2015 Atomic Spectrometry Update – a review of advances in X-ray fluorescence spectrometry and their applications

Contents

• 1 Introduction and reviews

• 2 Instrumentation
  • 2.1 Hand-held, mobile, on-line XRF techniques and planetary exploration
  • 2.2 Laboratory instruments and excitation sources
  • 2.3 Synchrotron and large scale facilities
  • 2.4 TXRF and related techniques
  • 2.5 X-ray detectors
  • 2.6 Quantification and data processing

• 3 Applications
  • 3.1 Geological and climate change
  • 3.2 Industrial minerals and consequences from mining activity
  • 3.3 Industrial materials and consequences of industrial activity
  • 3.4 Environmental
  • 3.5 Archaeological and cultural heritage
  • 3.6 Forensic
  • 3.7 Clinical
  • 3.8 Drugs
  • 3.9 Biological
  • 3.10 Thin films, coatings and nanomaterials
  • 3.11 Chemical state and speciation

• 4 Abbreviations

• References

---

Abstract

This review describes advances in the X-ray fluorescence (XRF) group of techniques published approximately between April 2014 and March 2015 and is therefore restricted to a selection of papers featuring developments in the XRF armoury. An active topic during this review period was that of imaging techniques and, more particularly of micro XRF spectrometry. Silicon-based semiconductor X-ray detectors such as SDD and Si(PIN) continue to reflect the maturity and widespread routine use of such devices. The significant expansion in studies evaluating the field use of portable XRF instrumentation in geological applications, often still proving the quality of the data, rather than adopting the technique in routine applications. New synchrotron beamlines offer previously unavailable spatial resolution and throughput for the characterisation of advanced energy materials and devices under varying temperatures and gas atmospheres. Nanomaterials feature extensively this year such as the use of nanoparticles in cancer imaging and therapy. Synchrotron radiation has become a preferred technique for the analysis of a wide range of archeological samples, artwork, museum specimens and environmental studies. There has been a substantial rise in the number of Chinese researchers investigating objects of cultural heritage, especially porcelain, glazes and glass. Advances in TXRF and related techniques continue to feature with studies on thin films and nanomaterials. Feedback on this review is most welcome and the review coordinator can be contacted using the email address provided.

---

1 Introduction and reviews

This review describes advances in the XRF group of techniques published approximately between April 2014 and March 2015 and continues the series of Atomic Spectrometry Updates in X-ray Fluorescence Spectrometry¹ that should be read in conjunction with other related reviews in the series.^2–6

Imaging techniques feature extensively this year with applications for studies related to micro- and nanotechnology. Thus Adams⁷ provided a review of developments over the last 50 years from straightforward measurements within a narrow wavelength range to complex higher-order interactions within multispectral and hyper-spectral imaging methods that produce massive amounts of data. This so called “data deluge” may occur not only from studies using large scale facilities such as synchrotron sources but also from laboratory analytical instrumentation. Such imaging applications suggest a need for increasingly complex data evaluation tools based on reliable statistics and chemometrics. The Denver X-ray Conference in 2013 celebrated the 100^th anniversary of X-ray spectroscopy with a number of retrospective oral presentations including that by Mantler whose subsequent paper⁸ summarised the development in electronics with focus on (mainly) energy-dispersive X-ray detectors and related data processing that find use in many fields including environmental, medical, archaeological, space, arts and industry. Also in 2013, the 15^th International Conference on total reflection TXRF Analysis and Related Methods was combined with the 49^th Annual Conference on X-ray Chemical Analysis (TXRF) as described by Misra⁹ in an overview of recent laboratory activities in India. Two important features of TXRF: requirements of very small amounts of sample (ng level) and multielement analytical capability were discussed in relation to handling nuclear samples. Micro-XRF with brilliant and micro-focused excitation sources were described for measurements on reactor fuel pellets that did not involve any sample preparation. New nuclear facilities in China and their analytical applications were described by Zhang et al.¹⁰ in a review covering a wide range of projects from life sciences to deep-space exploration. A critical review was offered by Wiedenbeck et al.¹¹ concerning analytical developments since 2012 in geo-standards and geo-analysis. Improvements in analytical hardware and innovative approaches to data acquisition and/or its interpretation were described in this biennial publication. Section 3.1 of this update features many other advances in applications for geological materials. Anzelmo et al.¹² published their second in a series of “Atomic Perspectives” that described the educational components and processes necessary in teaching and learning methods for sample preparation prior to analysis by the XRF range of techniques. Requirements for handling gases and liquids were discussed but the focus was mainly on solid samples and powders.

2 Instrumentation

2.1 Hand-held, mobile, on-line XRF techniques and planetary exploration

A number of researchers investigated the performance of field-portable XRF instruments in geochemical exploration applications during the current review period. Thus, Simandl et al.¹³ under-took an evaluation of hand-held XRF instrumentation for use in exploration with an emphasis on the determination of REE and related ‘speciality’ metals. They assessed precision and accuracy over a period of two years from determinations on three RMs: NIST SRM 2780 (hard rock mine waste), CGL 124 (a REE ore from the Central Geological Laboratory, Mongolia) and CANMET OKA-1 (niobium ore), and a silica blank. Most elements of interest could be determined to within 10% of the certified values of at least one CRM with differences of ≤17% and RSDs of ≤3.7% when concentrations of Ce, La, Nd, Pr and Y exceeded 1000 mg kg⁻¹. The authors concluded that this hand-held instrumentation was suitable for the exploration and development of carbonatite related REE deposits, apatite-monazite veins, peralkaline intrusions and ion absorption clay deposits with grades as low as 500 mg kg⁻¹ total REE. Arne and colleagues¹⁴ evaluated both hand-held and benchtop XRF spectrometers for the analysis of unsieved soil horizon samples from two gold exploration programmes in northern Canada, comparisons being made with laboratory ICP-MS determinations, following an aqua regia digestion. Acceptable data were obtained for As and Cu, but some correction was required to the Mo, Ni and Pb results to provide a reasonable fit to the laboratory data. The authors concluded that portable XRF of unprepared soil samples (other than drying) could be considered to be a robust method for gold exploration using particular elements such as As and Cu (with monitoring for consistency). Other pathfinder elements (Ag, Au, Bi, Sb, Te and W) occurred in soils at levels close to or below the detection limit of the current generation of field-portable devices. In two related publications, Ross et al.^15,16 investigated the use of portable XRF instrumentation to improve the lithological discrimination in the exploration analysis of drill cores, especially applied to the zinc–copper Matagami mining camp, Canada. The authors commented that for the three instruments tested, precision was quite good for a number of elements, but that accuracy was unsatisfactory using the factory calibrations. These problems were largely overcome by applying correction factors that were unique for each analyser and each project, although the effect of mineral heterogeneity in the in situ portable XRF analysis of 20 cm long core samples was much larger than that of instrumental precision. Yuan et al.¹⁷ used two portable XRF instruments to measure the geochemical distribution of Pb and Zn and other elements in hand specimens and at outcrop scale in an area of skarn formation. Statistical (PCA) analysis of the results was used to demonstrate successfully different stages of mineralisation.

Other developments in portable XRF instrumentation included a description by Figueroa and colleagues¹⁸ of a robust large area (up to 10 × 10 cm²) XRF imaging system, demonstrated in recording 2D elemental distributions in some biological samples. Guerra et al.¹⁹ described a combined portable XRF and Raman spectrometer for in situ analysis with examples of application in the fields of cultural heritage, geological and biomedical studies. Commenting on the fact that ⁵⁷Co radioisotope-excited portable XRF instrumentation has been available for over 30 years for the determination of Pb in paint with few reports documenting performance, Guimaraes et al.²⁰ evaluated ten units used in the State of New York, using NIST SRM 2579 (lead in paint) to assess precision and accuracy. On average, absolute bias was found to be within ±20% with a threshold value of 1.0 mg cm⁻² Pb. Incorrect positioning was claimed to contribute discrepancies reported by operators in the performance of some devices.

In the field of planetary science, Athiray and colleagues²¹ reported the first unambiguous evidence of enhanced levels of Na on the lunar surface, remotely sensed from the Chandrayaan X-ray spectrometer (which used X-ray emissions from the Sun as the excitation source). The authors presented abundances of the rock-forming elements, Al, Ca, Mg and Si (as well as Na) and provided a detailed description of analytical techniques, including the derivation of XRF line fluxes and the conversion to elemental abundances. From the same mission, Narendranath et al.²² reported elemental abundances in the Solar Corona using the Chandrayaan-1 X-ray Solar monitor. Turning now to the Mars Science Laboratory Curiosity Rover, Campbell et al.²³ reported changes to the basalt calibration target flown on the mission and used to monitor the performance of the alpha particle excited XRF spectrometer. These changes were caused, the authors believed, by a layer of dust (about 100 nm thick) deposited from surface fines mobilised by the engine plumes during the lander's terminal descent. The same mission has an observation tray made of titanium onto which Martian samples either drilled or scooped from the surface could be delivered for analysis by the alpha particle-excited XRF spectrometer. Berger et al.²⁴ undertook experiments in their laboratory with the flight equivalent instrumentation to evaluate limitations in the analysis of thin particulate samples deposited on this tray and reported that apart from a drop in Fe and Mn signals, no significant compositional differences were observed compared with surface measurements, except for a slight increase in Cl and S levels. Results from the Messenger X-ray spectrometer on the composition of Mercury's surface were reported by Weider et al.²⁵ during fifty-five large Solar flares. Results for the Fe/Si intensity ratio were presented with the expectation that the Fe was contained in sulfide phases and metallic iron, rather than within silicates (as on Earth). Terrestrial analogues of Martian rocks and minerals were of interest to Gazquez et al.²⁶ to better understand the formation of hydrated minerals that have recently been found on Mars and their characterisation using various portable instrumentation, including a combined XRD/XRF instrument. They reported that caves and mines in Spain and in the mining district of Iglesias-Carbonia (Sardinia) contained hydrated minerals, most of which have been detected on Mars, and which were suitable for the characterisation of instrument response. The authors considered that a Raman-LIBS combined instrument appeared to be the most powerful tool in this application. A number of multiauthor papers were published during the current review period presenting results of the analysis of stellar dust fragments collected by the Stardust Interstellar Dust Collector project. As an example of this research output, Flynn and a large number of colleagues²⁷ reported results from XRF mapping on six ‘interstellar candidates’. Enrichment of various elements was described as well as discrimination of secondary particles thought to originate from the space craft itself.

Once again, on-line applications of XRF spectrometry are restricted to a few contributions, including the publication from Choi et al.²⁸ who reported the use of a pilot-scale X-ray monitoring system for the on-line and real-time analysis of moisture in activated carbon powder, with measurements based on the total X-ray scattering intensities. Escarate and colleagues²⁹ presented a technique for the rapid and accurate XRF determination of Cu in minerals engineering applications. Measurement times were reported to be reduced from hours by current laboratory procedures to 30 s. In the on-line analysis of industrial raw materials, Gaft et al.³⁰ set out to prove that LIBS offered advantages over the XRF and prompt gamma NAA approaches, citing the ability of LIBS to give analytical results based on the surface rather than volume characteristics of the sample as one advantage of the technique, accepting that this surface would be affected by ablation.

2.2 Laboratory instruments and excitation sources

By far the most active topic in this section during the review period was that of imaging techniques and, more particularly of micro X-ray fluorescence spectrometry. The stated aim of a Swiss team³¹ was to have a high resolution μ-XRF set-up in an accessible laboratory in order to avoid the need for scarce and limited-access μ-SRXRF instruments at large facilities. The team combined an existing high resolving power von Hamos WD X-ray spectrometer with a 50 W tungsten target micro focus X-ray tube and used two half-lens polycapillary optics to provide a 50 μm beam spot on the sample. The small spot from the fully-focussing optics on the excitation side slightly improved the von Hamos WD spectrometer resolving power and also reduced spectrum background from the sample area around the beam spot, allowing the spectrometer to be operated in a slit-less mode, thereby increasing analytical sensitivity without reducing S/B. Full details of the novel instrument set-up and the spectrometer were provided and the basic performance was evaluated in a number of ways. Although the sensitivity of the focussed low power system was 5× lower than when a conventional collimated 3 kW water-cooled X-ray tube was used, the background in the new system was shown to be 50% lower, yielding comparable detection limits overall. The system was used to measure REE dopant inhomogeneity in glass fibre optic cables and was able to show inhomogeneity of Er, doped at the 1 atomic % level, although acquisition times were in the range 52 to 105 minutes per image. A study that also used a low power X-ray tube³² but in combination with a 2D CCD detector, provided both high-energy and high-spatial resolution. This novel benchtop, full-field X-ray pinhole camera system required no scanning of the beam or the sample. The small spot X-ray tube used had a tungsten target and could be operated at up to 50 kV and 100 W and was used to illuminate the whole of the sample at an incidence angle in the range 45 to 90°, depending on the experiment. The 2D detector was a commercially available back-illuminated pixellated CCD X-ray detector with 1024 × 1024 pixels each of 50 × 50 μm² and depletion depth of 40 μm, operated at −95 °C. The best energy resolution of the detector was 133 eV at 5.9 keV although the data acquisition setting used for application studies was of the order of 165 eV in order to increase the data rate. The XRF radiation from the sample to the detector was collimated with a 50 μm diameter pinhole that was laser drilled through 75 μm thick tungsten and then sandwiched between lead foils each 50 μm thick with a coaxial 50 μm diameter central hole. The camera was operated in two fixed magnifications, 6× and 0.35×, by changing only the position of the pinhole along the sample-detector beam axis. At the higher 6× magnification, samples up to 2.5× 2.5 mm² could be imaged with a spatial resolution down to 30 μm while at 0.35× demagnification, samples up to 4 × 4 cm² could be imaged with a spatial resolution of 140 μm. The detector readout rate was in the range of 100 Hz to 5 kHz and a software controlled shutter was used to block the radiation from the X-ray tube during the readout phase to avoid smearing in the image. Full details of the algorithms used to process the detector frames and to correct for effects such as charge sharing were provided by the authors. By operating under vacuum and removing beryllium and Kapton windows, the authors were able to measure X-ray energies <1 keV and showed impressive results for the separation of the N K and O K lines. However the low stopping power of the detector sensor material meant that energies >10 keV were only detected with low efficiency. The performance characteristics of the imaging analyser was demonstrated on ancient pottery samples for which a macro analysis performed with spatial and energy resolution of 170 μm and 180 eV respectively was obtained with an overall measurement time of 5000 s. A high resolution map, with 6× magnification and a field of view of 2.5× 2.5 mm² was obtained in a tedious 19 hours measurement time. The combined data were particularly useful for investigating the painting technique used for the decoration of the ancient pottery shards, making this table-top X-ray imaging camera a compact and valuable addition to the μ-XRF armoury. Although used for EDXRD, a benchtop system using a tungsten target microfocus X-ray tube operated at 160 kV and a 500 μm diameter pinhole coupled to a pixellated CCD ED X-ray detector, was also claimed by Egan et al.³³ to provide a simple full field EDXRD image with a spatial resolution of 600 μm. Data were compared with those collected on an SR beamline at the Diamond Light Source (Harwell, UK) and the authors indicated that, by means of attainable increases in X-ray tube power, useful imaging of crystallites could be achieved with this new compact benchtop pinhole camera system. A pixellated CCD X-ray detector was also used by Garrevoet et al.³⁴ for 3D XRF imaging. The detector comprised a sensor chip measuring 12.7 by 12.7 mm² with 264 × 264 energy dispersive pixels each of 48 × 48 μm² but having an impressive depletion depth of 450 μm, thereby providing much increased efficiency for high energy XRF lines, and an energy resolution of 156 eV at 5.9 keV. In this instrument, however, the excitation was by means of highly monochromatised (at 17.0 or 17.4 keV) SR on the hard X-ray beamline P06 at the 6 GeV electron storage ring PETRA III at DESY (Hamburg, Germany) making it anything but a compact and convenient μ-XRF setup. The pixellated detector provided the 2D image of the sample, which was step-scanned in the y direction through the flat and relatively wide incident fan-shaped beam to provide the 3D image and deliver an impressive spatial resolution of 8 μm with a field of view of nearly 2 × 2 mm². Impressive detection limits in the range 0.5 to 10 mg kg⁻¹ were obtained for elements in the Z range 19 to 38 in measurement times of 600 s. The impressive performance of this 3D XRF imaging set-up was demonstrated using images from sections of deep-Earth diamonds. In one case, the diamond was stepped through 41 steps each of 5 μm with a 30 minute acquisition per step, resulting in an overall measurement time of 20.5 hours. This measurement protocol produced a massive data set of 2 857 536 (41 × 264 × 264) XRF spectra, that were each deconvoluted using the AXIL software package before the images were combined into 3D elemental distribution images. The results were impressive and the authors claimed advantages over CT methods in that the individual spectra could be corrected for specific XRF effects, although that was not done in this case and would require substantial computer resources for such a huge data set. While the results were impressive, the value of resultant data would need to be very high to justify the financial cost and overcome factors such as limited beam access and substantial acquisition and data processing times. While on the topic of data correction in 3D XRF imaging, a group at TU Berlin³⁵ proposed a new method for reconstructing 3D elemental depth profiles obtained by the confocal μ-XRF spectrometry of well-defined stratified samples. The proposed method extended that previously used for the somewhat simpler case of monochromatic excitation at SR facilities to the use of polychromatic excitation by conventional laboratory X-ray tubes. The complexity of the energy distribution of Bremsstrahlung excitation and how that and the analysed volume were further affected by capillary optics led the authors to devise and present a simplified but very effective calibration protocol involving the change of probing volume with XRF peak energy. Full details of the instrument, which included a 30 W molybdenum target X-ray tube and two polycapillary lenses, were presented, as were the calibration process and the theory behind it. The authors showed that the corrected curve of sensitivity vs. XRF line energy was dominated by the transmission properties of the detection polycapillary optic and that the probing volume decreased with increasing XRF energy, due to the reduced critical angle for higher energy photons. It was concluded that using the new calibration method, one multielement glass reference sample, in addition to a set of thin foils, included enough interpolation points for a reliable calibration. The proposed method used a thick glass RM and a stratified polymer sample for validation purposes and the results showed uncertainties below 5% for thickness and concentrations could only be obtained in clearly stratified samples. In an unrelated paper³⁶ on the effect of geometry on confocal μ-XRF spectrometry, it was shown that changing the detection takeoff angle had an insignificant effect on the reconstruction of 3D maps of elemental composition and layer thicknesses. An interesting paper³⁷ on EDXRD of shotgun pellets used a commercially-available benchtop energy dispersive (ED)XRF instrument equipped with a 50 kV, 50 W X-ray tube, the output of which was collimated to 0.5 mm diameter, and a 50 mm² SDD. Data were collected by step scanning with a 5 μm step and a 50 ms dwell time resulting in a measurement time of 10 hours. X-ray spectra were stripped of characteristic XRF peaks, escape peaks, sum peaks and both Compton and Rayleigh scatter peaks, leaving only the spectrum background and low intensity scatter peaks due to diffraction effects. The authors were able to establish specific d-spacings and assign hkl values to the detected XRD peaks. Such data were shown to be a useful addition to the typical EDXRF/μ-XRF data in the study of three types of shotgun pellets of forensic interest. The non-destructive nature of μ-XRF spectrometry is particularly useful for the analysis of hazardous materials, none more so than spent nuclear fuel, which was the subject of a study by McIntosh and co-workers³⁸ who integrated microfluidic sample presentation devices into an in-house high resolution μ-EDXRF instrument. The novel spectrometer comprised a 50 W microfocus rhodium target X-ray tube coupled to a focussing optic and then to a doubly curved crystal (DCC) monochromator adjusted to pass only 20.14 keV X-rays. The detector, an SDD of 50 mm² active area and energy resolution of 130 eV at 5.9 keV, was also coupled to a DCC monochromator adjusted specifically to pass the Pu Lα line (14.279 keV). Rather than using solutions containing plutonium and other actinide elements, the authors used Sr, whose Kα lines are of similar energy (14.164 keV) to the Pu L X-ray series lines, to prove the instrument and sample presentation methodology. The microfluidics cells incorporated a sample channel with an estimated volume of 0.94 μL and a vent and were initially fabricated from readily available commercial polymer materials that were tested for their resilience to the aggressive acidic process liquids used in nuclear fuel reprocessing plants. The thin film of analyte was of the order of 0.5 mm thick, which, given the energy of the excitation and XRF lines could be regarded as an infinitely thin sample for quantitative calculation purposes, minimised absorption-enhancement effects thereby greatly simplifying calibration and quantification. Calibrations were performed using the prototype cells with Sr concentrations up to 0.1% m/v and linear calibration curves were produced in line with expectations. Repeatability for 60 s measurements were of the order of 5% RSD, being dominated by counting error but sufficiently encouraging to move to the use of commercially produced microfluidic cells that the authors showed varied by <2%. This new sample presentation and instrumentation yielded detection limits for Sr of 5 ng μL⁻¹ on a sample volume of just 1 μL and is an attractive method for hazardous materials although the technology still needs some improvement to eliminate bubble formation in the cells, which compromised the otherwise excellent accuracy of measurements. In what seems an unnecessarily elaborate excitation scheme, Ploykrachang et al.³⁹ used a 2.5 MeV proton beam to excite a copper foil target whose secondary fluorescence was then passed through a polycapillary half lens to deliver a rather uninspiring spot of 250 × 350 μm² onto the sample. Rather than use a large area, state of the art SDD, the authors used a conventional Si(PIN) energy dispersive detector in their mapping study of the distribution of Co in the leaves of an aquatic plant floating on a dilute Co solution. The authors were able to map the accumulation and distribution of Co in the leaves although the maps were only qualitative, which is disappointing given the complexity of the instrumentation. This reviewer cannot help but wonder how much better the results might have been with a simpler μ-XRF system using a state of the art copper target microfocus X-ray tube and a large area SDD. Disappointingly, as if to re-emphasise annually the fairly obvious benefits of using a plastic barrier (bag) for sample presentation in a vacuum μ-XRF system, Hardy & Scruggs⁴⁰ appear to have republished their 2013 paper.⁴¹

Although mainly used for clinical studies, developments continued during the review period on XRF Computed Tomography (CT) systems, analysis and simulation. A group at Stanford University (CA, USA) claimed an order of magnitude improvement in S/B for the benchtop XRF-CT of gold NP nodes in a water phantom.⁴² The authors developed a model for the prediction of sensitivity and background and used that in their MC prediction code for 3 different sources used for the excitation of Au and Pt K lines in gold NP and cisplatin respectively. The key improvement was to select detector positions with scatter angles >110°, which was high enough to shift the scattered Compton and Rayleigh peaks away from the analyte XRF peaks of interest, a common practice in the design of EDXRF instruments. Spectrum regions of interest were set to 1 keV and additional improvements in the reconstruction algorithms helped maintain the impressive improvements in S/B. Unsurprisingly, the highest S/B was obtained with a monochromatic source with an energy of 82 keV, which was optimum for the Au K edge, closely followed by a narrow Gaussian peak and then the broadband spectrum from a tungsten target X-ray tube operated at 110 kV. The authors showed good agreement between their MC model data and acquired CT images and suggested a way in which a further order of magnitude improvement in S/B might be achieved. Such a development would be sufficient to detect and image the typical concentration of tumour targeting agents such as gold NP and cisplatin used in chemotherapy. A similar finding was reported by Sjolin and Danielsson⁴³ who measured a six-fold improvement in S/B for Au and Pt K lines in XRF-CT by increasing the detector takeoff angle from 90 to 150° when using conventional X-ray tube pencil beam excitation. An order of magnitude improvement was obtained when the excitation was by means of a monochromatic source just above the Au K edge energy. Useful although unsurprising, the authors recommended that the designers of future XRF-CT systems bear in mind this conclusion. A MC simulation study was performed by Manohar and co-workers⁴⁴ into the optimisation of the incident X-ray spectrum from a conventional X-ray tube used in a benchtop XRF-CT system. The X-ray tube output spectrum was simulated at six settings in the range 81 to 110 kV and its cone beam was modified using foils of lead or tin with thicknesses in the range 1 to 3 mm. The authors reported that the best S/B and best overall performance was obtained with the highest kV of 110 kV and when using tin rather than lead of any given thickness. It seems obvious to this reviewer that such excitation improvements should continue, perhaps with compound primary filters or even secondary (detector) filters, and that higher resolution detectors, enabling narrower ROIs should be used to further improve S/B in XRF-CT systems, as was achieved over the years with EDXRF spectrometers. In a study also using MC simulation and actual XRF-CT measurements,⁴⁵ XRF was induced by means of a 7 mm wide incident 220 MeV proton beam and the Au K lines from gold NPs were detected by a 3 × 3 mm² CdTe ED detector at a takeoff angle of 90°. A simple linear calibration curve was observed of Au concentration vs. Au K line intensity in both experimental and MC simulation data and the authors claimed the first successful proton induced XRF-CT of aqueous solutions containing 3–5% m/v Au in vials within a small animal-sized water phantom, although vast improvements in sensitivity and S/B would be needed to reach the normal range for human clinical studies. The richness of data sets produced in 3D XRF mapping and particularly XRF-CT can be a problem to visualise in a user friendly manner. A new and very feature-rich software suite for spectrum analysis and composition imaging was described in detail⁴⁶ and its features explored briefly with a synthetic brass metal alloy and a water phantom containing insertions of barium, gadolinium and gold.

An interesting XRF imaging setup and application was described by Halls and co-workers⁴⁷ who used XRF spectrometry to determine tracer elements in two fluid jets in order to establish the nature of fluid mixing in binary liquid jets when optical techniques were no longer suitable due to the complex geometry and large effects caused by refractive index gradients. Due to the need for high sensitivity XRF and X-ray absorption measurements, the authors used high intensity SR from the APS (Argonne, IL) 7-BM bending magnet beamline that was tuned to a single energy in the range 5.1 to 12 keV by means of a double multilayer crystal monochromator. In order to observe the details of the mixing, the beam was focused to dimensions of 5 × 6 μm² FWHM, the transmitted X-ray signals being measured by a Si(PIN) diode detector and the XRF signals by an SDD mounted at a takeoff angle of 90°. The images were built up by scanning the jet across the beam and a spatial resolution of 25 μm was limited by the size of the raster steps. The authors described in detail their spectra and simple matrix correction procedures, as well as providing the details of the image reconstructions applied and recommended the technique for improving the understanding of mixing in liquid jets and for validating models describing the processes involved. Finally, Egan and colleagues⁴⁸ described the use of a CdTe sensor in a direct imaging, full field XRF colour camera. The X-ray imaging spectrometer consisted of a 1 mm thick CdTe single crystal pixellated detector (20 × 20 mm²) bump bonded to a large area ASIC packaged with a high performance data acquisition system. The detector was cooled to approximately 8 °C and the sensor had 80 × 80 pixels with a pitch of 250 μm and an energy resolution of around 800 eV at 59.5 keV. In the case of radiographic transmission (absorption) measurements a microfocus X-ray tube was operated at 80 to 100 kV and up to 3 W power, the detector being positioned behind a pinhole and with acquisition times of 2 minutes. For the XRF measurements, a medium power (200 W) X-ray tube was used and the detector was placed at a takeoff angle of 90° behind a 500 μm diameter tungsten pinhole that could be translated along the beam axis to provide an appropriate field of view and spatial resolution. The authors described in detail the spectrum processing and image reconstruction that was used to generate impressive 3D colour maps of simple, well-defined samples with image acquisition times of only a few minutes. The authors commented that faster and/or more-detailed XRF imaging could be performed by using either a higher power X-ray tube or SR.

2.3 Synchrotron and large scale facilities

Synchrotron radiation is undoubtedly the source offering the best possible conditions for X-ray analytical research and several publications dealing with new beamlines and instrument developments were published during the review period. At the Advanced Photon Source (APS) in Argonne IL, USA a suite of new beamlines was developed to study materials and devices across many length scales and under various conditions. The flagship beamline is the in situ nanoprobe (ISN) beamline described by Maser et al.,⁴⁹ which will provide in situ and in operando characterisation of advanced energy materials and devices under varying temperatures, gas atmospheres and applied fields, at previously unavailable spatial resolution and throughput. The authors gave examples covering: material systems including inorganic and organic photovoltaic systems, advanced battery systems, fuel cell components, nanoelectronic devices and advanced building materials. Nano-focusing mirrors in Kirkpatrick–Baez geometry provided increased photon flux by several orders of magnitude with a spatial resolution of 50 nm. Diffractive optics, such as zone plates and/or Laue lenses delivered a beam with highest spatial resolution of 20 nm. Coherent diffraction methods were used to study even small specimen features with sub-10 nm length scale. In this paper the optical concept, the technical design of the ISN beamline and, furthermore, the application of hard X-ray microscopy to study defects in multicrystalline solar cells were described. Using the APS nano-diffraction beamline Manickaraj et al.⁵⁰ unveiled through the use of nano XRD and XRF elemental mapping, that contrary to popular belief, Sr did not seem to interfere with the Twin Plane Re-entrant edge growth mechanism of eutectic silicon, but evolved as the Al₂Si₂Sr phase during the eutectic reaction at the boundary between the eutectic Si and Al grains. Zhang et al.⁵¹ described a novel efficient off-line sample positioning system for the hard μ-XRF beamline at the Shanghai Synchrotron Radiation Facility. The system comprised an off-line sample microscope, the on-line sample experiment system and a high precision positioning holder. The authors claimed that it was possible for the first time to achieve a sample off-line positioning at the μm scale. By comparing the differences of coordinates of the gold mesh nodes images from off-line microscope and images from XRF mapping, the accuracy of the positioning system was verified and found to be 1.3 μm for the x-axis and 2.5 μm for the z-axis. It was reported by Llorens et al.⁵² that the MARS beamline at SR facility SOLEIL (Paris, France) was dedicated to the characterisation of radioactive material with the great advantage that some 380 radionuclides could be investigated at the same place by these different X-ray techniques. This unique European facility offered a wide energy range from around 3.5 to 36 keV enabling the analysis of the elements with K-edges from K to Cs and with L-edge from Cd to Am and beyond and was optimised for XANES/EXAFS, powder diffraction and XRF spectrometry, high energy resolution fluorescence dejectedness, YES and μ-AS/XRD. A description of the beamline and some scientific examples of XAS studies were presented. The SIRIUS beamline at the SOLEIL synchrotron was devoted to surface X-ray scattering studies of organic molecular films as reported by Fontaine et al.,⁵³ most of the studies used grazing incident (GI) techniques. Among them GI-XRD, GI-SAXS and TXRF spectrometry may be applied to organised molecular films including Langmuir monolayers on liquid substrates. Thanks to high resolution Soller slits, the in-plane scattering wave vector resolution was better than 0.02 nm⁻¹ and the GI-SAXS scattering wave vector lower than 0.05 nm⁻¹. Special attention was paid to enable simultaneous or at least consecutive use of all these techniques on the same sample. Although SRXRF spectrometry is a powerful elemental analytical tool, yet synchrotrons are large multiuser facilities that are generally not amenable to modifications. Barberie et al.⁵⁴ was able to show how modifications could be made to improve SRXRF analysis by simply including X-ray filters or removing monochromators. In this study four easily applied beamline configurations for the analysis of three representative environmental samples, namely a thin aerosol sample, an intermediate thickness biological sample and a thick RE mineral specimen were evaluated. The results showed that the “white beam” configuration, was the optimal configuration for the thin film samples. The filtered white beam configuration removed the lower energy X-rays from the excitation beam yield better sensitivity for higher Z elements. A filter in front of the detector sacrificed the lower energy part of the spectrum for improved sensitivity in the higher energy end (26–48 keV). The use of a monochromatic beam, which tends to be the standard mode of operation for most SRXRF spectrometers gave surprisingly the least sensitive analysis. Also at the Brazil synchrotron in Campinas, a confocal XRF setup was established and used by Sosa et al.⁵⁵ from the group of the Cordoba National University in Argentina. This spectrometer was equipped with a polycapillary optic to focus the beam and a second polycapillary optic in confocal geometry. The parameters affecting the performance of the spectrometer were studied and a simplified calibration method was developed to perform depth profile analysis.

Synchrotron radiation was applied to various samples together with specific scientific questions such as the fascinating research experiment performed at APS presented by Harris et al.⁵⁶ The authors described degradation effects on solid oxide fuel cells (SOFC) after exposure to μg g⁻¹ levels of hydrogen sulfide at elevated temperatures. Performing SR based X-ray nano-CT and XRF techniques on a section of a SOFC Ni-YSZ (yttria-stabilised zirconia) anode, the XRF results provided elemental identification and coarse spatial mapping, further the nano-CT was used to map the detailed 3D spatial distribution of Ni, YSZ, and a nickel sulfur poisoning phase. The nickel sulfur layer was found to form a scale covering most of the exposed nickel surface, blocking most of the fuel reformation and hydrogen oxidation reaction sites. The results provided strong evidence of the detrimental effects of 100 μg g⁻¹ hydrogen sulfide on typical Ni-YSZ anode materials. The advantages of using SR for studies in fluid dynamics in many flowfields and especially multiphase flows was presented in a review by Kastengren and Powell.⁵⁷ The relevant properties of SR and merits, such as tuneability to a monochromatic X-ray beam with high flux plus types and capabilities of typical X-ray optics were discussed. Four major diagnostic techniques were described in detail: (1) X-ray radiography providing quantitative measurements of density in variable-density flows, (2) X-ray phase contrast imaging was used to visualise multiphase flows with high spatial and temporal resolution, (3) μ-XRF spectrometry showed significant promise to study mixing in single-phase and multiphase flows and (4) SAXS representing a powerful technique to examine small scale particles in flows. Wang et al.⁵⁸ introduced a versatile framework to identify targeted cellular structures from datasets, too complex for manual analysis like most XRF microscopy datasets. This approach allowed locating, identifying and refining positions and whole areas of cell structures based on element contents measured by XRF microscopy. It was shown that by initialising with only a handful of prototype cell regions, this approach could obtain consistent identification of whole cells without training by explicit annotation, even when cells were overlapping. Kashiwabara et al.⁵⁹ used the high energy SR at SPRING 8 (Harima, Japan) and determined by XAFS and μ-XRF spectrometry the host phase of La in deep sea mud, rich in REE and yttrium. The great advantage to excite the K-edge of La lay in the fact that the severe interferences of co-existing elements such as vanadium and titanium in XAFS and XRF were avoided. It was revealed for the first time that La was accumulated in apatite in this type of sediment. Micro-SRXRF beamline of Indus-2 in Indore, India was used by Misra et al.⁶⁰ to assess the distribution of U and Th in O₂ pellets covering the composition of advanced heavy water reactor fuel pellets prepared by powder metallurgical compaction (PMC) and coated agglomerate pelletisation (CAP) routes. The study revealed that the U distribution in pellets prepared by the PMC route was uniform, whereas the pellets prepared by the CAP route had a wide range of compositional variations. Although the CAP route of fuel pellet preparation required less exposure of personnel to high radiation doses, the non-uniformity in the fuel pellet must be considered when using them in reactors. An important study was made by Moini et al.⁶¹ using SR from the SOLEIL synchrotron in France, who investigated the impact of SR to cause damage on proteinaceous specimens at macro and molecular levels. The importance was recognised since SR has become a preferred technique for the analysis of a wide range of archeological samples, artwork and museum specimens. For the investigation on the molecular level four parameters were considered: (1) sample type irradiated e.g. silk, wool, parchment and rabbit skin glue, (2) SR energy, (3) beam intensity and (4) irradiation time (proportional to total dose). At the macroscopic level, colour change, brittleness and solubility enhancement were observed for several samples within only 100 s of irradiation time. At molecular levels the method allowed the authors to quantify significant amino acid modifications, which were ascribed to energy and time of irradiation. With the help of the results and their quantification by SRXRF spectrometry the change induced at the molecular level on proteinaceous specimens, thresholds to minimise the probability of damage occurring to cultural heritage objects were established.

De Jonge et al.⁶² from the Australian Synchrotron in Clayton Victoria reported that X-ray nanoprobes require coherent illumination to achieve optic-limited resolution, and so would benefit directly from diffraction-limited storage rings. The authors explained that XRF-CT is one of the most voracious demanders of coherent photons, since the signal detected is only a small fraction of the incident flux. Considering alternative schemes for beam delivery, sample scanning and detectors, the steps before and after the experiment, typically sample preparation and examination conditions and analysis complexity due to minimum dose requirements and self-absorption must be respected. By understanding the requirements and opportunities for nanoscale XRF-CT, the authors gained insight into the R&D challenges in optics and instrumentation needed to fully exploit the source advances that diffraction-limited storage rings offer.

2.4 TXRF and related techniques

Compared with the large number of contributions, that were reported in the last review period as a consequence of the TXRF Conference 2013, this year saw a smaller number of published papers. However, a significant event did occur, namely the publication of a second edition of the book by Klockenkaemper and von Bohlen Total Reflection X-ray Fluorescence Analysis and Related Methods.⁶³ In 7 chapters, the book provided a detailed and comprehensive description of the phenomenon and its applications. The first chapter included the fundamentals of XRF spectrometry and details of all types of X-ray sources from classical tubes to SR facilities. Attenuation and deflection of X-rays were described to give the reader the theoretical basis of total external reflection, the fundamental basis of TXRF spectrometry. The chapter on the principles of total reflection followed with details on TXRF instrumentation for both TXRF and GI-XRF, with the latter being a variant of TXRF. In GI-XRF the incident angle is changed in small increments from below to above the critical angle of total reflection to provide information from layers on the surface of a substrate or implants in a substrate. The chapter “Instrumentation” covered the large variety of available configurations from air-cooled low power sources to high power rotating anode X-ray tubes and SR facilities. Methods to modify the spectral distribution of the sources were described as well as sample positioning and the energy dispersive detection of the characteristic X-rays from the sample. A detailed description of the mathematical formulae and the physics unveiled the secrets of quantification procedures for the purely chemical analysis by TXRF and included the quantitative surface and thin layer analysis by GI-XRF. A large chapter was devoted to the different applications, ranging from environmental and geological, biological and biochemical, medical, clinical and pharmaceutical, industrial and chemical to art history and forensic sciences. The efficiency of the method including general costs, detection power and economical consideration were compared with other competing analytical techniques. Utility and competitiveness of TXRF and GI-XRF were reviewed as well as perception and propagation of TXRF methods. The authors finally discussed trends and future prospects, where in particular, combinations of preparation methods and time resolved in situ analysis were extremely promising for the understanding of dynamic processes in fundamental research and applications. Instrumental and method developments together with the use of combined techniques such as XRR, XANES or EXAFS showed the prospect for great potential in the future.

