Year-To Differentiation of Black Tea Through Spectroscopic and Chemometric Analysis

Abstract

The compositions of food products such as tea can vary significantly from one harvest year to another, primarily due to factors such as shifting climatic conditions, and plant periodicity. These fluctuations in composition can significantly affect the overall product quality. Spectral methods combined with chemometric techniques can provide efficient tools to monitor and assess these variations. In this study, 205 black tea samples from two consecutive harvest years were analyzed using mid-infrared, UV-visible, and fluorescence spectroscopy. Mid-infrared spectra were collected for both infused and powdered samples, while only the infused samples were used for the other spectroscopic methods. The study used partial least-square discriminant (PLS-DA) and orthogonal partial least-square discriminant analyses (OPLS-DA) to differentiate samples by harvest year. These models, applied after various data transformations, achieved high correct classification rates. Mid-infrared spectroscopic data yielded rates of 93.33% and 90.33% for powdered and infused samples, respectively. Fluorescence and UV-visible spectra also showed excellent prediction accuracy, with success rates of 98.3% and 100%. The results indicate that these spectroscopic methods, combined with chemometric differentiation, are valuable tools for monitoring year-to-year changes in black tea.