Simulation of surface processes using machine-learned potentials

Machine-learned models for atomistic simulations enable evermore applications in physics, chemistry, and material science, owing to the simulations' unparalleled accuracy/cost ratio. Our project aims to merge our recent developments in machine-learned potentials, surface science, and astrochemistry to tackle the question of the origin of organic chemistry in interstellar environments. Data science techniques will allow us to simulate the onset of organic chemistry on top of interstellar ices in star-forming regions in atomistic resolution. By the end of the project, we will be closer to determining the fate of organic molecules synthesized in space. Moreover, we will have the scientific tools to investigate surface processes in different fields of materials science.