Inverse Photoemission Experiments Research Articles

Possibility of a Mott-Hubbard ground state for the SiC(0001) surface

The possibility that large electronic correlation effects produce a semiconducting surface-state spectrum for the $\mathrm{Si}{\ensuremath{-}T}_{4}$ adatom model of the $\mathrm{SiC}(0001)\sqrt{3}\ifmmode\times\else\texttimes\fi{}\sqrt{3}$ surface is discussed. We argue that a Hubbard model, with parameters obtained from local-density calculations, provides a qualitatively correct account of the excitation spectra inferred from angle-resolved photoemission and inverse-photoemission experiments. The Hubbard $U$ parameter obtained in local-density-functional calculations for the Si-adatom dangling bond on SiC(0001) is 1.6 eV.

A comparison of two high performance inverse photoemission bandpass detectors

By performing inverse photoemission experiments on the same sample at the same time with two different detectors, their performance has been directly compared. The first detector is based on one of the most promising solid-state detector designs. It is comprised of a focused mesh electron multiplier and a CaF2 window. The second detector is a Geiger–Müller tube which uses dimethyl ether and a MgF2 window. Although it has already been demonstrated that detectors based on this design work, the dimethyl ether Geiger–Müller tubes are not widely used, and we show that it is essential to compensate for detector dead time effects for the detector to be practicably useful. Once this is done, the dimethyl ether Geiger–Müller tube has a sensitivity that is approximately 20 times greater than that of the solid-state detector. Furthermore, it is easy to operate and it does not appear to suffer from the problems that are normally associated with iodine Geiger–Müller detectors.

Magnetic polaron in ferro- and antiferromagnetic semiconductors.

The temperature-dependent quasiparticle spectrum of a single conduction electron exchange coupled to a ferro- or antiferromagnetically ordered localized-spin system (e.g., EuO, EuTe) is calculated by a moment-conserving Green function technique. In the weak coupling regime the exchange interaction leads to an almost rigid shift of the Bloch dispersion. The induced spin splitting of the conduction band states is proportional to the magnetization 〈${\mathit{S}}^{\mathit{z}}$〉 of the localized-spin system. As soon as the coupling constant exceeds a critical value an additional splitting of the quasiparticle dispersion for each spin projection sets in due to different elementary excitations. One is based on a repeated emission and reabsorption of a magnon by the conduction electron resulting in an effective attraction between magnon and electron. This gives rise to a polaronlike quasiparticle (``magnetic polaron''). Another excitation is due to a direct magnon emission or absorption by the electron thereby flipping its own spin (``scattering states''). For the exactly calculable special case of a ferromagnetically saturated spin system (T=0 K), the magnetic polaron appears only in the \ensuremath{\downarrow} spectrum and turns out to be a stable quasiparticle. For finite temperatures it gets a finite lifetime. In antiferromagnetic systems each quasiparticle band exhibits an additional ``Slater splitting'' due to the reduced magnetic Brillouin zone. The predicted strong correlation effects in the excitation spectrum require unconventional interpretations of respective inverse photoemission experiments. \textcopyright{} 1996 The American Physical Society.

Intrinsic periodicity associated with quantum-well states in a magnetic sandwich

From a simplified quantum-well model for a magnetic sandwich, it can be shown that in the variation of the number of occupied levels with the spacer layers, there exists an intrinsic periodicity, which leads the Fermi level to oscillate periodically. The intrinsic periodicity does not depend on the magnetic alignment but only on the quantum-well states themselves. The oscillation period can be approximately given by , where and are the effective masses of the electron in the lateral and perpendicular directions respectively, is the distance between the two neighbouring atomic layers, and is the electron density in the spacer layers. This makes one speculate that the long periodicity of the oscillatory coupling could be a result of the intrinsic periodicity, if and a small s-electron density are assumed. On the other hand, from the calculated energy bands for a thin film, it is found that the electronic structure is highly anisotropic, which is in agreement with this assumption. Therefore, it can be confirmed that the intrinsic periodicity plays an important role in the oscillatory coupling. An inverse photoemission experiment on Cu(100) films over Co can be explained quite well using this physical picture.

CO on Pd(110): determination of the optimal adsorption site

We present ab initio pseudo-potential plane-wave total-energy calculations for the geometric and electronic structure of the CO-covered Pd(110) surface. Our calculations were performed within the local-density approximation (LDA) of density functional theory (DFT). There has been some controversy as to whether CO prefers to adsorb at a bridge or on-top site when exposed to Pd(110). Total energy calculations for a CO monolayer adsorbed at the on-top and bridge adsorption sites revealed the bridge site adsorption to be favored by 0.59 eV per CO molecule. The preferential adsorption of CO to the bridge site was further corroborated by our band-structure calculations, with only the bridge site results being in good agreement with recent inverse photoemission experiments.

Spectral function of the 1D Hubbard model in the U-->+

We show that the one-particle spectral functions of the one-dimensional Hubbard model diverge at the Fermi energy like $|\omega-\varepsilon_F|^{-3/8}$ in the $U\to +\infty $ limit. The Luttinger liquid behaviour $|\omega-\varepsilon_F|^\alpha$, where $\alpha \to 1/8$ as $U\to +\infty $, should be limited to $|\omega-\varepsilon_F| \sim t^2/U$ (for $U$ large but finite), which shrinks to a single point, $\omega=\varepsilon_F$,in that limit. The consequences for the observation of the Luttinger liquid behaviour in photoemission and inverse photoemission experiments are discussed.

Inverse photoemission and Kelvin probe studies of the Au/GaP(110) interface

We performed angle-integrated inverse-photoemission experiments on the Au/GaP(110) interface in the lowcoverage regime. For a gold deposition of 0.5 Å we found the IPES spectrum shifting by 0.25 eV towards lower energies below the Fermi reference level. This result is attributed to an excess of positive accumulated charge at the interface induced by the impinging electrons. This shift is consistent with our calculations based on the thermionic field emission model for transport processes in the Schottky barrier. Contact potential difference measurements under different illumination conditions were used to correlate the above electron induced voltage and the photovoltage observed on the same sample.

Electronic and magnetic structure of rare-earth materials studied by high-resolution photoemission

Recent experimental results concerning the surface and bulk electronic and magnetic structure of rare-earth (RE) metals are presented. Surface core-level shifts in photoemission (PE) and inverse photoemission (IPE) experiments are discussed to summarize the phenomena that occur due to the energy lowering of 4f states at the surface. In Ce metal, a highly correlated electronic system, surface effects are influenced by the strong interaction of the 4f electron with the valence band, leading to essentially no observable shift in PE. Resonant PE experiments with varying degree of surface sensitivity have been performed to investigate the surface electronic structure in these systems. With recent progress in the growth of well-ordered epitaxial films of RE metals, detailed band-structure studies have become possible. The existence of a localized d-state at the surface is demonstrated and consequences for the surface properties are discussed. Single-crystalline RE surfaces open up the possibility to study their magnetic properties as shown for the case of Gd, where a strong magnetic circular dichroism has recently been observed.

Slave-particle studies of the electron-momentum distribution in the low-dimensional t-J model.

The electron momentum distribution function in the $t-J$ model is studied in the framework of slave particle approach. Within the decoupling scheme used in the gauge field and related theories, we treat formally phase and amplitude fluctuations as well as constraints without further approximations. Our result indicates that the electron Fermi surface observed in the high-resolution angle-resolved photoemission and inverse photoemission experiments cannot be explained within this framework, and the sum rule for the physical electron is not obeyed. A correct scaling behavior of the electron momentum distribution function near $k \sim k_{F}$ and $k \sim 3k_{F}$ in one dimension can be reproduced by considering the nonlocal string fields [Z. Y. Weng et al., Phys. Rev. B45, 7850 (1992)], but the overall momentum distribution is still not correct, at least at the mean field level.

Solid-state bandpass detector for inverse photoemission spectroscopy

This paper presents the details of a new arrangement for the bandpass detection of radiation in the vacuum ultraviolet region. The detector is based on two wide bandpass filters that share the diminishing quantum efficiency of a microchannel plate as their common low-energy cutoff, but have high-energy cutoffs determined by the transmittance limits of different crystal windows. The detector has been used to perform an inverse photoemission experiment on a sample of polycrystalline gold. From the results of this experiment the detector is estimated to have a bandpass centered at 11.4 eV and a resolution of 0.57 eV. (FWHM).

New calibration procedure for an inverse photoemission spectrograph

A new calibration procedure for determining the instrumental efficiency of a grating-based inverse photoemission spectrograph is described. Our results are in good agreement with those obtained by following another procedure previously adopted by other authors. The new procedure is particularly convenient for inverse photoemission experiments conducted in the so-called ‘‘isochromat’’ mode.

Dynamical effects on image-induced surface resonances at nearly-free-electron metal surfaces.

A mechanism for surface image-state formation in inverse-photoemission spectroscopy at free-electron-like metal surfaces is proposed and applied explicitly to Al(111). In the analysis, a multiple-scattering approach is used where the substrate (jellium) reflectivity to an electron virtual-photon field (or bremsstrahlung electromagnetic radiation) produced in the inverse-photoemission experiment is calculated via the Fresnel equations with the nonlocal surface corrections using the d-parameter formalism. In contrast to earlier statements, we assign the measured peak on inverse-photoemission spectra at Al(111) to the n=2 image surface state only.

Plasmon damping and surface interband transitions on Ag(001) and (011)

The dependence of surface-plasmon damping versus plasmon momentum and energy has been measured for the Ag(011) surface with angle-resolved electron energy loss spectroscopy. The data are compared with previous results obtained for Ag(001). At small transferred momenta, q‖, we observe a nearly linear dependence of the energy loss width with q‖ due to the growing phase space available for e-h pair creation. However the very efficient damping mechanism observed for q‖ beyond 0.10 Å−1 is not present for Ag(011). We demonstrate that it is due the onset of interband transitions between filled and empty surface states. Two such states are indeed present near the X point of Ag(001) as demonstrated by inverse photoemission experiments and electronic band structure calculations. The energy gap between the two bands can thus be determined by EELS with a greater precision than by photoemission.

QUANTUM MONTE CARLO SIMULATIONS FOR THE THREE-BAND HUBBARD MODEL

We present in this article results of Quantum Monte Carlo simulations for the three-band Hubbard model obtained both with the grand canonical approach as well as with the T =0 formalism. A direct comparison of Quantum Monte Carlo results with experimental values for the charge-transfer gap leads to a set of parameters for the three-band Hubbard model that consistently describes salient features of the high- T c superconductors. We find long-range antiferromagnetic order in the undoped case as well as an incommensurate structure for a doping concentration of δ≈0.20. Moreover the magnetic susceptibility as a function of doping δ is in a qualitative agreement with experiments. Concerning dynamical properties, we have evaluated the spectral density [Formula: see text]. The dispersion of the state with the lowest excitation energy is in very good agreement with experimental photoemission and inverse photoemission experiments for a doping concentration of δ=0.25. In the superconducting regime, we find that the interaction vertex for the pairing correlation function in the extended s-channel shows a maximum for δ~0.20, resembling the dependence of T c on doping. This feature is not present for other symmetries of the order parameter or in other parameter regions.

Unoccupied orbitals of C and O on Fe(100)

Inverse photoemission experiments from c(2 × 2) carbon and 1 × 1 oxygen overlayers on Fe(100) reveal adsorbate states at 2.5 and 3.9 eV for C and at 1.7 eV for O. They are assigned to antibonding combinations of C 2p and O 2p states with Fe 3d states. For oxygen on Fe(100) the adsorbate state resonates with the minority spin Fe 3d band, but not with the majority band, in agreement with a recent first principles calculation. Together with previous photoemission results we obtain the energy levels for a complete surface reaction, i.e., dissociation/recombination of CO on Fe(100).

Why is photoemission better than inverse photoemission for studying high-T c oxides?

Within the model of hole superconductivity a fundamental asymmetry between electrons and holes exists. Mathematically this translates into a dependence of the superconducting energy gap function Δ k on the band energy ε k . We discuss herethe consequences of this dependence for high resolution photoemission and inverse photoemission experiments. In particular we find: (1) the angle-integrated spectrum in the superconducting state should be sharper in photoemission than in inverse photoemission, and there should be a net gain (loss) in spectral weight in going from the normal to the superconducting state in the former (latter) technique. (2) High resolution angle-resolved spectra should exhibit small qualitative differences with usual BCS behavior. In particular, in a range of k values the peak in the photoemission spectrum in the superconducting state should move closer to the Fermi energy with k while the corresponding peak in the normal state is moving away from it. In the photoemission spectrum the peaks should stay close to the Fermi energy for a larger range of k values than in the inverse photoemission case.

Theory for photon emission from a scanning tunneling microscope

We have calculated the rate of light emission from a scanning tunneling microscope with an Ir tip probing a silver film. In the calculation we model the tip by a sphere. We find a considerable enhancement of the light emission compared with for example inverse photoemission experiments. This enhancement is explained as the result of an amplification of the electromagnetic field in the area below the microscope tip due to a localised interface plasmon. We estimate that one out of 104 tunneling electrons will emit a photon in the visible range. Due to an electromagnetic decoupling of the sphere from the sample the enhanced emission is lost for photon energies above a certain value. We also find that the experimentally observed maximum in the light emission as a function of bias voltage is related to the behavior of tip-sample separation versus bias voltage.

Anomalous dynamical behavior of the Cu(110) surface.

The thermal vibrations of surface atoms on Cu(110) have been studied as a function of temperature with ion-scattering spectroscopy in the impact-collision mode. By comparing the experimental data to Monte Carlo simulations of the scattering process, mean-square displacements of surface atoms are obtained that agree well with the results of helium diffraction and inverse photoemission experiments. Below 550 K the atom vibrations can be described quite well by a Debye temperature of 224 K obtained from phonon slab calculations. Above 550 K, strong anharmonic vibrations are observed. In this temperature region we find thermally activated adatoms to be produced with an activation energy of 0.27\ifmmode\pm\else\textpm\fi{}0.07 eV. Both effects lead above 1000 K to structural changes of the surface that can be described in terms of a rough surface or a quasiliquid layer on the surface.

Many-body calculation of the surface-state energies for Si(111)2 x 1.

The surface-state excitation energies for the Si(111)2\ifmmode\times\else\texttimes\fi{}1 surface have been calculated in the GW approximation. The energy position and the dispersion of the occupied and empty surface states are in excellent agreement with photoemission and inverse-photoemission experiments. The calculated quasiparticle surface-state band gap, 0.62 eV, is 0.15 eV larger than the measured onset energy for electron-hole pair excitations. The possibility that excitonic effects are responsible for this difference is examined.

Surface states of GaP (110)

Chadi's tight-binding total energy algorithm was used to study the reconstruction of the GaP (110) surface and its intrinsic occupied and unoccupied states. Assuming a rotation-relaxation model, the optimal surface atomic geometry was obtained by minimizing the total energy of the system. The angle of rotation obtained, ω = 22°, is near the experimental value of 27.5°. The features of the surface electronic states were determined and described. The calculated intrinsic occupied surface states are in excellent agreement with the experimental data and those for the unoccupied states are about 30% higher than the results of inverse photo-emission experiments.