Automated Self-Optimising Continuous Flow Reactors for the Process Development of Generic Medicines

Continuous flow chemistry is an area of increasing interest in chemical research and development, in part due to its significant benefits over batch chemistry for some applications. Flow boasts improved heat- and mass-transfer, which increases the safety of performing hazardous reactions, and allows safe process intensification with high pressures and temperatures. In an age of digitisation, there are significant additional benefits to continuous flow chemistry over traditional batch chemistry: the ease of automation and optimisation. Automated self-optimising systems can perform hundreds of reactions in days using a combination of algorithms, remotely controllable equipment, and process analytics. This use of machine learning algorithms allows for highly efficient investigation of experimental space and finds optimal reaction conditions for different metrics in a minimal amount of time and wasted material, allowing for a more sustainable and affordable research and development process.
This thesis aims to utilise and develop upon existing continuous flow technologies to increase the efficiency of generic API synthesis. The work herein describes the implementation and testing of an automated optimisation platform in a new laboratory, complete transference of a batch reaction into a novel continuous flow synthesis and subsequent optimisation of the process, a collaborative design and complex optimisation of a two-step mixed homogeneous and heterogeneous reaction with inline gas separation, and the intensification and multi-objective optimisation of an existing flow process to minimise impurities following the methodology for optimisation outlined throughout the thesis.