Real-time Evaluation And Monitoring Of Fluidized Bed Coating Process By Near Infrared Spectroscopy

Fluidized bed coating is a typical unit operation in the pharmaceutical production of solid preparations.The diversity of coating products and critical quality attributes(CQA)leads to the lack of standards for quality control,so it is imperative to implement applicable technical standards for evaluation.The advantage of process analytical technology(PAT)is that the production process can be analyzed in real time.By establishing the in-line monitoring method of CQAs in the fluidized bed coating process,the closed invisible problem in fluidized bed production process can be solved and the quality of the production process can be monitored in real time.Near Infrared Spectroscopy(NIRS)can be used as A PAT tool for in-line process monitoring to reflect the real state of the production process.Fluidized bed pellet coating is formed by spraying coating film on the substrate surface.The particle size and the uniformity of coating film thickness,the uniformity of Active Pharmaceutical Ingredients(API)content,and the release characteristics of coated pellets should be determined to meet the quality standards.A micro-NIR spectrometer was used to monitor the process of fluidized bed coating in this research.The in-line quantitative model of CQAs(pellet coating film thickness,API concentration and dissolution behavior of pellets in slow release coating process)was established during fluidized bed coating process,and the prediction results of NIR model were analyzed and evaluated.The specific research contents of this paper are as follows:(1)NIRS In-line monitoring of coating film thickness during fluidized bed coating process.In this study,a micro near infrared probe was integrated into a fluidized bed equipment to monitor the coating thickness of pellets during the coating process.A method of measuring the coating film thickness with a microplate reader was developed and tested.The method was used to identify the actual coating film thickness of the pellet.A Partial Least Squares Regression(PLSR)model for in-line quantitative analysis of coating agent concentration and coating film thickness in fluidized bed was established.After pre-processing the original data,three spectral band selection methods are compared to improve the prediction ability of the model.The optimal in-line quantitative model determination coefficient of R2cv,R2p,RMSECV and RMSEP were 0.9749,0.9911,0.9843μg/mL and 1.203μg/mL for coating agent concentration,0.9749.0.9911,0.2800μm and 0.3422μm for film thickness and the RPD value was 3.689.The RMSEP values of external validation were 1.331μg/mL and 0.3786μm,respectively.The feasibility of NIRS in measuring film thickness during fluidized bed coating was verified in this paper.(2)NIRS In-line monitoring of API content during fluidized bed coating process.In this study,the process of pellet coating in experimental fluidized bed was studied,and the In-line model for API content in the process of coating was established.After pre-processing the original data,three spectral band selection methods are compared to improve the prediction ability of the model.The optimal in-line quantitative model determination coefficient of R2cv,R2p,RMSECV and RMSEP were 0.9874,0.9973,8.509μg/mL and 5.060μg/mL,respectively,and the RPD value was 12.9696.The RMSEP value of externally validation was 8.896μg/mL.The experimental results verified the feasibility of NIRS for In-line monitoring of API concentration during fluidized bed coating.(3)Prediction of cumulative dissolution of coated pellets by NIRS during fluidized bed coating process.In this chapter,a micro near infrared spectrometer was used to monitoring the coating process of the sustained release functional layer in a pilot fluidized bed,and the cumulative dissolution degree of pellets at three coating time was monitored In-line,and a quantitative model was established.After pre-processing the original data,three spectral band selection methods are compared to improve the prediction ability of the model.The optimal in-line quantitative model determination coefficient of R2cv,R2p,RMSECV and RMSEP were 0.9755,0.9823,3.233%and 4.500%for first dissolution point,0.9358,0.9965.2.598%and 0.7939%for second dissolution point.0.9867,0.9927.0.4085%and 0.4726%for third dissolution point,and the RPD value were 5.201 5.991 7.047 respectively.In this part of the study,the model was validated in batch,and the prediction results were good,which proved the feasibility of In-line prediction of the dissolution rate of fluidized bed coated pellets.Here are the innovation points of this paper.The micro near infrared spectrometer was introduced into the fluidized bed coating process,combined with the new coating thickness measurement method and stoichiometric analysis.The In-line quantitative analysis model of pellet coating thickness monitoring was established and monitored;The change of API content during the coating process was monitored In-line by the micro near infrared spectrometer.The micro near infrared spectrometer was used to predict the dissolution behavior of pellets during fluidized bed coating.