Density functional theory has become the workhorse for simulations of catalytic reactions and computational design of novel catalysis. The generally applied semi-local exchange-correlation functionals however generally failed for strong correlated systems such as oxides and molecule/transition metal interface, showing systematic and non-canceling errors. Recently, the use of random phase approximation (RPA) for the correlation energy, combined with the exact, self-interaction free exchange energy, offers a promising non-empirical way for accurately describe the adsorption and reaction energies as well as dispersion forces. We have recently ported the RPA in the GPAW code, an electronic structure package using projector augmented wave method and real space grids, to graphics processing units (GPUs) and gained a speed up of up to 160. In this talk I will present our techniques for the GPU speed up and the applications on metal oxides and molecule/metal interfaces.

Predicting catalytic reactions using random phase approximation

246th National Meeting (2013)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.