目的采用近红外光谱(NIR)法对制川乌提取、浓缩液中的单酯型生物碱类成分(MAs)进行快速定量分析。方法应用近红外透射光谱法搜集103个制川乌样品的NIR数据,采用偏最小二乘回归(PLS)法建立NIR信息与MAs之间的定量分析模型。结果制川乌提取、浓缩液中MAs的近红外透射光谱模型建模的光谱范围为9264.35～7274.11cm?1,校正模型的内部交叉验证均方根(RMSECV)为1.171,相关系数(r)为0.9994;经外部验证,预测均方根误差(RMSEP)为1.321,r为0.9921;经相关性统计学分析,MAs的预测值与HPLC参考值的r为0.9990,P<0.001,说明NIR法与HPLC法测定MAs相关性较好,可以较准确预测其覆盖范围的MAs的量。结论本法具有方便、快速、准确、环保等特点,可用于制川乌提取、浓缩过程中MAs量的在线监测,进而较准确地判定提取、浓缩终点。 Objective To investigate the rapid quantitative analysis of monoester alkaloids (MAs) extracted and concentrated by using near infrared spectroscopy (NIR). Methods The NIR data of 103 samples of Radix Aconiti Kusnezoffii collected by Near Infrared Transmittance Spectroscopy, and the quantitative analysis model of NIR information and MAs was established by Partial Least Squares Regression (PLS). Results The spectral range of the Nansu extract of Mizuho Ubisumai was 9264.35 ~ 7274.11cm -1. The RMSECV of the calibration model was 1.171, and the correlation coefficient (r) was 0.9994; RMSEP was 1.321 and r was 0.9921 by external validation; r was 0.9990, P <0.001 by correlation analysis between predicted value of MAs and reference value of HPLC, indicating that NIR method and HPLC The correlation of MAs determined by the method is better, and the amount of MAs in the coverage area can be more accurately predicted. Conclusion This method is convenient, rapid, accurate and environmentally friendly. It can be used for on-line monitoring of MAs during the extraction and concentration of Radix Aconitum and then determine the endpoint of extraction and concentration more accurately.