Elsevier

Surface Science

Volume 405, Issue 1, 12 May 1998, Pages 121-137
Surface Science

H2O adsorption on alkali (Li, Na and K) precovered Ni(775)

https://doi.org/10.1016/S0039-6028(98)00061-2Get rights and content

Abstract

The coadsorption of H2O and alkalis (Li, Na and K) on a stepped Ni(s)(111) surface with nominal (775) orientation was studied by UPS (UV-light induced photoelectron spectroscopy), LEED, TDS (thermal desorption spectroscopy) and work function change measurements (ΔΦ). On clean Ni(775) five H2O desorption states denoted as A (Tm=160 K), B (175 K), C (225 K), D (260 K) and E (340 K) were detected. The A and B states are also observed at similar desorption temperatures on flat Ni(111) and are therefore attributed to H2O adsorption on the (111) terraces of Ni(775) (B state) or to the desorption of ice multilayers (A state). Consequently, the C, D and E states are ascribed to the adsorption and a partial dissociation at step sites. The C state is associated with a “five peak” UPS spectrum interpreted as due to the overlap of the 1b1, 3a1 and 1b2 orbitals of molecular water and 2σ and 1π emission from OH groups.

A preferred adsorption of the alkalis at step sites for low coverages is deduced from the ΔΦ changes. For Li adsorption the saturation of the step sites is accompanied by a significant change in the slope of the ΔΦ versus coverage curve. The increase of the effective dipole moment (less negative for low coverages, adsorption at step sites) is caused by screening of the positively charged alkalis in a step-down adsorption geometry. For the coadsorption of H2O and alkalis (at low alkali precoverages) only H2O desorption states C, D and E from step sites are influenced. The C state shifts to lower peak temperature, whereas the D and E states (recombination from dissociation products) decrease in intensity with increasing alkali coverage. Alkali coadsorption on Ni(775) is accompanied by the dissociation of water, depending on the nature and the precoverage of the alkali. The tendency to induce dissociation is highest for Li and decreases with increasing ionic radius from Na to K.

Introduction

The adsorption properties of water on transition metal surfaces are of fundamental interest in different areas of science as well as technology. For example, in electrochemistry the water structure and eventually also the formation of specific adsorbed ions on the electrodes play an important role. In recent years, studies employing surface analytical methods have demonstrated that UHV studies of the adsorption of H2O and the coadsorption of H2O with ions (obtained for example from alkali adsorption at low coverages) can be used as model systems contributing to the clarification of structures in the double layer in the interface between metal and the bulk water 1, 2, 3.

Furthermore, water adsorption and desorption plays an important role in a large number of technical important processes. Examples are given in Ref. [4]for the catalytic oxidation of organic molecules.

A large experimental body exists about the adsorption of water on low indexed transition metal surfaces as for example Ni(111) and Ru(0001). A very good overview of these investigations is given in a recent review article by Thiel and Madey [5]. On the flat Ni (111) surface the interaction between H2O and the substrate is rather weak. Water desorption occurs on this surface with two desorption maxima in TDS (thermal desorption spectroscopy) the peak maxima at ≈160 K (ice multilayer) and ≈175 K (water clusters in contact with the substrate) 6, 7. No dissociation is observed to occur on this surface. Recent experimental work, especially on flat Ni(111) 6, 7and Ru(0001) 8, 9, 10, 11has been directed to the clarification of the H2O adsorption geometry and the built up of hydrogen bonded water clusters even for fractional monolayer coverages.

