Search

Low-energy electron diffraction (LEED) I – V has been utilized to determine the surface structure of Cu(2 1 0). The surface structure is found to exhibit multilayer relaxation following the trend − − + (− is contraction and + is expansion in the interlayer spacing). The magnitude of interplanar relaxation is found to be damped, where | Δd 12 |>| Δd 23 |>| Δd 34 |, etc. This kind of behavior is quite different from that observed on Al(2 1 0) [Phys. Rev. B 38 (1988) 7913], where the magnitude of interplanar relaxation in the third interlayer is larger than that in the second interlayer (i.e., | Δd 23 | Δd 34 |). The difference of damped multilayer relaxation behavior of Cu(2 1 0) and Al(2 1 0) could be related to a charge density oscillation perpendicular to the surface.

The surface structure of Ru(0 0 0 1) has been studied between 300 and 1870 K using X-ray diffraction. At 300 K, the outermost layer is contracted by 2.2(4)%, in agreement with the value determined by low-energy electron diffraction studies, but about half that predicted by ab initio calculations. H-adsorption appears to introduce disorder through an increased surface Debye–Waller factor and an increased step density, but does not significantly alter the interlayer spacing. Temperature-dependent studies of the clean surface structure indicate a thermal expansion of the first three interlayer distances that is slightly smaller than that of the bulk, accompanied by increasing vibrational motion comparable to that of the bulk.

The growth morphology and electronic structure of Ni on Ag(100) has been studied with scanning tunneling microscopy (STM) and synchrotron based angle resolved photoemission spectroscopy. At deposition temperatures at or below 300 K, STM reveals Ni cluster growth on the surface along with some subsurface growth. Upon annealing to 420 K, virtually all Ni segregates into the subsurface region forming embedded nanoclusters. The electronic structure of Ni d bands in the unannealed surface shows dispersion only perpendicular to the surface whereas the annealed surface has Ni d bands that exhibit a three-dimensional-like structure. This is a result of the increased Ni d–Ag sp hybridization bonding and increased coordination of the embedded Ni nanoclusters.

We present X-ray scattering studies of the structure and growth of the heteroepitaxial interface Cu–Ru(0001) between 300 and 870 K. For Cu depositions of between one and two monolayers (ML), a bilayer stripe-phase coexists with domains pseudomorphic to the Ru substrate. At the completion of 2 ML, the stripe-phase wavevector abruptly locks to that of the substrate with a commensurability δ=1/16 or 1/18, depending on the temperature. Above 2 ML, a cooperative growth process leads to three-phase coexistence, including the bilayer stripe-phase and rotated and unrotated bulk-like Cu islands.

Electron-energy-loss spectroscopy (EELS) data are presented to illustrate line shape changes that occur as a result of oxygen interaction with metal surfaces. The metals were aluminum, beryllium and nickel. Core-level EELS data were taken for excitations from Al(2p), Be(1s), Ni(3p/3s) and O(1s) levels to the conduction band (CB) density of states (DOS) of the materials. The primary beam energies for the spectra were 300, 450, 300, and 1135 eV, respectively. The data are presented in both the (as measured) first-derivative and the integral forms. The integral spectra were corrected for coherent background losses and analyzed for CB DOS information. These spectra were found to be in qualitative agreement with published experimental and theoretical studies of these materials. One peak in the spectra for Al oxide is analyzed for its correlation with excitonic screening of the Al(2p) core hole. Similar evidence for exciton formation is found in the Ni(3p) spectra for Ni oxide. Data are also presented showing o...

We present the results of X-ray scattering studies of the structure and phase behavior of the Ir(001) surface between room temperature and T=2000 K. Two distinct phases have been identified. For temperatures above T=1950 K, the Ir(001) surface is rough, similar to what has been observed for Pt(001) at elevated temperatures. Below about T=1950 K, the surface reconstructs to form a quasi-hexagonal unit cell, which is commensurate with the substrate lying beneath it. The measured X-ray intensities, both inplane and along the specular rod, are consistent with a buckled, (5×1) description of the structure, including a 15% expansion of the top layer spacing and an enhanced surface Debye–Waller factor. A detailed comparison is made of the phase behavior determined here for Ir(001) to that found earlier for Pt and Au(001).

