The Research On In-line Monitoring Of The Fluidized Bed Granulation And Drying Process By Using Near Infrared

Solid dose pharmaceutical is one of the most important forms of drug at present.The factors that affect the quality of drugs are very complex,including material factors and process factors.At present off-line analysis methods are used to determine the critical material attributes(CQAs)in the processes of fluidized bed granulation and drying.However,the off-line analysis methods are destructive and time-consuming.Besides it will make the determination of parameters behind of the production,so,the final products will be of poor stability,uniformity,safety and quality.Therefore,in order to ensure safety,effectiveness,stability and uniformity of the products,the process analysis technology(PAT)was introduced to fluidized bed granulation drying process in real-time monitoring and control of critical quality attributes of the pharmaceutical production process.As a widely used PAT tool,near infrared spectroscopy(NIRS),due to the advantages of rapid,non-destruction and environment friendly,has potential applications in inline monitoring and quality analysis of manufacturing process in the field of pharmaceutical industry.In the fluidized bed granulation and drying process,the moisture content,particle size and bulk density of particles are the critical material attributes of the process.The moisture content of particles would affect the mobility,compressibility,and stability of the final products,and the quality of product particles would have influence on uniformity in tablet weight and constancy in the hardness and friability properties of the product.Besides,bulk density represents the flow ability and compressibility of the granules.Therefore,in this research,NIRS combined with chemometrics methods was applied in fluidized bed granulation drying process in monitoring moisture content,particle size and bulk density of particles.Moreover,the accuracy and effectiveness of models was investigated detailedly.The details are listed as below:(1)Feasibility study on NIRS monitoring of fluidized bed granulating drying process.In this study,the feasibility of in-line monitoring of the fluidized bed granulation and drying processes was investigated by using a micro near infrared spectrometer.First,the partial least squares regression(PLSR)off-line quantitative analysis models of particle moisture content,particle size and bulk density were established.During the establishment of the PLSR models,different spectral preprocessing methods and variable selection methods were compared detailedly.Finally,the off-line quantitative models of moisture content,particle size(D10,D50,D90)and bulk density weredeveloped,and the determination coefficient of calibration(Rc2),prediction(Rp2),root mean square error of cross validation(RMSECV),root mean square error of prediction(RMSEP)were 0.992,0.988,0.1069%,0.1463%for water content,0.981,0.986,1.059?m,0.8416 ?m for D10,0.981,0.990,2.339 ?m,1.737 ?m for D50,0.973,0.986,3.109 ?m,1.712 ?m for D90,and 0.988,0.967,0.008620 g/ml,0.009859 g/ml for bulk density.In addition,to evaluate the off-line quantitative models,moisture content,particle size(D10,D50,D90)and bulk density quantitative models of ratio of performance deviation(RPD)were 7.04,7.64,8.08,9.76,6.59,respectively.It showed the the feasibility of in-line monitoring of the moisture content,particle size and bulk density in the fluidized bed granulation and drying processes.(2)Research on developing in-line quantitative models of critical material attributes(CQAs)during fluidized bed granulating drying process by using NIRS.In this study,in-line quantitative models of moisture content,particle size and bulk density were developed by NIRS during fluidized bed granulating drying process During the establishment of in-line quantitative PLSR models,the models would be further optimized by comparing different pretreatment and variable selection methods.The in-line quantitative models of moisture content,particle size(D10,D50,D90)and bulk density were developed,and the determination coefficient of Rc2,Rcv2,RMSECV,RMSEP were 0.971,0.907,0.1697%,0.1239%for moisture content,0.981,0.986,0.8416 ?m,0.6984 ?m for D10,0.981,0.990,1.737 ?m,1.819 ?m for D50,0.973,0.986,1.712 ?m,2.565 ?m for D90,and 0.988,0.967,0.009859 g/ml,0.01716 g/ml for bulk density.The results showed that the established models have good prediction ability.(3)The application research of NIRS used in real-time monitoring of CQAs in fluidized bed granulation and drying processIn this study,the quantitative models of moisture content,particle size and bulk density which developed in the second and the third chapter used for real-time monitoring the process of fluidized bed granulation and drying.Firstly,the accuracy and effectiveness of in-line quantitative models and off-line quantitative models for real-time monitoring of the CQAs changes were analyzed and evaluated during fluidized bed granulation and drying process,respectively.The evaluation results showed that the in-line quantitative model has good accuracy and high predictive ability for real-time in-line monitoring.It indicates that NIRS can replace traditional measurement methods for monitoring and control the production process in real time.However,the accuracy of the off-line quantitative models applying in monitoring the fluidized bed granulation and drying process was poor,so it could not used for real-time monitoring.Realizing real-time in-line monitoring of fluidized bed granulation and drying process is conducive to real-time control the fluctuation of product quality in production process,so as to guide production and ensure the consistency of drug quality.The innovation points are listed as below:(1)The micro near-infrared spectrometer was introduced into the fluidized bed granulation and drying process for the first time,and in-line quantitative analysis models for the CQAs(particle water content,particle size and bulk density)were established.(2)The research real-time monitored of the CQAs in the process of the fluidized bed granulating and drying by micro near infrared spectrometer for the first time.