X‐ray fluorescence analysis with a pyroelectric x‐ray generator

Several types of handy x‐ray fluorescence spectrometers are presented. The results obtained with a Niton spectrometer are presented as a goal to develop a laboratory‐made spectrometer using an Amptek Cool‐X pyroelectric x‐ray generator. A small and cheap charge‐up x‐ray emitting device and its spectrum are also presented. Handy x‐ray spectrometers are now progressing rapidly and the detection limits are in the range of a few ppm for certain elements. Copyright © 2005 John Wiley & Sons, Ltd.     

Portable x‐ray fluorescence spectrometer with a pyroelectric x‐ray generator

A portable x‐ray fluorescence spectrometer was assembled with an x‐ray generator that was driven by a 9 V dry electric battery. Several possible optimum geometries of the x‐ray generator and detector were evaluated, and the results showed that the intensity of fluorescent x‐rays was strong enough when the angle between the x‐ray generator and detector was as small as 30°. The geometrically optimized x‐ray spectrometer was applied to the analysis of paints, plastics and aluminum foils. Pigments in paint and toxic elements in plastic could be easily detected with on‐site analysis. Fe in aluminum foil was quantitatively determined down to the sub‐% level. The detection limit of Fe was 180 ppm for 100 s of measurement. Copyright © 2005 John Wiley & Sons, Ltd.     

Semiconductor Peltier‐cooled detectors for x‐ray fluorescence analysis

We discuss recent results obtained in the development of Si(Li), Si p–i–n, CdTe p–i–n and CdZnTe x‐ray detectors with Peltier coolers for fabrication of laboratory and portable XRF analyzers. The characteristics of Si(Li) Peltier‐cooled detectors are close to those of detectors cooled with the liquid nitrogen and remain the most preferred type of detectors for the tasks of x‐ray fluorescence analysis. Considerable success was obtained in the improvement of the characteristics of CdTe p–i–n detectors and CdZnTe detectors with a metal–semiconductor–metal structure, effective in the energy range up to 100 keV. The spectra of all detectors are presented. Copyright © 2004 John Wiley & Sons, Ltd.     

The capabilities of an electrically cooled energy‐dispersive x‐ray detector for quantitative x‐ray fluorescence analysis

The capabilities of the Si PIN diode x‐ray detector were determined and compared with those of a standard Si(Li) detector. The x‐ray fluorescence (XRF) analysis systems assembled with these two detectors included annular radioisotope excitation sources of Cd‐109 and Fe‐55. The systems were calibrated for sensitivity and quantification was performed with fundamental parameters software. Based on the analysis of the standard reference material NIST 2710 (Montana soil), the elemental sensitivities and the limits of detection of both systems were obtained. The elemental sensitivities of the Si PIN detector for fluorescence x‐rays in the energy range up to 10 keV were comparable to those of the Si(Li) detector. At higher fluorescence x‐ray energies the sensitivity of the Si PIN detector gradually decreased and was smaller by a factor of ～4 at 20 keV. The reason was mainly the small thickness of the sensitive volume of the Si PIN diode (0.2 mm) and therefore the smaller relative efficiency of this detector. The assessed limits of detection (LODs) were comparable for the two detectors, which was mainly due to the lower spectral background of the Si PIN detector in excitation with the Cd‐109 source as a result of its smaller sensitive thickness. The accuracy of elemental determinations for the two detectors was comparable and within the limits of the assessed uncertainties, which were calculated considering all the steps of the analysis, i.e. spectrum measurement and analysis, sensitivity calibration and quantification. Copyright © 2004 John Wiley & Sons, Ltd.     

Calculation of attenuation and x‐ray fluorescence intensities for non‐parallel x‐ray beams

With the nowadays widespreaded use of x‐ray optics in x‐ray fluorescence analysis, large convergence or divergence angles can occur. This experimental situation violates a basic assumption of the usual fundamental parameter quantification procedure. In order to take beam divergences in micro x‐ray fluorescence analysis into account, a way of calculating fluorescence intensities numerically by Monte Carlo integration is described. For three examples of typical micro‐XRF set‐ups the fluorescence intensities and their deviation from the parallel beam geometry are calculated. Furthermore, we propose a new approach with ‘equivalent angles’ which correct for the beam divergences in fundamental parameter methods. Copyright © 2003 John Wiley & Sons, Ltd.     

