Predicting lateral surface interactions through density functional theory: application to oxygen on Rh(100)

First-principle density functional quantum chemical calculations are used to predict the binding energies and lateral interactions of adsorbed atomic oxygen on Rh(100). The DFT results are subsequently used to parameterize two general models for the a priori prediction of adsorbate interactions on surfaces. The first model is based on pairwise interactions where the lateral effects are assumed to be additive. The second model is based on bond-order conservation (BOC) principles. The parameters for these models are determined by performing a series of DFT calculations on Rh cluster and slab models of the Rh(100) surface Rh(8,3), Rh(9,4), and Rh(12,17) clusters are used to determine the O/Rh(100) binding energy and lateral interaction energies for different adsorbate-surface configurations. The O/Rh(100) binding energy on the Rh(100) surface is calculated to be 117 ± 5 kcal/mol. Lateral interactions involving metal atom sharing are repulsive, ranging from 2 to 7 kcal/mol per oxygen on a pairwise additive basis for both clusters as well as extended surfaces. We develop and analyze pairwise additive and bond-order conservation models for use in predicting lateral interaction energies. An unaltered BOC approach is found to overestimate the interaction energies. However, a modified BOC approach in which the basic interactions are scaled against DFT predictions proved to be significantly better in its predictions. Comparison of the modified BOC model with experiments suggests that this is a viable mean for predicting lateral interaction energies.