Simultaneous PIXE Research Articles

Application of positive matrix factorization, multi-linear engine and back trajectory techniques to the quantification of coal-fired power station pollution in metropolitan Sydney

Over 900 fine particle Teflon filters were collected within the Sydney Basin between 1 January 2001 and 31 December 2011 and analyzed using simultaneous PIXE, PIGE, RBS and PESA techniques to determine 21 different elements between hydrogen and lead. These elements were used in positive matrix factorization (PMF) and multi-linear engine (ME) techniques together with HYSPLIT wind back trajectory techniques to quantitatively determine source fingerprints and their contributions from coal-fired power stations. The power stations were many kilometers outside the greater Sydney metropolitan area but still had a significant impact on the fine particle mass loadings measured at the sampling site within this metropolitan area. The PM2.5 eleven year average mass at the sampling site was 6.48 μg m−3. The corresponding ammonium sulfate estimate was 1.65 μg m−3 or 26% of the PM2.5 mass. By applying back trajectory data and (ME) analysis methods, two power related fingerprints, secondary sulfate (2ndrySPower) and aged industrial sulfur (IndSagedPower) were determined. These two power related fingerprints were responsible for between 14 and 18% of the total PM2.5 mass and 34–47% of the total sulfate measured at the sampling site. That is on average somewhere between a third and a half of all the sulfate measured in the greater Sydney region could be attributed to coal-fired power station emissions.

New approaches for investigating paintings by ion beam techniques

Abstract Up to now, among the IBA techniques, only PIXE has been used for analyzing paintings. However, quantitative PIXE analysis is sometimes difficult to interpret due to the layered structure, the presence of varnish and organic binder and, in some cases, discoloration of the pigments has been observed due to the interaction of the ion beam with the compounds. In order to improve the characterization of paintings, we propose some alternative experimental procedures. First of all, backscattering spectrometry (BS) and PIXE are simultaneously combined in order to collect complementary information such as layer thickness and organic compound quantification. The simultaneous PIXE and BS experiments also have the advantage of being able to analyze the same area in one experiment. This combination, implemented with an external beam, was directly applied on paintings and on painting cross-sections for the study of Italian Renaissance masterpieces. We have obtained valuable results not only on the pigment itself but also, for the first time, on the binder to pigment proportion which is not well documented in the ancient recipes. Moreover, in order to restrain beam damages due to the ion stopping power, we propose to analyze very thin painting cross-sections by a combination of PIXE–RBS and Scanning Transmission Ion Microscopy (STIM).

μ-Erda developments in order to improve the water content determination in hydrous and nominally anhydrous mantle phases

The precise knowledge of the water content in geological samples such as mantle minerals, volcanic glasses and glassy inclusions proves to be crucial information for the Earth Sciences because water has a considerable influence on physical and chemical properties of Earth’s mantle and crust. Among the nuclear microbeam techniques, elastic recoil detection analysis (ERDA) has been used at the nuclear microprobe of the Pierre Süe Laboratory (LPS) with a very good reliability for a long time to determine the water content of various materials. Previous ERDA measurements gave the total H content at a micrometric scale in the bulk of hydrous geologic samples. However, it was more difficult to characterize nominally anhydrous phases. Recent efforts allowed us to significantly improve the different steps of the ERDA analysis, from the sample preparation to the determination of the uncertainties of the resulting H concentration. A large series of very different geological samples has been measured: volcanic glasses and glassy inclusions, synthetic and natural nominally anhydrous minerals. Our new sample preparation protocol limited the thickness of the surface layer of H-pollution, leading to an easier differentiation of the bulk contribution. Simultaneous PIXE and RBS measurements allow the precise location of the interesting areas and also give information on the chemical characterization of the investigated samples, with respect to the major and minor elements. The processing of the list-mode acquisition files is made using the RISMIN software, allowing among other interesting features to check the absence of water loss under the beam. We will present the results of the H-content measurements and show the good agreement between our present results and those obtained by Fourier transform infrared spectroscopy (FTIR) for the same set of samples. The detection limit was measured on a dehydrated natural San Carlos olivine ((Mg,Fe)2SiO4) equal to 130wt ppm H2O (15wt ppm H or 300at ppm H) and the relative uncertainties on the water concentration range from 10% to 15%.