Micro‐beam x‐ray fluorescence analysis of individual particles with correction for absorption effects

A method of correction for absorption effects in micro‐beam x‐ray fluorescence analysis is described. A fast, energy‐dispersive, silicon drift detector (SDD) was used to measure the primary x‐ray beam transmitted through the sample. The absorption factors were calculated using the data acquired with the SDD. The possibility of using the coherently, incoherently and multiple scattered primary radiation for determining the mass of individual particles was examined. The proposed methods were validated with the use of NIST K3089 glass micro‐spheres of known composition. Copyright © 2006 John Wiley & Sons, Ltd.     

Empirical coefficients models for x‐ray fluorescence analysis of intermediate‐thickness samples

Empirical coefficients models for correction of matrix effects in intermediate‐thickness samples are presented. Two proposed mathematical models correct simultaneously matrix effects and various masses per unit area of the samples. In the first model, it is necessary to know the masses per unit area of the samples and in the second model masses are predicted from the scattered radiation. To calculate coefficients of the models, a series of multi‐element intermediate‐thickness standards with various compositions and various masses per unit area are required. The models were tested experimentally using synthetic samples pressed into pellets and geological certificate reference materials collected on a membrane filter. The calculated concentrations were in good agreement with the certificate values. Copyright © 2004 John Wiley & Sons, Ltd.     

Application of selected fundamental parameters in x‐ray fluorescence analysis

The effectiveness of selected fundamental parameters in x‐ray fluorescence analysis previously proposed by the authors was examined with a few examples. Some synthetic fused binary disks were prepared and measured to obtain the measured binary coefficients. The theoretical influence coefficients were also calculated based on the synthetic specimens and measuring conditions using those selected fundamental parameters and compared with the measured values. It was found that the theoretical influence coefficients calculated with selected fundamental parameters agreed better with measured values than those calculated with original parameters. The synthetic fused disks, some Au–Ag–Cu alloys and some Chinese rock reference materials were measured and quantitatively calculated with the program NRLXRF, that uses original parameters, and with NRL301, that uses selected fundamental parameters, using only one calibration standard for fused disks and only pure elements for alloys. The relative errors of the results from NRL301 are much smaller than those from NRLXRF. Data for the example of pressed powders of cement published in an NBS Technical Note were also calculated with NRL301 with one calibration standard. The results are comparable to those from NBSGSC obtained with seven calibration standards. Copyright © 2003 John Wiley & Sons, Ltd.     

Performance of polycapillary x‐ray lens for x‐ray sources with various spot sizes

The performance of a polycapillary x‐ray lens (PCXRL) is related to the spot size of the x‐ray source. The transmission efficiency of the PCXRL decreases generally with the increase of the x‐ray source spot size. Both the output focal spot size of a focusing PCXRL and the divergence of a parallel PCXRL increase with increasing source spot size. A large source spot size results in more ‘escape halo’, which affects the measured transmission efficiency and the measured output focal spot size. Copyright © 2007 John Wiley & Sons, Ltd.     

Improvements of the low‐energy performance of a micro‐focus x‐ray source for XRF analysis with the SEM

X‐ray Fluorescence (XRF) with a scanning electron microscope (SEM) is a valuable completion of the analytical capabilities of SEMs. Small and compact micro‐focus x‐ray sources are mounted to the microscope chamber, and the x‐ray spectra are monitored with conventional EDS systems. Up to now the x‐ray tubes used for the micro‐focus x‐ray sources are equipped with beryllium windows about 100 µm thick. The poly‐capillary x‐ray lenses have their transmission maximum at photon energies around 10 keV. It drops down in both low‐ and high‐energy ranges. Hence, L‐radiation from an Mo or Rh target will be strongly attenuated, and the excitation of fluorescence in the soft x‐ray range becomes very ineffective. A new micro‐focus x‐ray source was developed. It is characterised by a lower self‐absorption in the tube target, thin beryllium windows and an x‐ray optics having a large distance between its foci and the maximum of transmission at about 5 keV. Thus K line fluorescence of light elements becomes effectively excited by the L‐radiation from Mo or Rh tube targets. The detection limit for sodium oxide in glass was found to be below 1 mass%. Copyright © 2009 John Wiley & Sons, Ltd.     

