丹参多酚酸盐柱层析过程的近红外光谱在线检测及质量控制

Near-infrared online monitoring and quality control of Salvianolate column separation

目的:研究利用近红外光谱(NIR)结合HPLC含量测定对丹参多酚酸盐生产的关键工艺——柱层析进行在线质量分析及控制的可行性。方法:对NIR透射光谱和丹酚酸B的含量进行PLS回归分析,并系统考察光程、波数选择、预处理方法对建模效果的影响。结果:2 mm光程优于1 mm光程,在2 mm光程下,最优的波数范围为6 102.1-5 446.3cm-1、预处理方法为矢量归一化,对预测集样本的均方根误差为0.234 mg/mL、R2为0.995 2。结论:该方法快速、简便、准确,可用于生产过程在线检测及质量控制。

AIM ： To study the feasibility of online monitoring and controlling of the key process of column chromatographic separation in the production of Salvianolate by combining near-infrared （NIR） spectrum and HPLC fingerprinting. METHODS ： Partial Least Square （PLS） regression was used to model the correlation of NIR spectra with the concentrations of salvianolic acid B, and the influences of light path, wavelength selection, and preprocessing method on the PLS model were investigated systematically. RESULTS： A 2 mm light path was better than 1 mm one, and with 2 mm light path, the optimal wave-number was in the range of 6 102.1 -5 446.3 cm^-1 and the optimal preprocessing method was the vector normalization. The root mean-square error of PLS model on test samples was 0. 234 mg/mL, and R^2 was 0. 995 2. CONCLUSION： This method is proved to be fast, convenient, and precise. It can be used to online monitoring and quality control of the manufacturing of Salvianolate.