(252d) Application of Fundamental Relationships and Models to Predict Spray-Dried Dispersion Particle Size

Atomization models can assist spray drying formulators in estimating ranges of droplet size that may be achieved for a given spray solution formulation and drying process. Droplet size can then be related to the final spray-dried dispersion (SDD) particle size by accounting for the fundamentals of droplet solidification and final particle density [1]. Specifically, for film-forming polymeric SDD systems, droplet drying kinetics and thermodynamics must also be considered due to their resulting particle morphological variations [2]. By combining these fundamental relationships with empirical correlations, a rational process development strategy can be employed early in the product lifecycle and give initial insight into future scale-up constraints or risks. Using basic solution properties as model inputs, multiple empirical approaches to predicting droplet size were evaluated. This study showed excellent correlation of predicted droplet size to the resulting final SDD particle size. In addition, use of a simplified porous particle approach that relies on the viscous gel point to predict particle size from droplet size was effectively demonstrated.

Methods:

Viscosity, surface tension and density were evaluated for a range of solids concentrations using two different dispersion polymers, HPMCAS-M (1-12 wt%) and PVP/VA (1-20 wt%) prepared in either acetone or methanol. Additionally, the viscous gel point (skinning concentration) was determined using a gravimetric method. Droplet size was measured sing a phase Doppler particle analyzer (PDPA) to study the effects of solids composition, nozzle geometry, and atomization parameters. Spray-drying experiments were then conducted at lab and feasibility scale to obtain final SDD particle size and evaluate potential impact of this approach on scale-up. SDDs were characterized for particle size distribution using light scattering methods, bulk/tap density, and morphology by scanning electron microscopy.

Results:

Droplet size predictions were conducted using an empirical model developed by Lefebvre:

These predictions correlated well with measured droplet size (PDPA) and measured SDD particle sizes (light scattering) for the formulations studied [3]. HPMCAS-M showed a steeper response of particle size to droplet size than PVP/VA possibly due to higher viscosities observed in solution at similar concentrations. Differences were observed between the relationship of predicted droplet size and SDD particle size based on the spray solvent used for both PVP/VA and HPMCAS-M polymers. A shrinking sphere model was initially applied to understand the relationship between droplet size and final particle size however, this approach was not found to demonstrate the desired predictability. Therefore, a porous particle approach was evaluated that accounts for the differences in skinning concentration of the polymers in various solvent systems. This was found to significantly improve the predictability of final SDD particle size.

Conclusions:

Application of simple models and fundamental relationships was shown to provide a high level of accuracy for predicting both droplet size and final SDD particle size. The utility of this approach can be further expanded to aid in scale-up predictions to rationally choose a droplet size to target a desired SDD particle size during initial process development and optimization on lab-scale spray dryers. Key differences between the dispersion polymer or solvent chosen were explained through either solution properties or relationship of polymer solidification.

[1] Singh, A. and Van den Mooter, G. Spray drying formulation of amorphous solid dispersions. Adv. Drug Deliv.

Rev. 2015.

[2] Vehring R. Pharmaceutical particle engineering via spray drying. Pharmaceut Res. 2008;25(5):999-1022.

[3] Lefevbre, Arthur. Atomization and Sprays. CRC Press, 1988.