Application Of Near Infrared Spectroscopy Technology In The Production Process Of Yinlantiaozhi Capsule

Object:Drugs, a special commodity as a treatment for diseases, its quality is directly related to the human health and life safety. Quality control and evaluation is essential to ensure the quality of medicines and thus guarantee safe and effective of drugs, moreover it is one of focusesin production and regulatory. However, current quality control of chinese medicine processing and production can not meet the modern industrial production needs which ask for fast, simple, convenient and nondestrure. Firstly, this is because medicine raw materials come from natural sources. At the same time, most manufactures control the quality with more experience and off-line detection methods.In addition, CHM(Chinese Herbal Medicine)quality control system is imperfect, so it is difficult to effectively control and evaluation the quality of medicines, meanwhile conventional detection methods cumbersome and time-consuming. These cause that the clinical efficacy of CHM and product quality of different batches is not guaranteed. Eventually these reasons restrict industrialization and modernization of CHM production.These problems may be effectively resolved with the advent of PAT (Process Analysis Technology). As one of PAT techniques, NIR Spectroscopy contain rapid, clean, accurate, non-destructive and suitable for remote analysis advantages which meet such requirements for monitoring pharmaceutical process as an alternative of reference method. In recent years, it has been widely used in agriculture, pharmaceutical, petrochemical and other fields.According to what have argued above, this paper take the Yinlantiaozhi capsule production process for the study, researching around the raw ingredients, extraction, concentration, recovery of ethanol, mixed materials and end products and other processes, establishing multivariate calibration model between NIR spectra and quality control indicators, and then NIR rapid quality control is carried out for Yinlantiaozhi capsule production processes in order to guarantee the clinical efficacy and product quality. Finally these results provide a paradigm for supporting the standardization and modernization of CHM production.Method and results:1. In this study, different batches of raw herbs in Yinlantiaozhi Capsule (Because propolis is a direct crushing medicine in this process, other conventional analytical methods are currently used for its quality control) were collected from various parts of the country, crushed under the same conditions, and over 80 mesh sieve. Then the selected spectral data were combined with chemical analysis measured data using PLS or PLS-DA to construct the optimal models(containing exocarpium qualitative identification model; exocarpium naringin, moisture content; ginkgo leaf total flavonol glycosides, terpene lactones, moisture content and fingerprint similarity; Gynostemma gypenosides, moisture content quantitative calibration models). In addition, verification and evaluation of the models show that they could accurately determine the indicators in corresponding CHM by using FT-NIR spectroscopy in combination with integrating sphere. These models can provide technical support for pharmaceutical raw materials procurement and inspection before feeding.2. This paper take the extraction process of Yinlantiaozhi capsule for the study, extracting relevant information between NIR spectra and indicators during extraction process by chemometric software, and establish naringin, rutin, solid content and conductivity quantitative NIR calibration models to monitor the changes of these indicators during the extraction process. The verification show that average relative error of prediction (AREP) and average recovery of prediction(ARP)of naringin model is 5.88% and 102.14%, rutin model is 5.61% and 98.81%, solid content model is 4.42% and 102.16% and conductivity model is 3.31% and 101.64% in the 1st extraction process. On the 2nd extraction process, the verification reveal that AREP and ARP of naringin model is 3.18% and 101.00%, rutin model is 5.30% and 100.32%, solid content model is 4.30% and 101.49% and conductivity model is 4.45% and 100.98%. All these results i ndicate that the predicted value and the reference value have good correlation. The constructed model provides a method for fast, real-time, dynamic monitoring the changes of indicators during the extraction process, finally ensures the quality of the end product uniformity and stability.3. In this work, the concentration process of Yinlantiaozhi capsule are taken as research object, quantitative prediction models based on PLS1 are constructed to extract the information between NIR spectral and chemical values. The research results show that the AREP of naringin is 8.34%, rutin is 8.82%, relative density is 0.17% and conductivity is 7.28%, respectively. The ARP of naringin is 99.32%, rutin is 104.55, relative density is 99.94% and conductivity is 99.03. All these results indicate that the constructed models can predict index components content accurately during concentration process. It can be achieved in real-time, dynamic, online quality monitoring, feedback control when these models are applied to large-scale production.4. This paper take Yinlantiaozhicapsule ethanol recovery process for the study, combining NIR technology with chemometrics software to establish relative density, conductivity, and the active ingredient content (naringin, rutin) quantitative calibration model in this unit. At last, these developed models provibe an effective, fast, real-time and comprehensive method for online monitoring recovery of ethanol process. Additionally, these results also prove that NIR analysis can predict the relevant index accurately in complex chemical system (eg:water-ethanol system).5. In this work, we determine the end point of blending process according to changes of the continuous scanning material mixing NIR spectra properties. The results show that the calculating of continuously scanning NIR spectrum similarity, the average absorbance of standard deviation spectra, and the difference between maximum and minimum peak of standard deviation spectra can accurately determine the end of mixing. This method not only can ensure mixing uniformity of materials, ensure the stability of drug quality. Furthermore, it can also improve equipment utilization and savingenergy.6. In this work, we establish the NIR quantitative calibration model for rapidly detecting of the moisture content in final product. The performance of constructed model reveals that RMSECV is 0.115, RMSEC is 0.0807, r is 0.9038, and AREP as well as ARP is 1.49% and 100.03%. These results reveal that the model has a good predictive ability and rubustness. At the same time, we randomly select finished products for NIR spectral scanning, and compare properties of these NIR spectra to determine the stability of product quality. The results show the quality of three batches samples is homogeneous and stable.Conclusion:According to the results obtained in this work, NIR spectroscopy technique can be used for monitoring the key units in preparation of Yinlantiaozhi capsules. In the actual production process, we can quickly and accurately learn the content of indicators when NIR spectra of the unknown samples are improted into the established models. Given the fast determination, nondestuction of samples and no reagent waste of these analytical methods, these models may be useful for rapid quality control in production process as well as the parameter optimization of production in corporation with feedback control in industry. Finally this method can ensure the quality of the end product uniformity and stability.