Nonadiabatic rare events from transition-path sampling of MASH trajectories

Rare nonadiabatic reactions are a key component of many important molecular processes, but are challenging to capture with direct dynamical simulations. In this paper, we combine our recently developed mapping approach to surface hopping (MASH) with transition-path sampling to create a framework to efficiently simulate these rare events. This is possible because MASH trajectories are Markovian, time-reversible, and obey Liouville's theorem. The combined approach generates nonadiabatic reactive pathways without biasing the underlying dynamics. The resulting ensemble allows for a detailed analysis of reaction mechanisms and the unraveling of statistical and dynamical properties, including rate constants. We apply the method to study a spin-boson model in thermal equilibrium over a wide range of diabatic coupling strengths. Our results demonstrate how this approach provides a practical and systematic tool for investigating rare nonadiabatic processes, potentially beyond the reach of brute-force simulations.