The segregation and ordering behavior of the (100), (110), (210), (310) and (111) surfaces of the ordered intermetallic compound FeAl were investigated by means of AES and LEED. Starting from preferentially sputtered, i.e. Al-depleted surfaces sequential annealing leads to segregation of aluminum for all five surfaces. For intermediate annealing temperatures several ordered superstructures are found for the low index surfaces, which can partly be attributed to metastable Fe 3 Al films developing at the surfaces. All well-annealed surfaces exhibit Al concentrations beyond bulk stoichiometry. The (100) surface is found to be Al terminated, additionally the second layer is enriched in Al. For all other FeAl surfaces no 1×1 diffraction patterns are observed in the whole temperature range up to 1000°C, indicating the absence of bulk terminated surfaces, in contrast to the equivalent surfaces of NiAl. An incommensurate surface alloy of FeAl 2 stoichiometry finally forms on the most densely packed FeAl(110) surface. The other more open surfaces are characterized by multilayer segregation of Al. While (111) and (210) surfaces develop long-range surface reconstructions (3×3 and 1×3), the (310) surface is unstable and facets. The differences in the segregation behavior between NiAl and FeAl can qualitatively be explained by a combination of different segregation models using the heat of formation of the alloy.

The effects of adsorbed H on the Mo1−xRex(110), x=0, 0.05, 0.15, and 0.25, surfaces have been investigated using low-energy electron diffraction (LEED) and high-resolution electron energy loss spectroscopy (HREELS). For the x=0.15 alloy only, a c(2×2) LEED pattern is observed at a coverage Θ∼0.25 ML. A (2×2) pattern is observed for H coverages around Θ∼0.5 ML from surfaces with x=0, 0.05, and 0.15. Both c(2×2) and (2×2) patterns are attributed to reconstruction of the substrate. At higher coverages, a (1×1) pattern is observed. For the alloy surface with x=0.25, only a (1×1) pattern is obtained for all H coverages. Two H vibrations are observed in HREELS spectra for all Re concentrations, which shift to higher energies at intermediate coverages. Both peaks exhibit an isotopic shift, confirming their assignment to hydrogen. For Re concentrations of x=0.15 and higher, a third HREELS peak appears at 50 meV as H (D) coverage approaches saturation. This peak does not shift in energy with isotopic substitution, yet cannot be explained by contamination. The intrinsic width of the loss peaks depends on the Re concentration in the surface region and becomes broader with increasing x. This broadening can be attributed to surface inhomogeneity, but may also reflect increased delocalization of the adsorbed hydrogen atom.

Information about the occupied portion of the surface density of states of materials can be derived from electron-excited Auger electron spectroscopy (AES), which is a standard experimental technique in most surface science laboratories. Surface sensitive experimental techniques that provide information regarding the unoccupied portion of the surface density of states are often not standard and are not so readily available. Here we explore the possibility of utilizing the same experimental equipment as in AES to derive information about the unoccupied portion of the surface density of states from a characteristic loss spectroscopy, in particular, from core-level inelastic electron-scattering spectroscopy (CLIESS). An important application of this technique is in comparative studies. CLIESS spectra from clean surfaces of aluminum, beryllium and nickel are presented. These data were taken in the first-derivative mode using the reflection of monoenergetic primary beams of 450 eV energy for Be, and 300 eV for...

