molasses is listed in Table 2.3

Table 2.3 Typical Composition of Louisiana’s Cane Molasses

Dry matter 80%

Total sugar 52-58%
Sucrose 16-17%
Reducing sugars 34-38%
Total nitrogen 2-3%
Organic acids 3%
Gums (insoluble in alcohol) ≤2%
pH value 5-6
Ash 12-18%

Source: ASI molasses survey (2000).

The chemical composition basically determines the quality of molasses. In

general good molasses should have a fresh and sweet aroma, and a pH value of 5-6.

Molasses should be stored cooled as otherwise self- ignition becomes possible. Certain

bacteria (exothermic) cause degradation reactions, which affect the quality of molasses.

12
Molasses can be used for animal feed, for the production of alcohol, rum, compressed

yeast, and other organic compounds such as citric acid, acetone, butanol, lactic acid,

itaconic acid, kojic acid, aconite acid and aconitates (Chen and Chu, 1993).

2.2 Rum

2.2.1 Rum History

The word rum showd up for the first time in 1612. The personal physician of the

Duke of Meckenburry (Germany) had to travel to Hamburg. Arriving late he found the

gates to the city closed. Some vendors were selling products, which he checked out. “A

Dutch woman was selling rum, which tasted quite good”, he recorded in his diary. The

rum was made from molasses imported from the West Indies. When the Dutch colonized

a portion of Brazil (the Recife area) they produced rum. When the Portuguese conquered

this area, the Dutch left and settled on different Caribbean Islands taking their “know-

how” of making rum with them.

In 1650 Father du Terre, a French Priest visited the West Indies and gave the first

description of a cane spirit. By 1722 Father Labat also described the spirit and said that

the fermentation was spontaneous and scum cakes were added to molasses during

fermentation which yielded so acid a wash that the house was called “vinagreirie”. He

also stated that distillation was accomplished by using pot-stills and several distillations

were done to produce a spirit before selling it. By this time the cane spirit was called

‘rum’, was produced in Barbados, Jamaica, Virgin Islands, and Santo Domingo, and was

exported mainly to the United Kingdom (Clutton, 1974).

The origin of the word rum is uncertain, some possible sources were (1) derivation

from the Latin word saccharum (sugar), (2) a corruption of the term ‘rombustion’

13
meaning ‘a strong liquid’, (3) from the Devonshire word ‘rumbullion’ brought to the

West Indies by British sellers, and (4) from the Spanish ‘ron’, because it is likely that

Spanish were distilling in the West Indies before the British arrived.

Even though rum was largely produced since the early 1600’s, its fist legal definition

did not come until 1909 by Sir Algernon Aspinall; ‘Rum is a spirit distilled direct from

sugarcane products, in sugar growing countries’. This definition was accepted by the

Royal Commission on Whiskey and other Potable Spirits (Clutton, 1974). The definition

of rum according to the labeling regulations of the U.S. Internal Revenue service (27

CFR 5:21) is “any alcoholic distillate from fermented juice of sugarcane syrup, sugarcane

molasses or other sugarcane products, distilled at less than 190° proof (whether or not

such proof is further reduced prior to boiling to no less than 80° proof) in such manner

that the distillate possesses the taste, aroma and characteristics generally attributed to rum

and includes mixtures solely of such distillates.”

2.2.2 Rum Manufacture

Rums can be produced from molasses, syrup and/or cane juice, with molasses being

the chief raw materials. The type of rum produced depends largely on the type of raw

materials, the treatment of the raw material, and the fermentation yeast used. Cane juice

is obtained after pressing finely ground sugarcane. For the production of rum the juice

can be used as it is obtained by pressing or after some concentration (Garnier-Larroche

and Cottrell, 1975). Rum produced in continuous stills, whether from cane juice or

molasses, typically posses a light aroma. Rum produced from long-duration

fermentations, batch distilled and aged in oak wood for several years, possess a much

heavier aroma. Molasses also contains a number of compounds that impact the aroma of

14
rum. Lehtonen and Suomalainen (1977) listed the compounds identified in sugarcane

molasses, the list includes aliphatic and aromatic esters, aldehydes, alcohols, furan

derivatives, nucleic acids, sugar alcohols, amino acids and other organic acids (Table

2.4).

Table 2.4 Aromatic Compounds Present in Molasses

Acids Esters Nitrogenous compounds
Formic acid Ethyl formate (cont.)
Acetic acid Ethyl acetate Proline
Propionic acid Isoamyl acetate Serine
n-Burtyric acid Methyl benzoate Threonine
n-Valeric acid Ethyl benzoate Tryptophan
Aconic acid Benzyl formate Tyrosine
Benzoic acid Phenethyl acetate Valine
Citric acid
Glycolic acid Ethers Phenolic compounds
Lactic acid Anisole Phenol
Malic acid Phenetole m-Cresol
Mesaconic acid Benzyl ethyl ether Guaiacol
Succinic acid Furfuryl ethyl ether Salicylic acid
Tricarballyc acid Resorcinol
Nitrogenous compounds Vainillic acid
Alcohols Acetyl pyrrole Syringic acid
Ethanol Alanine p-coumaric acid
Propanol β-Alanine Vainillin
2-Methyl-1-propanol γ- Aminobutyric acid
2-Methyl-2-butanol Asparagine Sugar alcohols
3-Methyl-1-butanol Aspartic acid D-Arabitol
Furfuryl alcohol Cystine D-Erythriol
Melissyl alcoho l Glucosamine Myo-Inositol
Pheylethyl lcohol Glutamic acid D-Mannitol
Glutamine
Carbonyl compounds Glycine Miscellaneous
Acetaldehyde Histidine 2-Acetylfuran
Furfural Homoserine 4-Methyl-2-propyl furan
5-Methylfurfural Isoleucine
Acetyl benzaldehyde Leucine
o-Methoxybenzaldehyde Lysine
Furfuryl methyl Ketone Methionine
δ- Valerolactone Phenylalanine
(-)-2-Decano-5-lactone Pipecolic acid

Source: Lehtonen and Suomalainen (1977)

15
 Compounds found in molasses as hydroxymethylfurfural, and some fatty acids

can adversely affect fermentation. They may retard or inhibit yeast fermentation

depending upon its concentration and other factors such as pH (Lehtonen and

Suomalainen, 1977). However, molasses is a rich and relatively inexpensive substrate for

fermentation.

2.2.2.1 Pre -treatments

2.2.2.1.1 Clarification

The initial pre-treatment is clarification by precipitation to remove the colloidal

matter especially calcium sulfate, which is the main component responsible for the

blockage of the stripping columns during distillation. Clarification also ensures that

fermentation and distillations processes are carried out at high efficiency. It can be

accomplished by physical and chemical effects; by heating, acidification, and cooling of

molasses. Molasses clarification consisted of diluting to 40 brix, heating to 92ºC, and

reducing the pH with sulfuric acid to promote precipitation of insoluble materials. After

clarification, insoluble impurities are removed by decantation and or

centrifugation/desludging. Clarification can also be achieved by a special distillation

method. Molasses is clarified to improve the fermentable sugar to non-sugar ratio to

approximately 6.5.

2.2.2.1.2 Dilution with Water and Nutrient Addition

Yeasts require a number of nutrients to grow. They should be provided with a

carbon source, a nitrogen source, minerals, and growth factors. Molasses possess a

number of nutrients required by yeasts; however, further additions of nutrients are

required by yeast to have an optimum fermentation. Sucrose, and further glucose and

16
fructose sugars are the main carbon source in molasses yeast fermentation. Molasses also

provide a number of amino acids as organic nitrogen containing compounds. The

presence of amino acids for rum production is important because they are used by yeasts

to produce a number of flavor active and flavor precursor compounds. However,

molasses has a low content of inorganic nitrogenous compounds, therefore ammonium

phosphate, ammonium sulfate, or urea are commonly added to molasses (Bluhm, 1983).

The minerals required by yeast for growth and fermentation are phosphorus,

potassium, sodium, calcium, sulphur, and magnesium. Cane molasses supply the

requirement of those minerals for yeast growth. The phosphorus utilized by yeast will be

mainly from phosphate. Sulphur taken by yeast is mainly from inorganic sulphate, but

can be replaced partially or totally by other inorganic or organic sulphur sources (Rose

and Harrison, 1971). Magnesium is the most common enzyme activator, particularly

important is the activation of phosphate transferases and some decarboxylases (Bowen,

1966). Even though calcium is not required for yeast growth, it stimulates growth and

fermentation (Rose and Harrison, 1971). Additionally cane molasses contain good

amount of minerals.

Table 2.5 Composition of the Mash After Molasses Pre-treatments

Density (ºBrix) 15-17

pH value 5.5-5.8

Total sugar content (g/100ml) >15

Nitrogen content (g/100ml) 2.0-2.5

Titratable acids (ml 0.1 N alkali/10 ml) 1.5-2.0

Phosphoric acid content (mg/100 ml) 600-750

17
 Other additions made to the mash prior to fermentation may be ‘dunder’ and

‘skimmings’. ‘Dunder’ is the slop from a previous distillation, and ‘skimmings’ is the

foam resulted after boiling of cane juice (Ruter, 1975) The typical composition of the

mash prior to fermentation is shown in Table 2.5 Pasteurization is another pretreatment

applied to molasses to destroy unwanted microorganisms.

2.2.2.2 Fermentation Conditions

Fermentation conditions will largely affect the quality of the rum and ethanol

produced. Fermentation is normally carried out at 32-35ºC. Lower temperatures will

result in low fermentation kinetics, and higher temperatures may destroy yeast. As an

exception to this, the temperature in which rum production by wild fermentations is done

can be as high as 37ºC (Lehtonen and Suomalainen, 1977). The pH value influences the

fermentation and the metabolic products formed. A lower pH will yield relatively higher

amounts of alcohol, while the production of a good aroma is achieved between pH 5.5-

5.8. Low pH inhibits bacteria, and therefore when using a combination of yeast and

bacteria for fermentation, the pH should not be much lower than 5 (Lehtonen and

Suomalainen, 1977). Additionally, sulfuric acid will catalyze reactions within the

fermentation such as esterification (Kampen, 1975).

Fermentation time is dependent on the quantities of yeast used in the inoculum,

the quantities of bacteria present, the type of yeast used, the fermentation temperature,

sugar concentration, and presence of nutrients. The normal fermentation time is 48-72 h

but can be longer. It has been reported than fermentation can take as long as 14 days;

however, only part of that time is necessary to complete fermentation, the rest of the time

is used for the formation of aroma (Ruter, 1975). The reaction that takes place during

18
fermentation to convert sugars to ethanol and CO2 is exothermic. Because of this effect,

cooling is normally necessary to prevent the temperature from exceeding 35ºC.

During fermentation, compounds such as sugars can move into the yeast from the

medium, and the metabolites move from the yeast to the medium, the movements to and

from the yeast cell are regulated by the membrane. Some characteristics of the

compounds such as size, branching, and lipophilic nature determine penetration through

the membrane. Short chain compounds can pass easily through the membrane (Lehtonen,

1983).

2.2.2.2.1 Yeast

The selection of the type of yeast used will affect the quality of the rum obtained.

For the production of heavy aromatic rums Schizosaccharomyces strains, fission-type top

yeasts are best suited, while the quick- fermenting budding-type Saccharomyces strains

are better for the production of light rums (Patrau, 1969). The best yeasts for rum

production are isolated from molasses (Lehtonen and Suomalainen, 1977). Fermentation

can be infected by the action of microflora present in the raw materials, equipment or air.

That is why the fermenters used for light rums are of the closed type. Some 34 species of

wild yeast have been isolated from molasses and cane juice (Table 2.6).

A successful natural- microflora fermentation occurs when aereophilic alcohol

intolerant yeasts are active at the start of the fermentation, but gradually become

dominated by fermentative yeast such as Saccharomyces.

Most modern day distilleries use proprietary cultured yeasts, some may use

recycled yeast from previous fermentations. The yeast usually used is a strain of S.

19
cerevisiae (Lehtonen and Suomalainen, 1977; Watson, 1985), although some rum

distilleries use Schizosaccharomyces pombe for the production of heavy bodied spirits.

Table 2.6 Yeasts Isolated from Molasses or Cane Juice

Candida guillermondii Saccharomyces cerevisiae

Candida intermedia var. ethanophila Saccharomyces chevalieri
Candida krusei Saccharomyces delbrucki
Candida mycoderma Saccharomyces marxianus
Candida parapsilosis var. intermedia Saccharomyces microellipsoides
Candida pseudotropicalis Saccharomyces rosei
Candida saccharum n. sp. Saccharomyces rouxii
Candida tropicalis Saccharomycodes ludgwigii
Endomyces magnusii Schizosaccharomyces pombe
Hanseula anomala Torulopsis candida
Kloeckera apiculata Torulopsis glabatira
Pichia fermentan Torulopsis glabrata
Pichia membranaefaciens Torulopsis globosa
Saccharomyces aceti Torulopsis saccharum n. sp.
Saccharomyces acidifaciens Torulopsis stellata
Saccharomyces carlsbergensis Torulopsis stellata va. Cambresieri
Saccharomyces carlsbergensis var.
alcalophila n. Var.

Source: Lehtonen and Suomalainen (1977).

Many distillers purchase commercial yeast, which is produced similarly to baker’s

yeast on molasses (Harrison and Graham, 1970; Rosen, 1989). Some indicators of the

fermentative ability of yeast are the trehalose and glycogen content (Pollock and

Holmstrom, 1951; Gadd et al, 1987; Pearce et al, 1989; Quain and Tubb, 1982). Yeast

manufacturers must design cultivation programs that lead to longer fermentative activity

in yeasts. This causes the fermentative action to occur not instantaneously (Watson,

1993).

20
 Some distillers propagate their own yeasts using stirred-tank reactors with a

substrate such as molasses. They generally operate in simple batch rather than a fed-batch

mode. The production of distiller’s yeast in closed aseptic systems to ensure inoculation

will be free of a pure yeast culture. Inoculation can also be made with a previously

finished yeast mash, therefore reducing laboratory propagation (Peppler, 1979). Table 2.7

shows the factors that have to be considered for the selection of distilling yeast.

Table 2.7 Desirable Properties of Distilling Yeast

1. Optimal conversion of carbohydrates to ethanol, e.g. by reduction of conversion

to biomass, glycerol, and other side products; increasing range of utilizable

substrates.

2. Rapid fermentation rate, generally at higher temperatures, thereby minimizing

fermentation, fermentation plant required for a particular distillation capacity.

3. High tolerance to osmotic pressure, ethanol, low pH, high temperature. All

contribute to achieve points 1 and 2.

4. Production of correct aroma compounds or their precursors.

5. High biomass yield and retention of high activity from propagation prior to

fermentation.

6. Flocculation, in cases of use of particular fermentation systems.

Source: Johnston (1990).

21
 Another option for the production of distilled beverages is to use mixed cultures.

By this means, a richer flavor can be obtained, with the down side of having more

problems controlling fermentations.

2.2.2.2.2 Fermentation Rate and Efficiency

It is important to maximize the ethanol production while maintaining the

desired product quality. The yeast used should be able to ferment the sugar at a fast rate.

Most commercial strains of S. cerevisiae can utilize a number of sugars. Glucose,

fructose galactose, maltose and maltotriose are transported across the cell membrane

(Barnett, 1976), while sucrose is extracellularly converted to glucose plus fructose by β-

D-fructofuranosidase (invertase). Since sugarcane molasses has a high sucrose content,

good yeasts for rum production should have high invertase activity (Watson, 1985). In

anaerobic conditions glucose and fructose are transported across the plasmalemma by

constitutive facilitated-diffusion systems, galactose by inducible facilitated diffusion, and

maltose and, in some strains, maltotetrose by inducible permease systems (Barnett, 1976)

Hexose monosaccharides are then metabolized by the Emdben-Mayerhof-Parnas pathway

(Figure 2.2) to pyruvate (Gancedo and Serrano, 1989). The pyruvate not required for

yeast growth is then converted to acetaldehyde and ethanol, regenerating NAD+ from

NADH, and thus maintaining the redox balance in the cell (Van Dijken and Scheffers,

1986). Other products of the EMP pathway are succinate and glycerol, and possibly lactic

acid. These products remain mainly in the stillage after distillation. Glycerol is formed in

order to regenerate NAD from the NADH formed in yeast growth reactions. It is believed

that succinate is formed from oxaloacetate via malate and fumarate as a substrate for

biosynthetic reactions in the tricarboxylic acid cycle (Gancedo and Serrano, 1989).

22
 HOCH2 O CH2OH

H HO Fructose
H OH

HO H

CH2 OH CH2 O—P

O O P—OCH2 O CH2 OH Fructose-1,6-diphosphate
H H H H P—OCH2 O CH2 O—P
ATP H HO ATP
OH H ADP (1) OH H H OH ADP (3) H HO
HO OH HO O (2) H OH
HO H
H OH H OH HO H
Glucose Glucose-6-phosphate Fructose-6-phosphate
CH2O—P
¦
C- O
¦
CH2OH
Dihydroxyacetone phosphate
O
2-Phosphogly ceric acid ¦
COOH COOH ATP ADP C - O ~P NADH 2 NAD P CHO
¦ ¦ ¦ ¦
H- C- O- P H- C- OH H— C—OH H—C—OH
¦ (7) ¦ (6) ¦ (5) ¦
CH2 OH CH2 O- P CH2 O- P CH2 O- P
3-Phosphoglyceric acid 1,3 Diphosp hoglyceric acid Glyceraldehyde-3-phosphate

H—OH
(8)

Phosphoenol- COOH ATP COOH CO2 CHO NADH2 CH2OH

¦ ADP ¦ ¦ ¦
pyruvic C—O~P C- O (10) CH3 (11) NAD CH3
¦ (9) ¦
acid CH2 CH3 Acetaldehyde Ethanol
Pyruvic acid

NADH2 (12) Oxidative pathways

NAD

COOH
¦
H- C- OH Lactic
¦ acid
CH3

The enzymes involved in the reactions are: (1), glucokinase; (2), phosphoehexoisomerase; (3),
phosphofructokinase; (4), aldolase; (5), glyceraldehydes-3-phosphate dehydrogenase; (6),
phosphoglycerokinase; (7), phosphoglyceromutase; (8), enolase; (9), pyruvic kinase; (10), pyruvic
carboxylase; (11) alcohol dehydrogenase; and (12), lactic dehydrogenase.

Figure 2.2 Embden-Mayerhof-Parnas Pathway of Glycolysis.

23
The overall fermentation effect is:

Glucose + 2 ADP à 2 Ethanol + 2 CO2 + 2ATP

Theoretical fermentation yield from 1 g glucose is 0.51g ethanol and 0.49 g CO2 .

In practice 10% of the glucose is utilized for biomass production, thus the ethanol and

CO2 yield is 90% of that theoretical value. ATP formed will be used to supply cell energy

requirements.

Saccharomyces cerevisiae preferentially metabolizes glucose to fructose. Many

factors affect yeast fermentation such as infection, available sugars, and genetic

complement of the strain. Other environmental factors and the composition of the

substrate also affect the fermentation rate. Lack of essential nutrients, especially available

nitrogen, will limit cell growth, and thus slow fermentatio ns. Low pH values, low

temperatures, high contents of sulfur dioxide, minerals and certain fatty acids, and the

presence of agricultural fungicides will inhibit fermentation rates (Jones et al., 1986;

Maiorella et al., 1984; Rose, 1987; Cantarelli, 1989; Laure and Lafon- Lafourcade, 1989).

Yeast inoculation as well as physiological health of the yeast will also affect the initial

fermentation rate (Watson, 1984 and 1985). Aerobically grown yeast tends to ferment

more rapidly and survive unfavorable cond itions better than anaerobically grown yeasts

because of its higher reserves of cell constituents and ATP, and higher contents of sterols

and unsaturated fatty-acyl residues (Rose, 1977; Korhola et al., 1989; Rosen, 1989).

Some of the techniques used to improve ethanol yield are vaccum, rapid and continuous

24
fermentation or immobilization of yeast, also low cost supplements can be used in

conventional fermentation to improve the alcohol yield.

2.2.2.2.3 Environment

In the initial stage of the fermentation, with high sugar concentrations, the yeast is

subjected to high osmotic pressure, while at the end of the fermentation, the yeast is

subjected to high alcohol concentrations as well as elevated temperatures due to

exothermic reactions. Molasses fermentations generally have high osmotic pressures,

under these conditions, strains of Schiz. pombe are more tolerant than S. cerevisiae

(Haraldson and Bjorling, 1981). The normal response of yeast to osmotic pressure is to

produce glycerol, to work as osmoregulator (Meikle et al., 1988). The glycerol produced

is leached out of the cell once the osmotic pressure drops (Panchal and Stuart, 1980).

In the presence of high ethanolic concentrations, especially under high

temperatures, yeast cells may die, and can disrupt fermentation causing inefficient sugar

conversion to ethanol (Casey and Ingledew, 1986; Pamment, 1989; van Uden, 1989).

High alcohol concentrations also affect the fermentation by inhibiting and denaturing

enzymes in yeast (Scopes, 1989). Yeast strains rich in sterols and double-unsaturated

fatty-acyl residues in the plasma membrane have shown better resistance to the toxic

effects of alcohol (Thomas and Rose, 1979; Casey and Ingledew, 1986).

2.2.2.2.4 Bacteria

In addition to the yeast used for fermentation, the mash may introduce significant

levels of bacteria (Sharp and Watson, 1979). They may influence the fermentation

characteristics and contribute to the sensory profiles of the final spirit.

25
 The presence of bacteria during fermentation normally comes from raw materials

(cane juice, molasses or process water). The bacteria will have metabolic activity parallel

to yeast fermentation and ethanol production, affecting the kinetics and biochemistry as

well as the sensory properties of rums (Fahrasmane and Ganou-Parfait, 1998). The

presence of bacteria can be controlled by using thermal treatment, acidification,

antibiotics, and by the fermentation yeast. The degree of infection during fermentation is

normally kept well under 10% [(live yeast cells/live bacteria) X 100). Some of the

compounds produced by bacteria parallel to alcohol fermentation by yeast is chiefly lactic

acid, acrolein. They will not only affect the sensory characteristics, but they may add

undesirable toxicity (Fahrasmane and Ganou-Parfait, 1998).

