Skip to main content

Final Report 2002

Surface Structure Determination using Photoelectron Diffraction at a Third-generation Synchrotron Radiation Source

Final Report

Background

This project was concerned with the continuing development and exploitation of the technique of scanned-energy mode photoelectron diffraction (PhD) through a Warwick-Berlin collaboration. In PhD one exploits the coherent interference of the directly emitted photoelectron wavefield from an atom adsorbed on a surface with other components of the same wavefield elastically scattered by the surrounding (mainly substrate) atoms. This causes modulations in the measured photoemission intensity in any given direction as a function of photoelectron energy (and thus photoelectron wavelength), and proper interpretation of these modulations provides quantitative information on the local structure around the emitter atom. The Berlin end of this collaboration has been based at the Fritz Haber Institute (FHI) in the Department of Professor Alex Bradshaw, although with his move to be Director of the Max Planck Institute for Plasma Physics near Munich in 1999, the Warwick grant-holder (Professor Phil Woodruff – DPW) has taken over the direction of both ends of the collaboration. The regular (monthly) visits to the FHI were possible because DPW has held an EPSRC Senior Research Fellowship throughout this period. The experiments have been performed in a purpose-built UHV surface science chamber owned by the FHI taking light from the BESSY synchrotron radiation facilities. The current grant period covered the switch-over from the old second-generation BESSY facility using bending magnet sources to the new third-generation undulator-based BESSY II facility, and it is this changeover which is reflected in the title of the grant. Nevertheless, the project always envisaged extensive use of the old BESSY facility during this transitional period.

The actual switch to BESSY II actually was rather more sudden than anticipated because of a decision by the facility, apparently on financial and political grounds, to close BESSY at the end of 1999 when only a small number of beamlines were in operation on BESSY II. The year 2000 was thus a year of very sparse beamtime using newly commissioned beamlines with all the attendant problems, but proved far more successful than had been anticipated. Nevertheless, bearing in mind the long data analysis times associated with these experiments, the results of experiments performed at BESSY II are only now being submitted for publication or in the final stages of preparation. Overall, the project has been extremely successful; the continuing programme has resulted in 28 published papers [1-28] since the start of the grant, a series of technical problems associated with the transfer to BESSY II have been overcome, and a series of studies exploiting the special capabilities of the new source are nearing publication.

In the remainder of this report the achievements of the last three years are summarised, organised according to the type of adsorbate/substrate combination or the specific extension of the methodology or range of application of the method, broadly along the lines laid out in the original grant proposal. Only one topic of this original proposal is not included; the slow implementation of polarisation switching of the circularly-polarised radiation beamline on BESSY II, combined with the difficulty of mounting a new set of experiments on magnetic thin films in the short beam-time allocations at this new facility, has meant that the experiments to investigate the applicability of PhD to investigations of local magnetic order have not yet been initiated.

Molecular adsorption and coadsorption on metal surfaces

A series of structure determinations of small molecular species on metal surfaces, mainly of Cu and Ni, continued to develop on the old BESSY facility. For example, we have investigated the adsorption of both pyridine [2] and methyl pyridine [14] on Cu(110); the primary information from the N 1s PhD is the (near-atop) bonding site on the surface, but intramolecular scattering also provides some information on the molecular orientation and the results reveal, as expected, that the addition of a methyl species on one of the C atoms adjacent to the bonding N atom adds additional steric constraints on the adsorption geometry, forcing the molecule into a twisted orientation. We have also extended our studies of other simple hydrocarbons, establishing the influence of different coverages of ethylene on Ni(110) [4], and determining the structure of benzene on this same surface [10], providing a measure of the influence of the reduced symmetry of the substrate on the structure of the benzene ring, which shows a statistically significant expansion due to the adsorption. As an extension of our very early investigation of the formate and acetate species on Cu(110), we have also applied our far more detailed modern methodology to determine the structure of the related benzoate species on this surface, confirming that the local bonding geometry through the carboxylate O atoms is similar [26].

