Surface Core-level Shifts Research Articles

Coverage-dependent hydrogen adsorption site determination onRh(100)by high-resolution core-level spectroscopy

We show that high-resolution real-time x-ray photoelectron spectroscopy can be used to determine hydrogen adsorption sites as a function of coverage on $\mathrm{Rh}(100)$. The measurement of the surface core-level shifts does not suffer from the lack of direct sensitivity of other surface probes due to the low scattering cross section and high mobility of atomic hydrogen. At low temperatures $(70--140\phantom{\rule{0.3em}{0ex}}\mathrm{K})$ and coverage (below $0.25\phantom{\rule{0.3em}{0ex}}\mathrm{ML}$), we find that hydrogen adsorbs in fourfold hollow sites on $\mathrm{Rh}(100)$, while at higher coverage the bridge site is preferred. Using Monte Carlo simulations, we unequivocally associate each surface component of the $\mathrm{Rh}\phantom{\rule{0.3em}{0ex}}3{d}_{5∕2}$ core level with a specific adsorption configuration. We obtain a value of $0.74\ifmmode\pm\else\textpm\fi{}0.08$ for the hydrogen initial sticking coefficient, in very good agreement with previous reports.

Rumpled relaxation and surface core-level shift in PbS(001)

The surface relaxation of the PbS (001) surface has been calculated by the all-electron full potential linear augmented-plane-wave method. Surface core level shifts have been estimated at different atomic sites of the rumpled surface of the crystal. The rumpling is such that the cations are slightly below the anions in the surface layer. Comparison with recent experimental measurements by x-ray photoemission spectroscopy shows good agreement. The initial state approximation slightly overestimates the value of the surface core level shifts (SCLS) whereas the final state effect reduces the SCLS. It is shown that the Madelung potential cannot explain the shifts.

Assignment of the surface core-level shifts to the surface layers ofBe(101¯0)

Photoemission studies and layer resolved Korringa-Kohn-Rostoker (KKR) multiple scattering calculations are used to find the assignment of the surface core-level shifts to the top layers of the Be(10 (1) over bar0) surface. Striking similarities between experimental and calculated data make it possible to assign the largest shift to layer two, the second largest shift to layers three and four, and the smallest shift to layer one.

Experimental and theoretical surface core-level shifts of aluminum (100) and (111)

The surface core-level shifts of Al(111) and Al(100) have been measured using high-resolution core-level photoemission spectroscopy and calculated using density functional theory (DFT). For Al(100), the 2p core-level shift of the first (second) layer was determined to be –75 meV (+20 meV) from experiment and –71 meV (+20 meV) from the DFT calculations. For Al(111), the corresponding values are –27 meV (0 meV) from experiment and –14 meV (–) from the DFT calculations. Core-level splittings caused by the low-symmetry crystal fields at the (111) and (100) surfaces have also been studied. These splittings turn out to be much smaller than previously reported provided proper care is taken of the influence of the core hole screening and of core–valence exchange beyond the DFT level. Finally, the experimental Al 2p line shape was found to contain structure caused by a sharp no-phonon line and a broad and weak phonon replica. (Less)

XPS study of the oxidized pyrite surface

The core level spectra of pristine and oxidized mineral pyrite FeS2 surfaces were studied using synchrotron excited photoelectron spectroscopy. In addition to the bulk feature, three surface core level shifted components were identified in the sulphur 2p spectra. After measuring the spectra of the clean pristine surface, the sample was oxidized for 12 h in O2 atmosphere. The effect of this treatment of the specimen, besides causing weakening in the signal intensities, was observed as changes in the surface core level shifts (SCLS). The surface components of the oxidized pyrite spectra were shifted towards higher binding energies compared to those of the pristine sample. The reason for this phenomenon is briefly discussed.

A surface core-level shift study of hydrogen interaction with Rh(1 1 1)

The interaction of hydrogen with the Rh(1 1 1) surface was studied by means of high-resolution real-time X-ray core-level photoemission. The analysis of the intensity of the Rh3d 5/2 core-level components was used to determine the hydrogen adsorption sites as a function of the hydrogen dose. While at low coverage only three-fold adsorption sites are occupied, at higher hydrogen concentrations bridge species are also present. The temperature dependence of the core-level spectra indicates that during desorption, surface depopulation occurs in an almost reverse order with respect to the adsorption process.

