Atomic spectrometry update – a review of advances in environmental analysis

Jeffrey R. Bacon *a, Owen T. Butler b, Warren R. L. Cairns c, Olga Cavoura d, Jennifer M. Cook e, Christine M. Davidson f and Regina Mertz-Kraus g
a59 Arnhall Drive, Westhill, Aberdeenshire AB32 6TZ, UK. E-mail: bacon-j2@sky.com
bHealth and Safety Executive, Harpur Hill, Buxton, UK SK17 9JN
cCNR Institute of Polar Sciences, Università Ca' Foscari, Via Torino 155, 30172, Mestre, Italy
dDepartment of Public Health Policy, University of West Attica, Leof Alexandras 196, 115 21 Athens, Greece
eBritish Geological Survey, Keyworth, Nottingham, UK NG12 5GG
fDepartment of Pure and Applied Chemistry, University of Strathclyde, Cathedral Street, Glasgow, G1 1XL, UK
gInstitut für Geowissenschaften, Johannes Gutenberg-Universität, Becher-Weg 21, 55099 Mainz, Germany

Received 14th October 2020

First published on 6th November 2020


Abstract

In the field of air analysis, highlights within this review period included: new isotopic data for reference dust materials; a new device for preparing and mounting delicate air filter samples for X-ray analysis and use of oxygen-mediated mass-shift ion chemistry for measuring sulfur in aerosol samples by ICP-MS. The use of DGT and ionic liquids for preconcentrating trace elements and other analytes from waters has become more established. The resurgence of interest in As speciation has been noted with extensive reviews as well as methods for the challenging determination of thioarsenic species. Improvements in methodologies have achieved LODs with GC-MS that were once only possible with ICP-MS, thereby making it possible for a larger number of laboratories to undertake speciation analysis. Field preconcentration methods and hand-held XRF instruments have made it easier to screen contaminated waters, thereby allowing sampling sites to be selected more effectively. There has been renewed interest in atomic emission sources such as the arc, GD and plasmatron for the analysis of plants and soils. The upward trend in publication of LIBS methods continued but many lacked validation through comparison with established methods or analysis of CRMs. There is a clear need for closer collaboration between the physicists driving fundamental developments in plasma spectroscopy and analytical geochemists, who understand the complexities of environmental samples and the requirement to implement robust QC procedures. Interest in multivariate analysis of pXRFS spectra to predict soil properties related to fertility has increased. Much research effort continues to be devoted to characterisation of matrix-matched geological RMs, both synthetic and natural samples, particularly for in situ analysis by microanalytical techniques, such as LA-MC-ICP-MS and SIMS. Such RMs are essential to compensate for matrix effects and need to be available in sufficient quantities to enable interlaboratory comparisons based on the same RM. The increased access to MC-ICP-MS instrumentation, especially in China, is reflected in the diverse range of isotopic systems now being studied. Most studies present incremental improvements to existing separation procedures or measurement protocols.


1. Introduction

This is the 36th annual review of the application of atomic spectrometry to the chemical analysis of environmental samples. This Update refers to papers published approximately between August 2019 and June 2020 and continues the series of Atomic Spectrometry Updates (ASUs) in Environmental Analysis1 that should be read in conjunction with other related ASUs in the series, namely: clinical and biological materials, foods and beverages;2 advances in atomic spectrometry and related techniques;3 elemental speciation;4 X-ray spectrometry;5 and metals, chemicals and functional materials.6 This review is not intended to be a comprehensive overview but selective with the aim of providing a critical insight into developments in instrumentation, methodologies and data handling that represent a significant advance in the use of atomic spectrometry in the environmental sciences.

All the ASU reviews adhere to a number of conventions. An italicised phrase close to the beginning of each paragraph highlights the subject area of that individual paragraph. A list of abbreviations used in this review appears at the end. It is a convention of ASUs that information given in the paper being reported on is presented in the past tense whereas the views of the ASU reviewers are presented in the present tense.

General reviews applicable to all areas of environmental analysis included a tutorial review7 (67 references) on the fundamentals and new approaches to calibration in atomic spectroscopy. In the context of direct solid sampling8 (255 references), recent advances in different techniques such as FAAS, ETAAS, HR-CS-AAS, ETV, LA, LIBS, XRFS, GD AES and MS and arc/spark AES, were discussed.

2 Air analysis

2.1 Sampling techniques

A new disposable personal air sampler for workplace aerosol sampling, designed to match the ISO/CEN/ACGIH inhalability convention, incorporated9 design features from two existing and widely-used samplers – the closed face cassette and IOM samplers. Initial wind tunnel evaluations were promising although further refinements are required to improve sampling efficiencies for particles >40 μm in size. The ultrasonic personal aerosol sampler, reviewed previously,1 was upgraded10 to include interchangeable sampler inlet options for either respirable- or thoracic-particle sized sampling. Cyclonic samplers, commonly used to sample respirable-sized particles, can demonstrate sampling biases due to particle-size dependencies and can overload if high dust concentrations are sampled for extended time periods. A new virtual cyclonic design demonstrated:11 a better fit with the ISO/CEN/ACGIH respirable convention curve; better ability to handle high dust loadings; and the ability to run at flow rates of up to 21.5 L min−1. The ability to collect more sample mass per unit time and thereby improve method sensitivities will become more important as regulatory workplace exposure-limits for many chemical agents become lower. An apparatus for measuring aerosol deposition in lungs, fabricated12 by 3D printers, simulated the human tracheobronchial airway to provide a better understanding of the potential for deposition of welding fume in welders’ airways. In simulated exercises, 9–31% of welding fume particles in the size range 10–100 nm was deposited in this simulated airway.

Sampling the regulatory PM2.5 and PM10particle size fractions in ambient air requires air sampler devices to meet performance specifications embodied in particle-size-collection efficiency curves that are characterised by their cutpoint (d50) and the steepness of the curve (σg). Whereas performance specifications for PM10 samplers are well established, those for PM2.5 samplers are less so. A Chinese-US research collaboration concluded13 that for an ideal PM2.5 size separator, the cutpoint tolerance should be 2.5 ± 0.2 μm and σg should be ≤1.3, whereas for a PM1 size separator, an emerging measurand of interest, values of 1.0 ± 0.02 μm and ≤1.2, respectively, were suggested. A cascade air sampler device, designed to sample PM10, PM2.5, PM1 and PM0.5 particles simultaneously, was modified14 to collect the residual PM0.1 particle size fraction uniformly on a filter. The aim was to facilitate either direct-on-filter analysis using beam techniques such as XRFS or to enable a representative portion of filter to be removed for chemical analysis.

The development of new air-sampling devices included an electrostatic precipitator (ESP) sampler15 for dust collection at a high flow rate and an ambient air sampler16 for dust collection at a low flow rate. Deployment of the ESP sampler over monthly intervals enabled sufficient mg quantities of indoor dusts to be collected for analysis. Similarly, the ambient air sampler collected sufficient sample mass for analysis when deployed over 1–2 month sampling intervals. The authors concluded that as only low flow rates (0.5 L min−1) were required, such samplers could be produced at relatively low cost allowing widespread deployment for sampling.

Deposition sampling can be an alternative approach to pumped sampling for collecting airborne particles and is attractive because it requires no power. Concerns about sampling performance still persist, however. A comparison of dry deposition samplers, based on17 single-particle SEM characterisation of collected dust, concluded that differences in deposition rates between sampler types did indeed exist. A side-by-side comparison exercise, in which passive UNC samplers were worn by mine workers alongside pumped respirable air samplers as reference, demonstrated18 that there was a ca. 30-fold over-estimation of sampled particle mass. It was concluded that performance might be improved by adjusting the positioning of the UNC sampler on the worker and grounding the device so as to minimise static charging.

2.2 Reference materials and calibrants

Small variations in the stable isotopic compositions of certain metals can identify the provenance of certain airborne dusts. Studies that provided new isotopic data for reference materials included one19 on the isotopic composition of Hg in NIES CRM 28 (urban aerosols) and one20 on new isotopic data for Cu, Fe, Nd, Pb, Sr and Zn in IRMM CRM BCR-723 (road dust) and in Powder Technologies RM ATD (Arizona test dust). Complementary to this work, the Cu, Pb and Zn isotopic compositions in representative windblown mineral samples collected from both African and Asian dust events were determined and tabulated.21

In a continuation of work reported in last year’s ASU,1 two approaches for the assignment of elemental values to new candidate filter-based (thin-layer) XRFS calibrants were evaluated.22 Consensus values, derived from an interlaboratory comparison exercise, agreed with a priori values determined during batch homogeneity testing of filter samples.

The reliability of monitoring mercury in stacks is constrained by the availability of accurate and metrologically traceable calibrants. Researchers at NIST,23 in collaboration with colleagues from the US EPA, devised a new calibration chain with traceability to the SI. This work involved certifying the Hg0 output of a calibration-gas generator (termed NIST prime) as the primary calibrant that then could be used to cross-calibrate vendor calibrant systems (secondary calibrant) by side-by-side testing conducted at NIST. This gave vendors a means of conferring traceability to their customer units (tertiary calibrant) installed alongside Hg0 analysers at stacks. Certification of this NIST prime unit involved sampling defined volumes of Hg0 on sorbent tubes and analysis using an accurate and precise ID-cold vapour-ICP-MS method (see Section 2.4.2.1). Traceable calibrants are also required for Hg2+ species, which can be emitted from combustion processes together with Hg0. Within the MercOX project, funded under the European EMPIR programme, two new calibrant systems were evaluated.24 In both systems, a Hg2+ solution was dosed into a heated chamber at a constant rate and the resultant gaseous Hg2+ species diluted with air to provide the required gas calibrant standard. If there were no Hg losses within the system, the resultant Hg concentrations could be calculated from the Hg concentration in the starting solution, the liquid dosing rate and the gas dilution flow rates. Further tests were planned, in which defined volumes of gas standards would be collected on sorbent traps and analysed, to check the accuracy of the theoretical outputs for these two prototype devices.

In an innovative approach, NO calibration gas was generated25 on demand by the photolysis of N2O, which is readily available in disposable gas cartridges used, for example, in the catering industry as an aerosol propellant for generating whipped cream. By combining this new device with an existing O3 calibrant generator, NO2 gases could be generated using the stoichiometric gas-phase reaction of NO with O3 to result in a new portable device that could calibrate NO, NO2 and O3 air monitoring instruments up to 1000, 500 and 1000 ppb respectively.

2.3 Sample preparation

The solubility of trace metals deposited into the oceans from aerosols is a key factor in phytoplankton growth. In an evaluation (97 references) of published leaching protocols, it was concluded26 that harmonised testing guidelines were required to better understand the biogeochemical impacts of such depositions because comparing solubility data generated in different studies by different research groups using different analytical protocol remained difficult.

Sample losses or contamination are omnipresent challenges when preparing air filter samples for trace element analysis. A substantial 6–10× increase in Hg recoveries from air filter samples was noted27 when lithium tetrathiafilvalene carboxylate (LICTTF) was added prior to MAD. The average airborne particulate Hg concentration determined for filter samples by ICP-MS was 0.4 ng m−3 with addition of LICTTF and 0.05 ng m−3 without. The BCR sequential extraction procedure, originally devised for soil samples, was adapted28 for the fractionation of PTEs in airborne particles collected on air filter samples. Simulant test samples were prepared by spiking air filters with small test portions (ca. 63 mg) of IRMM BCR CRM 701 (lake sediment), a CRM certified for its extractable analyte content using the BCR protocol. An endogenous Zn contaminant was noted, consistent with air sampling by glass-fibre based filters, which are notorious for being contaminated with Cu, Fe and Zn. After subtraction of results for blank filters, elemental fractionation patterns mirrored the certified patterns, and total elemental recoveries (combining elemental data from each of the 4 leaching steps of the BCR protocol) were 84–113% of the certified values.

A new automated system enabled29 radiocarbon analysis of carbonaceous aerosols collected on filters to be performed more efficiently. The initial step involved combustion, in which organic or elemental carbon moieties were converted to CO2 by heating filter pieces to defined temperatures in a stream of either He or O2 carrier gases, with on-line verification of complete combustion by NDIR spectroscopy. Combustion impurities such as NOx, SO2 and halogens were removed together with H2O. The purified CO2 was then cryogenically collected in glass ampoules under N2 and sealed for future off-line 14C measurement. The system was capable of preparing very small samples of 10–50 μg C. The rapid 235U/238U analyses of material deposited on cotton swabs was possible30 using an Advion® plate express reader on an Orbitrap® MS instrument equipped with a liquid-sampling GD microplasma ionisation source. Rapid (30 s) desorption of U species from swab samples into a stream of 2% (v/v) HNO3 was possible. The measured 235U/238U ratios (0.053–1.806) were accurate to within <10%. The system portability had potential application in nuclear non-proliferation survey exercises.

For air sampling purposes, a stretching ring is required to keep a PTFE filter taut and hence flat, but the thickness of the ring can subsequently impede the irradiation of a filter by X-rays when analysed. It would be preferable to remove this ring, but the PTFE filters would then deform and so present an uneven surface for analysis. A new sample preparation device, termed the Smart Store®, enabled31 a PTFE air sampling filter to be prepared for TXRFS analysis by removing this support ring and encapsulating the filter between two sheets of laminated polythene film to keep it flat. There was some signal attenuation due to absorption of X-rays by this film, but nevertheless this device offers potential for labour-saving in the preparation of delicate air filters for analysis.

2.4 Instrumental analysis

2.4.1 Atomic absorption and emission spectrometries. Two approaches for the speciation analysis of mercury in flue gases were evaluated.32 In the off-line approach, a gold-coated alumina sorbent for trapping Hg0 and an alumina sorbent for trapping oxidised forms of Hg were coupled in sequence for analysis by TD-AAS. In the on-line approach, flue gases were introduced directly into a quartz atomiser via a heated transfer line from the stack. The ability to vary the cell temperature enabled Hg0 to be determined selectively at ambient temperature, but the determination of oxidised Hg species required 900 °C to facilitate dissociation to Hg0. The former approach offers better sensitivity because preconcentration is involved, useful for benchmarking against regulatory emission limit requirements, but the latter approach offers speed because near instantaneous measurements were possible, useful for checking performance of Hg trapping technologies installed on stack.

The rapid elemental profiling of individual particles sampled from air was possible33 by combining bright-field microscopy with LIBS. An ultra-thin polythene film had the following advantages: it usefully immobilised selected particles for interrogation, vaporisation of the thin film was rapid so quenching effects were minimised, and the simple polymer composition minimised the potential for spectral artifacts. Measurement of unburnt C in fly ash can be a useful indicator of combustion efficiency within coal-fired power plants. A new two-stage cyclone enabled34 better C measurements by on-line LIBS. Fly ash and CO2 were separated effectively from a combustion gas stream. Application of a plasma-temperature correction-protocol that involved interrogation of Mg II/I emissions provided results that agreed well with those for ash samples tested off line.

2.4.2 Mass spectrometry.
2.4.2.1 Inductively coupled plasma mass spectrometry. A new ICP-MS/MS procedure for determining 67 elements in size-segregated particulate matter collected on air filters was reported.35 Of particular interest was the successful determination of S as SO+ at m/z 48 by using mass-shift ion chemistry involving O2. The S recoveries were 96 ± 4% when aliquots of NIST SRM 1648 (urban particulate matter) were analysed. A new method for determining 236U/238U isotope ratios also employed36 O2-mediated mass-shift ion chemistry to reduce significantly the 235U1H+ isobaric interference on 236U. By cleverly exploiting the fact that the hydride form of UO+ (UOH+) is less prone to formation in the plasma than UH+, together with the use of a desolvating nebuliser, it was possible to constrain the 235U16O1H+/235U16O+ formation rate to ca. 10−7 so 236U/238U ratios of <10−8 could be determined successfully for material deposited from the Fukushima Daiichi nuclear incident.

Siloxanes in gaseous fuels, even at low concentrations, can be problematic because, upon combustion, amorphous Si can deposit and cause damage within combustion systems or fuel cells. The LOQ for Si of ca. 0.01 mg m−3 for a new GC-ICP-MS approach to the analysis of fuels was37 below the benchmark limit of ≤0.1 mg m−3 designed to protect machinery. The Hg0 concentration output from the NIST prime calibrator (see Section 2.1) was certified38 using ID-CV-ICP-MS at selected span points over the range 0.25 to 38 μg m−3. Two procedures were used, a direct gas analysis approach and a preconcentration method that involved trapping defined gas volumes on activated carbon. The direct measurement approach yielded expanded MU ranging from 5.5% at 0.5 μg m−3 to 1% at 38 μg m−3 with a LOQ of 0.06 μg m−3, whereas sample preconcentration yielded an expanded MU of 1% across this range with a LOQ of 0.001 μg m−3.

The single particle ICP-MS analysis of atmospheric particles deposited in ice-core samples was performed39 for the first time using CFA coupled to ICP-TOF-MS. The fact that Al and Mg signals were associated with Fe signals emanating from Fe-rich particles suggested that clay minerals such as illite were the dominant components in the particles being examined. Use of a dry aerosol provided40 a significant gain in ion extraction from a plasma thereby making it possible to now size silver and titanium NPs at 3.5 and 12.1 nm, an improvement of 29 and 37% over that achievable under wet plasma conditions.

In the LA-MC-ICP-MS determination of isotopic ratios in sub-μm sized UOx particles a new small-dead-volume ablation-cell produced41 a better S/N, but alas a comparable improvement in precision was not achieved, suggesting that there was an unknown source of measurement imprecision. Upon investigation, the authors found that some sample ions were not being detected within a measurement window, which they called detector “blind time”. This issue was traced to the design of the instrumental data acquisition system, originally conceived to integrate the steady state ion signal arising from a continuous liquid sample nebulisation process, rather than rapid transient signals arising from the laser ablation process. By setting a signal integration window of 500 ms, this “blind time” effect was minimised. Development of new fast data acquisition systems for LA applications can therefore be anticipated.


2.4.2.2 Other mass spectrometry techniques. A novel real-time system employed42 an extractive ESI source coupled to a HR TOF-MS instrument for determining water-soluble metals present in aerosols. Laboratory experiments involved nebulisation of simultant metal-EDTA-chelate test samples as dry aerosols using either an in-house fabricated nebuliser with a silica gel drier or a commercially available desolvating HEN. The resultant ions were detected using negative ionisation. Preliminary results were encouraging with good linear mass responses, low ng m−3 LODs and a fast single-second response. An ultrafine-particle concentrator employed43 a water-based condensation tube to increase the size of small 10 nm-sized particles with a solvation shell so ensuring their effective capture prior to analysis by aerosol MS. Inlet sampling was configurable between 1 and 1.7 L min−1 with a resultant output flow into the aerosol MS of between 0.08 and 0.12 L min−1. As a consequence, aerosol samples were concentrated 8 to 21 times. Initial field experiments were carried out with an aerosol MS set-up equipped with a switching valve, so that aerosols could be sampled either through this new preconcentrator inlet or through a standard inlet. No alterations in the particle chemistry were observed for the new system.
2.4.3 X-ray spectrometry. Collection of size-segregated PM10 particles directly onto quartz reflectors acting as impaction plates facilitated44 subsequent analysis by TXRFS. Absolute mass-response calibration was performed using Si wafers impregnated with 4 or 8 ng Cr (±5%). Analysis of a Si wafer spiked with known aliquots of a multi-elemental standard allowed RSFs (relative to Cr) to be determined thereby alleviating the need, common with many TXRFS applications, for an IS to be spiked into each sample. The LODs of, for example, 0.09 (As), 0.24 (Ni) and 0.20 (Pb) ng m−3 for a 1 m3 sample volume and 3000 s integration made the procedure suitable for EU ambient air regulatory measurement purposes, for which the limit values (yearly averages) are 6 (As), 20 (Ni) and 500 (Pb) ng m−3. Furthermore, this inherent method sensitivity made it possible to monitor short-lived pollution episodes at high time-resolution. Further planned developments included a portable TXRFS instrument for measurements in the field.
2.4.4 Combustion-based techniques. Thermal-optical combustion-based measurements are widely used to classify carbonaceous aerosols captured on filters from measurements of their organic (OC) and elemental (EC) carbon moieties and, by summation, their total carbon (TC) content. Whereas, in general, results are in agreement for TC measured using various combustion protocols and/or instruments, results for EC can vary greatly because OC can be misclassified as EC. Accurate EC measurements are required because EC-containing particles contribute to climate change through radiative forcing effects. In addition, EC in the workplace is used as a marker of exposure to diesel fumes. An interesting comparision45 of EC and OC mass measurements was conducted concurrently by three long-term monitoring networks (the US IMPROVE and the Canadian CAPMoN and CABM networks) at one North American ambient-air monitoring site. Whereas the values for EC, OC and TC measured by IMPROVE were 5–75% higher than those measured by CAPMoN, they were 15–80% lower than those for CABM. The authors concluded that regular inter-comparisons between monitoring networks should be made and, additionally, that there is a need for new filter-based RMs because different operationally-defined procedures used in these networks give different answers. The formation of so-called pyrolysed carbon arising from OC charring during the combustion process is one reason why such measurement challenges exist. Upon further heating, pyrolysed carbon can be oxidised to CO2 and contribute to measured CO2 values arising from EC combustion, hence generating a false-positive EC value. A detailed report46 described the structural changes in simulant carbonaceous samples prepared with a soot generator during a typical thermal-optical measurement cycle.

A two-step thermal-oxidative analysis enabled47 EC and TC species emitted from turbine engines to be determined. The established NIOSH 5040 thermal-optical method requires filters to be sectioned so the ability of this new procedure to analyse a complete filter was a distinct advantage as it eliminated any potential sample-heterogeneity issues. Another advantage over the NIOSH approach was the ability to analyse filters with high carbon mass loadings (>90 μg cm−2) such as those expected from the direct sampling of engine exhausts. A constraint, however, was that OC was not measured directly but by the difference between TC and EC. For many tailpipe measurements, this may not be an issue as freshly emitted particles typically contain high EC/OC ratios (e.g. 9[thin space (1/6-em)]:[thin space (1/6-em)]1), but it would be a disadvantage for studies of aged carbonaceous emissions with lower EC/OC ratios (e.g. 1[thin space (1/6-em)]:[thin space (1/6-em)]1) that can occur over time upon mixing with other organic pollutants in the atmosphere, resulting in an increased potential for the formation of a pyrolytic carbon artifact.

Optical methods are attractive for measuring black carbon in carbonaceous aerosols because they can be faster alternatives to the laboratory-based combustion approaches and can be made portable. For quantification purposes, however, knowledge of the light-absorption properties of black carbon are required, so tabulation48 (63 references) of reported mass absorption cross-sectional values was timely. Lower cost alternatives to the expensive black-carbon sensors currently available were based on the use of cameras. In one approach, black carbon on filters was estimated49 using a smartphone camera and a calibration algorithm facilitated by the analysis of 1878 filter samples for which reference black carbon values were available. When used in the field, this smartphone approach demonstrated good predictability against reference black carbon on filter measurements with an R2 of 0.904 and a coefficient of variation (RMSE) of 25.3%. In another study, black carbon values estimated from filter images taken with a digital camera correlated50 well (normalised RMSE <10%) with data derived from two reference black carbon measurement approaches (smokestain reflectance and hybrid integrating plate and sphere methods) and from an EC measurement method.

The measurement of workers’ exposure to carbonaceous aerosols remains topical. Aethalometer-derived black carbon measurements of large-diameter (50–80 nm) CNTs sampled onto filters were51 comparable to reference EC on filter measurements by NIOSH method 5040, but black carbon measurements were lower than EC measurements when small-diameter (<8 nm) CNTs were assessed. Instrumental drift also occurred when the aethalometer was challenged with elevated CNT airborne concentrations (>30 μg m−3) that could arise from activities such as cleaning and powder bagging. Continuous monitoring of diesel particulate matter (DPM) in underground mine air is desirable to assess workers’ exposure to pollution. An evaluation52 of four DPM monitoring devices (the FLIR Airtec monitor, the Magee Scientific AE33 aethalometer, the Sunset Laboratory OC–EC field analyser and Airwatch, a new prototype continuous-monitor based upon the Airtec monitor) in a laboratory and in a mine setting concluded that both the AE33 and the Sunset instruments had potential for unattended monitoring underground provided that there was reliable access to power and that periodic routine maintenance was performed. The prototype Airwatch instrument had potential but required further development to achieve reliable and consistent measurements.

2.4.5 Other techniques. Advances in vibrational spectroscopic techniques offer potential for measuring airborne pollutants. An innovative optical-trapping Raman method examined53 single suspended CNTs, an approach that eliminates measurement artifacts arising from sampling onto filters. A Raman method for the analysis of respirable crystalline silica (RCS) collected on silver filters had54 a LOD of 0.26 μg, which was an order of magnitude better than that achievable by established XRD and FTIR approaches. In a continuation of work, reported in last year’s ASU,1 on measuring mine workers’ personal exposure to RCS by portable FTIR, researchers at NIOSH demonstrated55 that results for filter samples analysed by four commercially available instruments varied by <5%. Variations in average filter measurements, conducted over an extended time period, were not statistically significant.

3 Water analysis

3.1 Certification of reference materials and metrological investigations

An interlaboratory comparison assigned56 values for B, Cs, Ga, Ge, Hf, Li, Nb, P, Rb, Re, Rh, S, Sc, Se, Si, Sn, Th, Ti, Tl, W, Y, Zr and REEs in the NRCC CRM SLRS-6 (river water) together with an indicated Sr isotopic composition by mainly ICP-AES and ICP-MS. The concentrations were generally lower than those reported previously for the SLRS-5 RM.

A full validation approach was applied57 to the ID-cold vapour-ICP-MS determination of Hg in marine biota, sediments and coastal waters. The blanks, selectivity, working range (1.2–240 ng kg−1), linearity (0.9991), recovery (97–103%), repeatability (<2.5%), intermediate precision (<3.5%), LOD (0.72 ng kg−1) and LOQ (1.10 ng kg−1) were systematically assessed following ISO/IEC 17025 and Eurachem guidelines. The relative expanded uncertainty of the total Hg mass fractions in coastal seawater samples was 27.2–32.8%. Recovery correction for the ID spike contributed 60–75% to the uncertainty budget because the spiked IRMM CRM BCR 579 (coastal seawater) itself had a relative expanded uncertainty of 26%. Another significant contribution (10–35%) to uncertainty was correction of the procedural blanks, which was particularly important for the low Hg-mass-fractions typically found in coastal seawaters. The repeatability of the CVG ICP-MS measurements in seawater samples was 5–10%.

3.2 Sample preconcentration

The use of DGT samplers as passive solid-phase preconcentration devices in the measurement of labile trace metal concentrations in marine environments was reviewed58 (50 references). These devices have even been successfully mounted on autonomous underwater vehicles to probe labile metal concentrations in coastal and open ocean waters. Other uses of DGT samplers and significant advances in the SPE of trace elements from water samples are summarised in Table 1.
Table 1 Preconcentration methods using solid-phase extraction for the analysis of waters
Analytes Matrix Substrate Coating or modifier Detector LOD in μg L−1 (unless stated otherwise) Validation Ref.
AlIII, CoII CrIII, CuII, FeIII, MnII, NiII, PbII, ZnII Water 0.1–0.2 mm particle size silica gel Linear polyhexamethylene guanidinium and 1-nitroso-2-naphthol-3,6-disulfonic acid or 2-nitroso-1-naphthol-4-sulfonic acid ICP-AES 0.75 (Cu) to 1.35 (Al) Sample spike recovery 319
AgI, CdII, CoII, CrIII, CuII, NiII, PbII Environmental water samples Fe3O4-GO NPs SiO2 ICP-MS 2 (Co) to 14 (Ag) ng L−1 Sample spike recovery and CRM GSBZ 50009-88 (environmental water) 320
Ag, Cd, Cr, Cu Pb River water Gold NPs DDTC LIBS 1.5 (Cu) to 4.5 (Cd) Spike recovery and comparison with ICP-AES results 321
AsIII Ground, river, waste and drinking water Carbon sheets MnFe2O4 NPs ICP-AES 0.03 Spike recovery and NIST SRM 2669 (frozen human urine) 322
AsV River water MWCNTs Branched polyethyleneimine ICP-MS 0.05 Spike recovery and comparison with HPLC-ICP-MS results 323
Ba, Cd, Co, Cu, Mn, Ni Lake and natural water samples Silica gel N,N′-bis(4-methoxysalicylidene)-1,3-propanediamine ICP-AES 0.19 (Ni) to 0.36 (Cu) NWRI CRM TMDA-53.3 (fortified Water) and IRMM ERM-CA022a (soft drinking water) 324
Cd Petroleum production waters Low-density polyethylene semipermeable membrane DDTC complexes in solution ETAAS 0.08 NMIA CRM MX014 (acidified coastal seawater) 325
CdII, CoII, CrIII, CuII, FeIII, MnII, NiII, PbII River water SiO2 Ta2O5 FAAS 5.3 (Cd) to 56.0 (Cu) Spike recovery 326
CdII, CoII, PbII, PdII Well water, wastewater, soil MWCNTs Glutaric dihydrazide FAAS 0.12 (Pb) to 0.19 (Cd) Spike recovery 327
Cd, Cu, Pb plus carbamate and triazole pesticides Water In pipette tip polymer-based monoliths from allylthiourea and 1-allyl 3-methylimidazole difluoromethanesulfonylamide salt ETAAS 0.13 (Cd) to 1.1 (Pb) ng L−1 Spike recovery and comparison with ICP-MS results 328
CrIII, CuII PbII, ZnII Lake, river, spring and mineral water as well as seawater GO Modified by reaction between in situ mixed phosphoric–carboxylic anhydrides and Na2S with the graphene oxide surface EDXRFS 0.06 (Cu) to 0.10 (Cr) Spike recovery and NIST SRM 1640a (natural water) 329
CrIII, SbIII Environmental waters Carboxyl-functionalized organic–inorganic hybrid monolithic column ICP-MS 0.004 (CrIII) and 0.002 (SbIII) Spike recovery 330
Dissolved Cu fraction Seawater Polypropylene Accurel® PP S6/2 hollow fibres Di-2-pyridylketone benzoylhydrazone ETAAS 0.62 nmol L−1 IRMM CRM BCR-403 (seawater (trace elements)) 331
Hf, Nb, Ta, Zr Seawater NOBIAS® Chelate-PA 1 ICP-SF-MS 0.0008 (Ta) to 0.9 (Zr) pmol kg−1 Spike recovery and NRCC CRMs CASS-5 and 6 (near shore seawater), and NASS 5 and 6 (seawater) 332
Hg Lake and river water Selective laser sintering 3D printing with a mixture of polyamide-12 powder with thiol-functionalized silica ICP-MS 0.037 ng L−1 Spike recovery 333
Hg Seawater and river water Quartz glass dip sticks coated with 3-aminopropyltrimethoxysilane and 300 nm diameter SiO2 particles Gold NPs Thermal desorption AFS 0.18 ng L−1 Spike recovery and IRMM CRM ERM-CA400 (seawater (Hg)) 334
Hg, MeHg Tap and river waters Fe3O4 NPs Nanocellulose GC-pyro-AFS 5.6 pg mL−1 (Hg) and 4.0 (MeHg) pg mL−1 Spike recovery 335
Hg, MeHg, EtHg Tap and river waters Silica particles GO HPLC-ICP-MS 5 (Hg2+) to 9 (EtHg) pg L−1 Spike recovery 336
Nd/Th ratio Seawater Nobias® Chelate-PA1 ICP-SF-MS Blank reported as <10 pg of 232Th for 10 L of sample Internal RM BATS2000A 337
Pd Environmental water Mesoporous silica 3-Mercaptopropyltrimethoxysilane ETAAS and ICP-MS/MS 0.06 (ETAAS) or 0.2 ng L−1 (ICP-MS/MS) Spike recovery 338
Pd, Pt, Rh Natural waters DGT samplers Comparison between Purolite® S914, S920 and S985 resins ICP-MS MDL (14 day deployment) 0.007 (Pt) to 0.153 (Pd) ng L−1 Spike recovery 339
SeIV, TeIV Environmental water SiO2-coated Fe3O4 NPs Polyaniline ICP-MS 1.2 (Te) and 5.3 (Se) ng L−1 RM GBW(E)080548 (Te in water) and CRM GSBZ 50029-94 (environmental water) 340
TlI and TlIII River water AG1-X8 SAX resin DTPA (to complex TlIII) ICP-MS LOD not reported lowest sample concentrations reported 0.01 (TlI) and 0.22 (TlIII) Spike recovery 341
U Water and seawater GO TXRFS 0.04 Spike recovery 342


In their review (265 references) on the use of ionic liquids for liquid-phase extraction of trace analytes, Pletnev et al.59 defined ionic liquids as “salts being liquid at room temperature or not very high temperatures”, “not very high” was considered to be ≤100 °C. For water analysis only those that are liquid at room temperature are of interest. The expanding use of ionic liquids was confirmed60 by a review (130 references) on the use of ionic liquids and ionic-liquid-modified sorbents for the preconcentration of heavy metal ions and organic pollutants from water samples.

