FUNDAMENTAL ASPECTS OF SURFACE SCIENCE

Euroconference:

Fax.: 33/388366987

Email : euresco@esf.org

Contractor :
European Science Foundation
"E.S.F"

Contract N°:
ERBFMMACT 960081

FUNDAMENTAL ASPECTS OF SURFACE SCIENCE

I. Synchrotron Radiation and Surfaces

II. Elementary Processes in Surface Reactions

I. Synchrotron Radiation and Surfaces

The European Research Conference in the series "Fundamental Aspects of Surface Science" was held in Castelvecchio Pascoli from the 6^th to 11^th June 1997 on the topic "Synchrotron Radiation and Surfaces". The scientific opportunities provided by the new, third generation of synchrotron radiation sources provided the impetus for this meeting. Ten such sources, three covering the hard x-ray region and seven for the soft x-ray range, are now in operation, under construction or in the advanced planning stage. Six (ESRF, ELETTRA, MAX II, BESSY II, SLS and SOLEIL) are in Europe. Compared to sources of the second generation, where the bending magnets mainly provide the radiation, third generation machines are characterised by a large number of straight sections containing undulators. These periodic magnetic structures provide light of exceptionally high spectral brilliance (which may be defined as the flux of photons emitted per unit source area and per unit solid angle). In surface science experiments the higher spectral, spatial or temporal resolution, or simply the higher flux, which are then available, can be used to perform a whole range of experiments which have hitherto not been possible, or perhaps only in an incomplete way. The purpose of the meeting in Castelvecchio Pascoli was to review progress in surface science experiments requiring synchrotron radiation, to identify the various research highlights from the new sources already in operation, and to establish future directions. These aims were largely achieved. The organisers and participants are extremely grateful to the EU for support from their Euroconferences activity within the framework of the TMR programme.

In the area of surface structural research D. Chandesris and D.P. Woodruff described recent progress in surface EXAFS and photoelectron diffraction, respectively. Chandesris’ talk concentrated on the use of multishell analysis in the determination of the structure of epitaxially grown Fe layers on Cu single crystals. Advances in methodology and data analysis in photoelectron diffraction were described by Woodruff who also showed several applications to various adsorbate systems. He emphasised the need for new third generation sources in order to exploit "chemical shift" photoelectron diffraction, which requires high spectral resolution at high flux. G. Materlik, C. Norris and G. Renaud concentrated in their presentations on the use of hard x-rays in surface structural research. Unfortunately, one prominent figure in this field, R. Feidenhans’l had to cancel at the last minute due to an unfortunate accident. (The conference sent him its best wishes for a speedy recovery). Materlik’s talk dealt with the x-ray standing wave technique as well as the new and exciting method of x-ray holography (Tezge and Feigel, Nature 380 (1996)49). New data from the ESRF were described. Norris reported the use of "conventional” grazing-incidence x-ray diffraction to follow epitaxal growth in metal-on-metal systems similar to those treated by Chandesris. Renaud’s talk was concerned with the application of the same technique to oxide surfaces, in particular a-Al₂O₃(0001) and MgO(100), thus opening up new perspectives in the study of ceramic surfaces.

Surfaces electronic structure played a central role at the meeting. Photoemission investigations on magnetic systems were described by G. Rossi and C.M. Schneider. Rossi showed how spin polarisation of the photoemitted electrons can be used to follow changes in the exchange-split valence bands of iron as a function of state of the surface. Schneider gave examples of the use of photoelectron emission microscopy (PEEM) to image the magnetic domain structure by utilising the contrast, which derives from magnetic circular dichroism (MCD). Further uses of PEEM were described by M. Kiskinova who has applied the spectromicroscopy variant of the experiment at ELETTRA to perform chemical mapping. A more sophisticated instrument for this purpose was reported by Th. Schmidt and first results were presented.

The talks of K. Horn and C. Papageorgopoulos were concerned with semiconductor interface formation studied at BESSY. Horn showed how the concepts of commutativity and transitivity in heterojunctions can be investigated by determining valence band discontinuities in photoemission spectra. He also reported strain effects in thin metal films on silicon which lead to marked changes in the occupation of the metallic surface states, demonstrating the subtle interplay between morphology and surface electronic structure. The influence of alkali metal adsorption on the surface properties semi-conducting transition metal chalcogenides was the subject of Papageorgopoulos’ talk. The relevance of these systems for solar energy conversion was briefly discussed. Also on the theme of electronic structure, M.C. Asensio described experiments at SUPER-ACO on Fermi surface mapping and its combination with photoelectron diffraction. Measurements on pseudomorphically grown thin films of Co and Fe on Cu(100) as well as on the evolution of the Fermi surface of the Ag/Si(111), Pb/Ge(111) and Sn/Ge(111) were reported. Bulk Fermi surface mapping of both graphite intercalation compounds and high Tc superconductors was also mentioned. T. C. Chiang showed startling effects in the surface and bulk photoemission from nearly free electron metals, which had gone unnoticed for many years. In particular, direct and indirect (surface) emission channels were found to interfere leading to Fano-type line shapes in the photoemission spectrum.

