(658g) Three-Compartmental Population Balance Model and Experimental Validation of a Continuous Twin-Screw Wet Granulation Process
A novel modelling methodology for a continuous wet granulation process is presented, combining population balance model (PBM) and multi-compartment concept to predict the critical quality attributes of the granule product, such as granule size distribution and mean granule size, along with the barrel length. By dividing the granulator screw barrel into three compartments along barrel length: wetting compartment (WC), mixing compartment (MC) and steady growth compartment (SC) and assuming different granulation mechanisms on each compartment, the proposed model could consider spatial heterogeneity of the twin screw granulation system and improve prediction accuracy in estimating particle size distribution. It was assumed that each compartment is homogeneous and can be described by a one-dimensional PBM. DOE with 36 experiments were carried out for model development and model validation. 16 combinations of aggregation and breakage mechanisms are compared in predicting the experimental particle size distribution and it is found that aggregation is heavily dependent upon particle length/area and all the four breakage mechanism do not have influence on the model prediction capability. The proposed three-compartmental PBM is compared with the one-compartmental PBM in predicting granule critical quality attributes (CQAs) at different experimental operating conditions and is proved to have advantage over one-compartmental PBM. Average residence time of granules on each compartment has been investigated pertaining to their influence on granules properties. It was shown that the longer residence time generates the larger granules on WC and SC, while generating small average particle and large fraction of trivial particle on MC. This novel three-compartmental population balance model has advantages of both considering the system heterogeneity and affordable computational cost, which can facilitate process optimization and model-based control for the TSWG process.