A rather different investigation of an adsorbed hydrocarbon on a metal surface was an investigation of the methyl species on Cu(111), adsorbed in isolation from free methyl radicals produced by the pyrolysis of azomethane, and coadsorbed with I by the decomposition of methyl iodide. We find [29] the coadsorbed I does not have any significant effect on the local geometry of the methyl adsorbate which occupies ‘fcc’ hollow sites on the surface directly above third layer. At first site this seems rather surprising, as one might expect this species to adopt a singly-coordinated atop site; this caused us to undertake a separate DFT calculation [33] which confirmed the clear preference for the hollow sites, but also revealed a strong preference for an azimuthal orientation in which the methyl H atoms eclipse the three nearest-neighbour Cu atoms, showing that the nature of the bonding is quite different from that in an organic molecule. In fact DFT calculations indicate this hollow site is the preferred geometry on several close-packed metal surfaces (e.g. also on Ni(111)) although also indicate that in Pt(111), for example, the atop site is preferred.

Three other simple molecular adsorption systems studied were Cu(111)/NH3 [11], Cu(210)/CO [18] and Pd(110)/CO [30]. In the first two cases the expected atop adsorption sites were found, but in the former case we explored the potential of the technique to determine the H atom positions which, as might be expected, proves rather marginal, and certainly not capable of providing chemically significant structural information; the Cu(210)/CO system provided a first exploration of adsorption on a high-index or vicinal (‘stepped’) surface. The Pd(110)/CO system, which coincidentally explores our ability to extend PhD to studies of second transition-row elements as mentioned in the original proposal, resolves a very long-standing controversy over whether this system involves atop or bridge adsorption sites, clearly showing the bridge site favoured by DFT calculations to be correct. Further molecular systems studied were two coadsorption systems on Ni(111), NO+benzene [19] and CO + O [21]. Having established some years ago that pure benzene on Ni(111) adopts different local adsorption sites and different azimuthal orientations at different coverages, we were keen to confirm (or otherwise!) the single local site adopted in the NO coadsorption phase which, having long-range order, can also be determined by LEED; the question of which hollow site (if either) is adopted in the coadsorption phase is also of interest relative to its behaviour on Ni(111) in the absence of the coadsorbed benzene. The CO/O coadsorption system was studied, on the other hand, to resolve a long-standing controversy which had quite recently been revitalised. Early vibrational spectroscopy, notably by Yates and coworkers, indicated that in the presence of pre-adsorbed O, CO adopts atop sites on this surface, but our original investigation of this system some years ago indicated that this did not occur. More recently, XPS spectral fingerprinting by the group of Steinrück indicated that the pre-coverage of oxygen may be crucially important. Our re-investigation provides quantitative confirmation of this idea, revealing a temperature dependence in the relative occupation of hollow and atop sites which really stems from the different desorption temperatures from these two sites which are co-occupied when the initial O pre-coverage is less than the exact coverage of the ordered (2x2) phase.

Adsorption on silicon surfaces

One of our stated objectives of the present programme was to develop studies of a range of different substrates rather than concentrate entirely on the Cu and Ni materials which are not only valuable model systems but also have very favourable properties for PhD in terms of their elastic backscattering cross-sections. One specific substrate mentioned was Si, which is obviously of great practical interest from the point of view, especially, of electronics rather than heterogeneous catalysis and in this regard the adsorption of simple C2 hydrocarbons provides a model for larger molecular adsorbates of interest in the context of molecular electronics and sensors. We have so far concentrated in this area on the adsorption of ethylene [6] and acetylene [15] on Si(100). Conventional wisdom, based on various spectroscopies and total energy calculations, was that both species adsorb atop surface Si dimers in a so-called di-s configuration, although some work indicated that the dimer bonds were broken. Our results provided quantitative confirmation of these sites and clearly showed that the dimers are intact. However, as a result of some quite different conclusions from a ‘photoelectron holography’ study of acetylene on Si(100) by the group of G.J.Lapeyre we have recently taken new measurements on this system with the enhanced spectral resolution of BESSYII to look for evidence of multiple adsorption sites, especially at room temperature. We do, indeed, find evidence of this added complexity, although do not support the interpretation of the Lapeyre group. This particular adsorption system has generated a lot of controversy in the last couple of years, and we are now writing up the results of our far more exhaustive experimental structural study for publication.

