A Compartmental Population Balance Modelling of Continuous Twin Screw Wet Granulation Using Mechanistic Process Kernels

Wet granulation is a critical process to create larger stable granules from fine powders in pharmaceutical manufacturing. Twin screw granulation (TSG) becomes increasingly popular as a method of wet granulation in the pharmaceutical industry as part of continuous manufacturing flowsheet due to its customized configuration, short residence time, relative insensitivity to changes in formulation properties and ability to produce porous granules suitable for downstream compression. TSG is very different from traditional batch high shear wet granulation (HSWG) and there is evidence that the balance of rate processes within is quite different. For example, breakage of large wet granules largely controlled by screw element geometry, and subsequent layering of unwet fines is a very important process in the TSG. Thus, rate expressions used in population balance models (PBM) for HSWG, or even for ball mills, may not be applicable to the TSG. This paper presents a compartmental population balance modelling of twin screw wet granulation with a mechanistic understanding of the rate processing kernels. To develop and validate the rate expressions, we use experiments explicitly designed to study a single rate process in a single type of screw element. As granule breakage is a key mechanism in controlling the granule size, breakage-isolated experiments are carried out in the conveying and distributive mixing elements. The dominant breakage mechanism in the conveying and distributive mixing elements is chipping and fragmentation respectively. Based on the breakage results in different elements, the breakage kernels including the selection function and the breakage function are mathematically formed. Other rate process kernels were also clearly specified in each compartment via mechanistic approach. Model results from the compartment based PBM are compared with experimental results for TSG of a model formulation over a range of liquid to solid ratios. Avenues to improve the predictive capacity by coupling PBM with DEM (Discrete element method) are also discussed.