Search

X-ray phtoemission spectra of palladium prepared by vacuum evaporation and by argon ion sputtering are compared. The former show a close correlation with calculated band structures but the latter exhibit much less detail, strongly suggesting that the sputtered surface is amorphous. The narrow core lines of the evaporated samples show the asymmetry predicted to arise from the hole-conduction electron interaction.

Born-Haber cycles which relate core-electron binding energies of adsorbed isolated atoms and monolayers to adsorption enthalpies and monolayer cohesive energies are used to interpret the measured shifts for Au on Ag, Au, and Pt substrates and to deduce the cohesive interaction within the adlayer. For one monolayer of Au on Ag(111) the core-electron binding-energy shift is negative, but smaller in magnitude than that of surface atoms on bulk gold. The 5d band of the Au monolayers is, however, significantly narrower than that of the first monolayer on bulk gold, but not as much as expected on the basis of the reduced gold coordination number.

Core-line asymmetries have been observed in the x-ray-photoemission spectra of a number of metals. They are ascribed to the interaction of the suddenly created potential of the photohole with the conduction electrons. A comparison of the experimental line shapes with those predicted by a theoretical calculation gives reasonable agreement. The procedure makes it possible to measure the term that quenches the x-ray absorption threshold singularity and shows that it can be substantial.

The charge transfer in HfPt{sub 3} is discussed on the basis of the 4{ital f} core-electron binding-energy shifts of the constituent elements. Based on binding energies measured relative to the Fermi levels, both Hf and Pt have positive shifts in the compound. According to a Born-Haber cycle analysis these shifts are directly attributable to the large enthalpy of formation of HfPt{sub 3}. Corrections for the differences in reference level and final-state screening between the element and the compound make the shift of the Hf larger and that of the Pt smaller, and presumably negative, indicating a net flow of electronic charge from Hf to Pt. This is in accord with expectations based on the electronegativity difference and with the semiempirical model of Miedema {ital et} {ital al}., but counter to the direction of charge flow in the Engel-Brewer analysis of bonding.

Core-electron binding energies of atoms and monolayers adsorbed on metals are expressed in terms of atomic ionization potentials and chemical or thermodynamic quantities using various Born-Haber cycles. The relationship of these parameters to the microscopic quantities commonly used to interpret binding energy shift is explored. This type of analysis is applied to the transition from the isolated adsorbate atom to full monolayer coverage, and from mono- to multi-layer.

The effects of structure in the density of electron-hole pair excitations is examined by numerical integration of the many-body screening formalism. The results are applied to Ir and Pt.

We report XPS experiments on Ag 0.50 Cd 0.50 in different crystallographic modifications. Our results confirm earlier conclusions that the apparent splitting of the Cd 4d band is mainly caused by d—d overlap rather than by crystal field interactions.

X-ray photoemission spectroscopy (XPS) data for the valence bands of Rh, Pd, Ag, Ir, Pt, and Au are compared with densities of states calculated from interpolated band structures. The parameters of the interpolation scheme were obtained in an earlier paper by fitting first-principles augmented-plane-wave calculations, and then making adjustments, where necessary, to the width of the $d$ bands in order to improve agreement with ultraviolet photoemission spectroscopy (UPS) data. In the $4d$ metals, Rh, Pd, and Ag, there is excellent agreement between the energy positions of peaks in the occupied density of states and the XPS valence bands. In the $5d$ metals, Ir, Pt, and Au, the agreement is reasonable, but inferior to that for the $4d$ metals particularly in the middle of the $d$-band region. It is argued that the main discrepancies in peak locations are not significant and are most likely due to the approximations in the interpolated band structures, particularly those involved in the spin-orbit splitting and hybridization. The relative intensity of the XPS data in the lower-energy region of the $d$ bands is consistently lower than that in the density of states, indicating appreciable modulation of the spectra by optical-transition strengths.