Atomic adsorbates and substrate reconstruction

While the PhD method is especially effective in determining the structure of molecular adsorbate systems, which commonly lack long-range order, it is equally applicable to atomic adsorbate systems, with or without long-range order. A particular strength of a third-generation source in this regard is the ability to study low adsorbate coverages, and in particular to determine the local structure of adsorbates in systems in which high coverage generates reconstruction; this allows us to probe the structural character of the precursor to the reconstruction. So far we have only performed such experiments at the ALS in Berkeley, but this has allowed us to gain important insight into two such adsorption systems, C on Ni(100) [5,9] and O on Cu(100) [18]. In the former case 0.5 ML of C produces a ‘clock’ reconstruction of the Ni surface, and STM images had been interpreted in terms of local relaxation of the near-neighbour Ni atoms around the adsorbed C at low coverage. Our results show this is not correct, and that this is an imaging artefact; they also give further insight into which bond-length change, and which stay constant in the transition. In the case of Cu(100)/O the local adsorption structure prior to creation of the 0.5 ML missing row structure has also been a matter of dispute, and our PhD data show rather clearly that simple hollow sites are occupied, albeit complicated by the effects of the very small locally ordered domains which form.

In this general category of problem we have also used PhD to clarify the local structure of N-induced structures on both Cu(111) [12] and Cu(100) [25]. In both of these cases of adsorbate-induced surface reconstruction the combination of the Berlin PhD measurements and Warwick-based STM investigations have proved especially effective in elucidating these structures. Our results confirm the pseudo-(100) reconstruction of the (111) surface (but indicate there are probably two reconstructed layer), while on the (100) surface we find a novel ‘rumpling’ reconstruction which we believe is strongly related to the novel island self-organisation of this system. A third system, also exploring our ability to extend PhD to studies of second transition-row elements, is the Ag(110)/O missing row reconstruction [16]

Adsorption structures on oxide surfaces

A somewhat different substrate system which we have explored in oxide surfaces. Although these surfaces are widely recognised to be of great importance, there is a remarkable dearth of experimental structural information. By growing thin epitaxial oxide films on metallic substrates one can overcome the problem of charging in electron spectroscopy of insulating oxides, and the system we have studied is NiO(100) grown on Ni(100). Our first experiments, conducted even before the start of the current programme, were on NO adsorption [3] which proved the viability of the method but led us to realise in the course of open discussions (see discussion associated with [7]) that current theoretical treatments appeared to be woefully inadequate in describing the strength of bonding and the substrate-adsorbate bondlengths. Further studies on CO and NH3 adsorption on NiO(100) have shown that this problem is quite general on NiO [23, 28] and appears to have initiated a significant amount of new theoretical work. This general area is one we plan to extend in the future.

