Atomic spectrometry update: review of advances in elemental speciation

Contents

• 1 Books and topical reviews

• 2 CRMs and metrology

• 3 Elemental speciation analysis
  • 3.1 Antimony
  • 3.2 Arsenic
  • 3.3 Cadmium
  • 3.4 Chromium
  • 3.5 Copper
  • 3.6 Gadolinium
  • 3.7 Halogens
  • 3.8 Iron
  • 3.9 Mercury
  • 3.10 Molybdenum
  • 3.11 Phosphorus
  • 3.12 Platinum
  • 3.13 Selenium
  • 3.14 Silver
  • 3.15 Tellurium
  • 3.16 Vanadium
  • 3.17 Zinc

• 4 Biomolecular speciation analysis

• Abbreviations

• Conflicts of interest

• References

---

Abstract

This is the 15^th Atomic Spectrometry Update (ASU) to focus on advances in elemental speciation and covers a period of approximately 12 months from January 2022. This ASU review deals with all aspects of the 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. However, because speciation analysis is inherently focused on the relationship between the metal(loid) atom and the organic moiety it is bound to, or incorporated within, atomic spectrometry alone cannot be the sole analytical approach of interest. For this reason molecular detection techniques are also included where they have provided a complementary approach to speciation analysis. This year the number of publications covered has fallen again along with the number of elements covered, most likely due to effects of the Covid-19 pandemic restricting laboratory access. The most popular elements remain as As, Hg and Se and the range of matrices under study is still broad with biological, environmental and foods being the most popular.

---

1 Books and topical reviews

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 applications⁵ and advances in the analysis of metals, chemicals and materials.⁶

Several relevant books or book chapters have appeared in the current review period, though none are devoted entirely to elemental speciation. Some of the texts are concerned with applications of a particular sub-set of analytical techniques. Volume 97 of Wilson and Wilson's Comprehensive Analytical Chemistry, “ICP-MS and trace element analysis as tools for better understanding medical conditions”, consists of 5 chapters covering sample preparation, new statistical methods with machine learning, homeostasis studies in neurological disorders, single-cell biomedical analysis, and developments in imaging and chemical speciation.⁷ The 39 chapters in “X-ray fluorescence in biological sciences: principles, instrumentation and applications”, are divided into five sections: (1) a general introduction that includes a comparison with other techniques, as well as the basic principles of the various X-ray techniques and some applications, (2) SR-XRF spectroscopies, (3) TXRF spectroscopy (4) instrumentation, (5) applications to biological samples, (6) special topics and comparison with other methods, a section that includes a chapter devoted to the determination of As.⁸ Despite the title of the book, X-ray absorption methods are mentioned. Coincidentally, there are 39 mentions of speciation, featuring several elements, including Ag, As, Cd, Fe, Pb, Se, and Zn in “Analytical techniques for trace elements in geochemical exploration”, published by the Royal Society of Chemistry.⁹ Atomic spectrometry techniques are the main focus in the 16 chapters, with chapters also devoted to nuclear and electrochemical methods. Speciation is allocated a separate chapter with sub-sections devoted to each of the 22 elements designated the most commonly studied. The book also contains chapters on statistical and on-site analysis. Analytical techniques are also covered (one chapter out of the 31) in “Clinical chemistry: principles, techniques, and correlations.”¹⁰ Elemental speciation is mentioned briefly in a chapter devoted to trace elements, toxic elements and vitamins. The remainder of the texts are primary concerned with particular analytes, with a focus more on the application rather than the supporting analytical methodology. “Arsenic in plants: uptake, consequences, and remediation techniques” contains an entire chapter on analytical methods, in which atomic spectrometry techniques predominate, in addition to a section on speciation in Chapter 1 (an introduction to arsenic: sources, occurrence and speciation).¹¹ A text covering somewhat similar topics, “Bioremediation of toxic metal(loid)s”, is even more focused on the application rather than analytical methodology.¹² There are separate chapters devoted to As, Cr, Hg and U and other radionuclides. The final two chapters in “Toxicity of nanoparticles in plants: an evaluation of cyto/morpho-physiological, biochemical and molecular responses” deal with aspects of the measurements needed to characterise NPs, techniques used to detect the presence of NPs in treated plant tissues, and the current status of and future directions for examining NPs in plants.¹³ This last chapter covers various imaging techniques, as well as those based on ICP-MS.

