Characterize and explore the dynamic changes in the volatility profiles of sauce-flavor baijiu during different rounds by GC-IMS, GC-MS and GC×GC-MS combined with machine learning.
Journal:
Food research international (Ottawa, Ont.)
Published Date:
May 5, 2025
Abstract
The production process of sauce-flavor baijiu (SFB) involves seven distillations, yielding base baijiu of 7 rounds (RSFB), which are then blended to form the final product. Therefore, the quality of the base baijiu is closely related to the quality of the final product. Each of the seven rounds of RSFB has a distinct flavor profile. However, the volatiles and their specificities in base baijiu remain unclear. In this study, GC-MS, GC × GC-MS, and GC-IMS were comprehensively used to investigate the volatile compounds and their dynamic changes in RSFB. A total of 469 volatile compounds were identified, with 17 overlapping compounds. Using a Random Forest (RF) model, 35 key biomarkers were selected across the three methods. Key compounds such as ethyl isovalerate, ethyl butyrate, benzaldehyde, 2-ethyl-6-methylpyrazine, and phenylacetaldehyde played significant roles in flavor formation. The model's performance was validated using receiver operating characteristic (ROC) curves, with areas under the curve approaching 1. Additionally, the metabolic pathways of these volatiles are primarily involved in carbohydrate and amino acid metabolism. This study reveals the specificity of volatiles in RSFB and provides valuable insights for baijiu quality control.