Atomic spectrometry update: review of advances in elemental speciation

Contents

• 1 Introduction

• 2 Books and topical reviews

• 3 CRMs and metrology

• 4 Elemental speciation analysis
  • 4.1 Antimony
  • 4.2 Arsenic
  • 4.3 Cobalt
  • 4.4 Chromium
  • 4.5 Copper
  • 4.6 Gold
  • 4.7 Halogens
  • 4.8 Iron
  • 4.9 Lead
  • 4.10 Manganese
  • 4.11 Mercury
  • 4.12 Nickel
  • 4.13 Phosphorus
  • 4.14 Platinum
  • 4.15 REEs
  • 4.16 Selenium
  • 4.17 Silver
  • 4.18 Sulfur
  • 4.19 Technetium
  • 4.20 Tellurium
  • 4.21 Thorium
  • 4.22 Tin
  • 4.23 Uranium
  • 4.24 Vanadium
  • 4.25 Zinc

• 5 Biomolecular speciation analysis

• 6 Abbreviations

• Conflicts of interest

• References

---

Abstract

This is the 17th Atomic Spectrometry Update (ASU) to focus on advances in elemental speciation and covers a period of approximately 12 months from January 2024. This ASU review deals with all aspects of analytical atomic spectrometry speciation methods developed for: the determination of oxidation states; organometallic compounds; coordination compounds; metal and heteroatom-containing biomolecules, including metalloproteins, proteins, peptides and amino acids; and the use of metal-tagging to facilitate detection via atomic spectrometry. As with all ASU reviews, the focus of the research reviewed includes those methods that incorporate atomic spectrometry as the measurement technique, although molecular detection techniques are also included where they have provided a complementary approach to speciation analysis. The number of publications covered this year has increased since last year but remains relatively low compared to many of the years that this ASU has been published for. However, there is a good breadth of elements covered, with the most popular elements still being As, Hg and Se, although there is little novelty in the analytical approaches employed, which reflects the maturity of elemental speciation analysis, and over 25 elements are covered in total, including a growing number of papers covering essential macro and trace elements, such as Fe and Mn and P, and S as a heteroatom in proteins. The advent of ICP-MS/MS instrumentation, allowing for the cost-effective quantification of S as SO at m/z 48, thus negating the ¹⁶O₂ interference is probably behind the rise in papers on this element once again, the quality of the abstract for many of the papers is poor, with details on the methodology, key results with data included, conclusions and implications thereof missing. This is likely to lead to fewer researchers reading the article or considering it for reviews such as those produced by the ASU.

---

1 Introduction

This latest update adds to that from last year¹ and complements the five other annual Atomic Spectrometry Updates: advances in environmental analysis;² advances in the analysis of clinical and biological materials, foods and beverages;³ advances in atomic spectrometry and related techniques;⁴ advances in X-ray fluorescence spectrometry and its special applications;⁵ and advances in the analysis of metals, chemicals and materials.⁶ To mark the 40^th anniversary of the ASU, an overview of its history, topic coverage and a call for writers, has also been published this year.⁷

All the ASU reviews adhere to a number of conventions. An italicised phrase close to the beginning of each paragraph is intended to highlight the subject area of that individual paragraph. A list of abbreviations used in this review appears at the end. It is a convention of an ASU 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.

2 Books and topical reviews

No textbook devoted entirely to elemental speciation, or book chapters or sections dealing with the topic have been published this year.

A number of review articles have appeared concerned with aspects of elemental speciation analysis, several of which were focused on the instrumental techniques involved, with heavy emphasis on ICP-MS. De Pauw et al. reviewed (88 references) the capabilities of three types of WD-XRF spectrometers: in-house laboratory constructed, commercially available, and synchrotron facility-based instruments.⁸ The reviewers pointed out that WD-XRF has considerably better resolution than that of ED-XRF and can distinguish between characteristic X-ray lines separated only by a few eV. The focus of the review was very much on the design of the various spectrometers and not so much on the applications; however, three examples of speciation analyses performed at synchrotron facilities were discussed. These included speciation of Fe sulfides, S, and Si. They also cited a 2018 report of speciation applications with a commercial WD-XRF spectrometer. They tabulated relevant information about all commercially available instruments, which were listed by manufacturer and type of instrument. For each instrument, at least one example of an application was described in the text.

Quantitative bioanalysis by ICP-MS for clinical diagnosis was reviewed (187 references) by Du et al.⁹ The bulk of the review was concerned with the applications of exogenous tags (either NPs or lanthanides) and, based on the different binding affinities between detectable elemental tags and biomarkers, the relevant bio-interactions were categorised into 5 groups: antibody–antigen affinity binding, aptamer–affinity binding, nucleic acid–base pairing, enzyme–substrate interactions, and their cooperative bio-interactions. Each of these five types of interaction was considered in some detail. The reviewers also considered application of time-resolved ICP-MS to single-cell analysis, to follow cellular uptake and to assessment of cell-to-cell variations. Multiparametric single-cell analysis by mass cytometry, a specific application of ICP-TOF-MS, was also discussed. The review concluded with a brief section on LA-ICP-MS, which the reviewers described as a quasi-non-destructive microanalytical technique for solids that provides in situ quantitative determination of elemental/isotopic composition without complicated sample preparation. Although they qualify this characterisation by pointing out that the technique can only perform semi-quantitative analysis of biomarkers in pathological tissue samples. The reviewers concluded that there are still some barriers to the widespread adoption of ICP-MS in the clinical area. They note that the current sample introduction system requires, for a single analysis, sample volumes >1 mL, whereas most clinical samples are too small (without dilution). Also, in the diagnosis of disease or in prognosis monitoring, most key biomarkers or drug metabolites are often present in complex blood or tissue matrices, and so more defined and efficient sample pretreatments, to retain the integrity of the clinical sample, are needed. The reviewers also noted that the instrumentation is large and expensive, and called for miniaturisation with the development of “gasless” plasmas. The determination of trace elements in biological samples by procedures involving preconcentration prior to quantification by ICP-MS was reviewed (75 references) by Oviedo et al.¹⁰ They described their study as an exhaustive overview of liquid- and solid-phase preconcentration techniques from 2000 to the present. The vast majority of the articles they discussed were concerned with the analysis of clinical materials (urine, blood, plasma, hair) with only a few papers concerned with other biological matrices (such as food). The reviewers identified 6 preconcentration procedures, tabulated the respective advantages and disadvantages and gave examples of the application of several of these, with something of an emphasis on the use of magnetic nanoparticulate materials. Just 6 of the articles covered dealt with speciation analysis, mostly of compounds of As, Hg and Se. The reviewers acknowledged the limitation of the number of species that can be determined by procedures involving liquid-phase extraction or SPE, but noted that there were several examples in which the preconcentration step was followed by a chromatographic separation. A review (98 references) by Wang covers applications of metallomic and metalloproteomic techniques in biomedical research.¹¹ Although the original article is in Chinese, the abstract, references, and figure captions are in English and AI could be used to translate the text. The review focused on recent developments in metalloproteomics applications, particularly in identifying metal–drug interactions at the single-cell level, mapping metal distributions in tissues and organs, and analysing metal–protein interactions through bioinformatics approaches. The reviewer stressed the importance of ICP-MS as an element-specific detector for various separation techniques including CE, GE and HPLC. Examples of the applications of XRF, SRXF, and LA-ICP-MS for imaging of metals in tissues and cells were discussed with reference to the study of metal-related diseases, cancer, and neurodegenerative disorders. Other topics included (a) applications of single-cell analyses in cancer drug research, environmental toxicology, and studies of metal NP uptake, and (b) the contribution of bioinformatics tools and databases to the study of metal-protein interactions including the prediction of metal-binding sites.

The role of ICP-MS in the toxicological studies of metallic NPs was reviewed (121 references) by Fernández-Trujillo et al.¹² The reviewers covered various types of measurement including total elemental determination, single-particle or single-cell analyses, coupling with separation techniques, as well as the potential of LA for spatially resolved sampling. Applications to both in vitro and in vivo toxicological studies were examined in considerable detail with the help of two summary tables, containing information from 31 and 22 references, respectively. A further summary table (19 references) covered both in vitro and in vivo toxicological studies of NP transformations. A considerable portion of the review was devoted to future developments, under which heading the reviewers discussed problems associated with isobaric interferences from matrix components; the need for separation, clean-up, purification and preconcentration; the lack of harmonisation and standardisation of data acquisition and processing conditions; the limited analytical throughput for applications involving complex samples; the replacement of the typical quadrupole detector with a “multiple mass” analyser (such as TOF or SF); the lack of adequate NP suspension standards for classical approaches, such as external calibration; and the possible benefits of IDA. Finally, the reviewers pointed out the critical need to distinguish between NP material internalised in a cell from that bound to the cell surface.

In a tutorial overview (39 references), Feldmann et al. described the contributions of ICP-MS to non-target screening (NTS) in environmental monitoring.¹³ The authors claim that, among other characteristics, ICP-MS provides “unambiguous chemical composition information” that can enhance the confidence of compound identification. Given that all compounds introduced into an ICP are completely dissociated to atoms, which are subsequently ionised, it is hard to follow the argument about the provision of information relating to chemical composition. Towards the end of the article, the authors showed the power of running ES-MS in parallel with ICP-MS as detectors for HPLC. The authors also considered that responses for a given element are compound-independent, a claim that is easily refuted by examination of the literature in which researchers have provided values for the slopes of the calibrations of various species. However, they did demonstrate for four As compounds (As^III, As^V, MMA and DMA) separated by HPLC (no details were given) that the slopes of the calibrations obtained were probably (no ± terms were given) not significantly different. They pointed out that mass balance calculations for a given element allows assessment of the recovery of those elements and hence evaluation of methodologies for potential losses at the various stages. This feature was nicely illustrated by an example involving the determination of F-containing compounds in an unspecified solid environmental sample. The authors acknowledged the inability of ICP-MS to determine C, N, O, which is not strictly true as some workers are now detecting microplastics using either the ¹²C or ¹³C signal, and the difficulties with the detection of halogens, particularly F. They also noted that not all organic solvents are compatible with the ICP source, and that the challenge of dealing with co-eluting compounds can be addressed by running an ES-MS detector in parallel with the ICP. In a follow-up tutorial (104 references), the same research group expanded on the capabilities of simultaneous high-resolution molecular and atomic mass spectrometries for speciation determinations of both target and non-target analytes.¹⁴ They highlighted, in a summary table, 15 publications describing separation by HPLC and noted the increased application of HILIC, the mobile phases for which are typically water plus a less polar solvent (such as ACN, acetone, or MeOH) that are ideal for coupling with ES-MS and are also suitable for applications to biomolecules, such as proteins. The authors noted the significant improvement in performance now available from orbital ion trap mass spectrometers. The methodology was illustrated by the analysis of the livers of sea turtles of Caretta caretta species, found dead on the Brazilian coast, for species containing As, P and S. The sample preparation, instrumental set up and operation were described in detail. The data analysis for each of the elemental species detected was also described in considerable detail, illustrating the various MS modes available with the Orbitrap instrument (the molecular mass spectrometer). The article includes two useful flow diagrams that summarise the steps in characterising novel heteroatom-containing biomolecules for both targeted and non-target analytes. The reviewers concluded with a list of the advantages of the parallel atomic and molecular MS approach to speciation analysis, together with a slightly longer, list of drawbacks and challenges.

Dakova et al. reviewed (186 references) the applications of SPE for elemental speciation analysis in which analytes were retained on polymeric solids.¹⁵ The scope in terms of element coverage was rather limited, being confined to mostly inorganic species of As, Cr, Sb and a small number of organic species of As and Hg. The period reviewed was the last 15 years (up to 2023) with a significant number of the articles coming from the period 2020 to 2023. The applications of various types of SPE materials were summarised in five tables, with the vast majority of samples being environmental waters. It should be borne in mind that there is a large (and growing) literature related to the use of such materials for preconcentration of a much wider range of elements mainly from food matrices (see, for example, the Atomic Spectrometery Update devoted to clinical and biological materials, foods and beverages¹⁶). The reviewers highlight recent developments in the field of SPME, magnetic SPE, and uses of IIPs to which was devoted considerable space in the review as the various methods of preparation (chemical immobilisation, trapping, surface imprinting) were covered in detail. The reviewers also covered monolithic columns, MOFs, covalent organic frameworks, as well as various biopolymers (chitosan-based; cellulose-based; β-cyclodextrin, dextran and alginates; and peptides). Magnetic materials were dealt with in a separate section, and although there was a section entitled “Polymeric Composites with Nanoparticles”, most of the work in which NPs were used was dispersed throughout the review. The reviewers pointed out the favourable green characteristics of such SPE methodologies and predicted continuing growth in the development of new materials.

3 CRMs and metrology

Researchers at the Division of Chemical Metrology and Analytical Sciences, National Institute of Metrology, Beijing have produced two trimethyllead in aqueous solution calibrant CRMs, GBW(E)080971 and GBW(E)080972, that contain 92.73 ± 3.15 μg g⁻¹, and 0.740 ± 0.030 μg g⁻¹ (as Pb), respectively.¹⁷ The combined uncertainties used a coverage factor, k, of 2. The researchers included two important steps in the certification protocol, namely (1) the establishment of a traceability chain to the CRM GBW08619, which is a primary Pb^II calibrant solution, and (2) incorporation of the Pb isotope variations in both trimethyllead and GBW08619 into the certified values and combined uncertainties. The paper described the details of the homogeneity and stability studies, the impurity assessment (with NMR spectrometry and HPLC-ICP-MS), the traceability assurance protocol and the inter-laboratory scheme involving five independent laboratories.

4 Elemental speciation analysis

4.1 Antimony

Interest in Sb speciation methods continues at the modest level seen in the recent past. Han et al. separated inorganic Sb species by HPLC with detection by HG-AFS.¹⁸ As the results are reported in a Chinese language journal, the details are hard to decipher. The novelty would appear to be in the chromatography, which was performed on a Hamilton PRP-X100 column with a mobile phase of 5 mmol per L EDTA (as the dipotassium salt) and 2 mmol per L KHP, at pH 4.5 and 40 °C. For a flow rate of 1.5 mL min⁻¹, the separation time was 7 min and the LOD was 0.05 ng mL⁻¹. No information about the post-column HG could be discerned in the paper. The method was applied to one rainwater and one groundwater sample; the table of results is difficult to interpret, but it appears that Sb^V was detected in the groundwater. Recoveries for Sb^III and Sb^V spiked into the rainwater were 107 ± 1% and 98 ± 1%, respectively. The researchers also measured column recoveries, at several concentrations, which ranged from 96 to 108%. They also showed, by analysis with a MC-ICP-MS instrument, that no isotopic fractionation occurred during the separation. Other Chinese researchers, Ma et al. reported almost exactly the same HPLC separation (with detection by ICP-MS) to measure inorganic Sb species in wetland plants (Phragmites australis and Potamogeton crispus).¹⁹ Although they did not cite previous work in regard to their MAE procedure for total Sb, they did cite previous work in regard to the extraction of Sb species from plant tissue and the chromatography. For the extraction, 100 mg of sample was shaken with 20 mL 100 mmol per L citrate acid for 120 min, and the mixture sonicated for 1 h, centrifuged (10 min) and filtered. The final volume was 50 mL of which 50 μL was injected into a mobile phase of 10 mmol per L EDTA and 2 mmol per L KHP at pH 4.5. Separation was on a Hamilton PRP-X100 column (fitted with a guard column) at 40 °C. They also determined the Sb species in solution by a selective HG-AFS method (again a citation was provided) in which Sb^V was reduced to Sb^III by a mixture of thiourea and ascorbic acid, thereby allowing total Sb to be determined. Then the Sb^V was masked with citric acid allowing the Sb^III to be determined and hence Sb^V by difference. The researchers also described a method for determination of Fe (by ICP-OES) and Sb (by AFS) in iron plaque as well as a gradient centrifugation procedure to define subcellular distribution as a cell wall fraction, a cytosol fraction and a cytoplasmic organelle fraction. Each fraction was subjected to MAE with HNO₃ and H₂O₂ and measured by ICP-MS. None of the procedures was validated, other than to measure the recovery of species spiked at 20 μg L⁻¹ into method blanks for the HPLC method, for which values of 87 and 90% were obtained for Sb^III and Sb^V, respectively. Spanu and coworkers have updated their frontal chromatography ICP-MS method.²⁰ Following a rigorous optimisation, they decreased the analysis time to 150 s and the solution LOD values to 0.5 and 0.7 ng kg⁻¹ for Sb^III and Sb^V, respectively, and provided information about the washout to regenerate the column for the next analysis: 0.5 mol per L HNO₃ for 150 s at 1.45 mL min⁻¹, for a total analysis time of 5 min. They applied their method to the species extracted from seven PET samples, three of which were collected on the shore of Lake Como and four of which came from bottles purchased locally. To 160 mg of sample were added 20 g of 0.5 mol per L HNO₃ and the mixture sonicated for 10 min at room temperature. Total Sb was determined following MAE, with HNO₃ and H₂O₂, for which the sample mass was 50 mg, and the final solution mass was 30 g. The extraction procedure was validated by spike recoveries of both species at a concentration of 1 μg kg⁻¹, which were 93 ± 9% for and 95 ± 5% for Sb^III and Sb^V, respectively. They found total Sb in concentrations ranging from 200 to 340 mg kg⁻¹ and that the Sb leached corresponded to 1.4–6.5 μg per kg Sb^III and 10–39 μg per kg Sb^V. More Sb species were leached from the environmental samples. They calculated that their method scored 75 (out of 100) on the Analytical Greenness metric scale, which was higher than two other (HPLC-based methods) they examined.

4.2 Arsenic

The speciation of As has again received considerable attention this year, although the number of publications has fallen compared to previous years, and the scope of the reported studies is less diverse. No new analytical methods for As speciation have been reported, although several papers have described modification to existing techniques or new applications of existing methods for As speciation. The use of an ionic liquid, ferrofluid, for the preconcentration and simultaneous ultra-trace determination of iAs and Se species in waters by ICP-MS has been reported.²¹ A UAE sol–gel method was used for the synthesis of silica and titania coated and N-(2-aminoethyl)-3-aminopropyltrimethoxysilane functionalised magnetic nanoparticles (SCTCMNPs-AEAPTMS). The structural features of the SCTCMNPs-AEAPTMS were characterised by FTIR, SEM with ED-XRF, XRD, and TEM. Experimental conditions, including the sample solution pH, elution time, and eluent concentration, were optimised. After oxidation of As^III and Se^IV to As^V and Se^VI using H₂O₂, the total concentrations of As and Se were determined and those of As^III and Se^IV were obtained through subtraction of the concentration of As^V and Se^VI from the total concentrations. Under the optimal experimental conditions, the LOD values for As^V and Se^VI were between 0.3 ng L⁻¹ and 0.2 ng L⁻¹, respectively. The accuracy of the method was verified by analysing a CRM (NIST 1568a Rice Flour) and the measured As and Se concentrations agreed with the certified values based on a Student's t-test at the 95% confidence level. The proposed method was also successfully applied to the preconcentration and ultra-trace determination of As and Se species in different water samples. A liquid sampling APGD ionization source operating at a nominal power of 30 W and solution flow rate of 30 μL min⁻¹ and supported in a He sheath gas flow rate of 500 mL min⁻¹ has been interfaced to an Orbitrap MS and assessed for use in the rapid identification of As^III, As^V, MMA, DMA, and AB in a 2% (v/v) HNO₃ medium.²² Mass spectral acquisition in low-resolution mode, using only the ion trap analyser, provided detection of protonated molecular ions for AB (m/z 179), DMA (m/z 139), MMA (m/z 141), and As^V (m/z 143). The As^III present was oxidised to As^V, likely due to in-source processes. Typical fragmentation of the molecular compounds resulted in the loss of either water or methyl groups, as appropriate, i.e., introducing DMA also generated ions corresponding to MMA and As^V as dissociation products. Structure assignments were also confirmed by high-resolution Orbitrap measurements. Spectral fingerprint assignments were based on the introduction of solutions containing 5 μg mL⁻¹ of each As compound. Several papers have reported modifications to existing approaches to determine specific As species by AFS. A vortex-assisted LLME approach was used with a capric acid : lauric acid (CAP : LAU) (2 : 1) hydrophobic natural deep eutectic solvent by Botella et al. to determine iAs in complex samples such as nut-based beverages.²³ The As^III species were complexed with APDC and then extracted using 50 μL of a hydrophobic natural deep eutectic solvent, the preparation of which is described in a cited reference. The As^III species were then determined by HG-AFS. Optimisation of the sample volume, hydrophobic natural deep eutectic solvent volume, vortex stirring time, and cooling time was performed using a multivariate study. The method offered 100% As^III extraction efficiency and LOD values of 0.11 μg L⁻¹ for As^III and 0.30 μg L⁻¹ for As^V. The use of AFS has also been reported for the detection of As^III by selective electrolysis of As on nano-flower-like structured materials.²⁴ Four transition metal modifiers were evaluated and the Fe-modified Ni foam electrode with nano-flower structure gave the best efficiency for inducing As reduction and species selectivity. Following optimisation of the electrode materials and electrolysis conditions, good sensitivity and a wide linear range (0.1–50 μg L⁻¹) were reported.

