UV-Vis spectral fingerprinting and chemometric method applied to the evaluation of Camellia sinensis leaves from different harvests

Abstract

UV-Vis spectral fingerprinting was used to discriminate Camellia sinensis leaves of two different harvests and multivariate data analysis was applied to determine the relevant metabolites for separation. First statistical mixture designs of pure ethanol, ethyl acetate, dichloromethane and chloroform solvents as well as their binary, ternary and quaternary mixtures extracted larger varieties and amounts of C. sinensis leaf metabolites than would be obtained from classical solvent extractions. UV-Vis spectral fingerprints of crude extracts were subjected to Orthogonal Signal Correction and Partial Least Squares-Discrimination Analysis (OSC-PLS-DA) for classification. The spectra were all correctly identified and classified, showing that the OSC-PLS-DA model possesses a good predictive ability to separate spectral fingerprints of different harvests. VIP score values showed that bands at 272, 410 and 663 nm were responsible for separation. These metabolites were identified by HPLC-DAD as caffeine and pheophytin a. According to the mixture model, the maximum values of relative abundances of both caffeine and pheophytin a can be extracted with pure dichloromethane.