Mossbauer spectroscopy, X-ray photoemission spectroscopy and magnetization studies of EuCu2−xFexSi2, EuCu2Si2−xGex, EuCu2−xSi2+x, and EuNi2−xSi2+x have been performed. In EuFe2Si2 and EuNi2Si2 the Eu ion is trivalent, in EuCu2Si2 it is of intermediate valency. In all systems withx=0, many inequivalent Eu sites of intermediate valency are formed. In EuCu2−xFexSi2, the average valency of Eu moves nonmonotonically toward Eu3+. In EuCu2Si2−xGex, EuCu2−xSi2+x and EuNi2−xSi2+x, the Eu average valency moves quickly towards Eu2+. In all three systems, whenx=1, all Eu ions are already stable divalent. The systems EuCuSi3 and EuNiSi3 order magnetically at 40 K and 34 K, and exhibit hyperfine fields of 323 kOe and 458 kOe, respectively. XPS and Mossbauer isomer shift determinations of the average Eu valence as a function of temperature in EuCu1.5Fe0.5Si2 are in good agreement. The experimental observations concerning the Eu valencies can not be explained solely in terms of the local volume available to the Eu ion; the nature of chemical environment plays a dominant role.

Core-electron, valence-, and conduction-band spectra of the cubic sodium-tungsten bronzes ${\mathrm{Na}}_{x}\mathrm{W}{\mathrm{O}}_{3}$ have been obtained by x-ray photoelectron spectroscopy. The samples were vacuum-cleaved single crystals spanning the entire cubic composition range. The band spectra are consistent with the theoretical band structure, and the filling of the conduction band with increasing $x$ is explicitly demonstrated. Sodium and oxygen core-electron spectra have binding energies fixed relative to the bottom of the conduction band and exhibit the plasmon satellites expected on the basis of optical and energy-loss measurements. The unique satellite structure of the tungsten $4f$ electrons is discussed in terms of an excitation of the conduction-band electrons during the photoemission process.

Carbon 1s X-ray photoemission spectra of stage two lithium graphite contain information about the distribution of Li ions in the intercalant layers and of π-electron charge in the graphite layers. The salient observation is an asymmetrical broadening of the C 1s line in the stage 2 compound which is a direct consequence of the existence of carbon sites unequivalent by virtue of their Li neighbor environment. Analysis using a lattice gas model indicates that the charge donated to the graphite π band remains localized near the Li ion. More detailed analysis of the data, however, shows that the Li ions are not strictly confined to the six-fold sites of the graphite layer.

It is shown that fractional parentage coefficients calculated in intermediate coupling provide a quantitative representation of the final-state population produced by photoemission from 4f13. Calculations in LS coupling show significant deviations for both 4f12 and 4f13.

X-ray photoemission spectra of the d-bands of Cu, Ag, Au, Rh, and Pd are compared with corresponding U.V. photoemission data and with APW band structure calculations. It is found that there is a reasonable correspondence between the two experimental techniques, as well as between theory and experiment.

High-resolution x-ray photoemission has been used to study the electronic structure of the Ce-pnictides (CeN, CeP, CeAs and CeSb). This series of isostructural compounds allows us to follow the evolution of the 4f level as the distance between Ce atoms varies. Core level spectra show clearly that the 4f state is localized in all the compounds. In the valence band region, the width of the 4f peak is strongly influenced by the energy overlap with the extended states originating from anion p states. The spectra of CeN reveal the existence of a mixed configuration.

Submonolayer coverages of iodine deposited on clean Ag(111) have been studied by x-ray photoelectron spectroscopy. The adsorbate core-electron binding energy is found to be 0.1-eV smaller at saturation than in the dilute limit, indicating that the iodine carries only a small negative charge. The low-energy electron diffraction evidence for clustering of the adatoms into the $\sqrt{3}\ifmmode\times\else\texttimes\fi{}\sqrt{3}$ structure at coverages considerably less than $\frac{1}{3}$ monolayer is confirmed by the observation, at those coverages, of a constant $14d$ binding energy. The excitation of silver surface plasmons by the core ionization of the adsorbate has also been observed.

Vacuum-cleaved single crystals of mixed valence TmSe were studied by x-ray photoemission. The divalent and trivalent part of the Tm $4f$ spectrum were fitted using optical multiplet energies and fractional parentage intensities. From the areas of the divalent and trivalent components we obtained the average valences for samples with four different stoichiometries. Comparison of these results with other estimates of the average valence suggest that the x-ray photoemission spectroscopy data are significantly affected by a divalent surface layer. However, attempts to demonstrate the reality of this layer by varying the take-off angle were inconclusive. It was verified that the ${\mathrm{Tm}}^{2+}$ ground state lies at the Fermi energy, giving a direct proof of the homogeneous character of the mixed valency. The location of this multiplet can therefore be used in general to distinguish between homogeneous and inhomogeneous mixed valency.

