Predicting reactive graphene edge-adsorbed Pt single-atom catalyst structures for oxygen reduction via machine learning-enhanced DFT calculations

Single-atom platinum on graphitic materials can enhance catalysis of the oxygen reduction reaction (ORR). The local atomic structure shapes catalytic activity, but is not yet sufficiently understood due to its complexity. In this study, a machine learning-enhanced global structure search algorithm was combined with density functional theory calculations to explore previously unknown configurations of single-atom platinum adsorbed at graphene edges. Here, some of the novel stable structures demonstrated superior ORR catalytic performance compared to the conventional Pt(111) surface. This article is part of the theme issue 'Surfaces, interfaces and heterogeneous catalysis'.