Showing papers in "X-Ray Spectrometry in 1992"

Rayleigh and compton scattering contributions to x‐ray fluorescence intensity

Abstract: The enhancement of XRF intensity due to Rayleigh and Compton scattering in the target (which may occur after or before the photoelectric interaction) is commonly neglected in XRF spectrometry. However, this contribution can modify the XRF intensities to an appreciable extent, becoming a source of uncertainty for the computational methods of quantitative analysis. Analytical expressions are presented to describe the intensities of the corresponding four double interactions (which mix the photoelectric effect with the dominant types of scattering) that enhance the XRF intensity: Rayleigh-photoelectric, Compton-photoelectric, photoelectric-Rayleigh and photoelectric-Compton. These relationships are obtained by using an iterative solution of the Boltzmann transport equation for photons, suitable for the study of multiple scattering processes. It is shown that the first three chains contribute with discrete enhancements of the same energy of the XRF line. The last, in contrast, contributes with a short-range continuous spectrum that overlaps the low-energy tail of the peaks. This contribution can modify the line shape in a non-symmetric way, making the unfolding of peaks from energy-dispersive spectra more imprecise. The corrective terms to the XRF intensity are computed for both pure and composite materials. A common property to all of them is their azimuthal symmetry, a consequence of the isotropic behaviour of the photoelectric effect. As a rule, the chains with Rayleigh scattering predominate in the low-energy regime and contribute between 1 and 10% of the primary XRF line. The chains with Compton scattering prevail at higher energy and their extent may exceed several tens percent. All the corrections show a broad range of variation depending on the incident energy, the composition of the target, which defines its attenuation properties for all the energies, and the incidence and take-off angles.

X-ray fluorescence analysis of rocks by the fundamental parameter method

Abstract: Improvements to the fundamental parameter algorithm of x-ray fluorescence analysis are considered. The algorithm involves the following effects: primary and secondary x-ray absorption, secondary and tertiary fluorescence, excitation by scattered x-rays, excitation by photoelectrons and excitation by the divergent primary beam. The algorithm allows one to calculate the spectral intensity of x-ray tubes with both thick and electron transmission targets. A version of the algorithm that can be handled on a low-capacity computer is proposed. The algorithm was tested experimentally on the analysis of rocks, but its algorithm may be applied to different substances and materials.

Accurate description of surface ionization in electron probe microanalysis: An improved formulation

Abstract: A new analytical expression of the surface ionization Φ(0) is derived in terms of atomic number and overvoltage. This expression is obtained by means of an analytical description of the energy distribution of backscattered electrons. The Φ(0) data for Pt, W, Sb, InSb, Ni, Co and Mg have been measured using the tracer technique. The Φ(0) values obtained by using the new expression compare more favourably with experimental results, Monte Carlo simulation and backscattering data from the literature than those obtained with the previous analytical models.

Measurement of the spectral distribution of a diffraction x-ray tube with a solid-state detector

Abstract: A knowledge of the spectral distribution method emitted by an x-ray tube is important for all calculations in quantitative x-ray fluorescence analysis. A simple way of directly measuring the primary spectrum of an x-ray tube with a solid-state detector is presented. For samples which meet thin-film criteria sensitivity factors based on the measured primary spectrum were calculated and compared with experimental values.

Formulation of a universal electron probe microanalysis correction method

Abstract: A new correction method for obtaining quantitative electron probe microanalysis results is described It utilizes data provided by Monte Carlo simulations and tracer measurements to obtain a detailed picture of the way in which the x-ray emission process is influenced by the relevant experimental parameters Then empirical equations are derived which model the intensity of x-ray emission for an extremely wide range of experimental conditions, encompassing electron beam energies from 5 to 40 keV, characteristic x-ray wavelengths from 114 to 008 nm, angles of electron incidence from 90° to 40° and elements ranging from beryllium to uranium in the Periodic Table The method treats separately, as with the traditional approach, the effects of atomic number difference in specimen and standard, x-ray absorption and x-ray fluorescence The correction method was tested on an extensive range of microanalysis data and shown to work well under all conditions, providing chemical compositions which are accutate to within a few percent relative The capability of handling ultra-light element (beryllium to sodium) analysis is particularly impressive and the rms error of 38% for carbon analysis on carbides is very close to the errors incurred in measuring actual x-ray intensities The correction program is currently being developed for commercial use on IBM-compatible computers and the final stage in the evolution of the universal correction will be to extend it to deal with the analysis of thin surface films

