Adsorption energies and structural/electronic descriptors of *N₂, N₂H and NH₃ on 47 intermetallic compounds from high-throughput DFT

This dataset contains DFT-calculated adsorption energies and structural, atomic and electronic descriptors for key nitrogen reduction reaction (NRR) intermediates (*N₂, *N₂H and *NH₃) on 47 bimetallic intermetallic compounds (IMCs) composed of Co, Ni, Al, Zn, V, Fe, Cu and Pt. It accompanies the article "Going Beyond the d-band Center to Design Intermetallic Catalysts for Nitrogen Reduction" (Agharezaei and Ghuman), which uses these data to train interpretable machine learning models of adsorption energy. The dataset includes 1,268 relaxed adsorption configurations, one per adsorbate placed on one surface site of an IMC, across the following files: df_N2.csv: 392 configurations df_N2H.csv: 426 configurations df_NH3.csv: 450 configurations For each configuration, the files provide: the adsorption energy (eV) orientation of adsorbate (only N₂ and N₂H) on the surface (ebd-on or side-on adsorption) upon relaxation the bulk crystal system and space group intrinsic atomic properties, Bader charges and distances of the 21 surface atoms nearest the adsorbate s-, p- and d-projected DOS band centres and integrated band sums (band fillings), both local and for the whole structure All calculated configurations are included, including those where the adsorbate dissociated or desorbed. The machine learning models in the article were trained only on intact, adsorbed configurations. README.md describes every column, and the units. DFT settings are given in the Methods section of the associated article. Please cite both the article and this dataset when using these data.