We report an X-ray scattering study of the structure and morphology of a stepped Au(001) surface between room temperature and bulk melting. At high temperatures, between 1170 and 1220 K, the vicinal Au(001) surface appears rough. Below T0 = 1170 K, there is a facetting transition: hexagonally reconstructed (001) facets appear in coexistence with unreconstructed, stepped domains. The tilt angle of the stepped domains increases with decreasing temperature, exhibiting a power-law behavior versus reduced temperature. The phase behavior of the hexagonally reconstructed domains follows that observed on nominally unstepped Au(001) surfaces. At T1 = 992 K, there is a second facetting transition at which the stepped domains reconstruct to form a rotated and corrugated, quasi-hexagonal structure. The corrugation is threefold commensurate with the average terrace width, consistent with a 2° magic vicinal phase. The orientational phase diagram of vicinal Au(001) is deduced and found to be qualitatively similar to that of vicinal Pt(001), although with intriguing differences.

The oxidation of the (100), (110) and (111) surfaces of the intermetallic compound FeAl has been investigated using LEED and XPS. On all three surfaces, oxidation at room temperature leads to the formation of an amorphous oxide film on top of an Al-depleted interlayer. The film growth can be divided into two regions of differing kinetics, i.e. the initial formation of a closed oxide film and a subsequent thickening. In the first region, the oxygen-uptake rate varies significantly with surface orientation, while in the thickening regime the uptake is the same for all surfaces. The maximum thickness as well as the composition of the oxide films were found to depend on the initial oxidation rate. At higher oxidation temperatures, ordered oxide films of around 5–8 A in thickness are formed, very similar to those observed on NiAl. Photoemission spectra from these ordered phases showed evidence for Al atoms in two different chemical environments, i.e. the well-known oxide species in the interior of the film and an additional species present at the oxide/alloy interface.

The adsorption of hydrogen on the Mo(110) surface has been investigated by both low-energy electron diffraction (LEED) and high-resolution electron energy loss spectroscopy (HREELS). Results fall into three regimes. For low coverages, the LEED pattern observed remains (1 × 1), similar to the clean surface. At hydrogen coverages near 0.5 ML, a substrate reconstruction alters the observed pattern to (2 × 2). For higher coverages, fractional-order diffraction beams are extinguished, leaving a (1 × 1) pattern. Two vibrational peaks are observed in HREELS spectra for both low- and high-coverage regimes at loss energies of 88 and 136 meV, and at 99 and 152 meV, respectively. A quasi-trigonal adsorption site is consistent with the vibrational energies and diffraction patterns observed at low and saturation coverages. At intermediate coverages, a vibrational peak with differing symmetry indicates the occupation of a long bridge-site during the transition from (2 × 2) to the hydrogen-saturated (1 × 1). No evidence is found for the hydrogen “liquid” phase reported by Balden et al. [Phys. Rev. Lett. 73 (1994) 854] for H W (110) ; consequently this phase is not responsible for the giant phonon anomalies observed in both systems. Published by Elsevier Science B.V.

As a consequence of bulk truncation, the atomic structure and composition within the surface region of an ordered or random alloy may differ significantly from that of the bulk. Specifically, the response to the rearrangement resulting from the reduction of surface energy may cause elemental redistribution within the outermost layers and two- or three-dimensional atomic reordering. Furthermore, the presence of adsorbates may have a pronounced effect on these properties and cause significant changes. These properties for some surfaces of transition-metal aluminides, transition-metal carbides, and random alloys will be presented.

Clean random alloy surfaces usually exhibit non-bulklike concentration profiles. Adsorption at low temperatures does not modify these profiles as there is thermal activation for interdiffusion of surface atoms required. For strongly chemisorbed species like oxygen or carbon, however, the activation barrier can be overcome prior to desorption leading to a new equilibrium distribution of the surface constituents. This paper summarizes the results of different LEED-I(V) studies on the adsorption of hydrogen, oxygen and carbon on the (100) face of the random alloy Mo 0.75 Re 0.25. Hydrogen, which desorbs near room temperature, is found to affect the interlayer spacings only, whilst the concentration profile remains completely unchanged. In contrast, enhanced molybdenum segregation at temperatures above 600° C is shown to happen with oxygen and carbon adsorption leading to exclusive Mo-C and Mo-O bonds, respectively. The surface structures formed for high coverages differ substantially from each other. Oxygen induces a missing row structure and is bound in a “quasi” three-fold coordinated site. In contrast, carbon is adsorbed in five-fold hollow sites at the surface and in six-fold coordinated octahedral subsurface sites.

