Quantum Monte Carlo prerequisites for industrial catalysts. Accurately assessing H atom and H$_2$ physical adsorption energy on Pt(111) (1806.05068v1)

Abstract: The yardstick of new first-principles approaches to key points on reaction paths at metal surfaces is chemical accuracy compared to reliable experiment. By this we mean that such values as the activation barrier are required to within 1 kcal/mol. Quantum Monte Carlo (QMC) is a promising (albeit lengthy) first-principles method for this and we are now beyond the dawn of QMC benchmarks for these systems, since hydrogen dissociation on Cu(111) has been studied with quite adequate accuracy in two improving QMC studies \cite{hog0, kdd2}. Pt and Cu require the use of pseudo-potentials in these calculations and we show that those of Pt are less problematic than those for Cu, particularly for QMC work. This work determines physisorption energies for hydrogen atoms and molecules on Pt(111). These systems are used as asymptotes to determine reaction barriers. As such, they must be referred to clean Pt(111) surfaces and the isolated hydrogen atom or molecule. Previous work gave the activation barrier to hydrogen dissociation on Pt(111) using the bridging geometry. The method used is state-of-the-art ({\it ab initio}) QMC. This barrier agrees to better than chemical accuracy with a recent Specific Reaction Parameter (SRP-DFT) calculation, fitted to measurement \cite{iran}. The results give the dissociation barrier for hydrogen on Pt(111) as 5.4 kcal/mol (QMC and 6.2 by SRP-DFT) with a QMC standard error below 0.99 kcal/mol. This is encouraging for establishing less well-known benchmark values of industrial reaction barriers on Pt(111).