High-shear wet granulation is a particle size enlargement process widely used in industries such as the pharmaceutical, food and agricultural industries. Despite its predominance, knowledge on several of the key mechanisms of granulation is lacking, driving up the costs of the design of new processes, scale-up and control. For the rate mechanism of consolidation and layered growth, models can be found in literature, one of which has been validated for the case of growth in a static powder bed. However, these models remain to be experimentally validated for application in an actual granulator. This study is the first to develop a method to investigate consolidation and layered growth under such dynamic conditions, and presents a detailed investigation of the kinetics, as well as a model to describe them.
Initially, a high-shear mixer with three-bladed impeller was used for the method development of the study of the kinetics of consolidation and layered growth. However, experiments quickly showed that a dedicated piece of equipment was needed in order to isolate consolidation and growth from the other granulation phenomena, breakage in particular. Therefore, a novel, consolidation-only granulator (COG) was designed. Using the COG, the growth kinetics of a variety of powder-binder systems was evaluated. The granule mass was found to increase linearly with the square root of time, until a critical-packing liquid volume fraction had been achieved. This behaviour corresponds with surface tension-driven growth models. However, breakage and attrition were found to be prevalent for long granulation times, making it impossible to determine both the critical-packing liquid volume fraction and the final granulation time. Additionally, an unexpected rapid increase in initial granule mass was observed. Remarkably, the overall granule porosity was found to be constant. Tomography revealed that the granules developed a core-shell structure, with the higher-density shell becoming increasingly thick during granulation and the core becoming less dense.
A further study using a high-shear mixer with flat plate impeller was successfully performed to determine the critical-packing liquid volume fraction and final granulation time. Although the qualitative kinetic behaviour was found to match that predicted by the surface tension-driven growth model, the quantitative behaviour varied. Efforts to incorporate the observed core-shell structure into the existing model revealed that such an extension did not represent the observed behaviour. As such, no predictive expression was found for the critical-packing liquid volume fraction and final granulation time. However, these parameters can be obtained from experimental work. Finally, the initial rapid increase in granule was addressed, and it was deemed probable that this effect would not have a significant impact on in-situ nucleation in a granulator.
Finally, the results from all the studies were combined to adjust the existing model and convert it into two different population balance models (PBMs). The first, a three-dimensional PBM, was simply proposed. The second, a one-dimensional PBM, was solved by discretisation and compared to the experimental results to evaluate its performance. It was found that the discretisation method showed some deviation from the experimental results, but that this error could be reduced by decreasing the bin width.
This work has successfully identified the underlying kinetics of layered growth, elucidated the consolidation and growth behaviour of granules, and contributed to the modelling and design of granulation processes.
CORE (COnnecting REpositories)
CORE (COnnecting REpositories)