Epitaxial CoSi2 terminated with a thin Si bilayer has been studied by synchrotron radiation photoemission. This well‐defined surface phase of CoSi2 is obtained by high‐temperature annealing in the range 450–650 °C. The samples used were prepared by molecular‐beam epitaxy techniques with a film thickness of ∼800 A on Si(111) substrates and capped with a ∼100 A film of a‐Si to act as an oxide barrier while samples were transferred to the photoemission chamber. The surfaces studied by photoemission were prepared by careful sputtering to remove the oxide and a‐Si layer and then anealed to ∼450–650 °C for 1 min. Photoemission spectra were taken with total resolution in the range 200–237 meV and show evidence for four spin–orbit doublets. Our samples and results are somewhat different from those of previous studies of thin CoSi2 films or studies of bulk CoSi2 samples. We find three surface Si(2p) components: one shifted to higher binding energy and two shifted below the main Si(2p) spin‐orbit doublet in CoSi2. ...

By considering the results of M\"ossbauer studies and x-ray photoelectron spectroscopy, we obtain a phenomenological model of the intermediate valence state in the homogeneous system Eu${\mathrm{Rh}}_{2}$. The interconfiguration excitation energy is 1550 \ifmmode\pm\else\textpm\fi{} 150 K, and the mixing width, $2\ensuremath{\Delta}$, is 0.10 \ifmmode\pm\else\textpm\fi{} 0.02 eV. In $\mathrm{Eu}{\mathrm{Rh}}_{2\ensuremath{-}x}{\mathrm{Pt}}_{x}$ the Eu valence is determined by the local environment, including neighbors beyond the nearest shell.

The thermal shift of the $^{57}\mathrm{Fe}$ M\"ossbauer $\ensuremath{\gamma}$ ray in rhombohedral Fe${\mathrm{F}}_{3}$ gives a weak indication of a magnetization-dependent Debye temperature. A small isomer-shift discontinuity exists at the transition to cubic structure at high temperature.

The transition in Fe 2 Ti from ferromagnetism in iron-rich to antiferromagnetism in titanium-rich material is ascribed to excess iron atoms which enter titanium sites. The effects of the resulting ferromagnetic clusters on magnetic and Mossbauer measurements are discussed.

It is suggested that the valence-band structure observed in x-ray photoemission from certain insulating $3{d}^{n}$ compounds can be identified with the crystal-field states of the $3{d}^{n\ensuremath{-}1}$ final state.

The density of states of the AgPd disordered alloy system has been measured over the whole composition range by the XPS method. An analysis in terms of the coherent potential approximation is presented.

The recombination equations for a system containing an arbitrary number of Shockley-Read recombination centers are formulated. Transient solutions are obtained for the decay following the injection of a pulse of carriers into a system containing one or two centers. The specific cases considered include (1) the simple one-center case, which enables us to discuss (a) the situation for a large injection of carriers, (b) recombination through donors in $n$-type or acceptors in $p$-type material, and (c) recombination through centers in the presence of direct recombination; (2) the case of a recombination center with a temporary trap, and (3) the case of two recombination centers.

The $^{57}\mathrm{Fe}$ M\"ossbauer effect has been used to measure the sublattice magnetization $M(T)={M}_{0}D{(1\ensuremath{-}\frac{T}{{T}_{N}})}^{\ensuremath{\beta}}$ of Fe${\mathrm{F}}_{3}$ in the region $0.85l\frac{T}{{T}_{N}}l0.9995$. The behavior in the critical region yields $\ensuremath{\beta}=0.352\ifmmode\pm\else\textpm\fi{}0.005$ and $D=1.207\ifmmode\pm\else\textpm\fi{}0.015$. The critical exponent differs appreciably from those reported in Mn${\mathrm{F}}_{2}$ and Fe${\mathrm{F}}_{2}$ but is similar to those reported for liquid-vapor transitions and for other ferromagnets.

