Characterization of the key differential aroma compounds in five dark teas from different geographical regions integrating GC–MS, ROAV and chemometrics approaches

Dark tea (DT) holds a rich cultural history in China and has gained sizeable consumers due to its unique flavor and potential health benefits. In this study, headspace solid-phase microextraction (HS-SPME) combined with gas chromatography-mass spectrometry (GC-MS), relative odor activity value (ROAV), and chemometrics approaches were used to detect and analyze aroma compounds differences among five dark teas from different geographical regions. The results revealed that the five DTs from different geographical regions differed in types, quantities, and relative concentrations of volatile compounds. A total of 1372 volatile compounds of were identified in the 56 DT samples by HS-SPME-GC-MS. Using ROAV and chemometrics approaches, based on ROAV>1 and VIP>1. Eighteen key aroma compounds can be used as potential indicators for DT classification, including dihydroactinidiolide, linalool, 1,2,3-trimethoxybenzene, geranyl acetone, 1,2,4-trimethoxybenzene, cedrol, 3,7-dimethyl-1,5,7-octatrien-3-ol, β-ionone, 4-ethyl-1,2-dimethoxybenzene, methyl salicylate, α-ionone, geraniol, linalool oxide I, linalool oxide II, 6-methyl-5-hepten-2-one, α-terpineol, 1,2,3-trimethoxy-5-methylbenzene, and 1,2-dimethoxybenzene. These compounds provide a certain theoretical basis for distinguishing the differences in five DTs from different geographical regions. This study provides a potential method for identifying the volatile substances in DTs and elucidating the differences in key aroma compounds. Abbreviations: DT, dark tea; FZT, Fuzhuan tea; LPT, Guangxi Liupao tea; QZT, Hubei Qingzhuan tea; TBT, Sichuan Tibetan tea; PET, Yunnan Pu-erh tea; ROAV, Relative odor activity value; OT, Odor threshold; HS-SPME, Headspace solid-phase microextraction; GC-MS, Gas chromatography-mass spectrometry; PCA, Principal components analysis; PLS-DA, Partial least squares-discriminant analysis; HCA, Hierarchical clustering analysis.