CHAPTER 6

Analysis of Edible Oils

Apostolos Kiritsakis and
William W. Christie

CONTENTS

6.1 Introduction
6.2 Constituents of Edible Oils
6.3 Analysis of Main Olive Oil Constituents
6.4 Analysis of Constituents of Other Edible Oils
6.5 Measurement of Deterioration
6.6 Determination of Adulteration
6. 7 Advanced Methods of Oil Analysis

6.1 INTRODUCTION

Edible oils are mainly vegetable oils that have undergone several processes to
remove undesirable constituents. Most crude vegetable oils are subjected to a
refining process (neutralization, bleaching, and deodorization) to make them suit-
able for consumption. Only olive oil, which is a natural juice, can be consumed
without any refining process.
Vegetable oils contrast with animal carcass fats, mammalian butters, and fish
oils in that they have fewer component acids and a simpler triacylglycerol com-
position. Although edible vegetable oils have relatively straightforward composi-
tions, they vary so much from one to another that the group as a whole encom-
passes a vast range of chemical and physical properties (Rossell 1991 ). For most
uses, the suitability of an oil depends as much on its quality as on its composi-
tion. In determining the purity of an oil besides the fatty acid composition, more
sophisticated tests, such as analysis of the acids at the 2-position of the triacyl-

129
J. Harwood et al. (eds.), Handbook of Olive Oil
© Springer Science+Business Media New York 2000
130 HANDBOOK OF OLIVE OIL

glycerol molecule; analysis of the triacylglycerol molecular species; and determi-

nation of the tocopherol, sterol, and other constituents of the unsaponi:fiable part
of the oil, might be appropriate.
This chapter briefly describes the main methods for analysis of edible oils that
can be applied to olive oil, as described more fully in chapters 7 through 9. The
analysis is particularly concentrated on the natural constituents of olive oil and on
those constituents formed during the hydrolytic and oxidative deterioration of the
oils, thus the chapter is essentially an introduction to issues developed in chapters
12 and 13. Techniques and methods for detecting adulteration are also described,
but more information about this subject is given in chapters 10 and 14.

6.2 CONSTITUENTS OF EDIBLE OILS

Edible oil constituents can be divided into the saponifiable (e.g., triacylglyc-
erols, free fatty acid, phosphatides) and the unsaponi:fiable (e.g., hydrocarbons,
fatty alcohols) fractions. The unsaponi:fiable constituents of virgin olive oil
account for 0.5-1.5 percent of the oil. Some of them (e.g., phenols) contribute to
its flavor quality and also increase its oxidative stability.

Triacylglycerols

Edible vegetable oils seldom contain branched-chain or odd-numbered fatty

acids or unsaturated fatty acids with fewer than sixteen or more than twenty car-
bon atoms. Their triacylglycerol compositions generally follow a pattern in which
the fatty acids at the central or 2-position of the glycerol molecule are unsatu-
rated, with linolenic acid being favored more than oleic or linoleic acids. Satu-
rated acids are found at the 2-position only when there is a very high overall satu-
rated fatty acid concentration in the fat.
Most of the fatty acids of edible oil (Table 6-1) are present as triacylglycerols
(triglycerides). Fedeli (1977) reported the major olive oil triacylglycerols as POO
(18.4 percent), SOO (5.1 percent), POL (5.9 percent), 000 (43.5 percent), OOL
(6.8 percent) (P, palmitic acid; 0, oleic acid; S, stearic acid; L, linoleic acid). The
three major triacylglycerols of olive oil were confirmed as OOL, 000, and POO
(Phillips eta/. 1984). According to Santinelli, Damiani, and Christie (1992), the
1-random, 2-random, 3-random distribution theory is not applicable to olive oil.
Olive oil, like other vegetable oils, shows a high concentration of oleic acid and a
low concentration of saturated fatty acids in position sn-2.

Minor Nonglycerol and Glycerol Constituents

Several minor nonglycerol (unsaponi:fiable) constituents are present in olive

oil. They include hydrocarbons, sterols, triterpene alcohols, tocopherols, phenols,
 Analysis of Edible Oils 131

Table 6-1 Percent of Fatty Acids of Some Edible Oils

Fatty Olive
Acid Oil* Canola Coconut Corn Cottonseed Sunflower

C14:0 :::;0.05 0.1-0.2 16.5-20.8 <H>.3 0.6-1.0 Q-0.1

C16:0 7.5-20.0 3.0-5.0 8.2-10.2 9.1-16.8 21.0-26.8 5.5-7.7
C16:1 0.3-3.5 0.2-Q.6 <H>.3 0-1.3 <H>.3
C17:0 :::;0.3
C17:1 :::;0.3
C18:0 0.5-5.0 1.Q-2.0 2.3-3.4 1.4-3.0 2.0-3.3 2.8-6.5
C18:1 55.Q-83.0 52.Q-67.0 4.3-8.1 20.Q-38.0 14.5-22.0 14.Q-38.0
C18:2 3.5-21.0 16.Q-24.8 0.7-2.0 39.5-65.0 46.5-58.0 48.2-74.2
C18:3 :::;0.9 6.5-14.0 0-tr 0.6-1.4 Q-0.4 <H>.1
C20:0 :::;0.6 0.2-Q.B 0.1 0.3-Q.7 0.2-0.5 0.2-Q.4
C20:1 :::;0.4 0.9-2.4 Q-tr 0.2-Q.4 Q-0.1 <H>.2
C22:0 :::;0.2* 0.1-Q.5 o-o.5 Q-0.6 0.7-1.3
C24:0 :::;0.2 0-0.2 0-0.3 <H>.4

*Limit is ::50.3 for olive-pomace oils.

Note: tr, trace.

chlorophylls, :flavor compounds, and polar phenolic compounds. Mono- and dia-
cylglycerols, phosphatides, waxes, and esters of sterols are also considered minor
constituents.
Hydrocarbons
Squalene, a biochemical precursor of sterols, is an important hydrocarbon of
both virgin and refined olive oils. Olive oil contains the largest amount of squa-
lene among vegetable oils (2500-9250 IJ.g/g) compared with other edible oils
(16-370 IJ.g/g) (Gutfinger & Letan 1974). Other hydrocarbons, such as polycyclic
aromatics (e.g., phenanthrene, pyrene, :fluoranthrene, 1,2-benzanthracene, chry-
sene ), are also found in olive oil (Fedeli 1977).

Tocopherols
Olive oil contains a-tocopherol in higher amounts than other tocopherols; it
varies from 125 to 200 ppm (Aparicio & Mcintyre 1998). The determination of
the tocopherol content of olive oil can be used for detecting adulteration with
seed oils. Table 6-2 shows the content of tocopherols in other edible oils.

Aliphatic Alcohols
Saturated straight-chain aliphatic alcohols with even-numbered carbon atoms
(C 18 to C28) are found in olive oil. The main linear aliphatic alcohols present in
olive oil are hexacosanol, octacosanol, and tetracosanol. Also, tricosanol, penta-
132 HANDBOOK OF OLIVE OIL

Table 6-2 Tocopherol Content of Some Edible Oils (mg/kg)

Oil a-Tocopherol {3- Tocopherol y- Tocopherol a- Tocotrienol Total

Canol a 100-400 Q-150 18Q-780 40Q-2700

Coconut 0-18 Q-11 Q-15 0-44 Q-50
Cottonseed 13Q-690 Q-37 14Q-740 Q-30 38Q-1500
Corn 20-600 0-370 6Q-2500 Q-250 30Q-3810
Raw palm 2-190 0-240 0-500 2-350 9Q-1500
Soybean 10-360 Q-50 90-2400 560-3400
Sunflower 400-1000 Q-60 Q-60 400-1600

cosanol, and heptacosanol might be present in traces. Solvent extracted oil (olive-
pomace oil) contains more aliphatic alcohols than virgin olive oil. Tacchino and
Borgoni (1983) determined the aliphatic alcohols in pressed and solvent-extracted
olive oils and found that it varied from 10 to 70 (mg/lOOg) and from 224 to 434
(mg/100g), respectively. Esters of fatty alcohols with fatty acids (waxes) are also
found in olive oil. The main waxes are C-36, C-38, C-40, C-42, C-44, and C-46
esters. Olive-pomace oil contains more waxes than virgin olive oil.

Sterols
The main sterols found in olive oil are ~-sitosterol, Ll 5 -avenasterol, and
campesterol. Stigmasterol, cholesterol, 24-methylene-cholesterol, Ll 7-campes-
terol, ~ 5' 23 -stigmastadienol, sitostanol, ~ 5'24 -stigmastadienol, ~ 7-stigmastenol,
and ~ 7-avenasterol are also present in olive oil but in smaller quantities (Boskou
1978; Boskou & Morton 1975; Fedeli 1977; Itoh et al. 1981).
Olive oil contains four methylsterols: (1) 4a-methyl-24-methylene-Ll 7 -
cholesten-3 ~-ol (gramisterol); (2) 4a, 14a-dimethyl-24-methylene-Ll8-cholesten-
3~-ol (obtusifoliol); (3) 4a,14a-dimethyl-9, 19-cyclopropane-24-methylene-
cholesten-313-ol ( cycloeucalenol); and (4) 4a-methyl-(24Z)-24-ethylidene-
~7-cholesten-313-ol (citrostadienol).
4,4-Dimethylsterols (triterpene alcohols) are also found in olive oil (Paganuzzi
1982). They are a-and ~-amyrin; cycloartenol; 24-methylenecycloartanol;
taraxerol; dammaradienol; and 24-methylene-24-dihydroparkeol. Olive oil con-
tains triterpene alcohols in a concentration ranging from 100 to 150 mg/1 OOg of
oil (Kiosseoglou, Vlachopoulou & Boskou 1987). Two main triterpene dialcohols
(erythrodiol and uvaol) have been identified (Fedeli 1977). Table 6--3 shows the
content of sterols in other edible oils.

