Machine-Learning Modeling for Ultra-Stable High-Efficiency Perovskite Solar Cells

摘要

Understanding the key factor driving the efficiency and stability of semiconductordevices is vital. To date, the key factor influencing the long-term stabilityof perovskite solar cells (PSCs) remains unknown because of the manyinfluencing factors. In this work, through machine learning, the influencesof five factors, including grain size, defect density, bandgap, fluorescencelifetime, and surface roughness, on the efficiency and stability of PSCs havebeen revealed. It is found that the bandgap has the greatest influence on theefficiency, and the surface roughness and grain size are most influential to thelong-term stability. A mathematical model is given to predict efficiency basedon fluorescence lifetime and bandgap. Guided by the model, four groups ofexperiments are conducted to confirm the machine-learning predictions anda PSC with 23.4% efficiency and excellent long-term stability is obtained, asmanifested by retention of 97.6% of the initial efficiency after 3288 h aging inthe ambient environment, the best stability under these conditions. This workshows that machine learning is an effective means to enrich semiconductorphysical models.