A study of a fundamental topic was presented by Fittschen et al.⁶⁴ who addressed the problem of absorption effects and the impact of specimen shape on TXRF analysis. Model calculations indicated that a ring-shaped specimen should give better results in terms of higher counts per mass, compared with a filled rectangle or circle shaped samples. One major reason for the difference in signal was the shading effect, thus absorption of the excitation radiation on the path through the sample from different morphology. It was expected that sample material first encountered by the primary beam shades material behind it. The generation of the tailored specimen was achieved by drop-on-demand technology, allowing the generation of uniform, microscopic deposits of elements in the desired geometry for these studies. An optimisation study was presented by Misra et al.,⁶⁵ who compared the trace element determination of low Z atomic number elements (Al, Ca, K, Mg, Na, P) in air, helium and in vacuo by varying the excitation source with the respective characteristic energies of Cr Kα, Mo Kα, W Lβ. The best results were achieved in the ideal combination Cr Kα together with a vacuum environment and the UTW detector yielding detection limits for of Na 7048 pg and for Ti of 83 pg.

The analysis of nanoparticles (NPs) is of growing interest. A report by Romero et al.⁶⁶ described the trapping of Hg and Se vapours using silver NPs immobilised on quartz reflectors prior to their determination by TXRF spectrometry. The authors described in detail how the quartz reflectors were prepared and how the NPs were immobilised on the surface. The characterisation of the NPs on the surface was carried out by SEM and TEM measurements. The different parameters involved in both the silver NP synthesis and the continuous flow system for vapour generation and preconcentration were optimised. Detection limits were reported as 0.18 μg L⁻¹ for Se and 0.55 μg L⁻¹ for Hg and the obtained enrichment factors were 265 and 175 for Se and Hg, respectively. This new method was successfully validated against a CRM and applied to several seafood samples. The same group⁶⁷ described a novel TXRF method for the analysis of metal hydride vapours immobilised on a surfactant-free catalyst. An ethanol water mixture was used as the reducing agent. Ethanol was added in large excess to reduce the ionic Pd and stabilise the resulting palladium NPs. Freshly prepared palladium NPs were immobilised onto quartz substrates modified with 3-mercaptopropyltri-methoxysilane. Different parameters affecting the synthesis of palladium NPs and their immobilisation on the quartz were evaluated. TXRF spectrometry was applied in order to evaluate their catalytic activity for solid-gas reactions. Quartz substrates coated with nanosized palladium served as a novel sampling platform for TXRF and for example As was detected in situ in natural water with an impressive limit of detection of 80 ng L⁻¹.

The following papers described technical and instrumental developments for modern TXRF spectrometers. An interesting approach to create a laboratory TXRF instrument for the analysis of low concentration samples was suggested by Hampai et al.⁶⁸ using X-ray optics. For samples with concentration levels of ng mm⁻², a very bright SR source in total reflection geometry was required. Combining a conventional source with a polycapillary half-lens provided a quasi-parallel beam which was intense enough for a desktop TXRF spectrometer for low concentration samples containing less than 1 ng mm⁻². From the COPPE nuclear instrumentation laboratory, Rio de Janeiro, Brazil, the group of da Costa et al.⁶⁹ presented the use of two perspex™ parallel plates constituting a waveguide, a Si(PIN) detector and a 15 W X-ray tube with a gold anode and a quartz optical flat to establish a portable TXRF spectrometer. With this low cost compact and easy to handle instrumentation, absolute detection limits of 450 ng for P down to 3.5 ng for Zn were achieved. The X-ray source was rather unconventional comprising a liquid metal jet anode tube where the liquid gallium was pressed through a hole of 200 μm diameter producing a 75 m s⁻¹ high speed stream of gallium representing the anode. With an electron gun directing electrons on to the surface of the liquid gallium stream, the operating conditions were 70 kV with a maximum of 2.8 mA limited by the small size of the electron spot. The brilliance could be changed by adjusting the focal size of the electrons ranging from a few μm to 100 μm diameter. Maderitsch et al.⁷⁰ from the Atominstitut Vienna group described their application of a liquid jet anode source to a versatile TXRF spectrometer with vacuum chamber. The beam coming from the source was collimated by a 50 μm slit and directed to the multilayer monochromator selecting the intensive Ga Kα radiation. From the monochromator there was only a short distance in the chamber to the samples on quartz or silicon reflectors where the excitation was performed. Excellent results regarding geometric beam stability, high fluorescence intensities and low background were achieved leading to detection limits in the high fg range for Ni. A SDD with 80 mm² active area was used to collect the XRF radiation contributing to the high sensitivity and excellent detection limits for a laboratory scale instrument. The last contribution concerning instrumentation dealt with another variant of TXRF based on the inversion of the physical principle, leading to grazing exit XRF (GE-XRF). Ashida and Tsuji⁷¹ from the Osaka City University Japan, presented a compact GE-XRF spectrometer developed in their laboratory. An aluminium cylindrical collimator for the primary X-rays was placed just above the sample stage serving a dual role, that of the collimator but also the exit slit to detect the fluorescent X-rays at small grazing exit angles. The compactness of this instrument was impressive, being only 80 × 200 × 170 mm³. The background was drastically reduced at grazing exit angles, enabling trace element analysis. A detection limit for Ga was extrapolated to the impressive value of 20 ng g⁻¹.

Turning now to results from applications, Dhara et al.⁷² determined low Z elements, such as Al and Mg in uranium oxide samples in a TXRF spectrometer equipped with a vacuum chamber. The challenging task was to find measuring conditions for light elements in a heavy element matrix. Not surprisingly, it was found that the analysis of elements like Al or Mg required an almost complete removal of the uranium matrix. Two CRMs containing uranium and trace elements in different concentrations were dissolved in a minimum amount of high purity nitric acid. The uranium matrix was separated by solvent extraction using tri-n-butyl phosphate as reagent. The determination of Al and Mg used Sc as an internal standard and values for Mg were found to be in good agreement with the certified values. However, the Al values differed from the reference values, most probably due to interference from the U M escape peak with the neighbouring high intensity Si K line from the substrate. Using an alternative reflector material and a higher resolution detector this problem could be much reduced. The absolute determination of P and S in organic samples was of growing interest during the review period, as the ratio of P and S in proteins allowed the determination of the degree of phosphorylation of proteins. Rauwolf et al.⁷³ used the low Z TXRF spectrometer at the Atominstitut Vienna to analyse the P and S content of proteins. In comparison to ICP-QMS this TXRF method took less effort and allowed easy quantification. Although the sample preparation proved to be more difficult than expected, the lower limits of detection (1000 s) for P and S in proteins, were 34 pg and 19 pg respectively. One technical note and one article dealing with an interesting problem in TXRF spectrometry were presented by Vander Hoogerstraete et al.,^74,75 for the determination of halide impurities in ionic liquids. The detection of halide ions in solution by TXRF has been problematic because volatile hydrogen halide (HX) compounds are formed and lost when the sample was mixed with the acidic metal calibration solution. The authors developed a new method based on the preparation of a stable copper halide, Cu(NH₃)₄(NO₃)₂ for the determination of Br, Cl, and I in ionic liquids. Results validated this method successfully. An extremely informative article was presented by Klockenkamper and von Bohlen,⁷⁶ who carried out a survey of users and manufacturers of TXRF instrumentation in order to demonstrate the worldwide distribution of TXRF equipment and the different fields of application. The TXRF instrumentation was identified in more than 50 countries on six continents and at about 200 institutes and laboratories. The number of desktop operating instruments comprised nearly 300 units, about 60 beamlines use TXRF and about 300 floor mounted instruments were in use in the semiconductor industry. A total of 13 different fields of applications could be identified.

During the review period papers on related techniques exclusively dealt with GI-XRF and GE-XRF spectrometries. Nowak et al.⁷⁷ performed GI-XRF and GE-XRF experiments to characterise periodic structures on silicon and silica surfaces. Apart from the characteristics, which were typical for particle- and layer-like samples, the measured angular intensity profiles showed additional periodicity-related features. A novel theoretical approach based on simple geometrical optics considerations was used to explain the latter. The authors claimed their new calculations yielded results in good agreement with experiment and also in cases where other theoretical approaches were not successful, e.g., periodic particle distributions with an increased surface coverage. Brucher et al.⁷⁸ investigated nanoscale elemental distributions at surfaces with GI-XRF. In this paper a comparative characterisation of different instrumental concepts was investigated. Angle scans recording the fluorescent signal with a TXRF spectrometer were compared with data obtained using SR. The structures investigated included gold on silicon surfaces in the form of layers and particles, sub-micrometre-sized droplets, a liquid film, and ions implanted into a Si wafer. Ingerle et al.⁷⁹ combined the evaluation of GI-XRF and XRR data for improved profiling of ultra-shallow depth distributions. The authors emphasised that both techniques used similar measurement procedures and data evaluation strategies, i.e., optimisation of a sample model by fitting measured and calculated angle curves. Moreover, the applied sample models could be derived from the same physical properties, such as atomic scattering/form factors and elemental concentrations; a simultaneous analysis was therefore a straightforward approach. This combined analysis in turn reduced the uncertainties of the individual techniques, allowing a determination of the dose and depth profile of the implanted elements with significantly increased confidence levels. Silicon wafers implanted with arsenic at different implantation energies were measured by XRR and GI-XRF spectrometry. The data were processed using a self-developed software package “JGIXA”, designed for simultaneous fitting of GI-XRF and XRR data. The results were compared with depth profiles obtained by SIMS and agreement was found to be good. The same group published their combined measurements⁸⁰ from one table-top spectrometer for the improved characterisation of thin layers and implants on/in silicon wafers. As the GI-XRF spectrometer already used an X-ray beam impinging under grazing incidence with the ability to vary the angle of incidence, it was adapted with an XRR unit to obtain data from the angle-dependent XRF radiation as well as data from the reflected beam. A theta-2 theta goniometer was simulated by combining the translation and tilt movement of a SDD, which enabled detection of the reflected beam intensity over 5 orders of magnitude. Hafnium oxide layers, as well as As implants in silicon wafers in the nanometre range were characterised using this new setup. The combined approach of JGIXA was used when fitting the experimental data from both GI-XRF and XRR, with the simulated data. A significant discrepancy was observed for values within the range of 0.5 nm, highlighting the high sensitivity of the presented method.

2.5 X-ray detectors

The most active topic in detector publications this year was again that of pixellated X-ray detectors used in a variety of ED X-ray imaging applications. The characteristics of a new detector called DOSEPIX, based on CERN's existing MEDIPIX and TIMEPIX technologies, was described by Ritter and co-workers.⁸¹ The sensor itself comprised a 16 × 16 array with segments of both 55 μm and 220 μm pixels. The authors reported early results from their characterisation of the detector for which they used both XRF measurements and analog test pulses. A team at Paul Scherrer Institute (Villigen, Switzerland) reported⁸² the development of the EIGER single photon counting X-ray detector, based on the successful PILATUS detector already used in a number of SR and SRXRF spectroscopy instruments. The basic technology comprised a pixel size of 75 × 75 μm² coupled to electronics with an impressive frame rate of 22 kHz with an equally impressive claimed inter-frame dead time of only 4 μs. The initial 500 000 pixel module sensor was already deployed and much larger multimodule arrays were planned. The two impressive DEPFET active pixel detector modules for the Mercury Imaging and X-ray Spectrometer (MIXS) instrument destined for the BepiColombo mission to Mercury and the flight spare were fully characterised and calibrated using the PTB SR beamlines at BESSY II in Berlin.⁸³ Each detector module comprised an array of 64 × 64 active pixels, each of 300 × 300 μm² providing an active area of 1.92 × 1.92 cm² and the custom readout electronics enabled the full detector matrix to be read out within 165.6 μs per frame (which was a frame rate of about 6 kHz). Interestingly, the DEPFET design was claimed to eliminate the effects of shared charge across more than one pixel, which is a major advantage for spectroscopic performance due to the elimination of pixel cross-talk artefacts. The beamline monochromators provided 10 calibration energies in the range 500 eV to 10 keV, most of which were selected to be at the photon energy of XRF emission lines of elements of particular interest to the study of Mercurian surface geology. The measured energy resolution of all three modules with the sensor operated at −40 °C was in the range 133 eV to 134 eV at 6.404 keV (Fe Kα) and 75 to 77 eV at 1.041 keV (Na Kα). The very careful measurement and data analysis were described in detail and the three modules showed excellent consistency of both gain and quantum efficiency. These impressive MIXS detector modules for the BepiColombo mission are set to launch in 2015 and to arrive in the Mercury orbit in 2022. A new spectrum processing software package was developed by Lauf and Andritschke⁸⁴ and its utility was tested specifically on DEPFET detector data. The embodied algorithms allowed detector parameters to be changed by means of the graphical user interface and supported off-line recalculations. The authors claimed that the modular design of the package made it straightforward to apply to other types of ED detector. In a study covering the energy range 1.75 to 10 keV and also using the PTB beamlines at BESSY II, the X-ray and geometric properties were reported⁸⁵ of a hybrid pixel PILATUS detector destined for use in small angle scattering X-ray spectroscopy. The device had a very large active area of 179 × 169 mm² and a sensor thickness of 320 μm, yielding a quantum efficiency of 80% over the range from 3.4 keV to 10 keV, with a maximum of 96% at 8 keV. The device could be operated down to 1.75 keV although at a quantum efficiency as low as 5%, which was limited by the sensor structure and the electronic threshold level applied. Full details of the vacuum detector module and the authors' careful measurements were provided although the detector was not characterised for spectroscopy but was shown to be highly suitable for imaging signals from GI-SAXS experiments. The greater X-ray stopping power of germanium makes it an attractive detector material for higher energy photons and a linear segmented germanium detector, coupled with a custom ASIC was characterised by Rumaiz et al.⁸⁶ The sensor was fabricated on a 3 mm thick germanium crystal with the 64 sensor pixels, each of 0.5 mm by 5 mm, arranged as a strip separated by a trench 50 μm wide and 50 μm deep. Each sensor pixel was wire bonded through to the custom ASIC, albeit with rather long wires and acute angles due to the geometric mismatch to the ASIC that was usually used for silicon strip detectors with a different pitch. A useful description of the detector module and its experimental setup was provided and the energy dispersing and geometric characteristics of the sensor pixels were characterised using a 20 μm wide beam and a knife edge and were found suitable for EDXRD studies in the energy range 20–100 keV at a new SR beamline at the Brookhaven National Laboratory (USA). The geometry and limited energy resolution of this device makes it less useful for XRF imaging although its stopping power could be attractive for higher energy XRF spectrometry and the authors expected it to be much more attractive than the existing systems comprising many individual cryo-cooled HPGe sensors with discrete conventional nucleonics. Cadmium telluride is also a material well-suited to higher energy X-ray detection, prompting Veale et al.⁸⁷ to study the performance of an 80 × 80 array of CdTe pixels each 1 mm thick and arranged on a pitch of 250 μm. The sensor was flip-chip bonded to a HEXITEC custom ASIC and was shown to produce an energy resolution of <1.8 keV at 59.5 keV. The main problem with the penetrating radiation and the small pixel size was that of charge sharing across multiple pixels. In fact a total of 36.4% charge sharing was measured, which, without corrections, resulted in substantial degradation in spectroscopic performance. The authors investigated two correction algorithms, one using charge sharing discrimination (CSD) and the other, charge sharing addition (CSA). In CSD, events detected by multiple pixels were rejected from the data set, resulting in a good retention of energy resolution but a significant reduction (36.4%) in total counts and loss of useful spectrum information. The CSA method involved summing of data from sharing pixels which retained the highest counts (93% retention vs. 63% for CSD) but degraded the energy resolution but only from 1.8 to 1.9 keV. The authors provided an excellent description of charge sharing and escape events, the latter being a particular problem with such small pixels and higher photon energies. Finally, staying with materials well-suited to higher energy photon detection, Veale and co-workers⁸⁸ used a 0.5 mm thickness gallium arsenide (GaAs) crystal that had been compensated with chromium to yield detector-grade material. As for the CdTe sensor referred to above, an 80 × 80 array of pixels arranged on a pitch of 250 μm was bonded to the HEXITEC readout ASIC yielding a detector whose energy resolution was measured at 280 K to be in the range 2.8 to 3.3 keV at 60 keV. The detector's electrical characteristics were also carefully measured as was the charge sharing, which was significant, as for the CdTe detector. When shared events were eliminated from the data set, 20% of the counts were eliminated but the low energy background was substantially reduced, as was the tailing from the 60 keV photopeak. The authors also concluded that the high contribution of holes to the detected charge for GaAs meant that the 0.5 mm thick sensors were suitable only up to 40 keV, above which thicker sensors would be needed to retain attractive spectroscopic properties.

The low number of publications available for review this year on silicon-based semiconductor X-ray detectors such as SDD and Si(PIN) continues to reflect the maturity and widespread routine use of such devices. Improvements to their so-called Swept Charge Device (SCD) were reported by Smith and co-workers⁸⁹ in which these soft X-ray detectors, suitable for the energy range 0.5 to 10 keV, had been increased in active area from a few tens of mm² to 20 × 20 mm². These SCDs were based on CCD technology but have no real pixels and are not suited to X-ray imaging applications. The authors fully characterised the quantum efficiency of these devices, which had a depletion depth of only around 50 μm, using the facilities at the PTB beamlines at BESSY II in Berlin. The data were expected to be invaluable for spectrum data integrity when used on the proposed Indian Chandrayaan-2 and Chinese HXMT spacecraft. Also designed for use in X-ray spectrometers for space exploration, a very large active area (18 × 70 mm²) linear SDD, was described by Rachevski et al.⁹⁰ This device had many interesting design features, including 256 readout anodes but was not specifically used for X-ray imaging. Full design details of the module were provided along with details of how the anode pitch and operating parameters affected spectroscopic and physical performance. Great care was taken to optimise the sensor and readout systems to minimise the power requirements to make them suitable for use in spacecraft X-ray instruments. The reported energy resolution was impressively <200 eV at 6 keV and the design was specially suitable for tiling many such modules together to create very large X-ray detector arrays that the authors indicated would provide the highest sensitivity and widest view X-ray detectors available for future space missions. Finally, returning to routine of laboratory XRF spectrometry, Cornaby⁹¹ briefly described the performance of a typical, run-of-the-mill Si(PIN) ED detector connected to a standard digital pulse processor. Such systems have been largely supplanted by larger, faster and higher resolution X-ray detectors although the Si(PIN) still has a place at the least expensive end of the XRF semiconductor detector market.

The experimental determination and theoretical modelling of energy dispersive X-ray detector response characteristics remain important for new types of detectors and applications. A mathematical model for the response function of a Si(PIN) detector was proposed⁹² in which there were four components: a Gaussian peak; a truncated shelf; an exponential tail and a Gaussian silicon escape peak. The proposed model was used in a MC simulation of spectra from a number of pure elements and one of those was compared to experimental measurements on a copper sample. Such a model and parameters are already well-known and are widely used and this one was only validated using a single measured Cu K spectrum. A study by Mulware and colleagues⁹³ used PIXE and a number of pure element RMs to establish the detector efficiency of a commercially available 100 mm² HPGe detector over the energy range 1.25 to 20 keV. Interestingly, the authors used a calibrated particle detector to measure the proton backscatter and establish the effective charge deposited. A method for establishing the energy response function of Si-strip photon counting detectors used for XRF-CT was reported by Ding et al.⁹⁴ who used a tungsten target X-ray tube operated at 80 kV, various beam filters and samples containing Ag, Ba, Gd or I. The measured XRF data were used to establish a detector response model, which used four parameters, and also the gain factor for each pixel in the strip. The authors reported an average charge sharing of 36% although good agreement was obtained between simulated spectra using the new model and the experimentally measured XRF spectra. A similar method was used by Cho et al.⁹⁵ for the energy response calibration of a CZT ED photon counting X-ray detector. In this study, the XRF signals from 11 samples containing elements with XRF lines in the 20–45 keV energy range were detected by a small (3 × 3 mm²) single pixel CdTe detector in order to verify the methodology and MC simulations performed. An investigation of geometric positioning was performed using the CZT detector and the optimum XRF peak to background ratio was established to be when the detector was placed at an angle of 120° to the incident beam. The authors claimed that this method provided a simple and reliable method for the calibration of this type of detector for use in XRF-CT instruments. Taking spectrum simulation one step back in the detection chain, Zeng and colleagues⁹⁶ described an electronic spectrum generator for use in optimising signal processing electronics. The system used a reference spectrum then provided an analog signal stream on a ramp that simulated the output of a Si(PIN) or SDD preamplifier. Given the wide commercial availability of reliable plug and play Si(PIN) and SDD detectors with compatible digital pulse processes it seems unlikely that such a simulator will be widely needed.

Ultra high resolution cryogenic X-ray detectors remained at the margins of X-ray spectrometry during this review period although Carpenter and co-workers⁹⁷ described their recently developed pixellated tantalum STJ-based detector designed for experiments at SR facilities. This impressive cryogen-free system used an ADR coupled to a two-stage pulse tube cooler capable of maintaining sensor temperatures below 100 mK. The detector array comprised 36 tantalum-based STJ sensors, each 208 × 208 μm², chosen for their greater stopping power and better energy resolution than niobium-based STJ sensors. The energy resolution achieved was 9 eV at the 525 eV oxygen K line. The sensor array in this modern user-friendly set-up was in a 42 cm long shielded snout, making it only suitable for experiments where very high intensity X-ray excitation was available, such as at the newer SR beamlines. A 100-pixel STJ-based detector was described⁹⁸ in which the claimed sensor active area was an impressive 1 mm² and the measured energy resolution was 12 eV at 400 eV. The detector was used for XAFS studies and was able to clearly separate the weak N K line from the dominant C K line in the spectrum of a compound semiconductor (4H–SiC).

An interesting paper⁹⁹ on the passivation of CdZnTe detector chips using two popular chemical treatments, hydrogen peroxide and a mixture of ammonium fluoride and hydrogen peroxide, studied in detail the surface leakage current and the long-term stability of the resulting surface. The chemical species formed were investigated in detail using XPS and surface homogeneity of the material was measured using μ-XRF spectrometry. The sensors were characterised for their electronic response and then as X-ray detectors and both treatments were found to give lower leakage current but the effectiveness of both deteriorated over time.

2.6 Quantification and data processing

There are many complex excitation interactions to be taken into account in Fundamental parameter corrections, which prompted Pavlinsky & Portnoy¹⁰⁰ to study in detail the role of Compton scattering in the excitation of XRF from low-Z elements. The authors calculated the various contributions from: primary photons; Compton direct, Compton electrons; K photoelectrons and L photoelectrons to excitation of XRF radiation from elements in the Z range 5 to 10. The authors described the equations used and tabulated the results, which showed how dominant the electron excitation became as the energy of the primary photon increased from 20 to 80 keV. In fact, the contribution to XRF excitation from Compton electrons was by far the most dominant at an incident photon energy of 80 keV although such high energy excitation would only typically be used in wavelength dispersive (WD)XRF systems. In the case of sources typically used for benchtop EDXRF configurations, excitation was principally due to primary photons and K photoelectrons. The valuable empirical relationship between Compton–Rayleigh peak intensity ratio and the mean Z of the excited sample is well known leading Hodoroaba and Rackwitz¹⁰¹ to confirm that it was possible to discriminate samples differing in mean Z by as little as 0.1. The authors introduced their system for careful measurement of Compton–Raleigh ratios and also the theory of this important relationship. The mathematical problems associated with the inclusion of an effective and efficient model and algorithm for the incorporation of the incident spectrum into FP calculations was tackled by Delgado,¹⁰² who used a Volterra integral equation of the first kind in place of a full integral of the exciting spectrum. The model required measurement from a few pure element samples and the approach was claimed by the author to offer a stable and robust calculation process for the inclusion of the primary excitation spectrum in matrix effect calculations. The impressive X-ray instrumentation and facilities at BAM (Berlin, Germany) enabled Rackwitz and co-workers¹⁰³ to measure directly the spectrum emitted from a low power microfocus X-ray tube of the type typically used in benchtop μ-EDXRF systems. The fully calibrated instrumentation enabled the measured output to be recalculated to yield data in units of photons eV⁻¹ msr⁻¹ nA⁻¹ s⁻¹.

Although FP methods are well advanced and have been in use for many years, there is still space for semi-empirical correction models as concluded by Esbelin¹⁰⁴ in the case of the challenging but well-defined determination of transuranic elements in nuclear processing liquids. The author described their custom-built EDXRF instrument that used a collimated 3 kW rhodium target X-ray tube operated at 50 kV whose output was made quasi-monochromatic by means of a rhodium foil primary beam filter of 200 μm thickness. The XRF beam from the sample passed through a custom annular graphite monochromator tuned to pass energies in the range 13 to 15 keV through to the HPGe detector. This system was optimised for the measurement of the transuranic element L series lines but had the disadvantage that the useful Compton and Rayleigh scatter peaks were no longer available to assist in matrix corrections. That being the case, the author developed a useful semi-empirical correction algorithm, which required straightforward calibration and data fitting and provided accurate and reliable determinations of U and Pu in aqueous and organic process liquids.

In the area of chemometrics, Scapin et al.¹⁰⁵ used WDXRF spectrometry to measure major and minor analytes in U₃SiO₂ samples and compared results obtained by a conventional FP method to those from a multivariate calibration method. The latter method used spectrum scans and concentration data from seven RM samples and applied PCA to establish that a single principal component (PC) was sufficient to describe the system and used that component in a Principal Components Regression (PCR) model. The authors reported that the PCR approach yielded better accuracy than the FP method although both provided adequate LOQ values for the highly regulated nuclear application. The results from the chosen PCR method were found to be comparable with those from conventional volumetric analysis or ICP-OES spectrometry, leading the authors to conclude that the proposed PCR-WDXRF method would enable laboratories to be in compliance with ISO/IEC 17025 for the determination of U and impurities in nuclear fuel samples and with the significant advantages of being non-destructive and of producing minimal hazardous waste. In a study of the heavily overlapping Na K and Zn L lines in the WDXRF spectrometry of mineral samples, Ghasemi et al.¹⁰⁶ used the chemometric approaches of PLS, PCR and support vector machine (SVM). Results from these multivariate calibration methods were compared to the on-board FP standardless software on their commercially available WDXRF spectrometer. Spectra from WDXRF scans of 80 s each over the 2θ range 18 to 60° were collected from pressed pellet samples under typical WDXRF spectrometer settings. The authors used mixtures of pure compounds to produce 35 calibration samples with concentrations of Na₂O and ZnO in the range 1–5% m m⁻¹ and used PCA to establish that only 2 PCs were needed to give reliable calibrations for Na with the heavily overlapping Zn L lines present. The results showed that all three chemometric methods gave more accurate results than the instrument's FP correction algorithm using overlap factors and that the PCR and SVM methodologies gave the best performance. The best accuracy was observed when using the SVM algorithm and that was attributed to the fact that it incorporated a non-linear regression model, unlike the linear one of their PCR and PLS methods. Working with EDXRF spectrometry, Akbulut,¹⁰⁷ used the FP software on a commercially available EDXRF spectrometer and compared results to those using PLS and PCR methods for the determination of 10 trace elements in 76 soil RMs. The author reported that only 1 PC was needed for the elements Co, Cu, V and Zn but that 7 PLS and 10 PCR components were needed for Rb, an element that often needs only a simple linear regression empirical model when a rudimentary Compton corrected intensity is used. It was concluded that the PLS model provided better accuracy than the PCR and FP methods and required the fewest PCs although the use of 76 calibration samples is a daunting prospect for those wishing to follow this approach.

A novel application of MC simulation was described by Brunetti and Golosio,¹⁰⁸ who developed a new MC code for the simulation of spectra from irregular-shaped sample surfaces such as those typically encountered in the analysis of precious artworks or cultural heritage artefacts. The authors described fully the theoretical basis of their model and showed the equations used to describe the effects of surface roughness on the detected spectrum in a typical EDXRF setup. The effectiveness of their model was shown using two examples: a lead single layer and a copper–gold alloy two-layer system. Measurements were made on an EDXRF system whose X-ray tube was operated at 50 kV, 5 mA and provided a 20 mm diameter spot on the sample, in conjunction with an SDD coupled to a polycapillary optic at a takeoff angle of 90°. Each simulation took about 10 minutes and the early results indicated promise although the interfacial roughness in the copper–gold multilayer sample was particularly challenging. The authors intended to apply their new approach to other sample types where surface roughness can be a significant contribution to inaccuracy, such as when lower energy lines are measured. The problem of surface topography affecting XRF measured intensities was also the subject of a paper by Geil and Thorne¹⁰⁹ in the field of 3D XRF imaging and CT. The authors proposed a new correction procedure that was suitable for the determination of trace elements in a uniform matrix. With the increasing importance of confocal XRF measurements and the number of instruments being made available, it is pleasing to see the report¹¹⁰ by researchers at Atominstitut (TU Vienna) on the modelling of the measurement of layered structures in a benchtop confocal μ-XRF system. A particular benefit of this simulation tool, which used analytical functions rather than MC simulation, was that it allowed the authors to optimise the instrument setup prior to making measurements, thereby improving data quality and greatly reducing instrument time. The proposed calculation method was very fast and the experimental and simulation data agreed well, enabling the authors to conclude that the confocal volume diameter must be smaller than the investigated layer (high spatial resolution) and that the layer must be smaller than the information depth of the investigated element in the sample. The challenge of long computation times for MC simulations of XRF measurements, particularly when XRF imaging is concerned was confronted by Golosio and co-workers¹¹¹ who used variance reduction techniques to reduce computation times by a claimed several orders of magnitude compared with general purpose MC codes. The software and associated data were made available to potential users and the simulation was suitable for use with monochromatic and polychromatic excitation sources. Execution times were in the impressive range of <1 s to a few minutes. Similarly concerned by lengthy computation times, Tickner¹¹² sought to reduce the times, particularly when arbitrary and small changes of operating parameters or geometry were needed for such tasks as instrument optimisation. The author provided the mathematics behind both his perturbation methods and the modified Woodcock neutral-particle tracking algorithm that was used. Notwithstanding this new approach, a single computation still took 500 minutes, although that was much better than the several computations, each of 250 minutes that would otherwise have been needed without the proposed modifications. A useful online software routine was provided¹¹³ for the construction of a Compton–Rayleigh ratio vs. sample mass attenuation coefficient, which should prove useful to those not wishing to establish their own curve empirically. The authors claimed the new simulation tool provided coefficients that agreed to within a worthy 15% of literature values. An extension to the EGS4 MC code used for simulations in XRF-CT was described¹¹⁴ that incorporated corrections for photon transport through the sample and the instrument components. Simulations of Pt in head and neck tumours gave Pt Kα intensities within 2.3% of measured peak intensity and for the Compton scatter peak a difference of 2.5% was observed.

Many of the foregoing chemometric and MC methods embody within them either explicit or non-explicit algorithms for spectrum processing to yield net peak intensities. While many algorithms take into account spectrum background in a non-specific way, Sanchez and Leani¹¹⁵ studied in detail the contribution from resonant Raman scattering to the background immediately under the photopeak of interest. In the cases of trace Mn in iron and 1% m m⁻¹ V⁻¹ in 99% m m⁻¹ titanium, the authors established that the resonant Raman contribution could cause a distortion in the wings of an XRF peak and could in specific cases be of greater magnitude than secondary fluorescence and hence should not be ignored. A rediscovery was made¹¹⁶ of the use of cubic splines fitting to model and remove the background in EDXRF spectra while Zeng et al.¹¹⁷ applied a chemometric genetic algorithm approach to spectrum processing in EDXRF spectrometry. The authors claimed the GA approach to be suitable for EDXRF spectra although they only applied it to a very low resolution (8%) NaI (Tl) scintillation counter, so its wider adoption in EDXRF spectrometry is particularly unlikely.

Finally, the important matter of estimating limit of detection was considered by Borkhodoev¹¹⁸ who recommended the repeat measurement of samples in which all elements but the analyte were present or at least only a very small amount of the analyte was present. The author provided the mathematical background to this proposed approach, which appears to be much the same as that proposed by Currie as long ago as the mid-1970's. Of course, both approaches have the same interesting dilemma of finding suitable materials that contain everything but the analyte of interest!

3 Applications

3.1 Geological and climate change

This review period is notable for the significant number of publications describing exploration and geochemical applications of field-portable XRF instrumentation, indicating that the technique may have come of age in practical field investigations. In one of a series of publications, Simandl et al.¹¹⁹ reported the use of hand-held XRF instrumentation for the exploration of primary carbonatite-hosted niobium deposits. Correlation in the analysis of pulps with laboratory data was good for Ce, La, Nb, P and Y, but not surprisingly, more disappointing for elements that approached the XRF instrument's detection limit. Nevertheless, the authors considered that portable XRF spectrometry was a robust tool for facilitating exploration-related decision making in the field, based on elements that could be determined satisfactorily. Simandl et al.¹²⁰ also undertook the characterisation by portable XRF instrumentation of REE enriched sedimentary phosphate deposits from the Fernie deposit, British Columbia, Canada, demonstrating that if sample preparation (to pulps) and site-specific recalibration were carried out, portable XRF could identify zones of phosphate rocks that were enriched in REE and delimit zones with unacceptable levels of deleterious elements such as uranium. Gazley et al.¹²¹ described the use of portable XRF spectrometry for the characterisation of dolerite dykes at the Plutonic Gold Mine, Western Australia. These authors showed that plots of Ti versus Zr combined with PCA allowed a new interpretation of the area to be devised, based on four geochemically distinct suites of dolerite with a potential increase in the identification of gold host rock. Quiniou and Laperche¹²² showed that a very good correlation could be achieved between portable XRF results for the determination of Fe and Ni in saprolite and laterite samples and conventional laboratory determinations, when samples were dried, pulverised and pelleted. This good performance probably resulted from the calibration of the portable XRF instrumentation using representative field samples with field measurements being more reliably made on the laterite layer to reduce sample heterogeneity problems. Le Vaillant and co-workers¹²³ reported that the on-site geochemical analysis of rock powders and drill core in the lithogeochemical exploration for komatiite-hosted nickel sulfide deposits could be applied successfully, provided that strict quality control protocols were followed, especially concerning instrument calibration.

A number of other geological applications of portable XRF instrumentation were reported during the current review period. Weindorf et al.¹²⁴ used field-portable XRF spectrometry to investigate the influence of ice on the elemental characterisation of soil affected by permafrost in central and northern Alaska, comparing in situ measurements with measurements on soil refrozen in the laboratory, melted water/soil in the laboratory and moisture-corrected soil. The elements examined were Ba, Ca, Cr, Fe, K, Mn, Pb, Rb, Sr, Ti, Zn and Zr and results showed that the majority of readings from in situ, refrozen and melted samples significantly underestimated compositions compared with oven dried samples, owing to the expected influence of the moisture content. The authors found that measurements on the group of samples with <40% moisture performed better than those with >40% moisture. Shuttleworth and colleagues¹²⁵ assessed the performance of in situ measurements for Pb in surface peat samples from the South Pennines (UK) by field-portable XRF spectrometry, and undertook a comparison with measurements by the same technique on dried samples in the laboratory and with ICP-OES determinations. Field and laboratory measurements were reported to be directly comparable when corrected for moisture, making the in situ technique cost effective for the determination of Pb in contaminated peat land. Moisture content also influenced the in situ XRF measurements of Lemiere et al.,¹²⁶ whose interest lay in the determination of As, Cu, Pb and Zn in dredged waterway sediments to facilitate their safe reuse or treatment. Measurements were made on site on raw wet sediments containing 50–70% water and although the results were 2 to 3 times lower than laboratory analyses, data was found to relate sufficiently closely to absolute concentrations to allow samples to be ranked. The authors developed a technique for using a hand press to reduce the water content and produce a pellet containing 30–50% water. Results from these materials were found to be less than 20% lower than laboratory measurements for As and Pb, although higher differences were observed for Cu at concentrations approaching the detection limit. This simple partial dehydration procedure clearly has potential application for field XRF measurements of other types of ‘wet’ sample. A less sanguine report of the performance of portable XRF instruments was made in a paper by Brand and Brand¹²⁷ who undertook a comparison of commercial instrumentation, reporting that although repeatability was typically very good, accuracy was generally poor to very poor. To demonstrate this point, the authors provided examples of measurements on NIST SRM 2709a (agricultural soil) for Al and Si elements that in this reviewer's experience suffer serious air and window attenuation of their low energy fluorescence X-rays when measurements are made in situ and require careful sample preparation and reliable matrix correction if determinations are to approach the reliability of conventional WDXRF measurements, for example. The authors also reported that a deterioration in the performance of the portable XRF instruments evaluated was observed during a routine battery pack change (a problem that may be instrument specific). Following an evaluation of field-portable XRF for exploration lithogeochemistry, Piercey and Devine¹²⁸ considered that the technique should be considered as a preliminary screening tool for sample selection and not a substitute for laboratory-based XRF methods, as well as fusion ICP-OES and ICP-MS, particularly when important economic decisions were to be made using such data. Analytes reported to have very good to excellent correlation between portable XRF and conventional laboratory data were As, Ba, CaO, Co, Cu, Fe₂O₃, K₂O, MnO, Mo, Nb, Pb, Rb, S, Sr, TiO₂, U and Zr.