Most environmental studies focusing on As speciation have been marine-related this year, particularly relating to the determination of the As species in seaweed. A review of recent developments in the speciation and determination of As in marine organisms using different analytical techniques has been presented by Sadee et al.²⁵ The review covered the extraction techniques and analytical methodology used to determine the 53 As species identified in the 224 publications covered in the review. The authors conclude that quality control remains one of the biggest challenges when reporting new As species that have been identified in marine samples. Seaweed is becoming increasingly popular in the Western diet as consumers opt for more sustainable food sources. However, seaweed is also well known to accumulate high levels of As, which may include iAs. Sim et al. have proposed a routine method to measure iAs in seaweed using an HPLC-ICP-MS method which excludes the co-elution of any arsenosugars that may complicate quantification.²⁶ The developed method was optimised using the statistical design of experiments (DOE) method and tested on a range of CRMs including TORT-3 (0.36 ± 0.03 mg kg⁻¹), DORM-5 (0.02 ± 0.003 mg kg⁻¹), and DOLT-5 (0.07 ± 0.007 mg kg⁻¹). The use of HNO₃ in the extraction solution allowed for the successful removal of interferences from arsenosugars due to degradation to an unretained arsenosugar species. A recovery of 99 ± 9% was obtained for iAs in CRM Hijiki 7405-b when compared with the certified value. The method was found to be suitable for high-throughput analysis of iAs in a range of food and feed matrices including Asparagopsis taxiformis seaweed, grass silage, and insect proteins. The method was, however, found to be limited with regards to the quantification of DMA in seaweed, as the acidic extraction led to overestimation of this analyte due to causing degradation of lipid species that are typically more abundant in seaweed than other marine matrices (i.e. arsenophospholipids). However, the concentrations of DMA quantified using this method may provide a better estimation with regard to exposure after ingestion and subsequent digestion of seaweed. Although arsenosugars are the predominant species of As in most seaweeds, the analysis of these compounds is often hampered by the lack of calibration standards needed for unambiguous identification and quantification. Morales-Rodríguez et al. have reported on the feasibility of centrifugal partition chromatography (CPC) for the isolation and purification of arsenosugars from algae extracts.²⁷ Several biphasic solvents systems were studied to elucidate the distribution of the As species. The physical characteristics of the compounds, the presence of strong acids, the formation of ionic pairs and the presence of salts at high ionic strength were all considered. A system consisting of 1-BuOH/EtOH/sat. (NH₄)₂SO₄/water at a ratio of 0.2 : 1 : 1 : 1 facilitated the separation. The total As content and the As speciation of the eluted solutions from CPC were determined by ICP-MS and IC-ICP-MS, respectively. The developed CPC procedure allowed the authors to obtain three arsenosugars with a purity of 98.7% in PO₄-Sug, 90.4% in SO₃-Sug and 96.1% in SO₄-Sug. Other workers have looked at the bioavailability of As species from seafood and the influence of choice of culinary treatment on dietary intake.²⁸ An in vitro digestion model using a dialysis membrane was used to assess the bioavailability. The As species in the seafood samples: Carcharhinus limbatus (blacktip shark), Penaeus monodon (giant tiger prawn), Loligo opalescens (squid), Crassostrea virginica (oyster), and Argopecten irradians (scallop) were determined by HPLC-ICP-MS-MS. A total of 8 culinary treatments, with and without seasoning, were used prior to the digestion study. The CRM DORM-3 was also analysed using the same procedure. Arsenobetaine was found to be the most bioavailable form of As (up to 711 ± 36, n = 20). The other As species were below the LOQ in the bioavailable fraction for all samples (1.0–3.0 ng per g As^III, 0.9–2.0 ng per g As^V, 0.4–1.0 ng per g MMA, and 0.2–0.6 ng per g DMA). The results showed that the type of seafood was the most influential factor in the bioavailability of the As species and that the culinary treatments did not affect the bioavailability or interconversion of the As species.

Two papers report on the detection of lipophilic As compounds this year. Raab et al. determined studied these As species in a cultured freshwater alga.²⁹ The alga, Chlamydomonas reinhardtii, was treated with 20 μg per L As^V and both fractionation and ICP-MS/ES-MS analyses used to make a comparison with the known As metabolite profile of wild-grown Saccharina latissimi. While the total As accumulation in C. reinhardtii was about 85% lower than in S. latissima, the relative percentage of arsenolipids was significantly higher in C. reinhardtii (57.0% vs. 5.01%). As-containing hydrocarbons and phospholipids dominated the hydrophobic As profile in S. latissima, but no As-containing hydrocarbons were detectable in C. reinhardtii. Instead, for the first time, an arsenoriboside-containing phytol (AsSugPhytol) was found to dominate the hydrophobic arsenicals of C. reinhardtii. This compound and its relatives have so far only been identified in green marine microalgae, open-sea plankton (mixed assemblage), and sediments but not in brown or red macroalgae. The authors suggest that this compound family might therefore relate to differences in the As metabolism between the algae phyla. The bioaccumulation of arsenolipids in the brains of pilot whales has been studied by Kopp et al.³⁰ The authors reported the identification of two As-hydrocarbons, as well as 3 arsenosugar phospholipids, in the brains of a pod of stranded long-finned pilot whales (Globicephala melas). In addition, they found an as absence of arsenobetaine (AsB), often the dominant As species in fish. The data, obtained using HPLC-ICP-MS-MS and HPLC-ES-MS, suggested that there was an age dependent accumulation of arsenolipids in the brains of the animals. The results also showed that, in contrast to other organs, the total As, in addition to the arsenolipids, accumulated in an asymptotic pattern in the brains of the animals. Total As concentrations were found to range from 87 to 260 μg As per kg wet weight and between 0.6 and 27.6 μg As per kg was present in the form of AsPL958 in the brains of stranded pilot whales, which was the most dominant lipophilic species present. The asymptotic relationship between total As, as well as arsenosugar phospholipids concentration in the brain and whale age suggested that the accumulation of these species may take place prior to the full development of the blood-brain barrier in young whales. A comparison between the organs of local squid, a common source of food for pilot whales, highlighted a comparable arsenolipids profile indicating a likely bioaccumulation pathway through the food chain.

The analysis of plant materials for As species also continues apace. A method developed specifically for the determination of As speciation in plants using 2D-RP-HPLC separation with ICP-MS and ES-MS detection has been reported by Izdebska et al.³¹ Different extraction procedures and HPLC methods were explored to assess their efficiency, determine mass balance, and improve the resolution of compounds in the chromatograms. Conventionally applied AEC facilitated the separation of well-documented As compounds in the extracts, which comprised of 19 to 82% of As present in extracts. To gain insight into the compounds which remained undetectable by anion chromatography (18–81% of As in the extracts), but still possibly metabolically relevant, the authors explored an alternative chromatographic approach further refining the online 2D-RP-HPLC system with both ICP-MS and ES-MS. As^III-phytochelatins, and other arseno-thio-compounds were detected and identified in extracts derived from the tree roots of seedlings grown in the presence of As^III and As^V. A group of arsenolipids was also detected in the roots of plants exposed to As^V. The speciation of As in rice has once again been widely studied. The performance of a N₂-sustained microwave inductively coupled atmospheric-pressure plasma (MICAP) – QMS was evaluated as an alternative to the traditional Ar equivalent for As analysis in rice.³² The object of the study was to avoid the well-known Ar-based spectral interferences (specifically ⁴⁰Ar³⁵Cl⁺) found with some sample matrices. When using the N₂-sustained MICAP-QMS, the major plasma background species are clearly N₂-based. A high chloride matrix was simulated using up to 1% (w/w) NaCl solution. The study used a slightly modified extraction procedure for the As-species based on the Food and Drug Administration (FDA) protocol EAM 4.11. In all rice samples, As^III, DMA, MMA, and As^V were successfully determined without polyatomic interferences. The LOD values were 0.26, 0.26, 0.13 and 0.22 μg kg⁻¹, respectively. The SRM NIST 1568b rice flour was used to determine recoveries (85–110%). The determination of iAs in polished rice, and Cd using ID, in combination with HPLC-ICP-MS has been reported following a study to develop a rapid screening method for As in rice.³³ The iAs in the rice samples was extracted using a heat-assisted HNO₃–H₂O₂ extraction. The aim was to develop an accurate and precise method which could also be applied as a rapid screening test to identify contaminated rice samples. The developed technique was applied to multiple rice CRMs, and in all cases the measured results were within the uncertainty range of the certified values. To avoid problems with As extraction from large grain rice, the rice sample was crushed into powder of <850 μm. The As speciation in rice grain grown in microwave- and biochar-treated soils has been studied in order to access if contaminated soils could be pretreated to lower As uptake and thus reduce As accumulation in the grain.³⁴ Microwave treatments have previously been reported to promote crop development and help immobilise various metal(loid)s in soil. Similarly, biochar has been reported to aid the immobilisation of As in soil when applied in the right conditions. In this study, soils contaminated with 0, 20, 40, 60, and 80 mg per kg As were used and 0, 10, 20 t ha⁻¹ of pine sawdust biochar added with microwave irradiation times of 0, 3 and 6 minutes. Two methods, AFS and HPLC-ICP-MS were used to determine the total As and As speciation, respectively. With increasing soil As concentration both the iAs (as As^III) and DMA concentration in rice grains increased significantly. However, rice grain As^III and DMA concentrations were up to 74% lower in microwave treatments (both 3 and 6 minutes) compared to the untreated control. The As in the samples was found to be present mainly as DMA (70.6%) and the remainder was As^III, accounting for 29.4% of the total As. In the case of the biochar treatments, the As^III increased in both cases compared to the untreated control. The authors suggested that the microwave treatment decreased microbial richness but did not completely eliminate the soil microbes. However, further studies are needed to determine the recovery rates for soil microbiota in situ and for different soil types. Potential changes to the As speciation in rice as a result of cooking has again been reported this year.³⁵ Two cooking methods (excess water and parboiling with absorption where the rice is initially parboiled and then drained and cooked until all the cooking water is absorbed by the rice) were evaluated. The As species were determined by LC-ICP-MS. A range of other essential nutrient elements (Cu, Fe, K, Mg, Mn, Mo, P, Se and Zn) were also determined. White, parboiled and brown rice were used with As-safe (0.18 μg L⁻¹) and As-spiked (10 and 50 μg L⁻¹) tap water. The exposure risk was calculated using the margin of exposure for both low (UK) and high (Bangladesh) rice per capita consumption scenarios. The results obtained showed that the excess water and the parboiling with absorption methods produced similar efficacy of iAs removal (54–58%) for white and brown rice. However, the excess water method was better at removing iAs from parboiled rice (about 50%) than parboiling with absorption (about 39%). The study found that cooked brown rice was superior to other rice types in many essential nutrient elements, and that the choice of cooking methods could significantly affect the loss of Cu, Fe, K, and Mo. For both cooking methods, cooking with iAs-spiked water significantly increased iAs in all rice types: white > parboiled > brown. However, when using As-spiked water, the parboiling with absorption method retained more iAs than excess water. The risk evaluation calculations showed that cooking rice with 50 μg per L As significantly raised the As-exposure of the Bangladesh population due to the high per capita rice consumption rate, reinforcing the importance of accessing As-safe water for cooking.

The speciation of As in mushrooms has again been reported this year.³⁶ Four different mushroom species were studied using HPLC-ICP-MS. The As mass fractions were determined in the mushrooms and found to range from 0.3 to 22 mg As per kg dry mass. The As species AB, DMA and iAs which are commonly detected in mushrooms were reported, but the more recently discovered homoarsenocholine was also present in Amanita muscaria and Ramaria sanguinea samples. In addition, a previously unidentified As species was isolated from Ramaria sanguinea and identified as trimethylarsonioacetamide (AB amide). The new arsenical was synthesised and verified by spiking experiments to be present in all of the investigated mushroom samples. The authors suggest that this AB amide could be an important intermediate in the biotransformation pathway of As in the environment.

The speciation of As in milk and milk-based products has been reported. Imitation milks based on vegetables have become more widely used in recent years as a substitute for cow's milk. The As speciation in both cow's milk and plant-based imitation milks (based on peanuts, nuts, soy, rice, vetch, almond and quinoa) has been investigated.³⁷ The researchers used an optimised HPLC-ICP-MS method employing AEC to determine As^III and As^V. Samples were lyophilised for a period of 24 hours at a pressure of 0.25 mbar at −53 °C. In order to reduce the quantities of both sample and acid employed, a micro digestion was then used employing 180 mg of sample and 800 μL of fluid samples with 600 μL of HNO₃ (65%) in a sonic bath for 15 min. Finally, 500 μL of H₂O₂ (30%) was added and placed in a water bath at 60 °C for 90 min. The LOD values for the method ranged from 0.2 to 0.7 μg L⁻¹. For the imitation milk samples based on peanuts and nuts, the iAs content was between 1.7 and 1 μg L⁻¹, whereas higher values were found in imitation milks based on soy, rice and vetch, i.e. 27.3, 13 and 8 μg L⁻¹ respectively. The cow's milk samples presented intermediate levels of iAs (4 μg L⁻¹). The almond- and quinoa-based imitation milks had iAs values below the LOD and so these imitation milks may be regarded as the safest to consume with respect to As content. The speciation of As in milk from cows fed on seaweed has been studied.³⁸ The total As in milk of control diets (9.3 ± 1.0 μg As per kg), n = 4, dry mass) was significantly higher than a seaweed-based diet (high-seaweed diet: 7.8 ± 0.4 μg As per kg, P < 0.05, n = 4, dry mass; low-seaweed diet: 6.2 ± 1.0 μg As per kg P < 0.01, n = 4, dry mass). Speciation of the As by HPLC-ICP-MS showed that the main species present were AB and As^V (37% and 24% of the total As, respectively). Trace amounts of DMA and AC were also detected in the milk. Apart from As^V being significantly lower (P < 0.001) in milk from seaweed-fed cows than in milk from the control group, the other As species showed no significant differences between groups. In conclusion, the paper stated that the lower total As and As^V in seaweed diet groups indicated a possible competition of uptake between As^V and phosphate, and the presence of AC indicated that a reduction of AB occurred in the digestive tract. Feeding a seaweed blend (91% Ascophyllum nodosum and 9% Laminaria digitata) did not raise As-related safety concerns for milk.

Several papers have reported on As speciation in geological samples. The identification of As species produced in the dissolution of crystalline orpiment under anoxic conditions has been studied by Zhang et al.³⁹ Orpiment (As₂S₃) is an important secondary mineral in the geochemical process of As in the environment. The dissolution of orpiment has a close relationship with the migration and transformation of As and the dissolved species of As₂S₃ closely relate to sulfide in anoxic and sulfidic environments. The study, using XAS, LC-HG-AFS and Raman spectroscopy, focused on the As species formed when As₂S₃ dissolved in the presence and absence of excess S^II under anoxic conditions. The results showed that the As produced when As₂S₃ dissolved in excess S^II, contained a mixture of As^III and thioarsenite. Based on a linear combination fitting, thioarsenite was the dominant As species (88.2%) with As^III as the remaining component. However, the percentage of thioarsenite decreased to 43.7% if As₂S₃ was dissolved in the absence of excess S^II, indicting thioarsenite was favoured under sulfidic conditions. The findings gave further insights into the role and formation mechanism of thioarsenite in the dissolution process of As₂S₃, and may be important when considering the remediation of As contaminated soils and waters. A study to characterise the association between borate minerals and high As concentrations in brines and thermal waters, based on the Puna region's borate deposits in the central Andes of Argentina, has been published.⁴⁰ Five borate samples were collected from the Olaroz salt flat nucleus and thermal springs, together with associated water samples. Various analytical techniques including ICP-MS, ICP-OES, SR-μXRF, XRPD, Rietveld analysis, μFT-IR, and XPS, were employed to determine bulk and surface chemical compositions, mineral identification, and the solid-state speciation of As and B. The study revealed that under oxidising conditions and in the absence of organic matter, aqueous As species interacted with ulexite through a stepwise process involving charge neutralisation, cationic bridge formation, and surface complex formation with polyborate and As^V oxyanions. However, in environments associated with microbial mats or organic-rich sediments, the dissolved As^V is reduced to As^III, which formed complexes with functional groups of organic matter. The coexistence of As^III and As^V in specific layers suggested potential remediation strategies targeting organic matter for the removal of the more toxic As^III may work in similar geological settings.

The study of As in groundwater has attracted attention again this year. High levels of As in groundwater are influenced by a combination of processes: reductive dissolution of Fe minerals and formation of secondary minerals, metal complexation and redox reactions of organic matter and formation of more migratory thioarsenate, which together can lead to significant increases in As concentration in groundwater. A study of a typical S- and As-rich groundwater site within the Datong Basin in China has been conducted by Zhang et al.⁴¹ The aim of the study was to explore the conditions for thioarsenate formation and its influence on As enrichment using a range of techniques, including HPLC-ICP-MS, hydrogeochemical modelling, and excitation–emission matrix fluorescence spectroscopy. The shallow aquifer used in the study exhibited a highly reducing environment, marked by elevated sulfide levels, low concentrations of Fe^II, and the highest proportion of thioarsenate. In the middle aquifer, an optimal ΣS/ΣAs led to the presence of significant quantities of thioarsenate. In contrast, the deep aquifer exhibited low sulfide and high Fe^II concentration, with As primarily originating from dissolved iron minerals. Redox fluctuations in the sediment driven by S–Fe minerals generated reduced S, thereby facilitating thioarsenate formation. The organic matter played a crucial role as an electron donor for microbial activities, promoting Fe and sulfate reduction processes and creating conditions conducive to thioarsenate formation in reduced and high-S environments. In further research, microbially-mediated As speciation in aquifers governed by the As^V-reductase enzyme, ArsC, has been studied.⁴² Two forms of a bacterial strain highly resistant to As, Citrobacter youngae IITK SM2 (CyIITKSM2), ArsC1 (119aa) and ArsC2 (141aa), were biochemically characterised. The bacteria were isolated from a mixed-oxic groundwater obtained from the middle-Gangetic Plain in India. Coupled-As^V reductase assay and IC-ICP-MS analysis confirmed that ArsC2 showed higher As^V reduction than ArsC1 in the dissolved phase, which was consistent with the prominent structural changes found in ArsC2 as identified using circular dichroism spectroscopy. Furthermore, the two ArsCs were able to mobilise As from solid-bound As^V-loaded goethite, AsG predominantly as As^III. However, the total As released in the presence of ArsC2 was about 38% and 88% higher, respectively, when compared to the ArsC1-containing and ArsC-free conditions. A process-based model that considered ArsC-mediated As^V reduction to As^III in the dissolved phase, and surface complexation of As^V and As^III on goethite, suggested that the extent of As^V binding with ArsC was not affected by whether As^V was dissolved or was sorbed. However, the catalytic reduction rate was at least an order of magnitude lower in sorbed As^V than in dissolved As^V. Mutants of ArsC2 exhibited variable but reduced efficiencies compared to the wild-type ArsC2. This reduction was attributed by the authors to the C-terminal loop observed in the AlphaFold-predicted structure of ArsC2, which was absent in ArsC1.

The speciation of As in soils continues to be a popular topic of research. Extensive excavations for urban and subterranean construction can lead to soil and groundwater contamination with geogenic As, requiring effective management strategies. The potential mobility and geochemical speciation of geogenic As in the deep subsurface soil of the Tokyo metropolitan area in Japan, has been investigated.⁴³ The study examined the chemical speciation and solubility of geogenic As in soil samples collected at 25 cm intervals from boreholes extending up to 16 m deep within the alluvial Yurakucho Formation and the terrestrial Kanto Loam Formation in the Tokyo metropolitan area. Soils from the Yurakucho Formation exhibited significantly higher total As concentrations (10.5 ± 3.26 mg kg⁻¹) compared to those from the Kanto Loam Formation (5.58 ± 1.88 mg kg⁻¹), with notably elevated levels of water-soluble As throughout the profile. Analysis by XANES revealed that As-bearing sulfide species, including As₂S₃ and FeAsS, were the predominant forms in the Yurakucho Formation, while As^V species were more prevalent in the Kanto Loam Formation. Studies using μXAFS combined with μXRF analysis identified framboidal pyrite, characterised by micron-sized grains (about 10 μm), as the primary sink for As sequestration in the Yurakucho Formation, where As occurs mainly in sulfide-associated forms. The findings from the study highlighted the importance of characterising geogenic As speciation and the need to assess its leaching potential to avoid the associated environmental risks posed by As in excavated soils. The reaction kinetics of As sorption and oxidation in natural Mn-oxide-enriched soils has been reported.⁴⁴ The authors quantified the oxidative kinetics and adsorptive capacity of five soils rich in pedogenic Mn- and Fe-oxides through As^III oxidation batch reactions over a range of pH and temperatures to mimic diverse environmental conditions. The two A horizons were found to be less reactive and enriched in Mn^IV, compared to the B horizons, particularly the subsoil containing the Mn-rich wad material. The reaction kinetics fitted a pseudo-first-order reaction with distinct fast and slow phases. The baseline reactions were at pH 7.2 at 23 °C. Adjusting the pH to 4.5 or 9.0 increased the reaction rates. Decreasing the temperature to 4.0 °C reduced the reaction kinetics, while raising the temperature to 40 °C increased the As^III oxidation rate. The pH and temperature changes altered the reaction kinetics due to shifts related to the point of zero charge, the total system energy, and surface passivation from adsorbing As and Mn species. Mapping with SXRF indicated that As only penetrated the surficial layers of most Mn oxide-containing nodules found in the soil. After the As^III oxidation reaction in the pedogenically weathered subsoil, the use of XAS demonstrated significant differences in the average Mn oxidation number between the nodules' outer layers compared to the soil matrix and nodule centres.

A number of clinical studies focusing on As speciation have been published. Trace element concentrations in toenail clippings have increasingly been used to measure trace element exposure in epidemeological research. Conventional methods such as ICP-MS and HPLC-ICP-MS are commonly used to measure trace elements and their speciation in toenails. However, the impact of the removal of potential external contamination on trace element quantification have not been so well reported. Faidutti et al. have studied the micro distribution of trace elements (As, Ca, Co, Cu, Fe, K, Mn, Ni, Rb, S, Sr, Ti, and Zn) in dirty and washed toenails and the speciation of As in situ in toenails using synchrotron XRF and laterally resolved XANES.⁴⁵ The XRF analysis showed different distribution patterns for each trace element, consistent with their binding properties and nail structure. External (terrestrial) contamination was identified and distinguished from the endogenous accumulation of trace elements in toenails using metallic markers for terrestrial contamination. The XANES spectra showed the presence of one As species in washed toenails, corresponding to As bound to sulfhydryl groups. In dirty specimens, a mixed speciation was found in localised areas, containing As^III–S species and As^V species. The As^V was thought to be associated with surface contamination and exogenous As. In another study, the feasibility of using toenails as biomarkers for estimating iAs exposure in Japanese adults has been reported by Oguri et al.⁴⁶ Three sets of 7-day diet records and toenail clipping samples were collected from 39 healthy adult participants at intervals of 1–6 months over 4–8 months from June 2019 to March 2020. The analysis sample sets comprised 113 sample sets obtained from 38 subjects: 56 samples from 19 males and 57 samples from 19 females. The speciation of As in the toenail samples was performed using HPLC-ICP-MS. The sum of the iAs and MMA or sum of As species concentrations in the toenail samples was used as an index of iAs exposure. The geometric mean concentration of iAs + MMA or sum-As in toenails was 0.180 μg As per g or 0.284 μg As per g. The estimated geometric mean of daily dietary iAs exposure was 0.147 μg per kg per day. log-transformed iAs + MMA or sum-As concentrations in toenails did not predict dietary iAs exposure levels from rice and hijiki consumption in both males and females. Similarly, toenail iAs + MMA or sum-As concentrations showed no correlation with dietary iAs exposure levels from rice or hijiki consumption. The study concluded that human toenail clippings were not a suitable biomarker for assessing long-term iAs exposure levels in Japanese individuals based on the observed range of iAs and its metabolite concentrations in toenails.