Theoretical study on the temperature-induced structural transition of the Si(1 1 3) surface

The temperature-induced structural transition of the Si(1 1 3) surface is investigated by ab initio calculations. In this study, it is found that the room-temperature phase and the high-temperature phase have the 3 × 2 interstitial structure and the 3 × 1 interstitial structure, respectively. The existence of the 3 × 2 and 3 × 1 interstitial structures is supported by the analysis of scanning tunneling microscopy (STM) images and the calculation of surface core level shifts using final state pseudopotential theory. The theoretical STM images of interstitial structures are in good agreement with the STM images suggested by experiments. The analysis of STM images provides an insight into the characteristics of domain boundaries observed frequently in experiments. It is also found that the domain boundary can be formed by local 3 × 1 interstitial structures on the 3 × 2 interstitial surface. The theoretical analysis of the surface core level shifts reveals that the surface core levels in experiment originate from the interstitial structures. The lowest values in the surface core level shifts are found to be associated with the 2p core level shifts of the interstitial atoms.

A high resolution photoemission study of surface core-level shifts in clean and oxygen-covered Ir(2 1 0) surfaces

High resolution soft X-ray photoemission electron spectroscopy (SXPS), using synchrotron radiation, is employed to investigate 4f core-level features of four differently-prepared Ir(2 1 0) surfaces: clean planar, oxygen-covered planar, oxygen-induced faceted, and clean faceted surfaces. Surface and bulk peak identifications are supported by measurements at different photon energies (thus probing different electron escape depths) and variable emission angles. Iridium 4f 7/2 photoemission spectra are fitted with Doniach–Sunjic lineshapes. The surface components are identified with core levels positioned at lower binding energies than the bulk components, in contrast to previous reports of binding energy inversion on Ir(1 0 0) (1×1) and (5×1) surfaces. For clean planar Ir(2 1 0) three surface Ir 4f 7/2 features are observed with core-level shifts of −765, −529, and −281 meV, with respect to the bulk; these are associated with the first, second and third layers of atoms, respectively, for atomically rough Ir(2 1 0). Adsorption of oxygen onto the planar Ir(2 1 0) surface is found to cause a suppression and shift of the surface features to higher binding energies. Annealing at T≥600 K in oxygen produces a faceted surface as verified by low energy electron diffraction (LEED). A comparison of planar and faceted oxygen-covered surfaces reveals minor differences in the normal emission SXPS spectra, while grazing emission spectra exhibit differences. The SXPS spectrum of the clean, faceted Ir(2 1 0) exhibits small differences in comparison to the clean planar case, with surface features having binding energy shifts of −710, −450, and −230 meV.

XPS study of the sulphur 2p spectra of pyrite

The surface core level shifts have been measured and interpreted for cleaved FeS 2{1 0 0}. A new feature has been observed at about 2.0 eV from the S 2p bulk line on its low binding energy side. This small peak is the most surface sensitive component. The origin of this feature is thought to be the breakage of sulphur–sulphur bonds. The strongest surface peak is ascribed to the fracture of the Fe–S bond. An estimate for the effective attenuation lengths of photoelectrons in pyrite at kinetic energies 40, 180 and 610 eV supports the interpretation of the features.

High-resolution photoemission spectroscopy study of clean and Au-nanowire-decorated Si(5 5 12) surfaces

We have examined the Si $2p$ photoemission line shape of the clean $\mathrm{Si}(5512)2\ifmmode\times\else\texttimes\fi{}1$ surface in detail and have investigated the evolution of the Au-induced nanowires on this surface at different temperatures with low-energy-electron diffraction and photoemission spectroscopy. The Si $2p$ spectra from the clean $2\ifmmode\times\else\texttimes\fi{}1$ surface exhibit very complex line shapes composed at least of five different surface-related components. Possible origins of the most distinct surface component with a surface core-level shift of $\ensuremath{-}0.62\mathrm{eV}$ are discussed within the three different structure models proposed so far. Upon Au deposition, the intensity of that particular surface component was largely reduced indicating that the Au adsorbates interact dominantly with the specific Si surface atoms related. With an annealing above $500\ifmmode^\circ\else\textdegree\fi{}\mathrm{C}$ after Au deposition of 0.2 ML, the surface drastically changes into a metallic phase, as shown clearly in the valence band photoemission spectra, along with the gradual formation of the $(337)\ifmmode\times\else\texttimes\fi{}2$ facets observed in LEED. In this transition, a new Au $4f$ component appears while the Si $2p$ core-level spectra keep alike. This indicates that the surface metallization involves the rearrangement of Au adsorbates, possibly the formation of new Au chains on the surface.