The effect of adsorbed H on the Mo 0.95 Re 0.05 (110) surface has been investigated. Results obtained from low-energy electron diffraction (LEED), high-resolution electron energy loss spectroscopy (HREELS) and angle-resolved ultraviolet photoemission spectroscopy (ARUPS) are presented. A (2 × 2) LEED pattern is observed for H coverages around Θ ∼ 0.5 ML and is attributed to reconstruction of the substrate. At higher coverages, a (1 × 1) pattern is observed. Two peaks are observed at loss energies of 99 and 153 meV in the HREELS spectra for the H-saturated Mo 0.95 Re 0.05 (110) surface consistent with a quasi-trigonal adsorption site. Both peaks show an isotopic shift, confirming that they are due to hydrogen vibrational modes. A two-dimensional Fermi surface was determined using ARUPS for the H-saturated Mo 0.95 Re 0.05 (110) surface. Fermi-surface nesting vectors were observed at the momentum predicted by theoretical calculations for H-saturated Mo(110) where a phonon anomaly is observed.

The absolute coverages of atomic deuterium adsorbed on the Cu(111) surface at a temperature of 100 K have been determined using nuclear reaction analysis (NRA), and correlated with thermal desorption spectra (TDS) and deuterium-induced low energy electron diffraction (LEED) patterns for Cu(111). It was found that the adsorption of deuterium saturates at a coverage of 0.67 ± 0.03 ML, significantly lower than that observed for fcc(111) transition metal surfaces. The D coverage for a (2 × 2) LEED pattern was measured to be 0.5 ML, in common with those of H Ni (111) and H(D) Ag (111) . No bulk absorption was observed. The low saturation coverage of D on Cu(111) is attributed to the combination of a low heat of adsorption and D-D repulsion on the surface.

The temperature dependence of the equilibrium surface morphology of a stepped Pt(111) surface, miscut by 1.8° towards [112], has been studied by X-ray scattering techniques between 300 and 1900 K. Below Tc = 1389 K, the surface is found to be uniformly stepped and unreconstructed. Above Tc, the surface transforms into a periodic array of wide terraces and narrow terraces. Unlike macroscopic facets, the wide terraces are mesoscopic in extent, ranging from 230 A at 1389 K to 94 A at 1900 K. Reflectivity measurements suggest that multiple-height steps are not prevalent. The wave vector characterizing the array (δ) follows a power law versus reduced temperature [t = (T - T c ) T c ] : δ = δ 0 t \ gb with \ gb = 0.66 ± 0.05 .

The structure and composition of the (100) and (110) oriented surfaces of the ordered alloy FeAl were investigated by LEED and AES. Following sputtering and annealing, c(2×2) and (1×1) LEED patterns are observed on the (100) surface. The c (2 × 2) is attributed to an ordered arrangement of Al atoms on top of a complete Fe layer, while the ( 1 × 1 ) corresponds to the Al terminated surface, as determined previously. On the ( 110) oriented surface, a ( 2 × 1 ) diffraction pattern, an incommensurate superstructure, and severe streaking are observed with LEED upon annealing. For the (2 × 1) structure, a missing-row model is compatible with a stoichiometry of Fe2Al. The incommensurate superstructure and streak phase are associated with a near-surface aluminum concentration > 0.5, indicating segregation of Al. In contrast to the NiAl (110) surface, no (1 × 1) , which would indicate the formation of the bulk-like terminated surface, is observed with LEED. Annealing both surfaces over 800°C leads to evaporation of Al.

