近红外光谱透射与透反射法测定扶正止痒方提取液中黄柏酮等定量模型的比较分析

Comparative analysis of quantitative models of obacunone and other components in Fuzheng Zhiyang prescription extract by the methods of transmission and transmission&reflection of near infrared spectroscopy

目的:采用近红外光谱(NIRS)的透射与透反射2种方法,分别建立并比较扶正止痒方提取液中指标成分的定量校正模型,为液体样品采用NIRS测定方法的选择提供参考。方法:利用高效液相色谱(HPLC)法测定提取液中黄柏酮、落新妇苷及3个异构体(新落新妇苷、新异落新妇苷、异落新妇苷)的总量,采用NIRS的透射与透反射2种方法分别采集样品光谱,利用偏最小二乘法(PLS)和交叉验证法(CV)等化学计量学方法,尝试建立指标成分含量与近红外光谱间的定量校正模型。结果:采用透射与透反射建立的黄柏酮定量校正模型的相关系数r分别为0.9484、0.8883,验证集的r分别为0.9407、0.9210,交叉验证均方差(RMSECV)分别为0.0009、0.0013,预测均方差(RMSEP)分别为0.0011、0.0011;落新妇苷及3个异构体(简称落新妇总苷)定量校正模型的r分别为0.9795、0.9372,验证集的r分别为0.9821、0.9669,RMSECV分别为0.0109、0.0188,RMSEP分别为0.0113、0.0204。采用2种方法所建模型的预测值与实测值之间具有良好的线性关系,两者无显著性差异(p>0.05),预测效果良好。结论:采用NIRS的透射与透反射所建指标成分的校正模型均能达到定量要求,但应用透反射所建模型的预测值与实测值的相关性更好,测定误差更小,预测准确度更高,预测值更接近实测值,本文可为液体样品采用NIRS测定方法的选择提供参考。

Objective:To establish and compare respectively the quantitative calibration models for the index components in the extract of Fuzheng Zhiyang prescription by the methods of transmission and transmission&reflection of near infrared spectroscopy(NIRS),which provides reference for the selection of determination methods for liquid samples using NIRS.Methods:High performance liquid chromatography(HPLC)was used to determine the total contents of obacunone,astilbin and three isomers(neoastilbin、neoisoastilbin、isoastilbin)in the extract,and two methods of NIRS,the transmission and transmission&reflection,were used to collect respectively the spectra of samples.Chemometrics such as partial least squares(PLS)and cross-validation(CV)were used to establish the quantitative calibration models for the contents of index components and near infrared spectra.Results:The correlation coefficients of the quantitative calibration models of obacunone established by the methods of transmission and transmission&reflection were 0.9484 and 0.8883 respectively,and those of the validation sets were 0.9407 and 0.9210 respectively.The root mean square errors of cross validation(RMSECVs)were 0.0009 and 0.0013 respectively,and the root mean square errors of prediction(RMSEPs)were 0.0011 and 0.0011 respectively.The correlation coefficients of the quantitative calibration models of astilbin and three isomers(total astilbin for short)were 0.9795 and 0.9372 respectively,and those of the validation sets were 0.9821 and 0.9669 respectively.RMSECVs were 0.0109 and 0.0188 respectively,and RMSEPs were 0.0113 and 0.0204 respectively.There was a good linear relationship between the predicted value and the measured value of the model constructed by the two methods,and there was no significant difference between the two(p>0.05),so the prediction effect is good.Conclusion:The correction models of the index components which were established by the methods of transmission and transmission&reflection of NIRS can both meet the quantitative requirements.However,the models established by the method of transmission have better correlation between the predicted and measured values,less measurement error,higher prediction accuracy,and the predicted values of the model established by the method of transmission are closer to the measured value.This paper can provide reference for the selection of determination method for liquid samples using NIRS.