SPME-GC-MS/MS结合多元统计分析研究不同储藏年份玉米风味物质差异

风味物质是玉米食用品质和营养价值的重要衡量指标，玉米储藏过程中的品质劣变与其风味物质含量密切相关。采用固相微萃取-气相色谱-三重四极杆质谱联用技术对2019—2022年共4个年份收获的玉米样品进行检测，并采用主成分分析和偏最小二乘法判别分析法对风味物质相对含量进行多元统计分析。结果表明，玉米中共检出挥发性风味物质129种，包括醇类、醛类、酮类、酯类、苯系物、杂环化合物、酸类、醚类、烃类和酚类物质。固相微萃取的最优条件为样品用量6 g、萃取温度70 ℃、萃取时间60 min。多元统计分析结果显示，玉米挥发性风味物质与储藏年限存在一定的相关性。根据变量投影重要性，有47种化合物在不同年份玉米间具有显著性差异。其中1-辛烯-3-醇、丁酸橙花酯和2-正戊基呋喃相对含量随玉米储藏年份延长而增加，DL-泛酰内酯、辛酸甲酯和2-乙酰基呋喃相对含量随玉米储藏年份延长而减少，这6种化合物被鉴定为特征差异化合物。由此构建偏最小二乘判别分析分类模型，该模型对玉米储藏年份具有良好的区分性（R2=0.99，Q2=0.97，预测精确度为1），可为玉米储藏品质的动态监测提供技术手段，对保障国家粮食安全和节粮减损具有重要意义。

Diversity Analysis of Volatile Flavor Compounds of Corn with Various Storage Years Based on SPME-GC-MS/MS and Multivariate Statistical Analysis

Flavor compounds are important indicators of edible quality and nutritional value of corn. The quality deterioration of corn during storage is closely related to its flavor compounds. Corn samples harvested in four years from 2019 to 2022 were detected by solid phase microextraction-gas chromatography-triple quadrupole mass spectrometry (SPME-GC-MS/MS), and multivariate statistical analysis of flavor compounds relative content was carried out by principal component analysis and partial least squares discriminant analysis. The results showed that 129 volatile flavor compounds were detected in corn, including alcohols, aldehydes, ketones, esters, benzenes, heterocycles, acids, ethers, hydrocarbons and phenols. The multivariate statistical analysis showed that volatile flavor compounds and storage years of corn were correlated. According to the importance of variable projection, 47 compounds had significant diversity between corn samples with various storage years. Among them, the relative contents of 1-octen-3-ol, neryl butyrate and 2-pentylfuran increased with the extension of corn storage years, while the relative content of pantolactone, methyl octanoate and 2-acetylfuran decreased with the extension of corn storage years. These 6 compounds were identified as characteristic differential compounds. The partial least squares discriminant analysis classification model was constructed, and had good discrimination for corn storage years (R2=0.99, Q2=0.97, prediction accuracy was 1). The model will provide technical means for dynamic monitoring of corn storage quality, which is of great significance for ensurance national food security and reduce grain loss.