Some bacteria strains contribute favorably to the fermentation. The presence of

Clostridium saccharobutyricum in the fermentation is found to accelerate fermentation

due to an increased production of fatty acids and further corresponding esters (Arroyo,

1945). Clostridium butyricucum and Clostridium kluyvery can produce unsaturated fatty

acids as metabolites (Scheuerbrant and Bloch, 1962). The ideal ratio of bacteria to yeast

to produce the best rum yield is 1:5 (Arroyo, 1945).

2.2.2.3 Centrifugation

After fermentation the mash is allowed to settle for a few hours so that the yeast

cells and other solids can precipitate. Centrifugation helps remove all the insoluble solids.

Centrifugation is normally done with desludger (disk/nossle) centrifuges. Centrifugation

also removes yeast cells that otherwise, at the prevailing high bottom temperatures in the

stripping column (215 +°F), will decompose and may cause off-odors in the distillate.

26
2.2.2.4 Distillation

Separation of compounds can be achieved by creating various zones that differ in

the conditions such as temperature, pressure, composition, and or phase state. Each

compound reacts to different conditions, and therefore each compound establishes a

different concentration at different zones. Distillation is a method of separation that uses

the vapor and liquid phases at the same temperature and pressure for all the coexisting

zones. These various zones can be created using packing or plates or trays that will put

the two phases vapor and liquid in contact. These trays are stacked one over another in a

column. After the flow material is introduced, the liquid, because of gravity will flow

down, while the vapor flow up the column. The liquid is then reheated to vaporize after

reaching the bottom of the column. The liquid that remained in the bottom is then

removed. In the case of rum distillation, the remaining liquid after distillation is called

“stillage” or “dunder”. The vapor reaching the top of the column is cooled and

condensed; it will then be removed as distillate, while some may go back as reflux.

During distillation, the lighter components (light boiling) tend to concentrate in

the vapor phase, while the heavier (high-boiling) components will tend to remain in the

liquid phase. Separation depends on the relative volatility of the components, the number

of trays, and the ratio of the liquid-phase flow rate to vapor-phase flow rate (Perry et al.,

1984). In the cases where feed is introduced at a single point, the upper section of the

column will be the rectifying sections, and the lower part the stripping section.

The distillation step can have a major effect on the aroma composition of the final

rum product. Distillation can affect the composition by formation of new compounds

such as products of pyrolysis. Another way distillation affects the aroma is by altering the

27
proportion of the compounds formed during fermentation. Also the presence of yeast

during distillation can alter the ester content of the alcoholic beverage (Lehtonen and

Jounela-Eriksson, 1983).

Rum can be distilled in batch (pot-still) and continuous distillation systems. The

selection of the distillation procedure will depend on the type of product wanted (Ruter,

1975). Rum obtained by pot-still distillation is generally heavier and is mainly produced

in English and French speaking areas of the West Indies. Continuous distillation produces

lighter rum and is largely used for Cuban and Puerto Rican rums (Lehtonen and

Suomalainen, 1977).

2.2.2.4.1 Pot-still Distillation

The simplest form of batch distillation consists of a single pot, a condenser, and

one or more receiving tanks. In this case, no trays or packing are provided. The feed is

charged into the pot and the mash is then brought to boiling. The vapors are condensed

and collected in a receiver. Another variation of batch distillation contains pot stills in

series. The fermented mash is pumped into the first vessel and distilled. The condensed

distillate or “low wine” is collected in a second vessel. The “low wines” are subjected to

distillation to obtain “high wines”. “High wines” are rectified in a short rectifying

column. “Dunder” or stillage remaining from distillation and fermented by the action of

bacteria can be added back to the fermented mash prior to distillation.

2.2.2.4.2 Continuous Distillation

Continuous distillation normally consists of three columns (Figure 2.3). The

purpose of the first column or ‘stripping column’ is to remove all the alcohol from the

fermented mash. The overhead product is about 40-60 % alcohol by volume. At the head

28
of the second column or “purifying column” all the volatiles or low-boiling impurities

are taken off, while the water/alcohol/congeners mixture purified will be taken from the

bottom to the rectifying column. In the third or “rectifying column”, the alcohol is

concentrated and the remaining heavy compounds or fusel oils are removed. The upper

trays of the column will contain the rum distillate while the above trays will contain fusel

oils. The feed tray will contain (~70%) ethanol. This procedure can be modified by

feeding water at the head of the purifying column and make it work as an extractive

column (Unger and Cofey, 1975).

CWR

reflux CWS

CWR
CWS

Purifying head CWR

CWS
reflux
column

steam
T Rum
distillate
Rectifying
Stripping column
T
column ν

steam

Fusel oil

steamν ν
stillage

Figure 2.3 Example of a 3-Distillation Column System for Rum.

The distinctive taste from most flavored spirits comes from the acquired flavor

rather than the base spirit, which is commonly a neutral spirit (Watson, 1985). Early

29
continuous still consisted of two columns. Modern stills contain six inter-connected

columns. Continuous still can distill some 3,000 to 4,000 gallons per hour. The six

columns of modern continuous still are: (1) The purifying column, which is a column

filled with some 16 trays with hooded rectangular bubble-caps. The heated mash is fed at

the top of the column, steam is introduced with a sparge from the bottom. As the mash

flows down the trays, CO2 , other gases and aldehydes are removed for further

condensation. The purified mash is removed from the bottom to the top plate of the

second column; (2) The analyzing column is the second, and consists of 16-20 plates of

grater diameter. The mash is fed from the top and flows down while the steam flows up,

driving ethanol upwards. The mash is removed from the bottom with an ethanol content

of 0.01%. The obtained liquid is known as stillage; (3) The aldehydes column can be a

separate column or located at the top of the analyzing column. It concentrates aldehydes

from the purifying column and returns ethanol to process; (4) The rectifying column is

where the ethanol vapors are purified and concentrated. It removes congeners or

impurities according to the distiller’s requirements; (5) The preconcentrating column is

used for neutral spirit production to remove impurities, mainly aldehydes before passing

to analyzing or rectifying column. 6) The hydroselection column distills the mixture of

the preconcentrating column with pure hot water to remove all the impurities from the

alcohol. This produces an extra fine neutral spirit (L’Anson, 1976).

2.2.2.5 Aging

During maturation a series of many favorable changes occur steadily including

removing harsh aromas in the rum distillates and developing a pleasant aroma and taste.

It is necessary to have an initial distillate with good amounts of congener precursors in

30
order to have the opportunity of a series of chemical reactions to occur. This reactions

will determine the progress of the aging process.

Rum is typically aged in oak barrels for at least 2 years, where many changes

occur. According to Nishimura et al (1983), during maturation in oak casks, rum can

overcome several changes. Those changes may be due to:

1. Wood components extraction;

2. Reaction of wood or wood components with the distillate;

3. Reactions within the aged distillate such as oxidation and condensation;

Some of the reactions that occur within the rum distillates are: 1) esterification,

which requires the presence of acids and ethanol and other high alcohols; 2)

polymerization and condensation, which requires aldehydes; and 3) formation of new

acids and alcohols; which requires the presence of esters (Vargas Guzman, 1975). The

amounts of acids increases rapidly during the first two years reaching concentrations up

to ten times higher than the initial; however the increase in esters occurs at a slower rate.

Higher alcohols will increase about 15-20% after 6 years, and aldehydes will be double

the initial amounts during the same period of time (Vargas Guzman, 1975).

2.2.2.5.1 Oak Wood for Maturation

Although many woods have been used for spirit maturation, only oaks are

important as maturation barrels (Singleton, 1974). Some of the qualities that make oaks

suitable for liquid containers are the presence of multiseriate rays, tyloses, durable, tough,

bendable wood with high extract to inhibit rotting organisms. The oaks used for

cooperage are white oaks from the Leucobalanus or Lepidobalanus subgroups. European

species are, Quercus rubor, and Q. petraea, in North America there are many species but

31
the most important one is Quercus alba. Red oaks are not suitable for cooperage due to

the poor tylose formation. Tyloses are spheres of 28% lignin that expand from the cells to

plug the sprinwood pores of the heartwood, this prevents the liquid from leaking out of

the barrel. Q. alba is more fully plugged than the European species. Some other

differences from American and European oaks are that European oaks have more

extractable phenols, while American oaks have more aromatic compounds. In general the

components of different oak species are similar, varying in concentration and relative

amounts. Some conditions that may influence the composition of the oaks, and the

maturation that can be achieved with the barrels are tree age, growing rate, and such.

Older trees have larger medullary rays which are barriers to the migration of

liquids. This may be a reason why old, larger trees are better for staves. Wood chemical

and physical composition varies between and within trees. The youngest heartwood,

nearest to the sapwood has the highest and most diverse content of extractable

components. As the tree becomes older, the heartwood near the pith diminishes in

extractable and phenol compounds. Wood of slow grown trees is easier to work and bend,

and has greater extracts, this is why coopers prefer it to faster growing wood. European

staves have higher extractable solids (161%) than American oak, and have more phenol

per unit (154%) (Singleton, 1974). The general composition of dry heartwood is 50%

cellulose, 20% hemicellulose, and 30% lignin. Cellulose is the framework of the wood,

hemicellulose the matrix, and lignin the encrustant. Cellulose is a polymer of glucose,

hemicellulose is an heterogeneous polymer that includes xylose and other sugars,

especially pentoses. Lignin in oak is a three-dimensional polymer of phenylpropane

derivatives of guaical and syringyl units substituted in the four th position with the

32
aliphatic side chain and cross- linked by oxidation. Some other compounds such as

tannins and carbohydrates are attached to lignin (Rowell, 1984). In sufficient quantity,

the extractives of the wood compounds affect the flavor of the final beverage. Water and

ethanol extracts of oak wood contain color, carbohydrate derivatives, gallo- and

ellagitannins, lignin fragments, together with their precursors and degradation products

(Singleton, 1995). These compounds are flavorful and modify the aroma of the beverage

during maturation. The soluble extractives of oak are depleted readily from the layer

extracted, while the major structural polymers are broken down, if at all, during the

process of maturation (Singleton, 1995).

The ideal ethanol concentration that yields the maximum solid extraction is

approximately 55% (v/v) (Singleton and Draper, 1961). For the production of lighter

products, lower proofs can be used yielding a product with less extractives and having a

different composition. A barrel for 200 L has as much as 90 cm2 /L of wood surface.

Evaporation, extraction, oxidation and component reactions are maturation effects related

to the conditions of the barrels, and they will increase as more wood surface is in contact

with a unit of beverage (Singleton, 1995). Each mm penetration of the beverage to the

wood will contribute some 9 cm3 of extracted oak or 5.4 grams of soluble oak solids,

assuming a density of 0.6 g/cm3 , and some 10X flavor threshold (Singleton, 1995).

Once maturation results suffice, a common practice is to reuse or refill the barrels.

After refilling, barrels never seem to lose their physical effects, and never seem to be

depleted, even when they are not shaved off. This may be due to the faking and cracking

that exposed some new wood (Singleton, 1995).

33
 Some consideration has to be given to cooperage as the rays have to be tangential

to the barrels circumference. The staves should not be dried too rapidly, in order to

prevent checking. The ideal moisture content of the staves for cooperage is 8% to 10 %

(Rassmussen and McMillen 1956). According to Francis et al (1992), woods that are

seasoned in warmer, drier climates possess enhanced vanilla, buttery, caramel, and cedar

notes. These notes plus nutty, and a decreased raisiny note are present in wood that is

thermally treated at 105°C for 24 h. Other considerations are that due to tree-to-tree

differences, the staves used for a single barrel should be randomly selected from several

trees. This will decrease the barrel-to-barrel variation (Singleton, 1995).

Charring and toasting of the barrel wood produce various effects in the extractives

which are transferred to the beverage during maturation. Charring is a more radical

treatment and produces mainly carbon and some pyrolysis products as furfural and HMF.

It may transfer some undesirable smoky flavor to the beverage. Toasting is a less drastic

treatment and can be accomplished to different degrees and depths (Singleton, 1995).

High temperature treatments cause a reduction in hemicellulose and increased furfural

and HMF derivatives unless the heat treatment is so severe that the compounds

polymerize (Hillis, 1984).

During maturation water and ethanol losses occur in the barreled spirits. About

2% to 7% loss in volume can be exp ected per year during maturation in standard barrels

in the tropics. Many factors such as container size, temperature, relative humidity, and air

circulation influence evaporation, other factors include water’s liquid state, heat of

vaporization, self association, adsorption to barrel carbohydrates and vapor pressure

(Singleton, 1995). Given a certain temperature higher humidity lowers the water loss rate.

34
Ethanol is also lost during maturation. For a constant ethanol concentration in the

beverage, the relative humidity should be between 65% and 70%. While water and

ethanol are lost, the content of non-volatile compounds increases during maturation.

2.2.2.5.2 Contribution of Oak to the Aroma of Rum

Some of the components found in oak wood that contribute to the aroma of spirits

are 4-methyl-γ-octalactone, γ-nonalactone, and eugenol. Table 2.8 lists compounds

identified in oak wood. Acetic acid increases during maturation in oak casks, it is a main

component of pyrolegneous acid, and it derives from the acetyl group of hemicellulose by

degradation (Nishimura et al., 1983).

The products of lignin degradation obtained from charred oak are much larger

than those obtained from uncharred oak. These components include extracts, with

aromatic aldehydes such as vanillin, syringaldehyde, and sinapaldehyde. The higher the

toasting temperature, the higher the amounts found. Aromatic aldehydes are directly

extracted, while vanillin and syringaldehyde are products of the oxidation of

coniferaldehyde and sinapaldehyde (Nishimura, 1983). Nishimura et al (1983), found that

the ethanolysis of lignin yields more vanillic acid and syrinaldehyde, and less

sinapaldehyde and coniferaldehyde than oak chips. Native lignin of low molecular weight

yields about the same amounts than with oak chips. However, due to the relative low

amounts of native lignin in wood, they suggested that this is not the main source of

lignin-derived compounds. Later they studied oxidation and esterification of these

aromatics, and suggested that the lignin-derived compounds are formed by the following

pathways (Figure 2.4, and 2.5).

35
Table 2.8 Compounds Identified in Oak Wood.

Aliphatic hydrocarbons Aromatic hydrocarbons Other aromatic

Tetradecane Naphthalene compounds
Pentadecane α-methyl naphthalene 1-indanone
Hexadecane β-methyl naphthalene benzothiazole
Heptadecane α-ethyl naphthalene methyl salicylate
Octadecane β-ethyl naphthalene benzoic acid
Nonadecane dimethyl naphthalenes phenyl acetic acid
Eicosane trimethyl naphthalenes cinnamic acid
tetramethyl naphthalenes
Aliphatic acids biphenyl Furan compounds
Acetic acid acenaphthene Dibenzofuran
n-butyric acid acenaphthylene 2-furoic acid
i-butyric acid 1,1-6-trimethyl-1-1,2- 3-furoic acid
n-valeric acid dihydronaphthlene
i- valeric acid fluorine Terpene compounds
caproic acid 1,2-dmethyl-4-allyl α-Muurolene
heptanoic acid benzene γ-muurolene
caprylic acid β-bisabolene
nonenoic acid Phenols α-cadiene
nonanoic acid Phenol γ-cadiene
caprilic acid guaiacol δ-cadiene
decanoic acid o-cresol δ2 -cadiene
undecanoic acid p-cresol α-curcumene
lauric acid p-ethyl phenol calamene
myristic acid 4-methyl guaiacol α-calacorene
tetradecanoic acid eugenol
pentadecanoic acid cadalene
i-eugenol terpineol
pentadecenoic acid chavicol borneol
pentadecadienoic acid syringol
palmitic acid myrtenol
4-methyl syringol elmol
4-ethyl syringol epi-cubenol
Other aliphatic 4-allyl syringol
compounds β-eudesmol
vainillin
Cis- 4-methyl-γ-octalactone α-eudesmol
propiovainillone
Trans-4- methyl-γ- γ-eudesmol
octalactone α-cadinol
Other aromatic
γ-nonalactone T-cadinol
compounds
γ-decalactone Benzyl alcohol Verbonene
Geranyl acetate
1,1-dimethoxynonane phenethyl alcohol
1,1-dimethoxydecane phenethyl acetate
β-Ionone acetopehenone

Source: Nishimura et al., (1983).

36
 Other effects of maturation on the composition of beverages are an increase in

acetaldehyde, derived from ethanol + acetic acid. The already minor quantities of sulfur

compounds, such as dimethyl sulfide, dimethyl disulfide, methionyl acetate, ethyl

methionate, and dihydro-2-methyl-3(h)-thiophane decreased with aging. Toasting of oak

wood produced other non- lignin volatile compounds such as furfural, 2-methylfurfural,

guaiacol, and 4- methyl guaiacol (Nishimura et al., 1983).

Acetovainillone Propiovainillone
Vainillin Vainillic acid Ethyl vainillate

Coniferaldehyde Ferrulic acid

Sinapaldehyde Syringaldehyde Syringic acid Ethyl syringate

Figure 2.4 Reactions of Lignin Components During Storage in 60% Ethanol Solutions.
(Nishimura et al, 1983).

Vainillin+
Vainillin+ Vainillin +
Extraction Syrinaldehyde +++ Oxidation Syringaldehyde ++++
Syringaldehyde +++ Coniferaldehyde +
Coniferaldehyde +++ Coniferaldehyde ++
Esterification Sinapaldehyde ++
Vainillic acid + Sinapaldehyde Vainillic acid +
Syringic acid Vainillic acid Ethyl vainillate
B D
Acetovainillone Ethyl syringate
Other phenols Other

O2 Charring A Ethanolysis C
Vainillin +++
Syrinaldehyde +++
Ethanol lignin Coniferaldehyde +
Lignin Native lignin Sinapaldehyde +
Vainillic acid ++
Others

Figure 2.5 Pathways for Lignin Derived Compounds Formation (Nishimura et al., 1983).

37
2.2.3 Flavor of Rum

The flavor of alcoholic beverages is composed of various volatile and non-volatile

organic compounds. These compounds give the typical odor and taste to the beverage

(Lehtonen and Jounela-Eriksson, 1983). Many of these compounds have been identified,

and can be classified in several groups according to their chemical nature. Most

compounds responsible for the aroma of distilled beverages are volatiles, and the typical

flavor and chemical composition will be closely related to the manufacturing process

used (Nykänen and Nykänen, 1991).

The fermentation stage is chiefly responsible for the basic aroma formation. The

yeast metabolism is affected by the fermentation conditions. Distillation can largely

affect the proportion of compounds recovered. Maturation influences the aroma of rum

by adding new compounds formed in chemical reactions, while other compounds are

condensed. The final rum product will be related to all the production stages.

2.2.3.1 Higher Alcohols

Higher alcohols are the most abundant aroma compounds in rum (Nykänen and

Suomalainen, 1983). Higher alcohols such as n-propanol, and isobutanol are formed from

their correspondent keto-acid, following the way ethanol is converted from pyruvate.

Keto acids are formed from amino acids by transamination reactions, and then to alcohols

(Watson, 1985). Strongly smelling aliphatic alcohols, such as 1-propanol, 2- methyl-1-

propanol, 2- methyl-1-butanol, 3-methyl-1-butanol, and the aromatic alcohol phenethyl

alcohol, may be formed from sugars by an anabolic process via the pathways the amino

acids are synthesized. This formation seems to be suppressed in a medium with

abundance of assymilable nitrogen (Nykänen, and Nykänen 1983). 3-Methyl-1-butanol

38
is the most abundant fusel oil in rum, followed by 2- methyl-1-propanol (Lehtonen and

Jounela-Eriksson, 1983). Ter Heide et al. (1981) determined the presence of the

homologous series of 1-alkanols from methanol to decanol in rum. The choice of yeast

strain seems to be the main factor to control the formation of 1-alkanols (Parfait and

Jouret, 1975), and the nature and quantity of the nitrogen source may also play a role

(Parfait and Sabin, 1975). A fermentation mash low in inorganic nitrogen, and especially

at high temperatures yields high fusel oil yields due to aminoacid deamination and

decarboxylation (Figure 2.6).

AMINOACIDS
ALA,VAL, LEU ,
ILE, PHE, TYR SUGARS
YEAST CELL WALL

BIOSYNTHESIS PROTEIN
REGULATION

KETO ACID INTERNAL

AMINOACID
POOL (S) POOL

DECARBOXYLATION
AND REDUCTION

HIGHER ALCOHOLS

Figure 2.6 Diagram of the Formation of Higher Alcohols in Yeast Cells (Meilgard, 1975).

39
 The distillation procedure is known to have an effect on the content of higher

alcohols in rum. According to I’Ansosn (1971), larger amounts of higher alcohols were

found in rums distilled in pot-stills, than in continuous stills. Some higher alcohols found

in rum are included in Table 2.9.

Table 2.9 Higher Alcohols in Rum

3-methyl-1-butanol (isoamyl alcohol) 2-methyl-1-butanol

2-methyl-1-propanol (isobutanol) Propanol
2-butanol Isopentanol
1-propanol

Source: (Nykänen and Suomailanen, 1983).

The presence of 2-butanol appears to be related to the action of bacteria during

fermentation (Nykänen and Nykänen 1983). Aliphatic long-chain alcohols have hardly

any effect on the aroma of rum (Nykänen and Nykänen 1983).

2.2.3.2 Organic Acids

Organic acids such as acetic and even-chained fatty acids are produced by strains

of Sacch. cerevisiae through the fatty-acid synthesis pathway (Berry and Watson, 1987).

Acids are important precursors of esters as well as being flavor-active by themselves

(Berry and Watson, 1987). Fatty acids are produced during fermentation, and are easily

transferred to the distillate during distillation. The total acid content in heavy bodied rums

is about 100 to 600 mg per liter, with acetic acid being the predominant volatile acid.

Heavy-bodied rums contain more volatile acids than light rums (Nykänen, and Nykänen

1983). The amount of acids was found to increase during maturation. The increase in

acids was total, fixed and volatile acidity (Lehtonen and Jounela-Eriksson, 1983). Table

40
2.10 shows some of the fatty acids present in rums. Acids extracted with alcohol from

oak casks include salicylic acid, 4-hydoxy-cinnamic acid, gallic acid and chlorogenic

acid (Nykänen and Suomalainen, 1983). Table 2.11 includes other acids that may be

identified in rum.

