Core Hole States Research Articles

Observation of post-collision interaction (PCI) in HBr using two-dimensional photoelectron spectroscopy

The creation and subsequent Auger decay of the 3d-1 core-hole state in HBr is studied using two-dimensional photoelectron spectroscopy. In this technique the yield of electrons is collected as a function of both photon and electron kinetic energy. The effects of post-collision interaction (PCI) observed in the energy shift of the HBr+(3d-1) photoelectron lines. The energy shifts are compared with previous experimental results on Kr. This comparison suggests that the molecular nature of HBr has little effect on the PCI energy exchange mechanism.

Spin-orbit interaction and molecular-field effects in theL2,3VVAuger-electron spectra of HCl

Ab initio calculations based on quantum chemical methods and the one center approximation have been carried out for the ${L}_{2,3}\mathrm{VV}$ normal Auger-electron spectrum of HCl. Spin-orbit and molecular field effects were explicitly included in the description of the intermediate $2p$ core hole state. The results are compared with the experimental spectrum obtained with synchrotron radiation. The corresponding atomic ${L}_{2,3}\mathrm{MM}$ Auger transitions are studied in the isoelectronic argon atom. The calculated Auger electron spectra are in good agreement with the experimental and to previous theoretical results. For HCl the calculations predict substantially different total Auger transition probabilities of 126, 99, and 113 meV for the three nondegenerate spin-orbit and molecular-field-split ${2p}^{\ensuremath{-}1}$ states: ${2p}_{3/2}^{\ensuremath{-}1}{(}^{2}{\ensuremath{\Pi}}_{3/2}),$ ${2p}_{3/2}^{\ensuremath{-}1}{(}^{2}{\ensuremath{\Sigma}}_{1/2}^{+}),$ and ${2p}_{1/2}^{\ensuremath{-}1}{(}^{2}{\ensuremath{\Pi}}_{1/2}),$ respectively. Furthermore, each of these core hole states gives rise to remarkably different intensity distributions. These effects are explained by (i) the partial orientation of the chlorine $2p$ core hole by the molecular field, (ii) a decrease of the net population of the chlorine $3p$ orbital in the bond direction due to the chemical bond, and (iii) the clear (97%) preference of the Auger decay of an oriented $2p$ core orbital to produce final states with at least one hole in the $3p$ orbital with the same spatial orientation.

Local magnetic moment coupling of Gd on Fe(100) studied by magnetic dichroism in angular-dependent photoemission

We measured the magnetic linear dichroism in the angular distribution (MLDAD) of photoemission of thin Gd layers on Fe(100). At low photon energies large MLDAD asymmetries, up to 40%, in the $\mathrm{Gd}4f$ photoemission were observed. The line shape and the photon-energy dependence of the measured MLDAD are in good agreement with theoretical results. Analysis of the $\mathrm{Gd}4f$ and $\mathrm{Fe}3p$ magnetic signals indicates an antiferromagnetic coupling between Gd and Fe, confirming previous findings. We also demonstrate that the MLDAD plus-minus feature is governed by the orbital magnetic moment of the core hole state.

Interpretation of the N2,3N4,5O2,3 Coster–Kronig spectrum of xenon

Abstract The N 2,3 N 4,5 O 2,3 Coster–Kronig spectrum of Xe has been measured with synchrotron radiation. It arises from the decay of the 4 p core hole states that are split by strong configuration interaction to numerous close-lying components. Their binding energies are obtained from a high-resolution 4 p photoelectron measurement. The interpretation of the Coster–Kronig spectrum is further hampered by the lack of accurate binding energies for the 4 d −1 5 p −1 final states. The latter are, however, also reached in the M 4 , 5 N 4 , 5 O 2 , 3 Auger spectrum which has been measured in this work with electron impact ionization and assigned with the aid of calculations. With this new information the N 2 , 3 N 4 , 5 O 2 , 3 Coster–Kronig spectrum can be interpreted semi-empirically. It is also compared with calculations.

