C. Jeynes

Simulated annealing analysis of Rutherford backscattering data

The combinatorial optimization simulated annealing algorithm is applied to the analysis of Rutherford backscattering data. The analysis is fully automatic, i.e., it does not require time-consuming human intervention. The algorithm is tested on a complex iron-cobalt silicide spectrum, and all the relevant features are successfully determined. The total analysis time using a PC 486 processor running at 100 MHz is comparable to the data collection time, which opens the way for on-line automatic analysis.

Evaluation of non-Rutherford proton elastic scattering cross-section for magnesium

The experimental data available for magnesium (p,p) elastic scattering cross-section at angles and energies suitable for Ion Beam Analysis have been evaluated using the theoretical model approach together with additional measurements and benchmark experiments. The results obtained provide the evaluated differential cross-sections for magnesium (p,p) elastic scattering in the energy region up to 2.7 MeV.

The thermoluminescence response of doped SiO2 optical fibres subjected to photon and electron irradiations.

Modern linear accelerators, the predominant teletherapy machine in major radiotherapy centres worldwide, provide multiple electron and photon beam energies. To obtain reasonable treatment times, intense electron beam currents are achievable. In association with this capability, there is considerable demand to validate patient dose using systems of dosimetry offering characteristics that include good spatial resolution, high precision and accuracy. Present interest is in the thermoluminescence response and dosimetric utility of commercially available doped optical fibres. The important parameter for obtaining the highest TL yield during this study is to know the dopant concentration of the SiO2 fibre because during the production of the optical fibres, the dopants tend to diffuse. To achieve this aim, proton-induced X-ray emission (PIXE), which has no depth resolution but can unambiguously identify elements and analyse for trace elements with detection limits approaching microg/g, was used. For Al-doped fibres, the dopant concentration in the range 0.98-2.93 mol% have been estimated, with equivalent range for Ge-doped fibres being 0.53-0.71 mol%. In making central-axis irradiation measurements a solid water phantom was used. For 6-MV photons and electron energies in the range 6, 9 and 12 MeV, a source to surface distance of 100 cm was used, with a dose rate of 400 cGy/min for photons and electrons. The TL measurements show a linear dose-response over the delivered range of absorbed dose from 1 to 4 Gy. Fading was found to be minimal, less than 10% over five days subsequent to irradiation. The minimum detectable dose for 6-MV photons was found to be 4, 30 and 900 microGy for TLD-100 chips, Ge- and Al-doped fibres, respectively. For 6-, 9- and 12-MeV electron energies, the minimum detectable dose were in the range 3-5, 30-50 and 800-1400 microGy for TLD-100 chip, Ge-doped and Al-doped fibres, respectively.

Patterned low temperature copper-rich deposits using inkjet printing

A PZT piezoelectric ceramic research drop-on-demand inkjet print head operating in bend mode was used as a means of delivering a copper precursor, vinyltrimethylsilane copper (+1) hexafluoroacetylacetonate, in a controlled and placement accurate fashion. The reagent disproportionates at low temperature (<200 °C), to deposit copper on glass. These deposits are shown to be more than 90% copper by weight by electron probe microanalysis and microbeam Rutherford backscattering spectroscopy. Microscopy shows a deposit diameter and three-dimensional profile that suggests a complex deposition and conversion mechanism. Our findings represent an important step towards the manufacture of electronic devices by entirely nonlithographic means.

Accurate RBS measurements of the indium content of InGaAs thin films

In x Ga 1~x As been analysed and values of x obtained with an estimated accuracy of ?1% in most cases. The observed variation in two measurements of a set of nine samples with a range of values of x has a mean of 1.000 and a standard deviation of 2.2%. This observed error is not inconsistent (at the 5% signi‐cance level) with the estimated error. The analytical method described is valid for many compound thin ‐lms. 1997 by John Wiley & Sons, Ltd.