Table 2.10 Fatty Acids Present in Rum

Short-chain fatty acids Long-chain fatty acids

Acetic acid Octanoic acid
Propionic acid Decanoic acid
Isobutyric acid Lauric acid
Butyric acid Myristic acid
Acrylic acid Palmitoleic acid
Isovaleric acid Palmitic acid
Valeric acid Linoleic acid
2-ethyl-3-methyl Oleic acid
butyric acid Stearic acid
Hexanoic acid
Heptanoic acid

Source: Nykänen and Nykänen (1983).

Table 2.11 Aromatic and other Carboxylic Acids Identified in Rum

Benzoic acid Gallic acid

Salicilic acid Vainillic acid
p-Hydroxybenzoic acid p-Hydroxycinnamic acid
Gentisic acid 4-Hydroxy-3,5-dimethoxy-benzoic acid
o-Pyrocatechuic acid 4-Hydroxy-3-methoxy-cinnamic acid

Source: Nykäne n and Suomalainen (1983).

2.2.3.3 Esters

Esters are the key compounds in rum because of their contribution to the aroma.

Esters are responsible for the presence of fruity- like aroma in rums. The main esters

formed in rums (Table 2.12) are acetate-esters of alcohols (e.g. ethyl acetate), and ethyl

41
esters of fatty acids (e.g., ethyl decanoate). They are mainly from acid:alcohol esters, but

there is a proposed mechanism of formation from enzymatic reaction between free

alcohols and acetyl CoA derivatives of fatty acids (Watson, 1993)

The ester content of rum varies from 4 to 64% w/v, with ethyl acetate typically

being the predominant ester. The ester content in rum is related to many factors. The

esters formed during fermentation are related to the distribution of esters between the

yeast cells and in the medium, and to factors affecting ester concentration. When yeast is

present during distillation, the content of long-chain carboxylic acid esters derived from

the yeast cells increases. The ester content of the beverage of rum also depends on the

yeast.

Table 2.12 Esters of Aliphatic Monocarboxylic Acids in Rum

Ethyl formate Ethyl valerate Ethyl undecanoate

Methyl acetate Hexyl acetate Isobutyl dodecanoate
Ethyl acetate Isopentyl propionate Isobutyl duodecenoate
Isobutyl formate Isobutyl butyrate Ethyl laurate
Propyl acetate Ethyl hexanoate Isopentyl decanoate
Ethyl propionate Ethyl heptenoate Ethyl mysistate
Isopentyl formate Isopentyl butyrate Isopentyl laurate
Isobutyl acetate Ethyl heptanoate Ethyl pentadecanoate
Butyl acetate Methyl octanoate Methyl palmiatete
Sec-Butyl acetate Phenethyl acetate Phenetyl decanoate
Propyl propionate Isopentyl valerate Ethyl-9-hexadecenoate
Ethyl isobutyrate Ethyl octanoate Ethyl palmitate
Ethyl butyrate 2-Methylbutyl hexanoate Propyl palmitate
Isopentyl acetate Isopentyl hexanoate Ethyl linoleate
Isobutyl propionate Ethyl nonanoate Ethyl oleate
Propyl butyrate Methyl decanoate Ethyl stearate
Ethyl 2-methylbutyrate Ethyl decanoate Isopentyl palmitate
Ethyl isovalerate Isopentyl octanoate

Source: Nykänen and Suomalainen, (1983).

42
 Temperature affects the rate of ester formation in the medium (Nykänen, and

Nykänen 1983). 2-ethyl-3-methylbutyric acid has been determined in rum and is believed

to come from 3- methylbutyric acid, a compound found in molasses (Lehtonen et al.,

1977).

During distillation the middle aroma fraction consists of esters from ethyl

hexanoate to ethyl laurate, esters of long-chained carboxylic acid with the acid part

longer than 12 carbon atoms belong to the heavy aroma fraction of rum (Nykänen and

Nykänen, 1983). Table 2.12 includes some of the most common esters of aliphatic

monocarboxylic acids that have been identified in rum.

2.2.3.4 Carbonyl Compounds

It is believed that aliphatic carbonyl compounds are related to aroma nuances in

distilled alcoholic beverages, especially when present in large amounts (Nykänen and

Nykänen, 1983). Aldehydes can total up to 5 to 9% w/v in Jamaican, Puerto Rican, and

Martinique rum, acetaldehyde being the predominant compound (Nykänen and Nykänen,

1983). Some of carbonyl compounds present in rum are shown in Table 2.13.

Table 2.13 Carbonyl Compounds in Rum

Higher aliphatic aldehydes Ketones Di-ketones

Propionaldehyde Acetone 2,3-butanienone
Isobutyraldehyde 2-butananone 2,3-pentanedione
2-methylbutyradehyde 3-penten-2-0ne
isovaleraldehyde 2-pentanone
4-ethoxy-2-butanone
4-ethoxy-2-pentanone

Source: Nykänen and Nykänen (1983).

43
2.2.3.5 Acetals

As many as 43 acetals have been identified in rum (Nykänen, and Nykänen 1983).

Most of the acetals are formed during distillation by reaction of a molecule of aldehyde

and a molecule of alcohol, to form a hemiacetal. The hemiacetal formed is labile, and

once it combines with another alcohol molecule it forms a stable acetal (Figure 2.7).

The reaction of the formation of acetals is reversible. Both, the acetal formation

and hydrolysis reactions are catalyzed by acids (Nykänen, and Nykänen 1983). In rums,

the amount of acetals present can be from 4.3 to 41.6 % w/v (Lehtone n and Suomalien,

1977). The most abundant acetal is the diethyl acetal of acetaldehyde.

Ö: :ÖH OCH3
|  +H +

CH3 CH + CH3 OH CH3 COCH3 CH3 COCH3
CH3 OH 
H

Acetaldehyde + ethanol ethyl hemiacetal of +H+ diethyl acetal of

acetaldehyde CH3 OH acetaldehyde

Figure 2.7 Formation of Acetal

2.2.3.6 Phenols

Phenols are compounds present in small amounts, but the contribution they make

to the aroma of alcoholic beverages is significant (Lehtonen and Jounela- Eriksson, 1983).

Phenols are mainly formed during maturation via alcoholic extraction from the oak casks,

but they can also be formed during other production stages. During fermentation, p-

ethylphenol and p-ethylphenol can be formed through decarboxylation of p-coumaric and

44
p-ferulic acid. Table 2.14 contains some of the phenols identified in rum. High

temperatures, and some microorganisms as yeast and bacteria have been found

responsible for the decarboxylation. Other phenols are formed from the lignin of oak

barrels. Guaiacol, eugenol, vanillin, and m-creosol can be extracted from oak chips

(Lehtonen and Jounela-Ericksson 1983).

Table 2.14 Volatile Phenols Identified in Rum

Phenol 2,4-Dimethylphenol
o-Cresol p-Ethylphenol
m-Cresol p-ethylguaiacol
p-Cresol Eugenol
Guaiacol p-(n-Propyl)guaiacol
o-Rthylpehnol p-Methylguiacol

Source: Lehtonen and Jounela-Eriksson (1983).

2.3 Rum Aroma and Taste Perception

Alcoholic beverages contain numerous compounds that contribute to their aroma.

“Olfaction is the ability to recognize and discriminate with great accuracy and sensitivity

myriads of airborne molecules either by themselves or as complex mixtures” (Anholt,

1992). The olfactory system provides important information for survival by detecting and

processing chemicals in the environment. The aroma is perceived when a volatile

compound contacts the olfactory mucosa in the epithelium. The olfactory receptors axon

through the cribform plate of the ethmoid bone terminating in the olfactory bulb. There is

a large but limited number of olfactory receptors that transduce a large number of

odorants. The olfactory receptors direct extensions of the olfactory nerve into the

45
environment. The stimulus signal of a molecule contacting the olfactory receptor cells

has chemical specificity and it is translated to a neuronal electrical response. The

transduction mechanism is still unknown (Thorngate, 1997). Odors can be perceived in

two ways; direct, the molecules enter the nose from the front through the nosetrils by

inhalation, and indirect via mouth and rear nasal passages (Pousias and Chabanon, 1974).

The temperature rise of substances in the mouth releases more odorous compounds which

reach the olfactory region by diffusion and through exhalation. These in- mouth odors are

an important part of what is called flavor (Amerine and Roessler, 1983).

Man’s olfactory system is far less sensitive than that of many mammals. The

olfactory thresholds vary depending on the substance. Some factors like respiratory

infections and migraine can increase olfactory thresholds. Individual threshold variations

depend on various factors such as sex; women between adolescence and menopause have

higher sensitivities, especially after ovulation. During stimulation by a particular odor,

the threshold value rises above the stimulation concentration causing adaptation.

Adaptation is caused by the higher centers of the brain, and it explains why professional

smellers can perceive odors far longer than untrained persons (Carr, 1974).

Taste is a sensation transduced by a taste receptor cell. Taste receptors are located

in the taste buds which are specialized aggregates of taste receptor cells. The distribution

of the taste receptor is highly localized in the lingual margins. There are four tastes:

sweet, bitter, salty, and acid. The upper edges of the tongue react to acids, the rear to

bitter compounds, the tip of the tongue to sweet, and the sides to acid tastes.

The transduction of the taste sensation involves the interaction of sapid chemicals

with receptor sites, which are believed to be localized at the apical microvilli membranes.

46
The stimuli must be dissolved in order to diffuse through the saliva and mucous which

coats the apical membranes. Once proximate to the membrane, the stimuli may either

bind to a membrane-bound receptor, block membrane on channels, or directly influx

through ion channels (Thorngate, 1997).

The chemical sensations are astringency and irritations which are properly tactile

sensations. They are perceived by the trigeminal nerve endings, located in the lingual

mucous, pharingueal mucous, and the nasal mucous, including the yellow olfactory spot

(Poisais and Chabanon, 1974).

In most vertebrates, the receptor cells are located in the olfactory epithelium, a

pseudoestratified structure that lines a small portion of the nasal cavity. The olfactory

epithelium is unique in that; 1) the receptor cell body is in direct contact with the external

environment; 2) the olfactory neurons have a single, unbranched axon that projects to the

telencephalon without intermediate neurons; and 3) the olfactory system replaces neural

elements normally and when injured. The continued neurogenesys of olfactory cells is

due probably to its constant exposure to the environment resulting in injure and damage

of cells (Morrison and Constanzo, 1992). The neuroepithelium contains cells of three

types: olfactory cells, supporting cells, and basal cells, arranged in a psudoestratified

columnar organization.

The olfactory receptors have axons that terminate in the olfactory bulb and thus

they have a more direct connection to the brain than do the receptors of any other senses.

Their location in the upper part of the nostrils still allows for direct contact with the

environment. As they are exposed to pollution, they have the ability to regenerate. There

are about 6 million olfactory receptors in each nostril, enough to detect significant

47
environmental odors. They differ in the way they respond to different odorants, with each

cell responding to more than one, but not all odorants. Therefore, each compound

activates different sets of cells and generates unique patterns of ne ural activity that

correspond to the odor quality and concentration (Anholt, 1992). The response to the

odorant terminates when they are removed via mucociliary clearance and absorbed into

the circulation (Getchell et al., 1984; Hornug and Mozell, 1997; and 1980). In addition,

the olfactory epithelium has a high activity of cytochrome, which inactivate odorants, and

may also transform odorants into compounds of altered odorous quality and potency,

therefore affecting the complexity of the stimulus and its perception (Dahl et al., 1982;

Nef et al., 1989; and 1990; and Dahl, 1988).

2.3.1 Odor Perception

Odor perception is multimodal, involving other sensations, including taste, and

the relatively mild irritation from stimulation of the trigeminal nerve. “The sensory

function of smell is overshadowed in man by his other abilities for interaction with the

environment. Therefore, man has regressed developmentally in his olfaction ability as

compared with most other higher life forms” (Estrem and Renner,1987). Human

performance in detecting and identifying odors is less reliable than sight and hearing.

However, the sense of smell is a sensitive impressionable faculty for assessing the

chemical environment and for storing information. People can recognize familiar odors

but are usually able to describe them only in general terms. One may recognize an odor

as familiar without being able to pin down its identity, leaving one in the condition of

uncertainty that is called tip-of the-nose (Lawless and Engen, 1977).

48
 The pleasant and unpleasant qualities are not in the odors themselves but in the

events or persons with which they are associated. Odor preferences are learned and they

function because of the memory associations. Only familiar and identifiable odors are

perceived as good. Odor memory is better than other types of memory because one can

recapture the past, including the feeling of the remembered event. However, we cannot

recall odors at will in the absence of such stimulation. A long-term odor memory can be

established with only one exposure. Odor associations can be formed automatically and

without conscious awareness. Having learned one association to an odor makes it

difficult for us to replace the association with another one. Odors do not have their own

names, but are described as smelling like something else. The special role of odors is to

retrieve significant events regardless on when they happened. The sense of smell is very

sensitive, learned quickly, and is not forgotten, but is not very discriminating and it has

no judgment about what is important to remember and what is best forgotten. The human

nose is in constant use because environmental odors must be in constant monitory. For

each odor detected, a memory search is made to determine its identity. Familiar odors are

hardly noticeable, only odors that are unusual or unexpected get conscious attention.

2.3.2 Persistence of Odors

Adaptation is caused by a decreased sensitivity to an odor because of a change in

the function of the receptors, and this decrement may spread to the perception of odors

through cross adaptation. However, depending on the pair of odors, exposure to one odor

may also increase the response to a second odor, this is called facilitation. When the odor

stimulus is removed or avoided, recovery from adaptation occurs. The concentration of

the molecules can exceed the ability of the receptors to handle them, this is perceived as

49
adaptation, fatigue, or system failure. The two parameters that affect adaptation are

strength of the stimulus and the duration of the exposure; the intensity and stability of

odor stimulation are controlled by sniffing. Odor processing is variable even though one

may be exposed to the same constant concentration of odor. It is believed that only 5 to

10% of the breathed air will reach the olfactory cleft, the rest is diverted below to the

pharynx, past the epiglottis to finally reach the lungs. Sniffing serves a way to control the

amount of air flowing trough the nostrils. By simulating gagging one can shut off the

access of odorous molecules through the pharynx, between the mouth and the cleft. This

prevents overexposure and keeps the sense of smell working. For any single odor, the

longer the exposure, the less sensitive one is to it. This is a result of a change in the

reaction of the olfactory receptors to odorous molecules, causing a rapid exponential

decrement in sensitivity, but except for the weakest odors, sensitivity does not diminish

to zero. (Berglund et al., 1978).

Naturally using the sense of smell will cause some fatigue, regarding some odors

more than others, and it will change and not abolish perception. A properly working

olfactory system prevents functional anosmia by avoiding extreme stimulation and by

controlling the amount of stimulation through sniffing. Constant exposure to various

odors affects sensitivity without interfering with the perception of significant odors. For

certain pairs of odors an opposite effect can occur, this is a prior exposure to an odor

increases the perceptual strength of a new one. This effect is called facilitation, synergism

or potentiation. It has not been widely explained but it is believed to be involved the

periferial olfactory mucosa, where an odor interacts with the receptors mucous. The

possible mechanisms involved may be: 1) An odor may alter the mucus around the

50
olfactory receptors making them more accessible to new molecules, 2) the two odors

involve the same receptor in a way in which their effects are summed, or 3) the first

effect may create a priming effect that enhances general chemical communication. The

sense of smell can recover from fatigue associated with the exposure to an odor after

some period of time.

51
 CHAPTER 3. DEVELOPMENT OF SENSORY DESCRIPTORS FOR RUM
AROMA, FLAVOR, AND TASTE EVALUATION BY SEMI-EXPERT JUDGES

3.1 Introduction

Centuries ago, lexicon was a language to serve poetic styles and subjects of social

conversations. Nowadays lexicon is used as expressing judgment and comparing products

of mostly similar origin, and simultaneously creating a glossary allowing for the guidance

and training of the less knowledgeable panelists (Boyazoglu, 1986.) These descriptors

for products have not only been us ed by scientists, but also by wholesalers and retailers

that employ the power of words to attract the attention of consumers.

The development of a meaningful lexicon is needed in order to have words that

adequately interpret the description of the product. It is necessary that when those words

are used, their meaning is clear to an audience, and finally, when a product is marketed,

these terms will be used through ads and labels to educate the public about the product.

Lexicon development is widely used in Descriptive Sensory Analysis, techniques

used for product development, quality control, and in laboratory practices (Amerine and

Roessler, 1983). Among the Descriptive Analysis Techniques, Flavor Profile,

Quantitative Descriptive Analysis (QDA), Texture Profile and Spectrum require a

consensus among judges for the use of terms and definitions used to describe the product.

Free choice profiling is a technique that does not require consensus among judges, but it

allows for more descriptive terms that are not necessarily meaningful to all people.

For accurately describing a product, the words used should be objective and not

subjective. They should not represent feelings (good, bad, etc.), because the interpretation

made can be varied depending upon personal background and experiences.

52
 Disagreement often takes place in applying terminology, therefore it is

recommended that the panelists agree in applying the same terms to products (Lehrer,

1983). Having a set of references that can illustrate the meaning of the terms is necessary

for stabilizing the panel’s use of vocabulary. It is also useful for assisting in the training

of new recruits on the use of terms. Therefore, suitable materials should be available to

use as a standard (Lawless and Klein, 1991). The standard should: 1) contain a true

representation of the favor note as it appears in the product, 2) contain the flavor note and

no other note, and 3) be readily available in the required purity. It can be a pure

compound or a complex material, the problem of using the later is that it might not be

stable, and therefore not suitable for use (Lawless and Klein, 1991).

As stated by Bloomfield, (1939), “standardization of scientific language is

obtained when a technical term is fixed by an agreement of definition which receives

explicit formulation”. When interpreting descriptors, it is important to consider that even

experienced judges (particularly from different regions) are apt to have very different

concepts about the terms (Amerine and Roessler, 1983). One must also consider that

meaning of terms may change with time, for example a sweet wine in the past may not be

so today (Ribéreau-Gayon, 1973). It is a mistake to think that every word has a fixed,

specific meaning, where words have intra- and extra- linguistic relationships (Lehrer,

1983). Therefore, the importance of a term is based not only on its simplicity and clarity,

but also mainly on its recognizable meaning (Amerine and Roessler, 1983).

One important thing to consider for developing a meaningful list of descriptors is

to have a consensus among the judges on the terms and definitions. A disadvantage of

53
working on a consensus is that bias can be introduced by a dominant (assertive) member

of the panel (Amerine and Roessler, 1983).

For the selectio n of descriptors it is important to consider that they should be

meaningful; they accurately and precisely describe all required characteristics of the

samples; they have to be understood by the assessors, and they should be agreed on their

meaning; they can be easily identified with reliable standards, and they are understood by

the recipients of the report (Lawless and Klein, 1991).

The aim of the study was to develop and evaluate a list of terms with their

descriptions to describe rum products prepared from different substrates using different

processing protocols.

3.2 Materials and Methods

For the development of a list of lexicon, 15 different commercial rums were

selected to cover the spectrum and for comparative framework. For the actual product

evaluation 6 commercial and one experimental sample were used. The experimental

sample was made from molasses, distilled in continuous still and aged under an

accelerated aging processing using oak chips (10g/L) for 2 ½ months at 30°C.

For the development of lexicon a panel of 5 semi-experts judges familiarized with

evaluation of rum products and previously screened was used. A total of 9 sessions and

16 hours were required for training and development of the list of terms and definitions.

Three sessions of 1 1/2 hours were used for evaluation of samples with the list of terms,

definitions and references. All the judges were present in all sessions. The sessions were

done in an air-conditioned room with around 67°F under fluorescent lighting with a

round table. The time for the evaluations was between 3:00 to 5:00 PM.

54
 The samples (1 oz) were presented coded and served in plastic sealed cups at

room temperature. For the development of the list of terms, the judges were presented

with 5 to 6 rum samples simultaneously. They were asked to evaluate them individually,

and report the perceived character note with definition in a ballot provided. After this was

done, the panelists had a group discussion to get consensus for the terms. From the

individual sessions a pooled list of terms with definitions was prepared. The panelists

were provided with the list of terms with definitions and the references, they were then

asked to evaluate samples based on those descriptors, and to indicate the order in which

they were perceived. In this case they evaluated a total of 7 samples including the

experimental sample having two or three samples per session presented individually. The

judges were ask to smell in short deep sniffs; first at a distance of 2.5 cm from the rim of

the cup, then at the rim of the cup, next with the nose inside the cup, and finally they

were asked to swirl the cup and sniff again with the nose inside it. Afterward the judges

were asked to take a small sip, let it go to the back of the tongue, then slowly forward to

the front of the mouth, take a deep breath, and swallow.

The perception of aftertaste was recorded after 1 minute. After evaluating each

sample, the judges were asked to rinse their palates with spring water. In this case each

judge checked the list of terms provided and determined the perceived notes, then a

general consensus and conclusion for each sample was done, then the judges proceeded

to evaluate the next sample.

55
3.3 Results and Discussion

3.3.1 List of Terms with Definitions for the Description of Rum Aroma, Flavor, and

Taste

The panel of 5 semi-expert judges was able to determine the presence of the

following attributes in the 15 rum samples. Definitions for some of the terms are based on

Civille and Lyon (1996).