Several review articles were concerned with aspects of sample preparation and pre-treatment. Mandal and Lahiri, in summarising the contents of 214 articles, devoted a significant part of their review of CPE to elemental speciation.¹⁴ Following a tutorial introduction, in which they cover several different types of CPE including micro, displacement, rapid synergistic, dual and acid-induced CPE, there are sections devoted to the speciation analysis of 12 elements (Al, As, Cd, Co, Cr, Cu, Fe, Hg, Ni, Pb, Sb, Se). The reviewers also summarised methods for the extraction of anionic species of non-metals/metalloids and for the extraction and speciation of Ag and Au NPs. The reviewers took a critical look at both the advantages (such as only using water as the solvent) and disadvantages (such as the difficulty of automation and the prolonged heating needed) of CPE and concluded that much more research is required in terms strengthening the theoretical foundations, implementing automation, and applying on an industrial scale. A review of SPE (245 references) contains a comprehensive coverage of the principles, but as the review is limited to the use of SPE in procedures that also require further chromatographic separation.¹⁵ The applications reviewed are mostly those featuring molecular analytes but there is a short section on metal speciation within the relatively long section on pesticides and environmental pollutants, and even shorter coverage in the section on foods and beverages. A more general review covering the past 5 years (167 references) of sample preparation for the elemental analysis of environmental samples, does contain substantial coverage of elemental speciation.¹⁶ The reviewers introduced a system of classification of sample preparation featuring three strategies: phase separation; field-assisted acceleration; and integration system. The subsequent coverage of relevant procedures is a bit patchy: CPE, for example, merited only one reference and a single 50-word paragraph. Laser ablation was erroneously classified as a derivatisation strategy, and readers were told that “gas chromatography is very powerful for the separation of gas samples.” The review badly needed the input of a native English-speaking editor, and the structure made it difficult to find coherent commentary relating to a particular analyte or sample type, though, of course, the article can be electronically searched for any given element or matrix.

Several reviews might be classified in the category of “specified technique and application area”. A comprehensive, two-part review examined the impact and potential of ICP-MS in the medical sciences.^17,18 In the first part (258 references), the reviewers covered fundamentals, instrumental basics, technological advances, as well as applications of ICP-MS to biomonitoring, bioimaging and elemental speciation. The review included a substantial section devoted to “elemental speciation analysis in the medical sciences” that, naturally enough, featured many applications of HPLC-ICP-MS, but hardly any in which separation was by CZE. All the electrophoresis applications discussed were in the bioimaging section, in which LA-ICP-MS was the key technique. There was considerable discussion of the roles of elemental speciation and bioimaging in pathologies, particularly cancers (both for diagnostic purposes and for drug studies). The determination of isotopes and isotope ratios were also covered. The reviewers were confident that ICP-MS will be increasingly used in the clinical area, both for routine analyses and to support research, but noted that such increased use was predicated on a “transdisciplinary approach” and cooperation between researchers in diverse disciplines. The second part (166 references with titles) of the review was devoted to nanomedicine, immunochemistry, mass cytometry, and bioassays. Elemental speciation, as such, was not featured in this second part, though progress in these areas is clearly supported by a wide range of indirect methods in which relevant elements are determined by ICP-MS, especially in single-event measurements of individual cells or particles.

Continuing the medical theme, Bazin et al. reviewed (211 references) the clinical applications of XANES methods.¹⁹ They indicated that the subcellular spatial resolution and the capacity to operate at room temperatures and pressures represent major advantages for medical research. Following a general introduction and details of the experimental and data reduction procedures, the reviewers considered several case studies, including the Fe oxidation states in the substantia Nigra region of the brain, drug delivery and investigation of tattoos. In the latter case, Zn plays a crucial role. Various forms of Zn oxide can be distinguished by spatially resolved XANES, allowing discrimination between topically applied zinc species (such as those found in sunscreens) and those derived from tattoo inks. The reviewers concluded with a survey of other techniques to probe the biological roles of metals that included EELS, Energy Loss Near Edge Structure (ELNES) and XPS. They emphasised the complementarity between FTIR spectroscopy (considered the gold standard for the chemical analysis of kidney stones) and XANES. They proposed that for completely unknown samples, XRFS should be the first-choice technique, followed by vibrational spectroscopy, such as FTIR or Raman, to determine more detailed chemistry, while XANES can be applied to describe the electronic state as well as the first coordination sphere of trace elements selected by their edge absorption. They concurred with other researchers who considered such combinations of synchrotron-based X-ray and FTIR microspectroscopies to be ideal for assessing the nature and role of trace elements in biology and medicine.

Speciation analyses, mostly by X-ray spectrometries, form the bulk of a discussion (55 references) of the work performed in the radiochemistry laboratory at the “Institut de Chimie de Nice” (France) regarding the environmental chemistry of radionuclides that started in response to the nuclear fallout from the Chernobyl incident.²⁰ The authors rejected the term “heavy metal” preferring to refer to the analytes as “trace radionuclides” or “metallic trace radionuclides” pointing out that for environmental studies the elements in question come from every part of the periodic table, including noble gases. They argued that because very large dilution factors occur in the environment, chemical toxicity might be considered negligible in comparison with long-term radiation effects (radiotoxicity) and that speciation information is essential for understanding the biochemical pathways at the origins of possible chemical and radiological toxicities. They considered that chemical reactivity and speciation depend on the concentration scale, whereas nuclear decay processes do not. Speciation is therefore key if chemical mechanisms underlying transfer and accumulation are to be elucidated. They highlighted the two difficulties that immediately arise, concentration and heterogeneity, and go as far as to characterise these as the bottleneck of speciation studies. The remainder of the review was devoted to (a) speciation in field studies (dealing with large dilution factors), (b) speciation in field studies (dealing with spatial heterogeneities) and (c) speciation of radionuclides involved in transfer to living species.

