目的 应用在线中红外光谱技术（MIR）对热毒宁注射液金银花Lonicerae Japonicae Flos、青蒿Artemisiae Annuae Herba（金青）醇沉过程中新绿原酸、绿原酸、隐绿原酸、异绿原酸A、异绿原酸B、异绿原酸C和断氧化马钱子苷的含量进行在线监控，提高该过程在线质量控制水平。方法 利用衰减全反射中红外光谱技术采集9个批次金青醇沉过程样本的光谱信息，组合间隔偏最小二乘法（synergy interval partial least squares，SiPLS）找到最佳波段，偏最小二乘法（PLS）建立该过程7种指标成分的定量模型，以决定系数（R²）、校正集误差均方根（RMSEC）、交互验证集误差均方根（RMSECV）、预测相对偏差（RSEP）为指标评价模型性能。结果 经模型优化，7种指标成分模型的R²值均大于0.96，RMSEC和RMSECV值均小于0.2 mg/mL，RSEP值和相对误差均小于10%。结论 在线中红外光谱分析技术可用于定量测定金青醇沉过程7种指标成分的含量，结果准确可靠。

Objective Online mid-infrared spectroscopy (MIR) was used to monitor the concentration of neochlorogenic acid, chlorogenic acid, cryptochlorogenic acid, isochlorogenic acid A, isochlorogenic acid B, isochlorogenic acid C, and oxidized loganin of alcohol precipitation process of Jinyinhua (Lonicerae Japonicae Flos, LJF) and Qinghao (Artemisiae Annuae Herba, AAH) (LA) of Reduning Injection (热毒宁注射液), so as to increase the level of online quality control. Methods Attenuated total reflectance infrared spectroscopy technology was used to acquire the spectral information of nine batches samples from LA alcohol precipitation. The optimal bands were found by using synergy interval partial least squares (SiPLS). The quantitative models of seven indicators of the process were established based on partial least squares method (PLS). Model performance was evaluated by the coefficient of determination (R²), root mean square error of calibration (RMSEC), root mean square error of cross validation (RMSECV) and relative standard error of prediction (RSEP). Results After modeling optimization, the model values of R² of the seven indicator components were all greater than 0.96, RMSEC and RMSECV values were less than 0.2 mg/mL, and the values of RSEP and relative error (RE) were less than 10%. Conclusion The results demonstrate that online MIR could determine the concentration of seven indicators of alcohol precipitation process of LA rapidly and accurately.

R283.6

国家“重大新药创新”科技重大专项“基于功效成分群的中药口服固体制剂先进制药与信息化技术融合示范应用”（2018ZX09201010-004）；国家中药标准化项目“热毒宁注射液标准化建设”（ZYBZH-C-JS-31）