1. Allspice: spicy characteristic of ground allspice

2. Almond: aromatic associated with almond/almond extract

3. Apple: aromatic characteristic of various apple varieties

4. Apple/pear; fruity- like aromatic of pome fruits which cannot be recognized as any

specific fruit

5. Artificial: aromatics or tastes considered artificial or not natural for rum

6. Astringent: chemical feeling factor

7. Banana: fruity aromatic characteristic of ripen banana

8. Bite: chemical burning sensation felt in tongue, mouth or throat

9. Burnt: chemical feeling factor associated with high concentration of irritants to

the mucous membranes of the oral cavity

10. Butter: aromatic associated with fresh butter

11. Butterscotch: sweet aromatic typically having both buttery and caramelized notes

12. Caramel: an overall term reminiscent of chewy caramel

13. Chemical: associated with compounds such as solvents, cleansing compounds,

and hydrocarbons

14. Cinnamon: sweet, woody aromatic of ground cinnamon bark

56
15. Pure ethanol: a pungent aroma associated with ethanol

16. Eucalyptus/mint: a sweet green aroma associated with fresh mint/eucalyptus

leaves

17. Floral: fragrant associated with flowers

18. Fruity (general): aromatic associated with non-specific fruits

19. Fusel oil: aroma note associated with isoamyl alcohol, butanol

20. Gasoline: chemical solvent of hydrocarbons, reminiscent of gasoline, kerosene.

21. Leathery: aromatic associated with tanned animal hides

22. Medicinal: aromatic associated with band-aids, disinfectant-like (phenolic)

23. Musty: aromatic associated with closed air spaces such as attics (dry) and

basements (wet)

24. Nutty: aromatic associated with nuts or nut meats

25. Pineapple: aromatic associated with fresh pineapple

26. Plastic: aromatic associated with plastic, polyethylene containers

27. Smoke: perception of any kind of smoke flavor

28. Smooth: clean, pleasant chemical feeling factor

29. Spicy: overall term associated with pungent spices

30. Sweet: taste on tongue stimulated by sugars and other high potency sweeteners

31. Syrupy: aroma associated with clean syrup, pancake syrup.

32. Vanilla: aromatic blend of sweet, vanillin, woody

33. Woody: aromatic associated with dry fresh cut wood, balsamic or bark- like.

57
 For the development of the list of terms, it was very useful to provide the judges with

a variety of references for each descriptor. This allowed the judges to have an open

discussion on their perception, to homogenize their criteria, and also helped them to

identify attributes they were not able to perceive before. In each session, the judges were

presented with a list of the terms previously identified in the rum samples by them. This

facilitated finding those descriptors in new samples, and allowed them to focus on the

perception of new descriptors. All the attributes on the list were created from a consensus

among all the judges.

3.3.2 Rum Evaluation Using Terms Developed by the Judges.

During evaluation sessions, the rums used were tested and checked for the presence

of the different aroma, flavor and taste descriptors developed by the judges. The judges

were also asked to determine the order of attributes perceived when sniffing the sample at

4 different instances: (1) at about 1-2 inch from the rim of the cup, (2) at the rim of the

cup, (3) with the nose inside the cup, and (4) again with the nose inside the cup after

having swirled the sample. Then the panelist were asked to take a small sip, bring it to the

back of the tongue, slowly to the tip of the tongue, followed by a deep breath, and finally

swallow the sample. Smelling at different levels allowed the judges to detect different

aroma components naturally separated by the variation in their volatility. The fact that the

judges were able to determine the instance they detected a specific descriptor allowed

other judges to verify their perception, focus on that specific attribute, and be able to

detect attributes easier. This also facilitated reaching a consensus among the judges for

that sample.

According to consensus, their perception of the rum samples was as follows:

58
o Rum form Haiti from sugarcane juice double distilled in copper-pot stills

and aged in Limousin oak casks for 15 years, with 86 Proof was perceived

as "astringent", "bite", "caramel", "fruity", "pure ethanol", "smoke",

"sweet", and "woody".

o Puerto Rican rum from molasses distilled in continuous columns, that

consisted of blends of different aged distillates, with 80 Proof the

following character notes: "apple/pear", "bite", "caramel", "cinnamon",

"pure ethanol", "sweet", and "vanilla".

o Rum from Barbados made from molasses, double distilled in single

distillations, blend of different distillates with 80 Proof. It was described

as "bite", "pure ethanol", "fruity", "leathery", "smoked", "spicy",

"vanilla", and "woody".

o Rum made in Jamaica from molasses distilled in copper stills, blend of

different distillates. Aged for 12 years with 86 Proof, had the character

notes: "bite", "fruity", "vanilla", and "woody".

o Rum made in Honduras from molasses distilled in continuous columns,

aged for 7 years with 80 Proof was described as "bite" "butterscotch",

"pure ethanol", "fruity", "sweet", "vanilla", and "woody".

o Cuban rum made from molasses distilled in continuous columns aged 7

years with 80 Proof was described as "butterscotch", "leathery", and

woody".

o Experimental sample made from molasses distilled in continuous

distillation, aged for 2 1/2 months with American oak chips at 30°C with

59
 80 Proof was described as "artificial", bite", "chemical", "pure ethanol",

"fusel oil", "leathery", "smoke", "vanilla", "woody".

3.4 Significance

There are a large variety of published terms for the description of wine and other

alcoholic beverages. However, it is important to have a list of terms that are useful and

specific for the sensory evaluation of rum. The aroma of rum is very complex and varies

greatly according to the processing protocol, therefore this study aimed to create a

lexicon for the aroma, flavor and taste of rum that includes attributes for the description

of all the different types of rum produced. Other requirements for the list of lexicon were

to be objective and easy to understand and perceive by different judges, this is the reason

for the stress in the consensus of the created descriptors.

A set of lexicons and their description from this study is useful to describe a broad

spectrum of rum products. It can be further used for quantification of sensory quality

through descriptive analysis techniques.

60
 CHAPTER 4. DESCRIPTIVE ANALYSIS OF THE AROMA OF RUM

4.1 Introduction

Rum is commonly examined and evaluated for quality control purposes and for

ideal blending by highly trained and experienced expert blenders that work for each

individual bottling company. Consumers usually based their criteria for the selection of

rum on the raw material used, the origin, color, aging time, or brand characterization. In

general, dark aged rums are believed to have more intense aromas than the light white

rums. The light rums are most commonly used as mixers while dark rums are sipped

straight. However, little is known on the differences in the aroma attributes perceived in

these products.

Several techniques that focus on the description and quantification of the aroma of

the products have been developed. Descriptive Analysis is the initial step of the

characterization of food products, and is the basis for product testing. It categorizes the

different senses, and provides a language useful for the communication of perceptions

(Moskowitz, 1983). Quantitative Descriptive Analysis (QDA)® developed by Stone

(Stone et al, 1974) allows for the statistical analysis of the obtained sensory evaluation.

Quantitation of sensory responses requires the use of trained panels using the same

vocabulary, as well as adequate definitions and standard references. In addition, it

requires the use of the same scale to unify the evaluation criteria. SpectrumT M method of

analysis consists of a complete description and accurate characterization of the sensory

attributes of the product (Muñoz and Civille, 1999). It provides tools for designing a

descriptive procedure for a given product category. It is a custom design approach to

panel development, selection, training, and maintenance (Meigaard et al., 1999).

61
Spectrum analysis uses a 15 cm scale anchored with several intensity reference points. It

requires the participation of trained panelists with acute senses previously screened, the

use of standard terminology and scales to rate the intensity of the perceived attributes

(Muñoz and Civille, 1992).

Descriptive Analysis Techniques involve the characterization of the product by

attributes and the intensity of those attributes. They have qualitative and quantitative

components, both are necessary for the effective performance of the data analysis

(Muñoz, and Civille, 1998). The references also have qualitative and quantitative

characteristics, they include the background information and the reference points for

intensity that evaluators will relate to when evaluating products. Without intensity

references the panelists use their own criteria to rate intensities, resulting in higher

variability (Muñoz and Civille, 1998).

Some specific descriptive analysis techniques have been developed or modified

for the special needs of the evaluation of different alcoholic beverages. In most cases

such methodologies are created for wine evaluation, such as the Traditional Quality

Assessment (TQA) for wine, or the University of California Davis descriptive analysis

(UCD DA) for the wine industry (Noble, 1998). Other methodologies have been applied

to beer such as modified QDA for sensory profiling of beer (Mecredy et al., 1974); to

whisky (Piggot et al 1980 and 1985); and to gin (Piggot et al 1983). However, there has

not been a methodology developed specifically for the descriptive analysis of rum

products. The objectives of this study were to develop a methodology suitable for the

descriptive analysis of rum, to describe the aroma of a variety of different rum products,

and to relate the perceived rum aroma to the processing conditions of the products.

62
4.2 Materials and Methods

4.2.1 Panelist Screening

All participants were required to perform a screening test before training. The

screening tests were completed in four days. The participants completed the screening

test at the most convenient time for them, although they were asked not to come right

before or after meals. The test was done in an air-conditioned room, in partitioned booths.

The test was divided into 4 different sections. In the first section the participants were

presented with 7 solutions and asked to identify the taste perceived in each of them

(sweet, sour, salty or bitter). The solutions were prepared with sucrose (0.5%), citric acid

(0.25%), salt (0.2%) and grapefruit juice (TexSun ruby red grapefruit juice, Citrus World

Inc., Lake Wales, Florida). The concentration of the solutions was above identification

threshold level. The samples were presented in capped plastic cups filled with ca. 20 ml,

and coded with 3 digit random numbers. The participants were asked to write down the

number of the solution on the line of the matching taste perceived. In the second part the

participants were presented with four different sets of solutions and asked to rank

intensities. Sucrose solution was used for sweet taste; white vinegar was used for sour

taste; salt was used for salty taste, and grapefruit juice was used for bitter taste. 4

dilutions of each basic taste were presented. The following solutions were given to the

participants; sugar solutions at 0.5%, 1%, 1.5%, and 2%; salt solutions at 0.2%, 0.4%,

0.6%, and 0.8%; white vinegar solutions at 0.25%, 0.45%, 0.65%, and 0.85%; and

grapefruit juice (TexSun Ruby Red Grapefruit juice, Citrus World Inc, Lake Wales,

Florida) diluted with water to 1/4, 1/3, ½, and pure grapefruit juice. The set of solutions

were randomly coded with a 3-digit number, and presented in capped plastic cups with

63
ca. 20 ml of solution each. The participants were asked to rank the 4 solutions presented

according to the intensity from the lowest to the highest. Each set was presented

individually for evaluation. For the bitter taste ranking, the evaluation was performed

under red light to minimize the stimulus error from the effect of color from the grapefruit

juice. The purpose of the third part of the screening was to check the ability of the

panelists to adequately describe perceived aroma of common food or ingredient samples.

The samples, contained in 10 closed coded vials with a paper strip saturated with

different extracts were given to the judges.

Figure 4.1. Screening

64
 The panelists were instructed to sniff from the vial and describe the perceived

aroma of the samples. The identity of the samples presented was: vanilla extract (Mixim

Labs., Calle Sur No. 6 P.O. Box 3, Naucalpan, Mexico, Mexico); smoke flavor

(Firmenich de Mexico, Luisiana No. 80, Napoles, 03810, Mexico City, Mexico); artificial

butter extract (Firmenich de Mexico, Luisiana No. 80, Napoles, 03810, Mexico City,

Mexico); cinnamon essential oil (AMCO Internacional, S.A. de C.V. Trigo No. 16

Granjas Esmeralda, 09810, Mexico City, Mexico); oak essential oil extract (Robertet de

Mexico, S.A. de C.V., Año de Juarez No. 65, Granjas San Antonio, 09070, Mexico City,

Mexico); artificial almond extract (Firmenich de Mexico, Luisiana No. 80, Napoles,

03810, Mexico City, Mexico); black pepper essential oil (AMCO Internacional, S.A. de

C.V. Trigo No. 16 Granjas Esmeralda, 09810, Mexico City, Mexico); artificial caramel

extract (Firmenich de Mexico, Luisiana No. 80, Napoles, 03810, Mexico City, Mexico);

and artificial banana extract (Firmenich de Mexico, Luisiana No. 80, Napoles, 03810,

Mexico City, Mexico). The final section of screening consisted of 2 different sets of

samples coded with random 3 digit numbers. The participants were asked to sniff the first

set of samples allowing time for rest after each sample, then sniffing the second set of

samples, match the samples from the second set with the first one, writing down their

corresponding number, and assigning to them a term from a provided list that best

describes each sample. The identities of the samples were eucalyptus essential oil

(AMCO Internacional, S.A. de C.V. Trigo No. 16 Granjas Esmeralda, 09810, Mexico

City, Mexico); clove essential oil (AMCO Internacional, S.A. de C.V. Trigo No. 16

Granjas Esmeralda, 09810, Mexico City, Mexico); white vinegar; rose essential oil

(AMCO Internacional, S.A. de C.V. Trigo No. 16 Granjas Esmeralda, 09810, Mexico

65
City, Mexico); artificial garlic extract (Firmenich de Mexico, Luisiana No. 80, Napoles,

03810, Mexico City, Mexico); peppermint essential oil (AMCO Internacional, S.A. de

C.V. Trigo No. 16 Granjas Esmeralda, 09810, Mexico City, Mexico); artificial banana

extract (Firmenich de Mexico, Luisiana No. 80, Napoles, 03810, Mexico City, Mexico);

artificial hazelnut extract (Firmenich de Mexico, Luisiana No. 80, Napoles, 03810,

Mexico City, Mexico); artificial strawberry extract (Firmenich de Mexico, Luisiana No.

80, Napoles, 03810, Mexico Cit y, Mexico), and nutmeg essential oil (AMCO

Internacional, S.A. de C.V. Trigo No. 16 Granjas Esmeralda, 09810, Mexico City,

Mexico). After the screening test, the selected participants were informed in a detailed

manner of the project, and briefly interviewed for their attitude towards rum and their

availability for being part of the project. All participants were required to be over 21

years of age, and having good health. Since there was no other incentive for their

participation in the project, the main criteria for the selection were their willingness and

commitment to participate, and their availability to come to all the orientation, training

and evaluation sessions.

4.2.2 Orientation

The orientation session took place in a conference room with a large central table.

The session started at 4:30 PM and lasted about one hour. All the selected panelists were

asked to come to an orientation session. There was a brief presentation on the objectives

of the study, the definition and production of rum, the perception of aroma, the sensory

technique and the methodology involved in this study. They were informed of the

development of training and evaluation sessions, including dates, time, and location.

During orientation all the participants were informed of the importance of their

66
commitment to the project and their presence during all the group-training sessions. Then

there was a group discussion where they agreed on the best times for the group training

sessions, and all their questions about the project were answered. Each of them was

provided with copies containing all the information discussed during orientation

including an overview of the project, the perception of aroma, and a work schedule.

Figure 4.2 Orientation Session.

4.2.3 Group Training Sessions

The objective of the first part of the training session was to develop an ability for

the panelists to describe objectively and precisely the aroma of different samples. All the

panelists were asked to be present during all training sessions, all the sessions started at

67
4:30 PM. The panelists were ask to work individually and in silence during evaluation,

and to share openly their responses during discussions, paying attention to the responses

of other panel members and checking the samples for new descriptors as the others

present them.

During the first and second training sessions the panelist were presented with

different samples contained in capped plastic cups. They were also given an additional

cup filled with toasted coffee beans for neutralizatio n. The panelists were instructed to

smell each sample with short deep sniffs, write down the descriptors in the order they

were perceived to describe the aroma/odor of the sample. Then they were asked to give a

brief definition for each descriptor. In between samples they were required to sniff

themselves (the collars of their shirts) the coffee beans cup, and themselves again, then

rest for about 1-2 minutes, and proceed to the next sample. After all the panelists

completed the evaluation, they were encouraged to share their responses with the group,

and they had a brief discussion on what they perceived. Finally they were informed on

the true identity of the samples and asked to smell all the samples again. The first session

lasted about 1 hour, and the samples used were: (1) pureed pealed golden delicious apple;

(2) fried apples with cinnamon and sugar; (3) liquid smoke (The Colgin Companies,

P.O.Box 7779. Dallas TX. 75209); (4) ground pecans; (5) cane syrup (The C.C. Steen

Syrup Mill Inc. P.O. BOX 339. 119 N. Main St. Abbeville LA. 70510); (6) molasses (The

C.C. Steen Syrup Mill Inc. P.O. BOX 339. 119 N. Main St., Abbeville LA. 70510); (7)

grapefruit juice; (8) mix of vanilla/butter extracts (McCormick & Co. Inc. Hunt Valley,

MD. 21031-1100); (9) Dark Rum: (10) Light Rum. For the second training session the

samples used were (1) butterscotch extract (Firmenich de Mexico, Luisiana No. 80,

68
Napoles, 03810, Mexico City, Mexico), (2) Caramel /almond extract (1:1) (Firmenich de

Mexico, Luisiana No. 80, Napoles, 03810, Mexico City, Mexico), (3) Mixed pureed fresh

banana and pealed Golden Delicious apple, (4) Mixed ground clove and black pepper, (5)

Mixed oak wood extract (Robertet de Mexico, S.A. de C.V., Año de Juarez No. 65,

Granjas San Antonio, 09070, Mexico City, Mexico) with smoke flavor (Firmenich de

Mexico, Luisiana No. 80, Napoles, 03810, Mexico City, Mexico) (6:1), 6) Vanilla/pecan

extracts (Firmenich de Mexico, Luisiana No. 80, Napoles, 03810, Mexico City, Mexico)

(1:1), (7) Diet Pepsi (Pepsico, Inc.), (8) Minced fresh mushrooms, (9) Red Wine Vinegar.

Spice Islands Gourmet Premium Red Wine Vinegar, containing red wine vinegar, water

and sulfites (Burns Philip Food Inc. San Francisco Ca. 94108), (10) Cough Syrup:

Tanafed Suspension with strawberry/banana fla vors. It was contained in a plastic bottle

(Horizon Pharaceutical Corporation Roswell, GA, 30076), (11) Floral Perfume extract

(AMCO Internacional, S.A. de C.V. Trigo No. 16 Granjas Esmeralda, 09810, Mexico

City, Mexico), and (12) Grapefruit Essential Oil. Expressed oil from fresh white

grapefruit. The samples were presented in capped plastic cups. The second training

session lasted 1hr 15 min.

In the third training session, the panelists were provided with 4 sets of samples

with 3 samples each. They were asked to evaluate the samples as in sessions 1 and 2, and

provide terms and their descriptions of terms to describe each set of samples. Then after a

brief group discussion, the panelist were asked to agree on some terms that would be used

to describe each set of samples, and rank the samples according to the intensity of each of

the selected terms. The identity of the samples was: set (1) Molasses: Steen’s Home Style

Molasses: Undiluted, diluted 1:1 with water, and diluted 1:3 with water; set (2)

69
Mushroom/pecan/almond. 5 medium mushrooms were minced with ca. 20 ml water, 3

drops of artificial pecan extract (Firmenich de Mexico, Luisiana No. 80, Napoles, 03810,

Mexico City, Mexico) and 3 drops artificial almond extract (Firmenich de Mexico,

Luisiana No. 80, Napoles, 03810, Mexico City, Mexico) were added. The mixture was

expressed and filtered with cheese-cloth. Dilutions at 100%, 50% and 33% were

prepared; set 3) Vanilla/apple/smoke: a Golden Delicious Apple was cut in pieces and

minced with 20 ml of water. 5 drops of vanilla extract: pure natural vanilla extract (3x)

(Mixim Labs. Calle Sur No. 6 P.O. Box 3, Naucalpan, Mexico, Mexico), and 1 drop of

smoke flavor (Firmenich de Mexico, Luisiana No. 80, Napoles, 03810, Mexico City,

Mexico)) were added. The mixture was mixed thoroughly and filtered using cheese-cloth

to obtain a clear liquid. Dilutions at 100%, 50% and 33% were used; (4) Rum: Ron Rico

Extra Smooth Premium Gold Label. Dilutions at 100%, 50% and 33% were presented for

evaluation. This session lasted for 1.5 hrs.

In the fourth training session a single rum sample was provided. The panelists

were asked to evaluate the sample as in previous sessions, but they were encouraged to

provide more terms and their corresponding definitions for each of them. They were then

asked to sniff the sample several times and at different distances from the cup; (i.e., at 1

inch from the rim of the cup, at the rim of the cup, with the nose inside the cup, and sniff

again with the nose inside the cup after swirling). The panelists were then asked to share

their responses and reevaluate the samples again. Finally they had to agree on the terms

that they were able to perceive. For the fifth training session, 35 samples contained in

glass vials were given to the panelists for evaluation. They were asked to evaluate the

samples by very quick sniffs, then they were asked to record terms for describing the

70
aroma and a brief definition for it. After they had evaluated all the samples a group

discussion took place to agree on the term(s) to describe each sample. The identity of the

samples used was: (1) ground allspice (Astor products Inc. Jacksonville, Florida, 32203)

in a capped plastic cup; (2) anise seeds (Burns Philips Food Inc., San Francisco, Ca,

94108) in a capped plastic cup; (3) clove essential oil (AMCO Internacional, S.A. de

C.V. Trigo No. 16 Granjas Esmeralda, 09810, Mexico City, Mexico) in paper strips

contained in a closed glass vial; (4) black pepper extract (AMCO Internacional, S.A. de

C.V. Trigo No. 16 Granjas Esmeralda, 09810, Mexico City, Mexico) in paper strips

contained in a closed glass vial; (5) cinnamon essential oil (AMCO Internacional, S.A.

de C.V. Trigo No. 16, Granjas Esmeralda, 09810, Mexico City, Mexico) in paper strips

contained in a closed glass vial; (6) artificial butter extract (Firmenich de Mexico,

Luisiana No. 80, Napoles, 03810, Mexico City, Mexico) in paper strips contained in a

closed glass vial; (7) artificial butterscotch extract (Firmenich de Mexico, Luisiana No.

80, Napoles, 03810, Mexico City, Mexico) in paper strips contained in a closed glass vial

; (8) artificial caramel extract (Firmenich de Mexico, Luisiana No. 80, Napoles, 03810,

Mexico City, Mexico) in paper strips contained in a closed glass vial; (9) vanilla extract

(Pure natural vanilla extract (3x) from Mixim Labs., Calle Sur No. 6 P.O. Box 3,

Naucalpan, Mexico, Mexico) in paper strips contained in a closed glass vial; (10) Burnt

sugar: 5 grams sugar with 5 ml water contained in glass vials, and cooked for different

times in a microwave oven (1, 2, 2.25, 2.5, 2.75, 3, 3.5, and 4 minutes); (11) eucalyptus

essential oil (AMCO Internacional, S.A. de C.V. Trigo No. 16 Granjas Esmeralda, 09810,

Mexico City, Mexico) in paper strips contained in a closed glass vial; (12) peppermint

essential oil (AMCO Internacional, S.A. de C.V. Trigo No. 16 Granjas Esmeralda, 09810,

71
Mexico City, Mexico) in paper strips contained in a closed glass vial; (13) spearmint-

flavored chewing gum (Wrigley’s Company, Chicago, IL, 60611); (14) Artificial pecan

extract (Firmenich de Mexico, Luisiana No. 80, Napoles, 03810, Mexico City, Mexico)

in paper strips contained in a closed glass vial; (15) artificial hazelnut extract (Firmenich

de Mexico, Luisiana No. 80, Napoles, 03810, Mexico City, Mexico) in paper strips

contained in a closed glass vial; (16) artificial almond extract (McCormirck & Co., Inc.,

PO. Box 208 Hunt Valley, MD, 21030-0208) in paper strips contained in a closed glass

vial; (17) finely crushed peanuts in a capped plastic cup; (18) artificial strawberry extract

(Firmenich de Mexico, Luisiana No. 80, Napoles, 03810, Mexico City, Mexico) in paper

strips contained in a closed glass vial; (19) artificial banana extract (Firmenich de

Mexico, Luisiana No. 80, Napoles, 03810, Mexico City, Mexico) in paper strips

contained in a closed glass vial; (20) artificial apple extract (Firmenich de Mexico,

Luisiana No. 80, Napoles, 03810, Mexico City, Mexico) in paper strips contained in a

closed glass vial; (21) artificial green apple extract (Firmenich de Mexico, Luisiana No.