Physical parameters for L x‐ray production cross‐sections

Abstract: Photon-induced L x-ray production cross-sections for the elements with 56 ≤ Z ≤ 92 were calculated at different energies in the range 11–60 keV using the photoionization cross-sections based on the relativistic Dirac–Hartree–Slater (RDHS) model, fluorescence yields and Coster–Kronig transition probabilities based on the RDHS model and two sets of emission rates based on the Dirac–Fock (DF) model and the Dirac–Hartree–Slater (DHS) model These calculated results were compared with the experimentally measured cross-sections to check the applicability of emission rates based on the DF and DHS models

A peak-to-background method for electron-probe x-ray microanalysis applied to individual small particles

Abstract: A peak-to-background (P/B) method was implemented for standardless analysis of bulk samples. Components of the model were selected from the literature on the basis of comparison with the present experimental results. Pouchou and Pichoir's characteristic x-ray intensity calculations combined with Small et al. bremsstrahlung model proved to be the most successful in this respect. An empirical modification seems to be necessary to improve the agreement between calculated and measured P/Bs for the Lα lines of single-element samples. For modelling the dependence of the measured P/B on the particle size, small particles are approximated as thin layers. Preliminary experimental results illustrate that even such a rough approximation works.

Background determination in wavelength-dispersive electron microprobe analysis: Some difficulties and presentation of a new analytical model

Abstract: A background (BG) model is proposed for the determination of minor and trace element compositions with an electron microprobe (wavelength-dispersive spectrometry). This model is particularly useful when a major band partially overlaps the peak corresponding to the element of interest. Its relationship is derived from a second-order polynomial and can be written as Bv = (B1 + B2)/2 + (1 + KBG)(B2 − B1)/4, where B1 and B2 represent BG measurements to be performed on two symmetrical positions with respect to the analysed peak and KBG is a coefficient to be determined on a standard (BG standard) free from the sought-for element but containing the overlapping peak. No BG standard with an approaching unknown matrix composition is required. An important feature is the shape of the overlapping band, which should be identical in both standard and unknown (at least within the energy range where measurements are performed). This last point is discussed using the example of elements whose analysed peaks are situated close to the Kβ1, 3 bands of transition metals. The shape of such bands, which are related to valence states in the x-ray emission processes, depends on the atomic chemical environment. Ti and Cr Kβ1, 3 profiles, emitted from pure metals and some oxides, are usefully compared in this respect. The capabilities of the model were tested on the determination of Mn in Cr-rich compounds (Cr Kβ1,3 strongly overlaps Mn Kα).

Kβ/Kα ratios and chemical effects in partially filled 3d‐shell elements

Abstract: The dependence of the Kβ/Kα ratio on the chemical constitution of the x-ray emitting atom of 3d-shell elements. V, Cr, Mn and Fe, was studied employing a high-resolution Si(Li) detector. A maximum deviation of 9% in the Kβ/Kα ratio for the elements V and Mn is found when compared with their respective compounds. The experimental results are found to be consistent with the theoretical values due to Brunner's model.

Lorentzian contributions to x-ray lineshapes in Si(Li) spectroscopy

Abstract: If the lineshape response of a Si(Li) detector is measured using monoenergetic x-rays provided by a monochromator, it must be convoluted by a Lorentzian of appropriate width in order to represent the lineshape encountered in directly measured x-ray spectra. It is shown that the Lorentzian component is responsible for a significant fraction of low-energy peak tailing, a phenomenon usually attributed to detector imperfections.

Measurement of relative intensities of L‐shell x‐rays of some heavy elements using Cd‐109 radioisotope source

Abstract: The relative L-shell x-ray intensities of Sm, W, Ir, Au, Hg, Pb and U were measured using a Cd-109 radioisotope source and a Si(Li) detector. The measured relative intensities were compared with the theoretically calculated values due to Scofield, computed for the present excitation energy of 22.6 keV. The experimental results were found to agree with theory in most cases.