In reviewing (121 references) metal contaminants of emerging concern in aquatic systems, Batley et al. stated that for PGEs, REEs, Ga, Ge, In, Li, Ni, Rh, Ta, Te and Tl, there is a reasonable body of toxicity data for most, but the quality is quite variable, and more data are needed.²¹ They noted that reliable toxicity data for Nb and Ta are missing, while the conflicting results for La toxicity need to be resolved, that for most of the elements, measured speciation information is scarce, and for those present in oxidation states of 3+ or higher, there is also a need to explore the links between speciation and bioavailability. The authors then systematically work through the analytes or groups of analytes, summarising the recent literature on (a) environmental concentrations and speciation and (b) aquatic toxicity. Although no details of how speciation data were obtained are provided, the reviewers do make some interesting concluding comments. Firstly, they observed that for many of the technology critical elements (TCEs), the major impact will probably not be from the individual metal finding its way into the environment at the point of use, but rather from the mining activities that are needed to unearth and process the ores from which the TCEs are extracted. Secondly, they acknowledged that the work of a previous generation of environmental chemists, who were largely working on divalent cations (Cd, Cu, Ni, Pb, Zn), benefitted greatly from the careful work carried out by analytical chemists around the world, establishing stability constants for the complexation reactions of these metals with model ligands. They indicated that, unfortunately, this type of work has fallen out of favour and very few new data are being generated. For example, the US NIST database of ‘Critically Selected Stability Constants of Metal Complexes’ was “unceremoniously discontinued” in 2004. As a result, the thermodynamic data needed to predict the speciation of many of the TCEs are incomplete.

2 CRMs and metrology

The certification of the iHg and MeHg mass fractions in a soil certified reference material, EnvCRM 03, has been reported on.²² Both Hg species were quantified by ss-ID-ICP-MS with either on-line HPLC or off-line AEC. The HPLC separations were undertaken using a C18 column with a mobile phase of 0.05% (v/v) 2-mercaptoethanol, 0.4% (w/v) L-cysteine, 0.5% (w/v) ammonium acetate, and 2% (v/v) methanol flowing at 1 mL min⁻¹ for 8 minutes. Two different extraction protocols, a multistage procedure involving H₂SO₄/KBr/Cu_SO₄/DCM/Na₂S₂O₃ or a single step extraction with 6 mol L⁻¹ HCl were used. The former extraction allowed the quantification and separation of iHg and MeHg by HPLC column in one chromatographic run giving mass fractions of 316 ± 10 (U = 3.2%, k = 2) and 0.53 ± 0.02 μg kg⁻¹ (U = 3.8%, k = 2), respectively. The extraction with HCl followed by AEC to separate MeHg from iHg gave an MeHg mass fraction of 0.54 ± 0.03 μg kg⁻¹ (U = 5.4%, k = 2). The method was validated using IAEA-405 marine sediment CRM. This would appear to be the first soil CRM certified for Hg species.

3 Elemental speciation analysis

3.1 Antimony

Most of the Sb speciation work reported in the past year has focused on the determination of the inorganic species Sb^III and Sb^V by non-chromatographic procedures. For the ETAAS analysis of rice wine, Chen et al. devised a procedure in which one species (Sb^V) was extracted by SPE and the other (Sb^III) by LLE.²³ Samples were degassed, diluted 5-fold, filtered and the Sb^V in 20 mL of sample (adjusted to pH 5) was extracted by repeated passage (7 times in 4 min) through the extractant, 40 mg of ZnFe₂O₄ nanotubes packed into the needle hub of a syringe. The trapped Sb^V was dissolved in 4, 300 μL portions of 0.5 mol L⁻¹ HNO₃. The Sb^III in 5 mL was extracted into DDTC (50 μL in 1-dodecanol) by shaking at 40 °C in a 10 mL syringe. After ice bath cooling the solidified organic solvent was removed, thawed, and diluted to 100 μL with EtOH. The LOD values were 4.5 and 3.2 ng L⁻¹ for Sb^III and Sb^V, respectively. The method was validated by the analysis of CRM GSB-07-1376-2001 (water) and by spike recoveries, and both species were found in 3 real samples. For the ICP-MS analysis of several types of waters (reservoir, tap, ground and coastal sea), Portugal et al. devised a FI procedure in which organoSb species were retained on a cation-exchange resin prior to selective HG of Sb^III.²⁴ Following reduction with KI, total Sb was determined by HG, and hence the Sb^V concentration was determined by difference. The researchers did not demonstrate directly that no signal was obtained from Sb^V under the optimised conditions selected but presented results for spike recoveries of both species into 7 different samples at 0.25 and 0.50 μg L⁻¹ that ranged from 90 to 111%. The LOD was 0.02 μg L⁻¹ and the throughput was 30 h⁻¹. Both Sb species were found in all samples. For the selective extraction of Sb^III (and As^III and Se^IV, covered in the sections for these elements) from freshwater and soils, Fang et al. developed a selective in situ DGT sampling procedure.²⁵ Selectivity over other forms of the elements in question was achieved through the incorporation of aminopropyl and mercaptopropyl bi-functionalised mesoporous silica spheres into the binding gel. The passive sampling devices were deployed in the river environments for 72 h and the collected analytes eluted with 3 mL of 1 mol L⁻¹ HNO₃ + 1% KIO₃ (m/v) and quantified by ICP-MS. For the measurement of these species in the rhizosphere of growing rice, the binding gel was deployed in the soil for 24 h after 3 weeks of cultivation and the accumulation of the target species in the gels was measured with LA-ICP-MS. Spatially resolved labile fluxes could be visualised because the 3.4 μm microspheres were homogenously distributed throughout the binding gel. The LOD values were 0.06, 0.02 and 0.12 μg L⁻¹ for As, Sb and Se, respectively. The researchers also took grab samples from the river, which were analysed by HPLC-ICP-MS (details were given in the 33-page ESI). They reported that the time-averaged species concentrations by DGT were reproducible compared with high-frequency sampling and measurement by HPLC-ICP-MS. It is not possible to discern a direct comparison of the results obtained by the DGT method with those of the HPLC-ICP-MS method as the HPLC-ICP-MS method was not able to detect any Sb^III in the grab samples. For Se^IV, 3 out of the 5 samples had concentrations >LOD, whilst As^III was detected in all samples. A plot of the ratio of the amount of As^III determined by DGT to that determined by HPLC-ICP-MS shows values ranging from about 0.7 to about 1.2. It is not possible to find any results for the concentrations of Sb^III in the river water as determined by the DGT method as these are not given anywhere in the article or ESI.