80, Napoles, 03810, Mexico City, Mexico) in paper strips contained in a closed glass

vial: (22) oak wood essential oil (Robertet de Mexico, S.A. de C.V., Año de Juarez No.

65, Granjas San Antonio, 09070, Mexico City, Mexico) in paper strips contained in a

closed glass vial; (23) expressed juice of ground spinach; (24) expressed juice of ground

cabbage; (25) ground fresh green beans; (26) mango type natural flavor (Flavors of North

America Inc. 525 Randy Rd. IL, 60188) in paper strips contained in a closed glass vial;

(27) expressed juice of minced mushrooms; (28) liquid smoke (The Colgin Companies,

P.O.Box 7779. Dallas TX. 75209) in paper strips contained in a closed glass vial; (29)

cough syrup (Tanafed Suspension with strawberry/banana flavors. Horizon

72
Phamaceutical Corporation Roswell, GA, 30076) in a capped glass vial; (30) paint

thinner: few drops in a closed glass vial; (31) ethanol: few drops contained in a closed

glass vial; (32) rubbing alcohol (isopropyl alcohol): a few drops contained in a capped

glass vial; (33) methanol: a few drops contained in a capped glass vial; (34) propanol: a

few drops contained in a capped glass vial; (35) Isoamyl alcohol: a few drops in a capped

glass vial; (36) ethyl butyrate: a few drops in a capped glass vial; (37) new closed plastic

bottle. This session lasted for 1hr 45 minutes.

For the sixth session the panelists were given 18 samples, they were asked to

evaluate the samples as conducted in previous sessions, and to provide a larger number of

terms with their definition for each sample. After they had finished, a group discussion

took place. The identity of the samples used was: (1) artificial pecan extract (Firmenich

de Mexico, Luisiana No. 80, Napoles, 03810, Mexico City, Mexico); (2) burnt sugar; (3)

ground fresh green beans; (4) expressed juice of minced mushrooms; (5) cinnamon

essential oil (AMCO Internacional, S.A. de C.V. Trigo No. 16 Granjas Esmeralda, 09810,

Mexico City, Mexico); (6) eucalyptus essential oil (AMCO Internacional, S.A. de C.V.

Trigo No. 16 Granjas Esmeralda, 09810, Mexico City, Mexico); (7) allspice (Astor

products Inc. Jacksonville, Florida, 32203); (8) artificial hazelnut extract (Firmenich de

Mexico, Luisiana No. 80, Napoles, 03810, Mexico City, Mexico); (9) artificial

butterscotch extract (Firmenich de Mexico, Luisiana No. 80, Napoles, 03810, Mexico

City, Mexico); (10) clove essential oil (AMCO Internacional, S.A. de C.V. Trigo No. 16

Granjas Esmeralda, 09810, Mexico City, Mexico); (11) artificial banana extract

(Firmenich de Mexico, Luisiana No. 80, Napoles, 03810, Mexico City, Mexico); (12)

vanilla extract; (13) artificial apple extract (Firmenich de Mexico, Luisiana No.8,

73
Napoles, 03810, Mexico City, Mexico); (14) ethyl butyrate; (15) ground pecans; (16)

Maple syrup (100% pure maple syrup, Maple Grove Farms Vermont); (17) clover honey;

(18) soy milk. The samples were placed in glass vials. For those samples that contained

pure extracts a paper strip with a few drops of the correspondent extract was placed in the

vial. For session 7 the panelists were asked to evaluate samples as in session six. The

identity of the samples provided was: (1) peppermint extract; (2) artificial strawberry

extract (Firmenich de Mexico, Luisiana No. 80, Napoles, 03810, Mexico City, Mexico);

(3) artificial almond extract (Firmenich de Mexico, Luisiana No. 80, Napoles, 03810,

Mexico City, Mexico); (4) black pepper extract (AMCO Internacional, S.A. de C.V.

Trigo No. 16 Granjas Esmeralda, 09810, Mexico City, Mexico); (5) vanilla extract (Pure

natural vanilla extract (3x) from Mixim Labs., Calle Sur No. 6 P.O. Box 3, Naucalpan,

Mexico, Mexico): (6) artificial caramel extract (Firmenich de Mexico, Luisiana No. 80,

Napoles, 03810, Mexico City, Mexico); (7) Natural Oak Flavor (Wild Flavors Inc. 1261

Pacific Ave nue, Erlanger, KY 41018); (8) paint thinner. A few drops of the extracts were

placed on a paper strip in a capped glass vial. The second part of the session consisted on

the evaluation of a single rum sample contained in a capped plastic cup. The panelists

were encouraged to provide as many terms as they could come up with and provide a

definition for each of them. At the end of the session they shared their responses,

reevaluated the sample, and created a pooled list of terms and definitions. This session

lasted for 1hr 15 min. For sessions 8, 9 and 10, different rum samples were given to the

panelists in capped plastic cups, 2-3 samples per session. The panelist were asked to

evaluate each sample as in previous sessions, they were reminded to sniff at different

times and distances from the cup. They were encouraged to provide a list of terms as

74
extensive as they could, and to give a definition for each term. Final discussion of the

terms assigned took place. Then the panelists were encouraged to sniff the samples again

to verify the presence of each of the assigned terms. The terms that all the panelists

perceived were kept on a final pooled list of terms with definitions. Training sessions 8, 9

and 10 lasted about 1 hr 15 min each. During the course of the group training and the

development of the list of attributes for describing rum, records of the attributes detected

by each panelist were taken as well as of those specific attributes some panelists had

problems detecting.

Figure 4.3 Group Training. Development of a List of Terms for Describing Rum Aroma.

75
 The final part of the group training consisted of presenting the panelist all the

possible references for the evaluation of rum. They were asked to decide whether the

aroma perceived in the references matches the aroma they perceived in rum. Different

options for some references were presented, and references were presented in 40%

ethanol solutions where possible. This is to mimic the whole complex rum sample. As a

group the panelists concluded on which references would be most appropriate for the

evaluation. They also agreed on the final definitions that would be used for describing

each term for rum aroma.

4.2.4 Selection and Preparation of Standards and Scale Setting

After the panel had approved and agreed on the proposed references for all the

rum attributes, the reference standards were prepared. For doing these, different solutions

were prepared and presented to the panel. The panel agreed that ethanol based standards

would best mimic the perceived complex aroma sensations in rum; therefore, the

standards were prepared in 40% (v/v) ethanol (80 Proof) solutions, where possible.

Different concentrations of each reference were prepared for every corresponding

attribute. Once a wide array of concentrations were prepared, a group of 3 expert sensory

analysts selected the concentrations that will be included as standards and assigned

intensity values to them.

The creation of the scale was done considering that the possible maximum

intensity for each attribute will have a value of 15 on a 15 cm line scale. All the rum

samples were evaluated and compared against each set of standard solutions to determine

the dilutions and intensity values that would be used as references.

76
 All the standards were contained in glass vials unless otherwise specified. A

volume of 5 ml was used in all glass vial standards. Ten samples of each of the standards

were prepared. This allowed having fresh standard samples required for training and

product evaluation. The standards were periodically checked and replaced with new vials

as needed to make sure the volatiles are present sufficiently. Constant checking on the

standards was a critical part as some of them were not stable under normal room storage

conditions.

Figure 4.4. Reference Standards Preparation.

77
Caramel. It was defined as the aroma characteristic associated with caramelized sugar,

and was prepared using artificial caramel extract (Firmenich de Mexico, Luisiana No. 80,

Napoles, 03810, Mexico City, Mexico) in a 40% ethanol (v/v) solution.

Table 4.1 Preparation of References for Caramel Aroma

Amount in 50ml (40% ethanol) Intensity on the scale

50 µl 4

75 µl 6

100 µl 8

12
120 µl

Medicinal. It was defined as a disinfectant- like aroma. A stock solution of 6.316g/l of

Thymol was prepared in 40% ethanol.

Table 4.2 Preparation of References for Medicinal Aroma

Amount Intensity on the scale

1/70 stock solution in 40% ethanol solution (90ppm) 3

1/50 stock solution in 40% ethanol solutio n (120ppm) 5

1/10 stock solution in 40% ethanol solution (631.6 ppm) 9

78
Butterscotch. It was defined as the aroma characteristic of butterscotch candy, and was

prepared using artificial butterscotch extract containing vainillin, diacetyl, benzodihydro

propane, valeric acid, butyric acid, ethyl formate, ethyl acetate, ethyl butyrate,

pyrolognous acid, methyl ciclobentenotone. Dilutions were made in a 40% ethanol (v/v)

solution.

Table 4.3 Preparation of References for Butterscotch Aroma

Amount in 50ml (40% ethanol) Intensity on the scale

15 µl 8

30 µl 10

50 µl 12

Honey. It was defined as the sweet aroma of clover honey, and was prepared using pure

clover honey (Deep South Products Inc. P.O. Box 1448 Fitzgerald, GA) in ethanol (40%).

Table 4.4 Preparation of References for Honey Aroma

Amount Intensity on the scale

10g and 50 ml (40%ethaol) 3

25g and 41.6 ml (40% ethanol) 6

35g and 35ml (40% ethanol) 12

Pure honey (undiluted) 15

79
Vanilla. Defined as the aroma characteristic of vanilla beans. Pure natural vanilla extract

(3x) from Mixim Labs (Calle Sur No. 6 P.O. Box 3, Naucalpan, Mexico, Mexico) in 40%

ethanol (v/v) solution was prepasred.

Table 4.5 Preparation of References for Vanilla Aroma

Amount in 50ml (40% ethanol) Intensity on the scale

75 µl 8

150 µl 10

300 µl 12

Woody. Defined as the aromatic characteristic of oak wood pieces. A stock solution of

10g of dried oak wood shavings in 100 ml of ethanol (40% v/v). The solution was

steeped for 48 hours and filtered using cheese cloth.

Table 4.6 Preparation of References for Woody Aroma

Amount in 50ml (40% ethanol) Intensity on the scale

5 ml 5

10 ml 10

20 ml 13

Stock (undiluted) 15

80
Fruity (artificial). Defined as an aroma associated with fruity characteristics that is

perceived as artificial, such as the aroma of a bag of mixed-candy. Prepared using ethyl

butyrate in 40% (v/v) ethanol solution.

Table 4.7 Preparation of References for Fruity-artificial Aroma

Amount in 50ml (40% ethanol) Intensity on the scale

1 µl 2

2 µl 4

3 µl 6

10
5 µl

Almond. Defined as the aroma of freshly crushed almonds. Prepared using McCormick

Imitation Almond extract, containing water, alcohol (36%), and benzaldehyde

(McCormick &Co. Inc. Hunt Valley MD). Dilutions were made in a 40% (v/v) ethanol

solution.

Table 4.8 Preparation of References for Almond Aroma

Amount in 50ml (40% ethanol) Intensity on the scale

100 µl 1

200 µl 2

500 µl 5

81
Plastic. It was defined as the aroma characteristic of plastic containers, and was prepared

using a new vinyl inflatable tire. The vinyl was cut to squares of ca. 1cm length. 16g of

vinyl squares were soaked in 50 ml 40% ethanol solution.

Table 4.9 Preparation of References for Plastic Aroma

Amount of solution Intensity on the scale

25 ml solution stepped for 5 h and 25 ml ethanol (40%) 5

Vinyl solution stepped for 5 h 8

Vinyl solution stepped for 24 h 10

Cinnamon. It was defined as the aroma of ground cinnamon bark, and was prepared

using McCormick ground cinnamon (McCormick & Co. Inc. Hunt Valley MD). A stock

solution of 1g ground cinnamon in 50 ml 40% ethanol (v/v) was prepared.

Table 4.10 Preparation of References for Cinnamon Aroma

Amount in (40% ethanol) Intensity on the scale

10 ml stock solution 3

20 ml stock solution 5

30 ml stock solution 7

Undiluted stock solution 8

82
Butter. It was defined as the aroma of fresh unsalted full- fat butter, and was prepared

using McCormick imitation butter flavor, containing water, propylene glycol, butyric

acid, and other organic acids, diacetyl and other ketones, ethyl propionate and other

esters, vainillin, and FD&C yellow #5 (McCormick &Co. Inc. Hunt Valley MD).

Dilutions were made in 40% ethanol solution.

Table 4.11 Preparation of References for Butter Aroma

Amount in 50ml (40% ethanol) Intensity on the scale

200 µl 3

400 µl 5

600 µl 7

Ethanol. It was defined as the pungent aroma characteristic of pure ethanol. Dilutions of

ethanol in water were prepared.

Table 4.12 Preparation of References for Ethanol Aroma

Amount of ethanol (v/v) Intensity on the scale

40% 4

60% 6

80% 12

100% 15

83
Green apple. It was defined as the aroma of green apples, and was prepared using green

apple artificial extract (Firmenich de Mexico, Luisiana No. 80, Napoles, 03810, Mexico

City, Mexico). Typically containing 6-hexanol and cis-3 hexenil acetate was diluted in

40% ethanol.

Table 4.13 Preparation of References for Green Apple Aroma

Amount in 50ml (40% ethanol) Intensity on the scale

1 µl 5

2.5 µl 8

3.75 µl 10

Smoke. It was defined as the aroma characteristic of smoke or smoked products, and was

prepared using Colgin Liquid Smoke. Containing water, natural hickory smoke flavor,

vinegar, molasses, caramel color and natural flavorings (The Colgin Companies, P.O.Box

7779, Dallas, TX., 75209). Dilutions in 40% ethanol (v/v) were prepared.

Table 4.14 Preparation of References for Smoke Aroma

Amount in 50ml (40% ethanol) Intensity on the scale

50 µl 2

150 µl 5

500 µl 8

Liquid smoke 15

84
Nutty. It was defined as the aroma of fresh ground pecans, and was prepared using

artificial pecan extract (Firmenich de Mexico, Luisiana No. 80, Napoles, 03810, Mexico

City, Mexico). Dilutions were made in 40% ethanol (v/v).

Table 4.15 Preparation of References for Nutty Aroma

Amount in 50ml (40% ethanol) Intensity on the scale

5 µl 2

10 µl 3

25 µl 4.5

Cardboard. Defined as the characteristic aroma of paper or cardboard. Prepared using

corrugated cardboard pieces of ca. 1 cm length. A stock solution was prepared with 40g

of cardboard pieces soaked in 100 ml ethanol (40% by vol.) for 20 h.

Table 4.16 Preparation of References for Cardboard Aroma

Amount in 50ml (40% ethanol) Intensity on the scale

25 ml 3

37.5 ml 5

Stock (undiluted) 10

85
Isopropanol. Defined as a chemical aroma foreign to rum. Prepared using 70% isopropyl

alcohol solution (by vol.) (Top Care 70% isopropyl alcohol, Topco Associates Inc.

Skokie, IL). Dilutions were prepared us ing water.

Table 4.17 Preparation of References for Isopropanol Aroma

Amount Intensity on the scale

15% 4

30% 7

50% 10

Prune -like. Defined as the aroma characteristics of dried fruit such as prunes and raisins.

Prepared using Sunsweet Bite Size Pitted Prunes. Dried Plums (Sunsweet Growers Inc.

Yuba City, CA). 10g plums and 50ml (40% ethanol by vol.) were pureed and filtered

with cheese cloth, the filtered liquid was recovered, and was assigned an intensity of 5 on

the scale. The second reference was prepared using 20g of prunes pureed with 50 ml of

ethanol (40% by vol.), the obtained pureed was assigned an intensity of 8 on the scale.

The third reference was 3 whole bite size pitted prunes in a covered plastic cup, the

assigned intensity on the scale was 12.

Pineapple. Aroma characteristic of canned pineapple. Prepared using Dole Pineapple

Slices in its Own Juice.(Dole Co). Dilutions of the decanted liquid were made using 40%

ethanol (by vol.) solution.

86
Table 4.18 Preparation of References for Pineapple Aroma

Amount in 50 ml (40% ethanol) Intensity on the scale

0.5 ml 3

0.75 ml 4

1 ml 5

Pepper. Defined as the characteristic aroma of ground pepper. Prepared using ground

white pepper (Deep South. Deep South Blenders Inc. Metaire, LA). Dilutions were

prepared with 40% ethanol (by vol.) solution.

Table 4.19 Preparation of References for Pepper Aroma

Amount in 50 ml (40% ethanol) Intensity on the scale

0.5 ml 3

0.75 ml 4

1 ml 5

Banana. It was defined as the aroma of a fresh-ground ripened banana, and was prepared

using mashed ripened banana. A stock solution was prepared using 20 g of banana puree

dissolved in 100 ml (40% ethanol) solution.

87
Table 4.20 Preparation of References for Banana Aroma

Amount of stock solution in 50 ml (40% ethanol) Intensity on the scale

20 ml 3

30 ml 5

undiluted 8

The banana references were stored in a freezer. The references could be

maintained in refrigeration and used the same day.

Ocean-like. It was defined as an aroma characteristic of sea breeze, and was prepared by

blending 1.5 g roasted seaweed (Nagai Nori Co, Ldt. 2-10 Ohmori-Naka 3Chome Ohta-

ku Tokyo 143 Japan) in 200 ml of 40% ethanol. That stock solution was diluted in 40%

ethanol (v/v) to produce the different reference intensities.

Table 4.21 Preparation of References for Ocean- like Aroma

Amount of stock solution in 50 ml (40% ethanol) Intensity on the scale

0.5 3.5

1 5

2 7

The reference for pungent was pure clover honey, for sweet aroma was brown

sugar, and for sour (acid) was white vinegar. These standards were contained in plastic

capped cups with ca. 5 ml each. No scaled reference for metallic aroma could be

88
produced given the instability of the samples. However, “metallic” aroma is defined as;

the aroma characteristic of metals or metal pieces such as keys, coins, and cans, and the

reference for the aroma were pulverized iron vitamin pills.

4.2.5 Individual Training Sessions. Use of References and Scales

The second part of the training consisted in individual training sessions. The

panelists were required to sniff all the references (Figure 4.5) that were used for the

evaluation of rum samples. The objective was that all panel members could become used

to and memorize the aroma of the references and the scale to be used for product

evaluation.

Figure 4.5 Standard References.

89
 For the individual training sessions, the panelists were allowed to come at their

most convenient time, but avoiding the times immediately before and after meals. They

were provided with all the standards, and coffee beans cups for aroma neutralization. The

panelists were asked to evaluate the set of standards for one attribute at a time, starting

with that standard with the lowest intensity value, and continuing with the one with

increasing intensity value. They were instructed to open the vials and take a with short

deep sniffs, tightly close the vial, then smell themselves, and the coffee beans cup before

moving to the next sample. Panelists were asked to evaluate the set of standards for all

the attributes every session, and to pay more attention especially on those set of standards

which corresponded to attributes they had problems detecting in the rum samples during

the group training. This part of the training consisted of at least 9 sessions that lasted 30

minutes to 45 minutes each, and took place in individual evaluation booths.

4.2.6 Panel

Of a total of 19 initial participants, 12 panelists finished all the group and

individual training required. The final composition of the panel consisted of 6 male and 6

female participants, ages 22 to 34, 75% of them were Hispanic, and 11 of them were

students of different departments at Louisiana State University, the other one was a

researcher.

4.2.7. Selection of Rum Samples for Evaluation

An extensive search over the internet, from rum literature, and in local markets

was done in order to select an array of 9 commercial rum samples produced from

different raw materials, using different processing protocols, produced in different

regions, and that are of all price ranges. The selected rum samples (Figure 4.6 and Table

90
4.22) represent the different types and qualities of rum evaluated in this study. The

selections include rums with no added spices or flavoring, as only true pure rum aroma

was intended to be evaluated. An additional experimental sample was also used for

evaluation. The experimental sample was made form molasses, and fermented over 72 hr.

The distillation method used was a continuous system with two columns, were the initial

“head” and final part “tail” of the distillation were removed. The raw distillate was aged

for 3 months with toasted American oak chips under accelerated conditions. After aging

it was stored in glass containers with plastic caps.

Figure 4.6 Rum Samples for Evaluation

91
Table 4.22 Rums Samples Selected for Evaluation

Code Name Origin Raw Distillation Proof Aging Aging conditions Mixing Refining Type
Material time
174 Mount-Gay Barbados Molasses Two types: 80 Not Charred American From different
Eclipse Continuous specific oak casks previously distillation
and Pot-stills (blend) used for Bourbon methods/ aged
up to 96% for different
EtOH lengths
217 St. James Martinique Cane Pot-stills 84 >10 Limousin oak casks Dark
Extra old juice/dunder
348 Barbancourt Haiti Cane juice Pot-stills 86 8 years Cognac method in Amber
double Limousin oak casks
distillation
565 Experimental Molasses 2 column 80 ~3 American oak chips
continuous months 35ºC
599 El dorado Demerara Local Continuous 80 12 45 gallon casks Continuous and Dark
Superior molasses and pot-stills batch
distillations
722 Castillo Gold Puerto Rico Molasses Continuous 80 ≥3 Amber
years
796 Flor de Cana Nicaragua Molasses 5 column 80 4 years Slow-aged White
Dry continuous (undisturbed) barrels
used once
813 Ronrico Extra Puerto Rico Molasses Continuous 80 >1 American oak casks Amber
Smooth
Premium Gold
Label
975 Myers Jamaica Molasses/ Pot-still 80 >5 White oak Dark
dunder
983 Ron Bacardi Puerto Rico Molasses Continuous 80 >1 American oak casks Double White
Dry charcoal
filter

92
4.2.8 Product Evaluation

A total of 9 commercial rums and an experimental samples were evaluated. For

each session of evaluation of rum samples, the panelists were provided with all references

used during training as well as 2 vials of two different rum samples (Figure 4.7). The rum

samples were placed in 30 ml closed glass vials coded with 3 digit random numbers with

10 ml of rum sample.

