建立了基于离子迁移谱( IMS)技术的葡萄籽油掺伪鉴别新方法。经优化,选取进样口温度170℃,迁移管温度60℃。正己烷50倍稀释油样后进样检测,分析时间20 s。为了建立高效的葡萄籽油掺伪鉴别模型,本研究采用递归支持向量机( R-SVM)方法对葡萄籽油和掺伪葡萄籽油的IMS谱图进行分类,建立葡萄籽油和掺伪葡萄籽油分类判别模型,采用十折交互检验对建立的模型进行评价,结果显示模型判别正确率为91.2%。本方法操作简单,分析速度快,为食用油真伪鉴别提供了一种新方法。
A rapid method for authenticity detection of grapeseed oil was proposed based on Ion Mobility Spectrometry ( IMS) . After optimization, the inlet temperature was set at 170 ℃, drift tube temperature was set at 60 ℃. In this method, the oil sample was diluted in hexane (50-fold, V/V) and then directly analyzed in IMS. The detection time was 20 s. To establish an adulteration detection model, recursive Support Vector Machine ( R-SVM) was applied to classifying pure and adulterated grape seed oils. The result of 10-fold cross validation showed that the accuracy of discrimination was up to 91 . 2%. The results in study indicate that IMS method is a new, fast and convenient technique for the adulteration detection of edible oil.