The most significant advances in the use of liquid-phase extraction, published in period covered by this ASU are summarised in Table 2.

Table 2 Preconcentration methods using liquid-phase extraction for the analysis of waters
Analytes Matrix Method Reagents Detector LOD in μg L−1 Method validation Ref.
Cd Ground water DLLME APDC and the DES ZnCl2[thin space (1/6-em)]:[thin space (1/6-em)]CH3CONH2 FAAS 0.046 Spike recovery and NIST SRM 1643e (trace elements in water) 343
Cd, Co, Ni, Pb High salinity oilfield production water DLLME DDTC, CH3OH and CCl4 ICP-AES 0.003 (Co) to 0.15 (Pb) Comparison with ICP-MS results and NRCC CRM NASS-5 (seawater) 344
Cd, Pb Ground and treated waters; hair DLLME L-Cysteine (2-amino-3-sulfhydrylpropanoic acid), 1-butyl-3-methylimidazolium hexafluorophosphate and hexafluorophosphate FAAS 0.05 (Pb) and 0.13 (Cd) NIST SRM 1643e (trace elements in water) and IRMM CRM BCR 397 (trace elements in human hair) 345
Cd, Zn Water and fruit juice Air-assisted LLME Sorbitol, menthol and mandelic acid FAAS 0.12 (Zn) and 0.15 (Cd) RM SPS-WW2 batch 108 (wastewater) 346
Cr Water LLME Thiomalic acid and ferric chloride in ethylene glycol medium FAAS 1.18 NWRI CRM TMDA-53.3 (fortified water) 347
In Lake water In syringe LLME DDTC, chloroform SQT-FAAS 19.2 Spike recovery and check against ICP-AES results 348
Ni Water, food and tobacco DLLME 1-Hexyl-3-methylimidazolium tris(pentafluoroethyl)trifluorophosphate and quinalizarin FAAS 0.6 NWRI CRM TMDA-51.3 and TMDA-53.3 (fortified water) 349
Pd, Pt Tap, fresh, and saline waters CPE Triton X-114, 2-mercaptobenzothiazole and SnCl2 ICP-AES 0.53 (Pd) and 0.75 (Pt) Spike recovery 350
Ti, TiO2 NPs Water and swimming pool water CPE Triton X-114 and NaCl ICP-MS 0.13 Spike recovery 351


3.3 Speciation analysis

A comprehensive review of arsenic speciation (240 references) in environmental samples, including surface, ground, and geothermal waters, covered61 the period 2004–2018. Although LC-ICP-MS was by far the most widely used technique, other techniques such as those involving XRF detection were increasingly reported. The review covered the analytical process from sample preparation to measurement and is recommended for all workers in the field of environmental analysis.

Aluminium complexes in water were investigated62 using CE with ESI-MS and ICP-MS detection. Both detectors were most sensitive at pH 3. Peaks were identified by ESI-MS. Using ICP-MS under cool plasma conditions, the peaks were baseline resolved so an Al LOD of 0.037 μM was achievable for a 35 nL injection.

The capabilities of modern instrumentation for arsenic speciation analysis in waters was demonstrated63 using a standard Hamilton PRP-X100 SAX column with diluted-phosphate-buffer gradient-elution to decrease blank values. The LODs were as low as 0.01 ng L−1 for arsenobetaine and 0.35 ng L−1 for AsV when the ICP-MS instrument was operated in KED mode. Analysts from the US EPA solved64 the problem of thioarsenite detection in groundwater by matching the pH of the eluent to that of the sample to minimise species transformations due to proton transfers. It was possible to resolve the differences between previous XAS and chromatographic data, thereby progressing our understanding of arsenic behaviour in sulfate-reducing environments.

The detection of perchlorate in river water by HPLC-ICP-MS was improved65 10-fold by using ICP-MS/MS instead of a single quadrupole instrument. The instrumental LOD for Cl was 0.3 μg L−1 when freeze drying was used to preconcentrate the sample. A method LOD of 50 ng L−1 for Cl was possible in the absence of large concentrations of carbonate, making the technique competitive with the standard HPLC-ESI-MS/MS method whilst avoiding the need for isotopically labelled ISs.

In order to make the US EPA Method 1630 for methyl mercury in water more easily applicable to several Hg species in petroleum production waters, it was modified66 by including an ultrasonic CPE step with an acidified solution of Triton X-114 (0.5% w/v, pH 3) and by using propylation rather than ethylation for derivatisation. This sample preparation eliminated reagent interactions with oil-based constituents in the matrix thereby preventing artefact formation during derivatisation and distillation. With GC-cold vapour-AFS detection, the LODs for mercury species were 5 (Hg2+), 8 (CH3Hg) and 11 (CH3CH2Hg) pg L−1 in this type of sample, which is generally difficult to analyse.

It is pleasing to note progress in the determination of marine biogenic activity tracers and their reaction products in environmental matrices, as they are used extensively as proxies for past sea ice cover. Low-pressure IC was coupled67 with MC-ICP-MS to follow the oxidation of 129I spikes in seawater to iodate. Use of a commercially available sparging interface, in which volatile I2 was generated after the addition of HNO3 as an oxidant, increased I transport to the plasma ten-fold in comparison to that for pneumatic nebulisation, thereby making it easier to monitor the 129I/127I isotope ratio. An HPLC-ICP-MS/MS method was developed68 to determine the oxidised breakdown products of dimethylsulfide in seawater after reaction with HOBr. Use of a Hypercarb™ mixed-mode column with a formic acid gradient made it possible to separate dimethylsulfoniopropionate, dimethylsulfoxide, dimethylsulfone, dimethylsulfide, methanesulfonic acid and methanesulfinic acid in under 12 minutes. Detection by ICP-MS/MS with H2 and O2 as reaction gases resulted in LODs that ranged from 1.7 (dimethylsulfide) to 136 (methanesulfinic acid) nM. The reaction rate constants determined could be incorporated in models of the ocean–atmosphere interface.

Tin species were separated69 by TLC on a glass slide coated with an ion-imprinted polymer created by copolymerising N-allylthiourea and ethylene glycol dimethacrylate in the presence of SnII. The two tin species (SnII and SnIV) were completely separated using a mobile phase of 1[thin space (1/6-em)]:[thin space (1/6-em)]1 (v/v) acetonitrile-ethanol at pH 6. After deposition of 10 μL of standard (0.8–900 μg L−1), the plate was scanned using LA-ICP-MS with a 50 μm spot size and a 20 Hz shot frequency. The method LOD was 0.3 μg L−1, sufficient to quantify Sn species in samples from the Caspian Sea, a local river and waste waters.

Although GC-MS methods for the determination of elemental species in water samples have existed for many years, only recently has modern instrumentation been capable of LODs at the level routinely achievable by GC-ICP-MS. Selenocyanate was quantified70 in waste water by GC-MS/MS after derivatisation with triethyloxonium tetrafluoroborate, to generate volatile ethylSeCN, and by back extraction into chloroform. The LOD for Se was 0.1 ng g−1. Using 80Se13C15N as an IS, spike recoveries from sea and river waters were quantitative at the ng g−1 level. Selected organoarsenic compounds were determined71 in mining and waste waters by SPME-GC-MS using 1,3-propanedithiol as a derivatising agent and a 65 μm polydimethylsiloxane divinylbenzene SPME fibre. The LODs of 0.4 to 5.9 μg L−1 were achieved after a 30 minute fibre equilibration time. Mercury species (Hg2+, EtHg and MeHg) were derivatised with sodium tetraphenylborate72 and trapped on a polydimethylsiloxane-coated SPME fibre in the headspace for subsequent GC-MS/MS analysis. For thermal desorption in the injection port at 250 °C, the LODs were 0.03 (MeHg and EtHg) and 6 (Hg2+) ng L−1.

Although most published articles concern known compounds, there is still scope for untargeted and fractionated analysis of trace elements and their compounds. An important example of this approach was73 the determination of perfluorinated compounds in surface and groundwaters. Current control methods for the nearly 5000 per- and polyfluoroalkyl substances with related CAS numbers measure only a limited number of these compounds, so an untargeted screening method for the measurement of total organic F extracted from ground and surface waters was developed. After extraction by SPE and the addition of Ga, F was detected in a CS-ETAAS instrument as molecular GaF, probably at 211.248 nm but this important information was not given. The method LOD was 5.3 ng L−1. Accuracy was verified by analysis of the EC CRMs MISSIPPI-03 and Battle-02 (both river water) certified for total F. Although NPs and colloids can be fractionated by size and easily and routinely detected by FFF-ICP-MS, the dissolved fraction is lost in the cross-flow across the membranes. An FFF instrument was modified74 with an interface designed to convert the discontinuous cross-flow into a continuous flow suitable for ICP-MS analysis. Results for the dissolved fractions of Al, Ca, Mg, P and Si in aqueous standards and water samples compared well with those obtained from discrete sampling and ultrafiltration.

3.4 Instrumental analysis

3.4.1 Atomic absorption spectrometry. The advent of the xenon short-arc lamp as a continuum source for a high-resolution atomic-absorption spectrometer has been the main advance in AAS in recent years. Eskina et al.75 reviewed (173 references) use of this instrumentation and included the analysis of waste and drinking waters amongst all applications considered. The HR scanning ability of this ETAAS instrumentation was exploited76 to remove matrix interferences in the determination of Pb directly in seawater. A mixed Ba(NO3)2–HF modifier suppressed the sulfate interference (by formation of refractory barium sulfate) and removed the chloride interference (by formation of volatile HCl), thereby allowing an LOD of 0.3 μg L−1 to be achieved. Unsurprisingly, the recovery error was only 6% when the accuracy was estimated at the rather unrealistic spike concentration of 50 μg L−1, some 1000 times higher than the level expected in seawater.
3.4.2 Vapour generation. Photochemical reduction of AsV and DMA to AsIII was used77 in the VG-FAAS speciation analysis of arsenic in soils, sediments and surface waters from a heavily contaminated area of Brazil. The LODs of 3.2 (AsIII) to 6.7 (DMA) μg L−1, obtained using ZnO NPs as photocatalysts and optimisation of the pH and formic acid concentration, were considered fit for purpose for waters impacted by mine waste leakage.

There is an increasing focus on bismuth in the environment as it becomes more widely used in industrial processes. The sensitivity of ICP-MS determination of this element was improved78 70-fold for the PVG of volatile (CH3)3Bi after reaction with formic and acetic acids in the presence of Co2+ ions compared with that of pneumatic nebulisation. The LOD was 0.3 ng L−1. A different approach for improving the sensitivity of AFS for Bi involved79 modification of the flame atomiser, optimisation of the optical path of the spectrometer and addition of a 307.1 nm interference filter. The LOD was 0.9 ng L−1 in water. The Bi concentrations of 22 and 24 ng L−1 determined for the NRCC CRMs NASS-7 and CASS-6 (seawater), respectively, for which there are no certified or indicative values, were within the range of expected values for seawater. Accuracy was demonstrated by analysis of NIST SRM 1643f (trace elements in natural water) for which the measured value of 12.8 ± 0.1 ng L−1 (n = 3) was in close agreement with the certified value of 12.62 ± 0.11 μg L−1.

The AFS determination of lead in water samples was achieved80 by forming volatile Pb chelates with ammonium O,O-diethyldithiophosphate. Although the reported VG efficiency of 12% doesn’t seem an advance on the efficiency of traditional HG, the LOD of 1.1 μg L−1 was sufficient to determine Pb in river water samples successfully.

3.4.3 Inductively coupled plasma mass spectrometry. A correction for M2+interferences is required when determining As and Se in REE-rich waters by the US EPA method 200.8. Smith et al.81 considered that corrections based on the “half mass” interferences generated by alternative, non-interfering isotopes (e.g. m/z 71.5 for 143Nd2+) could be more effective than estimating the “whole mass” interference (e.g. on m/z 75 for 150Nd2+) because other, non-REE-related interferences, could lead to an overcorrection in the latter approach. The new approach using single-quadrupole instrumentation reduced the number of overcorrected measurements and gave results closer to those obtainable by ICP-MS/MS or ICP-SF-MS. An additional advantage was that the measured intensities could be used simultaneously to correct for drift and differences in matrix composition.

The boron memory effect during isotope ratio measurements by MC-ICP-MS was significantly reduced82 by adding NaF to the wash solution. A solution of NaF (0.6 mg g−1) in 1% v/v HNO3 was as effective as a 0.3 M HF rinse solution but had no damaging effects on the instrumentation. The B counts returned to baseline levels in <4 minutes after sample analysis.

A method for the preconcentration of mercury from waters without causing any isotope fractionation used83 a 3 L bubbler for CVG before analysis. The sample was transferred to the bubbler and a BrCl solution added to oxidise the Hg species to Hg2+. The subsequent addition of SnCl2 produced Hg0, which was then purged with an Ar stream and trapped on a Cl-impregnated activated-carbon cartridge to give an enrichment factor of 1000. The trapped Hg was desorbed thermally and detected by MC-ICP-MS. Analysis of test samples spiked with either NIST SRM 3133 (mercury (Hg) standard solution) or NIST RM 8610 (mercury isotopes in UM-Almaden mono-elemental secondary standard) gave results that were not significantly different in isotopic composition to that of the original spike. Repeat analyses (n = 16) of real samples gave analytical precisions of 0.06‰ (Δ199Hg) to 0.13‰ (δ202Hg).

Accurate determination of the sulfur isotope ratios of sulfite, sulfate and thiosulfate in waters was achieved84 in one analytical run by coupling IC to MC-ICP-MS. An S-containing IS (trimethylsulfoxide) was added to correct for mass bias, and on-column fractionation was corrected by external calibration. Use of a linear regression slope to calculate the isotope ratios of the transient signals resulted in a combined uncertainty of <0.25‰ for δ34S in solution and a reproducibility of 0.5‰ for an injection of 1 μg of S. An anion-exchange membrane was evaluated85 for the extraction of S from fresh and marine pore waters before determination of S isotope ratios by MC-ICP-MS. Recoveries were >90% when concentrations of competing ions (bicarbonate, carbonate, chloride, nitrate and phosphate) were limited to <0.5 mM per cm2 of membrane. The lowest S concentration detectable without fractionation was 0.5 μM, but the recovery dropped at concentrations >0.01 mM probably as a result of reaching the breakthrough volume for the disc membrane.

3.4.4 Laser methods. A new mercury-specific matrix containing mefenamic acid and thymine allowed86 the detection of mercury ions in solution by MALDI-MS. The Hg2+-containing ions were generated so specifically that there was little background signal from other ions. The stated LOD in water of 1 nM is equivalent to 0.2 μg L−1 yet it was presented erroneously by the authors to be equivalent to 2 μg L−1.

Preconcentration is required to determine trace elements in water by LIBS. By electrospraying aqueous samples onto a heated substrate to generate a detectable solid residue Ripoll et al.87 achieved LODs of 17 (Ni)-57 (Cr) μg L−1. However, the recoveries were poor due to strong matrix effects, so a standard additions calibration was required to improve the recoveries to 91 (Ni)–110 (Cd). Problems of water splashing and plasma quenching in the handling of liquid samples made it preferable to analyse solid samples. In a study into the effect of sample pH after deposition onto different substrates, a series of substrates was investigated88 for the formation of surface precipitates. The formation of a fine uniform ZnCl2 precipitate on the surface when the sample pH was 3.5 to 6.5 and HCl was used as an acidifying agent resulted in more laser energy being transferred to the sample in the precipitate and to improved LODs of 0.009 (Cd) to 0.0006 (Cr) mg L−1 at pH 6.5. The procedure was therefore suitable for the analysis of sewage discharge and other waste waters according to Chinese environmental standards.

3.4.5 X-ray fluorescence spectrometry. The ISO Standard 20[thin space (1/6-em)]289 for total reflection X-ray fluorescence analysis of water was based89 on the analysis of a residue dried on an X-ray reflector and was intended for the routine analysis of large numbers of drinking, ground and surface waters samples. The method could also be applied to the analysis of waste waters after sample dilution.

The use of hand-held or portable XRFS instruments continues to be of interest. The LODs in the low mg L−1 range for 200 μL of mine waters dried onto a filter paper were90 sufficient for screening in the field. Trapping the Pb present in 2 L of tap water onto a disc of carbon felt and detection with a hand-held XRF instrument gave91 an LOD of 15 μg L−1. It was suggested that the method could be used in the field and could be applicable to a wider range of elements (e.g. Ca, Cu, Fe, Mn and Zn).

4. Analysis of soils, plants and related materials

4.1 Review papers

Review articles featuring specific elements included a comprehensive overview of developments61 (239 references) in As speciation analysis in atmospheric particulates, biota, sediment, soil and water published in the period 2004–2018, and a discussion92 (144 references) of As content and speciation in Chinese mushrooms. Butcher93 (60 references) summarised advances in atomic spectrometry methods for the determination of Ca, and also discussed the use of Ca as an IS. Ravipati et al.94 (131 references) described trends in the measurement of Pb in biological and environmental samples from 2000 to 2018. Etteieb et al.95 (128 references) included a comparison of analytical methods in their review of Se as an emerging contaminant in the mining industry, whilst Filella et al.96 (122 references) highlighted that the lack of suitable analytical methods (with sufficiently low LODs) and of CRMs seriously limits current understanding of the environmental chemistry of Te.

An evaluation (75 references) of the scope and limitations of analytical techniques for visualisation of elements in hydrated plant tissue included97 SEM-EDS, TEM-EDS, XFM, PIXE, ICP-MS, SIMS, autoradiography and confocal microscopy with use of fluorophores. This article provides a valuable introduction to these techniques for workers interested in applying them in their own research.

Nawar et al.98 (151 references) showcased developments in the exploitation of different types of spectral information obtained in the field to estimate PTE concentrations in soil. The techniques considered were LIBS, pXRFS and MIR and VIS-NIR spectroscopies. Interest in multi-sensor data-mining has expanded rapidly in recent years, not only to obtain analyte concentrations but also to predict general soil properties such as texture, pH and cation exchange capacity.

4.2 Reference materials

New information on RMs included that for Ba isotope ratios determined99 by MC-ICP-MS for 34 NRCGA geological RMs (rocks, soils and sediments). Methyl Hg concentrations were reported100 for four RMs (one river sediment and three types of biological sample) already certified for their total Hg content. Measurements were performed using species-specific ID and HPLC-ICP-MS. Water-extractable concentrations of 18 elements in the NCS CRM DC 77302 (formerly GBW 07410 (soil)) and the Analytika RMs Metranal-31 (light sandy soil (metals)) and Metranal-33 (clay loam soil (metals)) were provided101 as information values to aid future researchers. Similarly, information values were presented102 for δ114Cd/110Cd in NIST SRMs 2709 and 2709a (San Joaquin soil), 2710 and 2710a (Montana I soil) and 2711 and 2711a (Montana II soil) and IGGE CRM GSS-1 (soil).

4.3 Sample preparation

4.3.1 Sample dissolution and extraction. Various sample-milling methods were compared103 for the ICP-MS determination of Cu, Pb, Sb and Zn in soil from a small-arms range. Concentrations and precision were affected by both the method of milling and grinding time. Whereas puck mill grinding required only 5 minutes to obtain a homogeneous powder, ball mill grinding required 20 h. Ball mill, puck mill and ring and puck mill all gave RSD values <15%, whereas grinding by pestle and mortar or the analysis of un-milled samples gave high RSD values of 28–55% and 17–257%, respectively.

In the eternal quest for improved reagents for digestion and extraction, Santos et al.104 compared acidic (HBF4) and alkaline (NaOH) digestions of plants and fertilisers for the determination of Si by MIP-AES. The LOD of 0.03 g kg−1 obtained with acid digestion was superior to that (0.4 g kg−1) obtained with alkaline digestion. However, the problems arising from the formation of highly corrosive HF had to be addressed through the use of high dilution factors, addition of H3BO3 and an inert sample introduction system. Mokoena et al.105 evaluated dilute HNO3 hot-plate digestion as a means to avoid dilution prior to analysis by ICP-AES. Optimised conditions for 1.0 g of sample were 180 °C, 45 minutes and 10 mL of 5 M HNO3. Accuracies for As, Cd, Cr, Cu, Fe, Pb and Zn ranged from 98 to 111% for Supelco CRMs CRM015 (trace metals – fresh water sediment 2) and CRM052 (trace metals – loamy clay 1). Precision was 1.4 to 5.8%.

The use of mathematical models for optimisation of extraction parameters has increased. Nuapia et al.106 applied response surface methodology to evaluate pressurised hot water for the extraction of nutrients from dried Moringa oleifera leaves for ICP-AES determination. The main factor that influenced the extraction of macronutrients was extraction time, whereas extraction of micronutrients was more markedly affected by increasing temperature, at a constant flow rate (0.3 mL min−1) and pressure (105 bar). The optimum extraction for 5 g samples was achieved at 90 °C for 60 minutes. Recoveries ranged from 22% for Cr to 98% for Ca and K when results were compared with those obtained by MAD in HNO3/H2O2. In an ultrasound-assisted extraction procedure for the determination of Cu, Fe, Mn and Zn in plants by FAAS, the extractant composition was optimised107 by applying a simplex centroid design and the time, power and temperature optimised by applying the Box–Behnken design. Optimum extraction from a 250 mg sample was obtained with 10 mL of a 0.5 M HNO3 + 1.1 M HCl mixture at 40 °C for 10 minutes. Recoveries for the analysis of NIST SRM 1515 (apple leaves) were between 89 and 98%, RSDs were <6.7% and LOQs were 0.95 (Cu) to 2.42 (Fe) mg kg−1. A central composite design was used108 to optimise (3.0 mL HCl + 3.6 mL HNO3 + 2.78 mL HF) an ultrasound-assisted extraction method for the determination by FAAS of Al, Cd Cu, Ni and Zn in 0.1 g soil samples. Method LODs ranged from 1.3 (Zn) to 230 (Al) mg kg−1. Analysis of IRMM CRM BCR 142 (light sandy soil), RM-Agro E2002a (tropical soil) and RS-3 (river sediment) gave results that were mainly within 10% of certified values. The Box–Behnken design was applied109 in the optimisation of a closed block digestor for determination of elements in tobacco products (cigar, shredded and rope) prior to analysis by ICP-AES. For an extraction solution of 3.0 M HNO3 + 3.5% H2O2, the optimum conditions were 180 °C and 120 minutes. Accuracies for up to 17 elements in three CRMs INCT CTA-OTL-1 (oriental tobacco leaves), NIST SRM 1515 (apple leaves) and Agro C1003a (tomato leaves) were in the range 91 ± 4% (V) to 117 ± 4% (Sr). Product types could be classified according to element concentrations.

A novel microscale sample pretreatment method was proposed110 for the multi-element analysis of plants. Laser-capture microdissection was used to isolate as little as 400 ng tissue samples which were digested under a pressure of 40 bar in 50 μL of 2[thin space (1/6-em)]:[thin space (1/6-em)]1[thin space (1/6-em)]:[thin space (1/6-em)]1 HNO3–H2O2–H2O. The digests were aspirated into an ICP-MS instrument at a low flow rate of 50 μL min−1. Accuracy was assessed by analysis of 500 μg portions of NIST SRM 1515 (apple leaves) and ranged from 84% (Mo) to 128% (Ni). Precision (RSD, n = 8) was ≤11% for Al, B, Ca, Cu, K, Mg and Mn but poorer for Mo, Ni and Zn (32, 62 and 19%, respectively). The authors did not explain why sample microdissection and digestion might be preferred to direct mapping of tissue samples by established techniques such as LA-ICP-MS, XFM or LIBS.

Methods for assessment of trace element availability in soil included111 a procedure for the measurement of B isotopic composition as a means of assessing B availability. The B was extracted into hot water (100 °C, 30 minutes), recovered using three stages of ion exchange chromatography and analysed by MC-ICP-MS. The isotopic composition of B extracted within the first 50 minutes was constant; then δ11B values increased, indicating that less available forms of B were being released. The chemistry of I in soil solution was investigated112 using microdialysis for in situ passive sampling, HPLC-ICP-MS for speciation analysis and SEC coupled to both UV spectroscopy and ICP-MS for determination of molecular weight. The optimised microdialysis approach was particularly recommended for soils with low moisture content. More conventional methods such as centrifugation and soil-squeezing may require soil wetting, which can alter redox conditions in the soil solution.

A dual-frequency ultrasonic enzymatic procedure allowed113 fast, efficient extraction of As species from powdered herbs used in traditional Chinese medicine. Using cellulase in Tris–HCl buffer at 30 °C, samples were irradiated simultaneously at 40 kHz in an ultrasonic bath and at 20 kHz by insertion of an ultrasonic probe. A 95% extraction efficiency was achieved in just 6 minutes. There were no significant differences between measured and certified values for BMEMC CRMs GBW(E)090066 (Salvia) and GBW(E)090067 (Paeoniae Radix Rubra). The overall LODs were 0.7 and 2.5 μg kg−1 for AsIII and AsV, respectively.

Estimating the human bioaccessibility of As species in soil and plant-based foods is important for accurate exposure assessment. Tokalioglu et al.114 assessed species stability during application of the popular BARGE UBM by applying it to simulated saliva spiked with AsIII, AsV, DMA, MMA and a mixture of these spikes. There was no species interconversion in either gastric or gastrointestinal phases when analysed by HPLC-ICP-MS.

The operational nature of sequential extraction procedures was highlighted28 by an attempt to scale-down the well-established BCR protocol. Soil samples with masses as low as 0.0625 g were extracted in 2–3 mL of reagents. Although the fractionation patterns obtained were similar to those obtained using the full-scale protocol, they were significantly affected when the size of the extraction vessel was reduced from 50 to 15 mL. Klotzbucher et al.115 investigated the mechanisms underlying the Hedley fractionation procedure. They spiked pure mineral phases (ferrihydrite, goethite, amorphous aluminium hydroxide, allophane, montmorillonite and kaolinite) with different forms of P (NaH2PO4, phytic acid and RNA), mixed these with silt-sized quartz and then carried out the extraction. They concluded that the protocol characterised neither mineral sources nor binding strengths of P in soil and that it was therefore unsuitable for studying P bioavailability.

Optimisation of extraction procedures for the determination of NPs by sp-ICP-MS has continued to be of interest. Gao et al.116 found sodium pyrophosphate to be a more effective extractant than water, sodium carbonate or sodium chloride for the isolation of gold NPs from soil and sediment. Their optimised method was evaluated for gold NPs with different diameters and coatings. The extraction efficiency was improved by destruction of SOM with UV irradiation. Li et al.117 decreased the sample SOM content in their optimisation and validation of a previously published method for silver NPs in soil by diluting their samples with SiO2 prior to extraction. In a study on the persistence of foliar NP pesticides on lettuce, Laughton et al.118 recommended methanol digestion rather than the standard enzymatic procedure to prevent the dissolution of copper NPs.

Ermolin et al.119 compared a microcolumn to a rotating coiled column (RCC) for estimating the mobility of CeO2nanoparticles in spiked soil. Although the microcolumn simulated the structure of a real soil more closely, the RCC was useful for estimating maximum NP mobility.

4.3.2 Analyte separation and preconcentration. A review (93 references) on the preconcentration of radionuclides considered120 fully and partially automated flow methodologies for ICP-MS or radiometric analysis. Particular emphasis was placed on determination of Ra, Sr, Tc, Th, U and Y nuclides as required under the auspices of REVIRA, the Spanish Network of Radiological Surveillance. Methods developed121 for the simultaneous determination of 237Np and Pu isotopes employed 242Pu as a yield tracer for all the nuclides measured. A new type of extraction resin, TK200, was used122 for the isolation of Pu isotopes. The Pu recovery was 81–91% and the LOD was 0.13–0.24 pg kg−1 for a 2 g soil or sediment sample.

An on-line preconcentration device for Mn in tea leaves based on two knotted reactors and FAAS was automated123 using four solenoid valves controlled by an Arduino board to switch sample streams and reagents. Multivariate optimisation of the method resulted in an LOD of 0.070 mg kg−1, spike recoveries of 95–105% and a Mn concentration of 52.0 ± 5.0 mg kg−1 for NIST SRM 1515 (apple leaves) for which the certified value is 54.0 ± 3.0 mg kg−1.

Tables 3 (liquid-phase extraction methods) and 4 (solid-phase extraction methods) summarise other methods for the analysis of soils, plants or related materials as well as those developed for other sample matrices using soil or plant CRMs for validation.

Table 3 Preconcentration methods involving liquid-phase microextraction used in the analysis of soils, plants and related materials
Analyte(s) Sample matrix Method Reagent(s) Detector LOD (μg L−1, unless otherwise stated) RMs or other validation Ref.
AsIII, AsV Carrot, coriander, radish spinach, and soil VAME Deep eutectic solvent based on choline chloride and citric acid, DDTP chelating agent ETAAS 0.1 BMEMC CRMs GBW10014 (cabbage) and GBW10015 (spinach); spike recovery 352
As, Cd, Hg, Pb, Se, V Fish, forage grass, peach leaves, liver Ultrasound-assisted extraction, MAE Deep eutectic solvents based on citric acid, malic acid, and xylitol ICP-MS 0.002–8.1 μg kg−1 NIST SRM 1547 (peach leaves); NRCC CRM DORM-3 (fish protein) 353
Au Soil, water LLME N-methyl-N,N,N-trioctylammonium chloride ion-pair forming agent, 1-octanol extraction solvent FAAS 0.6 CRM-SA-C (sandy soil C) 354
Cd Dust, seawater, soil, spring water, tap water, wastewater Co-microprecipitation/flotation KI complexation, cetyltrimethylammonium bromide/sodium perchlorate neutralisation FAAS 0.18 Spike recovery 355
Cd, Cu, Mn, Ni Water Coprecipitation Co(OH)2 FAAS 0.07 for Cd, 0.2 for Cu, 0.3 for Mn, 0.2 for Ni NIES CRMs No.1 (pepperbush) and no. 7 (tea leaves) 356
Co, Cu, Ni Fish, oyster SFODME 1-(2-thiazolylazo)-2-naphthol complexing agent, 1-undecanol extraction solvent FAAS 0.03–0.04 mg kg1 NIST SRMs 1573a (tomato leaves) and 1577 (bovine liver) 357
CrIII, CrVI Wild leafy vegetables Tunable solvent system–DLLME 8-Hydroxyquinoline and APDC complexing agents, 1,8-Diazabicyclo[5.4.0]undec-7-ene and decanol tunable solvent solution ETAAS 0.048 for CrIII 0.072 for CrVI Spike recovery from an aqueous soup of Digera arvensis 358
Mo Beef, grass DLLME Potassium ethyl xanthate chelating agent, acetonitrile dispersive solvent, carbon tetrachloride extraction solvent ETAAS 0.03 μg kg−1 NIST SRMs 1573a (tomato leaves) and 1577c (bovine liver) 359
Ni Cabbage, potato, spinach, tobacco tomato, water UA-IL-DLLME Quinalizarin complexing agent, 1-hexyl-3-methylimidazolium tris-(pentafluoroethyl)trifluorophosphate solvent FAAS 0.6 NWRI CRMs TMDA 51.3 (fortified water) and 53.3 (fortified water); NIST SRM 1570a (spinach leaves) 349
V Apple, banana, mushroom, tomato, spinach, water LLME Deep eutectic solvent based on zinc chloride and acetamide, Triton X-114 non-ionic surfactant, ammonium pyrrolidine complexing agent ETAAS 0.01 NWRI CRM TMDA-53.3 (Canadian lake water); NIST SRM 1573a (tomato leaves) 360


Table 4 Preconcentration methods involving solid-phase (micro) extraction used in the analysis of soils, plants and related materials
Analyte(s) Matrix Substrate Substrate coating Detector LOD (μg L−1) RMs or other validation Ref.
Cd, Co, Pb, Pd Soil, water MWCNTs Glutaric dihydrazide FAAS 0.12–0.19 NIST SRMs 2556 (used catalyst pellets) and 1570a (spinach leaves) 327
Co Soil Zirconium NPs SQT-FAAS 2.2 Spike recovery 362
Co Apple and orange juices, black tea, rice, saliva, urine, water, wheat Magnetic graphene oxide ETAAS 0.023 HPS RM TMDW (drinking water) and NIST SRM 1570a (spinach leaves) 363
Pb Cigarette samples Silica-coated magnetic nanodiamonds FAAS 40.8 NACIS CRM NCSDC 73349 (bush branches and leaves) 364
Pd Soil Magnetic NPs SQT-FAAS 6.4 Spike recovery 365
TeIV, TeVI Sediment, soil, water Ionic liquid nanocomposite based on Fe3O4@ SiO2@NH2 magnetic NPs HG-AFS 1.9 ng L−1 for TeIV 3.7 ng L−1 for TeVI Spike recovery 361
U Rock, water MWCNTs/Cu2O–CuO hybrid material ICP-MS 0.52 NWRI CRMs TMDA-62.2 and TMDA-70.2 (water); BMEMC CRM GBW 07423 (soil) 366


4.4 Instrumental analysis

4.4.1 Accelerator mass spectrometry. Progress has been made to address various challenges in the use of AMS. The Fe isotopic compositions of sediments are used by astrophysicists to gain insight into processes such as stellar evolution, so the availability of new 60Fe standards124 should allow greater accuracy to be achieved in such studies. Two new materials, one wood and one mammoth bone, were characterised125 and made available to fulfil the need for appropriate, radiocarbon ‘dead’, procedural blanks in dating applications. An analytical procedure was successfully adapted126 to allow 242Pu to be used as a yield tracer in the determination of 237Np in clays. Although not yet proven through the analysis of real environmental samples, the proposed new method127 for the preparation of calcium- or neodymium-fluoride-based targets for the determination of 236U is welcome because the higher ion yields obtained, relative to those of traditional iron oxide-based targets, should improve sensitivity.
4.4.2 Atomic absorption spectrometry. A method for the determination of Cd in lichens by SS-ETAAS involved128 pipetting 10 μL of a mixed matrix modifier (2 g L−1 Pd + 1 g L−1 Mg(NO3)2 + 0.1% Triton X-100) over 2 mg of sample. The LOD of 0.9 μg kg−1 was considerably better than that (2.1 μg kg−1) for conventional solution analysis. Results obtained for four plant CRMs were in agreement with certified values (t-test at 95% CI). In the analysis of mint, a molybdenum-coated T-shaped SQT atomiser for FAAS gave129 signals 75% higher than those for an uncoated quartz tube. The optimal trapping time was 360 s. Trapped Cd was released by introducing H2. Recovery from spiked mint leaves was 98% and a concentration of 1.56 ± 0.16 mg kg−1 was determined for NIST SRM 1573a (tomato leaves), for which the certified value is 1.517 ± 0.027 mg kg−1.