The determination of the As speciation in urine remains a popular subject for study. A novel urine self-sampling device with an automated preparation-sampler has been designed for the analysis of As metabolites by HPLC-ICP-MS following collection at home.⁴⁷ When the samples were collected, they were immediately stored in a sampling vial and Ar was pumped in to displace O₂, thereby inhibiting the oxidation of MMA^III and DMA^III. After transportation to the laboratory, the sampling vial was loaded directly onto the automated preparation-sampler device, where the urine sample could be automatically mixed with diluent, filtered, and loaded into the HPLC-ICP-MS for As speciation analysis under anaerobic conditions. For a single sample, the sampling time and the analysis time are <8 and <18 min, respectively. The recoveries of MMA^III and DMA^III in urine over 24 h at 4 °C were 86 and 67%, surpassing the conventional sampling method which gave 28 and 67%, respectively. The LOD values for AC, As^III, MMA^III, DMA^V, MMA^V, DMA^III, and As^V were in the range 0.03 to 0.10 μg L⁻¹ with precisions of <10%. The simultaneous speciation analysis of As and I in human urine has also been reported.⁴⁸ The study used HPLC-ICP-MS to simultaneous determine six As species (i.e. AB, As^III, DMA, AC, MMA, and As^V), and four I species (i.e., iodate, 3-iodo-tyrosine, 3,5-diiodo-tyrosine, and iodide) in the urine. The chromatographic separation was performed using AEC on a Dionex IonPac As7 column. The separation was initiated with a mobile phase of 0.5 mmol per L ammonium carbonate solution, followed by 50 mmol per L ammonium carbonate at 100 mmol per L ammonium nitrate solution (with 4% methanol). The LOQ values obtained ranged from 0.045 to 2.26 μg L⁻¹. At three spiked levels (10.0, 20.0, 50.0 μg L⁻¹), the average recoveries (%) ranged from 87.4 to 113%, with RSDs from 0.4 to 17.2%. The ratio of the sum of six As species to the total As measured by ICP-MS ranged from 77.4 to 121%, and the ratio of the sum of the four I species to the total I ranged from 70.7 to 115%. The speciation analysis of As in human urine has also been used as part of large-scale epidemiological study on sources of exposure and human metabolism. Glabonjat et al. have reported on the Multi-Ethnic Study of Atherosclerosis (MESA) in the USA, designed to assess chronic low-moderate As exposure and health effects in an ethnically diverse US population.⁴⁹ Quantification of the As species in 7677 MESA spot urine samples was by HPLC-ICP-MS. The method was capable of detecting 11 individual As species with LOD values of 0.02–0.03 μg As per L, with a method accuracy of 98% and precision of 5%. The main As species (AB, MMA, DMA, and total iAs), were detectable in 95% of urine samples. Two other minor urinary As species, dimethylarsinoylacetic acid and dimethylarsinoylpropionic acid, thought to be potential metabolites of seafood-related arsenicals, were also found. The study found differences in individual As species excretion by race/ethnicity, with Asian-American participants featuring 3–4 times higher concentrations compared to other participants. The study also found differences relating to site, body mass index, smoking status, rice intake, and water As levels, potentially indicating different exposures or related to individual bio-metabolism.

Finally, the determination of As species on microplastics ingested with food has been reported. A dynamic stomach model integrated with CE-ICP-MS was used in a study of the release and transformation of As species from microplastics during digestion.⁵⁰ The microplastics are thought to act as carriers of environmental As species into the stomach with food, which are then released during digestion. A 3D-printed dynamic stomach model, with a soft stomach chamber enabled the behaviours of gastric peristalsis, gastric and salivary fluid addition, pH adjustment, and gastric emptying to be controlled by an in-house program after oral ingestion of food containing As-microplastics was used in this work. The gastric extract from digestion was introduced into the spiral channel to remove the large particulate impurities and then filtered on-line to obtain the clarified As-containing solution for subsequent speciation analysis by CE-ICP-MS. The digestion conditions and pretreatment processes of the dynamic stomach model were tracked and validated. The release rates from the As-microplastics digested by dynamic stomach model were also compared with those digested by the static stomach model and dynamic stomach model without gastric emptying. The release rate of iAs on microplastics was higher than that of organic As throughout the gastric digestion process, and 8% of As^V was reduced to As^III. The dynamic stomach model with CE-ICP-MS facilitated LOD values for As^III, DMA, MMA, and As^V of 0.5–0.9 μg L⁻¹, with precisions of ≤8%.

4.3 Cobalt

The vitamin cyanocobalamin (Vitamin B12) in milk powder samples has been determined by using HPLC-ICP-MS.⁵¹ Separation was performed on a Luna C18 column (250 × 4.6 mm; 5 μm) with a mobile phase consisting of 1.6 mmol per L EDTA and 0.4 mmol per L KH₂PO₄ in 60% methanol solution (pH = 4.6). Under these conditions, free Co and cyanocobalamin were separated in 4.0 minutes. The selection of the mobile phase was optimised to obtain the best separation of free Co and cyanocobalamin as the former could seriously interfere with the latter. Optimisation of the ICP-MS collision parameters enabled the determination of vitamin B12 in dairy products without the need of applying a previous SPE step. The LOD obtained for vitamin B12 was 0.63 mg kg⁻¹ and the method was applied to infant milk powder samples after performing protein precipitation by using a mixture of K₄Fe(CN)₆ solution and Zn(CH₃CO₂)₂ solution under UAE for 40 minutes. Mass balance evaluation of free Co and cyanocobalamin accounted for 90–97.5% of the total Co present in the sample. An additional paper which reports a multielement speciation method for Co, Cu, Fe, Mn and Zn by HPLC-ICP-MS is covered in Section 4.5, Copper.

4.4 Chromium

Interest in Cr speciation has declined somewhat over recent years, possibly because it is now clear that concentrations of Cr^VI in foods and beverages are below the LOD values of SSID-ICP-MS methodologies. The benefits of SSID were emphasised by Chung in a review (18 references) of recently published work (2018 to 2023), who also considered whether the LOD values achieved were suitable for assessing dietary exposure to Cr^VI.⁵² Details of matrices, sample preparation, techniques for analytical detection and analytical performances (LOD, LOQ and recoveries) were summarised in a table. In most of the 11 studies featuring off-line detection, (ETAAS, FAAS, ICP-MS, molecular fluorescence, surface-enhanced Raman and voltammetry) the determination of one of the species was calculated by subtraction of the concentration of other from the total Cr concentration. The reviewer considered that strong oxidising agents in the sample preparation step would give erroneous results by this method. In the seven studies featuring on-line detection, the reviewer highlighted a series of four papers by Saraiva and coworkers from 2021 describing a procedure in which samples were spiked with ⁵⁰Cr^III and ⁵³Cr^VI, followed by selective complexation of Cr^III with EDTA and then the reduction Cr^VI to Cr^III using 1,5-diphenylcarbazide to form a DPC-complex and determination of the two complex species using HPLC-ICP-MS, in KED mode with He as the collision gas. In the Saraiva papers, results were presented for the analysis of several foods (bread, breakfast cereals, dairy products and meat), Cr^VI was found in none, and for the effects of cooking, showing that Cr^III was not oxidised. However, they did find that Cr^VI spiked into some cooked foods was stable for up to a week. Chung suggested that the LOD for Cr species should be 0.03 μg kg⁻¹ (or lower) and recommended detection by SF-ICP-MS or ICP-MS/MS.

Just one paper describing an HPLC-ICP-MS procedure has appeared in this review period. Song et al. applied an alkaline UAE (a mixture of phosphate buffer, NaOH, Na₂CO₃ and MgCl₂) to a variety of foodstuffs (yoghurt, milk powder, rice flour, orange juice, green tea, white vinegar, and whole wheat bread) prior to a novel separation (in just 1.5 min) on a short, 50 mm, weak anion-exchange column with 70 mmol per L NH₄NO₃ at pH 7 as eluent.⁵³ It is not clear whether EDTA was added in all cases. The LOD was 0.1 μg kg⁻¹ and Cr^VI was not detected in any of the samples. To study the parameters affecting the rate of reduction of Cr^VI, the researchers spiked 6 food components (vitamin C, whey proteins, tea polyphenols, fructose gelatin and cellulose) and found, not surprisingly, that the rate of reduction increased with increasing acidity, temperature and reducing nature of the matrix. They also spiked the seven real food samples at concentrations of 5, 25 and 50 μg kg⁻¹ and were able to detect conversion at times as short as 2 min for some samples. Interestingly, Cr^VI was most stable in rice flour, with almost no conversion after 3 h.

Two SPE procedures for Cr speciation have been described. For the analysis of natural and wastewater, Ahmed and Soylak separated the Cr^VI as the APDC complex by DSPME on a MWCNT composite material with CuAl₂O₄@SiO₂ that was synthesised by sol–gel and calcination methods.⁵⁴ Any Cr^III was oxidised with KMnO₄ by heating at 70–80 °C for 30 min, allowing total Cr to be determined and hence the Cr^III by difference. Following absorption from 50 mL of sample, the retained Cr was dissolved in 3 mL of 3 mol per L HNO₃ in 10% acetone and hence a preconcentration factor of 51 was calculated. The Cr was determined by CS-FAAS and the method LOD was 6.2 μg L⁻¹. The accuracy was evaluated by the analysis of CRMs NWRI Canada TMDA-64.3 (fortified water) and INCT OBTL-5 (oriental basma tobacco leaves), which are certified for total Cr at concentrations of 283 μg L⁻¹ and 6300 μg kg⁻¹, respectively. Spike recoveries at 500 μg per L (total Cr) from distilled, tap and mineral waters ranged from 89 to 107%. The procedure was applied to three wastewater samples that contained between 589 and 2260 μg L⁻¹ total Cr. Faiz et al. synthesised a novel amine- and carboxyl-bifunctionalised zwitterionic magnetic nanocomposite material by grafting 3-(2-aminoethylamino)propyltriethoxysilane and carboxyethylsilanetriol onto silica-coated Fe₃O₄ nanoparticles via a sol–gel process.⁵⁵ They showed that the material selectively extracted Cr^III at pH 10 and Cr^VI at pH 3. To a sample volume of 45 mL adjusted to the appropriate pH were added 15 mg of the nanoparticulate material extractant and the mixture shaken for 30 min. After magnetic separation, the Cr species was dissolved in 3 mL of 10% HNO₃ in 10 min, and determined by ICP-MS. The LOD values were 0.009 and 0.008 μg L⁻¹, for Cr^III and Cr^VI, respectively. The accuracy was evaluated by spike recoveries at 0.4 and 0.8 μg L⁻¹ to three water samples (tap, river and lake). Both species were found in all three samples and recoveries ranged from 93 to 99%.

There are two reports of speciation based on selective DLLME extraction into a deep eutectic solvent (DES). Sapyen et al. extracted Cr^III as the complex with triethanolamine into a choline chloride-thymol hydrophobic DES followed by back-extraction into 0.2 mol per L HNO₃ and determination by ICP-OES.⁵⁶ The details of the overall procedure were difficult to find in the paper, but it appears as though any Cr remaining in the aqueous phase after extraction was determined by ICP-OES and considered to be Cr^VI. The best extraction efficiency for Cr^III was 97.5%, and so any Cr^III not extracted would be measured as Cr^VI; on the other hand, the extent of extraction of Cr^VI was 5%, which would contribute to the Cr^III concentration measured. The LOD was 7.5 μg L⁻¹, a value that the authors indicated was “determined using linear regression”, but visual inspection of the calibration plot provided in the supplementary information does not reveal how the calculation was done. The lowest standard in the sequence is about 6 μmol L⁻¹, corresponding to 312 μg L⁻¹. The accuracy of the procedure was assessed by spiking water samples (tap, ground, surface), none of which contained measurable concentrations of either species, at 2600 μg L⁻¹ (50 μmol L⁻¹). The recoveries ranged from 83 to 110%. Yang et al. developed a similar DLLME method in which the Cr^III was extracted as a thenoyltrifluoroaceton complex into a DES of menthol and decanoic acid.⁵⁷ They achieved preconcentration by extracting 10 mL of sample solution into 50 μL of DES, 20 μL of which was taken for determination by ETAAS. Following the dispersion and extraction steps (vortexing for 10 min), the mixture was centrifuged (5 min at 4000 rpm) and placed in a freezer. The frozen extractant was removed with a spatula and allowed to melt at room temperature. To determine Cr^VI, a reductant (hydroxylamine hydrochloride) was added and after re-analysis, the Cr^VI concentration was calculated from the difference. The procedure was applied to the analysis of cucumber, tomato, spinach, carrot, mushroom, lettuce, cabbage and coriander. For speciation analysis, 1.0 g of each dried and crushed sample was extracted with 10 mL of 0.2 mol per L HNO₃ at 100 °C for 30 min, the mixture was centrifuged and the supernatant (adjusted to pH 3.5) diluted to 50 mL of which 10 mL was taken for analysis. For total Cr determination, 1 g of sample was subject to MAE with 5 mL of concentrated HNO₃ and 2.5 mL of concentrated H₂O₂ at 150 °C for 10 min. The digest was diluted to 50 mL, the pH adjusted to 3.5 and 10 mL taken for extraction. No reducing agent was added, and so the digestion procedure did not appear to oxidise any Cr^III. The dilution due to sample preparation was 50-times, and the preconcentration factor (based on the volume ratio) was 200. However, as the authors give the value as 277, it is deduced that the final volume of DES was only 36 μL. There is ambiguity over the LOD, which was given as 0.2 μg kg⁻¹ in the abstract and 0.2 μg L⁻¹ in the text, and so if the value given in solution concentration units applies to the digest, the LOD in the solid would be 10 μg kg⁻¹, and the corresponding concentration in the solution introduced to the spectrometer would have been 40 μg L⁻¹, which seems rather high. So, it seems more likely that the LOD in the solid samples is 0.2 μg kg⁻¹, meaning the corresponding concentrations in the digest solution and the solution introduced to the spectrometer would have been 0.004 and 0.1 μg L⁻¹, respectively. The accuracy was checked by the analysis of a CRM GBW 10052 (green tea), which contains 920 ± 200 μg per kg total Cr, for which values of 311 ± 30 for Cr^III and 521 ± 70 for Cr^VI, respectively, were obtained. Both species were found in all 8 samples, with Cr^VI concentrations ranging from 9 to 26 μg kg⁻¹. Recoveries of spikes (5–50 μg kg⁻¹) ranged from 91 to 109%. The enhancement factor (based on the slopes of the calibrations) was 335. No information was provided about moisture content, nor did the authors discuss the discrepancies between their results and those of other researchers concerning the presence of Cr^VI in foods. For example, Song et al. whose results are discussed earlier, did not find any Cr^VI in green tea.

A method in which Cr species were selectively immobilised on a chemically modified electrode surface prior to determination by LIBS was devised by Huang et al.⁵⁸ The procedure involved the synthesis of two reduced graphene oxide (rGO) nanocomposites, NH₂–Co₃O₄–rGO and COOH–Co₃O₄–rGO, that were spin coated on to a polished aluminium electrode. At pH 5, the species were considered to be Cr(OH)⁺ and HCrO₄⁻, which were collected by the combined action of electrostatic interaction with the functionalised surface and the applied electric field. For the collection of Cr^III, the working electrode was held at a negative potential and for the collection of Cr^VI it was held at a positive potential. For each species, the collection period was 10 min. No details of the LIBS instrumentation was provided, but the LOD values were 0.12 and 0.1 μg L⁻¹ for Cr^III and Cr^VI, respectively. The procedure was applied to one real sample, a reservoir water that did not contain measurable concentrations of either species. Recoveries of spikes at 100 and 200 μg L⁻¹ ranged from 98 to 109%.

4.5 Copper

Three papers report on Cu speciation this year. A pioneering application of CE-ICP-MS/MS was developed for controlling the formation of liposomes and Cu-tripeptide complex (GHK-Cu)-containing liposomes by monitoring the ³¹P¹⁶O⁺ and ⁶³Cu⁺ signals.⁵⁹ The encapsulation efficiency (EE) was determined by dividing the peak area of the signal from ⁶³Cu⁺ (at the same retention time as ³¹P¹⁶O) by the sum of the peak areas of all ⁶³Cu⁺signals present in the electropherogram. Separation was achieved by means of a fused-silica capillary column (70 cm length, o.d. 375 μm and i.d, 75 μm) at a separation voltage of +15 kV and sample loading of 30 mbar × 5 s. The background electrolyte used was composed of 0.5 mmol per L Tris–HCl adjusted to pH 7.4. Several parameters, including liposome formation and composition and instrumental measurement conditions were studied. The LOD values of 0.027 mmol L⁻¹ and 0.71 mg L⁻¹ for P and the GHK-Cu complex, respectively were attained. An EtOH injection method was selected as the most rapid and effective to produce lipid-vesicle formation. The highest EE value (425) was obtained when liposomes were prepared for DSE-PEG2000/HSPC/Chol in a molar ratio 6/62/32. CE-ICP-MS/MS allowed straightforward determination of EE without the need of applying filtration and dialysis steps where losses of lysosomes and metal complexes on the filter and dialysis membrane are frequently reported. The use of CE-ICP-MS/MS seems to be a promising approach to monitor liposomal formulations to be applied in many fields. An interesting piece of work describes the capabilities of a novel experimental set-up based on in-line fibre separation-ICP-MS to control metal content variation in cell culture media (CCM) of Chinese hamster ovary cells which are commonly employed to produce monoclonal antibodies.⁶⁰ Separation of inorganic and ligated species of Co, Cu, Fe and Mn and Zn was achieved by using a microbore polypropylene (PPY) C-CP fibre column. Solutions were injected into the fibre in water containing 0.1% TFA at a flow rate of 500 μL min⁻¹. The adsorbed ligated species were subsequently eluted from the fibre by using a mixture of 80% ACN and 0.1% TFA, under gradient elution. To reduce the organic content of the eluate a 1 : 10 flow splitter and a T fitting for 1 : 10 diluted was employed. The resulting flow was then supplemented with 2% HNO₃ and carried out to the nebuliser of the ICP-MS for quantification. The paper includes a detailed evaluation of the effect of the composition of the mobile phase on ICP-MS stability, ionisation efficiency and spectral effects. The LOD values ranged from 65 pg per mL (Co) to 26 ng per mL (Fe). Accuracy was tested by employing an in-house mixture of the metals and recoveries ranging from 92.9 to 105% were obtained. The utility of the proposed methodology was demonstrated in two cases of study: chemical contamination of CCM and species variability due to shelf-life of CCM. The effect of geographical location and bamboo variety on trace element content and chemical forms in bamboo samples was investigated by coupling HPLC with ICP-MS and ES-MS/MS dual detection.⁶¹ Screening analysis by ICP-MS of acid digests of bamboo shoots and extracts of soils collected at five different locations indicate the presence of Ba, Cu, Fe, Mn, Ni, Rb, Sr and Zn. The application of an ANOVA test evidenced no significant differences on metal content between sites and bamboo species. Total metal content and chemical speciation was also performed in bamboo sap. The metallome profile in the bamboo sap using ANOVA showed a strongly significant difference between bamboo species. Speciation analysis was accomplished by SEC and HILIC separation. The SEC-ICP-MS analysis showed no differences in chromatographic profiles for Cu, Mn and Zn, regardless of bamboo variety or location whilst Fe and Ni-containing chromatograms varied among bamboo species. The results were compared and validated using HILIC separation. Significant efforts were undertaken by the authors to get the best conditions for the HILIC separation. Three types of HILIC columns and mobile phases were tested: Kinetic HILIC column (150 × 2.1 mm, 2.6 μm), SeQuamt ZIC-cHILIC (150 mm × 2.1 mm, 3 μm) and BEH amide (100 mm × 2.1 mm, 1.7 μm) and NH₄Ac:ACN or NH₄formate:ACN as mobile phases depending of the type of column. The BHE amide column gave the best results for Cu, Ni and Zn species whereas the ZIC-cHILIC column was selected for Fe and Mn speciation. Again, Ni exhibited greater chromatographic variation as a function of bamboo species. The metal species detected by ICP-MS were further identified by ES-MS detection. Metals were associated to LMW metallophores such as nicotinamide and deoxymugenic acid. Two metals, Fe and Ni, were also found as histidine and malate–citrate complexes. The paper is a good example of speciation analysis by combining elemental and molecular mass spectrometry.

4.6 Gold

The use of simultaneous multi-angle angular-resolved (AR) XANES for the surface-sensitive speciation of Au nanolayers of a Nano Gold Gel (NGG) has been reported on this year.⁶² The NGG, which consists of Au NPs stabilised with natural organic polymers, is used for Au layer conservation in gilded Au surfaces of cultural heritage items and was synthesised during this study. The experimental validation performed showed that the method could be successfully applied in differentiating between Au species, such as Au and Au₂O₃ in nanolayered structures with Fe₂O₃ and CaSO₄ base layers. The paper contains full details of the experimental approach employed.

