Classification of the alcoholic beverages by using electronic nose technique based on ultra-fast GC

Abstract

Each alcoholic beverage has its own distinct aroma and unique flavor. An Electronic nose (E-nose) is an instrumental tool which allows one to analyze the aroma of samples. This work illustrates the use of E-nose for the determination of characteristic aromas of seven different alcoholic beverages (bourbon, brandy, cognac, vodka, liqueurs made from pears, and fruit spirits made from lemon). The following chemometrics methods for the data analysis were employed for the classification of alcoholic beverages: principal component analysis (PCA), discriminant factorial analysis (DFA, and soft independent modelling of class analogies (SIMCA). The results show that chemometrics methods PCA and DFA allow us the determination of the individual alcoholic groups of fruit spirits, bourbon, cognac, and brandy but there is a difficulty with separation of liqueur, gin, and vodka points from one group. Only the SIMCA method allowed to distinguish the point belonging to all groups without any approximation of graph fragments.