Comparison of two data processing approaches for aroma marker identification in different distilled liquors using comprehensive two-dimensional gas chromatography-time-of-flight mass spectrometry dataset.

Abstract

Whisky, brandy, rum, and Chinese baijiu are popular distilled spirits globally, and the volatile components play an essential role in the aroma of these distilled liquors. Volatile compounds in whisky, brandy, rum, and three main aroma types of Chinese baijiu (strong, light, and sauce) were investigated using comprehensive two-dimensional gas chromatography-time-of-flight mass spectrometry (GC×GC-TOFMS). Two significant variable detection approaches, the variable importance in the projection (VIP) and nonparametric test (Mann-Whitney U test), were compared to determine the volatile markers among these samples. It was found that the VIP model was more efficient in screening significant variants than the U test. A total of 117 common markers were selected by both the VIP and U test methods with potential aroma contributions. Esters and acids were the main aroma markers for baijiu, diethyl esters were the main aroma markers for brandy, whereras pyrazines, lactones, and furans were the main aroma markers for whisky. Based on the chosen markers, different unknown distilled liquors were successfully classified in the model validation. This study provided a feasible methodology for spirit sample speculation based on volatile composition obtained by GC×GC-TOFMS.