W.D. Robertson

Surface sensitivity and angular dependence of X-ray photoelectron spectra☆

Abstract By using an ultra-high vacuum X-ray photoelectron spectrometer designed specifically for surface studies, the surface sensitivity (ratio of surface to volume signal) of the technique was found to be enhanced by approximately one order of magnitude at low angles of electron emission measured from the surface plane. While the Mo(3d 5 2 ) peak intensity from a clean polycrystalline target of molybdenum was independent of angle, the Cs(3d 5 2 ) intensity from an adsorbed cesium layer on molybdenum increased and the Mo(3d 5 2 ) signal from the molybdenum substrate decreased as the angle of emission was decreased. A model is presented to account for the angular dependence and it is proposed that the angular dependence of spectral intensities can be used to differentiate surface from bulk species.

The order-disorder transformation at a (100) surface of Cu3Au: Theory and experiment☆

Abstract The temperature dependence of long range order at a (100) surface of Cu3Au has been determined by low energy electron diffraction. Auger electron spectroscopy revealed that the composition of the alloy near the surface is changed by ion bombardment, but the stoichiometric composition is reconstituted by annealing. The order parameter of the surface, which was obtained from the temperature dependence of a superlattice beam, appears to be a continuous function of temperature, unlike the bulk order parameter. The disordering process at the surface begins about 60°K below a critical temperature which is common to both the surface and the volume. Monte Carlo calculations of the order parameter as a function of temperature for a model incorporating nearest- and next-nearest-neighbor interactions reproduce the experimental results and also provide an evaluation of the change in order parameter as a function of layer number for boundary conditions corresponding to a copper or a copper-gold surface.

X-ray photoelectron spectroscopy of oxygen adsorption on clean Ni(100) surfaces

We have recorded photoelectron spectra of oxygen (1s) and nickel (2p32) and measured work functions of a Ni(100) surface as they depend on exposure to oxygen from 1 to > 500 × 10−6 Torr sec. O(1s) intensity as a function of exposure shows a fast chemisorption at low coverage, a region of slow oxygen adsorption in the 14 monolayer region [p(2 × 2) LEED periodicity] followed by an increasing rate of adsorption, and finally saturation. Sticking coefficients estimated from these data vary with coverage by more than an order of magnitude. Two adsorbed states of oxygen are apparent in the O(1s) spectra; a low coverage state with EbF = 531.6 eV and a high coverage state at EbF = 530.0 eV, but the Ni (2p32) spectrum does not reflect the presence of adsorbed oxygen until the coverage exceeds one monolayer. Comparison of the vacuum referenced binding energy in the condensed state (EbV = EbΓ + φ) to Evb for gas phase atomic oxygen shows that Evb[O(1s)] is reduced by 6.3–10.2 eV on adsorption. This change in measured binding energy is due in part to extra atomic relaxation (polarization) and in part to electron transfer to oxygen, both of which lower the binding energy. The shift of O(1s) binding energy at low oxygen coverage compared to that at saturation, however, appears to reflect increased charge transfer at saturation.

Low energy electron diffraction profiles from aluminum (100): Reproducibility and an evaluation of intensity averaged at constant momentum transfer

Abstract LEED intensity profiles have been obtained from the clean (100) surface of aluminum at small angular and energy intervals for azimuthal angles (0° to 45°) and incident angles (5° to 21°). These data have been compared with the data of Jona, and of Burkstrand and Propst, for reproducibility of independent experiments. The effect of oxygen on intensity, line width and the position of prominent peaks has been evaluated. Averages of intensity over azimuthal rotation, at two angles of incidence, and averages at one azimuth for increments of 2° in angle of incidence have been taken to evaluate the prescription of Lagally, Ngoc and Webb for extracting structural (geometric) information from multiple scattering profiles when applied to aluminum (100), instead of the original evaluations on (111) surfaces of nickel and silver. The principal conclusions are: (1) comparison of data from independent measurements of profiles from clean (oxygen-free) aluminum surfaces indicates that reproducibility of peak position and intensity is comparable with agreement obtained between independent computations of intensity profiles; (2) prominent multiple scattering maxima in profiles from Al (100) are not removed by averaging intensity in the range of incident angle available in most display-type LEED instruments, but a “kinematic-like” average is obtained from the (111) plane of aluminum; (3) comparison of the results of averaging intensity profiles from different metals and different crystal planes shows that additional exploration and a priori justification of the procedure are required to achieve a reliable (approximate) method of extracting surface structure from multiple scattering profiles, in the general case.

