Authentication of olive oils from the Siurana and Les Garrigues PDOs through fluorescence measurements coupled with chemometric tratments

Identifier:  TFG:3668

Handle: https://hdl.handle.net/20.500.11797/TFG3668

Authors:  Miravent Sans, Albert

Abstract:
Due to the high prices of the extra virgin olive oils (EVOO) and the virgin olive oils (VOO) and the positive effects of their nutritional values, it is widely known that some adulterations are carried out. There is an interest about it for several reasons, for example to decrease the producer price and get more profit in the sales or obtain more quantity of oil without considering the quality. These adulterations are produced by blending with other oils. Some examples can be sunflower or soybean oils. In fact, the European Parliament cites olive oil as one of the first on the list of foods that suffer from fraudulent activity in its January 2014 resolution [reference 2013/2091 (INI)]. The main objective of this work is to classify and differentiate adulterated samples of olive oils, both extra virgin (EVOO) and virgin (VOO) from the Arbequina variety of two Catalan certificates of origin (Les Garrigues and Siurana) and subsequently, to obtain quality parameters of these classification models. Adulterations were produced with refined sunflower oil from the brand of Borges. Samples have been analysed in the laboratory of the Chemistry Faculty of the Rovira i Virgili University (URV) in the Analytical and Organic Department (QAiQO) using fluorescence spectroscopy as experimental technique. The classification methods have been Soft Independent Modelling of Class Analogy (SIMCA) and Partial Least Squares – Discriminant Analysis (model PLS-DA). To obtain some quality semi-quantitative parameters, specifically CCα, CCβ and the unreliability region, performance characteristic curves (curve PCC) have been used. To the first models, the validation was done with a training and a test set. To the second models, another group of samples was used. Also some quality parameters were calculated such as sensibility and specificity for the purpose of validating.