Ethylene on Cu(1 1 0) and Ni(1 1 0): electronic structure and bonding derived from X-ray spectroscopy and theory

The bonding of ethylene to Cu(1 1 0) and Ni(1 1 0) is analyzed in detail using symmetry-resolved X-ray absorption (XAS) and emission (XES) spectroscopies in conjunction with density functional theory (DFT) calculations of geometric structure and spectra. XES, which probes the occupied valence states, reveals the formation of bonding and non-bonding orbitals of π-3d as well as π∗-3d character. Additional mixing of σ and π states indicates rehybridization upon adsorption. The anti-bonding π-3d and π∗-3d combinations are unoccupied and seen in XAS. A lower intensity of the π∗ transition for Ni is evidence of larger π∗ occupancy upon bonding. The position of the σ∗ shape-resonance indicates a 0.02 Å longer C–C bond on Ni than on Cu, in good agreement with the DFT structure optimizations. The XE spectra are well-reproduced both by specific spectrum calculations based on cluster models and by the carbon p-density of states calculated using periodic boundary conditions. The contribution of both π and π∗ levels to the new, surface-induced occupied states close to the Fermi-level lends support to the traditional Dewar–Chatt–Duncanson picture of the bonding. Theoretical charge-density difference plots support an alternative view of ethylene bonding in terms of the specific involvement of the excited molecular triplet state. Based on the variation in XE intensities the main difference between ethylene bonding to Cu and Ni is found to be an about two times larger occupancy of the π∗ orbital upon chemisorption on the transition metal, which comes along with C–C bond elongation and stronger σ–π rehybridization.