Caracterització, Classificació i Autenticació de Te mitjançant empremtes de FIA-MS

Treballs Finals de Grau de Química, Facultat de Química, Universitat de Barcelona, Any: 2022, Tutors: Oscar Núñez Burcio, Sònia Sentellas MinguillonNowadays, many food products have been subjected to some kind of fraudulent practices, including incorrect labelling, adulteration or substitution of undeclared compounds, among others. The main purpose of these practices is mainly to obtain illegal economic benefits, although there is a great concern about their increase due to the problems that the presence of undeclared allergenic or toxic compounds may entail for public health. Often, these frauds cannot be recognized visually or detected using simple methods, so the development of more advanced analytical techniques has become an urgent necessity. One of the techniques that is becoming increasingly important in this field is flow injection analysis coupled to mass spectrometry (FIA-MS) working in a non-targeted (fingerprinting) approach, which is characterized as a fast, simple, and effective technique for analyzing a large number of samples without the need to know the identity of their components. This work has focused on evaluating the use of FIA-MS fingerprints as chemical descriptors to study the characterization, classification, and authentication of different tea varieties, as well as the detection and quantitation of one of the most common adulterants in this beverage, chicory. The data obtained have been subjected to multivariant chemometric methods, such as principal component analysis (PCA), partial least squares-discriminant analysis (PLS-DA), and partial least squares regression (PLS). The results obtained have been very promising, demonstrating that the proposed method is able to discriminate perfectly between different types of tea and chicory, as well as to quantify different levels of adulteration in two adulterated tea varieties (black and green tea); in fact, low calibration and cross-validation errors have been obtained (0.7-5.8% and 6.7-8.5%, respectively), and quite acceptable errors regarding to prediction (7.8-16.4%) have been obtained, demonstrating the good ability of this method to address tea authenticatio