Darjeeling tea profiling using 1H NMR and multivariate analysis

Abstract

The current study emphasizes the metabolic fingerprints of tea’s considerable dependency on the plucking season. The taste, color, and aroma represent the quality of tea in general, and chemical constituent concentration levels in tea are greatly impacted by seasonal changes, such as temperature, humidity, precipitation, sun exposure and rainfall. 1H NMR spectral datasets of tea, in conjunction with multivariate techniques, can be used to distinguish differences in tea metabolic signatures influenced by seasonal changes. In the present study, the multivariate techniques PCA (Principal Component Analysis) and LDA (Linear Discriminant Analysis) are coupled to classify 'India’s premium Darjeeling tea based on plucking seasons. A total of 164 NMR spectra of Darjeeling tea samples from the first, second and autumn flushes are taken into consideration. The PCA followed by LDA has demonstrated a clear classification by minimizing variability within the samples of clusters as well as maximizing variability between the clusters with a good F1 score (0.95), indicating that there are significant differences in the metabolic fingerprints between the tea flushes and developed classification module helps in authenticating Darjeeling tea geographical origin.