近红外光谱技术测定黄连中6种生物碱含量的新方法

A New Method for Analysis of Six Alkaloids in Coptis chinensis Franch.by Near Infrared Diffuse Reflectance Spectroscopy

目的应用近红外光谱技术对黄连中药根碱、非洲防己碱、表小檗碱、黄连碱、巴马汀和小檗碱含量进行快速无损测定。方法选择在1 308～2 393 nm范围内的近红外光谱,比较了光谱不同预处理方法对校正结果的影响,并比较了主成分分析法(PCR)、偏最小二乘法(PLS)、改进偏最小二乘法(MPLS)建模的效果,并对模型进行内部和外部验证。结果采用SNV and Detrend散射处理技术,2.4.4.1数学处理方法及改进偏最小二乘法回归建立的定标方程模型的准确性最好。6种生物碱成分定标模型的校正决定系数(R2)为0.8047～0.9674,校正标准误差(SEC)为0.0312～0.2471。对预测集样品进行外部验证,预测值与真值的相关系数(Rv2)为0.744～0.929,预测标准误差(SEP)为0.043～0.346。结论建立了利用近红外光谱法直接测定黄连中6种生物碱含量的新方法。该方法快速简便,准确可靠,适合于黄连中主要生物碱类成分的快速测定分析。

ObjectiveTo establish a method for rapid determination of gatrorrhizine,columbamine,epiberberine,coptisine,palmatine and berberine in Coptis chinensis Franch.by near infrared diffuse reflectance spectroscopy. MethodsThe range of 1308-2393nm of near infrared spectra（NIRS） was selected.Different spectra pretreatment methods and mathematics methods such as PCR,PLS and MPLS were compared and the calibrations of the alkaloid content were performed. ResultsThe results showed that SNV＋Detrend,2,4,4,1 pretreatment and MPLS could be the best factors for the model.The average determination coefficient of validation（R2） and the square error of cross（SEC） of the models of six alkaloids were respectively 0.8047～0.9674 and 0.0312～0.2471.For the models based on prediction samples,the determination coefficient of independent validation（Rv2）and standard error of prediction（SEP） were 0.744～0.929 and 0.043～0.346. ConclusionThe establishment of the new near infrared diffuse reflectance spectroscopy method can be used to analyze the six alkaloids,which can provide a rapid,exact and reliable method for determining the main alkaloids in C.chinensis Franch.