GC-IMS技术结合化学计量学方法在食用植物油分类中的应用

建立了一种快速、无损分析食用植物油中挥发性有机物质的顶空进样/气相色谱-离子迁移谱(GC-IMS)联用方法。以芝麻油、菜籽油、山茶油共56个样品为研究对象,量取2 mL待测油样于标准样品瓶中,并用磁帽密封,直接进行GC-IMS分析检测。结果表明,基于GC-IMS三维谱中对应挥发性有机物质的特征峰强度可以有效表征不同类植物油的样品信息,选取对应三维谱中40个特征峰的强度作为变量,进行主成分(PCA)信息降维后,采用k最近邻(kNN)算法建立植物油种类的判别模型,训练集的识别率达到100%,预测集中仅有1个山茶油样品被误判成芝麻油样品,预测集的识别率达到94.44%。GC-IMS联用分析技术简单、快速、无损,可用于食用植物油等其他食品、农产品种类的快速分类识别。

Application of Headspace/Gas Chromatography-Tandem Ion Mobility Spectrometry with Chemometrics in Classification of Edible Vegetable Oils

A rapid and nondestructive method was established for the analysis of the volatile organic compounds in edible vegetable oils by headspace/gas chromatography-tandem ion mobility spectrometry(GC-IMS).56 sesame oil,rapeseed oil and camellia oil samples were taken as the research objects.2 mL tested sample was put into the standard sample bottle which was sealed with a magnetic cap for direct GC-IMS analysis.The results showed that the characteristic peaks in GC-IMS 3D spectra corresponding to volatile organic substances can effectively characterize the sample information areas of different types of vegetable oils.With the sectional intensities of 40 characteristic peaks in corresponding three-dimensional spectra as variables,the principal components analysis(PCA) algorithm was used to reduce information dimensionality,establishing the training model for discriminating the type of vegetable oil by kNN algorithm,and the recognition rate was up to 100%,only 1 camellia oil sample was judged as sesame oil in the predicted set,and the prediction set recognition rate was 94.44%.GC-IMS technique is simple,rapid and nondestructive,and could be used for the rapid identification of agricultural products and other foods such as edible vegetable oils.