In the development of a low-cost method for arsenic speciation analysis, a heterogeneous photocatalysis procedure was optimised77 by using ZnO/UV irradiation to reduce AsV and DMA to AsIII prior to HG-FAAS detection. Spike recoveries from soil, sediment and water samples were 89–104% and the LODs were 3.2, 3.9 and 6.7 μg L−1 for AsIII, AsV and DMA, respectively. The method is particularly suitable for studying environmental contamination in resource-poor regions of the world where techniques such as HPLC-ICP-MS are not widely available.

4.4.3 Atomic emission spectrometry. Developments in element-specific analysis included130 a multicomponent spectral-fitting algorithm that could reduce interference from Cu in the determination of P in soil extracts. A new method131 for the measurement of Hg and MeHg in biological tissue and sediment by ETV-CCP-AES was based on a prototype system previously proven for other analyses. Following sample digestion or extraction, 10 μL of solution was pipetted onto a Rh filament for atomisation into an argon microplasma. In a procedure for the determination of Si in plants by MIP-AES, the introduction of 1 L min−1 of air to the N2 plasma overcame104 matrix interference associated with the presence of high levels of Na resulting from use of alkaline digestion.

In multi-element analysis, the performances of Bi and Pt as ISs were compared132 for the determination of 17 elements in IRMM CRMs BCR 482 (lichen) and BCR 670 (aquatic plants) and NIST SRM 1515 (apple leaves) by ICP-AES with an axially viewed plasma. In general, best accuracy was obtained with Pt as the IS but the use of Bi also gave better results than analysis without an IS. An investigation into interference effects in the analysis of sediment extracts obtained by the BCR sequential extraction procedure highlighted133 the need for reagent-matched calibrants, robust plasma conditions and use of an IS.

Improvement in glow discharge emission sources included134 a new HG-APGD system for the determination of As and Se in environmental samples. Optimised GD conditions were: 10 mm electrode gap; He plasma gas flow rates of 65 (As) and 55 (Se) mL min−1; and discharge currents of 30 (As) and 25 (Se) mA. The LODs were 0.087 and 0.13 ng mL−1 for As and Se, respectively. Results for analysis of BMEMC CRM GBW 07381 (stream sediment) were within 6% (RPD) of certified values. A novel tantalum porous-cage-carrier improved135 signal stability and sensitivity in the determination of trace elements in soil. The fabricated device gave reasonable accuracy (relative error 0.7–17%) for the determination of 21 (mainly REE) analytes in BMEMC CRM GBW 07430 (soil). The LODs were 0.04–1.31 mg kg−1.

Arc emission spectrometry was recommended136 as an alternative to pXRFS for the direct determination of Cd in soil because its LOD (0.01 mg kg−1) was far superior than that of pXRFS (ca. 20 mg kg−1). Accuracy, estimated as RPD with respect to certified values for three BMEMC soil CRMs (GBW 07311, GBW 07308a, and GBW 07446) was better than 13% and the precision (n = 10) was <15%. This approach was therefore considered fit-for-purpose for rapid screening analyses.

The renewed interest in the use of the two-jet argon plasmatron emission source was highlighted137 by the publication of a method for the multielement analysis of powdered plant samples using a powder-introduction device in which a spark between zirconium electrodes agitated the sample for transport to the plasma in a stream of argon. Calibration was based on the trace element content in graphite powder CRMs. Whereas leaf and grass samples could be analysed directly, analysis of matrices with higher starch content (maize, potato, rice and wheat) benefitted from heating in a furnace for 15 minutes at 250 °C followed by 30 minutes at 350 °C.

In the direct analysis of solid samples by ETV-ICP-AES, Al Hejami and Beauchemin138 showed that addition of a small amount of H2, N2 or water vapour as a sheath gas around the sample aerosol increased sensitivity and lowered LODs for 13 elements in soil. Best performance was obtained with N2. This work was extended by Scheffler et al.139 who recommended addition of 0.4 L min−1 of N2 to the plasma gas flow and 20 mL min−1 of N2 sheath gas to the central channel of the ICP in order to form a mixed gas plasma. Results obtained for three soil CRMs were within 20% of certified values. There were140 no significant differences (student t-test at 95% CI) between results obtained for Ca, Cd, Cu, Fe, Mg, Mn, Sr and Zn when determined by either ETV-MIP-AES or LA-MIP-AES when the same spectrometer system was used with both sample introduction methods. The LODs for ETV-MIP-AES of 0.1–1.2 μg kg−1 were similar to those of LA-MIP-AES but the precision (n = 3) of 2–4% was slightly better than that (3–7%) of the latter technique. Accuracy was tested through analysis of NIST SRM 2711 (Montana soil) and NRCC CRMs PACS-2 (marine sediment) and TORT-2 (lobster hapatopancreas). For most of the elements a t-test (at 95% CI) showed no statistically significant difference between found and certified values. In an ETV-ICP-AES method for multielement analysis of plants, a high plasma operating power (1600 W) increased141 sensitivity, thereby allowing smaller sample masses to be analysed.

4.4.4 Atomic fluorescence spectrometry. Liu et al.142 reported the first combination of PVG (20 s UV irradiation at 12 W from a mercury lamp) and gas-phase analyte preconcentration on the surface of a dielectric barrier detector (DBD) for the determination of trace Se in water and soil by AFS. A detailed mechanistic study indicated that, in the presence of 10% (v/v) formic acid and 0.2% (m/v) NaNO3, the species generated by PVG were SeCO and H2Se which reacted under the Ar/O2 atmosphere inside the DBD and were trapped on the quartz surface as SeO2 or selenite. Subsequent introduction of Ar/H2 generated H radicals that released a pulse of Se atoms from the surface for quantification. The LOD was 0.004 μg L−1 and the precision was 4% (RSD) at 20 μg L−1 (n = 11). Results for the analysis of BMEMC soil CRMs GBW 07449 and GBW 07405 were within 10% of certified values or results obtained by ICP-MS.
4.4.5 Inductively coupled plasma mass spectrometry. The determination of Ra by ICP-MS is challenging because the presence of polyatomic interferences on 226Ra (the longest-lived isotope of Ra with t1/2 1600 y) dictates purification of samples prior to analysis. Ben Yaala et al.143 successfully developed a method for 226Ra measurement in sediments that did not require pre-treatment. Either the signal at m/z 226 from a matrix-matched blank was subtracted or a mathematical interference correction was used. The LOD was 0.11 ng kg−1 and results obtained were in broad agreement with those obtained by α-particle- or γ-ray-spectrometry.

The capabilities of flow FFF-UV-ICP-MS and sp-ICP-MS to obtain information on particle size distribution of colloidal iron oxyhydroxides in soil solution were compared144 with approaches such as DLS, TEM, filtration, centrifugation and dialysis. Flow FFF-UV-ICP-MS was recommended due to its ability to measure smaller particles than sp-ICP-MS (down to 5 nm hydrodynamic diameter) and because of the wide size range detectable.

Chromatographic separation coupled to ICP-MS is becoming more routine but new or improved methods continue to be reported. Worthy of note were: an HPLC-ICP-MS method145 for As speciation in seaweed, sediment and seawater; a SEC-ICP-MS method146 to study Cd and Cu complexation by humic acids in soil; an IC-SF-ICP-MS method147 for determination of lanthanides, Pu and U in spent nuclear fuel and soil; SEC-UV-ICP-MS and HPLC-ICP-MS methods112 that were used together for identification of I species in samples of soil solution obtained by microdialysis; a HPLC-ID-ICP-MS method148 for Hg speciation in marine sediment; and an IC-ICP-MS approach149 for S speciation in foodstuffs based on orange daylily and wolfberry.

In an ETV-ICP-MS method for the direct determination of iodine in rocks, soils and sediments the use of pre-reduced Pd prevented150 analyte loss during ashing and improved analyte transfer into the plasma. Subsequently, ionisation efficiency was improved with sodium citrate. The method LOQ was ca. 10 μg kg−1 and results similar to certified values were obtained for a suite of geological CRMs.

Researchers successfully developed151 a slurry nebulisation ICP-MS method for screening 16 elements in plant-based foods. A mean particle size of 0.8 μm was achieved in 90 s by wet milling with 1.5 mL 0.5% polyethylene imine in water. Results for eight CRMs were within 10% of target values when aqueous calibration standards were used.

An attempt was made152 to obtain 2D images of U distributions across oak tree rings by HR-LA-ICP-MS using pressed-pellet calibrants prepared from both U-doped cellulose and NIST SRM 1570a (spinach leaves). The IS was 13C. The high uncertainty and poor agreement with results obtained by HR-ICP-MS analysis of solutions was attributed to heterogeneous distribution of the analyte within the sample.

Interest in the measurement of isotope ratios has led to the development and application of new MC-ICP-MS isotope methods for the measurement of the following: 11B/10B in irrigation water and bell pepper plants to study B fractionation during uptake and translocation;153114Cd/110Cd in rock, soil and manganese nodule CRMs;102208Pb/206Pb and 207Pb/206Pb in soil and wine as a potential geographic tracer for authentication of Lambrusco PDO wines from Modena;154130Te/126Te in mine tailings, soil and sediment by HG-MC-ICP-MS as indicators of Te mobility and environmental redox conditions;155 and 234U/238U and 235U/238U in soils from Odisha, India and Fukushima, Japan.156 Although it was reported157 that MC-ICP-MS was not necessary for measurement of 67Zn/66Zn in soil-fertiliser-plant systems as the precision obtained with standard quadrupole ICP-MS was adequate, this study featured the addition of a nutrient solution containing highly enriched 67Zn and so did not concern the measurement of natural isotopic variations.

New ICP-MS/MS methods continue to be developed as instrumentation becomes more widely available. Key features of a procedure for determination of 236U/238U isotope ratios below 10−8 in lake sediments potentially affected by the Fukushima Daiichi nuclear power plant accident were36 O2 as the reaction gas and use of the 234U/235U ratio measured by MC-ICP-MS to correct for bias in the ICP-MS/MS measurements. The combination of chemical separation and use of a CO2–He reaction gas mixture allowed158 accurate measurement of trace Pu in soil samples containing high levels of U (U/Pu atom ratios up to 1012). A method for the measurement of non-metallic elements in herbal teas used159 O2 and H2 sequentially as reaction gases. First, O2 was introduced to the cell and Br, I, P and S measured as their oxide ions; then H2 was introduced and Cl measured as H235Cl+ and Si as 28Si+. Results for NIST SRMs 1515 (apple leaves) and 1547 (peach leaves) agreed (95% CI) with the certified values. The development160 of an approach to distinguish engineered zero-valent iron NPs from naturally-occurring Fe-rich colloids by ICP-MS/MS trace element profiling is welcome as it may help in the development of optimised strategies for the remediation of contaminated soils and waters.

4.4.6 Laser-induced breakdown spectroscopy. Useful two-part LIBS reviews have been published this year. One featured research in the period 2010–2019 on agricultural materials, with part 1 (ref. 161) (76 references) covering soils and fertilisers and part 2 (ref. 162) (82 references) crop plants and food products. Another two-part review focussed exclusively on soil. Part 1 (ref. 163) (74 references) covered the principles of LIBS and its use to estimate soil properties including pH, degree of humification of SOM and texture. Part 2 (ref. 164) (97 references) covered elemental analysis, trace element mapping and soil classification.

Improvements in the LIBS analysis of soil were achieved through study and optimisation of the laser energy,165,166 pressure,167 lens-to-sample distance166 and time delay between irradiation and signal acquisition.165–167 Application of a magnetic field perpendicular to the plasma increased168,169 plasma temperature and hence analyte emission intensities and thereby lowered LODs. Spatial confinement of the plasma within a cylindrical cavity achieved170 the same result. Ablation of the sample generated a shock wave, reflection of which from the cavity wall compressed the plasma plume, increased collision probability and thereby enhanced the number of atoms in excited states.

Numerous publications featured LIBS methods for the determination of trace elements in soil. This is a challenging field and genuine improvements are welcome. However, unfortunately, not all proved their accuracy through analysis of CRMs or by comparison of results with those obtained by an established technique. It cannot be emphasised strongly enough that taking a soil sample, spiking portions with different concentrations of analyte, dividing these into a training set and test set and then demonstrating that, with appropriate chemometric processing the model can accurately predict the analyte concentrations in the test set, does not necessarily lead to a robust analytical method widely applicable across a range of different soil types because the nature of the sample matrix markedly affects the signal obtained. Sun et al.171 illustrated the differences in the slope of calibration curves obtained for Ag in spiked CRMs and soil samples before using machine-learning algorithms based on BPNN to propose a soil-independent multivariate model with REP in the range 5–6%. Wu et al.172 developed a standard addition method for quantification of Pb in soils from the vicinity of a Pb/Zn smelter. Addition of uncontaminated soil taken from 3 m depth at each of the individual sampling locations facilitated production of a standard addition curve (perhaps more accurately a “matrix dilution” curve) for each sample analysed. Results were generally within 18% of those obtained by ICP-AES so the method was considered suitable for screening analysis. Other studies that used ICP-AES data for comparison included a multivariate LIBS method developed by Lu et al.173 for Sr and V and a method for the determination of Ca, Fe, K and Mg developed by Costa et al.174 Guo et al.175 measured 13 elements in 17 BMEMC ‘GBW’ soil CRMs to compare univariate calibration with PLSR and SVR. The PLSR model gave more accurate predicted concentrations than did the SVR model when GBW07448 was analysed as an ‘unknown’. An article with an eye-catching title apparently offered176 rapid detection of Cr in different oxidation states in soil by LIBS. Disappointingly, however, no solid-state speciation was performed. Instead, a soil was mixed with a solution containing CrIII and CrVI and the filtrate analysed directly (2.5 μL on a zinc target) to give total Cr concentration. The CrVI in a 1 mL aliquot was preconcentrated on 0.1 g of anion-exchange resin which was then attached to a glass slide for analysis. The CrIII concentration was estimated from the difference between the total Cr and CrVI concentrations. At best, the method is therefore applicable to Cr speciation analysis in the soil solution provided this can be recovered without species interconversion.

There has been continued interest in the prediction of soil properties based on chemometric processing of LIBS data. Methods were reported for cation exchange capacity177 based on PLSR, soil pH178 based on the RF model, and soil texture179 (proportions of sand, silt and clay) based on PLSR and the elastic net algorithm.

The same research group that applied a magnetic field to increase plasma temperature in the analysis of soil, as already described in this section, also carried out180 a similar study for plants by using an external electrical field for enhancement of emission line intensity. Results for Ca, Cr, K, Li, Mg and Na were broadly similar to those obtained by LA-TOF-MS in the root, stem and leaf of Euphorbia indica.

Notable for the determination of trace elements in plants was a method181 for measurement of Ca, K, Mg, Na and P concentrations in edible seeds that combined LIBS, ICP-AES, hyperspectral imaging and PCA to explore analyte correlations. A calibration-free LIBS procedure produced182 results within 5% of ICP-AES values for Ca, K, Fe, Mg, Mn, Na, P and S in Moringa oleifera leaves. Agreement for Cu and Zn was, however, poorer. The addition of copper powder and the use of Cu emission lines to correct for self-absorption effects in the determination of Ca in fennel, bay, dandelion, spinach and parsley leaves improved183 markedly the agreement with ICP-AES data. Without correction, the LIBS results were almost four times the ICP-AES values but with correction they were within 3% of the ICP-AES results. A slope ratio calibration procedure used successfully for the determination of major and trace elements in plant leaves was based184 on the relationship between emission intensity, ablated sample mass and number of laser pulses. A key advantage of the approach was that it required only a single solid calibrant. An optimised collinear dual-pulse LIBS method for the determination of Cr in rice leaves used185 an inter-pulse delay of 1.5 μs, energy ratio of 1[thin space (1/6-em)]:[thin space (1/6-em)]3 and a total energy of 80 mJ. Further work was carried out186 to visualise the distribution of Cr in the leaves.

There was continued interest in chemometric processing of LIBS data for plant classification. Liu et al.187 successfully determined the geographical origin of Mentha haplocalyx from five provinces in China by a LS-SVM method. Feng et al.188 discriminated between leaves of three types of plant growing at the same site (Ligustrum lucidum, Viburnum odoratissinum and bamboo) using PLS-DA.

The use of a secondary laser for excitation of the analyte atoms in the ablation plume for quantification by AFS can improve both selectivity and sensitivity in LIBS analysis. An optimised LIBS-LIF method incorporated189 a tuneable dye laser for the determination of Pb in medicinal herbs. Results for analysis of five samples of Rheum officinale were within ±15% of those obtained by ICP-MS. The LOD was 0.13 mg kg−1. An LA-LEAF procedure for As based on excitation by an ArF excimer laser was intended190 primarily for the analysis of steel but a successful preliminary analysis of Japanese seaweed was also reported with a LOD of 1.0 mg kg−1.

4.4.7 X-ray spectrometry. An informative IUPAC technical report (170 references) on advances in X-ray techniques for trace element analysis included191 recent applications involving atmospheric particulates, geological samples, living organisms, sediments, soils, waters and waste materials. A useful review (110 references) highlighted192 the scope and limitations of pXRFS when applied to soil analysis. Users were cautioned to be cognisant of the sources of uncertainty and of the influence of the matrix.

In the analysis of soils, She et al.193 selected calibration standards for samples of unknown composition based on Kβ/Kα ratios. The Kβ/Kα ratios for Al, Ca, Fe and Si measured in 14 CRMs deviated 1.5% (Si) to 21% (Al) from expected values according to sample type. More accurate results were obtained for “unknown” samples when the Kβ/Kα ratios for sample and CRM were similar. A TXRFS method for the chemical classification of clays was developed194 and optimised using a full factorial experimental design. Optimised conditions for a data acquisition time of 1000 s were: sample size 50 mg; 2.5 mL 1% Triton X-100 solution as dispersing agent; deposition of 5 μL of the slurry on a siliconised quartz reflector; and drying at 50 °C for 5–10 minutes. Validation was performed using the SARM-CRPG RM ZW-C (zinnwaldite). Recoveries were 80–100% for 12 out of 15 elements. Another TXRFS study explored195 different calibration strategies for the determination of Cd, Pb and Zn in soil suspensions. Whereas the analysis of a suite of CRMs gave an accuracy of 80–100% when using an IS of Ga (for the Mo X-ray tube) or Pd (for the W X-ray tube), results for environmental samples from a mining area were <70% of the values obtained by ICP-AES. Accuracy was not improved when spiked soils were used as external calibrants. However, when results were normalised using a calibration curve obtained by the ICP-AES analysis of a set of soils with similar matrix composition to that of the unknown samples, accuracies were in the range 95–98%. Mitsunobu et al.196 developed a ‘live soil imaging chamber’ for simulation experiments to observe metal(loid) behavior at the redox layer of waterlogged soils. The chamber consisted of a thin (6 mm) rectangular box made of corrosion-resistant titanium with a low permeability quartz glass window to allow observation and prevent oxidation. The chamber was mounted directly on the sample holder in the μXRF–μXAFS beamline. In paddy-field simulation experiments, As-spiked soil was incubated in the chamber for one month at 24–26 °C and then the chamber was mounted directly into the synchrotron allowing element abundance and speciation to be determined in situ.

Chemometric processing of pXRFS data to infer information on general soil properties continues to attract attention. Numerous researchers197–202 employed different machine learning algorithms in their proximal sensing approaches for investigating various soil fertility parameters. Mancini et al.203 compared RF, SVM and LDA models, built with and without PCA, to predict soil parent material. In an evaluation of the RF, SVM and GLM models for prediction of soil texture, the RF algorithm predicted204 sand content best but the SVM model was superior for predicting silt and clay contents.

Developments in WDXRFS included a study on As bioavailability in soil by An et al.205 who analysed directly the sorbent material (chelex and TiO2) recovered from a DGT device and dried at room temperature for 2 h before analysis. A LOQ of 0.06 μg As was achieved. Li et al.206 prepared samples of rock, soil or sediment by high-pressure (2000 kN for 30 s) pelleting followed by coating of the pressed pellets with a 3.6 μm film of polyester. The determined Cl content was stable under repeat analyses whereas the content of pellets prepared without a film was not. Accuracy, assessed using pellets prepared from 8 CRMs, was within 10% for major elements and 25% for trace elements.

Methods for the analysis of plant samples continue to evolve. Shao et al.207 used the argon fluorescence peak in air as a normalisation standard in a low-power TXRFS method for the rapid determination of the low-Z elements Ca, K, P and S and the medium-Z element Fe. The procedure required only a small amount a sample (20 mg) and a short measurement time (10 minutes). Results were in good agreement with those obtained using Ge and V as ISs. For the Chinese CRMs GSB 11 (citrus leaves), GSB 14 (porphyra), GSB 16 (spirulina), and GSB 19 (astragalus), recoveries were in the range 80–120% and RSDs were <10%. The spatial resolution (down to 5 μm) of μXANES was208 sufficient to differentiate leaf tissues in a method for the assessment of cell-specific metal speciation in plants. Metal coordination was determined by comparing the spectra with the XANES information for relevant reference compounds. Use of shock-frozen hydrated samples in a cryostream reduced the sample degradation observed for other sample preparation methods. Cell-specific changes in Zn ligands could be observed in infected plants.

5 Analysis of geological materials

5.1 Review articles

Several reviews covered the analysis of geological materials by a variety of techniques. That of Balaram et al.209 (207 references) on the REEs, critical components in a range of modern technologies, not only reviewed methods for the determination of the REEs but also summarised the whole life cycle of these metals. Tjabadi and Mketo210 (176 references) outlined the spectrometric, chromatographic and electroanalytical techniques that have been applied over the past decade for the determination of total elemental S and the halogens in various matrices including environmental materials. The challenges and merits of analytical techniques in which samples could be analysed directly were compared with those for methods requiring sample pretreatment.

A special issue of Chemical Geology (volume 537, March 2020) devoted to calcium isotope geochemistry is essential reading for anyone interested in this subject. It contained contributions that highlighted a range of applications and identified interesting and fundamental problems that could be the subject for future work. The introduction211 (63 references) to this special issue provided a brief historical perspective on the analysis and interpretation of stable Ca isotopes in the geosciences and identified future directions and approaches for their increased utility and broader acceptance by the geosciences community.

Reviews that focussed on a specific type of instrumental analysis can be found in the relevant sections.

5.2 Reference materials and data quality

The synthesis and characterisation of microanalytical RMs is an ongoing priority for matrix-matched calibration and QC for in situ microbeam technologies such as SIMS, LA-ICP-MS, LIBS and EPMA. A set of three new synthetic reference glasses with an andesitic major element composition (ARM-1, 2 and 3) was prepared212 by fusing mixed-oxide powders containing 54 trace elements at levels of 500, 50, and 5 μg g−1. Because the material was viscous, a prolonged melting procedure was needed. This involved maintaining the melt at 1550–1600 °C for 4 h and then stirring for 5 h. The ARM glasses were considered to be homogeneous for all major elements at a spatial resolution of 10 μm for EPMA and for trace elements at a spot size of 50 μm for LA-ICP-MS. Reference and information values were presented for 56 elements, including major elements. Ke et al.213 developed a new method for preparing a matrix-matched RM for the LA-ICP-MS analysis of scheelite, an accessory mineral common in various hydrothermal deposits. Single crystals of CaWO4 doped with several REEs were grown from a polycrystalline material synthesised from a solid-state reaction at 1200 °C over four days between CaCO3, WO3, Na2CO3 and REE oxides. The structure and matrix composition of the crystals were verified by XRD and EMPA and the homogeneity of the REE distribution assessed by methods such as LA-ICP-MS and solution ICP-MS. The variations in REE concentrations were of the order of 4% RSD. The CaWO4 single crystals were deemed to satisfy the requirements of a matrix-matched RM for LA-ICP-MS measurements of REEs in natural scheelite. Hydroxyapatite pellets spiked with Co, Mn, Ni, Sr and V were synthesised214 using a co-precipitation procedure in which the dried precipitate was calcined at 450 °C for 3 h prior to pelletisation. These potential RMs for LIBS mimicked the ablation properties of biogenic hard tissues such as bones and teeth much better than other pressed powder pellets of comparable composition. Although the pellets were chemically homogeneous, only 50 to 60% of the expected analyte concentrations were recovered after calcination so solution ICP-MS analysis was necessary to characterise the materials. In a feasibility study on the production of RMs suitable for SIMS and EPMA, 47Ti or 48Ti were implanted215 into synthetic, ultra-high-purity silica glass. This technique of high-energy ion implantation generated a known concentration of the implanted isotope in three dimensions, a so-called ‘box-profile’. The procedure involved multiple implantation steps with varying ion energies, thereby producing several ‘box-profiles’ with mass fractions ranging from 10 to 1000 μg g−1 Ti and homogenous Ti distributions to depths of 200 nm to 3 μm. The SIMS depth profile measurements demonstrated that the Ti distribution in the ‘box-profile’ was within the target uncertainty of ±5%. This multi-energy approach was considered promising also for the production of RMs for EPMA; 1000 μg g−1 Ti at an implant depth of 3 μm could be determined accurately by EPMA. An added advantage was that all elements from H to U could be implanted individually, with no restrictions regarding the matrix.

As well-characterised carbonate reference materials for in situ Sr isotope analysis by LA-ICP-MS were only available with Sr concentrations >1000 μg g−1, Weber et al.216 prepared a new RM with a Sr mass fraction of ca. 500 μg g−1. This nanopowder RM, called NanoSr, had a 87Sr/86Sr ratio of 0.70756 ± 0.00003 (2SD) as determined by TIMS and MC-ICP-MS and was homogeneous at the tens of μm scale. A study by Jochum et al.217 investigated whether the homogeneity of several CaCO3 RMs could be improved by further processing to produce fine grained nano-powders. Analysis of nano-pellets of USGS RM MACS-3 (synthetic CaCO3) and the natural GSJ RMs JCp-1 (coral) and JCt-1 (giant clam) by ns- and fs-LA-ICP-MS clearly showed that the nano-pellets were 2–3 times more homogeneous than pellets of the original material, so making them much more suitable as microanalytical RMs for LA-ICP-MS. For MACS-3, the mass fractions determined for all certified elements in the nano-pellets except that for Si agreed with those in the original sample within the LA-ICP-MS repeatability of several percent. However, for JCt and JCp very small but significant differences were found for some trace elements at low concentrations, indicating the need for a re-certification of these materials. The study proposed more robust reference values for MACS-3 and Sr isotope data for all three materials.

Several contributions focused on natural mineral RMs for in situ trace element and isotope analysis. Batanova et al.218 characterised the olivine MongOL Sh11-2 from central Mongolia by EPMA, LA-ICP-MS, SIMS, ID-ICP-MS, XRFS and ICP-MS at six institutions worldwide. The homogeneity was sufficient for reference and information values to be reported for 27 major, minor and trace elements. An examination of the chemical composition and homogeneity of two Mg-rich olivines (355OL and SC-GB) by EPMA, LA-ICP-MS and solution ICP-MS concluded219 that these olivines could be used as primary standards. Unfortunately, these materials are in limited supply so can only be distributed for characterising new olivine RMs in other laboratories. The importance of matrix-matched calibration was emphasised but, if this was not possible, settings for LA-ICP-MS parameters such as spot size, fluence and number of total shots were recommended to minimise fractionation effects when using a non-matrix-matched silicate glass RM. An apatite RM (Eppawalla-AP) for high precision Cl isotope measurements was obtained220 from a mega-crystal from the Eppawalla carbonatite (Sri Lanka) and characterised by IRMS and SIMS. Neither technique showed any variation in the δ37Cl values within the analytical uncertainties, so Eppawalla-AP was considered to be homogeneous at the 10 μm scale. The recommended δ37Cl value for Eppawalla-AP was −0.74 ± 0.15‰ (2SD) as determined by IRMS. It was proposed that this apatite could be used as a matrix-matched RM for in situ Cl isotope studies of apatites and as a QC material for bulk analysis. Various microanalytical techniques were used221 to confirm a uraninite from Utah, USA as a potential RM for the determination of REEs in U-rich matrices. The CaO content determined by EPMA was sufficiently homogeneous (2.70 ± 0.38 m/m%, 2SD) for Ca to be employed as an IS for LA-ICP-MS. Major element and REE compositions were homogeneous at the cm and μm scales, respectively. Ma et al.222 investigated various natural titanite crystals with ages ranging from ca. 20 Ma to ca. 1840 Ma as potential RMs for in situ U–Pb and Sm–Nd isotopic measurements by LA-(MC)-ICP-MS. The RMJG rutile from Hebei Province, China was introduced223 as a new RM for U–Pb dating and Hf isotope determinations by LA-ICP-MS. This rutile has very low Th (<0.003 μg g−1) and common Pb proportion of <0.5% but has high contents of Hf (102 μg g−1), radiogenic Pb (20 μg g−1) and U (61 μg g−1). Isotopic homogeneity was established by ID-TIMS, LA-MC-ICP-MS, LA-SF-ICP-MS and LA-ICP-MS; the recommended U–Pb age was 1750 ± 8.4 Ma and the 176Hf/177Hf value 0.281652 ± 0.000007 (2SD).

Several natural zircons have been characterised to assess their potential as microanalytical RMs. Huang224 proposed a zircon megacryst (SA01-A) as a new RM for microbeam U–Pb geochronology and Hf and O isotope geochemistry. A mean 206Pb/238U age of 535.08 ± 0.32 Ma was determined by CA-ID-TIMS, a δ18O value of 6.16 ± 0.26‰ by laser fluorination and a mean 176Hf/177Hf ratio of 0.282293 ± 0.000007 by solution MC-ICP-MS. Although the megacryst was homogeneous for these measurements, it had significant spatial variations in Th/U and Li isotope ratios. A huge dataset of more than 10[thin space (1/6-em)]000 analyses of the well-known Mud Tank zircon (MTZ), consisting of trace element, U–Pb and Hf isotope data was compiled225 from QC measurements made between 2000 and 2018. Based on the U–Pb data, a Concordia age of 731.0 ± 0.2 Ma (2SD, n = 2272) was proposed as the age of crystallisation for MTZ. It was noted that some grains had lower concordant to slightly discordant ages, probably reflecting minor Pb loss. It was concluded that MTZ is a suitable RM for the QC of U–Pb and Hf-isotope analyses if care is taken to select grains that have been tested for homogeneity.

