(725g) Predictive Drying Recipe Development for Drying a Heat-Sensitive and Sticky Powder: Scale-up from Lab to Manufacturing By Modeling and Simulation

Residual solvent content of powdered product from crystallization/filtration or spry drying is a quality aspect that related to CQA and ICH threshold. The residual solvent down to ppm level may cause form change, stability, and impose toxicity or bioavailability alteration. The secondary dryers in types of double cone dryers, agitated mixing dryers, fluidized bed, rotary and so are common unit operations in the industry. The transport phenomena involved in the drying of a power bed is very complex and incorporate the heat and mass transfer and mixing, which all are scale dependent. The scale dependency becomes a challenging problem in scale-up and technology transfer, where hundreds of kilograms of expensive and sensitive material should be processed. Under- or over-processing of the material may cause thermal degradation (impurity formation), agglomeration or breakage (change of PSD), or solvent-mediated phase change. The conventional scale-up method is still experimental trial and frequent sampling at the large manufacturing scale. This is a very risky and expensive [blind] process that is developed based on no knowledge of the material or equipment or process. In contrary, mechanistic modeling can be used to simulate all the involved transport phenomena in a predictive fashion.

This presentation will demonstrate a manufacturing scale drying process development and prediction for an Ekato 1600 SolidMix dryer with the capacity of 800kg, for drying a sticky and heat sensitive material to reduce Methanol and Dichloromethane from percentage level to PPM over 50 hours. The diverse heat transfer models, diffusion and evaporation, inner-bed mixing, stickiness, and residual solvent content trend will be discussed for a range of process parameters scenarios. The models are multiscale, multidomain, and Multiphysics with different levels of fidelity and computation expenses.