A COMPARISON OF MIDINFRARED AND RAMAN SPECTROSCOPIES FOR THE AUTHENTICATION OF EDIBLE OILS

Robust, routine, and rapid instrumental methods for the determination of the authenticity of edible oils, and the detection of adulteration, are continually being sought. In this paper, we compare mid-infrared and Raman spectroscopies for their ability to discriminate between oil s of differing botanical origin and for their ability to detect added adulterants. Furthermore, we used sufficient numbers of samples to per mit a comparison of some of the chemometric methods (linear discrimina nt analysis, artificial neural networks) available and looked at the r esults obtained when the two spectroscopic datasets were combined. We show that mid-infrared spectroscopy, in combination with linear discri minant analysis, gave the best classification rates and adulteration d etection levels compared to Raman or combined data.