To assess their potential as matrix-matched RMs for the calibration of O isotope measurements by SIMS, five olivine, three clinopyroxene and three orthopyroxene mineral samples were studied.226 Homogeneity at the μm scale was investigated on multiple grains using SIMS and the O isotope compositions determined by laser fluorination IRMS. All eleven minerals were considered suitable as RMs. A calcite from the Oka carbonatite complex (Quebec, Canada) was presented227 as a new Chinese national RM (GBW04481) for carbonate C and O isotopes microanalysis by SIMS. Homogeneity was demonstrated by hundreds of SIMS analyses and the recommended values of δ13CVPDB = −5.23 ± 0.06‰ and δ18OVPDB = −23.12 ± 0.15‰ were established by conventional IRMS. After measuring O isotope ratios in a range of zircon RMs with high-precision using SHRIMP, Avila et al.228 concluded that Temora 2 was a highly satisfactory RM for O isotope determinations if the provenance of the grains was well established. Repeated analyses over nine sessions and seven different mounts agreed within analytical uncertainty for zircons Temora 2, FC1, R33, QGNG, Plešovice and 91500 when normalised to Mud Tank zircon, which was regarded as a useful QC material with typical repeatability of ≤0.3‰ (2SD). Caution was sounded when using other zircons from the Duluth Complex (FC1, AS57 and AS3) as RMs for this purpose as they had an excessive scatter of δ18O values associated with low-U zircon grains. Yang et al.229 assessed the O isotope homogeneity of six well-known apatite RMs and two in-house apatites by SIMS. The O isotope data for all the apatites studied were normally distributed with precisions of between 0.38 and 0.47‰ (2SD), only slightly worse than the precision of 0.36‰ (2SD) for the Durango 3 apatite which was used for QC during the study. Application of a homogeneity index (H, the ratio of the measurement uncertainty to the expected total combined uncertainty) showed that none of the apatites had significant O isotopic heterogeneities. However, on consideration of all the evidence, the three apatites GEMS 203, Kovdor and McClure were considered to be the most suitable to act as RMs for in situ oxygen isotope analysis.

Four new selenium-rich rock RMs (GBW07397 to GBW07400) with Se mass fractions from 38.5 to 1030 μg g−1 were prepared230 by the China University of Geosciences following ISO guidelines and analysed in 10 laboratories by a variety of methods. Certified values were assigned for As, Cd, Cu, Mo, Pb, Se, V and Zn and the materials were approved as national CRMs.

Many RMs with established total elemental contents continue to be characterised for specific isotope systems. For example, the Ba isotopic compositions of 34 geological RMs encompassing a wide range of matrices (silicates, shale, carbonates, river and marine sediments, and soils) with Ba mass fractions between 6.4 and 1900 μg g−1 were determined99 by MC-ICP-MS. The variation of δ138Ba/134Ba in these RMs was up to 0.7‰. The highest ratio was in a shale that had been subjected to a high degree of weathering. An interlaboratory comparison231 involved measuring all currently available Mg isotope RMs and artefact standards with natural Mg isotope compositions with the aim of establishing SI traceability and the comparability between different Mg δ-scales. The RMs were cross-calibrated with expanded measurement uncertainties of <0.03‰ for δ25/24Mg and <0.04‰ for δ26/24Mg. The authors recommended retaining the established scale based on DSM3 (Mg solution) and anchoring it with European RM ERM-AE143 (Mg solution) at −1.681‰ for δ25/24MgDSM3 and -3.284‰ for δ26/24MgDSM3. This would allow a laboratory to use any of the Mg isotope RMs in its research and convert the δ values obtained to any other scale. The Ga isotope ratios of 10 geological RMs (silicates, shales and ferromanganese nodules) and two pure Ga RMs were measured232 by MC-ICP-MS with the aim of providing data to improve interlaboratory calibration. Difficulties encountered were the lack of consistency in the use of RMs by each laboratory for defining the Ga δ zero and the paucity of reliable uncertainties in published Ga isotope data for geological RMs. Following normalisation of all available δ71Ga data for geological RMs to a single RM, the results were in agreement with previously reported values. Kuessner et al.233 demonstrated the effectiveness of their automated IC separation method by obtaining a δ7Li value of 30.99 ± 0.50‰ (2SD) for NRCC RM NASS-6 (seawater) and then reported the first δ7Li values for CRPG RM GS-N (granite) and NIST SRM 2709a (soil). The δ65Cu values for 10 geological and biological RMs were determined234 for the first time using MC-ICP-MS. In addition, δ65Cu values measured for the USGS RMs BIR-1 (Icelandic basalt) and W-2a (diabase) agreed with previously published values. These RMs had complex and varied matrices with Cu mass fractions between 32.2 and 53.3 μg g−1 and δ65Cu values relative to NIST SRM 976 (copper isotope) ranging from −0.10‰ to 0.29‰. Molybdenum stable isotope compositions (δ98Mo relative to NIST SRM 3134 solution) and Mo mass fractions in a suite of NIST, USGS, GSJ and BAS geological RMs were determined235 by MC-ICP-MS using a double-spike method. The study focussed on low-temperature silicate and carbonate sedimentary materials for which Mo isotopic information can be a useful geochemical tool. These RMs had Mo contents between 0.076 and 364 μg g−1 with δ98Mo in the range −1.77 to 1.03‰.

Chinese reference materials have featured in the continuing search for suitable matrix-matched RMs for isotope ratio measurements. The homogeneity of a set of Chinese Geological Standard Glasses (CGSG-1, CGSG-2, CGSG-4 and CGSG-5) for the isotopic analysis of Hf, Nd, Pb and Sr was assessed236 by TIMS and MC-ICP-MS over a period of almost three years. It was concluded that all four glasses were sufficiently homogeneous for these analyses and would be suitable RMs for related geochemical measurements. Differences between the isotopic composition of the glasses and the powdered RMs from which they had been prepared resulted from the addition of flux during the glass preparation. Yang et al.237 provided the first comprehensive study of Hf and Lu mass fractions and Hf isotopic data for 13 Chinese rock RMs (GBW07 103–105, 109–113 and 121–125) representing a broad compositional range of volcanic, plutonic and ultramafic rocks. Values determined were in agreement with the limited data available from previous studies, so the RMs were considered suitable for Lu–Hf isotopic analysis. In order to expand the number of geological RMs available for Ni isotope ratio measurements, high-precision δ60Ni/58Ni values were determined238 by double-spike (61Ni–62Ni) MC-ICP-MS on 16 IGGE RMs for the first time. The intermediate precision for NIST SRM 986 (Ni isotope solution) was 0.05‰ (2SD, n = 69) and typically 0.06‰ for geological RMs. The δ60Ni/58Ni values of the 16 IGGE RMs varied from −0.27‰ to 0.52‰. Because of their characteristic Ni isotope compositions, GSS-1, GSS-7, GSD-10 and GSB-12 were proposed as potential RMs for QC and interlaboratory comparisons. Wu et al.239 measured the δ53Cr values of 22 geological RMs with values in the range −0.44‰ to 0.49‰. They suggested that IGGE RMs GSS-7 (soil), GSS-4 (soil) and GSD-10 (stream sediment) were suitable RMs for interlaboratory comparisons as their Cr isotope compositions complemented existing RMs, most of which had isotopic signatures similar to that of bulk silicate Earth.

To address the problem of there being no internationally-recognised sulfur isotope RM available for Δ33S and Δ36S data normalisation, essential for interlaboratory comparisons, two sodium sulfate materials (S-MIF-1 and S-MIF-2) artificially enriched in 33S were synthesised240. Four of the five laboratories that characterised these materials used conventional IRMS whereas the fifth used a newly-developed MC-ICP-MS method. Isotopic homogeneity and consistency of data led to the conclusion that the Δ33S values derived for S-MIF-1 (9.54 ± 0.09‰) and S-MIF-2 (11.39 ± 0.08‰) could be adopted to calibrate Δ33S measurements. In particular, they could be used to establish a calibration curve spanning a large Δ33S range (0–11‰) by mixing them with other sulfur RMs with zero Δ33S, such as IAEA-SO-5 and IAEA-SO-6 (BaSO4).

In an informative editorial, Meisel241 discussed the use of the δ and Δ notations to express variations in isotope ratios. In order to express the very small variations in stable isotope ratios relative to a reference isotope ratio (δ = 0), notations such as ε and μ were introduced into the scientific literature. However, Meisel argued that there is no need for these when expressing isotope ratio differences because δ is a relative difference and therefore dimensionless. Thus it could be expressed in %, ‰, parts per ten thousand or ppm, depending on how large the variation is; it should not be automatically assumed that δ values are ‰.

Participation in proficiency testing programmes enables laboratories to monitor, assess and improve the quality of their analytical data. In addition, proficiency tests can also provide a route to characterising RMs and CRMs. Potts et al.242 discussed how the GeoPT proficiency testing scheme, established about 25 years ago, could meet the requirements of ISO Guide 35:2017 for the certification of geological RMs. Following a detailed assessment of the metrological properties of GeoPT-assigned values in relation to Guide 35 recommendations, they demonstrated that these values could be regarded as certified values, provided a number of criteria were met. A related study provided243 a critical comparison of results from two rounds of the GeoPT proficiency testing scheme, in which the same material, an andesite, was distributed 18 years apart. A comparison of consensus values for over 50 determinands measured in the two rounds of testing showed remarkable consistency. This demonstrated not only the stability of the material but also the robustness of the procedures adopted by the scheme, given the changes in laboratory practices over the intervening 18 years. The paucity of geological RMs with reliable data for As, Bi, Sb, Se and Te at the μg g−1 level prompted244 a study of 34 samples from various GeoPT rounds and 10 geological RMs covering different rock types. Precautions were taken during the aqua regia digestion to avoid loss by volatilisation before measurement by HG-AFS. Although there were no assigned or even provisional values for many of the GeoPT materials because of the wide range of results reported, the authors calculated a medium value from the data submitted to the GeoPT scheme. Where the As, Bi, Sb and Te mass fractions were above their respective LODs, their HG-AFS results agreed with the calculated GeoPT median values suggesting that these GeoPT medians could be considered as information values. This was not the case for Se, for which it was noted that for Se mass fractions <0.5 μg g−1 many of the results reported to GeoPT may be systematically high.

5.3 Sample preparation, dissolution, separation and preconcentration

An authoritative review245 (180 references) of recent advances in sample preparation for elemental and isotopic analysis of geological samples focussed on: (1) acid digestion methods in open vessels, high-pressure bombs and microwave ovens; (2) alkali fusions; (3) high-pressure ashing; and (4) bulk analysis by LA-ICP-MS. Limitations and applications of the different sample digestion methods were discussed, with special emphasis on sample digestion with NH4F and NH4HF2, both of which are currently attracting much attention. Although LA-ICP-MS is a powerful in situ microanalytical technique, it is frequently used for bulk analysis and this review provided a balanced discussion of procedures for the preparation of pressed powder pellets and fused materials for this purpose.

Various contributions offered improvements in digestion and combustion techniques. He et al.246 developed a method based on NH4HF2 digestion and subsequent dilution with NH4OH for the simultaneous SF-ICP-MS determination of Br, Cl and I in geological materials. The halogens were retained during digestion in open Teflon vessels at temperatures of 200–240 °C for 0.5 to 12 h because the alkaline atmosphere produced during the digestion process suppressed their volatilisation. Most of the results for the 12 geological RMs analysed were consistent with literature values, for which, however, there is a significant spread. For the quantification of REEs in geological samples by ICP-AES, a digestion method using condensed phosphoric acid instead of HF was re-evaluated.247 Results for six international RMs and three Brazilian ore samples indicated that although many potential REE-bearing minerals were dissolved by this method, zircon and xenotime remained in the undigested residue so the method should be used with caution depending on the sample mineralogy. Rondan et al.248 developed a digestion method for the ultra-trace determination of Se and Te in coal by ICP-MS using microwave-induced combustion under O2 at 20 bar without the use of HF. Of the various combinations of mineral acids evaluated as the absorbing solution, a mixture of HNO3–HCl (1 + 1) gave the most accurate results. Negligible blank levels resulted in very low LOQs of 0.002 mg kg−1 for Se and 0.007 mg kg−1 for Te. To test whether desilicification with HF improved the extraction of elements with variable chalcophile affinities (Ag, Cd, Cu, In, PGEs, Re, S, Se and Te) from geological RMs, digestion procedures involving HF–HNO3 in bombs were compared249 with those using HNO3–HCl in Carius tubes. The extraction efficiency of HF-desilicification varied for different elements in different RMs; whereas a significant increase (30–70%) was observed for Cd and In mass fractions after HF-desilicification, there was negligible increase for other strongly chalcophile elements in many of the samples. The variabilities of the host matrix and of the chalcophilic nature of the elements were thought to influence the efficiency of element release from the matrix.

Developments in sample purification included233 a robotic pipetting arm to automate analyte separation by IC in the determination of Li isotope ratios in geological matrices. When compared with manual sample processing, the robotic system reduced sample processing time without compromising accuracy, precision and effectiveness of the chromatographic purification. Another semi-automated procedure, for the measurement of stable and radiogenic isotopes of alkali and alkaline-earth elements in silicate rocks, involved250 decomposition by borate fusion and purification with an IC system equipped with a fraction collector. Each analyte was completely separated from other elements in the elution profile without isotope fractionation and resulted in the complete removal of interference matrices from reagents and samples. Values for δ7Li, δ26Mg, δ88Sr, 87Sr/86Sr determined for various GSJ geochemical RMs were consistent with published values.

A modified method251 for determining very low concentrations of gold in rocks (<0.01 ng g−1) involved Carius tube digestion with reverse aqua regia, chromatographic separation to remove most of the sample matrix and measurement by SF-ICP-MS. Quantification was by external calibration with internal standardisation using Au/Pt ratios, which were precisely determined by ID. Procedural blanks were very low (<6 pg) and the LOD was <0.8 pg L−1. The results were indistinguishable (<5–10%, 2SD) from those obtained by a standard addition technique on the same solution. Reverse aqua regia was an efficient reagent for Au extraction under the high temperature conditions (240–270 °C) employed and was preferable to HF-aqua regia because it only released limited amounts of Hf and Ta, whose oxides would otherwise cause significant interferences. Daniel et al.252 evaluated procedures for preparing gold ores using various types of mills in common use. Optimal performance required the use of a grinding aid such as silica flour or bauxite to avoid caking; a 1 + 1 mixture of silica and bauxite proved to be as effective as grinding in 100% silica. The grinding charge mass was limited to ≤50% of the nominal capacity of the mills. Under these modified conditions, gold particles of mm size could be comminuted to ≤100 μm in <5 minutes; particle sizes of <50 μm could be achieved for 95% of the material with extended grinding times. These results made it viable to reduce the sample masses from the commonly used 25–50 g to 5 g or less for the routine determination of Au in geological samples, thereby decreasing reagent consumption in the subsequent aqua regia digestions or fire assay.

5.4 Instrumental analysis

5.4.1 Laser-induced breakdown spectroscopy. This technique has a wide variety of applications and a short tutorial review (58 references) summarised253 issues related to sample preparation, qualitative and quantitative approaches, signal enhancement, calibration strategies and data processing. Because LIBS is often regarded as only a semi-quantitative technique, Syvilay et al.254 (23 references) published some guidelines for improving the quality of LIBS analyses. Their recommendations such as the use of control charts, optimisation of measurement conditions and evaluation of the various data processing models available are in fact nothing more than good analytical practice. The authors, rather optimistically, claimed that based on their “rigorous” methodology, “it will become possible to compare the LIBS results obtained by different analysts and pave the way towards inter-laboratory comparison.”

Reviews of the LIBS analysis of geological samples included255 a critical account (127 references) of its application to the sourcing and discrimination of minerals and gems and the analysis of slurry and drill cores in mineral exploration. The determination of elements of economic importance such as Ag, Au, the REEs and several light elements (C, F, Li) were also highlighted. Coal analysis by LIBS was the subject of a comprehensive review256 (206 references) which will be invaluable to any reader working in this area. It offered up-to-date information on the progress of LIBS analysis from the fundamentals to industrial developments. The experimental and instrumental challenges that need to be addressed to realise the wide commercialisation of LIBS for coal analysis were also identified. A review (126 references) on advances in the remote detection capability of LIBS considered257 three basic configurations: stand-off LIBS, remote LIBS with optic fibre and compact-probe LIBS. The characteristics of these techniques were described together with a range of applications which included geological investigations and planetary exploration.

Handheld LIBS instruments are becoming increasingly popular. A recent application was the fast quantification258 of F during the purification of fluorite (CaF2) from low-grade and fine-grained ores. Particle size played a key role as there were significant differences in LIBS intensities exhibited for the 10–150 μm and <20 μm powders, with no discernible correlation between the particle size and magnitude of the signal. Matrix effects displayed a non-linear relationship with the F contents measured independently by ISE. By adopting a multi-variate approach based on the signal intensities at two CaF molecular bands to overcome the matrix effects, the on-line LIBS analyser was able to meet the target of a ≤2% error in the F content of CaF2-rich samples. Handheld LIBS was also evaluated259 in a field laboratory for measuring Ag in gold as a rapid method for gold provenance studies. Based on the results for nine training samples of commercial gold alloys (five distinct populations) from French Guiana, the best calibration model was a quadratic univariate model. Subsequently, the origins of four “unknown” samples of gold from the same region were correctly identified in this preliminary study.

Several contributions investigated the use of LIBS in the mining and ore processing industries. A total of 162 sulfide rocks were analysed260 by LIBS and chemometric methods to identify and classify minerals relevant to the copper industry. Of four different chemometric methods assessed for sensitivity, precision and accuracy, the nonlinear classifier artificial neural networks (ANN) proved to be the most reliable method for the identification of seven sulfur minerals (bornite, chalcopyrite, covellite, chalcocite, enargite, molybdenite and pyrite) in untreated rock samples. Another study demonstrated261 that LIBS combined with laser-induced fluorescence (LIBS-LIF) had great potential in the search for high-grade uranium deposits because of the selectively enhanced intensity of the U spectrum which minimised spectral line interferences. Parameters such as the slope of the calibration curve, coefficient of determination (R2) and precision (average RSD) were all significantly better than those of conventional LIBS. This was the first report of a U LOD of <100 μg g−1 in an ore matrix being achieved using LIBS. A multi-energy calibration strategy was successfully applied262 to the fully quantitative LIBS determination of Al2O3, Fe2O3 and TiO2. Two calibrants were required: (1) a mixture of the sample and a RM; and (2) a mixture of the sample and a blank; both in the same proportions. The NIST SRMs 679 (brick clay) and 2703 (sediment for solid sampling) were used as the RM and sample, respectively, and were prepared as fused discs by borate fusion to minimise sample heterogeneity. Boron and Li were used as ISs to compensate for matrix effects. The relative differences from the reference values were −4 to 15%, with LODs between 0.4 and 0.6% for all analytes.

Analysis by LIBS has great potential for the determination of coal properties such as calorific value, ash, volatile content and C and H contents. Zhang et al.263 developed a set of calibration schemes with the aim of improving the figures of merit of such measurements to meet industrial needs. The selection of an appropriate spectral pre-processing method combined with multivariate calibration models improved the accuracy and precision of each index of coal properties. Two methods of sample preparation were compared264 for the LIBS analysis of semi-coke (a special coal with relatively high C content and low volatility). Although painting semi-coke powders onto a tape was a simple way of presenting the samples for on-line monitoring, mixing the coal powder with a binder and pressing into a ‘slice’ for ablation improved the measurement precision significantly, reduced matrix effects and enhanced the stability of the spectrum. A new algorithm that combined SVM with PLSR was utilised to obtain an effective prediction model for determining the C content in this type of coal sample with high accuracy.

The shapes of LA craters produced by a LIBS system employing different numbers of laser shots and pulse energies were studied265 using high resolution X-ray CT. Accurate measurements of crater volume, width, depth and cone angle in aluminium and rock (gold ore) samples closely agreed with those produced by a theoretical simulation model. This method of 3D characterisation of LIBS crater geometry was considered useful for optimising LA setups to produce a constant ablation rate or known depth profile resolution.

5.4.2 Dating techniques. In recent years, in situ U–Th–Pb geochronology has been applied to a range of different minerals. Five allanite samples with significant variations in Fe and Th contents were analysed266 by SIMS to address the issue of matrix effects caused by variable amounts of Th that had been observed in a previous study. Changes in calibration parameters evident for allanites with a large range of Th contents were overcome by applying the power law relationships both between Pb+/U+ and UO2+/U+ and between Pb+/Th+ and ThO2+/Th+. No matrix effects were evident using this strategy for SIMS U–Th–Pb dating on allanites with FeO contents ranging from 12.8 to 16.1% and ThO2 contents between 300 μg g−1 and 2%. It was concluded that the accuracy was mainly controlled by the proportion of common Pb in the analysed material. Luo et al.267 assessed the potential of LA-ICP-MS for U–Pb dating of wolframite minerals which often occur in tungsten deposits and other hydrothermal ore deposits. Lack of a matrix-matched high-quality wolframite RM was a major limitation; the ages determined for wolframite samples when calibrated against the well-characterised zircon 91500 were ca. 12% younger than the corresponding ID-TIMS ages. If water vapour were added to the He carrier gas before the ablation cell, this discrepancy was eliminated and accurate wolframite U–Pb ages were obtained using ns-LA-ICP-MS. Several of the wolframite samples analysed in the study were promising candidate RMs. Bastnäsite is the end member of a large group of carbonate-fluoride minerals whose LREE contents make them important commercially. A study of 47 bastnäsite samples by LA-(MC)-ICP-MS emphasised268 the importance of a correction for the presence of common Pb in order to obtain reliable U–Th–Pb ages. The relatively high Th contents in most bastnäsites meant that Th–Pb ages were preferred to U–Pb ages as they could be determined more precisely. It was suggested that some of the bastnäsites analysed could be potential RMs for calibration or QC. Fission track dating of apatites by LA-ICP-MS is susceptible to small but systematic variations in apatite U contents so Cogné et al.269 adopted a zeta-based approach to correct for this bias. Instead of counting fission tracks in large numbers of zeta-standard grains (e.g. Durango apatite) for every LA-ICP-MS session – clearly a time-consuming procedure – they determined a single, high-precision zeta factor ζICP in an initial LA-ICP-MS session. This factor was reused for subsequent LA-ICP-MS sessions during which the unknowns were analysed and some of the Durango grains were reanalysed. In this way, a session-specific zeta fractionation factor could be calculated to account for differences in 238U/43Ca fractionation resulting from variations in LA-ICP-MS tuning between sessions.

To reduce fractionation effects in U–Pb age determinations of zircons by LA-ICP-MS, Corbett et al.270 applied multiple 1 Hz shots to a single sample location in a standard ablation cell (volume ca. 8 cm3). A short washout time (3 s) maintained an elevated signal between laser pulses. The extremely shallow craters (aspect ratio of ≪1) significantly reduced the effect of ‘downhole’ fractionation and allowed age determinations to be made on a μm to sub-μm scale. This ability to integrate and collate signal pulses for a small number of consecutive laser shots rather than pulsing the laser continuously at 5–20 Hz produced precise age determinations (ca. 1% reproducibility, 2 RSD) often indistinguishable from those determined by ID-TIMS for the same zircon. In addition, this approach reduced thermally induced effects such as substrate melting, plasma loading and signal mixing with depth in a heterogeneous sample. Liu et al.271 introduced a 2% v/v ethanol solution into the ICP to increase the sensitivity and to suppress any isotopic fractionation during the dating of zircons by LA-ICP-MS at high spatial resolution. In combination with a shielded torch system, the mixed gas plasma significantly improved the precision, accuracy and uncertainty of 206Pb/238U ages for small (10 and 16 μm) spot diameters. However, the effect was insignificant for intermediate (24 and 32 μm) spot diameters. Mean weighted 206Pb/238U ages of zircon RMs (Plešovice, GJ-1 and 91500) determined by this method agreed within 2SD with literature values obtained by ID-TIMS and LA-ICP-MS. Anderson et al.272 showed that Raman spectroscopy could be used to produce maps of radiation damage in zircon crystals as an aid to the interpretation of (U/Th)/He dates obtained by LA-ICP-MS. These Raman spectroscopy maps could be used to visualise intracrystalline variations in zircon properties, which had implications for the analytical strategy adopted when dating ancient, zoned zircons by LA-ICP-MS.

One of the challenges faced in the isotopic analysis of single grains of zircon is the small amounts of U they contain. By improving dissolution, purification and measurement methodologies previously developed to determine 238U/235U ratios in small samples, Tissot et al.273 demonstrated that it was possible to measure this ratio in single zircon crystals by MC-ICP-MS with a precision (±0.04 to ±0.25‰) sufficient to resolve U isotopic differences between grains from the same location. It was proposed that this method could be used to improve the accuracy and precision of U–Pb and Pb–Pb dates and to enable accurate re-evaluation of U decay constants. The in situ U–Pb dating of the U-rich mineral uraninite by SIMS or LA-ICP-MS is hardly ever carried out because a matrix-matched RM is required to correct for Pb/U fractionation. A procedure developed274 for U–Pb dating of single-grain uraninite by ID-TIMS had the advantage that no matrix-matched RM was necessary. Because <1 μg of material was required for the determination of ages with high precision, this method provided the basis for the highly spatially resolved analysis of uraninite grains in samples such as thin sections. Prior to the ID-TIMS analysis, it was important to use EPMA and LA-MC-ICP-MS to identify minerals, to determine chemical composition and to check age homogeneity.

In the context of dating speleothem carbonate samples, Perrin et al.275 proposed an integrated petrographical and geochemical approach for optimising the subsampling of speleothems for U/Th dating. Non-destructive in situ trace element screening by portable EDXRFS was combined with petrographical and mineralogical information to produce elemental distribution maps which were used to identify growth discontinuities and primary and secondary (diagenetic) carbonate phases on the sample surface and thereby facilitate the selection of areas suitable for subsampling for radiometric dating. A rapid procedure for extracting Pb and U from carbonate minerals combined276 the commonly used extraction of Pb on AG1-X8 anion-exchange resin with that of U on Eichrom TRU-resin to create a single-column sequential extraction. Experiments with large speleothem calcite samples showed that it was possible to process 200 mg of material in only one day, half the time required for separate sequential extractions based on the same chemistry. Because low blanks (ca.10 pg Pb) could be maintained, the procedure was ideally suited for high-precision U–Pb dating of speleothems.

Acid leaching is widely used in Pb–Pb geochronology to separate radiogenic Pb from non-radiogenic Pb in samples. A study of acid-leached minerals from meteorite samples demonstrated277 the potential pitfalls of acid leaching for this purpose. High-Ca lamellae within single pyroxene grains were more affected by leaching with dilute HF than were low-Ca lamellae so inaccurate Pb–Pb age estimations and scattered Pb isotopic data resulted. It was considered that leaching with HF should be utilised with great caution for Pb–Pb dating of pyroxenes and pyroxene-bearing materials.

Improvements in the 40 Ar/ 39 Ar dating of samples rich in volatile elements were achieved278 by modifying the gas purification protocol so that suppression of the Ar signals caused by incomplete cleaning of the gases extracted from the sample by a laser heating device was minimised. The modified preparation configuration improved the precision of the age determinations by MC-noble gas MS by an order of magnitude resulting in geologically plausible 40Ar/39Ar ages for small masses of rocks of Quaternary age. In a quest to find suitable RMs for inter-laboratory and inter-technique comparisons of ages produced by the 40Ar/39Ar and U–Pb geochronometers, sanidine and zircon crystals from the Carboniferous Fire Clay tonstein, a large ash bed in the Appalachian Basin (USA), were assessed.279 The preferred mean 40Ar/39Ar date from the sanidine crystals of 315.36 ± 1.10 Ma (2SD) was consistent with the weighted mean 206Pb/238U zircon age of 314.629 ± 0.35 Ma (2SD). Based on the good single-crystal reproducibility of the sanidine data and the overall consistency between the two geochronometers, the Fire Clay tonstein was considered to hold promise as a RM of Palaeozoic age.

5.4.3 Inductively coupled plasma mass spectrometry.
5.4.3.1 Instrumentation. A new micro-flow liquid sample introduction system developed for the direct nebulisation of small samples into the plasma at flow rates as low as 5 μL min−1 consisted280 of a demountable direct injection high efficiency nebuliser (dDIHEN), a FIA valve and a gas displacement pump connected to a mass-flow controller to deliver a very stable liquid flow. The μ-dDIHEN plugged directly into the ICP torch in place of the injector. The system was successfully tested on four different types of ICP-MS instrument with sample loops of 10 and 50 μL and flow rates of 5–50 μL min−1. Signal sensitivity increased with the liquid uptake rate up to 30 μL min−1 but poor aerosol quality and reduced ionisation under wet conditions affected performance at higher flow rates. Three different applications were demonstrated: B isotope ratio measurements in geological samples; trace element analysis of natural water samples; and gold NP characterisation by single particle ICP-MS.
5.4.3.2 Trace element determinations. New approaches to trace element determinations by LA-ICP-MS included281 the use of single-shot measurements (200 nm, 1 Hz repetition rate) to determine Mg/Ca in μm-sized calcareous chambers of foraminifera using a fs-LA system and SF-ICP-MS. Precision of the Mg/Ca ratio was improved through almost simultaneous measurement of doubly charged 44Ca2+ (m/z = 22) and singly charged 25Mg+ (m/z = 25) ions. Low fluence settings of 0.3–0.6 J cm−2 enabled a high depth resolution of 50–100 nm per pulse down to a depth of 10–20 μm. The precision was ca. 5% RSD. The procedure was applied to the detailed analysis of single chambers and shell-depth profiles in planktic and benthic foraminifera shells for paleoclimate reconstruction studies. In a novel approach to the analysis of geological samples, solutions of digested material were ablated282 to produce an aerosol for measurement by ICP-MS. In the procedure, 50 mg of material was first digested in NH4HF2 in a screw-top PFA vial at 230 °C for 3 h; then 15 μL of digest was pipetted into tiny pits on a Teflon sheet which was sealed with parafilm and ablated with a 193 nm ArF excimer LA system. Water-related interferences were reduced by 1–2 orders of magnitude compared to those for solution aspiration and were similar to those for sample introduction via a desolvating nebuliser. Other advantages were greatly reduced matrix effects and signal sensitivities improved 70–250 times. The NIST SRM 610 (silicate glass) was employed as the calibration material and a multi-element standard solution for QC. Using In as the IS, the concentrations obtained for major and trace elements in geological RMs ranging from mafic to felsic rocks were within 10% of the reference values for 45 elements; precisions were <7%. This method was regarded as comparatively environmentally friendly because the consumption of acid and ultrapure water during sample preparation was 20–100 times less than that of conventional liquid nebulisation. It would be interesting to see how the performance criteria and green credentials of this method compare with those of ablating a pressed pellet.

Matrix effects that can arise when elemental abundances in sulfides are determined by ns-LA-ICP-MS using non-matrix-matched silicate RMs for calibration were quantified283 using three different ISs (Cu, Fe, Ni). Individual fractionation indices (Fi values) for Fe-rich sulfides were significantly different from those derived for Fe-rich metal alloys. Nickel was the preferred IS for the measurement of volatile elements, whereas Cu or Fe was recommended for transitional and/or refractory elements. As shown in previous studies, the magnitude of the matrix effects for sulfides was strongly correlated with elemental volatility but remained constant for each element with increasing concentrations. This finding was used to derive a new model for predicting Fi values for Fe-rich sulfides and to assess any discrepancies between measured and true sulfide liquid–silicate melt partition coefficients which may be under- or over-estimated by up to 0.15 and 0.2 log units, respectively, if matrix effects were not taken into account.

The majority of software programs for elemental imaging by LA-ICP-MS were developed to work with rastered data and so are limited in their ability to handle signals generated from individual laser shots of short duration (<10 ms). To process baseline-separated peaks produced by the latest generation of LA systems, a new standalone software application called “LA-ICP-MS Image Tool” was developed284 for converting raw LA-ICP-MS data into images using a ‘pixel-by-pixel’ approach. This freeware tool located peaks within raw data files and used the peak locations to segment the data at appropriate intervals, converting the data into a matrix of colour-coded pixels. Raw data were converted into a 60[thin space (1/6-em)]000 pixel image within 2 minutes making it a viable approach for high-throughput imaging tasks. Both continuous signals and baseline separated peaks could be processed. The software also had the capability to identify peaks in single-shot or single-particle ICP-MS experiments to assist in the alignment of line scans. The image matrix could be exported as an Excel-compatible file, allowing further processing to be carried out off-line if required. van Elteren et al.285 explored the strengths and weaknesses of LA-ICP-MS imaging when the data were generated in single or multiple pulsed ablation modes. Depending on the LA-ICP-MS instrumentation and imaging conditions applied, various imaging artefacts such as smear, blur, aliasing and noise can degrade the image quality. An understanding of the potential sources of these artefacts was used in the development of a computer simulation model and metrics for the objective assessment of the image quality to aid the optimisation of LA-ICP-MS imaging parameters for fast and high-quality 2D mapping.

Recent developments in the elemental analysis of individual fluid inclusions by LA-ICP-MS were summarised286 in a short review (68 references) in which strategies for improving the success rate of ablation, accurate quantification and selection of suitable fluid inclusions were discussed. Tuba et al.287 addressed the problem of analysing assemblages of fluid inclusions in orogenic gold deposits in which the inclusions are often too small and densely populated to be measured individually by LA-ICP-MS. For this application, inclusion-rich areas of the quartz host were analysed using a single continuous LA profile. The signals generated were converted into time-slice datasets and plotted as element ratios in ternary diagrams to reconstruct specific major and trace-element ratios. The method had high spatial and chemical resolution and the estimated compositions were in good agreement with results from previous analyses of the same material. Separate inclusion populations could be distinguished on the basis of their major- and minor-to trace-element concentration ratios.


