The mathematical prediction model for the oxidative stability of vegetable oils by the main fatty acids composition and thermogravimetric analysis

Thermal oxidation stabilities of four kinds of vegetable oils (palm oil, rapeseed oil, sunflower oil, linseed oil), their characteristic fatty acids (FA, palmitic acid, oleic acid, linoleic acid and linolenic acid) and corresponding fatty acid methyl esters (FAME, methyl palmitate, methyl oleate, methyl linoleate and methyl linolenate) were quantified by the parameter of onset temperature (Ton) of the thermal oxidation. The parameter Ton was obtained by the non-isothermal pattern of TGA method in the oxygen atmosphere at different heating rates (1, 5, 7.5, 10, 15, 20 °C/min). The results showed that the stability order of four kinds of vegetable oils was: palm oil > rapeseed oil > sunflower oil > linseed oil. For FAs, palmitic acid and oleic acid were more stable than linoleic acid and linolenic acid. And the same situation was found in FAMEs because of the difference in the unsaturation degree. Furthermore, based on the composition of four most essential fatty acids in the oils, a mathematical model based on the stability of FAs and FAMEs was found to predict the stability of vegetable oils. The mathematical predictions of oil oxidative stability by FAME system were more accurate with lower average deviation below 2%.