Structures involving photoelectron binding energy ‘chemical’ shifts

An important aspect of the high spectral resolution available of the third generation synchrotron radiation source is the ability to use quite small ’chemical’ shifts in core level photoelectron binding energies to provide a spectral fingerprint of changes on a surface and to even obtain chemical-state specific PhD data from atoms of the same element co-existing in different chemical states. One of our last experiments at BESSY I [13] exploited this ability, studying the structure of propyne (HCº CCH3) on Cu(111), although because the low resolution beamline in this case could not resolve the methyl and acetylenic C atoms, addition experiments were performed with the fluorinated species HCº CCF3 for which the energy separation is much larger. At BESSY II we have pursued this idea further and several such studies have been completed or are very near completion. Two which have recently been written up are the systems Ag(110)/CO3 [31] and Cu(111)/CH3S- [32]. In the case of the carbonate the chemical state resolution proves essential to ensure that it is the carbonate, and not chemisorbed CO2 which is studied; this was a challenging system both in terms of the substrate scattering and the low symmetry of the adsorbate, but the PhD results clearly define an optimum structure. The problem of methanethiolate on Cu(111) is one which we have studied over several years at Warwick by other methods, but the local structure of the unreconstructed phase had not been determined unambiguously. The PhD data, on the other hand, show that both bridge and hollow species are present, a conclusion which can be rather effectively reconciled with all previous results form other methods. Also in the final stages of analysis are data on Ni(100)/N2 and Ni(100)/H/CO. In the case of the di-nitrogen the end-on bonding to the surface renders the two N atoms inequivalent, and separate PhD spectra from each N atom has allowed us to obtain a particularly clear and precise structure; these results show quite clearly that a previous structural study of this system by the group of D.A.Shirley in Berkeley involves an error in the N-Ni distance of 0.3 Å, a difference of huge chemical significance. The H/CO coadsorption system, on the other hand, appears to confirm (and quantify) the structural assignments (atop and bridge) previously made for the two different C and O 1s chemically-shifted states.

 

Papers published in this project

    • P.Baumgartel, J.J.Paggel, M.Hasselblatt, K.Horn, V.Fernandez, O.Schaff, J.H.Weaver, A.M.Bradshaw, D.P.Woodruff, E.Rotenberg and J.Denlinger, 'Structure determination of the (Ö 3xÖ 3)R30° boron phase on Si{111} using photoelectron diffraction', Phys.Rev.B. 59 (1999) 13014-13019
    • T.Gieb el, O.Schaff, R.Lindsay, R.Terborg, P.Baumgartel, J.T.Hoeft, M.Polcik, A.M.Bradshaw, A.Koebbel, D.R.Lloyd and D.P.Woodruff, 'Adsorption site and orientation of pyridine on Cu(110) determined by photoelectron diffraction' J.Chem.Phys. 110 (1999) 9666-9672
    • R.Lindsay, P.Baumgartel, R.Terborg, O.Schaff, A.M.Bradshaw and D.P.Woodruff, 'Molecules on oxide surfaces: a quantitative structural determination of NO adsorbed on NiO(100)' Surf.Sci. 425 (1999) L401-406
    • T.Gieb el, R.Terborg, O.Schaff, R.Lindsay, P.Baumgärtel, J.T.Hoeft, K-M.Schindler, S.Bao, A.Theobald, V.Fernandez, A.M.Bradshaw, D.Chrysostomou, T.McCabe, D.R.Lloyd, R.Davis, N.A.Booth and D.P.Woodruff, ‘Determination of the adsorption geometry of ethylene on Ni{110} using photoelectron diffraction’, Surf.Sci.440 (1999) 125-141
    • R.Terborg, J.T.Hoeft, M.Polcik, R.Linsay, O.Schaff, A.M.Bradshaw, R.Toomes, N.A.Booth, D.P.Woodruff, E.Rotenberg and J.Denlinger, 'Structural precursor to adsorbate-induced reconstruction: C on Ni(100)' Phys.Rev.B 60 (1999) 10715-8
    • P.Baumgärtel, R.Lindsay, O.Schaff, T.Giessel, R.Terborg, J.T.Hoeft, M.Polcik and A..M.Bradshaw, M.Carbone, M.N.Piancastelli, R.Zanoni , R.L.Toomes and D.P.Woodruff, ‘The dimers stay intact: a quantitative photoelectron study of the adsorption system Si{100}(2x1)-C2H4New.J.Phys 1 (1999) 20.1-20.15

     

