关键词: 酱香型白酒；智能化酿造；传统手工酿造；轮次与醅层；风味物质

Abstract: Liquid-liquid microextraction-gas chromatography-mass spectrometry, direct injection-gas chromatography-flame ionization detection, and multivariate statistical methods were employed to systematically compare the difference in flavor between intelligently and traditionally brewed Jiangxiangxing baijiu base liquors. A total of 42 base liquor samples from the upper, middle, and lower layers of fermented grains were collected from fermentation rounds 1–7 in two types of workshops within the same enterprise. The results indicated that base liquors from the intelligent workshops were generally rich in aldehydes, ketones, acids, pyrazines, furans, and long-chain fatty acid ethyl esters, while those from the traditional manual workshops prominently contained medium- and short-chain esters and alcohols. Based on odor activity value > 1 and variable importance in the projection (VIP) > 1, 17 key differential compounds (such as 2,5-dimethylpyrazine, 2-methylbutanal, and DL-leucine ethyl ester) were identified, with significantly higher concentrations in the samples from the intelligent workshops. Evolutionary trends across fermentation rounds and fermented grain layers were consistent between the two types of base liquors, but some substances in the base liquors from the intelligent workshops peaked in earlier fermentation rounds and exhibited lower coefficient of variation among fermented grain layers, indicating a more uniform distribution. This study provides a theoretical basis for quality evaluation and process optimization in the intelligent production of jiangxiangxing baijiu, offering guidance for the intelligent transformation and upgrading of the industry.

Key words: jiangxiangxing baijiu; intelligent brewing; traditional manual brewing; fermentation rounds and fermented grain layers; flavor compounds