Other areas of research into Sb speciation has included the possible application of XANES for the identification of organoSb compounds in environmental samples was extended by a theoretical analysis of the features in the spectra of Sb^III compounds with two model ligands, tartrate and EDTA.²⁶ The researchers identified the obvious difference in the intensity of the shoulder peak at the white-line region as one spectroscopic feature to be studied. They simulated the spectra with time-dependent density functional theory and assigned the shoulder feature to a metal–ligand charge transfer transition from the Sb 1s orbital to the unoccupied states of the carbon chains. In the EDTA complex, this transition would not be favoured due to the long distance between the ligand and metallic centre in leading to a decrease in intensity of the shoulder peak. In a study of a real environmental application, the removal of Sb^III and Sb^V from mining wastewater by phytosynthesised Fe NPs, a variety of techniques were used.²⁷ The oxidation of Sb^III to Sb^V by Fe^III on the NP surface was confirmed by XPS, HPLC-AFS (no details given) and amperometry; SEM-EDS indicated that both Sb^III and Sb^V were adsorbed onto the NPs. The researchers showed that in a batch extraction the concentration of Sb^III was decreased from 1.48 to 0.1 mg L⁻¹. They also showed that As, Cd, Pb and Sn were also removed, but with variable efficiencies; only Pb, at 95%, showed an efficiency similar to that of Sb (94%).

