CHARGE STATES OF SPUTTERED ATOMS/IONS

The charge states of sputtered cesium and oxygen ions/atoms from metal surfaces have been discussed with the time-dependent Newns-Anderson model. The final charge state distribution has been calculated as a function of work function of the surfaces. The calculated results fit the experimental data well. Besides, the dependence of the final charge states of the sputtered oxygen particles on the surface temperature is also predicted.     

铯离子/原子与金属表面电荷交换的计算

用Ｎｅｗｎｓ－Ａｎｄｅｒｓｏｎ量子模型，计算了低能铯离子与涂有铯原子的钨表面（Ｃｓ／ｗ（１１０））碰撞时，由于电荷交换引起的末态电荷态的分布与Ｃｓ／Ｗ（１１０）表面功函数及表面温度的关系，计算结果与相应的实验符合得较好。同时，也给出了铯离子和原子与Ｃｓ／Ｗ（１１０）表面相互作用过程中粒子电荷态变化的全过程，并对两者作了比较． 关键词：     

Evidences of a double resonant ionization mechanism in sputtering of metals

We present a theory of resonant charge exchanges, between sputtered atoms and metal surfaces, in which surface effects occur as quasi-molecular correlations in the diatomic systems formed, in the collision cascade, between secondary emitted atoms and their nearest-neighbor substrate atoms that have provided the last impulse for ejection. We set up a generalized Anderson-Newns Hamiltonian, from first principles, using a truncated and orthonormal set of states obtained from the valence orbitals of the diatomic system and from a continuous basis of jellium wave functions. We calculate the one-electron matrix elements appearing in the equations of motion for the annihilation operators of the truncated set in comparison with those resulting from the basic theory of resonant charge transfer. We determine the ionization probability of secondary emitted atoms versus their final emission velocities and we find it to be in good agreement with experimentally derived data on the Cu⁺/Cu-system. We support the hypothesis that the bare Anderson-Newns hopping mechanism needs to be completed with another charge transfer channel at the low energies of secondary ion emission.     

Deflection of sputtered ions by surface charges

The trajectories of ions sputtered from an insulating surface carrying a surface charge have been computed. The results show that yields measured in particular directions can be in serious error as the trajectories of low energy sputtered ions depend critically on surface charge and point of origin of the ion. The distorted yield curves for 1.2 eV ions and 10 eV ions leaving a surface with a ten volt surface potential are calculated.     

Non-adiabatic charge transfer process of oxygen on metal surfaces

The dynamics of charge transfer processes of oxygen on metal surfaces are investigated. The analysis is based on a set of diabatic potential energy surfaces, each representing a different charged oxygen species. Empirical universal potential energy functions have been constructed that mimic the oxygen–silver, oxygen–aluminum and oxygen–cesium systems. The differences between the work functions of these metals are reflected in the potential parameters. The dynamics are followed by solving the multichannel time-dependent Schrödinger equation starting from oxygen in the gas phase. Only the direct short time part of the dynamics is followed leading to the creation O⁻₂ in the gas phase as well as dissociative chemisorption. A large portion of the wave function is trapped in molecular chemisorption charged states. It is found that the position of the crossing seam between potentials has a profound influence on the outcome.     

Electronic properties of TiC(100) and polar TiC(111) surfaces

The energy bands of films of TiC have been calculated using the linear-combination-of-atomic-orbitals method with parameters obtained by a fit to the bulk band structure. The Madelung potentials and charge redistribution have been determined self-consistently. For the neutral TiC(100) surface, the density of states (DOS) is similar to that of the bulk. For the non-neutral Ti-covered TiC(111) surface, Ti 3d-derived surface states appear around the Fermi energy E_F. The long-range electric field produced by the polar surfaces is screened by the charge redistribution, and the polar surfaces are stabilized. Characteristic features of TiC(111) compared to other surfaces of TiC are attributed to the high surface DOS at E_F.     

Ion-induced secondary electron and negative ion emission from Mo/O

Absolute yields of secondary electrons and negative ions resulting from collisions of Na⁺ with Mo(100) and a polycrystalline molybdenum surface have been measured as a function of the oxygen coverage of the surface for impact energies below 500 eV. The sputtered negative ions have been identified with mass spectroscopy, and O⁻ is found to be the dominant sputtered negative ion for the surfaces at all oxygen coverages and impact energies. Both the electron and O⁻ yields have an impact energy threshold at about 50 eV and exhibit a strong dependence on oxygen coverage. The kinetic energy distributions of the secondary electrons and sputtered O⁻ were determined as functions of the oxygen coverage and impact energy. The distributions for O⁻ are characterized by a narrow low-energy peak (at 1–2 eV) followed by a low-level high-energy tail. The secondary electrons have a narrow (FWHM 1–2 eV) kinetic energy distribution, centered approximately at 1–2 eV. The shapes of the distributions and their most probable energies are essentially invariant with impact energy, oxygen coverage and the nature of the Mo surface. The emission is explained and analyzed in terms of a simple model which involves a collision-induced electronic excitation of the MoO⁻ surface state. The decay of this excited state leads to the production of both secondary electrons and O⁻ with energy distributions and yields comparable to those observed.     

Density of electron states near surfaces of layered copper oxides

Abstract

The electron structure of the semi infinite CuO₂ layer has been calculated to understand peculiarities of the layered oxides surfaces. All computations have been performed by means of the Green function method along with the simple tight-binding model. Fourier transforms of the Green functions were derived analytically. The potential at the atoms nearest to the boundary was directed to infinity. This formal maneuver led to dividing the ideal CuO₂ layer into two semi infinite layers. The density of states near the surface has been obtained by integrating the imaginary part of the diagonal element of the Green function over the wave vector. It has been shown that the charge on the surface Cu atoms is greater than on the bulk one. Changes of the density of states via the energy and the distance from the surface have been obtained.     

First-principles study of the interaction of oxygen with the SnO2(1 1 0) surface

First-principles calculations based on density functional theory in the generalised gradient approximation, together with pseudopotentials and plane-wave basis sets, have been used to investigate the energetics of oxygen adsorption on stoichiometric and weakly and strongly reduced SnO₂(1 1 0) surfaces. It is shown that, if the surface species formed by oxygen adsorption are restricted to be charge neutral, then oxygen cannot be exothermically adsorbed from the gas phase on the stoichiometric surface. A variety of molecular and dissociative modes of adsorption are examined on the reduced surface produced by removing all bridging oxygens and on the weakly reduced surface that results from removal of only a fraction of these oxygens, with the adsorbed species being in both the singlet and the triplet states, and we identify a number of modes not discussed before in the literature. We use the calculated adsorption energies to propose a tentative assignment of these adsorption modes to the peaks observed in temperature programmed desorption experiments on the SnO₂ and TiO₂(1 1 0) surfaces.     

Retention of the potential energy of multiply charged argon ions incident on copper

The retained fraction of the potential energy of argon ions incident on copper has been measured using stationary calorimetry at charge states up to 9+ and kinetic energies ranging from 75 to 240 eV per ionic charge. An average fraction of 30% to 40% is found with little dependence on the charge number and on the kinetic energy. The retention of the total energy ranges from 60% to 75% and can mainly be accounted for by the retained fraction of the potential energy and the collisional energy lost by reflected ions and sputtered target atoms.     

The sputtering process and sputtered ion emission

In this review an attempt is made to delineate those physical characteristics of the sputtering event which are of importance for the discussion of sputtered ion emission. Emphasis is placed on the grossly disordered nature of the sputtering site as the sputtered atoms depart. The electronic nature of a disordered site is discussed and attention is drawn to the inapplicability of bulk density of states concepts to this region. In particular, it is argued that the concept of a band gap in the surface state density is inappropriate. Sputtered ion emission models are reviewed with respect to their consistency with the physics of the sputtering event and with the general features of sputtered ion emission. It is argued that a self-consistent model can be developed in the framework of which sputtered ion emission from oxidized surfaces, from clean alloy surfaces, and from the pure elements can be understood.     

First-principles optical spectra for F centers in MgO

The study of the oxygen vacancy (F center) in MgO has been aggravated by the fact that the positively charged and the neutral vacancy (F+ and F0, respectively) absorb at practically identical energies. Here we apply many-body perturbation theory in the G0W0 approximation and the Bethe-Salpeter approach to calculate the optical absorption and emission spectrum of the oxygen vacancy in all three charge states. We observe unprecedented agreement between the calculated and the experimental optical absorption spectra for the F0 and F+ center. Our calculations reveal that not only the absorption but also the emission spectra of different charge states peak at nearly the same energy, which leads to a reinterpretation of the F center's optical properties.     

Stimulated desorption studies of defect structures on TiO2

The influence of the surface geometric structure on the electron and photon-stimulated desorption (ESD, PSD) ion yield from TiO₂(001) and TiO₂(110) has been studied. For both surfaces, angle-integrated ESD yields have been measured as a function of annealing temperature ranging from room temperature sputtered surfaces to 1200 K annealed surfaces. These measurements imply that the local geometry of the desorption site must be considered in order to adequately explain the observed ion yield variations. The ESD ion angular distribution (ESDIAD) patterns have been obtained after annealing in the temperature range 300–1200 K. These patterns show the presence of short-range structural order, even on highly defective, sputtered surfaces, and they are interpreted in terms of plausible models of surface structure. Changes in both the surface Ti-cation valence state and its coordination number as a function of annealing temperature are proposed to give rise to the observed ESDIAD patterns and the total ion yield behavior.     

Oxygen adsorption on the LaB6(100), (110) and (111) surfaces

Oxygen adsorption on the LaB₆(100), (110) and (111) clean surfaces has been studied by means of UPS, XPS and LEED. The results on oxygen adsorption will be discussed on the basis of the structurs and the electronic states on the LaB₆(100), (110) and (111) clean surfaces. The surface states on LaB₆(110) disappear at the oxygen exposure of 0.4 L where a c(2 × 2) LEED pattern disappears and a (1 × 1) LEED pattern appears. The work function on LaB₆(110) is increased to ～3.8 eV by an oxygen exposure of ～2 L. The surface states on LaB₆(111) disappear at an oxygen exposure of ～2 L where the work function has a maximum value of ～4.4 eV. Oxygen is adsorbed on the surface boron atoms of LaB₆(111) until an exposure of ～2 L. Above this exposure, oxygen is adsorbed on another site to lower the work function from ～4.4 to ～3.8 eV until an oxygen exposure of ～100L. The initial sticking coefficient on LaB₆(110) has the highest value of ～1 among the (100), (110) and (111) surfaces. The (100) surface is most stable to oxygen among these surfaces. It is suggested that the dangling bonds of boron atoms play an important role in oxygen adsorption on the LaB₆ surfaces.     

ESCA measurement of surface atom oxidation states on chemically modified Mo(100) surfaces

Quantitative surface atom oxidation states are measured in a conventional ESCA spectrometer. Spectra that solely reflect the chemical nature of the surface atoms are generated from a linear combination of spectra measured at two take-off angles. Proper selection of the multiplicative constants results in removal of bulk atom effects from the measured spectra. A simple method which is based on photoelectron attenuation lengths is used to calculate these coefficients. The surface Mo atom oxidation numbers for a series of chemically modified Mo(100) surfaces have been obtained by this procedure. Charge transfer effects resulting from the adsorption of electronegative surface modifiers (B, C, O and CO) are observed. At constant adatom coverage, the Mo oxidation number is linearly dependent on the adatom electronegativity. For a series of partially oxidized Mo(100) surfaces the extent of charge transfer to the metal atoms is not a simple function of the surface oxygen coverage. A model for the initial oxidation of Mo(100) is presented which is in agreement with previous structural and electronic state studies. At oxygen coverages below one monolayer, a chemisorbed oxygen phase exists which produces a constant Mo oxidation number of 1.3. At higher coverages, a rapid increase in the Mo surface atom charge indicates the formation of a surface oxide layer. The electron deficient Mo surface atoms are acidic sites for adsorption of gas phase Lewis bases. The relative strengths of the acidic sites correlate with the metal atom oxidation numbers. This result suggests that the observed adsorption selectivity and reactivity of these surfaces can be discussed in a traditional acid-base context.     

Dependence of the intrinsic line width of surface states on the wave vector: The Cu(111) and Ag(111) surfaces

The dependence of the intrinsic line width Γ of electron and hole states due to inelastic scattering on the wave vector k _∥ in the occupied surface state and the first image potential state on the Cu(111) and Ag(111) surfaces has been calculated using the GW approximation, which simulates the self-energy of the quasiparticles by the product of the Greens’s function and the dynamically screened Coulomb potential. Different contributions to the relaxation of electron and hole excitations have been analyzed. It has been demonstrated that, for both surfaces, the main channel of relaxation of holes in the occupied surface states is intraband scattering and that, for electrons in the image potential states, the interband transitions play a decisive role. A sharp decrease in the intrinsic line width of the hole state with an increase in k _∥ is caused by a decrease in the number of final states, whereas an increase in Γ of the image potential state is predominantly determined by an increase of its overlap with bulk states.     

Cluster calculations of the electronic structure of transition metal surfaces

The electronic structure and charge distributions for 13-atom clusters of 3d transition metals (Fe, Ni. Cu) have been calculated using a scattered wave technique. The cluster geometry is chosen to display features of stepped and flat surfaces. Many states in the band show pronounced charge lobes in the neighbourhood of edge and corner atoms, suggesting that these are active sites in these metals. These lobes are more extensive for lighter elements in the series and, for a given element, more diffuse at lower energies in the d-band. Charge expansion from the centre of the cluster to the edge atoms is a general feature, and the stepped surfaces show a variety of bonding orbitals not present on the flat surface.     

First-principles study of the influences of oxygen defects upon the electronic properties of Nb-doped TiO_2 by GGA + U methods

The influence of oxygen defects upon the electronic properties of Nb-doped TiO2has been studied by using the general gradient approximation(GGA)+U method. Four independent models(i.e., an undoped anatase cell, an anatase cell with a Nb dopant at Ti site(NbTi), an anatase cell with a Nb-dopant and an oxygen vacancy(NbTi+VO), and an anatase cell with a Nb-dopant and an interstitial oxygen(NbTi+Oi)) were considered. The density of states, effective mass, Bader charge, charge density, and electron localization function were calculated. The results show that in the NbTi+VOcell both eg and t2glevels of Ti 3d orbits make contributions to the electronic conductivity, and the oxygen vacancies(VO) collaborate with Nb-dopants to favor the high electrical conductivity by inducing the Nb-dopants to release more excess charges. In NbTi+Oi, an unoccupied impurity level appears in the band gap, which served as an acceptor level and suppressed the electronic conductivity. The results qualitatively coincide with experimental results and possibly provide insights into the preparation of TCOs with desirable conductivity.     

Comparison of site-specific densities of states of Ga and As in cleaved and sputtered GaAs(110) by means of Auger line shapes

Site-specific densities of states for the Ga and As sites in GaAs(110) are derived from the M₁M_4,5V Auger lines and compared for the cleaved and sputtered surfaces. X-ray photoelectron spectroscopy, scanning electron microscopy, and LEED results for these surfaces are also presented. The results are interpreted in terms of a structure for the sputtered surface that consists of a two-phase mixture of Ga and disordered GaAs.     

Critical temperature of a van Hove superconductor with spin–charge separation

We studied a spin–charge separation superconductor with a van Hove density of states. The critical temperature and the influence of the Coulomb interaction have been calculated in the mean field model. We also calculated the Ginzburg–Landau coefficients as a function of temperature.