The external beam microprobe facility in Florence: Set-up and performance

An external beam microprobe facility, based on a quadrupole doublet supplied by Oxford Microbeam Ltd, has been installed on a new beamline at the 3 MV single-ended Van de Graaff accelerator in Florence. The goal was to obtain a beam with a spot size on target of 10–20 μm and a current in the order of at least 1 nA, in order to allow PIXE, PIGE and RBS elemental analysis in air or in a helium atmosphere. The beam was extracted from the vacuum lines through a 0.1 μm thick Si₃N₄ window to minimise lateral straggling. The design goals have been successfully achieved; the measurements of the beam spot characteristics in vacuum as well as in air and in helium atmosphere, are here reported.     

M-shell X-ray production cross sections for PIXE applications

M-shell X-ray production cross sections by protons of energies 0.1–4.0 MeV are reported for the most intense M_β(M_4,5N_6,7), M_γ(M₃N_4,5) and M₃O_4,5 M-X ray transitions appearing in PIXE spectra. The cross sections have been measured systematically for selected heavy elements between Ta and Th (Z₂=73–90). Measured M-X-ray production cross sections were found to be universal with respect of M-shell scaled velocity ξ_M. The data are compared with available theoretical calculations of M-shell ionization by charged particles based on the plane-wave Born approximation (PWBA) and the semiclassical approximation (SCA), as well as the ECPSSR theory and relativistic RPWBA-BC which are going beyond the first order treatment. Simple parameterization of experimental proton induced M-X-ray cross sections is proposed for PIXE applications. This parameterization, being accurate within ±5%, can be used for precise determination of heavy metal concentrations by PIXE technique.     

On the accuracy of element concentrations and masses of micron sized samples determined with the Heidelberg proton microprobe

We have already tested the reliability of element-normalized PIXE (Proton Induced X-ray Emission) data on small (10–100 μm) particles that we routinely obtain with the Heidelberg proton microprobe. Thus, we here discuss the accuracy of quantitative results, i.e., absolute concentrations inferred from PIXE analyses of such particles. We investigated and reduced the effects of mechanical vibrations and of the instability of the electronic devices on the achievable minimum beam spot. We implemented into our computer code a mapping software for qualitative element distributions (PIXE) and quantitative mass images using STIM (Scanning Transmission Ion Microscopy). The STIM images determine the area density required to calculate from PIXE spectra the absolute element concentrations in thin samples. The accuracy of absolute PIXE concentrations is tested by measurements on 15 μm soda lime glass microspheres. Finally, the complete results of PIXE and STIM analyses of an interplanetary dust particle (IDP) are described.     

PESA as a complementary tool to PIXE at CTU Prague

Proton Elastic Scattering Analysis (PESA) is a simple convenient method for hydrogen analysis in thin samples. A Proton Induced X-ray Emission (PIXE) target chamber was equipped with a PIPS detector for detection of forward-scattered protons. One of the objectives was to perform PIXE and PESA analyses of air particulate targets simultaneously. Tests and calibrations were fulfilled mainly with Mylar foils 1.5–6.5 μm thick in the proton energy region between 1.35 and 2.3 MeV. The energy dependence of scattering cross section is different from the Rutherford formula. Comparison of PIXE/PESA analysis with the Guelph Scanning Proton Microprobe Laboratory at the University of Guelph, Canada on seven aerosol samples was carried out. The intercomparison results validated our PESA/PIXE quality assurance protocol. In addition, repeated measurements of Gelman Teflo^TM filters indicated a gradual increase of hydrogen content by 1 ng/cm² per 1 μC/cm² proton dose.     

Nuclear microprobe analysis of U-doped (Bi,Pb)2Sr2Ca2Cu3Oy/Ag superconducting tapes

Superconducting (Bi,Pb)₂Sr₂Ca₂Cu₃O_y/Ag tapes are doped with uranium compounds to introduce flux pinning defects from neutron-induced fission. The composition and distribution of elements in cross sections of the tapes were probed with a scanned 3 MeV proton microbeam using proton induced X-ray emission (PIXE). Distributions of the constituent elements were found to be heterogeneous on a scale of 10 μm. By combining the PIXE analysis with Rutherford backscattering spectrometry, the stoichiometry of the superconductor within the tape was measured, including oxygen from elastic scattering, revealing a departure from the desired 2223 composition. In one of the tapes, PIXE elemental maps of the Ag distribution showed diffusion of Ag into the superconductor from the enclosing jacket. Crystals of the same material, not fabricated into tapes, did not contain the contaminants and had a more ideal stoichiometry. Correlation maps between the constituent elements, deduced from the elemental maps, indicate the presence of secondary or unreacted phases.     