Accuracy of calibration procedure for energy-dispersive x-ray fluorescence spectrometry

Abstract: The quality of a calibration procedure was studies in terms of the accuracy of the calibration factors. In particular, the influence of the uncertainties of the incident and emerging angles and of the mass absorption coefficients on the calibration factors was described theoretically for a calibration procedure based on standard samples of any thickness. The calculations were carried out for an x-ray tube and a radioisotope excitation set-up. Practical consequences of the inaccuracies of the calibration factors were also demostrated for a few test samples of intermediate thickness analysed by the emission–transmission method.

Figures-of-merit, their philosophy, design and use

Abstract: Figures-of-merit (FOMs) is a term which is used to define the quantitative measure of performance based on parameters which can be determined by the experimentalist. The FOM may be intuitive, based on previous experience of the analyst or his personal assessment of the importance of certain parameters, or derived, based on measurable parameters or on mathematically deduced data, mostly involving statistics. The FOM may also be a combination of intuitive and derived factors. In this review, a number of commonly used FOMs, due to several authors, are considered. They are categorized by their originators, their type (intuitive, derived or a combination therof), their primary application and the factors on which they are based. The applications discussed include both x-ray spectrometry and x-ray diffraction, but many of the principles apply equally to both.

Monte Carlo simulation of the matrix and geometrical effects in x-ray microfluorescence analysis of individual particles

Abstract: A Monte-Carlo model was developed for the simulation of x-ray fluorescence (XRF) from spherical micro-samples including primary and secondary fluorescence events. The interelement and geometrical effects in XRF microanalysis of individual mineral grains were studied. Calculations were performed for grains of Fe2O3 containing minor concentration of Ti and for ZnS grains. The results of calculation indicated that the intensities of excited characteristic x-rays depend strongly on the diameter of the irradiated particle, the incident microbeam profile and its diameter.

Defects in glasses examined by backscattered electron imaging and by x-ray wavelength and energy dispersive spectroscopy

Abstract: Typical optical defects in glasses with a high lead content were examined by backscattered electron imaging (BSI), quantitative energy-dispersive x-ray spectroscopy (EDS) and windowless EDS (WEDS) in a scanning electron microscope (SEM) and by quantitative wavelength-dispersive spectroscopy (WDS) in an electron probe micro-analyser (EPMA). Most of the images were obtained by using a backscattered electron (BSE) detector of high efficiency and excellent atomic number (Z) resolution (ΔZ = 0.3). The observations showed the presence of cords of varying shape, diameter (10–200 μm) and composition, spherical inclusions and fractures; the fractures were found to be associated with regions of materials of low Z which have developed gas under electron beam irradiation. Quantitative WDS and EDS microanalyses performed on the cords showed the presence of about 3 wt.% of Al2O3 and 2 wt.% of ZrO2, originating from the interaction, at the interface, of the glass and the refractory material used to hold the melted glass, and a lead content lower than the matrix. A high lead content was found, however, in the spherical inclusions. WDS and WEDS showed the presence of carbonates in the low-Z regions, which probably developed CO2 under electron beam irradiation. These morphological and microchemical results are of great importance in clarifying the origin, and hence finding the means of avoidance, of defect formation.

Matrix correction in x-ray fluorescence analysis by the effective coefficient method

Abstract: Up till now, the well known effective coefficient method was used by referring to a single standard. The SFP (Siemens fundamental parameter) program additionally allows this method to be employed with several standards and to analyse samples containing an element that is not analysed. A modified algorithm, called the trace model, yields an improvement in the precision of the trace element analysis. The different variants are illustrated by practical applications.

Caveat emptor: Gross refraction effects pose problems in the use of multilayers for soft x-ray spectroscopy

Abstract: First- and higher order reflections from a molybdenum-boron carbide multilayer (nominal 2d 200 A) have been studied. In the first order the apparent 2d spacing was found to increase, and then to decrease, with increasing wavelength and angle of incidence, whereas for higher orders it increased with order, converging to a limit (2d∞) of about 210.5 A. Strong refraction effects were invoked to rationalize this behaviour, and the refraction constant δλ was found to vary from 2 × 10−3 to 50 × 10−3 for the wavelength range 23.5−135 A. Experimental procedures to calibrate multilayers for soft x-ray spectroscopy, by first determining 2d∞ and then δλ values for various wavelengths, are described.

