Development and validaton of multivariate strategies for food quality control

Identifier:  TDX:4446

Handle: https://hdl.handle.net/20.500.11797/TDX4446

Authors:  Rovira Garrido, Glòria

Abstract:
This thesis aims to develop and validate qualitative multivariate methods for the detection of adulteration or authentication of foods, including the next issues: i) using different molecular spectroscopic instruments such as Near-Infrared (NIR), Attenuated Total Reflectance Fourier Transform Infrared (ATR-FTIR), Fluorescence, and Low-Field Nuclear Magnetic Resonance (LF-NMR); ii) defining an uncertainty region from the semi-quantitative information in adulteration cases; iii) optimizing the performance parameters obtained from the multivariate analysis, and iv) applying a transfer method technique to demonstrate the applicability of models created under different conditions.Three different adulteration problems are presented in this work; olive oil adulterated with sunflower oil, cashew nuts adulterated with other nuts (Brazilian nut, Pecan nut, Macadamia nut, and Peanut), and honey adulterated with syrups (Inverted sugar, corn and rice syrup). In each case of study, different strategies to extract semi-quantitative information have proved its usefulness, including the use of Performance Characteristic Curves (PCC) and Receiver Operating Characteristic (ROC) curves, and the proposed strategy of setting two class limits to define uncertainty regions.In the authentication problem of extra virgin olive oils from two Protected Denominations of Origin (PDO), a standardization approach is proposed to manage the variations between harvests. The usefulness of the standardization techniques has been demonstrated to maintain the performance of a multivariate classification model.The research carried out in this thesis significantly contributes to advancing the detection and prevention of food fraud, ensuring safer consumer products, and preserving market integrity. The developed methods improve over traditional protocols in speed, sustainability, and non-destructiveness.