Microbeam line with 1.5 MeV helium ions and protons at Osaka

A microprobe for Rutherford backscattering (RBS) and particle-induced X-ray emission (PIXE) measurements has been realized by focusing 1.5 MeV helium-ion or proton beams with a demagnification system consisting of piezo-driven objective slits and a magnetic quadrupole doublet. Minimum beam spot sizes of 1.3 × 2.2 μm² for helium ions and 2.2 × 4.0 μm² for protons have been achieved. The factors which may limit the minimum spot size are discussed using a Monte Carlo simulation procedure. Rutherford backscattering image mapping of 3-dimensional structures is demonstrated.     

Aerosol composition and source apportionment in Santiago de Chile

Santiago de Chile, São Paulo and Mexico City are Latin American urban areas that suffer from heavy air pollution. In order to study air pollution in Santiago area, an aerosol source apportionment study was designed to measure ambient aerosol composition and size distribution for two downtown sampling sites in Santiago. The aerosol monitoring stations were operated in Gotuzo and Las Condes during July and August 1996. The study employed stacked filter units (SFU) for aerosol sampling, collecting fine mode aerosol (dp<2 μm) and coarse mode aerosol (210 mass of particles smaller than 10 μm) and black carbon concentration were also measured. Particle-Induced X-ray Emission (PIXE) was used to measure the concentration of 22 trace elements at levels below 0.5 ng m⁻³. Quantitative aerosol source apportionment was performed using Absolute Principal Factor Analysis (APFA). Very high aerosol concentrations were observed (up to 400 μg/m³ PM₁₀). The main aerosol particle sources in Santiago are resuspended soil dust and traffic emissions. Coarse particles account for 63% of PM₁₀ aerosol in Gotuzo and 53% in Las Condes. A major part of this component is resuspended soil dust. In the fine fraction, resuspended soil dust accounts for 15% of fine mass, and the aerosols associated with transportation activities account for a high 64% of the fine particle mass. Sulfate particle is an important component of the aerosol in Santiago, mainly originating from gas-to-particle conversion from SO₂. In the Gotuzo site, sulfates are the highest aerosol component, accounting for 64.5% of fine mass. Direct traffic emissions are generally mixed with resuspended soil dust. It is difficult to separate the two components, because the soil dust in downtown Santiago is contaminated with Pb, Br, Cl, and other heavy metals that are also tracers for traffic emissions. Residual oil combustion is observed, with the presence of V, S and Ni. An aerosol components from industrial emissions is also present, with the presence of several heavy metals such as Zn, Cu and others. A factor with molybdenum, arsenic, copper and sulfur was observed frequently, and it results from emissions of copper smelters.     

Study of basic mechanisms of semiconductor devices using ion beam induced charge (IBIC) collection

A change of wave form of current transients induced by a single heavy ion was investigated around a pn junction with 8 μm width and 10 μm length as a function of the ion incident position. Three pn junctions were made on a 3 μm thick Si epilayer (1 × 10¹⁶/cm³) grown on Si substrate and were in a line along an aluminum electrode with 10 μm spacing between the adjacent junctions. The elements of a pn junction array were irradiated with a 1 μm diameter 15 MeV C⁺ heavy ion microbeam spacing steps by 3 μm. At a bias voltage of − 10 V, 148, 91, and 54 fC were collected at the pn junction center, and at 3 μm and 4 μm from the edge of the electrode, respectively. Internal device structure was examined by IBIC (ion beam induced current) method by using a 2 MeV He⁺ ion microbeam. From the IBIC spectrum and the IBIC image, the charge collected from the open space by the diffusion process was observed in addition to the charge collected from the depletion layer of the pn junction.     

Analysis of metallic pigments by ion microbeam

Metallic paints consist of metallic flakes dispersed in a resinous binder, i.e. a light-element polymer matrix. The spatial distribution and orientation of metallic flakes inside the matrix determines the covering efficiency of the paint, glossiness, and its angular-dependent properties such as lightness flop or color flop (two-tone). Such coatings are extensively used for a functional (i.e. security) as well as decorative purpose. The ion microbeam analysis of two types of silver paint with imbedded metallic flakes has been performed to determine the spatial distribution of the aluminum flakes in paint layer. The average sizes of the aluminum flakes were 23 μm (size distribution 10–37) and 49 μm (size distribution 34–75), respectively. The proton beam with the size of 2×2 μm² at Ljubljana ion microprobe has been used to scan the surface of the pigments. PIXE mapping of Al K map shows lateral distribution of the aluminum flakes, whereas the RBS slicing method reveals tomograms of the flakes in uppermost 7 μm of the pigment layer. The series of point analysis aligned over the single flake reveal the flake angle in respect to the polymer matrix surface. The angular sensitivity is well below 1 angular degree.     

