E. Steinbauer

A particularly fast TRIM version for ion backscattering and high energy ion implantation

Abstract It was an unresolved problem to theoretically treat the details of RBS spectra, e.g. the influence of double or triple scattering events, or the effects of multiple scattering at glancing incidence or exit angles. The same statement holds true for high energy implantation, if more details than just the main peak are of interest, e.g. the lateral or backward low concentration tails which result from large angle scattering. The present work explains the necessary steps to make the TRIM code even more time efficient: (1) consider only large scattering angles individually; (2) account for the remaining small angle collisions in a global way by (a) subtracting the “partial” nuclear energy loss resulting from these small angle collisions, and (b) adding up the multiple scattering angular spreads caused by such small angle deflections; (3) replacing the “magic” procedure by a fully analytic formula of good precision for large angle scattering at high energies (4) modifying the electronic energy loss subtraction to account for the variation of Se(E) along the extended free flight paths; and (5) improving the random number generator to provide in a fast way uncorrelated numbers for billions of ion histories without repetition. In some cases it was possible to increase the program speed by an order of magnitude. Examples of backscattering spectra and high energy range profiles are shown, which could not be obtained by previous existing theories or numerical codes within reasonable computing times.

The width of an RBS spectrum: influence of plural and multiple scattering

Abstract The width of RBS spectra from thin films depends on film thickness and stopping power, hence it is possible to determine one of these quantities when the other is known. For low ion energies or large film thicknesses, however, the spectrum shape significantly differs from that obtained by the single scattering calculation normally used. In this contribution, we determine quantitatively the influence of plural and multiple scattering on the width of RBS spectra: we calculate RBS spectra for protons and He ions by Monte Carlo simulation and by an analytical single scattering model using the same stopping power and interaction potential. As a result, quantitative agreement of the Monte Carlo spectra with measurements is found, while the single scattering model fails, especially near the low energy edge and below. Not only the spectrum shape but also the position of the low energy edge is affected by plural and multiple scattering. We determine this shift by varying projectile and target parameters and we find it is mostly governed by the angular width α 1 2 of the multiple scattering distribution. In bad cases, the shift may be as large as 3–4% for α 1 2 ≅ 20°.

Monte Carlo simulation of RBS spectra: Comparison to experimental and empirical results

Abstract For low projectile energies, Rutherford backscattering (RBS) spectra from thin targets show a low energy background which can be explained by plural and multiple scattering of the projectiles. Knowledge about the influence of plural and multiple scattering on the shape of RBS spectra is important in many fields (e.g. RBS depth profiling, determination of projectile energy loss). Analytic calculation of RBS spectra with full inclusion of plural and multiple scattering is impossible. Therefore we used Monte Carlo computer simulations to calculate RBS spectra for 100–200 keV protons backscattered by thin gold films (1000 A). In order to limit the number of scattering events, a nonzero minimum scattering angle (cutoff angle) was used in the simulations. The cutoff angle must be chosen so small that a further reduction does not change the shape of the total spectrum to a significant extent. Therefore, for various cutoff angles, the contribution of single, double and plural scattering to the total RBS spectrum is discussed. Furthermore, simulated spectra are compared to experimental data and to the empirical formula presented by Weber and Mommsen [2].

Rutherford backscattering beyond the single scattering model

Abstract This contribution deals with the influence of plural and multiple scattering on Rutherford backscattering (RBS) spectra, investigated by Monte Carlo simulation. Multiple scattering leads to a spread in path length and scattering angle of the projectiles scattered into the detector. Thus, the low energy edge of the RBS spectrum may be shifted noticeably. Also the kinematic energy loss and the shape of the RBS peak are influenced. Due to plural scattering, excessive projectile path lengths may occur, producing the “low energy background”, which also extends within the RBS peak. In cases, where the kinematic energy loss is large, projectiles are detectable at energies higher thin the expected RBS high energy edge (“high energy background”). Plural scattering also increases the number of projectiles contributing to the RBS peak.

Kinetic electron emission yield induced by H+ and He2+ ions versus stopping power for Al, Cu, Ag and Au☆

For H+ and He2+ ions impinging on Al, Cu, Ag and Au targets we measured simultaneously the yield, γ, of emitted electrons and the electronic energy loss, Se, in the energy range 0.5 to 4.8 MeV. The targets were prepared under high-vacuum conditions before they were transferred to an ultra-high-vacuum chamber without breaking the vacuum. The targets were sputter cleaned and their composition was examined by Auger electron spectrometry. The values of γ were obtained by current integration and Se was determined from the energy width of Rutherford backscattering spectra. For H+ ions impinging on Cu, Ag or Au and He2+ on Al and Cu, the expected proportionality between γ and Se was found within the experimental errors of 2%. For H+ ions on Al and He2+ ions on Ag and Au targets, significant deviations were observed.

