Energy dependence of the ionization probability of sputtered Cu and Ni

Despite its great sensitivity, the usefulness of secondary ion mass spectrometry (SIMS) for many applications has been limited by an inadequate understanding of the probability of sputtering an atom in an ionized state. To determine this ionization probability for clean Cu and Ni surfaces, I have measured the energy distribution of sputtered neutrals and ions by quadrupole mass filtering and retarding potential analysis using potential modulation differentiation. Analysis of sputtered neutrals was accomplished by electron impact ionization. Because the neutrals outnumber the ions by at least two orders of magnitude, the ratio of sputtered ions to neutrals is an accurate measure of the ionization probability. For energies below 20 eV the dependence of the ionization probability on energy goes as P(E) α Eⁿ, where n = 0.65 for clean Cu. The absorption of oxygen on the Cu surface increases the total ion yield while causing a reduction in the value of the exponent n. Similar results are found for nickel, where n = 0.54 for the clean surface.     

Secondary-ion emission from clean and oxygen-covered beryllium surfaces: II. Energy dependence

The secondary-ion energy distribution obtained by sputtering clean and oxygen-covered Be has been analyzed in terms of competing processes in secondary ion emission. The ion energy distribution N⁺(E) has been separated into an ionization coefficient R⁺(E) and a total energy distribution, N(E), i.e. N⁺(E) = R⁺(E) N(E). Experimentally, the dependence of R⁺(E) on both energy and oxygen coverage indicated a linear superposition of adiabatic tunneling and resonanance ionization processes from clean and oxygen-covered portions of the surface with no contributions to the secondary-ion yield from regions of intermediate coverage. Total energy distributions of sputtered Be atoms have been deduced and the principal features agree with the predictions of the collision cascade sputtering model. Variations of the energy distributions with oxygen coverage are in accord with small changes expected in the surface binding energy as a result of surface oxidation.     

Sputtering—a review of some recent experimental and theoretical aspects

After a brief outline of the present sputtering theory for a random solid, recent results of the sputtering yieldS for polycrystalline targets are discussed, in particular in view of the influence of the projectile mass and the bombarding angle. The angle dependence ofS at low bombarding energies, and results on the angular distribution of sputtered particles for oblique ion incidence point out necessary modifications of present sputtering theories with respect to the anisotropy of the collision cascades in the solid and the influence of the target surface. The energy distribution of the neutral particles ejected along the target normals is related to the theoretically predictedE ^?2-distribution of low energy recoils in the Recent mass spectrometric studies of postionized sputtered neutrals are discussed in view of the formation of sputtered molecules and the application of sputtered neutral mass spectroscopy for surface analysis. Finally, the paper deals with ion-induced surface effects on non-elementary sputtering targets, and the protracted removal of foreign atoms from a matrix.     

High-resolution (e,e') study of isovector M1 and M2 transitions in the oxygen isotopes: (I). 16O

The M1 and M2 transition strength distribution for ¹⁶O in the excitation energy range from 16 to 20 MeV has been measured in a high-resolution electron scattering experiment. The M1 strength is concentrated in three sharp states at E_x = 16.22, 17.14 and 18.79 MeV (± 0.01 MeV) with B(M1, k)↑ = 0.20 ± 0.02, 0.32 ± 0.03 and 0.13 ± 0.03 μ_N², respectively. An additional strength of 0.35 ± 0.09 μ_N², distributed over eight weakly excited states with excitation energies E_x = 17.4 to 18.0 MeV, brings the total measured M1 strength to B(M1, k)↑ = 1.0 ± 0.1 μ_N². The experimental M2 strength is distributed over states at E_x = 16.82, 17.78, 18.50 and 19.0 MeV (± 0.01 MeV) with B(M2, k)↑ = 19 ± 2, 13 ± 2, 59 ± 7 and 341 ± 51 μ_N² · fm², respectively. Electric transitions were also measured to states at E_x = 16.45 MeV (2⁺, E2), 17.30 MeV (1⁺, E1) and 18.20 MeV (2⁺, E2). Calculations were performed using the modified surface delta interaction in a 2p-2h shell model for the M1 transitions and the random phase approximation for the M2 transitions. The results show the sensitivity of the M1 strength as a measure of ground-state correlations and compare well with results from the ¹⁵N($\text{p}$, γ) reaction.     

