Tailoring API Physical Properties to Provide for Continuous Manufacturing of Drug Product and Flexibility of Dosing in a Spray dryer

Abstract

A mechanistic model was developed to predict the drying history of solution droplets as they dry in a spray dryer. The model combines equations that describes the liquid and gas phases, and couples this with a population balance equation to describe the particle formation process. In the first part of this work, a carefully designed experimental system based on single droplet drying acoustic levitation was developed. This was used to evaporate solution droplets under controlled conditions including relative saturation of the gas phase and temperature. During the drying process its speed was monitored using a camera and automated image analysis. The final dried product morphology, solid form (polymorph) and microstructure was then characterised with a variety of analytical techniques. The type of formulation, concentration of the components and temperature were found to affect the size, shape, roughness, crystallinity and porosity of the particles to varying degrees. Relevant correlations were developed to relate the formulation and drying condition to the particle properties. In the second part, the model was implemented and the simulated result of the drying behaviour was compared to experiments for drying up to the point of shell formation. The model was applied to formulations containing varying proportions of crystalline solute (D-mannitol), excipient (PVP) and solvent and dry- ing conditions (temperature and relative saturation) and the drying history, the time of shell formation and particle size matches well with experiments. In the third part, the model was extended to describe drying following shell formation. The moisture profile during the drying of the particle was predicted. The model further predicts the drying of a particle to either a solid rigid porous or hollow particle, and also the porosity profile of the particle. The predicted results matched well with those ob- served from experiments. Overall, the model demonstrates excellent predictability for the model systems and range of conditions considered.

Date of Award	1 Aug 2021
Original language	English
Awarding Institution	The University of Manchester
Supervisor	Roger Davey (Supervisor) & Thomas Vetter (Supervisor)