    • M.Polcik, R.Lindsay, P.Baumgärtel, R.Terborg, O.Schaff, A.M.Bradshaw, R.Toomes and D.P.Woodruff, Structure determination of molecular adsorbates on oxide surfaces using scanned-energy mode photoelectron diffraction Faraday Disc. 114 (1999) 141-155
    • R.L.Toomes, D.P.Woodruff, M.Polcik , S.Bao, Ph.Hofmann, K-M.Schindler and A.M.Bradshaw, ‘Is PEXAFS Really PhD?’, Surf.Sci. 445 (2000) 300-308
    • R.Terborg, J.T.Hoeft, M.Polcik, R.Lindsay, O.Schaff, A.M.Bradshaw, R.L.Toomes, N.A.Booth, D.P.Woodruff, E.Rotenberg and J.Denlinger 'The coverage dependence of the local structure of C on Ni(100): a structural precursor to adsorbate-induced reconstruction' Surf.Sci. 446 (2000) 301-313
    • J-H.Kang, R.L.Toomes, J.Robinson, D.P.Woodruff, O.Schaff, R.Terborg, R.Lindsay, P.Baumgärtel and A.M.Bradshaw, 'The local adsorption geometry of benzene on Ni(110) at low coverage', Surf.Sci. 448 (2000) 23-32
    • P.Baumgärtel, R.Lindsay, T.Giessel, O.Schaff A.M.Bradshaw and D.P.Woodruff ‘Structure determination of ammonia on Cu(111)’ J.Phys.Chem.B, 104 (2000) 3044-3049
    • R. L. Toomes, J.Robinson, S.M.Driver, D. P. Woodruff, P. Baumgärtel, T. Geib el, R. Lindsay, O. Schaff and A. M. Bradshaw, ‘Photoelectron Diffraction Investigation of the Local Adsorption Site of N on Cu(111)’, J.Phys.:Condens.Matter 12 (2000) 3981-3991
    • R.L.Toomes, R.Lindsay, P.Baumgärtel, R.Terborg, J.-T.Hoeft, A.Koebbel, O.Schaff, M.Polcik, J.Robinson, D.P.Woodruff, A.M.Bradshaw and R.M.Lambert ‘Structure determination of propyne and 3,3,3-trifluoropropyne on Cu(111)’ J.Chem.Phys. 112 (2000) 7591-7599
    • R.Terborg, M.Polcik, J-T.Hoeft, M.Kittel, M.Pascal, J.H.Kang, C.Lamont, A.M.Bradshaw and D.P.Woodruff, ‘The local adsorption geometry of 2-methyl-pyridine on Cu(110) determined by photoelectron diffraction’ Surf.Sci. 457 (2000) 1-10
    • R.Terborg, P.Baumgärtel, R.Lindsay, O.Schaff, T.Gieb el, J.T.Hoeft, M.Polcik, R.L.Toomes, S.Kulkarni A.M.Bradshaw and D.P.Woodruff, ‘The local adsorption geometry of acetylene on Si(100)(2x1), Phys.Rev.B. 61 (2000) 16697-16703
    • M.Pascal, C.L.A.Lamont, P.Baumgärtel, R.Terborg, J.T.Hoeft, O.Schaff, M.Polcik, A.M.Bradshaw, R.L.Toomes and D.P.Woodruff, ‘Photoelectron diffraction study of the Ag(110)(2x1)-O reconstruction’ Surf.Sci. 464 (2000) 83-90
    • M.Kittel, M.Polcik, R.Terborg, J.-T.Hoeft, P.Baumgärtel, A.M.Bradshaw, R.L.Toomes, J.-H.Kang and D.P.Woodruff, M.Pascal, C.L.A.Lamont, E.Rotenberg, ‘The structure of oxygen on Cu(100) at low and high coverages’ Surf.Sci.470 (2001) 311-324
    • R.Terborg, M.Polcik, R.L.Toomes, P.Baumgärtel, J.-T.Hoeft, A.M.Bradshaw and D.P.Woodruff, ‘Photoelectron diffraction determination of the local adsorption geometry of CO on Cu(210)’ Surf.Sci 473 (2001) 203-212
    • S.Bao R.Lindsay, M.Polcik, A.Theobald, T.Geiβel, O.Schaff, P.Baumgδ
    rtel, R.Terborg, A.M.Bradshaw, N.A.Booth and D.P.Woodruff, ‘Local structure determination for benzene/NO coadsorption on Ni(111) using scanned-energy mode photoelectron diffraction’ Surf.Sci.478 (2001) 35-48
    • D.P.Woodruff, ‘Chemical-state specificity in surface structure determination’, Appl.Phys.A. 72 (2001) 421-428
    • J.-H.Kang, R.L.Toomes, J.Robinson, D.P.Woodruff, R.Terborg, M.Polcik, J.T.Hoeft, P.Baumgärtel and A.M.Bradshaw ‘The local structure of CO coadsorbed with O on Ni(111): a temperature-dependent study’ J.Phys.Chem.B 105 (2001) 3701-3707
    • D.P.Woodruff ‘Chemical-State-Specific Surface Structure Determination’ Surf.Sci. 482-485 (2001) 49-59
    • J.-T.Hoeft, M.Kittel, M.Polcik , S.Bao, R.L.Toomes, J.-H.Kang, D.P.Woodruff, M.Pascal and C.L.A.Lamont ‘Molecular Adsorption Bondlengths at Metal Oxide Surfaces: Failure of Current Theoretical Methods’ Phys.Rev.Lett. 87 (2001) 086101-(1-4)
    • D.P.Woodruff ‘Photoelectron diffraction and surface structure’, in Encyclopedia of Materials: Science and Technology Eds. K.H.J.Buschow, R.W.Cahn, M.C.Flemings, B.Ilschner, E.J.Kramer and S.Mahajan (Elsevier Science, Amsterdam, 2001)