The formation of hydrogen bonded H2O clusters (bilayer) is supposed to be the result of: (1) the weak interaction between H2O and the substrate; and (2) the good match of the bilayer hexagonal structure with the geometry of the substrate. If we introduce steps in the substrate structure we therefore would expect a considerable influence of the step structure on the built up of water bilayer structures. As we demonstrated recently 12, 13the presence of steps on Ni(111) creates new H2O adsorption sites which occur in TDS experiments at higher desorption temperatures. These states are attributed to the higher binding energy of H2O molecules at step sites and to a partial dissociation. The increase of the H2O/Ni binding energy was discussed in terms of a simple Lewis acid–base chemistry. Adsorbed water on transition metal surfaces donates electrons to the metal and acts as the Lewis base whereas the metal acts as the Lewis acid (electron acceptor). Ni atoms at step sites are positively charged owing to the Smoluchowsky smoothing effect [14]. The existence of Niδ+ at step sites increases the activity of these sites. This leads to the conclusion that H2O molecules at step sites are more strongly bound and probably occupy positions at the tops of the steps. This interpretation was supported by the high contribution of H2O from the step adsorption state to the negative change of the work function (at acid adsorption sites the electron donation from H2O to the Ni substrate is increased) [15].

On Ni(221) and Ni(775) we have found evidence from TDS measurements that the adsorption at step sites is associated with a partial dissociation [13]. High temperature desorption states detected on these surfaces at 260 and 340 K are attributed to the recombination of dissociation products. Further, we found hints for an influence of the H2O molecules adsorbed on steps on the formation of the bilayer structure on the (111) terraces. This influence is suggested to be due to the decoration of steps at low coverages. Water molecules adsorbed at low coverages on the terraces diffuse to step sites and are trapped there due to the higher binding energy. At higher coverages, the built up of bilayer clusters starts from H2O step molecules at step sites and extend then onto the terraces.

With LEED measurements we could demonstrate an influence of the terrace width on the long range ordering of the H2O bilayer on the terraces. On Ni(11 11 9) where the terrace width is 9.5 atoms we observed, similar to the flat Ni(111) surface [7], a (3×3)R30° overlayer near the saturation of terrace adsorption sites. On Ni(111) the (3×3)R30° overlayer had been attributed to the formation of a H2O bilayer with a coverage of ΘH2O=2/3. In contrast, on Ni(775) with a smaller terrace width of 5.5 atoms, no (3×3)R30° overlayer could be detected. Instead we observed a (2×2) LEED structure for an intermediate coverage lower than the saturation of the bilayer, attributed to ΘH2O=0.5. On Ni(11 11 9) the same (2×2) overlayer could be observed for lower coverages (ΘH2O=0.5) after moderate heating of the (3×3)R30° structure with a partial desorption of H2O. The lack of the (3×3)R30° overlayer on Ni(775) demonstrates the influence of the terrace width on the long range ordering of the adsorbate.

In the recent past, extensive work has been devoted to the role of surface additives for the adsorption and reaction of gases on transition metal surfaces including the coadsorption of alkalis and H2O. It is well established in catalysis that the modification of surfaces by addition of promoters or poisons can lead to changes in reaction rates, pathways and product distributions. Similar to our interpretation for the chemical activity of steps, the modification of a surface by additives can be discussed with the Lewis acid and base concept proposed for surface reaction by Stair [15]. Adsorption of electropositive additives such as alkalis decreases the acidity of the neighboring substrate atoms. Lang et al. [16]have shown further that the presence of alkalis is expected to change the bonding energies of polar molecules through electrostatic interaction. Finally, experimental results (for example by Madey [17]and Bornemann et al. [18]) demonstrate that, owing to the electrostatic interaction, the orientation of the polar H2O molecule can be changed in the neighborhood of electropositive or electronegative additives.

The present investigation of H2O adsorption on the stepped Ni(775) surface is aimed towards the modification of the steps by alkali (Li, Na and K) adsorption. Theoretical calculations by Thompson [19]predict for the alkali adsorption on stepped transition metals a preferred adsorption at step sites in step-down positions. We therefore expect a modification of the H2O step adsorption state for low alkali precoverages. These experiments should help to clarify and support our interpretations derived for H2O adsorption on clean Ni(775) and other stepped Ni(s)(111) surfaces, where we postulated the existence of water monomers and a lowering of the dissociation barrier at step sites.

