(695e) In-Line Monitoring of the Extraction Process of Panax Notoginseng (Burk.) F. H. Chen Based On the Quantitative Use of Near-Infrared Spectroscopy

    Panax notoginseng (Burk.) F. H. Chen, a well-known medicinal herb that belongs to the ginseng family, has a long history of use in China. It has been traditionally used for its hemostatic and cardiovascular effects, which help arrest internal and external bleeding, reduce swelling and pain, as well as to disperse blood clots, eliminate blood stasis and promote blood circulation. The pharmacological properties of Panax notoginseng are generally attributed to its saponins, including ginsenosides, notoginsenosides and gypenosides. Among these saponins, arasaponin R₁ and ginsenoside Rg₁, Re, Rb₁, Rd are considered to be the major active compounds.

    The extraction process of Panax notoginseng was the first step of XueShuantong injection manufacturing. It was considered to be a crucial process as it defined the quality and quantity of the effective constituents in the final product.

    Monitoring extraction solutions with NIR spectroscopy will produce an absorption spectrum of the chemicals present in the extraction solution in a rapid and non-destructive way. These spectra can then be manipulated using multivariate data analysis techniques to develop calibration models for each measured variable. Although the initial model building will require reference data based on the traditional destructive analytical methods, a robust model can thereafter be used to predict the quality attributes in a non-destructive way.

    For NIR measurements, the fiber optic probe was immersed in the extracting solution to collect NIR spectra in real-time. Ultra performance liquid chromatography (UPLC) was used as a reference method to determine the concentrations of arasaponin R₁ and ginsenoside Rg₁, Re, Rb₁, Rd in the extraction solution. Partial least squares regression (PLSR) calibration models of arasaponin R₁ and ginsenoside Rg₁, Re, Rb₁, Rd were built separately, using different spectral regions and pretreatments. The results showed that the correlation coefficients of inner cross validation were 0.9962, 0.9902, 0.9915, 0.9951 and 0.9876, and the root mean square error of cross validation (RMSECV) were 0.057, 0.232, 0.028, 0.106 and 0.084 (mg/ml), respectively. The established models were used for in-line and real-time monitoring of a new extraction batch, and the root mean square error of prediction (RMSEP) were 0.0568, 0.233, 0.0411, 0.168 and 0.0621 (mg/ml), respectively. The NIR spectroscopy was proved to be an effective technology in the real-time monitoring of extraction process of Panax notoginseng, providing real-time information of the extraction process.