Study of the influence of triacylglycerol composition on DSC cooling curves of extra virgin olive oil by chemometric data processing

In this work, a coupled high-pressure liquid chromatography-differential scanning calorimetry-partial least square (HPLC-DSC-PLS) procedure was applied to clarify the influence of triacylglycerol composition on the shape of cooling curves for extra virgin olive oil (EVOo), as chemometric processing of digitized DSC curves was previously reported to be an attractive alternative to the application of the common procedure of parameter extrapolation for the analysis of thermal transition. 69 samples of EVOo were analysed to obtain triacylglycerol (TAG) composition by means of HPLC and DSC cooling profiles. Results obtained by the application of PLS algorithm on TAG concentration (%) and digitized DSC curves showed that cooling transitions were markedly influenced by OOO, OLO and OOP + SLO that are the most representative TAG for EVOo. Other TAG as LLP + OLnO, LLL + LLPo and POPo developed good PLS models, appearing to influence EVOo cooling curve, although less markedly than the further. Otherwise, the other identified TAG did not render appropriate models. Finally, grouping TAG according to different unsaturation degree, high correlation coefficients (>0.80) and low relative standard deviations (<11 %) were found for sum of tri-unsaturated triacylglycerols in both calibration and validation sets. Starting from these encouraging results, this new and fast coupled approach may be applied to a wider set of EVOo samples to tentatively discriminate among oils according to different geographical and/or botanical origins taking into account relation established among TAG and cooling curves.