Mono- and Diacylglycerols

The presence of mono- and diacylglycerols is related in part to incomplete
biosynthesis but mainly to hydrolysis of the oil. When diacylglycerols are pre-
Table 6-3 Sterol Content of Some Edible Oils (mg!kg)

Oil Cholesterol Brassicasterol Campesterol Stigmasterol ~-Sitosterol .15 -Avenasterol .1 7 -Stigmasterol ,1 7 -Avenasterol

Canota 3Q-110 10Q-1100 800-4200 Q-320 250Q-6000 17Q-580 Q-350 Q-150

Coconut 4-28 Q-8 3Q-118 52-156 197-568 13Q-346 Q-35 Q-24
Cottonseed 3Q-120 Q-40 22Q-680 3Q-150 220Q-5400 70-400 Q-100 5Q-180
Corn 2Q-100 Q-30 1700-5400 500-1000 5000-13000 400-1900 0-550 20-400
Palm 1Q-30 0 7Q-160 3Q-70 200-400 Q-16 1-10 Q-25
Soybean 20-40 Q-15 35Q-850 30Q-700 950-2220 4Q-160 3Q-200 2Q-180
Sunflower 6-50 Q-10 130-460 20Q-400 900-3000 Q-200 15Q-500 5Q-250 ~
5
~
c;,j
!;;•

~
~
s:.
~
0
~

......
w
w
134 HANDBOOK OF OLIVE OIL

sent, olive oil is oflow quality (Mariani & Fedeli 1985). Determination of diacyl-
glycerols can be used for evaluating olive oil quality.

Pigments
The unique color of virgin olive oil is due to pigments, such as chlorophylls,
although carotenoids also make a contribution to the color of olive and other veg-
etable oils (Serani & Piacenti 1992).

Phenolic Compounds
Olive oil contains both simple and complex phenolic compounds that increase
its oxidative stability and considerably improve its flavor (Fedeli 1977; Vazquez,
Del Valle & Del Valle 1976). According to Cinquanta, Esti, and Notte (1997), it is
not the cultivar but the ripeness of olive fruit and the soil and climate conditions
that mainly affect phenol composition of virgin olive oil.

Flavor Compounds
Aroma and flavor are distinctive features of olive oil in comparison with
other edible oils. They are generated by a number of volatile compounds pre-
sent at extremely low concentrations (Aparicio, Calvente & Morales 1996;
Blekas & Guth 1995; Blekas, Guth & Grosch 1994; Fedeli 1977; Kiritsakis
1998a, 1998b; Kiritsakis & Min 1989; Montedoro, Bertuccioli & Anichini
1978; Morales, Aparicio & Calvente 1996; Morales, Aparicio & Rios 1994;
Morales et al. 1995).

6.3 ANALYSIS OF MAIN OLIVE OIL CONSTITUENTS

Determination ofTriacylglycerol Structure

Triacylglycerol structures have been determined by five procedures involving

separation into molecular species:
1. reversed-phase-high-perf ormance liquid chromatography (RP-HPLC)
(Goiffon, Remniac & Furon 1981; Perrin & Naudet 1983)
2. silver-ion-thin-layer chromatography (SI-TLC), in combination with pan-
creatic lipase hydrolysis (Goldberg Federico et al. 1969) or with an oxida-
tion procedure
3. high-temperature-capillary gas chromatography (HT-CGC) (Frega, Bocci
& Lercker 1990; Motta et al. 1987)
4. SI-TLC in combination with RP-HPLC and high-resolution gas chro-
matography (HRGC) (Damiani & Burini 1980)
5. silver-ion-high-performance liquid chromatography (SI-HPLC) (Toschi,
Christie & Conte 1993)
 Analysis ofEdible Oils 135

Triacylglycerols from olive oil show an asymmetry in the distribution of

fatty acids among positions sn-1, sn-2, and sn-3 (Christie et al. 1991; Toschi,
Christie & Conte 1993), but a single symmetric molecule (triolein) represents
approximately half of the total. By combining fractionation by silver-ion (SI)
chromatography with stereospecific analysis, it proved possible to evaluate the
degree of asymmetry in other fractions. Thus, Santinelli, Damiani, and Christie
(1992) utilized a combination of SI-HPLC and chromatographic resolution of
diacyl-sn-glycerol derivatives as part of a stereospecific analysis procedure to
obtain the distribution of fatty acids in positions sn-1, sn-2, and sn-3 of the
main triacyl-sn-glycerol species of extra virgin olive oil. Marked asymmetry
was observed in particular species but was not apparent from analysis of the
total oil.

Analysis of Phenolic Compounds

Many analytic methods, such as thin-layer chromatography (TLC) and gas-

liquid chromatography (GLC) have been used for the analysis of the phenolic
compounds of olive oil. More recently, liquid chromatography (LC) with electro-
chemical detection has been employed for the analysis of compounds that can be
oxidized or reduced. The procedure involves separation of the oil constituents by
LC prior to their oxidation at a glassy carbon electrode in a thin-layer electro-
chemical cell.
Akasbi, Shoeman, and Saari Csallany (1993) applied the RP-HPLC technique
with isocratic elution to separate and quantify the major phenolic compounds of
the hydroalcoholic extracts of olive oils. Hyroxytyrosol, tyrosol, caffeic acid,
p-hydroxyphenylacetic acid, and homovanillic acid were analyzed on a IJ.Bonda-
pak C 18 column with an acetonitrile-water-acetic acid (20:90:0.18 by volume) as
mobile phase. According to these researchers, electrochemical detection provides
selectivity, as well as sensitivity. This method can be applied for the analysis of
small quantities and of the most important phenolic compounds in olive oil.
Tsimidou, Papadopoulos, and Boskou (1992) separated the phenolic compounds
of virgin olive oil by using four external standards (tyrosol, vanillic acid, syringic
acid, and o-coumaric acid) to minimize errors of analysis.
Figure 6-1 shows the phenolic compounds in olive oil (A) and in olive oil mill
wastewater (B) by HPLC. Some of the phenolics present in olive oil appear in
wastewater but in lower amounts.

Analysis of Flavor Compounds

Efforts have been made to establish instrumental analysis of the flavor and
aroma components of olive oil in order to eliminate the disadvantages of the
136 HANDBOOK OF OLIVE OIL

40

35

30

GJ2S
""c:: 1
~20
.... -~
~ 1 d
·s ·5a
"ij IS
"0 0
"i3

i -~~1
"0
t:.::

l
10

u
"lib c.
> -~
c ~
5

1J~ Ill.. ....

·-
~

0 l
10 20 30 40 so liO 70 80
Retention time ( min )

(a)

50

40

~ 30
t:=
0
0..
v. "C
·o
....
~

~
.~20
-a!
."'
u
·;:
E
'i3 8
~ e
10 :g <> ."9

~ ."..,
"' "'
=
k
·;:
>" ~ ~
0 A ~ A

10 20 30 40 50 60 70 80
Retention time ( min )

(b)

Figure 6-1 A: Phenolics determined in olive oil by HPLC; B: Phenolics determined in

olive oil mill wastewater by HPLC (Kiritsakis, 1998b).
 Analysis ofEdible Oils 137

sensory evaluation by the panel test. Tateo and coworkers ( 1993) used gas
chromatography-mass spectroscopy (GC-MS) in order to analyze the flavor com-
ponents of extra virgin olive oil samples with different flavor characteristics.
Guth and Grosch (1993) applied aroma extract dilution analysis (AEDA) to
evaluate the flavor components of olive oil and determined fifteen compounds as
potent odorants of four olive oil samples with different flavor profiles. The fol-
lowing compounds contributed mainly to the flavor, given in parentheses: (Z)-3-
hexenal (green); ethyl 2-methylbutyrate, ethyl isobutyrate, ethyl cyclohexanoate
(fruity); and (Z)-2-non-enal (fatty) and 4-methoxy-2-methyl-2-butanethiol (black
currant-like).
Aparicio et al. (1994), Aparicio, Morales, and Alonso (1996), and Aparicio and
Morales ( 1995) proposed another methodology for the analysis of the virgin olive
oil volatiles and designed a sensory wheel that represents the whole virgin olive
oil matrix. This wheel is composed of seven sectors that show the basic percep-
tions produced by the oil: green, bitter-pungent, undesirable, ripe olives, ripe
fruity, fruity, and sweet. A gas chromatography/sniffing method (Morales, Apari-
cio & Rios 1994) was applied to assess the aroma notes that correspond to olive
oil volatile compounds and to verify the relationships found by the sensory wheel
procedure (Aparicio, Calvente & Morales 1996).
Sacchi and coworkers (1996) applied high-field (600 MHz) nuclear magnetic
resonance (NMR) spectroscopy for direct analysis of olive oil. NMR spectra of
virgin olive oils show a large number of structural and compositional data related
to olive oil composition and quality. The analysis of aldehyde composition, which
gives a picture of the flavor profile of the oil, seems to be interesting. The possi-
bility of determining additional minor components, such as phenols or alcohols
present in the oil, without extraction or chemical derivatization opens new appli-
cation possibilities for instrumental flavor quality assessment of virgin olive oil
(Sacchi et al. 1996).

