Abstract:

Fatty acids are the major nutrients in vegetable oils in different amounts depending on the type of oil. A predictive
method based on near infrared spectroscopy (NIR) was proposed to determine the contents of four fatty acids in adulterated
sesame oil samples. The chemical values were analyzed by GC. The calibration set consisted of 122 samples and the number
of samples in the validation set was 38. The results showed that linolenic acid (C18:3), arachidonic acid (C20:0), tetracosanoic
acid (C24:0) and myristic acid (C14:0) were the characteristic NIR absorption peaks of the adulterated oils. The optimum
conditions for mathematical modeling of the four components were studied with respect to sample set selection, chemical
value analysis, detection methods and detection conditions. The correlation coefficients (R2) between the NIR-predicted data
and the chemically measured data for the samples of the calibration set were R2 (C14:0) = 0.996, R2 (C18:3) = 0.989, R2 (C20:0) =
0.995, and R2 (C24:0) = 0.993, respectively, and 0.984, 0.949, 0.956, and 0.988 for the validation set samples, respectively.
For the validation set, 9 unknown samples were selected to be analyzed by NIR. Our results demonstrated that the error
between the predicted values and the chemical values was less than 6.0%.

Key words: near infrared spectroscopy, fatty acid, adulterated sesame oil, identification