Automated Optimization of a Multistep, Multiphase Continuous Flow Process for Pharmaceutical Synthesis.

Flow synthesis is becoming increasingly relevant as a sustainable and safe alternative to traditional batch processes, as reaction conditions that are not usually achievable in batch chemistry can be exploited (for example, higher temperatures and pressures). Telescoped continuous reactions have the potential to reduce waste by decreasing the number of separate unit operations (e.g., crystallization, filtration, washing, and drying), increase safety due to limiting operator interaction with potentially harmful materials that can be reacted in subsequent steps, minimize supply chain disruption, and reduce the need to store large inventories of intermediates as they can be synthesized on demand. Optimization of these flow processes leads to further efficiency when exploring new reactions, as with a higher yield comes higher purity, reduced waste, and a greener synthesis. This project explored a two-step process consisting of a three-phase heterogeneously catalyzed hydrogenation followed by a homogeneous amidation reaction. The steps were optimized individually and as a multistep telescoped process for yield using remote automated control via a Bayesian optimization algorithm and HPLC analysis to assess the performance of a reaction for a given set of experimental conditions. 2-MeTHF was selected as a green solvent throughout the process, and the heterogeneous step provided good atom economy due to the use of pure hydrogen gas as a reagent. This research highlights the benefits of using multistage automated optimization in the development of pharmaceutical syntheses. The combination of telescoping and optimization with automation allows for swift investigation of synthetic processes in a minimum number of experiments, leading to a reduction in the number of experiments performed and a large reduction in process mass intensity values.

Bayesian Self-Optimization for Telescoped Continuous Flow Synthesis.

The optimization of multistep chemical syntheses is critical for the rapid development of new pharmaceuticals. However, concatenating individually optimized reactions can lead to inefficient multistep syntheses, owing to chemical interdependencies between the steps. Herein, we develop an automated continuous flow platform for the simultaneous optimization of telescoped reactions. Our approach is applied to a Heck cyclization-deprotection reaction sequence, used in the synthesis of a precursor for 1-methyltetrahydroisoquinoline C5 functionalization. A simple method for multipoint sampling with a single online HPLC instrument was designed, enabling accurate quantification of each reaction, and an in-depth understanding of the reaction pathways. Notably, integration of Bayesian optimization techniques identified an 81 % overall yield in just 14 h, and revealed a favorable competing pathway for formation of the desired product.

Automated stopped-flow library synthesis for rapid optimisation and machine learning directed experimentation.

For the discovery of new candidate molecules in the pharmaceutical industry, library synthesis is a critical step, in which library size, diversity, and time to synthesise are fundamental. In this work we propose stopped-flow synthesis as an intermediate alternative to traditional batch and flow chemistry approaches, suited for small molecule pharmaceutical discovery. This method exploits the advantages of both techniques enabling automated experimentation with access to high pressures and temperatures; flexibility of reaction times, with minimal use of reagents (µmol scale per reaction). In this study, we integrate a stopped-flow reactor into a high-throughput continuous platform designed for the synthesis of combinatory libraries with at-line reaction analysis. This approach allowed ∼900 reactions to be conducted in an accelerated timeframe (192 hours). The stopped flow approach used ∼10% of the reactants and solvents compared to a fully continuous approach. This methodology demonstrates a significantly improved synthesis success rate of smaller libraries by simplifying the implementation of cross-reaction optimisation strategies. The experimental datasets were used to train a feed-forward neural network (FFNN) model providing a framework to guide further experiments, which showed good model predictability and success when tested against an external set with fewer experiments. As a result, this work demonstrates that combining experimental automation with machine learning strategies can deliver optimised analyses and enhanced predictions, enabling more efficient drug discovery investigations across the design, make, test and analysis (DMTA) cycle.

A Hybridised Optimisation of an Automated Photochemical Continuous Flow Reactor.

A new hybridized algorithm that combines process optimisation with response surface mapping was developed and applied in an automated continuous flow reaction. Moreover, a photochemical cascade CSTR was developed and characterised by chemical actinometry, showing photon flux density of ten times greater than previously reported in batch. The success of the algorithm was then evaluated in the aerobic oxidation of sp³ C-H bonds using benzophenone as photosensitizer in the newly developed photo reactor.

Kinetic Treatments for Catalyst Activation and Deactivation Processes based on Variable Time Normalization Analysis.

Progress reaction profiles are affected by both catalyst activation and deactivation processes occurring alongside the main reaction. These processes complicate the kinetic analysis of reactions, often directing researchers toward incorrect conclusions. We report the application of two kinetic treatments, based on variable time normalization analysis, to reactions involving catalyst activation and deactivation processes. The first kinetic treatment allows the removal of induction periods or the effect of rate perturbations associated with catalyst deactivation from kinetic profiles when the quantity of active catalyst can be measured. The second treatment allows the estimation of the activation or deactivation profile of the catalyst when the order of the reactants for the main reaction is known. Both treatments facilitate kinetic analysis of reactions suffering catalyst activation or deactivation processes.

Catalytic sp3 -sp3 Functionalisation of Sulfonamides: Late-Stage Modification of Drug-Like Molecules.

A new application of Pd-catalysed allylation is reported that enables the synthesis of a range of branched sp3 -functionalised sulfonamides, a compound class for which few reported methods exist. By reacting benzyl sulfonamides with allylic acetates in the presence of Pd0 catalysts and base at room temperature, direct allylation was efficiently performed, yielding products that are analogues of structural motifs seen in biologically active small molecules. The reaction was performed under mild conditions and could be applied to nanomolar sigma-receptor binders, thus enabling a late-stage functionalisation and efficient expansion of drug-like chemical space.