The electronic band structure of the monolayer graphite/TaC(111) system was examined using XPS, AES, and ARUPS and the results were compared with predictions of the rigid-band model to determine its validity. The measured graphite σ1 band energy relative to the Fermi energy is −23.0 eV at \gG and the π1 band energy is −10.3 eV at \gG and −3.4 eV at K K. By comparing these values to the results of a linearized augmented-plane-wave calculation of the electronic structure of an isolated graphite monolayer, the qualitative picture of charge transfer from the substrate into the graphite overlayer is confirmed, but the rigidband model is not sufficient to explain the observed binding energy differences between experiment and theory. The degree of hybridization between the electronic bands of TaC and the graphite monolayer is therefore not negligible and should be included in descriptions of the electronic band structures of such systems.

We have determined absolute coverages of deuterium on Cu(100) using the reaction D(3He, p)4He and found a saturation coverage of one deuterium atom per surface copper atom. LEED observations reveal no superlattice structures below $ 0.45 ML and four different superlattice structures above this coverage.

We present and apply an extension of the perturbation method tensor LEED in order to handle surface stoichiometry, i.e. the substitution of surface atoms by chemically different species. This avoids the need for new full dynamical calculations for surfaces of different chemical composition and allows fast access to their LEED intensities. This “chemical tensor LEED” is applied to the chemically disordered alloy Mo x Re 1− x (100) for x = 0.75, 0.85 and 0.95 for which LEED intensities were measured. Oscillatory layer dependent segregation is found in agreement with conventional LEED analyses. Even small Re concentrations strongly modify the structure of the pure Mo(100) surface, i.e. the surface reconstruction is lifted and multilayer relaxation effects are increased.

A simple model is established to describe diffraction phenomena (LEED) from a unidirectionally reconstructed TaC(110) surface. By comparing diffraction data with results of detailed simulations, the reconstruction is found to consist of alternate (100) and (010) facets propagating periodically along the [110] direction like an irregular ridge-and-valley grating, with (110) terraces completely absent. The irregularity is due to variation in facet sizes, and the mean separation is determined to be L = 6 lattice spacings. The validity of this model is supported by STM imaging results which confirm the structure obtained from the LEED-data analysis.

We summarize the results of x-ray scattering studies of the Au(001) and Pt(001) surfaces between 300 K and their respective bulk melting temperatures (T{sub m}). Both surfaces exhibit three distinct structural phases. At high temperatures (0--88 T{sub m} < T < T{sub m} ) both surfaces are disordered. At a temperature of T/T{sub m} {minus} 0.88 there are reversible phase transformations to incommensurate, corrugated-hexagonal phases. Below T/T{sub m} {minus} 0.8 hexagonal domains rotate with respect to the substrate orientation. In Pt, the rotational transformation is continuous and shows mean-field behavior. In Au, the rotational transformation is mainly discontinuous.

Using medium‐energy ion scattering with channeling and blocking, we have investigated the (1×2) reconstructions of Ni(110) and Au(113) (stabilized by submonolayer amounts of K and Ca, respectively). Our data contain evidence that the two surface structures correspond to missing‐row type reconstructions. Monte Carlo simulations show that the first interlayer spacing of the (1×2)‐Au(113) surface undergoes a large contraction (−18.6%), while the first interlayer spacing of (1×2)‐Ni(110) undergoes a small, unexpected expansion (+3%). Although the symmetry of Au(113) allows a lateral registry shift, none is found. Our structural models are discussed in terms of the energy balance between competing surface electronic charge densities.

We present N7VV (4f7/25d5d) Auger spectra from TaC(111) taken in time coincidence with photoemitted bulk and surface‐shifted 4f7/2 core‐level electrons, and compare these data with self‐convolutions of calculated valence band d densities of states (SCDOS) for the surface (first) and bulk (exponential sum of second, fourth, and sixth subsurface) atomic layers of a 13‐layer slab. The Ta 4f levels of TaC(111) and Ta(100) both exhibit large surface core level shifts, but of opposite sign. Therefore, we compare the TaC(111) data and calculations with analogous N7VV Auger coincidence spectra and calculated SCDOS that we have previously published for Ta(100). The bulk Ta(100) and surface TaC(111) spectra show favorable agreement with the calculations; the surface Ta(100) and bulk TaC(111) spectra, on the other hand, exhibit marked disagreement with the calculations. These discrepancies are consistent with and provide added support for the presence of our proposed1 surface‐to‐bulk (bulk‐to‐surface) interatomic Au...

