First-principles study of benzene on noble metal surfaces: Adsorption states and vacuum level shifts

We have studied the interaction of benzene with Cu(1 1 1), Ag(1 1 1) and Au(1 1 1) surfaces using density functional theory (DFT) within a generalized gradient approximation (GGA) and the van der Waals density functional vdW-DF; M. Dion, H. Rydberg, E. Schröder, D.C. Langreth, B.I. Lundqvist, Phys. Rev. Lett. 92 (2004) 246401]. The adsorption energies using vdW-DF are significantly more accurate than those using GGA, while the equilibrium adsorption distances between benzene and metal substrates () calculated by both GGA and vdW-DF are almost identical. The work function changes induced by the adsorption of benzene are significantly underestimated compared with the experimental values, as a result of the overestimation of by both GGA and vdW-DF. Instead of determining the values from first-principles calculations, we deduced the most probable adsorption distances in such a way as to reproduce the experimentally-observed work function changes. The deduced adsorption distance () is shortest on Cu(1 1 1) while it is longest on Ag(1 1 1), reflecting the strength of the interactions between benzene and the metal surfaces. It turns out that the substrate dependence of the work function change is mainly ascribed to the difference in the benzene–metal distance (Z_C). Charge transfer and work function changes by the adsorption of benzene were analyzed by means of the induced density of interface states (IDIS) model H. Vázquez, R. Qszwaldowski, P. Pou, J. Ortega, R. Pérez, F. Flores, A. Kahn, Europhys. Lett. 65 (2004) 802], and compared with the self-consistent GGA calculations. The vacuum level shifts estimated by the IDIS model agree with the GGA results for . On the other hand, the discrepancy between the two methods becomes larger for , where the back donation from the metal substrates to the adsorbate becomes significant. We show that the IDIS model reasonably works well for benzene on Cu(1 1 1), Ag(1 1 1) and Au(1 1 1) surfaces because on all surfaces. However, our analysis reveals that the actual charge density redistribution induced by the adsorption of benzene is more complicated than that assumed in the IDIS model.