An Integrated Framework for Green Tea Infusion Chemical Characterization: Process Optimization, PCA-Based Chemometric Evaluation, and Untargeted Metabolomic Profiling

This study investigated the optimization and quality evaluation of green tea infusions prepared from three commercial green tea samples from different production origins, namely Anhui Huangshan Maofeng (AH), Shaanxi Ziyang (SX), and Yunnan Xuelong (YN), using three extraction techniques: hot-water extraction (HWE), enzyme-assisted extraction (EAE), and cold brewing (CB). We used water-extractable content (WEC) as the sole response variable in a response surface methodology framework to optimize physical solid yield rather than chemical quality. We developed a principal component analysis (PCA) model that incorporates WEC with six essential functional constituents—epicatechin (EC), epigallocatechin (EGC), epicatechin gallate (ECG), epigallocatechin gallate (EGCG), caffeine, and L-theanine, to thoroughly evaluate infusion quality. Untargeted metabolomics was subsequently performed on three PCA-selected representative SX samples corresponding to PCA ranks 18, 58, and 107. Our findings demonstrated that WEC was significantly influenced by the investigated extraction variables, and validation of the BBD-RSM models showed absolute relative deviations of 0.12–0.81% between the predicted and experimentally measured WEC values. Six targeted components differed significantly among the three tea samples and extraction methods. The PCA score plots showed partial separation among the sample–process combinations, with the first two principal components explaining 78.45% of the total variance. Additionally, the three SX infusion groups were clearly distinguished by orthogonal partial least-squares discriminant analysis (OPLS-DA) in conjunction with KEGG pathway mapping; KEGG pathway mapping shows that the differentially enriched metabolites are primarily annotated to flavonoid, phenolic acid, amino acid derivative, and purine-related pathways. By combining RSM-based optimization, targeted HPLC measurement, PCA scoring, and non-targeted metabolomic profiling, this study created a strong framework for chemical evaluation. This integrated framework offers a valuable reference for selecting appropriate extraction processes and ensuring consistent quality in green tea infusions.