Particle Breakage in Agitated Drying Conditions

Particle breakage is an undesired phenomenon in the production chain of the pharmaceutical industry as it can strongly impact the product quality of an active pharmaceutical ingredient (API) e.g. dissolution rate and bioavailability. Moreover, most of the API crystals are organic and of high aspect ratio, thus more prone to break in agitation-based drying processes. A novel methodology combining experimental and modelling techniques is presented to predict the breakage of elongated particles in an agitated bed from the process environment and the particles properties. The breakage of glutamic acid crystals in their elongated form (β-LGA) is observed experimentally in lab-scale agitated filter dryer (AFD). Measurements of particle size distribution allows to quantify the fragmentation of crystals throughout agitation in both wet and dry beds. The breakage strength distribution of the β-LGA crystals is assessed introducing a novel 2-point bending method using atomic force microscopy. In total, 52 individual crystals are broken. A shear cell is built using distinct element modelling (DEM) to mimic the stress conditions in an agitated dryer and filled with particles modelled utilising the experimentally measured physical and mechanical properties of the β-LGA crystals. After optimising the simulation setup, a full-factorial study of uniaxial compressions and Couette flows for different normal stresses and particle elongation is performed allowing the calculation of the particles internal stress. Correlations between key variables are examined to understand the behaviour of the particle bed to its mechanical environment, and mathematical models are then created to estimate the particles internal stress and calibrated against simulation data. Lastly, a breakage kernel is built using a probabilistic approach combining the obtained breakage strength of β-LGA crystals and the particles internal stress. The breakage of the crystals in a lab-scale AFD is estimated with a designed population balance model and the results are validated against experiments. The calibrated model is finally used to predict particle breakage for the scale-up of agitation.