3.2 Arsenic

The development of methodology for As speciation has reached the stage where quantitative results may be obtained routinely for many species using HPLC-ICP-MS. However, the separations are often lengthy or do not resolve all the species of interest. Quarles et al.²⁸ have developed methodology for the rapid separation of AB, As^III, As^V, DMA, AC, MMA, and TMAO using an automated two column system, with (NH₄)₂CO₃ as the eluent and ICP-MS detection, for the separation of As^III, As^V, DMA, AC, MMA. This separation was optimised to 2 minutes using a gradient step of 0.5 mmol L⁻¹ ammonium carbonate followed by 80 mmol L⁻¹ ammonium carbonate with AEC. In this separation AB and TMAO both eluted in the void volume. Thus, to distinguishing between AB and TMAO a second stage was required. The separation of AB, TMAO, As^III, DMA, AC, MMA, and As^V was achieved using a combination of AEC and a C18 column. The separation was done by sending the sample through the anion exchange column first; the AB/TMAO peak eluted from the column and passed onto the C18 column. Both columns were on switching valves, which allowed for the C18 column to be then switched offline at this point in the method, trapping the AB and TMAO species. The As^III, DMA, AC, MMA, and As^V were eluted from the anion exchange column, followed by a switch back to eluent 1 and the C18 column brought back online and elution of the retained TMAO and AB. TMAO eluted first, due to the As=O bond that makes it slightly more polar than AB. Two options were possible with the method: to bypass column 1 or keep column 1 inline. The advantage of having column 1 inline is that eluent 1 then passed through the anion exchange column providing extra conditioning prior to the next sample. No obvious differences in data quality were detected between the two options when eluting TMAO and AB. The total separation time for the two-column method was optimised to 4.5 minutes. The methods were evaluated for accuracy by analysing proficiency samples from the Centre de Toxicologie du Quebec and New York Department of Health. Correlation between the measured values and reference values was good, with a <4.5% difference in results. The LOD values in a urine matrix ranged from 2.8–6.0 ng L⁻¹ As and 4.1–9.1 ng L⁻¹ As for the one- and two-column methods, respectively. The use of 1,2-hexanediol as a new eluent for HPLC-ICP-MS to avoid the problems associated with organic solvents has been reported.²⁹ 1,2-Hexanediol was found to be compatible with ICP-MS at high flow rates of 1.5 mL min⁻¹ and concentrations of at least 30% v/v, using the standard conditions and instrumental setup normally used with 100% aqueous media. Sensitivity for As was enhanced (improvements were also noted with Br, Cl, P, S, and Se). Concentrations of 1,2-hexanediol at ≤30% v/v were found to offer superior elution strength when compared to concentrations >90% v/v of common organic phases such as MeOH. This greatly decreased the amount of carbon required to elute highly hydrophobic compounds such as lipids and steroids, enabling detection at ultra-trace levels. The proposed approach was applied to the detection of As containing fatty acids in spiked human urine, and LOD values of <0.01 μg As L⁻¹ were achieved, >100-fold lower than those previously reported using conventional organic eluents. A simple and quick procedure has been described for the separation of As species from water samples at the sampling site to prevent changes in speciation prior to determination by ICP-MS.³⁰ For the separation of As^III anions, a C18 column was saturated with APDC. Acidification of the water sample to pH 2 ensured 100% retention of As^III anions and effective separation from As^V, MMA and DMA. The C18 columns modified with APDC could be prepared and stored in the laboratory for up to 24 hours before use on the sampling site. The influence of Cd, Cr, Pb, and Tl cations was limited by the addition of EDTA and anions such as SO₄²⁻ and PO₄³⁻ did not significantly affect the efficiency of As^III anion retention. The determination and speciation of trace iAs species in water samples using a batch adsorption process on metal–organic framework mixed-matrix membranes (miniaturised MIL-101(Fe)) and ED-XRF spectrometry has been reported.³¹ The quantitative adsorption of As^V was achieved at pH 3–6 from 30 mL samples and 120 min of equilibrium time employing a membrane with a monolayer adsorption capacity of 1.953 mg g⁻¹. The adsorbed As^V was determined directly by ED-XRF with a LOD of 0.094 μg L⁻¹. Water samples with 3 levels of As^V (5, 10 and 50 μg L⁻¹) gave recoveries in the range of 98 to 105%.

Research into As speciation in foodstuffs remain popular. Most studies use some form of analyte extraction from the food matrix, but the use of X-ray spectroscopies to provide in situ, non-destructive analyses of solid samples without perturbation to the As species therein can be attractive. Jahrman et al. have evaluated the use of As K-edge XANES measurements collected from 3 CRMs (BCR 627 tuna fish, CRM 7405b seaweed and SRM 3232 kelp powder which is value-assigned for As species).³² Traditionally, when using this technique, studies have relied on linear combination fitting of a minimal subset of empirical standards selected by stepwise regression. However, this is known to be problematic for compounds with collinear spectra and can affect the accuracy of the analysis. In this study the LASSO regression method was used to reduce the risk of overfitting and increase the interpretability of statistical inferences. As this is a biased statistical method, results and uncertainties were estimated using a bootstrap method accounting for the dominant sources of variability. This method does not separate model and data selection from regression analysis, and the authors present a survey of spectral influences including changes in the state of methylation, state of protonation, oxidation state, coordination geometry, and sample phase. The quantification of As species in wheat flour using MAE with IC-HR-ICP-MS has been reported.³³ An aliquot of each sample (ca. 0.5 g) was accurately weighed into a quartz capped vessel and 10 mL of H₂O was added as the extractant prior to a three-step MAE program. The resulting mixtures were allowed to cool to room temperature, transferred to 50 mL centrifuge tubes, diluted to the volume with H₂O and centrifuged at 3500 rpm for 10 minutes. The supernatants were then filtered (0.45 μm). The As species in the extracts were separated by IC using a gradient elution method with 0.5 mmol L⁻¹ HNO₃ (pH 3.4) and 50 mmol L⁻¹ HNO₃ (pH 1.4). Three CRMs, NIST SRM 1568b (rice flour), NMIJ CRM 7533-a and ERM BC211 (both brown rice flour) were used to evaluate the method. The LOD values for As^III, As^V and DMA were between 0.3–2.6 pg g⁻¹ and the LOQ values were 1.1–8.6 pg g⁻¹, respectively. The concentrations of the iAs in the South African wheat flour samples used in the study were low (6.8 to 17.8 ng g⁻¹).

