(652b) Modeling of Drug Substance Manufacturing in a Multi-Purpose Plant

This work presents different modeling strategies to address the manufacturing of drug substances in a multi-purpose plant. In multi-purpose plants the process equipment needs to have flexibility to tackle a wide range of chemical and physical operations, often with multi-phasic systems, and a broad range of solvents, temperature and pH. It is not uncommon that such process equipment in multi-purpose plants are not fully adequate to specific highly demanding chemical processes (eg a given impeller may be adequate for intense local mixing, but not fit for axial mixing of solids). Hence, it is crucial to use adequate modeling tools and methodologies to optimize the process performance as much as possible considering the equipment constraints; such strategies are aligned with the Quality by Design initiave being promoted by the regulators.

The case-study to be shown comprises the scale-up of the manufacturing of two distinct drug substances from the laboratory to the same full scale reactor. The kinetics of both processes is assessed in detail in the laboratory scale through the fitting of the transient differential equations to the analytical data (using Dynochem); the intrinsic kinetics of both processes is determined through decoupling of the fluid dynamics portion of the mass transfer resistance in the kinetic constants using both Dynochem and CFD (Computational Fluid Dynamics) - this is made possible since the laboratory reactor geometry, impeller and baffle system are relatively simple. In constrast, the geometry, impeller and baffle system of the full scale reactor are not simple, so this is addressed differently. The fluid dynamics of both processes in the full scale reactor is assessed solely with CFD, which enables the determination of the local mass transfer coefficient distribution. It is then shown how the use of the local mass transfer distribution along with the intrinsic kinetics of each process enables an improved understanding of the scale-up to the full scale.