X-ray K-absorption edge studies in a-Ga30Se70 and a-Ga30Se70 -xInx

Abstract: The shift of the K-absorption edge (ΔEk) was measured in ternary alloys of Ga30Se70-xInx (x = 0, 5, 10, 15, 20) with respect to the binary Ga30Se70 alloy. It was observed that the K-edge of selenium shifted progressively towards the lower energy side as the indium concentration increased from 0 to 10 at.%. However, at higher concentrations, the relative shift started to decrease. These results are in accordance with Pauling's concept of electronegativity and indicate that the nature of the bond is iono-covalent in these glasses, as found in many crystalline solids. The composition dependence of the edge shift is also discussed in terms of the structure of the SeGa system.

A compton peak method for incident angle determination in XRF with annular excitation

Abstract: Established methods for measuring the incident angle related to XRF when radioisotope annular sources are involved are briefly discussed. As an alternative technique, a direct measurement of the incident angle from the measured spectrum of the Compton scattered energy was developed. Analysing published experimental data, the limit of validity of this new approach is defined and it is established that for the sensitivity determination only the spectral data are required.

Comprehensive major, minor and trace element analysis of a submarine polymetallic nodule by wavelength-dispersive X-ray fluorescence spectrometry

Abstract: A method was developed for the determination of major, minor and trace elements in a submarine polymetallic nodule by x-ray fluorescence spectrometry. Fusion of the sample was performed in gfraphite crucibles using a 1:3 dilution ratio with an Li2B4O7Li2CO3LiNO3 flux. Calibration was carried out with synthetic standards. The absorption-enhancement effects for 28 analytes were each corrected using influence coefficients generated by the COLA equation. The results obtained were in relatively good agreement with the recommended values for the Chinese reference material GSPN-1. The technique of sample preparation and the use of synthetic standards can also be extended to the analysis of other polymetallic ores.

Determination of impurities in nuclear-grade uranium compounds by x-ray fluorescence spectrometry

Abstract: A method in described for the simultaneous determination of trace concentrations of Ca, Cr, Cu, Fe, Mn, Ni, Mo, As, Zn, Co, Th, V and Ti in nuclear-grade uranium oxides or other uranium compounds by x-ray fluorescence spectrometry without the use of chemical treatment. The limits of detection are Mo 25, As 40, Cu 5, Ni 29, Co 17, Fe 5, Mn 30, Cr 8, Th 7, Zn 5, V 52, Ti 10 and Ca 80 ppm. Samples are prepared in the form of double-layer pellets with starch as a binding agent. Standards are prepared in a U3O8 matrix, which is more chemically stable than other uranium oxides. The dynamic ranges of the calibration graphs are very large because, for all impurity elements, they are linear over three orders of magnitude of concentration. The method is fast, since it requires no chemical treatment and only limited manipulations. The method has good precision for the impurity elements.

X-ray analysis of thin GexCyOz: H films

Abstract: Investigations by means of x-ray photoelectron spectrometry, electron probe microanalysis, x-ray fluorescence analysis, imaging x-ray fluorescence analysis, x-ray diffraction and secondary ion mass spectrometry were performed to quantify thin GexCyOz: H films. The films were prepared by r.f. plasma deposition of tetraethylgermanium in a parallel-plate system and in a two-rod discharge system with magnetron enhancement. The results of these investigations indicated that depth profiling delivers a homogeneous composition, with the exception of the outermost surface. At a constant flow-rate of tetraethylgermanium, the film density and the Ge/C atomic ration increase with increasing r.f. power density towards a maximum of 3.5 g cm−3 and 0.5, respectively. The hydrogen content does not change significantly. At low r.f. power densities there is a considerable increase in oxygen. There is no evidence for crystalline germanium carbide. Magnetron enhancement causes an increasing deposition rate and an inhomogeneous deposition in the lateral direction of the films.