Accurate determination of Quantity of Material in thin films by Rutherford backscattering spectrometry

Ion beam analysis (IBA) is a cluster of techniques including Rutherford and non-Rutherford backscattering spectrometry and particle-induced X-ray emission (PIXE). Recently, the ability to treat multiple IBA techniques (including PIXE) self-consistently has been demonstrated. The utility of IBA for accurately depth profiling thin films is critically reviewed. As an important example of IBA, three laboratories have independently measured a silicon sample implanted with a fluence of nominally 5 × 10(15) As/cm(2) at an unprecedented absolute accuracy. Using 1.5 MeV (4)He(+) Rutherford backscattering spectrometry (RBS), each lab has demonstrated a combined standard uncertainty around 1% (coverage factor k = 1) traceable to an Sb-implanted certified reference material through the silicon electronic stopping power. The uncertainty budget shows that this accuracy is dominated by the knowledge of the electronic stopping, but that special care must also be taken to accurately determine the electronic gain of the detection system and other parameters. This RBS method is quite general and can be used routinely to accurately validate ion implanter charge collection systems, to certify SIMS standards, and for other applications. The generality of application of such methods in IBA is emphasized: if RBS and PIXE data are analysed self-consistently then the resulting depth profile inherits the accuracy and depth resolution of RBS and the sensitivity and elemental discrimination of PIXE.

Amorphous-iron disilicide: A promising semiconductor

We report here the synthesis and the measurements of the microstructural and optical properties of a promising semiconductor, amorphous-iron disilicide. The material was obtained by ion-beam mixing of Fe layers on Si, with Ar8+ ions, at 300 °C. Optical absorption measurements indicate a semiconductor with a direct band gap of 0.88 eV. The significance of this discovery is that it demonstrates the existence of such a material. It should be possible to synthesize by other techniques and could be applied in large-area electronics.

Accurate determination of the stopping power of 4He in Si using Bayesian inference

Abstract Knowledge of the energy loss of ions in matter is fundamental to many applications dependent on the transport of ions in matter, particularly ion implantation and ion beam analysis techniques such as Rutherford backscattering (RBS). We have devised a new method for the experimental determination of these stopping powers: we extract both the stopping powers and their uncertainties from RBS measurements with a Markov chain Monte Carlo method based on Bayesian inference. Experimentally the method is simple, requiring only the collection of bulk RBS spectra at different energies. We have validated the method on a technologically important system in which the stopping powers are already known to a good accuracy: namely the backscattering of 4 He off Si. We have re-analysed previously reported spectra with incident beam energy between 1 and 3 MeV [Nucl. Instr. and Meth. B 161–163 (2000) 293], and determined the stopping power of 4 He in Si in an energy range 0.5–3 MeV with an absolute accuracy of 2%. We give new recommended values for the stopping power parameterisation widely used for this system valid for both H and He beams up to energy of 12 MeV.

An intercomparison of absolute measurements of the oxygen and tantalum thickness of tantalum pentoxide reference materials, BCR 261, by six laboratories

Abstract Samples of the NPL/BCR reference material of 30 nm and 100 nm of tantalum pentoxide on tantalum foil have been analysed by six independent laboratories. Four laboratories used nuclear reaction analysis involving the 16O(d, p1)17o∗ reaction, traceable to weighing and charge transfer standards, three used Rutherford backscattering analysis traceable to calculated cross sections and stopping powers, and one used elastic recoil detection analysis, again traceable through calculated parameters. Although the assessed mean result is beyond one standard deviation of the estimated error from some of the individual results, the mean results for oxygen contents and the thickness ratio agree very well with the results from a previous paper (ref. [1]: M.P. Seah et al. Nucl. Instr. and Meth. B30 (1988) 128). Individual results of the Ta and O contents of the films confirm the Ta2O5 stoichiometry within 2.5% and the thickness ratio of 0.332 to better than 2%. The average thicknesses and standard deviations of the measurements of the two films, in oxygen atoms m 2 from this study, are (1.77 ± 0.05) × 1021 and (5.38 ± 0.21) × 1021, respectively, in close agreement wi values of (1.80 ± 0.04) × 1021 and (5.42 ± 0.10) × 1021 of the previous paper. An overall analysis gives final values as (1.79 ± 0.04 and (5.40 ± 0.10) × 1021 atoms of oxygen m 2 , respectively, where the uncertainties are the 95% confidence limits.

Unambiguous automatic evaluation of multiple Ion Beam Analysis data with Simulated Annealing

Abstract We present, for the first time, simultaneous analysis of multiple Ion Beam Analysis (IBA) spectra of the same sample, which may be Rutherford Backscattering (RBS), non-Rutherford elastic scattering, and/or Elastic Recoil Detection Analysis (ERDA) spectra, ensuring a fully consistent analysis of all data. Any combination of incident ion at any energy, with or without a stopping foil before the detector, can be treated, and plural scattering corrections can be introduced. The program uses the Simulated Annealing global optimisation algorithm, and is fully automatic, with no input from the user except for the data and the experimental conditions. Two concrete examples are given, for which the analysis time was comparable to the data collection time.