The degradation process of edible oils of different nature, submitted to heating at 170°C, 190°C and 210°C with aeration, was studied by means of (1)H nuclear magnetic resonance spectroscopy (NMR). In this study, secondary products such as aldehydes were detected and monitored over time. Two complementary analytical approaches were adopted to characterize the kinetics of the appearance of aldehydes in the heated oils. This first was a classical kinetic approach based on the assumption that the overall degradation reaction to form aldehydes follows a rate law of order 1. This approach allowed us to calculate a thermal stability criterion for classifying the oils according to their heat stability. A second approach was to use the spectral fingerprint corresponding to aldehydes in a multivariate data analysis procedure in order to give the major trend in the studied phenomena, taking into account the multiway nature of recorded data. The application of different 3-way and 4-way Tucker3 models led to a better understanding of the chemical stability of the oils studied and was used to determine the order of stability of these oils. This multiway approach provides additional information that 2-way processing (PCA) does not provide clearly, such as the overall contribution of the heating time factor on the chemical evolution of oils. In conclusion, this work shows that a fully chemometric study of NMR spectra allows to order the oils according to their thermal stability and to achieve a result in good agreement with existing analytical and kinetic studies in the literature.
Copyright © 2011 Elsevier B.V. All rights reserved.