A new setup for elastic recoil analysis using ion induced electron emission for particle identification

Abstract We describe a new setup for elastic recoil detection analysis (ERDA) using our recently developed particle identification method. Before the ions and elastic recoil atoms from the target reach a silicon surface barrier detector for energy analysis, they penetrate a set of thin foils (e.g. carbon). The ion induced electron emission yield from the foils depends on the nuclear charge of the penetrating ion and it is roughly proportional to the energy loss in the foil. The emitted electrons are accelerated towards a microchannel plate (MCP), which gives a signal amplitude proportional to the number of emitted electrons. This signal is measured in coincidence with the energy signal from the surface barrier detector using our dual-parameter multichannel analyzer system M2D. Since the energy resolution is not measurably deteriorated by the particle identification our setup offers optimum depth resolution for light elements. Due to the compact design large solid angles for high sensitivity can be achieved. A new measuring chamber has been built which offers considerable improvements. The ERDA scattering angle (30° or 45°) and the target orientation can be selected for optimum depth resolution or sensitivity. Element separation for light elements has been enhanced by several improvements: A new geometry of the foil setup improves the collection efficiency for ion induced electrons onto the MCP, coating of the carbon foils with insulators enhances the electron emission yield. Finally, a new data evaluation procedure has been developed in which the pulse height spectrum of the MCP is considered to be a linear combination of individual spectra from the incident ion and of the recoil atoms. The normalized shapes of these spectra are taken from calibration measurements, the intensities are then calculated using a linear fitting algorithm and finally give the depth profiles of the elements in the target. For hydrogen in near surface layers even isotopic separation is possible. Examples for 1H and 2H in a Be matrix will be given.

Electron emission yield from thin Al and insulating layers induced by 3 MeV He2+ and 3 keV electron impact

Abstract The kinetic electron yield was measured for the impact of 3 MeV He 2+ ions and 3 keV electrons on thin layers of Al on Cu, of Al 2 O 3 on Al, and of CeO 2 and CaF 2 on Si backings. The dependence of the yield on the layer thickness was determined. For Al on Cu a decreasing yield was observed for increasing Al layer thickness, since the yield of Cu is higher than that of Al. For insulating layers increasing yields were measured for increasing layer thickness. The observed yield dependencies were fitted by sums of exponential functions. The characteristic lengths of the exponential terms were interpreted as emission lengths. For Al and the metal oxides a small emission length in the range of 2–3 nm was found, for CeO 2 and electron impact on Al a second exponential term with a large emission length, 20–100 nm, was necessary to describe the measured yield dependence, which is probably due to backscattered projectile electrons and δ-electrons. The Al results are compared with computer simulations. The increasing yield of CaF 2 layers could be fitted by a function with only one exponential term and an emission length of 10.8 nm for He impact and 20.6 nm for electron impact.

Elastic recoil detection analysis for large recoil angles (LA-ERDA)

Abstract In this paper, elastic recoil detection (ERD) measurements at recoil angle of 60° using ion-induced electron emission (IEE) for particle identification are presented. In our IEE system for particle identification, recoiled target atoms and scattered projectiles penetrate a set of thin carbon foils before their energy is analyzed in a solid state detector. Particle identification is based on the fact that the total number of electrons emitted from the foils depends on the particle nuclear charge. This method is characterized by its low minimum detectable energy, which stimulated us to study ERDA at 60°. Due to collision kinematics and due to the angular dependence of the scattering cross-sections, it is expected that the sensitivity can be significantly improved. In this work, the detection efficiency of the IEE particle identification system for H recoils at energies below 1 MeV was determined. LA-ERDA measurements were performed with 4He and 12C projectiles using two different types of samples with a well-known amount and depth distribution of H atoms near the surface. Sample 1 consisted of a 50 μg/cm2 melamine layer evaporated on a flat Si substrate, sample 2 was a Si wafer with implanted H. Sensitivity and depth resolution were measured using LA-ERDA with a recoil angle of 60° and ERDA with recoil angles of 30° and 45°. The results for different recoil geometries and projectiles are discussed and compared with theoretical predictions.

Simulation of particle-induced electron emission in aluminum and copper

Abstract A Monte-Carlo computer code for the simulation of particle transport in metallic solids has been developed. Electrons or bare ion projectiles can be used. The code is able to calculate a wide variety of phenomena such as electronic energy loss, electronic energy loss straggling, particle-induced yield of emitted electrons or the statistical distribution of the number of emitted electrons per incident projectile. The theoretical models used in the simulation partially follow the basic work of Ganachaud and Cailler. However, for the loosely bound outer electrons of copper, the classical model of core ionization as it has been used by previous authors breaks down. Therefore, a fully quantum- mechanical description has been used in this work. For aluminum and copper the simulation results are compared with experimental and theoretical data. Excellent agreement is found.

Thin film analysis by low-energy ion scattering by use of TRBS simulations

Low energy ion scattering (LEIS) spectra of thin film structures are analyzed by Monte-Carlo simulations using the TRBS code. Although originally developed for the analysis of Rutherford backscattering (RBS) spectra, the TRBS code can be used to obtain valid simulations of LEIS data, which take place at energies several orders of magnitude lower than in RBS, when the appropriate adjustments are made. Experimental results from a set of Al2O3/HfO2 thin film stacks are shown, and their analysis by means of TRBS simulations is demonstrated. The authors show that the simulations yield valuable insights, especially in the case of ultrathin films <1 nm, where traditional evaluation methods using empirical models can be misleading.