An X‐Ray Photoelectron Spectrometer Designed for Surface Research

An x‐ray photoelectron spectroscopy (ESCA) apparatus designed specifically for surface research is described. The features, included in one vacuum envelope, are ultrahigh vacuum, a high intensity x‐ray source, a 10 cm mean radius hemispherical electron energy analyzer, sample cleaning by argon ion bombardment, sample heating by electron bombardment, and the facility for the collection of electron spectra as a function of angle relative to the surface plane of the sample. It is shown that (1) sample cleaning increases net signal strengths by factors as great as 14; (2) less than 2% of a monolayer of cesium on a clean molybdenum substrate can be readily detected; and (3) that ultrahigh vacuum is essential for investigation of well‐defined surfaces. Surface sensitivity is increased by an order of magnitude when spectra are collected at low electron take‐off angles relative to the surface plane, and the importance of measuring angular dependence of spectra for interpretation of data from surfaces is demonstrated.

Electron binding energies of core levels in caesium adsorbed on a nickel (100) surface

Electron binding energies obtained from X-ray photoemission and workfunctions obtained from UV photoemission have been recorded for caesium adsorbed on nickel (100) at coverages from 0.1 to >1.0 monolayer. Analysis of coverage as a function of Cs+ dose suggests a sticking coefficient of 0.5 for Cs+ on nickel (100) in submonolayer range and 0.05 above a monolayer of deposited caesium. The binding energy of the caesium 3d5/2 level, referred to the Fermi energy of the adsorption system, is 726.0+or-0.2 eV, independent of coverage. Comparison of binding energies of atomic caesium and caesium adsorbed at saturation coverage shows that the total change in electron binding energy, Delta EB, depends on the core levels studied (5s to 3s); subtracting a modified relaxation energy yields a solid state shift in binding energy of 2-4 eV.

LEED from Re(0001); And in comparison, from Be(0001)

Abstract Intensities of five beams of low-energy electrons diffracted from the (0001) surface of rhenium (Z=75), are presented over a wide range of energy and angle of the incident beam. Comparison of these intensities with Bakers intensities from the (0001) surface of beryllium (Z=4), which has the same hcp lattice and nearly the same c a ratio as rhenium, reveals no apparent qualitative agreement, as might be expected because of the very different scattering characteristics of the two atoms. However, quantitative comparison of the specular-beam intensities near normal primary-beam incidence in the energy range of ∼50–200 eV, in the context of s-wave calculations with the Beeby-Duke-Tucker inelastic-collision model, reveals a shift from dominant inter layer scattering effects to dominant intralayer scattering effects as the atomic-scattering cross-section increases.

Temperature dependence of specular and non-specular leed beams from Cu3 Au

Abstract The temperature dependence of intensities of the prominent diffraction peaks in the specular (00) and one non-specular ( 11 ) beam from the (100) surface of Cu3Au have been measured in the range of 300 to 673 K. The effective Debye temperature associated with the specular beam appears to increase continuously with energy below 50 eV but varies discontinuously at high energies. For the ( 11 ) beam, which was available only in the higher energy range of 65 to 136 eV, the effective Debye temperature varies discontinuously with energy. Parallel and normal components of the Debye temperature were deduced from the two sets of data from which it appears that the two mean square displacements are approximately equal, compared to the harmonic approximation which indicates a difference of 30 percent. The log of peak intensity versus temperature (Debye plot) deviates from the straight line at 60° below the disordering temperature for all beams and all energies.

Constant Stress Creep Apparatus

A simple creep apparatus for approximating a constant stress in tensile specimens of large cross sections, where large applied loads are required, is described. The apparatus consists of a lever with a large built‐in mechanical advantage, together with means for removal of incremental weights from the applied load.