The lifetime of higher valence states produced by Auger effect can be obtained from the Mossbauer effect in the case of some favorable isotopes. It is shown that valence states higher than Fe2+ which are produced following the electron capture decay of divalent 57Co in ZnF2 have lifetimes less than 10−8 sec. A similar limit is obtained for the lifetimes of higher charge states of iron produced by divalent cobalt diffused into NaF with the aid of charge‐compensating sodium vacancies. Trivalent iron has lifetimes greater than 3×10−8 sec in NaF, if charge compensation is accomplished by oxygen. The lifetime in hydrated ferrous ammonium sulfate is 5×10−8 sec. The short lifetimes even of Fe3+ in some divalent lattices indicate that the lifetimes of higher valences will generally be too short to be observed with the present technique. It also suggests that valence states which do not occur in stable compounds will generally have short lifetimes and be difficult to detect.

The valence bands of Cu, Ag and Ni have been obtained by X-ray photoemission spectroscopy (XPS) with a resolution of ≈ 0.5 eV. Comparison with calculated densities of states show good agreement in the cases of Cu and Ag. For Ni satisfactory agreement was not obtained.

The resonance absorption of 14.4 keV gamma rays in Fe57 (Mossbauer effect) has been observed in Fe(III) and Co(III) acetylacetonate, the former used as an absorber against a source of Co57 diffused into metallic chromium, the latter as a Co57 labeled source against an K4Fe(CN)6 (38%Fe57) absorber. The Fe(III) acetylacetonate absorber spectrum shows a single broad line (Γ½=1.83±0.09 mm sec—1) with an isomer shift of 0.50±0.09 mm sec—1. The Co(III) acetylacetonate source shows a complex spectrum which can be resolved into a single peak assigned to Fe(III) and a quadrupole split pair of lines assigned to Fe(II). These data are interpreted in terms of the chemical consequences following K capture in Co57, and suggest a twofold preference for the stabilization of the higher charge state in the Co(III) acetylacetonate matrix. No evidence for ionic charge states higher than 3+ is observed, and it is assumed that these species—if formed as a consequence of the prior nuclear event—have been reduced to the observed states in times short compared to 10—7 sec.

High-field Mossbauer measurements show that the a-site iron in Fe2Ti has a vanishing magnetic moment. This accounts for the absence of a-site magnetic hyperfine splitting in the antiferromagnetic compound below the Neel temperature.

The hyperfine structure in the Mossbauer spectrum of Fe57 in ferrocene and in related compounds shows that the bonding of the iron atom is not affected by ring substitution. The temperature dependence of the fraction of recoil‐free absorption events in these compounds indicates that the Debye temperature is low (∼150°K) and that the optical modes make only a minor contribution to the reduction of the recoil‐free fraction at low temperature.

The magnetic properties of iron in cubic Laves-phase compounds having the Mg${\mathrm{Cu}}_{2}$ structure have been investigated using the ${\mathrm{Fe}}^{57}$ M\"ossbauer effect. The field at the nucleus was found to be 230\ifmmode\pm\else\textpm\fi{}5 koe in all compounds except Ce${\mathrm{Fe}}_{2}$, where two distinct hyperfine structures were resolved. The results indicate that the configuration of the iron atoms is the same in all these compounds and that the contribution to the hyperfine field by the conduction electron polarization is either very small or else dominated by the iron $d$ shell. The more complex behavior in Ce${\mathrm{Fe}}_{2}$ is associated with the transfer of the cerium $4f$ electron to the iron $d$ band. The isomer shift indicates that this electron is not localized on a particular iron atom, while the existence of two distinct hyperfine fields suggests that there may be a spatially periodic spin density fluctuation in the $d$ band.