5.4.3.3 Isotope ratio determinations by ICP-MS. The considerable research effort to improve isotope ratio determinations by MC-ICP-MS and other techniques is reflected in Table 5. Because the range of elemental isotope ratios now being measured in geological materials is so diverse, a table is provided as a starting point for readers to explore the systems of most relevance to them. It is difficult to discern any major breakthroughs as many of the studies provided modest improvements to existing separation procedures or analytical protocols.
Table 5 Methods used in the determination of isotope ratios in geological materials by solution ICP-MS and TIMS
Analyte Matrix Sample treatment Technique Analysis and figures of merit Ref.
Ag Natural and processed gold Dissolution of mg-sized gold samples in aqua regia and Au removed by AEC on AG1-X8 resin. Solutions doped with NIST SRM 3138 (Pd solution) to generate Pd/Ag ratios close to 1.9 (108Pd/107Ag ca. 1) MC-ICP-MS Combination of SSB method and Pd-doping for mass bias correction using the exponential law. δ109Ag values expressed relative to bracketing NIST SRM 987a (Ag solution). Combined analytical uncertainty (2SD) was better than 0.016‰ 367
B Boron RMs, loess, sediment Fine-grained loess and sediment subjected to chemical leaching with acetic acid and B separated by ion chromatography using Amberlite IRA 743 resin. Procedural blank 0.25 ± 0.03 ng and average recovery 99.1% B MC-ICP-MS Different rinse solutions tested to minimise B memory effects; 0.6 mg g−1 NaF in 1% HNO3 reduced B signals to blank levels within 4 minutes SSB method for instrumental drift and mass bias correction. Long-term reproducibility for NIST SRM 951a (B isotopes) was 0.01 ± 0.06‰ (2SD, n = 27) 82
Ba Barite RMs Investigation of the effect of sample dissolution using the Na2CO3 reaction method on Ba isotope measurements. Ba purified by ion-exchange chromatography on AG50W-X12 resin; Ba yield >99%. A 135Ba–136Ba double spike was added to sample solutions prior to analysis MC-ICP-MS 131Xe, 134Ba, 135Ba, 136Ba, 137Ba, and 140Ce collected simultaneously. SSB and double spike used to correct for instrumental drift and mass bias. Ba isotope data reported relative to NIST SRM3104a (Ba isotopes), long-term external precision of δ137Ba/134Ba <0.05‰ (2SD) 368
Ca Ca-rich minerals and geological RMs Study to see if accurate Ca isotope ratios can be achieved by TIMS using 42Ca–43Ca double spike technique without column chemistry. All samples dissolved in 1.6 mol L−1 HCl. Two aliquots taken and mixed with 42Ca–43Ca double spike: one was measured directly and the other loaded onto Biorad AG MP50 resin to perform column chemistry TIMS Samples were loaded on a single Ta filament and H3PO4 was added as activator. Monitored masses included 40Ca, 41K, 42Ca, 43Ca and 44Ca. Instrumental mass bias correction by the double-spike technique and reported relative to NIST SRM 915a (CaCO3). Insignificant differences (−0.04 to +0.07‰) between with and without chemistry, smaller than the measurement precision of ±0.12‰ over 6 years (2SD, n = 515). Conclusion: bias-free Ca isotope ratios can be achieved on Ca-rich materials without chemical separation 369
Ca, Mg Geological and biological RMs, Digested materials in 4 M HNO3 purified in two-step CEC method using DGA and AG50W-X12 resins to separate Ca and Mg. Ca and Mg re-dissolved in 2% (v/v) HNO3, for isotope analysis by MC-ICP-MS. Effect of acidity and concentration mismatch as well as matric effects evaluated for Ca isotope analysis MC-ICP-MS SSB method applied. Ions collected at 42Ca+, 43Ca+, 87Sr2+ and 44Ca+; ratios reported relative to NIST SRM 915a (CaCO3). 24Mg, 25Mg and 26Mg measured and Mg isotopic ratios based on the DSM3 standard. Repeated measurements of USGS BHVO-2 (basalt) and NIST SRM 1400 (bone) was better than ±0.08‰ and ±0.06‰ (2SD), respectively, for δ44Ca/42Ca. For δ26Mg, the external precision was <0.11‰ (2SD) 370
Cd Soil and rock RMs, Mn nodule RM Different digestion schemes adopted according to sample matrix. All digested samples in 6 mol L−1 for one-step anion-exchange separation on AG1-X8 resin. Procedure blank <75 pg. Molecular and isobaric interferences studied MC-ICP-MS A 111Cd–113Cd double spike used for instrumental mass bias correction. Signals for 6 stable Cd isotopes collected plus 117Sn, 120Sn, 105Pd and 115In for isobaric corrections. All Cd ratios normalised to NIST SRM 3108 (Cd solution). Intermediate measurement precision of a Cd solution was better than ±0.05‰ (2SD) for δ114Cd/110Cd. Data for 15 RMs reported 102
Cd Geological RMs, Mn nodule RMs Following digestion, 111Cd–113Cd double spike added to the samples, before improved Cd purification scheme for low-Cd samples using polypropylene mini column with AGMP-1M resin. Recoveries ≥90%, with blanks of ≤0.1 ng MC-ICP-MS Cd isotope measurements on FCs using three different MC-ICP-MS instruments and data expressed relative to NIST SRM 3108 (Cd solution). External precision <0.064‰ (2SD). Cd isotope data for various geological RMs consistent with previously published results 371
Ce Geological RMs, uranium ore samples After sample digestion, Ce separated by modified two-stage chromatographic procedure the LREE fraction eluted from AG50W-X12 resin in the first stage still contained some Ba. During the second step, KBrO3 oxidised CeIII to CeIV, which was retained on LN resin. Ce yield was about 78%; blank <60 pg Ce TIMS Newly-developed film porous ion emitter (Pt/Re alloy with a porous structure) enhanced ionisation of Ce+ ions and use of TaF5 as an activator significantly suppressed Ba+ interference signal. Mass fractionation factor determined using the exponential law. Reproducibility of 138Ce/140Ce better by a factor of ca. 10 compared to previously published Ce+ results and comparable with that of the CeO+ technique 372
Cr Geological RMs Digested sample mixed 50Cr–54Cr double spike and heated overnight at 130 °C to homogenise. Modified two-step separation procedure on AG50W-X12 resin, which was preferred to AG50W-X8, as it gave a better separation efficiency TIMS Cr loaded onto Re single filaments and high purity silica gel and saturated H3BO3 added to the sample drops. Cr double spike used to correct for isotopic fractionation during column chemistry and TIMS measurement. 53Cr/52Cr ratios expressed relative to NIST SRM 979 (Cr isotope solution). Long-term measurement precision for BHVO-2 ≤0.031‰ (2SD) 373
Cr Geological RMs Sample digests containing 200–300 ng Cr spiked with 50Cr–54Cr double spike before 3-step ion-exchange scheme MC-ICP-MS Sensitivity improved by ≥1.5 times by cooling waste gas trap bottle of desolvating nebulizer to 5 °C. Empirical method to correct for effect of Fe interference on δ53Cr. Precision on δ53Cr measurements <0.06‰ (2SD); δ53Cr values reported on 19 new RMs ranging from −0.44‰ to +0.49‰ 239
Cu, Zn Geological RMs New separation and purification procedure involving a single pass, triple-stack column. This reduced volume of acids required by ca. 50%, thus shortening duration of separation and lowering blanks MC-ICP-MS Cu and Zn external normalisation in addition to SSB to correct for instrumental bias. New method of inter-calibrating Cu and Zn isotope fractionation coefficients by measuring mixed Cu–Zn solutions with enhanced mass bias variation generated by varying the sample gas flow rate 374
Er, Yb Rock RMs New column chemistry separation technique to separate heavy lanthanoids using an ultra-fine-grained LN resin (20–50 μm) and flash column chromatography, which accelerated the elution speed by 10 times compared to gravity flow. Recovery yields ca. 100% TIMS Ta and zone-refined Re tested as filaments. Er and Yb isotopes acquired separately by dynamic multi-collection and multi-static methods with a two-line cup setting to reduce FC deterioration. Mass fractionation corrected with exponential law 375
Eu Rock RMs Two-step CEC using AG50W X-8 resin. Complete separation of Eu from other REE (>99.99% purity) MC-ICP-MS External mass bias correction using 150Sm–154Sm double spike and exponential law to estimate 151Eu/153Eu ratio. Method applied to five geological RMs and commercial Eu reagents 376
Ga Geological RMs Two column separation method using AG MP-1M and AG 50-X8 resins; recoveries >99% Ga, procedural blanks ≤0.1 ng Ga MC-ICP-MS Instrumental mass bias corrected using combined SSB and internal normalisation. Interlaboratory discrepancies in δ71Ga data for geological RMs resolved by normalisation to a single Ga isotope RM. Highlighted the need for an internationally-agreed δ zero RM 232
Hf Geological RMs Method developed for samples with very low Lu and Hf contents (Hf < 0.1 μg g−1). 176Lu–180Hf enriched spike added before digestion, one-step chemical purification on Ln Spec resin. Hf recovery >90% MC-ICP-MS Lu and Hf concentrations also determined. Corrections for mass bias using enriched spike and normalisation to 179Hf/177Hf = 0.7325 using the exponential law. All data reported relative to JMC 475 Hf standard solution 377
Li Coal (also rock and seawater RMs) Two-step microwave-assisted digestion with HNO3–HF–H3BO3, Li separation on AG 50W X-12 cationic resin, recovery >99.3% MC-ICP-MS Precision better than ±0.30‰. Values for coal RMs SARM18, SARM19 and SARM20 reported relative to NIST L-SVEC (Li2CO3 powder) 378
Li Geological RMs Modified separation procedure on AG 50W-X12 cation-exchange resin using a micro column (to remove Ca and REE) and a long column (to separate Li from Na) arranged in series, to reduce time and volume of acid required for purification MC-ICP-MS Collected eluent introduced directly into the ICP without further processing. SSB method employed with NIST L-SVEC (Li2CO3 powder) as the standard. Long-term precision <±0.47‰. Experiments to find the best zone in the ICP for Li isotope measurements. 379
Li Geological and seawater RMs Modified two-step column purification on 50W-X8 resin to completely separate Na from Li in complex matrices MC-ICP-MS SSB method employed with IRMM-016 (Li2CO3 powder) as the standard. δ7Li values for a range of geological RMs consistent with published values 380
Li Geological and seawater RMs Dual column system for Li separation on AG50W-X8 designed for high matrix tolerance (Na/Li <100). Procedural blank <0.004 ng Li and final Na/Li ratio <1 Q-ICP-MS SSB method with IRMM-016 (Li2CO3 powder) as the standard. Li concentrations in samples and standards matched to within 5%. Long-term precision 1.1‰ (2SD). Quadrupole ICP-MS method with hot plasma tolerant of Na/Li in samples up to 100[thin space (1/6-em)]:[thin space (1/6-em)]1 381
Li Geological RMs Automated one-column separation using robotic pipetting with HCl and 2 mL resin (AGW50-X12) volume. Aqua regia digestion step to destroy any organic matter released from resin MC-ICP-MS SSB method with NIST L-SVEC (Li2CO3 powder) as the standard. δ7Li values for a range of geological RMs consistent with published values and first reported values for NIST SRM 2709a (soil) and CRPG GS-N (granite) 233
Mg Carbonate RMs Method for carbonates avoiding column chromatography. HsSO4 added to generate a CaSO4 precipitate and MgSO4 supernatant, which was diluted for Mg isotope measurements MC-ICP-MS Matrix effects evaluated and method validated by repeat measurements of GSJ RM JDo-1 (dolomite); δ26Mg of JDo-1 was −2.32 ± 0.11‰ (2SD, n = 34), consistent with recommended value of −2.35 ± 0.15‰. The Ca/Mg ratio in solution should be <0.5 for accurate Mg isotope ratios 382
Mg Geological RMs Three-step separation protocol designed especially for high-K and low-Mg rocks: (i) K removed by precipitation; (ii) Fe and Ca separated using 2 mL of AG50W-X12 resin; (iii) Al, Fe, Na and Ti separated on 0.5 mL of AG50W-X12 resin MC-ICP-MS SSB to correct for instrumental mass bias, and normalised to DSM3 international Mg isotope solution. Long-term reproducibility was ±0.06‰ (2 s) 383
Mo Geological RMs Single column extraction protocol using TRU Spec resin; Mo absorbed on resin in 1.5 M HCl while interfering elements are poorly retained. Time for separation procedure approx. 4 h MC-ICP-MS 97Mo–100Mo double spike combined with SSB methodology with NIST SRM 3134 (Mo isotope solution) for mass bias correction. Long-term precision of δ98Mo was 0.082‰ (2SD, n = 334) 384
Mo, W Geological RMs Microwave-assisted heating for rapid decomposition of rocks and sediments. Mo and W separated from sample matrix using chelating resin NOBIAS Chelate-PA1 and anion-exchange resin AG1-X8 MC-ICP-MS For mass bias correction and determination of concentrations, SSB and an external correction method using Ru for Mo and Fe for W was employed. Reproducibilities (2SD) were 0.10‰ for δ98Mo and 0.05‰ for δ186W. Data for 12 geological RMs presented 385
Nd Depleted basic and ultra-basic rocks 3 column purification scheme: (i) 1.5 mL of TRU resin to extract REE, Th and U; (ii) 1 mL DGA resin to separate Nd from Sm; (iii) 1 mL of LN2 resin for further purification from residual Ce and Pr. Takes about 12 h with blanks <50 pg and Nd recoveries >90% TIMS Nd isotopes measured in static multi-collection mode. Within run precisions (2 RSD) 3–9 ppm for 142Nd/144Nd and 2–8 ppm for 143Nd/144Nd; external precisions within a factor of 2 of within-run precisions. USGS RM BIR-1 (ocean basalt) and CRPG RM UB-N (ultramafic rock) used to assess accuracy 386
Nd Nd isotope RM, rock RMs, U ores Relatively simple two-stage extraction and ion-exchange chromatography using TRU resin and LN resin. Recoveries ca. 82% with blanks of 25 pg TIMS 143Nd/144Nd measured as Nd+ ions using a Pt/Re film porous ion emitter attached to the centre of a single Re filament ribbon; ion yields were 10× higher than those of traditional Nd+ ion analysis. External precisions of 35 ppm (2SD) achieved for 1 ng Nd samples. GSJ and USGS RMs used to validate procedure 387
Nd Rock RMs Stable and radiogenic Nd isotope ratios obtained simultaneously by addition of 145Nd/150Nd, plus 149Sm tracer for ID Sm concentrations. Standard chromatographic procedures on AG50W-X8 resin to separate REEs and Ln-Spec to collect Nd and Sm fractions TIMS Nd double spike used correct for mass dependent fractionation. Ratios normalised to GSJ RM JNdi-1, long-term precision (2SD) better than ±0.015‰ for 146Nd/144Nd and ≤11 ppm for 143Nd/144Nd 388
Nd, Sm Rock RMs Single column scheme based on TODGA resins for efficient separation of Nd from Ce, Pr and Sm. Yields >90% Nd and 95% Sm TIMS Measurements as NdO+ and Sm+. USGS Basalt RMs BCR-2, BHVO-2, BIR-1a, ultramafic RMs CRPG UB-N and IAG RM MUH-1 used to access accuracy of Nd and Sm isotope ratios and concentrations 389
Ni Geological RMs and samples Novel 4-step purification scheme on five columns (4 containing AG50W-X8 resin and one with AG1-X8) using only dimethylglyoxime and acetone as organic reagents. Ni yield >92%, blanks 0.4–1.2 ng MC-ICP-MS 61Ni–60Ni double spike used to correct for instrument mass bias and fractionation during purification. Long-term precision 0.06‰ (2SD, N = 18) for geological RMs; δ60Ni values in good agreement with previous studies 390
Os Geological RMs Rock powder dissolved a 1 + 3 mixture of HNO3–HCl (inverse aqua regia) and Os extracted by a conventional two-stage column separation NTIMS Static FC collectors with 1013 amplifiers. Of the 6 RMs analysed, USGS RM BIR-1 (basalt) and CCRM WPR-1 (altered peridotite) were the most homogeneous with respect to Os isotopic composition for test portions of 0.5–1 g 391
Re Wide range of rock RMs (also waters) Modified column chemistry involving loading dissolved sample onto AG1-X8 resin, removing the sample matrix in 3 steps, eluting the Re, evaporating and refluxing the Re fraction, and repeating the whole procedure twice. Samples doped with W to give W/Re ratio of 20 MC-ICP-MS Very low uptake rate nebulizer (ca. 37 μL min−1) and detectors measuring Re isotopes fitted with 1013 amplifiers. SSB method with external normalisation to W. 187Re/185Re reported with precision of ±0.05‰ (2SD) for a mass of >3 ng Re 392
Re, Os Shale RMs 3 protocols for digesting organic shales in Carius tubes were compared; a 1 + 3 mixture of HNO3–HCl (inverse aqua regia) preferred. 190Os and 185Re spikes added prior to digestion. Chemical separation and purification by published methods NTIMS, MC-ICP-MS Os measured as OsO3 by NTIMS, Re by MC-ICP-MS. USGS black shale RMs SGR-1b and SBC-1 considered suitable matrix-matched RMs for the determination of Os and Re isotopes in organic-rich sedimentary rocks 393
Sn, Cd Meteorites, sediments 3-Stage chromatography method: (i) Sn, Cd and Zn separated from matrix elements on AG1-X8 anion-exchange resin; (ii) Eichrom TRU resin used to separate Sn from Cd and remaining Zn; (iii) Eichrom pre-filter resin to remove organic compounds introduced by TRU resin. Cd fraction further purified in 2-stage anion-exchange procedure MC-ICP-MS Sn and Cd isotope ratios measured in dynamic mode, to allow isobaric interference corrections, using SSB method and normalised to either 116Sn/120Sn = 0.4460 or 116Cd/11Cd = 0.578505 using the exponential law to correct for mass bias. Removal of U shown to be critical because of interference from U2+. Data for NIST SRM 3161a (Sn solution) in good agreement with previous data 394
Sr Geological RMs Method developed for measurement of pg sample sizes (30–100 pg Sr). Digestion of 50 mg of silicate powder in HF–HNO3–HClO4 at 180 °C for 4 days. Purification on AG50W-X12 cation-exchange resin. Yields 75–81%, blanks <200 pg TIMS Use of single Re filaments with silicotungstic acid as the ion emitter produced 3-fold enhancement of Sr ionisation efficiency compared to that for a classical Ta emitter. Analysis of NIST SRM 987 (SrCO3 powder) gave precision of ≤0.013‰ (2RSD, n = 8). Accuracy verified by analysis of a suite of geological RMs 395
Sr Geological RMs To improve Sr yields and separate Rb satisfactorily from samples with high Rb/Sr, separation scheme devised involving HF coprecipitation combined with cation-exchange on AG50W-X12 resin. Method most suitable for samples with high Rb/Sr and low Ca and Mg contents TIMS Sr isotope ratios determined using a double Re filament geometry and 87Sr/86Sr ratios normalised to 88Sr/86Sr using the exponential law for mass fractionation correction. Results for GSJ RM JR-2 (Rb/Sr = 37.36) consistent with published data 396
Te Geological RMs, mine tailings, sediments Samples digested in a 1 + 3 mixture of HNO3–HCl (inverse aqua regia). Separation of matrix elements using AG1-X8 anion-exchange resin columns, then further purification on AG50W-X8 resin MC-ICP-MS 120Te/124Te double spike used to correct for mass bias and any fractionation during sample preparation. HG sample introduction so no Ba correction required but correction of Sn-based interferences essential. Data normalised to NIST SSRM 3156 (Te solution). Precision of ca. 0.09‰ (2SD) for δ130Te/126Te similar to other methods 155
Ti Geological RMs, minerals Purification using a dual-column loaded with Ln-Spec and AG50W-X12 resins, resulting in nearly 100% Ti recoveries with very low matrix element concentrations MC-ICP-MS SSB method with NIST SRM 3162a (Ti solution), precision for δ49Ti 0.047‰ (2SD, N = 130). 14 geological RMs analysed and results in good agreement with published data 397
U–Pb Cassiterite Dissolution in HBr in Parr bomb vessels with addition of 202Pb–205Pb–2233U–236U tracer. After refluxing to ensure reduction of Sn4+ to Sn2+, TRU-spec resin used to obtain U and Pb aliquots that were purified separately using HBr-anion-exchange chromatography ID-TIMS U Isotopes measured as UO2+ species, mass fractionation corrected online using the measured Pb and U isotope ratios in the tracer, an exponential fractionation law and assuming a 18O/16O of 0.0020485 398
W W standard solutions Regression model to obtain absolute isotope ratio measurements of an element by using an isotopic standard of another element, in this case NIST SRM 989 (Re isotope solution) was used to calibrate W in candidate RM WOLF-1 MC-ICP-MS Test solutions containing W and Re introduced into ICP and plasma power incrementally increased to induce a shift in the mass bias in a relatively short period of time, providing sets of W and Re isotope ratios within 10 minutes regression model based on 159 sets of W–Re sets of data and traceable to SI units. W and Re in same solution so any matrix effect was eliminated 399
Zn Geological and biological RMs Validation of SpinChem™ technique, which involves placing loaded chromatographic columns in 50 mL centrifuge tubes into a large- volume centrifuge; the centrifugal force generated enhanced reagent flow rates of up to ×10 faster than gravity protocols. Case study based on Zn isotopes purified using AG1-X8 resin in a two-pass protocol. MC-ICP-MS SSB employed to correct for instrument mass bias. Both δ66Zn and δ68Zn reported for USGS RMs BCR-2 and BHVO-2 (basalts). Figures of merit such as blanks, yields and analytical precision similar to those achieved by purification under gravity flow 400
Zr Geological RMs Separation of Zr achieved using single Eichrom DGA resin column after addition of 91Zr–96Zr double spike. Purification procedure took <4 h. Compromise was to reduce Zr yield to ca. 95% to reduce Mo recovery to ca. 33% TIMS Isobaric interference from Mo largely eliminated during filament heating and also off-line Mo correction. Long term precision for δ94Zr ≤ 0.06‰ (2SD). Accuracy confirmed by analysing USGS RMs BHVO-2 (basalt) and AGV-2 (andesite) 401


Reports of advances in isotope ratio methodology by LA-ICP-MS covered many different matrices and isotope systems. The determination of Hf isotopes in zircon is not trivial as it requires five corrections for mass bias and interferences and measurements accurate to the 5th decimal place. With this in mind, Spencer et al.288 discussed the fundamentals of Lu–Hf analyses of zircon and provided some novel techniques for data visualisation, integration of geographic information and statistical evaluation. Their recommended workflow was proposed as best practice for assuring that steps taken in data correction and interpretation were robust and rooted in fundamental geological, isotopic and analytical constraints. A new isobaric interference correction model was proposed289 for in situ determinations of Hf isotope ratios in zircons, especially those with high Yb/Hf ratios. This LA-MC-ICP-MS procedure employed a specified zircon RM to calculate the mass bias factors that were then applied to other samples. In contrast to previous correction models, it was not necessary to determine the “natural” Yb isotopic composition and the model was equally applicable to zircons with low as well as high Yb/Hf ratios. The practical lower limit for the 180Hf intensity was set at ca. 1 V so that meaningful 176Hf/177Hf ratios and <1.5 ε unit internal errors could be achieved simultaneously. Zircon was also the focus of a study which reported290 a protocol for in situ determinations of stable Zr isotope ratios by LA-MC-ICP-MS. Potential interferences from 89Y1H+ and 180Hf2+ were insignificant. Addition of N2 to the central gas flow increased the Zr sensitivity by a factor of ca. 2. The use of laser spot sizes of 16–32 μm and a low pulse frequency of 1 Hz, together with signal-smoothing, improved the analytical precision by a factor of ca. 61 times compared to that without signal-smoothing. Data were reported relative to the GJ-1 zircon and typical analytical precisions for δ94Zr//90Zr and δ96Zr/90Zr were 0.11 and 0.18‰ (2SD). Accuracy was confirmed by comparison with δ94Zr//90Zr data obtained by the well-established double spike solution method for zircon RMs 91500, Plešovice, Penglai and Mud Tank. Lugli et al.291 provided a user-friendly tool for processing large outputs of Sr isotope data generated by LA-MC-ICP-MS. This Excel-based interactive data reduction spreadsheet could be easily customised for user-specific data-acquisition protocols. Raw data files in a specific folder could be imported and the background and analysis cycles selected before corrections for the main Sr isobaric and polyatomic interferences and instrumental biases were applied. The results were automatically exported into a table. The performance of the spreadsheet was demonstrated by application to analysis of materials such as teeth, shells, speleothems and mineral phases. Further developments and improvements of in situ isotope ratio determinations by LA-ICP-MS included Ca isotope measurements in CaCO3 and CaPO4 materials,292 Fe isotope analysis of glassy cosmic spherules,293 Li isotope analysis in tourmalines294 and in glass RMs and zoned olivines,295 and Os isotope ratios in sulfides.296

Hydride generation MC-ICP-MS can be a valuable technique for isotope analysis because of the potential to isolate the element of interest from interfering species. A method to measure Te isotopes by HG-MC-ICP-MS utilised155 a 120Te–124Te double spike for mass bias corrections. Sensitivities were similar to those achieved with a desolvating nebuliser, and δ130Te/126Te precisions of 0.09‰ (2SD) were obtained for <8.75 ng of natural Te. Although HG avoided the need for a Ba correction and allowed analysis of samples without chemical separation of Te for simple matrices, a modified ion-exchange procedure was nevertheless employed to make the approach more universally applicable. Analysis of a range of USGS RMs, mine tailings, ancient sediments and soils revealed the largest spread in terrestrial Te isotopic composition to date (δ130Te/126Te = 1.21‰), indicating that isotopic fractionation of Te is prevalent in low-temperature marine and terrestrial environments. Selenium stable isotopes are regarded as having great potential as a tracer of redox processes and chemical cycling of chalcophiles and volatile elements. A procedure for the measurement of Se stable isotopes in samples with low Se contents employed297 a novel 76Se–78Se double-spike with HG-MC-ICP-MS detection. Sample requirement was typically 25 ng of natural Se and the sensitivity was >1 kV per 1000 μg L−1 for the total Se signal. Corrections were made for interferences from Ar, As and Ge. The results were expressed as δ82Se/78Se relative to NIST SRM 3149 (Se solution). The long-term external reproducibility was 0.040‰ (2SD, n = 93) so the method should be applicable to the measurement of the Se composition of a wide variety of geological samples.

5.4.4 Secondary ion mass spectrometry. Although TOF-SIMS cannot achieve the same precision as SF instruments, it is possible to record secondary ions for all elements and achieve high sensitivity analysis of a sample surface with minimal damage. Impressive new instrumentation included298 a TOF-SIMS instrument for the in situ microanalysis of geological materials with complex structural and chemical features. The aim was to achieve high spatial and mass resolution with an O2 beam of ca. 5 μm diameter and intensity of ca. 5 nA. The mass resolution was >20[thin space (1/6-em)]000 and the LODs <0.2 μg kg−1. The performance characteristics were demonstrated by the analysis of zircons to obtain REE and Ti compositional information. A group at the US Naval Research Laboratory combined299 a full SIMS instrument with a molecule-filtering single-stage AMS instrument used as a “detector” to make a single unified instrument for spatially resolved trace isotope analysis. This instrument was called NAUTILUS and was capable of collecting molecule-free raster ion images for rapid analysis of trace elements in complex, heterogeneous matrices with a trace element sensitivity at least 10 times better than commercial SIMS instruments due to the near-zero background. The novel capabilities of NAUTILUS were described and its performance demonstrated with relevance to nuclear materials analysis, cosmochemistry and geochemistry.

Various improvements in isotope measurements by SIMS included300 an investigation into how the so-called topography effect could be eliminated in order to obtain high precision Si isotope measurements in quartz samples. A tight linear correlation between measured δ30Si values and a secondary-beam centring parameter (DTCA-X value) was observed so the external repeatability and accuracy could be improved by correcting for this parameter. An external precision of ±0.10‰ (2SD) was considered achievable by using high primary-beam intensities (10–14 nA), a long acquisition time (160 s), sample mounts prepared as flat as possible and a correction based on the DTCA-X parameter. Villeneuve et al.301 measured Si isotopes in a set of 23 natural and synthetic olivine RMs and three natural low-Ca pyroxene RMs by SIMS using two quartz RMs as QCs. All results were normalised to data for NIST SRM 8546 (quartz). The Si ion yields and IMFs varied with the analytical settings and chemical composition of the samples. The magnitude of the IMF in olivine varied in a complex manner resulting from variations in MgO and FeO content so a comprehensive set of RMs was required to avoid inappropriate corrections. In contrast, ion yields and IMFs in low-Ca pyroxenes showed limited variations and were thus more predictable. Matrix effects in the determination of Mg isotopes in olivines and pyroxenes by SIMS were quantified302 by analysing 17 olivine and 5 pyroxene RMs by MC-ICP-MS and MC-SIMS. For olivines, the magnitude of the SIMS instrumental mass bias in δ25Mg was ca. 3.3‰ and was a complex function of the fosterite contents which ranged from 59.3 to 100%. A correction procedure based on a combination of Mg/Si ratios and fosterite content was proposed. On the other hand, the pyroxene RMs showed a smaller range of instrumental bias (ca. 1.4‰ in δ25Mg) but no smooth function with enstatite content (48.6–96.3%), indicating that additional factors such as minor element abundances may contribute to the matrix effects. Vho et al.303 assessed the matrix effects on SIMS O isotope measurements in garnet which appeared to be correlated to the relative proportions of the grossular, andradite and spessartine components present. To supplement the available RMs, three new garnet RMs were characterised (GRS2, GRS-JH2 and CAP02) that had grossular contents of 88.3 ± 1.2% (2SD), 83.3 ± 0.8% and 32.5 ± 3.0%, respectively. Micro-scale homogeneity in O isotope composition was established from multiple SIMS and reference δ18O values obtained by CO2 laser fluorination. A SIMS protocol for in situ B isotopic microanalysis of basaltic glass was developed304 with the aim of identifying different degrees of alteration of basalt glasses during magma generation and evolution in the mantle. The USGS RM BCR-2G (basalt) was chosen as the calibration material as its B content and matrix composition were well matched to those of natural basalt glasses. Its δ11B value of −5.44 ± 0.55‰ (2SD) was determined by solution MC-ICP-MS and the performance of the proposed method evaluated by the analysis of other USGS and MPI-DING glasses, for which the results were consistent with those obtained by other methods. For the natural glasses, distinct δ11B values were obtained along profiles from the grain core towards the rim.

Xia et al.305 investigated a suite of zircon RMs for their suitability for water content determinations in zircon by SIMS and described a modified analytical procedure to acquire data on the zircon water content and O isotopes simultaneously. Features of the method included mounting the samples in a tin-based alloy to reduce degassing and the introduction of liquid N2 to cool the analysis chamber and improve the vacuum, thereby limiting the atmospheric water vapour background to <10 ppm. While 16O and 18O ions were collected in FCs, 16O1H was measured simultaneously with an EM. The 16O1H/16O ratio was converted into water content using a calibration based on FTIR water content determinations. The reproducibility of the water content determination for the suite of zircon RMs was <5% (2SD) except for zircon 91500 for which it was 7.84% (2SD). Taking into account the large variation of water content in natural zircons (<55 to 1212 μg g−1) it was considered that the homogeneity of these zircon RMs was acceptable. A set of five natural white mica RMs was developed306 for in situ measurements of water content by SIMS. The water content of the RMs was obtained independently by thermal combustion elemental analysis. It was suggested that a matrix effect which correlated with the FeO content (1.13 to 3.67 wt%) of the RMs should be corrected by including at least two RMs with FeO contents that bracket those of the unknown white micas. It was argued that an analytical precision of 0.02–0.08% (1 RSD) was expected as the final uncertainty on measurements of water content in unknown white micas.

5.4.5 Other mass spectrometric techniques. Atom probe tomography (APT) is a TOF-MS technique capable of 3D nanoscale chemical mapping of individual atoms. An excellent review (257 references) described307 its development and application to the geosciences. A significant advantage of APT was its ability to provide quantitative, 3D images of all elements with sub-nm spatial resolution without the need to specify in advance which isotopic peaks to quantify. Because APT involves the detailed analysis of a very small volume of material, typically <0.002 μm3, the importance of characterising samples using other analytical techniques prior to APT was stressed. Since the development of laser-assisted APT around 2005, the technique has been applied to a broad range of geological materials, including the nanoscale analysis of accessory minerals, and has provided new insights into the processes by which trace elements may become mobile in geological systems.

