Pseudohalide (PH) anion engineering has become a surface passivation strategy of interest in the field of perovskite based optoelectronics; But so far, PH anions have led to insufficient passivation of defects, resulting in undesired deep impurity states. So far, the size of the pH anion chemical space (>106 molecules) has limited attempts to explore the entire candidate molecular family.
Edward H. Sargent et al. from the University of Toronto in Canada and Northwestern University in the United States created a machine learning workflow that uses full density functional theory calculations to train models to accelerate the discovery process. A physics based machine learning model enables us to accurately locate promising molecules, with the head group preventing lattice distortion and the formation of anti site defects, and the tail group optimized to firmly adhere to the surface.
Researchers identified 15 potential bifunctional pH anions that can passivate donors and receptors, and found through experiments that sodium mercaptoacetate is the most effective passivator. This strategy achieves a power conversion efficiency of 24.56% and an open circuit voltage of up to 1.19 V for inverted perovskite solar cells (the quasi steady state voltage certified by the National Renewable Energy Laboratory is 24.04%). The packaged device maintained an initial photoelectric conversion efficiency of 96% during a daily operation period of 900 hours at maximum power point.
Xu, J., Chen, H., Grater, L. et al. Anion optimization for bifunctional surface passivation in perovskite solar cells. Nat. Mater. (2023).
https://doi.org/10.1038/s41563-023-01705-y