Adsorption Energy Correlations at the Metal-Support Boundary

The emergence of supercell density functional theory (DFT) over the past few decades has enabled great progress in rational catalyst design for extended surfaces of transition metals. However, insights from such metal-only models may not translate directly to metal nanoparticles dispersed on high surface area supporting phases, where metal and support may both participate in catalysis. To quantify these differences, we investigate the adsorption behavior of common catalytic intermediates at the boundaries of late transition metals supported on MgO(100). We show that the oxide can either strengthen or weaken adsorption at the metal oxide boundary, depending on the metal adsorbate combination. Using a thermodynamic cycle, we trace the origins of these stabilization/destabilization effects to a combination of multiple structural and electronic perturbations, including strain and ligand effects, geometric reorientation, and charging of the adsorbate. These perturbations in some cases result in departures from the linear scaling relations developed on metal-only models. Computational screening studies based on typical scaling relations may thus miss potential catalyst materials where bifunctional gains are possible.