Evaluation of key taste components in Huangjin green tea based on electronic tongue technology

Abstract

A random forest algorithm combined with correlation analysis, was employed to identify key taste compounds in Huangjin green tea by using an electronic tongue. A total of 45 commercial Huangjin green teas were analysed for their amino acids, catechins, gallic acid, and caffeine using an amino acid analyser and HPLC. In this study, taste compounds of 30 were quantified, and 16 of these compounds exhibited taste activity values greater than 1 in the tea samples, including 6 amino acids, 8 catechins, as well as gallic acid and caffeine. Among these compounds, 5 compounds showed relatively high importance and strong correlations (P < 0.05) with the response of electronic tongue; specific for theanine, glutamic acid, lysine, gallocatechin, and catechin. A taste reconstruction experiment further confirmed that these 5 compounds significantly contributed to the overall taste profile, particularly enhancing the umami flavour.

Keywords: Amino acids; E-tongue; Green tea; Random Forest; Taste reconstruction; Umami taste.