近红外光谱技术在灵芝提取物品控中的应用研究  

Research on the authentication and adulteration of the Gandoderma lucidum extracts by Near Infrared Spectroscopy

作　　者：王钧[1] 邵平[1] 王星丽 瞿亮 孙培龙[1] 

机构地区：[1]浙江工业大学食用菌精深加工团队,浙江杭州310000 [2]浙江百山祖生物科技有限公司,浙江丽水323000

出　　处：《食药用菌》2016年第6期362-367,共6页Edible and Medicinal Mushrooms

基　　金：国家科技支撑计划课题食用菌深加工产品开发与产业化(2013BAD16B07)

摘　　要：采用近红外光谱漫反射分析技术对灵芝提取物进行了光谱测定和定性分析,通过计算样品的杠杆值、学生化残差和马氏距离剔除异常样品,根据原始图谱和一阶导数差谱,选取不同的波数区间和有效的预处理方法,分别采用主成分回归(PCR)和偏最小二乘法(PLS)对灵芝提取物真伪进行识别模型分析。结果表明:灵芝提取物样品中有一个为异常样品,在建模前予以剔除;在PCR模型中,用于识别灵芝提取物的最佳波数区间预处理分别为7 654~6 987 cm^(-1)、5 534~4 000 cm^(-1)和多元散射校正结合二阶导数(MSC+SD),相关系数达到0.946 5。在PLS模型中,用于识别灵芝提取物的最佳波数区间预处理分别为9 150~4 000 cm^(-1)、一阶导数(FD),相关系数达到0.965 6。运用所建立的识别模型预测灵芝提取物样品,其预测识别率都达到100%。同时对PCR和PLS模型的预测值进行配对T检验结果,两者无显著差异。Near infrared spectra of Gandoderma lucidum extracts （GLE） samples were obtained and a qualitative analysis was carried out. Leverage value, studentized residue and sample＇s Mahalanobis distance were applied to detect the outlier sample, and different wavenumber and the effective pretreatment method were selected by the original and the first derivative spectra. The near infrared qualitative analysis models were performed by principal component regression （PCR） and partial least square （PLS） regression. The results showed that one sample was an outlier and should be deleted. The best PCR model gave the correlation coefficient of 0.946 5, when the best factor, pretreatment and the optimum wave number range were 4, MSC＋SD, and 7 654-6 987 cm-1. 5 534-4 000 cm-1, respectively. In the PLS model, it also gave the correlation coefficient of 0.965 6, when the best factor, pretreatment and the optimum wave number range were 6, FD, and 9 150-4 000 cm-1, respectively. Both the forecast recognition rates of the two models reached 100%, according to the paired sample test, the result of PCR prediction had no significant difference compared with those by PLS.

关 键 词：近红外光谱 灵芝提取物 PCR PLS 真伪 

分 类 号：R284[医药卫生—中药学] O657.33[医药卫生—中医学]