Computational Fluid Dynamics (CFD) Simulations of Spray Drying: Linking Drying Parameters with Experimental Aerosolization Performance

Abstract

Purpose The purpose of this study was to develop a new computational fluid dynamics (CFD)-based model of the complex transport and droplet drying kinetics within a laboratory-scale spray dryer, and relate CFD-predicted drying parameters to powder aerosolization metrics from a reference dry powder inhaler (DPI). Methods A CFD model of the Buchi Nano Spray Dryer B-90 was developed that captured spray dryer conditions from a previous experimental study producing excipient enhanced growth powders with L-leucine as a dispersion enhancer. The CFD model accounted for two-way heat and mass transfer coupling between the phases and turbulent flow created by acoustic streaming from the mesh nebulizer. CFD-based drying parameters were averaged across all droplets in each spray dryer case and included droplet time-averaged drying rate (kappa(avg)), maximum instantaneous drying rate (kappa(max)) and precipitation window. Results CFD results highlighted a chaotic drying environment in which time-averaged droplet drying rates (kappa(avg)) for each spray dryer case had high variability with coefficients of variation in the range of 60-70%. Maximum instantaneous droplet drying rates (kappa(max)) were discovered that were two orders of magnitude above time-averaged drying rates. Comparing CFD-predicted drying parameters with experimentally determined mass median aerodynamic diameters (MMAD) and emitted doses (ED) from a reference DPI produced strong linear correlations with coefficients of determination as high as R-2 = 0.98. Conclusions For the spray dryer system and conditions considered, reducing the CFD-predicted maximum drying rate experienced by droplets improved the aerosolization performance (both MMAD and ED) when the powders were aerosolized with a reference DPI.