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9 264.35～7 274.11 cm^?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0.999 4；经外部验证，预测均方根误差（RMSEP）为1.321，r为0.992 1；经相关性统计学分析，MAs的预测值与HPLC参考值的r为0.999 0，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 rapidly and quantitatively analyze the monoester alkaloids (MAs) in the extracting and concentrating process of Aconiti Radix Cocta (ARC) by near infrared spectroscopy (NIR). Methods Using NIR to collect the data of 103 ARC samples and using partial least squares (PLS) regression method to establish the quantitative analysis model of MAs between the information of NIR and MAs. Results The spectral range of MAs model of ARC was 9 264.35—7 274.11 cm^?1. The root mean square error of cross validation (RMSECV) was 1.171 and correlation coefficient (r) of the calibration model was 0.999 4. Through the external validation, the root mean square error of prediction (RMSEP) was 1.321 and r of the validation model was 0.992 1. According to the results of statistical analysis, r of the predicted value and the reference value of MAs was 0.999 0. The value of P is less than 0.001. This revealed that the NIR and HPLC methods had a good correlation in determining MAs and could accurately predict the amount of MAs in the covered range. Conclusion This method is convenient, rapid, accurate, and environment protective, and could be used for the on-line determination of MAs in the extracting and concentrating process of ARC and for the determination of the extraction and concentration endpoint of ARC.

广东省省部产学研结合科技创新平台项目（2011A091000005）；广东省中医药局科研课题（20121241）