4.7 Halogens

The main focus for recent developments in the speciation analysis of halogens and compounds containing halogens has been on the environmental biogeochemistry of inorganic I, or I together with Br. This work has used a range of LC chromatographic separation approaches coupled to advanced inorganic MS instrumentation. In seawater, inorganic I is found predominantly in its reduced and oxidised anionic forms, iodide (I⁻) and iodate (IO₃⁻). However, the rates, mechanisms and intermediate species by which I cycles between these inorganic pools are poorly understood. There is also very little information on the presence and concentration of dissolved organic I (DOI) containing compounds. In a sampling campaign to elucidate I cycling in the upper 1000 m of the Pacific Ocean, a set of analytical approaches for the quantitation of I species used a combination of direct separation on short off-line packed columns or preconcentration onto SPE cartridges followed by capillary LC separation, both with detection by ICP-MS.⁶³ The former approach was used for inorganic I⁻ and IO₃⁻ speciation whilst the latter was used to investigate the presence of the much lower concentration DOI-containing compounds. Both approaches were adapted for taking samples during long term sampling voyages followed by later, on-shore analysis. Water samples from various depths were spiked with ¹²⁹I, a long-lived radioactive isotope as a tracer to both study I redox transformations and for quantification of the species by IDMS. Separation of inorganic I⁻ and IO₃⁻ was performed by packing acid-cleaned quartz-glass columns (5 mm id, length not given) with AG-1X8 AE resin. Separate columns for each species were eluted with a mixture of HNO₃ (2.0 mol L⁻¹) and TMAH (18% v/v) solution (15 mL), collected and refrigerated. The role of TMAH was to keep the I stable in solution and prevent its volatilisation. The different I species were measured using both Q-ICP-MS and MC-ICP-MS. The results for inorganic I speciation in surface waters indicated a deficit of approximately 8% compared to the total I content, which was attributed to DOI species. To investigate the presence of DOI species a large volume of seawater (4 L) was extracted using SPE cartridges, which were stored frozen (−20 °C) for transport back to the on-shore laboratory. Following a rinse with high-purity water to remove salts, the analytes were eluted using MeOH (6 mL), which was concentrated to a smaller volume (500 μL) using vacuum centrifugation. The resulting sample was spiked with cobalamin as an IS and separated on a C18 column (150 × 0.5 mm, 3.5 μm particle size) using a ammonium formate–ammonium formate/methanol gradient over 30 min with a 10 min isocratic step. Detection using Q-ICP-MS required infusion of O₂ into the plasma to avoid carbon build-up on the cones. The results indicated the presence of at least 3 DOI peaks with one peak common to each of the depths the samples were taken (between 5 to 100 m), although their identification remained elusive.

Work looking at the simultaneous speciation of halogens, in particular I and Br containing compounds in environmental samples, has measured these analytes in seaweed species from the Antarctic and snow samples from the Arctic. For the seaweed samples a range of fractionation steps were used to investigate halogen species present in brown and red macroalgae.⁶⁴ The sequential extraction approaches developed identified the distribution of halogens among fractions containing: lipids (extraction with petroleum ether); water-soluble species; proteins (dissolved with protease and SDS); carbohydrates (dissolved with driselase, with enzyme activities of cellulose, endo-1,3-β-glucanase, and xylanase); as well as the identification of the iodinated amino acids (MIT and DIT). The F content of the macroalgae investigated, was below the LOQ, and the high blank values for Cl did not allow for their determination in any of the fractions. Neither Br nor I was found in the lipid fraction primarily because of the low lipid content of Antarctic seaweeds. All seaweeds contained water-soluble Br and I species. The species present in each fraction were studied in detail by different HPLC techniques, including SEC and RP separations, coupled to both elemental and molecular MS. Some unidentified Br- and I-containing species were observed in the different extracts evaluated. Furthermore, 25 halogenated polyphenols were identified in the seaweeds using RP LC (150 × 2.0 mm id) coupled to an Orbitrap ES-MS instrument with structural identification using the included manufacturer's software. Of the 25 halogenated species, only four were already reported in the literature. The difference between the total element concentration in the samples and the sum of concentrations obtained for the different fractions was used to determine the extraction efficiency. For water-soluble and protein bound Br and I the recovery was between 10–60% for Br and 20–90% for I, and it was suggested that better methods for Br extraction were required. The column recovery for the different species was also assessed and shown to be acceptable for both halogens. A study on the speciation of inorganic I and Br in Arctic snow (Svalbard) has used IC separation coupled to SF-ICP-MS to deliver results in the pg g⁻¹ range, which is necessary due to the low concentrations of IO₃⁻ and BrO₃⁻ in such a remote location.⁶⁵ Bromine, mainly sourced from the ocean, plays a crucial role in polar regions due to its autocatalytic reactions that produce radical gas species, like BrO and other Br radicals, which are responsible for ozone depletion. Chromatographic separation was carried out using an AE column (250 × 2.0 mm) thermostated at 30 °C. A NaOH gradient produced by an eluent generator with a 0.25 mL min⁻¹ flow rate was used, with a suppressor unit which removed NaOH prior to the SF-ICP-MS. The samples (1.0 mL) were manually injected using a PEEK Rheodyne valve rather than an autosampler, which was acceptable as only 4 samples were analysed in the study. The reported method achieved LOD values (pg g⁻¹) ranging from 0.4 for I⁻, 0.8 for IO₃⁻, 4.0 for Br⁻, and 1.0 for BrO₃⁻ respectively, which were between 10- to 30-fold lower than similar studies and probably reflect the use of a SF instrument as the detector. The method also revealed further I- and Br-containing compounds in the samples, which remained unidentified.

A HPLC-ICP-MS method was developed for the simultaneous determination of four I and six As species in human urine.⁴⁸ The I species of interest included, IO₃⁻, 3-iodo-tyrosine, 3,5-diiodo-tyrosine, and I⁻ and the As species studied were, AB, As³⁺, DMA, AC, MMA, and As⁵⁺. The separation used an AE column and an [NH₄]₂CO₃–[NH₄]₂CO₃/NH₄NO₃ gradient at 1.0 mL min⁻¹. To validate the method accuracy, the As and I species in a commercially available urine reference material (Urine L-2 Seronorm) were measured and the sum of each species was shown to be in agreement with the certificate for total As of 261 ± 53 μg L⁻¹ and total I of 297 ± 60 μg L⁻¹. The LOQ values for the analytes ranged from 0.045 to 2.26 μg L⁻¹. At three spiked levels (10.0, 20.0, 50.0 μg L⁻¹), the average recoveries ranged from between 87.4 to 113%, and the short-term precision (RSD%) ranged from between 0.4 to 17.2%. The ratio of the sum of six As species to the total As measured by ICP-MS ranged from between 77.4 to 121.2%, and the ratio of the sum of the four I species to the total I ranged from between 70.7 to 114.7%, indicating a good agreement between these two methods for both As and I. Whilst analytically the method was shown to be accurate, precise and robust, it is unclear in what clinical or toxicological circumstances it would be applied, as I is usually measured for nutritional reasons, whilst As is of interest when there is concern that an individual has been exposed to this metal(loid).

4.8 Iron

Two papers report on the use of solid-state techniques for Fe speciation. One paper reports on Fe speciation (Fe^II/Fe_total ratio) in geological samples using WD-XRF.⁶⁶ The method is of analytical relevance as WD-XRF has been seldom used for determination of oxidation states. As WD-XRF does not require sample dissolution, pressed powder pellets were used, thus minimising sample manipulation and preventing oxidation of Fe^II to Fe^III. The FeKα_1,2 fluorescence line was employed to determine total Fe content, whereas FeLα_1,2 and FeLβ₁ fluorescence lines were selected for speciation as both lines are sensitive to the chemical state with a secondary dependence on the Fe^II/Fe_total ratio. Measurement of the intensity of L-lines was challenging due to their close proximity in the XRF spectrum. Fitting all the XRF spectrum to a Lorentzian function allowed authors to precisely determine the intensity of FeLα_1,2 and FeLβ₁ lines with an error as low as 0.002. With the aim of comparing results from samples with different iron content, peak intensities of FeLα_1,2 and FeLβ₁ were subsequently normalised by dividing their intensities by the total Fe content. Moreover, the methodology also entails the introduction of a chemical index factor to account for changes in fluorescence intensity due to the chemical environment of Fe. This index enabled calculation of the Fe^II/Fe_total ratio in the geological samples, regardless of their matrices. Results evidenced that only the normalised intensity of the FeLα_1,2 line showed a linear relationship with the Fe^II/Fe_total ratio and therefore it was selected for speciation of Fe and quantification of their oxidation states. Accuracy of the proposed method was determined by analysing total iron and FeO and Fe₂O₃ concentrations in geological CRMs of different nature (Fe ores, rocks, minerals) as well as in mixed CRMs, with only 6 out of 55 determinations showing a relative error more than 5% of the certified values. A study which focussed on Fe and P speciation in sediments from organic-rich lakes amended with Fe to reduce P content has been reported on.⁶⁷ Sediments were collected at different depths and submitted to sequential extraction and bulk and micro Fe K-edge and P K-edge XAS and micro-XRF spectroscopy analysis. The largest extracted Fe pool was Fe^II extracted with HCl accounted for 32% of the total Fe followed by ascorbic extractable Fe, with variations with depth. Regarding P, organic P was the dominant P phase over the entire depth accounting for up to 69% of total P. Additional information from XAS and micro-XRF measurements evidence the inhibition of the formation Fe^III(oxyhydr)oxides, responsible for P retention in sediment layers, for the presence of organic matter and Ca. Instead, nanoscale Ca–(Mn–)P–OM coprecipitates are produced. These coprecipitates can bind P but are highly redox sensitive and responsible for the rapid release of P under reducing conditions. Consequently, Fe remediation could have adverse effects when applied to organic-rich lakes as it can lead to P remobilisation under hypoxia conditions. There are also two reports of multielement speciation published this year: Co, Cu, Fe, Mn and Zn in cell culture media, and Cu, Fe Mn, Ni and Zn in bamboo and soil samples. Both reports are covered in Section 4.5, Copper.

4.9 Lead

Sowers et al. applied X-ray spectrometric methods to the speciation of Pb in paired house dust and soil samples collected from homes in the US built before 1978, when Pb-based paint was banned, with the primary goal of assessing the chemical factors driving Pb exposure from residential media.⁶⁸ They also measured total Pb, by ICP-MS, and determined in vitro bioaccessibility (a type of operationally defined speciation) by US EPA Method 1340. The results from bulk XAS as well as XFM and μ-XANES showed that Pb species in dust were hydrocerussite (such as hydroxycarbonate) phases that are commonly found in Pb-based paint. On the other hand, the Pb in soils was found to be predominantly associated with organic matter or iron. They found that Pb from house dust was >60% bioaccessible because of the high proportion of readily soluble species that are derived from paint, and the researchers stressed the need for speciation analysis in support of assessing the likely hazards of Pb exposure in the home.

Guo et al. developed several HPLC-ICP-MS methods for the simultaneous determination of Pb and Sn species in Antarctic krill and fish.⁶⁹ The analytes comprised the inorganic species together with two organolead compounds (trimethyl and triethyl) and four organotin compounds (trimethyl, triethyl, tributyl and triphenyl). The researchers investigated a number of possibly stationary phases, which they packed themselves, eventually devising a gradient cation-exchange procedure and an isocratic HILIC procedure. The latter was adopted for the analysis of the samples, as it was considered simpler with a predominantly aqueous mobile phase (4 mmol per L SDBS at pH 2.0), although the separation time of 16 min was slightly longer than that of the cation-exchange method. The LOD values ranged from 0.013 to 0.24 μg L⁻¹. Samples (4.0 g) were extracted by UAE with 10 mL of 1 mol per L HCl in 50% MeOH. After centrifuging and separating, the residue was re-extracted and the two extracts combined (for a total volume of 20 mL), the pH adjusted to 7 and filtered. The accuracy was evaluated by the analysis of a CRM (GBW10068 oyster tissue), which is certified for only total Pb and which did not contain measurable concentrations of the two organolead compounds. On the other hand, the CRM has certified or reference values for the MBT, DBT, TBT, and TPhT content. The value measured for TBT (2.60 ± 0.24 mg kg⁻¹) was significantly higher than the certified value (1.85 ± 0.34 mg kg⁻¹). The accuracy was also assessed by spiking both the CRM (at 0.5 mg kg⁻¹) and 24 krill and 8 fish samples at 0.1 mg kg⁻¹. The recoveries ranged from 89 to 98%. None of the species was detected in any of the samples. To improve the LOD, the same researchers devised a SPE preconcentation procedure that they applied prior to the HPLC analysis of ‘uncontaminated’ seawater.⁷⁰ Based on considerable prior experience of SPE (as evidenced by six publications), they evaluated ‘typical’ SPE adsorbents including C18, SAX, SCX, GO@SiO₂ as well as a HILIC stationary phase (Amphion II). They also evaluated a number of preconditioning reagents, including benzoic acid, Cys, SDBS, sodium 1-pentanesulfonate (SPS), sodium 2,3-dimercapto-1-propanesulfonate (DMPS), 2-mercaptoethanol (ME), 4-phenyl-3-aminothiourea (PATU) and tetra- butylammonium hydroxide (TBAH). The best performance was obtained with a column of GO@SiO2 (5 μm, 10 mm × 4.6 mm i.d.) that was preconditioned with 8 mL of 1 mmol per L benzoic acid. The procedure was carried out in a flow manifold consisting of two switching/injection valves that delivered the preconditioning solution, the sample (10 mL) and eluent 10 μL of the HPLC mobile phase (4 mol per L SDBS at pH 2.0). Enhancement factors (based on calibration slope ratios) ranged from 552 to 2848, leading to LODs of 2–8 pg L⁻¹. The procedure was applied to a CRM (National Institute of Metrology, Beijing, BWQ7001-2016; Cd, Cr, Cu, Pb and Zn in seawater) and to 13 real coastal seawaters. Only Pb^II was detected in the CRM, whereas in the seawaters both Pb^II and TBT were detected at concentrations of between 6–32 and 0.23–0.38 ng L⁻¹, respectively. Spike recoveries at concentrations of 5.0 ng L⁻¹ for Sn species and 50 ng L⁻¹ for Pb species ranged from 89 to 104%.

A magnetic SPE procedure for the simultaneous collection and preconcentration of Pb and Hg species prior to quantification by HPLC-ICP-MS was devised by Zhang et al.⁷¹ The extractant was a covalent organic framework (COF) to which thiocarbonyldihydrazide (TCH) was linked anchored to a magnetic Fe₃O₄ nanoparticulate core. The material was synthesised in a lengthy three-step procedure in which Fe₃O₄ and Fe₃O₄@NH₂ were first produced, then used in the synthesis of Fe₃O₄@2,4,6-trihydroxybenzene-1,3,5-tricarbaldehyde. Finally, Fe₃O₄@COF-TCH was prepared by a procedure in whose final step the reaction mixture was sealed in a Schlenk tube and heated at 120 °C for 3 days. The procedure was applied to the determination of Hg^II, MeHg, EtHg, Pb^II and trimethyllead in tap, surface, ground and sea waters. To a 200 mL sample (adjusted to pH 5–6 if needed) was added 1 mL of a 20 mg L⁻¹ adsorbent suspension. After mixing (15 min) and magnetic separation, the analytes were dissolved in 0.5 mL of 1% HCl (v/v) and 0.6% (m/v) thiourea. Following magnetic separation, 100 μL was injected into the HPLC system in which the analytes were separated on a C18 column (5 μm, 15.0 mm × 6.0 mm) with a mobile phase of 5% (m/v) MeOH, 1.2 g per L L-Cys and 60 mmol per L NH₄CH₃CO₂. Enrichment factors (based on concentration ratios) of between 360 and 389 were obtained leading to LOD values of between 0.08 (Pb^II) and 0.9 (Hg^II) ng L⁻¹. None of the analytes was found in the tap water; only Pb^II was found in the surface and ground waters, and Hg^II and Pb^II were found in the seawater. The accuracy was assessed by the recoveries of spikes at 5, 20 and 100 ng L⁻¹ to all four matrices, which ranged from 94 to 103%.

4.10 Manganese

There are two reports of multielement speciation published this year: Co, Cu, Fe, Mn and Zn in cell culture media and Cu, Fe, Mn, Ni and Zn in bamboo and soil samples. Both of these reports are covered in Section 4.5, Copper.

4.11 Mercury

Three papers report on fundamental aspects of Hg speciation this year. When measuring isotope ratios it is essential to accurately estimate the mass bias (the unequal transmission of different isotopes through the MS) factor for the system under study. This has been investigated in a comprehensive study by Suárez-Criado et al. for the measurement of Hg species-specific isotopic composition by GC-MC-ICP-MS.⁷² Separations were performed on a DB-5MS capillary column (30 m × 0.53 mm i.d. × 1.0 mm) with a Tl solution (90 ng L⁻¹) simultaneously nebulised into the Ar flow transporting the gaseous analytes from the GC into the MC-ICP-MS. The work compared the use of different mass bias correction models for achieving the best accuracy and precision of the Hg(II)-specific isotope ratios (IRs) and δ values and two different protocols for measuring mass bias: the standard-sample bracketing (SSB) approach, and an alternative bracketing procedure taken from a cited reference. Two NIST SRMS, the primary standard, NIST 3133, and the secondary standard, NIST 8610, were analysed using the various analytical procedures. Ethylation of Hg^II was with NaBEt₄ (2% w/v) followed by extraction into hexane.

The results obtained showed that δHg values are affected during chromatographic elution of Hg^II as this induces variations in the IR of Tl and, hence, in the mass bias-corrected Hg^II IR values. The point by point (PbP) method was found to be worst option for mass bias correction in terms of precision and accuracy. The internal precision of absolute Hg^II IRs was better when the range of points selected for a linear regression (LRS) approach is greater (321 instead of 27) regardless of the mass bias correction model used (Russell or Baxter). The evaluation of SSB and the Baxter bracketing procedure showed that both approaches provide similar accuracy and precision in δHg values with a slightly better accuracy and precision found when applying the LRS calculation in combination with the Russell model and the SSB approach. The paper contains a wealth of detail and is recommended reading for workers in this field. Non-chromatographic Hg speciation is often undertaken using thermal desorption-based instruments, which relies on the differing volatilities of the Hg compounds under study. This has been covered for four Hg species, HgCl, CH₃HgCl, HgS and HgSO₄, using a programmable thermal desorber in combination with ETAAS.⁷³ The Hg species, ranging from 0.1 to 1000 mg kg⁻¹, were mixed with Al₂O₃ to give the sample used for this work. The thermal desorption chamber was heated until a species began to be detected and the temperature was then held until the signal intensity fell back to the baseline signal followed by progression to the next desorption temperature. Desorption temperatures were found to be 165, 230, 295 and 355 °C and the LOD values were 0.005, 0.03, 0.003 and 0.02 for CH₃HgCl, HgCl, HgS and HgSO₄, respectively. Recovery of CH₃HgCl from ERM CE-464 (Tuna fish) was 104% and spiking into the same material gave a recovery of 99%. The linear ranges varied from 0.02–20 μg for CH₃HgCl to 0.1–50 μg HgSO₄. The method was then applied to primary and oxidised mine waste material, which were found to contain both CH₃HgCl and HgS and spike recoveries for the four Hg species in both materials ranged from 90 to 104%. Fang et al. report on the transformation of Hg^I species during standard preparation and analysis and discuss the implication for this on the analysis of sample of environmental origin.⁷⁴ It was shown that Hg^I, produced by dissolving Hg₂(NO₃)₂ and Hg₂SO₄ in water, was stable at 1000 mg L⁻¹, yet completely disproportionated when in a diluted solution (10 μg L⁻¹) after analysis by HPLC-ICP-MS (C18 column, 5 μm, 50 mm × 4.6 mm with a mobile phase of 0.5% (v/v) 2-mercaptoethanol with 5 μg per L Bi as an internal standard) flowing at 1.6 mL min⁻¹ for 9.5 minutes. In this analysis a peak at 200 s was identified as Hg⁰ based on comparison with the signal intensity of the for Hg²⁺, noting that Hg⁰ is preferentially transported to the plasma as a vapour. Isotopically enriched Hg species (¹⁹⁹Hg⁰ and ²⁰²Hg^II) were used to monitor for species transformations using both HPLC-ICP-MS and purge-trap-ICP-MS detectors. It was found that the use of polypropylene containers, increasing headspace, decreasing pH, and increasing dissolved oxygen significantly enhanced the disproportionation or redox transformations of Hg^I. Thus, using a glass container without headspace and maintaining a slightly alkaline solution are recommended for stabilisation of dilute Hg^I solutions. Analysis of water samples from a creek flowing through an abandoned Hg smelting plant showed of Hg^I (4.4–6.1 μg L⁻¹), accounting for 54–70% of total dissolved Hg. The reductive formation of Hg^I in Hg^II spiked environmental water samples, was also observed. The paper and associated supplementary information contain a wealth of detail in a complex read.

