摘要:

Faraday Discuss. Chem. SOC., 1990, 89, 259-273 Core-level Shift Spectroscopy on Tungsten Surfaces Overlayer and Underlayer Adsorption G . P. Derby and D. A. King Department of Chemistry, University of Cumbridge, Lensjield Road, Cumbridge CB2 1 E W Core-level binding energy shifts in the 4f,,, level of W previously reported from a range of clean single-crystal surfaces are consistent with a simple linear relationship to an effective coordination number. Here we report data for the adsorption of H2, O2 and CO on W{110}, where conventional overlayers are formed, and for N2 on W{ 1 lo}, where an underlayer is formed. The observed adsorbate-induced shifts are consistent with a model which accounts for changes in the coordination of surface W atoms due to adsorp- tion, invariably leading to an increase in binding energy, and also changes due to charge transfer between adsorbate and substrate.It is pointed out that these two effects are additive for electron-withdrawing adsorbates, but tend to cancel each other with electron-donating adsorbates. Recent observa- tions for alkali-metal atom adsorption are consistent with this model, and do not imply non-ionic adsorption. At clean and adsorbate-covered metal surfaces the electronic core-level bonding energies of surface atoms are shifted with respect to those of bulk atoms.'-3 At a microscopic level, these surface core-level shifts (SCLS) originate from both initial-state effects, including purely environmental contributions and differences in electronic configurations between bulk and surface atoms, and final-state effects arising from modifications of the screening of the core hole following photoemission.For a wide range of clean W crystal planes the dependence of the magnitude of the core-level shift was found to be almost monotonically dependent on the effective coordination of atoms in the surface region, provided that the third coordination shell is i n ~ l u d e d . ~ With adsorbates, large shifts from clean surface binding energies are observed for electronegative species, consistently to higher binding energy, while for electropositive species such as the alkali metals relatively small shifts, to higher or lower binding energies, have been observed.5-8 The latter observation has recently led to the conclusion that the alkali metals are covalently bonded at all coverages on W{llO}, in paradoxical contrast with 0 on the same surface.' In the present paper we examine in some detail the core-level shift behaviour on W{ 110) with electronegative adsorbates, including H, 0, P-CO and virgin CO, in conventional overlayer sites and one case (P-N) where an underlayer is formed.The results are examined in terms of available microscopic models. Experimental The experiments were performed on the SRS at Daresbury in a VG ADES 400 ultra-high vacuum chamber. The core levels studied are the 4f,,,7, which have proved to be very surface sensitive at kinetic energies ranging from 20 to 70eV. A torroidal grating monochromator provided radiation at photon energies between 60 and 95 eV, well suited to the present work.The base pressure was better than 1 x lO-'"Torr,l- and a typical working pressure was less than 2 x lO-"'Torr. An angle-resolving analyser (4" half-width) was used to collect the electrons emitted normal to the surface. The light was incident i 1 Tom= 101 325/760 Pa. 259260 Core-level Shqt Spectroscopy on Tungsten Surfaces r 1 0.01 ' 1 32.4 32.0 31 .O 30.4 binding energy/eV Fig. 1. W 4f,,, core level spectrum taken from clean W(110) at 80 K, at normal emission. The photon energy is 70 eV, and the incidence angle is 70 O to the normal. The crosses represent raw data points. The data have been fitted, and the full-line peaks are the Doniach-Sunjic lines without any instrument or phonon broadening included. The full line through the data points is the sum of the individual components inclusive of all broadening factors.( a ) Bulk, ( b ) clean surface. at 70" to the surface normal, to maximize the relative surface sensitivity.6 The total resolution of the system, monochromator and analyser, was estimated to be 0.16 eV from the Fermi edge: this value is in good agreement with the calculated resolution for the system. The crystal was cut to within 0.5 O of the required plane, and cleaned by repeated heating to 1800 K in Torr oxygen and flashing in vacuum to 2400 K. The surface was considered to be clean when no further increase in the amplitude of the surface peaks, relative to the bulk, could be seen in the core-level spectra. The spectra from the clean surface were taken as multiple scans (170 s per scan), the crystal being flashed to 2500 K before the start of each scan to reduce surface contamination to a minimum.Temperature measurement was effected by use of a W-Re (3%)/W-Re (25%) thermo- couple. Results and Discussion The Clean Surface: Reference Spectrum On the closely packed { 110) surface of tungsten, only two, well separated, core levels can be seen. Data are shown in fig. 1. To extract the core-level binding energies from these data a non-linear least-squares fitting technique has been used, in which each displaced core level is represented by a single peak. The lineshape of the peak is influenced by several processes: the underlying shape is Lorentzian, arising from decay of the core hole, but excitations of electron-hole pairs at the Fermi level contribute to an asymmetry, resulting in a Doniach-Sunjic (DS) lineshape.Bulk and surface phonon excitations caused by the photoemitted electron additionally produce a Gaussian broad- ening, and inhomogeneities on the surface may also contribute to Gaussian broadening.50( 400 - I v) v) - G 300 : S eD .- w. 200 100 G. P. Derby and D. A. King * : , : : . : : : : : - : : ; :, 26 1 I . , . i . I . 33.0. 32.6 32.2 31.8 31.4 31.0 30.6 binding energy/eV Fig. 2. W 4f7,2 core-level spectrum taken under the same conditions as for fig. 1 from a fully oxidised W{ 1 lo} surface, showing an isolated bulk peak. ( a ) Oxide, ( b ) bulk. Finally, the finite resolution of the experimental system used to measure the spectrum must be included and this is also taken to be Gaussian. From the earlier work of Wertheim et al.,9 it was concluded that the bulk and surface peaks on W{ 110) have the same Lorentzian widths and asymmetry parameters (Lorent- zian FWHM, 50 meV; asymmetry parameter, 0.06).These values were assumed in the analysis depicted in fig. 1, and the binding energy of the bulk peak was fixed to the value determined from the peak position observed in the heavily oxidised surface (fig. 2), where the bulk peak is completely separzited from the surface peaks, with all selvedge peaks shifted to higher binding energy by +0.97 eV relative to the clean surface peak. On the clean surface, the surface-to-bulk peak separation is 300 meV, in agreement with the data of Duc et a1.l’ and Purcell et a1.,l1 and rather smaller than that reported recently by Riffe et al.(321 meV).’’ As reported below, we do observe that the gap between the surface and bulk peaks is narrowed with hydrogen adsorption, but great care was taken to avoid hydrogen contamination on the clean surface spectrum. The bulk peak in fig. l ( b ) is well fitted with a DS Lorentzian FWHM of 50 meV, asymmetry parameter 0.05, and Gaussian broadening (phonon + instrument) of 190 meV. In their recent paper, Riffe et al.” report a Lorentzian FWHM of 60 meV and asymmetry parameter of 0.035. Significantly, they also report difleerent values for the surface peak: 84meV and 0.063. While these differences are important to the physics of the emission process, they do not significantly affect the analyses from singular and stepped surfaces previously published, or those presented here.The Lorentzian broadening is swamped by the Gaussian broadening at the instrumental resolution deployed. Overlayer Adsorption We compare core-level shift data obtained on the W{ 110) surface after varying exposures to Hz, 02, and CO. The results are summarised in fig. 3.262 I I I Core-level Shift Spectroscopy on Tungsten Surfaces I I I I I I I P D I I I I I P s I I B I I I I I , I B I VI. ( f ) N, Ni B ] N: S binding energy/eV Fig. 3. Summary of W 4f,,, core-level shift data, showing the surface (S) and bulk ( B ) components of the clean-surface spectrum, and the positions of all lines fitted to the spectra with H, 0, p-CO (dissociated), virgin CO (molecular) and P-N adsorbed on W{ 1 lo}, respectively. Not all these peaks are observed at any one coverage (see text).( a ) Clean W( 1 lo), (6) H/ W( 1 lo), ( c ) O/ W( 1 lo), ( d ) P-CO/ W( 1 lo), ( e ) V-CO/ W( 1 lo), cf) P-N/ W( 110). Hydrogen on W{ 110) As hydrogen is adsorbed on W(110) to half-monolayer coverage at 300 K the surface core-level peak at a binding energy of 31.08 eV is rapidly attenuated and replaced by a component estimated to be ca. 110meV below the binding energy of the bulk peak, which is at 3 1.38 eV. An increase in the bulk peak intensity indicates a second component at the same binding energy as bulk W atoms. At low coverages a further peak is estimated at ca. 200 meV below the bulk peak binding energy. The results may be interpreted in terms of the structural model of Holmes and King," shown in fig. 4. It is suggested that the low-coverage peak shifted by -200meV from the bulk peak is due to a surface W atom bonding to a single H adatom; while the two peaks observed at higher coverage at -1 10 and 0 meV from the bulk peak are assigned to surface W atoms bonded to two and to three H adatoms, respectively.The shift ofG. P. Derby and D. A. King 263 Fig. 4. Hard-sphere representation of the Holmes and KingI3 model for the (2 x 1) overlayer structure formed by H on W{110}, showing two distinct adsorption sites, threefold hollow and short bridge. (Open circles represent top-layer W atoms.) the surface layer W 4f binding energy from that of the clean surface is therefore roughly proportional to the coordination with H adatoms, -100 meV per H atom. that the work function decreases almost linearly with hydrogen coverage on W{110}, the maximum observed change at saturation being ca.-500 meV. There is general Oxygen on W( 110) The influence of adsorbed oxygen is demonstrated in fig. 5 . The clean surface peak in the W 4f,,* spectrum is rapidly attenuated with exposure to 02, being eliminated after an exposure of one Langmuir. At very low exposures a peak appears to build in between the clean-surface and bulk peaks, shifted by ca. -80 meV with respect to the bulk peak. However, when the well known half-monolayer p(2 x 1) structure is formed, this peak is replaced by higher binding energy peaks, with a dominant peak at +370meV with respect to the bulk peak. This is attributed to surface W atoms bound to two oxygen adatoms. The deconvolution is consistent with two further peaks, shifted by +200 and +610meV, respectively, from the bulk peak, which are attributed to surface W atoms singly and triply bound to 0 adatoms.Once again, the shift appears to be roughly proportional to the coordination with adatoms, -200 meV per 0 adatom. When oxygen is adsorbed on W{110}, there is a small work function increase, of +0.20 eV at 8 = 0.5. At low coverages the work function actually decreases, which may be associated with adsorption at defects, reconstruction, or a pre-island phase. Carbon Monoxide on W( 1 10) The adsorption of CO on W{110} produces a diverse range of states, depending on substrate temperature and exposure.”-”’ At 100 K adsorption produces only a single molecular adsorption state, the virgin state, bound through the C end of the molecule.Heating the surface causes some desorption and some conversion into a dissociated state, the p state. A molecular state is coadsorbed with the p state at temperatures264 Core-level Shiji Spectroscopy on Tungsten Surfaces h c) v1 .- c 2 0.01 . J 32.2 32.0 31.8 31.6 31.4 31.2 31.0 30.8 30.6 0.oL . 32.2 32.0 31.8 31.6 31.4 31.2 31.0 30.8 30.6 binding energy/eV Fig. 5. Fitted core-level spectra from oxygen on W(l10) at room temperature for exposures of ( a ) 0.05 L and ( b ) 0.75 L. The narrow spectrum below each curve is the difference spectrum between the data points and the best fit line. below 500 K. The molecular virgin and dissociated p states are readily distinguished using XPS; for example, the C 1 s peak exhibits a binding energy at 285.5 eV for virgin CO which is shifted to 283.0eV in the dissociated state.” Core-level shift spectra for W4f,,, obtained on exposure to CO at a surface tem- perature of 100 K are shown in fig.6. The initial doses were not started until the crystal temperature was below 120 K after having been flashed more than once to greater than 2300 K to remove any contaminating residue from the previous dose. After each ‘cold’ surface spectrum was collected, representing the virgin CO overlayer alone, the crystal was heated to 600 K and re-scanned at this temperature to avoid any complications dueG. P. Derby and D. A. King 265 -__ - .-- 1.41 1.3j 1.21 1 la'! .o 0.9.- 0.8 .- 0.7 -- 0.6 -- : o.olpa+--+--L . *-- * --.--+- --*- % 33.0 33.2 33.4 33.6 33.8 34.0 34.2 34.4 34.6 34.8 o.o-------+--- ~ 33.0 33.2 33.4 33.6 ,333 34.0 34.2 34.4 34.6 34.8 kinetic energy/eV Fig.6. Fitted core-level spectra obtained after ( a ) adsorbing CO at 100 K, where the adlayer is in a molecular state, and ( b ) after heating to 600 K, where the adlayer is dissociated. ( a ) 100 monolayers, 100 K; ( b ) 600 K.266 Core-level Shift Spectroscopy on Tungsten Surfaces h Y m .- E 3 E E .- - a E 0.0 0.5 1.0 0.0 0.5 surface coverage/monolayer 1 .o Fig. 7. Variation of intensities of each of the adsorbate-induced components in the core-level spectra for CO on W(110) with surface coverage: ( a ) the /3 components from the 600 K spectra; (i) P s , (ii) Po, (iii) P I ; ( b ) the virgin components from the 100 K spectra; ( i ) V,, (ii) V,, (iii) PT.to a-CO readsorption. The latter spectra therefore correspond to the pure p-CO state. Fig. 6 includes examples of the fitted spectra for both the initial adsorbed layer and for the associated heat-treated layer. The effects of phonon broadening were included in the fitting procedure for the 600 K data by inclusion of extra Gaussian broadening" within the instrument broadening function parameters. From a comparison between fig. 6 and 5, there are strong similarities between the p(2 x 1) structure formed by the p-CO state and the p(2x 1) overlayer structure formed as the result of pure oxygen adsorption. We can therefore be confident about drawing comparisons between the respective core-level spectra. In effect, it would appear that C and 0 adatoms from p-CO produce a similar influence to an adlayer of 0 adatoms alone.Three peaks are clearly derived from the p-CO state, labelled P , , ps and pD. The variation of each component with surface coverage is illustrated in fig. 7 ( a ) . The PI peak is assigned to surface W atoms bound to a single adatom; the ps and pD peaks are assigned to a (2 x 1) structure with C and 0 atoms occupying three-fold hollow sites. This analysis follows the analysis of the O/W{110} system very closely. The pT. peak, lying on the high-binding-energy edge of the spectra, has low intensity but appears to be a real feature; by comparison with the O/ W{ 1 lo} analysis, this is assigned to surface W atoms triply coordinated to adatoms. From data based on a range of different techniques,'5-20 the adlayer formed at 100 K is composed of virgin CO alone.The two main peaks associated with this layer, V, and V, (fig. 6), develop in intensity with coverage as shown in fig. 7 ( b ) . The V, peak develops rapidly at very low coverages, and then decays as VB grows to become the dominant feature. Additionally, the binding energy attributed to V, increases with coverage, from +lo8 meV shifted from the bulk peak at 8 =0.1 monolayers to +280 meV at 0.7 monolayers. As it does so, it also increases in Lorentzian half-width from 210 to 328 meV. At all coverages this component is significantly broader than that observed for other components, e.g. 50 meV for the bulk peak. At saturation coverage in the virgin CO state on W{ 110) the work function is increased by 0.6 eV, showing net charge transfer to the adlayer. We tentatively assign VL and V, to two different molecularly adsorbed CO species: a linear species V,, with a binding- energy shift of -200 meV with respect to the bulk peak; and a bridged species, VB, with a shift of +lo0 to +280 meV, depending on coverage.This is consistent, for example, with the overall increase in work function of 0.6 eV observed at saturation coverage in the virgin state, when the VB peak is dominant: the bridged species is dominated byG. P. Derby and D. A. King 267 d.rr* backbonding, and hence by charge transfer to the adsorbate. On the other hand the linear species is dominated by 0- donation to the metal. The shift in the binding energy for VB can be attributed to crowding in the adlayer, with W atoms being coordinated to a single bridged species at lower coverages; to one bridge and one linear or two bridge species at higher coverages; and even to three bridged species at full coverage. The enhanced width of the peak attributed to the bridge-bonded species is attributed to a combination of two sources.First, the existence of a range of coordination possibilities for surface W atoms to linear and bridged species will lead to inhomogeneous broadening, and the increased half-width with increasing coverage appears to be con- sistent with this. Secondly, in the bridge-bonded position the CO 27r* level is partially occupied, straddling the Fermi level. Its occupancy will vary with the vibrations of the surface molecule, leading to a variable charge on the adsorbate, and hence an increased spread of core-level binding energies for the substrate atoms involved in the bonding.Underlayer Adsorption Based on ion-scattering spectra, Auger depth-profiling and LEED data Somerton and King concluded that dissociatively adsorbed nitrogen, P-N, on W( 1 lo} forms an under- layer structure, sandwiched between the first and second W layers at the surface.‘2 The adsorption rate of nitrogen on tungsten exhibits a large crystallographic ani~otropy,~’ with an initial sticking probability that varies from 0.73 on the (310) surface to 5 x lo-’ on W(110) at 300 K. Formation of the p-state is believed to occur via dissociation of the molecule at adjacent pairs of four-fold (100) sites, which on the close-packed { 1 lo} surface are only to be found at defects, such as steps.” The exothermicity of the chemisorption process produces ‘hot’ adatoms which are then free to migrate and populate the entire surface.At 300 K this process is inefficient but may be enhanced by repeated heating to 600 K and readsorption. The heating assists migration away from the region adjacent to the defect sites, allowing further dissociative adsorption to occur on cooling. This ‘cycling’ is carried out in an ambient pressure of 2 x lop6 mbar of N2 until a saturation coverage of 8=0.25 monolayer, as indicated by a (2x2) LEED pattern,23 is reached. We found the best quality LEED pattern to be achieved after a total exposure time of 3-5 min, involving 6-7 heating-cooling cycles. Core-level spectra taken from the (2 x 2)-N structure show a rather unnatural flat- topped distribution of photoelectron binding energies, as shown in fig.8. In this figure a direct comparison is made between the spectrum from the clean surface and that from W{ 1 10) (2 x 2)-N, obtained under identical conditions, and a further unusual feature is immediately obvious. In addition to almost total attenuation of the clean-surface surface peak S, the bulk peak B is attenuated to roughly half the clean-surface intensity. This contrasts with all of the overlayer cases discussed above, where virtually no attenuation of the bulk peak was observed. Qualitatively, the underlayer model provides a relatively simple means of understanding the spectrum. From our previous study of N on W{ an overlayer system, it was concluded that N induces shifts from the bulk position in surface W atoms of -100meV (coordinated to one N adatom), +200meV (two N adatoms) and +400 meV (three N adatoms).In the present system, each tungsten atom can be coordinated to a single surface W atom only, since the maximum coverage is only 0.25 monolayer, and yet there is clearly a component with a shift of ca. +500 meV from the bulk peak. This is readily understood in terms of the rumpled underlayer model for W{ 110) (2 x 2)-N, depicted in fig. 9. The coordination of W atom C’ to the rest of the surface is significantly reduced, and charge transfer from this atom to the N atom beneath it could be enhanced due to a reduction of screening by valence electrons of the substrate. Furthermore this N adatom is also coordinated to the underlayer atoms D, E and K (fig.9) which may be expected to produce a small core-level shift to higher268 Core-level Shift Spectroscopy on Tungsten Surfaces 33.0 34.0 35.0 kinetic energy/eV Fig. 8. A comparison of raw core-level spectra from ( a ) the clean W{ 1 lo} surface, and ( b ) W{ 1 lo} with a ( 2 x 2 ) nitrogen underlayer structure. The spectra were collected and recorded under identical conditions. ( i ) Bulk, (ii) surface. binding energies. These atoms would, in the clean surface, contribute to the intensity of the bulk peak, and we therefore also have a ready explanation for the observed attenuation of the bulk peak (fig. 8). Attempts to quantify this picture are frustrated by our inability to resolve the component peaks in the flat-topped spectrum.The spectrum can only be consistent with a large number of component peaks, each from W atoms differently coordinated within the selvedge. The simplest fitting procedure was applied to the data, based on the assumption that the lineshapes of all component peaks were identical to that of the bulk, and then using the minimum number of components to fit the data. The result is shown in fig. 10. The bulk peak B, the clean surface peak S and the peak Ns are readily identified, the latter being attributed to the rolled surface W atom C’ (fig. 9). This peak is shifted by 650 meV to higher binding energy compared with the clean-surface surface peak. Within the (2 x 2) structure, a proportion of the surface atoms are unaffected by the presence of the underlayer N adatoms, which is consistent with a residue from the clean-surface peak S.The peaks labelled N$ and Ng could be assigned to, respectively, underlayer W atoms at J (fig. 9) interacting with the underlayer N adatom, and shifted by 170 meV from the bulk position, and atoms of type A and B in the surface, shifted by 130 meV from the clean-surface peak S. Two minor features remain to be assigned, one at each end of the spectrum. The feature with a binding energy of 30.9 eV may be due to the generation of low-coordination defect atoms at the surface, produced in the process of generating the underlayer structure. However, it is also possible that these two features are artefacts of our analysis, and would be removed by allowing peaks Nk and Ng to broaden.G. P.Derby and D. A. King 269 Fig. 9. Hard-sphere representation of the surface rumpling generated at the W{ 110) surface by an underlayer N atom (filled circle) in the (2x2) structure. ( a ) Clean surface, ( b ) nitrogen underlayer. When a surface oxide is formed on W{ 1 lo}, the core-level spectrum, described above [fig. 2( b ) ] , shows a clear distinction between bulk and oxide W atoms. The centre-of- gravity of the oxide peak is shifted by 150 meV, to higher binding energy, from the bulk peak. This compares, for example, with a shift of 610 meV for a surface W atom bonded to three chemisorbed 0 atoms, as described above. The larger shift is a measure of the ionicity of the W-0 bond in the oxide layer. Note that the shift of the surface W atom peak associated with nearest-neighbour bonding to a single underlayer N atom is 620 meV with respect to the binding energy of the clean surface peak, considerably more than singly coordinated W atoms in the W{100}-N overlayer, of 300 meV.The implication is that the charge on the relatively unscreened surface W atom C’ in the presence of N is considerably larger than that on an embedded W atom bonded to a single N atom in the overlayer system N/ W{ 100). Discussion Adsorption of hydrogen, oxygen, nitrogen and carbon monoxide on metals invariably produces a shift in the core-level binding energy of surface metal atoms towards higher binding energies, as also demonstrated in the present work for these adsorbates on W(110). These results have led almost universally to an explanation based on electron transfer of metal d electrons to the adsorbate.However, Apai et aL5 studied core-level shifts in Pt single-crystal surfaces produced by the electron-donating adsorbates NH3 and K, and still reported shifts towards higher binding energies (as with electron- withdrawing adsorbates), apparently contradicting this model. They concluded that K causes a rehybridisation of surface Pt atoms, leading to a decreased d population.270 Core-level Shgt Spectroscopy on Tungsten Surfaces 0.8 0.6 0.4 .+ ' /i i . . 0 . o L . * . . '1 32.6: 32.2 31.8 31.4 ' 31.0 30.6 0.08 it binding energy/eV Fig. 10. Fitted core-level spectrum obtained from the W{ 1 lo} (2 x 2)-N structure; the components have been fitted on the assumption that they all have the same half-width as the bulk peak. Core-level shifts are more sensitive to d than to s and p occupancy, because the latter are spatially more diffuse.Referring to self-consistent calculations of Wimmer et ~ 1 . ~ they also argue that the valence electrons of the alkali-metal atom are polarized, but not significantly ionized: Cs is believed to be covalently bonded to the surface. For K on Pt and Au, Duckers and Bonze125 reported small shifts towards higher binding energy. Soukiassian et al.' reported shifts to lower binding energies, for the first time, for Cs on W{ 100) and Ta(100). The shifts were small: 50 meV for W and 100 meV for Ta, but nevertheless the authors concluded that the observations were consistent with donation into the surface transition-metal d band.Now Riffe et ul.' have reported results for Cs, K and Na on W{110} which show small shifts to lower binding energy, in the range 12-28 meV at monolayer coverage. Once again quoting the band structure calculations of Wimmer et ~ l . , ~ ' they conclude that there is little if any charge transfer to the substrate. This non-ionic bonding picture extends to low coverages. On the other hand, the results of Riffe et al.' for 0 on W{ 1 lo}, producing a shift to higher binding energy in agreement with the present results, are attributed by them to a substantial degree of charge transfer. The authors are led to conclude that 0 adatoms, which produce a small work-function increase of 0.2 eV on W{ 1 have a high degree of ionic character, while Cs, which produces a much larger work-function decrease of 3.9eV, is non-ionic.We shall see that this conclusion is incorrect. Adsorption of an atomic layer on W which is neither electron-withdrawing nor electron-donating must itself lead to an increase in the core level binding energy of top-layer substrate atoms. Not surprisingly, we have found this with Au adsorption,' and, trivially, we note that a W layer on its own lattice clearly shifts the underlying core-level energies towards the bulk position. Thus, Purcell et aL4 report a simpleG. P. Derby and D. A. King 27 1 empirical relationship, based on data from six different crystal planes of tungsten, between the core-level shift A and the surface coordination: Asn(eV) = 0.953 - 0.044 C N,/ rJ J = 1 where E is the core-level binding energy, and the effective coordination is considered important out to the third shell j at distance rJ and with an atomic occupancy Nj.When considering the effect of an adsorbate on the core-level binding energies of the substrate, two effects therefore need to be accounted for: increased coordination of substrate surface atoms, which will increase the core-level binding energy, and charge redistribu- tion. Eqn (1) does, of course, contain contributions from both effects even for the case where adatom and substrate are the same element. But we point out that a layer of tungsten adsorbed on W(110) causes a shift in the core-level binding energy of the original top layer to higher binding energy, of 300 meV. This does not imply that W is an electron-withdrawing adatom on its own lattice.When a shift is induced by an adsorbate, it is necessary to define the quantity where E A is the core-level binding energy of the top-layer substrate atom after adsorption. It is useful to consider separately the contribution to the surface atom core-level binding energy arising from changes in the environment external to the atom,29 Aenv, and changes arising from alterations in electronic configuration. The effective one-electron potential necessarily increases at a surface, decreasing the core-level binding energy of surface layer atoms with respect to their bulk counterparts, even if no charge redistribution is accounted for. Adsorption will cause an increase in Aenv, towards the value Ebulk. Charge redistribution in the initial state takes two forms: a configurational change within the atom involving interconversion between s and d states, producing a shift Asd, and a redistribution involving net charge donation between atoms, producing a shift A,, .In moving an atom from the bulk to the surface, a conversion occurs from d to s states, producing an increase in the core-level binding energy. Adsorption of an electron- donating species can be expected to increase the d state occupation. Finally, we must account for a shift due to final state relaxations, Arelax. Thus the total shift is given by: Benesh and Haydock’” show that for surface layer atoms on W{lOO} the terms Aen, and Asd are both large (0.6-0.8 eV) and of opposite sign. From their analysis and more recent results for W{lOO}’* the conclusion is inescapable that there is a large final-state relaxation shift, ca.0.4eV, contributing to the data from clean surfaces. This may be a result of a high density of states at the Fermi level, associated with surface states and resonances. Now we note that for the shift induced in top layer substrate atoms by electronegative adsorbates A,, should be positive, while for electropositive adsorbates it should be negative. Since the overall coordination shift is necessarily positive (towards higher binding energy), charge flow between adsorbate and substrate tends to cancel the shift in the case of electropositive adsorbates, and reinforce it in the case of electron- withdrawing adsorbates. This is in full agreement with the general observation that the observed shifts for alkali metals are usually very small, and can be towards higher or lower binding energies, never exceeding the bulk binding energy.The shifts for elec- tronegative adsorbates are large, often exceeding the bulk value. These small shifts for alkali metals on W{ 1 lo} are due to a cancellation of two effects, and not to the non-ionic nature of the bond: on the contrary, if the bond were non-ionic a large shift should be observed towards the core-level binding energy of bulk atoms. AAS = Aen, + Awi i- Ael+ Arelax272 Core-level Shift Spectroscopy on Tungsten Surfaces The data obtained for H, 0 and CO overlayer structures on W{ 1 lo}, summarised in fig. 3, can be readily rationalised in terms of this model.H adatoms, with a W,-H bond energy of ca. 280 kJ mol-' (2.9 eV), produce shifts of ca. +lo0 meV per bond from the clean-surface peak S. The charge redistribution at this surface actually produces a small work-function decrease, but the results are consistent with a coordination shift combined with a charge-transfer shift involving some transfer to the H adatom. For 0 and C adatoms, formed from 0, or from CO, where the bond energy to the substrate is ca. 800 kJ mol-' (8.3 eV), i.e. comparable to the bond energy of W to its own lattice, the observed shifts from the clean-surface peak are considerably larger, consistent with a large coordination shift and a significant donation of substrate d electrons to the adatoms. The molecular virgin CO state appears to produce two core level peaks, one at a binding energy between clean surface and bulk peaks and a second, broad peak at higher binding energy.We have assigned these, respectively, to linear and bridge-bonded species bonded to surface W atoms. Although the bridge-bonded species is less stable, since it develops at higher exposures, the W2-C bond strength is higher than that of W-C in the linear species: increased d v " backdonation strengthens this bond, while weakening the C-0 bond. This may lead to a larger coordination shift for W atoms bonded to the bridged species. The underlayer case, (2x2)-N, is significantly different. Here, at a low coverage (0.25 monolayer) a large shift is observed (peak N,) indicating that top layer W atoms are poorly screened: this will influence both the initial-state charge transfer and final-state relaxation, combining to give the observed large shift.In additiori, the multicomponent nature of the spectrum demonstrates bonding to second-layer W atoms, causing also a decrease in the bulk peak intensity. We are grateful to the S.E.R.C. for the provision of beamtime at the Daresbury Syn- chrotron Radiation Facility, and for a Studentship to G. P. Derby. We also acknowledge useful discussions with Dr Greg Benesh. References 1 P. H. Citrin and G . K. Wertheim, Phys. Rev. B, 1983, 27, 3167. 2 D. Spanjaard, C. Guillot, M. C. Desjonqukres, G . Triglia and J. Lecante, S u r - Sci. Rep., 1985, 5, 1. 3 K. G. Purcell, J. Jupille and D. A. King, in Solvuy Conference on Sucfuce Science, ed. F. W. d e Wette (Springer, Berlin, 1988), p.477. 4 K. G. Purcell, J. Jupille and D. A, King, Surf: Sci., 1989, 208, 245. 5 G . Apai, R. C. Baetgold, P. J. Jupiter, A. J. Viescas and I . Lindau, Sue- Sci., 1983, 134, 122. 6 P. Soukiassian, R. Riwan, J. Cousty, J. Lecante and C. Guillot, Surf: Sci., 1985, 152/3, 290. 7 D. M. Riffe, G. K. Wertheim and P. H. Citrin, Phys. Rev. Lerr., 1990, 61, 571. 8 A. Awobode, G . P. Derby and D. A, King, t o be published. 9 G. K. Wertheirn, P. H. Citrin and J. F. van der Veen, Phys. Rev. B, 1984, 30, 4344; 1988, 38, 7820. 10 T.-M. Duc, C. Guillot, Y. Lassailly, J. Lecante, Y. Jugnot and J. C. Vedrine, P h ~ x Rev. Lett., 1979, 43, 789. 11 K. G. Purcell, G. P. Derby and D. A. King, J. PhJs. Condensed Matter, 1989, I , 1373. 12 D. M. Riffe, G. K. Wertheirn and P. H. Citrin, Phys. Reu. Lett., 1990, 63, 1976. 13 M . W. Holrnes and D. A. King, Sue$ Sci., 1981, 110, 120. 14 €3. D. Barford and R. R. Rye, J. Chem. Phys., 1974, 60, 1046; E. W. Plummer and A. E. Bell, J. Vuc. 15 C. Kohrt and R. Gomer, SurJ Sci., 1971, 24, 77. 16 C. Leung, M. Vass and R. Gomer, Sur$ Sci., 1977, 67, 67. 17 R. Opila and R. Gomer, Surf Sci., 1983, 129, 563. 18 Ch. Steinbruchel and R. Gomer, Surj: Sci., 1977, 67, 21. 19 A. M. Bradshaw, D. Menzel and M. Steinkilberg, C'hem. Ph~t.5. Lett., 1974, 28, 516. 20 E. Umbach, J. C. Fuggle and D. Menzel, J. Electron Spectrosc. Relat. Phenorn., 1977, 10, 15. 21 E. Urnbach and D. Menzel, Surf Sci., 1983, 135, 565. 22 C . Somerton and D. A. King, Surf: Sci., 1979, 89, 391. 23 S. P. Singh-Boparai, M. Bowker and D. A. King, Surf: Sci., 1975, 53, 55. 24 J. Jupille, K. G . Purcell and D. A. King, Solid Stute Cornmun., 1986, 58, 529. Sci. Techno/., 1972, 9, 583.G. P. Derby and D. A. King 273 25 K. Duckers and H. P. Bonzel, Surf: Sci., 1989, 213, 25. 26 E. Wimmer, A. J. Freeman, M. Weinert, H. Krakauer, J. R. Hiskes and A. M. Karo, Phys. Rev. Lett., 27 E. Wimmer, A. J. Freeman, J. R. Hiskes and A. M. Karo, Phys. Rev. B, 1983, 28, 3074. 28 C . R. Brundle and J. Q. Broughton, in The Chemical Physics of Solid Surfaces and Heterogeneous Catalysis, ed. D. A. King and D. P. Woodruff (Elsevier, Amsterdam, 1990), Vol. 3A. 29 G. A. Benesh and R. Haydock, J. Phys. C, 1984, 17, L83. 30 J. Jupille, K. G. Purcell and D. A. King, Phys. Rev. B, 1989, 39, 6871. 1982,48, 1128. Paper 0/016526; Received 12th April, 1990

