Metal-Coordinating Single-Boron Sites Confined in Antiperovskite Borides for N-2-to-NH3 Catalytic Conversion

The concept of single-atom catalysts (SACs) takes our understanding of active sites of heterogeneous catalysts to greater heights. 'While the knowledge about the SACs containing isolated metal centers has rapidly enriched over the last decade, a few examples of the SACs that feature single-nonmetal centers are studied, and their application potentials are also poorly explored. Herein, we report a family of 358 model surfaces featuring metal-coordinated single-boron sites for investigating the effects of boron-centered local environment on N-2-to-NH3 catalytic conversion. We demonstrate these model surfaces with (001) surfaces of antiperovskite borides (M3M'B, M, and M' represent different metal atoms), where the surface-isolated boron atoms are spacially separated by M atoms, but they do not directly bond to M' atoms. The surface-isolated boron atoms are found to be necessary for the adsorption and activation of dinitrogen, and the types of M atoms determine dinitrogen adsorption configurations. When M atoms are the late transition metals (e.g., Ni), dinitrogen prefers to adsorb on the top-site of the surface B atom with an end-on configuration; when M atoms are alkaline-earth metals, early transition metals and lanthanide elements (e.g., Sr, Zr, and La) and dinitrogen bonds to M-B-M three-center site with a side-on configuration. Additionally, during the catalytic conversion of dinitrogen to ammonia, the M atoms cooperate with the isolated boron atom to costabilize NxHy intermediates (x = 1, 2; y = 0-4), and the M' atoms finely regulate the electronic structure of B - M ensembles and thereby tune the adsorption property and catalytic activity. Furthermore, the adsorption free energy of *N is found to be applied as a descriptor of the limiting potential of dinitrogen reduction reaction, and its value is correlated with orbital hybridization between M and B atoms apart from the inherent electronic properties of M and B atoms. Finally, we predict Rh3GeB, Rh3SbB, and Ni3LiB to be promising ammonia synthesis catalysts with high activity that exceed that of the stepped Ru(0001) surface, a prominent benchmarking surface for ammonia synthesis.