Despite the plethora of chromatographic separations for Hg species reported in the literature, work continues on the optimisation of these types of procedure. For HPLC, separations are usually performed with ion pair RP or CEC approaches so it is interesting to see that a report on a L-cysteine-mediated separation of Hg²⁺ and MeHg⁺ by AE HPLC-ICP-OES has appeared this year.⁷⁵ A Hamilton PRP-X100 column, commonly used for As speciation studies, was used with a mobile phase of mixtures of 100 mmol per L solutions of Na₂HPO₄ and NaH₂PO₄, to give the desired pH ranges (5–8), and 10 mmol per L L-cysteine flowing at 1 mL min⁻¹. Also included in the mobile phase was MeOH in the range of 0–20% and 10 μg of each Hg species was injected onto the column with the Hg signal monitored at 253.652 nm. The optimal mobile phase comprised 15% MeOH in 100 mmol per L phosphate buffer with 10 mmol per L L-cysteine at pH 5.0 with a run time of 700 s. This produced a reasonably sharp Hg²⁺ peak, albeit with a long tail, but a very broad MeHg signal of about 350 s in duration. Some of this broadening could be due to the size or poor wash-out characteristic of the spray chamber used, but even so this broadening is much worse than for separations using RP chromatography. The paper does contain a good discussion of the chromatographic mechanisms involved though. For GC separations, Hg species are usually derivatised by ethylation, which has been shown to induce artefact formation, or propylation. This year sees a report on phenylation as the derivatisation method for GC-Pyro-AFS.⁷⁶ Phenylation was investigated as the more usual reagents for this purpose are now no longer available in Chile. The derivatisation of MeHg⁺ was evaluated using two reagents: (NaBPh₄) and (PhB(OH)₂) in a buffer of NaOH, and CH₃COOH with a total concentration of 0.4 mol L⁻¹ at a pH of 4.5 followed by extraction into C₂H₄Cl₂ after Hg species extraction using MAE which was optimised using a Plackett–Burman design followed by a Box–Behnken design. The variables evaluated during this screening phase included the type of organic solvent, buffer pH, buffer concentration, amount of derivatisation reagent, agitation time and speed, and LPME temperature. Both experimental designs were carried out with 18 experiments using synthetic MeHg samples. After this optimisation the following figures of merit were obtained: LOD and LOQ values of 0.0045 and 0.113 mg L⁻¹, respectively, absolute LOD of 0.4 pg MeHg with a preconcentration factor of 35 ± 2 and an extraction level of 100 ± 6%. The authors state that the results indicate that the sensitivity of the method is comparable to that reported in the literature, but with notable advantages in terms of speed and cost-effectiveness. The method was validated using the CRM ERM CE-464 (tuna fish) for which the found value of 5.7 ± 0.3 mg MeHg⁺ per kg compared with the certified MeHg⁺ mass fraction of 5.5 ± 0.17 mg MeHg per kg. The method was subsequently applied to determine the MeHg content in mice brain tissues, from control and exposed animals, which ranged from 0.1 to 50 mg kg⁻¹ as MeHg. A study into the optimisation and validation of a procedure for Hg speciation analysis in fish, rice and soil using different extractants, followed by HPLC ICP-MS detection has been reported.⁷⁷ Separations were on a C-18 column (4.6 × 150 mm, 5 μm) with a mobile phase of 0.06 mol per L NH₄CH₃CO₂, 0.1% 2-mercaptoethanol and 5% methanol at pH 6.8 with a flow rate of 1 mL min⁻¹ with detection by ICP-MS. After optimisation, assessing HCl and KOH at varying concentrations and MAE irradiation durations, MAE with 25% KOH was found to be the most suitable in terms of accuracy. A LOQ value of 12 μg kg⁻¹ was obtained and spike recoveries into a variety of matrices, fish, rice and soil gave recoveries described as acceptable in the range of 75.3–98.1% and RSD values for MeHg of 3.76–5.78. A central composite design, with five levels and four factors, has been used for the optimisation of the extraction and HPLC separation of Hg species in cooked fishery products.⁷⁸ Optimal separations were performed on a RP column (2.7 μm, 2.1 × 150 mm) and a mobile phase containing 2-mercaptoethanol at 0.25% (v/v) and methanol at 1% (v/v) with detection by ICP-MS.

The LOQ values were 2.5 μg per kg (wet weight, w/w) for MeHg and 1.2 μg per kg (wet weight, w/w) for Hg²⁺. The intermediate reproducibility in terms of coefficient of variation (CVR) was <6% and the bias (%) obtained for the analysis of four CRMs, TORT-3 (lobster hepatopancreas), SRM 1566-b (oyster tissue), SQID-1 (cuttlefish) and NMIJ CRM 7402-a (cod fish tissue) was <7%. The MeHg and Hg²⁺ content of the cooked fish products analysed ranged from 3.14 to 191 and <LOQ-43, 4 μg kg⁻¹, respectively. Finally in this section, a magnetic SPE procedure for the simultaneous collection and preconcentration of Hg and Pb species prior to quantification by HPLC-ICP-MS has been developed by Zhang et al.⁷¹ The extractant was a covalent organic framework (COF) to which thiocarbonyldihydrazide (TCH) was linked anchored to a magnetic Fe₃O₄ NP core. The material was synthesised in a lengthy three-step procedure in which Fe₃O₄ and Fe₃O₄@NH₂ were first synthesised, then used in the synthesis of Fe₃O₄@2,4,6-trihydroxybenzene-1,3,5-tricarbaldehyde. Finally, Fe₃O₄@COF-TCH was prepared by a procedure in whose final step the reaction mixture was sealed in a Schlenk tube and heated at 120 °C for 3 days. The procedure was applied to the determination of Hg^II, MeHg, EtHg, Pb^II and trimethyllead in tap, surface, ground and sea waters. One mL of the 20 mg L⁻¹ adsorbent suspension was added to a 200 mL sample (adjusted to pH 5–6 if needed). After mixing (15 min) and magnetic separation, the analytes were dissolved in 0.5 mL of 1% HCl (v/v) and 0.6% (m/v) thiourea and the extracted species were separated on a C18 column (5 μm, 15.0 mm × 6.0 mm) with a mobile phase of 5% (m/v) MeOH, 1.2 g per L L-Cys and 60 mmol per L ammonium acetate. Enrichment factors (based on concentration ratios) of between 360 and 389 were obtained leading to LOD values of between 0.08 (Pb^II) and 0.9 (Hg^II) ng L⁻¹. None of the analytes was found in the tap water; only Pb^II was found in the surface and ground waters, and Hg^II and Pb^II were found in the seawater. Method validation was assessed by the spike recoveries, at 5, 20 and 100 ng L⁻¹ to all four matrices, which ranged from 94 to 103%, and the linear range was 0.26–1000 ng L⁻¹ with reported RSD values of <4.5%.

Table 1 shows examples of other applications of Hg speciation presented in the literature during the time period covered by this ASU.

Table 1 Applications of speciation analysis: Hg

Analyte species	Technique	Matrix	Sample treatment	Separation	LOD	Validation	Reference
MeHg, Hg	ID-GC-ICP-MS	Sediment, waters	Sediments: extraction with KBr/H2SO4/CuSO4^/ChCl2	Aqueous ethylation	0.01 ng L^−1	Recoveries: ECM-CC580 sediment, 99.0 ± 0.35% DORM-4 fish muscle, 98.0 ± 0.67%	79
			Waters: distillation
MeHg	HPLC-ICP-MS, GC-IPC-MS	Binding of MeHg-cysteine with DOM	Extraction with bromide and thiourea, GC: distillation and KOH/CH3OH digestion	HPLC: SEC and a biphenyl column. MP 2 mmol per L MSA (pH 5), 100 μg per L Au, MeOH:ACN 90 : 7.5 : 2.5 (v/v/v)	MeHg: 5 pmol L^−1, MeHG-Cys 5.5 nmol L^−1	DORM-4 fish muscle 90–1105 recovery	80
						MeHG spike recovery from cell supernatant 80–94%
MeHg, iHg	Frontal chromatography ICP-MS	Hair	25% TMAH, MAE, 75 W, 4.5 min. Aqueous ethylation	Short AE column, AmberChrom® 1 × 2 chloride form 200–400 mesh	250 mg sample: 0.22 μg kg^−1	NMID-01 human hair	81
					10 mg sample: 5.5 μg kg^−1	MeHg 107% recovery
iHg, MeHg	HPLC-CV-AFS	Blood, urine	Aliquots of blood or urine vortex-mixed with HNO3 or L-Cys solution, respectively	HPLC: ODS-SP (4.6 mm × 250 mm, 5 μm). Post column: reduction KBH4–KOH, oxidation K2S2O8–KOH	Hg(II), MeHg, EtHg and PhHg 1.5, 1.6, 1.7, 2.0 ng mL^−1, respectively	Not found	82

---

4.12 Nickel

There are two reports of multielement speciation published this year: Co, Cu, Fe, Mn and Zn in cell culture media and Cu, Fe, Mn, Ni and Zn in bamboo and soil samples. Both of these reports are covered in Section 4.5, Copper.

4.13 Phosphorus

The environmental issues surrounding the impact of high concentrations of phosphate leaching into surface and ground water from soils under different agricultural or forestry management regimes has driven a recent interest in the speciation of P in different soil and sediment types. The methods developed in this context have exclusively involved speciation in the solid phase, using methods based on X-ray spectroscopy. Szerlag et al. have studied the long-term use of poultry manure application to agricultural Mid-Atlantic soils and the build-up of P, contributing to significant eutrophication problems in Chesapeake Bay due to P-containing leachate.⁸³ Usually, bulk P K-edge XANES followed by linear combination fitting (LCF) can be utilised to determine the solid P phases in soil. However, this approach limits the results to only a few major soil phases. Additionally, XANES spectra for different P species can have very similar features, leading to an over- or underestimate of their contribution to LCF. In this study, P speciation was improved by pairing multimodal microbeam XRF mapping with microbeam XANES analysis, to directly speciate major and minor P phases on the micron scale. By combining elemental maps of both tender (Al, P, S and Si) and hard energy (Ca, Fe and Mn) elements, to evaluate the elemental co-locations with P, it was possible to better predict P dissolution and mobility. In a similar study investigating the impact of fertiliser and manure on Mid-Atlantic soils, a combination of chemical extraction approaches, including Mehlich-3, water extractable P, and chemical fractionation, along with non-destructive methods such as XANES and XRF, were used to understand P dynamics in eight P-enriched soils with various management histories.⁸⁴ The Mehlich-3 test uses an extractant containing CH₃COOH, NH₄NO₃, NH₄F, HNO₃ and EDTA, and is widely used commercially in the USA and Canada to assess agricultural soils. It is popular because it can be used on different soil types and is effective for a wide range of elements of interest, including P. Chemical fractionation and XRF data were used to support XANES LCFs, allowing for the identification of various Al, Ca, and Fe phosphates and P-sorbed phases in soils amended with fertiliser, poultry litter, or dairy manure. These studies provide insights on how different management scenarios can impact soil P dynamics. The microscale heterogeneity of P species associated with secondary mineral phases in the B horizons of two boreal Podzols has been studied using SR X-ray microscopic techniques.⁸⁵ Poorly crystalline Al and Fe minerals have a high sorption capacity for PO₄²⁻ anions. However, there is still uncertainty about the molecular level distribution of P species between Al and Fe mineral phases, which can be relevant to their bioavailability in this soil type. The study aim was to distinguish between Al- and Fe-associated P phases in two specific Podzol B horizons (Flakaliden and Skogaby). The P species were clustered into major groups depending on the main element association (Al–P, Fe–P, Ca–P and organic–P) according to the colour of P-containing spots in the tricolour μXRF quantitative maps that combined P, Al, and Fe fluorescence intensities. The results were analysed using quantitative LCF of the P K-edge μXANES spectra. In cases where the results diverged, a careful visual comparison between the diagnostic features of the XANES spectra of the samples and the reference standards was made. In conclusion, Al–P, particularly imogolite-type materials (ITM, an aluminosilicate nanotube)-adsorbed PO₄²⁻, is the main P phase in the B horizons of the studied Podzols, which may have implications for bioavailability as ITM is more easily dissolved than Fe oxides. In a similar study on the complexity of eutrophication, but in this case focusing on the effects on lake chemistry, the mobilisation of legacy P from the sediments in shallow lakes was investigated to determine if this was the cause of persistent eutrophication after reduction of external P input into the lake.⁶⁷ By combining sequential extractions with bulk and micro X-ray spectroscopy it was possible to elucidate Fe and P speciation in sediments of an iron-treated peat lake. In this complementary study, bulk and micro Fe K-edge and P K-edge XAS and micro-focused XRF spectroscopy were applied to characterise the P hosting Fe^III pool. Combined with sequential extraction data, the SR X-ray analyses revealed that a continuum of co-precipitates of Fe^III with Ca, Mn, PO₄²⁻ and organic carbon within the organic matter matrix constitutes the reducible Fe^III pool. The complementary analyses also shed new light on the interpretation of sequential extraction results, demonstrating that pyrite was not quantitatively extracted by HNO₃ and that most of the Fe^II extracted by HCl originated from phyllosilicate minerals. Formation of an amorphous inorganic–organic co-precipitate upon Fe addition constituted an effective P sink in the studied peaty sediments. However, the high intrinsic reactivity of this nanoscale co-precipitate and its fine distribution in the organic matter matrix makes it very susceptible to reductive dissolution, leading to P remobilisation under reducing conditions.

In comparison to solid-state speciation approaches, methods based on HPLC-ICP-MS/MS have also recently been developed for the speciation of naturally occurring P-containing organic compounds in soil samples and oil derived from vegetable and animal sources. The analysis of myo-, scyllo-, neo- and chiro-inositol hexakisphosphate isomers in soil by HPLC-ICP-MS/MS shows improvement over the conventional ³¹P NMR approach.⁸⁶ The method offered improved sensitivity and fine-scale mapping compared to the more complex ³¹P NMR spectroscopy, which requires deconvolution of overlapping resonances and their isolation from underlying features. The developed method used extraction of the P-containing inositol species from forest soils (0.5 g), fertilised with N for many years, using a mixture of NaOH (0.25 mol L⁻¹) and EDTA (0.05 mol L⁻¹) for 16 hours. The HPLC separation was performed on a high-resolution SAX column, which was originally developed for enhanced chromatography of oligosaccharides. It consists of a pellicular AE resin, offering a high separation efficiency. The column was eluted with a mobile phase comprising HCL, but no further details of the eluent composition or gradient were given. The analytes were calibrated using inorganic P standards and detection at ³¹P⁺. The influence of the eluent on the detector response at m/z 47 for PO⁺ and at m/z 31 for P⁺ was tested using a matrix of KH₂PO₄ and compared to HCl delivered at 0.4 mL min⁻¹. The response for KH₂PO₄ differed by ≤7% at the concentrations and flow rates used. The effect of gradient elution on the detector response is an important consideration to make when development of the most accurate speciation methods is the goal and would seem to be particularly important with elements such as P, which are poorly ionised in an ICP. Another method based on HPLC-ICP-MS has been developed for the quantification of phospholipids in a wide range of different oil samples, including vegetable oils, animal fats, and phospholipid supplements.⁸⁷ It used a HILIC separation coupled to SF-ICP-MS in medium resolution mode (R = 4000) which allowed for the detection of ³¹P, while effectively overcoming spectral interferences from ¹⁵N¹⁶O⁺, and ¹⁴N¹⁶O¹H⁺. The separation used an HILIC column (130 Å, 3.5 μm, 4.6 mm i.d. × 100 mm length) with a mobile phase gradient combining aqueous ammonium acetate buffer (10 mmol L⁻¹, 0.1% formic acid, and 5% THF) (A) with THF (B). The separation and detector were coupled using a custom 1 : 10 splitter to reduce the flow rate and maintain plasma stability but still required addition of O₂ to reduce carbon deposition. The interface also used a modified DS-5 microflow total consumption nebuliser connected to a custom-made spray chamber, heated to 60 °C, which preheated the solvent before its introduction to the plasma. The direct analysis of phospholipids in this way eliminated the need for sample pretreatment for lipid fractionation and ensured full solubility of the samples. As with the previous publication, to improve quantification accuracy, the gradient effect on signal sensitivity was assessed and corrected using an experimentally measured time-dependent correction factor. The LOD was reported to be 15 pmol L⁻¹ as P for a 5 μL injection, which is comparable to other methods for phospholipid analysis. The developed methodology enabled direct lipid sample quantitation without the need for sample clean up using SPE. The reduction in extensive sample preparation not only saved time but also has environmental and economic advantages.

4.14 Platinum

Two very different and novel reports on Pt speciation have recently been published, one using the established HPLC-ICP-MS analytical work-flow for speciation of Pt-containing drugs in liquid samples and the second using X-ray methods for solid-state speciation to speciate Pt in historical metal alloy samples. The speciation of Pt-containing anti-cancer drugs in clinical samples usually focuses on the measurement of cis-platin and its metabolites, so the reporting of a method for the measurement of the third generation Pt-containing drug oxaliplatin and its metabolites in blood plasma samples by UHPLC-ICP-MS, is an important development.⁸⁸ The samples were processed immediately after collection from patients to preserve oxaliplatin speciation by using methanol-deproteinisation, followed by storage of the diluted supernatants (plasma : methanol 1 : 2 v/v) at −80 °C. The developed UHPLC separation of the intact oxaliplatin used carboplatin (a second generation chemotherapy agent) as an IS and a C18 micro-column (SB-C18, 2.1 × 50 mm, 2.7 μm particle size) with linear elution gradient (mobile phase A: water–methanol (97 : 3 v/v), 0.075 mmol per L sodium dodecyl sulfate, 9.79 nmol per L thallium, pH 2.5 with trifluoromethanesulfonic acid; B: 100% methanol (v/v)) with ICP-MS detection to monitor Pt and Tl at m/z 195 and 205, respectively. The LOQ for Pt was 50 nmol L⁻¹ in methanol-deproteinised diluted plasma (1 : 2 v/v). The intra-day and inter-day accuracy ranged from 96.8 to 103% of nominal concentration and the precision from 0.62 to 2.49% (% CV). Recovery was complete for spiked samples and a matrix effect confirmed the requirement for matrix-matched standards to be used for best accuracy to be achieved. The method was applied to determining the plasma concentrations of intact oxaliplatin in patients undergoing cancer chemotherapy, and studies of oxaliplatin degradation in vitro. In a very specialist and unusual solid-state speciation application, high-energy resolution fluorescence detection XRF (HERFD-XRF) imaging and HERFD-XANES were used to quantify and characterise trace Pt, including speciation, in Au solidi from the Late Roman and Byzantine Empires.⁸⁹ Historically, the elemental analysis of coins has been pivotal in distinguishing authentic artifacts from forgeries, elucidating minting practices, and understanding economic shifts. The work demonstrates the effectiveness of HERFD techniques for determining the spatial resolution of minor components (Pt) in the presence of the main constituent (Au), but also the speciation of the minor constituent metals present, in this case Pt and its oxidation state, using XANES. Three Au solidi, minted between 654 and 659 CE, were analysed alongside reference Au materials with known Pt concentrations. The HERFD-XRF imaging revealed spatial distributions of Pt, highlighting non-uniformities within the coins. Additionally, HERFD-XANES spectroscopy identified the oxidation states and chemical speciation of Pt. The results demonstrate that Pt in the solidi primarily exists as metallic Pt, with some surface oxidation containing Pt oxide species.

4.15 REEs

With the current increase in demand for REEs in the high technology manufacturing and clean-energy sectors, interest in their availability and potential recovery from solid waste has become a focus for commercial companies. This requires the development and testing of methods for the determination of REE speciation in municipal solid waste incineration ash (MSWIA) to be developed, so that suitable extraction and recovery processes can be undertaken.⁹⁰ A recent study in this area used SR X-ray spectroscopy and microscopy techniques to elucidate the speciation of representative REE (Y, Ce, and Nd) in different MSWIA samples. Linear combination fitting of bulk XANES data indicated that Y-bearing Al/Fe oxides and phosphates are the primary Y-hosting phases. Analysis by μXANES of individual Y-containing particles identified by μXRF mapping revealed notably different Y speciation at micro-scale from the bulk, consistent with the highly heterogeneous nature of MSWIA samples. The main REE-bearing phases in different size-fractionated MSWIA were however quite similar: Y and Nd as oxides and xenotime/monazite, and Ce as apatite and monazite. The results from this work provide important insights for designing pre-screening processes (e.g., density separation) and optimising extraction methods (e.g., pH, use of ligands) for cost-effective REE recovery from MSWIA.

4.16 Selenium

The characterisation of Se-enriched yeast in terms of Se speciation has attracted the interest of many researchers and it is still of interest. Angaïts et al. have evaluated different Se extraction protocols to verify the presence of iSe in Se-yeast and to identify sources of Se losses during Se-speciation.⁹¹ For this purpose, selenised yeast (SELM-1) was submitted to different sample treatments, sequentially applied: (1) aqueous extraction to obtain the corresponding selenometabolome, (2) enzymatic hydrolysis with protease of the pellet obtained in step 1, (3) carbamidomethylation to derivatise cysteine and SeCys followed by a protease digestion of the pellet obtained in step 1, and finally (4) total Se using SDS extraction followed by an acidic digestion procedure. The total Se and Se species content in each step was quantified using ICP-MS and HPLC-ICP-MS, respectively. The Se species separation was performed on a C8 column (4.6 × 250 mm, 5 μm, 120 A) with a mobile phase composed of A: 0.2% formic acid and B: 0.2% formic acid in MeOH, under gradient elution. The study demonstrated the presence of about 10% of inorganic Se and a serious risk of losses of SeCys during derivatisation and proteolysis. Based on the information obtained, the authors recommend a protocol based on HPLC-ICP-MS analysis of the yeast extract obtained after simultaneous derivatisation and proteolytic extraction of the yeast once the selenometabolome has been extracted. This protocol allowed authors to quantify SeMet (as the sum of SeMet, SeMetO, and derivatised SeMet), SeCys (as carboxymethylated SeCys) and iSe (as SeCAM₂). The importance of performing a mass balance to verify the effectiveness of each extraction step is highlighted by the authors throughout the manuscript. The following Se values for SELM-1 were obtained: Se-metabolome fraction: 14.8 ± 0.7%, SeMet: 66.2 ± 2.7%, total SeCys 12.5 ± 1.5% and iSe 9.7 ± 1.7%, accounting for >99.8% of total Se. A second paper on this topic focused on the characterisation of two candidate CRMs (not yet named) for SeMet (a Se-enriched yeast and a Se supplement).⁹² Both HPLC-ICP-MS and HPLC-ES-MS/MS measurements were applied. The paper presents a detailed evaluation of different sample treatments and chromatographic conditions. Overnight enzymatic hydrolysis with protease and 0.1 mol per L Tris–HCl was employed to isolate Se compounds from yeast and supplement when performing HPLC-ICP-MS measurements while UAE enzymatic hydrolysis with protease but without Tris–HCl was selected when using HPLC-ES-MS/MS. The optimal chromatographic conditions for HPLC-ICP-MS were a C18 column with a mobile phase consisting of 10 mmol per L 1-butanesulfonic acid, 8 mmol per L TMAH, 5 mmol per L malonic acid and 5% MeOH at a pH of 4.0, whereas for the ES-HPLC-MS/MS system a C18 column (3.0 × 100 mm, 2.7 μm) with an isocratic mobile phase of 2 mmol per L ammonium formate and 5% methanol was used. Both methodologies provided consistent results for SeMet in yeast with values of 718 mg SeMet per kg and 715 mg SeMet per kg, for HPLC-ICP-MS and HPLC-ES-MS/MS, respectively, with a final certified value of 715 mg SeMet per kg ± 36 mg SeMet per kg. For the Se-supplement, the SeMet content was 59 ± 5 mg kg⁻¹. The homogeneity and stability under storage and transportation conditions of the CRMs were also monitored and their associated uncertainties were taken into consideration in total uncertainties calculations.

