Study On The Quality Control Method Of Critical Purification Processes Of Lonicerae Japonicae Flos Extract

Recently,with the rapid development of Traditional Chinese Medicines(TCMs)industry,more attentions have been paid to the safety,effectiveness and consistency of TCMs.In order to reduce the batch-to-batch variation,enhanced quality control should be applied to the pharmaceutical processes.In this thesis,the purification processes of Lonicerae Japonicae Flos extract in Tanreqing Injection were investigated under the guidance of Quality by Design concept.A quality control study on the critical purification process was conducted through several control strategies such as process parameters control,statistical process control and end-points determination using design space method and spectroscopy technique.The main contents and achievements of this thesis are summarized as follows:1.A data driven method for critical processes identification was raised.Three types of process performance indices,including the change of the yield,the change of the similarity and the change of the purity,were evaluated to assess the influence of each process on product chemical composition and process efficiency.Finally,extraction process,lime milk prcipitation-sulphuric acid ethanol solution dissolution process and n-butanol liquid-liquid extraction process were identified as the critical processes.2.Design space of lime milk precipitation process of Lonicerae Japonicae Flos aqueous extract was established.The evaluation indices for this process were organic acid purity and the amount of each organic acid obtained from per unit mass of medicinal materials.Four critical parameters including drop speed of lime milk,pH value after adding lime milk,settling time and settling temperature were identified according to the weighted standardized partial regression coefficient method.Quantification models between process evaluation indices and critical process parameters were established by a stepwise regression analysis.A design space was calculated by a Monte Carlo simulation method,and then verified.The verification results showed that the quantification models were accurate and the design space was reliable.An operation space was recommended to make it more convenient for application.3.A real-time monitoring model was established for lime milk precipitation process of Lonicerae Japonicae Flos aqueous extract by combining NIR spectroscopy technique and multivariate statistical process control(MSPC).First principal scores,Hotelling T2 statistics and DModX statistics were used to establish control limits.The process trajectories were utilized to reflect the running condition of process.This model accurately diagnosed the batches deviated from normal conditions.4.Two rapid quantification models for 7 organic acid contents in the aqueous phase during n-butanol liquid-liquid extraction process were established using NIR spectroscopy dataor UV-vis spectroscopy data.And the latter one was used to determine the extraction end-points.This end-point determination approach could be utilized for on-site measurements to improve the process efficiency.