Métodos multivariables cualimétricos para el aseguramiento y control de la calidad en el ámbito de la soberanía y calidad alimentaria

Resumo

The current Doctoral Thesis develops and applies diverse multivariate analytical methods based on the liquid chromatography and spectroscopic vibrational analytical techniques, together with chemometric tools, which are focused on the assurance and quality food control, specifically, in the quality authentication of virgin and extra virgin olive oil, and White Tequila. The paramount objective of this Doctoral Thesis is the development of different multivariate analytical methods, that are fast, reliable, efficient and useful for the food industry to improve and increase the food fraud detection. These methods imply the use of the fingerprinting methodology, particularly, spectroscopic and chromatographic fingerprints, and the use of the 'instrument-agnostizing' methodology, using chemometric tools to mine the useful analytical information from them. Initially, the utility of the instrument-agnostizing methodology is explored on chromatographic fingerprints of olive oil, obtained through HPLC-UV/Vis, to build mathematic models with PLS-DA and SIMCA capable to detect adulterations in such oils. Afterwards, the intrinsic potential of the instrument-agnostizing methodology to transfer multivariate analytical signals among laboratories is proved, using 'agnostic' chromatographic fingerprints of White Tequila, obtained through HPLC-DAD in different laboratories. From such signals, a global database and a single mathematic model were developed in order to differentiate among the '100% agave' and 'mixed' categories of White Tequila samples, analyzed in Spain and México. Thereupon, the FTIR, NIR and SORS spectroscopic techniques, and certain chemometric tools were applied together to develop different multivariate analytical methods capable to authenticate the quality of the categories of White Tequila samples using pattern recognition techniques, such as PLS-DA, SVM, SIMCA, kNN, as well as to quantify their alcoholic content using the PLSR and SVMR multivariate calibration techniques. Finally, the application of these multivariate analytical methods is suggested along the food chain of each of the olive oil and tequila industries.