Lorenz Romanera, Dmitrii Naboka, Peter Puschniga, Egbert Zojerb, Matthias Schefflerc, and Claudia Ambrosch-Draxla
aChair of Atomistic Modelling and Design of Materials, University of Leoben, Franz-Josef-Strasse 18, A-8700 Leoben, Austria.
bInstitute of Solid State Physics, Graz University of Technology, Petersgasse 16, A-8010 Graz, Austria.
cFritz-Haber Institut der Max-Planck Gesellschaft, Faradayweg 446, B-1495 Berlin-Dahlem, Germany.
Extending density functional theory (DFT) to also describe van der Waals (vdW) interactions is an urgently needed and challenging task for current computational materials science. Very promising approaches have been recently developed and we focus on the method developed by the groups of Lundqvist and Langreth, called vdW-DF, which has been applied to a variety of systems such as graphene sheets, molecular dimers, or molecules adsorbed on surfaces. The typical problem of the generalized gradient approximation, which yields negligible binding energies in such systems, can be largely removed. However, a closer look reveals that there is still room for improvements, as a systematic overestimation of binding distances is observed. While in many cases the error is not significant, we present evidence that the method fails more severely for a certain class of systems where the interatomic bonding is of mixed character. We focus on the bulk coinage metals and hybrid systems where a 3,4,9,10-perylene-tetracarboxylic acid dianhydride (PTCDA) monolayer is adsorbed onto the respective (111) surfaces. We show, that already for the bulk metals, vdW interactions give a substantial cohesive contribution, and that the standard vdW-DF method fails in yielding a good description of basic properties such as the lattice constant and cohesive energies. Also for PTCDA on the coinage metals a strong overestimation of binding distances is observed. We discuss possible reasons for these discrepancies and identify a way to correct for the shortcomings.
Author entry (protected)