Probability of ionization of sputtered particles as a function of their energy: Part I: Negative Si ions

In this study we have investigated how the probability of ionization of sputtered Si atoms to form negative ions depends on the energy of the atoms. We have determined the ionization probability from experimental SIMS energy distributions using a special experimental technique, which included de-convolution of the energy distribution with an instrumental transmission function, found by separate measurements.We found that the ionization probability increases as a power law ∼E^0.677 for particles sputtered with energies of 0-10 eV, then becomes a constant value (within the limits of experimental error) for particles sputtered with energies of 30-100 eV. The energy distributions of Si⁻ ions, measured under argon and cesium ion sputtering, confirmed this radical difference between the yields from low and high-energy ions.To explain these results we have considered ionization mechanisms that are different for the low energy atoms (<10 eV) and for the atoms emitted with higher energy (>30 eV).     

Image potential eigenstates and resonances on the (110) surfaces of noble metals: Energies and lifetimes

The energies and lifetimes of the image potential resonances at the ? point on the Cu(110), Ag(110), and Au(110) surfaces are studied, and the energies of the image potential states on the Pd(110) surface are analyzed. It is shown that every quantum number n corresponds to a pair of image potential states (resonances) n ⁺ and n ^? at the ? point. The average energy of a pair of eigenstates (resonances) at n ≥ 2 is well described by the formula ē _n = (E _n+ + E _n?) = E ₀ ? (0.85 eV)/(n + δ)², where δ is the quantum defect, E ₀ = (1/2m _e )(?π/a)², and a is the lattice parameter. The splitting energy of a pair of eigenstates (resonances) obeys the law ΔE _n = E _n+ — E _n? ∝ n ^?3. The lifetimes τ _n± of image potential resonances are proportional to n ³.     

Postcollision interaction upon electron excitation of the 5sns 1 S 0 and 5snd 3 D j levels of the cadmium atom

The optical excitation functions (OEFs) for two series of spectral lines of the Cd atom originating from the 5sns ¹ S ₀ (n = 6?11) and 5snd ³ D _{1, 2, 3} (n = 5 and 6) levels excited by an ultramonoenergetic (ΔE _1/2 < 0.05 eV) electron beam with energies exceeding the single ionization threshold are presented. In the energy range from 10.8 to 12.9 eV, the energy dependences of the excitation cross sections of the studied spectral transitions exhibit the effect of postcollision interaction of slow scattered and fast emitted electrons. This process leads to an additional population of the initial levels of the spectral transitions due to the capture of a scattered electron into an excited atomic level. The energies and widths of the electronic decay of autoionizing states are estimated in the classical approach by two methods, namely, by the least squares method and by direct calculation. Calculations are performed using approximate formulas valid for different relations between the postcollision shifts of the OEF maxima and the binding energies of the atomic levels. The terms of the autoionizing atomic states responsible for the maxima observed in the OEFs of the spectral transitions are determined.     

Level density parameters of47Ti and50V

The energy distribution of the protons andα-particles from the reactions⁴⁷Ti(α, α′),⁴⁷Ti(α, p) and⁵⁰V(p, α) was measured in the angular range from 60° to 150°. The energies of the incoming particles were 15.35 MeV forα-particles and 13.85 MeV for protons. The results can be described in the frame of the statistical model of the nuclear reactions. The level density parameters could be determined by comparing the experimental data with the theoretical results. The values for the backshifted Fermi gas model are⁴⁷Ti:a=6.6±0.6MeV^?1,Δ=?0.5±0.3 MeV⁵⁰V:a=6.3±0.6 MeV^?1,Δ=?1.0±0.3 MeV.     

Model of ionization of atoms sputtered from solids

A new microscopical quantum-mechanical model of ionization of sputtered particles is described which takes into account electron excitations due to the atomic motions in the substrate. The substrate is represented by a simple atomic, metal-like, chain. Numerical calculations yield ionization probabilities R⁺ which have magnitudes close to the experimental values and have realistic velocity dependence. The function of R⁺ versus the sputtered atom ionization energy can be well approximated by the Maxwell-Boltzmann law with an effective temperature of about 3500 K.     

Quantum beats in two-photon resonance fluorescence

The two-photon resonance fluorescence spectrum of a three-level atom is shown to consist of the low frequency modes in addition to the high frequency ones in the limit of high photon densities. The spectral function for the low frequency modes consists of two lorentzian lines describing: the peak occuring at the renormalized beat frequency Δ₊ and that of the zero-photon excitation at the frequency Δ_-, where Δ_±=Δ-3Ω²/2ω_a±Ω²u/2ω_a, u²=1+(2Δω_a/Ω ²)². Here, 2Δ is the energy splitting between the two excited states, ω_a is the photon energy of the pump field and Ω is the Rabi frequency. The peak at the renormalized beat frequency Δ₊ occurs provided that the condition (2Δω_a/Ω²)² > 1 is satisfied. The two-photon laser spectroscopy is expected to be a useful tool for the observation of the low frequency modes in question.     

Coulomb energies and nuclear radii in the energy density formalism

The relation between Coulomb displacement energies,ΔE _c , andΔr=r _n -r _p , the difference between the rms radii of neutrons and protons in nuclei, is investigated within the energy density formalism (EDF). The variational equation, obtained by minimizing the Coulomb plus symmetry energies, is solved assuming the symmetry interaction is a simple functional of the local nuclear matter density. Varying parameters of the model, rather unique relation betweenΔE _c andΔr is obtained (within ±50 keV).ΔE _c isindependent ofr _ex, the rms radius of the excess neutrons distribution. Using the experimental values ofr _p and adjusting the model to reproduce the recent data onΔr (Δr∽~0.05 fm for⁴⁸Ca and²⁰⁸Pb), which are significantly smaller than those obtained from current Hartree-Fock calculations, the calculatedΔE _c agree with the experimental results. Using the value ofΔr～0.05 fm and the experimental values ofr _ex, a small compression (<0.02 fm) of the proton core in the analogue state relative to its parent state emerges, thus contributing an additional electrostatic term to the Coulomb displacement energy. The size of this relative core-compression effect depends on the values assumed forΔr andr _ex, it increases with the decreasing ofΔr and the increasing ofr _ex. IfΔr～0.05 fm the effect is large enough to remove the long standing Coulomb energy anomaly. The main result of the present work is the correlation betweenΔE _c andΔr, suggesting that the difficulties of current Hartree-Fock calculations in reproducing isotope shifts ofr _p , the small value ofr _n ?r _p and the values ofΔE _c may all be different manifestations of some missing residualp n effective interaction.     

Cluster projectile ions used for the SIMS analysis of silicon

Comparative studies of the emission of Si _n ⁺ (n = 1–11) cluster ions and impurity-containing polyatomic ions under bombardment of B-doped single crystal silicon with A _m ^? (m = 1–5) cluster ions with an energy of E ₀ = 6–18 keV are carried out. The peculiarities of sputtering an adsorbed-particle layer with cluster ions are revealed. The possibility of determining the depth distribution of adsorbed particles by analyzing the yield of sputtered heteroatomic molecular ions upon bombardment is demonstrated.     

Ionization,charge exchange,and excitation during collisions of sodium ions with argon

The absolute values of cross sections of ionization, charge exchange, and excitation during Na⁺-Ar collisions are measured in the ion energy range 0.5–10.0 keV. Experiments are performed by a modified method of transverse electric field and by the optical spectrometry method. The mechanisms for realization of inelastic channels are explained qualitatively on the basis of a schematic correlation diagram of diabatic quasi-molecular energy levels of a system of colliding particles. The quasi-molecular nature of interaction is revealed for excitation processes. The excitation mechanisms of collision partners are established. A noticeable contribution to the excitation of the 4s state of the argon atom comes from the cascade transition from upper-lying 4p and 3d levels. An oscillatory structure is observed in the excitation functions of atomic lines of ArI (104.8 and 106.7 nm). The shape of the curve describing the dependence of the charge-exchange cross section on the collision energy is explained by the presence of two nonadiabaticity regions. In the low-energy region (up to E = 2 keV), charge exchange is due to electron capture to the ground state as a result of Σ–Σ transitions, while Σ–Π transitions associated with the rotation of internuclear axis play the leading role in charge exchange in the energy range E > 3 keV. The contribution of a number of inelastic channel is estimated for the ionization process. It is found that the main contribution to ionization is associated with the decay of self-ionization states in an isolated atom.     

Vibrational dynamics of Fe in amorphous Fe–Sc and Fe–Al alloy thin films

The atomic vibrational dynamics of ⁵⁷Fe in 800-Å thick amorphous (a-) ⁵⁷Fe_0.25Sc_0.75, a-⁵⁷Fe_0.67Sc_0.33 and a-⁵⁷Fe_0.14Al_0.86 alloy thin films has been investigated at room temperature by nuclear resonant inelastic X-ray scattering (NRIXS) of synchrotron radiation. The amorphous phase has been successfully stabilized by codeposition of Fe and Sc or Al in ultrahigh vacuum onto substrates held at –140 °C during deposition. The amorphous structure of the samples was confirmed by X-ray diffraction and conversion electron Mössbauer spectroscopy. The ⁵⁷Fe-projected partial vibrational density of states, g(E), has been obtained from the measured NRIXS vibrational excitation probability, together with thermodynamic quantities such as the probability of recoilless absorption (f-factor), the average kinetic energy per Fe atom, the average force constant, and the vibrational entropy per Fe atom. A plot of the reduced density of states, g(E)/E², versus excitation energy E proves the existence of non-Debye-like vibrational modes (boson peak) with a peak energy, E _bp , in the range of 3–7 meV. Both, the boson peak height H _bp and E _bp were found to depend on the composition. Above the boson peak, g(E)/E² exhibits an exponential decrease. Our results demonstrates that the features of the boson peak depend on the amount and type of element M (M = Al, Si, Mg, Sc).     

Study of neutron induced charged particle reactions on 40K: (I). Thermal neutrons

The (n_th, α) (n_th, p) and (n_th, γα) reactions spectroscopical data are reported on ⁴⁰K and ¹⁴³Nd. For ⁴⁰K, values of σ_α = 0.39 ± 0.08 b, σ_p = 4.4 ± 0.9 b, σ_γα = 26 ± 4 mb are obtained and accurate measurements of E_α0 = 3491.7 ± 7 keV and E_po = 2232.9 ± 3 keV have been performed, α₁, and p₁ transitions and the ⁴⁰K(n_th, γp) reaction have not been observed. The ⁴⁰K(n_th, γα)³⁷Cl and ¹⁴³Nd(n_th, γα)¹⁴⁰Ce spectra are presented. The multipolarity of the primary low energy γ-rays is deduced from the γ-α spectra.     

Precise excitation energy and γ-ray decay of the lowest T = 2 state in 40K

The excitation energy of the lowest T = 2 state in ⁴⁰K has been determined as E_x = 4384.0 ± 0.3 keV from n-γ and γ-γ coincidence experiments. The state was populated with the ⁴Ar(p,n)⁴⁰K reaction at E_p = 8.30 MeV. Gamma-gamma angular correlation measurements yield unambiguous spin assignments J = 0 and 1 for the 1.64 and 2.290 MeV states, respectively. The excitation energy of the T = 2, J^π = 0⁺ state leads to a calculated mass excess of ?9120 ± 150 keV for ⁴⁰Ti.     

Sputtering of metals at ion-electron irradiation

It has been found that, in contrast to the commonly accepted opinion, simultaneous irradiation by 15-keV Ar⁺ ions and 2.5-keV electrons at temperatures above 0.5T _m (T _m is the melting temperature) induces much larger sputtering of metallic copper, nickel, and steel than irradiation only by Ar⁺ ions. The effect increases with the temperature. At T = 0.7T _m, the sputtering coefficients in the case of ion-electron irradiation are more than twice as large as the sputtering coefficients in the case of irradiation by Ar⁺ ions. The experiments on the sublimation of copper show that the sublimation rate in the case of the heating of a sample by an electron beam is higher than that in the case of heating in an electric vacuum oven. The revealed effects are explained by the electron-induced excitation of adatoms (atoms stuck over the surface, which appear owing to ion bombardment). Excited adatoms have a smaller binding energy with the surface and are sputtered more easily.     

Shallow impurity states and the free exciton binding energy in gallium phosphide

A new set of donor and acceptor ionization energies in GaP is deduced from photoexcitation spectra. Energy spectrum of donor states confirms an existence of the “camel's back” with a conduction band minima displacement ≈ 0.08(2π/a) from the X, and the corresponding energy shift ≈ 3meV. The free exciton binding energy in GaP is correctly determined: E_ex = 21 ± 2 meV.     

Electron spectroscopy of surfaces by impact of metastable He atoms: CO on Pd(110)

Deexcitation of metastable He1 2¹S (excitation energy E1 = 20.6 eV) or 2³S (E* =19.8 eV) atoms at a clean Pd(110) surface proceeds through a two-stage process (resonance ionization + Auger neutralization, RI + AN). The measured electron energy distribution reflects the self-convolution of the local density of states of the outmost atomic layer. A CO adlayer suppresses the RI step and the spectra are caused by Auger deexcitation (Penning ionization). Comparison with corresponding UPS data allows identification of the valence orbitals of the adsorbate. Emission up to the Fermi level is ascribed to contributions from the 5σ level. The effectively available excitation energy in front of the adlayer is lowered by 0.5 eV. Extensive data on the variation of the intensities from the adsorbate valence levels with angle of incidence as well as of emission are presented and are analyzed in terms of an empirical model.     

Der Zerfall des125m Xe,des127m Xe und des127g Xe

The decay of^125m Xe produced by the reaction¹²²Te(α, n)^125m Xe using a target enriched in¹²²Te (95.4%) and the decay of^127m Xe produced by the reaction¹²⁷J(d, 2n)^127m Xe have been investigated: ^125m Xe decays with a half-life ofT _1/2=(56±3) sec by ayy- cascade withE _γ1=(140.4 ±0.5) keV andE _γ2=(110.5 ±0.5) keV. The experimental conversion coefficients yield multipolarities ofE3 for the 140.4 keV isomeric transition and predominantlyM1 for the 110.5 keV-transition. ^127m Xe decays with a half-life ofT _1/2=(71±2) sec. The decay also proceeds by aγγ-cascade with an isomeric E3 transition ofE _γ1=(172.5±0.3) keV and a predominantlyM1 transition ofE _γ2=(124.6±0.3) keV. In the decay of^127g Xe an additional branching of the electron capture to a level at (618.1±0.3) keV was observed. The relative probability forK-captureP _K618/P_K375=0.40 ±0.07 yields a total transition energyQ _EC=(664 ±4)keV. A spin of 1/2⁺ was assigned to the ground state.