Turning now to developments in the XRF technique, Ling et al.¹²⁹ described a lithium borate fusion technique for the preparation of sulfide ores for XRF analysis, which involved the addition of silica to ensure the glass disk was stable and homogeneous and lithium nitrate as a pre-oxidising agent to prevent the degradation of platinum–gold crucibles. The proportion of added silica varied from 1 : 5 to 1 : 20, depending on the sulfur content and the combined sample-to-lithium tetraborate ratio was 1 : 14. It is useful to revisit a scheme for the fusion of sulfide samples, versions of which have been described several times previously in the literature. Cooper et al.¹³⁰ were interested in the time resolved monitoring of fluvial suspended particulate matter and described a combined XRF and diffuse reflectance FTIR method to estimate various elemental and organic concentrations trapped on quartz fibre filters at masses as low as 3 mg. Pereira et al.¹³¹ determined Au in ore samples by EDXRF spectrometry after separation and preconcentration on polyurethane foam. Technique evaluation was undertaken using appropriate ore RMs and the detection limit for the direct EDXRF measurement on the solid polyurethane sorbent was reported to be 1.36 mg kg⁻¹. An instrumental investigation of the WDXRF performance of pentaerythritol (PET) crystals was undertaken by Kato and Suzuki.¹³² Their interest was to study the drop in intensity (‘background holes’) that occurred at certain specific 2 theta diffraction angles. Reference materials have a key role to play in demonstrating the quality of analytical data and Bedard and Neron¹³³ used a μ-XRF instrument to measure the homogeneity of selected RMs to contribute to a model that better characterised the grouping and magnitude of heterogeneities and provide information necessary to define the minimum mass of a test portion.

In addition to the above developments, core scanning by XRF spectrometry continued to make significant contributions to this field of study, especially in the analysis of sediment cores. Thus, Poto et al.¹³⁴ undertook a cross calibration of XRF and ICP-MS spectrometry for the high resolution analysis of ombrotropic peat cores for palaeo-climatic studies from the Danta di Cadore, northeastern Italy. Perfect positive correlations were reported for Cd, Cr, Pb, Sr, Ti and Zn, high positive correlations for Ba, Cu and Fe and moderate correlations for Ca and Ga. Finne et al.¹³⁵ asked (and then answered) the question: can XRF scanning of speleotherms be used as a non-destructive method to identify palaeo-flood events in caves? They concluded that in certain cases, individual flood layers (identified from petrographic thin sections) had a distinctive geochemical signal suggesting the possibility of distinguishing individual flood layers based on their geochemistry. Rodriguez-Germade et al.¹³⁶ investigated the capabilities of an XRF core scanner to detect trace Cd and Hg in the Ala de Pontevedra harbour, Northwest Spain, making comparisons with ICP-OES and CV-AAS to evaluate performance. The presence of organic material was also evaluated from the incoherent/coherent X-ray scatter ratio and from the Br : Cl ratio and the instrument was considered to be an efficient and fast option for monitoring trace Cd and Hg pollution.

One of the most popular applications of XRF spectrometry continues to be contributing to studies of environmental and climate change over geological time scales. These studies often involve XRF scanning of lake or ocean sediment cores in combination with other techniques usually related to the dating of sediment layers by fossil characterisation or isotopic techniques. Rather than review the large number of contributions available during the current review period, a few examples are featured here to demonstrate the contribution of XRF data. Stuut et al.¹³⁷ were interested in the record of aeolian activity near the North West Cape, Australia and used the XRF technique for scanning deep-sea sediment cores recovered from the off-shore continental slope to reveal the history of climate change over the last 550 ka. Two Aeolian and one fluvial input to sediments were identified and XRF log concentration ratios of Zr : Fe and Ti : Ca were used to indicate those sediments of terrigenous origin. The palaeo-environmental record of the evolution of the Sokli Ice Lake, Finland, was reconstructed by Shala et al.¹³⁸ by examining the sediment succession in Lake Loitsana, a remnant of this lake. The XRF-core scanning data was used to show changes in the influence of regional catchment geochemistry (involving Precambrian crystalline rocks) versus local catchment geochemistry (involving Sokli Carbonatite Massif rocks). These results, together with other data, demonstrated the value of this approach in reconstructing environmental conditions in North East Finland directly following deglaciation in the early Holocene period. In a publication by Kumpan et al.,¹³⁹ XRF spectrometry contributed Zr : Al, K : Al, Sr : Al and Mn : Al concentration ratios in the characterisation of sedimentary sequences associated with three Devonian-Carboniferous boundary sections in shallow-water carbonate rocks in the Namur-Dinant Basin (Belgium-France). The aim of this work was to describe the sea level and environmental changes associated with this section and identify regional and inter-regional influences. Li and Piper¹⁴⁰ contributed to an understanding of the influence of melt water on the Labrador Current during Heinrich event 1 (about 16 ka BP) and the Younger Dryas during the last deglaciation period. The XRF contribution was in the analysis of relevant sediments from Flemish Pass, about 560 km East of the coast of Newfoundland. Lenz et al.¹⁴¹ used high resolution SR EXAFS and XANES combined with LA-ICP-MS and μ-XRF spectrometry to investigate the nature of the Mn enrichments in sediments from the Holocene Thermal Maximum (8000–4000 years BP) at a site in the northern Gotland Basin (Baltic Sea). Their particular interest was in the redox conditions in the bottom waters during this period. Schittek et al.¹⁴² used an XRF core scanner to measure the inorganic geochemistry of high altitude peat samples from the western Andes of southern Peru to elucidate changes in climatic and environmental conditions over the last 8600 years; sufficiently recent to have archaeological as well as geo-environmental implications. A link to the local archaeological record was also made by Unkel and co-workers,¹⁴³ who used a high resolution XRF core scanner to analyse sediments from the Asea Valley in the central Peloponnese, Greece. Their results extending back over a 6500 year period and provided evidence of climatic conditions going back to the Late Bronze Age.

The use of XRF spectrometry in classical geochemical studies is still popular, and the diversity of contributions is illustrated by the work reviewed here. Ulrich et al.¹⁴⁴ investigated dissolution–precipitation processes governing the carbonation and silicification of serpentine from the New Caledonian ophiolite, using μ-XRF images performed on a cm scale to describe relationships between serpentinite and alteration products. In contrast, μ-SRXRF was used by Dyl et al.¹⁴⁵ to undertake quantitative trace element mapping of carbonaceous chondrites (Allende and Vigarano) with the ultimate aim of understanding both nebular and parent-body processes within meteorites. Their work was undertaken at the Australian SR facility using a detector array called Maia, which allowed data to be collected from a 2 μm spot with very short dwell times (0.1 to 0.5 ms), facilitating the recording of a map from an entire thin section in about 5 hours. The identification of the chemical form of sulfur compounds in Japanese pink coral was the topic of interest to Tamenori et al.¹⁴⁶ using μ-XRF spectrometry combined with soft X-ray photoabsorption speciation mapping. Results showed that sulfate was the primary species in the coral skeleton with minor amounts of organic sulfur and the authors described their mineral and skeletal associations. Both XANES and EXAFS were used by Kashiwabara et al.¹⁴⁷ to understand the chemical processes involved in the extreme enrichment of Te in marine ferromanganese oxides. Mechanisms for the scavenging of the dominant Te^VI species from seawater were described. Kelly et al.¹⁴⁸ investigated a submarine volcanic vent site (Foerstner, Pantelleria, Italy) which, during an eruption in 1891, produced lava balloons that carried gas to the surface before sinking to the sea floor. The XRF and ICP-MS techniques were used to analyse samples and contributed to a petrographic and geochemical description to explain the evolution of such features and to allow a comparison with other known Strombolian eruptions. Deseta et al.¹⁴⁹ undertook a study of rocks from the Eocene Schistes Lustres Complex in Corsica to provide insights into the mechanism of intermediate-depth earthquakes using analytical contributions from both XRF spectrometry and EPMA. Microtextural and geochemical results from this study confirmed that the presence of abundant H₂O-rich minerals in the slab exerted a strong rheological control during high strain deformation, facilitating thermally-triggered and localised shear instabilities. The Burgess Shale is well known for its exceptional preservation of soft tissue from the middle Cambrian period and Pushie et al.¹⁵⁰ reported the first chemical evidence of blood in the invertebrate fossil record. Data for this interpretation came from SRXRF imaging of fossil specimens that revealed a black stain commonly preserved with the arthropod Marrella splendens, which was preferentially enriched in Cu, noting that modern-day arthropods generally use a copper-containing hemocyanin in their blood for oxygen transport. Wynn et al.¹⁵¹ worked at the ESRF in Grenoble, France to map Zn and SO₄²⁻ seasonal patterns in speleothem calcite. Delivery of Zn to the speleothem was found to be dependent upon the presence of excess water, the condition of any overlying snowpack and soil solution pH as controlled by microbial activity. The current analytical resolution achieved at the Grenoble facility was thought to be capable of providing some of the first evidence linking event-based meteorological (temperature and precipitation) records to trace element content of speleothem calcite, building towards reconstruction of indices of climatic change.

Environmental geochemistry is potentially a broad topic illustrated by a variety of topics in this review. Over the years, the toxic effect of fluorine in ground water has been the subject of a number of studies, continued by Mondal et al.,¹⁵² who elucidated the controls on fluoride concentrations in groundwater from the alluvial aquifers of the Birbhum District, West Bengal, India. Data from XRF and XRD techniques and total F⁻ analysis of sediments from the aquifer zone revealed that intercalated zeolitic clay within the aquifer sand was the main source of F⁻ which was mobilised by sorption and desorption at different pH levels as well as ion exchange in the pre-and post-monsoon seasons. The effectiveness of absorption materials (high aluminium bauxite ore, fabricated zeolite and activated neem seeds) designed to reduce the high fluoride levels in drinking water derived from groundwater wells in northern Ghana were investigated by Buamah et al.¹⁵³ High aluminium bauxite ore was found to have the highest fluoride removal capacity and XRF results showed that its principal elemental components were Al, Fe, Si and Ti (with C and O). Alternative materials for the removal of fluoride from ground water were characterised by a range of techniques, including EDXRF spectrometry by Mohan et al.¹⁵⁴ These materials were corn stover biochar and magnetic corn stover biochar (corn stover is the leaves and stalks of maize, left in the fields after harvest) and both were considered to be an effective substitute for expensive commercial absorbents. The way in which clayey materials in river basin water could enhance microbial contamination of river water often resulting in the formation of biofilms was studied by Fosso-Kankeu et al.¹⁵⁵ using the XRF and XRD techniques to determine their mineralogical composition. The authors considered that the attachment of micro-organisms could be ascribed to the presence of muscovite, which facilitated adhesion. In a similar vein, As and Fe speciation and distribution in organic flocks from the streambed of an area of arsenic-enriched peat land at Gola di Lago, Switzerland were studied by ThomasArrigo et al.¹⁵⁶ Mössbauer and SR (μ-XRF, AS, XRD) techniques were used to characterise the species present and the authors concluded that the floc bio-organics primarily served as nucleation sites for the precipitation of nanocrystalline Fe^III-(oxyhydr)oxides, rendering these flocks as effective absorbents for metals and metalloids. Turning to another geochemical application, Ohsawa et al.¹⁵⁷ studied the brownish colouration that occurred in the summit crater of Mount Shinoe-dake, Kirishima Volcano, Japan using the XRF and FTIR techniques to show the presence of schwertmannite (Fe₈O₈(OH)₆SO₄), probably formed by iron-oxidising bacteria that use pyrite (FeS₂) as an energy source. Geothermal energy is often generated in volcanic regions and XRF, XRD and SEM instrumentation were used by Demir et al.¹⁵⁸ to characterise the type of scaling formed during the extraction of geothermal brine at 173 °C in the Tuzla area, north western Turkey. Down the brine extraction well, scaling was found to be mainly composed of galena (PbS) and aragonite or calcite (CaCO₃); however, along the surface pipeline, a heterogeneous mixture of saponite-like amorphous structures, and sub-μm sized amorphous silica particles, layered double magnesium and iron hydroxide and NaCl were found.

Sediments can be collected and analysed for a number of reasons, sometimes related to topics such as climate change covered in previous paragraphs, sometimes related to environmental and anthropomorphic investigations, as the following examples will show. Thus, Cooper et al.¹⁵⁹ addressed two developing issues in research associated with identifying the source of fluvial sediments, namely improving the temporal resolution in sediment sequences (whilst minimising analytical costs) and consistently quantifying all perceived uncertainties associated with the sediment mixing model procedure. The first issue was addressed using direct XRF spectroscopy and diffuse reflectance FTIR in the analysis of suspended matter on filter papers used in conjunction with automatic water samplers. A realistic assessment of uncertainty was achieved using a Bayesian mixing model procedure to provide full characterisation of spatial geochemical variability, instrumental precision and residual error. These procedures were applied to the River Wensum catchment, UK and allowed the apportioning of sediment contributions to eroding arable topsoils, damaged road verges and combined channel bank and agricultural field drain sources at 60 to 120 minute intervals during five precipitation events. Schreiber et al.¹⁶⁰ were interested in applying pattern recognition methods to data from two sediment cores taken from an urban freshwater lake in Copenhagen. A PCA of XRF scanning results allowed the identification of a geogenic component in the sediment (Si, Fe, K, Rb) and a contamination component (Cu, Pb, Zn) with a temporal variance in contamination whereby Pb was superceded by Zn. In a completely different field of study, Liu et al.¹⁶¹ used XRF as one of the techniques to identify storm surge deposits in a coastal wetland along Lake Pontchartrain, southern Louisiana, USA caused by Hurricane Isaac in August 2012. Analysis of core and surface samples allowed the identification of two distinct sedimentary signatures for Hurricane Isaac deposits. Naquin et al.¹⁶² were also interested in storm deposition induced by hurricanes, this time associated with a marsh adjacent to a back-barrier lake along the Gulf of Mexico Coast of Louisiana (USA). Loss on ignition and XRF analysis were used to construct lithological and geochemical core profiles that resulted in the identification of three distinct sand layers deposited by recent hurricanes. The geochemical behaviour of coastal surface sediments offshore of the Cauvery Delta, Southeast India were studied by Silva et al.¹⁶³ using conventional XRF analysis to determine the major and trace elements in 16 surface sediment samples. The presence of enrichment and/or contamination by elements such as Cr, Cu and Zr was reported and a correlation between Cu and Zn with CaCO₃ was interpreted as showing the role of carbonates in precipitating these elements from the overlying water column, possibly related to agricultural pollution. Bataillard et al.¹⁶⁴ investigated As and Fe speciation in marine sediments following a re-suspension event, using the XRD, μ-XRF, ED-SEM, Raman and XANES techniques. They analysed sediment from L'Estaque Marina in France located near an industrial plant that had processed arsenic-bearing ores for several decades. The authors reported that unperturbed sediment was anoxic, retaining As in the 3+ state. After oxidation during a re-suspension event, As was present in the 5+ state. However, in the presence of a bacterial mat which consumed oxygen so preserving the sediment from total oxidation, both oxidation states of As could be present in the top of the sediment. A high temporal resolution record of pollution in recent sediments from Lake Windermere in the English Lake District was reported by Miller et al.¹⁶⁵ Core scanning XRF, radiochronological techniques and lead isotope measurements were used to identify a significant increase in Cu, Pb and Zn concentrations since the 1930s and three major sources of anthropomorphic lead were identified: gasoline Pb, coal combustion Pb (from coal-fired steam ships) and Pb derived from Pb–Zn mining activities. Care was recommended by Schillereff et al.¹⁶⁶ in the interpretation of μ-XRF core scanning due to variable down core water and organic content affecting X-ray attenuation, in a review of the use of lake sediment data to elucidate the frequency and magnitude of flooding events over centennial and millennial time scales. As part of this review, the authors presented a protocol illustrating the analytical techniques available to palaeoflood researchers.

The analysis of soils is another classical application of XRF spectrometry, this year exemplified by the work of Di Giuseppe et al.,¹⁶⁷ who used WDXRF data coupled with multivariate statistics to characterise the alluvial soils of the eastern-most Po Plain in Italy. The resultant data base was interpreted to develop a soil classification method that could be used to constrain the depositional environment of the alluvial sediments. Jayawardana et al.¹⁶⁸ used XRF spectrometry to measure the major and trace elements in agricultural soils from North-central Sri Lanka in a study designed to identify whether there were any links between the presence of contaminants and health hazards (such as chronic kidney disease) found in the area. Their results indicated that there was no significant threat from As and most other trace elements, but that there was a potential environmental risk from Pb and V due to their significant enrichment in soils. Challenges in developing sensors for the in situ quality assessment of soil to contribute to precision agriculture, for example, was discussed by Kaniu and Angeyo.¹⁶⁹ These authors described a chemometrics procedure based on EDXRF measurements as a ‘point of care’ sensor that inferred soil properties and intelligently modulated precision agriculture. The assessment of soil cation exchange capacity was the topic of interest to Sharma et al.¹⁷⁰ using portable XRF spectrometry. A large number (450) of soils from California and Nebraska, USA, were analysed and multiple linear regression was used to establish the relationship between laboratory-determined cation exchange and portable XRF elemental data, demonstrating that this approach was capable of the accurate prediction of soil cation exchange capacity. Acid sulfate soils, which arose when several elements in a sediment were mobilised, were investigated by Proske and co-workers¹⁷¹ based on measurements on two sediment cores from Western Australia. Their work demonstrated how XRF core scanning together with detailed sediment descriptions could be used to assess different zones of acid sulfate soils in areas where this phenomenon is suspected to occur. Portable XRF instrumentation was used by Badia et al.¹⁷² to determine total CaCO₃ in quaternary soils on the terraces of the River Alcanadre in the semi-arid Ebro Basin, Northeast Spain. Comparison of results with calcimetry measurements in the laboratory showed a sufficiently good correlation to confirm the efficiency of the in situ XRF approach for measuring this species.

As a contribution to other geological applications, in the Strandflat area of Nordland, northern Norway, Olesen et al.¹⁷³ argued that there was a causal link between deep weathering and increased seismic activity. They used XRF analyses to show leaching of the major elements, which, with a detailed analysis of geophysical measurements, contributed to this study. Regmi and colleagues¹⁷⁴ studied the weathering and mineralogical variations in gneissic rock and their effect in the Sangrumba Landslide, Nepal, one of the largest and most active landslides in the Nepal Himalaya. Elemental data on rock and soil samples by XRF spectrometry contributed to this study, which concluded that the mechanical strength of rock/soil was drastically decreased as the weathering intensity increased; the latter being dominantly influenced by rainfall than by rock type. As a contribution to regional geochemical mapping, Cheng and colleagues¹⁷⁵ presented a new QC method (‘visualised standard mapping’) to determine mean and background values and present the distribution of 76 elements in southern China, determined on stream sediments by XRF and ICP techniques. Dantu¹⁷⁶ presented the spatial distribution and geochemical baselines of major and trace elements in soils of the Medak district, Andhra Pradesh, India, partially to establish background values for future soil surveys. In a project to map bedrock lithologies in the Agnew-Lawlers area, Western Australia, Barnes et al.¹⁷⁷ reported that by taking advantage of the potential to determine several elements of interest with adequate precision and accuracy using portable XRF technology, new opportunities were available for the rapid geochemical mapping of bedrock in residual terrains. Portable XRF instrumentation was also used for mapping weak geochemical anomalies by Cheng¹⁷⁸ to obtain data from drill holes into Mo–W and Mo–Ag deposits in eastern Inner Mongolia in a project designed to evaluate the way elements associated with mineralisation decay in concentration with increasing distance from underlying rocks or saprocks. The author presented a new non-linear differential equation to model this trend. Gazley et al.¹⁷⁹ used elemental data from portable XRF analyses of core and underground working face samples from the Plutonic gold mine in Western Australia to develop a scheme for objective data logging. In their scheme, a mineralisation code was allocated based on the K : V concentration ratio (degree of alteration) and As + 12Cu (the degree of sulfide enrichment). In a different geological application, Edwards and co-workers¹⁸⁰ used SRXRF elemental mapping and XAS to characterise fossil leaf material from the Green River Formation (USA). Results showed that the fossils contained organometallic and organosulfur compounds that had been preserved in situ for 50 Ma. The performance of hand-held XRF instrumentation for the bulk analysis of iron meteorites was described by Gemelli et al.,¹⁸¹ analytical precision and accuracy being evaluated against metal alloy CRMs and iron meteorites of known chemical composition. The authors reported that most elements of higher atomic number than Mg could be determined accurately and that this technique represented a valuable tool for identifying extra-terrestrial origin of metallic samples rapidly and reliably. This would allow a preliminary classification of iron meteorites, identify mis-labelled or unlabelled specimens in museum or private collections and assist in the determination of the bulk composition of iron meteorites.

3.2 Industrial minerals and consequences from mining activity

As a contribution to the development of XRF techniques in this sector, Fisher et al.¹⁸² expressed concern about the lack of consistency in portable XRF data used in applications in the minerals industry and developed a workflow for the collection of portable XRF data in exploration and mining. The reliability of this workflow was demonstrated by a good correlation between conventional laboratory and portable XRF data in the analysis of drill hole samples at the Plutonic Gold Mine, Western Australia. Furthermore, fine scale lithostratigraphic variations were recognised in portable XRF data on grade control pulps from Agnew Gold Mine, Western Australia, sufficient to distinguish alteration signals from the background lithology and to discern which of these signals were associated with gold mineralisation. Hall et al.¹⁸³ undertook a study to evaluate the strengths and weaknesses of five portable and two benchtop EDXRF instruments in mineral exploration and mining. Their methodology involved the analysis of forty-one CRMs using factory calibrations and they categorised elemental performance as very good (As, Cu, Nb, Pb, Rb, Sr and Y), good (Ba, Mo, Sn, Zn and Zr), moderate (Cr, Sb, Se, Th and U), poor (Ag, Cd, Co, Ni and V) and very poor (Au, Bi, Cs, Hf, Hg, Pd, Sc, Ta, Te and W), providing an analytical evaluation for these and other elements of interest. Hunt et al.¹⁸⁴ presented mass fraction data for a comprehensive array of major and trace elements by EDXRF and WDXRF spectrometry and INAA for three fire clay RMs manufactured by Mittal Steel Ostava a.s., namely NH137, NH138 and NH139, as part of a characterisation study.

XRF spectrometry continues to play a significant role in the assessment of hazards from mining activities. Thus, Lortzie et al.¹⁸⁵ used XRF as one of the techniques to investigate the long term environmental impact from a gold–silver enrichment plant and associated mining activities in Mitsero, Cyprus, which was abandoned 70 years ago without remediation. Tailings samples were found to contain high levels of the leachable metals, Cr, Cu, Fe, Mn, Pb and Zn with high concentrations of cyanide. With the potential for acid runoff mobilising these contaminants, the authors suggested that there was a compelling argument for the remediation of legacy sites. Smith and co-workers¹⁸⁶ investigated mercury pollution associated with calcined samples from a closed mine in San Benito County, California, USA, using optical microscopy, μ-XRF, μ-AS and Hg isotope analysis to characterise the different layered structures in these samples. Their main observation was that incomplete roasting of mercury sulfide ores could cause extreme mass-dependent mercury isotopic fractionation, an important property for use as a tracer for the release of mercury from closed mercury mines. In another speciation study, Yang et al.¹⁸⁷ used the EXAFS and XANES techniques to determine Cu in soil associated with a mining site that was contaminated since the 1950s. Results showed that Cu was associated with iron oxides, not organic matter, an observation that facilitated the prediction of the reactivity and environmental fate of this contaminant. Tuovinen et al.¹⁸⁸ used portable XRF spectrometry for the in situ determination of Th and U in ores and mill tailings from a former pilot-scale phosphate mine in Sokli, northern Finland and a nickel mine at Talvivaara and undertook comparisons with laboratory methods of analysis (including gamma and alpha spectrometry and ICP-MS). They reported that the portable XRF technique gave comparable results for Th, even without prior sample preparation. The effect of wind-blown contamination originating from an abandoned manganese mine in the arid Metropolitan Las Vegas, Nevada, USA was of concern to Park et al.¹⁸⁹ XRF determinations were used to show high concentrations of the toxic elements As, Mn and Pb in tailings samples, with the furthest distance before levels were considered to reach background being 4.8 km. Perlatti et al.¹⁹⁰ used the XRF and XRD techniques as part of a multitechnique study to evaluate the geochemical speciation and dynamics of Cu in tropical semi-arid soils that were exposed to metal-bearing mine wastes associated with an open pit copper mine in Brazil. Their interesting conclusion was that the use of plants and organic amendments in mine sites with wastes containing high concentrations of copper carbonate should be viewed with caution as the practice might enhance the mobilisation of copper due to an increase in the rate of dissolution of the carbonate.

A number of contributions during the current review period investigated the use of minerals for the remediation of acid mine drainage and other forms of pollution. For example, Vadapalli et al.¹⁹¹ investigated the use of fly ash to neutralise acid mine drainage with a particular interest in the effect of particle size distribution. XRF spectrometry was used to demonstrate that the fine fraction was most effective in reducing the Mn content of acid mine drainage and that contra-intuitively, it was the coarse fraction that tended to increase the viscosity of the resultant suspension. Iakovleva and co-workers¹⁹² tested the properties of two limestones and two solid wastes for the removal of Cl⁻ and SO₄²⁻ from synthetic and real alkaline process mining water, analysing the composition and surface structure using EDXRF and a number of other techniques. For these two ions, absorption efficiency was found to be 96% and 99% respectively for synthetic solutions and 74% and 85% respectively for real process waters. The potential of crushed seashells as an absorbent for treating acid mine drainage was investigated by Masukume et al.¹⁹³ and after characterising absorption parameters using the XRF and FTIR techniques, the authors concluded that this material had great potential as an alternative low cost medium for the treatment of these drainage waters. The remediation of soil and groundwater contaminated with chlorinated hydrocarbons was the topic of interest to Rosales et al.¹⁹⁴ and especially the delivery of active nanoscale zero-valent iron particles to contaminated hot-spots. The authors used μ-XRF spectrometry to investigate the sub-surface behaviour of these NPs and the parameters that effected their transport in clay. Alshameri et al.¹⁹⁵ investigated the role of TiO₂/Yemeni natural zeolite for the enhanced removal of phosphate ions from water. XRF data contributed to this study which concluded that a TiO₂/zeolite composite was a highly efficient and economical material in this application with good regeneration properties. Geopolymers made from kaolin from the Sinai in Egypt, activated with blast furnace slag were reported by El-Naggar¹⁹⁶ to be a highly efficient material for the immobilisation of radioactive waste containing ⁶⁰Co. The XRF, XRD, FTIR and SEM techniques were used to characterise the raw materials used in this study and their alkali activated products.

It goes without saying that the properties of minerals dictate their industrial use, as illustrated by the work of de Menezes et al.,¹⁹⁷ who used XRF spectrometry and XRD to investigate the parameters that affected the colour and shade of ultramarine zeolite pigments that were synthesised from kaolin waste. Colour additives (sulfur and sodium carbonate) were calcined with zeolite A and the authors reported that the cooling rate after calcination was important in inducing changes in the colour of the product. Musyoka et al.¹⁹⁸ reported on the conversion of South African clays into high quality zeolites. The authors used XRF as one of the techniques for characterising the as-received clays as well as the resulting zeolitic phase. The influence of pH on the membrane behaviour of bentonite amended clay from Fa, China was investigated by Tang et al.¹⁹⁹ using the techniques of XRF, XRD and SEM to measure membrane properties. The relevance of this work is the use of this material as liners for landfill sites to provide a barrier against the migration of contaminants. Both XRF and XRD analyses were used by Picanco et al.²⁰⁰ to characterise Portland cement activated with pozzolanic zeolite sandstone from the Parnaiba Basin, Brazil. This study was designed to evaluate the ideal amount of thermally activated zeolitic sandstone to be incorporated into the Portland cement to improve its mechanical and mineralogical properties (10% was recommended). Sutcu²⁰¹ investigated the influence of expanded vermiculite on the physical properties and thermal conductivity of clay bricks, using analytical data obtained by the XRF, XRD, TGA and SEM techniques. The results showed that brick samples, with the addition of vermiculite, were suitable as insulating material in construction applications. Measurements by XRF spectrometry contributed to a study by Engidasew and Barbieri²⁰² of the properties of basaltic rock from Termaber, Central Ethiopia as a stone construction material and as an aggregate in concrete and asphalt. Although deleterious constituents were found within the middle basaltic flow layers, distinct flow layers were found to be suitable for this application. Baker and Strawn²⁰³ coupled experimental and analytical data (by Fe K-edge XAFS) and field observations to better understand the conditions for the formation of ferric smectite nontronite, comparing data from synthetic material with natural samples collected from weathered Columbia River basalt flows. They concluded that natural nontronite crystallisation in this region must have occurred at ambient near-surface temperatures over timescales of up to millions of years.

Minerals have many industrial uses such that here it is possible to feature only a small selection to illustrate the role of XRF techniques in their characterisation. Gliozzo et al.²⁰⁴ undertook the first review of the ceramic properties of thirty clays from Tuscany, Italy, and included measurements of their geochemical (XRF) and mineralogical (XRD) properties. Results distinguished Ca-poor clays suitable for the production of red stoneware and dense and frost-resistant masonry materials from Ca-rich clays suitable for the production of red earthenware as well as hard and porous masonry materials. Morsy et al.²⁰⁵ studied the modification of low quality Egyptian kaolinite to make it suitable as a pigment for the coating of paper. The modification methods were sedimentation, chemical bleaching, the preparation of organo-kaolinite and of nano-kaolinite. The XRF, XRD, FTIR and SEM techniques were used to characterise products and showed that paper coated with nano-kaolinite had a higher print density and higher print gloss than coatings formed from the original and commercial kaolinite. West et al.²⁰⁶ provided a practical guide for the assessment of glass-making limestones and dolomites providing practical advice and methodologies for the chemical analysis by XRF spectrometry, ICP-OES and chemical methods. A form of industrial mineralisation that is unwelcome was described in a report by Nitschke et al.,²⁰⁷ who used the XRF, XRD and SEM techniques to investigate the formation of alternating layered scaling comprising (Ba,Sr)SO₄ and PbS in a geothermal plant at Soultz-sous-Forets, Germany. The authors concluded that bacterial sulfate reduction occurred, initiating sulfide precipitation from sulfate-rich fluids and that the layered structure correlated with the operating state of the power plant. Sulfate layers were precipitated during normal operating conditions and sulfides during the start-up and shut-down phases. Particle size analysis, XRF spectrometry and XRD were used by Azdarpour et al.²⁰⁸ to study the direct carbonation of red gypsum (a waste product of titanium dioxide production) to produce solid carbonates. The authors reported that this mineral carbonation technology could potentially contribute to a reduction in CO₂ emissions and the remediation of waste material of environmental concern. Mohapatra et al.²⁰⁹ investigated the use of bastnasite ore as a phosphor material, with EDXRF results revealing the presence of Th and traces of Sm in the CeO₂ mineral matrix. Overall optical emission from this material was in the magenta region of the visible spectrum.

Selected from a number of contributions considering coal mining, the processing of coal products and the environmental legacy, Chen et al.²¹⁰ investigated the response of a coal deposit at the Pansan Mine in the Huainan coalfield (Anhui, China) to an igneous intrusion using a battery of techniques, including XRF. The authors provided a detailed assessment of the changes that occurred, caused by contact metamorphism, interaction with hydrothermal fluids, volatilisation and interaction with groundwater. Kelloway and co-workers²¹¹ demonstrated the benefit of automated EDXRF core scanning for profiling the abundance of inorganic elements through exploration cores of coal seams using samples from the Goonyella Middle seam, northern Bowen Basin (Queensland, Australia). Calibration was achieved using pressed powder pellets of reference coal samples and the validity of measurements confirmed by comparison with conventionally-determined chemical and mineralogical data for a representative core. A high pressure pelletising technique for coal was used to prepare samples for WDXRF analysis by Li et al.²¹² Coal samples were directly briquetted without binder at a pressure of 1400 kPa, producing a more compact and smooth surface that produced lower detection limits than a pellet produced by the routine process. XRF was one of the techniques used by Tremain et al.²¹³ to analyse char created from coal tailings sourced from two Australian coal mines. Their interest was to investigate the effects of different pyrolysis conditions as part of a novel waste management strategy for this type of waste material. Wang et al.²¹⁴ investigated the recovery of vanadium by flotation from low-grade stone coal, measuring the mineral composition and microstructure of the source material using the XRF, XRD and SEM techniques. The influence of coal spoil tips on Tetraena mongolica, a Chinese nationally protected plant, was published by Wang et al.²¹⁵ using data from the XRF, AFS and ICP-MS techniques. They reported that the elements As, Cu, Mo and Pb were highly enriched in soil, coal and coal gangue, with leaching and volatilisation being the two main pathways whereby spoil tips affected the environment. Jablonska-Czapla et al.²¹⁶ researched the long term impact of the mine waste stored at the coal waste dump at Halda Ruda on the heavy metal content of bottom sediments of the Bytomka River, Poland. EDXRF spectrometry was used to determine total element content and sequential extraction was used to measure geochemical fractionation of the elements As, Cd, Cr, Cu, Mn, Ni, Pb and Zn. The elements Mn, Pb and Zn were considered to be the most hazardous, because of their high total concentration and because they were associated with the most mobile fractions (ion exchange and carbonate) of the sediment. Duane²¹⁷ reported on a major hazard associated with 19^th century lignite mining around the cities of Halle and Leipzig, Germany, installing XRF as one of the techniques that contributed to a mobile laboratory used to investigate soil-gas and liquid samples and other complex samples from sub-surface waste pits and water basins. The conclusions of this investigation were that the clay layer liners containing contaminants had been breached over time due to overloading of the pits and/or structural damage, allowing chemical materials to permeate into the environment and into aquifers. Concerning a related topic, Mumford et al.²¹⁸ were concerned about the potential for acidic leachate from pyrite-bearing gravel pads that are required to elevate workings in the exploration of sub-arctic oil shales above the heaving effects of ground ice. They used the TXRF technique to analyse small liquid samples withdrawn from the pads and reported that there was significant potential for the leachate to contain Cu, Fe, Mn, Ni, S and Zn.

3.3 Industrial materials and consequences of industrial activity

The conventional XRF system is well established as an analytical tool for the analysis of steel samples. Three contributions from the research group of Tsuji extended its capabilities by developing a confocal 3D-XRF instrument with a vacuum chamber for elemental depth imaging to monitor the corrosion process of steel sheets. This instrument had a spatial resolution of about 10.9 μm at an energy of 17.4 keV. In a first publication²¹⁹ each layer of a painted steel sheet with a zinc phosphate conversion coating was observed non-destructively from the depth profile analysis of the characteristic elements. The painted steel sheet was then scratched to expose the substrate surface, and the sheet was immersed in a NaCl solution for 10 days. The 3D elemental map of Ti from the corroded steel sheet revealed blister-type corrosion around the scratch. The 3D elemental maps of Zn and P suggested that the zinc phosphate layer dissolved from the scratched region into the NaCl solution. In a second contribution,²²⁰ a low-carbon steel sheet was placed in artificial seawater (NaCl 3.5% m/v) and the corrosion process was observed in the solution by conducting a depth profile of the Fe Kα intensity in the NaCl solution with corrosion time by sequential line analysis at a fixed lateral position. The corrosion of Fe from the steel sheet, the diffusion of Fe ions into solution, and their condensation near a polymer window were successfully monitored. In their third contribution,²²¹ the confocal 3D-XRF technique was described for analysing the corrosion process under the coating layer of the steel sheet as well as for observing the Fe distribution in the NaCl solution during corrosion of Fe from the steel sheet. Dissolution, migration and deposition processes were successfully monitored. Pickering and Holland²²² recognised the value of the hand-held XRF system for mapping long-range variations in the chemical composition of metal components, without the need for extensive sample preparation. The applicability was successfully demonstrated by mapping macrosegregation in low alloy steel slabs.

TXRF spectrometry was used to determine the residual metal content in metallurgical slag after sulfuric acid extraction.²²³ Slurry sampling was evaluated as a simple sample preparation procedure for the determination of major (Ca, Fe, K and Si) and trace elements (As, Cu, Mn, Ni, Pb, Rb, Sr, Ti and Zn) in two certified reference materials (lake and river sediments). The precision calculated from three replicates was usually lower than 10% and the detection limits were in the range from 1.1 to 1079 μg g⁻¹. A two-element internal standard (Ga + P) was employed to determine all the elements. It was found that after sulfuric acid treatment As and Pb remained accumulated in the slag, whereas Cu, Fe and Zn were extracted at a percentage >90% from the slag. Bernal et al.²²⁴ performed nanoprobe SRXRF mapping to show, for the first time, that discrete iron-rich, titanium-rich and manganese/silicon-rich particles were present in blast furnace slag grains, and it was observed that these particles remained intact when the slag was used as a precursor for alkali-activated slag binders. However, there was no evidence of chemical interaction between these particles and the alkali-activated slag binder, which mainly comprised calcium silicate hydrates. Wu et al.²²⁵ studied the separation of a phosphorus-containing phase from modified basic oxygen furnace slag using XRF spectrometry among other analytical techniques. The results reflected that, compared with the original slag, the modified slag not only contained a phosphorus concentrating phase with a phosphorus content three times higher than in the original slag, but also that the separation performance improved when undertaken in a weak magnetic field by precipitation of the spinel-ferric carrier mineral.

Although the analysis of alloy samples using the XRF technique is well established, still new progress is made in this field. Gorewoda et al.²²⁶ improved the sample preparation method for the analysis of tin-lead solder by WDXRF spectrometry, to overcome the problem of smearing and to obtain the smallest and most reproducible microstructures. The optimal method involved melting the sample at a temperature of 350 °C, quenching it immediately with distilled water, followed by a washing step with acetone and the preparation of a flat disc with a diameter of 30 mm using a hydraulic press. Bouchard et al.²²⁷ simplified the ISO 9516-1 standard method describing the borate fusion method for iron ore with WDXRF analysis. A method applicable for a wide range of samples from different origins could be implemented by using selected CRMs instead of preparing pure oxide standards and applying the same automated fusion method. Validation of the adapted method was conducted, by comparing results with data from the ISO 9516-1 method, and all deviations from the prevailing standard method parameters were pointed out. Sieber and Mortensen²²⁸ prepared validation RMs, applicable for XRF as well as ED-SEM measurements, by evaporative deposition of thin films of NIST SRM 1729 Tin Alloy (97Sn–3Pb). Films were created on high-purity nickel foil to mimic some typical electronics structures and prevent charging during ED-SEM measurements. Maximum thickness of these films was kept below approximately 1 μm to ensure that the entire thickness could be probed by the primary X-ray or electron beam and that measured X-rays originated from the entire thickness of all films. The authors indicated that this approach would enable a laboratory to demonstrate competence in a controlled manner. Barriobero-Vila et al.²²⁹ combined laboratory EDXRF and μ-SRXRF analysis to trace the microstructural distribution of alloying elements during heating. Microstructural evidence of the role of V during the alpha–beta phase transformation kinetics of a bimodal Ti–6Al–6V–2Sn alloy was found. This approach provided an advance in the current knowledge of the phase transformation kinetics of the alloy.