A practical guide (50 references) on the double spike technique for correcting mass-dependent fractionation in isotope ratio measurements was written308 particularly for researchers tackling new isotopic systems. The measurement of Ca isotopes by TIMS was used as an example of how to achieve high quality results. Emphasis was placed on points to consider when selecting the optimal pair of isotopes for the spike and the importance of accurate calibration of the spike pair. The technique’s advantages and limitations were discussed.

A study of the determination of the stable isotope composition of C in carbonates by isotope ratio mass spectrometry compared309 three different approaches: dual inlet (DI); elemental analyser (EA); and continuous flow (CF). All methods were considered to be suitable for the determination of 13C/12C but DI-IRMS offered the most precise, accurate and sensitive instrumentation for this purpose. Extensive off-line sample preparation was, however, required. The EA-IRMS approach provided rapid and cost-effective determinations that may be sufficiently precise to distinguish natural trends in some applications requiring high sample throughput. In contrast, CF-IRMS yielded more precise and accurate results but was considered to be very time-consuming and equally expensive as DI-IRMS. Velivetskaya et al.310 developed an improved LA fluorination method for the GC-IRMS measurement of S isotopic anomalies in sulfides. The new gas purification system was based on temperature-controlled flow traps for the cryogenic separation of SF6 gas from other fluorinated products. The new method was tested with IAEA RMs and a natural pyrite with known isotope composition. Overall precisions of ±0.2‰ for δ34S, ±0.27‰ for Δ36S and ±0.03‰ for Δ33S were obtained when the optimal amount of 12–13 nmol SF6 was generated. These were significant improvements over results obtained using a previous purification system. The revised protocol was therefore considered suitable for measuring Δ33S and Δ36S in Archean sulfides.

New procedures were reported311 for the extraction of Ne from different mineral phases (quartz, pyroxene, hematite, apatite, zircon, topaz and fluorite) and measurement on a static vacuum noble gas mass spectrometer. Neon was extracted at 1100 °C by lithium-borate-flux fusion under vacuum and purified by a cryogenic method capable of separating Ne from He. The noble gas mass spectrometer was operated at its highest mass resolving power of ca. 10[thin space (1/6-em)]300 thereby permitting isobar-free measurement of all three Ne isotopes, albeit at reduced sensitivity. Cosmogenic 21Ne and 22Ne concentrations obtained for two internationally distributed Antarctic RMs, Cronus-A quartz and Cronus-P pyroxene, were in excellent agreement with previously obtained results. The same was true for nucleogenic 21Ne and 22Ne concentrations in two other RMs. However, as had been found for He, U and Th concentrations in previous studies, the Durango apatite was heterogeneous in Ne concentrations.

A new high-performance laser ablation and ionisation mass spectrometer for the analysis of solid samples was based312 on a fs LA ion source coupled to a TOF mass spectrometer, thus combining high mass and high spatial resolving powers in one instrument. With a mass resolution of 10[thin space (1/6-em)]000, the instrument was capable of separating isobaric interferences from clusters, molecules and multiple charged ions, thereby significantly improving quantitative analysis of complex samples. The analysis of various NIST SRMs demonstrated LODs in the ppm range and quantitative isotopic analysis with accuracies at the per mil level. Advantages of the technique included acquisition of spectra over the full mass range in a fraction of a second and with minimal sample preparation.

5.4.6 X-ray spectrometry and related techniques. A novel method of pellet preparation for the analysis of geochemical samples by WDXRFS involved206 weighing powdered samples (without any binder) into a polyethylene cup which was covered with a 3.6 μm thick polyester film before pressing at high pressure (2000 kN). The polyester film prevented changes in the Cl content after multiple analyses of the same pellet and reduced contamination from external sources such as dust. The film was not damaged by exposure to irradiation from a 4 kW X-ray tube for 180 minutes. Sixty-two rock, soil and sediment RMs were used to create the calibration curves and the method was evaluated by analysis of another 8 IGGE CRMs for major and trace elements. The relative differences between the results obtained and the certified values were <10% for major elements and <25% for trace elements except for those elements with concentrations close to the LODs. This method of preparation was particularly suitable for samples with high chlorine and sulfur contents to enable repeat analysis and preservation for future use.

Several calibration strategies for the determination of element concentrations in carbonate matrices by portable XRFS were proposed. Arenas-Islas et al.313 prepared 11 gravimetric mixtures of NRCC CRM PACS-3 (marine sediment) with reagent grade CaCO3 in different proportions to act as calibration standards for the analysis of carbonate sediments. The CaCO3 reagent and 11 mixtures were also analysed by FAAS and ETAAS after total digestion to check their trace metal contents. Of the 31 elements detected in the solid mixtures, only 20 exhibited significant linear regressions (p < 0.001). This calibration technique was cost-effective because only one RM was required. It was concluded that reliable determinations of Al, As, Ca, Cu, Fe, Hg, Mn, Mo, P, Pb, S, Si, Sr, Ti, Zn, K, V, Rb, Y and Zr concentrations in carbonate sediments could be made and that the method could be adapted for other mineral matrices. In a different approach for analysing carbonate rocks (limestones and dolomites), Al-Musawi et al.314 used a pXRFS instrument which was calibrated using a set of 43 carbonate samples previously analysed by WDXRFS, ICP-MS and ICP-AES. This carbonate-specific calibration yielded more accurate results than the procedure provided by the instrument manufacturer developed for siliciclastic mudrocks. The concentrations of 13 elements (Al, As, Ca, Cu, Fe, K, Mg, Mn, P, Rb, Si, Sr, Ti, Y and Zr) could be accurately quantified as long as the unknown samples were prepared using the same protocol as the calibration materials to minimise any matrix effects.

Core scanning XRFS is a well-established technique for rapid semi-quantitative analysis of sediment cores at sub-mm resolution. A valuable review (89 references) of current perspectives on the capabilities of high resolution XRF core scanners was published315 as the editorial in a special issue of Quaternary International dedicated to advances in data quantification and applications of this technique. The review summarised the historical evolution of high resolution XRF scanners, approaches to calibration and validation and gave examples of applications related to environmental, sedimentological and seismological studies as well as mineral exploration and forensic geochemistry. Several existing calibration methodologies were assessed316 in a study in which 100 freshwater sediment samples were analysed by core scanning XRFS and ICP-MS to determine which method gave data with the best accuracy and precision and was most cost-effective. Although calibration using multivariate analysis of elemental log-ratios provided the most accurate results relative to the ICP-MS data, this strategy was considered most appropriate for studies involving large numbers of sediment samples (n >100) or when it was crucial to obtain absolute concentrations. Otherwise, either normalising core scanning XRFS data to the X-ray scatter signal or converting the results to dry mass concentrations were regarded as suitable strategies for studies for which absolute geochemical values were less important.

Synchrotron X-ray techniques are increasingly used to study processes at a molecular level, particularly with the advent of ultrahigh-brilliance fourth-generation synchrotron-light-sources. A review (221 references) of the application to studies of ore deposits of various synchrotron X-ray techniques emphasised317 their importance as tools to study trace element distributions at μm and smaller scales, the structure and chemistry of poorly-crystalline ore materials and the chemical nature of the fluids that give rise to ore formation. In an assessment of the accuracy of synchrotron XRFS, quantitative elemental maps of pyrite samples were compared318 to those obtained by LA-ICP-MS and EPMA. A well-characterised, highly homogeneous pyrite sample (CX-15) was employed as the RM for quantification. The accuracy and reliability of the synchrotron-XRFS data were strongly dependent on careful data processing. An overall positive correlation with datasets produced by EPMA and LA-ICP-MS was demonstrated with any differences being attributed to the heterogeneous nature of some of the pyrite grains and the different spot sizes employed. It was concluded that synchrotron XRFS offered complementary capabilities to those of EPMA and LA-ICP-MS. The rapid acquisition of quantitative elemental distributions over a wide range of concentrations enabled large areas (tens of mm to tens of cm) to be studied, something that could not be achieved by either EPMA or LA-ICP-MS independently.

Glossary of abbreviations

2Dtwo dimensional
3Dthree dimensional
AASatomic absorption spectrometry
ACGIHAmerican Conference of Governmental Industrial Hygienists
AECanion exchange chromatography
AESatomic emission spectrometry
AFatomic fluorescence
AFSatomic fluorescence spectrometry
AMSaccelerator mass spectrometry
ANNartificial neural networks
APDCammonium pyrrolidine dithiocarbamate
APGDatmospheric pressure glow discharge
ASUAtomic Spectrometry Update
BARGEBioaccessibility Research Group of Europe
APTatom probe tomography
BASBureau of Analysed Samples
BCRCommunity Bureau of Reference (of the Commission of the European Communities)
BMEMCBeijing Municipal Environmental Protection Monitoring Center
BPNNback-propagation neural network
CAchemical abrasion
CABMCanadian Aerosol Baseline Measurement
CAPMoNCanadian Air and Precipitation Monitoring Network
CASChemical Abstracts Service
CCPcapacitively coupled plasma
CCRMCanadian Certified Reference Material
CEcation exchange
CECcation exchange chromatography
CENEuropean Committee for Standardisation
CFcontinuous flow
CFAcontinuous flow analysis
CIconfidence interval
CNTcarbon nanotube
CPEcloud point extraction
cpscounts per second
CRMcertified reference material
CRPGCentre de Recherches Pétrographiques et Géochimiques (France)
CScontinuum source
CTcomputer tomography
CVGcold vapour generation
DAdiscriminant analysis
DBDdielectric barrier detector
DDTCdiethyldithiocarbamate
dDIHENdemountable direct injection high efficiency nebuliser
DDTPdiethyldithiophosphoric acid
DESdeep eutectic solvent
DGTdiffusive gradients in thin films
DIdual inlet
DLSdynamic light scattering
DLLMEdispersive liquid liquid microextraction
DMAdimethylarsonic acid
DPMdiesel particulate matter
DTPAdiethylenetriamine pentaacetate
EAelemental analyser
ECelemental carbon
EDSenergy dispersive (X-ray) spectrometry
EDTAethylenediaminetetraacetic acid
EDXRFSenergy dispersive X-ray fluorescence spectrometry
EMelectron multiplier
EMPIREuropean Metrology Programme for Innovation and Research
EPMAelectron probe microanalysis
ERMEuropean reference material
ESIelectrospray ionisation
ETAASelectrothermal atomic absorption spectrometry
EtHgethylmercury
ETVelectrothermal vaporisation
EUEuropean Union
FAASflame atomic absorption spectrometry
FCFaraday cup
FFFfield flow fractionation
FIAflow injection analysis
FTIRFourier transform infrared
GCgas chromatography
GDglow discharge
GFgraphite furnace
GLMgeneralised linear model
GOgraphene oxide
GSBZInstitute for Environmental Reference Materials, Ministry of Environmental Protection of China, Beijing, China
GSJGeological Survey of Japan
HENhigh efficiency nebuliser
HFSEhigh field strength element
HGhydride generation
HPLChigh performance liquid chromatography
HPSHigh Purity Standards (USA)
HRhigh resolution
IAEAInternational Atomic Energy Authority
IAGInternational Association of Geoanalysts
ICion chromatography
ICPinductively coupled plasma
IDisotope dilution
IECInternational Electrotechnical Commission
IGGEInstitute of Geophysical and Geochemical Exploration, China
ILionic liquid
IMFinstrumental mass fractionation
IMPROVEInteragency Monitoring of Protected Visual Environments
INCTInstitute of Nuclear Chemistry and Technology (Poland)
IRMMInstitute for Reference Materials and Measurements
IRMSisotope ratio mass spectrometry
ISinternal standard
ISEion selective electrode
ISOInternational Organisation for Standardization
JMCJohnson Matthey Corporation
KEDkinetic energy discrimination
LAlaser ablation
LCliquid chromatography
LDAlinear discriminant analysis
LEAFlaser excited atomic fluorescence
LIBSlaser induced breakdown spectroscopy
LIFlaser-induced fluorescence
LLMEliquid liquid microextraction
LODlimit of detection
LOQlimit of quantification
LREElight rare earth element
LSleast squares
MADmicrowave-assisted digestion
MAEmicrowave-assisted extraction
MALDImatrix-assisted laser desorption and ionisation
MCmulticollector
MDLmethod detection limit
MeHgmethylmercury
MIPmicrowave-induced plasma
MIRmid infrared
MMAmonomethylarsonic acid
MPI-DINGMax Planck Institute
MSMass spectrometry
MS/MSTandem mass spectrometry
MTZMud Tank zircon
MUmeasurement uncertainty
MWCNTmultiwalled carbon nanotube
m/zmass to charge ratio
NACISNational Analysis Centre for Iron and Steel
NCSNCS Testing Co., Ltd. (China)
NDIRnon dispersive infra red
NIESNational Institute for Environmental Studies
NIOSHNational Institute of Occupational Safety and Health
NISTNational Institute of Standards and Technology
NMIANational Measurement Institute of Australia
NPnanoparticle
NRCCNational Research Council of Canada
NRCGANational Research Centre for Geoanalysis (China)
NTIMSnegative thermal ionisation mass spectrometry
NWRINational Water Research Institute
OCorganic carbon
PCAprincipal component analysis
PDOprotected designation of origin
PFAperfluoroalkoxy alkane
PGEplatinum group element
PIXEparticle-induced X-ray emission
PLSpartial least squares
PLSRpartial least square regression
PM0.1particulate matter (with an aerodynamic diameter of up to 0.1 μm)
PM0.5particulate matter (with an aerodynamic diameter of up to 0.5 μm)
PM1particulate matter (with an aerodynamic diameter of up to 1.0 μm)
PM2.5particulate matter (with an aerodynamic diameter of up to 2.5 μm)
PM10particulate matter (with an aerodynamic diameter of up to 10 μm)
PTEpotentially toxic element
PTFEpolytetrafluoroethylene
PVGphotochemical vapour generation
pXRFSportable X-ray fluorescence spectrometry
QCquality control
RCCrotating coiled column
RCSrespirable crystalline silica
REErare earth element
REPrelative error of prediction
RMreference material
RMSEroot mean square error
RNAribonucleic acid
RPDrelative percentage difference
RSDrelative standard deviation
RSFrelative sensitivity factor
SARMService d’Analyses des Roches et des Minéraux (France)
SAXstrong anion exchange
SDstandard deviation
SECsize exclusion chromatography
SEMscanning electron microscopy
SFsector field
SFODMEsolidified floating organic drop microextraction
SHRIMPsensitive high resolution ion microprobe
SISystème International (d’unités)
SIMSsecondary ion mass spectrometry
S/Nsignal to noise ratio
SOMsoil organic matter
spsingle particle
SPEsolid-phase extraction
SPMEsolid-phase microextraction
SQTslotted quartz tube
SRMstandard reference material
SSsolid sampling
SSBsample standard bracketing
SVMsupport vector machine
SVRsupport vector regression
TCtotal carbon
TDthermal desorption
TEMtransmission electron microscopy
TLCthin layer chromatography
TIMSthermal ionisation mass spectrometry
TOFtime-of-flight
TXRFtotal reflection X-ray fluorescence
TXRFStotal reflection X-ray fluorescence spectrometry
UAultrasound-assisted
UBMunified bioaccessibility method
UNCUniversity of North Carolina
US EPAUnited States Environmental Protection Agency
USGSUnited States Geological Survey
UVultraviolet
VAMEvortex-assisted microextraction
VIS-NIRvisible near infrared
VPDBVienna Pee Bee Belemnite
WDXRFSwavelength dispersive X-ray fluorescence spectrometry
XAFSX-ray absorption fine structure
XANESX-ray absorption near edge structure
XASX-ray absorption spectroscopy
XFMX-ray fluorescence microscopy
XRDX-ray diffraction
XRFX-ray fluorescence
XRFSX-ray fluorescence spectrometry
Zatomic number

Conflicts of interest

There are no conflicts to declare.

