Using offline HPLC-GC-FID 4-Desmethylsterols Concentration Profiles, Combined with Chemometric Tools, to Discriminate Different Vegetable Oils

In order to improve the separation of the sterol classes present in vegetable oils, an isocratic off-line HPLC fractionation of unsaponifiable has been developed using a cyano-bonded phase column as stationary phase and hexane/tert-butylmethyl ether (80:20 (v/v)) as mobile phase, which permits to obtain them in less than 12 min. With the aim of evaluating the developed procedure, different samples of edible oils, such as extra virgin and pomace olive oil, sunflower, soybean, etc., were analyzed by GC-FID. The concentration of each phytosterol was found by means of an internal calibration at one level, using alpha-cholestanol as surrogate and theoretical correction factors. Their concentration profiles have been used to provide a pattern capable of distinguish among different vegetable oils. The chemometric study based on the Hierarchical Cluster Analysis and Factor Analysis (FA) using principal components and Varimax rotation was applied to the Box-Cox transformed data set to obtain the natural classes used in the final Linear Discriminant Analysis. Two discriminant functions explaining the 94.89 % of the total variance has been considered enough to classify the studied samples.