Different submonolayer phases of bismuth on copper (100) have been investigated with low‐energy electron diffraction and Auger electron spectroscopy. The room temperature observations, made after depositing Bi onto well annealed Cu, agreed with those reported previously. Below room temperature, a low coverage c(2×4) phase was observed. With increasing Bi coverage, this phase was observed to coexist with the c(2×2) phase, until finally, only the c(2×2) phase was observed. Restructuring of the Cu substrate into {210} facets, at 26°34’ angle with the original (100) surface, was obtained by depositing Bi onto roughened Cu(100), and subsequently annealing. A possible explanation for the nucleation and growth of these facets is given.

X‐ray diffraction is used to study the phase behavior and structure of the Au(001) surface between 300 K and the triple point. For T>1170 K, the x‐ray reflectivity is consistent with a thin, disordered surface film. At T=1170 K, there is a reversible transition to an incommensurate, hexagonal overlayer that resides on top of the square substrate. At 970 K, the orientation of the hexagonal layer rotates discontinuously by 0.81 ° with no observable change in incommensurability.

Using angle-resolved ultraviolet photoemission spectroscopy with synchrotron radiation, the absolute work functions of four (100) surfaces having estimated compositions TaC1.0, TaC0.5, HfC1.0, and HfC0.6 were determined to be 4.38, 4.73, 4.63, and 3.87 eV, respectively. The results are interpreted qualitatively by considering the work function as a sum of two terms, one which arises from the presence of a surface dipole layer and another which is associated with the Fermi level. The difference in work-function behavior (its increasing with carbon content x for HfCx but decreasing with x for TaCx) is shown to be associated with a difference in Fermi-level behavior.

Chemisorption of nitrogen on Cu(110) has been studied using XPS, LEED, AES, and TDS. Gas phase nitrogen was activated and accelerated to 200 eV to promote interaction with the surface. Coverages were obtained from XPS N 1s areas and calibrated using a p(2 × 1)O structure. After exposures at 300 K, nitrogen coverages saturated near 1 ML. The width of the N 1s core indicates that only one site is occupied for coverages below 2 3 ML, but additional site(s) are occupied at higher coverages. Annealing the crystal to 650 K produced a stable (2 × 3) structure consisting of 2 3 ML of nitrogen in a single site. The lineshape of the nitrogen Auger signal is independent of coverage or surface structure. Annealing the crystal to 800 K removes all nitrogen from the surface, by a combination of desorption and absorption into the bulk.

We describe our recent X-ray scattering studies of the structure and phases of the clean Au(111) and Pt(111) surfaces. Below 0.65 of their respective bulk melting temperatures, the Au(111) surface has a well-ordered chevron reconstruction and the Pt(111) surface is unreconstructed. Above these temperatures, both surfaces reconstruct to form layers that are isotropically compressed and have only short-range order. Throughout their reconstructed phases, the densities of the Au and Pt(111) surfaces increase with increasing temperature.

The results of X-ray scattering studies of the (111) and (001) surfaces of Au and Pt are summarized. Each of these surfaces exhibits, in some temperature range, a reconstruction which approximates a close packed hexagonal layer, denser than the (111) planes of bulk Au or Pt. The results are discussed in the context of recent theories of surface reconstruction and of two-dimensional phase transformations.

