Study On The Quality Control In Water Extraction And Concentration Process Of Tianshu Capsule

Traditional Chinese medicine(TCM)is a multi-component complex system.The fluctuations in the qualithe of Chinese materialy medica,the complexity of chemical ingredients and the complexity of chemical ingredients and the complexity of the production process make that it is necessary to realize process quality control in the TCM production process.Near-infrared(NIR)spectroscopy has the advantages of convenient,speed,efficient and accurate,non-destruction,the test samples without complicated pretreatment,consumption of chemical reagents,no environmental pollution and other advantages.It can monitor the content of cricricritical attributes quickly,thus it is more and widely applied in the process quality control field.However,the accuracy of models is one of the critical problems in the TCM production process.Migraine,with the character of palpitating pain by one or two sides,is one of the most common diseases that recurs frequently.Because of its high morbidity and ratio of recurring,Migraine disturbs human living quality severely.Tianshu capsule come from the ancient prescription of Dachuanxiongfang,and it included two traditional Chinese herbs,Rhizoma Ligustici Chuanxiong(CX)and Rhizoma Gastrodiae(GA).It it has the effect of promoting blood circulation to remove blood stasis and alleviating pain by calming endogenous wind.Tianshu capsule has been proved to be an effective formula in treating Migraine clinically.In this paper,taking water extraction and concentration process of Tianshu capsule as the research object,established a water extraction,concentration and other critical process quality offline monitoring system which is based on the active ingredient,to lay the foundation for on-line monitoring studies.It will behelpful to improve the quality control level of Tianshu capsule,improve the batch-to-batch quality uniformity and stability,provide technical support for later automation production of Tianshu capsule.Study includes the following two parts:Part one:Study on the water extraction process of Tianshu capsule to establish NIR monitoring models and the relationships between the pivotal parameters and the quality attributes.Taking the water extraction process of Tianshu capsule as the research objects,NIR monitoring models were constructed to predict the contentof effective ingredients real-timely in the actual production process.Partial least squares(PLS)algorithm was used to establish a rapid quantitative analysis model for content of gastrodin and otal solid content in the first and the second water extracting process of Tianshu capsule,which could timely reflect its state.The samples of the first and the second water extracting process of Tianshu were collected,after the rejection of abnormal samples,the pretreatment of the spectra and choice of the corresponding wave band,the multivariate calibration models were respectively established using partial least squares algorithm between spectra and content of gastrodin and total solid content determined by the reference method.The samples in the first and the second water extracting process,which did not participate in modeling were used to test the performance of the models.The firs water extracting process correlation coefficients of gastrodin content and total solid content were respectively 0.9622 and 0.9900,and the predicted coefficients were respectively 0.9846 and 0.9895,Root Mean Square Error of Calibration(RMSEC)were 10.434and 0.0662 respectively,Root Mean Square Error of Prediction(RMSEP)were 5.9682 and 0.0787 respectively,and the relative standard errors of predictions(RSEP)were 4.78%and 3.91%respectively.The second water extracting process correlation coefficients of gastrodin content and total solid content were respectively 0.9680 and 0.9772,and the predicted coefficients were respectively 0.9659 and 0.9786,Root Mean Square Error of Calibration(RMSEC)were4.6857 and 0.0525 respectively,Root Mean Square Error of Prediction(RMSEP)were 3.6975 and 0.0343 respectively,and the relative standard errors of predictions(RSEP)were 4.85%and 3.07%respectively.After multi-batch verification,the relative standard errors of prediction were lower than 5%for each compound,indicating all the established models exhibited satisfactory predictiveperformance.Taking the water extraction process of Tianshu capsule as the research objects,study on the correlation between process parameters and quality attributes.Take the extraction rate of gastrodin and total solid content as evaluation indexs,I investigated the times of water volume,decocting time and decocting number influence on the water extractioneffect by the single factor and response surface experiments to find the key process parameters and establish relation equations between the key process parameters and quality attributes:Y1=87.77+1.28×A+8.65×B+0.42×C-0.42×A×B+0.38×A×C-0.63×B×C-0.20×A2-8.25×B2+0.28×C2,R2=0.9931;Y2=14.81+0.82×A+2.52×B+0.30×C-0.13×A×B+0.18×A×C+0.24×B×C-0.56×A2-2.44×B2-0.20×C2,R 2=0.9656.After verification,the equations are accurate and useful,which can lay the foundation for automatic control and feedback control in the process.Part two:Study on the concentration process of Tianshu capsule to establish NIR monitoring models.Taking the concentration process of Tianshu capsule as the research objects,NIR monitoring models were constructed to predict the contentof effective ingredients real-timely in the actual production process.Partial least squares(PLS)algorithm was used to establish a rapid quantitative analysis model for content of gastrodin and density in concentrating process of Tianshu Capsule water extraction concentration,which could real-timely reflect its state.The samples of concentrating process of Tianshu water extraction were collected,after the rejection of abnormal samples,the pretreatment of the spectra and choice of the corresponding wave band,the multivariate calibration models were respectively established using PLS algorithm between spectra and content of gastrodin and density determined by the reference method.The samples in the concentrating process,which did not participate in modeling were used to test the performance of the models.The correlation coefficients of content of gastrodin and density were respectively 0.9356、0.9646,and the predicted coefficients were respectively 0.9911、0.9447,Root Mean Square Error of Calibration RMSEC)were 0.2966 and 0.0107 respectively,Root Mean Square Error of Prediction(RMSEP)were 0.1422 and 0.0154 respectively,and the relative standard errors of predictions(RSEP)were 4.52%and 0.91%respectively.After multi-batch verification,the relative standard errors of prediction were lower than 5%for each compound,indicating all the established models exhibited satisfactory predictiveperformance.