Continuing with the speciation of As in rice, Raab et al.³⁴ have identified a previously unreported As species, dimethylarsonyl-dimethylarsinic acid, during a routine survey by the US FDA. The compound was identified using a combination of HPLC-ICP-MS and HPLC-ES-MS and chemical derivatisation. The origin of the species in the sample was unknown. Most studies of As speciation in rice report only iAs and DMA. Although DMMTA is also occasionally reported, Dai et al.³⁵ have found that it is widely detectable in rice worldwide and they suggest that it is often unknowingly determined as DMA. Since DMA is far less toxic than DMMTA such errors have implications with respect to food safety. Using an enzymatic extraction followed by HPLC-ICP-MS analysis with a C18 column, the DMMTA was detected in rice grains (n = 103) from China and in polished rice grains (n = 140) from a global market-basket survey. The concentration of DMMTA ranged from <0.20 to 34.8 μg kg⁻¹ (median 10.3 μg kg⁻¹), accounting for up to 21% of total As. A strong linear correlation was observed in all rice samples between DMA and DMMTA (30 ± 8% of DMA) concentrations. This robust relationship allows an estimation of DMMTA in rice grains from the DMA data reported in previous market-basket surveys, showing a global geographical pattern with DMMTA concentration increasing from the equator toward high-latitude regions. A UAE enzymatic hydrolysis method has been reported for As species in rice flour.³⁶ Several solvents and techniques were evaluated to extract As^III, As^V, MMA, DMA, AB and AC from the rice flour CRM NMIJ-7532a prior to determination of the As species by HPLC-ICP-MS. The results indicated that UAE at 60 °C for 1 h and then heating at 100 °C for 2.5 h in an oven with a thermostable alpha-amylase aqueous solution was effective in liberating the As species. The recoveries of iAs and DMA in NMIJ-7532a were 99.7% ± 1.6% (n = 3) and 98.1% ± 2.3% (n = 3), respectively. The extraction method effectively maintained the integrity of the As species. Under optimum conditions, the LOD values obtained were 0.47 ng g⁻¹, 1.67 ng g⁻¹ and 0.80 ng g⁻¹ for As^III, As^V, and DMA, respectively with LOQ values of 1.51 ng g⁻¹, 5.34 ng g⁻¹ and 2.57 ng g⁻¹ for As^III, As^V, and DMA, respectively. There have been further studies on the influence of irrigation methods on As speciation in rice grain.³⁷ The concentrations of As^III, As^V, DMA, and MMA in grains from 26 different rice genotypes irrigated either with continuous flooding, periodic saturation or sprinkler irrigation were reported. Speciation was performed using IC-ICP-MS and PCA was also performed on a dataset of 78 rice samples (26 for each irrigation method) and 5 variables i.e., the total amount of As and the percent amounts of As^III, of As^V, of DMA, and of other As species, respectively. The last variable was the difference between total extracted As and the sum of the As species measured in the IC-ICP-MS analysis. Since MMA was always found below the LOD for all irrigation methods, it was not used in the statistical analysis. The results indicated that rice irrigated by continuous flooding was mainly described by high amounts of total As and DMA, whereas As^III was the only representative species to describe rice irrigated by periodic saturation, and As^V the sprinkler irrigated rice. A number of regional studies focusing on As in rice have also been reported. Rice and soil samples collected from 16 different districts (non-mining and mining sites) of Khyber Pakhtunkhwa (Pakistan) were assessed with respect to potential human health risks.³⁸ Total As concentrations were determined by ICP-MS, whilst As^III, As^V, AB, DMA, and MMA were determined by HPLC-ICP-MS. The results showed higher concentrations of iAs than organic As species in rice, and rice samples collected from the mining districts had higher value of As (0.28 mg kg⁻¹ of total As) as compared to non-mining areas (0.072 mg kg⁻¹ of total As). Arsenic speciation in rice grains collected from wet, intermediate and dry zones in Sri Lanka have also been studied.³⁹ Field rice (brown rice) and market rice (polished rice) samples were analysed for their total elemental profile (10 elements) and As speciation using ICP-MS and IC-ICP-MS, respectively. The total amounts of the 10 elements determined was found to vary across climatic zones. The major As species detected in all field collected samples was iAs, DMA was detected as the major organic As species and MMA was below the LOD. The maximum As content in field rice was 0.258 mg kg⁻¹ with only 0.3% of field rice samples exceeding the maximum allowable limit for iAs in brown rice (0.25 mg kg⁻¹). However, 18% of the field samples exceeded maximum recommended level of As in infant food (0.1 mg kg⁻¹). None of the market rice samples exceeded the maximum tolerable iAs levels in white rice (0.2 mg kg⁻¹).

