Abstract
Understanding the changes in tea aroma and non-volatile substances during roasting is essential for optimizing tea processing and enhancing tea quality. In this study, the Carbon Module Labeling (CAMOLA) technique was employed to simulate the roasting conditions of Qidan, thereby elucidating the formation pathway of the theanine-glucose Maillard system. Combined with sensory evaluation, the results indicated that the floral and fruity aromas of Qidan tea decreased, while the woody, roasted, smoky, and herbal aromas increased with prolonged roasting time. Kinetic modeling demonstrated that higher temperatures favored the production of benzaldehyde, which was directly proportional to the heating temperature. In contrast, pyrazines exhibited zero-order kinetics, influenced by both temperature and time. An increasing trend in furans was observed with rising temperature and extended heating time. The kinetic equations effectively describe the changes in aroma compounds associated with merad, highlighting the differences in the production patterns of aroma compounds under varying roasting conditions. This study provides a theoretical foundation for optimizing roasting parameters to enhance tea quality.