Food Chemistry

668 14 Edible Fats and Oils

Table 14.27.Analytical aspects related to the deter-
mination of the extent of oxidation of unsaturated
fatty acids: relative sensitivities of spectrophoto-metric
proceduresa

Method Autoxidized fatty acid
methyl esters
18:2 18:3
(9,12)b (9,12,15)c

UV-Absorption 234 nm 1. 01. 0
270 nm 0. 10. 3
Fe^2 +/Thiocyanate
(rhodanide) 9. 46. 3
Thiobarbituric 452 nm 0. 10. 5
acid test 530 nm 0. 11. 0
Kreis-test 0. 10. 1
Anisidine value 0. 30. 75
Heptanal value 0. 10. 1
aRelated to UV absorption at 234 nm.
bPeroxide value: 475.
cPeroxide value: 450.

determined. Since the proportion of aroma-active
and sensory neutral carbonyls is not known, any
correlation found between the carbonyl value and
aroma defects is clearly coincidental.

Thethiobarbituric acid test(TBA) is a preferred
method for detecting lipid peroxidation in biolog-
ical systems. However, the reaction is nonspecific
since a number of primary and secondary prod-
ucts of lipid peroxidation form malonaldehyde
which in turn reacts in the TBA test. In food con-
taining oleic and linoleic acids, the TBA-test is

not as sensitive as the Fe^2 +-test outlined above.

The gas chromatographic determination of
individual carbonyl compounds appears to be
a method suitable for comparison with findings
of sensory panel tests. Analytical methods for
the odorants causing aroma defects is still in the
early stages of development because only a few
fats or fat-containing foods have been examined
in such detail that the aroma substances involved
are clearly identified.
The well studied warmed-over flavor of cooked
meat (cf. 12.6.2.1) is an example. It can be
controlled relatively easily because the easy-to-
determine hexanal has been identified as the most
important off-flavor substance. On the other hand,
the easily induced rancid aroma defect of rape-
seed oil is primarily caused by the volatile hy-

droperoxides (1-octen-3-hydroperoxide, (Z)-1,5-

octadiene-3-hydroper-oxide) and (Z)-2-nonenal

which can be quantitatively detected only by us-

ing isotopic dilution analyses (cf. 5.2.6). This lim-

itation also applies to 3-methyl-2,4-nonandione,

which appears as the most important off-flavor

substance in soybean oil on exposure to light.

To simplify the analytical procedure, individual

aldehydes (e. g., hexanal, 2,4-decadienal), which

are formed in larger amounts during lipid peroxi-

dation, have been proposed as indicators. In most

of the cases, however, it was not tested whether

the indicator increases proportionally to the off-

flavor substances which cause the aroma defect.

14.5.3.2.2 Shelf Life Prediction Test

To estimate susceptibility to oxidation, the fat

or oil is subjected to an accelerated oxidation

test under standardized conditions so that the

signs of deterioration are revealed within several

hours or days. Examples of such tests are the

Schaal test(fat maintained at 60◦C) and the

Swift stability test(fat kept at 97. 8 ◦C and aerated

continuously). The extent of oxidation is then

measured by sensory and chemical tests such as

peroxide value (cf. 3.7.2), ultraviolet absorption

(suitable for fats and oils containing linoleic or

linolenic acids) or oxygen uptake. There are also

methods based on the fact that in the process of

triglycerol oxidation, when the initiation period

is terminated, large amounts of low molecular

weight acids are released. They are then deter-

mined electro-chemically. During oxidation of

a given fat or oil sample, a good correlation exists

between the length of the induction period and

the shelf-life.

14.5.3.3 Heat Stability

The behavior of a frying oil, when heated, is as-

sessed from the content of oxidized fatty acids

which are insoluble in petroleum ether and from

the smoke point (cf. 3.7.4) of the fat or oil. The

smoke point of a fat or oil is the temperature

at which its triacylglycerols start to decompose

in the presence of air. Smoke is the sign of de-

composition. The smoke point of a fat or oil is

normally in the range of 200–230◦C during pro-