Excitation-emission matrix fluorescence spectroscopy coupled with chemometric methods for characterization and authentication of Anhua brick tea

The present study sought to characterize and authenticate Anhua brick tea by coupling the excitation-emission matrix fluorescence spectroscopy with chemometric methods. A total of 230 samples belonging to 23 brands were collected. Four primary fluorophores were identified in the brick tea for the first time, including humic and fulic-like, heme-like, tryptophan-like substances, and theabrownin through the parallel factor analysis (PARAFAC) of the raw and theabrownin-spiked tea infusion data. The different types and brands of brick tea were processed respectively based on the PARAFAC factor scores and fluorescent data using four classifiers, including linear discriminate analysis, k-Nearest neighbors, partial least square discriminate analysis, and N-way partial least square discriminate analysis (NPLS-DA). The full fluorescent data combined with NPLS-DA showed the best classification results (96.5% for tea types and 98.3% for brands). In summary, the proposed methodology could be a potential tool for brand recognition of Anhua brick tea.