Determination of the Bitterness of Olive Oil

Garcia Mesa, de Castro, and Valcarcel (1993) developed an automatic method

based on flow-injection analysis principles for determining the bitterness of
olive oil. The ultraviolet (UV) absorption spectra of aqueous alcoholic extracts
of bitter oils show a sharp maximum at 225 nm and a smaller maximum at 278
nm. The oil tastes bitter when its K 225 absorbance value exceeds 0.30. The
absorbance is given as K225 per concentration of the oil (g/1 00 mL ). Evaluation
of the bitter taste of virgin olive oil also can be made in a sensorial manner
(panel test) or by HPLC analysis of an oil extract and quantification of its com-
ponents, which correlate well with the intensity of bitterness determined senso-
rially (Gutierrez et al. 1992).
13 8 HANDBOOK OF OLIVE OIL

Determination and Analysis of Cblorophylls

The chlorophyll content of olive oil can be determined by measuring its

absorbance at 630, 670, and 710 run, according to the American Oil Chemists
Society method (AOCS 1987). Although the method is proposed for the Beck-
man B spectrophotometer, the Beckman DU or other spectrophotometers can be
used. When the Beckman DU or a similar instrument is available, the following
formula must be applied:

chlorophyll (ppm) = [A670 - (A630 + A 710) X 0.5]/0.101 L

where A = absorbance, L = thickness of cuvette ( 1 em).

Rahmani and Csallany (1985) proposed a rapid and precise HPLC method for
analysis of the chlorophylls a and b and pheophytins a and b in olive oil. The oil
sample is diluted twenty times (w/v) with the solvent mixture (hexane-iso-
propanol, 98.5:1.5 by volume) and then injected into a IJ-Porasil column. The four
pigments are qualitatively separated at 430 run. A quantitative determination is
obtained at 430 run for chlorophyll a, 452 run for chlorophyll b, 409 nm for pheo-
phytin a, and 432 nm for pheophytin b. These wavelengths correspond with
absorption maximum of the pigments in the mobile phase.

6.4 ANALYSIS OF CONSTITUENTS OF OTHER EDIBLE OILS

Gutfinger and Letan (1974) analyzed the unsaponifiable fraction of several edi-
ble oils (soybean, cottonseed, coconut, avocado, and olive) and found that the oils
differed in the contents of total unsaponifiable squalene, tocopherols, and sterols
and also in the composition of the tocopherol and sterol fractions. The quantita-
tive determination of individual tocopherols was achieved by the colorimetric
method, whereas the contents of total sterols and squalene and the composition of
the sterol fractions were determined by GLC.
Choo et al. (1996) separated triacylglycerols, diacylglycerols, free fatty acids
(FFAs ), carotenoids, tocopherols, and tocotrienols from crude palm oil by super-
critical fluid chromatography (SFC) that used the programmed extraction elution
method. FFAs present in palm oil were separated into five components by GLC;
tocopherols and tocotrienols were also separated into four components by SFC
analysis, and the carotenes were obtained by preparative SFC. Thus, by using
SFC, the components of an edible oil can be separated and fractionated, based on
the differences in their functional groups. The separation of the minor compo-
nents of olive oil by the above technique is recommended. Singleton and Stike-
leather (1995) describe an HPLC method for simultaneous concentration, separa-
tion, and analysis of phospholipids and other trace compounds in edible oils. The
method greatly reduces analysis time (40 percent), solvent use, and the complex-
ity of solvent arrays normally required to separate phospholipids.
 Analysis ofEdible Oils 139

Near-infrared (NIR) spectroscopy has been recently recognized as one of the

most powerful analytical techniques for determining various constituents in edi-
ble oils.

6.5 MEASUREMENT OF DETERIORATION

The methods described here for olive oil quality analysis are not necessarily
those suggested by the International Olive Oil Council (IOOC) or by the Com-
mission of the European Communities (EC).

Determination of Hydrolysis

Diacylglycerol determination can be used as an alternative (to titration of acid-

ity) method for estimating the degree of hydrolysis of olive oil and other edible
oils. Acidity is a conventional expression, frequently used instead of percentage
of free fatty acids. There is a linear correlation between diacylglycerols and FFA
(acidity). The acidity can be expressed as acid value (AV).
May and Hume (1993) developed an automated GLC method for the analysis
ofFFA in canola oil. The FFAs were measured in a capillary chromatograph (GC)
with a flame ionization detector. An internal FFA standard was used (1 g heptade-
canoic acid, C-17, in 1 L heptane). Individual FFA peaks were identified by
means of pure standards of palmitic, stearic, oleic, linoleic, and linolenic acids.
The relationship between the GLC method and the titration method was
described as follows:

Ytitration method = 1.090 X GLCmethod + 0.095

This procedure permits the analysis of a large number of samples in a short

time. The method could be applied for the determination of FFAs in other edible
oils. A sample chromatogram of the FFA determination in edible oils is given in
Figure 6-2.
HRGC also can be used for the determination of FFAs and diacylglycerols
(Capella 1993). Ismail and coworkers (1993) proposed a rapid quantitative deter-
mination of FFA of oils by Fourier transform infrared (FTIR) spectroscopy. The
determination is based on both transmission and attenuated total reflectance
approaches, and it covers an analytical range of 0.2-0.8 percent FFA. Calibration
curves are prepared by adding oleic acid to the oil chosen for analysis and mea-
suring the C = 0 band at 1711 em -I after ratioing the sample spectrum against
that of the same oil free of fatty acids.
Recently, a new pH-metric method for the determination of AV in vegetable
oils without titration has been developed (Berezin et al. 1997; Tur'yan et al.
1996). The extraction of FFA from the oil sample into a reagent containing tri-
140 HANDBOOK OF OLIVE OIL

C-17

L
p
Ln
LLJ u_/t

2 4 6 8
Minutes

Figure 6-2 A gas liquid chromatogram of FFA from an edible oil. Note: P, palmitic;
C-17, a1 7-carbon FFA; S, stearic; 0, oleic; L, linoleic; Ln, linolenic.

ethanolamine (TEA), isopropanol, and water is necessary. The oil is mixed with
the reagent and the pH is measured before (pH 1 ') and after (pH/) the addition of
hydrochloric acid using a commercial pH meter with an aqueous reference elec-
trode. The AVis calculated from the difference between pH 1 ' and pH2 ':

AV = 56.11Nst X Ys/[(1011PH -1) m], in mg KOH!g oil

where 56.11 is the molecular weight of KOH; Nst is the standard acid concentra-
tion (M); Yst is the volume of the standard acid added (mL), which is considerably
less than the volume of the reagent; ~pH= pH 1 ' - pH2 '; and m is the weight of
the oil (g).
According to Berezin and coworkers (1997), an AV of 0.02 mg KOH/g oil is
accepted as the limit of quantitation. Because refined edible oils usually have AV
of 0.05 mg KOH/g oil or more, the method should be suitable for AV analysis of
these oils.
 Analysis ofEdible Oils 141

Determination of Oxidation

Among the several methods for estimating the degree of oxidative deteriora-
tion of edible oils are peroxide value (PV), thiobarbituric acid value (TBA), and
total carbonyls. PV is the most commonly used chemical method, despite the fact
that it is time consuming and requires special skills.
The determination of the hydroperoxides (initial products of oxidation) can be
accomplished by the International Union of Pure and Applied Chemistry
(IUPAC) (IUPAC 1987) or other official method. The oil is treated in a solution of
acetic acid and chloroform by a potassium iodide solution. PV is expressed in
terms of milliequivalents of active oxygen per kilogram of oil and gives the quan-
tity of peroxides that oxidizes potassium iodide under the operating conditions.
Terao and Matsushita (1977) proposed a colorimetric method for measuring
hydroperoxides in photooxidized oils. Cadmium acetate was used, and the
absorbance of the color formed was measured at 350 nm. Hydroperoxide concen-
trations were calibrated by using a standard solution of benzoyl peroxide in
ethanol.
van de Voort and coworkers (1994) proposed the determination of the degree
of oxidation of oils by FTIR spectroscopy, which provides a simple and rapid
means to follow the complex changes that take place as a lipid oxidizes. The crit-
ical absorption bands associated with common oxidation end products are identi-
fied by relating them to those of spectroscopically representative reference com-
pounds. A quantitative approach is proposed in which the standards used are
spectroscopically representative of oxidative end products and by which the
oxidative state of the oil can be defined in terms of percent hydroperoxides, per-
cent alcohols, and total carbonyl compounds. FTIR analysis provides a rapid
means of evaluating the oxidative state of an oil or of monitoring changes in an
oil undergoing thermal stress. It is necessary, however, to characterize the spec-
tral changes occuring as oil oxidation proceeds, to assign wavelengths to the
more common molecular species produced, and to access potential spectral cross-
interferences (van de Voort eta/. 1994).
Extra weak photon emission, also termed low-level chemiluminescence, is well
known as one of the physical chemical phenomena observed during oil deteriora-
tion (Miyazawa et al. 1994). Chemiluminescence generally originates from elec-
trically excited states, such as singlet molecular oxygen and triplet carbonyls pro-
voked by radical chain reactions in lipid peroxidation. The chemiluminescence
method detects peroxyesters and dialkylperoxides as well, whereas the iodiomet-
ric method is not suitable to determine these compounds because of their redox
potential (Matthaus 1996). For the chemiluminescence method, a small quantity
of oil is dissolved in a mixture of acetone-ethanol (2: 1) and 1 mL of a luminal
solution (consisting of luminal, hemin, and sodium carbonate) is added. A photo-
multiplier below the sample vessel receives the light and converts it to an electri-
142 HANDBOOK OF 0LNE OIL