The evaluation of different Se dietary sources for preparing Se enriched food (of animal or vegetable origin) has been the subject of two papers. Selenium nano/microparticles (SeNPs) were evaluated as additive for feeding O. macrolepis (Onychostoma macrolepis) and grass carp fish species.⁹³ The SeNPs were biogenically synthesised by culturing Bacillus subtillis in presence of 8 mmol per L Se^IV and subsequently characterised by electron microscopy. The SEM micrographs showed the presence of SeNP agglomerates with average sizes greater than 100 nm (570 nm). Fish were supplemented with SeNPs, Se^IV and SeMet at three concentration levels (0.1, 0.3 and 0.9 mg kg⁻¹) for 60 days. At the experiment's conclusion the fish were sacrificed and the serum and fish fillet samples collected. The results obtained revealed the SeNPs as the most suitable Se additive as they provided the strongest antioxidant capacity in fish serum and the best flesh quality by reducing the concentration of As, Cd and Hg and increasing PUFAs levels. Se speciation analysis of fish fillets was performed by HPLC-ICP-MS. Samples were enzymatically hydrolysed using a mixture of protease and trypsin in 100 mmol per L Tris–HCl. The SeNP supplementation significantly increased the levels of MeSeCys, SeMet and SeCys₂. Unfortunately, the paper lacks information on the chromatographic conditions employed and the protocol used for identifying the Se compounds in the resulting chromatograms. In a different paper, both ICP-MS and HPLC-ICP-MS were applied to determine the total content of Se, Se species distribution and Se in vitro bioaccessibility in organs and tissues of piglets fed with 3 Se supplements: Se^IV, Se-enriched yeast and Cardamine hirsuta.⁹⁴ The results revealed kidney, liver and spleen as the organs with the highest concentration of Se; however, muscle (peroneal longus and longissimus dorsi) exhibited the highest incremental increase of Se after supplementation. Speciation of Se was performed in liver and longissimus dorsi muscle by HPLC-ICP-MS after enzymatic hydrolysis with a mixture of protease, pancreatin and lipase. A C18 column (250 mm × 4.6 mm, 5 μm) and a mobile phase composed of 10 mmol per L citric acid with 5 mmol per L sodium 1-hexane sulfonate and 1% methanol was employed for Se species separation. Two species, SeMet and SeCys, were predominant in longissimus dorsi muscle and liver, respectively. Unfortunately, no information on the protocol used to stabilise SeCys is given. The method was validated by using CRMs for total Se and by spiking experiments for Se speciation with average recoveries ranging from 91% to 109%. Results from Se in vitro bioaccesibility in liver evidenced variations among different feeding groups with values ranging from 60 to 89%, and with SeMet being the most bioaccessible Se species.

Moving on to Se speciation in Se-enriched plants, Se species have been analysed in rapeseed (Brassica napus L.) seedlings and flowering stalks obtained from hydroponic cultures containing iSe.⁹⁵ To achieve the best analytical conditions, different types of proteases (protease XIV, protease E, protease K, trypsin and alcalase) and chromatographic conditions (an AE column and a C18 RP column) were considered. Protease XIV under UAE gave the highest hydrolysis efficiency. The use of 10 mmol per L Tris–HCl, buffer commonly used with proteases, provided quantitative average recovery for all Se-species except for SeCys₂ with recovery values (7%) lower than that with ultrapure water (74%). The authors justified the observed results on the conversion of SeCys₂ to SeMet. All five Se compounds (SeCys₂, MeSeCys, SeMet, Se^IV, and Se^VI) were baseline resolved under the two different chromatographic modes and a quantitative mass balance of around 99% was attained. OrganoSe compounds were the dominant species on both seedling and flowering stalks. Of special relevance is the high concentration of MeSeCys found, with a content as high as 5 mg kg⁻¹. Another paper describes the use of natural deep eutectic solvents (NADES) for iSe speciation in nuts.⁹⁶ The iSe in defatted and dried nuts was extracted by using 0.2 mol per L H₂SO₄ under UAE. The Se^IV in the resulting acidic extract was selectively complexed with APDC and extracted with 50 mL of NADES (thymol : decanoic acid, 1 : 2) followed by vortexing for 5 minutes. The Se^IV-containing NADES phase was solidified in an ice-bath for 10 minutes and separated, with Se^VI remaining in solution. The Se^VI and total Se concentrations were measured by ICP-MS and the Se^IV content was calculated by difference. The complexity of the NADES phase (high viscosity, organic content) made direct determination of Se^IV by ICP-MS not feasible. The accuracy of the proposed method was evaluated by spiking known standards of both Se^IV and Se^VI in different types of nut samples (almond, hazel nut, peanut, walnut, cashew nut, pecan nut and Brazil nut) with recoveries ranging from 97 to 105%. Furthermore, the AGREE (Analytical GREEnness Metric Approach) Software was applied to compare the greenness of the method with those of other existing methods. The proposed method achieved a greenness score of 0.55 which is comparable or even better than previously reported scores.

Concerning the biological role of Se, variations in Se levels in cerebrospinal (CFS) fluid from 10 patients treated with tofersen, developed for the treatment of Amyotrophic Lateral Sclerosis (ALS) were monitored by HPLC-ICP-DRC-MS.⁹⁷ For each patient, two samples of CFS were collected (one right before starting tofersen treatment and the second after 6 months of the tofersen treatment). Selenium species in CFS samples were separated by using an AE column under gradient elution with a mobile phase composed of eluent A: 10 mmol per L Tris–HCl 5% MeOH (pH = 8) and eluent B: 50 mmol per L Na₂CO₃, 20 mmol L⁻¹, 5% MeOH (pH = 8). The following species were determined: Se^IV, Se^VI, SeMet, SeCys₂, thioredoxin reductase-bound Se (Se-TxNRD), glutathione peroxidase-bound Se (Se-GPX), selenoprotein P (SelP) and human serum albumin-bound Se (Se-HAS). The paper contains a detailed description of the procedure used to synthesise SelP and Se-HAS. The Se-HAS was prepared by mixing selenite and HAS followed by incubation for 14 days whilst SelP was isolated from human serum, purified with a Heparine affinity column and preconcentrated by freeze-drying. After tofersen treatment, an important increase in all Se species in CFS was detected, specifically in Se-TxNRD, Se-HAS and in the three major neurotoxic Se species: SeMet, Se^IV and Se^VI. These findings showed the influence of tofersen treatment on the distribution of Se species in the central nervous system (CNS). Although the amount of data from which to draw conclusions is limited (only 10 patients are involved in the study), the information obtained is a starting point to get a deeper knowledge on the role of Se in the CNS. The authors hypothesise that tofersen treatment may induce changes in the brain and blood barrier or in the redox status of neuronal cells thus producing an upregulation of Se-containing antioxidant enzymes.

Several papers focus on Se species determination in environmental samples. A comprehensive report covers the evaluation of factors affecting Se atmospheric deposition by integrating Se and S speciation analysis and atmospheric dynamic modelling.⁹⁸ Samples were collected at the high-altitude Pic du Midi Observatory over 5 years (2015–2020). The Se speciation in aerosol and wet deposition samples (cloud water and precipitation) was conducted by HPLC-ICP-MS/MS, using an AE column (2.1 × 50 mm) under gradient elution with ammonium citrate (5–13 mmol L⁻¹, 2% MeOH at pH 5.2), whereas S speciation was undertaken using a C18 column (100 × 4.6 mm, 5 mm) with formic acid (24 to 240 mmol L⁻¹) as the mobile phase. The main Se species in all samples tested was Se^IV followed by Se^VI, with organic Se in lower amounts. For S, sulphate, methanesulfonic acid, dimethylsulfone and hydromethanesulphonate were identified. The speciation results combined with weather dynamics allowed authors to identify Se sources. No clear seasonal differences were observed on proportions of Se^IV and Se^VI, but the proportion of Se^IV was found to be significantly lower in aerosol samples collected in summer (7.7%) than in autumn (11.9%). The sources of inorganic Se were linked to continental anthropogenic sources. The iSe levels were highly affected by atmospheric processes, such as oxidation and pH changes, whilst Se^IV seems to be linked with local sources and the presence of Se^VI is more related to the transport of Se over longer distances. Organic Se was first detected in wet deposition samples and it is postulated by the authors as a potential biomarker of biogenic Se of marine origin. In an interesting and detailed paper, five Se species (SeNPs, Se^IV, Se^VI, SeMet and CuSe) were determined in Se-spiked soils using a methodology consisting of one-step extraction followed by separation of SeNPs by a nylon membrane (0.45 μm) and analysis of Se species by using HPLC-ICP-MS.⁹⁹ Several extracting solutions were tested: ultrapure water, 5 mmol per L TSPP, 2.5 mmol per L TMAH, 10 mmol per L KNO₃ and 10 mmol per L NaNO₃. The highest extraction efficiency (87%) was obtained at 5 mmol per L TSPP under shaking for 1 hour followed by UAE for 30 minutes. Separation of the extracting solution and soil matrix phases was accomplished by sedimentation as this allowed SeNPs to remain in the upper later of the extraction solution. The main novelty of the paper is the separation of SeNPs from the other Se-species by membrane filtration. More than 91% of SeNPs (different coatings were tested) were retained on the 0.45 μm nylon membrane while the adsorption of other extractable Se species was below 10%. Even in those cases where SeNP sizes were smaller than the membrane pore size the adsorption efficiencies were kept higher than 90%. Once SeNPs were removed, the Se species in the filtrate were separated and quantified by HPLC-ICP-MS by means of using an AE (250 mm × 4.1 mm, 10 μm) chromatographic column and 40 mmol per L NH₄Ac as the mobile phase flowing at 1 mL min⁻¹. The metal selenide content was further calculated as the difference between Se extractable content (SeNPs, Se^IV, Se^VI, and organic Se) and total Se. The analytical approach gave LOD values of 0.05 mg kg⁻¹ for Se^IV and Se^VI, 0.25 mg kg⁻¹ for SeMet and 0.02 mg kg⁻¹ for SeNPs. The accuracy of the proposed method was evaluated in Se-enriched soils with spike (1 mg kg⁻¹) recoveries of Se species of 80% or greater. The method was subsequently implemented in the determination of Se-species in five soil samples from a Se-enriched area in China. Metal selenide was the major Se-species found in all soil samples tested. Interestingly, SeNPs were detected in one soil sample at a concentration of 0.61 mg Se per kg. The natural presence of SeNPs was attributed by the authors to the reduction of SeI^V or Se^VI to Se⁰ by soil microorganisms as a detoxification mechanism. A paper covering Se oxidation state determination in 16 samples of coal fly ash by XANES and μ-XRF has also been published.¹⁰⁰ The XANES analysis indicated a wide distribution of oxidation states, including Se⁰, Se^IV and Se^VI, and associations with Fe and Ca. Compared to previous studies reported in the literature, a high percentage of Se⁰ was obtained with two of the samples analysed containing percentages higher than 89% of total Se and two others with values higher than 45% of total Se. The μ-XRF and μXANES measurements support the results obtained by bulk XANES. Elemental maps by μ-XRF were used to identify Se-containing particles with an average size of 40 μm and containing multiple oxidation states. The Se speciation results were correlated with parameters such as elemental composition (Al₂O₃, SiO₂, FeO and CaO), loss on ignition, average particle size and selective catalytic reaction. PCA analysis suggests selective catalytic analysis used to limit NOx emissions during coal combustion as the main responsible for the presence of reduced Se oxidation states.

4.17 Silver

Most papers covering Ag speciation involve the separation of ionic Ag species from Ag nanoparticles (NPs), usually involving some form of chromatography or field flow fractionation coupled with ICP-MS, whilst the use of X-ray-based techniques is less commonly reported. This year, Zhao et al. have reported a method for the characterisation and quantification of Ag complexes with dissolved organic matter by size exclusion chromatography coupled to ICP-MS.¹⁰¹ Complexes of Ag with dissolved organic matter (DOM) were prepared at three different concentration levels (2, 20 and 100 μg L⁻¹), three different water samples (DI water at pH 6.7, medium hard water at pH 7.9 and artificial seawater at pH 8.2), and three different DOM materials (Suwannee River humic acid (SRHA), Suwannee River fulvic acid (SRFA) and Suwannee River natural organic matter (SRNOM)). Samples were incubated for three days prior to dialysis to remove Ag⁺ followed by a further three days of incubation prior to sample analysis. Chromatographic separations were by SEC with a mobile phase of 0.5 mmol per L ammonium acetate flowing at 0.5 mL min⁻¹ with detection by ICP-MS. Quantification of the Ag/Dom complexes was by suIDMS using a ¹⁰⁹Ag spike added post column via a syringe pump at a flow rate of 0.1 mL min⁻¹, whilst ICP-MS or ICP-OES was used for total Ag measurements. The results obtained by this approach showed that Ag⁺ was weakly bound to all three DOM types in all of the sample matrices, which resulted in Ag/DOM complex dissociation during the chromatographic separations and a full discussion of the factors behind this is contained in the paper.

4.18 Sulfur

There has been one review of S measurements by atomic spectroscopy published this year.¹⁰² The paper covers covering articles published between January 2015 and April 2023, with 117 citations, and instruments covered were classified as Q-ICP ICP-MS, TQ-ICP-MS, SF-ICP-MS and MC-ICP-MS. The data is presented mainly in well-designed tabular format which gives, e.g., instrument type and model, any hyphenation of techniques, sample matrix, analysis type such as speciation or isotope ratio measurements, although it does lack the actual analytes when speciation is concerned. Further tables cover potential spectral interferences, with the text giving a brief outline of some strategies employed to minimise these, and the LOD values achieved in the cited works. Greater than 30% of the articles covered related to elemental speciation and 30% report on analysis by LA. Due to the tabulated format, the work provides a good starting point for those new to this field though the text only gives brief highlights of the work covered.

Two papers form the same broad group of researchers cover the coupling of CE with ICP-MS for S speciation purposes. The first of these compares a vibrating capillary nebuliser (VCN) and other commercially available CE-ICP-MS interfaces for this.¹⁰³ The VCN, the design of which is fully covered in the text, is described as a low-cost, non-pneumatic nebuliser based on the design of capillary vibrating sharp-edge spray ionisation with a piezoelectrically driven nebulisation source. This is said to create an aerosol independent of gas flows without a low-pressure region at the nebuliser orifice. A SF-ICP-MS instrument was used in this work with SO₄²⁻ in river water as the analyte and the method optimisation for both the VCN and conventional nebulisers is fully covered. The CE-VCN-ICP-MS method achieved about a two-to-four-fold reduction insensitivity compared to the pneumatic nebuliser data but, due to lower noise levels and an improved linear correlation for the calibration standards, gave comparable LOD and LOQ values to this latter approach. This is a good example of the need to optimise instruments for stability and sensitivity rather than simply the latter. A breakdown product of some aquatic anti-fouling paints, 2-pyridinesulfonic acid (PSA), was also spiked into the river water to demonstrate the capability of the CE separation and the method proved linear for both PSA and SO₄²⁻ in river water up to 250 mgL⁻¹ for both analytes. The second paper reported on here is a proof of concept study for the online measurement of the isotope ratios of S in proteins via CE-MC-ICP-MS.¹⁰⁴ For this work the target analyse was bovine serum albumin (BSA) and three IAEA AgS CRMs (S1, S2 and S3), which have δ³⁴S_VCDT values ranging from 22.62 to −32.49‰. The solutions for analysis were diluted to a final S concentration of approximately 1 mg L⁻¹ in 0.3 mol per L HNO₃, containing 3 mg L⁻¹ of Na to improve the transmission of S through the membrane desolvation system used to negate O₂-based interferences. Mass bias correction was by the standard-sample bracketing approach using a multiple injection procedure so that data could be acquired in one run. For the IAEA S2 and S3 CRMs the found data was in statistical agreement with the certified values, albeit with greater uncertainty values which arise from the measurement of transient signals compared with those from continuous measurements used in the certification procedure. The δ³⁴S_VCDT value for the BSA was calculated to be 3.59 ± 2.24 (n = 6) by CE-MC-ICP-MS. The uncertainty of the measurement is greater than that of a cited value obtained using a similar instrumental setup and this difference is attributed to the poorer sensitivity obtained compared with that of the cited reference, 4.5 vs. 0.3 V, respectively, and the authors suggest the need for better tuning of the CE-MC-ICP-MS, probably with respect to the impact of the effects ¹⁶O₂ interference on the ³²S signal. The authors suggest that the current set up is suitable for their current needs with and for more complex sample types but further improvements to the CE separation parameters to reduce run times (background electrolyte composition, pH, capillary length and inner diameter); other types of CE capillary coating to minimise the interaction of proteins with the CE capillary wall; and potentially employing large volume sample stacking approaches within the CE capillary to increase the amount of injected protein is also mooted.

Two further papers also report on S speciation. In one paper GC-ICP-MS was used to determine the δ³³S and δ³⁴S values in a range of organic compounds, whilst an elemental analyser (EA) was used to determine bulk values.¹⁰⁵ The compounds studied, C₄H₈S (tetrahydrothiophene, THT), C₄H₄S (thiophene, THI), C₄H₁₀S (diethyl sulfide, DES) and C₂H₆S₂ (dimethyldisulfide, DMDS), were separated on an 60 m × 0.32 mm × 1 μm column (Zebron ZB-1) using an isothermal temperature of 150 °C coupled to the MC-ICP-MS instrument via a transfer line heated to 280 °C. For the bulk measurements by EA-ICP-MS, the found δ³⁴S‰ values were in good agreement with the certified values for a suite of 7 CRMs of varying compositions but less so for the δ³³S values which were compared with literature values as the CRMs are not certified for δ³³S. For the organic compounds, the δ³⁴S‰ values were determined to be −2.13 ± 0.10, +3.56 ± 0.13, +8.23 ± 0.15 and +6.17 ± 0.05 for THI, THT, DMDS and DES, respectively, whilst the δ³³S‰ values were −1.74 ± 0.18, +4.13 ± 0.17, +8.60 ± 0.17 and +6.36 ± 0.20 for THI, THT, DMDS and DES, respectively. All of these values were in agreement with those obtained from EA-ICP-MS analysis of the same samples. The final paper here investigates the factors affecting Se atmospheric deposition by integrating Se and S speciation analysis and atmospheric dynamic modelling.⁹⁸ The paper, which gives full coverage to sample collection and preparation, mainly focuses on Se speciation, which is covered in detail in Section 4.16. Sulfur species were separated on a C18 column, 100 × 4.6 mm, 5 μm, with a gradient mobile phase of 24–240 mmol per L formic acid and 1% MeOH at pH 2.1 flowing at 1 mL min⁻¹ and a run time of 11.5 minutes coupled to a TQ-ICP-MS with O₂ used as a reaction gas. A solution of 30 μg per L of Y and 14% v/v TMAH was added post-column as an internal standard and calibration was made using dimethyl sulfoxide (DMSO), dimethyl sulfone (DMSO₂), methanesulfonic acid (MSA), methanesulfinic acid (MSIA), hydroxymethanesulfonate (HMS), and SO₄²⁻. The only time S speciation is mentioned in the discussion is when it can be related to the Se speciation data evaluation.

4.19 Technetium

Two papers report on the speciation of Tc this year, continuing the series of papers on this topic from the same research group, and this was carried out by using HPLC-ICP-MS and HPLC-ES-HR-MS. Due to its use in nuclear medicine, ⁹⁹Tc needs to be monitored in urine and wastewaters. For detection in urine, an AE-ICP-MS method, with aerosol desolvation of the column eluent, was developed.¹⁰⁶ An IonPac AG9-SC column (4 × 50 mm) was used for the separations with 150 mmol per L NH₄NO₃ at pH 9 and high-purity water used in a gradient elution followed by column rinsing with 500 mmol per L NH₄NO₃ then 20 mmol per L HNO₃, all flowing at 0.65 mL min⁻¹. Quantification was by an approach termed isobaric dilution analysis (IBDA), which uses a Ru spike added post-column and 0.05 mL min⁻¹, with a sensitivity correction factor calculated by comparison of ⁹⁹Tc and ⁹⁹Ru signals obtained using this system. The ⁹⁹Ru eluted at 50 s, whilst the Tc compound eluted at 150 s, thus avoiding analytical errors due to co-elution of the two isobaric species. Column recovery experiments, 10 to 1000 ng per L Tc, ranged from 84 to 103%, whilst the LOD and LOQ values obtained were 0.67 ± 0.04 ng and 2.2 ± 0.1 ng kg⁻¹, respectively. The method was used to determine ⁹⁹Tc species in undiluted urine ⁹⁹TcO⁴⁻ (19.6 ± 0.5 ng L⁻¹). Another ⁹⁹Tc-containing species was also detected (9149 ± 4 ng L⁻¹) and the latter species was most likely the injected ⁹⁹Tc-MDP tracer used for bone scintigraphy. For the detection of ⁹⁹TcO⁴⁻ in wastewater, a TK201 resin was used for preconcentration followed by washing with 0.1 mol per L HNO₃ and elution, with a 2-step gradient with NH₄OH (0.5 mol L⁻¹) followed by a mixture of NH₄NO₃ (0.15 mol L⁻¹ at pH 9.2) and NH₄OH onto the same AE column as used for the work involving urine.¹⁰⁷ Quantification was again by IBDA with ⁹⁹Re. When the additional (manual) preconcentration step was used, via filter discs impregnated with the TK201 resin, the sensitivity was further increased by a factor of 9, giving an LOD of 0.7 ± 0.1 fg kg⁻¹ and a preconcentration factor as high as 4615. A sample of wastewater from the retention basin of a hospital radioisotope therapy ward was found to contain a concentration of 89 ± 4 fg kg⁻¹ of ⁹⁹Tc.