Figure 4.7 Evaluation of Rum Samples.

93
 For the evaluation of samples the judges were provided with a ballot containing

all the attributes. The attributes that needed to be quantified had a 15 cm scale marked

with decimal fractions; on the scale there were also marks for the intensities of the

references provided. See appendixes for an example.

Before the evaluation, each panelist was reminded of the evaluation procedure as

well as the use of the scale. Additional instructions were placed in each evaluation booth.

Evaluation took place under red right to minimize the effect of color on the samples

judging. The room was air-conditioned with positive air flow to prevent accumulation of

aromas. Panelists evaluated the samples at the most convenient time for them, avoiding

times before and after meals, and while in good health conditions. Each sample was

evaluated 3 times in random order. A total of 15 sessions were required to finish all the

evaluations. Each evaluation session lasted for approximately 40 minutes, resting for

about 5 minutes in between samples. After each session was completed, the evaluation

ballots were then checked for errors or missing data. Panelists were allowed to evaluate

another session after 15 to 20 minutes. The first ten sessions were completed after 3

weeks. Some panelists were able to complete sessions 11-15 immediately after the

previous ones, while others completed them 4 weeks after.

A total of 22 attributes were evaluated using the 15-cm line with the provided

references. The evaluation of the presence of the chemical aroma sensations; pungent,

sweet, and sour (acid), and the metallic aroma in the rum samples consisted on a yes/no

section of the evaluation ballot.

94
4.2.9 Analysis of Data

The panelists evaluated all the samples and assigned intensities to each of the

aroma attributes in every sample by marking a vertical line on 15 cm scale. The

perception of chemical sensations (pungent, sweet, sour (acid)), and metallic aroma in the

rum samples was analyzed by comparing the frequencies. This was done by assigning a

number 1 to a positive detection and 0 when the panelist did not perceived the sensation.

The obtained data was analyzed using analysis of variance (ANOVA) (SAS Institute Inc.

1999-2001) to check for significant differences in the intensities of each of the selected

aroma attributes in the rum samples. The varia tion in the responses for the intensities of

the aroma attributes was analyzed. The aim was to determine the reliability and

uniformity of the judgment criteria used by the panelist. This is also a good parameter to

determine the quality of the training sessions.

Multivariate Analysis of Variance (MANOVA) was performed to determine

overall differences in the rum samples based on all the aroma attributes used to describe

them. Further descriptive discriminant analysis was used to determine if the panelists

were able to discriminate the rum samples using the selected attributes. The attributes that

appear to have a major influence in the discrimination of the rum samples (ocean- like,

cardboard, artificial- fruity, honey, caramel, woody, cinnamon, vanilla, and buttery) were

selected.

The data for the selected attributes was used for further data analysis by Principal

component analysis (PCA) (SAS Institute Inc., 1999-2001). The PCA plots that indicate

the rum samples that are discriminated by the panelist from the group of samples were

95
made. The product-attribute plots were used to determine the aroma attribute that is

responsible for the discrimination. All statistical analyses were conducted at alpha=0.05.

4.3 Results and Discussion

The attributes selected for the training and evaluation sessions were based on the

consensus of the panel, and included the following terms woody, ethanol, buttery,

butterscotch, caramel, honey, plastic, smoke, vanilla, almond, cinnamon. artificial- fruity,

banana, prune, medicinal, pineapple, pepper, green apple, nutty, isopropanol, cardboard,

ocean- like, metallic, and the chemical aroma sensations; pungent, sweet, sour. Some of

the terms developed by the panel that did not reach consensus include berries, chemical,

clove, coconut, grape, green, peppermint, clove, and coconut.

4.3.1 ANOVA Results

The data for the intensity of the aroma attributes in all the rum samples was

analyzed using analysis of variance (SAS Institute Inc., 1999-2001) to determine if there

were significant differences in the judgments for the intensity. The deviations in

judgment from the mean intensity assigned to each attribute, and every sample were

reported (Tables 4.23, 4.24, and 4.25).

The rum with the longest aging time (12 years) was perceived significantly more

“woody” (9.53) than light rums aged for 2 years such as the two rum samples from

Puerto Rico (6.28 and 5.98). The lowest woody intensity (5.10) was assigned to a white

rum from Nicaragua aged for 4 years; it had a significantly lower woody aroma as

compared with a rum from Martinique aged for 10 years, a rum from Demerara aged for

12 years, and a dark rum from Jamaica aged for at least 5 years. From these observations

96
we can conclude that the aging type as well as the time of rum (i.e., dark vs. amber or

white) are related to the intensity of the “woody” aroma.

Table 4.23 Intensity of the Aroma Attributes (Part I) a

Woody Ethanol Buttery Butter- Caramel Honey Plastic Smoke Vanilla Almond Cinnam.
scotch
ABC BA A
174 7.44 8.42 3.71 4.68A 4.49BA
2.91BCA
2.36 A
1.25 BA
5.91BA
1.44 A
1.68A
(3.94) (3.94) (2.39) (3.44) (2.58) (2.40) (2.50) (1.57) (3.44) (1.65) (1.69)
BA BA A A BA BAC A BA B A
217 8.4 8.85 2.69 2.82 4.32 3.68 2.68 1.54 4.48 1.37 2.32A
(4.16) (3.33) (2.42) (2.84) (2.71) (3.10) (2.76) (1.65) (3.16) (1.49) (1.95)
BCA BA A A BA BAC A BA BA A
348 6.42 7.91 2.66 3.51 4.24 2.59 2.75 1.67 4.49 1.28 1.26A
(3.76) (3.34) (2.01) (2.96) (2.96) (2.46) (2.89) (1.55) (3.27) (1.55) (1.42)
BCA B A A A BA A BA A A
565 7.87 6.65 3.64 4.91 5.63 4.43 2.75 1.24 7.05 1.6 1.83A
(4.17) (3.62) (2.57) (3.25) (3.09) (3.08) (2.87) (2.58) (3.83) (1.87) (1.75)
A BA A A BA BAC A A BA A
599 9.53 8.18 2.31 3.61 4.77 3.59 2.33 1.91 4.99 1.54 2.22A
(3.89) (3.29) (2.01) (3.08) (2.63) (3.11) (2.54) (2.03) (2.98) (1.73) (1.81)
BCA BA A A BA BAC A B BA A
722 7.48 7.93 2.46 2.92 4.03 3.31 2.14 0.7 4.57 1.4 1.49A
(4.17) (2.35) (2.15) (2.66) (2.71) (2.25) (2.66) (2.71) (2.25) (2.27) (0.66)
C BA A A B C A B B A
796 5.10 8.8 2.19 2.87 3.14 1.72 2.3 0.72 4.05 1.03 1.15A

(3.5) (3.48) (2.25) (2.85) (2.77) (1.67) (2.17) (1.18) (2.88) (1.29) (1.37)
BC A A A BA BC A BA BA A
813 5.98 9.54 2.11 3.09 3.92 2.39 1.86 0.89 4.69 0.98 1.30A
(3.64) (3.48) (1.79) (3.48) (3.12) (2.55) (2.13) (1.11) (3.06) (1.24) (1.55)
BA BA A A BA A A BA BA A
975 8.76 8.38 2.76 3.41 5.22 4.59 3.71 1.52 5.43 1.41 2.18A
(4.49) (3.33) (2.37) (3.46) (3.07) (4.11) (4.12) (1.78) (3.37) (1.67) (1.81)
BC BA A A B BC A BA B A
983 6.28 8.08 2.06 2.53 3.28 2.32 1.99 1.18 3.81 1.18 1.25A
(4.11) (3.12) (1.92) (3.07) (2.53) (2.24) (2.09) (1.44) (3.39) (1.53) (1.62)

a
Numbers in parenthesis are the standard deviations. For each column, means with the
same superscript letter are not significantly different (p≥0.05).

The highest ethanol aroma intensity (9.54) corresponded to that of an amber rum

from Puerto Rico distilled in continuous distillation and aged for over 1 year. The second

highest intensity was assigned to a rum from Martinique, that had the highest ethanol

content. It was expected that the highest ethanol concentration would be related to the

perception of ethanol aroma, but also a very light rum that does not have a complex

97
aroma mixture may allow for a more obvious detection of ethanol. The perception of

lowest ethanol intensity (6.65) was in the experimental sample. A possible explanation

for this was that it possessed a very complex aroma.

There was not a significant difference in the intensity of the “buttery” aroma of

the samples, however the highest intensity (3.71) was perceived in a rum from Jamaica

made from molasses, pot-stilled and aged for over 5 years. The lowest intensities (2.06,

2.11, and 2.19 respectively) were assigned to a white rum from Puerto Rico, an amber

rum from Puerto Rico, and a white rum from Nicaragua, all of them made from molasses

and distilled in continuous systems, those samples typically have short fermentation and

aging times.

The experimental sample was significantly higher in the “caramel” aroma

intensity (5.63), followed by a rum made in Jamaica, from molasses and pot stilled, and a

rum from Demerara made from molasses with a blend of continuous and batch

distillations and aged for 10 years (5.22, and 4.77, respectively). Lower intensities were

perceived in a white rum, and amber rum both from Puerto Rico (3.28, and 2.89). The

lowest “caramel” aroma intensity belonged to a white rum from Nicaragua. Similar

tendencies were observed for the “honey” aroma attribute, where the dark Jamaican rum

was perceived as having a statistically higher value than those of the white rums (1.72,

and 2.32 for the Nicaraguan and Puerto Rican rums, respectively), as well as the rum

from Haiti made from cane juice, pot stilled and aged for over 3 years (2.39). Higher

“honey” aroma intensity was also observed in the experimental sample (4.43). A

significantly higher “smoke” aroma was perceived in the rum from Demerara aged for 12

years, as compared to the amber Puerto Rican rum aged for over 1 year (0.7), and the

98
white rum from Nicaragua (0.72). The experimental sample (aged with toasted American

oak chips) had significantly higher intensity of “vanilla” aroma (7.05) than those of the

Puerto Rican white rum (3.81), the Nicaraguan white rum (4.05), and the rum from Haiti

aged for over 3 years (4.48). The perception of “vanilla” attribute seems to be related to

the aging time. However it is not aging time alone, but also aging conditions responsible

for the “vanilla” aroma intensity. This is the case of the rum from Demerara aged for 12

years. A possible reason for these may be the use of uncharred barrels. As it was stated in

the introduction, lignin degradation in the barrels that leads to the formation of aldehydes

(such as vainillin) is higher in charred barrels as compared to uncharred ones (Nishimura,

1983). The toasting temperature also plays a role, as the temperature used increases so

does the formation of those compounds. The type of oak wood used may also influence

the formation of oak degradation compounds. This coincides with the idea that wood

treated at higher temperatures will possess stronger vanilla notes as well as others

(Francis et al., 1992). It was observed that the intensity of the “vanilla” aroma in those

rums aged in Limousin oak casks (the ones from Martinique and Haiti) was in general

lower than that of the rums aged in American oak casks. In the perception of the almond

aroma we observed the same pattern, where the amber Puerto Rican rum aged for 1 year,

the white rum from Nicaragua, and the white Puerto Rican rums were given the lower

intensities (0.98, 1.03, and 1.18 respectively), and while the differences were not

statistically significant, the higher values corresponded to the experimental sample (1.6),

and the rum from Demerara aged for 12 years. “Cinnamon” aroma also appears to be

related to the aging conditions. The higher cinnamon aroma intensities were perceived in

the rum from Martinique aged for over 10 years, and the rum from Demerara aged for 12

99
years, followed by the rum from Haiti aged for 8 years, all of them aged in Limousin oak

casks using the cognac method. As expected, the lower intensities were observed in the

white Nicaraguan and Puerto Rican rums (1.15 and 1.25, respectively).

There was no observed significant difference in the perception of plastic aroma,

but the higher value was assigned to the Jamaican rum (3.71), followed by the

experimental sample (2.75). Both samples were contained in bottles with a plastic cap,

therefore that may be a possible source for the off-aroma transfer.

The rum sample having the statistically strongest “artificial fruity” aroma was the

experimental sample. This sample was produced from molasses, with long fermentation

times and distilled in a double column continuous system. The processing method had

less controls and lacked the rectification of the composition one would expect in a

commercial setting. This may have allowed for the formation of a distillate containing a

larger variety of aroma compounds such as esters or their acid precursors, and in

quantities larger than those in other distillates. Such composition will promote the

formation of fruity- like aromas. There were not significant differences in the artificial

fruity aroma of the other rums, however the weakest perceived aroma corresponded to

that of the white Nicaraguan and Puerto Rican rums (1.15, and 1.25 respectively).

Typically light rums are produced in short fermentation times, therefore less aroma

compounds are formed, and they are distilled in continuous systems, where they are

stripped of many congeners. The experimental sample was also assigned the highest

banana aroma intensity (1.41), and it was significantly higher than those of the rum from

Haiti made from cane juice and aged for over 3 years (0.31), as well as the ones for the

100
Nicaraguan and Puerto Rican white rums (0.39 and 0.31, respectively), both of them

made from molasses and distilled in continuous systems.

Table 4.24 Intensity of the Aroma Attributes (Part II)a

Artif. Banana Prune Medici - Pine- Pepper Green Nutty Isopro Card- Ocean-
Fruity nal apple apple panol board like
174 2.03B 0.84 BA
2.56BA
0.91A 1.57A 1.12 A
2.64A 1.58 A
1.72A 0B 0B
(2.68) (1.62) (3.21) (1.01) (1.54) (1.08) (2.73) (1.29) (2.54) (0.0) (0.0)

217 1.77B 0.41BA 3.69A 1.42A 1.63A 1.62A 1.96A 1.85A 2.31A 0B 0B
(1.84) (1.01) (4.04) (1.59) (1.73) (1.49) (2.29) (1.37) (2.67) (0) (0)
B BA BA A A A A A A A
348 1.75 .67 2.29 1.34 1.18 1.43 1.62 1.41 2.23 2.44 0B
(1.92) (1.24) (3.09) (1.64) (1.28) (1.1.9) (2.21) (1.44) (2.31) (2.49) (0)
A A BA A A A A A A B
565 4.33 1.41 2.04 1.68 1.85 1.24 2.67 1.75 1.24 0 0B
(3.12) (2.12) (4.17) (3.62) (1.66) (1.37) (2.15) (1.31) (1.98) (0) (0)

599 1.54B 0.31B 3.82A 1.25A 1.19A 1.48A 2.05A 1.84A 1.59A 0B 0B
(2.07) (0.67) (4.4) (1.53) (1.60) (1.39) (2.4) (1.22) (2.35) (0) (0)
B BA BA A A A A A A B
722 2.13 0.52 1.98 1.05 1.66 1.09 2.72 1.5 1.81 0 0B
(3.17) (1.42) (2.5) (1.31) (1.65) (0.97) (2.37) (1.36) (1.97) (0) (0)
B B BA A A A A A A B
796 2.05 0.39 1.54 1.58 1.74 1.09 2.44 1.26 1.92 0 2A
(2.18) (0.7) (1.77) (1.59) (1.60) (1.23) (2.06) (1.27) (1.93) (0) (1.83)
B BA B A A A A A A B
813 2.09 0.95 1.19 1.44 1.20 1.15 1.98 1.18 1.55 0 0B
(2.24) (1.53) (1.88) (1.70) (1.69) (1.16) (2.85) (1.51) (1.53) (0) (0)

975 1.37B 0.77BA 3.69A 1.82A 1.59A 1.62A 2.35A 1.44A 1.81A 0B 0B
(1.79) (1.74) (4.01) (2.58) (1.9) (1.46) (2.82) (1.29) (2.39) (0) (0)
B B A A A A A A B
983 1.64 0.31 2.05 1.89 1.24 1.60 2.11 1.12 2.48 0 0B
(1.83) (0.47) (2.39) (1.78) (1.30) (1.32) (1.77) (1.25) (2.61) (0) (0)

a
Numbers in parenthesis are the standard deviations. For each column, means with the
same superscript letter are not significantly different (p≥0.05).

Those rums that had the higher “prune” aroma intensity were the rum form

Demerara, blend of continuous and batch distillations and aged for 12 years (3.82), and a

Jamaican rum distilled in pot-stills rum aged for over 5 years, both of them made from

molasses. A possible reason for the high “prune” or “raisiny” aroma could be that the

101
wood used for maturation was not seasoned in a warm and dry climate, or it was not

subjected to any additional high temperature treatment. Francis et al, (1992) stated that

treatment of the wood using high temperatures reduces the “raisiny” notes of the wood.

The prune aroma in the Demerara and Jamaican samples was statistically higher than that

of the amber Puerto Rican rum aged for 1 year (1.19). The values in the Nicaraguan white

rum (1.54) and the other Puerto Rican Gold rum aged for 3 years (1.98) also appeared to

be low. The experimental sample was perceived as having the highest intensity for

pineapple aroma among the rum samples, however the difference was not significant.

Table 4.25 Frequency of the Perception of some Aroma Sensations a

Pungent Sour Sweet Metallic

BA BA BA
174 0.73 0.36 0.79 0.45BA
(0.45) (0.49) (0.42) (0.51)
A BA B
217 0.88 0.45 0.58 0.64BA
(0.33) (0.51) (0.50) (0.49)
BA BA B
348 0.7 0.45 0.61 0.42BA
(0.47) (0.51) (0.5) (0.5)
B B A
565 0.53 0.16 1 0.25B
(0.51) (0.37) (0) (0.44)
BA BA BA
599 0.79 0.36 0.76 0.7 A
(0.42) (0.49) (0.44) (0.47)

722 0.64BA 0.36BA 0.82BA 0.45BA

(0.49) (0.49) (0.39) (0.51)

796 0.76BA 0.58A 0.64B 0.48BA

(0.44) (0.5) (0.49) (0.51)
BA BA BA
813 0.82 0.42 0.67 0.39BA
(0.39) (0.5) (0.48) (0.5)
BA BA BA
975 0.82 0.33 0.76 0.45BA
(0.39) (0.48) (0.44) (0.51)
BA BA B
983 0.64 0.52 0.64 0.39BA
(0.49) (0.51) (0.49) (0.5)

a
Numbers in parenthesis are the standard deviations. For each column, means with the
same superscript letter are not significantly different (p≥0.05).

102
 There were no observed significant differences in the intensities of the

“medicinal”, “peppery”, “green apple”, “nutty”, and “isopropanol” aromas. The panelists

perceived a slight (2.4) cardboard aroma in the rum from Barbados made from molasses,

and distilled in pot-still, while the white rum from Nicaragua had an ocean- like aroma (2)

(Table 4.25)

The sample more frequently perceived as having a pungent aroma was the rum

from Martinique, which is the dark rum having a ethanol content of 84 %. This suggests

that there may be a relationship between the alcohol content and the pungent aroma.

However, the presence of other compounds in the rum may contribute to the pungent

aroma either by masking the pungency of alcohol, or by increasing that perceived aroma.

Dark rums have very complex aromas and are normally perceived as more pungent than

white rums. White rums are in general characterized by a light taste. The rum most

frequently perceived as having a “sour” aroma was the white rum from Nicaragua,

followed by the white rum from Puerto Rico. The former one was statistically more

frequently perceived as “sour” than the experimental sample, and this latter one having

the least sour aroma perception. This may be directly related to their acid make-up. As

with those short aging times, most of the acids that will normally be transformed into

other chemical compounds after long maturation times (i.e., esters) may have remained

intact. The aroma of the experimental sample was perceived as “sweet” by all panelists.

Its sweet aroma was significantly different than that from the rums from Martinique and

Haiti, both made from cane juice and distilled in pot-stills, and it was also different from

that of the white rums from Nicaragua and Puerto Rico. It can be concluded that the raw

material used has an effect on the perceived “sweet” aroma of the samples. Rums made

103
from molasses and aged for a long period are more frequently perceived as having a

sweet aroma as compared to tho se made from cane juice. White rums typically have short

aging times and lower concentration of congeners as compared to aged amber or dark

rums. The aroma of rum from Demerara made from molasses and aged for 12 years was

perceived as more “metallic” than the experimental sample. The experimental rum was

aged by an accelerated method using American oak chips for three months. This method

allowed for the extraction of large amounts of wood components into the distillate;

however, it undoubtedly had the shortest aging time, and it was not aged in casks, as the

other distillates were. The aroma of the experimental sample was less frequently

perceived as “metallic”, and it was significantly different in its’ “metallic” aroma from

those rums aged for longer periods. There appears to be a direct relationship between the

time the distillate was maturating in the oak casks and the metallic aroma of the final

rum.

Several problems contributed to the large variations in the results. One possible

cause is the high complexity of the sample; rum contains about 40% ethanol, the ethanol

content, as well as the general chemical composition will affect the aroma perception of

specific aroma attributes. The complexity of the aroma of rum was also noticeable as

seen by in the large amount of possible aroma descriptors the panelists assigned to rum.

Handling such a large number of variables represented a problem for the analysis and

interpretations. It also represented a problem during training and evaluation itself; the

panelist had to smell the rum samples and the references for a very extended period of

times to complete each session. In such conditions, it is very possible that the panelists

were fatigued and may have been less sensitive due to affect the perception of aroma by

104
adaptation. Another possible cause of variation is the mere task of evaluating aroma.