     

    • J.T.Hoeft, M.Polcik, M.Kittel, R.Terborg, R.L.Toomes, J.-H.Kang and D.P.Woodruff ‘Photoelectron diffraction structure determination of Cu(100)c(2x2)-N’ Surf.Sci.492 (2001) 1-10
    • M. Pascal, C.L.A. Lamont, M. Kittel, J.T. Hoeft, R. Terborg, M. Polcik, J.H. Kang, R.Toomes and D.P. Woodruff
    , ‘Quantitative structural determination of the high coverage phase of the benzoate species on Cu(110)’ Surf.Sci. 492 (2001) 285-293
    • D.P.Woodruff, P.Baumgärtel, J.T.Hoeft, M.Kittel and M.Polcik ‘Direct methods in photoelectron diffraction; experiences and lessons learnt based on the use of the projection method’ J.Phys.:Condens.Matter 13 (2001) 10625-10645
    • J.-T.Hoeft, M.Kittel, M.Polcik , S.Bao, R.L.Toomes, J.-H.Kang, D.P.Woodruff, M.Pascal and C.L.A.Lamont ‘The local adsorption geometry of CO and NH3 on NiO(100) determined by scanned-energy mode photoelectron diffraction’ Surf.Sci.499 (2002) 1-14

     

    Papers in press or submitted for publication

    • M. Pascal, C.L.A. Lamont, M. Kittel, J.T. Hoeft, L. Constant, M. Polcik, A.M. Bradshaw, R.Toomes and D.P.Woodruff, ‘Methyl on Cu(111) – structural determination including influence of co-adsorbed iodine’ Surf.Sci in press
    • M.Kittel, R.Terborg, M.Polcik, A.M.Bradshaw, R.L.Toomes, D.P.Woodruff and E.Rotenberg, ‘The structure of the Pd(110)(2x1)-CO surface’, Surf.Sci in press
    • M. Kittel, D.I. Sayago, J.T. Hoeft, M. Polcik, M. Pascal, C.L.A. Lamont, R.L. Toomes, and D.P. Woodruff ‘Quantitative determination of the adsorption structure of carbonate on Ag(110)’ Surface Sci. submitted
    • R.L.Toomes, M. Polcik, M. Kittel, J.-T Hoeft, D. Sayago, M. Pascal, C. L. A. Lamont and D. P. Woodruff ‘Structure determination of methanethiolate on unreconstructed Cu(111) by scanned-energy mode photoelectron diffraction’ Surf.Sci. submitted

     

    Other references

    • J.Robinson and D.P.Woodruff ‘The local adsorption geometry of CH3 and NH3 on Cu(111): a density-functional theory study’ Surf.Sci. .498 (2002) 203-211