Using comprehensive two-dimensional liquid chromatography to unveil and mitigate the bidirectional quantification biases of non-volatile compounds in Chinese black tea processing.

To better understand Chinese black tea processing, the quantification of non-volatiles is essential, yet conventional one-dimensional liquid chromatography (1D-LC) struggles with increasing matrix complexity as fresh tea leaves undergo enzymatic oxidation and polymerization. This study evaluated comprehensive two-dimensional liquid chromatography (LC × LC) against 1D-LC for quantifying catechins and theaflavins across fresh tea leaves and four processing stages (withering, rolling, oxidation, and drying). Non-volatiles profiling with 1D-LC, revealed extensive co-elution across the chromatographic space, with unidentified features accumulating particularly in late-retention regions. The primary quantitative finding was that 1D-LC exhibited a severe bidirectional bias due to co-elution: catechins were overestimated (up to 4-fold), while theaflavins were underestimated (64%). Crucially, LC × LC mitigated these quantification errors by resolving theaflavins (TF3G and TF3'G) and minimizing matrix effects (ME), narrowing the ME range from 0.60 to 2.20 (1D-LC) to 0.75-1.25 (LC × LC). These results demonstrate LC × LC as a robust method for better quantification in complex food matrices.