Towards truly simultaneous PIXE and RBS analysis of layered objects in cultural heritage

For a long time, RBS and PIXE techniques have been used in the field of cultural heritage. Although the complementarity of both techniques has long been acknowledged, its full potential has not been yet developed due to the lack of general purpose software tools for analysing the data from both techniques in a coherent way. In this work we provide an example of how the recent addition of PIXE to the set of techniques supported by the DataFurnace code can significantly change this situation. We present a case in which a non homogeneous sample (an oxidized metal from a photographic plate – heliography – made by Niépce in 1827) is analysed using RBS and PIXE in a straightforward and powerful way that can only be performed with a code that treats both techniques simultaneously as a part of one single and coherent analysis. The optimization capabilities of DataFurnace, allowed us to obtain the composition profiles for these samples in a very simple way.

IBA of iridescent Art Nouveau glass – comparative studies

Simultaneous PIXE, PIGE, and RBS in air were employed to characterise the surface structure of iridescent Art Nouveau glass artefacts produced around 1900 by Tiffany, USA and Loetz, Austria. Using PIXE and PIGE, the chemical composition of the bulk glass and the overlays was determined in a non-destructive manner. Furthermore, the combination of PIXE and RBS enabled the layer structure of the analysed glasses (bulk, overlays, and iridescent layers) and the thicknesses of the thin layers in the near-surface domain to be determined. The measurement and evaluation procedure is demonstrated on blue iridescent glass fragments of Tiffany and Loetz by way of example. The initial results showing similarities but also differences in the layered glass structure of Tiffany and Loetz objects are presented.

Development of a fast multi-parameter data acquisition system for microbeam analyses

A commercially available Personal Computer (PC)-based multi-parameter data acquisition system has been developed for ion microbeam analyses. The PC is equipped with a recent fast processor, a large volume of memory and a general multi-channel Analogue to Digital Converter (ADC) interface board. When a signal from an X-ray detector (PIXE), from a particle detector (RBS)or from others triggers this ADC for data acquisition. X-Y beam scanning control signals which indicate the beam position are also digitized at the same time. These data are addressed to the 3D matrices in the memory space, that consist of 1024 channels for the energy spectra and 150 × 150 pixels for corresponding the beam scan area. Real time data processing can be done in addition to this data acquisition by the fast processor with the large memory. Simultaneous PIXE and RBS elemental mapping and ion beam induced charge (IBIC) imaging on fine structure diodes have been demonstrated with this system.

Elemental micro-PIXE mapping of hypersensitive lesions in Lagenaria sphaerica (Cucurbitaceae) resistant to Sphaerotheca fuliginea (powdery mildew)

Genotypes of the Southern African cucurbit, Lagenaria sphaerica, that are resistant to powdery-mildew ( Sphaerotheca fuliginea) exhibit foliar hypersensitive (HS) lesions on inoculation with this fungal pathogen. Elemental distributions across radially symmetrical HS lesions, surrounding unlesioned leaf tissue and uninoculated leaf tissue, were obtained using the true elemental imaging system (Dynamic Analysis) of the NAC Van de Graaff nuclear microprobe. Raster scans of 3 MeV protons were complemented by simultaneous PIXE and BS point analyses. The composition of cellulose (C 6H 10O 5) was used as constant matrix composition for scans, and the sample thickness was found from BS spectra. Si and elements heavier than Ca contributed to matrix composition within HS lesions and the locally elevated Ca raised the limits of detection for some trace metals of interest. In comparison to uninoculated tissue, inoculated tissue was characterised by higher overall concentrations of all measured elements except Cu. Fully developed, 6 day-old HS lesions and the surrounding tissue could be divided into five zones, centred on the fungal infection site. Each zone was characterized by distinct local elemental distributions (either depletion, or accumulation to potentially phytotoxic levels).