LEED observations and XPS measurements were used to study Cu(110)c(6×2)-O prepared by two procedures: (i) exposing Cu(110) to about 5×104 L of molecular oxygen at a surface temperature Ts in the range 300⩽Ts⩽320K, and (ii) exposing the surface to as little as 1.9 L of molecular oxygen at Ts=83±1K, and then annealing it to between about 230 and 530 K. The resulting surfaces exhibit c(6×2) LEED patterns but appear to be different, having oxygen densities of 56 and 23 adatom per substrate atom for procedures (i) and (ii), respectively. Two possibilities are presented to account for this difference. The first is that the two surfaces consist of different phases. The second is that both surfaces consist of the same phase, the density difference being explained by adatom vacancies. Structural models are presented as candidates for the phases.

Angle‐resolved ultraviolet photoemission spectroscopy (ARUPS) has been used to study the p(2×1) oxygen‐induced reconstruction of Cu(110). While there is agreement between experimental groups that the O atoms sit in the long‐bridge site along the 〈001〉 Cu directions, there are two proposed models for the reconstruction: the missing‐row model, in which alternating 〈001〉 rows of Cu atoms are missing, and the buckled‐row model, in which the O‐containing 〈001〉 rows are displaced outwards from their bulk position. Our ARUPS data show oxygen‐induced changes in the Cu d‐band photoemission features that are inconsistent with the missing‐row model, thus favoring the buckled‐row picture. In addition, the O–Cu bond appears to be highly directional and oriented along the 〈001〉 Cu rows, consistent with O sitting in the long‐bridge site. Moreover, the O–Cu bond is primarily a result of the O(2pxy)–Cu(3dx2−y2) interaction.

XPS was used to study oxygen adsorption on Cu(110) and to follow changes in oxygen coverage θ resulting from isochronal anneals at up to 880 K. Clean surfaces or surfaces preexposed to oxygen and then annealed to effect adsorbate ordering were exposed at a surface temperature of 83 ± 1 K (denoted as T l or between 300 and 320 K (denoted as T h ) to up to 10 5 L of room-temperature molecular oxygen. Spectra for clean surfaces exposed at T l or T h exhibited peaks near 530.15 ± 0.2 eV BE having FWHM's of about 1.9 or 1.6 eV, respectively. These peaks are attributed to photoemission from 1s levels of chemisorbed oxygen. Normalized areas under these peaks were used to obtain curves of θ versus exposure and sticking probability versus θ. Adsorption at both T l and T h is describable in terms of a mobile-precursor kinetics, with initial sticking probabilities of about 0.77 and 0.23, respectively. Chemisorption at T l was greatly inhibited by the presence of a preadsorbed layer of ordered atomic oxygen. The broadest O 1s XPS spectra remained broad even after anneals to 180 K but began to sharpen during anneals at 230 K. Anneals at up to 480 K produced no measurable change in θ. Higher-temperature anneals reduced θ to about 0.50 if θ ≳ 0.50 initially; no change in θ was observed if θ ≲ 0.50 initially.

Electron energy distributions have been measured for Fq+ ions (∽q×10 keV, q=2−8) interacting with a Cu surface at 5° grazing incidence. The measurements show evidence for K-shell excitation in Fq+ (q≤6) ions. The resultant inner-shell vacancy production is understood in terms of calculated molecular orbital diagrams which qualitatively describe the close collisions between the incident ion and the target atoms. When the K-shell vacancy is created in the collision, the mean KLL Auger electron energy shows a shift to higher energies relative to that observed for K-vacancies carried into the collision. This shift occurs since the KLL Auger decay for a vacancy created in the collision is constrained to take place late in the neutralization process, where the K- and L-shell energies differ from those in the less screened ion.

Carbon-K characteristic loss, carbon-KVV Auger, and valence-band X-ray photoelectron spectra were measured for a well-ordered, nearly stoichiometric (100) surface of tantalum carbide. These spectra compare well with curves that were generated by using a calculated total density of states.