cal pulse. Matthaus (1996) observed a good correlation (r = 0.9865) in using the
Rancimat apparatus and chemiluminescence measurements for the determination
of oil oxidation. According to Miyazawa et al. (1994), the chemiluminescence
method has the advantage of estimating the peroxide value of an edible oil in a
short time and does not require any special skills.
Min (1983) analyzed the flavor compounds formed when edible oils were
exposed to fluorescent light by GC and compared the results with the sensory
scores. Sensory qualities of the different oils were evaluated with a hedonic scale
of 1-10, where 1 indicated the poorest flavor quality and 10 the highest flavor
quality. The correlation coefficients of the flavor scores between sensory evalua-
tion and GC analysis for all tested edible oils were higher than 0.95. According to
Min (1983), instrumental GC could be used to evaluate the flavor quality of vege-
table oils. Snyder and King (1994) used a supercritical fluid extraction (SFE)
technique for recovery of edible oil volatiles as a mild method to analyze lipid
samples that are easily decomposed. Both volatile and semivolatile compounds
were successfully collected on Tenax sorbent. The SFE technique is suitable for
the desorption and analysis of high-molecular-weight compounds. This provides
a unique analysis method for determining both volatile and semivolatile com-
pounds, as well as executing desorption under nonoxidative, low-temperature
conditions that do not contribute to the degradation of lipid components (Snyder
& King 1994).
Viinanen and Hopia (1994) analyzed the oxidation products oftriacylglycerol
standards (trilinolein and triolein) and of natural triacylglycerols of vegetable oils
(rapeseed oil) by an RP-HPLC method. The oxidation products were detected
with an ultraviolet and evaporative light-scattering detector (ELSD). The ELSD
is a mass detector and can be used to detect all high-molecular-weight autoxida-
tion products. In contrast, an UV detector, which is often used in autoxidation
studies, can detect conjugated dienes at wavelengths of 232-235 nm. The ELSD
detector is particularly useful in studying the oxidation products of monounsatu-
rated compounds at different stages of oxidation. As the consumption of oils high
in oleic acid (e.g., olive oil, canola oil) is increasing, it is necessary to use meth-
ods that detect all oxidation products. Therefore, the ELSD is promising to be a
good universal type of detector for monitoring autoxidation products of edible
oils (Viinanen & Hopia 1994). Static headspace and capillary chromatography-
infrared spectroscopy-mass spectrometry were used to collect, separate, identify,
and quantify the oxidative and thermal decomposition products of heated triolein
(Mahungu, Hansen & Artz 1994). This technique favors the detection of
low-molecular-weight components associated with the flavor development of
fried lipid foods. It is a useful technique for the identification and quantification
of volatiles in heated triacylglycerols.
Oligopolymer determinations in oils and fats have been the subject of research
because of the importance of oligopolymers in ascertaining the level of oxidative
 Analysis ofEdible Oils 143

and thermal degradation during the exposure of fatty substances to heat. According
to Gomes (1992), small amounts of oligopolymer compounds can be determined
by direct oil analysis without concentration. High-performance, size-exclusion
chromatography can be used for the determination of oxidized triacylglycerols. The
column packing materials are highly cross-linked styrenedivinylbenzene copoly-
mers. Polystyrene standards of known molecular weights (4000 and 2000 g/mol)
and tristearin, distearin, monostearin, and stearic acid standards are used for peak
identification. The absence of oligopolymers in virgin olive oils can be used as a
quality standard. Conversely, edible virgin olive oils should not contain oligopoly-
mer compounds. The above technique has been reviewed elsewhere (Dobarganes &
Marquez-Ruiz 1993).
For evaluating olive oil quality, the determination of the specific absorption
coefficients (specific extinctions) in the ultraviolet region is required (IOOC
1996b; EC 1995). These measurements also provide information on the state of
preservation of the oil and on the changes that are due to the refining process. The
absorption at the specified wavelengths is due to the presence of conjugated diene
and triene in the systems. The oil is dissolved in the solvent, and the extinction of
the solution is determined at the specified wavelengths using pure solvent as
blank.
The values of the specific coefficients are expressed as:

K/~:'n 232, Kl~':n270

The 1 percent indicates the solution of the oil in the specified solvent, and the 1
em is the thickness of the cuvette.
The absorption coefficients at wavelengths around 270 nm are also determined
by calculating the aK value:

..iK = Kn, - (K.t, _4 + Km+4) X 0.5

where Km is the specific extinction at wavelength m, which is the wavelength

around 270 nm with the maximum absorbance.

6.6 DETERMINATION OF ADULTERATION

Olive oil is often illegally adulterated with other, cheaper vegetable oils. Oils
used widely for this purpose include olive-pomace oil, com oil, peanut oil, cot-
tonseed oil, sunflower oil, soybean oil, and poppy seed oil (Aparicio, Alonso &
Morales 1996; Kiritsakis 1998b).
Several physical and chemical tests have been used to analyze and detect adul-
teration of olive oil by low-grade olive oils and seed oils (Codex Alimentarius
Commission 1993; Fedeli 1977; IOOC 1997; Kiritsakis & Markakis 1991).
144 HANDBOOK OF OLIVE OIL

Analysis of the entire unsaponifiable fraction (e.g., sterols, tocopherols, triter-

penic alcohols) provide additional tests for detecting adulteration and sharpening
the differentiation among vegetable oils (Fedeli 1977; Firestone & Reina 1996;
Itoh et al. 1981; Pallotta 1976). According to Fedeli (1977), quantitative analysis
of certain classes of minor components is one of the chief methods for determin-
ing the authenticity of olive oil or for distinguishing between pressed and solvent
extracted oils.
Zamora, Navarro, and Hidalgo (1994) studied the unsaponifiable matter from a
number of olive and olive-pomace oils by high resolution 13 C NMR spectroscopy.
Their spectra show characteristic peaks that correspond to molecular substruc-
tures, rather than to the individual constituents present in the unsaponifiable mat-
ter. The presence of squalene and other hydrocarbons, sterols, and triterpenic
alcohols, in addition to other groups of minor components of the unsaponifiable
fraction, can be determined. Based on the analysis of these spectra, it would be
possible to distinguish among different grades of olive oils by using stepwise dis-
criminant analysis. Analyses of some components of the unsaponifiable and
saponifiable fraction by different techniques follow.

Sterol Analysis

The presence of seed oils in olive and olive-pomace oils can be detected by
analysis of sterols (IOOC 1997). The oil is saponified with potassium hydroxide
in ethanolic solution, and the unsaponifiables are then extracted with diethyl
ether. The sterol fraction is separated from the unsaponifiable extract by chro-
matography on a basic silica gel plate. Sterols are converted to their trimethylsi-
lylether (TMS) derivatives and are analyzed by capillary-column GC. According
to Gutfinger and Letan ( 197 4), large quantities of stigmasterol indicate the pres-
ence of soybean oil in olive oil. Under standard conditions, virgin olive oil has a
high 13-sitosterol content, whereas refining (mainly neutralization and bleaching)
causes the destruction of sterols.

Alkane Analysis

McGill et al. (1993) analyzed a number of edible oils for n-alkane content.
They found that n-alkanes with chain lengths of C27 , C29 , and C31 predominated
in all vegetable oils except olive oil, where C23 , C25 , and C27 were the most abun-
dant. According to these researchers, the n-alkane content may be useful in char-
acterizing olive oil and other vegetable oils.

Wax and Aliphatic Alcohol Analysis

A reliable means to detect olive-pomace oil in olive oil is the analysis and
determination of the wax esters. According to the method recommended by EC
 Analysis ofEdible Oils 145

(1991) for the analysis of wax esters, the oil containing eicosanol as internal stan-
dard is saponified with methanolic potassium hydroxide and the unsaponifiable
matter is extracted with diethyl ether. The alcoholic fraction is separated from the
unsaponifiable matter by chromatography on a layer of potassium-hydroxide
impregnated silica gel; the alcohols recovered from the silica gel are transformed
into TMS esters and are analyzed by capillary GC.
Bianchi et al. (1994) determined the chemical structure of long-chain esters
present in lower-grade olive oil by GC-MS and by comparison with synthetic
authentic model compounds. Three classes of esters composed the hexane-diethyl
ether (99: 1) extract of the wax fraction from olive-pomace oil: ( 1) esters of oleic
acid with C 1-C6 alcohols, (2) esters of oleic acid with long-chain aliphatic alco-
hols in the range C22-C28 , and (3) benzyl alcohol esters of the very-long-chain
saturated fatty acids (C26 and C28 ). By applying analysis of the chemical structure
of long-chain esters, the adulteration of high-grade olive oil by less expensive
olive-pomace oil and seed oils can be detected.
Extra virgin olive oil is characterized by the virtual absence of waxes with
forty to forty-six carbon atoms, but these waxes are found in comparatively large
amounts in refined and olive-pomace oil. Thus, it is a good index for identifying
solvent-extracted oil (pomace oil) in virgin olive oil (Mariani & Fedeli 1993). A
chromatogram ofthe waxes of a virgin olive oil sample is shown in Figure 6-3.
Aliphatic alcohols are found in significantly larger quantities in olive-pomace
oils than in olive oils. Gracian and Cota (1984) used GC to analyze the aliphatic
alcohol fractions of olive oil and olive-pomace oil and found linear-chain saturated
alcohols and their mono- and diunsaturated derivatives with eighteen to thirty car-
bon atoms. Olive oil contains a lower amount of alcohols and a higher percentage
of low-molecular-weight alcohols than olive-pomace oil. This information can be
used for distinguishing these two categories of oil (Gracian & Cota 1984).