References

  1. J. R. Bacon, O. T. Butler, W. R. L. Cairns, J. M. Cook, C. M. Davidson, O. Cavoura and R. Mertz-Kraus, J. Anal. At. Spectrom., 2020, 35(1), 9–53 RSC.
  2. A. Taylor, A. Catchpole, M. P. Day, S. Hill, N. Martin and M. Patriarca, J. Anal. At. Spectrom., 2020, 35(3), 426–454 RSC.
  3. E. H. Evans, J. Pisonero, C. M. Smith and R. N. Taylor, J. Anal. At. Spectrom., 2020, 35(5), 830–851 RSC.
  4. R. Clough, C. F. Harrington, S. J. Hill, Y. Madrid and J. F. Tyson, J. Anal. At. Spectrom., 2020, 35(7), 1236–1278 RSC.
  5. C. Vanhoof, J. R. Bacon, U. E. A. Fittschen and L. Vincze, J. Anal. At. Spectrom., 2020, 35(9), 1704–1719 RSC.
  6. S. Carter, R. Clough, A. Fisher, B. Gibson, B. Russell and J. Waack, J. Anal. At. Spectrom., 2019, 34(11), 2159–2216 RSC.
  7. G. L. Donati and R. S. Amais, J. Anal. At. Spectrom., 2019, 34(12), 2353–2369 RSC.
  8. R. C. Machado, D. F. Andrade, D. V. Babos, J. P. Castro, V. C. Costa, M. A. Speranca, J. A. Garcia, R. R. Gamela and E. R. Pereira, J. Anal. At. Spectrom., 2020, 35(1), 54–77 RSC.
  9. F. Ben Borsh, D. K. Sleeth, R. Handy, L. F. Pahler, R. Andrews and K. Ashley, J. Occup. Environ. Hyg., 2019, 16(9), 634–642 CrossRef.
  10. D. Leith, C. L’Orange, J. Mehaffy and J. Volckens, J. Occup. Environ. Hyg., 2020, 17(6), 274–282 CrossRef.
  11. S. H. Huang, Y. M. Kuo, C. W. Lin, T. J. Chen, J. G. Liu, H. Q. Gui and C. C. Chen, J. Occup. Environ. Hyg., 2019, 16(12), 785–792 CrossRef.
  12. W. C. Su, Y. Chen, M. Bezerra and J. Wang, J. Occup. Environ. Hyg., 2019, 16(10), 694–706 CrossRef CAS.
  13. X. H. Li, B. Ruan, P. K. Hopke and T. Mehmood, Aerosol Air Qual. Res., 2019, 19(10), 2173–2184 CrossRef CAS.
  14. N. Kumsanlas, S. Piriyakarnsakul, P. Sok, S. Hongtieab, F. Ikemori, W. W. Szymanski, M. Hata, Y. Otani and M. Furuuchi, Aerosol Air Qual. Res., 2019, 19(8), 1666–1677 CrossRef CAS.
  15. G. J. Parker, C. H. Ong, R. B. Manges, E. M. Stapleton, A. P. Comellas, T. M. Peters and E. A. Stone, Aerosol Air Qual. Res., 2019, 19(11), 2387–2395 CrossRef CAS.
  16. M. Catrambone, S. Canepari, M. Cerasa, T. Sargolini and C. Perrino, Aerosol Air Qual. Res., 2019, 19(10), 2160–2172 CrossRef CAS.
  17. A. Waza, K. Schneiders, J. May, S. Rodriguez, B. Epple and K. Kandler, Atmos. Meas. Tech., 2019, 12(12), 6647–6665 CrossRef.
  18. M. Shirdel, I. A. Bergdahl, B. M. Andersson, H. Wingfors, J. N. Sommar and I. E. Liljelind, J. Occup. Environ. Hyg., 2019, 16(10), 675–684 CrossRef.
  19. A. Yamakawa, S. Berail, D. Amouroux, E. Tessier, J. Barre, T. Sano, K. Nagano, S. Kanwal, J. Yoshinaga and O. F. X. Donard, Anal. Bioanal. Chem., 2020, 412, 4483–4493 CrossRef CAS.
  20. A. Vanderstraeten, S. Bonneville, S. Gili, J. de Jong, W. Debouge, P. Claeys and N. Mattielli, Geostand. Geoanal. Res., 2020, 44(2), 307–329 CrossRef CAS.
  21. N. J. Schleicher, S. F. Dong, H. Packman, S. H. Little, R. O. Gonzalez, J. Najorka, Y. B. Sun and D. J. Weiss, Front. Earth Sci., 2020, 8 DOI:10.3389/feart.2020.00167.
  22. S. Yatkin, K. Trzepla, N. P. Hyslop, W. H. White, O. Butler, T. Ancelet, P. Davy, M. Gerboles, S. D. Kohl, A. McWilliams, L. Saucedo, M. Van Der Haar and A. Jonkers, Microchem. J., 2020, 158 DOI:10.1016/j.microc.2020.105225.
  23. S. E. Long, J. E. Norris, J. Carney, J. V. Ryan, G. D. Mitchell and W. D. Dorko, Atmos. Pollut. Res., 2020, 11(4), 639–645 CrossRef CAS.
  24. S. Sari, R. Timo, H. Jussi and H. Panu, Meas. Sci. Technol., 2019, 31(3) DOI:10.1088/1361-6501/ab1084d1068.
  25. J. W. Birks, A. A. Turnipseed, P. C. Andersen, C. J. Williford, S. Strunk, B. J. Carpenter and C. Ennis, Atmos. Meas. Tech., 2020, 13(2), 1001–1018 CrossRef CAS.
  26. M. M. G. Perron, M. Strzelec, M. Gault-Ringold, B. C. Proernse, P. W. Boyd and A. R. Bowie, Talanta, 2020, 208,  DOI:10.1016/j.talanta.2019.120377.
  27. M. Budanovic, M. E. Tessensohn and R. D. Webster, Environ. Pollut., 2019, 252, 637–643 CrossRef CAS.
  28. B. S. Sagagi, C. M. Davidson and A. S. Hursthouse, Int. J. Environ. Anal. Chem., 2019 DOI:10.1080/03067319.03062019.01674847.
  29. U. Dusek, M. M. Cosijn, H. Ni, R. J. Huang, H. A. J. Meijer and S. van Heuven, Aerosol Air Qual. Res., 2019, 19(11), 2604–2611 CrossRef CAS.
  30. H. W. Paing, B. T. Manard, B. W. Ticknor, J. R. Bills, K. A. Hall, D. A. Bostick, P. Cable-Dunlap and R. K. Marcus, Anal. Chem., 2020, 92(12), 8591–8598 CrossRef CAS.
  31. L. Borgese, F. Bilo, A. Zacco, S. Federici, A. W. Mutahi, E. Bontempi, K. Trzepla, N. Hyslop, S. Yatkin, P. Wobrauschek, J. Prost, D. Ingerle and L. E. Depero, Spectrochim. Acta, Part B, 2020, 167 DOI:105810.101016/j.sab.102020.105840.
  32. J. Svehla, R. Zidek, T. Ruzovic, K. Svoboda and J. Kratzer, Spectrochim. Acta, Part B, 2019, 156, 51–58 CrossRef CAS.
  33. D. Chen, Z. X. Huang, T. Wang, Y. W. Ma, Y. Zhang, G. W. Wang and P. F. Zhang, Anal. Chim. Acta, 2020, 1095, 14–19 CrossRef CAS.
  34. Z. Z. Wang, R. W. Liu, Y. Deguchi, S. Tanaka, K. Tainaka, K. Tanno, H. Watanabe, J. J. Yan and J. P. Liu, Energy Fuels, 2019, 33(8), 7805–7812 CrossRef CAS.
  35. A. Ari, P. E. Ari and E. O. Gaga, Talanta, 2020, 208,  DOI:10.1016/j.talanta.2019.120350.
  36. S. Diez-Fernandez, H. Jaegler, C. Bresson, F. Chartier, O. Evrard, A. Hubert, A. Nonell, F. Pointurier and H. Isnard, Talanta, 2020, 206,  DOI:10.1016/j.talanta.2019.120221.
  37. D. Foppiano, M. Tarik, J. Schneebeli, A. Calbry-Muzyka, S. Biollaz and C. Ludwig, Talanta, 2020, 208,  DOI:120310.121016/j.talanta.122019.120398.
  38. S. E. Long, J. E. Norris, J. Carney and J. V. Ryan, Atmos. Pollut. Res., 2020, 11(5), 909–919 CrossRef CAS.
  39. T. Erhardt, C. M. Jensen, O. Borovinskaya and H. Fischer, Environ. Sci. Technol., 2019, 53(22), 13275–13283 CrossRef.
  40. M. Hadioui, G. Knapp, A. Azimzada, I. Jreije, L. Frechette-Viens and K. J. Wilkinson, Anal. Chem., 2019, 91(20), 13275–13284 CrossRef CAS.
  41. G. Craig, M. S. A. Horstwood, H. J. Reid and B. L. Sharp, J. Anal. At. Spectrom., 2020, 35(5), 1011–1021 RSC.
  42. S. Giannoukos, C. P. Lee, M. Tarik, C. Ludwig, S. Biollaz, H. Lamkaddam, U. Baltensperger, A. S. H. Prevot and J. Slowik, Anal. Chem., 2020, 92(1), 1316–1325 CrossRef CAS.
  43. S. Saarikoski, L. R. Williams, S. R. Spielman, G. S. Lewis, A. Eiguren-Fernandez, M. Aurela, S. V. Hering, K. Teinila, P. Croteau, J. T. Jayne, T. Hohaus, D. R. Worsnop and H. Timonen, Atmos. Meas. Tech., 2019, 12(7), 3907–3920 CrossRef CAS.
  44. J. Osan, E. Borcsok, O. Czompoly, C. Dian, V. Groma, L. Stabile and S. Torok, Spectrochim. Acta, Part B, 2020, 167 DOI:10.1016/j.sab.2020.105852.
  45. T. W. Chan, L. Huang, K. Banwait, W. Zhang, D. Ernst, X. L. Wang, J. G. Watson, J. C. Chow, M. Green, C. I. Czimczik, G. M. Santos, S. Sharma and K. Jones, Atmos. Meas. Tech., 2019, 12(8), 4543–4560 CrossRef CAS.
  46. T. Haller, C. Rentenberger, J. C. Meyer, L. Felgitsch, H. Grothe and R. Hitzenberger, Atmos. Meas. Tech., 2019, 12(7), 3503–3519 CrossRef CAS.
  47. C. D. Klingshirn, Z. J. West, M. J. DeWitt, A. Higgins, J. Graham and E. Corporan, J. Air Waste Manage. Assoc., 2019, 69(8), 1003–1013 CrossRef CAS.
  48. F. S. Liu, J. Yon, A. Fuentes, P. Lobo, G. J. Smallwood and J. C. Corbin, Aerosol Sci. Technol., 2020, 54(1), 33–51 CrossRef CAS.
  49. G. Chen, Q. Wang, Y. F. Fan, Y. Q. Han, Y. W. Wang, B. Urch, F. Silverman, M. Tian, Y. S. Su, X. H. Qiu, T. Zhu and A. W. H. Chan, Atmos. Environ., 2020, 224 DOI:10.1016/j.atmosenv.2020.117291.
  50. M. Jeronimo, Q. Stewart, A. T. Weakley, J. Giacomo, X. L. Zhang, N. Hyslop, A. M. Dillner, M. Shupler and M. Brauer, Atmos. Environ., 2020, 223 DOI:10.1016/j.atmosenv.2020.117300.
  51. P. O’Shaughnessy, A. Stoltenberg, C. Holder and R. Altmaier, J. Occup. Environ. Hyg., 2020, 17(6), 262–273 CrossRef.
  52. C. Barrett, E. Sarver, E. Cauda, J. Noll, S. Vanderslice and J. Volkwein, Aerosol Air Qual. Res., 2019, 19(11), 2367–2380 CrossRef.
  53. Z. Y. Gong, Y. L. Pan, G. Videen and C. J. Wang, Appl. Spectrosc., 2019, 73(8), 910–916 CrossRef CAS.
  54. P. Stacey, F. Clegg, J. Morton and C. Sammon, Anal. Methods, 2020, 12(21), 2757–2771 RSC.
  55. E. L. Ashley, E. Cauda, L. G. Chubb, D. P. Tuchman and E. N. Rubinstein, Ann. Work Exposures Health, 2020, 64(5), 536–546 CrossRef.
  56. D. Yeghicheyan, D. Aubert, M. Bouhnik-Le Coz, J. Chmeleff, S. Delpoux, I. Djouraev, G. Granier, F. Lacan, J. L. Piro, T. Rousseau, C. Cloquet, A. Marquet, C. Menniti, C. Pradoux, R. Freydier, E. V. da Silva and K. Suchorski, Geostand. Geoanal. Res., 2019, 43(3), 475–496 CrossRef CAS.
  57. E. Vassileva, A. A. Krata and S. Azemard, Accredit. Qual. Assur., 2020, 25(3), 221–231 CrossRef CAS.
  58. Y. Gao, C. Y. Zhou, C. Gaulier, A. Bratkic, J. Galceran, J. Puy, H. Zhang, M. Leermakers and W. Baeyens, TrAC, Trends Anal. Chem., 2019, 116, 92–101 CrossRef CAS.
  59. I. V. Pletnev, S. V. Smirnova and N. V. Shvedene, J. Anal. Chem., 2019, 74(7), 625–658 CrossRef CAS.
  60. M. L. Tian, L. W. Fang, X. M. Yan, W. Xiao and K. H. Row, J. Anal. Methods Chem., 2019, 2019 DOI:10.1155/2019/1948965.
  61. F. Ardini, G. Dan and M. Grotti, J. Anal. At. Spectrom., 2020, 35(2), 215–237 RSC.
  62. M. Tanaka, Y. Hoshino and D. Nakamoto, Rapid Commun. Mass Spectrom., 2019, 33(19), 1527–1536 CrossRef CAS.
  63. H. F. Hou, W. J. Cui, Q. Xu, Z. H. Tao, Y. F. Guo and T. L. Deng, Int. J. Anal. Chem., 2019 DOI:10.1155/2019/3280840.
  64. R. T. Wilkin, R. G. Ford, L. M. Costantino, R. R. Ross, D. G. Beak and K. G. Scheckel, Environ. Sci. Technol., 2019, 53(20), 11684–11693 CrossRef CAS.
  65. B. Lajin and W. Goessler, Anal. Chim. Acta, 2020, 1094, 11–17 CrossRef CAS.
  66. J. R. Miranda-Andrades, S. Khan, M. J. Pedrozo-Penafiel, K. D. B. Alexandre, R. M. Maciel, R. Escalfoni, M. L. B. Tristao and R. Q. Aucelio, Spectrochim. Acta, Part B, 2019, 158 DOI:10.1016/j.sab.2019.105641.
  67. D. S. Hardisty, T. J. Horner, S. D. Wankel, J. Blusztajn and S. G. Nielsen, Chem. Geol., 2020, 532,  DOI:10.1016/j.chemgeo.2019.119360.
  68. E. Muller, U. von Gunten, S. Bouchet, B. Droz and L. H. E. Winkel, Environ. Sci. Technol., 2019, 53(22), 13146–13157 CrossRef CAS.
  69. A. S. Kisomi, T. Alizadeh and A. Shakeri, Microchim. Acta, 2020, 187(5) DOI:10.1007/s00604-00020-04260-00604.
  70. E. Pagliano, K. L. LeBlanc and Z. Mester, Anal. Chem., 2019, 91(19), 12162–12166 CrossRef CAS.
  71. M. Czaplicka, K. Jaworek and J. Klyta, Desalin. Water Treat., 2019, 172, 386–394 CrossRef CAS.
  72. J. X. Li, Q. Q. He, L. J. Wu, J. Sun, F. Zheng, L. Li, W. Y. Liu and J. Liu, Microchem. J., 2020, 153 DOI:10.1016/j.microc.2019.104459.
  73. M. Metzger, P. Ley, M. Sturm and B. Meermann, Anal. Bioanal. Chem., 2019, 411(19), 4647–4660 CrossRef CAS.
  74. P. Tan, J. J. Yang and V. Nischwitz, J. Anal. At. Spectrom., 2020, 35(3), 548–559 RSC.
  75. V. V. Eskina, V. B. Baranovskaya, Y. A. Karpov and D. G. Filatova, Russ. Chem. Bull., 2020, 69(1), 1–16 CrossRef CAS.
  76. N. A. Sobolev, N. L. Ivanchenko and A. Y. Kozhevnikov, J. Anal. Chem., 2019, 74(5), 444–448 CrossRef CAS.
  77. B. S. Pinheiro, A. J. Moreira, L. L. S. Gimenes, C. D. Freschi and G. P. G. Freschi, Environ. Monit. Assess., 2020, 192(6) DOI:10.1007/s10661-10020-08316-z.
  78. Y. Yu, Q. Zhao, H. L. Bao, Q. Mou, Z. M. Shi, Y. L. Chen and Y. Gao, Geostand. Geoanal. Res., 2020, 44(3), 617–627 CrossRef CAS.
  79. B. Stadlerova, M. Kolrosova, J. Dedina and S. Musil, J. Anal. At. Spectrom., 2020, 35(5), 993–1002 RSC.
  80. S. Liu, J. S. Sun, J. Z. Huo, X. C. Duan and Y. Liu, Anal. Sci., 2019, 35(5), 499–504 CrossRef CAS.
  81. S. W. Smith, N. Hanks, P. A. Creed, K. Kovalcik, R. A. Wilson, K. Kubachka, J. A. Brisbin, J. L. Figueroa and J. T. Creed, J. Anal. At. Spectrom., 2019, 34(10), 2094–2104 RSC.
  82. M. Y. He, L. Deng, H. Lu and Z. D. Jin, J. Anal. At. Spectrom., 2019, 34(5), 1026–1032 RSC.
  83. K. Li, C. J. Lin, W. Yuan, G. Y. Sun, X. W. Fu and X. B. Feng, J. Anal. At. Spectrom., 2019, 34(11), 2303–2313 RSC.
  84. M. Martinez, J. Garcia-Alonso, C. Parat, J. R. Encinar and I. Le-Hecho, Anal. Chem., 2019, 91(15), 10088–10094 CrossRef CAS.
  85. L. W. Yang, C. H. Liu and T. Yang, RSC Adv., 2019, 9(54), 31224–31232 RSC.
  86. H. N. Abdelhamid and H. F. Wu, J. Am. Soc. Mass Spectrom., 2019, 30(12), 2617–2622 CrossRef CAS.
  87. L. Ripoll and M. Hidalgo, J. Anal. At. Spectrom., 2019, 34(10), 2016–2026 RSC.
  88. S. X. Ma, Y. Tang, Y. Y. Ma, D. M. Dong, L. B. Guo, H. H. Zhu, J. G. Liu and Y. F. Lu, J. Anal. At. Spectrom., 2020, 35(1), 198–203 RSC.
  89. L. Borgese, F. Bilo, S. Federici, E. Margui, T. Hase, Y. Y. Huang, B. Beckhoff and L. E. Depero, Surf. Interface Anal., 2020, 52(3), 119–123 CrossRef CAS.
  90. S. Pessanha, E. Margui, M. L. Carvalho and I. Queralt, Spectrochim. Acta, Part B, 2020, 164,  DOI:10.1016/j.sab.2019.105762.
  91. M. Tighe, M. Bielski, M. Wilson, G. Ruscio-Atkinson, G. F. Peaslee and M. Lieberman, Anal. Chem., 2020, 92(7), 4949–4953 CrossRef CAS.
  92. J. Zhang, D. Baralkiewicz, Y. Z. Wang, J. Falandysz and C. T. Cai, Chemosphere, 2020, 246,  DOI:10.1016/j.chemosphere.2019.125685.
  93. D. J. Butcher, Appl. Spectrosc. Rev., 2020, 55(1), 60–75 CrossRef CAS.
  94. E. S. Ravipati, N. N. Mahajan, S. Sharma, K. V. Hatware and K. Patil, Crit. Rev. Anal. Chem., 2019 DOI:10.1080/10408347.10402019.11678381.
  95. S. Etteieb, S. Magdouli, M. Zolfaghari and S. Brar, Sci. Total Environ., 2020, 698,  DOI:10.1016/j.scitotenv.2019.134339.
  96. M. Filella, C. Reimann, M. Biver, I. Rodushkin and K. Rodushkina, Environ. Chem., 2019, 16(4), 215–228 CrossRef CAS.
  97. P. M. Kopittke, E. Lombi, A. van der Ent, P. Wang, J. S. Laird, K. L. Moore, D. P. Persson and S. Husted, Plant Physiol., 2020, 182(4), 1869–1882 CrossRef CAS.
  98. S. Nawar, S. Cipullo, R. K. Douglas, F. Coulon and A. M. Mouazen, Appl. Spectrosc. Rev., 2019, 55(7), 525–557 CrossRef.
  99. Y. J. An, X. Li and Z. F. Zhang, Geostand. Geoanal. Res., 2020, 44(1), 183–199 CrossRef CAS.
  100. A. A. Krata, M. Wojciechowski and E. Bulska, Microchem. J., 2019, 147, 674–681 CrossRef CAS.
  101. V. Roje and F. Galinec, Environ. Monit. Assess., 2019, 191(9) DOI:10.1007/s10661-10019-17677-10668.
  102. M. S. Liu, Q. Zhang, Y. N. Zhang, Z. F. Zhang, F. Huang and H. M. Yu, Geostand. Geoanal. Res., 2020, 44(1), 169–182 CrossRef CAS.
  103. J. L. Clausen, S. A. Beal, T. Georgian, K. H. Gardner, T. A. Douglas and A. M. Mossell, Microchem. J., 2020, 154 DOI:10.1016/j.microc.2019.104583.
  104. M. A. Santos, A. B. S. Silva, R. C. Machado, E. A. Dias, J. Barros and A. R. A. Nogueira, Spectrochim. Acta, Part B, 2020, 167 DOI:10.1016/j.sab.2020.105842.
  105. D. P. Mokoena, S. V. Mngadi, M. Sihlahla, M. K. Dimpe and P. N. Nomngongo, Soil Sediment Contam., 2019, 28(3), 323–333 CrossRef CAS.
  106. Y. Nuapia, S. Al-Hamimi, P. G. Matshediso, E. Cukrowska, H. Tutu, C. Turner and L. Chimuka, Anal. Bioanal. Chem., 2020, 412, 2495–2503 CrossRef CAS.
  107. F. S. Costa, R. V. P. Leal, C. S. V. Pacheco, F. A. C. Amorim, R. M. de Jesus, L. N. Santos and E. G. P. da Silva, Anal. Methods, 2020, 12(19), 2509–2516 RSC.
  108. D. Medeiros, F. Piechontcoski, E. Watanabe, E. S. Chaves and S. D. Inglez, Environ. Monit. Assess., 2020, 192(2) DOI:10.1007/s10661-10020-18065-10660.
  109. H. S. Ferreira, S. S. Oliveira, D. Santos, K. B. Fontana, T. A. Maranhao, T. S. Almeida and R. G. O. Araujo, Microchem. J., 2019, 151 DOI:10.1016/j.microc.2019.104196.
  110. A. L. Chen, T. H. Hansen, L. I. Olsen, M. Palmgren, S. Husted, J. K. Schjoerring and D. P. Persson, Plant Methods, 2020, 16(1) DOI:10.1186/s13007-13020-00566-13009.
  111. A. D. Sun, D. D. Gou, Y. L. Dong, Q. C. Xu and G. J. Cao, J. Agric. Food Chem., 2019, 67(25), 7183–7189 CrossRef CAS.
  112. O. S. Humphrey, S. D. Young, E. H. Bailey, N. M. J. Crout, E. L. Ander, E. M. Hamilton and M. J. Watts, Chemosphere, 2019, 229, 41–50 CrossRef CAS.
  113. L. Cheng, X. A. Yang, M. T. Shi and W. B. Zhang, J. Chromatogr., 2020, 1619,  DOI:10.1016/j.chroma.2020.460915.
  114. S. Tokalioglu, R. Clough, M. Foulkes and P. Worsfold, Biol. Trace Elem. Res., 2020 DOI:10.1007/s12011-12020-02066-12012.
  115. A. Klotzbucher, K. Kaiser, T. Klotzbucher, M. Wolff and R. Mikutta, J. Plant Nutr. Soil Sci., 2019, 182(4), 570–577 CrossRef.
  116. Y. P. Gao, Y. Yang, L. Li, W. J. Wei, H. Xu, Q. Wang and Y. Q. Qiu, Anal. Chim. Acta, 2020, 1110, 72–81 CrossRef CAS.
  117. L. Li, Q. Wang, Y. Yang, L. Luo, R. Ding, Z. G. Yang and H. P. Li, Anal. Chem., 2019, 91(15), 9442–9450 CrossRef CAS.
  118. S. Laughton, A. Laycock, F. von der Kammer, T. Hofmann, E. A. Casman, S. M. Rodrigues and G. V. Lowry, J. Nanopart. Res., 2019, 21(8) DOI:10.1007/s11051-11019-14620-11054.
  119. M. S. Ermolin, N. N. Fedyunina, V. K. Karandashev and P. S. Fedotov, J. Anal. Chem., 2019, 74(8), 825–833 CrossRef CAS.
  120. V. Cerda, TrAC, Trends Anal. Chem., 2019, 118, 352–367 CrossRef CAS.
  121. Z. Y. Huang, Y. Y. Ni, H. Wang, J. Zheng, S. Yamazaki, A. Sakaguchi, X. G. Long and S. Uchida, Microchem. J., 2019, 148, 597–604 CrossRef CAS.
  122. Z. T. Wang, Z. Y. Huang, Y. Xie and Z. Y. Tan, Anal. Chim. Acta, 2019, 1090, 151–158 CrossRef CAS.
  123. J. A. Barreto, V. A. Lemos, D. M. de Oliveira, U. Cerqueira, L. A. Meira and M. A. Bezerra, J. Braz. Chem. Soc., 2020, 31(1), 15–24 CAS.
  124. D. Schumann, N. Kivel and R. Dressler, PLoS One, 2019, 14(6) DOI:10.1371/journal.pone.0219039.
  125. H. A. M. De La Torre, A. V. Reyes, G. D. Zazula, D. G. Froese, B. J. L. Jensen and J. R. Southon, Nucl. Instrum. Methods Phys. Res., Sect. B, 2019, 455, 154–157 CrossRef.
  126. T. Renz, M. Plaschke, F. Quinto, A. Bauer, M. Lagos, H. Taubald and H. Geckeis, Clays Clay Miner., 2019, 67(3), 183–189 CrossRef CAS.
  127. T. Prasek, M. Nemec, P. Steier, M. Kern, M. Honda, K. Hain and X. Zhang, Nucl. Instrum. Methods Phys. Res., Sect. B, 2020, 472, 64–71 CrossRef CAS.
  128. P. Coufalik, A. Uher, O. Zverina and J. Komarek, Environ. Monit. Assess., 2020, 192(4) DOI:10.1007/s10661-10020-18186-10665.
  129. N. A. Kasa, C. Buyukpinar, A. F. Erulas and S. Bakirdere, Food Chem., 2020, 319,  DOI:10.1016/j.foodchem.2020.126572.
  130. L. D. Santos, H. Bley, L. P. R. Seau, T. Tiecher and C. Gianello, Commun. Soil Sci. Plant Anal., 2019, 50(14), 1671–1682 CrossRef CAS.
  131. S. B. Angyus, E. Darvasi, M. Ponta, D. Petreus, R. Etz, M. Senila, M. Frentiu and T. Frentiu, Talanta, 2020, 217,  DOI:10.1016/j.talanta.2020.121067.
  132. G. Drava and V. Minganti, J. Anal. At. Spectrom., 2020, 35(2), 301–306 RSC.
  133. G. Heltai, Z. Gyori, I. Fekete, G. Halasz, K. Kovacs, A. Takacs, L. Khumalo and M. Horvath, Microchem. J., 2019, 149 DOI:10.1016/j.microc.2019.104029.
  134. X. X. Peng and Z. Wang, Anal. Chem., 2019, 91(15), 10073–10080 CrossRef CAS.
  135. J. L. Dong, R. Qian, S. J. Zhuo, P. F. Yu, Q. Chen and Z. Q. Li, J. Anal. At. Spectrom., 2019, 34(11), 2244–2251 RSC.
  136. M. X. Wan, W. Y. Hu, M. K. Qu, K. Tian, H. D. Zhang, Y. Wang and B. Huang, Ecol. Indic., 2019, 101, 583–594 CrossRef CAS.
  137. N. P. Zaksas, Appl. Spectrosc., 2019, 73(8), 870–878 CrossRef CAS.
  138. A. Al Hejami and D. Beauchemin, J. Anal. At. Spectrom., 2019, 34(7), 1426–1432 RSC.
  139. G. L. Scheffler, D. Pozebon and D. Beauchemin, J. Anal. At. Spectrom., 2019, 34(5), 891–898 RSC.
  140. H. Matusiewicz and M. Slachcinski, Ecol. Chem. Eng. S, 2019, 26(3), 429–441 CAS.
  141. L. Sera, M. Loula, S. Matejkova and O. Mestek, Chem. Pap., 2019, 73(12), 3005–3017 CrossRef CAS.
  142. M. T. Liu, J. X. Liu, X. F. Mao, X. Na, L. Ding and Y. Z. Qian, Anal. Chem., 2020, 92(10), 7257–7264 CrossRef CAS.
  143. H. Ben Yaala, R. Fniter, D. Foucher and O. Clarisse, J. Anal. At. Spectrom., 2019, 34(8), 1597–1605 RSC.
  144. C. Moens, N. Waegeneers, A. Fritzsche, P. Nobels and E. Smolders, J. Chromatogr. A, 2019, 1599, 203–214 CrossRef CAS.
  145. M. K. Park, M. Choi, L. Kim and S. D. Choi, Environ. Monit. Assess., 2019, 191(8) DOI:10.1007/s10661-10019-17675-x.
  146. M. Radaelli, E. Scalabrin, G. Toscano and G. Capodaglio, Molecules, 2019, 24(17) DOI:10.3390/molecules24173201.
  147. N. N. Wanna, K. Van Hoecke, A. Dobney, M. Vasile, T. Cardinaels and F. Vanhaecke, J. Chromatogr., 2020, 1617,  DOI:10.1016/j.chroma.2019.460839.
  148. A. A. Krata and E. Vassileva, Talanta, 2020, 217,  DOI:10.1016/j.talanta.2020.121113.
  149. W. Y. Chu, Y. M. Weng, S. J. Jiang and A. C. Sahayam, Spectrochim. Acta, Part B, 2019, 157, 63–67 CrossRef CAS.
  150. Y. Cui, L. L. Jin, H. L. Li, S. H. Hu and Y. Lian, At. Spectrosc., 2020, 41(2), 87–92 CrossRef.
  151. H. Cui, W. Guo, L. L. Jin, Y. E. Peng and S. H. Hu, J. Anal. At. Spectrom., 2020, 35(3), 592–599 RSC.
  152. Y. H. Loni, K. David, S. Larrue, B. Grambow, C. Corona, S. Ribet, P. Chardon and G. Montavon, Spectrochim. Acta, Part B, 2019, 161,  DOI:10.1016/j.sab.2019.105709.
  153. S. Geilert, J. Vogl, M. Rosner and T. Eichert, Rapid Commun. Mass Spectrom., 2019, 33(13), 1137–1147 CrossRef CAS.
  154. L. Lancellotti, S. Sighinolfi, A. Marchetti and L. Tassi, Molecules, 2020, 25(7) DOI:10.3390/molecules25071641.
  155. N. L. Wasserman and T. M. Johnson, J. Anal. At. Spectrom., 2020, 35(2), 307–319 RSC.
  156. N. Veerasamy, A. Takamasa, R. Murugan, S. Kasar, T. Aono, K. Inoue, M. Fukushi and S. K. Sahoo, Molecules, 2020, 25(9) DOI:10.3390/molecules25092138.
  157. T. Durr-Auster, M. Wiggenhauser, C. Zeder, R. Schulin, D. J. Weiss and E. Frossard, Front. Plant Sci., 2019, 10 DOI:10.3389/fpls.2019.01382.
  158. X. L. Hou, W. C. Zhang and Y. Y. Wang, Anal. Chem., 2019, 91(18), 11553–11561 CrossRef CAS.
  159. L. Fu, H. L. Xie, J. H. Huang and L. Chen, Biol. Trace Elem. Res., 2020 DOI:10.1007/s12011-12020-02175-y.
  160. P. Skacelova, P. Lebed, J. Filip, D. Oughton and R. Zboril, Environ. Sci.: Nano, 2020, 7(2), 623–633 RSC.
  161. G. Nicolodelli, J. Cabral, C. R. Menegatti, B. Marangoni and G. S. Senesi, TrAC, Trends Anal. Chem., 2019, 115, 70–82 CrossRef CAS.
  162. G. S. Senesi, J. Cabral, C. R. Menegatti, B. Marangoni and G. Nicolodelli, TrAC, Trends Anal. Chem., 2019, 118, 453–469 CrossRef CAS.
  163. P. R. Villas-Boas, M. A. Franco, L. Martin-Neto, H. T. Gollany and D. Milori, Eur. J. Soil Sci., 2020, 71(5), 789–804 CAS.
  164. P. R. Villas-Boas, M. A. Franco, L. Martin-Neto, H. T. Gollany and D. Milori, Eur. J. Soil Sci., 2020, 71(5), 805–818 CAS.
  165. Y. Yu, N. J. Zhao, Z. G. Lan, D. S. Meng and M. J. Ma, Plasma Sci. Technol., 2020, 22(1) DOI:10.1088/2058-6272/ab1084c1086a.
  166. K. Q. Yu, Y. R. Zhao, Y. He and D. J. He, Chemom. Intell. Lab. Syst., 2019, 195,  DOI:10.1016/j.chemolab.2019.103891.
  167. R. X. Yi, X. Y. Yang, F. R. Lin and S. Ren, Appl. Opt., 2019, 58(32), 8824–8828 CrossRef CAS.
  168. M. Akhtar, A. Jabbar, S. Mahmood, Z. A. Umar, R. Ahmed and M. A. Baig, Anal. Lett., 2019, 52(14), 2312–2328 CrossRef CAS.
  169. M. Akhtar, A. Jabbar, N. Ahmed, S. Mahmood, Z. A. Umar, R. Ahmed and M. A. Baig, Appl. Phys. B, 2019, 125(6) DOI:10.1007/s00340-00019-07225-00349.
  170. S. Y. Zhao, X. Gao, A. M. Chen and J. Q. Lin, Appl. Phys. B, 2019, 126(1) DOI:10.1007/s00340-00019-07354-00341.
  171. C. Sun, Y. Tian, L. Gao, Y. S. Niu, T. L. Zhang, H. Li, Y. Q. Zhang, Z. Q. Yue, N. Delepine-Gilon and J. Yu, Sci. Rep., 2019, 9 DOI:10.1038/s41598-41019-47751-y.
  172. C. Wu, D. X. Sun, M. G. Su, Y. P. Yin, W. W. Han, Q. F. Lu and C. Z. Dong, J. Anal. At. Spectrom., 2019, 34(7), 1478–1484 RSC.
  173. C. P. Lu, M. Wang, L. S. Wang, H. Hu and R. J. Wang, Appl. Opt., 2019, 58(27), 7510–7516 CrossRef CAS.
  174. V. C. Costa, S. D. Ferreira, L. N. Santos, M. A. Speranca, C. S. da Silva, G. A. Sodre and E. R. Pereira, J. Appl. Spectrosc., 2020, 87(2), 378–386 CrossRef CAS.
  175. G. M. Guo, G. H. Niu, Q. Shi, Q. Y. Lin, D. Tian and Y. X. Duan, Anal. Methods, 2019, 11(23), 3006–3013 RSC.
  176. X. L. Fu, S. X. Ma, G. L. Li, L. B. Guo and D. M. Dong, Spectrochim. Acta, Part B, 2020, 167 DOI:10.1016/j.sab.2020.105817.
  177. M. C. Pelagio, D. A. Navarro, L. J. Janik and R. B. Lamorena, Chemistryselect, 2020, 5(13), 3798–3804 CrossRef CAS.
  178. M. J. Zhao, C. H. Yan, Y. H. Feng, J. Xue, H. S. Tang, T. L. Zhang and H. Li, Plasma Sci. Technol., 2020, 22(7) DOI:10.1088/2058-6272/ab1086ac1082.
  179. C. L. Goueguel, A. Soumare, C. Nault and J. Nault, J. Anal. At. Spectrom., 2019, 34(8), 1588–1596 RSC.
  180. A. Jabbar, M. Akhtar, S. Mehmood, N. Ahmed, Z. A. Umar, R. Ahmed and M. A. Baig, J. Anal. At. Spectrom., 2019, 34(5), 954–962 RSC.
  181. R. R. Gamela, M. A. Speranca, D. F. Andrade and E. R. Pereira, Anal. Methods, 2019, 11(43), 5543–5552 RSC.
  182. R. K. Aldakheel, M. A. Gondal, M. M. Nasr, M. A. Almessiere and N. Idris, Talanta, 2020, 217,  DOI:10.1016/j.talanta.2020.121062.
  183. S. E. Ghoreyshi, H. Shirvani-Mahdavi and S. Z. Shoursheini, Appl. Phys. B, 2019, 125(7) DOI:10.1007/s00340-00019-07234-00348.
  184. L. C. Nunes, F. R. P. Rocha and F. J. Krug, J. Anal. At. Spectrom., 2019, 34(11), 2314–2324 RSC.
  185. J. Y. Peng, Y. He, J. D. Jiang, Z. F. Zhao, F. Zhou and F. Liu, Food Chem., 2019, 295, 327–333 CrossRef CAS.
  186. J. Y. Peng, Y. He, Z. F. Zhao, J. D. Jiang, F. Zhou, F. Liu and T. T. Shen, Environ. Pollut., 2019, 252, 1125–1132 CrossRef CAS.
  187. X. N. Liu, X. Q. Che, K. Y. Li, X. B. Wang, Z. Z. Lin, Z. S. Wu and Q. S. Zheng, Plasma Sci. Technol., 2020, 22(7) DOI:10.1088/2058-6272/ab1087eda.
  188. Z. Q. Feng, D. C. Zhang, B. W. Wang, J. Ding, X. Y. Liu and J. F. Zhu, Plasma Sci. Technol., 2020, 22(7) DOI:10.1088/2058-6272/ab1084ed.
  189. Y. H. Jiang, J. Kang, Y. R. Wang, Y. Q. Chen and R. H. Li, Appl. Spectrosc., 2019, 73(11) DOI:10.1177/0003702819857740.
  190. P. D. Tribbett, C. A. Jones, A. G. Anders, B. Rougeau and J. A. Merten, J. Anal. At. Spectrom., 2020, 35(1), 78–83 RSC.
  191. R. Terzano, M. A. Denecke, G. Falkenberg, B. Miller, D. Paterson and K. Janssens, Pure Appl. Chem., 2019, 91(6), 1029–1063 CAS.
  192. R. Ravansari, S. C. Wilson and M. Tighe, Environ. Int., 2020, 134 DOI:10.1016/j.envint.2019.105250.
  193. X. M. She, Z. X. Zhu, J. Gao, R. Qian, C. Sheng, R. X. Shen and S. J. Zhuo, X-Ray Spectrom., 2019, 48(6), 664–673 CrossRef CAS.
  194. I. Allegretta, B. Ciasca, M. D. R. Pizzigallo, V. M. T. Lattanzio and R. Terzano, Appl. Clay Sci., 2019, 180,  DOI:10.1016/j.clay.2019.105201.
  195. F. Bilo, L. Borgese, G. Pardini, E. Margu, A. Zacco, R. Dalipi, S. Federici, M. Bettinelli, M. Volante, E. Bontempi and L. E. Depero, J. Anal. At. Spectrom., 2019, 34(5), 930–939 RSC.
  196. S. Mitsunobu, T. Hiruta, J. Fukudo, Y. Narahashi, N. Hamamura, N. Matsue and Y. Takahashi, Geoderma, 2020, 373,  DOI:10.1016/j.geoderma.2020.114415.
  197. S. Nawar, N. Delbecque, Y. Declercq, P. De Smedt, P. Finke, A. Verdoodt, M. Van Meirvenne and A. M. Mouazen, Geoderma, 2019, 350, 29–39 CrossRef CAS.
  198. R. Andrade, W. M. Faria, S. H. G. Silva, S. Chakraborty, D. C. Weindorf, L. F. Mesquita, L. R. G. Guilherme and N. Curi, Geoderma, 2020, 357,  DOI:10.1016/j.geoderma.2019.113960.
  199. R. Andrade, S. H. G. Silva, D. C. Weindorf, S. Chakraborty, W. M. Faria, L. F. Mesquita, L. R. G. Guilherme and N. Curi, Geoderma, 2020, 357,  DOI:10.1016/j.geoderma.2019.113957.
  200. T. R. Tavares, J. P. Molin, L. C. Nunes, E. E. N. Alves, F. L. Melquiades, H. W. P. de Carvalho and A. M. Mouazen, Remote Sens., 2020, 12(6) DOI:10.3390/rs12060963.