ISSN:0301-7249    

DOI:10.1039/DC9908900259

出版商:RSC    

年代:1990

数据来源: RSC

摘要:

Faraday Discuss. Chem. SOC., 1990, 89, 275-290 Characterization of Metal/ Organic Molecule and Metal/ Polymer Interfaces by NEXAFS Spectroscopy G . Tourillon, D. Guay and A. Fontaine LURE-Bat. 2090, 91405 Orsay, France R. Garrett and G. P. Williams NSLS, BNL, Upton, N Y , 11973, USA NEXAFS spectroscopy is a technique well suited t o probing metal/organic molecule o r metal/polymer interfaces. Two systems, related to the Pt/organ- ic conducting polymer and Ni/polyacry lonitrile interfaces, are discussed in detail. For poly-3-alkylthiophene (and selenophene) thin films electrochemically deposited onto Pt, the C K-edge characteristics reveal that ( i ) the polymeric chain unit is composed of the same architecture as the monomer and ( i i ) the doping proceeds via a narrowing of the band gap with the appearance of metallic-like behaviour.7 he same tendencies are observed by varying the structure of the monomer, the nature of the dopant and the film thickness, which evidences a similar conduction mechanism for these five-membered polyheterocycles. The NEXAFS spectra reveal that the polymeric chains are well ordered on the metallic surface: in its undoped state, the layers lie flat on Pt with strong interactions between the electron states of the metal and the antibonding T* band of the polymer; when doped, the NEXAFS characteristics exhibit changes as a function of thickness. The first layer lies flat on the Pt surface while the other ones are randomly oriented, due to the intercalation of anions during the oxidation process. Finally, the orienta- tion of the polymeric chain switches from a 'lying-down' to an 'on-edge' configuration when a long alkyl chain is added on the ring, this alkyl chain is oriented perpendicular to the Pt surface.Polyacrylonitrile (PAN) thin films electrochemically deposited on Ni have been studied by NEXAFS as a function of the film thickness and annealing treatment. For 20 A thick films, thz polymer chains are oriented perpendicular to the surface with the CN groups parallel to it. Below a few A, no polymerization occurs, but molecules are adsorbed perpendicular to the surface. Annealing at 300 "C results in the loss of the majority of the N content of the film in contrast to the admitted mechanism for bulk PAN. Introduction The interaction of unsaturated molecules and/or polymers with a metallic surface is of great interest both from a theoretical point of view'-3 (charge-transfer mechanisms between the substrate and the compound, modification of the molecular orbital distribu- tion) and from a practical point of view"-' (catalysis, surface chemistry).A considerable amount of work has thus been devoted to the adsorption of benzene,' pyridines or ethylene'' on various substrates. Very little work has been done on the adsorption of five-membered heterocycles."'-'' However, these molecules are of great interest because upon electrochemical oxidation they produce a new class of materials: the organic conducting polymers.' ' Among this class of compounds, polythiophene, polyseleno- phene and derivatives appear the most promising because of their high stability against oxygen and moisture.'" However, controversies still exist with regard to the interpretation 275276 Interface Characterization by NEXAFS Spectroscopy of the large change in conductivity during the doping. Three models have been pro- posed: "-" (i) creation of a degenerate semiconductor, (ii) formation of a closed-gap metallic state, or (iii) appearance of localized spinless bipolaron levels in the gap.Regarding the nature of the metal/polymer interface, apart from preliminary ellipsometry work done on Pt-poly-3-methylthiophene (which concludes with the deposition of oligomers on the metal surface") no precise information was available. Bulk polyacryl- onitrile (PAN) has been the object of numerous studies in recent year^'^-^' because its intrinsic conductivity can be increased with annealing treatment up to 1 R-' cm-'.When mixed with transition-metal salts, PAN appears to become an excellent candidate for catalysis.'2 Moreover, this polymer can be electrochemically deposited as thin films on various metallic surfaces, which is of technological importance in tribology or micro- electronics. Questions regarding (i) the nature of the interaction and/or the chemical bond between these thin films and the electrode and consequently (ii) the organization of the first layers on the surface are of great interest since all the applications in the field of catalysis, adherence or lubrication depend on the charge-transfer mechanism which occurs at this interface. In view of getting a better knowledge of the conduction mechanism and of the orientation of the thin films, a systematic NEXAFS study has been performed on these two interfaces.Near-edge X-ray absorption fine structure (NEXAFS) and extended X-ray absorption fine structure (EXAFS) spectroscopies are well adapted to probe these parameters, since they allow the determination of the unoccupied density of states, the bond lengths and the local environment of a particular atom embedded in a matrix.'> X-Ray Absorption Spectroscopy (NEXAFS, EXAFS): Principles This spectroscopy is based on the photoelectric effect, where the energy of the impinging X-ray photon is used to promote a core electron of an atom to the continuum. Within the dipole approximation, the Fermi golden rule yields the probability of the absorption processes: Since the X-ray photon energy is large, the only initial states li) to consider are the core levels, typically 1s or 2p,,,, 2p, for K or L edges.Thus, among the three contributions in the matrix element (ilE- .if), the non-trivial part to determine in the ab initio computation is the final state If). Two different ways have been developed to calculate this ~arameter:?~ ( i ) the band structure approach, which requires the triple periodicity; (ii) the scattering formalism, where If) is a solution of the excited potential, including the single back-scattering from neighbouring atoms, the spatial variation of the scattered wavelets and the multiple- scattering processes between the excited atom and its neighbours. It has been clearly demonstrated that the X-ray absorption spectrum can be split in two main regions (XANES and EXAFS) above the edge according to the kinetic energy of the photoelectron.'' EXAFS EXAFS is a final-state interference effect involving scattering of the outgoing photoelec- tron from the neighbouring atom.'" For a monoatomic gas such as Kr" with no neighbouring atoms, the ejected photoelectron will travel as a spherical wave with a wavelength, A = 27r/ k whereG.Tourillon et al. 277 Here E is the incident photon energy and E,, is the threshold energy of that particular absorption edge. The p versus E curve follows the usual smooth A' decay. In the presence of neighbouring atoms,?' this outgoing photoelectron can be back-scattered, thereby producing an incoming wave, which can interfere either constructively or destructively (depending on the nature of the neighbouring atoms) with the outgoing wave, resulting in the oscillatory behaviour of the absorption coefficient. This simple picture of EXAFS has been formulated into the generally short-range single-electron single-scattering The modulation of the absorption rate in EXAFS, normalized to the background absorption, po is given by X ( E ) = [tL(E)-puo(E)I/puo(E).( 2 ) In order to relate x ( E ) to structural parameters, it is necessary to convert the energy E into the photoelectron wavevector k. This transformation gives rise to x ( k ) in k space where where N, is the number of neighbours at the distance R; from the central atom, a, is the Debye-Waller factor to account for thermal vibration and static disorder, and +,( k ) is the total phase shift experienced by the photoelectrm.The term exp [ - 2 R , / A ( k ) ] is due to inelastic loss in the scattering process, A being the electron mean free path. Near-edge Structure (XANES-NEXAFS) For molecules in the gas phase or adsorbed on a surface, it has been found that the XANES spectra are dominated by two kinds of transitions.'0.31 (i) In molecules with unoccupied T* orbitals, the first resonance which occurs at an excitation energy less than the 1s ionization potential ( Ei) corresponds to a transition of a 1 s electron into an empty or partially filled antibonding T* orbital. This resonance is lowered below the Ei by the Coulomb interaction with the core hole (bound-state transition).In a multiple scattering picture, the T* resonance may be envisioned as arising from an excitation into a final state which is trapped on the molecule by the valence electron potential. ( i i ) The second molecular resonance which is observed at higher energy than the n* resonance is usually referred to as a a-shape resonance. This resonance is arising from a scattering process where the photoelectron wave is resonantly scattered back and forth along the internuclear axis between the absorbing atom and its neighbours. This picture suggests that the shape resonance position should be sensitive to the intramolecular distance. For instance, a continuous shift in the energy position of the a-shape resonance has been observed with aliphatic molecules containing single, double and triple C-C bonds.Moreover, Sette et a!." have established that a linear relationship exists between the position, 6, of the cr resonance relative to the K shell ionization threshold (of B, C, N, 0 or F) and the internuclear distance, r, between the pair of atoms giving rise to the resonance. This empirical relationship can be used to determine i n t e r a t o m i c d i stances i n unknown structures . Taking advantage of the polarization of the synchrotron light, the orientation of an adsorbed molecule relative to the surface can be assessed because the photoabsorption process is governed by the dipole selection rules. It has been demonstrated'() that for a final state of pure a or 7~ symmetry, the absorption coefficient p is given by278 Interface Characterization by NEXAFS Spectroscopy where S is the angle between the polarisation vector, E, and the interatomic axis, M, of the molecule.The orientation and the chemical stability uersus the temperature of several compounds [ aliphatic3' or heterocyclic adsorbed on different monocrystalline surfaces (Pt, Ag, C U ) ~ ~ * ' ~ ] have been fully determined. Experimental Films of poly-3-alkylthiophene (or selenophene) were electrochemically deposited onto a flat Pt electrode immersed in an electrolytic medium composed of CH3CN-(5 x lo-' mol dm-3) LiC104-(5 x lo-' mol dmP3) m ~ n o m e r . ' ~ . ~ ' The electrode was first polar- ized at ca. +1.35 V us. SCE (saturated calomel electrode) to obtain doped films. Inter- mediately doped and fully undoped films were obtained by gradually reversing the potential from +1.35 V to -0.2 V us.SCE. Different thicknesses between one monolayer and 500A were deposited ,by varying the deposition time, according to the linear thickness relationship of 10 H mC-' cm-2. Film thicknesses could also be gauged from Pt 4f photoemission data as a secondary check. Films of PAN were electrochemically deposited on Ni electrodes in a CH3CN-(5 x lop2 mol dmP3) N( Et),ClO,-( 5 x lo-' mol dmp3)-acrylonitrile electrolyte. A two-step polarization (-0.8 to -2.8 to -0.8 V us. Ag-Agf) was applied to the working electrode and the film thickness was monitored by varying the electrolysis time between 10 and 500 ms. The NEXAFS experiments were carried out at the National Synchrotron Light Source at Brookhaven National Laboratory on the U14 beam line using a plane-grating mono- chromator. C K-edge spectra were recorded in both partial (CMA) and total (channel- tron) yield modes.The EXAFS characteristics were performed at LURE by using the synchrotron radiation from the ACO storage ring. The soft X-rays were monochromat- ized by two Ge( 1 1 1 ) single crystals. The spectra were obtained by collecting the total electron yield up to 300 eV above the S K-edge. Results and Discussion PtiOrganic Conducting Polymer Interface Electronic Structure Thiophene or selenophene is a planar molecule with C,, symmetry and a high degree of aromaticity. Theoretical calculations (INDO, CNDO/S3) performed by DukelS and Bredas et a/.'' show that the highest occupied molecular orbital, 1a2n, and the lowest unoccupied molecular orbital, 3b, T*, have pure carbon n character. Upon polymeriz- ation, the la,n and 3bzn* orbitals give n bonding and T* antibonding bands, the first one being located 1 eVt below the Fermi level, Er.This n band has never been observed experimentally on yolyacetylene and on polypyrrole, owing to structural defects. "-" In addition to the T* band, unoccupied levels of u symmetry from both the C-S (or C-Se) and C-C bonds are expected in the NEXAFS spectra. Our UPS data obtained from a thin undoped poly-3-methylthiophene (PMeT) clearly revealed the existence of the T bonding band which means that a long-range order exists along the polymeric Our measurement locates the upper n band edge at 1.05 eV below Ef-.Upon doping, this T bonding band extends further towards lower binding energy up to 0.25 eV below Er [fig. l ( a ) ] . Moreover, the C 1s and S 2p core levels become broad and asymmetric on the high binding-energy side [fig. l ( b ) ] . These modifications are completely reversible during a doping-undoping cycle and indepen- dent of the anion used during the oxidation, of the film thickness and of the structure 1 eV= 1 . 6 0 2 ~ l O - ' " J .G. Tourillon et al. 279 A 168 166 164 162 290 286 282 binding energy/eV B t 1 Fermi Level binding energy/eV Fig. 1. (A) C 1s (right) and S 2p (left) core-level XPS spectra of C104- doped (bottom) and undoped (top) poly-3-methylthiophene. (B) He I ( b ) and He I1 ( a ) UPS spectra of a 10008, thick poly-3-methylthiophene film electrochemically deposited on Pt.of the monomer. These results evidence that the origin of the conduction mechanism is similar for these poly-five-membered heterocycles. NEXAFS Characteristics. The C K-edge NEXAFS spectrum of a 1000 8, thick undoped poly-3-methylthiophene film deposited on Pt4"74' is very similar to that of thiophene in the solid phase33 which confirms the same structural architecture for the polymeric chains (fig. 2). Five transitions are observed at 286, 287.7,289,294.7 and 304eV. The first one corresponds to the 1 s - r " bound-state transition (table 1). The second transition at 287.7 eV arises from a (T* shape resonance due to the C-S bond. We assign the third feature as a T resonance arising from the transition of the C 1s core electron to the second antibonding n* band of the polymer.The other one at 294.7 eV is attributed to a a*(C-C) shape resonance. Finally, the last peak at 304eV could correspond to a a*(C=C) shape resonance or to a double excitation of T symmetry since no transition is expected at this energy in the multiple scattering X, c a l c ~ l a t i o n s . ~ ~ When the polymer280 Interface Characterization by NEXA FS Spectroscopy energy/eV Fig. 2. C K-edge NEXAFS spectra of 1000 A thick poly (3-methylthiophene) film electrochemically deposited on Pt. ( a ) Undoped form and ( b ) fully doped state. Table 1. Assignment of features in C K-edge NEXAFS spectra of thin poly-3- methylthiophene films on Pt undoped doped transition assignment 1’ ca. 284-284.5 new empty states 1 286 286 3b, T* 2 287.7 287.7 a * ( C - S ) 3 289 289 2a2,* 4 294.7 294.7 a*(C-C) 5 304 304 a*(C=C) or double excitation 284.6 284.3 c I S ionization energy X P S ~ is doped with C104-, no energy variations of the a ( C - S ) and o(C-C) shape resonances are observed [fig.2(6)], suggesting that no modifications in the bond lengths take place during the doping process according to the shape resonance-bond length relati~nship.~’ However, the T* antibonding band is greatly affected as evidenced by the continuous decrease in intensity with the doping level (fig. 3). Additionally, a transition appears at low energy whose intensity increases with the doping process. These empty states are still observed by changing the structure of the monomer (pyrrole, selenophene) and/or the nature of the dopant (BF,-, S03CF3--, PF6-).EXAFS Data. The Fourier transform of the EXAFS above the S K-edge for 3-methylthiophene is composed of two peaks at 1.6 and 2.5 A (uncorrected from the phase shift) corresponding to the S-C, and S-C, bonds, re~pectively~~ (fig. 4). When the polymer is synthesized, the same peaks are detected in both the undoped and doped states. The intensity of the second peak decreases by a factor of 2 with the appearance of a new peak at 2.63 A. The origin of these modifications is due to the a-a coupling of the monomeric units during the polymerization. To obtain quantitative estimates of the bond lengths and the coordination numbers around the S atom, we have fitted theG. Tourillon et al. 28 1 16 12 h c. v) .- c -2 - 8 x c c c..- v1 c. .- 4 0 I 1' I 283 288 energy/eV Fig. 3. Evolutions of the C K-edge NEXAFS spectra with the doping level. ( a ) Undoped, ( b ) 3% ClO, doped, ( c ) 15% doped and ( d ) 30% fully doped. C f I 5.500 m E: 'D c) .- - a 5 0. ooc " v - - 0.0 1u.u distance/8( methylthiophene) film deposited on R: ( a ) undoped, ( b ) 15% ClO, doped and ( c ) fully doped. Insert: the formula for poly(3-methylthiophene) indicates the a, a' and p positions for C.282 Interface Characterization by NEXAFS Spectroscopy Table 2. EXAFS parameters" Deb ye- Waller, distance/A coord. no. a/ energy/ eV 3-methylthiophene s-c, undoped PMeT S-C, 15% doped PMeT S-C,, 30% doped PMet s-c, 3-methylthiophene s-c, undoped PMet s-c, s-c,. doped PMeT s-c, s-c,. 1.72 1.74 1.738 1.74 2.57 2.60 2.67 2.60 2.67 2 1.95 0 0.2 2.03 0.001 0.25 1.97 0.003 0.76 2 0 0.05 2.01 0.0 1 0.56 1.97 1.99 0.02 0.67 2.02 (' Bond lengths, R, coordination number, N , Debye- Waller factor, a, and energy variation, E, obtained by fitting the different shells in the Fourier-filtered EXAFS spectra of undoped and doped PMeT.inverse-filtered Fourier transforms of the different peaks. These parameterso are given in table 2 and show that the S-C, bond length increases from 1.72 to 1.74 A from the monomer to the undoped polymer. No variation is observed when the thin film is progressively doped, in agreement with !he NEXAFS data. The Fourier transforms also exhibit a broad peak in the range 3-4 A, associated with the interaction of the dopant with the S atoms. The same evolutions have recently been observed on polyselenophene during an in situ doping-undoping electrochemical cycle.45 Several models have been suggested to explain the large increase of conductivity during the electrochemical oxidation of these polymers.Formation of a degenerate semiconductor associated with the extraction of electrons from the vbonding band and a shift of the Fermi level below the valence band.I7 With this scheme, the C K-edge NEXAFS spectra would exhibit a peak followed by a sharp threshold 2.2 eV higher in energy (band-gap value of polythiophene) due to transitions into the T* conduction band. Moreover, the intensity of this last transition should be constant during the doping process since the 7-P antibonding density of states is not affected.Evidently the NEXAFS characteristics do not agree with this model. Formation of local distorsions in the polymeric chain around localized charges associated with the evolution of the aromatic structure to the quinoid one. This leads to the appearance of two localized empty levels in the band gap: bipolaron Structural modifications are thus expected, especially in the bond lengths for the polymeric backbone; so energy variations of peaks 2 and 4 should be observed, which is not the case experimentally. The NEXAFS spectra would also show two new peaks at low energy, corresponding to the C 1s transitions toward the bipolaron bands. The evolution of the C K-edge NEXAFS characteristics combined with the Se K-edge dispersive EXAFS results obtained during an in situ doping-undoping electrochemical cycle which again do not reveal any modification in the C-Se bond length47 are not in agreement with the bipolaron predictions.Moreover, it is not yet clear how the bipolaron model can explain the r-bonding band extension towards the Fermi level, observed when the polymers are doped. Finally, if the conducting state corresponds to metallic-like behaviour, a simple step-like absorption edge 1.2 eV (midgap value) below the n* band edge would be ~ b s e r v e d . ' ~ In the fully doped state, the C l s + T* resonance intensity observed in the undoped form has almost vanished with the absorption threshold shifted to lower energy, the conduction band density of states being clearly altered in the heavily doped poly- selenophene. Thus, the NEXAFS characteristics suggest a closed-gap metallic stateG.Tourillon et al. 1 2 3 L 5 6 7 283 3 Table 3. Assignment of the peaks in the NEXAFS spectra of the poly-3- alkylselenophene series at the C K edge shown in fig. 5 peak energy/eV assignment ca. 284 285.3 287.1 288.4 292.8 293.8 295 302.0 new empty states (doped polymer) 3b, T* (polymer) a*(C-Se)( polymer) a*(C-H)(alkyl chain) and/or 2a2rr* (polymer) u*( C -C)( alkyl chain) a*( C-C)( polymer) a*( C-C)( polymer) cr*(C-C)'(polymer, alkyl chain) where the oxidation proceeds through an extraction of electrons from the n bonding band and a narrowing of the T-T* gap. Ordering Eflect The data for 20 A thick undoped polyselenophene (fig. 5A, table 3) reveal that peak 1 is more intense at grazing incidence than at normal in~idence.~' Assuming rotational symmetry, the following expression for radiation with a degree of linear polarization, P, holds:' ( 5 ) where 8 is the polar angle and a the tilt angle between the surface and the polymer I,* = [P(sin' a sin' e + 2 cos' a cos' ~ ) + ( l - P ) sin' a ]284 Interface Characterization by NEXAFS Spectroscopy R z H, -CH3,-CH;CH3 I o r Fig.6. (A) Plan view of Pt surface with polyselenophene ( o r polythiophene) in lying-down position. This orientation was found for the polymers with small alkyl chains. ( B ) Elevation of Pt surface showing orientations of the polyselenophene (or thiophene) with long alkyl chains attached. plane. Least-squares fits according to eqn ( 1 ) with P = 0.85 yield a tilt angle of a = ( 0 * 5 ) O .The measured tilt angle is in good agreement with that found recently for polythiophene electrochemically deposited on Pt.44 This shows that the polymer lies flat on the surface. Reverse behaviour is expected and confirmed experimentally for the a bands (peaks 2 and 5 ) which come from the ring plane. For poly-3-methyl-and ethyl-selenophene, the same polarization dependence is observed which means that the polymeric chains also lie flat on the metallic surface. A change in orientation of the polymeric backbone with increasing alkyl length is observed for undoped poly-3-hexyl-, octyl- and decyl- selenophene (fig. 5B). A complete opposite polarization dependence to that obtained earlier is revealed. The polymeric T* (peak 1 ) and alkyl (peak 3) a*(C-H) transitions have maximum intensities at normal incidence while the a*( C-C) shape resonance intensities are stronger at grazing incidence.The polarization dependence least-squares fits yield an average tilt angle of a = (85 f 5)" for the n* transition. That means that the polymer and the alkyl chains are oriented normal to the Pt surface. These polyalkyl- selenophenes composed of a long hydrocarbon chain terminated at one end by the selenophene ring can be viewed as Langmuir-Blodgett films which form ordered layers on surface^.^' During the electrochemical synthesis, this alkyl chain will force the polymeric backbones to take an 'on-edge' configuration in which they become oriented perpendicular to the surface (fig. 6). The orientation of the polymer being imposed byG.Tourillon et al. 285 , 20 15 h c1 .- v) c 3 4 .2 10 - c 0-l c c1 .- E c cc a 9 I1234 6 2, I I I ) 275 285 295 305 31 5 energy/eV Fig. 7. C K-edge NEXAFS spectra of thin PAN ofilms electrochemically deposited on Ni ( a ) 140 A thick film; ( b ) 500 A; ( c ) and ( d ) the 500 A thick film after annealing treatment to 300 and 500 "C, respectively. the aliphatic chain when several layers are deposited on Pt, a head-to-tail configuration should occur as observed for Langmuir-Blodgett films. Similarly ordering effects have been obtained for poly-3-alkylthiophenes." In the case of poly-3-butylselenophene, almost no polarization dependence is detec- ted. This should be due to the polymeric chain tilting away from the surface. The alkyl chain is not long enough to impose a special configuration for the polymer but its steric effect prevents the polymeric chain lying flat on the Pt.For thicker films, the polarization dependence becomes weaker, vanishing in films a few hundred A thick, where the polymeric chains become randomly oriented. Upon doping, a thickness-dependent orientational transition is observed in the C K-edge NEXAFS characteristics of thin polyselenophene, poly-3-methyl- and ethyl- selenophene films between the first layer and subsequent overlayers: the first monolayer is still flat on the surface when the other ones become randomly oriented, due to the intercalation of anions inside the polymeric chains during the oxidation process. Finally, doped poly-3-hexyl-, octyl- and decyl-selenophene films exhibit a polarization depen- dence but become tilted from the Pt surface with a tilt angle of a = 70* 5" (averaged over multiple determination of angle). Ni/Polycrylonitrile Interface Characteristic C and N K-edge NEXAFS spectraS0 of a 140 A thick PAN film are shown in fig.7 ( a ) and 8 ( a ) . These NEXAFS spectra exhibit two main features: sharp T*286 Interface Characterization by NEXAFS Spectroscopy f IP energy/eV Fig. 8. As fig. 7 but for the N K edge. Table 4. Resonance positions and their assignments for the 20, 140 and 500 A thick PAN films shown in fig. 7-9 ~~ energy/eV peak Cedge N edge proposed assignment ca. 285 285.8 287.2 288.8 293.9 302.1 307.1 286 287.1 395.6 397.4 398.7 402.5 ca. 406 408 411.1 400.1 r * a ' ( C E N ) r*(C=C) and/or r*(C=N) r* a " ( C z N ) u*(C-H) alkyl chain u*(C-C) alkyl chain u*(C=C) and/or u* (C=N) a * ( C r N ) E,(C-C) from XPS E , ( N r C ) (T*(C-N) Ei(C=N) resonances at the threshold and broad a*-shape resonances at higher energies (table 4).The appearance of two T* components is a consequence of lifting the orbital degeneracy of the two CN T* levels.5' This film is sufficiently thick that the layer adjacent to the Ni substrate contributes indetectibly to the NEXAFS spectra. We therefore attribute the splitting of the T* levels to the intramolecular C N interactions, in agreement with HFSCF calculations performed for different oligomers." From the orientations effects for a 20 8, thick film (see below), peak 1 is due to the a' component while peak 3 corresponds to the a'' component.G. Tourillon et al.287 k., energy/ eV Fig. 9; Polarization dependence of the ( a ) N and ( b ) C K-edge NEXAFS spectra obtained from a 20 A thick PAN film at normal and grazing incidences; insert: the orientation of the polymer chains on the metallic surface. We attribute peaks 4 , 6 and 9 at the C edge to the shape resonances due to the C-H, C-C and C-N bonds, respectively. From the polarization dependence, we conclude that the 411 eV feature at the N K-edge corresponds to the a*(CN) shape resonance in agreement with the results of Kordesch et d5' The NEXAFS spectra obtained from a 500 A thick film [fig. 7( b ) and 8( b ) ] are very similar to those for the 140 8, film, except for a shoulder at 285.8 eV (C K edge) and 397.6 eV (N K edge). The features are attributed to chemical defects inside the chains (C=C and C=N bonds), as confirmed by infrared rneasurement~.~~ After annealing in vacuum to 300 and 500 "C for 2 h, noticeable changes occur in the NEXAFS spectra.Transitions 3 and 9 at the C edge [fig. 7(c), ( d ) ] and 1, 3, 5 and 9 at the N edge [fig. 8(c), ( d ) ] , associated with C N bonds, are reduced in samples annealed to 300°C and have almost vanished after annealing at 500°C. At the same time, peak 2 at the C edge associated with C=C and C=N r* resonances is greatly enhanced and there is a new feature at 302.4 eV due to C=N and C=C shape resonances. The N K-edge NEXAFS spectra are now dominated by a very broad (T' resonance, corresponding to the synthesis of C N single and double bonds. These results reveal that: ( i ) the majority of the N component of the polymer is removed in samples annealed to 300°C and ( i i ) after annealing to 500"C, the film is composed essentially of C=C and C-H bonds.These results differ significantly from288 Interface Characterization by NEXAFS Spectroscopy * ;p* .- . o 3 390 400 410 v I I hv 1 I 0 I * 275 2 05 295 305 31 5 energy/ eV Fig. 10. Polarization dependence of the ( a ) N and ( b ) C K-edge NEXAFS spectra from a few A thick layer at normal and grazing incidences; insert: the proposed arrangement of the layer adsorbed on Ni. the general mechanism proposed for bulk PAN where a cyclization process of the CN groups is proposed to occur around 300 "C, followed by the formation of a 'graphite-like' structure with a release of N atoms at 800°C." The C and N K-edge NEXAFS spectra of a 20 8, thick PAN film (fig.9) are dominated by the same transitions as those in fig. 7 and 8 but they exhibit a pronounced angular variation. The first T* and the a*(C-C) resonances are enhanced at grazing incidence and almost vanish at normal incidence. The opposite polarization dependence is observed for the second v*, the a*(C-H) and the u*(CN) resonances. Assuming a strong directional character for the CN resonance and using the previous relationship(sj, least-squares fits yield a tilt angle of 85O, which indicates the CN groups are parallel to the substrate. A similar analysis of the C-H and C-C u* components reveals that the polymer backbone is oriented norma! to the surface. NEXAFS spectra from a film a few A thick exhibit great changes (fig.10, table 5 ) as evidenced by ( i ) no splitting of the T* resonance, ( i i ) the absence of the CN resonance, replaced by a a*(C=N) transition and (iii) a strong polarization dependence of the n* feature which peaks at normal incidence. This indicates that in these very thin films, a strong interaction between the Ni surface and the nitrogen atoms occurs. We conclude from these results that for very short electrolysis times no polymerization occurs, but that a layer with molecular structure is adsorbed on Ni. If the electrolysisG. Tourillon et al. Table 5. Resonance positions and assignments for a few 8, thick PAN film from fig. 10 289 energy/eV peak C edge N edge proposed assignment ~ ~~~~ 1' 285.8 398.3 r*(C=N) 2' 288.8 a*(C-H) alkyl chain 3' 294.2 a*(C-C) alkyl chain 4' 301.8 406.9 a*(C=N) is persued, polymerization begins and the backbone orients normal to the surface with the CN groups perpendicular to it.For the organic conducting polymer films electro- chemically deposited on Pt, several main effects have been observed. First, a long-range order exists along the polymer, associated with the formation of the n-bonding and n*-antibonding bands. Secondly, the doping to the conducting state proceeds via a narrowing of the band-gap causing the appearance of metallic like behaviour. Thirdly, the polymeric chains are well ordered on the metallic surface but the orientation depends on the thickness, the structure of the monomer and the oxidation state. Finally, these NEXAFS results show the potential of the method for investigating electronic charac- teristics of thin organic films and metal/polymer interfaces.References 1 3 4 5 6 7 8 9 10 1 1 12 13 14 15 16 17 18 19 20 21 7 - 77 -- 23 24 25 26 27 28 J. L. Dehmer and D. Dill, Phys. Ret.. Lert., 1975, 35, 213. N. Padial, G. Csanak, B. V. McKoy a n d P. W. Langhoff, J. Chem. Phys., 1978, 69, 2992. J. Stohr a n d R. Jaeger, Phys. Reu. B, 1982, 26, 4111. P. C. Mitchell, in Catalysis, ed. C. Kemball (Chem. SOC., London, 1977), vol. 1, p. 233. N. R. Avery, Surj Sci., 1984, 146, 363. A. J . Gellman, M. H. Farias, M. Salmeron a n d G. A. Somorjai, Surj: Sci., 1984, 136, 217. J. C. Bertolini, G. Dalmai-lmelik a n d J. Rousseau, Surf: Sci.. 1977, 67, 478. A. L. Johnson, E. L. Muetterties, J.Stohr and F. Sette, J. Phys. Chem., 1985, 89, 407. R. J. Koestner, J . Stohr, J . L. Gland and J. A. Horsley, Chem. Phys. Lett., 1984, 105, 332. M. Salmeron, G. A. Somorjai, A. Wold, R. Chianelli a n d K. S. Liang, Chem. fh?x Lert., 1982, 90, 105. T. M. Thomas, F. A. Grimm, T. A. Carlson and P. A. Agron, J. Electron Spectrusc. Relat. Phenom., 1982, 25, 159. B. A. Sexton, Surf: Sci., 1985, 163, 99. Organic Conducting Polymers, ed. T. Skotheim (Marcel Dekker, New York, 1986), vol. 1 and 2. G . Tourillon and F. Garnier, J. Electroanal. Chem., 1982, 135, 173. C . B. Duke, Int. J. Quantum Chem. Svmp., 1979, 13, 267. J. L. Bredas, B. Themans, J. G. Fripiat, J. M. Andre a n d R. R. Chance, Ph?,.s. Rev. B, 1984, 29, 6761. J. J. Ritsko, f h y x Reu. Lett., 1981, 46, 849.P. Lang, F. Chao, M. Costa and F. Garnier, J. Electroanal. Chem., in press. H. Teoh, P. D. Metz a n d W. G. Wilheim, Mol. Crjat. Li9. Cryst., 1982, 83, 297. C. R. Wu, W. R. Salaneck, J. J. Ritsko and J. L. Bredas, Synth. Met., 1986, 16, 147. J. J. Ritsko, G. Crecelius a n d J. Fink, f h j * . v . Rev. B, 1983, 27, 2612. E. Yeager, Elecrrocata1~~si.s .fLr 0.yrgen Electrodes (Technology Base Research Project, University of California, Berkeley, 1988). EXAFS and Near Edge Structure, ed. P. Lagarde, D. Raoux and J. Petiau (Les Editions d e Physique, Les Ulis-France, 1986), T47, vol. 1 and 2. J . E. Muller and W. L. Schaich, Ph!-s. Rec. B, 1983, 37, 6489. M. Benfatto, C . R. Natoli, A. Bianconi, J. Garcia, A. Marcelli, M. Fanfoni and I . Davoli, Phys. Reo.B, in press. B. K. Teo, in EXAFS spectroscopy, ed. B. K. Teo and D. C . Joy (Plenum Press, New York, 1983) p. 13. B. M. Kincaid and P. Einsenberger, Phj9.s. Rev. Lett., 1975, 34, 1361. E. A. Stern, Phj-s. Rev. B, 1974, 10, 3027.290 Interface Characterization by NEXA FS Spectroscopy 29 ( a ) C. A. Ashley and S. Doniach, Phys. Rev. B, 1975, 11 1279; ( 6 ) P. A. Lee and G. Beni, Phys. Rev. B, 1977, 15, 2862; ( c ) P. A. Lee and J. B. Pendry, Phys. Rev. B, 1975, 11, 2795. 30 J. Stohr and R. Jaeger, Phys. Rev. B, 1982, 26, 41 11. 31 F. Sette, J. Stohr and A. P. Hitchcock, J. Chem. Phys., 1984, 81, 496. 32 J. Stohr, F. Sette and A. L. Johnson, Phys. Rev. Lett., 1984, 53, 1684. 33 J. Stohr, J. L. Gland, E. B. Kollin, R. J. Koestner, A. L. Johnson and E. L. Muetterties, Phys. Rev. Lett., 1984, 53, 2161. 34 G. Tourillon, S. Raeen, T. Skohteim, M. Sagurton, R. Garrett and G. P. Williams, Surj: Sci., 1987, 184, L345. 35 D. Arvanitis, K. Baberschke, L. Wenzel and U. Dobler, Phys. Rev. Lett., 1986, 57, 3175. 36 J. Haase, Appl. Phys. A, 1985, 38, 181. 37 G. Tourillon, C. Mahatsekake, C. Andrieu, G. P. Williams, R. Garrett and W. Braun, Surj: Sci., 1988, 38 Y. Jugnet, G. Tourillon and Tran Minh Duc, Phjis. Rev. Lett., 1986, 56, 1862. 39 G. Tourillon and Y. Jugnet, J. Chem. Phys., 1988, 89, 1905. 40 G. Tourillon, A. Fontaine, R. Garrett, M. Sagurton, P. Xu and G. P. Williams Phvs. Rev. B, 1987, 35, 41 G. Tourillon, A. Fontaine, Y. Jugnet, Tran Minh Duc, W. Braun, J. Feldhaus and E. Holub-Krappe, 42 A. P. Hitchcock, J. A. Horsley and J. Stohr, J. Chem. Phys., 1986, 85, 4836. 43 A. P. Hitchcock, S. Beaulieu, T. Steel, J. Stohr and F. Sette, J. Chem. Phys., 1984, 80, 3927. 44 G. Tourillon, A. M. Flank and P. Lagarde, J. Phys. Chem., 1988, 92, 4397. 45 G. Tourillon, E. Dartyge, A. Fontaine, A. Jucha and C. Andrieu, J. Phys. (Paris), 1986, 47, C8-551. 46 J. L. Bredas, R. R. Chance, R. Silbey, G. Nicolas and Ph. Durand, J. Chem. Phys., 1981, 75, 255. 47 G. Tourillon, E. Dartyge, D. Guay, C. Mahatsekake, C. Andrieu, S. Bernstorff and W. Braun, J. Phj9.r. 48 G. Tourillon, E. Dartyge, A. Fontaine, R. Garrett, M. Sagurton, P. X u and G. P. Williams, Europhj1.r. 49 D. A. Outka, J. Stohr, J. P. Rabe, J. D. Swalen and H . H. Rotermund, Phys. Rev. Letf., 1987, 59, 1321. 50 G. Tourillon, R. Garrett, N. Lazarz, M. Raynaud, C . Reynaud, G. Lecayor and P. Viel, J. Electrochem. Soc., in press. 51 M. E. Kordesch, T. Lindner, J. Somers, W. Stenzel, H. Conrad, A. M. Bradshaw and G. P. Williams, Spectrochim. Acta, Part A, 1987, 43, 1561. 52 M. Raynaud, Thkse, Universite Paris XI-Orsay, 1989. 53 S. Leroy, C. Boiziau, J. Perreau, C . Reynaud, G. Zalczer, G. Lecayon and C. Le Gressus, J. Mol. Struct., 201, 171. 9863. Phys. Rev. B, 1987, 36, 3483. Chem., submitted. Lett., 1987, 4, 1391. 1985, 125, 269. Paper 9/05384K; Received 19th December, 1989

ISSN:0301-7249    

DOI:10.1039/DC9908900275

出版商:RSC    

年代:1990

数据来源: RSC

摘要:

Faraday Discuss. Chem. SOC., 1990, 89, 291-300 Structure of the Surface Methoxy Species on Cu(ll1) Detlef E. Ricken, Joseph Somers*, Andrew W. Robinson and Alexander M. Bradshaw Fritz-Haber- Institut der Max- Planck-Gesellschaft, Faradayweg 4-6, D- 1000 Berlin 33, Federal Republic of Germany Angle-resolved photoemission spectroscopy has been used to identify four adsorbate-induced features from the surface methoxy species on Cu{ 11 l}. I n agreement with IR data, the application of dipole selection rules suggests that the 0-C axis is oriented perpendicular to the surface. The ordering of the 5a, and l e emission bands is reversed compared with the methoxy species adsorbed on the Cu{ 100) and Cu{ 110) surfaces. The interpretation of the X-ray absorption data is more difficult, owing to the occurrence of overlap- ping resonances.A clear polarisation dependence is observed at the oxygen edge, but indicates that the 0-C axis could be tilted by up to 30" from the surface normal. This apparent conflict with the photoemission data is attributed to the somewhat arbitrary nature of the standard curve-fitting procedures. Despite considerable progress in recent years in the identification of molecular fragments adsorbed on metal surfaces [ e.g. ref. (1)-(3)], precise information on their structure and bonding properties has generally been lacking. One reason for the very few structural studies is that such species seldom, if ever, form ordered overlayers and thus cannot be studied by conventional diffraction techniques such as LEED. A deeper understanding of the electronic structure, on the other hand, has been hindered not only by the lack of quantum chemical 'cluster' calculations for anything but the simplest adsorbates but also by often incomplete photoemission data.The surface methoxy species proves to be an exception probably because of its putative role as an intermediate in the synthesis of methanol over Cu/ZnO catalysts,"5 which is a reaction of considerable industrial i m po rt a n ce . S p e c t r o s co pi c tech n i q u e sh- () re q u i r i ng s y n c h ro t ro n rad i at i on have a 1 re ad y played an important role in the characterisation of the methoxy species and other adsorbed molecules and molecular fragments. On all three basal planes of copper, the methoxy fragment is readily formed by the deprotonation of methanol following pre-adsorption of oxygen."-'3 On Cu{ 1 lo} near- edge X-ray absorption spectroscopy (NEXAFS) indicated that the 0-C bond is tilted in an azimuthally random manner at an angle of 30" away from the surface On the other hand, an angle-scanned X-ray photoelectron diffraction study favoured different tilt angles in each principal azimuth." An early X-ray absorption investigation of the methoxy species adsorbed on Cu(100) also concluded that the molecule was inclined by 30" to the surface normal." A later measurement combined with an energy- scanned photoelectron diffraction investigation of this system concluded, however, that the methoxy species is in fact oriented with the 0-C axis perpendicular to the surface" and that a low-symmetry 'off-bridge' adsorption site is occupied.The adsorption of the methoxy species on Cu(ll1) has recently been investigated by infrared reflection-absorption spectroscopy ( I RAS)." Application of the surface selection rule and comparison of spectra from adsorbed CH30 and CD,O with those from a model compound containing the methoxy ligand, indicated that the adspecies is also perpendicular to the surface. In the present paper, we describe a series of angle-resolved photoemission and X-ray absorption investigations on this adsorption system undertaken with the aim of confirming this bonding geometry. 29 1292 Surface Methoxy Species on Cu(l11) In both types of spectroscopy (photoabsorption and photoemission) dipole selection rules can be applied.In the photoabsorption experiment, an adsorbate core electron is excited into the unoccupied molecular orbitals either below or above the photoionisation threshold. The resulting ‘near-edge’ spectrum may also contain other features due to Rydberg transitions (if they are not quenched) or to scattering from substrate atoms. The latter phenomenon gives rise to the well known EXAFS structure at higher energies above the threshold. Since chemisorption normally results in a single, fixed orientation for a molecule or molecular fragment, the dipole transitions from the core level into the unoccupied levels are polarised and can thus be used for orientation determination.6”0 The yield of Auger electrons or a high-energy fraction of the secondary electrons (partial electron yield) are normally used as measures of the absorption coefficient (partial photoionisation cross-section). In the photoemission experiment, or photoelectron spectroscopy, the primary excitation is a dipole transition from an occupied molecular orbital into a level in the continuum.In order to apply selection rules the symmetry of the final-state continuum wavefunction an be selected specifically in an angle-resolved experiment. (Thus, the angle-resolved experiment measures a differential partial photo- ionisation cross-section for the particular valence level.) When the molecular orientation is known, the symmetry of the initial state can be Alternatively, if the symmetry of the initial state is already known, i.e. the bands in the spectrum are already assigned, the selection rules can be used to identify the point group symmetry, and thus perhaps the orientation of the Experimental The surface methoxy species was prepared by first exposing the Cu(1ll) surface to 2 x mbar s of methanol at 200 K.The crystal was then briefly annealed to 300 K in order to desorb water formed by the deprotonation reaction. Following this procedure, no features attributable to chemi- sorbed oxygen were found in the photoemission spectra. The sample temperature was reduced to 100 K during the measurements but was periodically reheated to 300 K to ensure that no methanol was readsorbed. All experiments were performed in a VG ADES 400 electron spectrometer, equipped with facilities for low-energy electron diffraction (LEED), and Auger electron spectros- copy (AES).Synchrotron radiation was provided by the electron storage ring BESSY in Berlin. Photoemission measurements were performed using the 1 m Seya-Namioka monochromator with a 1200 lines per mm grating.” The overall resolution was ca. 200 meV.‘t‘ On this particular beamline the excellent focussing properties allow the light to be threaded through a small hole in the analyser, making it possible to perform an experiment at normal light incidence and at normal electron emission. As we have discussed previously,” this is a unique experimental geometry for the application of photoemission selection rules. The 1500 lines per mm grating on the high-energy toroidal grating monochromator” (HE-TGM1) provided light in the 250-600 eV energy range for the X-ray absorption measurements.The data were collected in the Auger yield mode using a ca. 12 eV wide kinetic energy window centred at 270 and 510eV for the carbon and oxygen KLL transitions, respectively. To minimise the intensity of direct photoemission features, the Auger electron emission was detected at an angle out of the incident plane. On rotating the sample with respect to the light beam, which is necessary in order to change the angle between the electric vector and the surface normal, the Auger emission angle was kept constant by simultaneously rotating the electron analyser with the sample. The mbar s of oxygen and subsequently to 2 x t 1 eV- 1 . 6 0 2 ~ lO-”J.D. E. Ricken et al. 293 D I I I I I 15 10 5 0 energy below E,/eV Fig. 1.Normal emission spectra of methoxy adsorbed on Cu{ 11 1) recorded at a photon energy of 28 eV. The incidence angle of the light, a, is 0" (lower spectrum) and 50" (upper spectrum). Difference spectra ( x 7), where the corresponding clean-surface spectra have been subtracted, are shown with dots. absorption spectrum of the clean surface was subtracted from that of the adsorbate- covered surface and the resulting difference spectra at each angle were normalised to the height of the edge jump. The photon energy was calibrated to an accuracy of k0.5 eV by measuring the photoemission spectrum of the Cu 3p level in first- and second-order light at a photon energy just below the absorption threshold. Angle-Resolved Photoemission In the photoelectron spectra at hv = 28 eV shown in fig.1 three adsorbate-induced features at 5.3, ca. 9.5 and 15.5 eV below the Fermi level, Ef, can be identified. Both spectra were recorded in normal emission but with incidence angles of a =O" (lower spectrum) and 50" (upper spectrum) of the radiation. Comparison with data for methoxy adsorption on other surface^'^-'^ and with molecular orbital calculations for this specie^,"^'^ indicate that the 15.5 eV feature must be due to emission from the 4a, molecular orbital. As can be seen in the spectra of fig. 1, this feature is not detected in normal emission with a = 0", i.e. with the E vector parallel to the surface, but is clearly observed at a = 50" when there are components of the E vector both perpendicular and parallel to the surface. Analysis of the photoemission matrix indicates that emission along the surface normal from an a, orbital is symmetry-forbidden for the parallel component of E, if the effective symmetry of the molecule is C3".The absence294 Surface Methoxy Species on Cu{ 1 1 l} c 9: \ b . 6 /C I I I I I 20 15 10 5 0 energy below E , / e V Fig. 2. Photoemission spectra recorded for off-normal emission with normal incidence radiation. The E vector is in the (211) azimuth. In the lower spectrum ( b ) the emission direction is perpendicular t o the direction of the E vector, i e . in the (110) azimuth. In the upper spectrum ( a ) the emission direction is coincident with the plane of the E vector, i.e. in the (211) azimuth. of the 4a, feature in the lower spectrum thus indicates that the 0-C axis is oriented perpendicular to the surface.At lower binding energy bands from three more orbitals are expected: le, 5a, and 2e. The presence of the two adsorbate-induced features at ca. 9.5 and 5.3 eV in the normal incidence/normal emission geometry enables these to be assigned to emission from the l e and 2e orbitals, respectively. On analysing the polarisation dependence of the ca. 9.5 eV feature, we also note a slight shift towards higher binding energy on introducing the perpendicular component of the E vector ( a = 50°, fig. 1). Hence the higher-energy component may be assigned to emission from the 5a, molecular orbital and the lower-binding-energy component to emission from the 1 e molecular orbital. This assignment is corroborated by data recorded at hv = 22 eV where the 5a, cross- section is higher relative to the le.The ionisation energies of the l e and 5a, components are estimated to be 9.2 f 0.2 and 9.7 f 0.2 eV, respectively. (The shift of the ca. 9.5 feature to higher binding energy for an experimental configuration which allows emission from totally symmetric states is also indicated in fig. 2). When deducing the orientation of an adsorbed species from the analysis of the polarisation dependence of the spectral features in a normal emission geometry, a lower effective symmetry cannot always be excluded. For example, if the 0-C axis were tilted with respect to the surface or the molecule adsorbed in a site with lower thanD. E. Ricken et al. 295 threefold symmetry, the nominal point group of the adsorbate-surface complex would be reduced to C,.In this case, the excitation would be polarised in the single mirror plane of the complex, i e . emission from the 4a, molecular orbital (4a' in C,) could be excited by the components of the electric vector both parallel and perpendicular to the surface. Were the effect of the parallel component small compared with that of the perpendicular component, the effective symmetry would still appear to be C3,, with the 0-C axis perpendicular to the surface, even though a substantial tilt may in fact be present. The data of fig. 2, however, exclude this possibility. These spectra were recorded at normal incidence (& = 0") and in an off-normal collection geometry, with the emission plane both coincident (upper spectrum) and orthogonal (lower spectrum) to the plane defined by the E vector and the surface normal.(These configurations are often referred to as the allowed and forbidden geometries, re~pectively.~~ Note that they are only applicable for the point groups C,,, when n = 3 , 4, 6 and a). On the assumption of C3, symmetry, the selection rules predict that emission from an a , orbital is allowed in the coincident (in-plane) geometry and forbidden in the orthogonal I out-of-plane) geometry. As expected, the 4a, emission is completely extinguished in the out-of-plane configuration. This would not be the case if the molecule were tilted (owing to the threefold symmetry of the surface) or if a low symmetry adsorption site were to reduce the effective point group to C,. Note also in Fig.2 that the ca. 9.5 eV feature is again shifted to higher binding energy in the in-plane configuration, indicative of the fact that this is a composite peak containing both l e and 5a, bands, with the latter at higher binding energy. We therefore conclude from the photoemission data that the 0-C axis of the adsorbed methoxy species is perpendicular to the Cu{ 11 1 ) surface, as indicated by the IRAS data." There are, however, two remarks to be made at this point. First, the orientation determination is dependent upon the correct assignment of a particular photoemission feature. In this case, there is little doubt that the band at 15.5 eV is due to the 4a, orbital. In more complicated adsorption systems this kind of approach may not be possible: in general, the determination of both the molecular orientation and the symmetry of the valence levels using the selection rules is not possible.The second point concerns the accuracy of the orientation determination. Whereas it is unlikely that the molecular axis is inclined by more than ca. 10" to the surface normal, photo- emission selection rules cannot give any real margin of error. In the present case the 0-C axis may indeed be slightly tilted (but certainly not as much as indicated by the analysis of the X-ray absorption data described below), and the a , photoemission matrix element could still be coincidentally zero in the critical geometries. A simple test of consistency is to vary the incident angle of light in fig. 1 and the photoelectron emission angle in fig.2, or the photon energy, both of which have been done in the present experiments with the same result. On comparing the ionisation potentials of the occupied molecular orbitals of the methoxy species adsorbed on Cu{ 11 l} with those for Cu{ loo}'" and Cu{ 1 10}25*26 it becomes apparent that the relative binding energies of the 5a, and l e orbitals are reversed on the surface. This is shown schematically in fig. 3 where the ionisation energies for all three surfaces are compared with the vertical ionisation potentials of the free CH,F mo1ecule30~3' which is isoelectronic with CH,O-. (Note that the energy scale for CH,F has been shifted so that the 4a, orbitals are at the same level.) Also shown are the calculated orbital energies of C H30- calculated by Rodriguez and Campbell2' using INDO/S.These authors have also treated the methoxy species adsorbed on zinc oxide clusters. Owing to the interaction with the substrate the relative ordering of 5a, and l e is also reversed compared with free CH30-. The 5a, orbital derives mainly from the 0 2s and 2p, orbitals and is expected to be most sensitive to the surface site. We might speculate that the relative stabilisation of 5a, relative to l e on Cu{ 11 l } reflects a stronger chemisorption bond than on the other two basal planes. In the only relevant calculation available so far, actually for Cu{100}-CH30, Zeroka and Hoffmann" find in their296 CH,F Surface Methoxy Species on Cu(ll1) INDO/S Cu{lll} Cu{lOO} CU{llO} CH 0- t l e 5 3 - l e 5al/le 3 le _. - t Fig. 3. Comparison of the ionisation energies for the methoxy species adsorbed on copper surfaces [for Cu(100) see ref.(24) and for Cu(ll0) see ref. ( 2 5 ) and (26)] with INDO/S orbital energies for the methoxy anion2' and ionisation energies of the isoelectronic molecule CH,F [see ref. (31) and (32)]. extended Huckel slab treatment a stronger interaction of the 2e orbital (mainly 0 2p, and 2pV) with the Cu d band for the on-top site compared with the fourfold hollow site. Although no pronounced effect was observed for the 5a, orbital, it would appear that there may be some slight stabilisation relative to l e for adsorption in the fourfold hollow site. Clearly, more calculations are required for these systems to clarify these rather subtle effects. X-Ray Absorption Spectroscopy X-Ray absorption spectra recorded at the oxygen K-edge are shown in fig.4 for two angles of the electric vector of the light with respect to the surface normal. Two broad resonances are observed at photon energies of 537 and 542 eV. As discussed above, such resonances can be assigned to excitations of the core 1s electron into the unoccupied orbitals of the r n ~ l e c u l e ' ~ ~ ~ ~ and/or the unoccupied states formed by the adsorbate- surface bond.33 Provisionally excluding the latter possibility, the adsorbed methoxy species might be expected to show three resonances corresponding to excitation into the 6a1, 7a, and 3e unfilled molecular orbitals. In order to use the polarisation depen- dence of the X-ray absorption spectrum for determining the orientation of the molecule on the surface, the symmetry of the resonances must in general be known.Conversely, and thus similar to the situation for photoemission, a known orientation can be used to assign the symmetry of the resonances. Data evaluation is also a non-trivial problem, as several approximations have to be made in order to fit the curves. Here we have adopted a procedure outlined by Outka and Stohr.34 The background has been approxi- mated by two Gaussian-broadened step functions, in order to simulate the edge jump, and the two resonances have been fitted with asymmetric Gaussian functions. Using this analysis procedure, the 542 eV resonance exhibits no pronounced depen- dence on the polarisation of the light. The 537 eV feature, however, exhibits a rather strong dependence on &, the angle between the E vector and the surface normal, as shown in fig.5. It is relatively straightforward to show that for C,\, symmetry the intensity of an a , resonance will vary as cos2 OE and that of an e resonance as sin' OE. The two calculated curves assume that the effective symmetry remains CJL and that this first resonance is due to an excitation into the 6al unoccupied level. The latter assumptionD. E. Ricken et al. 297 530 540 550 560 proton energy/eV Figure 4. 0 K-edge X-ray absorption spectra of methoxy on Cu(ll1) at two angles, O E , of the electric vector with respect to the surface normal. The spectra have been fitted with two Gaussian step functions to simulate the background and two asymmetric Gaussian functions at 537 and 542 eV to describe the resonance^.^^ is justified both by comparison with molecular orbital calculations for the adsorption of the methoxy species on both Cu{lOO}*' and ZnO'* as well as with X a calculations35, inner-shell electron energy loss'6 and X-ray a b ~ o r p t i o n ~ ~ measurements on CH,F. The dashed curve and the solid curve correspond to the 0-C axis inclined at angles of 10 and 30" to the surface normal, respectively, assuming a random azimuthal orientation.The 30" curve appears to be a reasonable fit to the experimental data. As far as the determination of the apparent tilt angle is concerned, one must be aware of the ambiguities present in the fitting procedure used. For an upright molecule, as indicated by the photoemission data, the intensity of the 537eV resonance should be zero for normal incidence of the light on the surface (OF = 90").As can be seen from the data of fig. 4, this is the very angle where the uniqueness of the fit must be questioned. It is doubtful whether a better signal-to-noise ratio or improved energy resolution would improve matters here. The problem lies primarily in the arbitrariness of the fitting procedure and leads us to doubt the suitability of the X-ray absorption technique itself for obtaining accurate structural information in this particular case. In our previous investigation on the Cu{ 100}-CH30 system17 some residual intensity on the a , resonance at the oxygen edge was also observed at 8, =90". Because the corresponding resonance at the carbon edge actually disappeared at BE =90", it was298 Suvface Methoxy Species on Cu{ 11 l} I I I I 20 40 60 80 6, / " Fig.5. Dependence of the 537 eV resonance on the angle O b , between the E vector of the light and the surface normal. The theoretical curves are the expected distributions for tilt angles of 10" (- - - ) and 30" (-). concluded that the remaining feature was due to a substrate scattering resonance. Thus, the polarisation dependence at the oxygen edge gave a tilt angle of 28 * lo", in good agreement with an earlier study,16 but the carbon-edge data definitely indicated a zero tilt angle (also to within *lo"). The present situation is more complicated because the residual feature comes at the same energy (assuming the correctness of the fit) and there is considerable overlap with the 542eV resonance.That the carbon-edge data do not help in this case is indicated by fig. 6: there is practically no polarisation dependence of the near-edge structure at all. We also note that the e-type resonance observed on Cu{lOO}" directly at the threshold in the C-edge data and assigned to a surviving Rydberg transition is absent on Cu{lll}. What is the origin of the resonance at 542eV? The absence of a polarisation dependence suggests that there are overlapping a , and e resonances. Their origin, however, must remain a matter of speculation: they could be the 7a, and 3e molecular resonances or a , and e substrate scattering resonances or, alternatively, combinations of both. The single resonance observed at the carbon K edge may have a similar origin.Conclusions Angle-resolved photoemission data for the surface methoxy species on Cu{ 1 1 l } show four adsorbate-induced features at 5.3, ca. 9.2, ca. 9.7 and 15.5 eV below EF. Since only the 4a, orbital gives rise to the emission peak at 15.5 eV, this can be used to determine the orientation of the molecule using selection rules. Although no definitive limits of accuracy can be established, owing to the nature of selection rules, the data show clearly that the 0-C axis is perpendicular to the surface. Having established the CZc geometryD. E. Ricken et al. 299 - 1 I I I 280 290 300 310 photon energy/eV Fig. 6. C K-edge X-ray absorption spectra of methoxy adsorbed on Cu{ 1 1 1) for two angles, Ot, of the E vector of the radiation with respect to the surface normal.of the adsorbate complex, the data further show that the 5a, ionisation energy (ca. 9.7 eV) is larger than the l e ionisation energy (ca. 9.2 eV). This indicates a reversal of order compared with the methoxy species adsorbed on the Cu{ 1 lo} and Cu{ 100) surfaces and reflects differences in the bonding schemes perhaps due to different adsorption sites. Koopmans' theorem permitting, new and more sophisticated cluster calculations may be particularly useful here in interpreting these data. The X-ray absorption results are unfortunately somewhat inconclusive. The problem may lie in the arbitrary nature of the fitting procedure, in that it is not clear exactly how many resonances are to be expected. Although one might be tempted to question the suitability of the technique itself for obtaining accurate structural information, it is fair to say that this is a particularly difficult example.At the carbon edge, where in the case of earlier work on the Cu{ 100) surface the definitive data were obtained, no polarisation dependence was observed at all. We are forced to conclude that the X-ray absorption simply precludes an accurate determination of the angle of tilt, but if such a tilt does exist, it is certainly no greater than 30". As in the case of the Cu{lOO}-CH,O system, we hope that photoelectron diffraction will provide the appropriate confirmation of the photoemission data and also give the adsorption site. These experiments are currently in progress. This work has been supported by the Deutsche Forschungsgemeinschaft through the Sonderforschungsbereich 6-8 1 as well as by the Bundesministerium f u r Forschung and Technologie through its support programme for experiments with synchrotron radiation.300 References Surface Methoxy Species on Cu(ll1) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 L.L. Kesmodel, L. H. Dubois and G. A. Somorjai, J. Chem. Phys., 1979, 70, 2180; Chern. Phys. Lett., 1978, 56, 267. D. Ying and R. J. Madix, J. Catal., 1980, 61, 48; M. Bowker and R. J. Madix, Surf Sci., 1981, 102, 542. M. E. Kordesch, Th. Lindner, J. Somers, W. Stenzel, H. Conrad, A. M. Bradshaw and G . P. Williams, Spectrochim. Acta, Part A, 1987, 43, 1561. H. H. Kung, Catal. Rev. Sci. Eng., 1980, 21, 275.K. Klier, Adv. Catal., 1982, 31, 243. J. Stohr, in Chemistry and Physics of Solid Surfaces, ed. V. R. Vaneslow and R. Howe (Springer-Verlag, Berlin, 1984), p. 231. J. Somers, Th. Lindner and A. M. Bradshaw, in ACS Symposium Series: Molecular Phenomena at Electrodes, ed. M. P. Soriaga (ACS, Washington, 1988) chap. 8, p. 111. M. Neumann and H. J. Freund, Appl. Phjls. A, 1988, 47, 3. A. M. Bradshaw, in Structure and Reactivity of Surfaces, ed. C . Morterra, A. Zecchina and G. Costa (Elsevier, Amsterdam, 1989), p. 201. A. M. Bradshaw and J. Somers, Phys. Scr., in press. I. E. Wachs and R. J. Madix, J. Catal., 1978, 53, 208. B. A. Sexton, Surf Sci., 1979, 88, 299. M. A. Chesters and E. McCash, Spectrochim. Acta, Part A, 1987, 43, 1625. M. Bader, A. Puschmann and J.Haase, Phys. Rev. B, 1986,33, 7336. E. Holub-Krappe, K. C. Prince, K. Horn and D. P. Woodruff, Surf: Sci., 1986, 173, 176. D. A. Outka, R. J. Madix and J. Stohr, Surf: Sci., 1985, 164, 235. Th. Lindner, J. Somers, A. M. Bradshaw, A. L. D. Kilcoyne and D. P. Woodruff, Surf; Sci., 1988, 203, 333. N. V. Richardson and A. M. Bradshaw, in Electron Spectroscopy: Theory, Techniques and Applications, ed. A. Baker and C . R. Brundle (Academic Press, London, 1982), p. 154. E. W. Plummer and W. Eberhardt, Adv. Chem. Phys., 1Y82, 49, 533. H. Petersen, E. Dietz and U. Sowoda, in BESSY Annual Report (Berliner Elektronenspeicherring- Gesellschaft fur Synchrotronstrahlung mbH, Berlin, 1983,) p. 192. Th. Lindner, J. Somers, A. M. Bradshaw and G . P. Williams, Surf: Sci., 1987, 185, 75. E. Dietz, W. Braun, A. M. Bradshaw and R. L. Johnson, Nucl. Instr. Methods A , 1985, 239, 359. J. L. Erskine and A. M. Bradshaw, Chem. Phjx Lett., 1980, 72, 260. P. Hofmann, C. Mariani, K. Horn and A. M. Bradshaw, in Proc. 4th Int. Conf Solid Surfaces (SOC. Francaise d u Vide, Paris, 19801, vol. 1, p. 1125. B. A. Sexton, A. E. Hughes and N. R. Avery, Surf Sci., 1985, 155, 366. P. Hofmann and D. Menzel, Surf Sci., 1987, 191, 353. D. Zeroka and R. Hoffmann, Langmuir, 1986, 2, 553. J. A. Rodriguez and C. T. Campbell, Surf: Sci., 1988, 194, 475. C. L. Allyn, T. Gustafsson and E. W. Plummer, Solid State Commun., 1978, 28, 85. K. Kimura, S. Katsamata, Y. Achiba, T. Yamazaki and S. Iwata, Handbook of Photoelectron Spectra qf Fundamental Organic Molecules (Japanese Scientific Society Press, Tokyo, 198 1 ). I . Novak, J. M. Benson and A. W. Potts, J. Electron Spectrosc, Relat. Phenom., 1986, 41, 225. J. Somers, A. Robinson, Th. Lindner, D. Ricken and A. M. Bradshaw, Phys. Rev. B, 1989, 40, 2053. Th. Lindner and J. Somers, Phys. Rev. B, 1988, 37, 10034. D. A. Outka and J. Stohr, J. Chem. Phys., 1988, 88, 3539. J. S. Tse, E. Pellach and M. G. Bancroft, Can. J . Chem., 1985, 63, 457. A. P. Hitchcock and C. E. Brion, J. Electron Spectrosc. Relat. Phenom., 1978, 13, 193. F. C. Brown, R. Z. Bachrach and A. Bianconi, Chem. Phys. Lett., 1978, 54, 425. Paper 9/05382D; Received 15th December, 1989

ISSN:0301-7249    

DOI:10.1039/DC9908900291

出版商:RSC    

年代:1990

数据来源: RSC

摘要:

Furuduy Discuss. Chem. SOC., 1990, 89, 301-310 Normal-incidence Standing X-Ray Wavefield Absorption and SEXAFS Studies of Adsorption Structures on Cu and Ni Surfaces N. P. Prince, M. J . Ashwin and D. P. Woodruff Physics Department, University of Warwick, Coventry CV4 7AL Naginder K. Singh, W. Walter and Robert G. Jones Chemistry Department, University of Nottingham, Nottingham NG7 5R D The relative insensitivity of the Bragg condition to the exact angle of incidence to the scattering plane near normal incidence means that the X-ray standing wavefield absorption method can be applied to typical (imperfect) surface-science single-crystal samples in this special geometry. The method is highly complementary to the SEXAFS experiment that can be performed with the same instrumentation and we demonstrate this with examples of structures associated with CI, CH3S-- and S on Cu( 1 1 1) and Hg on Ni( 100).The combined approach is particularly valuable in cases in which the adsorbed species induce substrate reconstruction, and in particular reason- ably definitive structures of considerable complexity have been identified for the two S-containing surfaces. The application of the EXAFS (extended X-ray absorption fine structure) method to the study of surface adsorption structures has now been persued for some 10 years and is a well established tool in the surface scientist's armoury of structural probes.'-3 As with most other surface structural techniques, it is possible to solve structural problems completely with surface EXAFS (SEXAFS) alone, under favourable circumstances, but complementary information from other techniques can often simplify this process or enhance its incisiveness.In SEXAFS the primary information content is the adsorbate- substrate nearest-neighbour scatterer distance and this can often be obtained quite reliably even from data of relatively poor signal-to-noise ratio. More complete site identification is possible if information on more distant shells of scattering atoms can be extracted, or if reliable first-shell amplitudes can be measured as a function of the polarisation direction of the incident radiation, leading to a determination of the polar angle of the nearest neighbours relative to the surface normal. A technique which provides rather complementary information in the form of adsorbate-substrate layer spacings is X-ray standing wavefield (XS W) absorption, and we have shown recently that this technique can be applied to metal single crystals prepared in the usual way by choosing to work close to normal incidence to the appropriate Bragg scattering planes (normal incidence XSW or NIXSW)."-' The XSW technique exploits the fact that at the Bragg scattering condition the incident and diffracted beams interfere to produce an X-ray standing wavefield with a periodicity equal to (or an integral fraction of) the scatterer plane spacing (fig.l ) , and that the phase of this wavefield relative to the scatterers shifts in a systematic and well understood fashion through the finite width of the (dynamical) Bragg condition.' Note, incidentally, that the intensity of the X-ray wavefield has twice the spatial frequency of the amplitude. As this X-ray wavefield extends outside the crystal surface the location of an adsorbed species relative to the substrate scatterer planes can be determined by monitoring the 301302 Adsorption Structures on Cu and Ni Surfaces \ / standing wave field amplitude intensity Fig.1. Schematic diagram of the X-ray standing wave pattern set up at a Bragg reflection condition by interference between the incident and diffracted beams. Also shown are the reflectivity ( a ) (for zero absorption) as a function of incidence angle or energy around the Bragg condition, and the absorption profiles for atoms on ( b ) , and mid-way between ( c ) , the atomic scatterer planes.X-ray absorption at the adsorbate site through the Bragg condition. Maximum absorp- tion (monitored by measuring the fluorescent or Auger electron yield from the adsorbate) occurs when the wavefield phase is such as to locate the standing wave antinodes on the adsorbate sites. A limitation of the method in its usual (general) form is that the typical angular ('rocking curve') width of the Bragg peak is very narrow (seconds of arc) necessitating the use of crystals of a very high degree of perfection (low mosaicity) if the effect is to be observed. At normal incidence to the Bragg planes; however, the Bragg condition becomes very insensitive to the exact incidence angle, leading to an effective rocking curve width as large as 1"; under these conditions we have shown that the method can be applied to normal metal crystals.In this paper we demonstrate the effectiveness of the NIXSW method, and its complementarity to SEXAFS, with four examples. The first two [CI on Cu(ll1) and Hg on Ni( loo)] are simple chemisorption systems, whilst the second two [S and CH3S- on Cu( 11 1 >] both appear to involve significant disruption of the metallic substrate. Simple Overlayer Chemisorption Systems As a test case for the initial investigation of the NIXSW technique4.' we investigated the ordered chemisorption structure Cu( 11 1 )(a x A)R3Oo-Cl which we had previously investigated by SEXAFS and photoelectron diffraction (PD) and found to correspond to simple chemisorption with CI atoms occupying the (f.c.c.) three-fold coordinatedN.P. Prince et al. 3 03 2 .o- C 0 .C Y 1 .5-- 0.51 I -4 -2 0 2 4 relative photon energy/eV Fig. 2. Calculated X-ray absorption profiles for normal-incidence ( 1 1 1 ) Bragg reflection from Cu as a function of energy relative to the nominal Bragg condition for different absorber layer spacings relative to the Cu( 1 1 1 ) scattering atom planes. ( a ) 0.0 A, ( b ) 0.5 A, ( c ) 1.0 A, ( d ) 1.5 A. The calculations include instrumental and sample broadening of 1.7 eV in energy and 0.23" in angle (FWHM). (- - -) Random filling of all possible layer spacings (incoherent absorption). hollow $tes above Cu atoms in the third layer, with a local CI-Cu top-layer spacing of 1.88 A.' Bragg reflection from the ( I 11) planes of Cu at normal incidence occurs at an X-ray energy of 2975 eV, conveniently close to (but above) the CI K-edge at 2820 eV.The experiments were conducted on the same synchrotron radiation beamline (station 6.3 on the SRS at the S.E.R.C.'s Daresbury Laboratory) used for the SEXAFS and PD investigations. At normal incidence to the Bragg planes the Bragg condition is most conveniently swept by scanning the photon energy, an arrangement which not only removes the need for a high-precision sample goniometer, but also avoids problems with the complex changes in the rocking curve which can occur close to normal incidence. In this mode the intrinsic Bragg peak width ( i e . the range of total reflectivity in the absence of absorption) for Cu( 11 1) is 0.87 eV. In fact the high-flux (focussed) beamline designed for SEXAFS has an energy resolution worse than this so some broadening of the reflectivity or absorption profile results from this instrumental effect.The angular convergence of the (focussed) incident beam, however, like the crystal mosaicity, has little effect. Fig. 2 shows calculated X-ray absorption profiles (and the reflectivity profile) including the actual instrumental limitations found for this experiment (1.7 eV FWHM energy resolution and 0.23" angular spread) for different locations of the X-ray absorbing species relative to the atomic scatterer planes ( i e . the Cu atom planes in this case). It is clear that we anticipate substantial sensitivity to the absorber locations in an NIXSW experiment despite these instrumental limitations. Experimental data from the Cu( 11 1 )(a x &)R3Oo-C1 structure are shown in fig. 3, the X-ray absorption at the Cu and CI atoms being measured by recording the intensities of the Cu L,VV and CI KLL Auger electron peaks relative to their respective secondary electron backgrounds.The Cu signal is fitted by the tbsorption profile expected for atoms on the scatterer planes (i.e. at a 'sQacing' of 0.0 A ) whilst the C1 signal is found to correspond to a layer spacing of 1.81 A with an estimated error of ca. 0.05 A. Note that the XSW technique determines the absorbate layer spacing relative to the extension of the (bulk) scatterer planes, whilst SEXAFS determines the layer spacing relative to the nearest-neighbour scatterers. As a result a difference between the SEXAFS and NIXSW layer spacings implies an expansion or contraction of the uppermoost substrate layer spacings.In the present case the implied contraction (0.07*0.06 A ) is of very marginal significance, but this does highlight an intrinsic complementarity of the two techniques, and indeed a recent study of CI adsorption on Cu(100) by these methods leads to a more experimentally significant conclusion of expansion.'304 Adsorption Structures on Cu and Ni Surfaces I I -6 -3 0 3 6 relative photon energy Fig. 3. Experimental CuL,VV (- . -) and C1 KVV ( a * .) Auger electron yields as a function of energy around the normal-incidence (1 11) Bragg reflection from a Cu( 11 1)(& x &)R30°-C1 surface compared with theoretical profiles for various absorber layer spacings ( A ) : ( a ) 0.0, ( b ) 1.74, ( c ) 1.81, ( d ) 1.88.I I I I I 1 0 -4 -2 0 2 4 relative photon energy/eV Fig. 4. Experimental Ni (-) and Hg ( - - - ) Auger electron yields as a function of energy around the normal-incidence (200) Bragg reflection from an Ni( 100)c(2 x 2)-Hg surface compared with theoretical profiles for various absorber layer spacings (A): ( a ) 0.0, ( b ) 0.4, ( c ) 0.6, ( d ) 0.8. One notable feature of the experimental data of fig. 3, when compared to the model calculations of fig. 2, is that the NIXSW profiles of Cu and C1 are quite similar relative to the wide range of possible absorption lineshapes. In fact very many simple atomic overlayer chemisorption systems appear to lead to occupation of the same (highly coordinated) site which would be occupied by the next layer of sustrate atoms, and becaus? typical adsorbate-substrate bondlengths are very similar (i.e.within a few tenths of an A) of substrate-substrate bondlengths, this result in XSW is likely to be common for such systems. One case in which we have found this not to be the case, however, is in the Ni( 100)c(2 x 2)-Hg adsorption system"'for which NIXSW data recorded around the Ni(100) Bragg condition are shown in fig. 4. In this case the best fit to the Hg absorption profile 1e:ds to an inferred Fg-Ni layer spacing of 0.6*0.1 A (or, more plausibly, 2.36* 0.10 A, adding a bulk Ni layer spacing). If the Hg-Nd nearest-neighbour spacing is assumed to be the same as in the compound NiHg (2.63 A), this leads to the surprising conclusion that the adsorbed Hg atoms occupy bridging sites on the surface.Unfortunately, this conclusion is not unambiguous because an alternative explanation based o n substrate expansion and a longer local Hg-Ni bondlength cannot be excluded. The absence of a direct measurement of the Hg-Ni bondlength by SEXAFS leads toN. P. Prince et al. 305 this ambiguity and again highlights the complementarity of the methods. Note, however, that it is possible to determine the adsorption site by NIXSW alone if the experiment is performed at several different Bragg conditions, leading to the determination of adsorbate-substrate plane spacings in different directions (not all perpendicular to the surface) and thus locating the exact adsorbate site by triangulation. Adsorbate-induced Reconstructive Systems Although the idea that surface adsorption may lead to reconstruction of the uppermost layer(s) of a surface is not new, it is only relatively recently that hard structural evidence of this effect as a common phenomenon has become available. In the early stages of amassing quantitative information on surface adsorption structures, particularly in the 1970s using LEED, simple overlayer chemisorption with no apparent substrate rearrange- ment appeared to be the norm, and some adsorption systems previously thought (on the basis of qualitative arguments alone) to involve mixed adsorbate-substrate atomic overlayers were found to involve no singificant substrate atom movement.' ' More recently clear evidence of distortion or reconstruction of transition-metal surfaces by light atomic adsorbates (C, N and 0) has emerged for several systems.'? The investigation of these more complex structural problems requires information not only on the local adsorbate-substrate coordination (such as that readily obtained from SEXAFS), but also longer-range information.In this situation the complementary aspects of SEXAFS and NIXSW may be particularly valuable. We have investigated two structural problems associated with the reaction of sulphur-containing species with Cu( 11 1) in which the combined information from the two methods has proved invaluable. The first of these concerns the structure of the species CH$- on Cu( 11 1 ) . This species can be formed on Cu surfaces either by deprotonation of methyl thiol, CH,SH or S-S bond scission in dimethyl disulphide, (CH,S), through direct catalytic decompo- sition by the appropriate Cu surface.'' There is some interest in such reactions both directly in terms of their relevance to desulphurisation reactions, and in the extent to which this surface species is an analogue of methoxy, CH30-, a surface intermediate in methanol formation and decomposition to produce formaldehyde.N IXSW measure- ments from this system,'" recorded around the (111) Bragg reflection using Cu L,VV and S KLL Auger electron yields, reveal an S absorption profile which is almost exactly the inverse of that from the substrate species (fig. 5) indicating that the S atoms lie approximately midway between the (extended) Cu lattice planes (with a spacing of 1.20*0.10 A determined by optimisation over several data sets). By coontrast the Cu-S nearest-neighbour distance is found, from SEXAFS, to be 2.38 f 0.03 A, closely similar to the Cu-CI distance in Cu( 11 1 ) ( a x fi)R3O0-CI of 2.34* 0.03 A.Simple overlayer chemisorption in a hollow adsorption site should therefore lead to a NIXSW profile like that seen for C1 in fig. 3. Clearly the two situations are very different. The dependence of the SEXAFS amplitude on the polarisation direction of the incident X-radiation confirms this difference, yielding a S-Cu bond angle relative to theBurface normal of 60*5", implying a local top layer Cu-S layer spacing of 1.2*0.2 A, identical to the NIXSW value. These parameters imply that the S atom in the CH7S- species partly penetrates the top Cu atom layer; if we assume that this occurs in the most favourable hollow sites, it Fecessitates movement of the nearest-neighbour Cu atoms parallel to the surface by 0.6 A, indicating major reorganisation of the metal surface layer.A more detailed picture of the surface structure in this system is difficult to obtain because it is not possible to determine any long-range order; attempts to perform LEED observations led to molecular decomposition and the production of the sulphur phase discussed below. Moreover, it is not clear that the surface overlayer is truely commensur- ate (ie. that there is a unique and well defined local structure relative to the substrate). Some further information is available, however, from the SEXAFS and NIXSW results.306 h C 0 .- Adsorption Structures on Cu and Ni Surfaces 0.51 I I , I I I I I -4 -2 0 2 re 1 a t i v e photon energy / e V Fig.5. Experimental Cu (- - - ) and S (- . -) Auger electron yields as a function of energy around the normal-incidence ( 1 1 1 ) Bragg reflection from Cu( 1 1 1 ) with an adlayer of CH3S- (A), and from Cu( 11 1 ) ( d x d ) R 1 9 " - S ( B ) . Various theoretical profiles are also shown corre- sponding to L-layer spacings (in A ) of ( a ) 0.0, ( b ) 1.05, ( c ) 1.15 and ( d ) 1.25. In the case of the S structure, the theoretical fit to the S emission corresponds to 20% of atoms at a layer spacing of 0.85 A, and SOo% at a spacing of 1.75 A. First, modelling of the SEXAFS data collected at normal and grazing incidence (fig. 6), indicates that while the EXAFS is dominated by the nearest-neighbour copper scatterers, some evidence of the influence of S and C back-scattering is seen at low energies a t these two angles.The C back-scattering is seen only at grazing incidence (and not at normal incidence), indicating that the S-C axis of the molecule is perpen- dicular to the surface (and thus perpendicular to the A-vector of the X-radjation at normal incidence). The S-C bond length foundo from this simulation (1.90 A ) is not too far from that found in methyl thiol (1.81 A ) , bearing in mind the absence of phase-shift optimisation and the short data range over which the C back-scattering is signifitant. The S . . . S distance inferred lrom the low-energy EXAFS at normal incidence is 3.4 A; again the error is difficult to estimate, but this value could be consistent with the C u - - - C u distance (3.57 A ) inferred from the Cu-S bond length and bond angle as discussed above.This agreement could be reconciled with a regular hexagonal Cu and CH3S- mesh of this periodicity (corresponding to a surface coverage of 0.5 1 monolayer). One point we have not considered so far, however, is that the SEXAFS and NIXSW Cu-S layer spacing agree (within the large experimental errors) despite the fact that we have proposed that the top C u atomic layer suffers substantial reconstruction; in particular, large movements of these top-layer Cu atoms parallel to the surface wouldN. P. Prince et al. 307 +O.OL 0.00 -4 2 -0.04 8- .- c) u C +0.04 z 2 - u rn g 0.00 c -0.04 I 1 I I 1 I I 3 4 5 6 7 8 9 1 0 1 1 electron wavevector, & / A - ' Fig.6. S K-edge SEXAFS fine structure functions recorded at normal [ 8 = 90" ( a ) ] and grazing [ 8 = 30" ( h ) ] incidence from Cu( 11 1 ) with an adlayer of CH3S-, compared with model calculations including first shell Cu ( - - -), S [(. . .) normal incidence] and C [ ( a * * ) grazing incidence] scat terers. be expected to be accompanied by a significant layer expansion perpendicular to the surface as these atoms 'ride-up' on the layer below. One possible solution to this problem is to identify local adsorption sites for which these parallel distortions can be accommo- dated with the minimum layer expansion. We have already shown'" that such a site is the 'h.c.p.' hollow (above a Cu atom in the second layer) in which the lateral expansion of the hollow site leads to the top-layer Cu atoms being displaced towards !ridging sites on the layer below, minimising the top-layer expansion to less than 0.12 A, which lies within the error bars of our experimental layer spacings and thus could reconcile the SEXAFS and NIXSW results. These sites, however, need to be well separated to allow all the adsorbed species to occupy such sites, whereas the photoemission and Auger electron yields indicate that the coverage is not particularly low; no precise measurement is possible, but we estimate a coverage in the range 0.3-0.6 monolayer.The need to accommodate this higher coverage leads us to consider variants of this model. One particular possibility is illustrated in fig. 7. If a CH3S- species is put into an h.c.p.hollow site, but with the Cu nearest neighbours opened up to a separation of 3.82 A, these displaced atoms occupy the bridge sites of the second metal layer exactly. An ordered hexagonal Cu overlayer of this spacing leads to a ( 3 x 3) structure with four C u overlayer atoms per unit mesh, three in bridge sites and one in atop. Filling the three-fold hollow sites of this overlayer leads to a coverage of CH3S- of 0.t4 monolayer, and if the Cu-S nearest-neighbour bond-length is maintained at 2.38 A, the Cu-S bond angle increases to 68" (close to our experimental limits), and the local Cu-S layer spacing (0.89 A) plus the average top-layer Cu expansion (0.23 A) leads to a predicted NIXSW spacing of 1.12 A, in good agreement with experiment.The S - . . S distance in this layer (3.82 A) is, however, rather long relative to the SEXAFS value.308 Adsorption Structures on Cu and Ni Surfaces Fig. 7. Schematic diagram of a possible ordered commensurate ( 3 x 3) structure for CH3S- on Cu( 11 l ) , illustrating the form of the substrate reconstruction induced by this molecular adsorbate. Although there is no direct evidence that this long-range ordered structure does occur, it serves to demonstrate that quite a small increase in the Cu-S bond angle and the Cu-Cu overlayer spacing relative to the mean experimental value does allow the local Cu-S layer spacing to fall sufficiently to allow significant Cu surface layer expansion to be accommodated without being in conflict with the NIXSW result. Of course an incommensurate ordered overlayer would be forced to display a larger average layer expansion than the (3 x 3) structure of fig.7, but if the need for local distortions from this long-range order is included, severe expansion regions can be avoided. This ordered structure thus provides a basis for understanding the form of the reconstruction induced by CH3S- on C u ( l l l ) , but is almost certainly an idealisation of the true situation. By contrast to this surprising major reconstruction induced by an adsorbed molecule, the idea that atomic sulphur may lead to the formation of a surface sulphide phase, with a mixture of S and Cu atoms in the uppermost layers(s) was proposed many years ago" to account for the observed complex coincidence-mesh LEED patterns, although no quaptitative confirmation of these structures has been made.The Cu( 1 1 l)(J? x fi)R19"-S structure may be formed by interaction of H2S with Cu( 1 1 I ) , although it is also observed as a result of segregation of atomic S from bulk Cu or, indeed, by electron-beam dissociation of the surface CH$- species described above. S K-edge SEXAFS from this system" indicates a nearest-neighbour Cu distance of 2.30 A, although there is some evidence from polarisation-dependent studies that there may actually be two slightly different distances close to this value; the latter measurements also indicate an average nearest-neighbour bond angle of 67 * 5", which we have already seen to be incompatible with simple overlayer chemisorption. In this case, however, we find that the NIXSW data (see also fig.5) cannot be reconciled with a single S-Cu layer spacing. Moreover, remeasurement of this system indicates that whilst the SEXAFSN. P. Prince et al. 309 Fig. 8. Schematic diagram of a model of the Cu( 11 1 )(w'? x v'?)RIY"-S structure consistent with the SEXAFS and NIXSW measurements from this system. nearest-neighbour spacing is not in doubt, NIXSW profiles do vary but can always be fitted with two layer spacings of ca. 0.8 and 1.8 A, although the relative occupations of these states do vary (the 0.8 8, spacing species being in the minority in the range of ca. 20-40%). The existence of two different S-layer spacings and of the large S-Cu bond angle are both consistent with a surface phase of CuS stoichiometry which is essentially a distorted ( 1 11 ) layer of cubic Cu,S with near-tetragonal bonding within the layer.A more complex question concerns the exact registry of this layer with the underlying surface, and the origin of the variability of S-layer spacing occupation. The model which we propose for this structure is shown in fig. 8. S atoms in the surface sulphide phase (three S atoms per overlayer surface mesh) occupy three different sites relative to the substrate; these are the two different hollow sites (f.c.c. and h.c.p.) and atop sites. In such a structure it is not obvious, Q priori, whether the S atoms lie above or below the overlayer (sulphide) Cu atoms relative to the underlying metal, but it seems that the S atoms above hollow sites can go below the Cu overlayer atoms (achieving six-fold coordination), whilst those above the atop sites must stay above the Cu overlayer.Quite small distortions could force hollow site S atoms to 'flip-up', however, thus providing a possible explanation for the variable layer occupation. Although this system is clearly too complex for us to be sure that the structure has been totally and uniquely defined, it is clear that the combination of the SEXAFS and NIXSW results is crucial in achieving the present structural assignment. Discussion Despite its considerable successes, the SEXAFS technique is not without its difficulties, most particularly with obtaining suficiently good signal-to-noise ratios to obtain unique310 Adsorption Structures on Cu and Ni Surfaces and reliable structural information beyond the nearest-neighbour scatterer environment.The use of complementary information from a range of methods is an essential feature of all high-quality surface-science studies but in the case of SEXAFS, other techniques which utilise a tunable source of X-radiation in the same energy range are particularly attractive complements, offering efficient in situ compatibility. One such method, photo- electron diffraction, is of particular value in cases in which local site information is in doubt’ (perhaps because a low local symmetry site is involved, leading to confusing pseudo-first-shell SEXAFS information”). Another such method is NIXSW. It has the advantage of simplicity and speed; a narrow energy range is scanned and the modulation of the photoionisation signal is much higher than in EXAFS, leading to less demands on the statistics of the data.It requires, of course, that the photoionisation cross-section of the adsorbate species of interest is reasonably large at the energy of the substrate Bragg reflection(s), a condition that probably excludes its use for adsorbates in the first full row of the periodic table. In those adsorbate-substrate systems in which it can be used, however, we find, as demonstrated in the examples above, that the additional and complementary structural information is of enormous value. Indeed, for those systems in which substrate reconstruction appears to occur (an increasingly large proportion of adsorption systems) it is unlikely that entirely reliable structural conclusions could have been obtained with SEXAFS alone.The authors are pleased to acknowledge the support of staff at the S.E.R.C.’s Daresbury Laboratory, without which many of the experimental results reported here could not have been obtained, and to thank the S.E.R.C. for financial support in the form of studentships and research grants to perform this work. N.K.S. thanks the Association of Commonwealth Universities for a studentship. References 1 J . Stohr, in X - R a y Absorption, Principles, Techniques, Applications cf EXAFS, SEXAFS and XANES, ed. R. Prins and D. C. Konongsberger (Wiley, New York, 1988). 2 P. H . Citrin, J. PbJ3.s. ( P a r i s ) , 1986, CS, 437. 3 D. P. Woodruff, Rep. Prog. PhIx, 1986, 49, 683. 4 D. P. Woodruff, D. L. Seymour, C. F. McConville, C. E. Riley, M . I). Crapper, N . P. Prince and R. G. Jones, Pl7y.s. Reii. Lett., 1987, 58, 1460. 5 D. P. Woodruff, D. L. Seymour, C. F. McC‘onville, C. E. Riley, M . 1). Crapper, N. 1’. Prince and K. G. Jones, SUI$ Sci., 1988, 195, 237. 6 B. W. Batterman, Phjx Rec. A, 1964, 133, 759. 7 M. I). Crapper, C. E. Riley, P. J . J. Sweeney, C’. F. Mcc‘onville, I>. P. Woodruff and R. G. Jones, Sirrf 8 A . A. McDowell, D. Norman, J . B. West, J. c‘. Campuzano and K. C . Jones. Nircl. It7.vlrum. Melhods 9 J. R. Patel, D. W. Berreman. F. Sette, P. H. C-itriti, J . E. Rowe, P. L. c‘owan, T. Jach and B. Karlin, 10 N. P. Prince, N . K. Singh, W. Walter. I). P. Woodruff and R. G. Jones. ./. (hndensc~tl Matter, 1989, 1, 1 1 M. A. Van Hove and S. Y. Tong, Surfuce (’t.~~stul/ogr~~I,II?. h!, LEEf> (Springer-Verlag, Berlin, 1979). 12 See, for example, the systems described in this Discussion by A. L. D. Kilcoyne, D. P. Woodruff, A. Robinson, Th. Lindner, J. Somers and A. M. Bradshaw. 13 S. Hao. C. F. McConville and I). 1’. Woodrul-f, Surf. Sci., 19x7. 187, 133. 14 N . P. Prince, D. L. Seymour, D. P. Woodrutf, R. G. Jones and W. Walter, .Surf. Sci.. 1989, 215. 566. 15 J. L. Domange and J. Oudar, Surf: &I., 1968, 11, 124. 16 N. P. Prince, L). L. Seymour, M. J. Ashwin, <’. F. McC’onville, I). 1’. W o d r u l ? . K. G. Jones and W. Walter, Surf: Sci., submitted for publication. 17 I). P. Woodruff, C ’ . F. McConbille, A. L. I>. Kilcoyne, Th. Lindner, J . Sorners, M. Surnian, G. Paolucci and A. M. Rradshaw, Sur. Sci., 19XX, 201. 218. .Sc,i., 1987, 182, 213. A, 1986, 246, 131. PIi?-x. Rec. R. 19x9, 40, 1330. sn2i.

ISSN:0301-7249    

DOI:10.1039/DC9908900301

出版商:RSC    

年代:1990

数据来源: RSC

摘要:

Faraday Discuss. Chem. SOC., 1990, 89, 31 1-321 Photoelectron Diffraction Study of 0, N and C Adsorption Structures on Ni(100) and Cu(ll0) A. L. D. Kilcoyne and D. P. Woodruff Physics Department, University of Warwick, Coventry CV4 7AL A. Robinson, Th. Lindner, J. Somers, D. E. Ricken and A. M. Bradshaw Fritz-Haber-Institut, Faradayweg 4-6, D-1000 Berlin 33, Federal Republic of Germany Scanned energy mode adsorbate 1 s photoelectron diffraction measurements have been taken from half-monolayer structures of 0 on Ni(100) and on Cu( 1 lo), and from C and N on Ni( 100). The sensitivity of these data to the various structural parameters of these surface phases is investigated. In particular, the fact that 180” scattering contributions typically dominate the spectra in this form of the photoelectron diffraction experiment leads to a strong sensitivity to adsorbate-second layer substrate spacings.Even in the case in which the adsorbate and top substrate layers are nearly coplanar, however, some sensitivity to top layer geometry is retained, except in the case of the Cu( 110) surface for which the lower rotational symmetry, com- bined with the plane polarisation of the incident radiation, leads to insensitiv- ity to certain sites. There are many examples of structural determinations of atomic adsorption structures which indicate that simple chemisorption involves adsorbed atoms forming a layer above an essentially undisturbed substrate surface, typically in the highest coordination sites.’ However, there is now an increasing awareness that adsorbates may often induce relaxation, distortion or even major reconstruction in the underlying substrate.In such cases the number of structural parameters to be determined may be large and individual techniques are typically much more sensitive to some of these parameters than others. A proper recognition of this fact, and clear identification of the structural parameters to which a technique is most sensitive is an important prerequisite to an accurate and reliable structural determination by a combination of methods. In this regard it is worth noting that many structural determinations in the past have not considered the possibility of substrate reconstruction; the apparent success of these determinations need not infer that such a reconstruction does not, in fact, occur.One particular structural technique which we have already applied with some sucess to the determination of local adsorbate-substrate registry is scanned-energy mode photoelectron diffraction’ (PD) in which the intensity of an adsorbate core-level photo- emission peak is measured at specific emission angles as a function of photoelectron kinetic energy, typically within the energy range 100-400 eV.3-x+ Coherent interference between the directly emitted component of the photoelectron wavefield, and components elastically scattered by surrounding atoms, provides information on the scattering pathlengths and thus on the structural environment of the emitter. We have shown that the technique can prove particularly incisive in the determination of the local registry of adsorbed molecular fragments (especially formate, HCOO-) for which other methods had proved controversial in their interpretati~n.~” One particular feature of this tech- nique, evidenced in studies by several groups, is that the most important elastic scattering + I eV= 1 .6 0 2 1 8 ~ I O - ” J . 31 1312 0, N and C Adsorption Structures on Ni( 100) and Cu( 110) events leading to PD modulations are those involving scattering at angles close to 180°, thus providing information on pathlengths to scatterers essentially directly behind the emitter relative to the collection direction. One reason for this effect is the peaking of the elastic scattering cross-section at this angle which normally occurs at elevated energies. This effect does mean that distances to different neighbours can be probed preferentially by performing experiments at appropriate take-off angles.In this paper we consider the results of experiments of this type from a series of structures in which there is believed to be substrate relaxation [ Ni( lOO)c(2 x 2)-01, substrate relaxation and distortion parallel to the surface [Ni( 100)(2 x 2)-N and Ni(100)(2 x 2)-C] and probable major substrate reconstruction [Cu( 110)(2 x 1)-01. In each case we have attempted a complete structural determination although we find our ability to do so is somewhat restricted; in particular, most of these systems appear to have an associated adsorbate location which is quite close to being coplanar with the top metal atom layer, and this makes it difficult to select a geometry which is highly sensitive to the location of this top layer of substrate atoms.By contrast, however, the location of the adsorbed atoms relative to the second substrate atom layer is found rather readily, and indeed, rather precisely. Experimental and Computational Methods The general method of performing scanned-energy mode photoelectron diffraction on the 1s photoemission signals from C, N and 0 has already been described in relation to our studies of molecular adsorbates.”” In particular, the necessary source of soft X-radiation in the energy range 300-900 eV was the Fritz Haber Institute’s high energy torroidal grating monochromator” taking light from the BESSY electron storage ring in West Berlin. The Ni(100) and Cu(ll0) samples were prepared by the usual methods of X-ray LauC orientation, spark machining, polishing and in situ cleaning with argon ion bombardment and annealing cycles.Surface order and cleanness were monitored by LEED and by XPS and NEXAFS performed with the synchrotron radiation. Oxygen adsorption structures were formed by exposure to molecular oxygen at room temperature. The carbon structure was formed by exposure at elevated temperature to ethene, and the nitrogen structure was formed by low energy (500 eV) ion bombardment with nitrogen followed by annealing; all these procedures are ones previously used by us in other studies of these Photoelectron diffraction spectra were obtained by recording kinetic energy scans of the angle-resolving (V.G. ADES) spectrometer in the region of the appropriate adsorbate 1s photoelectron peak at each photon energy which was stepped through the kinetic energy range of interest.These spectra were subsequently subjected to back- ground subtraction and integration to yield a spectrum of integrated 1s emission intensity against photoelectron energy. Model calculations of the anticipated photoelectron diffraction for possible adsorp- tion structures were performed using a spherical wave, double scattering method from a cluster of typically 500 scattering atoms as described el~ewhere.’.~ It was also found that much simpler calculations, based on single scattering only, from very much smaller clusters, gave rather similar results in the main amplitude variations and proved par- ticularly valuable in identifying the approximate value of the main structural parameters to which the method proves sensitive.Here, however, we present only the results of the full cluster calculations. Results and Discussion Ni( 100)c(2 x 2)-0; Simple Overlayer Chemisorption? One structural problem which has been studied in particular depth is the Ni( 100)c(2 xA. L. D. Kilcoyne et al. 313 - - - side bulk spacing n Fig. 1. Schematic diagram of the proposed structure of Ni( 100)c(2 x 2 ) - 0 . 2)-0 structure formed by chemisorption of 0.5 monolayer (ML) oxygen. Indeed, this system has also formed the basis of a previous scanned-energy mode photoelectron diffraction experiment (over a less extensive data range than the present study) which was successfully interpreted.2 It is now generally agreed that the oxygen atoms occupy the four-fold coordinated hollow sites on this surface at a layer spacing above the top Ni layer of around 0.9 (see fig.1). There remain, however, some controversial details to this structure. In particular, the exact spacing of the oxygen atomic layer relative to the top and second Ni layers is not completely clear, although some expansion of the top Ni layer spacing relative to that of bulk Ni seems to be favoured.16 In addition, there have been suggestions that the oxygen atoms may be displaced parallel to the surface by ca. 0.4A towards the two-fold coordinated bridge sites;” this ‘pseudo’ site is still favoured by some workers although there is increasing evidence that it is not Finally, it is possible that there is slight ‘rumpling’ of the second Ni layer, with the atoms below the oxygen adsorbates being at a different layer spacing to those below unoccupied sites.This rather subtle effect (probably <0.05 A) is not investigated here. Fig. 2 shows 0 1s photoelectron diffraction data collected at normal emission from this structure compared with model calculations for structures based on hollow site occupation but with slightly different O-Ni top layer spacings ( z - S z ) , and some associated Ni-Ni top layer expansions (62). On the basis of our comment that PD favours scattering events of 180” scattering angle, we might expect that the normal emission geometry would be most sensitive to the distance to the second layer Ni atom directly below each 0 emitter. (For this reason we have defined the distance z as the separation of the adsorbate layer and the location of the top metal layer on an ideally terminated solid; using this definition calculations at constant z correspond to the 0 overlayer being fixed relative to the underlying metal substrate below the top metal layer.) This effect is seen in the results which show a much greater sensitivity to z than to 6 z (indicating a relatively weak sensitivity to the true O-Ni top layer spacing z - S z ) .Nevertheless, some dependence on Sz is seen, and the detailed shape of the spectral features in the range 100-250 eV is best fitted by including a small top-layer expansion314 0, N and C Adsorption Structures on Ni( 100) and Cu( 110) 0.00 0.05 0.10 0.15 0.15 z/B, 0.90 0.90 0.90 0.9c 0.85 0 .8 5 0. 8E 0.85 I , , , I 1 100 200 300 400 photoelectron energy/eV Fig. 2. Comparison of experimental normal emission 0 1s PD spectrum from Ni( 100)c(2 x 2 ) - 0 with the results of theoretical calculations for different 0-Cu top layer spacings and Cu-Cu top layer spacing expansions, assuming the 0 atoms occupy four-fold coordinated hollow sites. (62 =0.10*0.05 A) with a true O-Ni top layer spacing ( z - 62) of 0.75 k0.05 A. We should, at this point, note that in comparing theoretical and experimental spectra we concentrate on peak positions and relative amplitudes of nearby peaks only; in a real experiment the amplitudes of more widely spaced features are influenced by collector efficiency, monochromator output and slow variation in the absolute photoionisation cross-section, none of which are included in the theory.One way of reducing this problem is to construct ‘fine structure f ~ n c t i o n s ’ ~ from the experimental data (as is done in EXAFS); we have not applied this procedure here, but rather show raw experimental data. Of course, the PD technique offers the possibility of changing the collection geometry to set different atoms ‘behind’ the emitter (i.e. close to the favoured 180” scattering angle condition). PD spectra were recorded from this system for off-normalA. L. D. Kilcoyne et al. 315 0 0 top layer Ni atom clean surface top layer Ni atom site C or N atom z - 62 bulk spacing + 62 side Fig. 3. Schematic diagram of the proposed structure of Ni( 100)(2 x 2)-C and Ni( 100)(2 x 2)-N.emission, but most of these experiments were performed in the (110) azimuth which also tends to favour scattering from second layer Ni atoms. For example, measurements were made at 45" emission in (1 10) but the favoured scattering atom is the second layer Ni atom adjacent to that lying directly below the emitter; these spectra show even less sensitivity to the top metal layer expansion, but confirm the optimum value of z. An independent measurement of z - 62, the local 0-Ni spacing, should be achieved by PD recorded at around 60" in (loo), favouring strong scattering by a nearest-neighbour atom, and an experiment of this kind is now being analysed. We should also note that, even in the absence of these additional measurements, comparison of the existing data and model calculations clearly favour the true hollow site over the 'pseudo-bridge' site.Ni( 100x2 x 2pC and -N; Adsorbate-induced Substrate Distortion By contrast to the essentially simple overlayer structure formed by oxygen on Ni(100) at 0.5 ML coverage, the same coverage of carbon or nitrogen, adjacent in the periodic table, appears to lead to significant substrate disruption. In particular, the new surface net is (2x2) indicating two (inequivalent) adsorbate atoms per unit mesh, although characteristic missing diffraction beams in LEED show the structure has a p4g space group, indicating that the two inequivalent sites are actually related by a glide symmetry operation. The structure proposed on this basis" is shown in fig. 3 and involves alternate clockwise and anticlockwise rotations of the groups of four nearest-neighbour atoms around the hollow sites of the surface, coupled with slight movement of these atoms away from the hollow site about which they have rotated.This distortion leads to an enlargement of half of the surface hollow sites, while the other half show a reduction in size and a change from four-fold to two-fold symmetry and coordination, producing a 'diagonal bridge' site. The two adsorption sites for C and N consistent with the p4g symmetry are the true (enlarged) hollow and the diagonal bridge position, but LEED," surface vibrational studies'' and SEXAFS20*21 all favour the hollow site, although a recent reconsideration of vibrational data favours the diagonal bridge site." As may be seen from fig.3, there are three significant structural parameters to be determined in this case. These are the spacing of the adsorbate atom plane above the metal surface (again described by z, the distance above the ideally situated top metal layer), the displacement of the top Ni atoms parallel to the surface (Sxy) and the316 0, N and C Adsorption Structures on Ni(100) and Cu(ll0) 100 200 300 400 z = 0.25 8, 6xy= +z5x -0.4 0.3 1 ' 1 ' 1 ' 1 100 200 300 400 photoelectron energy/eV Fig. 4. Comparison of experimental normal emission PD spectra recorded by C 1s emission from Ni(100)(2x2)-C and by N 1s emission from Ni(100)(2x2)-N with the results of theoretical simulations for different structures. In all cases the adsorbate is located above the four-fold symmetric hollows of fig.3 at a layer spacing relative to the position occupied by an ideally located top Ni layer of 0.25 A. The parameters Sxy and Sz are the amplitudes of the Ni atom displacements in the top layer parallel and perpendicular to the surface (see fig. 3 ) . expansion of the top metal layer spacing ( 6 z ) . Note that if there is significant parallel movement of the top Ni atom layer, we expect surface layer expansion simply on the basis of a hard-sphere model in which these surface atoms 'ride up' on the layer below. Normal emission PD spectra from both C and N 1s signals show, as expected, a strong sensitivity to the separation of the adsorbate atom and the second layer Ni atom directly below, favouring a value of z of 0.25 f 0.05 8, (i.e. indicating a value for this separation of the bulk layer spacing plus this value of z ) .This value, coupled with reasonable estimates of the Ni-C or Ni-N nearest-neighbour spacings (also determined by SEXAFS), clearly rules out the diagonal bridge site which would necessitate the adsorbed atoms lying l.OA or more above the top Ni layer. Some indication of the sensitivity of the normal emission data to the other structural parameters is shown in fig. 4, which compares both experimental spectra with the results of calculations in which z is fixed at 0.25 A but S z is varied for two different values of Sxy. Note that the N and C 1s spectra are essentially identical, indicating that the adsorption sites are the same. Fig. 4 shows significant improvement in the quality of the fit between experiment and theory if Sxy is increased from zero to 0.35 8, and the optimum associated value of Sz is in the range 0.1-0.2 8,. Broadly similar results emerge from an analysis of off-normal emission data collected in the (1 10) azimuth.These results are therefore somewhat similar to those of Ni( 100)c(2 x 2 ) - 0 in display- ing a strong sensitivity to the adsorbate - second metal layer spacing, but a much weaker sensitivity to the location of the top-layer metal atoms. In the case of the C and NA. L. D. Kilcoyne et al. 317 structures, the fit of theory and experiment without surface-layer atom movements is much worse than for the 0 structure, providing clearer evidence for substrate disruption. On the other hand, the need to investigate two different significant structural parameters for the top-layer distortion induced by C and N means that the precision with which these parameters can be determined is not high.We should, however, note that separation of these structural variables is not easy in other techniques either. In the case of Ni( 100)c(2 x 2)-0 we noted that a suitably chosen PD experiment (at some 60" emission in (100)) should allow the adsorbate-top-layer geometry to be determined more effectively. In the case of the C and N structures, however, this is likely to prove less successful for two reasons. First, the fact that the top adsorbate and metal atom layers are almost coplanar means that the preferred 180" scattering angle condition with one of the nearest-neighbour Ni atoms 'behind' the emitter, is almost impossible to achieve.In addition, however, the rotational distortion of the top metal layer means that the Ni-C and Ni-N nearest-neighbour bonds do not lie in the (100) azimuth, but out of this azimuth in two different directions. This means that even if the experiment could be performed, we would average over two different scattering paths, and thus obtain much less precise information. In summary, therefore, we see that PD measurements have a given definitive adsor- bate-underlying substrate spacing which clarifies the bonding site, and we find clear evidence for the overlayer distortion and can estimate its amplitude. On the other hand, we see that the technique is not well-suited to precise determination of top-layer coordination, particularly in a system of such high complexity, if the adsorbate is close to being coplanar with this top layer.We should, incidentally, note that the coplanar geometry is precisely the one that favours the application of another version of the PD phenomenon, based on forward-scattering effects in grazing angle XPS,23 although the specific problem of the N and C Ni( 100) distortions may remain troublesome because of the azimuthal averaging effect described above. Cu( 110x2 x 1)-0; Major Adsorbate-induced Substrate Reconstruction? The 0.5 ML structure formed by oxygen on Cu(ll0) is rather more complex than on Ni(100) in that there is ample evidence that the top metal layer undergoes a major rearrangement (reconstruction). The favoured structural model for this surface is the 'missing-row' model in which alternate (100) Cu atom rows from the top layer are removed, whilst the adsorbed 0 atoms occupy 'long-bridge' sites on the remaining (100) Cu atom rows (fig.5) [see, for example, ref. (24) and references therein]. SEXAFS indicate that 0 atoms lie 0.3 A above these top-layer Cu atoms, with the metal layer spacings unchanged from bulk values, although rather large surface-layer expansions have been proposed from some ion scattering Alternative pro- posals for the reconstruction are the 'saw-tooth' model, involving additional removal of alternate third layer metal atoms [a model favoured by a SEXAFS study of the closely related Ni( 110)(2 x 1)-0 str~cture'~], and a surface rumpling model [ e . g . ref. (29), sometimes referred to as a shifted row model] in which no row removal is required, but the layer spacing of alternate (100) Cu atom rows differs significantly.Note, incidentally, that this model is inconsistent with SEXAFS data. The main structural parameters of interest in investigating this structure are therefore the nature of the reconstruction (missing row, saw-tooth or rumpling), the Cu-0 top-layer spacing and the Cu-Cu top-layer spacing. PD spectra from this system were recorded in the (1 10) azimuth at several emission angles ( O O , 39", 57" and 67"). Comparison with calculations shows clearly that the 0 atoms do occupy the long-bridge site and if the substrate layer spacings are ynchanged from the bulk values, the spacing of the 0 above the top Cu layer is 0.23 A. Normal emission calculations for several different 0 atom heights z above a Cu(ll0) surface318 0, N and C Adsorption Structures on Ni(100) and Cu(ll0) side Fig.5. Schematic diagram of the proposed (missing row) structure of Cu( 110)(2 x 2 ) - 0 . with missing rows but bulk layer spacings are shown, together with the experimental spectrum, in fig. 6. Fig. 7 shows the effect of changing the top Cu-Cu layer spacing in terms of the expansion of this spacing, 6z, keeping the 0 emitter site fixed relative to the underlying bulk. These two figures show rather clearly that whilst the fit of theory and experiment are very sensitive to the height of 0 atom relative to the underlying bulk, they show essentially no dependence on the relative position of the top layer Cu atoms.This effect is also seen in the off-normal spectra. As in the Ni(100) case, the dominant scatterers are the pair of ('short-bridge') second layer Cu atoms lying below the adsorbed 0 atoms. At 39" emission one of these atoms lies very close to the 180" scattering condition; the PD spectrum looks particularly simple at this angle, being dominated by a single scattering path and the theoretical fit is particularly good. Evidently the insensitivity of the spectra to the location of the nearest-neighbour Cu atoms in the top layer suggests that distinctions between the different reconstruction models, which involve changes in the second nearest-neighbour Cu atom sites in the top layer, should be weak. Calculations confirm this, showing no significant differences between spectra calculated for the different models.The results therefore give a clear atom location for the adsorbed 0 atoms relative to the bulk, a location consistent with the SEXAFS results (which are more sensitive to the location of the nearest-neighbour atoms in the top Cu layer), but add no information regarding the details of the substrate reconstruction. The extreme insensitivity of the PD measurements to the top-layer Cu atom locations in this case [which seems to be much more serious than in any of the Ni( 100) systems described above] appears to be due to a combination of two factors. One of these is the fact that the structure evidently involves adsorbate atom locations which are quite close to being coplanar to the top metal layer, leading to the difficulty of choosing a collection geometry which properly enhances the role of top layer metal atom scatterers as discussed in the previous section.For example, if the substrate layer spacings really are unchanged from the bulk, one of the nearest neighbours would be in the optimum geometry at 83" emission in the (000) azimuth, but if the top metal layer is expanded, this angle would be even more grazing. A second reason, however, arises from the linearly polarised nature of the incident synchrotron radiaticn, and thus from the azimuth of incidence. Photoemission from a 1s initial state leads to an outgoing photoelectronA. L. D. Kilcoyne et al. 319 cu (110) ( 2 x 1 ) - 0 missing row 6 2 = 0.0 8, long bridge site .37 .27 I I I I I I I 1 0 0 200 300 400 photoelectron energy/eV I Fig.6. Comparison of experimental normal emission 0 1s PD spectra from Cu(110)(2~2)-0 with the results of theoretical calculations based on a model in which oxygen atoms occupy the long-bridge site (fig. 5 ) at different distances z above a surface represented by the missing-row model having a top metal layer spacing equal to that of the bulk. p-wave (change of angular momentum quantum number of unity), and this p-wave is directed along the A-ve$or of the incident radiation. In the extreme case of a Cu-0 top-layer spacing of 0.0 A, therefore, incidence in the (1 10) azimuth leads to an A-vector which is perpendicular to the nearest-neighbour Cu atoms, so that in this plane-wave- limited description these Cu atoms receive no photoelectron illumination, and so contribute zero to the PD signal.The fact that the 0 and top layer Cu atoms are nearly coplanar, coupled with the use of the (1 10) azimuth of incidence on this low-symmetry substrate (two-fold rotational) therefore produces this particularly poor sensitivity to near-neighbour locations. Conclusions The application of scanned energy mode photoelectron diffraction to the half-monolayer structures formed by 0 on Ni( 100) and on C u ( 110) and by C and N on Ni( 100) reinforces the view that scattering paths involving near 180” scattering are strongly represented in the data, making it relatively easy to determine the positions of atoms ‘behind’ the emitter (relative to the collection direction). The fact that these small atoms adopt320 0, N and C Adsorption Structures on Ni(100) and Cu(ll0) x Y .I 2 Y c .- c p.c u (110) ( 2 x 11-0 missing row, z = 0 . 2 3 1 long bridge site f . . - A 100 200 30 0 40 0 photoelectron energy/eV Fig. 7. Theoretical calculations of normal emission 0 1s PD spectra from Cu(110)(2x2)-0 for different Cu-Cu top layer spacings assuming oxygen atoms in the long-bridge site at a fixed distance above the metal substrate (corresponding to 0.23 A above the top layer in the case of a top metal layer spacing equal to that of the bulk). surface sites of reasonably high (two-fold or four-fold) coordination means that normal emission probes mainly adsorbate-second layer substrate spacings. Moreover, the short adsorbate-substrate bond lengths associated with these species typically lead to rather small adsorbate-top layer substrate spacings, making it difficult to choose experimental collection geometries which strongly favour scattering from top layer substrate atoms.Nevertheless, the role of these top layer scatterers is usually sufficient to allow reasonably accurate top-layer structural parameters to be determined, albeit with less precision than those associated with deeper layers. A special case arises, however, in the case of surfaces having less than three-fold rotational symmetry on which (as in SEXAFS) it is possible to select incidence and collection geometries which totally exclude the scattering influence of some near-neighbour atoms using plane polarised incident (sych- rotron) radiation and emission from an initial s-state. This special situation could, on the other hand, be turned to advantage to optimise the influence of specific (selected) structural parameters by the proper choice of experimental geometry.The Warwick and Berlin groups are pleased to acknowledge the financial support of the Science and Engineering Research Council (UK) and the German Federal Ministry of Research and Technology, respectively, and of the European Community SCIENCE programme, in the form of research grants.A. L. D. Kilcoyne et al. 321 References 1 M. A. Van Hove and S . Y. Tong, Surface Crystallography by LEED, (Springer-Verlag, Berlin, 1979). 2 S. Y. Tong, W. M. Kang, D. H. Rosenblatt, J. G. Tobin and D. A. Shirley, Phys. Rev. B, 1983, 27, 4632. 3 D. P. Woodruff, Vacuum, 1989, 39, 621. 4 J . J.Barton, S. W. Robey and D. A. Shirley, Phys. Rev. B, 1986, 34, 778. 5 C. F. McConville, D. P. Woodruff, K. C. Prince, G. Paolucci, V. Chab, M. Surman and A. M. Bradshaw, 6 M. D. Crapper, C. E. Riley, P. J. J. Sweeney, C . F. McConville, D. P. Woodruff and R. G . Jones, Surf 7 D. P. Woodruff, C. F. McConville, A. L. D. Kilcoyne, Th. Lindner, J. Somers, M. Surman, G. Paolucci 8 Th. Lindner, J. Somers, A. M. Bradshaw, A. L. D. Kilcoyne and D. P. Woodruff, Surf: Sci., 1988, 203, 9 E. Dietz, W. Braun, A. M. Bradshaw and R. L. Johnson, Nucl. Instrum. Methods A, 1985, 239, 359. Surf Sci., 1986, 166, 221. Sci. 1987, 182, 213. and A, M. Bradshaw, Surf Sci., 1988, 201, 228. 333. 10 A. L. D. Kilcoyne, D. P. Woodruff, J. E. Rowe and R. H. Gaylord, Phys. Rev. B, 1989, 39, 12604. 11 J. H. Onuferko, D. P. Woodruffff and B. W. Holland, SurJ: Sci., 1979, 87, 357. 12 C. F. McConville, D. P. Woodruff, S. D. Kevan, M. Weinert and J. W. Davenport, Phys. Rev. B, 1986, 34, 2199. 13 K. C. Prince, M. Surman, Th. Lindner and A. M. Bradshaw, Solid State Commun., 1986, 59, 71. 14 J. E. Demuth, D. W. Jepsen and P. M. Marcus, Phys. Rev. Lett., 1970, 31, 540. 15 M. A. Van Hove and S. Y. Tong, J. Vac. Sci. Technol., 1975, 12, 230. 16 S. R. Chubb, P. M. Marcus, K. Heinz and K. Muller, Phys. Rev. B, in press. 17 J. E. Demuth, N. J. DiNardo and G. S. Cargill 11, Phys. Rev. Lett., 1973, 50, 1373. 18 I . Stensgaard, Vacuum, in press. 19 M. Rocca, S. Lehwald, H. lbach and T. S. Rahman, Phys. Rev. B, 1987, 35, 9510. 20 M. Bader, C. Ocal, B. Hillert, J. Haase and A. M. Bradshaw, Phys. Rev. B. 1987, 35, 5900. 21 L. Wenzel, D. Arvanitis, W. Daum, H. H. Rotermund, J. Stohr, K. Baberschke and H. J. Bach, PhyF. Rev. B, 1987, 36, 7689. 22. J. Szeftel, F. Mila and A. Khater, Surf: Sci., 1989, 216, 125. 23 C. S. Fadley, Prog. Surf: Sci., 1984, 16, 275. 24. M. Bader, A. Puschmann, C. Ocal and J. Hease, Phys. Reu. Lett., 1986, 57, 3273. 25. E. van d e Riet, J. B. J. Smeets, J. M. Fluit and A. Niehaus, Surf: Sci., 1989, 214, 111. 26 J . A. Yarmoff, D. M. Cyr, J. H . Huang, S. Kim and R. S . Williams, Phys. Rev. B, 1986, 33, 3856. 27 H. Niehus and G. Comsa, Surf: Sci., 1984. 140, 18. 28 K. Baberschke, U. Dobler, L. Wenzel, D. Arvanitis, A. Baratoff and K. H. Rieder, Phys. Ret;. B, 1986, 33, 5910. 29 F. M. Chua, Y. Kuk and P. J. Silverman, Phys. Rev. Lett., 1989, 63, 386. Paper 9/05424C; Received 19th December, 1989

ISSN:0301-7249    

DOI:10.1039/DC9908900311

出版商:RSC    

年代:1990

数据来源: RSC

摘要:

Faraday Discuss. Chem. SOC., 1990, 89, 323-328 Study of Xenon Layers on a Cu(ll1) Surface Jacques Jupille," Jean-Jacques Ehrhardt, Daniele Fargues and Albert Cassuto CNRS Laboratoire Maurice Letort, BP 104, 54600 Villers-Les- Nancy, France The adsorption of xenon has been studied by SEXAFS on a Cu( 11 1) surface in the monolayer and in the multilayer range. At temperatures ranging between 18 and 47 K, the Xe monolayer has been shown to be incommensur- ate because (i) the Xe-Xe spacing differs from the distance between easy three-fold-hollow sites (a,,J3) and (ii) the coefficient of thermal expansion of the Xe adlayer is much higher than that of copper. In addition, we have found an Xe-Cu distance of 0.345 nm, which compares well with the value determined by a hard-sphere model.The behaviour of xenon overlayers adsorbed on metallic surfaces has been shown to depend very much on the substrate concerned. The two systems which have been studied in depth to date, namely Xe/Ag( 11 1)' and Xe/Pt( 11 1),2 show some substantial differen- ces. In both cases, the attractive adatom-adatom interactions induce island formation [however, these interactions are believed to be repulsive on the Pd(100) surface as no ordered structure is observed at a temperature as low as 10 K3]. A key role is played by the lateral modulations of the holding potential. They are quite effective for xenon adsorption on the P t ( l 1 l ) surface, which gives rise to several commensurate and incommensurate solid phases depending on temperature and coverage.- In contrast, these modulations have been assumed to be of marginal importance for the case of an Xe layer on an Ag(ll1) surface, since their only effect is possibly' to align to within a few degrees the commensurate Xe layer with the underlaying hexagonal structure (this has alternatively been suspected to come from extrinsic effects such as surface defects').If one considers the relative strength of the Xe-Xe interactions and the Xe substrate holding potential, the Cu( 11 1) surface offers an interesting opportunity. On this sub- strate, the Xe layer has been shown to form a commensurate ( J 3 x J 3 ) R30" structure at 77 K.4 This is consistent with the very small mismatch existing at this temperature between the Xe-Xe spacing in bulk xenon (0.4388 nm"') and the distance between the three-fold hollow sites separated by 43 times the Cu-Cu spacing (ac-UJ3 = 0.4414 nmi), which are the probable easy sites (the distance between these particular sites will hereinafter simply be referred to as 'the distance between three-fold hollow sites').However, at temperatures lower than 77 K, a greater mismatch is expected because, if the parameter ac.,J3 is virtually unchanged (0.4413 nm at 25 K'), the minimum Xe-Xe distance in bulk xenon decreases to below 0.434 nrn5*' at the 0 K limit. In these conditions, two distinct behaviours can be expected, depending on whether or not the xenon overlayer remains commensurateX: either the Xe-Xe spacing remains equal to the distance between two three-fold hollow sites of the Cu( 11 1 ) surface and the layer then remains commensurate, or it is governed by its own thermal expansion law and the layer becomes incommensurate (this will ensure that neither commensurate nor high- order incommensurate structuresX are formed).The purpose of this paper is to investigate which of these cases is realised. Measurements have been done by surface extended X-ray absorption fine structure (SEXAFS), this technique being used for the first time Distances between three-fold hollow sites on the Cu( 1 1 1 ) surface were determined using ( i ) a lattice parameter of 0.361 SO nm at 298 K'" and ( i i ) the coeficients of linear thermal expansion."' 323324 SEXAFS on Xe/Cu( 1 1 1 ) I ’ I I I I I I I h Y m .I c 3 i” L3 3.1 L2 i I I I I I I I I I 4800 4900 5000 5100 hvleV i Fig. 1.SEXAFS from an Xe multilayer at 20 K. to study rare gas adsorption on a monocrystalline substrate. The Xe-Cu distance is also determined and discussed. Experimental The L,-edge SEXAFS data have been collected at the Laboratoire pour 1’Utilisation du Rayonnement Electromagnktique (LURE) at Orsay, on the DCI storage ring, with a Si(311) or a Ge( 1 1 1) double-crystal monochromator, thus favouring either resolution or sensitivity, respectively. The Cu( 1 1 1 ) crystal was mounted in an ultra-high-vacuum (UHV) chamber.’ It could be cooled down to 18 K by circulating liquid helium in the sample holder. Temperatures were measured with a calibrated platinum resistor. The sample was treated by repeated argon-ion bombardment followed by annealings at 900 K until it showed a clean surface by Auger electron spectroscopy and a ( 1 x 1 ) pattern with sharp spots by low-energy electron diffraction (LEED).SEXAFS measurements were made using the total-electron-yield method. The SEXAFS signal was recorded at the L,-edge of xenon (4782eV) with the vector polarisation of the light parallel to the surface sample. In this way the SEXAFS oscillations are dominated by the in-plane coordination shells. Resu 1 ts To determine experimental phase shifts and amplitudes we have first examined the EXAFS from multilayers (fig. 1 ) which were deposited on the substrate between 20 and 35 K. As judged from the exposure, the deposits were 15-75 layers thick. The height of the Xe L,-edge jump reached a limit which was ca. 10 times that of the monolayer when the deposit was ca.15-20 layers thick, as estimated from the exposure. A 10% contribution of the first layer to the overall signal recorded from the semi-infinite solid corresponds to an escape depth of ca. 3.5-4 nm; in these conditions, 20 layers of xenon (ca. 7 nm) would give rise to a signal ca. eight times bigger than that of the monolayer, which is precisely what we have observed. The behaviour of the edge jump is thus consistent with a layer-by-layer growth up to 15-20 layers, and we have supposed that a thicker deposit would grow in the same way. EXAFS data were also collected above the temperature of deposit, up to 47 K. It was checked, from the height of the L3 edge, that no major desorption occurred during both the annealing and the data acquisitionJ.Jupille et al. 0.06 - 0.00- -3 -0.06 325 - - 4800 4900 E,IeV aJ '0 E on +-a .I E 0.8 0.6 0 . 4 0.2 0 .o 1 2 3 4 5 6 7 R I A I I I 1 1 1 2 3 4 5 6 k / k ' Fig. 2. SEXAFS from an Xe monolayer on a Cu(l11) surface at 30 K using an Si( 111) mono- chromator: ( a ) SEXAFS data; ( b ) Fourier transform of the normalized kx( k ) signal, experimental (-) and calculated ( * * * .); ( c ) inverse Fourier transform of Fourier-transformed data after multiplication by a window function including the Xe-Xe and the Xe-Cu peaks. time. The accessible data range above the L3 edge (4782 eV)? is limited by the occurrence of the L2 threshold (around 5100 eV, see fig. 1 ) . In any case, very little information can be collected beyond 200 eV because of the huge Debye-Waller damping of the SEXAFS oscillations related to the xenon layer.Fortunately, these limitations are somewhat compensated by the large size of the xenon atom, which shortens the period of SEXAFS oscillations, thus making the analysis tractable. Xenon monolayers have been prepared in two different ways: (i) by adsorption of Xe at 60-65 K under a pressure of (2-3) x Pa, i.e. far from the conditions at which the three-dimensional condensation of xenon takes place and within a temperature range where little desorption occurs from the Xe monolayer adsorbed on the Pt( 11 1 ) surface, even in UHV conditions; ( i i ) by adsorption of a multilayer followed by annealing to + 1 eV= 1.602 18x lo-'" J.3 26 SEXAFS on Xe/Cu( 11 1) 0.453 0.42 1 1 I I 10 20 30 40 50 T l K Fig.3. SEXAFS determination of Xe-Xe spacings as a function of temperature in three- dimensional xenon (stars and dotted line); also indicated are the Xe-Xe spacings as determined by X-ray measurement^^.^ (thin line). 60-65 K and cooling to the working temperature. An Xe (d3 x d3) R30" LEED pattern could be seen for more than 2 h when a monolayer was prepared in such conditions and kept in UHV. This corresponds to the longest collection time we needed to study monolayer coverage, i.e. when using an Si(3 11) double-crystal monochromator. The SEXAFS spectrum recorded from an Xe monolayer at 30 K is shown in fig. 2(a). It is worth noting that the data shown herein have been collected through three different beam times, thus increasing the confidence of their reproducibility. Discussion SEXAFS calculations were done using plane waves.For the case of the multilayer we have used theoretical phase shifts and amplitudes." The coordination number was fixed to 12. Values found in this way for the Xe-Xe spacing were very close to those from X-ray mea~urernents.~~~ However, SEXAFS values at different temperatures were correc- ted to bring their centre onto the X-ray curve. This resulted in a decrease of only 0.0007 nm in the calculated values of the Xe-Xe spacing. The variation of this spacing versus temperature is seen in fig. 3. The error bar corresponds to the scattering of the results when varying the parameters involved in the calculation. Our results agree with the X-ray measurements within experimental error (however, one can notice that EXAFS data show a possibly higher thermal expansion than the reference data; we do not explain this trend since the expectation for materials having a large amount of thermal disorder is a shrinkage of the actual thermal expansion"). The purpose of the examination of the multilayer was mainly to determine experimental phase shifts and amplitudes.In fact, that study has proved that the theoretical values of these parameters1' were fairly accurate. This may be connected with the fact that, in the case of a rare gas, one obviously avoids possible variations of phase shifts due to chemical changes (amplitudes are much less sensitive to chemical environment'"). In the forthcoming calculation, both kinds of parameters (theoretical and experimental) were used and the two sets of results did not show any significant difference (the correction of 0.0007 nm was done as in the case of the multilayer).The Fourier transform of the normalized k x ( k ) signal and the inverse Fourier transform of the Xe-Xe first shell are shown in fig. 2, the experimental spectrum being collected by means of an Si( 3 1 I ) double-crystal monochromator to obtain good resol- ution. As the Xe-Cu bond is not exactly normal to surface it contributes to some extent to the SEXAFS spectrum, despite the fact that the vector polarisation of the lightJ. Jupille et al. 2 ... C e, 0.44- * v) .- 'c3 2 0.43- X I 327 0.45 I 0 0.42 ! I I I I I 10 20 30 40 50 Fig. 4. SEXAFS determination of Xe-Xe spacings as a function of temperature and in a xenon monolayer (-0-) deposited on a Cu(ll1) surface; also indicated are the Xe-Xe spacings in bulk xenon as determined by SEXAFS measurements (.- * a , see fig. 1) and distances between three-fold-hollow sites on the Cu( 11 1) surface (a,,d3) (see footnote t ) Table 1. Xe-substrate distances (nm) surface measurement hard-sphere model reference Xe/Ag( 11 1) 0.35 f 0.01 0.361 Cohen et a/.' Xe/Cu( 11 1) 0.345 f 0.01 0.345 this work LEED SEXAFS is parallel to the surface. In these conditions, the Fourier transform of the kx( k ) signal shows two well separated peaks, above 0.4 nm and slightly below 0.3 nm, which are due to the Xe-Xe and the Xe-Cu (first-neighbour distances), respectively. These two peaks have been treated by means of theoretical phase shifts and amplitudes (as seen above this is a fairly good approximation for Xe scatterers), either separately or together, by introducing two types of scatterers (Xe and Cu) in the neighbourhood of the central Xe atom.The result of such calculation including the two peaks is shown by the dotted line in fig. 2(6). For the case of the Xe-Cu distance, the data could be fitted only by means of a large negative Eo value (ca. -12 eV) accounting for the internal potential of the copper substrate. In contrast, this value was only 2 or 3 eV when the Xe-Xe bond was considered. This can be qualitatively understood by the fact that the xenon levels are referenced to the vacuum level. The photoadsorption of L3 edge involves final states of both s and d symmetry," but, as data were collected with in-plane polarisation, the s-d cross term is zero and the coordination number amounts to a 7.5'' for the Xe-Xe interaction and 2.25 for the Xe-Cu interaction.Xe-Xe spacings measured for the case of the monolayer at temperatures ranging between 18 and 47 K are shown in fig. 4. They clearly characterise an incommensurate structure. First, Xe-Xe spacings are shorter than the distances between three-fold hollow sites of the Cu( 11 1) surface, which are also represented in fig. 4. Moreover, the expansion coefficient is much larger than that of the substrate. It means that even the occurrence of a high-order commensur- ate phase' can be discarded. As on the Ag( 11 1) surface, one can suggest that the lateral modulations of the holding potential are negligible with respect to the xenon-xenon interaction since in the temperature range under discussion (18-27 K) a very small mismatch can be preserved between the Xe overlayer and the copper substrate.However, the thermal expansion of the Xe layer varies in such a way that our observations do328 SEXAFS on Xe/Cu( 1 1 1 ) not contradict the finding that this layer could form a (J3 x J3) R30" commensurate structure at higher temperature and especially at 77 K.4 An average value of 0.345*0.01 nm was found for the Xe-Cu distance in the temperature range of interest. This measurement compares nicely with the value deduced from a simple hard-sphere model, the radius of the Xe and Cu atoms being 0.1275 and 0.2175 nm, respectively. A rather similar conclusion has been reached concerning the Xe monolayer on the Ag( 1 1 1 ) surface, as studied by LEED.' The results are summarised in table 1 .Conclusion The first SEXAFS study of a rare-gas monolayer has been performed on the Xe/Cu( 1 1 1 ) system. As evidenced by both the Xe-Xe distance and the thermal expansion of the adsorbate layer with respect to that of the substrate, xenon has been shown to form an incommensurate structure on the Cu( 1 1 1 ) surface. The average Xe-Cu distance was found to amount to 0.345 f 0.01 nm, in good agreement with a hard-sphere model. The authors are very grateful to D. Chandesris and J. Lecante (LURE), C. Brouder and G. Krill (Universite Nancy I) for helpful assistance and discussions about the treatment of the data. References 1 P. I . Cohen, J. Unguris and M. B. Webb, Surj Sci., 1976, 58, 429. 2 K. Kern, R. David, P. Zeppenfeld and G. Comsa, Surf: Sci., 1988, 195, 353. 3 E. R. Moog and M. B. Webb, Surj: Sci., 1984, 148, 338. 4 M. A. Chesters, M. Hussain and J . Pritchard, Surf: Sci., 1973, 35, 161. 5 G. L. Pollack, Rev. Mod. Phys., 1964, 36, 748. 6 M. L. Klein and J . A. Venables, Rare Gas Solids (Academic Press, London, 1977), vol. 11. 7 ( a ) Handbook of Chemistry and Physics (CRC Press, Boca Raton, Florida, 69th edn, 1988-1989); ( b ) 8 K. Kern, P. Zeppenfeld, R. David and G . Comsa, Phys. Rev. Lett., 1987, 59, 79. 9 P. Roubin, D. Chandesris, G. Rossi, J. Lecante, M. C. Desjonqukres and G . TrCglia, Phys. Rev. Lett., American Institute Handbook (McGraw-Hill, New York, 2nd edn, 1963). 1986, 56, 1272. 10 B-K. Teo and P. A. Lee, J. Am. Chem. Soc., 1979, 101, 2815. 1 1 P. A. Lee, P. H. Citrin, P. Eisenberger and B. M. Kincaid, Rev. Mod. Phys., 1981, 53, 769. 12 P. H. Citrin, fhys. Rev. B, 1985, 31, 700. Paper 91055161; Received 22nd December, 1989

ISSN:0301-7249    

DOI:10.1039/DC9908900323

出版商:RSC    

年代:1990

数据来源: RSC

摘要:

