典型品牌料酒风味物质的气相色谱/质谱指纹图谱构建及主成分分析和聚类分析

GC/MS Fingerprint Analysis of Flavor Substances and Its Principle Component Analysis,Cluster Analysis with Typical Brand Cooking Wine

本文采用气相色谱-质谱联用技术(Gas Chromatography-Mass Spectrometer,GC-MS)检测料酒中的风味物质,利用中药色谱指纹图谱相似度评价系统提取10种典型品牌料酒共有峰的信息,建立料酒风味物质的标准指纹图谱,采用夹角余弦法和相关系数法计算10种料酒的相似度,利用标准指纹图谱的特征峰检测4种未知料酒样品并推断是否归属该典型品牌料酒。将14种料酒的所有色谱峰依据主成分分析原理降维处理提取有效成分和进行聚类分析。结果表明,建立的标准指纹图谱能够区分典型品牌料酒与其他品牌料酒。主成分分析和聚类分析可以较好的评判料酒的品质并分类,且指纹图谱特征峰、主成分分析所有峰和聚类分析所有峰对料酒品牌的鉴定结果一致。

Gas chromatography/mass spectrometer(GC/MS)was used to detect the flavor substances in the cooking wine.Using the similarity evaluation system of chromatographic fingerprints of Chinese materia medica(Version 2004 A)to extract the common peaks information of 10 kinds of cooking wine of typical brand,the standard fingerprints of flavoring substances of cooking wine were established.The similarity of 10 kinds of cooking wine was calculated by cosine similarity and correlation coefficient.The characteristic peaks of the standard fingerprint spectrum were used to detect 4 kinds of unknown cooking wine.According to the principle of principal component analysis(PCA),the effective components were extracted from all chromatographic peaks of 14 kinds of cooking wine by dimension reduction treatment.Cluster analysis was carried out on all chromatographic peaks of 14 kinds of cooking wine.The results showed that the establishment of standard fingerprint could distinguish typical brand cooking wine from other ones and PCA and cluster analysis could better judge the quality of cooking wine and classify it.