Analytical techniques to determine rapidly hazardous substances for verification of the prescribed limit values of these hazardous substances in different EU regulations remains popular. In the recycling industry, methods for the rapid determination of the concentration of Br and subsequently of brominated flame retardants (BFRs) in WEEE plastics is definitely of interest. Taurino et al.²³⁰ proved the effectiveness of XRF spectroscopy in combination with μ-Raman spectroscopy as an effective tool for the rapid detection of BFRs in plastic materials. These two techniques in combination were considered as a promising method for QC applications in the recycling industry. Gallen et al.²³¹ also successfully developed a rapid and effective non-destructive testing strategy to identify BFRs in the plastics of consumer products. Non-destructive XRF analyses of 1714 samples rapidly identified bromine in 92% of products later confirmed to contain BFRs. Surface wipes of 137 products identified tetrabromobisphenol A, c-octa polybrominated diphenyl ethers (BDE) congeners and BDE-209 with relatively high accuracy (>75%) when confirmed by conventional destructive chemical analysis (n = 48). A relationship between the amount of BFRs detected in surface wipes and subsequent destructive testing showed promise in predicting not only the types of BFRs present but also estimating their concentration. Miller and Harris²³² used a hand-held XRF spectrometer to quantify the hazardous metal content in vintage plastic toys. Lead and Cd was found in 67% of these toys, frequently at concentration levels exceeding the current US and European limits, while As was detected at levels of concern in 16% of the samples. Thankfully, none of the contemporary vinyl toys contained detectable As, Cd and Pb concentrations. Given that vintage toys remain in widespread use by children in homes and other locations, the results illuminated a potential source of heavy metal exposure for children. An interesting study was presented by Hillyer et al.,²³³ who combined a multitechnique approach for the quantitative analysis of As, Cd and Pb in children's toys and toy jewelry, that included FAAS and XRF spectrometry. The results showed socioeconomic considerations, indicating that the origin of purchase, rather than cost, was a significant factor in the risk assessment of these materials with 57% of toys/toy jewelry items from bargain stores being non-compliant or suspect compared with only 15% from retail outlets and 13% if only low cost items from the retail stores were compared. While jewelry was found to be the most problematic product (73% of non-compliant/suspect samples), Pb (45%) and As (76%) were the dominant toxic elements found in non-compliant/suspect samples. The discrepancy between bargain and retail children's products, along with growing numbers of bargain stores in low-income and urban areas, exemplified an emerging socioeconomic public health issue. Fellin et al.²³⁴ also underlined the need for a fast method to quantify the elemental content of solid wood and wood-based materials in order to trace polluted (e.g. with chromated copper arsenate) samples. In a sampling study carried out mainly in north Italy, 336 wood waste specimens were collected and analysed with the EDXRF technique, implementing a method with short scan time. Roughly 84% of the specimens complied with the EU decision (Ecolabel) on heavy metals. Most detected heavy metals originated from furniture and building materials, whereas packaging and specimens of unknown origin presented no major concern. Very high concentrations of Cl, Cr and Pb were found.

On a regular basis, the impact of industrial discharges to the environment is studied. The XRF technique is routinely implemented to conduct these studies, as demonstrated by Ali and Ateeg,²³⁵ who evaluated soil polluted with heavy metals for 24 surface soil samples (0–10 cm in depth) from various locations that covered an industrial area in Omdurman city, Sudan. It was found that Cd, Cr and Cu concentrations at almost all studied sites were higher than the normal values, while most of the As, Co, Ni, Pb and Zn concentrations were lower. Using the Pb poisoning crisis in small-scale communities in Zamfara State, Northern Nigeria as an example, where more than 400 children have already died as a result of ongoing Pb poisoning since early 2010, Bartrem et al.²³⁶ reviewed the common toxic risk factors of the major heavy metals to advance the understanding of co-exposures and their common pathologies. Environmental contamination in Bagega village, examined by the XRF analysis of soils, included As, Cd, Hg and Mn. Co-exposure risk was explored by scoring common toxic risk factors and hazard indices to calculate a common pathology hazard risk ranking of Pb > As > Hg ≫ Cd > Mn. Although Zamfara presented an extreme picture of both Pb and multiple heavy metal mortality and morbidity, similar situations have become increasingly prevalent worldwide. Dutta et al.²³⁷ used the EDXRF and INAA techniques for two cases; (a) enrichment of Zn levels in the soil (1080 ± 76 mg kg⁻¹) and Zn uptake ranging between 40 and 628 mg kg⁻¹ in edible plants and cereals grown in agricultural fields near an active zinc smelter, Rajasthan and (b) the depth-profile distributions of Cr, Pb and other metals in the sediment core samples from the Sundarban wetland, West Bengal, associated with waste discharges from tanneries and other industries. The extent of contamination in sediments in the Balanagar industrial area, Hyderabad, Andhra Pradesh, India was studied by Machender et al.,²³⁸ who analysed sediment samples by XRF spectrometry. High concentrations of Cr, Co, Cu, Ni, Pb, V and Zn were found, attributed to some pharmaceutical and metal industries in the study area. These data were valuable in defining suitable remedial measures such as phytoremediation and bio-remediation for the reduction of heavy metals in sediments. As well as monitoring the impact of industrial pollution, it is also of importance to reduce the environmental impact by, for instance, the stabilisation of Pb by the addition of P to contaminated soils and mine spoil materials, as was studied by Baker et al.²³⁹ The authors hypothesised that differences in the efficacy of Pb stabilisation in contaminated soils following fluid or granular P addition was due to different P reaction processes in and around fertilizer granules and fluid droplets. A combination of several SR-based techniques (i.e., μ-XANES and μ-XRF) was used to measure the speciation of Pb over two incubation times in a smelter-contaminated soil after the addition of several fluid and granular P amendments. The results indicated that all fluid and granular P sources were able to induce Pb phosphate formation, but fluid phosphoric acid and triple super phosphate were the most effective. Nevertheless the authors recognised that caution needs to be exercised when adding large amounts of soluble P to the environment. Hayes et al.²⁴⁰ also described a study dealing with a reduction of the environmental impact by a Tailings Management Facility (TMF) with respect to the uranium ore processing operation. Their philosophy was based on the idea that elements of concern, such as Mo, would be controlled in the very long term through equilibrium with supporting minerals. To understand how these reactions evolved toward an equilibrium state, samples for Mo speciation were taken from the TMF during the 2008 sampling campaign and were analysed by XRD, μ-XRF spectrometry and XANES. These results showed, as could be expected, that only XANES was effective in measuring the speciation of Mo in the tailings samples, because it was both element-specific and sensitive enough to detect the low concentrations of Mo present.

Some interesting contributions extend the applicability of XRF instrumentation. The development by Hutton et al.²⁴¹ of a novel analytical technique, electrochemical-XRF, was described and applied for the quantitative detection of heavy metals in solution, achieving sub ng mL⁻¹ limits of detection. Electrochemical pre-concentration of a species of interest onto the target electrode was achieved by cathodic electrodeposition. Unambiguous elemental identification and quantification of metal concentration was then made using the EDXRF technique, resulting in an improved LOD by 4 orders of magnitude. During electrodeposition it was possible to vary both the deposition potential (E_dep) and deposition time (t_dep). For the metal ions Cu²⁺ and Pb²⁺ the highest detection sensitivities were found for E_dep = −1.75 V and t_dep = 4000 s with impressive LODs of 0.05 and 0.04 ng mL⁻¹, respectively. Herzog et al.²⁴² studied the ability of angle-resolved XRF analysis to detect non-destructively Ga concentration gradients in Cu(In, Ga)Se₂ solar cells, relevant for their performance. The authors presented the effect of different Ga concentration gradients on fundamental parameter-based calculations of the XRF spectra in comparison with their influence on the solar cell efficiency as derived by device simulations.

Nowadays, the XRF technique, often in combination with other analytical methods, is routinely applied for the characterisation of a huge diversity of materials, covering catalysts, recycling waste, construction materials, fly ash, alloys, glass, oil, etc. The obtained results always serve as a basis for the development or evaluation of different kinds of industrial processes. Some typical examples are presented in this review, but not all could be included. Kumar et al.²⁴³ used the XRF technique to analyse fly ash and determine the main elements and small quantities of As₂O₃, BaO, Br, Cr₂O₃, CuO, MnO, NiO, Rb₂O, SrO, TiO₂, ZnO and ZrO₂ as active ingredients. This fly ash, produced by a biomass-based thermal power plant, was considered as a potential catalyst for biodiesel production. Zeolite-based adsorbents impregnated with REE (Ce, La or Pr) were analysed with an XRF spectrometer by Jung et al.,²⁴⁴ and were further evaluated for their adsorption properties towards organic sulfur compounds in LPG. The feasibility of using urban river sediments as a primary raw material to produce high-insulation brick was proven by Xu et al.,²⁴⁵ who employed the XRF technique in combination with XRD and TG-DSC analyses to determine the chemical, mineralogical and thermal characteristics of the material. Gazulla et al.²⁴⁶ fine-tuned the methodology for the determination of B in ceramic frits and borates by WDXRF spectrometry, by optimising the sample preparation and the measurement conditions to obtain the highest S/N. The proposed method was less time consuming and more environmentally friendly than the traditional analytical methods. Fatoba et al.²⁴⁷ evaluated the effects of the co-disposal of brine and fly ash, a common practice at South Africa power utilities. The XRF results showed that the concentrations of Na, Cl, Mg, and SO₄²⁻ (as S) in the ash residues were somewhat higher than their concentrations in the fresh fly ash. This study revealed that some species contained in the brine solution could be captured by the fly ash through secondary mineralisation during co-disposal in a closed static environment, while many other elements could be significantly leached into the brine. A rapid determination of trace metals in oil was described by Chu et al.,²⁴⁸ who used a hand-held EDXRF analyser for this purpose. The method involved minimal sample preparation, the use of reliable standards for calibration, and resulted in μg mL⁻¹ detection limits. Finally, Sun et al.²⁴⁹ characterised precious metals in electronic waste of end-of-life ICT products, using XRF spectrometry, DSC and SEM-EDS. Due to the high heterogeneity of the material, special sample preparation procedures were introduced to minimise the discrepancies during compositional analyses. The gathered data were expected to lead to smarter decisions during further development of a clean and effective recovery process. Additionally, readers interested in more applications on these topics can consult our companion review of advances in the analysis of metals, chemicals and functional materials.⁵

3.4 Environmental

Synchrotron-based facilities again demonstrated the extent to which elemental mapping and μ-XRF techniques were increasingly used to study the spatial distribution and speciation of elements within plants. The Stanford Synchrotron Radiation Laboratory (SSRL), beamlines 2–3, for μ-XRF analysis were used by Tian et al.²⁵⁰ to investigate trace foliar applied elements within sunflower, a significant oil seed crop. Experiments were recorded at 13.5 keV with the incident X-ray beam focused using a pair of Kirkpatrick–Baez mirrors, and a Si (111) double-crystal monochromator. Micro-XRF maps were obtained by rastering the beam at 20 or 5 μm steps, with a count time of 200 ms per step, for the major and minor/trace elements: Ca, Cl, Cu, Fe, K, Mn, P, S and Zn. The authors were able to show that foliar applied zinc combined with an organic bio-stimulant increased the mobility of zinc within the sunflower. The SSRL facilities were also used by Lu et al.²⁵¹ where μ-XRF spectrometry and EXAFS were used to report speciation and localisation of zinc in the hyper-accumulator plant Sedum alfredi. Two-dimensional μ-SRXRF imaging revealed age-dependent differences between the zinc content of stems and leaves. In old leaves, Zn was high in the midrib, margin regions and the petiole, whereas the distribution of Zn was reported to be essentially uniform in young leaves. Most of the Zn was found to be complexed with malate in the leaves, but a shift to cell wall and citric acid Zn complexes was noted during transportation and storage in stems and roots. Shen²⁵² reported the distribution and speciation of Pb in Arabidopsis thaliana shoot and rhizosphere soil by in situ μ-SRXRF spectrometry where the overlap of spectrum lines had an enormous impact on the accurate selection of a region of interest before the distribution of elements in plant samples may be scanned. Adjacent elements are known to contribute to complex interferences leading to the production of inaccurate distribution characteristics in the 2D maps of Pb. Lead speciation in the same samples was also determined using μ-XANES. The author stressed that it was important to define a combination fitting range because different possible Pb combinations could emerge using different ranges. Different speciation were found in the root; PbAc₂ and PbSO₄, stem; PbAc₂ and Pb₃(PO₄)₂, leaf; Pb(OH)₂ and Pb₅Cl(PO₄)₃, and seed coat; Pb₃(PO₄)₂, Pb(OH)₂, and PbCO₃, between the fitting range of E-0 − 20 eV and E-0 + 70 eV. A more comprehensive Pb XANES database with additional references, especially organic Pb compounds, was found to be needed. The Grenoble facilities were also used by de La Rosa et al.²⁵³ who were interested in the potential use of Helianthus annuun L. for the remediation of Cr^VI polluted waters. Hydroponic experiments were devised to determine chromium uptake and tolerance under different Cr^VI sulfate conditions, and chromium bio-transformations. The authors claimed that their results were the first to show that, using μ-SRXRF spectrometry, chromium reached the root stele and was located in the walls of xylem vessels. Bulk and μ-XANES data also showed that chromium in the roots was mostly present as Cr^III phosphate (80%), with the remainder complexed to organic acids. Another “first” was claimed by Archibald et al.²⁵⁴ who showed how iron interacts with sporopollenin exine capsules. Their X-ray absorption investigation using Fe K-edge XANES and EXAFS represented the first direct structural data on the interaction of metals with sporopollenin exine capsules thereby revealing the structure-property relationships. Workers in Krakow, Poland²⁵⁵ studied Cr distribution in shoots of macrophyte Callitriche cophocarpa by means of two X-ray-based techniques: μ-XRF and EPMA. A comparative analysis of Cr^III-treated plants demonstrated high deposition of Cr in epidermal glands/hairs localised on leaves and stems of the plant shoots. Whereas, Cr in the Cr^VI-treated group was found in vascular bundles, suggesting differences in chromium uptake, transport and accumulation dependent on the oxidation state of the element. Chen et al.²⁵⁶ confirmed Typha angustifolia had an excellent ability to accumulate Cr^VI from highly contaminated waste water. Synchrotron mapping showed that Cr uptake was mainly found in the outer layer of the roots with a small proportion uniformly distributed in the fronds. The consequences of radionuclide discharge into the environment in the northeast region of Japan following the devastation at the Fukushima Daiichi nuclear power station in March 2011 continue to feature in the scientific literature. Kowata et al.²⁵⁷ used μ-SRXRF spectrometry to study a possible mechanism for Cs absorption in Egeria densa, a submerged vascular plant. A collection of plants, waters and sediment contaminated by radioactive fallout in the Fukushima Prefecture in September 2012 was assessed for ¹³⁴Cs, ¹³⁷Cs and ⁴ K using a germanium semiconductor detector. Due to its chemical similarity, K exerts an influence on the Cs uptake by plants, even though the Cs : K ratio may not be uniform within the plant. Potassium fertilisers are known to inhibit radioactive Cs absorption. Using a 0.7 μm X-ray beam at the SPring-8 facility in Hyogo, Japan, approximate metal distributions in the samples were first examined in 50 μm steps followed by higher resolution analyses with 5 to 2 μm step sizes. Measurement time was 5 s per point. Using stable ¹³³Cs, plants were cultivated in hydroponic baths to simulate active Cs uptake and show the predominant localisation of Cs and K, as well as several divalent metals, in the cell wall or apoplastic regions, suggesting a possible absorption mechanism of Cs in the plants. The authors concluded that E. densa plays an important role in the deposition of radioactive Cs in fresh water. Huanglongbing (HLB) is a highly destructive, fast-spreading disease of citrus, causing substantial economic losses to the citrus industry worldwide. Tian et al.²⁵⁸ used μ-SRXRF spectrometry for the spatial imaging of Zn and other elements to show that preferential localisation of Zn to phloem tissues was observed in the stems and leaves collected from healthy grapefruit plants, but was absent from HLB-infected samples. Quantitative analysis, using standard references, revealed that Zn concentration in the phloem of veins in healthy leaves was more than 10 times higher than that in HLB-infected leaves. It was thought that reduced phloem transport of Zn was an important factor contributing to HLB-induced Zn deficiency in grapefruit. A useful addition to the understanding of mineral localisation in cereal grains was offered by Singh et al.²⁵⁹ with spatial micro-imaging of the distribution of mineral nutrients (Ca, Cu, Fe, K, Mg, Mn, P, S and Zn) in both maternal and filial tissues in grains of two wheat cultivars. Majumdar et al.²⁶⁰ used μ-SRXRF spectrometry to map CeO₂ NPs in kidney beans to study disturbance in the plant's defence mechanisms. The plants were analysed for Ce accumulation after one, seven and fifteen days exposure. The primary indicators of stress were cited as lipid peroxidation, antioxidant enzyme activities, total soluble protein and chlorophyll content. The chemical forms were identified using μ-XANES. In the root epidermis, cerium was primarily shown to exist as CeO₂, although a small fraction (12%) was bio-transformed to a Ce^III compound. Cerium was found to reach the root vascular tissues and translocate to aerial parts with time. Prolonged exposure to 500 mg CeO₂ L⁻¹ showed that the root antioxidant enzyme activities were significantly reduced, simultaneously increasing the root soluble protein by 204%. Other workers also exploited synchrotron mapping techniques to show effects on plants; Cd in Arbuscular mycorrhizas,²⁶¹ Mn in the algae Chara coralline,²⁶² Zn in tomato and citrus,²⁶³ Ce in cucumber²⁶⁴ and Ca and Zn in Euphorbia.²⁶⁵

The impact of engineered nanomaterials on plants, which act as a major point of entry of contaminants into trophic chains, was reported by Larue et al.²⁶⁶ The uptake of Ag-NPs in the lettuce crop, Lactuca sativa, after foliar exposure and the possible biotransformation and phytotoxic effects was studied using μ-SRXRF spectrometry and μ-AS, TOF-SIMS and electron microscopy. Silver was found to be trapped on the lettuce leaves despite thorough washing; crucial information for those assessing the risk associated with products containing NPs. The building industry is known to use NPs, especially in paints and so the same research group²⁶⁷ also considered possible transfer of TiO₂ NPs via foliar transfer. In this study, lettuces were exposed to pristine TiO₂ NPs and aged paint leachate containing TiO₂ NPs and TiO₂ microparticles, offering another perspective on risk from atmospheric contamination. Zhao et al.²⁶⁸ reported the nutritional quality of cucumber fruit, Cucumis sativus, attributed to the impact of CeO₂ and ZnO NPs. Changes in functional groups were detected using FTIR while ICP-OES and μ-XRF spectrometry were used to quantify and map the distribution of nutrient elements, respectively.

The literature continues to demonstrate interest in rice, hence analysts will welcome the publication by Inui et al.²⁶⁹ on the preparation of Cd-containing brown rice grains for use as calibration samples for XRF techniques. Generally determined by AAS, after a lengthy and complex digestion stage, this work offered a rapid alternative X-ray technique. Calibration samples were prepared by adding 10 g of base rice grains (from Japan) to 100 mL of methanol containing 5 to 100 μg of Cd. The mixture was heated, cooled, and stored in a silica gel desiccator. Seven grams of each calibration sample were packed into a polyethylene cup (32 mm internal diameter and 23 mm height) covered with a 6 μm thick polypropylene film and then subjected to XRF analysis. The calibration curves for Cd in brown (unpolished) and white (polished) rice grains showed good linearity in the 0.50–10 mg kg⁻¹ range. The detection limits for Cd in brown and white rice grains were 0.14 and 0.12 mg kg⁻¹, respectively. The expected absorption effects were corrected in the time honoured way using the Compton scattered Rh Kα X-ray tube line as internal standard for the Cd Kα measurements thereby giving identical slopes for the calibration curves of Cd in brown and white rice. Kyriacou et al.²⁷⁰ were interested in the localisation of Fe in rice grain as only a small fraction (5–10%) is bioavailable in human diets because the extensive co-localisation of Fe in phytic acid, a strong chelator of metal ions, results in the formation of insoluble complexes that we cannot digest. Distribution maps, by μ-SRXRF spectrometry and nano-SIMS, of Fe, P and Zn from the aleurone and sub-aleurone layers of mature, wild type and Fe-enriched Olyza sativa L. grains indicated that most Fe was co-localised with P in the aleurone layer but that a small amount of Fe, often present as “hotspots”, extended further into the sub-aleurone and outer endosperm in a pattern that was not co-localised with P. It is thought that Fe in sub-aleurone and outer endosperm layers of rice grain could be bound to low molecular weight chelators such as nicotianamine and/or deoxymugineic acid. The threat from paddy rice grown in waters contaminated by mercury in China was investigated by Meng and colleagues²⁷¹ as the health risk is known to be related to the chemical speciation of Hg inside the grain. Using μ-SRXRF mapping and XANES data the work showed that inorganic Hg in the bran was primarily bound to cysteine, and was associated with phytochelatins, consequently, being largely immobile and restricted to the outer layers of rice grain. Methyl Hg in bran was primarily bound to cysteine and associated with proteins. However, this methyl Hg-cysteine association behaved like a mobile nutrient and was actively transported to the endosperm during seed ripening. Concentration of methyl Hg-cysteine in white rice has implications for public health. Evidence of speciation of As near rice roots was found by Yamaguchi et al.²⁷² to depend on spatial and temporal variations in the soil matrix. The As^V to As^III ratio decreased toward the outer-rim of sub-surface iron mottles where the soil matrix was not completely aerated.

Turning now to TXRF spectrometry, Ribeiro et al.²⁷³ determined trace elements in honey from different regions in Brazil. Up to 12 elements (Br, Ca, Cr, Cu, Fe, K, Mn, Ni, Se, Sr, Ti and Zn) were detected in 160 samples of honey from four regions of Rio de Janeiro State. The results showed that samples from the Teresopolis region had higher concentrations of essential and non-essential elements than samples from the other regions. The analytes, Cu, Ni, Se, Sr, and Zn, were identified in small concentrations (0.01 to 12.08 g g⁻¹) in all samples, indicating a low level of contamination in all the sampling regions. The dependency of the trace element content of honeys on their geographical origin may be used for their provenance. A preliminary investigation was made by Margui et al.²⁷⁴ into major, minor, trace and ultra-trace elements in several clam species that were used for human consumption in Portugal and worldwide. Two X-ray techniques, EDXRF and TXRF spectrometry were used showing, unsurprisingly, the advantage of TXRF spectrometry for trace and ultra-trace analyses. Concentrations of ten mineral elements (Ca, Cu, Fe, K, Mn, P, Rb, S, Sr and Zn) were determined²⁷⁵ in 27 genotypes of Chinese dwarf cherry by TXRF spectrometry. The results indicated that the plants were a rich source of mineral elements, especially Ca, Fe and Zn, with the highest concentrations being 524, 24.2 and 4.86 mg kg⁻¹ respectively.

Advances in less sophisticated X-ray techniques have also featured in literature during this review period with a comparison of analytical performance of benchtop and hand-held EDXRF systems for the direct analysis of plant materials. Thus Guerra et al.²⁷⁶ were interested in the direct and simultaneous determination of Ca, Fe, K, Mn, P, S and Si in comminuted leaves of 23 varieties of sugar cane. Both instrumental configurations were reported to present similar figures of merit and were able to provide useful data for plant nutrition diagnosis. Linear correlation between the elemental mass fractions in the test samples and characteristic X-ray intensities was obtained for all analytes with both types of equipment. Correlation coefficients from 0.9601 to 0.9918 and from 0.9094 to 0.9948 were attained for benchtop and hand-held EDXRF spectrometers, respectively. The coefficient of variation of measurements carried out in 3 different test samples was also appropriate, being lower than 13% for all analytes. Limits of detection were comparable for both systems (20 mg kg⁻¹ for Fe and Mn, and approximately 0.1 g kg⁻¹ for P, K, Ca, S and Si) and permitted the evaluation of the mineral nutrition status of sugar cane crop taking into account the critical levels of these elements. The XRF calibration models were built from ICP-OES analyses of the 23 sugar canes. Given the inherent non-destructive nature of EDXRF spectrometry, the same sample pellets were analysed by both the hand-held and laboratory units. The hand-held system was recommended as a cost-effective and appealing option for those who intend to perform in situ measurements to provide useful data for plant nutrition status, crop requirements and potential deficiencies leading to a more effective fertilisation programme.

Synchrotron excited XRF micro-tomography has emerged as a powerful tool for 3D visualisation of elemental distribution in biological samples. The mechanical stability of both instrument and specimen is paramount for this work. By combining the progressive lowering of temperature with femtosecond laser sectioning, Bourassa et al.²⁷⁷ were able to embed, excise, and preserve a zebrafish embryo at 24 hours post fertilisation in an X-ray compatible, transparent resin for tomographic elemental imaging. Based on a data set comprised of 60 projections, acquired with a step size of 2 μm during 100 hours of beam time, the authors reconstructed the 3D distribution of Cu, Fe, and Zn using the iterative maximum likelihood expectation maximisation reconstruction algorithm. The volumetric elemental maps, which entailed collection of over 124 million individual voxels for each transition metal, revealed distinct elemental distributions that could be correlated with characteristic anatomical features at this stage of embryonic development. A newly discovered unicellular micro-alga, Coccomyxa actinabiotis, was found to be highly radiation-tolerant and to strongly concentrate radionuclides, as well as large amounts of toxic metals. Leonardo et al.²⁷⁸ used nano-SRXRF spectrometry at the ID22 imaging beamline of the ESRF to understand the mechanisms involved in the accumulation and detoxification of Ag and Co by the alga. The high resolution and high sensitivity of this technique enabled the assessment of elemental associations and exclusions in subcellular micro-algae compartments. This work was thought to be the first report on element co-localisation and segregation at the sub-cellular level in micro-algae by means of nano-SRXRF spectroscopy. Giant clam shells gain their proportions by consuming sugars and proteins produced by billions of algae that live in their tissue. Yoshimura and colleagues²⁷⁹ used μ-SRXRF spectrometry and XANES to investigate Mg and S profiles in the inner layer of these clam shells. The observed S profile implied a clear change in calcifying fluid chemistry towards a less alkaline condition with age. Magnesium fluctuations suggested that Mg was incorporated into the shells at high growth rates during warm seasons. The spectrum of the Mg K-edge XANES structure and comparison of Mg and S-bearing amino acids profiles indicated that a pronounced effect of the organic fraction or disordered phases was observed in aragonitic clam shell rather than the regulated substitution that was observed into the aragonite crystal lattice. In an interesting application²⁸⁰ for conventional XRF spectrometry, the durability of insecticidal nets under operational conditions was investigated. Deltamethrin residue levels were evaluated on 189 nets used for three to six months from nine sites, 220 nets used for 14 to 20 months from 11 sites, and 200 nets used for 26 to 32 months from 10 sites in Ethiopia. Insecticide residue levels were found to decline by about one third between 3 and 6 months and 14 to 20 months, but remained relatively stable and above minimum requirements thereafter for up to 26 to 32 months.

A study of Pb levels in soils from weathering of lead bullets was needed following the introduction of legislation concerning dove hunting in Cordoba, Argentina. Rubio et al.²⁸¹ collected soil samples at depth of 50 mm from 315 pits referenced by GPS in accordance with local environmental authorities for XRF analysis. The average Pb concentration was reported to be 80 μg g⁻¹ in dry soil sieved through 200 mesh. The decayed bullet casings examined by SEM and XRD revealed three lead minerals in the corroded alloy. In a supplementary investigation²⁸² using μ-SRXRF spectrometry the authors reported a positive correlation between Fe and Sb in the crust material on the outer rim of the weathered bullets due to Sb adsorption to Fe oxyhydroxides in the soil. A spatial correlation between Cu and Sb and between Sb and Zn were also seen in the bullet crusts. The influence of phosphorus applied as P-based fertilizers to the soil was observed in the hydroxyl-pyromorphite phase detected in samples collected near agricultural land. Portable XRF spectrometers with tube excitation are established tools for the determination of elemental composition in soils in the field and laboratory. The majority of this work involved mineral soil measurements using soil applications provided by the instrument manufacturer. Shand and Wendler²⁸³ however, extended the technique's use to include organo-mineral soils and peat, recommending that modifications to the manufacturer's calibration should be made.

Silver NPs are often applied to consumer products for their antimicrobial properties, which are desired in fabric used for sportswear as well as cloth used for cleaning. Hazards to human health from airborne Ag NPs may occur when the NPs are inhaled. Such particulates are comparable in size to macromolecules and viruses and are able to penetrate deep into the lungs, e.g., the alveoli, where they may cause damage to cells and tissue due to their large surface area. Menzel and Fittschen²⁸⁴ collected aerosols released from fabrics treated with Ag NPs using a low pressure Berner impactor for subsequent analysis by TXRF spectrometry and SEM. The sample cell was purged with N₂ to eliminate Ar peak interference, offering an impressive LOD of 0.2 ng Ag. The authors found that the particulates were primarily released in the form of larger particles, mainly 0.13–2 μm, probably attached to fibrous material. Following these analyses, the aerosol samples were dissolved in nitric acid and analysed by ICP-MS to successfully confirm the results obtained by the TXRF measurements.

Other environmental applications reviewed this year included the determination of U in bituminised radioactive waste drums²⁸⁵ by self-induced XRF. The method used the spontaneous 661.7 keV gamma emission from the drums following ¹³⁷Cs decay produced by Compton scattering in the bituminised matrix giving an intense photon continuum around 100 keV i.e. in the uranium XRF region. The authors presented the experimental qualification of the method with real waste drums, showing a detection limit well below 1 kg of U in 20 minute acquisitions while the usual gamma rays of ²³⁵U or ²³⁸U were not detected. The relative uncertainty on the uranium mass assessed by this self-induced technique was about 50%, with a 95% confidence level, taking into account the correction of photon attenuation in the waste matrix. For a comprehensive review of advances in environmental analysis, readers are reminded of our companion ASU Update.⁶

3.5 Archaeological and cultural heritage

The provenancing of geological artefacts of archaeological significance was one of the earliest archaeological applications of XRF spectrometry (predating portable instrumentation) and it is a pleasure to feature here one of the most high profile applications of this approach, namely identifying the origin of the blue stones at one of the most important Neolithic sites in the UK, Stonehenge. It has long been recognised that many of these blue stones originated from the eastern Mynydd Preseli area of SW Wales and Bevins and colleagues²⁸⁶ reanalysed the earlier data set of Thorpe and colleagues of The Open University, emphasising the importance of data for Fe₂O_3, Cr, MgO and Ni elements compatible with the major crystallising minerals of spotted dolerite. They concluded that a single outcrop, Carn Goedog in the eastern Mynydd Preseli was the source of the majority of the blue stones. The WDXRF technique was used for more traditional provenancing by Rolett et al.,²⁸⁷ whose interest lay in elucidating the colonisation of eastern Polynesia from the distribution of stone adze implements that were shown to originate from the Vitaria Adze Quarry, Rurutu, Austral Islands. Portable XRF instrumentation was used in conjunction with Raman spectroscopy by Sepulveda et al.²⁸⁸ to show for the first time the use of K-jarosite and natrojarosite in pre-Hispanic times (about 2500 years BP) at the archaeological site of Playa Miller 7 in northern Chile on the coast of the Atacama Desert. Results were used to infer a hydrothermal source of this yellow pigment in a geothermal area in the Andes. Obsidian remains a popular material for provenancing studies and the work of Pappalardo et al.²⁸⁹ was interesting because of the complementary use of XRF instrumentation with beam stability controlled and PIXE as portable spectrometers for the analysis of 800 obsidian artefacts from quarries in Locodia Eubea, Sicily. The elements Nb, Rb, Sr, Y and Zr were considered to be the most effective discriminators by XRF spectrometry while PIXE contributed the major elements from Na to Zn. Analysis of the results allowed the authors to identify the islands of Lipari and Pantelleria as the only two sources of the selected samples. Frahm²⁹⁰ examined the notion that whereas portable XRF measurements are widely accepted for the analysis of obsidian artefacts, the technique is not considered to have the accuracy or reliability to analyse source samples, for which conventional laboratory techniques are preferred. Using Armenian obsidian sources as an example, the author compared the accuracy and reproducibility of a portable XRF instrument with five laboratory-based techniques (INAA, EDXRF, WDXRF, EPMA and LA-ICP-MS) and concluded that there was no reason to believe that the portable XRF data was inadequate and indeed the technique had the capability of analysing large numbers of specimens, demonstrating variability and elucidating field relationships. A different type of field study was undertaken using a portable XRF spectrometer by Grattan et al.,²⁹¹ whose interest was in the geoarchaeology of waste heaps from both ancient mining and the benefaction of copper-rich ores in the Wadi Khalid, southern Jordan. Two types of mining waste/spoil were identified. One resulted from the purposeful breaking and selection of copper-rich clasts as part of the benefaction of the mined ores (Khalid OP beds) and the other was the result of the widespread discarding of the more abundant final waste products (Khalid W beds).

The characterisation of building materials has increased in popularity in recent years and XRF spectrometry contributed to a study of wall plasters from the Neolithic town of Catalhoyuk in Turkey by Anderson et al.²⁹² Based mainly on mineralogical information, the authors concluded that foundation plasters originated mainly from a marl source, whereas finishing plasters contained dolomite revealing that softlime was used in its preparation, the nearest source of which was 6.5 km away. Neolithic inhabitants probably valued this finishing material because of its whiter colour. Reflecting the increasing trend for the XRF technique to be used as part of a multitechnique approach, Kamel et al.²⁹³ investigated the deterioration of stucco mihrabs (i.e., semi-circular niches in the wall of a mosque) in Islamic buildings in Egypt. Their aim was first to allow the original material to be reproduced and second to understand the cause of the deterioration, which was identified with the halite (sodium chloride) content of the stucco. Franceschi and Locardi²⁹⁴ claimed that strontium was a new marker for the use of natural gypsum in cultural heritage, especially for wall paintings. The authors observed that the determination by portable XRF instrumentation and the correlation observed between Ca and Sr concentration data, found in natural gypsum were valuable tools for conservation scientists, restorers and art historians when studying temera, frescos and Egyptian paintings. A combination of XRF analyses, petrographic characterisation and stable isotope ratios (C and O) were used by Columbu et al.²⁹⁵ to ascertain the provenance of white and greyish marbles used to decorate the Roman Heliocaminus Baths, Tivoli, Italy. Sources were identified mainly from the Apuan Alps Basin (Carrara), but also several Greek island and Turkish quarrying areas.

Conservation and indeed the prevention of deterioration are important aspects of the management of archaeological artefacts leading Krug and Hahn²⁹⁶ to demonstrate that this is not without a potential hazard to museum staff. They used portable XRF instrumentation to detect the presence of organo-chlorinated pesticides in textiles at the German Historical Museum, Berlin and found that the majority of objects also contained As, Hg and Pb, introduced as biocides or as part of the original manufacturing process; information that was essential to proper prediction of risk. Fierascu et al.²⁹⁷ were interested in the development of synthesised apatetic materials to protect artefacts of cultural or historic interest from fungal attack. The authors used EDXRF as one of the techniques to characterise several synthesised materials based on hydroxyapatite, which were tested for their effectiveness to provide protection against bio-deterioration. Fungal bio-deterioration of stained glass windows using reproductions produced by historically accurate methods was investigated by Rodrigues et al.²⁹⁸ The μ-EDXRF technique was used to check the chemical composition of 15^th and 17^th century window glass. Results after inoculating fungi onto representative glass surfaces showed that all surfaces displayed clear evidence of damage with no differences between initial corroded and non-corroded glass surfaces. Arizio and colleagues²⁹⁹ investigated the degradation of amalgam mirrors used between the 15^th and 19^th centuries, employing both ED-SEM and μ-EDXRF elemental mapping. Their interest was in the reflective surface of the tin amalgam, adhering to the glass sheet and they reported the presence of romarchite and cassiterite as degradation products growing to form hemispherical stratified calottes. In complete contrast, Fors et al.³⁰⁰ measured the Fe and S accumulations on 17^th century marine archaeological shipwrecks in the Baltic Sea using scanning XRF and S K-edge XANES. The authors found organic thiols, disulfides and elemental S produced by bacterial action but not iron sulfides due to the low Fe content of the wood. When near anaerobic conditions underwater, changed to a humid museum environment, atmospheric oxidation of iron sulfides formed in the presence of corroding iron could induce post-conservation degradation of the wood.

Concerning the analysis of human remains, Lanzirotti et al.³⁰¹ used high spatial resolution SR-based elemental XRF mapping, tomography and XAS to measure the Hg and Pb content of bone and hair samples from Ferrante II of Aragon (King of Naples, 1469–1496) and Isabella of Aragon (Duchess of Milan 1470–1524). The high sensitivity and spatial resolution of these techniques allowed the authors to report that the high levels of Hg in the Isabella hair samples was consistent with exposure during her life time, as were the high levels of Pb found in Ferrate's remains. However, the high levels of Hg found in the samples from Ferrante were considered to be the result of post mortem embalming. Goldenberg and colleagues³⁰² were interested in the detection of biological compounds preserved in ancient ceramics and proposed two simple and rapid methods based on the reaction of unsaturated compounds with iodine. The subsequent detection of I by XRF methodology was considered to be a potential pre-screening method for identifying ceramic samples suitable for organic residue analysis. The second method was based on mapping unsaturated molecules on ceramic surfaces using ED-SEM. Residues in bottle-shaped vessels of a different kind were of interest to Ay et al.³⁰³ They reported results using μ-XRF, μ-XRD and μ-Raman spectroscopy on a residue sample from an Early Bronze Age grave in Muslumantepe, south eastern Anatolia, Turkey, identifying the black residue as a manganese mineral, pyrolusite. The authors postulated that it was used for cosmetic purposes, the earliest reported use of cosmetics during this period in the Upper Tigris region.

