Scott R. Walker

Multivariate analysis method for energy calibration and improved mass assignment in recoil spectrometry

Abstract Heavy ion recoil spectrometry is rapidly becoming a well established analysis method, but the associated data analysis processing is still not well developed. The pronounced nonlinear response of silicon detectors for heavy ions leads to serious limitation and complication in mass gating, which is the principal factor in obtaining energy spectra with minimal cross talk between elements. To overcome the above limitation, a simple empirical formula with an associated multiple regression method is proposed for the absolute energy calibration of the time of flight-energy dispersive detector telescope used in recoil spectrometry. A radical improvement in mass assignment was realized, which allows a more accurate and improved depth profiling with the important feature of making the data processing much easier.

RBS and recoil spectrometry analysis of CoSi2 formation on GaAs

Abstract Mass and energy-dispersive recoil spectrometry has recently reached the state of development where it is possible to separately characterise Ga and As in GaAs samples. Since it is possible to simultaneously characterise several elements (light as well as heavy), e.g. C, O, Si, Co, Ga and As, the technique is suited for examining the depth distribution of metallisation contacts on GaAs. In a Swedish-Australian collaboration a recoil detector telescope was attached to a beamline of the FN tandem accelerator “ANTARES”, at Lucas Heights Research Laboratories, Australia. In the measurements presented here, 127 I 10+ at an energy of 77 MeV was employed to analyse GaAs samples with thin film overlayers — Si(220 nm)/Co(50 nm)/〈100〉-GaAs. A reference sample and samples annealed at 300 to 600°C were analysed. The measurements showed that CoSi 2 is formed during annealing at and above 500°C with no detectable reaction between the GaAs-substrate and the CoSi 2 overlayer.

Recoil spectrometry : ion accelerator based elemental characterisation of surface layers

Recoil Spectrometry covers a group of techniques that are very similar to the well known Rutherford backscattering Spectrometry technique, but with the important difference that one measures the recoiling target atom rather than the projectile ion. This makes it possible to determine both the identity of the recoil and its depth of origin from its energy and velocity, using a suitable detector system. The incident ion is typically high-energy (30–100MeV)35C1,81Br or127I. Low concentrations of light elements such as C, O and N can be profiled in a heavy matrix such as Fe or GaAs. Here we present an overview of mass and energy dispersive recoil Spectrometry and illustrate its successful use in some typical applications.

INTERFACIAL REACTION STUDIES OF CR, NI, TI, AND PT METALLIZATION ON INP

Interfacial reactions between (100) InP and thin films of the transition metals Cr, Ni, Pt, and Ti have been studied. A thin layer of metal was deposited onto the InP substrates using e‐beam evaporation and parts of the samples were then subjected to heat treatment in vacuum for 30 min at several temperatures up to 500 °C. Separate characterizations of the metal, In, and P depth distributions were carried out using mass and energy dispersive recoil spectrometry. The different crystalline phases observed were determined using x‐ray diffraction. The near‐noble metals (Ni, Pt) formed ternary phases, while Ti and Cr formed phosphides. The phases formed were generally stable up to 500 °C with the major exception being Pt where the ternary phase decomposed to form PtIn2, PtP2, and Pt3In7.

Empirical characterisation of mass distribution broadening in ToF-E recoil spectrometry

Abstract The mass broadening function in mass and energy dispersive recoil spectrometry using a detector telescope for time-of-flight and energy determination, has been characterised for a number of isotopes in the range A = 12 to 197. The broadening was well described by a Gaussian function where the standard deviation is given by the empirical equation: θ A ( E, A ) = C 1 + C 2 A 3/2 E − 1 + C 3 A 2 E − 2/3 + C 4 AE 1/2