Laboratory XAFS spectrometer for x-ray absorption spectra of light elements

Abstract: A new laboratory x-ray absorption fine structure (XAFS) spectrometer specially designed for measurements of the absorption spectra of light elements is presented. The spectrometer consists of a newly developed rotating-anode x-ray source of portable size, a power supply of low-voltage and high-current type, a vacuum x-ray path, a Johann-type curved monochromator, a cell unit for high-temperature measurements, and a solid-state detector (SSD) system with a single-channel analyser. The SSD system can eliminate x-rays of other orders to give x-ray absorption data containing no noise due to high frequencies. A current control method is employed to suppress characteristic x-rays emitted from filament and target materials. The performance of the spectrometer was evaluated by x-ray absorption measurements on a copper foil at the Cu K-edge at room temperature and on powdered gibbsite and aluminosilicates at the Al K-edge. The results showed that the spectrometer is capable of obtaining good quality XAFS and x-ray absorption near-edge structure (XANES) spectra at Al and Cu K-edges within about 2 h. Thus, both XAFS and XANES studies will be possible with the spectrometer for a wide range of elements from sodium to lanthanides by use of suitable monochromator crystals.

X-ray fluorescence determination of trace elements in geological materials : an iterative approach to compton scatter corrections for matrix absorption

Abstract: Compton scatter corrections are widely used for trace element determinations in geological materials where elements heavier than iron commonly occur only in low concentrations and whose effects on matrix absorption are usually ignored. Problems with the Compton correction for absorption can be seen if the concentrations of these trace elements become anomalously large, perhaps up to a few percent, although the effects can actually become significant at only a few hundred parts per million. When this occurs the simple absorption relationship, estimated at the Compton wavelength, becomes distorted. An iterative algorithm is presented that corrects absorption ratios at any wavelength as absorption edges are encountered during processing and in turn extends the Compton method to the accurate determination of anomalously high trace element concentrations.

Calculation of x-ray fluorescence and scattered primary radiation intensitie in x-ray fluorescence analysis of powder slurry-like materials

Abstract: Analytical equations for calculating x-ray fluorescence and scattered primary radiation intensities in x-ray fluorescence analysis of slurries and slurry-like materials are proposed. These equations are derived for the slurry model, in which the space distribution of the solid-phase particles is described by Poisson's law. The computation results are compared with literature data and the authors' experience.

Chemical analysis of ream defect in float glass

Abstract: To identify the source of a ream defect in float glass, the chemical composition of the ream has to be accurately determined, which can only be achieved by the electron probe microanalysis (EPMA) technique. Two ream samples were analyzed with EPMA. Because of the high mobility of the Na ion, caution had to be exercised to use a very low electron beam density during data acquisition. The results show that the reams in both cases are SiO 2 rich, but one has more Na whereas the other has less Na than the base glass, pointing to possible different causes to the ream problems. In view of the fact that the compositional difference between a ream and the base glass is very small, several approaches were used to confirm the analytical results: the statistical t-test, the digital elemental mapping and the compositional profiling. The differential etching analysis technique was also used to supplement the information on the chemistry and structure of a ream

Absorption correction in electron probe microanalysis

Abstract: The absorption correction factor f(χ) is calculated by using an empirical equation based on a new shape of the x-ray depth distribution function Φ(ρz). It assumed that Φ(ρz) increases firstly linearly with the depth ρz up to the depth ρzm at which the function Φ(ρz) reaches its maximum value Φm, and subsequently the function Φ(ρz) decreases exponentially with increasing ρz. The proposed f(χ) equation is a function of the parameters ρzm, Φm and the mean depth of x-ray production ρz. Based on Monte Carlo calculations, the variation of these parameters with the target atomic number Z, the incident electron energy E0 and the critical energy Ec was studied and empirical expressions were deduced. The calculated parameters were compared with experimental results. The factor f(χ) was calculated and compared with some measurements. The comparison gives an RMS error of 1.88%. The proposed equation was applied to 680 microanalysis data (including light elements). The results gave a relative deviation of about 3%, compared with 4.27% obtained by the quadrilateral model. For high absorbing systems (absorption factor ≤0.5), the present model gave an RMS error of 4.72% compared with 8.18% obtained by the quadrilateral model.

Compton peak shift in XRF study of graphite

Abstract: A significant shift from the theoretical position of the Compton peak was observed in a graphite sample using Rh Kα radiation. The magnitude of the Rh KαCompton peak shift depends on both the thickness of the graphite sample and the x-ray beam diaphragm used. The larger the specimen thickness or the beam diaphragm size, the greater is the magnitude of the Compton peak shift. This Compton peak shift is interpreted as being due to the change in the ‘effective’ scattering angle, and it occurs with very light elements only.