In the present investigation the assumption of a preferred alkali adsorption at step sites is supported by work function change measurements for low precoverages. Further support comes from thermal desorption experiments for H2O from alkali precovered Ni(775). These TDS measurements show that for low alkali precoverages only H2O step induced desorption states are influenced. Our general conclusion from H2O adsorption on Ni(s)(111) concerns the influence of surface defects (steps) on the structure of the hydrogen bonded H2O network and on new pathways for H2O dissociation.

Section snippets

Experimental details

The experiments were performed in two different UHV systems operating at base pressures below 1×10−10 Torr. Details of these systems were recently reported elsewhere 12, 20.

Therefore, we restrict ourselves to a brief description of the analytical methods employed in the present study. The first UHV system had facilities for thermal desorption spectroscopy (TDS), LEED and work function change measurements (ΔΦ). In order to clean the Ni(775) surface and to detect surface impurities, the system was

Alkali adsorption on Ni(775)

Fig. 1 shows a ball model and a schematic LEED pattern of the Ni(775) surface. According to the nomenclature by van Hove and Somorjai [21]it consists of (111) terraces and monoatomic steps of (111̄) orientation:(775)=66×(111)+11×(111̄).

The evaluation of the spot splitting from LEED yields a value of 5.66±0.12, which is quite close to the theoretical value of 5.54 [21].

The step height of the Ni(775) surface was obtained from the intensity changes of the split (00) spot with beam energy [22].

H2O adsorption on clean Ni(775) and Ni(221)

We mainly will concentrate on the H2O adsorption states attributed to the presence of steps. These states have been observed in TDS experiments on Ni(775) as well as Ni(221) at nearly the same desorption temperature. Due to the higher step concentration, however, on Ni(221) a higher population of step states (C+D+E) relative to the terrace state B was measured for Ni(221) compared with Ni(775). In UPS step associated states with the occurrence of a “five peak” spectrum could only be observed on

Conclusions

We briefly summarize the main conclusions from our present investigations.

  • 1.

    With UPS we found evidence for a thermal induced dissociation of H2O adsorbed at step sites at ≈160 K. A HOH⋯OH water hydroxyl complex is built which is identified in UPS by a “five peak” spectrum.

  • 2.

    The desorption states C, D and E associated with the H2O adsorption at step sites could be identified with photo electron spectroscopy (UPS) to be molecular water stabilized by OH (desorption state C at 220 K), recombination of OH

Acknowledgements

The financial support of this investigation by the Deutsche Forschungsgemeinschaft is gratefully acknowledged.

References (39)

  • F.T. Wagner et al.

    Surf. Sci.

    (1987)
  • E. Stuve et al.

    Surf. Sci.

    (1984)
  • P.A. Thiel et al.

    Surf. Sci. Rep.

    (1987)
  • R.H. Stulen et al.

    Surf. Sci.

    (1985)
  • G. Pirug et al.

    Surf. Sci.

    (1991)
  • G. Held et al.

    Surf. Sci.

    (1995)
  • C. Mundt et al.

    Surf. Sci.

    (1994)
  • C. Mundt et al.

    Surf. Sci.

    (1993)
  • N.D. Lang et al.

    Surf. Sci.

    (1985)
  • T. Bornemann et al.

    Surf. Sci.

    (1991)
  • M.A. Van Hove et al.

    Surf. Sci.

    (1980)
  • M. Henzler

    Surf. Sci.

    (1970)
  • K.J. Uram et al.

    Surf. Sci.

    (1986)
  • R.L. Gerlach et al.

    Surf. Sci.

    (1970)
  • S. Semancik et al.

    Surf. Sci.

    (1986)
  • J.K. Sass et al.

    Surf. Sci.

    (1991)
  • K. Bange et al.

    Vacuum

    (1983)
  • C. Nöbl et al.

    Surf. Sci.

    (1985)
  • J.K. Sass, K. Bange, Electrochemical Surface Science, Ed., in: M.P. Soriaga, ACS Symposium, vol. 378, Washington, DC,...
  • Cited by (0)

    View full text