Simulation of RBS spectra for quantitative mapping of inhomogeneous biological tissue

A computational algorithm, capable to simulate RBS spectra recorded during scanning proton microprobe analysis of organic specimens with inhomogeneous thickness has been developed. This program is based on an enhancement of the RUMP source code. It allows us to normalize in terms of dry specimen mass the elemental concentrations obtained with simultaneous PIXE and RBS microanalysis. This code includes non-Rutherford cross sections for proton scattering from carbon, nitrogen and oxygen. It permits the computation of matrix correction factors for the analysis of low energy X-ray emitters. In this paper, the first version of the simulation algorithm is presented as also the application to the analysis of reference organic material.

Simultaneous PIXE and PIGE analysis of metalloproteins separated by cellulose acetate electrophoresis

The Fe and N distributions of myoglobin bands prepared by cellulose acetate electrophoresis were simultaneously measured by detecting the proton induced Fe K X-rays and the γ-rays of 2313 keV energy coming from the 14N(p, p′ γ) 14N nuclear reaction at about 4 MeV proton energy. In the case of unstained bands the Fe and N distributions were very similar in shape and the integrated γ- and X-ray yields were found to be proportional to the amounts of myoglobin supplied for electrophoresis. In the cases of stained samples no quantitative conclusions could be drawn as a consequence of the drastic denaturation of the proteins caused by the staining.

Quantitative PIXE analysis using a scanning proton microbeam

PIXE analysis is often quoted as being simple, fast, accurate and multielemental. In practice, however, especially if a wide range of samples is analysed, the measurements are adversely affected by uncertainty about the sample matrix composition, thickness and homogeneity. This is particularly true for thick samples, where major corrections to the X-ray yield can depend on knowledge of the chemical composition of the specimen. This can often be circumvented by techniques such as determination of bulk composition by other methods, or comparison with standards, but all these additional processes add to the analysis time and complexity. In addition, it is necessary to assume that the sample composition is uniform over the full area of the beam spot, not necessarily true even for carefully homogenised materials. This paper describes how the use of a scanning proton microbeam (SPM) with simultaneous PIXE, backscattered particle spectroscopy and secondary electron detection can enhance the accuracy of PIXE analysis by allowing the selection of a uniform region of the sample and the determination of the matrix composition. System calibration procedures are described and examples of analyses of standard reference materials and experimental samples are presented.

Characterization of low-temperature MOCVD Cd 1−xMn xTe thin films

New organomanganese sources RMn(CO) 5 (e.g. R = CH 3 or C 6H 5CH 2) have enabled low-temperature MOCVD of Cd 1−xMn xTe(0 < x < 0.8) thin films on GaAs, GaAs/Si, InP, sapphire and glass substrates over an area > 50 cm 2 in a horizontal reactor at atmospheric pressure. Diluted magnetic Cd 1−yMn yTeCd 1−xMn xTe strained layer superlattices have been characterized by SIMS, PIXE, RBS, HRTEM of ultramicrotomed cross-sectional samples, double-crystal XRD rocking curve and satellite analysis and simulation, X-ray topography, WDX/EPMA, optical transmission and FTIR. Local thickness mapping by simultaneous PIXE and RBS, profilometry, SIMS sputter times and FTIR has provided uniformity information required for improved reactor design. Preferential (011) growth on (100) GaAs leads to anisotropic micrograting-like X-ray diffraction at glancing angles. Smoother materials is obtained on (111B), (311A), (311B), and (100) 2° off toward (110) GaAs.

Simultaneous pixe and pige analyses of aerosol samples collected in urban areas

Simultaneous pixe and pige analyses of aerosol samples collected in urban areas

Provenance studies of New Zealand obsidian artefacts

A large scale obsidian analysis program at the AAEC Research Establishment uses an automated facility with a multisample capacity and simultaneous PIXE and PIGME measurement. A pinhole filter in front of the PIXE detector ensures that up to 20 element concentrations can be obtained from the samples in a 5 min irradiation. Spectrum analysis and data handling procedures have been set up together with a variety of parametric and non-parametric statistical procedures. The clustering programs have been used with data from a previously measured S.W. Pacific and New Zealand obsidian source collection to identify the origin of artefact material. The example used to illustrate this provenancing program is of obsidian artefacts from the Kermadec and Chatham Islands.