4.20 Tellurium

Two papers report on Te speciation studies this year. The use of Te as additive for special steels and in electrical and electronic products has increased the amount of Te released into the environment. However, information about the impact of this metalloid on the environment and ecosystems is scarce. Thus, a method based on the use of HPLC-ICP-MS for the speciation of Te^IV and Te^VI in seawater samples collected at different depths and in leachates of electronic products (solar panels, DVD) has been presented.¹⁰⁸ Four types of chromatographic separation columns were used for comparison: anion-exchange column (PRPX-100, 150 mm × 4.1 mm), cation exchange column (TSKgel® IC-Cation I/II HR, 100 mm × 4.6), HILIC column (PC-HILIC 150 mm × 4.6 mm) and reversed-phase column ODS (150 mm × 4.6 mm). The Te^IV and Te^VI species were completely separated with the reversed-phase ODS column within 5 minutes by using 4 mmol per L cysteine (pH = 2.3) as eluent. L-cysteine efficiently binds Te^VI preventing its retention on the column. The developed method offers the novelty of performing iTe speciation in just a single run, which constitutes an important improvement over the existing methods where iTe speciation is achieved by calculating the difference between total Te concentration and the concentration of Te^IV or Te^VI. The analytical procedure offers LOD values for Te^IV and Te^VI of 1.4 ng g⁻¹ and 0.5 ng g⁻¹, respectively. Accuracy of the proposed method was evaluated by analysing standard solutions of known concentration of total Te. For all samples, including seawater, good agreement between the results and the certified or nominal values were obtained. The method was further applied to the determination of Te^IV and Te^VI in electronic product leachates. Four leaching agents were tested for their efficacy for this purpose: H₂SO₄, TMAH, citric acid and seawater. Two, H₂SO₄ and TMAH, provided the highest leaching efficiencies (52% and 66% of the total Te, respectively), whereas citric acid provided an extraction efficiency less than 20% of total Te. Seawater was not able to extract Te from the electronic devices. The percentage of Te^IV detected in H₂SO₄, TMAH and citric acid was 51.1, 60.2 and 73.3%, respectively. In a second paper, the Te metabolic pathway in broccoli plants exposed to TeO₄²⁻ has been tracked by HPLC-ICP-MS and LC-ES-Orbitrap-MS.¹⁰⁹ Broccoli plants were hydroponically grown with Te^IV for 14 days. After growing, plants were harvested and divided into leaves, stems and roots. Aqueous extracts from the collected parts of the plant were submitted to HPLC-ICP-MS on a SEC column (7.5 mm × 300 mm) maintained at 30 °C. The retained species were eluted with 50 mmol per L NH₄Ac (pH = 6.5) at a flow rate of 0.6 mL min⁻¹. Different Te compounds (the metabolite named by the authors as UKTe and hexonic acid tellurates complexes) were synthesised. Tellurate complexes of gluconic acid and gluonic acid were prepared by treating both acids with Te^IV for 24 h at room temperature, whereas galactonic acid and mannonic acid–Te complexes were obtained by bromide oxidation of galactose and mannose, respectively, followed by purification and subsequent incubation with Te^IV during 24 h at room temperature. The UKTe was prepared ex vivo by incubating an aqueous extract of broccoli florets with Te^IV for 12 h. All these solutions were employed in LC-ICP-MS speciation and in the determination of their exact mass and structure by LC-ES-Orbitrap-MS. The LC-ICP-MS analysis of leaf extracts evidenced the presence of two Te-containing chromatographic peaks at the retention times of 15 and 18.5 minutes. The second peak was assigned to Te^IV by retention time matching with the standard solution, whereas the first peak was assigned to the unknown metabolite UKTe, detected by the authors in Te-enriched garlic, alga and india mustard in previous works. The UKTe present was identified by LC-ES-Orbitrap-MS as gluconic acid 3-tellurate, and it was the major Te metabolite found in tellurate-exposed broccoli. Neither of the other hexonic acid–tellurate complexes or Te-containing amino acids were detected. This study is the first to identify the metabolite GA-3Te in plants and suggest the defensive role of gluconic acid against Te^IV toxicity in plants.

4.21 Thorium

Chadirji-Martinez et al. have investigated the Th speciation in ilmenite concentrates containing 133 mg per kg Th and 12.8 mg per kg U from the Mandena deposit, Madagascar.¹¹⁰ They employed a number of techniques including powder XRD, EPMA, SEM, ICP-MS, LA-ICP-MS, synchrotron XAS, microbeam synchrotron XRF mapping, and microbeam synchrotron Laue XRD analysis. The synchrotron-based measurements were made at the BioXAS-Main beamline of the Canadian Light Source. They showed that 55% of Th in the Mandena ilmenite concentrates could be accounted for as monazite-(Ce), mainly as discrete grains. That the remaining 45% was thorianite was deduced from the results of synchrotron Th L^III-edge XANES and EXAFS analyses and was correlated with the degree of ilmenite alteration. The researchers suggested that the results support the formation of thorianite from the alteration of ilmenite and precipitation on its surface. They deduced the presence of monazite-(Ce) and thorianite in the Mandena ilmenite concentrates by PCA of trace elements. They concluded that combined magnetic separation and acid leaching were effective for decreasing the Th (and U) concentrations in the Mandena ilmenite concentrates.

4.22 Tin

There have been few published reports on the speciation of Sn this year. Although the use of OTCs is widely regulated, they are still the most used organometallic compounds in industry. In the workplace, OTCs can be released as vapours or dust particles and can be absorbed by inhalation or skin contact. Occupational exposure thus represents a risk for the absorption of OTCs for employees, although monitoring is not routine. Cläsgens et al. have developed a separation method for 11 OTCs using HPLC-ICP-MS.¹¹¹ The method allows a near baseline separation of MMT, MBT, MOT, MPhT, DMT, DBT, DPhT, TMT, TBT, TPhT and tetramethyltin (TTMT) within 22 min on a C18 column. The method uses a flow rate gradient with methanol, acetonitrile, and ultrapure water, 6% (v/v) acetic acid and 0.17% (m/v) alpha-tropolone. Ten of the analytes showed linearity in the working range from 10 to 100 μg OTCs per L with R² > 0.999. Due to its high volatility, TTMT showed a quadratic relationship between concentration and signal intensity with R² = 0.9998. The LOD values were between 0.14 and 0.57 μg Sn per L and LOQ values were between 0.49 and 1.97 μg Sn per L. Over the course of the study, problems with thermal instability and cross reactivity of the OTC in solution became apparent. The formation of two reaction products in mixed OTCs solutions were observed, and these will be the subject of further study. Two studies looking at simultaneous determination of Pb and Sn species in marine samples have been reported by Guo et al. The first looked at ultra-trace Pb and Sn species in seawater by on-line SPE coupled with HPLC-ICP-MS.⁷⁰ A graphene oxide-modified silica conditioned with 1 mmol per L benzoic acid was used to extract the Pb and Sn species from 10 mL of sample. The LOD values were from 2 to 8 pg L⁻¹ with the large enrichment factors ranging from 522- to 2848-fold. The method was found to be applicable to the quantification of ultra -trace Pb and Sn species at pg L⁻¹ levels in uncontaminated seawater samples collected in Hangzhou Bay, China. In the second study, Pb and Sn species were determined in Antarctic krill and fish using HPLC-ICP-MS based on strong cation-exchange and a 10-cm Amphion II column (mobile phase: 4 mmol per L sodium dodecyl benzene sulfonate at pH 2.0). Inorganic Pb and Sn, four organotin and two organolead compounds could be analysed in 16 min with LOD values ranging from 0.02 to 0.24 μg L⁻¹.

4.23 Uranium

The geochemistry, environmental impact and biogeochemical cycling of U in the environment is a continuing area of interest, particularly in relation to the effect of climate change on its remobilisation from soil, but also the legacy of historical contamination and its effect on marine ecosystems. These studies have required the implementation of novel combinations of analytical instrumentation to understand U in the environment. The effect of geochemical processes on the speciation and mobilisation of U has been investigated in relation to the thawing of permafrost in NW North America.¹¹² The environmental mobility of U is closely tied to its oxidation state: U^VI prevails under oxidising conditions and has relatively higher mobility than U^IV, which forms poorly soluble phases. The analytical approaches used to elucidate the solid-phase speciation of U in permafrost soils and thawed porewaters collected from the subarctic study area included the use of XAS, μXRF and SEM-EDS analyses. Geochemical speciation modelling was also used to assess the role of aqueous complexation on U mobility in thawed permafrost. The results provide important insights into the geological and geochemical controls that govern U mobility during permafrost thaw and identified risks to water quality. Permafrost thaw produced circumneutral pH porewater (pH 6.2–7.5) with elevated dissolved uranium (0.5–203 μg L⁻¹). Geochemical modelling indicated that calcium–uranyl–carbonate complexes dominated the dissolved U speciation. The study highlighted the potential for the mobilisation of U from permafrost soil and that U fate is linked to dynamic biogeochemical reactions involving organic carbon and groundwater chemistry. In a study to understand the transfer and accumulation of U in marine biota, mussels were used as sentinel species because of their sedentary nature and ability to filter seawater.¹¹³ In the aquatic environment, U is found in two forms: tetravalent U^IV mainly in immobile colloidal forms and hexavalent U^VI that is a more soluble form. The uranyl form is ubiquitous in atmospheric conditions unless a specific reducing mechanism is activated. This is also the case in seawater where speciation studies have confirmed the occurrence of the uranyl tricarbonato complex. In the investigation, a combination of several analytical techniques were used including: ICP-MS, XAS, time-resolved LIF, and imaging (TEM, μXAS, SIMS) techniques. Two cohorts of mussels from the Toulon Naval Base and the Villefranche-sur-Mer location were investigated. The measurement of U Concentration Factor (CF) values showed a clear trend in the organs of M. galloprovincialis: hepatopancreas ≫ gill > body ≥ mantle > foot. Although CF values for the entire mussel are comparable for both sampling locations, hepatopancreas values showed a significant increase in those from Toulon versus Villefranche-sur-Mer. In further U speciation studies, the oxidation states of U in zircon and other minor mineral phases represent important time-capsules for studying the evolution of Earth and other planetary bodies, as these minerals can record both temporal and compositional information regarding their host rocks.¹¹⁴ In silicate melts, U can occur in either the U^IV, U^V, or U^VI valence state and its redox sensitive nature could, in principle, allow for a greater understanding of geological processes. In this study, conventional XANES spectroscopy was applied to a set of natural zircon (n = 140), titanite (n = 9), apatite (n = 7), baddeleyite (n = 7), and garnet (n =2) samples to determine the oxidation state of U in these crystals. The results from this work establish U oxidation states in zircon as a powerful new tracer of magma redox status. Also, because XANES is non-destructive and can be performed in situ, this technique can be utilised alongside other microanalytical methods (e.g., LA-ICP-MS, SIMS) to further expand the breadth of information that can be extracted from single mineral grains.

4.24 Vanadium

Two papers report on V speciation this year. Firstly, a SPE separation procedure for the determination of V^IV and V^V in seawater by ICP-MS was developed by Kurahashi et al.¹¹⁵ A Chelex-100 column (0.2 g) was conditioned with 24 mL of 0.1 mol per kg HCl followed by 60 mL of H₂O at 2.8 mL min⁻¹ (total time 30 min), then 10 mL of sample solution (without pH adjustment) were loaded at 2 mL min⁻¹ followed by washing with 20 mL acetate buffer (pH 4.5). The V^V was eluted with first 16 mL of 0.1 mol per kg NH₄OH (pH 11) followed by 10 mL of H₂O. Then, the V^IV was eluted with 20 mL of 0.2 mol per kg HNO₃ (pH 0.8). Fractions (4 mL) of the eluent were collected, evaporated and the residues dissolved in 2 mL of 0.5 mol per kg HNO₃. The total loading and elution time was 38 min. Finally, the column was cleaned with a six-step procedure that took at least a further 56 min. The LOD values were 0.47 and 0.87 nmol kg⁻¹ for V^IV and V^V, respectively (corresponding to 24 and 44 ng kg⁻¹). The method was validated by the analysis of three NRCC CRMs SLRS-6 (river water), NASS-7 (coastal seawater) and SLEW-3 (open ocean seawater) and was applied to the analysis of selected samples from open ocean and coastal seawater collected in the South-East Atlantic Ocean during GEOTRACES research cruise GA08 that had been kept frozen for 5 years. The researchers also showed that under the operating conditions of their procedure V^IV was stable for at least 24 h (and maybe even for 14 days) with respect to oxidation by atmospheric oxygen and therefore there was no need for an anaerobic chamber, a considerable benefit compared with their previous method, which had involved a pH adjustment prior to loading. In the second paper on V speciation, Volcheck et al. devised an HPLC separation with parallel ES-MS and ICP-OES detection for the characterisation of three phosphovanadotungstates, [PV_nW_12−nO₄₀]^{(3 + n)−} for values of n of 1, 2 and 3.¹¹⁶ This development is a continuation of their previous work on the speciation studies of niobium-substituted polyoxometalates by HPLC-ICP-OES. Although it is not clear why, they chose ion-pair chromatography over ion-exchange chromatography and conducted an extensive investigation of candidate ion-pair agents, as well as other relevant parameters, such as pH and flow rate. They chose TBAH as the ion-pair agent at 0.04% in 1 mmol per L HNO₃ acid (at pH 3) as mobile phase A, and ACN as mobile phase B. The species were separated on a C18 column (2 × 75 mm) with a four-step gradient elution program at a total flow rate of 0.2 mL min⁻¹. Following the 18-minute separation stage, the column was purged with ACN for 4 min, conditioned for 4 min by reversing the gradient to 100% mobile phase A, and then equilibrated for a further 6 min. The total run time was about 32 min. The flow exiting the UV detector was split between the OES and MS detectors (ratio not given) and the flow to the ICP-OES instrument diluted with H₂O (delivered by a peristaltic pump) just prior to the nebuliser (again the ratio was not given). Peak purity was confirmed by measuring the ratio of the V to W signal intensities across the three peak profiles, the corresponding species of which were baseline separated in the elution order of n = 3, n = 2 and n = 1. The paper and the supplementary information contained a detailed discussion of the interpretation of the ES mass spectra.

4.25 Zinc

There are two reports of multielement speciation published this year: Co, Cu, Fe, Mn and Zn in cell culture media, and Cu, Fe, Mn, Ni and Zn in bamboo and soil samples. Both of these reports are covered in Section 4.5, Copper.

5 Biomolecular speciation analysis

The development of new methods for the absolute quantification of proteins and metalloproteins is important in metallomic studies, as the majority of currently used calibration methods approaches are based on relative rather than absolute methods. Two approaches described recently use indirect calibration with elemental measurements and calibration standards that are not an exact match for the proteins being quantified. Tukmetova et al. have used online ID-CE-ICP-MS for the quantification of S in biological compounds via post-column addition of a ³⁴S spike.¹¹⁷ This approach was described as S speciation and further developed for the accurate measurement of the serum protein albumin in both human and bovine serum. The quantity of S was converted to the quantity of the compounds investigated, which included sulfate, methionine, cysteine, cystine, and albumin. This was possible because of the established S content of the molecules. The compounds were separated on a CE-fused silica capillary (90 cm length with 50 μm i.d.) coated with successive multiple ionic polymer layers. This reduced the adsorption associated with longer capillaries and improved separation efficiency and resolution, as well as improving the repeatability. The CE separation was interfaced to the SF-ICP-MS instrument using a MiraMist CE nebuliser with a sheath liquid, containing the ³⁴S-enriched isotope standard delivered at a constant flow rate via a syringe pump. The method included several improvements compared to previously published setups: (i) reduced adsorption of proteins onto the capillary, (ii) baseline separation of the compounds in less than 30 min, (iii) quantification of several S-containing species within one run, (iv) SI traceability of the quantification results through online ID, and (v) facilitated data processing of the transient signals using in-house developed software (IsoCor). The use of SAX chromatography coupled to ICP-MS/MS detection has implemented quantification via the use of EDTA-Cu complexes for the calibration of holo-ceruloplasmin in Wilson disease patients.¹¹⁸ The serum proteins were separated using gradient elution and the Cu monitored at m/z 63 (Cu) and 48 (SO) using oxygen reaction mode. The accuracy of the holo-ceruloplasmin measurement was established using two human serum CRMs (LGC8211 and ERM DA250A), with certified values for total Cu and published values for Cu bound to ceruloplasmin, giving a mean recovery of 94.2%. Regression analysis between the sum of the Cu-containing species and total Cu concentration in the patient samples was acceptable (slope = 0.964, intercept = 0, r = 0.987) and a difference plot gave a mean difference for Cu of 0.38 μmol L⁻¹. Intra-day precision for holo-ceruloplasmin at serum Cu concentrations of 0.48 and 3.20 μmol L⁻¹ were 5.2 and 5.6% CV and the intermediate precision at concentrations of 0.80 and 5.99 μmol L⁻¹ were 6.4 and 6.4% CV, respectively. The LOD and LOQ for holo-ceruloplasmin were 0.08 and 0.27 μmol L⁻¹ as Cu, respectively. The values for total Cu and ceruloplasmin Cu determined using this method were used to calculate two new clinical measurements, termed accurate non-ceruloplasmin Cu (ANCC) and Relative ANCC, which were shown to be important as new diagnostic and monitoring biomarker indices for Wilson disease.

In another application, CE coupled to ICP-MS/MS has been used in combination with LC-ES-MS/MS to investigate the reactivity of thioredoxin (Trx), a ubiquitous S-containing protein that maintains cellular redox balance, with the Au^I containing drug auranofin, as well as two other candidate therapeutic anti-cancer drugs containing Au^I complexes with N-heterocyclic carbene (NHC) ligands.¹¹⁹ Direct infusion ES-MS/MS was used to elucidate the structure, stoichiometry, and kinetics of formation of Trx-Au adducts in vitro. The fragmentation of the adducts in the gas phase gave insights into the exact Au binding site within the protein. The method showed issues with using RP-HPLC, including: the difficulty of elution of Au compounds; the instability of the metal-protein adducts on the column; and loss of the phosphine or NHC ligands from Au^I. These limitations were eliminated when using CE for the separations, which enabled the resolution of the gold compounds, Trx and the formed adducts. The ICP-MS/MS detection allowed for the simultaneous quantitative monitoring of the Au and S isotopes and the determination of the metallation extent of the protein. This combined analytical approach overcame issues such as formation of metal adducts in the MS source when a mixture of reagents was introduced, and changes in speciation when using RP chromatography. The recovery and resolution of the metal adducts formed was improved using CE separation and quantification of the extent of metallation was more reliable using ICP-MS detection compared to ES-MS/MS. However, fragmentation of the formed adducts and analysis by ES-MS/MS was indispensable for providing insight into the metal-binding site within the protein.

6 Abbreviations

2D	two-dimensional
AB	arsenobetaine
AC	alternating current
ACN	acetonitrile
AE	anion exchange
AEC	anion-exchange chromatography
AFS	atomic fluorescence spectrometry
ANOVA	analysis of variance
APDC	ammonium pyrrolidine dithiocarbamate
AR	angular-resolved
ASU	Atomic Spectrometry Update
ASV	anodic stripping voltammetry
BSA	bovine serum albumin
CCM	cell culture media
CE	capillary electrophoresis
CEC	cation-exchange chromatography
CFS	cerebospinal fluid
COF	covalent organic framework
CPC	centrifugal partition chromatography
CRM	certified reference material
CS	continuum source
CVR	coefficient of variation
DBS	dodecyl benzene sulfonate
DBT	dibutyltin
DES	deep eutectic solvent
DIT	3,5-diiodo-L-tyrosine dehydrate
DLLME	dispersive liquid–liquid microextraction
DMA	dimethylarsenic (include oxidation state if known)
DMSO	dimethysulfoxide
DMSO2	dimethyl sulfone
DMT	dimethyltin
DOE	design of experiments
DOI	dissolved organic iodine
DOM	dissolved organic matter
DPhT	diphenyltin
DSPME	dispersive solid phase microextraction
ED	energy dispersive
EDTA	ethylenediaminetetraacetic acid
EPA	Environmental Protection Agency
EPMA	electron probe microanalysis
ERM	European reference material
ES	electrospray
ETAAS	electrothermal atomic absorption spectrometry
EtHg	ethylmercury
EtOH	ethanol
eV	electron volt
FAAS	flame atomic absorption spectrometry
FTIR	Fourier transform infrared
GC	gas chromatography
GE	gas electrophoresis OR gel electrophoresis (but NOT in same review)
GPX	glutathione peroxidase
HAS	human albumin serum
HERFD	high-energy-resolution fluorescence detection
HG	hydride generation
HILIC	hydrophilic interaction liquid chromatography
HPLC	high-performance liquid chromatography
HR	high resolution
IAEA	International Atomic Energy Agency
iAs	inorganic arsenic
IBDA	isobaric dilution analysis
ICP	inductively coupled plasma
ID	isotope dilution
IDA	isotope dilution analysis
IIP	ion-imprinted polymer
KED	kinetic energy discrimination
KHP	potassium hydrogen phthalate
LA	laser ablation
LC	liquid chromatography
L-Cys	L-cysteine
LIBS	laser-induced breakdown spectroscopy
LIF	laser-induced fluorescence
LLME	liquid–liquid microextraction
LMW	low molecular weight
LOD	limit of detection
LOQ	limit of quantification
MAE	microwave-assisted extraction
MBT	monobutyltin
MC	multicollector
MeHg	methyl mercury
MeOH	methanol
MESA	multi-ethnic study of atherosclerosis
MeSeCys	methylselenocysteine
MICAP	microwave inductively coupled atmospheric-pressure plasma
MIT	3-iodo-L-tyrosine
MMA	monomethylarsenic
MMT	monomethyltin
MOF	metal–organic framework
MOT	monooctyltintrichloride
MPhT	monophenyltin
MS	mass spectrometry
MS/MS	tandem mass spectrometry
MSA	methane sulfonic acid
MSWIA	municipal solid waste incineration ash
MWCNT	multi-walled carbon nanotubes
NADES	natural deep eutectic solvent
NIST	National Institute of Standards and Technology
NMR	nuclear magnetic resonance
NP	nanoparticle
NTS	non-target screening
OES	optical emission spectrometry
OTC	organotin compounds
PCA	principal component analysis
PEEK	polyetheretherketone
PSA	2-pyridinesulfonic acid
Q	quadrupole
REE	rare earth element
RP	reversed phase
RSD	relative standard deviation
SAX	strong anion exchange
SCX	strong cation exchange
SDBS	sodium dodecylbenzene sulfonate
SDS	sodium dodecylsulfate
SEC	size exclusion chromatography
SeCys	selenocysteine
SeCys2	selenocystine
SelP	selenoprotein
SEM	monomethylmercury
SeMet	selenomethionine
Se-TxNRD	thioredoxin reductase-bound Se
SF	sector field
SIMS	secondary ion mass spectrometry
SPE	solid phase extraction
SPME	solid phase microextraction
SR	synchrotron radiation
SRM	standard reference material
SS	species specific
SSB	sample standard bracketing
SXRF	synchrotron XRF microscopy
TBT	tributyltin
TCH	thiocarbonyldihydrazide
TEM	transmission electron microscopy
TFA	trifluoroacetic acid
THF	tetrahydrofuran
TMAH	tetramethylammonium hydroxide
TMT	trimethyltin
TOF	time-of-flight
TPhT	triphenyltin
TSPP	tetrasodium pyrophosphate
TTMT	tetramethyltin
UAE	ultrasound-assisted extraction
UHPLC	ultra-high-performance liquid chromatography
VCN	vibrating capillary nebuliser
WD	wavelength dispersive
XANES	X-ray absorption near-edge structure
XAS	X-ray absorption spectroscopy
XFM	X-ray fluorescence microscopy
XPS	X-ray photoelectron spectroscopy
XRD	X-ray diffraction
XRF	X-ray fluorescence
XRPD	X-ray powder diffraction

Conflicts of interest

There are no conflict of interest to declare.

