Autonomous Development of Telescoped Catalytic Reactions

Multistep continuous flow processes are an example of uninterrupted reaction networks. They have a low space-time demand, as large inventories of intermediates do not need to be stored and transported between different manufacturing sites. This minimises the risk of supply chain disruptions, enabling reliable on-demand synthesis with a reduced ecological footprint. Catalysis plays a significant role in many of these transformations, offering more efficient and greener processing routes. However, telescoped catalytic reactions are complex and expensive to optimise. Automated optimisation could overcome this limitation by using machine learning techniques to efficiently explore the design space, however these systems are currently limited to simpler single step processes. Research at this interdisciplinary interface has thus far been limited by a lack of expertise in digital technology for chemical synthesis, combined with high equipment cost. The Institute of Process Research and Development (iPRD) at the University of Leeds houses state-of-the-art continuous flow technology. Using this facility, we will develop new multistep optimisation platforms for telescoping catalytic reactions. This will be achieved by incorporating multi-point on-line sampling to enable simultaneous optimisation of consecutive reaction steps. Different reactor configurations will be developed to facilitate the combination of different types of catalysis. For example, immobilised catalysts (e.g. chemical, enzymes) will be spatially separated in different reactors to overcome compatibility issues. Furthermore, in alignment with the drive for more sustainable processes, the short path length of meso- and microscale reactors will be used to enable inclusion of photocatalysed transformations. Software will also be developed to incorporate new features such as multi-point sampling and different reactor configurations, which will be distributed to collaborators to facilitate the industrial uptake of this technology. In summary, this research will develop new digital technology to enable the efficient automated optimisation of telescoped catalytic processes, thus streamlining the research and development timeline. 

Publications and outputs

Dixon, T. M., Williams, J., Besenhard M. O., Howard, R., Macgregor, J., Peach, P., … Bourne, R. (2024). Operator-free HPLC automated method development guided by Bayesian optimization. Digital Discovery.
doi: 10.1039/D4DD00062E

Arshad, Z., Blacker, A. J., Chamberlain, T. W., Kapur, N., Clayton, A. D., & Bourne, R. A. (2024) Droplet microfluidic flow platforms for automated reaction screening and optimisation. Current Opinion in Green and Sustainable Chemistry. doi:10.1016/j.cogsc.2024.100940

Aldulaijan, N., Marsden, J. A., Manson, J. A., & Clayton, A. D. (2024). Adaptive Mixed Variable Bayesian Self-Optimisation of Catalytic Reactions. Reaction Chemistry & Engineering. doi:10.1039/D3RE00476G
Clayton, A. D. (2023). Recent Developments in Reactor Automation for Multistep Chemical Synthesis. Chemistry-Methods. doi:10.1002/cmtd.202300021

Clayton, A. D. (2023). Recent Developments in Reactor Automation for Multistep Chemical Synthesis. Chemistry-Methods. doi:10.1002/cmtd.202300021

Clayton, A. D., Pyzer-Knapp, E., Purdie, M., Jones, M., Barthelme, A., Pavey, J., . . . Bourne, R. (2023). Bayesian Self-Optimization for Telescoped Continuous Flow Synthesis. Angewandte Chemie International Edition, 62(3). doi:10.1002/anie.202214511