Erythrodiol

Paganuzzi (1982) reported that the determination of erythrodiol and uvaol pro-
vides a good means of differentiating between pressed and solvent-extracted
olive oils. Today, the determination of erythrodiol is not sufficient to detect the
presence of olive-pomace oil in olive oil because a relatively large quantity of vir-
gin olive oils produced in certain regions contain a high amount of erythrodiol.
On the other hand, the erythrodiol content of olive-pomace oils can be reduced by
applying some technological procedures. According to IOOC (1996a), the ratio
of erythrodiol and uvaol to the total sterols for each category of olive oil should
not exceed 4.5.
The method recommended by EC (1991) for the determination of erythrodiol
and uvaol involves saponification with potassium hydroxide solution. The
unsaponifiable fraction is then extracted with diethyl ether and purified by pas-
sage through a column of alumina. The unsaponifiables are subjected to TLC on a
146 HANDBOOK OF OLIVE OIL

I.S

3
4

Figure 6--3 A gas chromatogram of the wax analysis of a sample ofvirgin olive oil. Note:
(I.S) Internal standard, (1) C36 ester, (2) C38 ester, (3) C40 ester, (4) C42 ester, (5) C44 ester,
(6) c46 ester.

silica gel plate until the bands corresponding to the sterol and erythrodiol frac-
tions are separated. The sterols and the erythrodiol recovered from the plate are
transformed into TMS ethers, and the mixture is analyzed by GC. The results are
expressed as the percentage of erythrodiol in the mixture of erythrodiol and
sterols.

Fatty Acid

The analysis of fatty acids of edible oils by GLC has become a routine analyti-
cal tool. The edible oil that is subjected to analysis is heated under reflux with
methyl alcohol and sodium methylate. The methyl esters obtained are extracted
with diethyl ether. The analysis of esters should be carried out by a capillary
column-GC and a flame-ionization detector (EC 1991). Currently, according to
Mariani and Fedeli (1993), GC is one of the most useful methods for analyzing
fatty acids and other components of olive oil, and it cannot be replaced by any
other instrumental method.
 Analysis ofEdible Oils 147

Saturated Fatty Acids at Position-2 ofTriacylglycerols

Palmitic and stearic acids are the main saturated fatty acids present in olive oil
and in olive-pomace oil. Stearic acid is almost never found in position-2 of virgin
oils but can be present at 0.2--0.3 percent in olive-pomace oils. The saturated fatty
acid level (sum of palmitic and stearic acids) at the position-2 for virgin olive oil
should be :::::;I .5 percent.
Determination of the saturated fatty acids at this position of the triacylglycerol
molecule can be accomplished by the method recommended by EC (1995). The
method involves possible neutralization, purification by passing the oil through
an alumina column, partial hydrolysis of the triacylglycerols by pancreatic lipase,
separation of the formed monoacylglycerols by TLC, and methanolysis of the
monoacylglycerols. Analysis of these methyl esters by GLC follows. This method
is applicable to all edible oils having a melting point below 45°C. It is not applic-
able unreservedly to oils containing substantial amounts of fatty acids with
twelve or less carbon atoms (coconut, palm kernel oil, butterfat), highly unsatu-
rated oils with more than four double bonds containing twenty or more carbon
atoms (fish and marine oils), or fatty acids containing oxygenated groups, other
than the acid group.
Sacchi and coworkers ( 1992) determined the saturated fatty acids in positions
1,3 and 2 of olive oil triacylglycerols by using NMR spectroscopy and avoiding
chemical procedures. The composition and positional distribution of the fatty
acids in the glycerol molecules of oils has been determined with the application
of 13 C NMR. This technique has the advantages of being nondestructive and
requiring no chemical manipulation.

Trans Fatty Acid

A large number of techniques and methods is available for analyzing trans

fatty acids in edible oils, including infrared (IR) and NMR spectroscopy, packed
and capillary column-GC, HPLC, and supercritical fluid chromatography (SFC).
Firestone and Sheppard (1992) described these in detail. van de Voort, Ismail,
and Sedman (1995) developed a rapid FTIR method to determine simultaneously
the cis and trans content of fats and oils while Baeten et al. ( 1996) described the
possibilities of FT-Roman spectroscopy.
According to Pallotta (1976), the detection of elaidic acid by IR spectroscopy
reveals adulteration of olive oil by olive-pomace oil. Further, the presence of a
large amount of trans octadecenoic acids in olive oil might be due to adulteration
with an oil high in oleic acid (e.g., canota oil) that has been bleached to lower the
sterol content. Raman spectroscopy (RS) has been used for determining the cis
and trans unsaturation of oils and could be useful in detecting the adulteration of
virgin olive oil. C=C stretch for a series of methyl ester and triacylglycerol refer-
148 HANDBOOK OF OLIVE OIL

ence compounds was considered to fall within the limits 1626 and 1691 em -I,
and 1,664 cm- 1 was designated as the distinctive absorbance of a trans bond
(Firestone & Sheppard 1992). FTIR spectroscopy also has been used for quanti-
fying the levels of total trans-triacylglycerols in oils (Mossoba, Yurawecz &
McDonald 1996). It requires about five minutes for spectroscopic measurement,
band area integration, and calculation of the trans content from a linear regres-
sion equation. The trans levels, determined by attenuated total reflection, were
close to those found by capillary column-GC. IOOC (1997) recommends the
determination of the trans fatty acid content of olive oil by capillary column-GC.

Triacylglycerol

Triacylglycerol analysis always has been of great interest in the food industry
(Antoniosi Filho, Carrilho & Lan9as 1993). In particular, the triacylglycerol pro-
file can give valuable information related to the authenticity and origin of olive
oils (Cortesi, Rovellini & Fedeli 1990; EC 1991; Stefanoudaki, Kotsifaki &
Koutsaftakis 1997; Synouri et al. 1995). According to the method recommended
by IOOC (1996a), prior to triacylg1ycerol analysis, oils should undergo purifica-
tion. The method involves separation and identification of the triacylglycerol
groups having the same carbon number by RP-HPLC. The determination of the
equivalent carbon number (ECN) is necessary in order to carry out triacylglyc-
erol analysis. Table 6--4 shows the content of the values of ECN in other edible
oils. The equivalent carbon number is

ECN = CN- 2n

where CN = the total number of carbon atoms of the fatty acids in the triacyl-
glycerol molecule, and n = the number of double bonds.
If n0 , n~> and n 1n are the numbers of double bonds attributed to oleic, linoleic,
and linolenic acids, respectively, the equivalent carbon number can be calculated
as:

where the coefficients d0 , db and d 1n can be calculated by means of the reference

triacylglycerols.
Under the conditions specified in the method recommended by EC (1991), the
ECN is calculated as:

ECN = CN - (2,60 1\,) - (2,35 n 1) - (2, 17 n 1n)

The method is applicable to all vegetable oils containing triacylglycerols of

long-chain fatty acids. It is especially applicable to the detection of the presence
 Analysis ofEdible Oils 149

Table 6-4 Equivalent Carbon Number Values of Some Edible Oils

Oil C48 cso C52 C54 C56

Canol a 0-3 10-18 68-80 4-7

Raw cottonseed Q-2 13-20 40-48 32-44 0-5
Raw corn Q-1 2-7 23-33 55-70 1-3
Raw palm 6-10 34-45 36-44 7-16 0-1
Soybean 2-4 25-30 60-70 1-5
Sunflower 1-2 16-19 75-80 1-3

of small quantities of semidrying oils (rich in linoleic acid with iodine values of
100-150) in vegetable oils containing oleic acid as the predominant unsaturated
fatty acid (e.g., olive oil). As stated above, the method is based on the separation
of triacylglycerols according to their equivalent carbon number by HPLC and
interpretation of the chromatograms.
Olive oil is characterized by four major peaks with ECNs of 44, 46, 48, and 50.
Triacylglycerols with ECN s of 40 are absent from olive oil, and those with an
ECN of 42 are present only in trace amounts. In contrast, all the common edible
oils rich in linoleic acid (corn, sunflower, and soybean) are characterized by a
large HPLC peak with an ECN of 42. Canola oil contains lower but significant
amounts of these triacylglycerols (Salivaras & McCurdy 1993). This oil shows a
triacylglycerol composition more similar to olive oil because the main peak rep-
resents triacylglycerols with an ECN of 48. Figure 6-4 shows the triacylglycerol
profiles of olive oil adulterated with soybean oil and canola oil.
Several methods have been applied for the determination of triacylglycerols.
Phillips et al. (1984) used RP-HPLC for analyzing the triacylglycerols of olive oil
and other oils. Tsimidou and Macrae (1987) also used RP-HPLC to determine the
triacylglycerol profile of oils obtained after low-temperature crystallization and
suggested that this procedure can be applied for detection of olive oil adulteration.
Stefanoudaki, Kotsifaki & Koutsaftakis ( 1997) used isocratic HPLC to study the
triacylglycerol fraction of a number of olive oil samples obtained from two differ-
ent cultivars in different locations and from a different maturity stage of olive fruit.
Analysis of triacylglycerols by RP-HPLC allow detection and even identifica-
tion of seed oils used to adulterate olive oil (Salivaras & McCurdy 1992, 1993).
Addition of as little as 2.5 percent of corn, sunflower, and soybean oils can be
detected in olive oil mainly because of the high content of triacylglycerols with an
ECN of 42 in these seed oils. Antoniosi Filho, Carrilho & Lanc;as (1993) applied
HT-CGC for fast quantitative analysis oftriacylglycerol profile. A low level (4 per-
cent) of soybean oil was detectable in olive oil by determining the trilinolein abun-
dance in soybean oil. Similarly, Capella (1993) determined diacylglycerols and tri-
150 HANDBOOK OF OLIVE OIL