The analysis of paints and pigments continues to be a huge application of XRF spectrometry, often using portable instrumentation in conjunction with other techniques, especially Raman and FTIR spectroscopy. Inevitably this review can only include a small selection of the available contributions. In the analysis of historic paintings, researchers at the University of Antwerp group have made some seminal contributions including this year an investigation by Monico et al.³⁰⁴ of the degradation of lead chromate in paintings by van Gogh, using the μ-XRF and μ-XANES techniques. Complications occurred from the subsequent application of varnish to conserve the paintings. Results showed the presence of Cr^III alteration products (sulfate, oxide, organo-metallic and chloride-based compounds) present inside the coating and homogeneously widespread on the painting surface. These products were considered to result from mechanical friction caused by the coating application brush redistributing Cr photo-degradation products already present on the surface of the painting. Studies of the palettes of historic artists are quite common, but Izzo et al.³⁰⁵ used the XRF technique, together with Raman, FTIR and GC-MS techniques to study 20^th century artists' oil paints used during the period 1960–1964 by Lucio Fontana (1899–1968) to demonstrate the impact of new synthetic materials. The authors reported the use of non-drying or slow-drying oils in paint formulations, which facilitated the development of appropriate conservation and restoration strategies. Moving now back to the 13^th century, Daveri et al.³⁰⁶ investigated a Medieval icon that was hidden beneath a 19^th century panel painting of the Madonna and Child attributed to an unknown Tuscan artist. Results of their work, using various non-invasive μ-techniques (including XRF spectrometry) and μ-destructive techniques (Raman and ED-SEM) revealed numerous gilding techniques, including the use of surrogate gold (disulfur tin) and orpiment as a further false gold and an area with an original silver-rich layer. In addition, the authors identified the palette of pigments including an animal anthraquinone lake, kermes, and an extremely rare chalcone pigment, possibly safflower. Sefcu et al.³⁰⁷ claimed a lateral resolution of 20 μm using a μ-XRF method to examine a 16^th century panel painting of St Anne with the Virgin Mary and Child attributed to the Master of the Litomerice Altarpiece in Bohemia, Czech Republic. As an example of the investigation of manuscripts, Bicchieri³⁰⁸ examined the invaluable purple Codex Rossanensis, one of the oldest surviving illuminated manuscripts of the New Testament using micro techniques (Raman, FTIR and XRF). The author reported what was believed to be the first use of the elderberry lake pigment in a 6^th century manuscript. Beck et al.³⁰⁹ described the first use of a portable system coupling XRD and XRF spectrometry for the in situ characterisation of pigments in prehistoric cave art at Rouffignac, France. They reported the use of two main manganese oxides, pyrolusite and romanechite. Van de Voorde and colleagues³¹⁰ also reported the use of state-of-the-art instrumentation: hand-held XRF, portable XRF/XRD and laser-based Raman instrumentation, to examine the ‘Mad Meg’ oil painting of the Flemish Renaissance artist, Pieter Bruegal the Elder. Their interest lay in whether or not this artist used a blue smalt pigment to paint the boat that appears towards the top of the painting, to confirm the hypothesis that Bruegal was economical in using this pigment for all blue areas in the painting. It is interesting to conclude this paragraph with the proposal by Aceto et al.³¹¹ that the use of UV and VIS diffuse reflectance spectrometry with optical fibres adds significant value in the preliminary examination of the main colorants used in illuminated manuscripts prior to the use of more selective methods, such as XRF or Raman spectroscopy.

Popularity in the use of XRF techniques in the characterisation of pottery and glazes almost matches the interest in paint and pigments reviewed in the previous paragraph with again a trend towards the use of μ-XRF techniques and often with accompanying XRD and Raman data. Portable XRF instrumentation offers significant flexibility in the analysis of archaeological sediments and ceramics as recognised by Hunt and Speakman,³¹² who proposed an analytical protocol and calibration strategy designed to optimise portable XRF performance, taking account of limitations of the technique in sample preparation, instrument calibration and the inability to quantify low-Z elements. Hao et al.³¹³ used the EDXRF and XANES techniques to investigate faience, a special type of brilliantly coloured glazed ceramic, from the Peng State Cemetery site, Western Zhou Dynasty (1046–771 BC), China. Results from a blue and a green faience bead showed that the colouring element in both was Cu²⁺; the variations in colour being caused by differences in the local chemical environment of this ion. Attaelmanan³¹⁴ used μ-XRF instrumentation to analyse inclusions in Mleiha ceramics (in the Emirate of Sharjah, United Arab Emirates) to furnish provenance data. The majority of inclusions were found to be silicates (with the majority of these being olivines) with 6% of them being calcite. Terra Sigillata Hispanica was an easily recognisable Roman pottery with a characteristic intense red colour. Compana et al.³¹⁵ applied WDXRF as one of a group of techniques to characterise samples of this material from known Baetica workshops in the Iberian Peninsula. The authors reported that chemical data used in conjunction with a macroscopic examination distinguished the production workshop, whereas mineralogical data showed the wide range and variation in the amorphous content and firing temperature (900–1000 °C) and finally ED-SEM results showed the coarseness of the slip starting material. Turning now to the study of glazes, Zhu et al.³¹⁶ analysed red decorated porcelain from 13^th–14^th century China to elucidate the colouration mechanism of the copper underglaze. They used SRXRF, XANES and reflection spectrometry of the red and orange regions to show that the Cu content of the former was higher but with similar valence states. It was suggested that differences in colour could be due to light scattering effects. Composition and chromaticity effects were of interest to Wu et al.³¹⁷ in relation to the purple-gold glaze of the Jingdezhen Imperial Kiln ceramics (Ming and Qing dynasties), based on an interpretation of results by EDXRF spectrometry and colourimetry. Differences in colour were reported to be caused by both variations in the Ca and Mg content (leading to the separation of iron crystals, so reducing the brightness and glossiness of the glaze), and to variations in the kaolin content of the ceramic body. Ming et al.³¹⁸ undertook an investigation of two ancient porcelains from the Hutain Kiln also in Jingdezhen City, Jiangxi Province, China using μ-EDXRF spectrometry and visible light microscopy. The porcelains in question were a bluish white glazed porcelain (Qingbai, Song Dynasties – 960–1279 AD) and an egg-white glazed porcelain (Luanbai, Yuan Dynasty – 1271–1368 AD) and results were interpreted as showing differences in the proportion of potassium-rich components, the presence of distinct micro-structures in Qingbai ware that affected its opacity and differences in the firing temperature and the time to heat and cool a sample. A hand-held XRF spectrometer and a mobile Raman instrument were used by de Voorde et al.³¹⁹ to analyse in situ a unique 16^th century majolica tile floor in Antwerp, Belgium. Results allowed the identification of the majolica elemental signature and the pigments used in the colourful glaze. In addition, the presence of Zn, an element not present in 16^th century majolica tiles, indicated the originality of parts of the floor. How do archaeological ceramics become altered after contaminated by seawater was a question answered by Montana et al.,³²⁰ who prepared ceramic test pieces designed to simulate ceramic materials found in ship wrecks in the western Mediterranean. The techniques of XRF and XRD were used to characterise these materials, which were monitored in the course of exposure to seawater; significant changes were detected in some major and trace element concentrations.

A more limited number of contributions was available during the current review period concerning the characterisation of glass. However, Zhao et al.³²¹ studied 46 stratified glass eye beads in the Hubei Provincial Museum, Wuhan, China using portable XRF and Raman spectrometers. The beads originated from the Warring States Period (about 433 BC) and were reported to be typical soda-lime-silicate glasses with a low content of K₂O and MgO and two types of colorants: antimony-based opacifiers/colorants and oxidised transition metals such as Co, Cu and Fe. The TXRF technique was used by Detcheva et al.³²² to study the colouration of medieval glass bracelets found in the necropolis of Stambolovo and the castle of Mezek in Bulgaria. The authors provided a description of their technique, which involved analysing samples as slurries in a detergent (Triton X114) and reported that colouration of the samples was caused by Co, Fe and Mn. Faience beads from the Western Zhou Dynasty (1046–771 BC) were of interest to Gu et al.,³²³ who used μ-SRXRF-CT and μ-EDXRF spectrometry to investigate manufacturing processes. The CT results showed that the beads could be classified as glazed faience and glass faience and the authors postulated that the former were formed on an organic cylinder before the direct application of glaze. Other aspects of composition led the authors to propose that (i) Western Zhou faience had an indigenous origin in China, rather than being influenced by Western developments and that (ii) this faience was not a precursor to early glass production in China.

In the analysis of metallic objects, Ferretti³²⁴ undertook a critical evaluation of the use of portable XRF instrumentation in an investigation of ancient metal artefacts, including the advantages and disadvantages of various measurement strategies. One issue of concern was the development of alternative methods of quantification to the removal of the patina, an approach that is rarely acceptable when examining ancient artefacts. Troalen et al.³²⁵ were interested in gold working practices associated with the Qurneh burial (a spectacular discovery of jewellery at a burial of a women and child discovered in Egypt by Flinders Petrie in 1908 and now resident in the National Museum of Scotland). The investigators used various techniques including XRF to demonstrate the coexistence of different levels of wear and colours of gold, the extensive use of hard soldering by the addition of Cu to gold-based alloys and the presence of PGE inclusions implying the use of alluvial gold or recycled alloys of alluvial gold. Bronzes from the necropolis of Palaepaphos (Southwest Cyprus) from the 11^th–18^th centuries BC were investigated by Charalambous et al.,³²⁶ who used portable XRF instrumentation to determine the alloy type used in their production. Results indicated the use of fresh rather than recycled Cu–Sn binary alloys, with the relative concentrations adjusted according to use to be made of the alloy. Interpretations were made of the presence and measured concentrations of Sn (higher levels result in a golden surface to the alloy), Pb (proposed to be a deliberate addition even at low concentrations), Fe and Zn (believed to be non-intentional additions to the alloy originating from copper ores used in smelting) and As (originating from Limassol Forest ores or the recycling of Early and Middle Bronze Age artefacts). As a result of the analysis of patinated metalwork from Japan, mainly using μ-XRF instrumentation, La Niece and co-authors³²⁷ suggested that it was possible to track the introduction to Japan of the liquidation process to de-silver copper by the identification of copper objects of increasing purity. In addition, their results indicated the deliberate addition of As to certain alloys to produce an enhanced patina. The work of de Figueiredo et al.³²⁸ demonstrated that sodium glycinate was more effective at removing the brownish-black tarnish on silver–copper artefacts than sodium acetylglycinate solution, with XRF and IR spectroscopies contributing to the characterisation of the products obtained. The problem that surface alteration effects the quantitative analysis by EDXRF spectrometry of artefacts such as those made of gold and silver was addressed by Cesareo.³²⁹ This arises because the penetration depth of the X-ray beam is a maximum of 10 μm. To overcome this problem, the author described an EDXRF instrument for the bulk analysis of such artefacts in which the X-ray beam from the X-ray tube was partially monochromatised using a tin secondary target, the K-lines of which bracket the silver-K absorption edge discontinuity. The sample is then positioned between this target and the detector permitting the determination of the thickness and/or composition of gold alloys (of thickness <120 μm) and silver alloys (of thickness >0.7 mm) by the differential attenuation of the Sn K-lines. The thickness of gold leaf applied to six Japanese folding screens from the Momoyama (1573–1603 AD) and Edo (1603–1868 AD) periods was reported by Pessanha et al.³³⁰ The authors made measurements in situ by EDXRF instrumentation using a methodology based on the differential attenuation of different characteristic lines of Au in the gold leaf layer. Account was taken of the fact that the gold leaf might not be pure and results showed that the gold leaf applied to one of the Endo screens was thinner, consistent with the development of beating technology in producing ever thinner gold leaf. In a second contribution, Pessanha et al.³³¹ reported a thickness of 1.6 μm for gold leaf applied to a priming layer of white lead on a Portuguese illuminated manuscript. Differential attenuation of the Pb L lines visible in the gold leaf spectra were used to make this estimate.

Platinum is of particular importance as a marker of provenance in ancient gold coins and Hinds et al.³³² investigated the determination of Pt in this matrix by the WDXRF method, reporting a detection limit of 20 μg g⁻¹. Bias caused by relief in the coin design (compared with a perfectly flat object) was largely overcome by making a 50 mm trough in the sample cup to lower the coin relative to the X-ray tube. In this way, Pt values on the side with an imprinted design were within 1–13% of those on the opposite side, which had been polished flat. The classification of analytical results is clearly an important aspect of any interpretation of coin data and Christie et al.³³³ were interested in developing chemometric methods of classifying Hungarian silver coins of the Arpad Dynasty (997–1301 AD). Their approach involved developing a ‘marker object projection-aided’ classification, claiming that this approach was capable of extracting information that was not accessible by classical pattern recognition methods.

The calibration of portable XRF instrumentation for the accurate analysis of geological archaeological artefacts was discussed by Conrey et al.³³⁴ To correct for matrix effects (absorption and secondary enhancement), the authors strongly recommended the influence coefficient approach (rather than fundamental parameters, Compton ratioing, multivariate statistics or dilution). Bias in WDXRF data was claimed to be less than 1% for obsidian and flint and less than 2% for mudbrick and sediment compared with up to 35% that the authors found using fundamental parameter or multivariate statistical methods.

3.6 Forensic

XRF measurements contributed to a number of forensic applications in the research publications available to this year's review. In an investigation of the provenance of soil collected from crime scenes in Japan, Maeda et al.³³⁵ started to construct a nationwide forensic soil sediment data base for the heavy mineral and heavy element (e.g., Cs, REE, W) compositions of stream sediments collected at 3024 locations across Japan. Samples were characterised using SRXRF and SRXRD techniques based on a high capacity facility capable of analysing 130 samples of a few mg each per day. Woods and co-workers³³⁶ compared the capabilities of LIBS, XRF and ED-SEM techniques to discriminate soil samples collected from Canberra and other parts of Australia. The authors reported that very good results were obtained by each technique and these data contributed to a protocol to differentiate soil specimens. Portable XRF instrumentation was used by Appoloni and Melquiades³³⁷ to undertake PCA of element data for the classification of money bills in counterfeiting investigations. Discrimination was based on the elements Ca, Cu, Fe, Pb, Sr, Ti, Y and Zr present in the printing pigments used. Feraru et al.³³⁸ were interested in the forensic analysis of blue ballpoint pen inks and related dyes using XRF and UV-VIS-IR spectrometry and TLC to provide data that might lead to the identification of the writing instrument used in a suspect document. Suzuki^339,340 used the techniques of FTIR, XRF and ED-SEM to determine the origin of vehicles painted with both bismuth oxychloride paint, which was used for a limited period by a motor vehicle manufacturer to provide a pearlescent lustre and bismuth vanadate, which produces a durable bright hue and was more widely used since 1985; (Raman spectroscpy also contributed to this study). On a different topic, Kinoshita et al.³⁴¹ used EDXRF spectrometry to identify high concentrations of Ti in the stomach contents of a drug overdose victim – the element apparently being added as a pharmaceutical filler to the ingested tablets. And finally in this section, Zapata et al.³⁴² undertook a review of emerging spectrometric techniques, including XRF, for the forensic analysis of body fluids prior to DNA extraction. However, the authors identified IR and Raman spectroscopies as showing the most promise in discriminating stains from bodily fluids.

3.7 Clinical

The use of nanoparticles in cancer imaging and therapy is an emerging field. In the first of two publications, Ren et al.³⁴³ successfully studied the feasibility of applying XRF spectrometry to early breast cancer diagnosis and treatment, using a novel approach for 3D-XRF microscopy of objects loaded with gold NPs in a physical phantom. The study was conducted using an X-ray pencil beam and a compact integrated X-ray spectrometer. The results showed that the Au Kα fluorescence X-rays had sufficient penetrability for this phantom study, and the reconstructed mapping data indicated that both the spatial distribution and relative concentration of gold NPs within the designed BR12 phantom could be well identified and quantified. The second contribution³⁴⁴ demonstrated the proof-of-concept of the XRF microscopy technique to identify accurately breast tumours located posteriorly, close to the chest wall musculature, a technically challenging measurement with conventional mammography. A 10 mm thickness breast tissue simulation phantom (BR12), attached to the back of a gel phantom as the region of interest (ROI), was embedded with two gold NP solutions to simulate varying gold NP uptake within breast lesions. A 2D mapping of the middle plane in the ROI demonstrated the feasibility of this approach and matched well with the known spatial distribution and different gold NP concentrations. A 3D reconstruction of the ROI was easily rendered by repeating the 2D mapping process. Ricketts et al.³⁴⁵ recognised the need to understand the mechanism of interaction between gold NP and the host tissue to maximise the potential in cancer imaging and therapy. The authors presented a novel NP uptake platform consisting of a tissue-engineered 3D in vitro cancer model (tumouroid), which mimicked the microarchitecture of a solid cancer mass and stroma. Furthermore, an XRF imaging system was developed to demonstrate 3D imaging of gold NPs and to determine uptake efficiency within the tumouroid. The authors indicated that this platform had implications for optimising the targeted delivery of NPs to cells to benefit cancer diagnostics and therapy. Liu et al.³⁴⁶ demonstrated the potential of SRXRF microscopy to advance the understanding of the penetration of targeted nanomedicines within tumour tissues. Quantitative mapping of the distribution of transferrin-conjugated gold NPs inside multicellular tumour spheroids was achieved using XRF microscopy and compared with qualitative data obtained using reflectance confocal microscopy. Gold NPs conjugated with human transferrin that had a narrow size-distribution and high binding affinity to tumour cells, were prepared as confirmed by cellular uptake studies performed on 2D monolayers. Although the prepared 100 nm transferrin-conjugated gold nanoparticles had high targeting capability to cancer cells, penetration inside multicellular spheroids was limited even after 48 hours as shown by the quantitative XRF microscopy measurements.

Dietary selenium has been implicated in the prevention of cancer and other diseases, but its safety and efficacy is dependent on the supplemented form and its metabolites. Weekley et al.³⁴⁷ continued their research on the speciation and distribution of Cu and Se in vivo, using XAS and XRF microscopy. In kidneys isolated from rats fed a diet containing 5 μg mL⁻¹ Se as selenite for 3 weeks, Se levels increased 5-fold. The XAS spectra suggested that most of the Se in the kidney was found as Se–Se species, rather than Cu-bound, and that most of the Cu was bound to N and S, presumably to amino acid residues in proteins. Hughes et al.³⁴⁸ assessed the Se status by measuring serum levels of Se using TXRF spectrometry and selenoprotein P (SePP) by immune-lumino-metric sandwich assay, and examined the association with the risk of colorectal cancer in a nested case–control design within the European Prospective Investigation of Cancer and Nutrition. The findings indicated that the Se status was sub-optimal in many Europeans and suggested an inverse association between colorectal cancer risk and higher serum Se status, which is more evident in women. It is well-known that the process of carcinogenesis might influence normal biochemical reactions leading to alterations in the elemental composition of the tissue. More and more, the TXRF technique is applied to verify the elemental content of different brain tumours, as shown by Lankosz et al.³⁴⁹ to determine Ca, Cu, Fe, K, Rb and Zn in neoplastic tissues. The results showed that the elemental composition of a relatively small fragment of tissue represented satisfactorily the biochemical “signature” of a cancer. On the basis of the element concentrations determined, it was possible to differentiate between some types of brain tumours. The Cu level in human serum of 142 chemotherapy patients and 44 healthy persons was determined by Kubala-Kukus et al.,³⁵⁰ using TXRF spectrometry. Serum Cu levels were significantly higher in the chemotherapy group than in the control group. The obtained results showed that the applied procedure could be used as a diagnostic tool for chemotherapy patients.

Neuroimaging of metals in a variety of intact brain cells and tissues is emerging as an important tool for increasing our understanding of the role of metal dysregulation in neurodegenerative diseases such as Alzheimer's. A review, presented by Braidy et al.,³⁵¹ discussed the advantages and challenges of using emerging elemental and molecular imaging techniques, and highlighted clinical achievements in Alzheimer research using bioimaging techniques, including PIXE, XRF microscopy, SRXRF, SIMS, MALDI imaging mass spectrometry and FTIR. Taken together, these techniques provided new opportunities to probe the pathobiology of Alzheimer's and paved the way for identifying new therapeutic targets. Collingwood and Davidson,³⁵² on the other hand, focussed the role of Fe in neurodegenerative disorders and considered how the field has evolved with regard to the study of Fe in proteins, cells, and brain tissue, and identified challenges in sample preparation and analysis. Selected examples were used to illustrate the contribution, and future potential, of SRXRF analysis for the characterisation of Fe in model systems exhibiting Fe dysregulation, and for human cases of neurodegenerative disorders. To understand better the link between brain venous drainage and neurodegenerative disorders, which has recently been found associated with jugular truncular venous malformations, a first attempt was conducted by Pascolo et al.³⁵³ to investigate possible morphological and biochemical anomalies at the jugular tissue level. The authors performed sequential XRF analyses on jugular tissue samples from two patients and two control subjects, using complementary energies at three different SR facilities. This investigation, coupled with conventional histological analyses, revealed anomalous micro-formations in the pathological tissues and allowed the determination of their elemental composition. In particular, rapid XRF analyses on large tissue areas excited at 12.74 keV by X-rays, showed an increased Ca presence in the pathological samples, while at lower energy the high Ca level corresponded to micro-calcifications, also containing P and Mg. Grubman et al.³⁵⁴ also underlined the strength of the current XRF detectors to measure rapidly the biometal content at subcellular resolution in cell populations using XRF microscopy. Subcellular biometal homeostasis in a cerebellar cell line isolated from a natural mouse model of a childhood neurodegenerative disorder, was investigated. Despite no global changes to whole-cell concentrations of Ca or Zn, XRF microscopy revealed significant subcellular mislocalisation of these important biological second messengers in cerebellar cells. Moreover, subtle differences in the Zn K-edge XANES spectra of control and the affected cells could be observed. The study highlighted the complementarities of bulk and single cell analysis of metal content for understanding disease states.

Bone strontium intake in osteoporotic females self-supplementing with strontium citrate was monitored by Moise et al.,³⁵⁵ who performed non-invasive bone measurements using an in vivo XRF ¹²⁵I radioisotope based system. Thirty minute measurements were taken at the finger and ankle, representing primarily cortical and trabecular bone, respectively. For analysis, the 14.2 keV Sr Kα peak normalised to the coherent peak at 35.5 keV was used. Once Sr supplements were started, a 24 h reading was taken, followed by frequent measurements ranging from weekly, biweekly to monthly. The longest volunteer participation was 1535 days. Bone Sr levels continued to increase throughout the length of the study. Furthermore, although the same research group had previously reported that finger bone Sr levels might plateau within two years, these results suggested otherwise, indicating that bone Sr levels would continue to rise at both bone sites even after 4 years of Sr intake. A feasibility study for the in vivo detection of Gd in bone was conducted by Mostafaei et al.,³⁵⁶ using a ¹⁰⁹Cd radioisotope based K-shell XRF spectrometer equipped with an array of four detectors, normally used for the non-invasive study of bone lead levels. The system was used to measure bone-simulating phantoms doped with varying levels of Gd and fixed amounts of Ca, Cl and Na. Layers of plastic were used to simulate overlying soft tissue and this permitted the prediction of a detection limit of 6.1 to 8.6 μg Gd per gram phantom for fingers with 2 to 4 mm of overlying tissue using a source of activity 0.17 GBq. With a new source of activity 5 GBq, the authors predicted that this system could achieve a detection limit of 4 to 5.6 μg Gd g⁻¹ Ca, which lies within the range of levels previously found in the bone of patients receiving Gd-based contrast imaging agents. The bone lead concentration measurement with K-shell XRF spectrometry is already well-established and regularly implemented in cumulative lead exposure studies. Ji et al.³⁵⁷ demonstrated that the job type significantly predicted cumulative lead exposure in a community-dwelling population and accounted for a large proportion of the association between educational attainment and bone lead, while the data presented by Eum et al.³⁵⁸ suggested that low-level lead exposure might contribute to menopause-related health outcomes in older women through the effects on age at menopause.

Beside occupational lead exposure, the XRF technique was also implemented to study occupational exposure of various toxic components in human tissues. The distribution of actinides in human lymph node tissues originating from individuals with documented occupational exposure was investigated by Vergucht et al.,³⁵⁹ who used μ-SRXRF analysis for the qualitative and semi-quantitative elemental mapping of these elements. Using external standards for calibration, more than 100 μg g⁻¹ of U could be measured in the detected actinide hot spots, while for the plutonium hot spots, concentration levels of up to 31 μg g⁻¹ were found. Once again, this case study on these samples illustrated the high potential of μ-SRXRF spectrometry for this type of elemental bio-imaging owing to its high sensitivity, high spatial resolution, and non-destructive character. Andujar et al.³⁶⁰ provided evidence that NPs found in welding fumes could be responsible, at least partially, for the pulmonary inflammation observed in welders. A combination of imaging and material science techniques including STEM, EDX, and μ-XRF spectrometry, was used to characterise the NP content in lung tissue from 21 welders and 21 matched control patients, demonstrating a link between human exposure to NPs and long-term pulmonary effects.

One of the reasons for total hip implant failures might be induced by metal ion-release caused by corrosion and wear of alloys. Thus, Jakobsen et al.³⁶¹ evaluated implant components from 52 revision cases with spot tests for free Cr^VI ions, Co and Ni. Implant composition was determined by XRF spectroscopy, and information on the reason for revision and complications in relation to surgery was collected from the medical charts when possible (72% of cases). The results showed that Co and Ni were released from some failed total hip arthroplasties, and corrosion was frequently observed. Calasans-Maia et al.³⁶² evaluated the effect of zinc-substituted calcium phosphate (CaP) on bone osteogenesis using a normalised ISO 10993-6 protocol. Measurements using μ-SRXRF spectrometry suggested that some of the Zn released from the biomaterial was incorporated into new bone near the implanted region. Despite the great potential of Zn-containing CaP matrices for future use in bone regeneration, the authors recognised the fact that additional in vivo studies must be conducted to explain the mobility of Zn at the CaP surface and its interactions with a biological medium. The composition of new Zr-binary alloys, useful for knee and hip replacement, and their biocompatibility, was evaluated by Totea et al.³⁶³ Three binary alloys (Zr10Nb, Zr2.5Nb and Zr12Ta) were made from high purity powders (>99.9%), using an induction furnace followed by an electric arc furnace for a perfect homogenisation. Their final composition was verified with a hand-held XRF analyzer. The hemolysis tests showed that the released hemoglobin quantity was extremely small, under 2%, in all cases, and the coagulation tests showed no risk for thrombosis. The electrochemical behaviour was also studied in biological fluid, human female serum, and showed a low corrosion rate.

The analysis of human fingernails and hair always reveals a wealth of information from their elemental composition. Although further work is needed, interesting results were collected by Ponomarenko et al.,³⁶⁴ who determined the concentration and chemical speciation of arsenic in the finger- and toenail clippings of volunteers living in the vicinity of Sackville, New Brunswick, Canada, as toxicity biomarkers of human exposure to elevated levels of As in drinking water. Arsenic species in clippings were represented by three main groups, distinguished by As–K NEXAFS spectra: (1) As^III type, which could be fitted as a mixture of As bound to thiols, and also to oxygen or methyl groups, with a small contribution from As^V species, (2) As^V type, best represented by fitting arsenate in aqueous solution and (3) As^III + As^V mixture type. The high proportion of sulfur-bound arsenic species most likely corresponded to binding between arsenic (in its trivalent and, to a lesser extent, pentavalent forms) and cysteine residues in the sulfur-rich fraction of keratin and keratin-associated proteins. A specially designed EDXRF spectrometer for in vivo scan analysis, applied to determine the concentration profile of elements present in human fingernails, was presented by Figueroa et al.³⁶⁵ In most of the samples Ca, Cu Fe, Pb, S and Zn were detected. A 2D (x, y) scan was also performed on a whole removed nail in which the 2D spatial distribution of the detected elements was observed, showing significant differences for some of the elements present. The exposure time did not exceed 15 s, and the calculated administered dose in the surface nail region was 0.1 mGy s⁻¹. Chikawa et al.³⁶⁶ examined the correlation between breast cancer and Ca levels along single hair strands by SRXRF spectrometry. In fact, patient hair has a memory of the entire cancer process from the state before cancer generation, and the pattern could be distinguished from concentration variations due to the chronic Ca deficiency without cancer, leading to a new and interesting criterion for cancer detection from the ratio of Sr/Ca. Majewska et al.³⁶⁷ applied the TXRF technique to define reference values of elements in human serum, urine and hair in the range of concentration from ng g⁻¹ to several hundred μg g⁻¹. The method of selection of the control group, the experimental setup and its calibration procedure were described, together with the sample preparation methods and the measurement procedure. Biomonitoring of Cr, Cu, Fe, Ni, Pb and Zn in scalp hair of the rural population in Punjab state of India (age group 30–45 years) was carried out by Yadav et al.³⁶⁸ using the TXRF technique, to understand their exposure to the trace elements from the local environment. The validity of the method was checked by analysing the CRMs IAEA-436 and INCT-MPH-2. The results of the hair samples showed no environmental gradient in the concentration of elements such as Cr, Cu, Fe, Ni and Zn, however Pb was observed to be higher in a few volunteers as compared to published values of concentrations in volunteers from Delhi.

In a preliminary study, the anatomical variations in the trace element content of primary teeth with known differences in Pb content were studied by μ-SRXRF spectrometry. De Souza-Guerra et al.³⁶⁹ determined variations in the spatial distribution of Ca, Cu, K, Mn, Pb, Sr and Zn in four anatomical locations: superficial enamel (SE, 0–10 μm), subsuperficial enamel (10–30 μm), primary dentin, and secondary dentin. Two teeth had low concentrations of Pb in the SE (<250 μg g⁻¹), while three contained very high Pb concentrations in the SE (>2000 μg g⁻¹). Teeth were sliced, and five spot measurements (20 μm beam diameter) were accomplished at each location. The obtained results highlighted the importance of the first 10 μm of the SE for the determination of some elements, such as Zn, Pb, Mn, and Cu. In two similar studies by the same authors,^370,371 a portable μ-EDXRF system, making use of a polycapillary optic to obtain a focal spot of 100 μm for Fe Kα, was used to assess whether the elemental content of Ca, P and Zn in tooth enamel would alter when treating the teeth with a bleaching gel. To obtain 8 × 2 mm samples, cuts of sound vestibular surfaces of anterior healthy teeth were made, and treated with a bleaching product accordingly to the manufacturer's instructions and stored in artificial saliva between each application. Before and after treatment, μ-EDXRF measurements were performed and the quantitative data, obtained by the WinAXIL compare mode method using four reference materials as calibrants, showed no significant statistical differences for the studied elements. Similar to knee and hip implants, corrosion of metals in dental implants might occur and is an issue of concern. Abraham et al.³⁷² presented an indirect study of corrosion of dental implants by analysing changes in the elemental concentration of metals such as Al, Ti and V in oral fluids (saliva and gingival fluids) by means of SR-TXRF spectrometry. While the Ti-ions reflected a different behaviour in the oral fluids, revealing higher concentrations in gingival crevice fluid than in saliva, no significant differences for Al and V were observed from normal levels in the oral fluids. To get more insight into the relationship between the mineralisation of peritubular dentin (PTD) and intertubular dentin (ITD), Deymier-Black et al.³⁷³ examined what organic or nanostructural signatures might indicate the end of ITD or the beginning of PTD mineralisation using advanced characterisation techniques. The XRF intensity (Ca, P, and Zn) and XRD patterns from carbonated apatite were mapped around dentintubules at resolutions ten times smaller than the feature size (200 nm pixels), representing a 36% increase in spatial resolution over earlier work. In the near-tubule volumes of near-pulp, root dentin, the Zn intensity was higher than in ITD remote from the tubules. This increase in Zn²⁺, as determined by XANES analysis, was thought to indicate the presence of metalloenzymes or transcription factors important to ITD or PTD mineralisation. The XRD maps indicated a lack of structural difference between in the Zn-rich near-tubular region and the remote ITD.

Other interesting applications in this section are brought to your attention. Yoshii et al.³⁷⁴ proposed a new methodology for the on-site and rapid evaluation of heavy-atom contamination in wounds using a portable XRF device, which might be applicable for workers decommissioning the Fukushima-Daiichi nuclear power plant damaged from the Great East Japan Earthquake and resulting tsunami. First, the wound model was developed by placing a liquid blood phantom on an epoxy resin wound phantom contaminated with Pb, which was used as the substitute model contaminant in place of radioactive heavy atoms. Next, the correlation between the concentration of contaminant and the XRF peak intensity was formulated taken account of the thickness of blood exiting the wound. At a maximum equivalent dose of 16.5 mSv to the wound and blood thickness of 0.5 mm, the method detection limit value for Pb was 1.2 μg g⁻¹ (3.1 nmol). Moreover, this method might be applicable for evaluation of Pu contamination in wounds. Centeno et al.³⁷⁵ also focussed their research on the chemical analysis of retained embedded fragments in war wounds of US military personal and their potential health impact. The physical and compositional properties of the foreign material surfaces removed from 4 patients with penetrating blast wounds to the limbs were examined by EDXRF, SEM-EDS, LA-ICP-MS and confocal laser μ-Raman spectrometry. Quantitative chemical analysis of both fragments and available tissues was conducted employing ICP-MS. The majority of the fragments were primarily composed of a single metal such as Al, Cu or Fe with traces of Pb, Sb, Ti and U. One case demonstrated W in both the fragment and the connected tissue, together with Pb. Capsular tissue and fragments from a case from the 1991 Kuwait conflict showed evidence of U, which was further characterised by U isotopic ratios analysis to contain depleted uranium. Lis et al.³⁷⁶ studied the influence of lipids on the progression of calcification by determining the distribution of selected elements in three distinct regions of calcified human aortic valves, representing successive stages of the calcific degeneration: normal, thickened (early lesion) and calcified (late lesion), using μ-SRXRF spectrometry for elemental composition and Oil Red O staining for the demonstration of lipids. In calcified valve areas, the accumulation of Ca and P was accompanied by enhanced concentrations of Sr and Zn. Calcifications preferentially developed in lipid-rich areas of the valves. Calcium concentration ratio between lipid-rich and lipid-free areas was not age-dependent in early lesions, but showed a significant increase with age in late lesions, indicating age-dependent intensification of lipid involvement in the calcification process. The results suggested that mechanisms of calcification changed with the progression of valve degeneration and with age. Schmeling et al.³⁷⁷ conducted TXRF measurements on the lens and aqueous humor of 14 cataract patients, to study the presence and concentration of selected metals in the eye. Most commonly found elements in both types of media were Cr and Mn, whereas Ba was found in the lens, but not in aqueous tissue, and Ni only in the aqueous humor. Concentrations were generally higher in aqueous samples. These findings demonstrated that further research was required to elucidate more accurately the relationship between systemic and ocular metal accumulation and the impact of metal accumulation on measures of visual function and ocular disease. Finally, Morrison et al.³⁷⁸ evaluated a class of mitochondrially-targeted Gd^III agents, designed for potential application in binary cancer therapies. Mitochondria uptake was determined by ICP-MS, while Gd uptake was determined by SRXRF imaging. This study reported the highest in vitro tumour selectivity for any Gd agent reported to date, with a tumour/normal cell ratio of up to 23.5 ± 6.6, and also highlighted the delicate balance required to minimise in vitro cytotoxicity and optimise in vitro tumour selectivity and mitochondrial localisation for this new class of mitochondrially-targeted binary therapy agents.

3.8 Drugs

Only a limited number of contributions in this area were published in this review period. Sogut et al.³⁷⁹ analysed several kinds of cigarette tobaccos to determine contents of essential elements, natural and artificial radioactivity. Samples were analysed by the WDXRF technique for the following elements: Al, As, Br, Ca, Cl, Cu, Fe, In, K, Mn, Ni, P, S, Sr, Ti and Zn, showing the highest contents in the cigarette tobaccos for K and Ca with a mean value of 48.9% m m⁻¹ and 38.4% m m⁻¹, respectively. Fernandes et al.³⁸⁰ determined trace mineral nutrients such as Cu, Fe, and Zn in eight different infant milk powders, representing the majority of commercial milk powder brands available in Portuguese pharmacies. The WDXRF results, collected in helium flush mode, suggested that the concentrations of some of the elements determined were not consistent with the values labeled by manufacturers. However, all fell inside the concentration ranges defined by Portuguese and European legislation, except for Al with an average concentration ranging from 8.25 up to 19.75 mg kg⁻¹. Skin penetration of silica NPs, currently used in pharmaceutical and cosmetic products, was a topic of interest. Iannuccelli et al.³⁸¹ aimed at elucidating the in vivo mechanism by which bare and lipid-coated silica NP enter human stratum corneum through the evaluation of the role played by the NP surface polarity and the human hair follicle density. Two types of silica sample, bare hydrophilic silica and hydrophobic lipid-coated silica, were applied on both the volar and dorsal side of volunteer forearms. Twelve repetitive stripped tapes were removed from the human skin and evaluated for their elemental composition by EDXRF spectrometry and for their Si content by ICP-MS. Finally, Pascolo et al.³⁸² successfully evaluated the applicability of μ-SR-FTIR and μ-SRXRF spectroscopies in the study of the in vitro interaction of poly-lactide-co-glycolide (PLGA)-based NPs with cells. Those PLGA NPs, sized around 200 nm, were loaded with superparamagnetic iron oxide NPs (PLGA-IO-NP, size, 10–15 nm). After exposing human mesothelial cells to PLGA-IO-NPs (0.1 mg mL⁻¹) and fixation, the μ-SR-FTIR measurement enabled the detection of PLGA NPs at single-cell level, allowing polymer detection inside the biological matrix by the characteristic band in the 1700 to 2000 cm⁻¹ region. Moreover, the SRXRF microscopy measurement performed on the same cells provided evidence for the presence of Fe in the cells and so emphasised the intracellular localisation of the PLGA-IO-NPs.

