Abstract
Initial water deposition on the moderately reactive precious metal surface Ru(0001) has been thought to produce an ice-like bilayer. However, calculations from first principles show that the wetting layer observed on Ru(0001) cannot be formed of undissociated water molecules. An energetically favorable alternative, consistent with the remarkable observation that the wetting layer's oxygen atoms are nearly coplanar, is a half-dissociated monolayer wherein water molecules and hydroxyl fragments are hydrogen-bonded in a hexagonal structure and hydrogen atoms bind directly to the metal.
References
44
Referenced
466
- J. P. Perdew in Electronic Structure of Solids '91 P. Ziesche H. Eschrig Eds. (Akademie Verlag Berlin 1991).
10.1103/PhysRevB.46.6671
10.1016/0039-6028(94)91131-2
10.1016/0039-6028(82)90331-4
10.1103/PhysRevLett.74.4221
10.1016/0039-6028(94)00836-1
10.1116/1.571895
10.1016/0167-5729(87)90001-X
10.1103/PhysRevB.47.558
10.1103/PhysRevB.49.14251
10.1016/0927-0256(96)00008-0
10.1103/PhysRevB.54.11169
10.1103/PhysRevB.41.7892
10.1103/PhysRevLett.69.1982
10.1103/PhysRevB.47.10142
10.1088/0953-8984/6/40/015
10.1103/PhysRevB.59.1758
10.1103/PhysRevB.50.17953
- With USPs and a basis cutoff of only 29.1 Ry VASP produces a PW‘91 sublimation energy for Hamann’s ice-Ih model (38 39) (neglecting zero-point motion) equal to 0.71 eV. Using PAW potentials instead and a basis cutoff of 51.5 Ry the sublimation energy is reduced to 0.67 eV. This latter value is within 0.01 eV of what Hamann recently obtained using an independently written total-energy code; hard norm-conserving pseudopotentials (NCPPs); and a basis cutoff of 130 Ry. With this agreement as a sign of convergence I have used USPs with the 29.1-Ry cutoff to select promising candidate water adsorption structures and PAWs with the higher cutoff to check and refine the results. This calculation was done only at the experimental ice-Ih volume (D. R. Hamann private communication). Because energy varies slowly with volume near a minimum the effect of not optimizing the lattice parameter in the NCPP calculation is no more than a few meV—and in the right direction relative to agreement with the present PAW result.
- Similar LEED I - V curves were reported in (5 6) for H 2 O and D 2 O adlayers on Ru(0001) implying similar local geometries for the two molecules. This means that the local geometry was not affected much by the mass of H versus that of D and that ignoring quantum effects is a reasonable approximation if one is trying to understand a local effect such as the near coplanarity of the adlayer O atoms.
- To sample the surface brillouin zone (SBZ) for the 3 ×3 supercell I used a 4 × 4 set of equally spaced k vectors and for the larger cell a 2 × 2 k vector set. In all the adsorption calculations discussed here I accelerated electronic relaxation using Methfessel and Paxton's Fermi-level smearing method (width = 0.2 eV) (40). I corrected for the contact potential difference associated with adsorbing water on only one side of an Ru slab using Kresse's adaptation (41) of Neugebauer and Scheffler's method (42). Because Held and Menzel's LEED results imply that water prefers bonding in atop sites (i.e. with the O atom directly above an Ru rather than in a two- or threefold site) (3) in agreement with numerous earlier calculations (8) I did not consider adlayer geometries in which O atoms bridge Ru atoms or reside in threefold hollows. Similarly because Held and Menzel report that D 2 O/Ru(0001) produces a sharp 3 ×3 LEED pattern I only considered geometries where this would occur.
- Using a 12-molecule per supercell Bernal-Fowler ice-Ih model the PW'91 exchange-correlation potential (1 2) a hard NCPP and a plane-wave basis cutoff of 130 Ry Hamann finds a sublimation energy of 0.66 eV. The experimental value removing zero-point energy to allow direct comparison is 0.61 eV (23 24). Hamann reported a theoretical E sub of 0.55 eV in (38). The new theoretical value 0.66 eV follows the correction of a problem with the variable grid used in that work.
- E. Whalley in The Hydrogen Bond P. Schuster G. Zundel C. Sandorfy Eds. (North-Holland Amsterdam 1976) vol. 3 pp.1425–1470.
- According to the analysis in (23) zero-point vibration reduces the 0 K sublimation energy of H 2 O by 120 meV and of D 2 O ice by 98 meV.
- PAW results are also listed in Table 2 to provide a convergence check. The PAW preference for ice is ∼0.17 eV compared with 0.19 eV for the same structures but with the use of USPs. Thus lack of basis convergence is not the source of weak binding of undissociated water structures to Ru(0001).
- This agrees qualitatively with the finding (43) of an energy cost of several tenths of an electron volt to move an adsorbed H to a subsurface octahedral site.
10.1021/jp003511g
10.1016/0039-6028(91)90089-B
10.1016/0039-6028(89)90626-2
10.1063/1.446561
10.1016/0039-6028(87)90126-9
10.1016/0039-6028(83)90012-2
10.1016/S0009-2614(00)00663-1
- As noted in (4) they should be attracted back to the surface by image forces.
- P. Feulner D. Menzel unpublished data.
- ESD of the dissociated H atoms of the half-dissociated structure should also be weak.
- A mosaic of 3 ×3 patches with H-up waters only in their central regions.
10.1103/PhysRevB.55.R10157
- The theoretical sublimation energy equals the energy of a single D 2 O molecule less 1/12 the energy of the 12-molecule supercell.
10.1103/PhysRevB.40.3616
- G. Kresse VASP Guide .
10.1103/PhysRevB.46.16067
10.1103/PhysRevB.39.5623
- I am grateful to D. R. Jennison for alerting me to the problem of accounting theoretically for Held and Menzel's LEED geometry; N. D. Shinn B. C. Bunker D. Menzel and T. E. Madey for helpful discussions; and D. R. Hamann D. M. Teter and G. Kresse for assistance in establishing a baseline for the accuracy of H bonding calculations. VASP was developed at the Institut für Theoretische Physik of the Technische Universität Wien. Supported in part by the DOE Office of Basic Energy Sciences Division of Material Sciences. Sandia is a multiprogram laboratory operated by Sandia Corporation a Lockheed Martin Company for the United States DOE under contract DE-AC04-94AL85000.
Dates
Type | When |
---|---|
Created | 23 years ago (July 27, 2002, 5:47 a.m.) |
Deposited | 1 year, 7 months ago (Jan. 9, 2024, 6:07 p.m.) |
Indexed | 2 weeks, 5 days ago (Aug. 6, 2025, 9:54 a.m.) |
Issued | 23 years, 7 months ago (Jan. 4, 2002) |
Published | 23 years, 7 months ago (Jan. 4, 2002) |
Published Print | 23 years, 7 months ago (Jan. 4, 2002) |
@article{Feibelman_2002, title={Partial Dissociation of Water on Ru(0001)}, volume={295}, ISSN={1095-9203}, url={http://dx.doi.org/10.1126/science.1065483}, DOI={10.1126/science.1065483}, number={5552}, journal={Science}, publisher={American Association for the Advancement of Science (AAAS)}, author={Feibelman, Peter J.}, year={2002}, month=jan, pages={99–102} }