RESUMO
In order to differentiate and characterize Chinese Luzhou-flavor liquor according to geographical origins, the volatile flavor compounds were analyzed for forty commercial Luzhou-flavor liquor samples from Sichuan, Jiangsu, and Hubei provinces. A total of 113 volatile flavor compounds were quantified; among them, 29 flavor compounds were quantified according to the internal standard method. The differences in flavor composition among different brands of Luzhou-flavor liquor were compared. A data matrix of 64 (flavor components) × 40 (samples) was studied and interpreted using chemometric analysis. The research object could be naturally clustered according to geographical origin (brand) based on the hierarchical cluster analysis (HCA), principal component analysis (PCA) and multivariate analysis of variance (MANOVA) methods. A 100% of predication ability was obtained by the application of K-nearest neighbor model (KNN) for study sample classification. The results demonstrate that the abundance of volatile flavor components in liquors combined with appropriate multivariate statistical methods could be used for the division and traceability of liquors from different geographic origins. PRACTICAL APPLICATION: This study can provide the basis for the identification of liquor authenticity and the traceability of liquor.