A 48 B 48

46
6

40

0 4 0 4 8 12 16

c 48 D 48
....c:: ......
c 46
0
..::0
til
-0
>
0
"'-l

444
40
44
40
50 50

0 4 8 12 16 0 4 8 12 16
Time (min) Time (min)

Figure 6-4 Triacylglycerol profiles of adulterated samples of olive oil with soybean oil
and canola oil. A: Virgin olive oil, B: Adulterated with 7.5% soybean oil, C: Adulterated
with 30% soybean oil, and D: Adulterated with 30% canola oil.

acylglycerols in olive oil by HRGC, and Frega, Bocci, and Lercker (1990) deter-
mined triacylglycerols in several oils by gas and liquid chromatography.

Stigmastadienes

Analysis of the hydrocarbon stigmasta-3,5-diene has been proposed (IOOC

1997; EC 1991, 1995) for detecting refined oils (olive oil, olive-pomace oil, sun-
flower oil, and palm oil) in virgin olive oil. According to the method recom-
mended by EC (1991), the determination of stigmastadiene involves saponifica-
tion of the oil and separation of the unsaponifiable fraction. The separation of the
steroid hydrocarbons is accomplished by TLC separation on silica gel plates and
the analysis and determination by GC with a capillary column. The method is
 Analysis ofEdible Oils 151

applied to all edible vegetable oils. The measurements are valuable, however,
when the values are 0.01-4.0 mg/kg.
Virgin olive oil and crude olive-pomace oil contain low levels of this particular
steroid hydrocarbon. Stigmasta-3,5-diene is formed by the removal of one mole-
cule of water from 13-sitosterol, which is facilitated during refining and mainly
during bleaching (Cert et al. 1994; Lanz6n et al. 1994). Table 6-5 shows the stig-
mastadiene content of different classes of olive oil. The R1 ratio of stigmasta-3,5-
diene/campesta-3,5-diene has to be applied when the stigmastadiene content is
greater than 4 ppm.

6. 7 ADVANCED METHODS OF OIL ANALYSIS

High-resolution 1H nuclear magnetic resonance eH NMR) lately has been

used for the direct, rapid, and automated determination of the iodine value (IV) of
edible oils, including hydrogenated oils (Miyake, Yokomizo & Matsuzaki 1998).
The total time required is about 3 minutes per sample. The IV is calculated from
the number of double-bonded protons and the average molecular weight derived
directly from the spectrum. To predict the proportions of saturated, monounsatu-
rated, and polyunsaturated acyl groups of edible oils, Guillen and Cabo ( 1997)
used Fourier transform infrared spectroscopy (FT-IRS) analysis. The bands of the
spectra were assigned to different functional group vibrations and were related to
the composition of the oil.
Goodacre, Kell, and Bianchi (1992) applied curie-point pyrolysis mass spec-
troscopy (PyMS) to detect adulteration of virgin olive oil with refined olive oil
and other vegetable oils. Neural network analysis (NNA) permits correct identifi-
cation ofthe oils used for adulteration on the basis of the mass spectra. A combi-
nation ofPyMS and artificial NNA is useful to assess adulteration of virgin olive

Table 6-5 Stigmastadiene Content of Olive Oils

Category of Olive Oil Stigmastadienes (ppm) R1

Edible virgin olive oils ~0.15

Lampante virgin olive oil ~0.50
Refined olive oil ~50* 2::12
Olive oil ~50* 2::12
Crude olive-pomace oil ~5* No limit
Refined olive-pomace oil ~120* 2::10
Olive-pomace oil ~120* 2::10

*Provisional limits.
152 HANDBOOK OF OLIVE OIL

oil with corn, peanut, soya, sunflower, or olive-pomace oil (Goodacre, Kell &
Bianchi 1993).
Wesley, Pacheco, and McGill (1996) applied NIR spectroscopy and principal
component analysis (PCA) to develop a discriminant analysis equation useful for
identifying the 90 percent of cases of extra virgin olive oil adulteration by seed
oils. Partial least-squares analysis is used to develop a calibration equation that
could predict the level of adulteration. The calibration predicts the purity of the
oil to an accuracy of ± 1.5 percent (Wesley, Barnes & McGill 1995). Figure 6-5
shows the NIR spectra of some edible oils.
FTRS was first used by Sadeghi-Jorabchi et al. (1990) to determine total unsat-
uration in edible oils and margarines without pretreatment. RS can measure cis
and trans isomers simultaneously (Firestone & Sheppard 1992) and could be

1100 1300 1500 1700 1900 2100

Wavelength (nm)

Figure 6-5 The near-infrared spectra of some edible oils. Note: a, rapeseed oil; b, soy-
bean oil; c, sunflower oil; d, extra virgin olive oil.
 Analysis ofEdible Oils 153

used to detect adulteration with oils containing trans fatty acids. C=C stretch for
a series of methyl esters and triacylglycerol reference compounds was considered
to fall within the limits 1226 and 1691 cm- 1, and 1664 cm- 1 was designated as
the boundary between cis and trans regions. FTRS was recently used to detect
adulteration of virgin olive oil with soybean oil, com oil, and olive-pomace oil.
According to Baeten et al. (1996), this technique is capable of detecting adulter-
ation of olive oil beyond the limits achieved with other methods.
Favier et al. (1998) used optothermal window spectroscopy (a spectroscopic
technique related to photoacoustic spectroscopy), at C02 laser infrared wave-
lengths to detect the extent of adulteration of extra virgin olive oil by sunflower
and safflower oil. A good linearity between the strength of optothermal signal and
the concentration of each adulterating compound was recorded. According to
Favier et al., the predicted limits of detection by this new technique were 6 (w/w)
and 4.5 (w/w) for safflower and sunflower oils respectively. Marigheto et al.
(1998), on the other hand, compared mid-infrared (MIR) and Raman spectro-
scopies for their ability to discriminate between oils of different botanical origin
to detect added adulterants. They denoted that the mid-infrared method in combi-
nation with linear discriminant analysis (LDA) gives the best classification and
adulteration detection levels, compared to Raman technique, although the authors
did not use selected Raman shifts (Baeten et al. 1996) but the whole spectra.

REFERENCES
Akasbi M., Shoeman, D. W. & Saari Csallany, A. (1993). High-performance liquid chromatography of
selected phenolic compounds in olive oils. JAm Oil Chern Soc 70, 367-370.
American Oil Chemists Society (AOCS) (1987). Official Methods and Recommended Practices, 4th
edn. Edited by D. Firestone. Champaign, IL: American Oil Chemists Society.
Antoniosi Filho, N. R., Carrilho, E. & Lan~as, F. M. ( 1993). Fast quantitative analysis of soybean oil
in olive oil by high-temperature capillary gas chromatography. JAm Oil Chern Soc 70, 1051-1053.
Aparicio, R., et al. (1994). Relationship between the COl test and other sensory profiles by statistical
procedures. Grasas Aceites 45, 26--41.
Aparicio, R., Alonso, V & Morales, M. T. (1996). Developments in olive oil authentication. In Pro-
ceedings of Food Authenticity '96: Method for the Measurement of Food Authenticity and Adulter-
ation. Norwich, England: Institute of Food Research.
Aparicio, R., Calvente, J. J. & Morales, M. T. (1996). Sensory authentication of European extra-virgin
olive oil varieties by mathematical procedures. J Sci FoodAgric 72,435-447.
Aparicio, R. & Mcintyre, P. ( 1998). Oils and Fats Commodity Group. In FAIM - Food Authenticity:
Issues and Methodologies. Edited by M. Lees. Nantes, France: Eurofins Scientific.
Aparicio, R. & Morales, M. T. (1995). Sensory wheels: A statistical technique for comparing QDA
panels. Application to virgin olive oil. J Sci FoodAgric 67,247-257.
Aparicio, R., Morales, M. T. & Alonso, M. V (1996). Relationship between volatile compounds and
sensory attributes by statistical sensory wheel. JAm Oil Chern Soc 73, 1253-1264.
Baeten, V, et a/. (1996). Detection of virgin olive adulteration by Fourier transform Raman spec-
troscopy. J Agric Food Chern 44, 2225-2230.
154 HANDBOOK OF OLIVE OIL