Rum contains a large number of volatile compounds that contribute to the aroma, in this

case, the judges had to smell the sample and inhale the composition of the headspace of

the bottle that contained the sample. It is not possible to provide a constant headspace

composition even for the same sample, given that it will be affected by settling time and

temperature, and eventually if opened and exposed to environment for long time, the

initial aroma of the sample will no longer be maintained. According to Cain et al., (1992)

the concentration of the headspace varies proportionally with the odorant in the solution

by a factor known as activity coefficient. This factor will differ among odorants diluted

with the same solvent, among solvents, and among concentrations of the same odorant-

solvent pair. Also, when subjects smell the headspace of a liquid solution contained in a

recently opened bottle, some amount of air will be inhaled, causing the original stimulus

to be diluted. According to Dravenkis (1975) a bottle left still will regain its headspace

after 30 minutes or faster if the bottle is shaken. Cain et al (1992) stated that the problem

of the equilibrium of the headspace is solved when 2 bottles are provided for alternative

use. In the case of this study, 5 evaluation booths with the complete set of standards were

used. Two sets of samples for evaluation were provided. Each set contained two

containers for the same sample, so that the panelists could use it alternatively while

evaluating for each and every attribute. The sets of samples were also used alternatively.

Both the rum samples and the reference standards were periodically checked and replaced

with new ones (refer to the methodology section). By providing many vials for the same

sample the problem of having a representative headspace of the sample was minimized.

The detailed instructions provided aimed to minimize the variation in the method of

105
evaluation used by each of the panel member. However, total control over those

differences is difficult to achieve. The characteristics of this study, using a sample with

such complex composition as in rum, and the complexity of the evaluation task itself

demand a more intensive and extensive training in order to obtain results that are more

significant, and that present less variation between panelists, and in the replications

within the same panel member.

4.3.2 Overall Samples Differences

MANOVA was performed to determine if the overall differences existed in the

rum samples differences based on all the aroma attributes used to describe them. The P

value on the Wilk’s Lambda (0.0001) revealed that the aromas of all 10 rum samples

evaluated was different (Table 4.26).

Table 4.26 Multivariate Statistics and F Approximations

MANOVA Test Criteria and F Approximations for the Hypothesis of No Overall Form

Effect

H= Type III SSCP Matrix for Form

E= Error SSCP Matrix

S=9 M=-0.5 N=151.5

Statistics Value F Value Num DF Den DF Pr>F

Wilk’s Lambda 0.15802275 8.08 81 1980 <.0001

Pillai’s Trace 1.39254391 6.37 81 2817 <.0001

Hottelling- Lawley Trace 2.57477145 9.65 81 1304.5 <.0001

Roy’s Greatest Root 1.19600156 41.59 9 313 <.0001

106
 Discriminant analysis was used to determine the attributes responsible for the

perceived differences. As it is shown in Table 4.27, in the first dimension of the pooled

within canonical structure, “ocean- like” aroma (0.931) significantly contributed to the

overall difference among rum samples. In the second canonical structure dimension,

“cardboard” aroma appears to be responsible for the differences among samples. Further

canonical structure dimensions reveal other aroma attributes that contribute significantly

to the differences among samples. The most important attribute that contributes to

differences among samples in the third canonical structure is “artificial fruity”.

Table 4.27 Canonical Structure r’s Describing Group Differences Among Samplesa

Variable Can1 Can2 Can3 Can4 Can5

Woody -0.201 -0.146 -0.256 0.538 0.241

Buttery -0.060 -0.012 0.297 0.342 0.755

Caramel -0.140 -0.061 0.121 0.643 0.286

Honey -0.169 -0.103 -0.057 0.687 0.169

Vanilla 0.081 -0.082 0.361 0.538 0.474

Cinnamon -0.102 -0.116 -0.222 0.483 0.228

Art. Fruit 0.011 -0.043 0.680 0.537 -0.141

Cardboard -0.093 0.994 -0.054 0.261 -0.082

Ocean- like 0.931 -0.007 -0.138 0.289 0.059

a
Based on pool-within class variances. Highlighted numbers indicate aroma attributes
which largely account for group differences in each dimension.

107
4.3.3. Principal Component Analysis

Figure 4.8. Plot of Prin2*Prin3.

*
The symbols correspond to the value of the product. Sample 813 is product A, sample
975 is product B, sample 565 is product D, sample 722 is product E, sample 174 is
product F, sample 983 is product G, sample 217 is product H, sample 599 is product K,
and sample 796 is product M.

108
Figure 4.9. Plot of Prin3*Prin1.
*
The symbols correspond to the value of the product. Sample 813 is product A, sample
975 is product B, sample 565 is product D, sample 722 is product E, sample 174 is
product F, sample 983 is product G, sample 217 is product H, sample 599 is product K,
and sample 796 is product M.

109
 Principal component analysis was performed with the data for all samples. In the

plots Prin2*Prin1 (Figure 4.8) and Prin3*Prin1 (Figure 4.9) we can observe that run 348

(D) and 796 (M) are different from the rest of the samples. Those rums appeared clearly

separated while rest of the samples are scattered in between in the middle section of the

plots. With the attributes plotted as vectors in the same principal component for sample

plots, it is possible to determine that there is one attribute responsible for the

discrimination of each of the two clearly distinct rums. Those attributes are “ocean- like”

aroma for rum 796 (M), and cardboard aroma for sample 348 (D). Sample 796 was a

white rum from Nicaragua made from molasses. What was peculiar of this sample was

the use of a continuous distillation using 5 columns, and an undisturbed “slow-aged”

method of aging that lasted for 4 years. Sample 348 was original of Haiti, was made of

cane juice, double distilled in pot-still distillations, and aged in Limousin oak casks for 8

years. It is important to note that not only that those rums where clearly discriminated

from the rest of the samples, but they appear to have the most differences in the perceived

aroma.

4.4. Conclusion

Evaluation of rum, as with any other highly alcoholic beverage faces many

problems, one of them is the impairment of the judges’ ability to perform due to such

quantities of ethanol will give the judge after tasting a large number of samples. This

study proves that the mere aroma perceived by controlled sniffing of the rum samples can

be used to evaluate and describe rums, and that such information can be used to

discriminate between rum samples. Other problem faced when evaluating of rum is the

overpowering effect that the aroma of ethanol may have on the samples. The use of

110
ethanol-based standards, with ethanol content similar to that in rum samples was an

effective way to train panelists on the aroma of rum, and was also useful for the creation

of adequate references that best mimic the perceived aroma of rum.

The developed method for descriptive analysis of rum products was effective in

analyzing a larger number of samples than an aroma and taste evaluation would have

allowed for, also it presented an advantage in recruiting panelists. People, who for

different personal, cultural, or religious reasons, would not be part of a rum evaluation

that required tasting can be part of an aroma evaluation test such as the one conducted for

this thesis research. The use of this method with adequate training can provide

comprehensive information on the aroma of rum.

With this descriptive analysis method and the developed list of aroma attributes, it

is possible to discriminate the aroma of different rum samples. Significant differences in

the intensities of attributes “woody”, “ethanol”, “caramel”, “honey”, “smoke”, “vanilla”,

“banana”, “prune”, “cardboard”, and “ocean-like” were found in the evaluated rum

samples. There were also significant differences in the frequency of the perception of

“pungent”, “sour” (acid), and “sweet” chemical sensations, and “metallic” aroma. The

most important attributes that have a significant effect in discriminating among samples

were: “ocean- like”, “cardboard”, “artificial- fruity”, “honey”, “caramel”, “woody”,

“cinnamon”, “vanilla”, and “buttery”.

It is possible to analyze the information on the differences between samples and

relate back the processing protocol of the different rum samples. This information can be

a very useful tool for the development of new rum products that outstand the rest of the

samples in a specific desirable aroma attribute, or that poses a unique aroma. The use of

111
rum descriptive analysis with the method used in this study can also be a good tool for

the quality control of rum products, can be useful for brand identification, and can also be

a helpful marketing tool.

112
 CHAPTER 5. GENERAL CONCLUSIONS

Rum production constitutes a value addition of molasses with low economic

value. It is important to determine the optimal conditions that yield desirable rum

production with desirable aroma quality. The number of compounds present in rum is so

large that it is impossible to single-out one component responsible for the aroma of rum,

but it is important to know how the composition of volatile compounds, of and thus the

aroma of rum will be impacted by the many steps involved in rum production.

Even though the raw material naturally influences the aroma of rum, the core of

the aroma is formed during fermentation by the action of yeast, and also affected by the

presence of bacteria. Other processing conditions have a major influence on the final

quality of rum. Distillation affects the aroma of rum by altering the proportion of the

aroma compounds originally present in the fermentation mash. The type of distillation, as

well as the distillation temperature largely influence the aroma composition of the raw

distillate. Reactions within the rum components, with wood components, and extraction

of wood components alter the composition of the distillate, and have a large impact on

the aroma and other sensory qualities of rum. The conditions of the maturation process,

such as aging temperature, type of wood, and condition of the wood, as well as the aging

time influence the perceived aroma of the final product.

As rum is a product with so many different compounds responsible for its aroma,

and some of those compounds are present in a very small amount; it presents a problem

for using chemical analysis alone for determining, achieving, and controlling the quality

of rum. The developed list of descriptors for the aroma, taste and flavor of rum can be a

good tool to describe a large variety of rum products.

113
 The descriptive analysis of rum as used in this study was useful in describing

different rum products, and to discriminate among rum samples. However, it is strongly

recommended, for further studies, to the use of a reduced number of attributes, and to

maintain the most important ones in order to make the training, evaluation, data handling,

and interpretation easier, and to provide more significant information. “Ocean- like”,

“cardboard”, “artificial- fruity”, “honey”, “caramel”, “woody”, “cinnamon”, “vanilla”,

and “buttery” are the most significant attributes responsible for the discrimination among

samples. However, significant differences in the intensity of attributes such as “smoke”,

“banana”, “prune”, and in the frequency of perception of “pungent”, “sour” and “sweet”

chemical sensations, and “metallic” aroma were also found.

The use of a larger panel size, and more intensive training is also recommended

for further studies, given the complexity of rum, which made the description of the aroma

a difficult task.

114
 REFERENCES

Amerine, M. A. and Roessler, E. B. 1983. Wines: Their Sensory Evaluation. W. H.

Freeman and Co. New York, NY.

Anholt, R.H. 1992. Molecular aspects of olfaction. In “Science of Olfaction” (Serby,

M.J., and Chobor, K.L. eds.). Springer-Vela. New York. pp 51-79.

Arroyo, R. 1945. The production of heavy bodied rum. Sugar, 40:34-39.

ASI, 2000. Molasses Survey. Audubon Sugar Institute. Lousiana State University
Agricultural Center.

Barnett, J.A. 1976. The utilization of sugars by yeast. Advances in Carbohydrate

Chemistry and Biochemistry, 32: 125.

Berry, D. R. and Watson, D. C. 1987. Production of organoleptic compounds. In “Yeast

Biotechnology” (Berry, D.R., Russell, I., and Stewart, G.G., eds.) Allen and Unwin,
London. pp. 345-368.

Berglund, B., Berglund, U., and Lindvall, T. 1978. Olfactory self- and cross-adaptation:
Effects of the time of adaptation on odor intensity. Sensory Processes, 2: 291-297.

Bloomfield, L., 1939. Linguistic aspects of science. In “ International Encyclopedia of

the United States” Vol. 14 (Neurath, O., Carnap, R., and Morris, C., eds.)University of
Chicago Press. Chicago. From Lehrer A. 1983. Wine and Conversation. Indiana
University Press. Bloomington, IN.

Blackburn, F. 1984. Sugar-cane. Longman. New York.

Bluhm, L. 1983. Distilled beverages. In “Biotechnology” (Rehm, H.J., and Reed, G, eds.)
Vol. 5.Vela Chimie, Weinheim, pp. 447-475.

Bowen H.J.M. 1966. “Trace elements in Biochemistry”. Academic Press, London.

Boyazoglu, J.C. 1986. Wine Talk and Linguistics. J. Am. Wine Soc. Winter 1986.

Cain, W.S., Comettio-Muñiz, E., and DeWijk, R. A. 1992. Techniques in the Quantitative
Study of Human Olfaction. In “Science of Olfaction” (Serby, M.J., and Chobor, K.L.
eds.). Springer-Verlag, New York, pp. 179-308.

Cantarelli, C. 1989. In “Biotechnology Applications in Beverages Production”

(Cantarelli, C., and Larzarini, G. eds.) Elsevier Applied Science, London, pp. 127-151.

Carr, J. G. 1974. Aroma and Flavour in Winemaking. Mills & Boon Limited, London.

115
Casey, G.P., and Ingledew, W.M. 1986. Ethanol tolerance in yeasts. CRC Critical
reviews in Microbiology. 13:219.

Chen, J.C.P, and Chu, C. 1993. Cane Sugar Handbook: a Manual for Cane Sugar
Manufactures and their Chemists. John Wiley and Sons, Inc. New York. pp 1090.

Civille, G. V. and Lyon, B. G. 1996. Aroma and Flavor Lexicon for Sensory Evaluation:
Terms, Definitions, References and Examples. ASTM Data Series Publication DS 66.
West Conshohocken, PA.

Clutton, D. W. 1974. Rum. The flavor Ind. Nov/Dec: 286-288.

Dahl, A. R. 1988. The effect of cytochrome P450-dependent metabolism and other

enzyme activities in olfaction. In “Molecular Neurobiology of Olfactory System:
Molecular, Membranes, and Cytological Studies” (Margolis, F.L., and Getchell, T.V.,
eds).Plenium Press, New York, pp. 51-70.

Dahl, A.R., Hadley, W. M., Hahn, F.F., Benson, J.M, and McCellan, R.O. 1982.
Cytochrome P450-dependent monoxygenases in olfactory epithelium of dogs. Possible
role in tumorigenecity, Sience. 216: 57-59.

Dravnieks, A. Instrumental aspects of olfactometry. In “Methods in Olfactometry

Research” (Moulton, D.G., Turk, A., and Johnston, J.W. eds.) New York. Academic, pp
1-58.

Estrem, S.A. and Renner, G. 1987. Disorders of smell and taste. Neurologiacl Disorders
in Otolaryngology, 20. 133-147.

Fahrasmane, L., Ganou-Parfait, B. 1998. Microbial flora of rum fermentation media. J.

App. Microb. 84: 921-928.

Fahrasmane, L., Ganou-Parfait, B., and Bazile, F. 1989. Le metabolisme du soufre dans
la rhumerie. MIRCEN J. 5 (5): 239-245.

Francis, I. L., Shefron, M. A., And Williams, P. J. 1992. A study by sensory descriptive
analysis of the effects of osk origin, seasoning, and heating on the aroma of oak model
wine extract. Am. J. Eno. Vitic. 43: 23-30.

Gadd, G.M., Chalmers, K., and Reed, R.H.1987. The role of trehalose in dehydration
resistance of Saccharomyces cerevisiae. FEMS Microbiology Letters 48: 249.

Gancedo, C. and Serrano, R. 1989. Energy-yielding metabolism. In “The Yeasts” (Rose,

A.H., and Harrison, J.S., eds.), 2nd edn. Academic Press, London, pp. 205-259.

116
Garnier-Larroche, G. and Cottrell, R. H., 1975. Paper presented in ‘International
Symposium on Rums, Alcohols and Alcoholic beverages from Sugar Cane’ held in
Martinique, 1975.

Getchet, T.V., Margolis, F.L., and Getchell, M.L., 1984. Perireceptors and receptor
events in the vertebrate olfaction. Prog. Neurobiol. 23: 317-345.

Haraldson, A., and Bjorling, T.1981. Yeast strains for concentrated substrates. Eur. J. of
App. Microbiol. Biotechnol. 13(1): 34-38.

Harrison, J.S., and Graham, J.C.J. 1970. In “The Yeasts” (Rose, A.H., and Harrison, J.S.
eds.) 1st edn. Academic Press, London, pp. 283-348.

Heide, ter R., Schaap, H., Wobben, H. J., Valois, P. J. de, and Timmer, R., 1981. Flavor
constituents in rum, In ‘The quality of Foods and Beverages’, (Charalambous, G., and
Inglett, G., eds.). Academic Press, New York. 183-200.

Hillis, W. E. 1984. High temperature and chemical effects on wood stability. Wood Sci.
Technol. 18: 281-293.

Hornugg. D.E. and Mozell M.M. 1980. Titrated odorants to monitor retention in the
olfactory and vermonasal organs. Brain Res. 181: 488-492.

Hunsigi, G. 1993. Production of Sugarcane. Springer-Verlag, Germany.

Johnston, J. R. 1990. Brewing and distilling yeast. In “Yeast technology” (Spencer, J. F.

T, and Spencer, J. F. T, eds.). Springer-Velag. Berlin. Germany.

Jones, R.M., Russell, I., and Stewart, G.G. 1986. American Society of Brewing Chemists
Journal 44:161

Jones, L. and Scard, F, 1921. “The manufacture of Cane Sugar”. Duckworth & Co.
Second edition.

Kampen, W.H. 1975. Technology of the rum industry. Sugar y Azucar. 36-43.

Korhola, M., Harju, K., and Lehtonen, M. 1989. In “The Science and Technology of
Wiskeys” (Piggot, J.R., Sharp, R., and Duncan, R.E. B, eds.) Longman, Harlow, pp. 89-
117.

L’Anson, 1971. Rum manufacture. Process Biochem. 6: 35-39.

L’Anson P. 1976. Diversification in the distilleries. Wine and Spirit. 106: 38-39, 42-43,
45.

117
Laure, F., and Lafon-Lafourcade, S. 1989. In “Alcohol Toxicity in Yeast and Bacteria”
(van Uden, ed.) CRC Press, Boca Raton, pp. 193-215.

Lawless H. T. and Engen, T. 1977. Associations to odors: Interference, memories, and

verbal labeling. J. Exp. Psycol. 3: 52-59.

Lawless, H.T., and Klein, 1991. Sensory Science: Theory and Applications in Foods. M
Decker, New York,

Lehrer, A. 1983. Wine and Conversation. Indiana University Press. Bloomington, IN.

Lehtonen, M. 1983. Volatile and non- volatile compounds in the flavour of alcoholic
beverages. In “Flavor of Distilled Beverages: Origin and development” (Piggot, J.R., ed.)
Ellis Horwood Ltd., Chichester, pp. 65-78.

Lehtonen, M. J., Greff, B. K., Puputti, E. V., Suomalien, H. 1977. 2-Ethyl-3-

methylbutyric acid, a new fatty acid found in rum. J. Agric. Food Chem. 25: 953-955.

Lehtonen, M., and Suomalainen, H. 1977. Rum. In “Alcoholic beverages” (Rose, A. H.

ed.), Academic Press, London., pp. 595-633.

Maiorella, B.L., Blanch, H.W., Wilke, C.R., 1984. Feed component inhibition in
ethanolic fermentation by Saccharomyces cerevisiae. Biotechnol. Bioeng. 26(9): 1155-
1166.

Manaohar Rao, P. J., 1997. Industrialization of Sugar Cane and its Co-products. ISPCK
Publishers and Distribuitors, Delhi, India.

Mecredy, J.M., Sonnemann, J.C., and Lehmann, S.J. 1974. Sensory profiling of beer by a
modified QDA methodology. Food Technoloogy. Nov. 1974: 36-41.

Meikle, A.J., Reed, R.H., and Gadd, G.M. 1988. Osmotic adjustments and the
accumulation of organic solutes in whole cells ad protoplasts of Saccharomyces
cerevisiae. Journal of General Miocrobiology. 134 (11): 3049.

Meilgaard, M.C. 1975. Flavor chemistry of beer. Part II. Flavor and Threshold of 239
Aroma Volatiles. Tech. Quart. Master Brew. Assoc. A. 12:152-168.

Meilgaard, M, Civille, G.V., and Carr, B.T. 1999. Sensory Evaluation Techniques, 3rd
CRC Press, Inc. Boca Raton, FL.

Morrison, E.E. And Constanzo, R.M., 1992. Morphology and Plasticity of the vertebrate
olfactory epithelium. In “ Science of Olfaction” (Serby, M. J., and Chobor, K.L. eds.)
Springer-Verlang. New York. pp 31-47.

118
Moskowitz, H. 1983. Descriptive analysis of perceptions. In “Product Testing and
Sensory Evaluation of Foods. Food and Nutrition Press, Trumbull, Connecticut. Pp 20-
78.

Muñoz, A.M., and Civille, G.V. 1992. The Spectrum descriptive analysis method. In
ASTM Manual Series MNL 13, “Manual on Descriptive Analysis Testing” (Hootman,
R.C., ed.) ASTM West Conshohocken, PA.

Muñoz, A.M., and Civille, G.V. 1998. Universal, product, and attribute specific scaling
and the development of common lexicons in descriptive analysis. J. Sensory Studies.
Food and Nutrition Press. Trumball, Connecticut. 13:57-75.

Muñoz, A.M., and Civille, G.V. 1999. The Spectrum descriptive analysis method. In “
Descriptive Analysis Testing for Sesnsory Evaluation” (Hootman, R.C., ed.) ASTM
Philadelphia, PA.

Nef, P, Heldman, J.,Lazard, D., Margalit, T., Jaye, M., Hanokoglu, I., and Lancet, D.
1989. Olfactory-specific chromosome p-450: cDNA cloning of a nobel neuroepithelial
enzyme possibly involved in chemoreception. J. Biol. Chem. 264, 6780-6785.

Nef, P., Larabee, T.M., Kagimoto, K., and Meyer, U.A. 1990. Olfactory-specific
cytochrome P-450 (P-450Olfl:IIG1): Gene structure and development regulation. J. Biol.
Chem, 265 2903-2907.

Nishimura, K. Ohnishi, M., Masuda, M., Koga, K., and Matsuyama, R. 1983. Reactions
of wood components during maturation. In “The flavor of distilled beverages; origins and
development” (Piggott, J.R., ed.). American ed. Chichester, West Sussex: E. Horwood
Ltd., pp 241-255.

Noble, A.C. 1998. Analysis of wine sensory properties. In “Modern Methods of Plant
Analysis” Vol. 6. (Linskin, A.F., and Jackson, J.F., eds.) Springer-Vela, Berlin.

Nykänen, L, and Nykänen, I. 1991. Distilled beverages. In “Volatile compounds in food

and beverages” (Maarse, H, ed.). Mercel Dekker, New York.

Nykänen, L., and Nykänen, I. 1983. Rum flavor. In “Flavor of Alcoholic Beverages;
Origin and Development’ (Piggott, J.R., ed.). American ed. Chichester, West Sussex: E.
Horwood Ltd., pp 49-63.

