Surface diffusion potential energy surfaces from first principles: CO chemisorbed on Pt{110}

Abstract

Lateral potential energy curves for the chemisorption of CO on Pt{110} (1×1) and (1×2) along different azimuthal directions have been calculated using density functional theory slab calculations. In contrast to the simple models almost universally used, the results along 〈11̄0〉 show that there is a barrier of ∼0.15 eV between bridge and atop sites. Both bridge and atop sites are local minima. Diffusion along 〈100〉 on the (1×1) surface is strongly inhibited by a barrier ⩾1.2 eV. Quasielastic helium atom scattering data require reanalysis in the light of these results. The free energy, determining the most stable site at finite temperatures, includes a significant vibrational entropy term in the atop site.

CO adsorption on Cu(111) and Cu(001) surfaces: Improving site preference in DFT calculations

“…Lewis and Rappe [20] report for the same site (top) a lower adsorption energy, but still too high (in absolute value) compared to experiment (E ads = −1.239 eV). Although Ge and King [21,22] found that bonding at the bridge site is stronger by 80 meV than in the top site, and a barrier between the top and the bridge site which was not observed in a helium atom scattering experiment (HAS) [23,24]. Finally, we know of only one attempt to study CO on the Cu(001) surface including the generalized gradient approximation (GGA) for the exchange-correlation functional [25].…”

Comment on “Surface diffusion potential energy surfaces from first principles: CO chemisorbed on Pt{110}” [J. Chem. Phys. 111, 9461 (1999)]

“…5,11 In particular, for CO/Pt͑111͒ the curvature of the potential provided by the T-mode frequencies in the bridge and on-top sites, coupled with the potential energy difference available via the temperature dependence of the occupation probabilities, 17 indicate a diffusion barrier of 130 meV, 13 the same as the value measured with QHAS 13 and calculated by Ge and King for CO/Pt͑110͒. 1 It is clear that their comparison of CO/Pt͑110͒ with CO/Cu͑001͒ is not appropriate since, unlike CO/Pt͑111͒, for which two adsorption sites clearly exist and contribute to the shape of the PES, CO is observed to occupy exclusively on-top sites on Cu͑001͒, even under compression. 18 In summary, the length scale information provided by QHAS rules out intrasite diffusional motion as an explanation of the measured QHAS broadening.…”

“…͑1͒ In their article Ge and King claim that very short, intrasite jumps are responsible for the quasielastic broadening observed in previous QHAS studies. 1 Motions within the unit cell ͑length scales less than lϷ1 Å͒ correspond to parallel momentum transfers of ⌬KϾ6 Å Ϫ1 and, consequently, do not contribute to the quasielastic broadening in the ⌬K ranges reported for the QHAS measurements (0Ͻ⌬Kр3 Å Ϫ1 , lϷ2 -ϱ Å͒. Further, in the small momentum transfer limit, corresponding to long length scales (lϭ2/⌬K) much greater than one lattice constant, the microscopic details of the diffusional motion are averaged out over many jumps.…”

“…In their paper, 1 Ge and King compare the potentials from extensive density functional theory ͑DFT͒ calculations for coverages of ⌰ϭ0.25 and ⌰ϭ0.5 of CO on platinum ͑110͒ with quasielastic helium atom scattering ͑QHAS͒ results on copper and nickel surfaces. They note that the QHAS diffusion barriers measured at low coverages ͑⌰ϭ0.05-0.15͒ for the ͑001͒ and ͑110͒ nickel surfaces, which are between 30 and 60 meV, are significantly lower than those reported previously and their values for CO/Pt͑110͒.…”

“…1 It is now well established, from the work of the King group, that the stable CO adsorption positions on Ni͑111͒ are the threefold hollow sites, 15 whereas the hollow site is not occupied on Pt͑111͒ faces even at high coverages. 16 Thus, when comparing diffusion along the ͓100͔ directions on Ni͑110͒ and Pt͑110͒, jumps via the more highly coordinated ͑111͒ trough sites are anticipated to be more facile on Ni͑110͒ than on Pt͑110͒.…”