Faraday Discuss. Chem. SOC., 1990, 89, 329-339 GENERAL DISCUSSION Dr K. C. Prince (Sincrotrone Triesfe, Trieste) began the discussion of Prof. King’s paper: There is an alternative way of interpreting the present data which does not cause any conflict. Specifically, the Born-Haber cycle pioneered by Johansson and Martensson’ provides a way of using the photoemission data to obtain useful thermody- namic information. It has the advantage that final-state effects are always included. In the present case these can be neglected because screening effects are roughly constant in the middle of a series of d metals. However, at the end of a series there is a change from screening by the filling of an (initial-state) d hole to sp hole filling. This is much less effective, and strongly influences the surface core-level shift (SCLS).Any initial state theory of SCLS therefore becomes more complicated at the end of a d series, where the charging screening contribution must be included. To test how well the Born-Haber cycle works, we measured the following values of adsorbate-induced SCLS,’ and compared them with predicted values. The calculated values are based on published ground-state thermodynamic data, such as enthalpies of adsorption. adsorbate SCLS/eV calculated SCLS/eV ~~ H -0.18 0 2 -0.15 co - 1.26 -0.17 -0.17 <-1.3 The agreement is good, indicating that the method is accurate. Note that the SCLS for CO is quite large, much larger than on W. This does not mean that large charge transfer occurs for CO/Pt; it is generally accepted that the bonding is covalent.This non- correlation between the magnitude of the SCLS and charge transfer supports the previous speaker’s conclusions. Applied to the present data, the Born- Haber cycle provides useful thermodynamic information. For instance, it can be concluded that the adsorption energy of all the adsorbates is similar on W and Re because the SCLS is small, and in particular, the adsorption energy of hydrogen on Re is 2.8 eV. Under certain conditions this approach can be extended to the case of semiconductor SCLS. 1 B. Johansson and N. Martensson, Pbps. Rev. B, 1980, 21, 4427 2 K. Duckers et al., Phys. Rev. B, 1987, 36, 6592. Prof. D. A. King (Uniuersify ofCarnbridge) replied: In our previous publications we have made good use of the Born-Haber cycle approach. However, it is a thermodynamic approach, and while it has distinct advantages it does not contain a microscopic picture of the process.We have attempted to outline a reasonably complete microscopic approach to the origin of the adsorbate-induced shifts. Dr J. Jupille (Laboratoire Maurice Letort, Villiers-les- Nancy) said: The core-level binding-energy shift induced by an adsorbate is described as a sum of components arising from both changes in environment and changes in electronic configuration. On this basis, the question of the very small amplitude of the core-level shifts induced on surface atoms by the adsorption of alkali metals is examined. The conclusion that such 329330 General Discussion small shifts provide evidence for the adsorbate-substrate charge transfer to be small is discarded.Moreover, the observed shift is said to result from the cancellation of two effects, one of these being a large electron flow between the adsorbate and the substrate (creation of an ionic bond). In the above sum, two terms are expected to be negative [the initial state (Ael) and the final state (Arelax) contributions due to the charge transfer] meanwhile one (Aenv) is expected to be positive. The value of the sum is the only available parameter, hence the conclusion that one of these terms (Ael) has a strong amplitude is not unambiguous. (Furthermore, the covalent picture of the alkali-metal adatom' predicts a strong polarisa- tion of this adatom, i.e. possible negative and A,, and Arelax contributions.) 1 E.Wimmer er al., Phvs. Rev. Left., 1982, 48, 1128; H. Ishida, Phjx Reu. B, 1988, 38, 8006. Prof. King replied: The only point where I really cannot agree with you is in your last, parenthetic, sentence. Wimmer and co-workers refer to a strongly polarised bond between surface metal atom and Cs adatom, and not to a polarisation of the adatom itself. In the case of rare-gas adsorption, e.g. xenon, we may talk sensibly about polarisation of the atom, but with Cs there is no disputing that a chemical bond is formed, with charge redistribution weighted away from the adatom. Prof. A. M. Bradshaw ( Fritz-Haber-Institut der MPG, Berlin) said: Although Professor King's explanation of the small shifts observed for alkali metals on W(110) is very convincing, I wonder in view of the four distinct contributions to the surface core-level shift whether it is wise to use this quantity at all as a probe of ground-state electronic structure? Prof.King replied: My main point is that we should not be drawn into over-simplified interpretations of core-level shift data. In the end you may be right, but at this stage I think it is important to broaden the data base so that, in line with self-consistent calculations, we have some confidence in developing good microscopic models. It is too soon to be pessimistic. Dr P. L. Wincott and Dr G. Thornton (University of Munchester) communicated: The results of some studies we have made on Cu(O0l) and TiN(001) Tamm surface states may be of interest in connection with the paper of Derby and King. In particular, our results cast doubt on their explanation for a small W4f surface core-level shift (SCLS) accompanying Cs adsorption on W{ 1 lo}.This involves the cancellation of several effects, namely the expected shift to lower binding energy (&,) caused by donation of Cs electrons into the substrate, and the shift to higher Eb due to the change in surface atom coordination and/or possible final-state relaxation effects. Tamm-type surface states are similar to surface core-level states in that their origin lies in the potential shift caused solely by the presence of the surface.' However, Tamm states are narrow surface bands consisting of surface-plane-localised orbitals which have negligible orbital overlap in the surface-normal direction. The Cu(OO1) M Tamm state originates from planar-localised Cud,, orbitals and is split ca.0.2 eV off the bulk d,, band at M owing to the shift in the (001) surface potential.' This state is very sensitive to any perturbation of the potential in the surface plane and shows dispersion towards higher & and a rapid intensity loss following sub-monolayer adsorption of oxygen.' It is thus rather surprising that following adsorption of up to 0.5 monolayers of Cs on the Cu(OO1) surface the Tamm state shows only a 50 meV shift to higher Eh (in contrast to the lower & shift expected from an electropositive adsorbate) and no change in intensity relative to the bulk Cu d band (see fig. 1). The fact that the Cu{OOl) M Tamm state results are in line with those observed for the W(110) W4f SCLS does not rule out an interpretation based on cancellation ofGenera 1 Discussion 33 1 Cs 0 0.02 0.07 0.09 0.12 0.18 0.27 0.54 1 - I I 2.25 2.05 1.85 binding energy/eV Fig.1. Angle-resolved photoemission spectra of Cu(OO1) recorded off-normal in the I'xWK plane as a function of Cs coverage. Cu(OO1) + Cs, He I, 293 K. 55" incidence, 61.5" emission angle. opposing effects. However, it would seem rather too fortuitous given that the electronic structures of the substrates are very different. Moreover, exactly the same effect of Cs adsorption is seen on the TiN(001) T Tamm state,' which has an N p , , . basis. In the absence of accurate calculations, the model used by Riffe et al? would currently seem to offer a more consistent explanation for this behaviour.In this model potential shifts (and consequent changes in bonding electron density) are localised to the plane between the Cs adsorbate and the substrate. 1 P. L. Wincott, N . €3. Brookes, D. S-L. Law and ti. Thornton, fhj-s. Rec. B, 1986, 33, 4373; P. L. Wincott, I>. S-L. Law, N . B. Brookes, B. Pearce and ti. Thornton, Surf Sci., 1986, 178, 300. 2 P. L. Wincott, P. A. P. Lindberg and L. I . Johansson, .I. f h j x : Condensed Matter, 1989, 1, SB225. 3 D. M. Riffe, G. K . Wertheim and P. H. Citrin, fhjis. Ret?. Lett., 1990, 64, 571. Prof. King replied: There are several earlier studies of the kind referred to by Wincott and Thornton which are of even greater relevance to the work reported in our paper on W{ 110). In particular, Soukiassian et al.' examined the behaviour of the r surface state on W{lOO), Mo(100) and Ta(100) as a function of Cs coverage.On W(100) and Ta{ 100) they found that the surface state, which is ca. 0.3 eV below the Fermi level on the clean surface, is shifted to higher binding by ca. 1 eV, without attenuation. On Mo(100) a shift to higher binding energy by 0.9 eV was observed, but in this case the shift was accompanied by attenuation. These experimentally observed shifts were in close agreement with the theoretical predictions of Wimmer et al.,? and the effect was attributed to a 'strongly polarised covalent bond between the d-like surface states and the Cs 6s-derived valence states'.3 It is difficult to avoid a semantic discussion here, but I believe that the descriptions of the bonding as 'covalent with a strong degree of ionic character' and a 'strongly polarised covalent bond' are effectively synonymous.332 General Discussion My main point is that the interesting data referred to by Wincott and Thornton on Cu{lOO} are irrelevant to the current discussion.While the W{lOO} surface state is shifted to higher binding energy with increased coverage of Cs, the W4f,,, level is shifted by a small amount to lower binding energy. The former is a bonding shift between atomic orbital states. Finally, in their last sentence Wincott and Thornton attempt to characterise the model used by Riffe et al., but I feel that I should point out that we have no difficulty with the model described in this sentence: of course the charge is localised in the plane between Cs and metal surface.This has been demonstrated consistently by a range of theoretical models. But I do wish to stress that ( a ) this charge will be associated at least in part with particular eigenstates of the surface metal atoms, thus affecting core-level binding energies on these atoms, and ( b ) the same model is applicable to electronegative adsorbates, such as oxygen, except that there is a charge dejciency in the plane between surface metal atoms and the adatoms. Riffe et al., and Wincott and Thornton, are inconsistent in ignoring these two points. 1 P. Soukiassian, R. Riwan, C. Guillot, J . Lecante and Y. Borensztein, Phys. Scri., 1983, T4, 110. 2 E. Wimmer, A. J. Freeman, J. R. Hisks and A. M. Karo, Phvs. Rev. B, 1983, 28, 3074. 3 P. Soukiassian, R.Riwan, J . Lecante, E. Wimmer, S. R. Chubb and A. J . Freeman, Phjls. Ret.. B, 1985, 31, 4911. Dr G. C. King (Manchester Uniuersity) said: You described the various effects that lead to core-level energy shifts. Does the possible presence of vibrational structure pIay a role in the observed energy shifts? Prof. King replied: This is certainly an interesting point, but there is no evidence that I know of that vibrational structure is observed in XPS data from, for example, the C or 0 1s levels from adsorbed CO on metals. Presumably vibrational loss peaks occur, but with a very low cross-section, making them difficult to observe. Prof. Y. H. Chung (University of Liverpool) said: It seems odd to me to see that several experimental papers presented in this Faraday Discussion did not show their experimental errors.Any measurement performed in the physical world could not be free from its own errors. Professor King stated his instrumental resolution, which enables me to ask him a simple question: The instrumental resolution, as stated, is much larger than the Lorentzian broadening. How can you be convinced that your results are reliable? Prof. King replied: There is little doubt that our analyses would be more reliable if the resolution could be improved. The natural linewidth, including the Doniach-SunjiE and phonon broadening (the latter is not included in the figures in our paper) for the clean surface of W{ 110) is ca. 80-100 meV, and our instrumental broadening is 160 meV. Nevertheless, the peak separation for the clean surface is ca.300meV, and the two peaks are resolved. As clearly stated in the paper, we do not have the same confidence in the resolution of the component peaks in the case of the W{llO}/N spectrum, however (fig. 10). Prof. Y. H. Chung said: Would you explain the differences between your fitting parameters and Riffe's? Riffe et al.' reported the different values for the surface and bulk peaks, while you used the same values in your analyses. 1 D. M. Riffe, G. K . Wertheim and P. H. Citrin, Phys. Rev. Left., 1990, 63, 1976. Prof. King replied: With an instrumental resolution of 160 meV the Lorentzian broadening, estimated by Riffe et al. to be 60meV (bulk peak) and 84meV (surfaceGenera 1 Discussion 333 peak), is swamped, and use of these values rather than the 50 meV we have used would not alter our analyses in any significant way.(This point is made in our paper.) Prof. Bradshaw began the discussion of Dr Fontaine’s paper: I would like to use this opportunity to point out that the ‘building-block’ approach to the interpretation of near-edge X-ray absorption spectra of (adsorbed) molecules can lead to erroneous Whilst this appears not to be the case in Dr Fontaine’s paper, he does base his description of the near-edge structure (NEXAFS or XANES) and, in particular, of its polarisation dependence on such a model. A more comprehensive is based on the realisation that both the quasi-bound and continuum states probed in this experiment correspond to the unoccupied molecular orbitals of the system and belong to the irreducible representations of the appropriate point group.The selection rules and polarisation dependences then follow from standard group-theoretical techniques. Moreover, it is readily apparent that in polyatomic molecules the observation of several continuum resonances makes it impossible to correlate bond length with resonance energy. In the case of an adsorbed molecule the point-group symmetry will be lower than for the free species, although this may not necessarily affect the polarisation dependence of the continuum resonance^.^ An example of the need for this more complete description of the near-edge structure is provided by the orientation determination of adsorbed Langmuir-Blodgett chains. The polarisation dependence of the ‘u*(C-C)’ resonance here is not a superposition of cr* resonances polarised along the individual C-C bond directions.Rather, it contains two components polarised parallel and perpendicular, respectively, to the chain direction, corresponding to the symmetry of the unoccupied one-dimensional band-like states. (The band structure of such chains is actually quite complicated since the one-dimensional space group is non-symmorphic). Since Dr Fontaine does not use the u*(C-C) resonances to determine the chain orientation, this problem does not arise in his paper. More relevant is perhaps a second example, namely, the polarisation dependence of the resonance(s) of an adsorbed aromatic molecule. If the ring plane is parallel to the surface, it is reasonable to suppose that a strong interaction takes place, involving the T states.This leads to a lowering of the symmetry of the molecule, with the result that the T* resonance(s) may no longer be polarised strictly perpendicular to the surface. The ring plane may then appear to be inclined at an angle. The converse situation of an inclined geometry giving rise to a polarisation dependence symptomatic of the ring plane parallel to the surface is theoretically possible but unlikely. In summary, these remarks are intended to show that proper account of the molecular orbital picture and of symmetry aspects must be made in the application of X-ray absorption spectroscopy to the determination of molecular orientation at surfaces. 1 J. Stohr and D. A. Outka, Phys. Rev. B, 1987, 36, 7891. 2 J. Somers, A. W. Robinson, T.Lindner and A. M. Bradshaw, Phj*s. Rev. B, 1989, 40, 2053 3 A. M. Bradshaw and J. Somers, Phys. Scr., 1990, T31, 189. 4 M. Bader el al., Europhys. Lett., 1988, 5 , 443. Dr A. Fontaine (LURE, Orsay) replied: The question is related to the capability of using the n* resonance at the C K-edge, as a probe to determine the orientation of the heterocycles with respect to the platinum surface. We can find two ways to envisage this question. The first aspect concerns the accuracy of this experiment to distinguish between the situation where the cycle lies parallel to the surface and the other situation where the cycle is tilted with respect to the surface. In the latter case many such domains should exist and therefore yield an average direction of the cycle parallel to the surface.Polarisation-dependent spectroscopy can distinguish between the two situations through the magnitude of the variation of the intensity of the n* resonance uersus the angle of334 General Discussion the electric field of the light and the normal to the cycle. Experimentally it is difficult to reach the zero glancing angle (where the T* resonance disappears if the cycle is strictly parallel to the surface), therefore the result should always be given with error bars of the order of 5”. It is obvious to recall that in case of cycles tilted at the magic angle the intensity of the T* resonance is no more angle-dependent. The second way to look at this question deals with the perturbation of the T* state introduced by the metallic surface. The electrode-induced perturbation of the (T* resonance is well documented. One can ask if such an effect can perturb the r* resonance as well? First, we found the T* resonance locked at the same energy as the thickness of the deposited polymer layer increases from one to 15 layers.(If the cycles begin to be bounded to the metallic electrode a shift of the empty states to higher energy is expected and broadening should occur). Point two, the polarisation dependence of the intensity of the T* resonance is kept unchanged. Point three, the powder-like situation occurs for thicknesses >20 layers. Again the distinction discussed in the previous point can be tackled only by the angle dependence of the intensity of the T* resonance. Clearly the one-monolayer signal is less accurate than that found for thicker films.Nevertheless, there is no discontinuity of the signal behaviour with increasing thickness as long as we keep the thickness < 15 layers. Dr Thornton then turned to Dr A. W. Robinson: I have two questions. First, would you not expect to see a larger 6 ~ 1 resonance at the C K-edge rather than at the 0 K-edge given that Rodriguez and Campbell’s calculations [ref. (28)] point to a 92% C 2s,H 1s character for the 6a1 orbital? Secondly, could you comment on the fact that in transition- metal alkoxides the M-0-C bond is bent, whereas your results point to a collinear configuration (or more accurately a C-0 bond parallel to the surface normal). Dr A. W. Robinson (Fritz-Haber-Znstitut der MPG, Berlin) replied: In answer to your first question: The 6 ~ 1 orbital has indeed strong C 2s character [ref.(28)]. The important point to be made, however, is not a comparison of the intensities of the resonance between 0 K-edge and C K-edge NEXAFS, but of the relative strength of this resonance with respect to other resonances/features in the same NEXAFS spectrum. As to your second comment, both our ARUPS results and those of a recent FT-IRAS study [ref. (13)] indicate that the methoxy species possesses C3\. symmetry on the Cu{ 11 l} surface. We do not yet have an adsorption site determination for methoxy/Cu(lll) and so cannot comment on the M-0-C bond angle. It seems possible that both experiments are probing the effective symmetry of the CH30 species with respect to the surface, and that the actual position of the CH30 molecular unit relative to the surface is not important. Prof.J. P. Simons (University of Nottingham) said: In the discussion of the likely electronically excited states of the surface-adsorbed methoxy species, isoelectronic comparisons are made with CH,F. This is isoelectronic with methoxyl. Is the ‘methoxy’ species physisorbed onto Cu{ 11 l} to preserve a C3, environment through a perpendicular bond to a single Cu atom (in which case methoxy is the best description), or is it chemisorbed in an ‘off-axis’ configuration (in which case methoxyl might be more appropriate)? In the latter case, the local environment would no longer be C3,. Prof. D. P. Woodruff (University of’ Warwick) said: I have two remarks concerning the possibility that the methoxy species bonds to copper surfaces with a bent C-0-Cu bond.The first concerns the fact that the symmetry of these species according to the UPS data is C3\., and this is inconsistent with a bent bond. On the other hand, it is not actually clear whether the symmetry probed by UPS is really sensitive to the substrateGeneral Discussion 335 interaction or not-i.e. the local symmetry is C,,, but this might relate to the CH30 species alone. The second point is that in the case of CH30 on Cu{lOO} we actually have a structural analysis' which indicates that there is an effective bent bond between the C-0 and 0-Cu axes. Nevertheless, the local symmetry in this case obtained from UPS is C3"! 