Non-volatile metabolomic differentiation of geographical origin and quality grade in pre-Qingming Xihu Longjing green tea using in-source fragment ion-assisted annotation

Xihu Longjing green tea (XHLJT) faces challenges in origin and grade authentication due to mislabeling and limited chemical interpretability of sensory evaluation. Here, an enhanced untargeted metabolomics workflow integrating in-source fragment ion (ISFI) identification was developed to improve annotation reliability and investigate the combined effects of geographical origin and grade on non-volatile metabolites. UHPLC–HRMS profiling of samples from four core production areas across three grades revealed three universal origin markers (theobromine, luteolin-4′-O-glucoside, and tyrosine), while additional origin-discriminative metabolites were mainly amino acids, flavonoids, and alkaloids. Tyrosine was the only universal grade marker, and grade differentiation was largely associated with umami-related amino acids. Geographical origin induced metabolic reprogramming, whereas grade variation primarily reflected quality-related amino acid accumulation. Cross-year external prediction across different years further supported good transferability of the developed models for geographical origin and grade discrimination. These findings provide a chemically interpretable basis for integrated origin–grade authentication of pre-Qingming XHLJT.