References

1. R. Clough, C. F. Harrington, S. J. Hill, Y. Madrid and J. F. Tyson, J. Anal. At. Spectrom., 2024, 39(7), 1629–1664.
2. J. R. Bacon, O. T. Butler, W. R. L. Cairns, O. Cavoura, J. M. Cook, C. M. Davidson and R. Mertz-Kraus, J. Anal. At. Spectrom., 2024, 39(1), 11–65.
3. M. Patriarca, N. Barlow, A. Cross, S. Hill, A. Robson and J. Tyson, J. Anal. At. Spectrom., 2024, 39(3), 624–698.
4. E. H. Evans, J. Pisonero, C. M. M. Smith and R. N. Taylor, J. Anal. At. Spectrom., 2024, 39(5), 1188–1211.
5. C. Vanhoof, J. R. Bacon, U. E. A. Fittschen and L. Vincze, J. Anal. At. Spectrom., 2024, 39(9), 2152–2164.
6. E. Bolea-Fernandez, R. Clough, A. Fisher, B. Gibson and B. Russell, J. Anal. At. Spectrom., 2024, 39(11), 2617–2693.
7. S. Hill, H. Goenaga-Infante and R. Garton, J. Anal. At. Spectrom., 2024, 39(9), 2150–2151.
8. E. De Pauw, P. Tack and L. Vincze, J. Anal. At. Spectrom., 2024, 39(2), 310–329.
9. H. Du, P. Yang, L. Y. Xia, J. B. Chen and X. D. Hou, TrAC, Trends Anal. Chem., 2024, 172, 117552.
10. M. N. Oviedo, C. E. Luján, A. A. Lemos, M. B. Botella, M. Llaver and R. G. Wuilloud, Anal. Bioanal. Chem., 2024, 416(11), 2641–2656.
11. Y. C. Wang, Prog. Chem., 2023, 35(10), 1492–1504.
12. S. Fernández-Trujillo, M. Jiménez-Moreno, N. Rodríguez-Fariñas and R. C. R. Martín-Doimeadios, Anal. Bioanal. Chem., 2024, 416(11), 2657–2676.
13. J. Feldmann, H. R. Hansen, T. M. Karlsson and J. H. Christensen, Environ. Sci. Technol., 2024, 58(29), 12755–12762.
14. L. S. Kato, V. H. C. da Silva, D. C. de Andrade, G. Cruz, J. H. Pedrobom, A. Raab, J. Feldmann and M. A. Z. Arruda, Anal. Chim. Acta, 2024, 1331, 343084.
15. I. Dakova, T. Yordanova and I. Karadjova, Molecules, 2024, 29(1), 187.
16. M. Patriarca, N. Barlow, A. Cross, S. Hill, A. Robson, A. Taylor and J. Tyson, J. Anal. At. Spectrom., 2023, 38(3), 496–577.
17. H. Lu, J. Wang, C. Wei, J. B. Chao, T. Zhou and M. T. Zhao, At. Spectrosc., 2024, 45(5), 415–421.
18. S. Han, G. Y. Sun, X. Y. Li and X. B. Feng, Chin. J. Anal. Chem., 2024, 52(10), 1536–1543.
19. C. L. Ma, D. J. Bai, C. L. Wu, Y. D. Li and H. J. Wang, Sci. Total Environ., 2024, 918, 170606.
20. A. R. López, G. Binda, G. Roncoroni, S. Recchia, D. Monticelli and D. Spanu, Molecules, 2024, 29(12), 2870.
21. Y. N. Gotame and D. Beauchemin, At. Spectrosc., 2024, 45(3), 164–172.
22. J. Goodwin, R. K. Marcus, G. McRae, R. E. Sturgeon and Z. Mester, Anal. Bioanal. Chem., 2024, 416(15), 3585–3594.
23. M. B. Botella, P. Y. Quintas and R. G. Wuilloud, J. Food Compos. Anal., 2024, 136, 106748.
24. Y. Y. Gao, X. A. Yang and W. B. Zhang, Talanta, 2024, 275, 126127.
25. B. A. Sadee, Y. Galali and S. M. S. Zebari, RSC Adv., 2024, 14(30), 21563–21589.
26. R. Sim, M. Weyer and A. H. Pétursdóttir, Anal. Bioanal. Chem., 2024, 416(12), 3033–3044.
27. A. Morales-Rodríguez, A. Sahuquillo, C. Minguillón, J. F. López-Sánchez and D. Barrón, Anal. Chim. Acta, 2024, 1315, 342814.
28. L. Schmidt, F. P. Balbinot, D. L. R. Novo, D. Santos, M. F. Mesko and E. M. M. Flores, J. Food Compos. Anal., 2024, 128, 106020.
29. A. Raab, J. Y. Zhang, Y. Ge, F. Fernández-Mendoza and J. Feldmann, Anal. Bioanal. Chem., 2024, 416(11), 2809–2818.
30. J. F. Kopp, L. Paton, Z. Gajdosechova, S. Sinawivat, A. Raab, A. Brownlow and J. Feldmann, Sci.Total Environ., 2024, 946, 173816.
31. A. Izdebska, S. Budzynska and K. Bierla, Molecules, 2024, 29(13), 3055.
32. S. Mukta and A. Gundlach-Graham, J. Anal. At. Spectrom., 2024, 39(2), 491–499.
33. T. Narukawa, S. Okabayashi, M. Tsuboi and K. Chiba, Anal. Methods, 2024, 16(10), 1523–1530.
34. M. H. Kabir, G. Brodie, D. Gupta, A. L. X. Pang, M. M. Rahman and R. Naidu, J. Food Compos. Anal., 2024, 136, 106715.
35. M. Menon, A. Nicholls, A. Smalley and E. Rhodes, Sci. Total Environ., 2024, 914, 169653.
36. M. Walenta, A. Raab, S. Braeuer, L. Steiner, J. Borovicka and W. Goessler, Anal. Bioanal. Chem., 2024, 416(6), 1399–1405.
37. I. S. Permigiani, N. K. Vallejo, P. E. Hasuoka, R. A. Gil and M. C. Romero, J. Food Compos. Anal., 2024, 126, 105898.
38. C. Xiong, A. H. Petursdottir, G. Rikhardsson, S. Stergiadis, A. Raab and J. Feldmann, J. Sci. Food Agric., 2024, 104(11), 6957–6965.
39. D. N. Zhang, Y. Wang, S. F. Wang, F. Xiao, X. Y. Guo, D. K. Xu, F. H. Wang and Y. F. Jia, J. Hazard. Mater., 2024, 476, 134989.
40. L. Borda, G. Bia, L. Borgnino, N. Chiaramonte and M. G. García, J. Hazard. Mater., 2024, 473, 134547.
41. Y. Y. Zhang, X. J. Xie, S. T. Sun and Y. X. Wang, Sci. Total Environ., 2024, 943, 173776.
42. A. Verma, H. V. Chinnasamy, B. Biswas, A. Singh and S. Matheshwaran, Biochem. Biophys. Res. Commun., 2024, 739, 150936.
43. Y. Hashimoto, Y. Imoto, M. Nishikata, H. Suga, S. L. Wang and T. Yasutaka, J. Hazard. Mater., 2025, 484, 136580.
44. M. H. H. Fischel, C. E. Clarke and D. L. Sparks, Geoderma, 2024, 441, 116715.
45. C. Faidutti, C. Doolette, L. Hair, K. R. van Daalen, A. Naheed, E. Lombi and J. Feldmann, Anal. Chem., 2024, 96(10), 4039–4047.
46. T. Oguri, N. Shinohara and S. F. Nakayama, Environ. Health Prev. Med., 2024, 29, 59.
47. X. Zhang, X. Y. Jiang, L. Zhao, S. Chen, Y. L. Yu and J. H. Wang, Anal. Chem., 2024, 96(4), 1742–1749.
48. S. M. Zhou, H. Wang, J. Tang, H. Wang and J. B. Yan, Anal. Sci., 2024, 40(3), 555–562.
49. R. A. Glabonjat, K. Schilling, V. N. Slavkovich, C. N. Izuchukwu, O. Balac, S. Dubey, R. Nandakumar, M. Galvez-Fernandez, W. A. Anderson, M. S. Spaur, M. H. Sobel, A. E. Nigra, T. R. Sanchez, M. R. Jones and A. Navas-Acien, Environ. Res., 2024, 262, 119833.
50. X. Zhang, X. Y. Jiang, J. Y. Cai, S. Chen, Y. L. Yu and J. H. Wang, Anal. Chem., 2024, 96(18), 7155–7162.
51. Y. Yang, B. Zhou and C. Y. Zheng, Molecules, 2024, 29(8), 1795.
52. S. W. C. Chung, Food Addit. Contam., Part A, 2024, 41, 782–789.
53. G. Song, H. Tan, C. Cheng, P. Li, X. Sun, Y. Zhou and Y. Fang, Toxics, 2024, 12(5), 325.
54. H. E. H. Ahmed and M. Soylak, Water, Air, Soil Pollut., 2024, 235(4), 217.
55. F. Faiz, Y. F. Pan, S. Zhang, J. X. Dai, J. Q. Qiao, H. Z. Lian, L. Mao and X. B. Cui, J. Hazard. Mater., 2024, 474, 134860.
56. W. Sapyen, J. Jirattisak, N. Bhawawet, N. Praphairaksit and A. Imyim, Microchem. J., 2024, 202, 110841.
57. M. E. Yang, J. Y. He, H. Y. Liu and J. J. Cao, Int. J. Environ. Anal. Chem., 2024, 1–13.
58. C. C. Huang, Z. Y. Song, X. Y. Xiao, X. Cai, Y. F. Yang, S. H. Chen, P. H. Li, M. Yang and X. J. Huang, Anal. Chem., 2024, 96(1), 179–187.
59. J. Zajda, E. Wadych, K. Ogórek, M. Drozd and M. Matczuk, Electrophoresis, 2024, 45(21–22), 1946–1954.
60. C. J. Stouffer, S. K. Wysor and R. K. Marcus, J. Anal. At. Spectrom., 2024, 39(6), 1600–1608.
61. N. Mayengbam, K. Bierla, A. Angaïts, J. Szpunar, K. B. Singh and R. Lobinski, J. Food Compos. Anal., 2024, 136, 106772.
62. M. Na'es, L. Lühl, D. Grötzsch, I. Mantouvalou, J. Baumann and B. Kanngiesser, J. Anal. At. Spectrom., 2025, 40, 238–247.
63. I. M. Streanga, D. J. Repeta, J. S. Blusztajn and T. J. Horner, Front. Mar. Sci., 2024, 10, 1272968.
64. F. S. Rondan, P. Pisarek, S. Godin, J. Szpunar and M. F. Mesko, J. Trace Elem. Med. Biol., 2024, 83, 127396.
65. S. Frassati, E. Barbaro, C. Rossetti, G. Cozzi, C. Turetta, F. Scoto, M. Roman, M. Feltracco, K. Kim, C. Barbante, A. Gambaro and A. Spolaor, J. Anal. Sci. Technol., 2024, 15(1), 49.
66. A. K. Maurya and P. D. Barman, X-Ray Spectrom., 2025, 54, 38–46.
67. M. A. Münch, A. Voegelin, L. C. C. Hurtarte, J. Göttlicher and T. Behrends, Environ. Sci.: Processes Impacts, 2025, 27, 563–585.
68. T. D. Sowers, M. D. Blackmon, R. T. Wilkin, M. Rovero, S. E. Bone, M. L. Jerden, C. M. Nelson and K. D. Bradham, Environ. Sci. Technol., 2024, 58(21), 9339–9349.
69. F. Guo, P. X. Zeng, J. H. Liu, H. M. Hu, W. B. Zhu, Y. C. Wang and H. Y. Cheng, Food Chem., 2024, 443, 138552.
70. F. Guo, P. X. Zeng, J. H. Liu, H. M. Hu, W. B. Zhu, Y. C. Wang and H. Y. Cheng, Anal. Chim. Acta, 2024, 1294, 342294.
71. X. L. Zhang, F. L. Li, J. B. Chao, Z. C. Li, G. M. Zhang, L. H. Zhai, L. G. Hu, H. P. Jiao and Z. H. Wang, Anal. Chim. Acta, 2024, 1307, 342622.
72. L. Suárez-Criado, S. Queipo-Abad, P. Rodriguez-González and J. I. G. Alonso, J. Anal. At. Spectrom., 2024, 39(2), 508–517.
73. O. V. Shuvaeva, I. A. Bekesha and D. Y. Troitskii, Microchem. J., 2024, 207, 111755.
74. Y. Y. Fang, G. L. Liu, Y. Wang, Y. W. Liu, Y. G. Yin, Y. Cai, A. M. Mebel and G. B. Jiang, Environ. Sci. Technol., 2024, 58(15), 6825–6834.
75. N. Pourzadi and J. Gailer, J. Chromatogr. A, 2024, 1735, 465309.
76. M. Verdugo, F. Avila, J. Ruiz, C. Vásquez, N. Roldán, M. G. Lobos and A. O. Ardiles, J. Anal. At. Spectrom., 2025, 40, 173–185.
77. A. S. Anil, S. Alam and L. K. Thakur, J. Food Compos. Anal., 2024, 129, 106092.
78. D. Ghaffour, A. Leufroy and P. Jitaru, Anal. Bioanal. Chem., 2024, 416(19), 4353–4367.
79. Y. D. Liu, X. Y. Guo, J. Q. Ju, H. T. Gong, H. J. Wang, L. F. Chen, Y. Q. Liu, P. Wang and Y. Liang, Anal. Chim. Acta, 2024, 1323, 343077.
80. Z. L. Shen, G. L. Liu, Y. Y. Guo, T. Jiang, Y. W. Liu, J. B. Shi, L. G. Hu, Y. Song, Y. G. Yin, Y. Cai and G. B. Jiang, Environ. Sci. Technol., 2024, 58(32), 14410–14420.
81. D. Spanu, L. Butti, S. Recchia, C. Dossi and D. Monticelli, Talanta, 2024, 270, 125612.
82. Z. L. Hu, X. Wu, Y. Du, W. Y. Zou, H. J. Dong, M. Song, T. J. Hang and Y. T. Lu, J. Trace Elem. Med. Biol., 2024, 83, 127398.
83. K. D. Szerlag, M. G. Siebecker, F. Izaditame, P. Northrup, R. Tappero and D. L. Sparks, Soil Sci. Soc. Am. J., 2024, 88, 1992–2012.
84. L. R. Mosesso, M. S. Reiter, K. G. Scheckel, N. M. Fiorellino, G. S. Toor and A. L. Shober, J. Environ. Qual., 2024, 53(3), 352–364.
85. J. R. M. Tuyishime, T. Florén, C. Rivard and J. P. Gustafsson, Chem. Geol., 2024, 655, 122083.
86. J. J. Carroll, C. Sprigg, G. Chilvers, I. Delso, M. Barker, F. Cox, D. Johnson and C. A. Brearley, Methods Ecol. Evol., 2024, 15(3), 530–543.
87. W. Ruiz, S. Fontagné-Dicharry, S. Verdier, D. C. Dayton, J. Guillemant, R. Moulian, P. Giusti, C. Barrère-Mangote and B. Bouyssiere, Anal. Chem., 2024, 96(38), 15142–15150.
88. J. Ho, C. Hartinger, M. McKeage and C. Han, J. Chromatogr. B: Anal. Technol. Biomed. Life Sci., 2024, 1243, 124211.
89. L. L. Van Loon, Y. Z. Finfrock, D. M. Meira, R. W. Burgess, G. Bevan and N. R. Banerjee, J. Anal. At. Spectrom., 2025, 40, 137–145.
90. Y. H. Wen, L. Hu, P. Liu, Q. Wang, E. Garcia, W. Y. Yan and Y. Z. Tang, Appl. Geochem., 2025, 178, 106239.
91. A. Angaïts, K. Bierla, J. Szpunar and R. Lobinski, Anal. Bioanal. Chem., 2024, 416(23), 5111–5120.
92. X. Li, L. Shi, P. S. Song, W. Cai, X. M. Luo and B. Zhao, Molecules, 2024, 29(1), 235.
93. C. Zhu, Q. M. Liu, Y. Wang, X. L. Wang, Y. X. Ma, F. X. Yang, W. Z. Dong and H. Ji, Food Chem.: X, 2024, 21, 101088.
94. Q. Y. Li, Y. W. Li, X. Cong, S. Z. Chen, Y. Liu, Y. L. Liu, L. P. Liu, Z. Y. Yan and Y. M. Wang, Food Biosci., 2024, 58, 103827.
95. J. P. Zhan, M. Qi, C. Wang, X. F. Wang, H. Z. Wang and X. L. Dun, J. Agric. Food Chem., 2024, 72(16), 9477–9486.
96. M. B. Botella, M. N. Oviedo, P. Pacheco and R. G. Wuilloud, Spectrochim. Acta, Part B, 2024, 211, 106832.
97. M. Vinceti, T. Urbano, T. Filippini, R. Bedin, C. Simonini, G. Soraru, F. Trojsi, B. Michalke and J. Mandrioli, Biol. Trace Elem. Res., 2025, 203, 2355–2364.
98. E. S. Breuninger, J. Tolu, I. Thurnherr, F. Aemisegger, A. Feinberg, S. Bouchet, J. E. Sonke, V. Pont, H. Wernli and L. H. E. Winkel, Atmos. Chem. Phys., 2024, 24(4), 2491–2510.
99. R. Yang, R. G. Zheng, J. Y. Song, H. Liu, S. J. Yu and J. F. Liu, Anal. Chem., 2024, 96(1), 471–479.
100. E. Garcia, P. Liu, S. E. Bone, Y. H. Wen and Y. Z. Tang, Environ. Sci.: Processes Impacts, 2024, 26(12), 2240–2249.
101. J. Zhao, X. J. Wang, B. W. Gao, X. H. Xia and Y. Li, J. Hazard. Mater., 2024, 466, 133645.
102. Y. B. Zhu, At. Spectrosc., 2023, 44(4), 267–281.
103. T. L. Taylor, D. Tukhmetova, T. P. T. Duong, A. M. Böwe, B. Meermann and A. Gundlach-Graham, Anal. Bioanal. Chem., 2024, 416(7), 1613–1621.
104. D. Tukhmetova, N. Langhammer, J. Vogl and B. Meermann, Electrophoresis, 2025, 46, 290–295.
105. A. Horst, M. Gehre, M. Fahle and S. Kümmel, Anal. Chem., 2024, 96(21), 8510–8517.
106. M. Horstmann, C. D. Q. Quarles Jr, S. Happel, M. Sperling, A. Faust, K. Rahbar, D. Clases and U. Karst, Anal. Bioanal. Chem., 2024, 416(11), 2849–2858.
107. M. Horstmann, C. D. Quarles Jr, S. Happel, M. Sperling, A. Faust, D. Clases and U. Karst, J. Anal. At. Spectrom., 2024, 39(11), 2774–2782.
108. T. Narukawa, A. Wada, C. Cheong and K. Chiba, Anal. Sci., 2024, 40, 2279–2286.
109. S. Takada, Y. Yamagishi, Y. Tanaka, Y. Anan, S. Nagasawa, H. Iwase and Y. Ogra, Chem. Res. Toxicol., 2024, 37, 1210–1217.
110. K. Chadirji-Martinez, G. Hudon, R. Chernikov, E. Heredia, R. F. Feng, A. Crawford and Y. M. Pan, Appl. Geochem., 2024, 160, 105872.
111. C. Cläsgens, T. Schwank and K. Pitzke, J. Chromatogr. A, 2024, 1718, 464695.
112. E. K. Skierszkan, V. A. Schoepfer, M. Fellwock and M. B. J. Lindsay, Environ. Sci. Technol., 2024, 58(38), 17058–17069.
113. R. Stefanelli, M. R. Beccia, P. L. Solari, D. Suhard, S. Pagnotta, A. Jeanson, J. U. Mullot, F. Vernier, C. Moulin, M. Monfort, J. Aupiais and C. Den Auwer, Environ. Res., 2024, 252, 118877.
114. S. K. Houchin, F. L. H. Tissot, M. Ibañez-Mejia, M. Newville, A. Lanzirotti, F. Pavia and G. Rossman, Geochim. Cosmochim. Acta, 2024, 381, 156–176.
115. E. Kurahashi, S. Poehle and A. Koschinsky, Deep-Sea Res., Part I, 2024, 206, 104261.
116. V. V. Volchek, D. G. Sheven and P. A. Abramov, Microchem. J., 2024, 207, 111985.
117. D. Tukhmetova, J. Lisec, J. Vogl and B. Meermann, Anal. Chem., 2024, 96(8), 3276–3283.
118. C. F. Harrington, G. Carpenter, J. P. C. Coverdale, L. Douglas, C. Mills, K. Willis and M. L. Schilsky, Clin. Chem. Lab. Med., 2025, 63, 320–328.
119. M. B. De Maria, M. Matczuk, D. Tesauro, M. Saviano, J. Sikorski, G. Chiappetta, S. Godin, J. Szpunar, R. Lobinski and L. Ronga, Anal. Bioanal. Chem., 2024, 416(11), 2819–2833.

---