气相色谱-离子迁移谱法结合多元统计学分析不同陈酿时间白兰地的挥发性香气成分差异

Analysis of volatile aroma component differences in brandy with different aging times by gas chromatography-ion mobility spectrometry combined with multivariate statistics

中文关键词:  气相色谱-离子迁移谱法  白兰地  陈酿  挥发性香气成分  主成分分析  偏最小二乘判别分析

英文关键词:gas chromatography-ion mobility spectrometry  aging brandy  aging in oak barrels  volatile aroma component  principal component analysis  partial least squares-discriminant analysis

中文摘要:

      目的 分析不同年份白兰地的挥发性香气成分差异, 建立陈酿白兰地年份酒快速判别预测模型。方法 基于气相色谱-离子迁移谱法(gas chromatography-ion mobility spectrometry, GC-IMS), 利用指纹图谱判别各年份白兰地挥发性香气成分差异和特征性标记物, 结合化学计量学方法对各物质吸收峰体积进行数据降维分析和可视化操作, 建立年份判别模型。结果 共定性出50种挥发性香气成分, 指纹图谱显示, 不同年份白兰地中挥发性香气成分含量及种类存在显著差异。主成分分析(principal component analysis, PCA)结果显示, PC1和PC2累计贡献率可达85%, 表明不同年份白兰地样品具有良好的聚类效果。通过偏最小二乘回归分析(partial least squares-discriminant analysis, PLS-DA)建立年份酒判别模型, 得到不同年份酒样的特征性标记物, 筛选出22种变量投影重要性(variable important for the projection, VIP)>1的对陈酿白兰地年份判别贡献较大的挥发性香气成分。结论 利用GC-IMS结合多元统计学可快速判别年份白兰地的特征性标记物, 并且建模效果良好, 可为白兰地的年份鉴别和质量控制提供经验借鉴和应用思路。

英文摘要:

      Objective To identify the volatile aroma component differences of brandy with different aging times, and establish a rapid discriminant and prediction model for aging brandy. Methods Based on gas chromatography-ion mobility spectrometry (GC-IMS), fingerprint was used to identify the differences of volatile aroma components and characteristic markers in aging brandy, Combined with the chemometrics methods, dimensionality reduction analysis and visualization of the absorption peak volume of each substance were performed, and the prediction model of aging brandy was established. Results A total of 50 kinds of volatile aroma components were identified by GC-IMS. The fingerprint showed that there were significant differences in the content and types of volatile aroma components in brandy with different aging time. The cumulative contribution rate of PC1 and PC2 of principal component analysis (PCA) was up to 85%, indicating that different aging brandy had good clustering effect. The discriminant model of aging brandy was established by partial least squares-discriminant analysis (PLS-DA), and the characteristic makers of each aging brandy were obtained, 22 kinds of volatile aroma components with variable important for the projection (VIP)>1 were screened for their significant contribution to the identification of aging brandy. Conclusion GC-IMS combined with multivariate statistics can be used to identify the characteristic makers in brandy with different aging times and the model is effective, which provides references and application ideas for the identification and quality control of aging brandy.