Four related iron carbonyls, Fe(CO)/sub 5/, Fe/sub 2/(CO)/sub 9/, Fe/sub 3/(CO)/sub 12/, and Fe(CO)/sub 4/I/sub 2/, were used as stationary absorb ers in a Mossbauer experiment in which Co/sup 57/, diffused into metallic chromium, was employed as a source. All four absorbers showed significant resonance absorption of the 14.4-kev gamma -ray of Fe/sup 57/ at 78 deg K. The observed isomer shifts were interpreted in terms of the chemical bonding in transition metal-carbonyl compounds. The two major Mossbauer parameters (isomer shift and quadrupole splitting) in the case of Fe(CO)/sub 5/ and Fe/sub 2/(CO)/sub 9/ were in agreement with a trigonal bipyramid structure and a (3,3,3) structure, respectively. The data for Fe/sub 3/(CO)/sub 12/ strongly suggested the linear (3,3,3,3) structure and could not be accounted for by the trigonal arrangement of the three iron atoms, which was proposed on the basis of x-ray data. The quidrupole splitting for Fe(CO)/sub 4/I/sub 2/ was essentially independent of temperature in the range 78 to 298 deg K, in contrast to the large temperature dependences, which were observed for some related Fe(II) compounds. The isomer shift was consistent with an assignment of 2+ for the oxidation state of iron in this compound.

The Mossbauer effect, the resonant absorption of nuclear gamma rays in solids, may be used to obtain the hyperfine structure of Fe57 in magnetic materials. Experiments are performed by observing the absorption by stable Fe57 of the 14.4-kev gamma ray coming from a source which contains radioactive Fe57 produced by the decay of Co57. The experiments are not limited to naturally iron-bearing materials; other substances can be studied, provided only, that small amounts of cobalt can be introduced into lattice sites of interest. The magnetic moments of the ground and first excited states of Fe57 are known and make possible direct determination of the field at the iron nucleus once the hyperfine structure has been measured. The magnetic field at iron nuclei has been determined in the ferromagnetic transition metals (Fe 3.42×105 oe, Co 3.12×105 oe, Ni 2.80×105 oe at 0°K), but no hyperfine structure has been observed down to 4°K in the antiferromagnetic transition metals, Mn and Cr. In the case of yttrium-iron g...

The hyperfine structure of ${\mathrm{Fe}}^{57}$ in yttrium-iron garnet [${\mathrm{Y}}_{3}$${\mathrm{Fe}}_{2}$ ${(\mathrm{F}\mathrm{e}{\mathrm{O}}_{4})}_{3}$] has been obtained through the M\"ossbauer effect. A 0.002-in. thick slice of a single crystal of yttrium-iron garnet, cut normal to a [110] direction, was used as an absorber of recoil-free gamma rays emitted by a stainless steel source. The iron in yttrium-iron garnet is located in two nonequivalent sites, tetrahedral and octahedral, each of which has an axially symmetric electric field gradient. Data were taken with the magnetization aligned in a [111] and in a [100] direction in order to produce the simplest absorption pattern. For each direction of magnetization, the absorption lines of ${\mathrm{Fe}}^{57}$ at both sites have been resolved. The magnetic field at an ${\mathrm{Fe}}^{57}$ nucleus has been found to be 3.9\ifmmode\times\else\texttimes\fi{}${10}^{5}$ oe at a tetrahedral site and 4.7\ifmmode\times\else\texttimes\fi{}${10}^{5}$ oe at an octahedral site when the crystal is at room temperature (\ensuremath{\sim}300\ifmmode^\circ\else\textdegree\fi{}K). The quadrupole coupling was found to be 7.5\ifmmode\times\else\texttimes\fi{}${10}^{\ensuremath{-}8}$ ev in the tetrahedral site and 9.0\ifmmode\times\else\texttimes\fi{}${10}^{\ensuremath{-}8}$ ev in the octahedral site.

>Neutron-bombardment damage in silicon is compared to electron- bombardment effects which have been previously analyzed. A discrete energy level 0.27 ev above the valence band previously ascribed to the acceptor member of a defect pair having a separation greater than 50 A, is found to be produced by both neutrons and electrons. A spectrum of energy levels ruaning from 0.16 ev below the conduction band toward the middle of the gap is ascribed to a defect pair with variable spacing, and related to the discrete level 0.16 ev below the conduction band which was found previously in electron-bombardment of silicon, and there ascribed to a close-spaced pair of defects. Lifetime effects are found to be dominated by levels near the middle of the energy gap, which may be related to the above spectrum. A very rapid decrease in mobility at low temperature is ascribed to bombardanent-induced imhomogeneities. (auth)