作 者 ：侯力璇，张京芳*，徐洪宇
期 刊 ：西北植物学报 2012年 9期 页码：1819~1828
关键词：酿酒红葡萄；HS-SPME-GC-MS；香气成分；聚类分析；判别分析；
Keywords:red wine grape, HS-SPME-GC-MS, aroma component, cluster analysis, discriminante analysis,
摘 要 ：利用顶空固相微萃取与气相色谱质谱联用(HS-SPME-GC-MS)技术测定了26种酿酒红葡萄的香气成分以及部分重要组分的含量，并采用聚类分析与主成分分析法，探讨利用香气成分含量对酿酒红葡萄品种的分类。结果显示：（1）在26种酿酒红葡萄品种中共检测到49种香气成分并分为6大类，分别为C6-醇和醛类化合物、醇类化合物（除C6-醇类）、酯类化合物、醛类（除C6-醛类）和酮类化合物、萜烯类化合物、有机酸类化合物。（2）聚类分析结果显示，26种酿酒红葡萄分成了三大类，其中：第Ⅰ类有赤霞珠等11个品种，香气成分以C6-醇和醛类化合物为主，其总含量高达66.4%；第Ⅱ类有宝石解百纳等6个品种，香气成分以C6-醇和醛类化合物以及醇类化合物（除C6-醇类）为主，分别为40.73%和30.21%；第Ⅲ类有梅鹿辄等9个品种，香气成分以酯类化合物为主，占总香气成分的71.06%；而且三类酿酒红葡萄中有机酸含量均最低，分别为0.18%、0.08%和0.005%。（3）主成分分析结果表明，第Ⅰ类酿酒红葡萄中可以用前6个主成分代表原有的变量信息，第Ⅱ类和第Ⅲ类均可以用前4个主成分代表原有的变量信息。（4）利用判别分析创建了酿酒红葡萄香气成分判别模型，并对其进行验证，其准确率达85.7%。研究表明，利用香气成分建立的判别模型可以对酿酒红葡萄进行分类，可为酿酒红葡萄品种改良及葡萄酒质量控制提供理论依据。
Abstract:The aroma components in 26 different kinds of wine grape and the content of the important component part were identified by the headspace-solid phase microextraction gas chromatography-mass spectrometry.This paper discussed the classification of wine grape according to the contents of aroma compounds with the cluster analysis and principal component analysis.The results showed that:（1）49 kinds of aroma compositions were detected in 26 different kinds of wine grapes and were divided into 6 types,C6-alcohols and aldehydes,alcohols (except C6-alcohols),esters,aldehydes (except the C6-aldehydes) and ketones,terpenes,organic acids.（2）The results of clustering analysis showed that 26 different kinds of wine grape were classified into three categories.Among them,the first kind of wine grapes included 11 varieties,such as cabernet sauvignon.C6-alcohols and aldehydes compounds were major properties in the first category,and its total content up to 66.4%.The second kind of wine grape includes 6 varieties,such as ruby cabernet.The contents of C6-alcohols and aldehydes and alcohols (except C6-alcohols) reached to 40.73% and 30.21%,respectively.The third kind of wine grape included 9 varieties,such as merlot.The content of esters was the supreme aroma components in the third class and it account for 71.06% of the total aroma components.And the organic acids was the lowest part in the three categories (0.18%,0.08% and 0.005% respectively).（3）The results of principal component analysis showed that the first six main components could represent its original information of variables in the first class of wine grape and the first four main components could represent their original information of variables in both the second and third classes of grape.（4）Based on the known aroma compounds,a discrimination model was established for assessing and ranking wine grape by discriminante analysis,and the accuracy of the model was 85.7%.This research showed that the discrimination model could be set up to classify wine grape and provided the theory basis for wine grapes variety improvement and wine quality control.