Structural evolution of the Si(113) surface: Ab initioand tight-binding molecular dynamics calculations

The structural evolution of the Si(113) surface is investigated by ab initio calculations and by tight-binding molecular dynamics calculations using the environment-dependent tight-binding Si potential. In this study, it is found that $3\ifmmode\times\else\texttimes\fi{}2$ and $3\ifmmode\times\else\texttimes\fi{}1$ phases have interstitial structures and the transition from $3\ifmmode\times\else\texttimes\fi{}2$ to $3\ifmmode\times\else\texttimes\fi{}1$ phase results from the increasing of interstitial atoms. The existence of the interstitial structures is proved by the analysis of scanning tunneling microscopy (STM) images and the calculation of surface core level shifts using final state pseudopotential theory. The study of adsorption energy clarified that the phase transition from the $3\ifmmode\times\else\texttimes\fi{}2$ to $3\ifmmode\times\else\texttimes\fi{}1$ interstitial surface plays an important role in the behavior of the ${113}$ facet in the selective epitaxial growth of Si(001). It is also found that the domain boundary observed frequently in the filled state STM images of Si(113) is formed by the local $3\ifmmode\times\else\texttimes\fi{}1$ interstitial structure on the $3\ifmmode\times\else\texttimes\fi{}2$ interstitial surface.

Optimized surface-slab excited-state muffin-tin potential and surface core level shifts

An optimized muffin-tin (MT) potential for surface slabs with preassigned surface core-level shifts (SCLS's) is presented. By using the MT radii as adjustable parameters the model is able to conserve the definition of the SCLS with respect to the bulk and concurrently to generate a potential that is continuous at the MT radii. The model is conceived for elastic electron scattering in a surface slab with exchange-correlation interaction described by the local density approximation. The model employs two data bases for the self-energy of the signal electron (after Hedin and Lundqvist or Sernelius). The potential model is discussed in detail with two surface structures $\mathrm{Be}(101\ifmmode\bar\else\textasciimacron\fi{}0),$ for which SCLS's are available, and $\mathrm{Cu}(111)p(2\ifmmode\times\else\texttimes\fi{}2)\mathrm{Cs},$ in which the close-packed radii of the atoms are extremely different. It is considered plausible that tensor LEED based on an optimized MT potential can be used for determining SCLS's.

Surface Core Level Shifts of Yb Films in the Multilayer, Monolayer and Submonolayer Range

AbstractSurface core level shifts in Yb and other rare earth elements, as well as in their compounds, arise from a reduced band width of the 5d band, caused by the reduced coordination at surface sites, with the strength and direction of the shifts depending on the occupation of the 5d band. If the coordination, or more exactly, the local symmetry is considered to be the decisive factor for changing the band width, it is obvious to look for core level shifts even at the interface of two solids. In order to investigate such interface core level shifts, thin Yb films of thicknesses from the multilayer down to the submonolayer range, grown on a polycristalline W substrat, were investigated by photoelectronspectroscopy. It was found, that for decreasing film thicknesses, the ratio of the surface to bulk related 4f intensities increases systematically, and finally, for thicknesses of one up to two monolayers, only the broad, surface related part of the Yb 4f doublet can be observed. For Yb coverages fewer than one monolayer, there is a sudden shift of this remaining doublet towards the Fermi energy, which must result from a reduced occupation of the Yb 5d band due to an electron transfer from Yb to W at the interface.

Relaxation of the [formula omitted] surface in binary Sc, Ti and V nitrides: a first principles density functional study

Density functional calculations with periodic boundary conditions of the structure of ScN, TiN and VN(0 0 1) surfaces are reported. Using slabs of different thickness to model the surface, the rippling relaxation of outermost layer is estimated to be 0.011, 0.178 and 0.256 Å, respectively. Calculations show that on nitrogen atoms there is a significant surface core-level shift arising from both the lack of periodicity and the rippling relaxation. For the TiN(0 0 1) surface, the rippling contribution to the N 1s surface core-level shift is estimated to be −0.20 eV (35%), with a final value of −0.56 eV, in excellent agreement with photoemission experiments. The analysis of the electronic band structure of TiN(0 0 1) surface along the Γ– X and X– M directions confirms that the experimentally surface state observed in the angle-resolved photoemission spectra at −2.9 eV is due to a Tamm surface state pulled off the top of the Δ 5 bulk band. The shift of the state is also clearly enhanced by the rippling relaxation.

