Authentication of the production season of Xinyang Maojian green tea using two-dimensional fingerprints coupled with chemometric multivariate calibration and pattern recognition analysis

Abstract

This study aimed to authenticate the production season of Xinyang Maojian green tea, and screen and identify its characteristic components through non-targeted metabonomics methods based on two-dimensional fingerprints combined with chemometric analysis. Firstly, two-dimensional fingerprints of spring and autumn teas were obtained through HPLC-DAD analysis to form a three-dimensional array (retention time × absorption wavelength × sample). Subsequently, the multiple co-elution peaks and spectral profiles in two-dimensional HPLC-DAD fingerprints were resolved by using alternating trilinear decomposition assisted multivariate curve resolution (ATLD-MCR) algorithm. We obtained the relative concentration matrix C (24 × 122), which was further used to distinguish the production season of Xinyang Maojian green tea through chemometric pattern recognition analysis. The evaluation results of both orthogonal partial least squares-discriminant analysis (OPLS-DA) and partial least squares-discriminant analysis (PLS-DA) models were better than those of PCA models, and could effectively distinguish the production season of Xinyang Maojian green teas. Moreover, 5 variables were selected through VIP method to build new UV-scaling and Par-scaling OPLS-DA models. In conclusion, the following characteristic components were identified in accordance with the analytical standards and published data: gallocatechin (GC), theobromine (THB), epigallocatechin gallate (EGCG), gallocatechin gallate (GCG), and epicatechin gallate (ECG).