LEED observations were used to study oxygen adsorbed on Cu(110) and to follow changes in surface structure produced by 5 min anneals at up to 880 K. Factors determining the surface structure include the surface temperature Ts at exposure, the amount of exposure, and the maximum Ts after exposure. For surfaces exposed at Ts = 83 ± 1 K to between 0.3 and 100 langmuir (L) of room-temperature molecular oxygen, the LEED patterns upon exposure and after anneals are described for the first time in a detailed and systematic way. Upon exposure the surface exhibits a (1 × 1) pattern, although diffuse third-order reflections corresponding to an increase in the [110] dimension of the unit mesh are visible for coverages θ ≳ 0.50. A (3 × 1) pattern, which may be due to a molecular configuration, is seen after annealing at 180 K. After annealing at 330 K or higher, a simple relationship exists between oxygen coverage and surface structure, as inferred from LEED patterns; if θ ≲ 0.50, a (2 × 1) pattern is seen; if 0.50 ≲ θ ≲ 0.67, a superposition of a (2 × 1) and c(6 × 2) pattern is seen; if θ ≅ 0.67, a c(6 × 2) pattern is seen; and if θ ≳ 0.67, a c(6 × 2) pattern with additional reflections is seen. For surfaces exposed at 300 ≲ Ts ≲ 320 K to up to 105 L oxygen, the LEED patterns are similar to those reported by others; here too, however, previously unreported behavior is presented. The observations are interpreted in light of UPS and recent XPS data, which give information regarding the chemical state (atomic or molecular) and amount of the adsorbate.

An investigation of (100) surfaces of TaC and HfC was performed using low‐energy electron diffraction (LEED) and angle‐resolved ultraviolet photoemission spectroscopy with synchrotron radiation. Detailed LEED I–V analyses show that, on both surfaces, carbon atoms displace outward and metal atoms inward, giving rise to (1×1) rippled relaxations, with the magnitude of the ripple being larger on TaC(100). Photoemission spectra of metal 4f levels from these surfaces show no obvious surface core‐level shifts. Using x‐ray photoemission spectroscopy (XPS) and Auger electron spectroscopy (AES), changes in (bulk) core‐level binding energies, valence‐band structures, and local carbon environments were determined for TaCx (0.5

A photoelectron diffraction study of Ta(100), employing the surface shifted 4 f core levels, has been performed to determine its first interplanar spacing. Photoemission intensities of the bulk and surface shifted 4 f core levels were measured as a function of photon energy at normal emission and near the 〈111〉 symmetry direction of the crystal. The binding energy of 21.7 eV (23.6 eV) for the bulk 4 f 7/2 (5/2) core level, and the large surface core level shift of 0.75 eV, result in well‐resolved peaks in the photoemission spectra with areas that are easily measured independently. We find that the ratio of the surface to bulk core level intensities at normal emission exhibits large oscillations as a function of photon energy due to diffraction effects. Intensity modulation is also seen in the off‐normal intensities, but the magnitude is much smaller. We have performed a multiple‐scattering photoemission calculation for several first interlayer spacings ranging from the bulk value to a 15% contraction. The...

Auger and ionization loss spectra have been obtained from clean and Si-contaminated Be surfaces both before and after oxidation. The Auger spectrum from the clean Be surface exhibits only one major minimum at 105 eV in the d N ( E )d E versus E spectrum. Si-contaminated Be surfaces produced corresponding spectra with two major minima at 92 and 105 eV. Partial oxidation of clean Be surfaces resulted in Auger spectra which resemble in kind those obtained from Be-Si surfaces. Following extensive oxidation, Auger spectra from both types of surfaces were indistinguishable, and Si L 23 level ionization losses could no longer be detected.

Two satellite peaks have been observed on the high energy side of the Be KVV Auger peak. The lower energy satellite is attributed to coupling of energy from bulk plasmon de-excitations with Auger electrons, and the higher energy event to Auger electrons ejected from Be atoms with doubly ionized K levels. Following oxidation, the ionization loss spectra of BeO were observed to have structure which is interpreted as being related to the density of unfilled electron states above the BeO valence band. In addition, the characteristic loss and the low energy (“true secondary”) spectra of Be and BeO were determined. Peaks in these spectra are discussed in terms of characteristic energies related to excited electron states in the solids.