Total reflection x‐ray fluorescence spectrometer with parallel primary beam

A new type of x‐ray optical device with two curved mirrors was tested experimentally for total reflection x‐ray fluorescence (TXRF). When focusing optics are used to increase the primary beam intensity at the sample position of TXRF spectrometers, it is always associated with an increase in the angular divergence, which is tolerable to only a limited extent. The possibility of improving the divergence by means of an additional curved mirror was reported in the past. One may hope that this additional mirror will correct some of the adverse characteristics of conventional x‐ray sources, such as the angular divergence and the intensity at the sample position. Copyright © 2005 John Wiley & Sons, Ltd.     

High‐resolution x‐ray analysis with multilayer gratings

Periodic multilayers are nowadays widely used to perform x‐ray analysis in the soft x‐ray range (photon energy lower than 1 keV). However, they do not permit to obtain high‐resolution spectra such as natural or synthetic crystals. Thus, multilayers cannot resolve interferences between close x‐ray lines. It has been shown and demonstrated experimentally that patterning a grating profile within a multilayer structure leads to a diffractive optics with improved resolving power. We illustrate the use of an Mo/B₄C multilayer grating in the Fe L and C K spectral ranges, around 700 and 280 eV, respectively. First, in the Fe L range, the improved spectral resolution enables us to distinguish the Fe Lα and Lβ emissions (separated by 13 eV). In addition, using a sample made of a mix of LiF and an iron ore, we show that it is possible to easily resolve the F K and Fe L emissions. These examples demonstrate that an improved x‐ray analysis can be obtained with multilayer gratings when there is the need to study samples having elements giving rise to close emission lines. Second, in the C K range, by comparing C Kα spectra from B₄C and cellulose, we show that the shape of the emission band is sensitive to the chemical state of the carbon atom. Copyright © 2009 John Wiley & Sons, Ltd.     

Total reflection x‐ray fluorescence analysis—a review

Total reflection x‐ray fluorescence analysis (TXRF) is a special energy‐dispersive x‐ray analytical technique extending XRF down to the ultra trace element range. Detection limits of picograms or nanograms per gram levels are reached with x‐ray tube excitation. Using synchrotron radiation as excitation source, femtogram levels are detectable, particularly important for Si wafer surface analysis. TXRF is specially suited for applications in which only a very small amount of sample is available, as only a few micrograms are required for the analysis. In this review, an overview of theoretical principles, advantages, instrumentation, quantification and application is given. Chemical analysis as well as surface analysis including depth profiling and thin‐film characterization is described. Special research results on extension to low‐Z elements, excitation with synchrotron radiation and x‐ray absorption spectroscopy (XAS) for chemical speciation at trace levels are reviewed. Copyright © 2007 John Wiley & Sons, Ltd.     

X‐ray absorption and x‐ray fluorescence for the characterization of titanium oxide‐modified HPLC silicas

Titanium oxide grafted on to the surfaces of chromatographic silica was investigated by x‐ray fluorescence (XRF) and x‐ray absorption (XAS) spectroscopy and the latter used before and after the extensive use of this material as a support for reversed‐phase high‐performance liquid chromatography (RP‐HPLC). XRF indicated the formation of a complete 2:1 monolayer whereas XAS suggested the presence of more than one titanium oxide structure. These structures show some slight modification after immobilization of PMOS and use in HPLC. Copyright © 2005 John Wiley & Sons, Ltd.     

