Post-harvest processing based on the steaming process significantly affects the quality of medicinal and edible Gastrodia elata Blume (G. elata). This study examined the effects of different processing methods on the bioactive components of G. elata. It proposed a process for identifying and evaluating the quality of various preparation methods using infrared spectroscopy combined with chemometrics. The results indicate that different processing methods led to an upward trend in parishin (A, B, C, E) and gastrodin, while 4-hydroxybenzyl alcohol showed a downward trend. The stir-frying (LP) preparation group exhibited the highest content of total characteristic components, reaching 11.93 mg/g. Furthermore, unlike previous studies that focused solely on quantitative composition analysis, we have been the first to employ two-dimensional correlation spectroscopy (2DCOS) to decipher the sequential changes in molecular dynamics of G. elata under different processing methods. This approach successfully visualized the continuous response of functional groups (such as O-H and C-H) to thermal agitation, providing new mechanistic insights into how processing alters product quality. It is noteworthy that the NIR-CARS-2DCOS-ResNet model achieved 100 % accurate identification across six processing methods, while also demonstrating improved efficiency compared to the full-spectrum ResNet model. For quantitative analysis, partial least squares regression (PLSR) models established between spectral data and HPLC reference values successfully predicted the content of individual characteristic components (except for parishin E). The residual prediction deviation (RPD) values for all optimal models under both strategies exceeded 2.00, indicating robust predictive performance, with the NIR strategy outperforming the FTIR strategy. This study provides a rapid, non-destructive, and effective strategy for both the authentication of processing methods and the quantitative prediction of key active constituents in G. elata, providing a powerful tool for its post-harvest quality control.
扫码关注我们
求助内容:
应助结果提醒方式:
