Search

Depth profiling of NbxO/W multilayered samples was carried out using O2+, Cs+ or Ar+ beams of different energies (2 keV, 1 keV and 0.5 keV). The obtained depth profiles showed that sputtering with O2+ or Ar+ beams can reveal the distribution of light elements, while sputtering with Cs+ cannot show correctly their distribution if lower energy of 0.5 keV is applied. SRIM simulations show that changes in primary energy of sputter ions have much stronger effect on deposited energy incl. vacancy/ion and sputter yields than for oxygen and argon ion beam etching. It is suggested that O2+ or Ar+ sputter sources should be used when both light and heavy species are present in nano-scale multilayered targets. Copyright © 2012 John Wiley & Sons, Ltd.

Ion reflection based on a two-target-atom model is considered. In this approximation simple estimates of the angular distribution for small sliding angles are obtained (with or without thermal vibrations of atoms). The energy spectra in this case are also discussed. [Russian Text Ignored].

Irradiation by ion beams, used alone or combined with electron beams/X-rays, is in the basis of most vacuum techniques for surface analysis of materials. The ion bombardment effects in solids are however still hard to predict and consequently the ion beam techniques are less quantitative than others. Lack of proper description of some mechanisms and the large number of processes that take place concurrently during the ion beam irradiation lead to difficulties in the interpretation of the experimental data that are collected in the measurements. Computer simulations are considered a promising approach to understand the main physical and chemical phenomena, and a valuable tool in processing and interpreting the analytical data. In this review we summarize the fundamentals and the numerical approximations in modelling the characterization of solid surfaces by ion beams. The approach and the limitations of two major groups of existing codes, known as Monte Carlo and molecular dynamics techniques, are described in more detail. Special attention is paid to recent achievements in the description of phenomena that were not included in earlier stages of computer modelling, such as electron–phonon inelastic losses, bombardment enhanced diffusion and segregation, selective sputtering among others. Examples of the simulation approach to ion scattering and sputtering are given, including sputtering of oxides and multielemental targets, the sputter yield amplification effect, dynamic target modification, reactive sputtering, and shallow depth profiles.

Experimental and simulated energy distributions of Ga+ and In+ secondary ions produced by 4 keV Ne+, Ar+ and Kr+ bombardment of the AIIIBV semiconductors (GaP, GaAs, GaSb, InP, InAs and InSb) are reported. The measurements were carried out for a wide range of initial energy (up to 1000 eV) in a small solid angle along the surface normal, without applying electric field to extract the ions into the mass-energy analyser. It is shown that the energy spectra are complex, with evident high-energy hump, whose relative intensity increases with the mass of the second component (P, As, Sb) of the compound. The Sigmund–Thompson distribution cannot fit reliably these data, and a satisfactory approximation of the measured spectra was obtained with a sum of two decaying exponential functions to describe the contribution of both, the isotropic linear collision cascades and the outward knock-on atoms. The experimental results are compared with simulations based on the MARLOWE computer code.

A two-dimensional classical dynamics model for calculating the reflection coefficient of keV-ion beams backscattered from a crystal surface at glancing incidence is presented. The target is considered as consisting of one or two layers of atomic strings. A string potential of cylindrical symmetry was constructed by using the well-known iuniversali pair potential or the Molire interatomic potential. The critical transverse energy needed for a 100% reflection of light ions from the surface is assessed for the most open (11 _ 0) direction of Cu(111), Ni(111), and Al(111) crystal surfaces. The distribution of the backscattered ions on the azimuthal scattering angle exhibits the multimode structure related to the zig-zag trajectories in surface semichannels. The energy spectra are compared to the experimental data. The simulations show that the reflection coefficient is substantially dependent on the interatomic potential used but is not much affected by the model adopted for description of the inelastic losses. The inelastic-interaction model affects mostly the energy peak position, so that an estimate based on OennRobinson empirical formulae compares better with the experimental data than the original Firsov model. Resume : Nous prOsentons un modle de dynamique classique en 2-D pour calculer le

The composition and structure of an InP{100} surface in both the (1 × 1) and the reconstructed (4 × 2) phases, prepared by ion bombardment and annealing, have been examined by time-of-flight scattering and recoiling spectrometry (TOF-SARS) and low-energy electron diffraction (LEED) . Time-of-flight spectra of scattered and recoiled neutrals plus ions were collected as a function of crystal azimuthal rotation angle δ and primary beam incident angle α. Compositional analyses of the surfaces were obtained from 4 keV Ar+ scattering and recoiling spectra. Structural analyses of the phases were obtained from the azimuthal anisotropy of the δ-scans and the features of the polar incident α-scans using 4 keV Ne+ scattering from In atoms and 4 keV Kr+ for recoiling of P atoms. These azimuthal δ-scans and incident a-scans were simulated by means of a shadow cone focusing model and a modified version of the MARLOWE code, respectively. The totality of this data leads indubitably to a model in which every fourth In 〈011〉 row is missing, the In atoms are trimerized along the 〈011〉 azimuth, and the 2nd-layer P atoms exposed in the 〈011〉 troughs are dimerized. This In missing-row-trimer P dimer model (MRTD) is consistent with all of the data, is autocompensated, and has In intratrimer spacings of 3.65 ± 0.20 A and P intradimer spacings of 2.95 ± 0.20 A. The results of the simulations suggest that the two end In atoms of the trimers are relaxed downward by a minimum of 0.15 A. Two other models were considered: (1) An In missing-row (MR) model without trimers or dimers in which every fourth In 〈011〉 row is missing. (2) An In missing-row-dimer (MRD) model, similar to that proposed for the GaAs (4 × 2) structure, in which every fourth In 〈011〉 row is missing and In dimers form along 〈011〉. These MR and MRD models are inconsistent with large portions of the experimental data and the simulations.

The azimuthal anisotropy of 3 keV Ar + ions scattered from a Ni(111) single crystal surface at relatively large angle (θ = 40°) is investigated. Shadowing mode (incidence angle, α = 5° measured from the surface) was used. Correlation between the azimuthal anisotropy and the energy distribution of the scattered ions is observed both experimentally and by computer simulations using a MARLOWE-based computer code. A noticeable dependence of the primary ion azimuthal angular distribution on the energy of the scattered particles is observed.

Time was introduced in the well-known TRIM code without changing the “event-driven” logic of the program. This modified TRIM code was used to simulate the slowing-down process of fast Cu ions in amorphous Cu targets. Distributions of the radial range and slowing-down time of the ions together with other time-dependent characteristics of the process were obtained at different projectile energies. A comparison was made with the results obtained by Robinson using the time-dependent MAR-LOWE computer code [Phys. Rev. B 40 (1989) 10717].

Wafers of (100) GaAs have been implanted with 140 keV Zn+ ions at off-channeling direction up to a fluence of 1×1014 cm−2. The target temperature was kept at 110±10°C. The defect clusters and the damage depth distributions were investigated by planar and cross-sectional high-resolution transmission electron microscopy (HRTEM and XHRTEM) and were modelled by a MAR-LOWE-based computer code. The experimental mean depth of the damage profile located at about 85 nm exceeds the calculated deposited energy and vacancy distributions by more than 50%. The correlation between the deposited energy density within the cascade and the inner structure of the observed defect clusters is discussed.

This work is part of a series of investigations started a few years ago to study high fluence ion beam effects. Experimental and computer techniques are applied here to the implantation of 40 keV As + in (111) Si samples at doses ranging from 1 × 10 16 to 2 × 10 17 cm −2 . The experimental evolution of the profile was analyzed by 2 MeV RBS spectroscopy. A special procedure was used to deconvolute the RBS data for highly nonhomogeneous layers. The atomic collision cascades were also simulated by a dynamic computer code, based on sputtering yield values and impurity depth distributions resulting from the well-known static code TRIM.SP. The distributions were used as input data to a procedure which takes into account the modification of the stopping power, the surface erosion by sputtering and the lattice relaxation.

A dynamic Monte Carlo program developed for the simulation of high-fluence effects in ion bombardment of multicomponent targets is described. The program includes a code for modelling the chemically guided motion of the slowed down primaries and recoils. Examples of the application of the simulation program to Ar + sputtering of CdHgTe and SiO 2 , O + implantation into Si and Ar + sputtering of Al/Si with concurrent adsorption of O 2 are presented and compared with experimental results.

Depth distributions of displaced atoms for boron and arsenic implanted (111) silicon have been studied both by RBS channeling measurements and by a Monte Carlo simulation technique. The experiments and calculations are compared for ion energies of 80 and 160 keV and different doses. The computer code based on TRIM depth distributions takes into account the accumulation, diffusion and migration of the implantation-induced defects.

Semi-insulating Cr-doped (100) GaAs was implanted with Bi+ with energy 40 keV at room temperature in a 7 ° off-axis direction. The implanted dose was varied between 2 × 1015 cm−2 and 1 × 1016 cm−2. The profile evolution was studied experimentally by RBS measurements and theoretically by a new dynamic computer code. It is based on sputtering yield values and impurity depth distributions, obtained by using the static simulation program TRIM.SP.

The effect of 12 MeV electron irradiation of p-type Si-SiO2 structures is studied by Deep Level Transient Spectroscopy (DLTS) measurements. The DLTS spectra of non-irradiated samples exhibit one peak only corresponding to a deep level located in the forbidden gap at 0.56 eV above the valence band edge of the Si matrix. Four additional, shallower levels are found in the spectra after bombardment with high-energy electrons, whereas the peak intensity is dependent on the irradiation dose. The corresponding activation energy of the created defects, as well as the density of the traps and the electron-capture cross sections are evaluated.

structures implanted with 50-keV boron ions is studied by deep-level transient spectroscopy (DLTS) measurements. The DLTS spectra of ion-implanted samples exhibit one peak which corresponds to a deep level located in the forbidden gap of the silicon matrix at Ec-0.40 eV below the conducting band edge. New additional shallower levels are found in the spectra following bombardment by high-energy electrons, the peak intensity being dependent on the irradiation dose. The corresponding activation energy of the created defects, the density of the traps, and the electron-capture cross sections are evaluated.

A method based on a thermocouple detector measuring simultaneously the total flux of energetic particles and the portion of the charged particles is proposed and studied. These data allow the neutral fraction of the ion beam to be easily calculated and the fraction is determined for different values of the working gas pressure in Ar and SF 6 . The results obtained are in good agreement with the theoretical model, for an ion beam accelerating bias of 500 V.

It has been shown [Todorov and Fossum, J Vac Sci Technol , B6 , 466 (1988)] that ultra-thin ( d ≈ 50 A) oxide films can be grown on silicon surfaces at reduced substrate temperatures using oxygen-containing low energy ion beams ( E

Computer simulation of the reflection of a keV ion beam from a single crystal is carried out for different target temperatures. Ar ions bombarding obliquely a Cu lattice with 2 keV, and a Ni crystal with 30 keV energy are considered. The potentials used are of Born-Mayer ( A e −r/b ) and Firsov ( Br −2 ) type respectively. Uncorrelated thermal vibrations of atoms with Gaussian distribution of their displacements are taken into account. The spatial and energy distribution of the reflected ions obtained for different temperatures are discussed. The energy spectra are compared with the experimental data for the same conditions.

A secondary ion mass spectrometer with high energy, angle and mass resolution is used to study the energy and angular distributions of ions sputtered from molybdenum (110) surface under 4 keV argon beam bombardment. The sputtering at different beam orientations is also simulated by the MARLOWE computer code. The sputtering yields, in addition to the angular distribution of sputtered particles are assessed as a function of a given angle and energy.

Transmission sputtering of thin (up to 300 A) gold films bombarded by 10 to 20 keV Ar ions is simulated by using the MARLOWE computer code. The energy spectra, angle of ejection and depth of origin distributions of ejected particles are presented. The results indicate that even for a thin film the cascading of the recoils is sufficient to produce an E−2 energy spectrum and a cosine angular distribution. The evaluated transmission-sputtering yields for various film thicknesses are compared with the amount of gold retained in a silicon substrate in dynamic recoil mixing experiments.

Sputtering of copper, nickel and alumium polycrystalline targets by 40 keV Ar + ions is simulated by a NIARLOWE-based computer code. Angle-of-incidence dependence of sputtering yields and depth-of-origin distributions are calculated. The angular distributions and energy spectra of sputtered particles are similar to those found earlier for low-energy sputtering.

Inelastic energy-loss in a single collisionT e and related stopping cross-sectionT e based on the Firsov model are evaluated for different screening functions in the elastic interaction potentials. Various approximations ofT e andS e for keV-ions in solids are discussed.

The binary-collision approximation (BCA) based on the MARLOWE computer code is used to simulate the energy and angular distributions of atoms sputtered from a Rh(111) crystal surface exposed to ion bombardment. The influence of the depth-of-origin and replacement collisions on these distributions is discussed. The results are compared to the molecular-dynamic (MD) simulations and the experimental data of Winograd et al , based on energy- and angle-resolved detection of sputtered neutral atoms.

The reflection of argon ions with an energy of 50 to 500 eV from a copper single crystal is simulated on a digital computer. Three models are used to describe the interaction of the projectiles with the crystal: 1. the ion interacts in each moment only with the nearest crystal atom, 2. the ion interacts with all atoms of some microcrystallite whose atoms are assumed to be free, 3. the ion interacts with a microcrystallite of bound lattice atoms. For the Ar+–Cu and Cu–Cu interaction we use the Born-Mayer potential. A comparison of the results on the spatial and energy distributions of the ions in the different models is made. [Russian Text Ignored].

The scattering of Ar, Ne, and He ions from the Cu {100} face has been investigated by a digital computer technique; only binary elastic collisions of ions with the lattice atoms have been considered; the angular and energy distribution at various directions of incidence are discussed. The scattering by surface atomic chains and its contribution to the general picture of ion reflection were also studied.

Many theoretical works deal with recoil mixing of thin or thick Layers on a substrate [1,2,3,4]. These studies were carried out within the assumption of isotropic angular distribution of the moving low-energy cascade atoms. This is true if E ” Eo, where E is the initial energy of the primary ions and Eo amounts to between 10 and 100 eV [5]. In this case one can expect a maximum recoil mixing efficiency at the depth of maximum energy deposition by the primary ions, i.e. where FE (\(E,\vec{\Omega },\vec{r}\)) has its maximum [3,5]. There are also many experimental works which assess this suggestion [6,7,8].