Encouraging the Exploration of Nickel Catalysis in Pharmaceutical Manufacturing Processes: Lessons Learned during Catalyst Screening and Process Optimization for the Monoarylation of Ammonia

Abstract Reports of new ligands or precatalysts based on non-platinum group metal (pgm) catalysis, as well as the number of publications describing new methodology that relies on non-pgm catalysts, are at an all-time high. Catalyst screening groups and process chemists are now beginning to seriously investigate and evaluate non-pgm-based processes when carrying out route scouting and process optimization. The million-dollar question is how many already established manufacturing processes and how many launched drugs may have benefited from the use of non-pgm catalysis, had it been available and considered at the time of process optimization and launch? We wanted to explore one such example and identified a previously reported palladium-catalyzed selective formal monoarylation of ammonia, from the process used to make abemaciblib, as a suitable model reaction. This article describes our lessons learned from evaluating this specific reaction in a high-throughput screening study to identify a suitable nickel-based catalyst for the monoarylation of ammonia (NH3), process optimization, and product isolation attempts. The preformed complex (NHP-DalPhos)Ni(o-tol)Cl) Cat5 was identified as the best-performing catalyst and provided the desired primary amine 2 in 82% conversion. The DalPhos family of ligands has been proven to provide good reactivity in cases where other ligands have failed, and the reason behind this has so far only been partially understood. Crucially, as part of our study, a structure-activity relationship (SAR) model based on three distinct catalyst descriptors (buried volume, P–Ni–P bond angle, and electron richness) was uncovered. This has previously not been described for nickel catalysis while being known in broad terms for palladium catalysis and could fill a gap as an up-until-now missing insight for the successful implementation of nickel-catalyzed reactions. Through this model, we achieved a better understanding of the ligand parameters that proved to be important for a highly active and selective catalyst. Overall, we expect that by sharing our lessons learned, our story can act as a trigger to encourage more researchers to evaluate nickel catalysis alongside palladium while avoiding having to tackle the same issues we encountered, and that the SAR model can provide a useful tool to aid and accelerate the identification of the most effective nickel catalyst by maximizing the coverage of chemical space.