Recoil spectrometry of thin film reactions in the Pd/InP system

Interfacial reactions between (100) InP and Pd were investigated as part of a systematic study aimed at investigating the stability of planar nonalloyed metallizations to InP. A 50‐nm‐thick Pd film was deposited on an InP substrate, and parts of it were subsequently thermally treated for 30 min at temperatures varying from 100 to 500 °C in steps of 50 °C. Separate characterizations of the Pd, In, and P depth distributions were obtained using mass and energy dispersive recoil spectrometry. The different phases were determined using x‐ray diffraction, and scanning electron microscopy was used to study the surface topography. It is assumed that the interaction starts in the as‐deposited sample, and definite formation of a ternary phase with the suggested composition Pd5In2P2 starts at an annealing temperature of 100 °C. At 250 °C all Pd is chemically reacted. Preferential outdiffusion of P leads to a loss of P from the surface at 500 °C, and the only phase observed in the x‐ray diffraction spectrum from the ...

Mass and energy dispersive recoil spectrometry of SixGe1−x grown by electron beam evaporation

Abstract Mass and energy dispersive recoil spectrometry employing 77 MeV 127 I ions from the recently commissioned ANTARES (FN Tandem) accelerator facility at Lucas Heights have been used to examine the isotopic composition of samples of Si x Ge 1− x grown by Electron Beam Evaporation (EBE) at the Australian National University. Analysis of RBS spectra using yield simulations indicated the presence of an unobserved element. Oxygen was considered a likely contaminant for incorporation into the Si x Ge 1− x layer by a reaction between Ge and residual oxygen in the EBE system. Determination of oxygen concentration and distribution is not normally possible with RBS but may be inferred from yield simulations using the code RUMP. Recoil spectrometry has shown the presence of oxygen in the Si x Ge 1− x layer at lower concentrations than inferred from RBS and has enabled the determination of energy spectra for individual elements.

Measurement of carbon contamination in barium strontium titanate films

Abstract Ferroelectric films are being developed for use in DRAMs. The films are often manufactured using spin-on techniques making contamination with residual carbon inevitable. The concentration of carbon is rarely known and its effect not understood. This paper describes the analysis of barium strontium titanate (BST) films for carbon content using both Mass and Energy Dispersive Recoil Spectrometry (RS) with 77 MeV 127 I ions and Nuclear Reaction Analysis (NRA) using the 12 ( 3 He,p 0 ) 14 N reaction.

Formation of thin films of CoSi2 on GaAs

CoSi2 exhibits the features of low resistivity and stability at elevated temperatures which make it interesting to employ for metallization on GaAs. The interfacial reactions in GaAs samples with thin film overlayers of Si and Co [Si(220 nm)/Co(50 nm)/(〈100〉‐GaAs)] were studied using x‐ray diffraction, scanning electron microscopy, x‐ray photoelectron spectroscopy, and mass and energy dispersive recoil spectrometry. Samples were vacuum furnace annealed for time periods between 1 and 8 h at temperatures ranging from 300 to 700 °C. It was found that a CoSi2 layer formed without observable reaction with the substrate at 500 °C and above. The excess Si (Si/Co atomic ratio of 2.41) remained near the surface as elemental Si and as SiO2 for the 500 and 600 °C annealings. For the 700 °C annealing the excess near‐surface Si was not observed.

Damage in semiconductor materials during heavy-ion elastic recoil detection analysis

Abstract Heavy-Ion Elastic Recoil Detection Analysis (HIERDA) is an analytical technique which has undergone rapid development in the past few years with the availability of high-energy Tandem accelerators for materials science applications. HIERDA has found application in the study of various semiconductor systems, particularly III–V compounds. The technique employs a high-energy heavy-ion analysing beam to eject constituent nuclei from the target material and a time of flight and energy (ToF-E) detector system to extract mass and depth of origin information from these recoiling nuclei. The present work examines the sample damage produced in semiconductor materials under typical analysis conditions. The depth distribution of damage induced by an 127 I analysing beam of varying energy (54–98 MeV) and dose, in GaAs and Si has been examined using RBS channelling and cross-sectional TEM. Channelling Contrast Microscopy (CCM) of the sample edge has been undertaken using a scanning proton microprobe.