The effects of absorption edge structure of atoms in x‐ray fluorescence analysis

The effects of the absorption edge structure in x‐ray fluorescence (XRF) manifest themselves as a very strong attenuation of the analytical line radiation when it is in the x‐ray absorption near edge structure (XANES) range. In this paper, this effect is exemplified by an ultra‐strong Ba‐caused attenuation of the Ce Lβ₁ line. This case was experimentally studied. Comparison of measured and calculated relative intensities has shown that the actual mass attenuation coefficient of the Ce Lβ₁ line in Ba is 1.6 times greater than known values. This is due to that the Ce Lβ₁ line is within the initial range of L3 Ba absorption edge, i. e. the XANES range. Such an effect of the absorption edge structure of atoms must be considered using the fundamental parameters method in quantitative XRF analysis. This paper also presents some other possible cases of this effect. Copyright © 2009 John Wiley & Sons, Ltd.     

The advantages of using digital signal processing in polarized x‐ray fluorescence analysis

Improving the specificity and productivity of XRF is of great relevance for the determination of trace elements in samples of diverse origin. The advantages of using digital signal processing in energy‐dispersive polarized x‐ray fluorescence analysis are demonstrated by comparing the instrumental sensitivities achieved with those obtained by using a conventional analogue signal processing‐based spectrometer. A compact geometry secondary target arrangement was designed to increase the effective solid angles and to reduce the distances between secondary target, sample and detector, thus achieving larger x‐ray fluxes for both the excitation and detection process, resulting in improved instrumental sensitivities. The performance of both spectrometers was evaluated for two different detectors: an Si(Li) detector and a thermoelectrically cooled passivated‐implanted planar silicon detector (X‐PIPS). The uncertainties achieved and accuracy are illustrated for the analysis of a group of sediment and organic‐origin certified reference materials using two different quantitative procedures. Copyright © 2006 John Wiley & Sons, Ltd.     

Random left‐censoring: a statistical approach accounting for detection limits in x‐ray fluorescence analysis

The use of x‐ray fluorescence techniques (XRF, TXRF, µXRF) in studies of trace element concentrations is limited by the detection limits (DL) of these methods. In this work we demonstrate that concentration measurements below the DL level, the so‐called ‘non‐detects’, can be included in data analysis by using the statistical concept of ‘censoring’, which is widely used, for instance, in survival analysis. This paper describes the non‐parametric methods of analysis of censored data using the random left‐censoring formalism, which can be used to account for the detection limits in XRF analysis. In particular, the application of the Kaplan–Meier and Nelson–Aalen estimators for the estimation of concentration distributions under censoring is discussed. These approaches are compared with the ‘reconstruction method’ developed earlier in our group to correct XRF data for the effect of detection limits. By using Monte Carlo simulations, the accuracy of the Kaplan–Meier and Nelson–Aalen estimators for censored data is discussed, in particular in the context of the estimation of the mean value and median of a concentration distribution. We demonstrate that for the ‘two‐group’ comparison of left‐censored concentrations, which is a problem of great practical importance, the log‐rank test can be used. The idea of random left censoring is applied to analyse the TXRF detection limit‐censored measurements of the concentrations of trace elements in biomedical samples. Copyright © 2004 John Wiley & Sons, Ltd.     

Determination of Mg and Ti in Ziegler–Natta catalysts by x‐ray fluorescence spectrometry

The metal content in Ziegler–Natta catalysts based on Ti and Mg derivatives could not be determined directly by x‐ray fluorescence (XRF) spectrometry owing to the instability of the catalyst. The samples were then calcined and the resultant Mg and Ti oxides were characterized by x‐ray diffraction (XRD). A series of synthetic standards were prepared with high‐purity MgO and TiO₂, since there were no available certified standards for the catalysts. Samples and standards were compressed with high‐purity H₃BO₃ flux. This protocol was shown to be reproducible, having a correlation coefficient of 0.9999. The metal contents determined by XRF were in accordance with those obtained by a complexometric titration (Mg) and a UV–visible spectrophotometric method (Ti). Copyright © 2004 John Wiley & Sons, Ltd.