Berezin, 0 Yu., et al. (1997). Practical limitations in determining vegetable oil acid values by a novel
pH-metric method. JAm Oil Chern Soc 74, 1339-1341.
Bianchi, G., et al. (1994). Chemical structure of long-chain esters from "Sansa" olive oil. JAm Oil
Chern Soc 71, 365-369.
Blekas, G. & Guth, H. (1995). Evaluation and quantification of potent odorants of Greek virgin olive
oils. In Food Flavors, in Generation, Analysis and Process Influence, pp. 419-427. Edited by G.
Charalambous. Amsterdam: Elsevier Science.
Blekas, G., Guth, H. & Grosch, W (1994). Changes in the levels of olive oil odorants during ripening
of the fruits. In Trends in Flavor Research, pp. 499-502. Edited by H. Maarse & D. G. Vander Heij.
Amsterdam: Elsevier Sciences.
Boskou, D. (1978). Stability of natural terpenoids in heated olive oil. Grasas Aceites 29, 193-195.
Boskou, D. & Morton, I. D. (1975). Changes in the sterol composition of olive oil on heating. J Sci
FoodAgric 26, 1149-1153.
Capella, P. (1993). Analisis del aceite de oliva. 0/ivae 45, 24-27.
Cert, A., et al. (1994). Formation of stigmasta-3,5-diene in vegetable oils. Food Chern 49, 287-293.
Choo, Y. M., eta/. (1996). Separation of crude palm oil components by semipreparative supercritical
fluid chromatography. JAm Oil Chern Soc 73, 523-525.
Christie, W W, eta/. (1991). Stereospecific analysis oftriacyl-sn-glycerols via resolution of diastere-
omeric diacylglycerol derivatives by high-performance liquid chromatography on silica. JAm Oil
Chern Soc 68,695-701.
Cinquanta, J., Esti, M. & Notte, E. L. (1997). Evolution of phenolic compounds in virgin olive oil dur-
ing storage. JAm Oil Chern Soc 74, 1259-1264.
Codex Alimentarius Commission (1993). Proposed Draft Standard for Named Vegetable Oils, CX
1993/16, issued by the Joint FAO/WHO Food Standards Program, via delle Terme di Caracalla
00100 Rome.
Cortesi, N., Rovellini, P. & Fedeli, E. (1990). Triglycerides of natural oils. Note 1. Riv Ita/ Sostanze
Grasse 67,69-73.
Damiani, P. & Burini, G. (1980). Determination of the triglyceride composition of olive oil by a mul-
tistep procedure. J Agric Food Chern 28, 1232-1236.
Dobarganes, M. C. & Marquez-Ruiz, G. (1993). Size exclusion chromatography in the analysis of
lipids. In Advances in Lipid Methodology-Two, pp. 113-137. Edited by W W Christie. Dundee,
Scotland: Oily Press.
European Communities (EC) (1991). Official Journal of the Commission of the European Communi-
ties. Regulation No. 2568/91, L248, September 5, 1991.
European Communities (EC) (1995). Official Journal of the Commission of the European Communi-
ties. Regulation No. 656/95, March 29, 1995.
Favier, J. P. eta/. (1998). C0 2 laser infrared optothermal spectroscopy for quantitative adulteration
studies in binary mixtures of extra-virgin olive oil. JAm Oil Chern Soc 75, 359-362.
Fedeli, E. (1977). Lipids of olives. In Progress on Chemistry of Fats and other Lipids, pp.l5-74.
Edited by E. Ralph & T. Holman. Paris: Pergamon Press.
Firestone, D. & Reina, R. J. (1996). Authenticity of vegetable oils. In Food Authentication, pp.
198-246. Edited by P.R. Ashurst & M. J. Dennis. London: Chapman and Hall.
Firestone, D. & Sheppard, A. (1992). Determining the purity of olive oil by gas chromatography. In
Advances in Lipid Methodology-One, pp. 273-322. Edited by W W Christie. Dundee, Scotland:
Oily Press.
Frega, N., Bocci, F. & Lercker, G. (1990). The HRGC determination oftriglycerides. !tal J Food Sci 2,
257-264.
 Analysis ofEdible Oils 155

Garcia Mesa, J. A., Luque de Castro, M. D. & Valcarcel, M. (1993). Coupled robot-flow injection
analysis system for fully automated determination of total polyphenols in olive oil. Anal Chern 65,
3540--3542.
Goiffon, J.P., Remniac, C. & Furon, D. (1981). Application de la chromatographie liquide haute per-
formance a I' analyse des triglycerides des corps gras. II- Grandeurs de retention des triglycerides.
Rev Fr Corps Gras 28, 199-206.
Goldberg Federico, L., et al. (1969). Indagini sperimentali sulla struttura gliceridica dei grassi natu-
rali. Nota III-Olea Europaea e Glycine Hispida. Riv /tal Sostanze Grasse 46, 102-105.
Gomes, T. (1992). Oligopolymer, diglyceride and oxidized triglyceride contents as measures of olive
oil quality. JAm Oil Chern Soc 69, 1219-1223.
Goodacre, R., Kell, D. B. & Bianchi, G. (1992). Neural networks and olive oil. Nature 359, 594.
Goodacre, R., Kell, D. B. & Bianchi, G. (1993). Rapid assessment of the adulteration of virgin olive
oils by other seed oils using pyrolysis mass spectrometry and artificial neural networks. J Sci Food
Agric 63, 297-307.
Gracian, J. & Cota, J. (1984). Estudio cualitativo y cuantitativo de la fracci6n de alcoholes grasos de
los aceites de oliva y de orujo. Grasas Aceites 35, 358-363.
Guillen, M. D. & Cabo, N. (1997). Characterization of edible oils and lard by Fourier transform
infrared spectroscopy. Relationships between composition and frequency of concrete bands in the
fingerprint region. JAm Oil Chern Soc 74, 1281-1286.
Gutfinger, J. & Letan, A. ( 1974). Studies of unsaponifiables in several vegetable oils. Lipids 9,
658-663.
Guth, H. & Grosch, W. (1993). Quantitation of potent odorants of virgin olive oil by stable-isotope
dilution assays. JAm Oil Chern Soc 70,513-518.
Gutierrez, F., et al. (1992). Bitter taste of virgin olive oil: Correlation of sensory evaluation and
instrumental HPLC analysis. J Food Sci 54, 68-70.
International Olive Oil Council (IOOC) (1996a). Trade Standard Applying to Olive Oil and Olive-
Pomace Oil. COIIT.l5 No 2/Rev. 5 Madrid, Spain.
International Olive Oil Council (IOOC) (1996b). Spectrophoto-Metric Investigation in the Ultravio-
let. COIIT.20 No 19. Madrid, Spain.
International Olive Oil Council (IOOC) (1997). Trade Standard Applying to Olive Oil and Olive-
Pomace Oil. COIIT.15 No 2/Rev. 6. Madrid, Spain.
International Union of Pure and Applied Chemistry (IUPAC). (1987). Standard Methods for the
Analysis of Oils, Fats, and Derivatives: Method 2501, Determination of the Peroxide Value (PV.).
7th ed., pp. 199-200. Edited by C. Paquot & A. Haufenne. Oxford, England: Blackwell Scientific
Publications.
Ismail, A. A., et al. (1993). Rapid quantitative determination of free fatty acids in fats and oils by
Fourier transform infrared spectroscopy. JAm Oil Chern Soc 70, 335-341.
Itoh, T., et al. (1981). Triterpene alcohols and sterols of Spanish olive oil. JAm Oil Chern Soc 58,
545-550.
Kiosseoglou, B., Vlachopoulou, I. & Boskou, D. (1987). Esterified 4-monomethyl-and 4'4-dimethyl-
sterols in some vegetable oils. Grasas Aceites 38, 102-103.
Kiritsakis, A. K. ( 1998a). Flavor components of olive oil-A review. JAm Oil Chern Soc 75,
673-681.
Kiritsakis, A. K. (1998b). Olive Oil. 2nd edn. Trumbull, CT: Food and Nutrition Press.
Kiritsakis, A. K. & Markakis, P. (1991). Olive oil analysis. In Modern Methods of Plant Analysis:
Essential Oils and Waxes, pp. 1-20. Edited by H. F. Linskens & J. F. Jackson. Berlin: Springer-Ver-
lag.
156 HANDBOOK OF OLIVE OIL