1 Th. Lindner, J . Somers, A. M. Bradshaw, A. L. D. Kilcoyne and D.P. Woodruff, Surf: Sci., 1988,203,333. Dr Prince said: It has previously been shown by polar-angle X-ray photoelectron diffraction that methoxy is tilted on Cu{llO}. Later NEXAFS work confirmed that it is tilted, but disagreed about the tilt angle. The present work shows nicely that the NEXAFS analysis is more complex than was previously thought, and this is probably the source of the contradiction. There is an alternative explanation of the reversal of the order of the 5a, and l e orbitals on Cu(ll1) relative to the other faces. Since the methoxy is tilted on Cu(ll0) but perpendicular on Cu(lll}, this may be a purely geometric effect. Dr A. W. Robinson replied: We are inclined to discount a purely geometric effect to account for the relative positions of the 5a1 and l e orbitals on the grounds that the C-0 axis of the methoxy species is upright on both Cu(lO0) and Cu{lll}, but the ordering of the orbitals is still reversed.Dr S. M. Heald (Brookhaven National Laboratory, Upton) began the discussion of the Dr Jones's paper: You mentioned that the fits require a coherent fraction less than 1. What were the coherent fractions needed for the fits in the paper and what is the reliability of these numbers? Is it possible that the coherent fraction could be used as a diagnostic of the surface system to determine how many of the adsorbed atoms are in the 'proper' sites? Dr R. G. Jones ( University ofNottingham) replied: For chlorine on Cu{ 11 1) a coherent fraction of 0.8 was used to fit the copper Auger data, the same number was then used for fitting the chlorine Auger data [see ref.(5) for a discussion of this]. For CH3S on Cu(ll1) the same value was used for both copper and sulphur atoms, but for sulphur on Cu(ll1) the data could not be fitted with a single-layer spacing and the coherent fraction of the substrate copper; hence we discuss possible reconstruction in which several Cu{ 11 1)-S layer spacings are involved, each with an individual coherent fraction equal to that of the substrate. For Hg on Ni(100), a single-layer spacing system, the coherent fraction was 0.9. The coherent fraction used for the adlayer cannot exceed that of the substrate [ref. ( 5 ) ] , while that of the substrate depends on thermal motion and the static disorder of the substrate within the penetration depth of the X-rays.Thus, until instrumental broadening effects, and inherent disorder in the substrate can be reduced to such a level that the coherent fraction for the substrate + 1 .O, it will not be possible to use the coherent fraction as a diagnostic tool for surface disorder. Dr 3. E. Macdonald (University of Wales College of Cardif) said: Since NlXSW gives the adsorbate position relative to the bulk lattice planes, whereas SEXAFS yields adsorbate bond lengths, relaxations in the surface region may be detected. However, it is not clear how relaxations normal to and parallel to the surface may be distinguished, particularly where relaxation occurs over several atomic layers. Dr Jones replied: Taking CH,S as an example; although the SXW layer distance and the SEXAFS bond length and angle agree rather well, when a model is introduced, it is found that a simple adlayer model is not possible.The only way to explain both sets of results is to invoke substrate reconstruction involving both lateral and perpen- dicular displacements of surface atoms.336 General Discussion Dr Thornton asked: To what extent does multiple scattering affect your photoelectron diffraction results? Prof. Woodruff replied: In fact there are two important refinements to the theory which we include which parailel similar developments in EXAFS-namely, ‘curved- wave’ rather than plane-wave scattering, and multiple scattering. As in EXAFS, multiple scattering is mostly important under conditions in which strong forward scattering can occur.In fact we include only double scattering but find that the largest difference between a single scattering plane-wave calculation, and one using the curved-wave double-scattering approach is in computational cost! There are certainly detailed quanti- tative changes to the spectra, but the only real error typical of using the simpler approach is of bond length changes in the 0.01-0.02 8, level. Prof. Bradshaw then said: You refer in your paper to the requirement that the photoionisation cross-section of the adsorbate atom has to be sufficiently high at the energy of the substrate Bragg reflection in order to apply the standing-wave technique. Have any attempts actually been made to see whether there is a measurable signal from low-2 adsorbate atoms such as C, N and O? Dr Jones replied: As far as I am aware, NISXW has not been attempted for adsorbate atoms such as C, N and 0.Not only would the cross-sections be low for h v = 3000 eV, but the Auger peaks would occur on top of a relatively large secondary electron background, making the experiment difficult, but perhaps not impossible. The technique should be readily applicable to the L and then M levels of the third, fourth etc. rows of the periodic table. The main experimental requirement is that the Auger peak should lie on a low background, well away from photoelectron peaks or their secondary electron tails. Prof. Woodruff then said: Prof. Bradshaw is certainly correct that the photoionisation cross-section of low-2 elements (C, N, 0) at typical Bragg-peak energies (in excess of 3 keV) is low; XSW experiments on these adsorbates might prove possible in the future.For example, we do know that C 1s XPS peaks can certainly be recorded with Mg Ka, a photon energy (ca. 1.5 keV) also rather well above the photoionisation threshold. Dr Thornton said: Just to add to the last point. The C 1s signal strength from CO on Fe(100) using a conventional X-ray source was sufficient to allow Fadley and co-workers to record photoelectron diffraction data. Prof. Woodruff replied: Dr Thornton is quite correct in stating that Fadley has performed X-ray photoelectron diffraction on CO adsorbed on Fe (and other surfaces) using conventional XPS laboratory sources. Indeed, we have also performed such experiments. This is an example of the visibility of C 1s XPS signals which I mentioned in my response to Prof.Bradshaw. Prof. King then turned the discussion to the paper of Dr Jupille: Your fig. 3 shows the apparent bond distance between Xe atoms in solid xenon as a function of temperature, as determined by SEXAFS, together with a comparison with data from X-ray diffraction. The larger expansion coefficient derived from the SEXAFS analysis suggests a problem in the analysis. The most obvious source of discrepancy lies in the assumption of a Gaussian pair distribution function: at temperatures of 40-50 K we should anticipate substantial anharmonic behaviour in xenon. However, I note that this would produce an error in the opposite direction to that which you observe. My main point is to question the reliability of the temperature dependence in Xe-Xe spacings in the monolayer regime (fig.4) from your SEXAFS analysis in the light of the discrepancy observed with bulk xenon.General Discussion 337 Dr Jupille replied: It is worth noting that the discrepancy between the X-ray curve and the SEXAFS data collected from the bulk xenon is only likely, since it is suggested by the good internal consistency of the SEXAFS data, although, within the precision of the measurements, SEXAFS and X-rays agree (fig. 3). In this respect, the above discrepancy is of secondary importance and should not affect our conclusions. However, even if the thermal expansion law of the Xe-Xe spacing in the monolayer regime was questioned, it must be taken as more pronounced than that of the bulk xenon (fig.4), which in turn (even in its X-ray version) clearly differs from that of the almost tem- perature-independent copper parameter. This indeed supports our main finding, i.e. the incommensurability of the Xe/Cu{ 1 1 1 ) adlayer (in the monolayer regime). Dr Heald said: Regarding anharmonic affects: there is a simple test for anharmonicity. If you take the phase difference between data taken at different temperatures, anharmon- icity would show up immediately as a non-linear curve. This would be a very interesting test to make as the behaviour of R would indicate an unusual type of anharmonicity if it exists. Dr Jupille replied: This is a good suggestion since an anharmonicity would generate terms depending on the high orders of the radial distribution.An example of that is given by the Kr-Grafoil system' where an extension to the third moment of the radial distribution was found to be sufficient to account for the observed anharmonicity. 1 C . Bouldin and E. A. Stern, Pbys. Rev. B, 1982, 25, 3462. Prof. D. C. Koningsberger (Eindhoven University of Technology) said: You have studied the adsorption of xenon layers on a Cu{ 1 1 1) surface. I was wondering whether it is possible to use this technique also for studying the surface structure of bimetallic particles deposited on a polycrystalline support. You can probe the structure from both sides, the X-ray absorption edges of the two metals and also the edge of the adsorbed xenon. Dr Jupille replied: A successful study of xenon adsorbed on bimetallic particles dispersed on a solid would of course offer a non-destructive tool for probing the surface of the particles. However, such an analysis first implies obtaining a selective adsorption of xenon on the metallic particles under study (so as to avoid adsorption on the support). Moreover, most of the xenon layers form incommensurate structures and, as far as the Xe-substrate bonding is concerned, the meaning of the number of nearest neighbours can be questioned (see the question from Prof.Woodruff). Prof. King commented to Prof. Koningsberger: I am sceptical about the idea of using SEXAFS analysis of xenon adsorbed on bimetallic catalysts to obtain information about sites. In a situation where every xenon atom occupies the same structural site, as may pertain in some situations on a single-crystal surface, each atomic shell, with its phase shifts and spacing, can be readily assigned to the site array.If more than one type of site were occupied by the xenon atom, as would occur on a heterogeneous catalyst, it would be virtually impossible to make a structural assignment of the various sites from the many atomic shells that would contribute to the SEXAFS spectrum. Prof. Bradshaw then said: Photoelectron diffraction might be more suitable for probing the structure of bimetallic surfaces. Owing to the core-level chemical shift it should be possible to distinguish between Xe atoms adsorbed on each component material. A reasonably homogeneous system with only one adsorption site of each type would, however, be necessary.338 General Discussion Dr Jupille replied: This approach offers various possibilities since, in addition to the valence (core)-level binding-energy shifts, the temperature of adsorption of the xenon adatoms can be used to characterise the adsorption sites.However, the suggestion of probing the surface structure by means of photoelectron diffraction implies the use of monocrystalline surfaces. Note that this would very greatly facilitate the EXAFS study of the Xe/bimetallic systems. Prof. Woodruff said: The problem of SEXAFS for an incommensurate overlayer referred to by Prof. King leads me to another point which is that I think the meaning of the value obtained for the adsorbate-substrate (Xe-Cu) bond lengths obtained from SEXAFS is far from clear. Even if the true nearest-neighbour Xe-Cu bond length is site-independent, the incommensurate overlayer will lead to a continuous distribution of ‘second-nearest-neighbour’ bond lengths extending from this true nearest-neighbour value.Thus, the true Xe-Cu distance distribution will have a strongly asymmetric peak at the ‘nearest-neighbour distance’, and SEXAFS will distort the distribution just as in the case of asymmetric distribution due to anharmonic effects. The interpretation of this value is therefore far from clear. Dr Jupille replied: This is quite an important comment since it is questioning the capability of SEXAFS to deal with incommensurate overlayers. The expected asym- metric Xe-Cu distribution of distances is peaking at the true nearest neighbours Xe-Cu ( N N Xe-Cu) bond length since any Xe atom has at least one true N N Cu atom in its coordination shell.The asymmetry involves a distribution of ‘second-nearest-neighbour’ bond lengths, i.e. bond lengths larger than true NN Xe-CJ bond lengths. This implies that the bond length derived from the SEXAFS analysis has a meaning: it is an upper limit of the genuine NN Xe-Cu bond length. In the present case ( i e . physisorbed rare-gas adatoms) it is reasonable to assume that the N N Xe-Cu bond length cannot be very much contracted with respect to the hard-sphere model. The upper limit of the Xe-Cu bond length (the SEXAFS measure- ment) which is equal to the distance calculated using the hard-sphere model (table l ) , is therefore likely to be a good estimate of that bond length.More insight should probably be gained into that question by an appropriate simulation. Dr Thornton said: Dr Jupille draws attention to the interference term between outgoing s- and d-waves in L,-edge SEXAFS, which has been discussed by Stohr and Jaeger,’ and Citrin.’ The presence of this effect can result in erroneous bond distances and coordination numbers unless suitable precautions are taken. Perhaps the easiest involves the use of the ‘magic-angle’ experimental geometry, in which the X-ray electric vector is at ca. 54.7” to the surface normal. In this geometry the cross-term is negligible, whatever the adsorption site. Other possibilities exist; criteria for choosing an experi- mental geometry in which the effect of the cross-term is effectively removed are displayed graphically in fig.2 [after ref. ( 1 )]. Inspection of fig. 2 shows that Jupille’s choice of a normal geometry should indeed result in a negligible cross-term for the back-scattering contribution from Xe atoms in the overlayer. However, this is not necessarily the case for the substrate back-scattering contribution. To illustrate this point, the appropriate positions for the first and second Cu nearest-neighbour backscatterers are marked on the cross-term ‘map’ in fig. 2. These have been derived assuming hollow site adsorption and a Xe-C‘u nearest-neighbour distance of 0.345 nm. These observations have three implications. First, the accuracy of the derived Xe-Cu distance should be significantly smaller than that normally obtained with SEXAFS, which is presumably the origin of the *0.01 nm error bar on the Xe-Cu distance.Secondly, cross-term effects could provide an alternative explanation for the unusually low E,, value (-12 eV) found for the Cu back-scatterer contribution. Finally, theGeneral Discussion 339 80 60 0 \ G2 40 20 0 - 0 20 40 60 80 P / " Fig. 2. Cross-term map for L,-edge SEXAFS [after ref. (l)]. In the unshaded region the cross-term is negligible. The appropriate position for the first @ and second 0 nearest-neighbour Cu atoms are indicated. second-nearest-neighbour Cu atoms may have an anomalously large back-scattering amplitude. Since the appropriate Xe-Cu bond distance is 0.43 nm, this may complicate analysis of the Xe back-scattering contribution. 1 J. Stohr and R. Jaeger, Ph-vs. Rev. B, 1983, 27, 5146. 2 P. H. Citrin, Phys. Rev. B, 1985, 31, 700. Dr Jupille replied: The polarisation vector was set parallel to the surface so as to maximise the in-plane back-scattered amplitude or to c p c e l out perfectly :he cross-term associated with it [this geometry leads to Cz, ( 1 -31E. = 0, where E is the polar- isation vector of the light and ?, the unit vector connecting the central atom and its neighbours in the considered coordinate shell]. * In these conditions, the scattering from the first-neighbour copper atoms involves an s-d cross term. However, in what follows, we shall see that this term affects only slightly the determination of the Xe-Cu bond length. Bond-length anisotropies due to the effect of cross-terms are maximum for atoms in the atop geometry ( ~ 0 . 0 0 2 nm) and less for other geometries (<0.0015 nm).' This is negligible compared to our error bar (0.01 nm). The cross-term correction, Ak, for the adsorbing-atom phase shift is difficult to evaluate for geometries other than atop and A k = 0 was found to be an appropriate approximation in most cases.' In addition, the value of the internal potential of the copper substrate? actually justifies the observed value of E, (ca. -12 eV). From a simple consideration of the scatterers (number and nature), the back-scattered amplitude from the second-nearest-neighbour Cu atoms is expected to be slightly smaller than that from the first-nearest-neighbour Cu atoms. The existence of a cross-term could perturb this, but, as seen in Thornton's fig. 2, the adsorption geometry and polarisation associated with the next-nearest-neighbour Cu atoms is not expected to give rise to any important cross-term effects. 1 P. H. Citrin, Ph!.y. Rev. B, 198S, 31, 700. 2 S. A. Lindgren, L. Wallden, J. Rundgren and P. Westrin, Phys. Reu. Lerr., 1983, 50, 368.

ISSN:0301-7249    

DOI:10.1039/DC9908900329

出版商:RSC    

年代:1990

数据来源: RSC

摘要:

I N D E X O F Ashwin, M. J., 301 Barrera, E. V., 21 Barrett, N. T., 51 Bradshaw, A. M., 1, 65, 66, 73, 143, 205, 206, 291, Cai, Z-h., 211 Carr, S. W., 119 Cassuto, A., 323 Catlow, C. R. A., 119 Chen, H., 21 Chung, Y. H., 332 Couves, J. W., 70 Cowley, R. A., 181, 207, 210, 249, 256 Derby, G. P., 259 Dooryhee, E., 119 Dring, I., 51 Earnshaw, J. C., 154, 249 Ehrhardt, J-J., 323 Evans, J., 107 Fargues, D., 323 Fontaine, A., 41, 65, 70, 73, 155, 157, 275, 333 Forty, A. J., 31 Garrett, R., 275 Gauntlett, J. T., 107, 149, 152, 157 Greaves, G. N., 51, 74, 119, 149, 151, 152, 156, 207 Guay, D., 41, 275 Hardman, P., 77 Heald, S. M., 21, 66, 67, 68, 71, 335, 337 Johnson, A. L., 77 Johnston, P., 91 Jones, R. G., 301, 335, 336 Joyner, R. W., 66, 72, 91, 145, 147, 148, 149, 150, Jupille, J., 68, 70, 323, 329, 337, 338, 339 Karnpers, F.W. H., 137 Kerkar, M., 31 Kern, K., 159 Kilcoyne, A. L. D., 311 King, D. A., 147, 154, 155, 206, 207, 259, 329, 330, 33 I , 332, 336, 337 King, G. C., 332 Koningsberger, D. C., 67, 69, 70, 137, 144, 146, 148, Lambooy, P., 249 Law, D. S-L., 77 Leadbetter, A. J., 153 Lindner, Th., 311 311, 330, 333, 336, 337 156 151, 152, 153, 154, 155, 157, 337 NAMES* Lucas, C. A., 181 Macdonald, J. E., 169, 191, 206, 210, 335 Mosselmans, J. F. W., 107, 147 Muryn, C. A., 77 Norris, C., 169 Oldman, R. J., 51, 67, 150, 152 Ottewill, R. H., 247 Parsons, R., 69 Pershan, P., 67, 206, 207, 210,231,248, 249, 250, 255 Pizzini, S., 51 Prabhakaran, K., 65, 68, 153 Prince, K. C., 329, 335 Prince, N. P., 301 Prakash, N. S., 77 Pudney, P. D. A., 91 Purdie, D., 77 Raiker, G. N., 77 Rice, S. A., 73, 211, 247, 248, 250, 256, 257 Ricken, D. E., 291, 311 Roberts, K. J., 51, 71, 72, 74 Robinson, A., 311 Robinson, A. W., 291, 334, 335 Robinson, I. K., 66, 71, 151, 159, 201, 204, 206, 207, Robinson, J., 31, 68, 69, 70, 143, 152 Robinson, K. M., 68, 157, 248 Shpiro, E. S., 91 Simons, J. P., 334 Singh, N. K., 301 Smith, J. V., 69, 151, 152 Somers, J., 291, 311 Steel, A. T., 119 Thomas, J. M., 68, 72, 119, 151, 152, 153, 156, 258 Thornton, G., 70, 77, 143, 330, 334, 336, 338 Tourillon, G., 41, 275 Townsend, R. P., 119 van der Veen, J. F., 67, 169, 206, 255, 257 van Silfhout, R. G., 169, 204 Vilfan, I., 202 Villain, J., 202 Vlieg, E., 159 Walter, W., 301 Williams, B. P., 91 Williams, G. P., 275 Wincott, P. L., 330 Woodruff, D. P., 65, 143, 203, 205, 301, 311, 334, 208 336, 338 * The page numbers in heavy type indicate papers submitted for discussion. 341

ISSN:0301-7249    

DOI:10.1039/DC9908900341

出版商:RSC    

年代:1990

数据来源: RSC