3.9 Biological

Element distribution maps of Ca, K, P and S in fungal samples were created by Rak et al.,³⁸³in situ, using the μ-SRXRF instrument at the ESRF beam line ID21, Grenoble, France, to study developing hyphae, septa, and conidiophores in the fungus Aspergillus nidulans, comparing the wild type with two cell wall biosynthesis gene deletion strains. Thus, S was examined as a proxy for the protein content, P for the nucleic acid content, as well as Ca and K, which also have important metabolic roles. The findings demonstrated that deleting a dispensable gene involved in galactose metabolism and one involved in the biosynthesis of a minor cell wall component led to changes in hyphal elemental distribution that might have resulted from compromised wall composition. Using the μ-SRXRF and μ-XANES set-up, Gonzalez-Chavez et al.³⁸⁴ aimed to identify the in situ localisation and speciation of As in the fungus Rhizophagus intraradices exposed to As^V. The findings suggested a dual protective mechanism by the plant to rapidly exclude As from the fungus and prevent As translocation to the plant root. The μ-XANES data showed that the gradual reduction of As^V took place to predominantly an As^III species complexed with a reduced iron^II carbonate compound. These results validated the hypotheses that R. intraradices directly participates in arsenic detoxification.

An interesting contribution, presented by Jurowski et al.,³⁸⁵ emphasised many different and complex aspects of analytical calibration problems in studies related to (bio)imaging/mapping of metallic elements in different kinds of biological samples using various analytical techniques including XRF. This state-of-the-art paper included as well as nomenclature, definitions, as examples of calibration strategies with analytical calibration procedures. Previous studies of the Se status indicated that Se was necessary for fertility but how precisely, was not known. Ceko et al.³⁸⁶ aimed to show that Se is important in the bovine female reproductive function. The elemental distribution in bovine ovary (n = 45 sections) was identified by XRF imaging. The first XRF imaging of mammalian ovaries showed that Se was consistently localised to the granulosa cells of large healthy follicles. The findings suggested that Se and selenoproteins were elevated in these large healthy follicles and might play a critical role as an antioxidant during late follicular development. The use of the aquatic plant Egeria densa was successfully evaluated as a potential Cr^VI biomonitor in aquatic environments by Hayashi,³⁸⁷ who performed in the laboratory in vivo time-resolved Cr and Ca XRF mapping over a period of 69 days on E. densa, immersed in 5 mM K₂CrO₄ aqueous solution. The temporal and spatial resolution of each XRF map were similar to 1.6 days and 1 × 1 mm², respectively. The obtained XRF maps exhibited characteristic localised Cr and Ca areas and indicated that once the Cr species were introduced, they substituted the original Ca-accumulated regions. The sizes and intensity distributions of the Cr hot spots were sensitive to the Cr^VI exposure approximately 1 week prior to each XRF measurement.

The distribution and concentration of trace elements within pathological and normal bone of both extant and extinct archosaurs (Cathartes aura and Allosaurus fragilis) was determined by Anne et al.,³⁸⁸ who recognised the strength of the rapid scanning-SRXRF technique, which had the advantage of combining high sensitivity (μg g⁻¹), excellent spatial resolution (20 to100 μm) and scanning of large specimens (dm scale) approximately 3000 times faster than other mapping techniques. Results revealed discrete chemical inventories within different bone tissue types and preservation modes. Chemical inventories also revealed the detail of histological features not observable in thin section, including fine structures within the interface between pathological and normal bone as well as a woven texture within pathological tissue. Kerschnitzki et al.³⁸⁹ investigated the microscopic and nanoscopic alteration of avian medullary bone architecture during the egg-laying cycle of hens, using a combination of SR X-ray tomography together with small angle X-ray scattering, wide angle XRD and XRF spectrometry. During eggshell calcification, the mineral content and the size of the trabeculae of medullary bone decreased markedly. Furthermore, the average mineral particle size increased during resorption, suggesting that the smaller mineral particles were preferentially removed. Medullary bone thus formed a fast-responding system exhibiting rapid alterations of the material at the micro- and nanoscale. Insights to these mechanisms are interesting in order to provide hens with Ca for their high metabolic calcium demand during eggshell mineralisation. To conclude, Feng et al.³⁹⁰ carried out an analytical comparison between XRF-CT and K-edge CT to quantify the distribution of gold NPs in a small animal for preclinical studies. They reported a theoretical analysis in terms of S/N for the two modalities, showing that XRF-CT had a better S/N than K-edge CT when gold NPs concentrations fell below a certain threshold value. Numerical tests were performed for XRF-CT and K-edge CT on two kinds of phantoms with multiple concentration levels and structural features. Experimental results illustrated that XRF-CT was superior to K-edge CT when contrast concentration was lower than 0.4%, which coincided with the theoretical analysis. Readers interested in additional applications can consult our companion review on clinical and biological materials, food and beverages.³

3.10 Thin films, coatings and nanomaterials

The trend over the last year of a significant number of papers dealing with the characterisation of thin films in the nanometre range continued within the current review period and is a particularly hot topic for the semiconductor industry. Muller et al.³⁹¹ characterised high-k nanolayers using GI-XRF spectrometry. The authors claimed that reference-free GI-XRF spectrometry based on SR could significantly contribute to the characterisation of such nanolayered systems. The reliable modelling of the X-ray Standing Wave (XSW) field in conjunction with the radiometrically calibrated instrumentation at the PTB (Berlin, Germany) allowed reference-free, FP-based quantitative analysis. Interestingly the authors did not use fitting procedures as a majority of other groups working in this field do. The capabilities of this method were demonstrated for two systems of transistor gate stacks, i.e., Al₂O₃ high-k layers grown on silicon or silicon/SiO₂ and Sc₂O₃ layers on InGaAs/InP substrates. Yakunin et al.³⁹² performed a XSW analysis of periodic multilayer structures. This model independent approach, based on the direct solution of the system of linear equations that characterises the XRF yield, allowed a fast reconstruction without fitting procedure of the atomic distribution function inside a sample. The new approach was applied to the reconstruction of the atomic distribution function for LaN/BN multilayers with 50 periods of 43 Å thick layers. The authors emphasised that the object was difficult to analyse with traditional methods as the estimated thickness of the interface region between the constituent materials was comparable to the individual layer thicknesses. Using the new approach, it was possible to show that the La atoms stay localised within the lanthanum nitrate layers and interfaces and do not diffuse into the boron nitrate layer. A review of an X-ray investigation of monolayers formed at the soft air/water interface (Gibbs or Langmuir monolayers) was provided by Stefaniu and Brezesinski,³⁹³ which highlighted recent achievements in the monolayer field with special emphasis on different Sr-based X-ray methods such as: GI-XRD, XRR and TXRF. Some examples of single-chain surfactants, special sugar lipids, and semi-fluorinated compounds were given and thin layers formed by peptides, polymers or NPs were highlighted. Lesnik et al.³⁹⁴ characterised AlInN/AlN/GaN FET structures using XRD, XRR and GI-XRF techniques. The authors explained that the AlN-interlayer (spacer) thickness played an important role to enhance the mobility and the density of the 2D electron gas. A layer analysis was performed for the determination of the aluminum nitrate interlayer thickness. A sample series of AlInN/AlN/GaN FETs on Si(111) was grown and analysed. The growth time of the aluminum nitrate-interlayer was varied up to 12 s and the AlInN barrier was grown nearly lattice-matched to GaN with a nominal thickness of 5 nm. By the combination of HR-XRD, XRR, GI-XRF techniques and simultaneous simulation of the data, the spacer thickness was determined successfully.

Several papers described the use of advanced SR-based X-ray analytical techniques with resolution in the nanometre range for material characterisation. In a very exciting application Laforce et al.³⁹⁵ published nano-XRF imaging of meteoritic particles and diamond inclusions. In this paper, the new ESRF ID16B-NA Nanoanalysis beamline was applied for the first time for XRF imaging with a resolution level down to a few tens of nanometres on rare geological materials: meteoritic fragments from achondrite NWA 6693 and diamond inclusions. The instrument proved to be an extremely valuable tool for mapping samples containing submicrometre heterogeneities. It was discovered that the track of bubble-like inclusions in NWA 6693 consisted mainly of Cr-rich phases, although some inclusions containing Ni and Ca were also detected. In diamond SL05, originating from the Juina region in Brazil, multiple inclusions with dimensions of about 0.5 μm² were analysed, detecting As, Ca, Fe, Ti with a step width of 40 nm and with a measuring time of 1 s per step. Bare et al.³⁹⁶ characterised a fluid catalytic cracking equilibrium catalyst (E-CAT) at the ensemble and individual particle level. At the ensemble level, the authors used μ-XRF-CT to determine the average size, shape, and respective distributions of over 1200 individual catalyst particles. This information was claimed to be important in order to determine performance in commercial operations. It could be shown that a major fraction of the particles contained large internal voids (5–80 μm diameter), which supported the accessibility to large hydrocarbon molecules. Using nano XRF-CT at the individual particle level, these voids were visualised at a much smaller scale (approximately 100 nm to 12 μm in diameter) and the individual phases present in the particle were visualised in 3D. Micro-XRF was used to map both the contaminant (Fe, Ni, V) and the inherent lanthanum catalyst elemental distributions. The distribution of zeolite Y in the E-CAT particle was inferred from the La XRF map. The authors concluded that this in-depth characterisation study at the ensemble and individual E-CAT particle level presented a robust methodology that provided an understanding of the E-CAT at both the micro- and nanometre scales. Segura-Ruiz et al.³⁹⁷ investigated phase separation in single In_xGa_1−xN nanowires revealed through a hard X-ray SR Nanoprobe. The authors reported the composition, short- and long-range structural order of single molecular beam epitaxy-grown In_xGa_1−xN nanowires. Nano-XRF mapping revealed an axial and radial heterogeneous elemental distribution in the single wires with Ga accumulation at their bottom and outer regions. Polarisation-dependent nano-XANES demonstrated that the tetrahedral order around the Ga atoms remained along the nanowires. The In-content along the single nanowires estimated from XRF spectrometry were in good agreement with XRD data. The authors anticipated that this methodology would contribute to a greater understanding of the underlying growth concepts not only of nanowires but also of many nanostructures in materials science. Ishiguro et al.³⁹⁸ visualised the heterogeneity of Ce oxidation states in single Pt/Ce₂Zr₂O_x catalyst particles by nano-XAFS using an X-ray nanobeam. Differences in the distribution of the Ce oxidation states between Pt/Ce₂Zr₂O_x single particles of different oxygen compositions (x) were visualised in the obtained 2D XRF mapping images and the Ce L3-edge nano-XANES spectra. Cristofolini³⁹⁹ reviewed SR X-ray techniques for the investigation of structures and dynamics in interfacial systems with the main emphasis on flat interfaces and on colloidal systems. The techniques for structural determinations were XRR, GI-XRD and GI-XRF. The systems reviewed were, in order of growing complexity, floating Langmuir monolayers, supported films of lipids and proteins, polymeric films, buried interfaces, colloidal systems and gels formed by colloids either in 3D or in the form of 2D interfacial layers. Recent results were critically discussed, and some interesting directions of development were outlined, having also in mind new technical developments such as X-ray-free-electron laser sources and μ-focused SR beamlines. Hu et al.⁴⁰⁰ combined 3D structure and chemistry imaging with nanoscale spatial resolution. A data fusion technique was presented that combined nano-SRXRF-CT and nano-SRXRF spectrometry to non-destructively investigate complex nanoscale materials and provide combined 3D renderings of microstructure and chemistry. The technique was named nano tomography-assisted chemical correlation (nTACCo) and its effectiveness was demonstrated on fly ash particles with nanoscale chemical inhomogeneities. The authors claimed that nTACCo was capable of providing the concentration and location of seven different nano-inclusions within a particle as well as direct observations of reactivity and chemical distribution on fly ash particles. They further anticipated that the ability to combine 3D structure and chemistry at the nanoscale will provide unprecedented tools for nanoscience in material science, biology, chemistry and medical science. Kuhn et al.⁴⁰¹ applied site-selective high spatial resolution XAS and YES to cobalt NPs. The authors pointed out that the special (macroscopic) properties of NPs were mainly due to their large surface-to-volume ratio and concluded that the separate characterisation of geometric and electronic properties of surface and bulk would be favourable, for a better understanding of the properties of NPs. Three types of wet-chemically synthesised cobalt NPs, of about 6 nm diameter with varying thicknesses of a protective shell, were investigated at the ID26 beamline of the ESRF. High resolution fluorescence detected (HRFD)-XAS spectra at the Co K-edge were recorded via detection of the K_β1,3 XRF at specific energies. As these spectra were only partly site-selective due to a strong overlap of the emission lines, a numerical procedure was applied based on a least-squares fitting procedure. To obtain additional information about ligands attached to Co, valence-to-core X-ray emission spectra (VTC-XES) using the K_{β 2,5} and the satellite structure K_β and VTC-XANES spectra were also recorded, by which the former results were confirmed. The authors concluded that VTC-XES and HRFD-XAS were suitable tools for the detailed specification of the core and the surface of NPs, in particular upon realising “real” site-selectivity for XANES and EXAFS with a general strategy applicable to a wide range of systems.

Several papers were published within the review period dealing with nanoparticles, nanowires and nanorods, emphasizing the growing importance of that topic. Cobalt-implanted ZnO nanowires were investigated using hard X-ray nanoprobe-based techniques by Chu et al.,⁴⁰² who reported the short and long range order of single room temperature-implanted and high temperature-implanted Co : ZnO nanowires. The XRF maps revealed a homogeneous distribution of the Co ions implanted along the nanowires, while the XANES data indicated a substitutional incorporation of Co²⁺ into the wurtzite ZnO host lattice. The room temperature-implanted nanowires presented a higher structural disorder around Co atoms compared with the high temperature-implanted ones. Johannes et al.⁴⁰³ reported enhanced sputtering and incorporation of Mn in implanted GaAs and ZnO nanowires. The authors simulated and investigated experimentally the sputter yield of ZnO and GaAs nanowires, which were implanted with energetic Mn ions at room temperature. The resulting thinning of the nanowires as well as the dopant concentration with increasing Mn ion fluence were measured by SEM and nano-XRF spectrometry, respectively. A clearly enhanced sputter yield for the irradiated nanowires compared with bulk was observed, which was also corroborated by simulations showing a maximum if the ion range matched the nanowire diameter. As a consequence of the thinning of the nanowire by erosion, the incorporation of the Mn dopants was also enhanced and increased non-linearly with increasing ion fluence. Martinez-Criado et al.⁴⁰⁴ studied the local structure of a crossed Ga₂O₃/SnO₂ multiwire architecture with nanometre resolution. The authors proposed that crossed nanowire structures were the basis for high-density integration for a variety of nanodevices and that due to the critical role of nanowire intersections in creating hybrid architectures, it had become a challenge to investigate the local structure in crossing points in metal oxide nanowires. By combining electron and SR nanoprobes, the authors could show experimental evidence of the role of impurities in the coupling formation, structural modifications and atomic site configuration. Filez et al.⁴⁰⁵ used XAS and XRF techniques to study in situ NP nucleation during Pt deposition. A nucleation delay indicative for NP formation was observed for Pt loadings below one equivalent monolayer. Both XAS and XRF spectra were recorded simultaneously at different catalyst loadings in this nucleation regime. Analysis of the combined in situ data yielded a quantitative picture of the evolution of the diameter, shape, lattice packing and density of the deposited platinum clusters. By linking the XRF and XAS data, a linear increase in cluster diameter with platinum loading was observed. Hertz et al.⁴⁰⁶ used laboratory XRF-CT for high spatial resolution NP bio-imaging. The authors demonstrated that NP XRF-CT in mouse-sized objects could be performed with very high spatial resolution at acceptable dose and exposure times with a compact laboratory system. The method relied on the combination of a high-brightness indium anode liquid-metal-jet X-ray source (emitting a 24 keV line), pencil-beam-forming X-ray optics, an ED detector and molybdenum NPs. Experiments on phantoms and simulations showed that the arrangement allowed 100 μm detail imaging at dose and exposure times compatible with small-animal experiments. The authors proposed that the method provided a possible path to in vivo molecular X-ray imaging at sub-100 μm resolution. The shape of CdSe nanorods was examined by Meyns et al.⁴⁰⁷ revealing hexagonal pyramids in a hot-injection synthesis under the influence of halogenated additives in the form of organic chlorine, bromine and iodine compounds. Synchrotron XPS measurements and TXRF results showed that the shape was accompanied by a modification in the chemical composition of the ligand sphere. The authors claimed that their results presented a new degree of freedom in NP shape control and highlighted the role of additives in NP synthesis and their possible in situ formation of ligands. Fernandez-Ruiz et al.⁴⁰⁸ applied TXRF spectrometry to the evaluation of bioaccumulation kinetics of gold nanorods (NR) in vital mammalian organs. The authors emphasised that their work presented the first application of TXRF spectrometry to the evaluation of the bioaccumulation kinetics of gold NRs in various tissues upon intravenous administration in mice. The detection limit was found to be 112 ng g⁻¹. The gold NR bioaccumulation kinetics were analysed in several vital mammalian organs such as liver, spleen, brain and lung at different times. Additionally, urine samples were analysed to study the kinetics of elimination of the gold NRs by this excretion route. The authors reported that gold NRs were quickly bioaccumulated by highly vascular filtration organs such as liver and spleen, whereas gold NRs did not show a bioaccumulation in brain and lung for the period of time investigated. In addition, urine also showed no gold NR accumulation. The authors concluded that TXRF spectrometry proved to be a powerful, versatile, and precise analytical technique for the evaluation of gold NR content in biological systems and, in a more general way, for any kind of metallic NP. Rashkow et al.⁴⁰⁹ compared SRXRF spectrometry with ICP-MS for quantification of NPs and distribution in single-cell populations. The SRXRF microscopy technique was employed to quantify and characterise the distribution of titanium dioxide nanosphere uptake in a population of single cells. These results were compared with the average NP concentrations per cell obtained by ICP-MS. The results showed that NP concentrations per cell quantified by SRXRF spectrometry were one to two orders of magnitude greater compared with ICP-MS. The authors drew attention to the limitation of methods available to determine physical parameters from large population averages leading to potentially misleading information and the lack of any information about the cellular heterogeneity. Choi et al.⁴¹⁰ estimated cellular SiO₂ NPs using flow cytometry combined with XRF measurements. In order to improve the understanding of the quantitative aspects of cell–NP interactions, flow cytometry, TEM and XRF experiments were carefully performed for the HeLa cells exposed to silica NPs with different core diameters, hydrodynamic sizes, and surface charges. A semi-quantitative method was proposed and applied for the determination of cellular silica NPs within their size-dependent linear ranges.

In addition to NPs, nanomaterials such as graphene were the topic of several publications. Using μ-AS and XRF mapping, Bozzini et al.⁴¹¹ investigated the electrodeposition of Co/CoO NPs onto graphene for oxygen reduction reaction (ORR) electrocatalysis. The authors combined spectroscopic methods with electrochemical measurements, allowing a molecular-level understanding of potentiostatic and pulsed-potential plating processes from the organic solvent onto a freestanding graphene film. The compositional and chemical-state distribution of Co were investigated ex situ by soft XAS and XRF mapping, showing that both spatial distribution and valence state were homogeneous and independent of the local current density. The pulsed deposit consisted of Co/CoO NPs with diameter of ca. 30 nm and the potentiostatic deposit of NPs with 200 nm diameter, with the latter offering better ORR performance. Nanomaterials have been successfully used for preconcentration by Kocot and Sitko.⁴¹² They used graphene nanosheets in dispersive micro solid-phase extraction (DMSPE) as a preconcentration method for trace and ultratrace determination of heavy metal ions by EDXRF spectrometry. The adsorptive properties of graphene nanosheets were used for simultaneous preconcentration of Co, Cu, Ni and Pb ions from water samples. Only 200 μL of suspension containing graphene, ammonium pyrrolidine dithiocarbamate (APDC) and Triton-X-100 was rapidly injected into a water sample. Then, graphene nanosheets with adsorbed metal–APDC chelates were collected on a membrane filter and measured by EDXRF spectrometry. Various parameters including pH, amount of APDC, sample volume, amount of Triton-X-100 and sorption time were optimised in order to obtain the best recoveries. The experiment showed that Co, Cu, Ni and Pb could be simultaneously preconcentrated at pH of 5 with recoveries of >96% and with very good precision (RSD about 3%). Due to the excellent enrichment factors ranging from 400 to 2500 the proposed DMSPE-EDXRF procedure offered low detection limits in the range of 0.08 ng mL⁻¹.

Though a confocal μ-XRF instrument is not commercially available, several publications concluded that confocal XRF was a suitable method to characterise layered material. Cordes et al.⁴¹³ reported the use of laboratory-based X-ray techniques for the non-destructive elemental quantification of polymer-embedded thin films. The authors explained that characterising highly uniform coatings for their thickness, elemental composition and uniformity could be difficult when the layers were subsurface and need to be interrogated non-destructively. Confocal μ-XRF and nano-CT were successfully coupled for these measurements in order to meet these sensitivity and spatial resolution specifications necessary for characterising thin films. Elemental composition, atomic percent, placement and uniformity could be measured in 3D. As the spatial resolution of confocal XRF is limited by the use of the second polycapillary lens between sample and detector, which is in the μm range, nano-CT had to be used to image the layers at very high spatial resolution (down to 50 nm) to measure precisely the embedded layer thickness. These two techniques must be used together if both the thickness and atomic density of a layer are unknown. The authors demonstrated the possibility to measure both the atomic percent of an embedded thin film layer and confirm its manufacturing quality. As a proof of principle, a 1.5 atomic percent, 2 μm thick germanium layer embedded within polymer capsules, used for laser plasma experiments at the Omega Laser Facility and National Ignition Facility, were successfully measured. Sun et al.⁴¹⁴ applied confocal μ-XRF spectrometry to the investigation of paint layers. Multilayered paint fragments of a car were analysed non-destructively to demonstrate that this confocal μ-XRF instrument could be used in the discrimination of the various layers in multilayer paint systems. Wrobel et al.⁴¹⁵ performed depth profiling of element concentrations in stratified materials by confocal μ-XRF spectrometry with polychromatic excitation. The authors extended a method, well established for monochromatic radiation, to polychromatic radiation by effective energy approximation. The reduction of the whole incident spectrum energy distribution into one effective value eliminated the necessity for integration over the primary beam energy range for a number of basic parameters. This simplification was claimed to be attainable without loss of accuracy in the analysis, which seems to be a surprise. The proposed approach was validated by applying it to the reconstruction of element concentration depth profiles of stratified standard samples measured with a table-top confocal μ-XRF setup. By comparing the obtained results from two independent algorithms, showing good agreement. Peng et al.⁴¹⁶ reported the in situ element-resolved determination of the uniformity of thickness in a multilayer film by confocal μ-XRF spectrometry. The relative error of measuring an iron film with a thickness of 16.3 μm and a copper film with a thickness of 24.5 μm were 7.3% and 0.4%, respectively.

Thin films as sample supports were used by Kanrar et al.⁴¹⁷ to improve detection limits in EDXRF spectrometry. The authors compared the results obtained with thin specimens deposited on thin transparent adhesive tape supports with those obtained by TXRF spectrometry. The detection limits was improved to 1050 pg for Cr and 320 pg for Y, compared with TXRF data of 320 pg and 168 pg, respectively.

3.11 Chemical state and speciation

Within the review period, a critical review from Grafe et al.⁴¹⁸ was published concerning the speciation of metal(loid)s in environmental samples by X-ray absorption spectroscopy. The authors emphasised that XAS was a widespread and accepted research tool in the environmental and geo-sciences due to its element specificity. They further commented that XAS provided information about speciation of metal(loid)s in complex environmental matrices independent of physical state, at environmentally relevant interfaces (e.g. solid–liquid) as well as in different media such as plant tissues, rhizosphere, soils, sediments, ores, mineral process tailings, etc. The advantages of using XAS for environmental investigations were stated to be: limited sample preparation requirements, ability to preserve original physical, chemical states and the independence from crystallinity. Through advances in optics, detectors, and data processing, XRF microprobes capable of focusing X-rays to μm or nm size have become competitive research tools for resolving the complexity of environmental samples. The authors' ability demonstrated that μ-XANES imaging allowed lateral resolution of chemical states over wide areas due to improved data processing and detector technology. Liu et al.⁴¹⁹ published a study of Zn relative concentration and chemical state in broiler duodena by μ-XRF spectrometry and μ-XAS. The objective was to determine the change in Zn quantity, distribution and chemical state (Zn²⁺, organic zinc) in intestinal wall using different zinc sources. Day old male broilers were randomly allotted to three perfusion groups (ZnMet, ZnLys, ZnSO₄). Micro-XRF spectrometry and μ-XAFS were used to analyse the relative quantity, distribution and chemical state of Zn in the intestinal wall, while flame AAS was used to verify the Zn concentration in the whole intestinal sac. The results showed that the ZnMet and ZnLys group fed samples had a greater amount of Zn in the intestinal wall than in the ZnSO₄ group. The Zn chemical state of organic Zn and ZnSO4 group were identical in intestinal wall specific region. In addition, the Zn concentration achieved using ZnMet was greater than that for ZnLys and ZnSO₄ fed samples. Further, it was found that organic zinc was more easily absorbed than inorganic Zn. Solid phase speciation of Cu, Ni and Zn, which plays an important role in the long-term stability of metals in a biosolid-amended soil, was reported by Mamindy-Pajany et al.⁴²⁰ using μ-XRF spectrometry and μ-XANES. Comparison of metal absorption edges on the biosolid-amended soil and the control soil sample showed that Cu, Ni, and Zn could be retained by both soil and biosolid components such as amorphous iron phases, organic matter and clay minerals. Linear combination fitting of K-edge XANES spectra of metal hot-spots indicated consistent differences in metal speciation between metals. The authors found that organic matter played a dominant role in Ni binding in the biosolid-amended soil, but was of lesser importance for Cu and Zn. The authors concluded that even if the metals could be associated with soil components (clay minerals and organic matter), biosolid application would increase metals retention in the biosolid-amended soil by providing reactive organic matter and iron oxide fractions. Among the studied metals, the long-term mobility of Ni could be affected by organic matter degradation while Cu and Zn were strongly associated with iron oxides, which were less mobile.

Abe et al.⁴²¹ reported the detection of U and chemical state analysis of individual radioactive microparticles emitted from the Fukushima Daiichi Nuclear Power Plant (FDNPP) accident using multiple SR X-ray analysis techniques. Micro-SRXRF spectrometry revealed the detailed chemical nature of radioactive aerosol micro-particles, resulting in better understanding of what occurred in the plant during the early stages of the accident. Three spherical particles containing radioactive Cs were found in aerosol samples collected in Tsukuba, 172 km southwest of the FDNPP. The following elements were detected in all three particles: Ba. Cs, Fe, Mo, Rb, Sb, Sn, Te, Zn, Zr, and in addition, U was found for the first time in two of the particles. Uranium L-edge XANES further confirmed that U fuel and its fission products were contained in these particles along with radioactive Cs. The authors claimed these results strongly suggested that the FDNPP was damaged sufficiently to emit uranium fuel and fission products outside the containment vessel as aerosol particles. Micro-XANES spectra of Fe, Mo, Sn and Zn K edges for the individual particles revealed that they were present in their higher oxidation states, i.e., Fe³⁺, Zn²⁺, Mo⁶⁺, and Sn⁴⁺ in the glass matrix, confirmed by μ-SRXRD measurements. The authors concluded that these radioactive materials in a glassy state might remain in the environment longer than those emitted as water-soluble radioactive Cs aerosol particles. Sun et al.⁴²² used SRXRF spectrometry and XANES to study the distribution and chemical speciation of Pb during corn seed germination concluding that these techniques were especially suitable for in situ non-destructive trace element analysis of biological samples. Thus XANES was used to analyse the chemical speciation of Pb in different parts of corn seed to understand the absorption and transformation mechanism of Pb by corn seeds. The results showed that the germination rate, bud length and root length decreased with the increasing contents of Pb. In addition, the inhibition effect on root growth was greater. The results of SRXRF analysis showed that Pb was mainly enriched in episperm and embryo, which would inhibit embryonic development into bud and root. Moreover, the results showed that all samples including root, shoot and the different parts of the seed had the same Pb speciation. The linear combination fitting results indicated that the lead phosphate chloride and lead stearate accounted for 74.3% and 24.2% respectively, suggesting that the main Pb speciation in corn was lead phosphate chloride deposited inside the corn. Polysulfide speciation and reactivity in chromate-contaminated soil was investigated by Chrysochoou and Johnston.⁴²³ Calcium polysulfide (CPS) was observed to maintain a reducing capacity for prolonged time periods when used to treat Cr^VI-contaminated soils. The study utilised bulk and μ-XANES to investigate S speciation in soil samples treated with CPS and to determine the source of the reducing potential. Bulk XANES spectra indicated the presence of two dominant S species: elemental S, which was the product of the sulfide-chromate redox reaction, and thiosulfate. Micro-XANES analysis confirmed these findings and showed that elemental S precipitated as large particles, while thiosulfate was diffused within the soil grains and thus available to react with chromate that leached from slowly dissolving PbCrO₄. Micro-XRF analyses indicated a close association of Pb and thiosulfate, so that PbS₂O₃ was a sink for thiosulfate, accounting for up to 20% of the total S added. The authors hypothesised that ferrous iron production was an additional mechanism to maintain reductive conditions in CPS-treated soils. Price et al.⁴²⁴ reported chemical imaging of single catalyst particles with scanning μ-XANES-CT and μ-XRF-CT. The physicochemical state of a catalyst was supposed to be a key factor in determining both activity and selectivity; however these materials were often not structurally or compositionally homogeneous. The authors presented 3D imaging of an industrial catalyst, molybdenum-promoted colloidal platinum supported on carbon. The distribution of both the active Pt species and Mo promoter were mapped over a single particle of catalyst using μ-XRF-CT. The use of XANES and EXAFS revealed a mixed local coordination environment, including the presence of both metallic Pt clusters and Pt chloride species, but also no direct interaction between the catalyst and Mo promoter. The authors also emphasised the benefits of scanning μ-XANES-CT for chemical imaging, allowing for 2D and 3D mapping of the local electronic and geometric environment, in this instance for both the Pt catalyst and Mo promoter throughout the catalyst particle.

The measurement of shifts in K X-ray peak energies of potassium and calcium in different compounds using EDXRF spectrometry was reported by Kaur and Mittal.⁴²⁵ The set-up consisted of a low-power X-ray tube photon source and Si(PIN) X-ray detector. A statistical procedure had been applied to determine the shifts and a t-test to find the statistical significance of the results. The shifts were determined in several potassium and calcium compounds using KNO₃ and CaO as reference and correlated with differences in electro-negativity, number of ligands, structural changes, type of bonding, axial distances, etc. in the compounds. The authors claimed that the shifts in potassium compounds have been evaluated for the first time while the same for calcium compounds almost agreed with literature data, which is surprisingly good.

Electrodeposition was highlighted by Li et al.⁴²⁶ as a simple and efficient preconcentration method for the determination of traces of Se^IV and Te ^IV. The presence of Pb^II improved the efficiency and reproducibility of the deposition of Se^IV on a polished substrate of high purity copper. The detection limits for the measurement of Se^IV and Te ^IV were 0.44 and 0.57 μg L⁻¹, respectively. The authors claimed that the method was validated by analysing CRMs and applied successfully to the determination of trace Se and Te in environmental water samples, but they did not report on recovery and selectivity as well as the species of the CRM.

4 Abbreviations

2D/3D	Two dimensional/three dimensional
AAS	Atomic absorption spectrometry
AD	Anno domini
ADR	Adiabatic demagnetisation refrigerator
AFS	Atomic fluorescence spectrometry
APDC	Ammonium pyrrolidine dithiocarbamate
APS	Advanced photon source
ASIC	Application-specific integrated circuit
ASU	Atomic Spectrometry Update
BAM	Federal Institute for Materials Research and Testing
BC	Before Christ
BDE	Brominated diphenyl ether
BFR	Brominated flame retardants
BP	Before present
CAP	Coated agglomerate pelletisation
CCD	Charge coupled detector
CdTe	Cadmium telluride
CERN	European Organisation for Nuclear Research
CPS	Calcium polysulfide
CRM	Certified reference material
CSA	Charge sharing addition
CSD	Charge sharing discrimination
CT	Computed tomography
CV-AAS	Cold vapour atomic absorption spectrometry
CZT	Cadmium zinc telluride
DCC	Doubly curved crystal
DEPFET	Depleted P-channel field effect transistor
DMSPE	Dispersive micro solid-phase extraction
DNA	Deoxyribonucleic acid
DSC	Differential scanning calorimetry
E-CAT	Equilibrium catalyst
ED	Energy dispersive
EDX	Energy dispersive X-ray analysis
ED-SEM	Energy dispersive scanning electron microscopy
EDXRD	Energy dispersive X-ray diffraction
EDXRF	Energy dispersive X-ray fluorescence
EPMA	Electron probe microanalysis
ESRF	European Synchrotron Radiation Facility
EU	European Union
EXAFS	Extended X-ray absorption fine structure
FAAS	Flame atomic absorption spectrometry
FDNPP	Fukushima Daiichi Nuclear Power Plant
FET	Field-effect-transistor
FP	Fundamental parameter
FTIR	Fourier transform infrared
FWHM	Full width at half maximum
GA	Genetic algorithm
GC-MS	Gas chromatography-mass spectrometry
GE-XRF	Grazing exit X-ray fluorescence
GI	Grazing incidence
GI-SAXS	Grazing incidence small angle X-ray scattering
GI-XRD	Grazing incidence X-ray diffraction
GI-XRF	Grazing incidence X-ray fluorescence
GPS	Global positioning system
HLB	Huanglongbing
HPGe	High purity germanium
HRFD-XAS	High-resolution fluorescence-detected X-ray absorption spectroscopy
HR-XRD	High resolution X-ray diffraction
HX	Hydrogen halide
ICP	Inductively coupled plasma
ICP-OES	Inductively coupled plasma optical emission spectrometry
ICP-MS	Inductively coupled plasma mass spectrometry
ICP-QMS	Inductively coupled plasma quadrupole mass spectrometry
ICT	Information and communications technology
IEC	International Electrotechnical Commission
INAA	Instrumental neutron activation analysis
IR	Infrared
ISN	In situ nanoprobe
ISO	International Organisation for Standardisation
ITD	Intertubular dentin
LA-ICP-MS	Laser ablation inductively coupled plasma mass spectrometry
LIBS	Laser induced breakdown spectroscopy
LOD	Limit of detection
LOQ	Limit of quantification
LPG	Liquefied petroleum gas
MALDI	Matrix-assisted laser desorption ionisation
MC	Monte Carlo
MIXS	Mercury imaging X-ray spectrometer
NAA	Neutron activation analysis
NEXAFS	Near edge X-ray absorption fine structure
NIST	National Institute of Standards and Technology
NP	Nanoparticle
NR	Nanorod
nTACCo	Nano tomography-assisted chemical correlation
ORR	Oxygen reduction reaction
PC	Principal component
PCA	Principal component analysis
PCR	Principal component regression
PCR-WDXRF	Principal component regression wavelength dispersive X-ray fluorescence
PET	Pentaerythritol
PGE	Platinum group element
PIXE	Particle-induced X-ray emission
PLGA	Poly-lactide-co-glycolide
PLGA-IO-NP	Poly-lactide-co-glycolide iron oxide nanoparticle
PLS	Partial least squares
PTB	Physikalisch-technische Bundesanstalt
PTD	Peritubular dentin
QC	Quality control
R&D	Research and development
ROI	Region of interest
RE	Rare earth
REE	Rare earth elements
RM	Reference material
RSD	Relative standard deviation
SAXS	Small angle X-ray scattering
SCD	Swept charge device
SDD	Silicon drift detector
SE	Superficial enamel
SEM	Scanning electron microscopy
SEM-EDS	Scanning electron microscopy-energy dispersive X-ray spectrometry
SIMS	Secondary ion mass spectrometry
Si(PIN)	Silicon PIN detector device
S/B	Signal-to-background ratio
S/N	Signal-to-noise ratio
SOFC	Solid oxide fuel cells
SRM	Standard reference material
SR	Synchrotron radiation
SR-FTIR	Synchrotron radiation fourier transform infrared
SR-TXRF	Synchrotron radiation-total reflection X-ray fluorescence
SRXRD	Synchrotron radiation X-ray diffraction
SRXRF	Synchrotron radiation X-ray fluorescence
SRXRF-CT	Synchrotron radiation X-ray fluorescence computed tomography
SSRL	Stanford Synchrotron Radiation Lightsource
STEM	Scanning transmission electron microscope
STJ	Superconducting tunnel junction
SVM	Support vector machine
SW	Southwest
TC	Technical committee
TEM	Transmission electron microscope
TG	Thermogravimetric
TGA	Thermogravimetric analysis
TLC	Thin layer chromatography
TMF	Tailing Management Facility
TOF-SIMS	Time-of-flight secondary ion mass spectrometry
TU	Technical University
TXRF	Total reflection X-ray fluorescence
UK	United Kingdom
US	United States
USA	United States of America
UTW	Ultra thin polymer window
UV	Ultraviolet
UV-VIS-IR	Ultraviolet-visible-infrared spectroscopy
VIS	Visible spectroscopy
VTC-XANES	Valence-to-core X-ray absorption near edge structure
VTC-XES	Valence-to-core X-ray emission spectroscopy
WD	Wavelength dispersive
WDXRF	Wavelength dispersive X-ray fluorescence
WEEE	Waste electrical and electronic equipment
XAFS	X-ray absorption fine structure
XANES	X-ray absorption near edge structure
XANES-CT	X-ray absorption near edge structure computed tomography
XAS	X-ray absorption spectroscopy
XES	X-ray emission spectroscopy
XPS	X-ray photoelectron spectroscopy
XRD	X-ray diffraction
XRF	X-ray fluorescence
XRF-CT	X-ray fluorescence computed tomography
XRR	X-ray reflectometry
XSW	X-ray standing waves
YSZ	Yttria-stabilised zirconia
Z	Atomic number