WDX-PIXE analysis of low energy X-rays using a microbeam

A high-energy resolution PIXE system developed at a heavy ion microbeam line was used to analyze low energy X-rays below 1 keV. The system is equipped with a plane crystal spectrometer with a gas flow position sensitive proportional counter (PSPC), which enables high-energy resolution PIXE analysis using a microbeam. In order to improve the detection efficiency for the low energy X-rays, the X-ray entrance window of the PSPC was replaced with a thin polymer film supported by a metal grid. As the result, the detectable energy range was extended to carbon K X-rays and chemical effect in Fe and Cu L X-rays could be detected. A preliminary result of high-energy resolution PIXE mapping of Cu mesh (#500) showed that it is possible to obtain the Cu L mapping image using a 2 MeV proton microbeam with the size of 20 × 20 μm.     

Analysis of lichen thin sections by PIXE and STIM using a proton microprobe

In order to better understand the distribution pattern of mineral elements in lichen tissues, thin sections (15 μm) of the foliose, vagrant soil lichen Xanthoparmelia chlorochroa were examined using proton microprobe Particle induced X-ray emission (PIXE). This technique was used to make two-dimensional scans, with 5 μm resolution, across tissue cross sections of the test species. Element maps for Si, P, S, Cl, K, Ca, Ti, Mn, Fe, Cu, Zn, and As have been prepared. Several elements are strongly localized in the element maps. PIXE data are complimented with STIM, light micrographs, and SEM images. Preliminary data suggest that nuclear microprobe techniques may be useful in elucidating element absorption and transport mechanisms in lichens.     

The nuclear microprobe at the University of Arizona

A nuclear microprobe has been operational at the University of Arizona since early 1994. It utilizes a magnetic quadrupole doublet (model QL-300 from Microscope Associates, Inc.) with an 11 mm diameter aperture and lens lengths of 6 cm each. The magnetic pole tips are electrically insulated to enable electric rotational alignment and beam rastering by application of varying voltages to the pole tips. Ion beams are obtained from a 5.5 MV model CN, High Voltage Engineering Corp. Van de Graaff accelerator with a Penning type ion source. Present mininum beam spot size is about 2 μm obtained with a 4 MeV H₂⁺ beam with a current of about 40 pA. To date, the instrument has been successfully used to map concentrations of Hg, Cr, and As in rabbit renal slices using PIXE.     

Quadrupole-doublet with permanent magnets for proton beam focussing

For PIXE studies on inhomogeneous samples with medium spatial resolution of about 50 μm two quadrupole doublets with Co₅Sm pole pieces for focussing 2.0 and 2.5 MeV proton beams, respectively, have been designed and tested. Technical data and first results are presented.     

Characterisation of ferromagnetic magnetic storage media surfaces by complementary particle induced X-ray analysis and time of flight-energy dispersive elastic recoil detection analysis

Thin (10 nm–1 μm) films of ferromagnetic material constitute an important class of materials that are difficult to analyse by conventional ion beam analytical (IBA) techniques because they are based on the ferromagnetic elements (Co, Fe, Mn, Ni, and Cr). The similar or overlapping isotope masses makes it difficult to separate the elemental signals using time of flight and energy dispersive elastic recoil detection (ToF-E ERD). In this exploratory study we have investigated the use of Particle Induced X-ray Emission (PIXE) measurements to refine the mass dispersive depth profile information from ToF-E ERD. The surfaces of two commercial magnetic media were investigated. One sample was a double density diskette with a coating of ferrite particles in an organic binder. The other sample was a complex C/Co/Cr/Ni–P/Al multilayer structure taken from a standard hard disc. The Lund nuclear microprobe with a 2.55 MeV proton beam was used for PIXE analysis. ToF-ERD measurements were carried out using a 55 MeV ¹²⁷I¹⁰⁺ ion beam incident at 67.5° to the surface normal. The time of flight and kinetic energy of recoils ejected at 45° to the ion beam direction was measured in a detector telescope. The findings demonstrate that by detailed analysis of the PIXE spectra it is possible to remove the ambiguities in mass assignment of the ToF-ERD data associated with the ferromagnetic elements.     

Study of particulate emissions near a steel plant in Genova by continuous sampling and PIXE hourly analysis

The particulate emissions near a large steel plant located in a densely inhabited suburb of the town of Genova (Italy) have been studied for a period of six months. We have used two-stage continuous streaker samplers and subsequent PIXE analysis with hourly resolution, to follow both seasonal and daily trends. The first streaker sampler remained installed very close to the plant cokery and furnaces, while another sampler was moved to different locations. Samples have been analysed by PIXE, deducing concentrations for elements from Na to Pb. During part of the campaign, the aerosol fractions with aerodynamic diameter (D_ae) < 2.5 μm (fine fraction) and with 2.5 < D_ae < 10 μm (coarse fraction) have been separately collected. We have measured and analysed about 8500 PIXE spectra: the steel plant emissions have been identified to some extent and resolved from other aerosol sources.     

