Quantum Monte Carlo studies of surface adsorptions

Surface adsorption is the first step to the study of surface catalytic reaction. The most common used tool is the Density Functional Theory (DFT) based on exchange-correlation approximations and the accuracy usually has not been checked carefully by highly accurate quantum many-body approaches. We have performed calculations of the surface adsorptions using the state-of-the-art diffusion quantum Monte Carlo (QMC) method to examine the accuracy of LDA and GGA (PBE) functionals in the study of surface adsorptions. The systems examined include the H$_{2}$O and OH adsorptions on various types of surfaces such as NaCl(100), MgO(100), TiO$_{2}$(110), graphene, Si(100)-(2x2) and Al(100). By comparing GGA (PBE) results with DMC, our results indicate that (i) for the H$_{2}$O adsorption, PBE predicts the correct adsorption energies; (ii) for the OH adsorption, PBE has predicted a large over-binding effect except on graphene and Si(100) surfaces. This fact indicates that one needs to be cautious when using DFT to study the surface adsorptions of OH free radical.