References

1. M. West, A. T. Ellis, P. J. Potts, C. Streli, C. Vanhoof and P. Wobrauschek, J. Anal. At. Spectrom., 2014, 29(9), 1516–1563.
2. E. H. Evans, J. Pisonero, C. M. M. Smith and R. N. Taylor, J. Anal. At. Spectrom., 2014, 29(5), 773–794.
3. A. Taylor, M. P. Day, S. Hill, J. Marshall, M. Patriarca and M. White, J. Anal. At. Spectrom., 2014, 29(3), 386–426.
4. R. Clough, C. F. Harrington, S. J. Hill, Y. Madrid and J. F. Tyson, J. Anal. At. Spectrom., 2014, 29(7), 1158–1196.
5. B. Gibson, S. Carter, A. S. Fisher, S. Lancaster, J. Marshall and I. Whiteside, J. Anal. At. Spectrom., 2014, 29(11), 1969–2021.
6. O. T. Butler, W. R. L. Cairns, J. M. Cook and C. M. Davidson, J. Anal. At. Spectrom., 2015, 30(1), 21–63.
7. F. Adams, J. Anal. At. Spectrom., 2014, 29(7), 1197–1205.
8. M. Mantler, Powder Diffr., 2014, 29(2), 127–132.
9. N. L. Misra, Spectrochim. Acta, Part B, 2014, 101, 134–139.
10. Z. Y. Zhang, X. He, Y. H. Ma, Y. Y. Ding and Z. F. Chai, Radiochim. Acta, 2014, 102(1), 3–11.
11. M. Wiedenbeck, L. P. Bedard, R. Bugoi, M. Horan, K. Linge, S. Merchel, L. F. G. Morales, D. Savard, A. K. Souders and P. Sylvester, Geostand. Geoanal. Res., 2014, 38(4), 467–512.
12. J. Anzelmo, M. Bouchard and M. E. Provencher, Spectroscopy, 2014, 29(7), 12.
13. G. J. Simandl, R. S. Stone, S. Paradis, R. Fajber, H. M. Reid and K. Grattan, Miner. Deposita, 2014, 49(8), 999–1012.
14. D. C. Arne, R. A. Mackie and S. A. Jones, Geochem.: Explor., Environ., Anal., 2014, 14(3), 233–244.
15. P. S. Ross, A. Bourke and B. Fresia, Geochem.: Explor., Environ., Anal., 2014, 14(2), 171–185.
16. P. S. Ross, A. Bourke and B. Fresia, Geochem.: Explor., Environ., Anal., 2014, 14(2), 187–196.
17. Z. X. Yuan, Q. M. Cheng, Q. L. Xia, L. Q. Yao, Z. J. Chen, R. G. Zuo and D. Y. Xu, Geochem.: Explor., Environ., Anal., 2014, 14(3), 265–276.
18. R. G. Figueroa, E. Lozano, F. Belmar, D. Alcaman, A. von Bohlen, C. A. B. Oliveira, A. L. M. Silva and J. Veloso, X-Ray Spectrom., 2014, 43(2), 126–130.
19. M. Guerra, S. Longelin, S. Pessanha, M. Manso and M. L. Carvalho, Rev. Sci. Instrum., 2014, 85(6), 9.
20. D. Guimaraes, T. M. Cleaver, S. F. Martin and P. J. Parsons, Anal. Methods, 2015, 7(1), 366–374.
21. P. S. Athiray, S. Narendranath, P. Sreekumar and M. Grande, Planet. Space Sci., 2014, 104, 279–287.
22. S. Narendranath, P. Sreekumar, L. Alha, K. Sankarasubramanian, J. Huovelin and P. S. Athiray, Sol. Phys., 2014, 289(5), 1585–1595.
23. J. L. Campbell, P. L. King, L. Burkemper, J. A. Berger, R. Gellert, N. I. Boyd, G. M. Perrett, I. Pradler, L. Thompson, K. S. Edgett and R. A. Yingst, Nucl. Instrum. Methods Phys. Res., Sect. B, 2014, 323, 49–58.
24. J. A. Berger, P. L. King, R. Gellert, J. L. Campbell, N. I. Boyd, I. Pradler, G. M. Perrett, K. S. Edgett, S. J. V. VanBommel, M. E. Schmidt and R. E. H. Lee, J. Geophys. Res.: Planets, 2014, 119(5), 1046–1060.
25. S. Z. Weider, L. R. Nittler, R. D. Starr, T. J. McCoy and S. C. Solomon, Icarus, 2014, 235, 170–186.
26. F. Gazquez, F. Rull, J. M. Calaforra, G. Venegas, J. A. Manrique, A. Sanz, J. Medina, A. Catala-Espi, A. Sansano, R. Navarro, P. Forti, J. de Waele and J. Martinez-Frias, Estud. Geol., 2014, 70(2), 16.
27. G. J. Flynn, S. R. Sutton, B. Lai, S. Wirick, C. Allen, D. Anderson, A. Ansari, S. Bajt, R. K. Bastien, N. Bassim, H. A. Bechtel, J. Borg, F. E. Brenker, J. Bridges, D. E. Brownlee, M. Burchell, M. Burghammer, A. L. Butterworth, H. Changela, P. Cloetens, A. M. Davis, R. Doll, C. Floss, D. Frank, Z. Gainsforth, E. Grun, P. R. Heck, J. K. Hillier, P. Hoppe, B. Hudson, J. Huth, B. Hvide, A. Kearsley, A. J. King, J. Leitner, L. Lemelle, H. Leroux, A. Leonard, R. Lettieri, W. Marchant, L. R. Nittler, R. Ogliore, W. J. Ong, F. Postberg, M. C. Price, S. A. Sandford, J. A. S. Tresseras, S. Schmitz, T. Schoonjans, G. Silversmit, A. Simionovici, V. A. Sole, R. Srama, F. J. Stadermann, T. Stephan, V. Sterken, J. Stodolna, R. M. Stroud, M. Trieloff, P. Tsou, A. Tsuchiyama, T. Tyliszczak, B. Vekemans, L. Vincze, J. von Korff, A. J. Westphal, N. Wordsworth, D. Zevin and M. E. Zolensky, Meteorit. Planet. Sci., 2014, 49(9), 1626–1644.
28. Y. S. Choi, J. Hwang and J. Y. Kim, Asian J. Chem., 2014, 26(13), 4067–4069.
29. P. Escarate, R. Hein, M. Duran and P. Ramaciotti, Miner. Eng., 2015, 71, 13–15.
30. M. Gaft, L. Nagli, Y. Groisman and A. Barishnikov, Appl. Spectrosc., 2014, 68(9), 1004–1015.
31. Y. Kayser, W. Blachucki, J. C. Dousse, J. Hoszowska, M. Neff and V. Romano, Rev. Sci. Instrum., 2014, 85(4), 9.
32. F. P. Romano, C. Caliri, L. Cosentino, S. Gammino, L. Giuntini, D. Mascali, L. Neri, L. Pappalardo, F. Rizzo and F. Taccetti, Anal. Chem., 2014, 86(21), 10892–10899.
33. C. K. Egan, S. D. M. Jacques, M. D. Wilson, M. C. Veale, P. Seller, P. J. Withers and R. J. Cernik, J. Appl. Crystallogr., 2015, 48, 269–272.
34. J. Garrevoet, B. Vekemans, P. Tack, B. de Samber, S. Schmitz, F. E. Brenker, G. Falkenberg and L. Vincze, Anal. Chem., 2014, 86(23), 11826–11832.
35. I. Mantouvalou, T. Wolff, C. Seim, V. Stoytschew, W. Malzer and B. Kanngiesser, Anal. Chem., 2014, 86(19), 9774–9780.
36. M. Czyzycki, P. Wrobel and M. Lankosz, Spectrochim. Acta, Part B, 2014, 97, 99–104.
37. D. W. Lane, A. Nyombi and J. Shackel, J. Appl. Crystallogr., 2014, 47, 488–494.
38. K. G. McIntosh, J. A. Neal, P. Nath and G. J. Havrilla, X-Ray Spectrom., 2014, 43(6), 332–337.
39. K. Ploykrachang, J. Hasegawa, K. Kondo, H. Fukuda and Y. Oguri, Nucl. Instrum. Methods Phys. Res., Sect. B, 2014, 331, 261–265.
40. J. Hardy and B. Scruggs, Spectroscopy, 2014, 6–7.
41. J. Hardy and B. Scruggs, Spectroscopy, 2013, 8–9.
42. M. Ahmad, M. Bazalova, L. Z. Xiang and L. Xing, IEEE Trans. Med. Imag., 2014, 33(5), 1119–1128.
43. M. Sjolin and M. Danielsson, Phys. Med. Biol., 2014, 59(21), 6507–6520.
44. N. Manohar, B. L. Jones and S. H. Cho, Med. Phys., 2014, 41(10), 10.
45. M. Bazalova-Carter, M. Ahmad, T. Matsuura, S. Takao, Y. Matsuo, R. Fahrig, H. Shirato, K. Umegaki and L. Xing, Med. Phys., 2015, 42(2), 900–907.
46. A. Amirkhanov, B. Frohler, J. Kastner, E. Groller and C. Heinzl, Comput. Graph. Forum, 2014, 33(3), 91–100.
47. B. R. Halls, T. R. Meyer and A. L. Kastengren, Opt. Express, 2015, 23(2), 1730–1739.
48. C. K. Egan, M. D. Wilson, M. C. Veale, P. Seller, S. D. M. Jacques and R. J. Cernik, Nucl. Instrum. Methods Phys. Res., Sect. B, 2014, 324, 25–28.
49. J. Maser, B. Lai, T. Buonassisi, Z. H. Cai, S. Chen, L. Finney, S. C. Gleber, C. Jacobsen, C. Preissner, C. Roehrig, V. Rose, D. M. Shu, D. Vine and S. Vogt, Metall. Mater. Trans. A, 2014, 45(1), 85–97.
50. J. Manickaraj, A. Gorny, Z. H. Cai and S. Shankar, Appl. Phys. Lett., 2014, 104(7), 4.
51. D. L. Zhang, C. X. Qiu, F. Han, B. Chen, P. R. Hua, D. Y. Yu and E. Y. B. Pun, IEEE Photonics Technol. Lett., 2014, 26(5), 524–527.
52. I. Llorens, P. L. Solari, B. Sitaud, R. Bes, S. Cammelli, H. Hermange, G. Othmane, S. Safi, P. Moisy, S. Wahu, C. Bresson, M. L. Schlegel, D. Menut, J. L. Bechade, P. Martin, J. L. Hazemann, O. Proux and C. Den Auwer, Radiochim. Acta, 2014, 102(11), 957–972.
53. P. Fontaine, G. Ciatto, N. Aubert and M. Goldmann, Sci. Adv. Mater., 2014, 6(11), 2312–2316.
54. S. R. Barberie, C. R. Iceman, C. F. Cahill and T. M. Cahill, Anal. Chem., 2014, 86(16), 8253–8260.
55. C. M. Sosa, H. J. Sanchez, C. A. Perez and R. D. Perez, Nucl. Instrum. Methods Phys. Res., Sect. B, 2014, 319, 171–176.
56. W. M. Harris, J. J. Lombardo, G. J. Nelson, B. Lai, S. Wang, J. Vila-Comamala, M. F. Liu, M. L. Liu and W. K. S. Chiu, Sci. Rep., 2014, 4, 7.
57. A. Kastengren and C. F. Powell, Exp. Fluids, 2014, 55(3), 15.
58. S. W. Wang, J. Ward, S. Leyffer, S. M. Wild, C. Jacobsen and S. Vogt, J. Synchrotron Radiat., 2014, 21, 568–579.
59. T. Kashiwabara, R. Toda, K. Fujinaga, T. Honma, Y. Takahashi and Y. Kato, Chem. Lett., 2014, 43(2), 199–200.
60. N. L. Misra, M. K. Tiwari, B. G. Vats, S. S. Kumar, A. K. Singh, G. S. Lodha, S. K. Deb, P. D. Gupta and S. K. Aggarwal, X-Ray Spectrom., 2014, 43(3), 152–156.
61. M. Moini, C. M. Rollman and L. Bertrand, Anal. Chem., 2014, 86(19), 9417–9422.
62. M. D. de Jonge, C. G. Ryan and C. J. Jacobsen, J. Synchrotron Radiat., 2014, 21, 1031–1047.
63. R. Klockenkamper and A. von Bohlen, Total Reflection X-ray Fluorescence Analysis and Related Techniques, John Wiley & Sons, Inc., 2015, ISBN 978-1-118-46027-6.
64. U. E. A. Fittschen, M. Menzel, O. Scharf, M. Radtke, U. Reinholz, G. Buzanich, V. M. Lopez, K. McIntosh, C. Streli and G. J. Havrilla, Spectrochim. Acta, Part B, 2014, 99, 179–184.
65. N. L. Misra, B. Kanrar, S. K. Aggarwal, P. Wobrauschek, M. Rauwolf and C. Streli, Spectrochim. Acta, Part B, 2014, 99, 129–132.
66. V. Romero, I. Costas-Mora, I. Lavilla and C. Bendicho, J. Anal. At. Spectrom., 2014, 29(4), 696–706.
67. V. Romero, I. Costas-Mora, I. Lavilla and C. Bendicho, Nanoscale, 2015, 7(5), 1994–2002.
68. D. Hampai, S. B. Dabagov, C. Polese, A. Liedl and G. Cappuccio, Spectrochim. Acta, Part B, 2014, 101, 114–117.
69. A. C. M. da Costa, U. B. de Araujo, E. F. O. de Jesus, M. J. Anjos and R. T. Lopes, Anal. Sci., 2014, 30(10), 955–960.
70. A. Maderitsch, S. Smolek, P. Wobrauschek, C. Streli and P. Takman, Spectrochim. Acta, Part B, 2014, 99, 67–69.
71. T. Ashida and K. Tsuji, Spectrochim. Acta, Part B, 2014, 101, 200–203.
72. S. Dhara, N. L. Misra, S. K. Aggarwal, D. Ingerle, P. Wobrauschek and C. Streli, X-Ray Spectrom., 2014, 43(2), 108–111.
73. M. Rauwolf, C. Vanhoof, K. Tirez, E. Maes, D. Ingerle, P. Wobrauschek and C. Streli, Spectrochim. Acta, Part B, 2014, 101, 118–122.
74. T. Vander Hoogerstraete, S. Jamar, S. Wellens and K. Binnemans, Anal. Chem., 2014, 86(3), 1391–1394.
75. T. Vander Hoogerstraete, S. Jamar, S. Wellens and K. Binnemans, Anal. Chem., 2014, 86(8), 3931–3938.
76. R. Klockenkamper and A. von Bohlen, Spectrochim. Acta, Part B, 2014, 99, 133–137.
77. S. H. Nowak, D. Banas, W. Blchucki, W. Cao, J. C. Dousse, P. Honicke, J. Hoszowska, L. Jablonski, Y. Kayser, A. Kubala-Kukus, M. Pajek, F. Reinhardt, A. V. Savu and J. Szlachetko, Spectrochim. Acta, Part B, 2014, 98, 65–75.
78. M. Brucher, A. von Bohlen, M. Becker and R. Hergenroder, X-Ray Spectrom., 2014, 43(5), 269–277.
79. D. Ingerle, F. Meirer, G. Pepponi, E. Demenev, D. Giubertoni, P. Wobrauschek and C. Streli, Spectrochim. Acta, Part B, 2014, 99, 121–128.
80. D. Ingerle, M. Schiebl, C. Streli and P. Wobrauschek, Rev. Sci. Instrum., 2014, 85(8), 5.
81. I. Ritter, G. Anton, R. B. Sune, F. Bisello, M. Campbell, T. Gabor, X. L. Cudie, S. Wolfel, W. S. Wong and T. Michel, J. Instrum., 2014, 9, 8.
82. I. Johnson, A. Bergamaschi, H. Billich, S. Cartier, R. Dinapoli, D. Greiffenberg, M. Guizar-Sicairos, B. Henrich, J. Jungmann, D. Mezza, A. Mozzanica, B. Schmitt, X. Shi and G. Tinti, J. Instrum., 2014, 9, 9.
83. P. Majewski, F. Aschauer, S. Aschauer, A. Bahr, B. Bergbauer, M. Hilchenbach, M. Krumrey, C. Laubis, T. Lauf, P. Lechner, G. Lutz, F. Scholze, H. Soltau, A. Stefanescu, L. Struder and J. Treis, Exp. Astron., 2014, 37(3), 525–538.
84. T. Lauf and R. Andritschke, Nucl. Instrum. Methods Phys. Res., Sect. A, 2014, 762, 142–148.
85. J. Wernecke, C. Gollwitzer, P. Muller and M. Krumrey, J. Synchrotron Radiat., 2014, 21, 529–536.
86. A. K. Rumaiz, T. Krings, D. P. Siddons, A. J. Kuczewski, D. Protic, C. Ross, G. de Geronimo and Z. Zhong, IEEE Trans. Nucl. Sci., 2014, 61(6), 3721–3726.
87. M. C. Veale, S. J. Bell, D. D. Duarte, A. Schneider, P. Seller, M. D. Wilson and K. Iniewski, Nucl. Instrum. Methods Phys. Res., Sect. A, 2014, 767, 218–226.
88. M. C. Veale, S. J. Bell, D. D. Duarte, M. J. French, A. Schneider, P. Seller, M. D. Wilson, A. D. Lozinskaya, V. A. Novikov, O. P. Tolbanov, A. Tyazhev and A. N. Zarubin, Nucl. Instrum. Methods Phys. Res., Sect. A, 2014, 752, 6–14.
89. P. H. Smith, J. P. D. Gow, N. J. Murray, J. H. Tutt, M. R. Soman and A. D. Holland, J. Instrum., 2014, 9, 14.
90. A. Rachevski, G. Zampa, N. Zampa, R. Campana, Y. Evangelista, G. Giacomini, A. Picciotto, P. Bellutti, M. Feroci, C. Labanti, C. Piemonte and A. Vacchi, J. Instrum., 2014, 9, 18.
91. S. Cornaby, Spectroscopy, 2014, 9.
92. Z. Li, X. G. Tuo, R. Shi and J. B. Zhou, Nucl. Sci. Tech., 2013, 24(6), 7.
93. S. J. Mulware, J. D. Baxley, B. Rout and T. Reinert, Nucl. Instrum. Methods Phys. Res., Sect. B, 2014, 332, 95–98.
94. H. J. Ding, H. M. Cho, W. C. Barber, J. S. Iwanczyk and S. Molloi, Med. Phys., 2014, 41(12), 11.
95. H. M. Cho, H. Ding, B. P. Ziemer and S. Molloi, Phys. Med. Biol., 2014, 59(23), 7211–7227.
96. G. Q. Zeng, C. J. Tan, Q. Luo, L. Q. Ge, C. Li, C. H. Gong and X. Y. Liu, Nucl. Sci. Tech., 2014, 25(3), 8.
97. M. H. Carpenter, S. Friedrich, J. A. Hall, J. Harris and R. Cantor, J. Low Temp. Phys., 2014, 176(3–4), 222–227.
98. S. Shiki, M. Ukibe, N. Matsubayashi, N. Zen, M. Koike, Y. Kitajima and M. Ohkubo, J. Low Temp. Phys., 2014, 176(3–4), 604–609.
99. A. Hossain, A. Dowdy, A. E. Bolotnikov, G. S. Camarda, Y. Cui, U. N. Roy, R. Tappero, X. Tong, G. Yang and R. B. James, J. Electron. Mater., 2014, 43(8), 2941–2946.
100. G. V. Pavlinsky and A. Y. Portnoy, X-Ray Spectrom., 2014, 43(2), 118–121.
101. V. D. Hodoroaba and V. Rackwitz, Anal. Chem., 2014, 86(14), 6858–6864.
102. V. Delgado, J. Math. Chem., 2014, 52(4), 1129–1136.
103. V. Rackwitz, A. Warrikhoff and V. D. Hodoroaba, J. Anal. At. Spectrom., 2014, 29(3), 458–463.
104. E. Esbelin, X-Ray Spectrom., 2014, 43(4), 198–208.
105. M. A. Scapin, C. P. Silva, M. E. B. Cotrim and M. A. F. Pires, J. Radioanal. Nucl. Chem., 2014, 302(3), 1077–1085.
106. J. B. Ghasemi, M. K. Rofouei and N. Amiri, X-Ray Spectrom., 2014, 43(3), 131–137.
107. S. Akbulut, J. Anal. At. Spectrom., 2014, 29(5), 853–860.
108. A. Brunetti and B. Golosio, Spectrochim. Acta, Part B, 2014, 94–95, 58–62.
109. E. C. Geil and R. E. Thorne, J. Synchrotron Radiat., 2014, 21, 1358–1363.
110. C. Huber, S. Smolek and C. Streli, X-Ray Spectrom., 2014, 43(3), 175–179.
111. B. Golosio, T. Schoonjans, A. Brunetti, P. Oliva and G. L. Masala, Comput. Phys. Commun., 2014, 185(3), 1044–1052.
112. J. Tickner, Comput. Phys. Commun., 2014, 185(6), 1628–1638.
113. C. C. Conti, M. J. Anjos and C. M. Salgado, Nucl. Instrum. Methods Phys. Res., Sect. B, 2014, 335, 61–65.
114. H. H. Fayez and A. L. Ghorabie, J. X-Ray Sci. Technol., 2015, 23(1), 57–64.
115. H. J. Sanchez and J. J. Leani, Radiat. Phys. Chem., 2015, 107, 160–163.
116. L. T. Yi, Z. G. Liu, K. Wang, M. Chen, S. Q. Peng, W. G. Zhao, J. L. He and G. C. Zhao, Nucl. Instrum. Methods Phys. Res., Sect. A, 2015, 775, 12–14.
117. G. Q. Zeng, Y. Y. Luo, L. Q. Ge, Q. X. Zhang, Y. Gu and F. Cheng, Spectrosc. Spectral Anal., 2014, 34(2), 562–564.
118. V. Y. Borkhodoev, J. Anal. Chem., 2014, 69(11), 1041–1046.
119. G. J. Simandl, S. Paradis, R. S. Stone, R. Fajber, R. D. Kressall, K. Grattan, J. Crozier and L. J. Simandl, Geochem.: Explor., Environ., Anal., 2014, 14(3), 211–221.
120. G. J. Simandl, R. Fajber and S. Paradis, Geochem.: Explor., Environ., Anal., 2014, 14(2), 161–169.
121. M. F. Gazley, C. M. Tutt, L. I. Brisbout, L. A. Fisher and G. Duclaux, Geochem.: Explor., Environ., Anal., 2014, 14(3), 223–231.
122. T. Quiniou and V. Laperche, Geochem.: Explor., Environ., Anal., 2014, 14(3), 245–255.
123. M. le Vaillant, S. J. Barnes, L. Fisher, M. L. Fiorentini and S. Caruso, Geochem.: Explor., Environ., Anal., 2014, 14(3), 199–209.
124. D. C. Weindorf, N. Bakr, Y. Zhu, A. McWhirt, C. L. Ping, G. Michaelson, C. Nelson, K. Shook and S. Nuss, Pedosphere, 2014, 24(1), 1–12.
125. E. L. Shuttleworth, M. G. Evans, S. M. Hutchinson and J. J. Rothwell, Water, Air, Soil Pollut., 2014, 225(2), 13.
126. B. Lemiere, V. Laperche, L. Haouche and P. Auger, Geochem.: Explor., Environ., Anal., 2014, 14(3), 257–264.
127. N. W. Brand and C. J. Brand, Geochem.: Explor., Environ., Anal., 2014, 14(2), 125–138.
128. S. J. Piercey and M. C. Devine, Geochem.: Explor., Environ., Anal., 2014, 14(2), 139–148.
129. M. X. Ling, Y. Liu, H. Zhang and W. D. Sun, Anal. Lett., 2014, 47(9), 1598–1605.
130. R. J. Cooper, B. G. Rawlins, B. Leze, T. Krueger and K. M. Hiscock, Hydrol. Processes, 2014, 28(13), 4042–4056.
131. C. D. Pereira, J. L. Mantovano, C. C. Turci and E. D. M. Ferreira, Microchem. J., 2014, 115, 121–125.
132. T. Kato and K. Suzuki, J. Mineral. Petrol. Sci., 2014, 109(3), 151–155.
133. L. P. Bedard and A. Neron, Geostand. Geoanal. Res., 2014, 38(2), 177–188.
134. L. Poto, J. Gabrieli, S. Crowhurst, C. Agostinelli, A. Spolaor, W. R. L. Cairns, G. Cozzi and C. Barbante, Anal. Bioanal. Chem., 2015, 407(2), 379–385.
135. M. Finne, M. Kylander, M. Boyd, H. S. Sundqvist and L. Lowemark, Int. J. Speleol., 2015, 44(1), 17–23.
136. I. Rodriguez-Germade, B. Rubio and D. Rey, Mar. Pollut. Bull., 2014, 86(1–2), 458–467.
137. J. B. W. Stuut, F. Temmesfeld and P. de Deckker, Quat. Sci. Rev., 2014, 83, 83–94.
138. S. Shala, K. F. Helmens, K. N. Jansson, M. E. Kylander, J. Risberg and L. Lowemark, Boreas, 2014, 43(2), 362–376.
139. T. Kumpan, O. Babek, J. Kalvoda, T. M. Grygar and J. Fryda, Sediment. Geol., 2014, 311, 43–59.
140. G. Li and D. J. W. Piper, Quat. Sci. Rev., 2015, 107, 129–137.
141. C. Lenz, T. Behrends, T. Jilbert, M. Silveira and C. P. Slomp, Chem. Geol., 2014, 370, 49–57.
142. K. Schittek, M. Forbriger, B. Machtle, F. Schabitz, V. Wennrich, M. Reindel and B. Eitel, Clim. Past., 2015, 11(1), 27–44.
143. I. Unkel, A. Schimmelmann, C. Shriner, J. Forsen, C. Heymann and H. Bruckner, Z. Geomorphol., 2014, 58, 89–107.
144. M. Ulrich, M. Munoz, S. Guillot, M. Cathelineau, C. Picard, B. Quesnel, P. Boulvais and C. Couteau, Contrib. Mineral. Petrol., 2014, 167(1), 19.
145. K. A. Dyl, J. S. Cleverley, P. A. Bland, C. G. Ryan, L. A. Fisher and R. M. Hough, Geochim. Cosmochim. Acta, 2014, 134, 100–119.
146. Y. Tamenori, T. Yoshimura, N. T. Luan, H. Hasegawa, A. Suzuki, H. Kawahata and N. Iwasaki, J. Struct. Biol., 2014, 186(2), 214–223.
147. T. Kashiwabara, Y. Oishi, A. Sakaguchi, T. Sugiyama, A. Usui and Y. Takahashi, Geochim. Cosmochim. Acta, 2014, 131, 150–163.
148. J. T. Kelly, S. Carey, M. Pistolesi, M. Rosi, K. L. Croff-Bell, C. Roman and M. Marani, Bull. Volcanol., 2014, 76(7), 18.
149. N. Deseta, L. D. Ashwal and T. B. Andersen, Lithos, 2014, 206, 127–146.
150. M. J. Pushie, B. R. Pratt, T. C. Macdonald, G. N. George and I. J. Pickering, Palaios, 2014, 29(10), 512–524.
151. P. M. Wynn, I. J. Fairchild, C. Spotl, A. Hartland, D. Mattey, B. Fayard and M. Cotte, Environ. Chem., 2014, 11(1), 28–36.
152. D. Mondal, S. Gupta, D. V. Reddy and P. Nagabhushanam, J. Geochem. Explor., 2014, 145, 190–206.
153. R. Buamah, R. A. Mensah and A. Salifu, Water Sci. Technol.: Water Supply, 2013, 13(2), 238–248.
154. D. Mohan, S. Kumar and A. Srivastava, Ecol. Eng., 2014, 73, 798–808.
155. E. Fosso-Kankeu, A. F. Mulaba-Bafubiandi and T. G. Barnard, Physics and Chemistry of the Earth, Parts A/B/C, 2014, 67–69, 236–241.
156. L. K. ThomasArrigo, C. Mikutta, J. Byrne, K. Barmettler, A. Kappler and R. Kretzschmar, Environ. Sci. Technol., 2014, 48(22), 13218–13228.
157. S. Ohsawa, K. Sugimori, H. Yamauchi, T. Koeda, H. Inaba, Y. Kataoka and T. Kagiyama, Bull. Volcanol., 2014, 76(5), 10.
158. M. M. Demir, A. Baba, V. Atilla and M. Inanli, Geothermics, 2014, 50, 1–9.
159. R. J. Cooper, T. Krueger, K. M. Hiscock and B. G. Rawlins, Earth Surf. Processes Landforms, 2015, 40(1), 78–92.
160. N. Schreiber, E. Garcia, A. Kroon, P. C. Ilsoe, K. H. Kjaer and T. J. Andersen, Water, Air, Soil Pollut., 2014, 225(12), 11.
161. K. B. Liu, T. A. McCloskey, T. A. Bianchette, G. Keller, N. S. N. Lam, J. E. Cable and J. Arriola, J. Coastal Res., 2014, 266–271.
162. J. D. Naquin, K. B. Liu, T. A. McCloskey and T. A. Bianchette, J. Coastal Res., 2014, 308–313.
163. J. D. Silva, S. Srinivasalu, P. D. Roy and M. P. Jonathan, Environ. Earth Sci., 2014, 71(5), 2043–2058.
164. P. Bataillard, S. Grangeon, P. Quinn, F. Mosselmans, A. Lahfid, G. Wille, C. Joulian and F. Battaglia-Brunet, J. Soils Sediments, 2014, 14(3), 615–629.
165. H. Miller, I. W. Croudace, J. M. Bull, C. J. Cotterill, J. K. Dix and R. N. Taylor, Environ. Sci. Technol., 2014, 48(13), 7254–7263.
166. D. N. Schillereff, R. C. Chiverrell, N. Macdonald and J. M. Hooke, Earth-Sci. Rev., 2014, 135, 17–37.
167. D. Di Giuseppe, G. Bianchini, B. Faccini and M. Coltorti, X-Ray Spectrom., 2014, 43(3), 165–174.
168. D. T. Jayawardana, H. Pitawala and H. Ishiga, Environ. Earth Sci., 2014, 71(9), 4097–4106.
169. M. I. Kaniu and K. H. Angeyo, Geoderma, 2015, 241, 32–40.
170. A. Sharma, D. C. Weindorf, D. D. Wang and S. Chakraborty, Geoderma, 2015, 239, 130–134.
171. U. Proske, H. Heijnis and P. Gadd, Soil Res., 2014, 52(8), 760–768.
172. D. Badia, C. Marti, J. Casanova, T. Gillot, J. A. Cuchi, J. Palacio and R. Andres, Geoderma, 2015, 241, 158–167.
173. O. Olesen, H. P. Kierulf, M. Bronner, E. Dalsegg, O. Fredin and T. Solbakk, Norw. J. Geol., 2013, 93(3–4), 189–213.
174. A. D. Regmi, K. Yoshida, M. R. Dhital and B. Pradhan, Environ. Earth Sci., 2014, 71(6), 2711–2727.
175. Z. Z. Cheng, X. J. Xie, W. S. Yao, J. Z. Feng, Q. Zhang and J. D. Fang, J. Geochem. Explor., 2014, 139, 183–192.
176. S. Dantu, Environ. Earth Sci., 2014, 72(4), 955–981.
177. S. J. Barnes, L. A. Fisher, R. Anand and T. Uemoto, Aust. J. Earth Sci., 2014, 61(2), 269–285.
178. Q. M. Cheng, Geochem.: Explor., Environ., Anal., 2014, 14(3), 277–289.
179. M. F. Gazley, C. M. Tutt, L. A. Fisher, A. R. Latham, G. Duclaux, M. D. Taylor and S. J. de Beer, J. Geochem. Explor., 2014, 143, 74–83.
180. N. P. Edwards, P. L. Manning, U. Bergmann, P. L. Larson, B. E. van Dongen, W. I. Sellers, S. M. Webb, D. Sokaras, R. Alonso-Mori, K. Ignatyev, H. E. Barden, A. van Veelen, J. Anne, V. M. Egerton and R. A. Wogelius, Metallomics, 2014, 6(4), 774–782.
181. M. Gemelli, M. D'Orazio and L. Folco, Geostand. Geoanal. Res., 2015, 39(1), 55–69.
182. L. Fisher, M. F. Gazley, A. Baensch, S. J. Barnes, J. Cleverley and G. Duclaux, Geochem.: Explor., Environ., Anal., 2014, 14(2), 149–159.
183. G. E. M. Hall, G. F. Bonham-Carter and A. Buchar, Geochem.: Explor., Environ., Anal., 2014, 14(2), 99–123.
184. A. M. W. Hunt, D. Dvoracek, M. D. Glascock and R. J. Speakman, J. Radioanal. Nucl. Chem., 2015, 303(1), 977–978.
185. K. Lortzie, M. Stylianou, D. Dermatas and K. Kostarelos, Eng. Geol., 2015, 184, 119–125.
186. R. S. Smith, J. G. Wiederhold, A. D. Jew, G. E. Brown, B. Bourdon and R. Kretzschmar, Geochim. Cosmochim. Acta, 2014, 137, 1–17.
187. J. J. Yang, J. Liu, J. J. Dynes, D. Peak, T. Regier, J. Wang, S. H. Zhu, J. Y. Shi and J. S. Tse, Environ. Sci. Pollut. Res., 2014, 21(4), 2943–2954.
188. H. Tuovinen, D. Vesterbacka, E. Pohjolainen, D. Read, D. Solatie and J. Lehto, J. Geochem. Explor., 2015, 148, 174–180.
189. J. H. Park, V. Hodge, S. Gerstenberger and K. Stave, Appl. Sci., 2014, 4(2), 240–254.
190. F. Perlatti, X. L. Otero, F. Macias and T. O. Ferreira, Sci. Total Environ., 2014, 500, 91–102.
191. V. R. K. Vadapalli, V. Fester, L. Petrik and P. Slatter, Desalin. Water Treat., 2014, 52(37–39), 6947–6955.
192. E. Iakovleva, E. Makila, J. Salonen, M. Sitarz and M. Sillanpaa, Chem. Eng. J., 2015, 259, 364–371.
193. M. Masukume, M. S. Onyango and J. P. Maree, Int. J. Miner. Process., 2014, 133, 52–59.
194. E. Rosales, J. P. G. Loch and C. Dias-Ferreira, Electrochim. Acta, 2014, 127, 27–33.
195. A. Alshameri, C. J. Yan and X. R. Lei, Microporous Mesoporous Mater., 2014, 196, 145–157.
196. M. R. El-Naggar, J. Nucl. Mater., 2014, 447(1–3), 15–21.
197. R. A. de Menezes, S. P. A. da Paz, R. S. Angelica, R. D. Neves and S. B. C. Pergher, Mater. Res.-Ibero-Am. J. Mater., 2014, 17, 23–27.
198. N. M. Musyoka, R. Missengue, M. Kusisakana and L. F. Petrik, Appl. Clay Sci., 2014, 97–98, 182–186.
199. Q. Tang, T. Katsumi, T. Inui and Z. Z. Li, Sep. Purif. Technol., 2015, 141, 132–142.
200. M. D. Picanco, R. S. Angelica and M. S. Barata, Materia, 2014, 19(2), 68–80.
201. M. Sutcu, Ceram. Int., 2015, 41(2), 2819–2827.
202. T. A. Engidasew and G. Barbieri, J. Afr. Earth Sci., 2014, 99, 581–594.
203. L. L. Baker and D. G. Strawn, Clays Clay Miner., 2014, 62(1–2), 89–101.
204. E. Gliozzo, F. Iacoviello and L. M. Foresi, Appl. Clay Sci., 2014, 91–92, 105–116.
205. F. A. Morsy, S. El-Sherbiny, M. S. Hassan and H. F. Mohammed, Powder Technol., 2014, 264, 430–438.
206. M. West, S. Jamieson, M. Flower and S. Jones, Glass Technol.: Eur. J. Glass Sci. Technol., Part A, 2014, 55(2), 37–48.
207. F. Nitschke, J. Scheiber, U. Kramar and T. Neumann, Neues Jahrb. Mineral., Abh., 2014, 191(2), 145–156.
208. A. Azdarpour, M. Asadullah, R. Junin, M. Manan, H. Hamidi and E. Mohammadian, Fuel Process. Technol., 2014, 126, 429–434.
209. M. Mohapatra, V. Natarajan, B. Rajeswari, A. R. Dhobale and S. V. Godbole, J. Lumin., 2014, 145, 105–109.
210. J. Chen, G. J. Liu, H. Li and B. Wu, Int. J. Coal Geol., 2014, 124, 11–35.
211. S. J. Kelloway, C. R. Ward, C. E. Marjo, I. E. Wainwright and D. R. Cohen, Int. J. Coal Geol., 2014, 128, 55–67.
212. X. L. Li, S. Q. An, Z. S. Yu, J. F. Bai and Q. Zhang, Chin. J. Anal. Chem., 2014, 42(2), 283–287.
213. P. Tremain, J. Zanganeh, L. Hugo, S. Curry and B. Moghtaderi, Energy Fuels, 2014, 28(12), 7609–7615.
214. L. Wang, W. Sun, R. Q. Liu and X. C. Gu, Trans. Nonferrous Met. Soc. China, 2014, 24(4), 1145–1151.
215. W. F. Wang, W. D. Hao, Z. F. Sian, S. G. Lei, X. S. Wang, S. X. Sang and S. C. Xu, Int. J. Coal Geol., 2014, 132, 94–102.
216. M. Jablonska-Czapla, S. Szopa and C. Rosik-Dulewska, Arch. Environ. Prot., 2014, 40(2), 3–19.
217. M. J. Duane, Kuwait J. Sci., 2014, 41(3), 191–202.
218. K. A. Mumford, B. Pitt, A. T. Townsend, I. Snape and D. B. Gore, Minerals, 2014, 4(2), 293–312.
219. K. Akioka, T. Nakazawa, T. Doi, M. Arai and K. Tsuji, Powder Diffr., 2014, 29(2), 151–154.
220. S. Hirano, K. Akioka, T. Doi, M. Arai and K. Tsuji, X-Ray Spectrom., 2014, 43(4), 216–220.