The speciation of As in seafood also continues to be a popular area of research. A method for the simultaneous speciation of total As, As^III, As^V, AB, AC, MMA and DMA in seafood has been reported⁴⁰ and MAE was used for the isolation of As species. The sample first extracted with a MeOH solution (3 : 1, v/v) with shaking (800 rpm, 15 min). Subsequently, the tubes were placed into PTFE vessels, half filled with water and subjected to a multistep MAE procedure then, after cooling, centrifuged, the supernatants removed and evaporated to about 0.5 mL at 70 °C under N₂. The evaporated extracts were then dissolved to form a 10 mL solution with H₂O and filtered. The separation and quantification of analysed compounds were performed by IC-ICP-MS in one chromatographic run using (NH₄)₂CO₃ based buffers as the eluent. The results of validation and proficiency tests confirmed the reliability, robustness, and applicability of the developed procedures for various types of matrices. Shrimp is one of the major sources of As for humans. A comparison of wild and farm shrimp in Brazil has found that wild shrimp have As concentrations an order of magnitude higher than farmed shrimp.⁴¹ Samples of wild (Farfantepenaeus brasiliensis) and farmed shrimps (Litopenaeus vannamei) from NE Brazil were fractionated into subsamples of carapace, muscle tissue and viscera. The whole shrimp as well as the animal tissue fractions were decomposed using MAE and the total As was determined by ICP-MS. The water-soluble As species were also extracted and then identified using HPLC-ICP-MS with an anionic and cationic column. The total As in wild shrimp samples exceeded Brazilian and USA food legislation by one order of magnitude, with concentrations of 11.5 ± 0.5 mg kg⁻¹, while farmed shrimp had significantly lower total As levels (0.53 ± 0.09 mg kg⁻¹). More than 60% of the As was in the edible fraction in the wild shrimp, while in farmed shrimp this was less than 50%. The speciation analysis showed that AB was the predominant As form and iAs was below the Chinese legislation levels (iAs <0.50 mg kg⁻¹) for shrimp in both species. The authors conclude that legislations that consider only total As were not appropriate to assess the toxicity of As in seafood. The speciation of As in sexually mature Chinese mitten crabs (Eriocheir sinensis) has been reported.⁴² Total As, As^III, As^V, AB, AC, DMA, and MMA were determined by HPLC-ICP-MS. The influence of three common cooking methods on the speciation of As and the concentrations in different edible parts of the crabs was then explored. The bioaccessibility of the As species in the crabs was also studied by a simulated in vitro gastrointestinal digestion. The results detected changes in the As content, and the authors speculate that there was a morphological transfer and morphological transformation in different parts of the crab during cooking and gastrointestinal digestion, thus increasing the likelihood of As consumption. The Target Hazard Quotient of As was less than 1, indicating no significant health risk, but the risk of carcinogenesis of As should not be overlooked.

The speciation of As in seaweed has continued to attract attention this year. A study describing the optimisation of the extraction conditions for the determination of As^III, As^V, AB, DMA, and MMA in Kappaphycus alvarezii, a carrageenan-producing red seaweed has been reported.⁴³ The study used HPLC-ICP-MS to evaluate a series of HNO₃ extraction solutions (0 to 2.0%) with various extraction times (0 to 240 min) at 90 °C. The proposed method was successfully validated using CRM NMIJ 7405-a (Hijiki seaweed) and sample spiking experiments. Matrix effects were also studied. Optimum results (86.8 to 94.2% recovery) were obtained using 0.2% HNO₃ and an extraction time of 60 min. Recoveries were between 88 and 107% for all 5 As species. As the largest seaweed producer, China contributes about 60% of the global seaweed production and a study investigated As species in 20 seaweed species collected from representative seaweed farming sites in the 6 provinces along the Chinese coastline.⁴⁴ The seaweeds included Saccharina japonica, Undaria pinnatifida, Neopyropia spp., Gracilaria spp., and Sargassum fusiforme, listed as the most consumed seaweeds by the FAO. Determination of AB, AC, MMA, DMA, As^III, and As^V and the arsenosugars AsSug-OH, AsSug-PO₄, AsSug-SO₃ and AsSug-SO₄ was by HPLC-ICP-MS-MS. The As species extraction was by a previously published method based on H₂O, although the extraction efficiencies were generally poor. The average extraction efficiency for brown algae species Saccharina japonica and Sargassum were above 60%, while the extraction efficiency in U. pinnatifida and P. australis was only approximately 30%. Higher extraction efficiency was found from Neopyropia spp. and G. furcate among the red algae, whereas the extraction efficiency for other red algae were all <33%. A low extraction efficiency was also found for green and blue algae, 16.3–24.0% and 42.3%, respectively. The total As concentration ranged from 15.3 to 150 mg kg⁻¹ for brown algae, 2.2–39 mg kg⁻¹ for red algae, and 1.5–28.3 mg kg⁻¹ for green algae. The estimated daily intake of iAs via seaweed consumption was generally below the EFSA CONTAM Panel benchmark dose lower confidence limit (0.3 μg per kg bw per day) except for all Sargassum species where the EDI was significantly higher. Obtaining reliable speciation data from algae is often hindered by the lack of available of suitable arsenosugar standard materials and chemical syntheses of such compounds have been reported to be complex and tedious. Morales-Rodriguez et al.⁴⁵ investigated the feasibility of using AE SPE cartridges (DSC-SAX and DSC-NH₂) as an alternative for the isolation and preconcentration of arsenosugars (PO₄-Sug, SO₃-Sug and SO₄-Sug). The effect of pH, ionic strength, type of salt and solvent on the elution protocols of the arsenosugars was studied. The eluted solutions from the SPE cartridges were analysed by ICP-MS for total As and IC-ICP-MS for As speciation. The developed SPE procedure facilitated the collection of a solution containing the three arsenosugars isolated from other As species with recoveries over 75% for SO₃-Sug and SO₄-Sug, and around 45% for PO₄-Sug.