Kiritsakis, A. K. & Min, D. ( 1989). Flavour chemistry of olive oil. In Flavor Chemistry ofLipid Foods,
pp. 196-221. Edited by D. Min & T. Smouse. Champaign, IL: American Oil Chemists Society.
Lanz6n, A., et al. (1994). The hydrocarbon fraction of virgin olive oil and changes resulting from
refining. JAm Oil Chem Soc 71, 285-291.
Mahungu S.M., Hansen, S. L. &Artz, WE. (1994). Quantitation ofvolatile compounds in heated tri-
olein by static headspace capillary gas chromatography/infrared spectroscopy-mass spectrometry. J
Am Oil Chem Soc 71, 453-455.
Mariani, C. & Fedeli, E. ( 1985). Determinazione delle forme gliceridiche presenti negli oli alimentari.
Nota 1: caso dell'olio di oliva. Riv !tal Sostanze Grasse 62, 3-7.
Mariani, C. & Fedeli, E. (1993). La cromatografia de gases en el analisis de aceite de oliva. Olivae 45,
34-39.
Marigheto, N. A. eta!. (1998). A comparison of mid-infrared and Raman spectroscopies for the
authentication of edible oils. JAm Oil Chem Soc 75, 987-992.
Matthaus, B. ( 1996). Determination of the oxidative stability of vegetable oils by rancimat and con-
ductivity and chemiluminescence measurements. JAm Oil Chem Soc 73, 1039-1043.
May, WE. & Hume, D. J. (1993). An automated gas-liquid chromatography method of measuring free
fatty acids in canola. JAm Oil Chem Soc 70, 229-233.
McGill, A. S., eta/. (1993). The composition and concentration of n-alkanes in retail samples of edi-
ble oils. J Sci FoodAgric 61, 357-362.
Min, D. B. (1983). Analyses of flavor qualities of vegetable oils by gas chromatography. JAm Oil
Chem Soc 60, 544-545.
Miyake, Y., Yokomizo, K. & Matsuzaki, N. (1998). Rapid determination of iodine value by 1H nuclear
magnetic resonance spectroscopy. JAm Oil Chem Soc 75, 15-19.
Miyazawa, T., et at. (1994). Rapid estimation of peroxide content of soybean oil by measuring ther-
moluminescence. JAm Oil Chem Soc 71, 343-345.
Montedoro, G., Bertuccioli, M. & Anichini, F. (1978). Aroma analysis of virgin olive oil by head-
space (volatiles) and extraction (polyphenols) techniques. In Flavor of Foods and Beverages:
Chemistry and Technology, pp. 247-281. Edited by G. Charalambous & G. E. Inglett. New York:
Academic Press.
Morales, M. T., et at. (1995). Virgin olive oil aroma: Relationship between volatile compounds and
sensory attributes by Chemometrics. J Agric Food Chem 43, 2925-2931.
Morales, M. T., Aparicio, R. & Calvente, J. J. (1996). Influence of olive ripeness on the concentration
of green aroma compounds in virgin olive oil. Flavour Fragr J 11, 171-178.
Morales, M. T., Aparicio, R. & Rios, J. J. (1994). Dynamic head-space gas chromatographic method
for determining volatiles in virgin olive oil. J Chromatogr A 668, 455-462.
Mossoba, M. M., Yurawecz, M.P. & McDonald, R. E. (1996). Rapid determination of the total trans
content of neat hydrogenated oils by attenuated total reflection spectroscopy. JAm Oil Chem Soc
73, 1003-1009.
Motta, L., et al. (1987). Considerazioni sull'analisi gascromatografica dei trigliceridi con utilizzo di
colonne capillari. Riv !tal Sostanze Grasse 64, 545-549.
Paganuzzi, V. (1982). Applicazione della CSS-AgN03 all'analisi degli alcoli triterpenici dell'olio
d'oliva. Nota II: oli vergini ed oli di sansa greggi. Riv Ita/ Sostanze Grasse 59, 415-422.
Pallotta, U. (1976). Analytic problems in the ascertainment of olive oil genuineness. In Lipid Technol-
ogy, vol. 2., pp. 387-409. Edited by R. Paoletti, G. Jacini and R. Porsellati. New York: Raven Press.
Perrin, J. L. & Naudet, M. (1983). Identification et dosage des triglycerides des corps gras naturels par
CLHP. Rev Fr Corps Gras 30,279-285.
 Analysis of Edible Oils 157

Phillips, F. C., et al. (1984). Quantitative analysis of triglyceride species of vegetable oils by high per-
formance liquid chromatography via a flame ionization detector. Lipids 19, 880-887.
Rahmani, M., & Csallany, A. S. (1985). Mise au point d'une methode de chromatographie liquide a
haute performance (CLHP) pour la determination des pigments chlorophylliens dans les huiles
vegetates. Rev Franc Corps Gras 32, 257-259.
Rossell, J. B. (1991). Vegetable oils and fats. In Analysis ofOilseeds, Fats and Fatty Foods, pp.
261-327. Edited by J. B. Rossell & J. L. R. Pritchard. London: Elsevier Science.
Sacchi, R., et al. (1992). Analysis of the positional distribution of fatty acids in olive oil triacylglyc-
erols by high resolution 13 C-NMR of the carbonyl region. Ita/ J Food Sci 4, 117-123.
Sacchi, R., eta/. (1996). A high-field 1H nuclear magnetic resonance study of the minor components
in virgin olive oils. JAm Oil Chern Soc 73, 74 7-758.
Sadeghi-Jorabchi, H., eta/. (1990). Determination of the total unsaturation in oils and margarines by
Fourier transform Raman spectroscopy. JAm Oil Chern Soc 67,483-486.
Salivaras, E. & McCurdy, A. R. (1992). Detection of olive oil adulteration with canola oil from tria-
cylglycerol analysis by reversed-phase. JAm Oil Chern Soc 69, 935-938.
Salivaras, E. & McCurdy, A. R. (1993) Adulteration of olive oil: detection through reversed-phase
high performance liquid chromatography. In Food Flavors, Ingredients and Composition, pp.
279-300. Edited by G. Charalambous. Amsterdam: Elsevier Science.
Santinelli, F., Damiani, P. & Christie, W W (1992). The triacylglycerol structure of olive oil deter-
mined by silver ion high-performance liquid chromatography in combination with stereospecific
analysis. JAm Oil Chern Soc 69, 552-556.
Serani, A. & Piacenti, D. (1992). Kinetics ofpheophytin-A photodecomposition in extra virgin olive
oil. JAm Oil Chern Soc 69, 469-470.
Singleton J. A. & Stikeleather, L. F. (1995). High-performance liquid chromatography analysis of
peanut phospholipids. I: Injection system for simultaneous concentration and separation of phos-
pholipids. JAm Oil Chern Soc 72, 481-483.
Snyder, M. J. & King, J. W (1994). Oilseed volatile analysis by supercritical fluid and thermal desorp-
tion methods. JAm Oil Chern Soc 71, 261-265.
Stefanoudaki, E., Kotsifaki, F. & Koutsaftakis, A. (1997). The potential of HPLC triglyceride profiles
for the classification of Cretan olive oils. Food Chern 60, 425-432
Synouri, S., eta/. (1995). Estudio de las caracteristicas de los aceites de oliva virgenes griegos. 0/ivae
57,27-33.
Tacchino, C. E. & Borgoni, C. (1983). Indagine sui contenuto di alcoli alifatici dell'olio d'oliva di
presione e di estrazione. Riv Ita/ Sostanze Grasse 60, 575-581.
Tateo, E., et a/. (1993). New trends in the study of the merits and shortcoming of olive oil in
organoleptic terms in correlation with GC/MS analysis of the aromas. In Food Flavors, Ingredients
and Composition, pp. 301-311. Edited by G. Charalambous. Amsterdam: Elsevier Science.
Terao, J. & Matsushita, S. (1977). Products formed by photosensitized oxidation of unsaturated fatty
acid esters. JAm Oil Chern Soc 54, 234-238.
Toschi, T. G., Christie, W. W & Conte, L. S. (1993). Capillary gas chromatography combined with
high performance liquid chromatography for the structural analysis of olive oil triacylglycerols. J
High Resol Chromatogr 16, 725-730.
Tsimidou, M. & Macrae, R. (1987). Detection and quantitative determination of adulteration of olive
oil. Intern Analyst 2, 29-34.
Tsimidou, M., Papadopoulos, G. & Boskou, D. (1992). Determination of phenolic compounds in vir-
gin olive oil by reversed-phase HPLC with emphasis on UV detection. Food Chern 44, 53-60.
158 HANDBOOK OF OLIVE OIL

Tur'yan, Va I., et al. (1996). pH-metric determination of acid values in vegetable oils without titration.
JAm Oil Chern Soc 73, 295-301.
van de Voort, F. R., et al. (1994). Monitoring the oxidation of edible oils by Fourier transform infrared
spectroscopy. JAm Oil Chern Soc 71, 243-253.
van de Voort, F. R., Ismail, A. A. & Sedman, J. (1995). A rapid automated method for the determina-
tion of cis and trans content of fats and oils by Fourier transform infrared spectroscopy. JAm Oil
Chern Soc 72, 873-880.
Vazquez, R., Del Valle, A. J. & Del Valle, J. M. (1976). Componentes fen6licos de la aceituna. Grasas
Aceites 21, 185-191.
Viinanen, E. & Hopia, A. (1994). Reversed-phase high-performance liquid chromatography analysis
of triacylglycerol autoxidation products with ultraviolet and evaporative light-scattering detectors.
JAm Oil Chern Soc 71, 537-539.
Wesley, I. J., Barnes, R. J. & McGill, A. E. J. (1995). Measurement of adulteration of olive oils by
near-infrared spectroscopy. JAm Oil Chern Soc 72, 289-292.
Wesley, I. J., Pacheco, F. & McGill, A. J. (1996). Identification of adulterants in olive oils. JAm Oil
Chern Soc 73, 515-518.
Zamora, R., Navarro, J. L. & Hidalgo, F. J. (1994). Identification and classification of olive oil by
high-resolution 13 C nuclear magnetic resonance. JAm Oil Chern Soc 71, 361-364.