A Decision Support System Integrating PAT and Population Balance Models for Pharmaceutical Material Granulation

Yajun LI,Xuezhong WANG
DOI: https://doi.org/10.11896/j.issn.1005-023X.2018.10.029
2018-01-01
Abstract:A decision support system (DSS)is developed for a high shear wet granulation process that has integrated near in-frared spectroscopy (NIR),on-line imaging and image analysis,and population balance (PB)models.The integrated PAT (process analytical technology)based on NIR and imaging allows multiple process variables and granule properties to be characterized in real-time,including mixing uniformity,binder content,granule size and size distribution,as well as granule growth,aggregation and breakage,and facilitates rapid identification of the operational spaces.The PAT measurements also provide data for estimation of the rate constants of kinetics kernels in the PB models and continuous improvement of the models.The models on the other hand can be used to guide PAT based experiments in searching for optimum operational space.A case study applying the DSS to the granulation of microcrystalline cellulose and mannitol powders using 3% aqueous solution of polyvinylpyrrolidone as the binder is described.The spraying period has shown to experience four phases:wetting,nucleation,rapid growth and moderate growth,but the boundaries separating the phases were different depending on the spray rate.At a high spraying rate,the binder contents at which the growth stepped into nucleation regime and rapid growth regime were lower.However,it didn't show significant influence on the median size of the granules formed.In addition,the mixing uniformity was monitored using NIR,allowing identification of the end point of pow-der mixing.The DSS combined the effective experiments based on PAT with process modelling,which enabled rapid characterization of the operational spaces and granule growth behavior.It also provided sufficient data collected real-time during the process to im-prove the robust and accuracy of the model continuously,consequently increasing the efficiency of optimization in granulation.
What problem does this paper attempt to address?