Core-level shifts for surface bimetallic systems from first-principles theory: Pd-Mn structures on Pd(100)

The surface core-level shift (SCLS) for Pd(100) and Mn-induced Pd core-level shifts (CLSs) for PdMn surface alloys with different configurations formed on Pd(100) are calculated from first-principles theory. The shifts are calculated within the complete screening picture, which includes initial (core-electron energy-eigenvalue) and final state (relaxation due to screening) effects in the same scheme. The theoretical results are compared to recent experimental results for 1 monolayer Mn/Pd(100), formed at low temperature, and the ordered Pd(100)-$c(2\ifmmode\times\else\texttimes\fi{}2)$-Mn surface alloy formed upon annealing. The calculated core-level shifts are in all cases in good agreement with the experimental values. It is furthermore shown that the Pd(100)-$c(2\ifmmode\times\else\texttimes\fi{}2)$-Mn surface alloy as prepared experimentally is metastable, and predictions are made on the CLSs for energetically more stable structures that could be reached with further annealing.

A Detailed Theoretical Study of TiC(001) Relaxed Surface

AbstractThe geometric and electronic structures of clean TiC(001) surface were investigated using first principles calculations based on generalized gradient approximation density functional theory. The structural optimization of the surface shows that the C and Ti atoms on the top layer move outward and inward, respectively. The results of the band structure calculation indicate that the “active” surface states of TiC(001) surface located around 2.3 eV below Fermi level are originated from 2pz orbital of surface C atoms. Furthermore, the influence of the surface relaxation on the electronic structures, including the charge densities, work functions and surface core level shifts are also discussed in the paper.

Surface XPS core-level shifts of FeS2

The XPS S 2p core-level spectra of pyrite FeS2 have been measured to study the surface core-level shifts (SCLSs). In addition to the previously found features on the low binding energy (BE) side of the line, evidence for a new structure on the high BE side is given. This feature seems to be surface sensitive. Also it is found that the final state effect may give a significant contribution to the SCLS.

Reinvestigation of the Si2pphotoemission line shape from a cleanSi(001)c(4×2)surface

We have reinvestigated the Si $2p$ photoemission spectra of a clean $\mathrm{Si}(001)c(4\ifmmode\times\else\texttimes\fi{}2)$ surface with an apparently higher resolution than previous studies. A systematic and extensive line-shape analysis of the spectra taken at various photon energies and emission angles indicates that the Si $2p$ spectra are composed of seven different components. With regard to the recent debate on the Si $2p$ component of the down-dimer atoms, we conclude that even with the present resolution the corresponding surface core-level shift (SCLS) cannot be determined accurately but is in the range of 30--130 meV between the bulk and the second-layer components. One of the two newly identified components, with a SCLS of 1.4 eV, is interpreted as a surface loss structure due to the interband transition between surface-state bands.

Surface core level shift observed on NiAl(1 1 0)

Using high resolution core level spectroscopy, a surface core level shift towards lower binding energy of −0.13 eV is determined for the 2p level of the outwardly relaxed Al surface atoms on NiAl(110). Density functional theory based calculations with inclusion of final state effects yield a value of −0.14 eV for this shift in excellent agreement with experiment. We show that the initial state approximation yields a value of +0.09 eV, i.e. the inclusion of final state relaxation effects is vital not only to obtain the correct value but even the correct sign for this shift.

First principles studies of the surface of galena PbS

Ab initio calculations of the surface structure and properties of the (001) surface of galena (PbS) have been performed using two complementary approaches. Both Gaussian and numerical basis functions have been utilized, as embodied within the programs CRYSTAL98 and SIESTA, with their associated implementations, in order to determine to what extent computational factors are influencing the results. Relaxations are presented based on calculations employing the HF, B3LYP, LDA, and GGA Hamiltonians. Significant relaxations (up to about 0.2 Å) are found to occur at the surface which decay with increasing depth into the bulk. The magnitude of the relaxation oscillates between layers, leading to pairing, with Pb ions being displaced outwards (by approximately 0.05 Å) relative to the S ions. We find that the LDA, GGA, and B3LYP Hamiltonians give results in reasonable agreement with each other, but that the structures computed with HF theory are vastly different. The present calculations also predict a surface core level shift of 0.3 eV for the sulphur 2 p level at the surface itself, in excellent agreement with a recent experimental determination.