(129d) CFD Modelling of Drying and Crystalline Transformation in a Co-Current Spray Dryer

Spray drying is widely applied in the pharmaceutical industry for the formulation of particulate materials such as solid dispersions, excipients, active pharmaceutical ingredient (API) as well as co-dried materials (API with excipient). It has a distinct advantage of converting pumpable feed into dry powder in a single unit operation. In spray drying, the atomised droplets typically ranging from 10 to 500 μm in diameter comprise dissolved solute in a solvent. The droplets undergo rapid evaporation of solvent upon contact with turbulent hot gas flow, favouring the formation of amorphous particles. The process is well suited for production of amorphous solid dispersions. However, amorphous solid dispersions can be unstable and the API can transform to crystalline form when exposed to high temperature and humid conditions, which reduces its efficacy. Furthermore, the size distribution of particles, density and moisture content also effect the product performance. By optimising the input operating conditions in the spray dryer and feed composition, it is possible to control the dried powder attributes.

Modelling can be used to assess the impact of feed composition and dryer operating conditions on the size distribution, density, moisture and crystallinity of spray dried particles in a cost effective manner. In this study, modelling of a lab-scale co-current spray drying tower utilising two-fluid nozzle for feed atomisation is carried out to predict dried powder characteristics. The turbulent air flow is modelled using the Computational Fluid Dynamics (CFD) approach and coupled with a detailed droplet drying model to predict drying rate as well as concentration gradients within the droplet/particle. A distributed modelling approach is applied to predict amorphous to crystalline transformation within the particles undergoing drying, utilising the glass transition temperature. The drying and crystalline transformation model is applied to feed containing lactose-water solution and the results are validated with data available in literature. The modelling approach can be applied to an industrial scale tower to predict optimised operating conditions.