  201. T. M. de Lima, D. C. Weindorf, N. Curi, L. R. G. Guilherme, R. M. Q. Lana and B. T. Ribeiro, Geoderma, 2019, 353, 264–272 CrossRef.
  202. A. F. D. Teixeira, M. H. P. Pelegrino, W. M. Faria, S. H. G. Silva, M. G. M. Goncalves, F. W. Acerbi, L. R. Gomide, A. L. Padua, I. A. de Souza, S. Chakraborty, D. C. Weindorf, L. R. G. Guilherme and N. Curi, Geoderma, 2020, 361,  DOI:10.1016/j.geoderma.2019.114132.
  203. M. Mancini, D. C. Weindorf, S. H. G. Silva, S. Chakraborty, A. F. D. Teixeira, L. R. G. Guilherme and N. Curi, Geoderma, 2019, 354,  DOI:10.1016/j.geoderma.2019.113885.
  204. S. H. G. Silva, D. C. Weindorf, L. C. Pinto, W. M. Faria, F. W. Acerbi, L. R. Gomide, J. M. de Mello, A. L. de Padua, I. A. de Souza, A. F. D. Teixeira, L. R. G. Guilherme and N. Curi, Geoderma, 2020, 362,  DOI:10.1016/j.geoderma.2019.114136.
  205. J. An, B. Jeong, S. Jeong and K. Nam, Spectrochim. Acta, Part B, 2020, 164,  DOI:10.1016/j.sab.2019.105752.
  206. X. L. Li, Z. S. Yu, J. L. Xu, Y. S. Pan, W. Bo, B. Liu, P. P. Zhang, J. F. Bai and Q. Zhang, X-Ray Spectrom., 2020 DOI:10.1002/xrs.3147.
  207. J. F. Shao, W. B. Jia, X. L. Zhang, Y. Liu, X. R. Tang, G. C. Xiong and Q. Shan, J. Anal. At. Spectrom., 2020, 35(4), 746–753 RSC.
  208. A. Mijovilovich, A. Mishra, D. Bruckner, K. Spiers, E. Andresen, J. Garrevoet, G. Falkenberg and H. Kupper, Spectrochim. Acta, Part B, 2019, 157, 53–62 CrossRef CAS.
  209. V. Balaram, Geosci. Front., 2019, 10(4), 1285–1303 CrossRef CAS.
  210. E. Tjabadi and N. Mketo, TrAC, Trends Anal. Chem., 2019, 118, 207–222 CrossRef CAS.
  211. E. M. Griffith and M. S. Fantle, Chem. Geol., 2020, 537,  DOI:10.1016/j.chemgeo.2020.119470.
  212. S. T. Wu, G. Worner, K. P. Jochum, B. Stoll, K. Simon and A. Kronz, Geostand. Geoanal. Res., 2019, 43(4), 567–584 CrossRef CAS.
  213. Y. Q. Ke, J. Z. Zhou, X. Q. Yi, Y. J. Sun, J. F. Shao, S. Y. You, W. Wang, Y. Z. Tang and C. Y. Tu, J. Anal. At. Spectrom., 2020, 35(5), 886–895 RSC.
  214. M. Martinez, C. Bayne, D. Aiello, M. Julian, R. Gaume and M. Baudelet, Spectrochim. Acta, Part B, 2019, 159 DOI:10.1016/j.sab.2019.105650.
  215. H. S. Wu, R. Bottger, F. Couffignal, J. Gutzmer, J. Krause, F. Munnik, A. D. Renno, R. Hubner, M. Wiedenbeck and R. Ziegenrucker, Geostand. Geoanal. Res., 2019, 43(4), 531–541 CrossRef CAS.
  216. M. Weber, F. Lugli, B. Hattendorf, D. Scholz, R. Mertz-Kraus, D. Guinoiseau and K. P. Jochum, Geostand. Geoanal. Res., 2020, 44(1), 69–83 CrossRef CAS.
  217. K. P. Jochum, D. Garbe-Schonberg, M. Veter, B. Stoll, U. Weis, M. Weber, F. Lugli, A. Jentzen, R. Schiebel, J. A. Wassenburg, D. E. Jacob and G. H. Haug, Geostand. Geoanal. Res., 2019, 43(4), 595–609 CrossRef.
  218. V. G. Batanova, J. M. Thompson, L. V. Danyushevsky, M. V. Portnyagin, D. Garbe-Schonberg, E. Hauri, J. I. Kimura, Q. Chang, R. Senda, K. Goemann, C. Chauvel, S. Campillo, D. A. Ionov and A. V. Sobolev, Geostand. Geoanal. Res., 2019, 43(3), 453–473 CrossRef CAS.
  219. Y. Bussweiler, A. Giuliani, A. Greig, B. A. Kjarsgaard, D. Petts, S. E. Jackson, N. Barrett, Y. Luo and D. G. Pearson, Chem. Geol., 2019, 524, 136–157 CrossRef CAS.
  220. Y. Li, Q. L. Li, G. Q. Tang, A. M. Gargano, Z. Sharp, A. Pitawala, L. Zhao, M. G. Zhai and X. H. Li, At. Spectrosc., 2020, 41(2), 51–56 CrossRef.
  221. C. Dorais, A. Simonetti, L. Corcoran, T. L. Spano and P. C. Burns, Geostand. Geoanal. Res., 2020, 44(1), 125–132 CrossRef CAS.
  222. Q. Ma, N. J. Evans, X. X. Ling, J. H. Yang, F. Y. Wu, Z. D. Zhao and Y. H. Yang, Geostand. Geoanal. Res., 2019, 43(3), 355–384 CrossRef CAS.
  223. L. Zhang, J. L. Wu, J. R. Tu, D. Wu, N. Li, X. P. Xia and Z. Y. Ren, Geostand. Geoanal. Res., 2020, 44(1), 133–145 CrossRef CAS.
  224. C. Huang, H. Wang, J. H. Yang, J. Ramezani, C. Yang, S. B. Zhang, Y. H. Yang, X. P. Xia, L. J. Feng, J. Lin, T. T. Wang, Q. Ma, H. Y. He, L. W. Xie and S. T. Wu, Geostand. Geoanal. Res., 2020, 44(1), 103–123 CrossRef CAS.
  225. S. E. M. Gain, Y. Greau, H. Henry, E. Belousova, I. Dainis, W. L. Griffin and S. Y. O’Reilly, Geostand. Geoanal. Res., 2019, 43(3), 339–354 CrossRef CAS.
  226. G. Q. Tang, B. X. Su, Q. L. Li, X. P. Xia, J. J. Jing, L. J. Feng, L. Martin, Q. Yang and X. H. Li, Geostand. Geoanal. Res., 2019, 43(4), 585–593 CrossRef CAS.
  227. G. Q. Tang, X. H. Li, Q. L. Li, Y. Liu and X. X. Ling, Surf. Interface Anal., 2020, 52(5), 190–196 CrossRef CAS.
  228. J. N. Avila, T. R. Ireland, P. Holden, P. Lanc, A. Latimore, N. Schram, J. Foster, I. S. Williams, L. Loiselle and B. Fu, Geostand. Geoanal. Res., 2020, 44(1), 85–102 CrossRef CAS.
  229. Q. Yang, X. P. Xia, L. Zhang, W. F. Zhang, Y. Q. Zhang, L. L. Chen, Y. N. Yang and M. H. He, Surf. Interface Anal., 2020, 52(5), 197–213 CrossRef CAS.
  230. M. L. Tang, B. L. Fan, L. Y. Yao, Z. Y. Bao, H. Tian, Z. Z. Ma, Z. Y. An, M. Ma, H. Y. Zhang and Z. B. Fu, Geostand. Geoanal. Res., 2020, 44(2), 375–384 CrossRef CAS.
  231. J. Vogl, M. Rosner, S. A. Kasemann, R. Kraft, A. Meixner, J. Noordmann, S. Rabb, O. Rienitz, J. A. Schuessler, M. Tatzel and R. D. Vocke, Geostand. Geoanal. Res., 2020, 44(3), 439–457 CrossRef CAS.
  232. L. P. Feng, L. Zhou, J. H. Liu, Z. C. Hu and Y. S. Liu, Geostand. Geoanal. Res., 2019, 43(4), 701–714 CrossRef CAS.
  233. M. L. Kuessner, A. Gourgiotis, G. Manhes, J. Bouchez, X. Zhang and J. Gaillardet, Geostand. Geoanal. Res., 2020, 44(1), 57–67 CrossRef CAS.
  234. K. Sullivan, D. Layton-Matthews, M. Leybourne, J. Kidder, Z. Mester and L. Yang, Geostand. Geoanal. Res., 2020, 44(2), 349–362 CrossRef CAS.
  235. N. Gaspers, T. Magna and L. Ackerman, Geostand. Geoanal. Res., 2020, 44(2), 363–374 CrossRef CAS.
  236. Y. H. Yang, M. Yang, K. P. Jochum, S. T. Wu, H. Zhao, L. W. Xie, C. Huang, X. C. Zhan, J. H. Yang and F. Y. Wu, Geostand. Geoanal. Res., 2020, 44(3), 567–579 CrossRef CAS.
  237. M. Yang, Y. H. Yang, N. J. Evans, L. W. Xie, C. Huang, S. T. Wu, J. H. Yang and F. Y. Wu, Geostand. Geoanal. Res., 2020, 44(3), 553–565 CrossRef CAS.
  238. W. H. Li, J. M. Zhu, D. C. Tan, G. L. Han, Z. Q. Zhao and G. L. Wu, Geostand. Geoanal. Res., 2020, 44(3), 523–535 CrossRef CAS.
  239. G. L. Wu, J. M. Zhu, X. L. Wang, T. M. Johnson and G. L. Han, Anal. Chem., 2020, 92(1), 1463–1469 CrossRef CAS.
  240. L. Geng, J. Savarino, N. Caillon, E. Gautier, J. Farquhar, J. W. Dottin, N. Magalhaes, S. Hattori, S. Ishino, N. Yoshida, F. Albarede, E. Albalat, P. Cartigny, S. Ono and M. H. Thiemens, J. Anal. At. Spectrom., 2019, 34(6), 1263–1271 RSC.
  241. T. C. Meisel, Geostand. Geoanal. Res., 2019, 43(4), 527–528 CrossRef.
  242. P. J. Potts, P. C. Webb and M. Thompson, Geostand. Geoanal. Res., 2019, 43(3), 409–418 CrossRef.
  243. P. C. Webb, P. J. Potts, M. Thompson, S. A. Wilson and C. J. B. Gowing, Geostand. Geoanal. Res., 2019, 43(3), 397–408 CrossRef.
  244. E. T. Mansur, S. J. Barnes, D. Savard and P. C. Webb, Geostand. Geoanal. Res., 2020, 44(1), 147–167 CrossRef CAS.
  245. W. Zhang and Z. C. Hu, Spectrochim. Acta, Part B, 2019, 160 DOI:10.1016/j.sab.2019.105690.
  246. T. He, Z. C. Hu, W. Zhang, H. H. Chen, Y. S. Liu, Z. C. Wang and S. H. Hu, Anal. Chem., 2019, 91(13), 8109–8114 CrossRef CAS.
  247. A. L. B. de Oliveira, J. C. Afonso, L. da Silva, A. A. Neto, M. C. Carneiro, L. I. D. da Silva and M. I. C. Monteiro, Geostand. Geoanal. Res., 2019, 43(4), 689–699 CrossRef.
  248. F. S. Rondan, A. S. Henn, P. A. Mello, M. Perez, L. A. Bullock, J. Parnell, J. Feldmann, E. M. M. Flores and M. F. Mesko, J. Anal. At. Spectrom., 2019, 34(5), 998–1004 RSC.
  249. Z. Q. Zou, Z. C. Wang, H. Cheng, T. He, Y. N. Liu, K. Chen, Z. C. Hu and Y. S. Liu, Geostand. Geoanal. Res., 2020, 44(3), 501–521 CrossRef CAS.
  250. D. Araoka and T. Yoshimura, Anal. Sci., 2019, 35(7), 751–757 CrossRef CAS.
  251. H. Cheng, Z. C. Wang, K. Chen, K. Q. Zong, Z. Q. Zou, T. He, Z. C. Hu, M. Fischer-Godde and Y. H. Liu, Geostand. Geoanal. Res., 2019, 43(4), 663–680 CrossRef CAS.
  252. L. Daniel, B. R. Clarke and D. F. Parsons, Geostand. Geoanal. Res., 2019, 43(3), 435–452 CrossRef CAS.
  253. D. F. Andrade, E. R. Pereira and D. Amarasiriwardena, Appl. Spectrosc. Rev., 2020 DOI:10.1080/05704928.05702020.01739063.
  254. D. Syvilay, J. Guezenoc and B. Bousquet, Spectrochim. Acta, Part B, 2019, 161,  DOI:10.1016/j.sab.2019.105696.
  255. C. Fabre, Spectrochim. Acta, Part B, 2020, 166,  DOI:10.1016/j.sab.2020.105799.
  256. S. Sheta, M. S. Afgan, Z. Y. Hou, S. C. Yao, L. Zhang, Z. Li and Z. Wang, J. Anal. At. Spectrom., 2019, 34(6), 1047–1082 RSC.
  257. W. T. Li, X. Y. Li, X. Li, Z. Q. Hao, Y. F. Lu and X. Y. Zeng, Appl. Spectrosc. Rev., 2020, 55(1), 1–25 CrossRef CAS.
  258. Y. Foucaud, C. Fabre, B. Demeusy, I. V. Filippova and L. O. Filippov, Spectrochim. Acta, Part B, 2019, 158 DOI:10.1016/j.sab.2019.1005.1017.
  259. A. Pochon, A. M. Desaulty and L. Bailly, J. Anal. At. Spectrom., 2020, 35(2), 254–264 RSC.
  260. J. Alvarez, M. Velasquez, A. K. Myakalwar, C. Sandoval, R. Fuentes, R. Castillo, D. Sbarbaro and J. Yanez, J. Anal. At. Spectrom., 2019, 34(12), 2459–2468 RSC.
  261. Q. Z. Li, W. Zhang, Z. Y. Tang, R. Zhou, J. L. Yan, C. W. Zhu, K. Liu, X. Y. Li and X. Y. Zeng, J. Anal. At. Spectrom., 2020, 35(3), 626–631 RSC.
  262. A. A. C. Carvalho, L. A. Cozer, M. S. Luz, L. C. Nunes, F. R. P. Roche and C. S. Nomura, J. Anal. At. Spectrom., 2019, 34(8), 1701–1707 RSC.
  263. Y. S. Zhang, M. R. Dong, L. H. Cheng, L. P. Wei, J. B. Cai and J. D. Lu, J. Anal. At. Spectrom., 2020, 35(4), 810–818 RSC.
  264. X. J. Xu, A. Z. Li, X. S. Wang, C. J. Ding, S. L. Qiu, Y. G. He, T. Q. Lu, F. He, B. S. Zou and R. B. Liu, J. Anal. At. Spectrom., 2020, 35(5), 984–992 RSC.
  265. A. H. Galmed, A. du Plessis, S. G. le Roux, E. Hartnick, H. Von Bergmann and M. Maaza, Spectrochim. Acta, Part B, 2018, 139, 75–82 CrossRef CAS.
  266. L. Xin, Q. L. Li, M. J. Whitehouse, Y. H. Yang and Y. Liu, J. Anal. At. Spectrom., 2020, 35(3), 489–497 RSC.
  267. T. Luo, X. D. Deng, J. W. Li, Z. C. Hu, W. Zhang, Y. S. Liu and J. F. Zhang, J. Anal. At. Spectrom., 2019, 34(7), 1439–1446 RSC.
  268. Y. H. Yang, F. Y. Wu, Q. L. Li, Y. Rojas-Agramonte, J. H. Yang, Y. Li, Q. Ma, L. W. Xie, C. Huang, H. R. Fan, Z. F. Zhao and C. Xu, Geostand. Geoanal. Res., 2019, 43(4), 543–565 CrossRef CAS.
  269. N. Cogné, D. M. Chew, R. A. Donelick and C. Ansberque, Chem. Geol., 2020, 531,  DOI:10.1016/j.chemgeo.2019.119302.
  270. E. P. Corbett, A. Simonetti, P. Shaw, L. Corcoran, Q. G. Crowley and B. C. Hoare, Chem. Geol., 2020, 544,  DOI:10.1016/j.chemgeo.2020.119568.
  271. S. H. Liu, X. Wang, H. X. Shi, H. Zhang and B. B. Liu, J. Anal. Methods Chem., 2020, 2020 DOI:10.1155/2020/1819639.
  272. A. J. Anderson, J. M. Hanchar, K. V. Hodges and M. C. van Soest, Chem. Geol., 2020, 538,  DOI:10.1016/j.chemgeo.2020.119494.
  273. F. L. H. Tissot, M. Ibanez-Mejia, P. Boehnke, N. Dauphas, D. McGee, T. L. Grove and T. M. Harrison, J. Anal. At. Spectrom., 2019, 34(10), 2035–2052 RSC.
  274. J. R. Tu, Z. B. Xiao, H. Y. Zhou, S. Q. An, G. Z. Li, Y. R. Cui, W. G. Liu and H. M. Li, Ore Geol. Rev., 2019, 109, 407–412 CrossRef.
  275. C. Perrin, R. Tilhac and L. Prestimonaco, Sediment. Geol., 2019, 389, 91–102 CrossRef CAS.
  276. J. Engel, R. Maas, J. Woodhead, J. Tympel and A. Greig, Chem. Geol., 2020, 531,  DOI:10.1016/j.chemgeo.2019.119311.
  277. K. T. M. Ito, Y. Hibiya, Y. Homma, T. Mikouchi and T. Iizuka, Chem. Geol., 2019, 525, 343–355 CrossRef CAS.
  278. J. Kim and I. H. Cho, J. Anal. Sci. Technol., 2020, 11(1) DOI:10.1186/s40543-40020-00207-40549.
  279. M. L. Machlus, E. K. Shea, S. R. Hemming, J. Ramezani and E. T. Rasbury, Chem. Geol., 2020, 539,  DOI:10.1016/j.chemgeo.2020.119485.
  280. P. Louvat, M. Tharaud, M. Buisson, C. Rollion-Bard and M. F. Benedetti, J. Anal. At. Spectrom., 2019, 34(8), 1553–1563 RSC.
  281. K. P. Jochum, A. Jentzen, R. Schiebel, B. Stoll, U. Weis, J. Leitner, J. Repschlager, D. Nurnberg and G. H. Haug, Geochem., Geophys., Geosyst., 2019, 20(4), 2053–2063 CrossRef CAS.
  282. X. H. Liao, Z. C. Hu, T. Luo, W. Zhang, Y. S. Liu, K. Q. Zong, L. Zhou and J. F. Zhang, J. Anal. At. Spectrom., 2019, 34(6), 1126–1134 RSC.
  283. E. S. Steenstra, J. Berndt, S. Klemme, W. van Westrenen, E. S. Bullock and A. Shahar, J. Anal. At. Spectrom., 2020, 35(3), 498–509 RSC.
  284. A. J. Managh and P. Reid, J. Anal. At. Spectrom., 2019, 34(7), 1369–1373 RSC.
  285. J. T. van Elteren, V. S. Selih and M. Sala, J. Anal. At. Spectrom., 2019, 34(9), 1919–1931 RSC.
  286. X. Lin, W. Guo, L. L. Jin and S. H. Hu, At. Spectrosc., 2020, 41(1), 1–10 CrossRef.
  287. G. Tuba, D. J. Kontak, Z. Zajacz and J. A. Petrus, Chem. Geol., 2019, 529 DOI:10.1016/j.chemgeo.2019.119326.
  288. C. J. Spencer, C. L. Kirkland, N. M. W. Roberts, N. J. Evans and J. Liebmann, Geosci. Front., 2020, 11(3), 843–853 CrossRef CAS.
  289. H. O. Gu, H. Sun, F. Y. Wang, C. Ge and T. F. Zhou, J. Anal. At. Spectrom., 2019, 34(6), 1223–1232 RSC.
  290. W. Zhang, Z. C. Wang, F. Moynier, E. Inglis, S. Y. Tian, M. Li, Y. S. Liu and Z. Hu, J. Anal. At. Spectrom., 2019, 34(9), 1800–1809 RSC.
  291. F. Lugli, M. Weber, T. Giovanardi, S. Arrighi, E. Bortolini, C. Figus, G. Marciani, G. Oxilia, M. Romandini, S. Silvestrini, K. P. Jochum, S. Benazzi and A. Cipriani, J. Anal. At. Spectrom., 2020, 35(5), 852–862 RSC.
  292. W. Zhang, Z. C. Hu, Y. S. Liu, L. P. Feng and H. S. Jiang, Chem. Geol., 2019, 522, 16–25 CrossRef CAS.
  293. C. G. de Vega, M. Costas-Rodriguez, T. Van Acker, S. Goderis and F. Vanhaecke, Anal. Chem., 2020, 92(5), 3572–3580 CrossRef.
  294. J. Lin, Y. S. Liu, Z. C. Hu, W. Chen, C. X. Zhang, K. D. Zhao and X. Y. Jin, J. Anal. At. Spectrom., 2019, 34(6), 1145–1153 RSC.
  295. L. K. Steinmann, M. Oeser, I. Horn, H. M. Seitz and S. Weyer, J. Anal. At. Spectrom., 2019, 34(7), 1447–1458 RSC.
  296. L. Y. Zhu, Y. S. Liu, S. Y. Jiang and J. Lin, J. Anal. At. Spectrom., 2019, 34(8), 1546–1552 RSC.
  297. M. L. Pons, M. A. Millet, G. N. N. Nowell, S. Misra and H. M. M. Williams, J. Anal. At. Spectrom., 2020, 35(2), 320–330 RSC.
  298. T. Long, S. W. J. Clement, H. Q. Xie and D. Y. Liu, Int. J. Mass Spectrom., 2020, 450,  DOI:10.1016/j.ijms.2019.116289.
  299. E. E. Groopman, D. G. Willingham, A. J. Fahey and K. S. Grabowski, J. Anal. At. Spectrom., 2020, 35(3), 600–625 RSC.
  300. Y. Liu, X. H. Li, G. Q. Tang, Q. L. Li, X. C. Liu, H. M. Yu and F. Huang, J. Anal. At. Spectrom., 2019, 34(5), 906–914 RSC.
  301. J. Villeneuve, M. Chaussidon, Y. Marrocchi, Z. B. Deng and E. B. Watson, Rapid Commun. Mass Spectrom., 2019, 33(20), 1589–1597 CrossRef CAS.
  302. K. Fukuda, B. L. Beard, D. R. Dunlap, M. J. Spicuzza, J. H. Fournelle, M. Wadhwa and N. T. Kita, Chem. Geol., 2020, 540,  DOI:10.1016/j.chemgeo.2020.119482.
  303. A. Vho, D. Rubatto, B. Putlitz and A. S. Bouvier, Geostand. Geoanal. Res., 2020, 44(3), 459–471 CrossRef CAS.
  304. M. H. He, X. P. Xia, X. L. Huang, J. L. Ma, J. Q. Zou, Q. Yang, F. Yang, Y. Q. Zhang, Y. N. Yang and G. J. Wei, J. Anal. At. Spectrom., 2020, 35(2), 238–245 RSC.
  305. X. P. Xia, Z. X. Cui, W. C. Li, W. F. Zhang, Q. Yang, H. J. Hui and C. K. Lai, J. Anal. At. Spectrom., 2019, 34(6), 1088–1097 RSC.
  306. C. Luisier, L. Baumgartner, G. Siron, T. Vennemann and M. Robyr, Geostand. Geoanal. Res., 2019, 43(4), 635–646 CrossRef CAS.
  307. S. M. Reddy, D. W. Saxey, W. D. A. Rickard, D. Fougerouse, S. D. Montalvo, R. Verberne and A. van Riessen, Geostand. Geoanal. Res., 2020, 44(1), 5–50 CrossRef.
  308. F. Liu, Z. F. Zhang, X. Li and Y. J. An, Int. J. Mass Spectrom., 2020, 450,  DOI:10.1016/j.ijms.2020.116307.
  309. J. Trubac, T. Magna, B. Cejkova, L. Vondrovicova and V. Rapprich, Rapid Commun. Mass Spectrom., 2019, 33(16), 1355–1362 CrossRef CAS.
  310. T. A. Velivetskaya, A. V. Ignatiev, V. V. Yakovenko and S. V. Vysotskiy, Rapid Commun. Mass Spectrom., 2019, 33(22), 1722–1729 CrossRef CAS.
  311. K. A. Farley, J. Treffkorn and D. Hamilton, Chem. Geol., 2020, 537,  DOI:10.1016/j.chemgeo.2020.119487.
  312. R. Wiesendanger, V. Grimaudo, M. Tulej, A. Riedo, R. Lukmanov, N. Ligterink, R. Fausch, H. Shea and P. Wurz, J. Anal. At. Spectrom., 2019, 34(10), 2061–2073 RSC.
  313. D. Arenas-Islas, M. A. Huerta-Diaz, C. O. Norzagaray-Lopez, K. G. Mejia-Pina, J. A. Valdivieso-Ojeda, X. L. Otero and F. Arcega-Cabrera, Sediment. Geol., 2019, 391,  DOI:10.1016/j.sedgeo.2019.105517.
  314. M. Al-Musawi and S. Kaczmarek, Appl. Geochem., 2020, 113 DOI:10.1016/j.apgeochem.2019.104491.
  315. I. W. Croudace, L. Lowemark, R. Tjallingii and B. Zolitschka, Quaternary Int., 2019, 514, 5–15 CrossRef.
  316. B. R. B. Gregory, R. T. Patterson, E. G. Reinhardt, J. M. Galloway and H. M. Roe, Chem. Geol., 2019, 521, 12–27 CrossRef CAS.
  317. B. P. von der Heyden, Ore Geol. Rev., 2020, 117 DOI:10.1016/j.oregeorev.2020.103328.
  318. I. del Real, L. Smieska, J. F. H. Thompson, C. Martinez, J. Thomas and D. Layton-Matthews, J. Anal. At. Spectrom., 2019, 34(8), 1724–1738 RSC.
  319. S. L. Didukh-Shadrina, V. N. Losev, A. Samoilo, A. K. Trofimchuk and P. N. Nesterenko, Int. J. Anal. Chem., 2019 DOI:10.1155/2019/1467631.
  320. L. Z. Suo, X. Y. Dong, X. Gao, J. F. Xu, Z. Huang, J. Ye, X. M. Lu and L. S. Zhao, Microchem. J., 2019, 149 DOI:10.1016/j.microc.2019.104039.
  321. X. J. Liu, Q. Y. Lin, Y. H. Tian, W. L. Liao, T. Yang, C. Qian, T. L. Zhang and Y. X. Duan, J. Anal. At. Spectrom., 2020, 35(1), 188–197 RSC.
  322. H. Ahmad, U. Haseen, K. Umar, M. S. Ansari and M. N. M. Ibrahim, Microchim. Acta, 2019, 186(9) DOI:10.1007/s00604-00019-03753-00606.
  323. M. R. Letsoalo, T. W. Godeto, T. Magadzu and A. A. Ambushe, J. Environ. Sci. Health, Part A: Toxic/Hazard. Subst. Environ. Eng., 2019, 54(9), 924–932 CrossRef CAS.
  324. F. Tokay and S. Bagdat, Int. J. Environ. Anal. Chem., 2019, 99(15), 1528–1539 CrossRef CAS.
  325. D. G. V. Nova, N. F. Robaina, K. D. do Amaral and R. J. Cassella, Microchem. J., 2020, 152 DOI:10.1016/j.microc.2019.104310.
  326. J. Sulejmanovic, M. Memic, E. Sabanovic and N. Ljubijankic, Anal. Methods, 2019, 11(48), 6184–6192 RSC.
  327. H. Tavallali, H. Malekzadeh, M. A. Karimi, M. Payehghadr, G. Deilamy-Rad and M. Tabandeh, Arabian J. Chem., 2019, 12(7), 1487–1495 CrossRef CAS.
  328. L. Chen, Z. Z. Wang, J. X. Pei and X. J. Huang, Anal. Chem., 2020, 92(2), 2251–2257 CrossRef CAS.
  329. K. Pytlakowska, M. Pilch, B. Hachula, J. E. Nycz, K. Kornaus and W. A. Pisarski, J. Anal. At. Spectrom., 2019, 34(7), 1416–1425 RSC.
  330. L. Y. Zhao, J. J. Fei, H. Z. Lian, L. Mao and X. B. Cui, J. Anal. At. Spectrom., 2019, 34(8), 1693–1700 RSC.
  331. R. J. González-Álvarez, J. J. Pinto, J. A. López-López, C. Mendiguchíaa and C. Moreno, Anal. Bioanal. Chem., 2020, 412(8), 1863–1870 CrossRef.
  332. Y. Tanaka, M. Tsujisaka, L. J. Zheng, S. Takano and Y. Sohrin, Anal. Sci., 2019, 35(9), 1015–1020 CrossRef CAS.
  333. S. Kulomaki, E. Lahtinen, S. Peramaki and A. Vaisanen, Anal. Chim. Acta, 2019, 1092, 24–31 CrossRef CAS.
  334. M. Schlathauer, J. Friedland, M. Linden and K. Leopold, Microchem. J., 2019, 147, 253–262 CrossRef CAS.
  335. F. Abujaber, M. Jimenez-Moreno, F. J. G. Bernardo and R. C. R. Martin-Doimeadios, Microchim. Acta, 2019, 186(7) DOI:10.1007/s00604-00019-03492-00608.
  336. S. W. Yang, D. Y. Zhang, H. Y. Cheng, Y. C. Wang and J. H. Liu, Anal. Chim. Acta, 2019, 1074, 54–61 CrossRef CAS.
  337. H. Perez-Tribouillier, T. L. Noble, A. T. Townsend, A. R. Bowie and Z. Chase, Talanta, 2019, 202, 600–609 CrossRef CAS.
  338. B. Lesniewska, Z. Arciszewska, A. Wawrzynczak, S. Jarmolinska, I. Nowak and B. Godlewska-Zylkiewicz, Talanta, 2020, 217,  DOI:10.1016/j.talanta.2020.121004.
  339. E. Abdulbur-Alfakhoury, S. Van Zutphen and M. Leermakers, Talanta, 2019, 203, 34–48 CrossRef CAS.
  340. M. He, S. W. Su, B. Chen and B. Hu, Talanta, 2020, 207,  DOI:10.1016/j.talanta.2019.120314.
  341. A. Rasool, T. F. Xiao, S. Ali, W. Ali and W. Nasim, Environ. Sci. Pollut. Res., 2020, 27(9), 9686–9696 CrossRef CAS.
  342. H. Takahashi, Y. Izumoto, T. Matsuyama and H. Yoshii, X-Ray Spectrom., 2019, 48(5), 366–374 CrossRef CAS.
  343. A. A. Lashari, T. G. Kazi, J. A. Baig and H. I. Afridi, Int. J. Environ. Anal. Chem., 2019 DOI:10.1080/03067319.03062019.01691185.
  344. L. F. D. Guedes, B. F. Braz, A. S. Freire and R. E. Santelli, Microchem. J., 2020, 155 DOI:10.1016/j.microc.2020.104714.
  345. R. Shaikh, T. G. Kazi, H. I. Afridi, A. Akhtar and J. A. Baig, Chemosphere, 2019, 237 DOI:10.1016/j.chemosphere.2019.124444.
  346. S. M. Sorouraddin, M. A. Farajzadeh, H. Dastoori and T. Okhravi, Int. J. Environ. Anal. Chem., 2019 DOI:10.1080/03067319.03062019.01686144.
  347. M. A. Habila, E. A. Alabdullkarem, Z. A. Alothman, E. Yilmaz and M. Soylak, Int. J. Environ. Anal. Chem., 2020 DOI:10.1080/03067319.03062020.01743831.
  348. T. Unutkan, T. Borahan, A. Girgin and S. Bakirdere, Environ. Monit. Assess., 2020, 192(2) DOI:10.1007/s10661-10020-18104-x.
  349. R. El Sheikh, W. S. Hassan, A. M. Youssef, A. M. Hameed, A. Subaihi, A. Alharbi and A. A. Gouda, Int. J. Environ. Anal. Chem., 2020 DOI:10.1080/03067319.03062020.01727461.
  350. E. J. D. Souza, C. D. B. do Amaral, N. Nagata and M. T. Grassi, Microchem. J., 2020, 152 DOI:10.1016/j.microc.2019.104309.
  351. J. Lopez-Mayan, M. C. Barciela-Alonso, M. R. Dominguez-Gonzalez, E. Pena-Vazquez and P. Bermejo-Barrera, Microchem. J., 2020, 152 DOI:10.1016/j.microc.2019.104264.
  352. M. Ataee, T. Ahmadi-Jouibari, N. Noori and N. Fattahi, RSC Adv., 2020, 10(3), 1514–1521 RSC.
  353. A. P. R. Santana, D. F. Andrade, T. G. S. Guimaraes, C. D. B. Amaral, A. Oliveira and M. H. Gonzalez, Talanta, 2020, 216,  DOI:10.1016/j.talanta.2020.120956.
  354. F. Salimi and M. Ramezani, Sep. Sci. Technol., 2019, 54(14), 2274–2282 CrossRef CAS.
  355. Z. Shahbazi-Manshadi, M. Kaykhaii, S. Dadfarnia and A. M. H. Shabani, Int. J. Environ. Anal. Chem., 2019, 99(13), 1365–1374 CrossRef CAS.
  356. S. Z. Mohammadi, T. Shamspur and Y. M. Baghelani, Arabian J. Chem., 2019, 12(7), 1751–1757 CrossRef CAS.
  357. F. D. Dias, S. Neto, L. D. Pires and V. A. Lemos, Anal. Methods, 2020, 12(6), 865–871 RSC.
  358. T. G. Kazi, N. S. Memon, S. A. Shaikh and S. S. Memon, Food Anal. Meth., 2019, 12(9), 1964–1972 CrossRef.
  359. I. Machado and F. Tissot, Talanta, 2020, 212,  DOI:10.1016/j.talanta.2020.120783.
  360. J. Ali, M. Tuzen and T. G. Kazi, Food Chem., 2020, 306,  DOI:10.1016/j.foodchem.2019.125638.
  361. M. Llaver, F. A. Casado-Carmona, R. Lucena, S. Cardenas and R. G. Wuilloud, Spectrochim. Acta, Part B, 2019, 162,  DOI:10.1016/j.sab.2019.105705.
  362. Z. Tekin, T. Borahan, O. Ozdogan and S. Bakirdere, Anal. Methods, 2020, 12(9), 1244–1249 RSC.
  363. A. AlKinani, M. Eftekhari and M. Gheibi, Int. J. Environ. Anal. Chem., 2019 DOI:10.1080/03067319.03062019.01659254.
  364. G. S. Kanberoglu, E. Yilmaz and M. Soylak, J. Iran. Chem. Soc., 2020, 17, 1627–1634 CrossRef.
  365. N. A. Kasa, S. Sel, B. C. Ozkan and S. Bakirdere, Environ. Monit. Assess., 2019, 191(11) DOI:10.1007/s10661-10019-17826-10660.
  366. F. Aydin, E. Yilmaz, E. Olmez and M. Soylak, Talanta, 2020, 207,  DOI:10.1016/j.talanta.2019.120295.
  367. G. Brugmann, M. Brauns and R. Maas, Chem. Geol., 2019, 516, 59–67 CrossRef.
  368. L. L. Tian, Z. Zeng, X. Y. Nan, H. M. Yu and F. Huang, J. Anal. At. Spectrom., 2019, 34(7), 1459–1467 RSC.
  369. F. Liu, X. Li, Y. J. An, J. Li and Z. F. Zhang, Geochem., Geophys., Geosyst., 2019, 43(3), 509–517 CAS.
  370. Z. Bao, C. L. Zong, K. Y. Chen, N. Lv and H. L. Yuan, Int. J. Mass Spectrom., 2020, 448 DOI:10.1016/j.ijms.2019.116268.
  371. D. C. Tan, J. M. Zhu, X. L. Wang, G. L. Han, Z. Lu and W. P. Xu, J. Anal. At. Spectrom., 2020, 35(4), 713–727 RSC.
  372. X. P. Shao, W. T. Bu, Y. C. Fan, K. M. Long, H. M. Yang, L. Tang, C. M. Cheng, X. M. Liu and F. H. Hao, J. Anal. At. Spectrom., 2020, 35(3), 467–477 RSC.
  373. C. Y. Liu, L. J. Xu, C. T. Liu, J. Liu, L. P. Qin, Z. D. Zhang, S. A. Liu and S. G. Li, Geochem., Geophys., Geosyst., 2019, 43(4), 647–661 CAS.
  374. H. Freymuth, C. D. R. Reekie and H. M. Williams, Geostand. Geoanal. Res., 2020, 44(3), 407–420 CrossRef CAS.
  375. T. Yokoyama, R. Fukai and M. Nakahara, Geostand. Geoanal. Res., 2020, 44(2), 265–285 CrossRef CAS.
  376. S. G. Lee and T. Tanaka, Spectrochim. Acta, Part B, 2019, 156, 42–50 CrossRef CAS.
  377. Q. Ma, M. Yang, H. Zhao, N. J. Evans, Z. Y. Chu, L. W. Xie, C. Huang, Z. D. Zhao and Y. H. Yang, J. Anal. At. Spectrom., 2019, 34(6), 1256–1262 RSC.
  378. M. Y. He, C. G. Luo, H. Lu, Z. D. Jin and L. Deng, J. Anal. At. Spectrom., 2019, 34(9), 1773–1778 RSC.
  379. Z. Y. Zhu, T. Yang and X. K. Zhu, J. Anal. At. Spectrom., 2019, 34(7), 1503–1513 RSC.
  380. L. Xu, C. G. Luo and H. J. Wen, Geostand. Geoanal. Res., 2020, 44(1), 201–214 CrossRef CAS.
  381. X. M. Liu and W. S. Li, J. Anal. At. Spectrom., 2019, 34(8), 1708–1717 RSC.
  382. Z. A. Bao, C. L. Zong, K. J. Huang, K. Y. Chen, N. Lv and H. L. Yuan, J. Anal. At. Spectrom., 2019, 34(12), 2469–2475 RSC.
  383. Z. A. Bao, K. J. Huang, T. Z. Huang, B. Shen, C. L. Zong, K. Y. Chen and H. L. Yuan, J. Anal. At. Spectrom., 2019, 34(5), 940–953 RSC.
  384. L. P. Feng, L. Zhou, W. F. Hu, W. Zhang, B. C. Li, Y. S. Liu, Z. C. Hu and L. Yang, J. Anal. At. Spectrom., 2020, 35(1), 145–154 RSC.
  385. M. Tsujisaka, S. Takano, M. Murayama and Y. Sohrin, Anal. Chim. Acta, 2019, 1091, 146–159 CrossRef CAS.
  386. C. Pin and A. Gannoun, J. Anal. At. Spectrom., 2019, 34(10), 2136–2146 RSC.
  387. X. P. Shao, W. T. Bu, K. M. Long, L. Tang, X. M. Liu, H. Yan and F. H. Hao, Spectrochim. Acta, Part B, 2019, 159 DOI:10.1016/j.sab.2019.105656.
  388. A. J. McCoy-West, M. A. Millet, G. M. Nowell, O. Nebel and K. W. Burton, J. Anal. At. Spectrom., 2020, 35(2), 388–402 RSC.
  389. Z. Y. Chu, M. J. Wang, C. F. Li, Y. H. Yang, J. J. Xu, W. Wang and J. H. Guo, J. Anal. At. Spectrom., 2019, 34(10), 2053–2060 RSC.
  390. G. L. Wu, J. M. Zhu, X. L. Wang, G. L. Han, D. C. Tan and S. J. Wang, J. Anal. At. Spectrom., 2019, 34(8), 1639–1651 RSC.
  391. G. Q. Wang, H. Vollstaedt, J. F. Xu and W. G. Liu, Geochem., Geophys., Geosyst., 2019, 43(3), 419–433 CAS.
  392. M. Dellinger, R. G. Hilton and G. M. Nowell, J. Anal. At. Spectrom., 2020, 35(2), 377–387 RSC.
  393. J. Li and L. Yin, Geostand. Geoanal. Res., 2019, 43(3), 497–507 CrossRef CAS.
  394. M. Friebel, E. R. Toth, M. A. Fehr and M. Schonbachler, J. Anal. At. Spectrom., 2020, 35(2), 273–292 RSC.
  395. C. F. Li, Z. Y. Chu, X. C. Wang, J. H. Guo and S. A. Wilde, Anal. Chem., 2019, 91(11), 7288–7294 CrossRef CAS.
  396. W. G. Liu, S. Wei, J. Zhang, C. Ao, F. T. Liu, B. Cai, H. Y. Zhou, J. L. Yang and C. F. Li, J. Anal. At. Spectrom., 2020, 35(5), 953–960 RSC.
  397. X. Y. He, J. L. Ma, G. J. Wei, L. Zhang, Z. B. Wang and Q. S. Wang, J. Anal. At. Spectrom., 2020, 35(1), 100–106 RSC.
  398. P. A. Carr, S. Zink, V. C. Bennett, M. D. Norman, Y. Amelin and P. L. Blevin, Chem. Geol., 2020, 539,  DOI:10.1016/j.chemgeo.2020.119539.
  399. R. X. Zhang, J. Meija, Y. Huang, X. J. Pei, Z. Mester and L. Yang, Anal. Chim. Acta, 2019, 1089, 19–24 CrossRef CAS.
  400. G. H. Zhu, J. L. Ma, G. J. Wei and Y. J. An, J. Anal. At. Spectrom., 2020, 35(5), 873–877 RSC.
  401. L. P. Feng, W. F. Hu, Y. Jiao, L. Zhou, W. Zhang, Z. C. Hu and Y. S. Liu, J. Anal. At. Spectrom., 2020, 35(4), 736–745 RSC.

This journal is © The Royal Society of Chemistry 2021