R. Gink and P. Rudolf described synchrotron radiation studies of the electronic structure of large adsorbed molecules. Fink’s lecture was concerned with very large organic molecules on silver single crystal surfaces; such systems show a surprisingly high degree of two-dimensional order. NEXAFS was used primarily for the characterisation and the data were interpreted with the help of Xa scattered-wave calculations. Rudolf showed how different synchrotron radiation techniques could be combined to obtain information on a specific adsorption system, namely, mono- and multilayers of C₆₀ on various clean and alkali-covered metal surfaces. Both photoemission, NEXAFS and autoionisation spectroscopy showed that the interaction of the monolayer with the substrate is not just Van der Waals in nature but that a chemisorption bond is formed. Infrared and STM data were also shown.

The talks of R. Uhrberg, J.N. Anderson and M. Methfessel were concerned with surface core-level shifts. Uhrberg reported data for silicon surfaces recently measured at low temperature at MAXlab. The various groups working in this area still do not agree on some of the finer points of assignment of the core-level lines for the most complicated example, namely, the 7x7-reconstructed Si(111) surface. Anderson showed how surface core-level shifts on clean and adsorbate-covered metal surfaces can be used to obtain information on surface structure. The examples he showed were CO on Rh(111) as well as the coadsorption of CO with oxygen and potassium. Further, a number of examples from alkali-metal adsorption on aluminium surfaces were cited to demonstrate that structural information on surface alloy formation can also be obtained using this method. Methfessel’s talk was devoted to the interpretation of such data using ab initio total energy density functional calculations. The latter were shown to reproduce surface core-level shifts to a relatively high accuracy.

Two further contributions from Invited Speakers covered the use of new techniques (at least in surface science). A. Nilsson reported recent data taken at the Advanced Light Source in Berkeley on x-ray emission spectroscopy and how it can be used to investigate adsorbed molecules. This method allows the contribution of the valence electronic states of the adsorbate to the chemisorption bond to be separated from those of the substrate. P. Hollins described the use of reflection-absorption infrared spectroscopy at synchrotron radiation sources and used several examples from recent work performed at Daresbury.

Altogether there were 45 posters, which were presented on two afternoons in very popular poster sessions. The abstracts of both the Invited presentations and the posters are included in the enclosed abstract book.

A topic, which concerned many participants in informal discussions, was the situation in Europe with regard to new sources. As noted above, six third generation machines are in various stages of operation, construction or planning, five of which are intended for the soft x-ray region. These machines have energies between 1.2 and 2.0 GeV and provide intensive undulator radiation between 50 and 2000 eV. Nothing is being done, however, for the VUV range between 5 and 50 eV. On the contrary, quite a number of existing second generation sources, used extensively for this energy range and including a few undulators, are likely to close down as the new facilities come on the stream, mainly to save running costs. This appeared to many participants to be rather strange, since one might have thought that in Europe the potential existed for a rational co-ordination of national science policies. We therefore have a situation in which in the next five years or so there will be a major growth in synchrotron radiation facilities, both in quality and in quantity, for photon energies above 50 eV, but a major reduction in low energy beamlines.

A.M. Bradshaw

II. Elementary Processes in Surface Reactions

The theme of this meeting was the identification and description on an atomic level of elementary dynamical processes underlying various surface phenomena in nature of which many are of direct technological importance. Some typical examples of such phenomena that were covered at the meeting are heterogeneous catalysis, oxidation, and growth. As demonstrated by numerous examples at this conference, the studies of such processes have now become very successful and fruitful thanks to recent developments of experimental and theoretical approaches. This progress builds on our detailed knowledge of the geometric and electronic structure of adsorbates on surfaces that has been gained by extensive and elaborate experimental and theoretical surface science studies over the last decades. More specifically, the approaches that formed the basis of the meeting included on the experimental side state-resolved molecular beam scattering, femtosecond laser techniques, atomic and molecular imaging and manipulation by scanning tunneling microscope (STM) and on the theoretical side reliable total-energy calculation schemes based on density functional theory (DFT) and classical and quantum molecular dynamics calculations.

