Testing the pairwise additive potential approximation using DFT:: Coadsorption of CO and N on Rh(100)

The interaction between adsorbates is a key issue in surface science, because these interaction can influence strongly the properties of chemisorbed species with consequences for the thermodynamics and kinetics of surface processes. The simplest representation of adsorbate-adsorbate interactions is based on the assumption that all interactions are pairwise additive. This approach has been satisfactorily used in the modeling of temperature-programmed desorption (TPD) spectra using both continuum and Monte Carlo methods. However, the energies estimated within the pairwise approximation have never been compared to the energies calculated using density functional theory (DFT) methods. We demonstrate that the pairwise additive potential approximation is indeed a good representation of the adsorbate-adsorbate interactions, and that we do not need to include three-body interactions or higher-order terms to estimate the perturbation of the adsorption energy of an adsorbate by the presence of other coadsorbates. Moreover, we show for the first time how DFT can be used to explain the desorption features that one finds in TPD experiments, thus linking the TPD desorption features with actual microscopic configurations.