Non-targeted Fingerprinting Methodologies for the Authentication of Tea. Application to the Detection and Quantitation of Frauds in Adulterated Tea Samples with Chicory

Abstract

In recent years, in an increasingly globalised world the food industry like so many others has undergone a paradigm shift. This increasing globalisation has generated high competition, and some industries have turned to fraudulent practices in order to reduce production costs. Food frauds are very diverse, ranging from the substitution and/or addition of components to fraud in product labelling. This malpractice for economic purposes becomes a serious problem for consumers by putting their health at risk. To avoid this kind of negligence, it is very important to develop new fraud detection methodologies, especially in a world where these practices are becoming increasingly sophisticated. This project focuses on tea adulteration with chicory. For this purpose, two methodologies have been developed simultaneously for the authentication of different tea varieties (black, red, green, white and oolong). These are two non-targeted fingerprinting methods that use the chromatograms obtained by high performance liquid chromatography with ultraviolet (HPLC-UV) and fluorescence (HPLC-FLD) detection. The obtained data have been subjected to multivariate chemometric treatments, such as Principal Component Analysis (PCA) and Partial Least Squares regression (PLS). The good results obtained have shown that both methodologies are effective for the detection and quantification of tea adulteration with chicory, exhibiting good linear regressions (R2 > 0.998) and calibration, cross-validation, and external validation errors below 1.4%, 6.4% and 3.7%, respectively. Moreover, very acceptable prediction errors have been obtained (less than 21.7%), except for white tea extracts which show higher errors due to the similarity of their fingerprints to those of chicory. In addition to these two methods, two new studies have been started to further advance the authentication of teas. The first one is a non-targeted FIA-MS fast-screening method and the second one is a non-targeted method using the fingerprints obtained by liquid chromatography coupled to a mass spectrometer (LC-MS) as chemical descriptors. Data from both methodologies have also been chemometrically treated with PCA and Partial Least Squares-Discriminant Analysis (PLS-DA) and the results are very promising since both techniques are able to distinguish between chicory and tea sample extracts (although further work is needed to optimise them). The most interesting approach is the FIA-MS method, which represents a great advantage over previous methods, as it entails significant time saving in tea authentication analysis