The purpose of the meeting was to give a broad overview and status of the research frontier in this field, to identify and establish future directions. These goals were largely fulfilled at the meeting. A summary of the 27 talks presented at the meeting are given below. 19 of these talks were invited, 5 short talks were selected from the contributed abstracts and the short talks by Bartels, Katz and Walker were selected from the poster contributions after a general ballot. The meeting included two lively and "hot" poster sessions with about 80 excellent posters in total. The book of abstracts is available through world-wide-web at http://www.phchem.uni-essen.de/gsd/ED4_abstracts/abstracts.html

The interactions of hydrogen molecules and atoms with surfaces provide relative simple prototype examples of elementary processes in surface reactions that are amenable for detailed theoretical and experimental studies. In particular, chemisorption of H₂ on Cu and Pd surfaces have become model examples of activated and non-activated dissociation dynamics at surfaces, respectively. As demonstrated by several contributions, we have now achieved a rather detailed understanding of the dissociation dynamics on these surfaces. For instance, J. Hafner showed in the case of dissociation of H₂ on Pd(100) that sticking probabilities obtained from six-dimensional quantum dynamics calculations on potential energy surfaces (PESs) that are based on DFT calculations are in quantitative agreement with measured data from molecular beam scattering experiments. Steering plays an important role in the dissociation dynamics for this case and its relation to precursor-like behavior was elucidated by G. Darling using quantum and classical dynamics calculations. Results from DFT total-energy calculations of dissociative chemisorption of H₂ on Cu in the presence of co-adsorbates were used by D. Bird to develop a theoretical understanding of their poisoning and promotion effects. One potentially important aspect of the dissociation dynamics that is not well understood is the role of substrate excitations. State-resolved molecular beam studies by G. Sitz showed that such excitations play a significant role in scattering of rovibrationally excited H₂ molecules from Pd surfaces. An observation that needs attention from theory. The dissociation dynamics of H₂ on Si surfaces is still a controversial issue. From experiments based on optical second-harmonic generation, U. Höfer clarified this issue by showing that dissociation at step-edges play an important role, which was corroborated by DFT total-energy calculations by P. Kratzer. Using effusive beams of hydrogen atoms, A. Winkler showed that hot-atom reaction pathways for the abstraction reaction of hydrogen atoms with hydrogen-covered adsorbate surfaces dominates over direct Eley-Rideal reaction pathways.

Our understanding of the interaction of a wide variety of other small and technologically important molecules such as O₂, NO, N₂, and CH₂ with surfaces has also made substantial progress as demonstrated by several contributions at the meeting. Based primarily on molecular beam scattering, I. Zoric showed that the initial step of oxidation of Al(111), the dissociative chemisorption, is activated in variance with results from DFT total-energy calculations and speculated that non-adiabatic charge transfer plays an important role in this process. However, as demonstrated by J. Hafner, DFT is able to provide a good description of binding for the two molecularly chemisorbed states of O₂ on Pt(111). The importance of dangling bonds in the site-selective reaction of Br₂ molecules with second layer Ga atoms on a GaAs surface -- "The first shall be the last and the last shall be the first"-- was demonstrated by A. C. Kummel from a combined molecular beam scattering and scanning tunneling microscopy study. The possibility to create molecular beams of highly vibrationally excited NO molecules with up to v=20 was exploited by D. Auerbach in a pioneering study of its effect on scattering from and dissociation on a Cu(111) surface. Another interesting aspect of NO in its interaction with surfaces is its unpaired spin. In an infrared vibrational spectroscopy study complemented by DFT electronic structure calculations, W. Brown showed that the formation of (NO)₂ on an Ag surface and N₂O on a Pt surface could be understood from DFT calculations in terms of the fate of the unpaired spin of the chemisorbed NO. That steering dynamics is not only restricted to the dissociation of H₂ molecules was demonstrated by A. Walker from a molecular beam study of CH₄ dissociation on Pt(110)-(1x2).

More complex gas-surface processes than encountered by small molecules with solid surfaces were also covered by the meeting. G. Nathanson has developed a novel experimental technique for the study of scattering of molecules from liquid surfaces. For instance, he showed how this technique could be used to study reactions of gaseous N₂O₅ with sulfuric acid, which is a problem of great interest for the understanding of the role of sulfate aerosols in the heterogeneous destruction of the stratospheric ozone. Scattering of clusters from surfaces is another example of a complex gas-surface process and was covered by A. De Martino's talk. For instance, an unexpected finding was that the translational distribution of scattered nitrogen molecules in nitrogen clusters impinging on graphite was thermal while the rotational distribution was non-thermal with a very cold component.