Selective bond rupture in nitrous oxide stimulated by core electron excitation

Dissociation of nitrous oxide following selective excitation of a 1 s electron either from the terminal nitrogen atom or from the central nitrogen atom to the 3π orbital is observed by single-event energy-resolved Auger-electron–multiple-ion coincidence mass spectroscopy. The ionic products, observed by time-of-flight mass spectroscopy, are correlated with the valence–hole states, which are reached by autoionization of the core-hole excited state, and with the energetically allowed ionization and fragmentation channels. Selectivity in bond rupture resulting from the atomic-site selected excitation is observed. This selectivity results from the different populations of the states reached in the Auger-like decay of the core holes and not from a proximity effect.

Resonant auger processes in adsorbates

We will present results of a very recent study of the electronic decay properties of resonant adsorbate core hole states performed under high resolution conditions, i.e. excitation with subnatural linewidth. Under these conditions two types of resonant Auger processes after core-to-bound excitation are observed for adsorbates, a coherent resonant `Raman'-type channel and an incoherent `normal' Auger-type channel. Their branching ratio can be used as a quantitative measure for the (sub-) femtosecond dynamics of charge transfer from an adsorbate resonance to the underlying substrate. Thereby the core hole lifetime serves as an `internal' clock for measuring the dynamics of relaxation phenomena.

Symmetry-forbidden magnetic circular dichroism in Auger emission

The Fe L 3 M 2,3 M 2,3 Auger spectrum measured with the light helicity vector perpendicular to the magnetization direction shows a strong magnetic circular dichroism (MCD). Such an effect is not expected in a two-step model when the photoelectron is excited into a non-interacting continuum state and appears to be absent in strongly correlated materials, such as nickel. We tentatively explain this phenomenon by a spin-dependent screening of the intermediate core-hole state. Supporting evidence for a breakdown of the independent-particle model for iron is found in the observed MCD of the Fe 2p photoemission.

X-ray magnetic circular dichroism of multielectron excitation by fluorescence spectroscopy

Abstract Fluorescence spectra of a Co67Gd33 amorphous thin film were measured for the Gd Lα1,2 lines using circularly polarized X-rays of energies tuned at 134, 154, and 174 eV above the Gd L3-edge (7244 eV). Multielectron excitation (MEE) effects and their magnetic circular dichroism (MCD) for the Gd Lα1,2 lines were investigated. The MEE effect for Gd Lα1 appeared as superimposed over the original fluorescence line, and the peak shift due to a two core hole state could not be detected.

Calculation for the charge-transfer effect of La compounds in the3d−1core-hole state

