【作者】 陈美丽； 张俊； 唐德松； 龚淑英； 张颖彬； 顾志雷；

【Author】 CHEN Mei-li;ZHANG Jun;TANG De-song;GONG Shu-ying;ZHANG Ying-bin;GU Zhi-lei;Tea Research Institute,Zhejiang University;

【机构】 浙江大学茶叶研究所；

【摘要】 选择有代表性的112个茉莉花茶茶样为实验材料,采用近红外光谱分析技术结合化学计量学方法对茶多酚总量(TP)、游离氨基酸总量(AA)、咖啡碱(CAFF)、水浸出物(water extract)、没食子酸(GA)、没食子儿茶素(GC)、表没食子儿茶素(EGC)、儿茶素(C)、表儿茶素(EC)、表没食子儿茶素没食子酸酯(EGCG)、没食子儿茶素没食子酸酯(GCG)、表儿茶素没食子酸酯(ECG)、儿茶素没食子酸酯(CG)这13种品质成分的含量建立定量分析模型,除CG的模型综合评价指标Q值为0.7702以外,其余均达0.8以上。经外部检验,CAFF和GCG的真实值与预测值的相关系数R为0.7以上,GC、EGCG、ECG、CG为0.8以上,其余均达0.9以上,模型效果较好,为简便快速地测定茉莉花茶品质成分的含量提供了新思路。更多还原

【Abstract】 In this paper,the quantitative calibration models for determining the contents of 13 chemical compositions such as total tea polyphenols(TP),free animo acids(AA),caffeine(CAFF),water extract,GA,GC,EGC,C,EC,EGCG,GCG, ECG,CG were established by NIRS combined with Chemometrics.Q-value taking all important statistical values into account can evaluate the quality of models.Results showed that almost all the Q-value of models were exceed 0.8,except for CG with a Q-value of 0.7702.Each of the 13 predictive models was applied for external inspection.Coefficients of correlation(R) between predicted value by NIR and the actual value were exceed 0.9,except CAFF and GCG were exceed 0.7 and GC、EGCG、ECG、CG were exceed 0.8.The established models present high stability and predictive accuracy.This paper provided a simple and convenient methodfor measuring the contents of major quality components in jasmine tea.更多还原