The availability of femtosecond lasers has opened up the possibility to study in real time and to control the electron and ion-core dynamics behind surface reactions. On metals surfaces, pulsed laser-induced reactions are predominantly driven by the generated hot-electron distribution. Results from interferometric time-resolved two-photon studies of the phase and energy relaxation of such distributions on metal surfaces were reported and discussed by H. Petek. He stressed the long d-band hole lifetime on Cu and its significance on surface photochemistry and laser processing. One impressing result was that he had been able to resolve the effects of the wave packet motion of a Cs atom on the excited state potential down to a 50 as (1as = 10^-18s) time scale. Using time-resolved two-photon photoemission, M. Wolf demonstrated the possibility of tailoring lifetimes of surface electron states by forming sandwich structures. He also reported results from the photochemistry of CO with O on a Pt surface induced by femtosecond pulses, for which oxidation and desorption was found to vary differently with pump-probe delay. From a theoretical point of view the description and modeling of electronically excited states is a difficult and unsolved problem. A important contributions towards the solution of this problem was provided by Th. Klüner for NO on NiO from quantum-chemistry calculations of the ground and excited state PESs and quantum molecular dynamics calculations on these PESs. The observed bimodality of the product distributions was shown to becaused by wave packet bifurcation on the excited state PES. Charge transfer processes in oxygen dissociation was modeled by G. Katz using semi-empirical diabatic PESs and wave packet dynamics.

The scanning tunneling microscope is a powerful tool to study surface diffusion and growth on an atomic scale. Using this technique F. Besenbacher presented results for the relative role of single and double jumps in diffusion of Pt adatoms, Pt dimer dissociation and association, and the enhancement of O and H on diffusion of Pt adatoms on Pt(110)-(1x2). In the latter case, he proposed an explanation in terms of "skyhooks" formed by H- and O-Pt ad-dimers. In particular he demonstrated that combined experimental and theoretical studies of surface reaction on atomic level have been used to design an improved surface-alloy catalyst for the steam forming reaction. A theoretical description and analysis of surface diffusion of ad-atoms and ad-dimers on Au(110) surfaces were presented by R. Ferrando using classical molecular dynamics and empirical many-body potentials. In particular, he corroborated the suggestion by F. Besenbacher that the observed double jumps were caused by metastable walks and identified various pathways for dimer diffusion, such as concerted jumps, dissociation-reassociation and leap-frog. Another useful technique to study growth is high resolution surface diffraction techniques such as, for instance, He scattering. B. Poelsema showed using such techniques that, for instance, the morphology of the growth of Cu on Cu(100) is strongly influenced by the angle of incidence of the Cu atoms. An effect that might involve transient mobility of the incident Cu atoms.

A most interesting development during the last years is the possibility by various means to manipulate and control the formation of nanoscale and atomic scale structures on surfaces. K. Kern demonstrated by several examples how nanostructures can be formed by exploiting the hierarchy of activation barriers in diffusion-controlled growth, using cluster softlanding on Ar-covered Pt(111) surfaces, and using vicinal Pt(117) and superstructures on Ag/Pt(111) as one- and two-dimensional templates. On structured-Ag(111) surfaces, he used scanning tunneling spectroscopy to characterize the behavior of the Shockley surface states. G. Dujardin showed how the STM can be used in a single atom manipulation mode to encode chemical reactivity of individual atomic sites on surfaces. In particular, he was able to extract single Ge atoms by tip contact on a Ge(111) surface and show that the vacancy sites are reactive to O₂ dissociation whereas the vacancy-free surface is non-reactive. The possibility to perform single molecule manipulation was demonstrated by L. Bartels through using the tip of a STM to perform tunneling electron-induced desorption of CO molecules from Cu(111).

The possibility to image the kinetics of surface reactions on an atomic scale were demonstrated by J. Wintterlin using variable temperature and a fast STM for the oxidation of CO and trimerization of acethylene on Pt(111). His animations of the STM images provided a nice and pedagogical illustration of the various kinetic processes involved in a surface reaction under a stationary situation. His results suggested that, for instance, only bridge-bonded CO is reactive. That STM images of O₂, CO and C₂H₂ molecules on Pd(111) can be simulated in unprecedented detail was demonstrated by P. Sautet using calculations based on the generalized Green function transfer matrix method. For instance, he could decompose the tunneling current into various molecular states.

The organizational and administrative help by the office in Strasbourg and provided by conference secretary were most valuable and the choice of the venue was in general very much appreciated by attendees. The international advisory committee thought it was very important to continue the series of European surface dynamics meetings and was so happy with the organization of the present meeting as a Euroconference that they decided to make an application to ESF for a new meeting.

Support from the European Commission Euroconference activity of the Training and Mobility of Researchers (TMR) Programme is gratefully acknowledged.

Mats Persson, October, 1998
Chairperson