We calculate the electronic structures of La halides and oxides from a first-principles calculation by the spin-unrestricted DV-$X\ensuremath{\alpha}$ molecular-orbital method of model clusters. The model clusters we use are $[{\mathrm{LaO}}_{6}{]}^{9\ensuremath{-}},$ $[{\mathrm{LaF}}_{6}{]}^{3\ensuremath{-}},$ $[{\mathrm{LaCl}}_{6}{]}^{3\ensuremath{-}},$ and $[{\mathrm{LaBr}}_{6}{]}^{3\ensuremath{-}},$ as model clusters of ${\mathrm{La}}_{2}{\mathrm{O}}_{3},$ ${\mathrm{LaF}}_{3},$ ${\mathrm{LaCl}}_{3},$ and ${\mathrm{LaBr}}_{3},$ respectively. The electronic structures of these clusters are calculated for the ground state and the ${3d}^{\ensuremath{-}1}$ core-hole states. The ${3d}^{\ensuremath{-}1}$ core-hole states can be regarded as the final states of La $3d$ x-ray photoemission spectroscopy (XPS), and the creation of ${3d}^{\ensuremath{-}1}$ core-hole states induces the charge-transfer effect. It is found from these calculations that, due to the charge-transfer effect, the numbers of unpaired $4f$ electrons in the ${3d}^{\ensuremath{-}1}$ core-hole states of $[{\mathrm{LaCl}}_{6}{]}^{3\ensuremath{-}},$ $[{\mathrm{LaBr}}_{6}{]}^{3\ensuremath{-}},$ $[{\mathrm{LaF}}_{6}{]}^{3\ensuremath{-}},$ and $[{\mathrm{LaO}}_{6}{]}^{9\ensuremath{-}}$ increase by 1.10, 1.15, 0.37, and 0.58, respectively, from the ground state. From this result, it is found that charge-transfer effect ordering is ${[\mathrm{LaBr}}_{6}{]}^{3\mathrm{\ensuremath{-}}}{=[\mathrm{LaCl}}_{6}{]}^{3\mathrm{\ensuremath{-}}}{g[\mathrm{LaO}}_{6}{]}^{9\mathrm{\ensuremath{-}}}{g[\mathrm{LaF}}_{6}{]}^{3\mathrm{\ensuremath{-}}}.$ $\mathrm{La} 3d$ XPS of La halides and oxides was reported [S. Suzuki, T. Ishii, and T. Sagawa, J. Phys. Soc. Jpn. 37, 1334 (1974); J. C. Fuggle et al., Phys. Rev. Lett. 45, 1597 (1980)], and the intensity ratios of the high- to low-binding-energy peaks of $\mathrm{La} 3d$ XPS for ${\mathrm{LaCl}}_{3},$ ${\mathrm{LaBr}}_{3},$ ${\mathrm{LaF}}_{3},$ and ${\mathrm{La}}_{2}{\mathrm{O}}_{3}$ are 1.4, 1.4, 0.3, and 0.8 in case of both ${3d}_{5/2}$ and ${3d}_{3/2}$ XPS, which means that the core-hole screening ordering is ${\mathrm{LaBr}}_{3}{=\mathrm{L}\mathrm{a}\mathrm{C}\mathrm{l}}_{3}{g\mathrm{La}}_{2}{\mathrm{O}}_{3}{g\mathrm{LaF}}_{3}.$ From these results, it is found that the intensity ratio of the high- to low-binding-energy peak of $\mathrm{La} 3d$ XPS among La compounds corresponds to the charge-transfer effect.

Dynamical core-hole screening in weak chemisorption systems

Adsorbate core-hole screening in weak chemisorption systems is studied by the example of the ${N}_{2}/Ni(100)$ system. Fourth-order Green's-function calculations have been carried out on the $\mathrm{N}1s$ x-ray photoelectron spectra (XPS) of a linear ${\mathrm{NiN}}_{2}$ cluster used as a model of the chemisorption system. Good agreement with the experimental spectra is obtained. A powerful method, based on exact effective Hamiltonians (EEH's) constructed by means of a technique for a unique block diagonalization of Hermitian matrices [J. Phys. A 22, 2427 (1989)] is used for a detailed analysis of many-electron core-hole states. Two lowest-energy states in the $\mathrm{N}1s$ XPS of the ${\mathrm{NiN}}_{2}$ cluster, split by 1.47 eV (the experimental splitting in the chemisorption system is 1.3 eV), are classified as $\stackrel{\ensuremath{\rightarrow}}{\ensuremath{\pi}}{\ensuremath{\pi}}^{*}$ shakedown satellites, whereas two states dominating the region of the so-called ``giant satellite'' are attributed to the main states. The EEH technique allows us to study in detail the dynamical screening of core holes, which is shown to be responsible for the formation of the strong shakedown satellites. A comparative EEH analysis of the core-hole states of the ${\mathrm{NiN}}_{2}$ and NiCO clusters explains in simple terms the origin of the very different intensities of the giant satellites observed experimentally in adsorbate spectra of weak and strong chemisorption systems.

MO and DFT approaches to the calculation of X-ray absorption/emission spectra of nitrogen atom adsorbed on Cu(100)

X-ray emission/absorption spectra are calculated for nitrogen chemisorbed on cluster models (up to Cu61) of the four-fold hollow site in the Cu(100) surface. We employ both Hartree-Fock and gradient corrected density functional (DFT) techniques to compute transition energies and intensities involving the fully relaxed core-hole state. Large covalent contributions between the N(2p) and Cu(3d) orbitals are found only at the DFT level, giving rise to bonding and antibonding states that are the main characteristics of the experimental spectra.

Long-lived core-hole excited states and high-energy thresholds in stimulated desorption processes: Cl/Si(100)-(2 × 1)

The valence-electron response to an adsorbate core-hole is studied by means of the configuration-Iattice approach in order to include correlation effects as accurately as possible. For the specific case of Cl adsorption on the reconstructed Si(100)-(2 × 1) surface, we find that removal of an electron from the Cl 3s orbital leads to long-lived excited states at 22 and 41 eV, in excellent agreement with the high-energy thresholds found in the stimulated desorption of Cl ions. The first excitation corresponds to a neutral valence-electron distribution (Cl+ desorption) while the second one, which leads to Cl2+, requires transfer of a valence electron from the adsorbate to the semiconductor accompanied by a strong shake-up of electron-hole pairs. These results are compared with those of a previous calculation where the core-hole was not included.

Auger decay of theC1s→2π*excitation of CO

Calculations on the Auger decay of CO after the $\mathrm{C}1s\ensuremath{\rightarrow}2{\ensuremath{\pi}}^{*}$ excitation have been performed utilizing a multichannel Schwinger approach. The bound orbitals used to expand the multichannel configuration-interaction wave functions of the ion state were the natural orbitals of a single excitation and double excitation configuration-interaction (CI) calculation on the $\mathrm{C}{1s}^{\ensuremath{-}1}$ core-hole state of CO. Fixed-nuclei photoionization cross sections were then obtained with target states represented by complete active space CI wave functions. An analysis of the resulting cross sections using Fano's spectral line profile and the lifetime vibrational interference theory has lead to photoabsorption profiles in good agreement with experimental measurements. Additionally, we have found good agreement with experimental data for the resulting photoelectron and photoemission spectra.

Direct calculation of valence-band Auger emission: Spin polarization of Auger electrons from a potassium (110) surface

We report calculations on spin-polarized Auger electron emission from the valence band of a potassium (110) film. In treating this process we use an $N$-electron scheme that is based on a generalized version of density-functional theory. The transition rate is determined by explicitly evaluating the transition matrix elements that contain the four states involved. The core and valence states are obtained from a self-consistent full-potential linearized augmented plane wave calculation on a K(110) multilayer. In the experiments that the present calculations refer to, the particular oriented core-hole state is created by photoexcitation using circularly polarized light. The observed energy spectrum and the angular dependence of the spin polarization of the emitted Auger electrons can be simply related to the character of the spinor-hole state. We compare our results for normal incidence of the light to the pertinent experiments on the Auger spin polarization referenced to the spin of the incoming photons. The results are in fair agreement with each other.

Natural circular dichroism in non-resonant x-ray emission

The possibility of observing natural circular dichroism in non-resonant x-ray emission spectroscopy is investigated by means of simulations of the chiral molecules twisted ethylene, propylene oxide and trans-1, 2-dimethylcyclopropane, in a two-step model and at the SCF level, with or without relaxation of the core-hole states. We observe both a chemical and an element dependence of the phenomenon and also an effect of electron relaxation. However, the latter is much less crucial than for circular dichroism in x-ray absorption. The calculations indicate that, at least for the decay of the carbon core-hole states, the effect could be detectable with the present or soon to be available experimental equipment.

Role of d screening in ion-excited electron emission from Ca

We report an analysis of the electron emission spectrum from Ca metal excited by rare-gas ion bombardment recorded at high-energy resolution. Energy considerations, assisted by electron-energy-loss measurements, indicate that the electron emission from Ca is dominated by the decay of neutral excited atoms from ${3\mathrm{p}}^{5}$ ${3\mathrm{d}4\mathrm{s}}^{2}$ states. The spectrum is explained in terms of initial 3p bombardment induced ionization followed by 3d neutralization as a result of d electron screening involving electrons from the conduction band in the solid. On emerging from the surface most Ca atoms are in the lowest ${3\mathrm{p}}^{5}$ ${3\mathrm{d}4\mathrm{s}}^{2}$ multiplets ${(}^{3}$ P and $^{3}$ F) leading to electron emission via autoionization at energies well below that observed in resonant photoemission from Ca atoms. The results give insight into the relationship between the d-screened core hole state in Ca and multiplets of core excited atomic states.

Auger - photoelectron coincidence spectra of gaseous : decay of selected ionized states

Auger spectra are generally interpreted as decay of a core-ionized state. However, not only the pure core-hole state but also shake-up and shake-off configurations contribute to conventionally excited spectra. The interpretation becomes difficult whenever the various Auger transitions are observed at the same kinetic energy. Using Auger - photoelectron coincidence spectroscopy we here demonstrate the possibility to obtain unique decay spectra of the pure core ionized single hole state and the singlet coupled shake-up satellite state of .

Correlation effects in the - MM Auger transitions of Ar

- MM Auger spectra of argon have been measured with very high resolution using excitation by monochromatic synchrotron radiation. The photon energy was selected slightly above the 2p ionization threshold for creation of the single 2p core hole state only. The Auger spectra are thus free from any satellite lines originating from two-hole initial states and therefore well suited to study the configuration interaction in the final ionic state of the decay. Electron correlation is investigated by calculating the energies and intensities of the transitions using the multiconfiguration Dirac - Fock method and by comparing them with the experimental results.

Altered photoemission satellites atCaF2- andSrF2-on-Si(111) interfaces

Bulk and interface photoemission satellite excitations, measured with x-ray photoelectron spectroscopy and x-ray photoelectron diffraction, are compared for thick, thin, and monolayer films of ${\mathrm{CaF}}_{2}$ and ${\mathrm{SrF}}_{2}$ on Si(111). Intrinsic satellites are observed for excitation of atoms in the first monolayer, both uncovered and at the buried interface, that differ from those associated with bulk atoms. For F 1s excitation, the bulk and interface satellites differ only in width and amplitude; for the cation core excitations (Ca 2p, Sr 3p, and Sr 3d), the observed excitation energies and intensities differ both from the equivalent bulk satellites and among the various core hole states. The results yield new information on the nature of the interface bonding, as well as on the origin of both bulk and interface satellites. Several models are considered, including differential screening, dielectric losses, crystal-field and multiplet effects, and interface excitation. The most likely explanation for the new cation satellites at the ${\mathrm{CaF}}_{2}$/Si(111) [${\mathrm{SrF}}_{2}$/Si(111)] interface is localized excitation of the interface bond, where the interface band gap is controlled by the collapse of the Ca 3d (Sr 4d) level in the presence of the Ca (Sr) core hole. \textcopyright{} 1996 The American Physical Society.

Simple correlation between Mn-K? X-ray emission and Mn 2p X-ray photoelectron spectra for high oxidation number or low-spin manganese compounds

The high-resolution Mn-Kα X-ray emission spectra (XES) and Mn 2p X-ray photoelectron spectra (XPS) for K[MnO4], K3[Mn(CN)5(NO)], K2[MnO4] and K3[Mn(CN)6] were measured. The first and second compounds which have no unpaired electrons, give symmetrical peaks in Mn-Kα1 XES and Mn 2p3/2 XPS spectra. The spectra of the other compounds show asymmetrical lines. When the lines are divided into two Lorenzian (for XES) or Gaussian (for XPS) curves, the energy separation and intensity ratio of the two peaks coincide with each other in both spectroscopies. These results are rationalized by spin-unrestricted DV-Xα molecular-orbital calculations of the anions. The electronic structures of these anions are calculated for their ground state and hole states in 1s and 2p levels. It is concluded that the electronic structures of the compounds in the ground states are conserved in the core hole states. Thus the satellite peaks due to charge transfer are not observed and the peak profiles are governed by the exchange splitting between 2p and valence electrons.