Identification of the dominant recombination process for perovskite solar cells based on machine learning

Over the past decade, perovskite solar cells have become one of major research interests photovoltaic community, and they are now on brink catching up with classical inorganic cells, efficiency reaching to 25%. However, significant improvements still achievable by reducing recombination losses. The aim this work is develop a fast easy-to-use tool pinpoint main losses in cells. We use large-scale drift-diffusion simulations get better understanding light intensity dependence open-circuit voltage how it correlates dominant process. introduce an automated identification using machine learning methods loss intensity-dependent performances as input. was trained >2 million gives accuracy prediction 82%.