Machine-Learning-Assisted\nDesign of Buried-Interface\nEngineering Materials for High-Efficiency and Stable Perovskite Solar\nCells

Buried-interface engineering is crucial to the performance\nof perovskite\nsolar cells. Self-assembled monolayers and buffer layers at the buried\ninterface can optimize charge transfer and reduce recombination losses.\nHowever, the complex mechanisms and the difficulty in selecting suitable\nfunctional groups pose great challenges. Machine learning (ML) offers\na powerful tool for screening and identifying effective structures\nfor interface modification. Our ML-driven approach led to the preparation\nof two promising organic molecules, PAPzO and PAPz, which exhibit\nsynergistic interactions with SnO<sub>2</sub> and perovskites. These\nmolecules decrease charge trap densities, elongate carrier lifetimes,\nand retard perovskite crystallization. PAPzO, with a stronger binding\nenergy and better aligned energy levels, enables a power conversion\nefficiency (PCE) of 26.04% and long-term stability, maintaining 91.24%\nof its original PCE after 1,200 h of continuous maximum power point\ntracking. This ML-integrated approach marks a significant advancement\nin the development of efficient and stable perovskite photovoltaics.