Measurement of the H2-D mass difference with a nuclear microprobe

A 1.24 MeV deuteron (D) beam mixed with a H₂ molecular beam was separated with a microslit system of a nuclear microprobe consisting of a 100 μm diameter object and a 1 mm diameter aperture diaphragm. D was distinguished from H₂ by Rutherford backscattering (RBS) on a thin Au film. By slightly changing the magnetic field strength of the beam steerer installed in front of the object diaphragm, the maximum and the minimum RBS D/H₂ ratios were found to be 50.3 and 1.5, respectively. M/ΔM = 3.9 × 10³ was obtained as the mass resolution of the nuclear microprobe. The transmission of this system was 2 × 10⁻³.     

Application of the micro-PIXE technique for analyzing arsenic in biomat and lower plants of lichen and mosses around an arsenic mine site, at Gunma, Japan

Microhabitats of bacteria (biomat) and lower plants, such as lichen and mosses, are known to accumulate hazardous elements. Since the concentration of hazardous elements in the environment is quite low, we have applied the in-air μ-PIXE (particle induced X-ray emission) system developed in the TIARA facility of JAERI, which has low concentration detection limit of ppm, to measure As, one of the hazardous elements, distributions in biomat, lichen and mosses observed around an abandoned As mine site in Gunma, Japan to elucidate the applicability of these biomat and lower plants as bio-indicators of As. Spatial distributions of As, Fe, Si and S in all biomat, lichen and moss collected within 3 m from the mine entrance indicate that As is localized, and is associated with silicate and Fe-containing compounds. In addition, the intensity ratio of peak area for As to Fe in μ-PIXE spectrum of the moss collected from the concrete wall at 3 m downstream of the mine water discharge position is different from those of the lower plants on the rock near the closed entrance, but is the same as that of biomat formed at the mine water discharge position. This indicates that As trapped by the moss on the concrete wall probably has the same origin as the biomat. It is concluded that application of μ-PIXE analysis to the measurement of As in the lower plants and biomat gives not only the distribution of the hazardous element of As, but also the information of the origin.     

New directions in the Bartol PIXE system for studies in archaeometry

This paper describes, firstly, the physical processes of PIXE that underly its advantages as an analytical tool and the practical limitation met in the compositional patterning of ancient bronzes, irons and glasses. Controlled variation of the energy of the proton beam and the use of selective filters in the X-ray detection system allows certain background effects from the X-ray output of the dominant elements to be significantly reduced. Secondly, plans for a new beam line for the Bartol PIXE facility are described. This system will provide for a beam diameter as small as 20 μm with a laser-based optical system that will allow for the precise location of the proton beam on the artifact's surface. This system will enable one to study inhomogeneous materials, such as metal residues trapped in smelted slag, solder joints in classical jewelry and islands of metal in heavily corroded artifacts.     

PIXE analysis of trace elements in northern fur seal teeth

Trace elements in teeth of northern fur seal Collorhinus ursinus were analyzed by PIXE and micro-PIXE (μ-PIXE). Trace elements such as Mn, Fe, Zn, Cu, As, Br, Sr, and Pb were detected in the teeth, which were composed of hydroxyapatite crystals. Among these elements, the concentrations of Fe and Zn were relatively higher in winter and lower in summer and also Zn varied along with growth of individuals. These elemental fluctuation seems to correspond to the life history of the northern fur seal. It suggests that the PIXE analysis will make a powerful tool to reconstruct the life history.     

The performance of the Ljubljana ion microprobe

In August 2000 the setup of the Ljubljana ion microprobe, based on OM 150 triplet, has been completed. The beam line is installed at the 10° exit port of the 2 MeV Tandetron accelerator. It is equipped with motor driven slits, a precise five-axis goniometer and a spherically shaped measuring chamber with detectors for PIXE, PIGE, PESA, SE and RBS. In order to understand the beam optics along the complete system, consisting of the tandem accelerator and the beam line optical elements, an interactive computer code, based on a linear approximation, has been developed. The program is used both to determine the optimal parameters of the tandem focusing system in its daily use and to develop new beam line configurations. Test measurements performed on a copper grid yielded a spatial resolution of 1.0×1.5 μm² in the high current mode (30–100 pA) and 0.5×0.9 μm² in low current mode (10⁴ counts/s). First analytical results confirmed excellent performance of the new Ljubljana ion microprobe.