Nykänen, L. and Suomalainen, H. 1983. Aroma of Beer, Wine, and Distilled Alcoholic
Beverages. D. Reidel Publishing Co. Berlin.

Pamment, N.B. 1989. In “Alcohol Toxicity in Yeast and Bacteria” (van Uden, N. ed.)
CRC Press, Boca Raton, pp 1-75.

119
Panchal, C.J., and Stewart, G.G. 1980. The effect of osmotic pressure in the production of
ethanol and glycerol by a brewing yeast. J. Inst. Brew. 86: 207-210

Parfait, A. and Sabin, G. 1975. Traditional fermentation of molasses and sugar cane juice
in the French Antilles. Ind. Aliment Agric. 92:27-34.

Parfait, A., and Jouret, C., 1975. Formatio n des alcohols supérieurs dans le rhums. Ann.
Technol. Agri., 24:421-436.

Patrau, J. M. 1969. “By-products of the Cane Sugar Industry’ p 153. Elsevier Publishing
Company, New York.

Pearce, D.A., Rose, A.H., and Wright, I.P., 1989 Yeast 5: S 443.

Peppler, H. J. 1979. Yeast technology. In “ Microbial Technology” Vol. 2 2nd edition.

Academic Press.New York.

Perry, R. H., Green, D. W., and Maloney, J. O. 1984. Perry’s Chemical Engineer’s
Handbook. 50th edition. Mc. Graw Hill.

Piggot, J.R., Carey, R.G., and Canaway, P.R. 1985. Sensory analysis and evaluation of
whisky. In “proceedings Symposium Alcoholic Beverages” (Birch, G.G., and Lindley,
M.G. eds.). Elsevier Applied Science Publishers, London, pp. 69-84.

Piggot, J.R., Hose, L., and Sharp, R.1980. Expressing flavor: a system for Scotch, Brew.
Distill. Int. 10: 48-49.

Piggot, J.R., and Holm, A.M. 1983. Descriptive sensory analysis of gin flavour. In
“Flavour of Distilled Beverages: Origin and Development” (Piggot, J.R., ed.), Ellis
Horwood, Ltd. Chichester, pp. 145-153.

Pollock, G.E., and Hollstrom, C.D. 1951. Cereal Chemistry. 28:498.

Pousias, J. and Chabanon, R.L. 1974. The Art of Wine Tasting. Interpublish Inc.
Madison, Wisconsin.

Quain, D.E., and Tubb, R.S.1982. The importance of glycogen in brewing yeasts. MBAA
Technical Quarterly 19, 498.

Rassmussen, E.F., and McMillen, J.M., 1956. Seasoning in white oak for tight cooperage.
For. Prod. Lab. Rept. 1784: 1-31.

Ribéreau-Gayon, P. 1978. Wine Flavor. In “Flavor of Foods and Beverages; Chemistry

and Technology”. (Charalambous, G., and Inglett, G., eds.) Academic Press. New York,
NY.

120
Rose, A. H., and Harrison, J.S. 1971. Physiology and biochemistry of yeasts. In “The
Yeast”. Vol. 2. Academic Press. London.

Rose, A.H, 1977. “Economic Miocrobiology” (Rose, A.H. ed.). Academic Press,
London., vol 1. Pp 1-41.

Rosen, K. 1989. In “Biotechnology Applications in Beverages Production” (Cantarelli,

C., and Lanzarini, G. eds.) Elsevier Applied Science, London, pp. 169-187.

Rowell, R. 1984. The chemistry of solid wood. Adv. Chem. Ser. ACS. 207: 1-614.

Ruter, P. 1975. Molasses utilization. Agricultural Services Bulletin. No. 25 Food and
Agricultral Organization of the United Nations, Rome, Italy. 1-41.

SAS Institute, 1999-2001. Statistical Analysis System. Version 8. The SAS Institute,
Cary, NC.

Scheuerbrant, G.,and Bloch, K. 1962. Unsaturated Fatty-acids in Microorganisms. J.

Biol. Chem. 267:2064-2068.

Scopes, R.K, 1989. In “Alcohol Toxicity in Yeast and Bacteria” (van Uden, N. ed.) CRC
Press, Boca Raton.

Singleton V. L. and Draper, D. E. 1961. Wood chips and wine treatment; the nature of
aqueous alcohol extracts. Am. J. Enol. Vitic. 12: 152-158.

Singleton, V. L. 1974. Some aspects of the wooden container as a factor of wine

maturation. Adv. Chem. Ser. (ASC) 137:254-277.

Singleton, V.L,. 1995. Maturation of wines and Spirits: comparisons, facts and
hypotheses. Am. J. Enol. Vitic. 46 (1): 98-115.

Stone, H., Sidel, J., Oliver, S., Woolsey, A., and Singleton, R.C. 1974. Sensory
evaluation by quantitative descriptive analysis. Food technology. 28:24-34.

Thomas, D.S., and Rose, A.H. 1979. Archives in Microbiology. 122:49.

Thorngate, J.H. III. 1997. The Physiology of Human Sensory Response to wine: a
Review. Am. J. Enol. Vitic. 48 (3) 271-279.

Unger, E. D., and Coffey, T. R. 1975. Ann. Technol. Agri. 24 (3/4):469.

Van Dijken, J.P., and Scheffers, W.A. 1986. Redox balances in the metabolism of
sugars by yeasts. FEMS Microbiological Reviews. 32(3-4):199-224.

121
Van Uden, N. 1989. In “Alcohol Toxicity in Yeast and Bacteria” (van Uden, N. ed.) CRC
Press, Boca Raton.

Vargas Guzman, D. 1975. Técnicas usadas en barriles de robles para disminuir mermas y
recuperación de los mismos. Annales de Technologie Agricole. 24 (3/4):297-305.

Walker, D.I.T, and Simmons, N.W. 1984. Breeding, selection and trials. In “Sugar-cane”
(Blackburn, F. ed.) Longman. New York.

Watson, D. C. 1984. Distilling yeast. Dev. Ind. Microbiol 25:213-220.

Watson, D. C. 1985. Current developments in the potable distilling industry. CRC Crit
Rev Biotechnol 2:147-192.

Watson, D. C. 1993. Yeast in distilled alcoholic-beverage production. In “The Yeast, vol.

5: Yeast technology” (Rose, A. H. ed.) Academic Press Limited.

122
 APPENDIX A. BALLOT FOR THE EVALUATION OF RUM USING THE LIST
OF TERMS BY SEMI-EXPERT JUDGES

Please mark the descriptor (if present) and indicate the order in which it was perceived.

867 432 258 093 449 389 758

Allspice
Almond
Apple
Apple/pear
Artificial
Banana
Bite
Burn
Butter
Butterscotch
Caramel
Chemical
Pure Ethanol
Eucalyptus/mint
Floral
Fruity (general)
Fusel oil
Gasoline
Leathery
Medicinal
Musty
Nutty
Pineapple
Plastic
Rubbing alcohol
Smoke
Spicy
Sweet
Syrupy
Vanilla
Woody

123
 APPENDIX B. BALLOT FOR PANELIST SCREENING. PAGE 1

Name:__________________________ Date:___________________
Phone:__________________________ e-mail:____________________

Screening Part I:

Match each solution to one of the perceived tastes (sweet, sour, salty, or bitter)

Taste Write down the solution number

Sweet ___________________________

Sour ___________________________

Salty ___________________________

Bitter ___________________________

Screening Part II:

1. Rank the sweetness intensity of the solutions from the least sweet to the sweetest.
Write down the solution numbers on the space below.
__________ ___________ ___________ __________

Least sweet Sweetest

2. Rank the saltiness intensity of the solutions from the least salty to the saltiest. Write
down the solution numbers on the space below.
__________ ___________ ___________ __________

Least salty Saltiest

2. Rank the sourness intensity of the solutions from the least sour to the most sour. Write
down the solution numbers on the space below.

__________ ___________ ___________ __________

Least sour Most sour

2. Rank the bitterness intensity of the solutions from the least bitter to the most bitter.
Write down the solution numbers on the space below.

__________ ___________ ___________ __________

Least bitter Most bitter

124
 BALLOT FOR PANELIST SCREENING. PAGE 2

Screening Part III:

Sniff each sample and describe in your own words the perceived aroma.

Sample Description

A _____________________________________________________

B _____________________________________________________

C _____________________________________________________

D _____________________________________________________

E _____________________________________________________

F _____________________________________________________

G _____________________________________________________

H _____________________________________________________

I _____________________________________________________

J _____________________________________________________

125
 BALLOT FOR PANELIST SCREENING. PAGE 3

Screening Part IV:

Sniff the first set of samples allowing time to rest after each sample. Then sniff

the second set of samples. Match the samples from the second set with the ones from the

first set. Write down the sample numbers of the second set next to the matching pair in

the first set. Then determine which term(s) from the list below best describes the aroma

of the samples

Fist set Second set Descriptors

473 ________ _________________________________

318 ________ _________________________________

941 ________ _________________________________

502 ________ _________________________________

079 ________ _________________________________

992 ________ _________________________________

467 ________ _________________________________

134 ________ _________________________________

756 ________ _________________________________

345 ________ _________________________________

List of terms or descriptors:

Floral Clove Strawberry Pepper

Hazelnut Peppermint Banana Garlic

Rose Cinnamon Eucalyptus Nutmeg

Pineapple Vinegar Allspice

126
APPENDIX C. RUM AROMA DESCRIPTIVE ANALYSIS. ORIENTATION
HANDOUT. PAGE 1

Objective

Determine differences in flavor of rum

Rum Flavor Descriptive Analysis products produced from different
processing protocols

1 2

Rum Rum Process

Raw material (cane juice, molasses)

• Alcoholic Distillate Fermentation (yeast, bacteria)

• Molasses, cane juice, cane syrup Distillation (continuous, batch)

• Distilled at < 190 Proof
Aging
• Posses tastes, aroma, and characteristics of
rum Blending

Other treatments

Additions (flavors, fruits, spices)

3 4

Spectrum Technique Requirements

• Expose panel to dimensions of flavor to
ensure accurate evaluation
• Discriminate
• Describe Flavor characteristics
• Provide similar frame of reference in
• Quantify
terminology and scaling

5 6

127
 RUM AROMA DESCRIPTIVE ANALYSIS. ORIENTATION HANDOUT.
PAGE 2

Flavor Perception Spectrum Method

Volatile compound olfactory receptor 1. Development of terminology

electrical response 2. Review references and evaluation

procedures
• Odors perceived:
– Inhalation 3. Product evaluation, discussion of results
– Via mouth, rear nasal passages; flavor

7 8

Review of References,
Development of Terminology
Evaluation Procedures
• Practice description of perceived flavor
attributes • Selection of adequate references
– General samples • Detailed definitions
– Samples related to rum (raw materials) • Group discussion Selection terms
– Evaluation products within rum category • Use of scale and references

none extreme

9 10

Practice Sessions
Product Evaluation
• Review samples/terminology
Individual

• Review references/evaluation procedure 9-10 sessions

• Most of it individual training Most convenient time

11 12

128
RUM AROMA DESCRIPTIVE ANALYSIS. ORIENTATION HANDOUT.
PAGE 3

Location
• Orientation
• Training Room 201

• Evaluation Individual booths

13

Results

14

129
 RUM AROMA DESCRIPTIVE ANALYSIS. ORIENTATION HANDOUT.
PAGE 4

Physiology of Aroma Perception

Rum products contain numerous products that contribute to their aroma. The
aroma is perceived when a volatile compound contacts the olfactory mucosa, in the
epithelium of the mucosa localized in the olfactory receptors, they axon through the
cribform plate of the ethmoid bone terminating in the olfactory bulb. There is a large but
limited number of olfactory receptors that transduce a large number of odorants. The
olfactory receptors direct extensions of the olfactory nerve into the environment. The
stimulus signal of the molecule contacting the olfactory receptor cells has chemical
specificity, and it is translated to a neuronal electronic response. The transduction
mechanism is still unknown (Thorngate, 1997). Odors can be perceived in two different
ways; direct, the molecules enter the nose from the front through the nosetrils by
inhalation, and indirect via mouth and rear the nasal passages (Pouisias and Chabanon,
1974). As the temperature of the substances in the mouth rises, more odorous compounds
are released. These compounds will then reach the olfactory region by diffusion and
through exhalation. The in- mouth odors are an important part of what is called flavor
(Amerine and Roessler, 1983.)
Man’s olfactory system is far less sensitive that that of many mammals. The
olfactory thresholds vary depending on the substance. Some factors like respiratory
infections and migraine can affect the olfactory thresholds. Individual threshold
variations depend on various factors such as sex; women between adolescence and
menopause have higher sensitivities, especially after ovulation. During stimulation by a
particular odor, the threshold vale rises above the stimulation concentration causing
adaptation. Adaptation is caused by the higher centers of the brain, and it explains why
professional smellers can perceive odors far longer than untrained individuals (Carr,
1974).

References:

Amerine, M.A., and Roessler, E.B., 1983. Wines: Their Sensory Evaluation. W.H.
Freeman and Co. New York, N.Y

Carr, J.G., 1974. Aroma and Flavour in Winemaking. Mills and Boon Ltd. London.

Pousias, J., and Chabanon, R.L. 1974. The Art of Wine Tasting. Interpublish Inc.
Madison, Wisconsin.

Thorngate, J.H. III. 1997. The physiology of human sensory responses to wine; a review.
Am. J. Enol. Viticult. 48 (3): 271-279.

130
 RUM AROMA DESCRIPTIVE ANALYSIS. ORIENTATION HANDOUT.
PAGE 5

Tentative Work Schedule

Sunday Monday Tuesday Wednesday Thursday Friday Saturday

1 2 3 4 5 6

7 8 9 10 11 12 13
Screening Screening Screening Screening Orientation

14 Group 15 Group 16 Group 17 Group 18 Group 19 20

training training training training training

21 Group 22 Group 23 Group 24 Group 25 Group 26 27

training training training training training

28 Indiv. 29 Indiv. 30
Training Training

Sunday Monday Tuesday Wednesday Thursday Friday Saturday

Indiv. 1 Indiv. 2 Indiv. 3 Indiv. 4

Training Training Training Training

5 Indiv. 6 Indiv. 7 Indiv. 8 Indiv. 9 Indiv. 10 11

Training Training Training Training Training

14 15 16 17 18 19 20
Evaluation Evaluation Evaluation Evaluation Evaluation Evaluation

21 22 23 24 25 26 27
Evaluation Evaluation Evaluation Evaluation Evaluation

28 29 30
Evaluation Evaluation

131
 APPENDIX D. BALLOT FOR GROUP TRAINING SESSION NO. 1 AND 2

Name:_______________________________ Date:_____________

Training Session No.__

Please smell each sample with short deep sniffs, and write down all the descriptors in the
order perceived to describe the aroma/odor of each sample. Then give a brief definition
for each term.

In between samples, sniff yourself, the coffee beans, and yourself again. Rest for about 1-
2 minutes, then move to the next sample.

Sample No.____

Term Definition
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________

Sample No.____

Term Definition
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________

Sample No.____

Term Definition
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________

132
 APPENDIX E. BALLOT FOR GROUP TRAINING SESSION NO. 3

Name:_______________________________ Date:_____________

Training Session No. 3

Please smell each sample with short deep sniffs, and write down all the
descriptors in the order perceived to describe aroma/odor of the set of samples. Then give
a brief definition for each descriptor.

In between samples, sniff yourself, coffee, and yourself again. Rest for about 1-2
minutes, then move to the next sample

Set No. ___

Term Definition
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________

Selected Group Terms

_________________
_________________
_________________
_________________

Rate samples according to the intensity of each of the selected terms

Term Rank Order

__________________ ___________ __________ _________

Less intense More intense

__________________ ___________ __________ _________

Less intense More intense

__________________ ___________ __________ _________

Less intense More intense

__________________ ___________ __________ _________

Less intense More intense

133
 APPENDIX F. BALLOT FOR GROUP TRAINING SESSION NO. 4

Name:_______________________________ Date:_____________

Training Session No. 4

Please smell each sample with short deep sniffs, and write down all the
descriptors in the order perceived to describe aroma/odor of the set of samples. Then give
a brief definition for each descriptor.

In between samples, sniff yourself, coffee, and yourself again. Rest for about 1-2
minutes, then move to the next sample

Term Definition

_____________ ________________________________________________
_____________ ________________________________________________
_____________ ________________________________________________
_____________ ________________________________________________
_____________ ________________________________________________
_____________ ________________________________________________
_____________ ________________________________________________
_____________ ________________________________________________
_____________ ________________________________________________
_____________ ________________________________________________
_____________ ________________________________________________
_____________ ________________________________________________
_____________ ________________________________________________
_____________ ________________________________________________
_____________ ________________________________________________
_____________ ________________________________________________
_____________ ________________________________________________
_____________ ________________________________________________

134
 APPENDIX G. BALLOT FOR GROUP TRAINING SESSION NO. 5

Name:_______________________________ Date:_____________

Training Session No. 5

Please smell each sample with a fast single short deep sniff, and write down one
descriptor to describe aroma/odor of the samples. Then give a brief definition for that
term, and rapidly proceed to the next sample.

After a few samples sniff yourself, the coffee beans, and yourself again. Rest for
about 1-2 minutes.

Sample Term Definition Assigned Term

1. _____________ _____________________________ _______________
2. _____________ _____________________________ _______________
3. _____________ _____________________________ _______________
4. _____________ _____________________________ _______________
5. _____________ _____________________________ _______________
6. _____________ _____________________________ _______________
7. _____________ _____________________________ _______________
8. _____________ _____________________________ _______________
9. _____________ _____________________________ _______________
10. _____________ _____________________________ _______________
11. _____________ _____________________________ _______________
12. _____________ _____________________________ _______________
13. _____________ _____________________________ _______________
14. _____________ _____________________________ _______________
15. _____________ _____________________________ _______________
16. _____________ _____________________________ _______________
17. _____________ _____________________________ _______________
18. _____________ _____________________________ _______________
19. _____________ _____________________________ _______________
20. _____________ _____________________________ _______________
21. _____________ _____________________________ _______________
22. _____________ _____________________________ _______________
23. _____________ _____________________________ _______________
24. _____________ _____________________________ _______________
25. _____________ _____________________________ _______________
26. _____________ _____________________________ _______________
27. _____________ _____________________________ _______________
28. _____________ _____________________________ _______________
29. _____________ _____________________________ _______________
30. _____________ _____________________________ _______________
31. _____________ _____________________________ _______________

135
 APPENDIX H. BALLOT FOR EVALUATION SESSION NO. 6

Name:_______________________________ Date:_____________

Training Session No. 6

Please smell each sample with a fast single short deep sniff, and write down one
descriptor to describe aroma/odor of the samples. Then give a brief definition for that
term, and rapidly proceed to the next sample.

After a few samples sniff yourself, the coffee beans, and yourself again. Rest for
about 1-2 minutes.

Sample Term Definition Assigned Term

1. _____________ _____________________________ _______________

2. _____________ _____________________________ _______________
3. _____________ _____________________________ _______________
4. _____________ _____________________________ _______________
5. _____________ _____________________________ _______________
6. _____________ _____________________________ _______________
7. _____________ _____________________________ _______________
8. _____________ _____________________________ _______________
9. _____________ _____________________________ _______________
10. _____________ _____________________________ _______________
11. _____________ _____________________________ _______________
12. _____________ _____________________________ _______________
13. _____________ _____________________________ _______________
14. _____________ _____________________________ _______________
15. _____________ _____________________________ _______________
16. _____________ _____________________________ _______________

136
 APPENDIX I. BALLOT FOR GROUP TRAINING SESSION NO. 7

Name:_______________________________ Date:_____________

Training Session No. 7

Please smell each sample with short deep sniffs, and write down a single term to
describe aroma/odor of the sample.

Remember to sniff yourself, coffee, and yourself again. Rest for about 1-2
minutes, then move to the next sample

Sample Term Assigned Term Group Definition

1. __________ ______________ _______________________________

2. __________ ______________ _______________________________
3. __________ ______________ _______________________________
4. __________ ______________ _______________________________
5. __________ ______________ _______________________________
6. __________ ______________ _______________________________
7. __________ ______________ _______________________________
8. __________ ______________ _______________________________

Please smell the sample with short deep sniffs, and write down all the descriptors
in the order perceived to describe aroma/odor of the sample. Then give a definition for
each descriptor.

Term Definition
_____________ ________________________________________________
_____________ ________________________________________________
_____________ ________________________________________________
_____________ ________________________________________________
_____________ ________________________________________________
_____________ ________________________________________________
_____________ ________________________________________________
_____________ ________________________________________________
_____________ ________________________________________________
_____________ ________________________________________________
_____________ ________________________________________________
_____________ ________________________________________________
_____________ ________________________________________________

137
 APPENDIX J. BALLOT FOR GROUP TRAINING SESSION NO. 8, 9, AND 10

Name:_______________________________ Date:_____________

Training Session No. ___

Please smell each sample with short deep sniffs, and write down all the
descriptors describe aroma/odor of the sample. Then give a brief definition for each term.

Remember to sniff yourself, coffee, and yourself again. Rest for about 1-2
minutes, then move to the next sample

Sample: _____

Term Definition
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________
________________ ________________________________________________

138
APPENDIX K. BALLOT FOR THE EVALUATION OF THE AROMA OF RUM.
PAGE 1

139
BALLOT FOR THE EVALUATION OF THE AROMA OF RUM. PAGE 2

140
BALLOT FOR THE EVALUATION OF THE AROMA OF RUM. PAGE 3

141
BALLOT FOR THE EVALUATION OF THE AROMA OF RUM. PAGE 4

142
BALLOT FOR THE EVALUATION OF THE AROMA OF RUM. PAGE 5

143
APPENDIX L. GUIDELINES FOR THE EVALUATION OF RUM AROMA

144
 VITA

The author was born in Mexico City, Mexico, on August 22, 1975. She graduated

from High School majoring in the chemical-biological area in 1993. She enrolled at the

La Salle University, Mexico City, and completed an exchange semester at Louisiana

State University in 1997. She graduated with a bachelor’s degree in Chemistry with an

emphasis in Food Science in 1998. She returned to Louisiana State University to pursue a

master’s degree. She will be awarded the degree of master of science in December 2002.

She plans to continue her doctoral program in the Department of Food Science at

Louisiana State University.

145