Stepwise Identification of Six Tea (Camellia sinensis (L.)) Categories Based on Catechins, Caffeine, and Theanine Contents Combined with Fisher Discriminant Analysis

Abstract

Six tea types were stepwise discriminated based on their catechins, caffeine, and theanine contents of total 436 tea samples collected worldwide, combined with Fisher classification pattern recognition. Those tea samples of six types (green, white, yellow, oolong, black, and dark teas) of commonly consumed teas with different processing methods were analyzed in this work. Five main catechins ((−)-epigallocatechin gallate (EGCG), (−)-epigallocatechin (EGC), (−)-epicatechin gallate (ECG), (−)-epicatechin (EC), and (+)-catechin (C)), caffeine, and theanine contents were accurately measured by high-performance liquid chromatography (HPLC). As a novel approach, stepwise identification combined with Fisher discriminant analysis was applied to develop an identification model. Several parameters, including model component factors, were optimized by cross-validation. The optimal Fisher model was achieved with caffeine, total catechins, theanine, theanine × theanine, EGCG/total catechins, and theanine × caffeine as component factors. The discrimination rates of black, dark, white, oolong, yellow, and green teas were 95.90, 100.00, 97.40, 95.70, 91.80, and 88.30 %, respectively. Compared with other pattern recognition approaches, the Fisher algorithm exhibited excellent performance in the final identification. The overall results show that this method is suitable to stepwise identify six tea categories, according to the measurements of main chemicals with catechins, caffeine, and theanine by HPLC and followed by the Fisher pattern recognition.