A Bayesian Optimization Strategy for Mechanochemical Nickel-Catalyzed Negishi Cross-Coupling Reaction

Bayesian optimization (BO) has emerged as a powerful strategy to accelerate reaction development by efficiently exploring large experimental spaces through probabilistic models¹. In parallel, mechanochemistry offers a sustainable approach for catalytic transformations under solvent-minimized ball-milling conditions². Within this context, Negishi cross-coupling remains as a valuable method for C–C bond formation, motivating the integration of BO with mechanochemical nickel catalysis³. Inspired by the work of Browne⁴, we sought to expand this concept through data-driven optimization by replacing Pd catalysis with a Ni based mechanochemical system for C(sp³)–C(sp²) Negishi cross-coupling. A large reaction condition space comprising Ni catalysts, ligands, additives, LAG solvents, grinding auxiliaries, and catalyst/ligand loadings were explored using EDBO, with yield defined as the objective. After only eight rounds of six parallel experiments, optimized conditions were identified for a mechanochemical Ni-catalyzed C(sp³)–C(sp²) Negishi cross-coupling, affording the desired product in up to 93% yield