Studies On Quality Control Of Danshen Injection Production Using Statistical Process Control

The modernization and internationalization of Traditional Chinese Medicine (TCM) are blocked by its poor consistence and instability. One of the reasons is lack of efficient on-line quality measuring and controlling methods. As one of the techniques for on-line monitoring, fast and non-destructive analyzing, near infrared spectroscopy (NIRS) has already been widely used in many areas, but scarily in TCM pharmacy process. Multiple statistical process control (MSPC) is one of the useful online data analysis tool, and it has been used for process controlling, information extraction and process diagnosis. In this thesis, application of NIRS technique and MSPC to on-line monitoring of Danshen injection manufacturing process, including concentration and ethanol precipitation process, had been researched. The main content was summarized as follows:Firstly, NIRS was used for on-line analysis of concentration process of Danshen. Spectra were collected on-line, reference values were measured by HPLC and lyophilization. Wavelength ranges 7200～9569 cm^-1 and 4500～7083 cm^-1 were selected, and partial least square (PLS) regression was used to establish the quantitative analysis models of Danshensu, protocatchuic aldehyde and solid content.Secondly, NIRS has been explored for fast analysis of multi-components in ethanol precipitation process of Danshen. Spectra were collected off-line. Samples were measured by HPLC as reference values. Spectra were pretreated with different methods, and wavelength was selected. Calibration models were developed by PLS regression technique to simultaneously measure Danshensu, protocatchuic aldehyde, caffeic acid, salvianolic acid A, salvianolic acid B, salvianolic acid D and solid content.Thirdly, principle component analysis (PCA) was used to analysis the alcohol precipitation of Danshen. The process trends during alcohol precipitation under normal operation were researched combined NIRS and PCA. The PCA monitoring model of alcohol precipitation was established and explained. Finally, the MSPC models of alcohol precipitation of Danshen were established. Spectra were collected on-line. Extracting the information contained in the spectra using PCA, Hotelling T² and SPE models were established, and were employed to on-line analyze ethanol precipitation process under different process conditions.