221. K. Tsuji, S. Hirano, R. Yagi, T. Nakazawa, K. Akioka, M. Arai and T. Doi, J. Iron Steel Inst. Jpn., 2014, 100(7), 897–904.
222. E. J. Pickering and M. Holland, Ironmaking Steelmaking, 2014, 41(7), 493–499.
223. I. de La Calle, M. Quade, T. Krugmann and U. E. A. Fittschen, X-Ray Spectrom., 2014, 43(6), 345–352.
224. S. A. Bernal, V. Rose and J. L. Provis, Mater. Chem. Phys., 2014, 146(1–2), 1–5.
225. X. R. Wu, G. M. Yang, L. S. Li, H. H. Lu, Z. J. Wu and X. M. Shen, Ironmaking Steelmaking, 2014, 41(5), 335–341.
226. T. Gorewoda, J. Anyszkiewicz, Z. Mzyk and L. Buzek, Spectrochim. Acta, Part B, 2014, 92, 29–33.
227. M. Bouchard, A. Milliard, S. Rivard and S. Ness, Powder Diffr., 2014, 29(2), 170–175.
228. J. R. Sieber and A. Mortensen, X-Ray Spectrom., 2014, 43(5), 259–268.
229. P. Barriobero-Vila, G. Requena, T. Buslaps, M. Alfeld and U. Boesenberg, J. Alloys Compd., 2015, 626, 330–339.
230. R. Taurino, M. Cannio, T. Mafredini and P. Pozzi, Environ. Technol., 2014, 35(24), 3147–3152.
231. C. Gallen, A. Banks, S. Brandsma, C. Baduel, P. Thai, G. Eaglesham, A. Heffernan, P. Leonards, P. Bainton and J. F. Mueller, Sci. Total Environ., 2014, 491, 255–265.
232. G. Z. Miller and Z. E. Harris, J. Environ. Health, 2015, 77(6), 8–13.
233. M. M. Hillyer, L. E. Finch, A. S. Cerel, J. D. Dattelbaum and M. C. Leopold, Chemosphere, 2014, 108, 205–213.
234. M. Fellin, M. Negri and R. Zanuttini, Eur. J. Wood Wood Prod., 2014, 72(2), 199–211.
235. I. H. Ali and A. A. Ateeg, Int. J. Environ. Res., 2015, 9(1), 291–294.
236. C. Bartrem, S. Tirima, I. von Lindern, M. von Braun, M. C. Worrell, S. M. Anka, A. Abdullahi and G. Moller, Int. J. Environ. Res., 2014, 24(4), 304–319.
237. R. K. Dutta, S. Sarkar, S. S. Ram, M. Sudarshan, R. Acharya and A. V. R. Reddy, J. Radioanal. Nucl. Chem., 2014, 302(3), 1519–1523.
238. G. Machender, R. Dhakate, S. T. M. Rao, B. M. Rao and L. Prasanna, Arabian J. Geosci., 2014, 7(2), 513–525.
239. L. R. Baker, G. M. Pierzynski, G. M. Hettiarachchi, K. G. Scheckel and M. Newville, J. Environ. Qual., 2014, 43(2), 488–497.
240. J. R. Hayes, A. P. Grosvenor, J. Rowson, K. Hughes, R. A. Frey and J. Reid, Environ. Sci. Technol., 2014, 48(8), 4460–4467.
241. L. A. Hutton, G. D. O'Neil, T. L. Read, Z. J. Ayres, M. E. Newton and J. V. Macpherson, Anal. Chem., 2014, 86(9), 4566–4572.
242. C. Herzog, O. Bakaeva, C. Boit, B. Kanngiesser and F. Friedrich, Appl. Phys. Lett., 2014, 105(4), 4.
243. P. Kumar, M. Aslam, N. Singh, S. Mittal, A. Bansal, M. K. Jha and A. K. Sarma, RSC Adv., 2015, 5(13), 9946–9954.
244. S. Y. Jung, J. M. Moon, S. C. Lee, S. C. Paik, K. S. Park and J. C. Kim, Adsorption, 2014, 20(2–3), 341–348.
245. Y. Xu, C. H. Yan, B. T. Xu, X. H. Ruan and Z. Wei, Ceram. Int., 2014, 40(6), 8833–8840.
246. M. F. Gazulla, M. Rodrigo, M. Orduna and M. J. Ventura, Glass Technol.: Eur. J. Glass Sci. Technol., Part A, 2014, 55(6), 177–182.
247. O. O. Fatoba, L. F. Petrik, R. O. Akinyeye, W. M. Gitari and E. I. Iwuoha, Int. J. Environ. Sci. Technol., 2015, 12(2), 551–562.
248. G. Chu, D. Guinea and P. T. Palmer, Spectroscopy, 2014, 29(7), 38.
249. Z. H. I. Sun, Y. Xiao, J. Sietsma, H. Agterhuis, G. Visser and Y. Yang, Waste Manage., 2015, 35, 227–235.
250. S. K. Tian, L. L. Lu, R. H. Xie, M. Z. Zhang, J. A. Jernstedt, D. D. Hou, C. Ramsier and P. H. Brown, Front. Plant Sci., 2015, 5, 9.
251. L. L. Lu, X. C. Liao, J. Labavitch, X. Yang, E. Nelson, Y. H. Du, P. H. Brown and S. K. Tian, Plant Physiol. Biochem., 2014, 84, 224–232.
252. Y. T. Shen, X-Ray Spectrom., 2014, 43(3), 146–151.
253. G. de la Rosa, H. Castillo-Michel, G. Cruz-Jimenez, J. Bernal-Alvarado, T. Cordova-Fraga, L. Lopez-Moreno and M. Cotte, Int. J. Phytorem., 2014, 16(11), 1073–1086.
254. S. J. Archibald, S. L. Atkin, W. Bras, A. Diego-Taboada, G. Mackenzie, J. F. W. Mosselmans, S. Nikitenko, P. D. Quinn, M. F. Thomas and N. A. Young, J. Mater. Chem. B, 2014, 2(8), 945–959.
255. J. Augustynowicz, P. Wrobel, B. J. Plachno, G. Tylko, Z. Gajewski and D. Wegrzynek, Planta, 2014, 239(6), 1233–1242.
256. Y. L. Chen, X. Q. Hong, H. He, H. W. Luo, T. T. Qian, R. Z. Li, H. Jiang and H. Q. Yu, Bioresour. Technol., 2014, 160, 89–92.
257. H. Kowata, Y. Nagakawa, N. Sakurai, A. Hokura, Y. Terada, H. Hasegawa and E. Harada, J. Anal. At. Spectrom., 2014, 29(5), 868–874.
258. S. K. Tian, L. L. Lu, J. M. Labavitch, S. M. Webb, X. E. Yang, P. H. Brown and Z. L. He, J. Exp. Bot., 2014, 65(4), 953–964.
259. S. P. Singh, K. Vogel-Mikus, P. Vavpetic, L. Jeromel, P. Pelicon, J. Kumar and R. Tuli, Planta, 2014, 240(2), 277–289.
260. S. Majumdar, J. R. Peralta-Videa, S. Bandyopadhyay, H. Castillo-Michel, J. A. Hernandez-Viezcas, S. Sahi and J. L. Gardea-Torresdey, J. Hazard. Mater., 2014, 278, 279–287.
261. K. Nayuki, B. D. Chen, R. Ohtomo and Y. Kuga, Microbes Environ., 2014, 29(1), 60–66.
262. A. Scholer, I. Zaharieva, S. Zimmermann, M. Wiechen, A. M. Manke, P. Kurz, C. Plieth and H. Dau, Eur. J. Inorg. Chem., 2014, 780–790.
263. Y. M. Du, P. M. Kopittke, B. N. Noller, S. A. James, H. H. Harris, Z. P. Xu, P. Li, D. R. Mulligan and L. B. Huang, Ann. Bot, 2015, 115(1), 41–53.
264. X. Gui, X. He, Y. H. Ma, P. Zhang, Y. Y. Li, Y. Y. Ding, K. Yang, H. F. Li, Y. K. Rui, Z. F. Chai, Y. L. Zhao and Z. Y. Zhang, RSC Adv., 2015, 5(6), 4554–4560.
265. J. Siritapetawee, W. Limphirat, C. Kantachot and C. Kongmark, Appl. Biochem. Biotechnol., 2015, 175(1), 232–242.
266. C. Larue, H. Castillo-Michel, S. Sobanska, L. Cecillon, S. Bureau, V. Barthes, L. Ouerdane, M. Carriere and G. Sarret, J. Hazard. Mater., 2014, 264, 98–106.
267. C. Larue, H. Castillo-Michel, S. Sobanska, N. Trcera, S. Sorieul, L. Cecillon, L. Ouerdane, S. Legros and G. Sarret, J. Hazard. Mater., 2014, 273, 17–26.
268. L. J. Zhao, J. R. Peralta-Videa, C. M. Rico, J. A. Hernandez-Viezcas, Y. P. Sun, G. H. Niu, A. Servin, J. E. Nunez, M. Duarte-Gardea and J. L. Gardea-Torresdey, J. Agric. Food Chem., 2014, 62(13), 2752–2759.
269. T. Inui, Y. Koike and T. Nakamura, X-Ray Spectrom., 2014, 43(2), 112–117.
270. B. Kyriacou, K. L. Moore, D. Paterson, M. D. de Jonge, D. L. Howard, J. Stangoulis, M. Tester, E. Lombi and A. A. T. Johnson, J. Cereal Sci., 2014, 59(2), 173–180.
271. B. Meng, X. B. Feng, G. L. Qiu, C. W. N. Anderson, J. X. Wang and L. Zhao, Environ. Sci. Technol., 2014, 48(14), 7974–7981.
272. N. Yamaguchi, T. Ohkura, Y. Takahashi, Y. Maejima and T. Arao, Environ. Sci. Technol., 2014, 48(3), 1549–1556.
273. R. D. R. Ribeiro, E. T. Marsico, E. F. O. de Jesus, C. D. Carneiro, C. A. Conte, E. de Almeida and V. F. do Nascimento, J. Food Sci., 2014, 79(4), T738–T742.
274. E. Margui, A. D. Marques, M. D. Prisal, M. Hidalgo, I. Queralt and M. L. Carvalho, Appl. Spectrosc., 2014, 68(11), 1241–1246.
275. W. D. Li, O. Li, C. Mo, Y. S. Jiang, Y. He, A. R. Zhang, L. M. Chen and J. S. Jin, J. Hortic. Sci. Biotechnol., 2014, 89(6), 674–678.
276. M. B. B. Guerra, E. de Almeida, G. G. A. Carvalho, P. F. Souza, L. C. Nunes, D. Santos and F. J. Krug, J. Anal. At. Spectrom., 2014, 29(9), 1667–1674.
277. D. Bourassa, S. C. Gleber, S. Vogt, H. Yi, F. Will, H. Richter, C. H. Shin and C. J. Fahrni, Metallomics, 2014, 6(9), 1648–1655.
278. T. Leonardo, E. Farhi, A. M. Boisson, J. Vial, P. Cloetens, S. Bohic and C. Rivasseau, Metallomics, 2014, 6(2), 316–329.
279. T. Yoshimura, Y. Tamenori, H. Kawahata and A. Suzuki, Biogeosciences, 2014, 11(14), 3881–3886.
280. G. Y. Anshebo, P. M. Graves, S. C. Smith, A. B. Wills, M. Damte, T. Endeshaw, E. B. Shargie, T. Gebre, A. W. Mosher, A. E. Patterson and P. M. Emerson, Malar. J., 2014, 13, 14.
281. M. Rubio, A. Germanier, M. F. Mera, S. N. Faudone, R. D. Sbarato, J. M. Campos, V. Zampar, E. Bonzi and C. A. Perez, X-Ray Spectrom., 2014, 43(3), 186–192.
282. M. F. Mera, M. Rubio, C. A. Perez, V. Galvan and A. Germanier, Microchem. J., 2015, 119, 114–122.
283. C. A. Shand and R. Wendler, J. Geochem. Explor., 2014, 143, 31–42.
284. M. Menzel and U. E. A. Fittschen, Anal. Chem., 2014, 86(6), 3053–3059.
285. P. Pin and B. Perot, IEEE Trans. Nucl. Sci., 2014, 61(4), 2131–2136.
286. R. E. Bevins, R. A. Ixer and N. J. G. Pearce, J. Archaeol. Sci., 2014, 42, 179–193.
287. B. V. Rolett, E. W. West, J. M. Sinton and R. Iovita, J. Archaeol. Sci., 2015, 53, 459–471.
288. M. Sepulveda, S. Gutierrez, M. Campos-Vallette, E. Clavijo, P. Walter and J. J. Carcamo, J. Chil. Chem. Soc., 2014, 58(3), 1836–1839.
289. L. Pappalardo, F. P. Romano, D. Bracchitta, A. Massimino, O. Palio and F. Rizzo, J. Geophys. Eng., 2013, 10(6), 11.
290. E. Frahm, J. Archaeol. Sci., 2014, 49, 105–125.
291. J. P. Grattan, D. D. Gilbertson, J. H. Waller and R. B. Adams, J. Archaeol. Sci., 2014, 46, 428–433.
292. E. Anderson, M. J. Almond and W. Matthews, Spectrochim. Acta, Part A, 2014, 133, 326–334.
293. A. M. A. Kamel, H. A. H. Marie, M. F. Ali and H. A. Mahmoud, Period. Mineral., 2014, 83(3), 313–327.
294. E. Franceschi and F. Locardi, J. Cult. Herit., 2014, 15(5), 522–527.
295. S. Columbu, F. Antonelli, M. Lezzerini, D. Miriello, B. Adembri and A. Blanco, J. Archaeol. Sci., 2014, 49, 332–342.
296. S. Krug and O. Hahn, Stud. Conserv., 2014, 59(6), 355–366.
297. I. Fierascu, R. C. Fierascu, R. M. Ion and C. Radovici, Rev. Rom. Mater., 2014, 44(3), 292–297.
298. A. Rodrigues, S. Gutierrez-Patricio, A. Z. Miller, C. Saiz-Jimenez, R. Wiley, D. Nunes, M. Vilarigues and M. F. Macedo, Int. Biodeterior. Biodegrad., 2014, 90, 152–160.
299. E. Arizio, E. F. Orsega, R. Falcone and M. Vallotto, Environ. Sci. Pollut. Res., 2014, 21(23), 13243–13251.
300. Y. Fors, H. Grudd, A. Rindby, F. Jalilehvand, M. Sandstrom, I. Cato and L. Bornmalm, Sci. Rep., 2014, 4, 6.
301. A. Lanzirotti, R. Bianucci, R. LeGeros, T. G. Bromage, V. Giuffra, E. Ferroglio, G. Fornaciari and O. Appenzeller, J. Archaeol. Sci., 2014, 52, 204–217.
302. L. Goldenberg, R. Neumann and S. Weiner, J. Archaeol. Sci., 2014, 42, 509–518.
303. E. Ay, M. Kibaroglu and C. Berthold, Archaeol. Anthropol. Sci., 2014, 6(2), 125–131.
304. L. Monico, K. Janssens, F. Vanmeert, M. Cotte, B. G. Brunetti, G. van der Snickt, M. Leeuwestein, J. S. Plisson, M. Menu and C. Miliani, Anal. Chem., 2014, 86(21), 10804–10811.
305. F. C. Izzo, B. Ferriani, K. J. van den Berg, H. van Keulen and E. Zendri, J. Cult. Herit., 2014, 15(5), 557–563.
306. A. Daveri, B. Doherty, P. Moretti, C. Grazia, A. Romani, E. Fiorin, B. G. Brunetti and M. Vagnini, Spectrochim. Acta, Part A, 2015, 135, 398–404.
307. R. Sefcu, S. Chlumska, A. Tregtikova, T. Trojek and L. Dragounova, Appl. Radiat. Isot., 2015, 95, 8–12.
308. M. Bicchieri, Environ. Sci. Pollut. Res., 2014, 21(24), 14146–14157.
309. L. Beck, H. Rousseliere, J. Castaing, A. Duran, M. Lebon, B. Moignard and F. Plassard, Talanta, 2014, 129, 459–464.
310. L. van de Voorde, J. van Pevenage, K. de Langhe, R. de Wolf, B. Vekemans, L. Vincze, P. Vandenabeele and M. P. J. Martens, Spectrochim. Acta, Part B, 2014, 97, 1–6.
311. M. Aceto, A. Agostino, G. Fenoglio, A. Idone, M. Gulmini, M. Picollo, P. Ricciardi and J. K. Delaney, Anal. Methods, 2014, 6(5), 1488–1500.
312. A. M. W. Hunt and R. J. Speakman, J. Archaeol. Sci., 2015, 53, 626–638.
313. X. X. Qi, M. Santosh, L. H. Zhu, Y. H. Zhao, Z. C. Hu, C. Zhang and F. B. Ji, Precambrian Res., 2014, 245, 207–224.
314. A. G. Attaelmanan, X-Ray Spectrom., 2014, 43(6), 325–331.
315. J. M. Compana, A. Cabeza, M. A. G. Aranda and L. Leon-Reina, J. Archaeol. Sci., 2014, 45, 26–35.
316. J. Zhu, H. P. Duan, Y. M. Yang, L. Guan, W. Xu, D. L. Chen, J. Zhang, L. H. Wang, Y. Y. Huang and C. S. Wang, J. Synchrotron Radiat., 2014, 21, 751–755.
317. J. M. Wu, Q. J. Li, M. L. Zhang, J. Wu, Y. Z. Ding, J. W. Cao and X. J. Liu, Spectrosc. Spectral Anal., 2014, 34(3), 827–832.
318. C. F. Ming, Y. M. Yang, J. Zhu, L. Guan, C. S. Fan, C. Q. Xu, Z. Q. Yao, J. M. Kenoyer, G. D. Song and C. S. Wang, J. Archaeol. Sci., 2014, 47, 78–84.
319. L. V. de Voorde, M. Vandevijvere, B. Vekemans, J. van Pevenage, J. Caen, P. Vandenabeele, P. van Espen and L. Vincze, Spectrochim. Acta, Part B, 2014, 102, 28–35.
320. G. Montana, L. Randazzo, C. M. Belfiore, M. F. la Russa, S. A. Ruffolo, A. M. de Francesco, A. Pezzino, R. Punturo and V. Di Stefano, Period. Mineral., 2014, 83(1), 89–120.
321. H. X. Zhao, Q. H. Li, S. Liu, L. Li and F. X. Gan, X-Ray Spectrom., 2014, 43(6), 316–324.
322. A. Detcheva, R. Velinova, E. Ivanova, J. Jordanov and M. Karadjov, C. R. Acad. Bulg. Sci., 2014, 67(6), 769–776.
323. Z. Gu, J. Zhu, Y. T. Xie, T. Q. Xiao, Y. M. Yang and C. S. Wang, J. Anal. At. Spectrom., 2014, 29(8), 1438–1443.
324. M. Ferretti, J. Anal. At. Spectrom., 2014, 29(10), 1753–1766.
325. L. G. Troalen, J. Tate and M. F. Guerra, J. Archaeol. Sci., 2014, 50, 219–226.
326. A. Charalambous, V. Kassianidou and G. Papasavvas, J. Archaeol. Sci., 2014, 46, 205–216.
327. S. La Niece, V. Harris and H. Uchida, ISIJ Int., 2014, 54(5), 1111–1116.
328. J. C. D. de Figueiredo, S. S. Asevedo and J. H. R. Barbosa, Appl. Surf. Sci., 2014, 317, 67–72.
329. R. Cesareo, X-Ray Spectrom., 2014, 43(6), 312–315.
330. S. Pessanha, T. I. Madeira, M. Manso, M. Guerra, A. le Gac and M. L. Carvalho, Appl. Phys. A: Mater. Sci. Process., 2014, 116(3), 1053–1058.
331. S. Pessanha, M. Guerra, S. Longelin, A. le Gac, M. Manso and M. L. Carvalho, X-Ray Spectrom., 2014, 43(2), 79–82.
332. M. W. Hinds, G. Bevan and R. W. Burgess, J. Anal. At. Spectrom., 2014, 29(10), 1799–1805.
333. O. H. J. Christie, A. Racz, J. Elek and K. Heberger, J. Chemom., 2014, 28(4), 287–292.
334. R. M. Conrey, M. Goodman-Elgar, N. Bettencourt, A. Seyfarth, A. van Hoose and J. A. Wolff, Geochem.: Explor., Environ., Anal., 2014, 14(3), 291–301.
335. I. Maeda, S. Furuya, W. S. K. Bong, Y. Abe, K. Osaka, M. Itou, T. Ninomiya and I. Nakai, Bunseki Kagaku, 2014, 63(3), 171–193.
336. B. Woods, K. P. Kirkbride, C. Lennard and J. Robertson, Forensic Sci. Int., 2014, 245, 195–201.
337. C. R. Appoloni and F. L. Melquiades, Appl. Radiat. Isot., 2014, 85, 92–95.
338. D. L. Feraru, M. Mihaly and A. Meghea, Color Res. Appl., 2015, 40(2), 169–177.
339. E. M. Suzuki, J. Forensic Sci., 2014, 59(2), 344–363.
340. E. M. Suzuki, J. Forensic Sci., 2014, 59(5), 1205–1225.
341. H. Kinoshita, N. Tanaka, M. Kumihashi, M. Jamal, A. Takakura, T. Umemoto, K. Tsutsui and K. Ameno, Rom. J. Leg. Med., 2014, 22(2), 117–118.
342. F. Zapata, M. A. F. de la Ossa and C. Garcia-Ruiz, TrAC, Trends Anal. Chem., 2015, 64, 53–63.
343. L. Q. Ren, D. Wu, Y. H. Li, G. Wang, X. Z. Wu and H. Liu, Med. Phys., 2014, 41(3), 12.
344. L. Q. Ren, D. Wu, L. L. Fajardo, Y. H. Li, B. Zheng and H. Liu, J. X-Ray Sci. Technol., 2014, 22(6), 785–796.
345. K. P. M. Ricketts, U. Cheema, A. Nyga, A. Castoldi, C. Guazzoni, T. Magdeldin, M. Emberton, A. P. Gibson, G. J. Royle and M. Loizidou, Small, 2014, 10(19), 3954–3961.
346. T. Q. Liu, I. Kempson, M. de Jonge, D. L. Howard and B. Thierry, Nanoscale, 2014, 6(16), 9774–9782.
347. C. M. Weekley, A. Shanu, J. B. Aitken, S. Vogt, P. K. Witting and H. H. Harris, Metallomics, 2014, 6(9), 1602–1615.
348. D. J. Hughes, V. Fedirko, M. Jenab, L. Schomburg, C. Meplan, H. Freisling, H. B. Bueno-de-Mesquita, S. Hybsier, N. P. Becker, M. Czuban, A. Tjonneland, M. Outzen, M. C. Boutron-Ruault, A. Racine, N. Bastide, T. Kuhn, R. Kaaks, D. Trichopoulos, A. Trichopoulou, P. Lagiou, S. Panico, P. H. Peeters, E. Weiderpass, G. Skeie, E. Dagrun, M. D. Chirlaque, M. J. Sanchez, E. Ardanaz, I. Ljuslinder, M. Wennberg, K. E. Bradbury, P. Vineis, A. Naccarati, D. Palli, H. Boeing, K. Overvad, M. Dorronsoro, P. Jakszyn, A. J. Cross, J. R. Quiros, M. Stepien, S. Y. Kong, T. Duarte-Salles, E. Riboli and J. E. Hesketh, Int. J. Cancer, 2015, 136(5), 1149–1161.
349. M. W. Lankosz, M. Grzelak, B. Ostachowicz, A. Wandzilak, M. Szczerbowska-Boruchowska, P. Wrobel, E. Radwanska and D. Adamek, Spectrochim. Acta, Part B, 2014, 101, 98–105.
350. A. Kubala-Kukus, D. Banas, J. Braziewicz, U. Majewska, M. Pajek, J. Wudarczyk-Mocko, G. Antczak, B. Borkowska, S. Gozdz and J. Smok-Kalwat, Biol. Trace Elem. Res., 2014, 158(1), 22–28.
351. N. Braidy, A. Poljak, C. Marjo, H. Rutlidge, A. Rich, T. Jayasena, N. C. Inestrosa and P. Sachdev, Front. Aging Neurosci., 2014, 6, 14.
352. J. F. Collingwood and M. R. Davidson, Front. Pharmacol., 2014, 5, 19.
353. L. Pascolo, A. Gianoncelli, C. Rizzardi, V. Tisato, M. Salome, C. Calligaro, F. Salvi, D. Paterson and P. Zamboni, Sci. Rep., 2014, 4, 10.
354. A. Grubman, S. A. James, J. James, C. Duncan, I. Volitakis, J. L. Hickey, P. J. Crouch, P. S. Donnelly, K. M. Kanninen, J. R. Liddell, S. L. Cotman, M. D. de Jonge and A. R. White, Chem. Sci., 2014, 5(6), 2503–2516.
355. H. Moise, D. R. Chettle and A. Pejovic-Milic, Bone, 2014, 61, 48–54.
356. F. Mostafaei, F. E. McNeill, D. R. Chettle and M. D. Noseworthy, Physiol. Meas., 2015, 36(1), N1–N13.
357. J. S. Ji, J. Schwartz, D. Sparrow, H. Hu and M. G. Weisskopf, J. Occup. Environ. Med., 2014, 56(4), 435–440.
358. K. D. Eum, M. G. Weisskopf, L. H. Nie, H. Hu and S. A. Korrick, Environ. Health Perspect., 2014, 122(3), 229–234.
359. E. Vergucht, B. de Samber, A. Izmer, B. Vekemans, K. Appel, S. Tolmachev, L. Vincze and F. Vanhaecke, Anal. Bioanal. Chem., 2015, 407(6), 1559–1566.
360. P. Andujar, A. Simon-Deckers, F. Galateau-Salle, B. Fayard, G. Beaune, B. Clin, M. A. Billon-Galland, O. Durupthy, J. C. Pairon, J. Doucet, J. Boczkowski and S. Lanone, Part. Fibre Toxicol., 2014, 11, 13.
361. S. S. Jakobsen, C. Liden, K. Soballe, J. D. Johansen, T. Menne, L. Lundgren, D. Bregnbak, P. Moller, M. S. Jellesen and J. P. Thyssen, Contact Dermatitis, 2014, 71(6), 319–325.
362. M. Calasans-Maia, J. Calasans-Maia, S. Santos, E. Mavropoulos, M. Farina, I. Lima, R. T. Lopes, A. Rossi and J. M. Granjeiro, Mater. Sci. Eng., C, 2014, 41, 309–319.
363. G. Totea, D. Ionita, I. Demetrescu and M. Mitache, Cent. Eur. J. Chem., 2014, 12(7), 796–803.
364. O. Ponomarenko, M. R. Gherase, M. S. LeBlanc, C. Y. Kim, E. D. Desouza, M. J. Farquharson, F. E. McNeill, S. Nehzati, I. J. Pickering, G. N. George and D. E. B. Fleming, Environ. Chem., 2014, 11(6), 632–643.
365. R. G. Figueroa, I. R. Chavez and E. Bonzi, X-Ray Spectrom., 2014, 43(6), 338–344.
366. J. Chikawa, Y. Mouri, H. Shima, K. Yamada, H. Yamamoto and S. Yamamoto, J. X-Ray Sci. Technol., 2014, 22(5), 587–603.
367. U. Majewska, D. Banas, J. Braziewicz, A. Kubala-Kukus, M. Pajek, I. Sychowska, J. Wudarczyk-Mocko, G. Antczak, B. Borkowska and S. Gozdz, Acta Phys. Pol., A, 2014, 125(4), 864–868.
368. V. B. Yadav, R. H. Pillay and S. K. Jha, J. Radioanal. Nucl. Chem., 2014, 300(1), 57–60.
369. C. de Souza-Guerra, R. C. Barroso, A. P. de Almeida, I. T. A. Peixoto, S. Moreira, F. B. de Sousa and R. F. Gerlach, J. Trace Elem. Med. Biol., 2014, 28(2), 186–193.
370. J. Godinho, J. Silveira, A. Mata, M. L. Carvalho and S. Pessanha, Nucl. Instrum. Methods Phys. Res., Sect. B, 2014, 337, 78–82.
371. J. Silveira, J. Godinho, A. Mata, M. L. Carvalho and S. Pessanha, X-Ray Spectrom., 2015, 44(1), 3–6.
372. J. A. Abraham, H. J. Sanchez, M. S. Grenon and C. A. Perez, X-Ray Spectrom., 2014, 43(4), 193–197.
373. A. C. Deymier-Black, A. Veis, Z. Cai and S. R. Stock, Scanning, 2014, 36(2), 231–240.
374. H. Yoshii, K. Yanagihara, H. Imaseki, T. Hamano, H. Yamanishi, M. Inagaki, Y. Sakai, N. Sugiura, O. Kurihara and K. Sakai, PLoS One, 2014, 9(7), 7.
375. J. A. Centeno, D. A. Rogers, G. B. van der Voet, E. Fornero, L. S. Zhang, F. G. Mullick, G. D. Chapman, A. O. Olabisi, D. J. Wagner, A. Stojadinovic and B. K. Potter, Int. J. Environ. Res. Public Health, 2014, 11(2), 1261–1278.
376. G. J. Lis, J. Czapla-Masztafiak, W. M. Kwiatek, M. Gajda, E. Jasek, M. Jasinska, U. Czubek, M. Borchert, K. Appel, J. Nessler, J. Sadowski and J. A. Litwin, Micron, 2014, 67, 141–148.
377. M. Schmeling, B. I. Gaynes and S. Tidow-Kebritchi, Powder Diffr., 2014, 29(2), 155–158.
378. D. E. Morrison, J. B. Aitken, M. D. de Jonge, F. Issa, H. H. Harris and L. M. Rendina, Chem.–Eur. J., 2014, 20(50), 16602–16612.
379. O. Sogut, A. F. Kocaer, O. S. Zorer, Y. Ozdemir and M. Dogru, Microchem. J., 2014, 115, 19–26.
380. T. A. P. Fernandes, J. A. A. Brito and L. M. L. Goncalves, Food Anal. Methods, 2015, 8(1), 52–57.
381. V. Iannuccelli, D. Bertelli, M. Romagnoli, S. Scalia, E. Maretti, F. Sacchetti and E. Leo, Colloids Surf., B, 2014, 122, 653–661.
382. L. Pascolo, B. Bortot, N. Benseny-Cases, A. Gianoncelli, G. Tosi, B. Ruozi, C. Rizzardi, E. de Martino, M. A. Vandelli and G. M. Severini, Int. J. Nanomed., 2014, 9, 2791–2801.
383. M. Rak, M. Salome, S. G. W. Kaminskyj and K. M. Gough, Anal. Bioanal. Chem., 2014, 406(12), 2809–2816.
384. M. D. A. Gonzalez-Chavez, B. Miller, I. E. Maldonado-Mendoza, K. Scheckel and R. Carrillo-Gonzalez, Fungal Biol., 2014, 118(5–6), 444–452.
385. K. Jurowski, B. Buszewski and W. Piekoszewski, Talanta, 2015, 131, 273–285.
386. M. J. Ceko, K. Hummitzsch, N. Hatzirodos, W. M. Bonner, J. B. Aitken, D. L. Russell, M. Lane, R. J. Rodgers and H. H. Harris, Metallomics, 2015, 7(1), 66–77.
387. H. Hayashi, X-Ray Spectrom., 2014, 43(5), 292–297.
388. J. Anne, N. P. Edwards, R. A. Wogelius, A. R. Tumarkin-Deratzian, W. I. Sellers, A. van Veelen, U. Bergmann, D. Sokaras, R. Alonso-Mori, K. Ignatyev, V. M. Egerton and P. L. Manning, J. R. Soc., Interface, 2014, 11(96), 10.
389. M. Kerschnitzki, T. Zander, P. Zaslansky, P. Fratzl, R. Shahar and W. Wagermaier, Bone, 2014, 69, 109–117.
390. P. Feng, W. X. Cong, B. A. Wei and G. Wang, IEEE Trans. Biomed. Eng., 2014, 61(3), 975–985.
391. M. Muller, P. Honicke, B. Detlefs and C. Fleischmann, Materials, 2014, 7(4), 3147–3159.
392. S. N. Yakunin, I. A. Makhotkin, R. W. E. van de Kruijs, M. A. Chuev, E. M. Pashaev, E. Zoethout, E. Louis, S. Y. Seregin, I. A. Subbotin, D. V. Novikov, F. Bijkerk and M. V. Kovalchuk, J. Appl. Phys., 2014, 115(13), 7.
393. C. Stefaniu and G. Brezesinski, Curr. Opin. Colloid Interface Sci., 2014, 19(3), 216–227.
394. A. Lesnik, J. Blasing, J. Hennig, A. Dadgar and A. Krost, J. Phys. D: Appl. Phys., 2014, 47(35), 6.
395. B. Laforce, S. Schmitz, B. Vekemans, J. Rudloff, J. Garrevoett, R. Tucoulou, F. E. Brenker, G. Martinez-Criado and L. Vincze, Anal. Chem., 2014, 86(24), 12369–12374.
396. S. R. Bare, M. E. Charochak, S. D. Kelly, B. Lai, J. Wang and Y. C. K. Chen-Wiegart, ChemCatChem, 2014, 6(5), 1427–1437.
397. J. Segura-Ruiz, G. Martinez-Criado, C. Denker, J. Malindretos and A. Rizzi, Nano Lett., 2014, 14(3), 1300–1305.
398. N. Ishiguro, T. Uruga, O. Sekizawa, T. Tsuji, M. Suzuki, N. Kawamura, M. Mizumaki, K. Nitta, T. Yokoyama and M. Tada, ChemPhysChem, 2014, 15(8), 1563–1568.
399. L. Cristofolini, Curr. Opin. Colloid Interface Sci., 2014, 19(3), 228–241.
400. Q. Hu, M. Aboustait, M. T. Ley, J. C. Hanan, V. Rose and R. Winarski, Acta Mater., 2014, 77, 173–182.
401. T. J. Kuhn, J. Hormes, N. Matoussevitch, H. Bonnemann and P. Glatzel, Inorg. Chem., 2014, 53(16), 8367–8375.
402. M. H. Chu, G. Martinez-Criado, J. Segura-Ruiz, S. Geburt and C. Ronning, Phys. Status Solidi A, 2014, 211(2), 483–487.
403. A. Johannes, S. Noack, W. Paschoal, S. Kumar, D. Jacobsson, H. Pettersson, L. Samuelson, K. A. Dick, G. Martinez-Criado, M. Burghammer and C. Ronning, J. Phys. D: Appl. Phys., 2014, 47(39), 6.
404. G. Martinez-Criado, J. Segura-Ruiz, M. H. Chu, R. Tucoulou, I. Lopez, E. Nogales, B. Mendez and J. Piqueras, Nano Lett., 2014, 14(10), 5479–5487.
405. M. Filez, H. Poelman, R. K. Ramachandran, J. Dendooven, K. Devloo-Casier, E. Fonda, C. Detavernier and G. B. Marin, Catal. Today, 2014, 229, 2–13.
406. H. M. Hertz, J. C. Larsson, U. Lundstrom, D. H. Larsson and C. Vogt, Opt. Lett., 2014, 39(9), 2790–2793.
407. M. Meyns, F. Iacono, C. Palencia, J. Geweke, M. D. Coderch, U. E. A. Fittschen, J. M. Gallego, R. Otero, B. H. Juarez and C. Klinke, Chem. Mater., 2014, 26(5), 1813–1821.
408. R. Fernandez-Ruiz, M. J. Redrejo, E. J. Friedrich, M. Ramos and T. Fernandez, Anal. Chem., 2014, 86(15), 7383–7390.
409. J. T. Rashkow, S. C. Patel, R. Tappero and B. Sitharaman, J. R. Soc., Interface, 2014, 11(94), 5.
410. S. Y. Choi, N. Yang, S. K. Jeon and T. H. Yoon, Cytometry, Part A, 2014, 85(9), 771–780.
411. B. Bozzini, P. Bocchetta, A. Gianoncelli, C. Mele and M. Kiskinova, Acta Chim. Slov., 2014, 61(2), 263–271.
412. K. Kocot and R. Sitko, Spectrochim. Acta, Part B, 2014, 94–95, 7–13.
413. N. L. Cordes, G. J. Havrilla, I. O. Usov, K. A. Obrey and B. M. Patterson, Spectrochim. Acta, Part B, 2014, 101, 320–329.
414. T. X. Sun, Z. G. Liu, G. F. Wang, Y. Z. Ma, S. Peng, W. Y. Sun, F. Z. Li, X. P. Sun and X. L. Ding, Appl. Radiat. Isot., 2014, 94, 109–112.
415. P. Wrobel, D. Wegrzynek, M. Czyzycki and M. Lankosz, Anal. Chem., 2014, 86(22), 11275–11280.
416. S. Peng, Z. G. Liu, T. X. Sun, G. F. Wang, Y. Z. Ma and X. L. Ding, Appl. Radiat. Isot., 2014, 90, 84–88.
417. B. Kanrar, K. Sanyal, N. L. Misra and S. K. Aggarwal, Spectrochim. Acta, Part B, 2014, 101, 130–133.
418. M. Grafe, E. Donner, R. N. Collins and E. Lombi, Anal. Chim. Acta, 2014, 822, 1–22.
419. H. W. Liu, D. S. Liu and L. X. Zheng, Livest. Sci., 2014, 161, 101–108.
420. Y. Mamindy-Pajany, S. Sayen, J. F. W. Mosselmans and E. Guillon, Environ. Sci. Technol., 2014, 48(13), 7237–7244.
421. Y. Abe, Y. Iizawa, Y. Terada, K. Adachi, Y. Igarashi and I. Nakai, Anal. Chem., 2014, 86(17), 8521–8525.
422. J. L. Sun and L. Q. Luo, Chin. J. Anal. Chem., 2014, 42(10), 1447–1452.
423. M. Chrysochoou and C. P. Johnston, J. Hazard. Mater., 2015, 281, 87–94.
424. S. W. T. Price, K. Ignatyev, K. Geraki, M. Basham, J. Filik, N. T. Vo, P. T. Witte, A. M. Beale and J. F. W. Mosselmans, Phys. Chem. Chem. Phys., 2015, 17(1), 521–529.
425. K. Kaur and R. Mittal, Eur. Phys. J.: Appl. Phys., 2014, 68(3), 6.
426. Q. Li, H. T. Zheng, Z. L. Zhu and Z. Y. Tang, Anal. Lett., 2014, 47(5), 843–854.

---

Footnote

† Review coordinator.

---