(373f) Multi-Layered Modelling Techniques for the Development of Continuous Manufacturing Processes

Donnellan, P., University College Dublin
Jones, R., University College Dublin
Roche, P., Synthesis and Solid State Pharmaceutical Centre (SSPC), School of Chemical & Bioprocess Engineering, University College Dublin
Glennon, B., University College Dublin
Traditionally the majority of active pharmaceutical compounds have been manufactured at large plant scales using batch methodologies. Continuous manufacturing is becoming an increasing attractive alternative however due to apparent benefits such as smaller manufacturing footprints, increased process safety, lower manufacturing costs and increased supply-chain flexibility. Realising all of the above benefits using conventional continuous equipment is however quite challenging and sometimes not achievable. In this project, a new approach to the design of continuous manufacturing platforms is being presented which is not restricted to using such conventional equipment but which is based around the bespoke design and manufacture of all unit operations, optimised to fit the specific process being examined. Processes are quantified in detail using first principle models which in turn are used to inform computational fluid dynamics (CFD) simulations leading to tailored equipment designs which can be manufactured with high precision using metallic 3D printing.

This presentation will focus on a number of case studies which demonstrate the linking of these various modelling techniques for several different unit operations required for the continuous manufacturing of pharmaceutical compounds including liquid-liquid separations, biphasic reactors and homogenous reactors. In particular however, the talk will focus on the development of a novel distillation column methodology, employing an inert gas to enable the concentration of solvent streams at low temperatures (required by temperature sensitive APIs). First principles theoretical models were used in order model the behaviour of such gas-liquid systems in bubble columns, enabling verification with experimental data. However as a bubble column does not represent the optimum design of such a gas-liquid contacting system, CFD was then employed to examine possible designs concepts suitable for 3D printing and to quantify evaporation rates in such possible configurations.

Following the modelling processes, these unit operations are 3D printed from stainless steel to enable their use as part of manufacturing processes. This approach leads to significant process intensification as all unit operations are designed specifically for the process being examined and are hence manufactured to ensure the process’s optimum performance under the required conditions. Utilising 3D printing informed by process models to manufacture the process train also leads to significant cost reductions as unit operations become essentially disposable, reducing both the upfront capital costs and also essentially eliminating ongoing cleaning costs.

Several results demonstrating this approach are presented, including a biphasic reactor which is able to reduce the required reaction residence time by approximately 100 fold compared to an agitated batch reactor and a continuous homogenous reactor with dimensions of 15cm x 7cm x 3cm which is capable of achieving a throughput of 580,000 kg/L/day, representative of manufacturing output levels for many modern pharmaceutical compounds.