The speciation of As in other marine based organisms has also received attention. Arsenic species in mesopelagic organisms and their fate during aquafeed processing has been studied.⁴⁶ Single-species samples of the five most abundant mesopelagic organisms in Norwegian fjords were used. In addition, As species were studied in mesopelagic mixed biomass and in the resulting oil and meal feed ingredients after lab-scale feed processing. Water-soluble As species were determined by IC-ICP-MS. This was supplemented by extracting the arsenolipids and determining the total As in this fraction. The dominant As form in mesopelagic crustaceans and fish species was AB, accounting for approximately 70% and 50% of total As, respectively. Other water-soluble species were present in minor fractions, including iAs which, in most samples, was below the LOQ. The fish species had a higher proportion of arsenolipids, approximately 35% of total As, compared to crustaceans which contained 20% on average. The feed processing simulation revealed generally low levels of water-soluble As species besides AB, but considerable fractions of potentially toxic arsenolipids were found in the biomass, and transferred to the mesopelagic meal and oil.

Environmental studies focusing on As speciation also continue to be published. Arsenic is commonly sequestered at the sediment–water interface in mining-impacted lakes through adsorption and/or co-precipitation with authigenic Fe-(oxy)hydroxides or sulfides. A report by Miller et al.⁴⁷ found that the accumulation of organic matter in near-surface sediments also influences the mobility and fate of As in sub-Arctic lakes. Sediment gravity cores, sediment grab samples, and porewater were collected from three lakes downstream of the former Tundra Au mine, Northwest Territories, Canada. Analysis of sediment using combined μXRF, μXRD, XANES, and organic petrography showed that As was associated with both aquatic (benthic and planktonic alginate) and terrestrially derived organic matter (e.g. cutinite, funginite). Most As was associated with fine-grained Fe-(oxy)hydroxides or sulfide minerals (e.g., goethite, orpiment, lepidocrocite, and mackinawite). However, grain-scale synchrotron-based analysis shows that As was also associated with amorphous organic matter. Mixed As oxidation states in porewater (median = 62% As^V, 18% As^III; n = 20) and sediment (median = 80% As-I and As^III, 20% As^V; n = 9) indicated the presence of variable redox conditions in the near-surface sediment and suggested that post-depositional remobilisation of As had occurred. Detailed characterisation of the As-bearing organic matter at and below the sediment–water interface suggested that organic matter played an important role in stabilising redox-sensitive authigenic minerals and associated As.

The speciation of As in soils and sediments has been further studied this year. Investigating the reaction mechanisms that control the mobility of nutrients and toxic elements in soil matrices is confounded by complex assemblages of minerals, non-crystalline solids, organic matter, and biota. A study of the chemical elements and solids that contribute to As binding in matrices of soil samples from different pedogenic environments at the micrometer spatial scale has been reported.⁴⁸ Arsenic was reacted with and imaged within thin weathering coatings on eight quartz sand grains separated from soils with different drainage and varying levels of Fe and Al (hydr)oxides, OC, and other elements. The grains were analysed using μXRF imaging and μXANES spectroscopy before and after treatment with 0.1 mmol L⁻¹ As^V solution. Partial correlation analyses and regression models developed from multi-element μXRF signals collected across 100 × 100 μm² areas of sand-grain coatings inferred augmenting effects for Cu, Fe, Mn, Ti and Zn on As retention. Significant partial correlations (r² > 0.11) between Fe and Al from TOF-SIMS analysis of most samples suggested that Fe and Al (hydr)oxides were partially co-localised at the microscale. Linear combination fitting results for As K-edge μXANES spectra collected across grain coatings typically included >80% of As^V adsorbed on goethite, along with varying proportions of As^V adsorbed on boehmite, As^III or As^V bound to Fe^III-treated peat, and DMA. Overall, the results inferred a dominance of Fe and possibly Al (hydr)oxides in controlling As immobilisation, with variable contributions from Zn, Ti, Cu, or Mn, both across the coating of a single sand grain and between grains from soils developed under different pedogenic environments. Thioarsenic species are known to exist in transitional environments with varying redox potential such as hot springs and paddy fields, where As^III, sulfide and thioarsenites may be exposed to ambient air. A study using kinetic experiments and computational information on the formation mechanisms of thioarsenic species from the sulfide-driven oxidation of As^III under an air atmosphere at typical groundwater pH (6.5) has been presented by Zhang et al.⁴⁹ The determination of non-thiolated As and thioarsenate species as well as indirect identification of thioarsenites at environmentally relevant As concentrations was by IC-HG-AFS. The results showed that the ambient air facilitated the thiolation and oxidation of As^III, during which thioarsenites were found as intermediates. The activity towards oxidation was in the order: monothioarsenite > dithioarsenite > trithioarsenite > As^III. The thiolation of As^III decreased with pH while the oxidation of thioarsenites by air increased with pH. In a second paper by the same group,⁵⁰ the species and fate of As with respect to sulfide (S^II) in anaerobic and sulfidic environments was reported. The mechanisms and kinetics of As^V reduction by S^II at different pH was studied. The S^II : As^V molar ratios, and the initial As^V, As^III, and mono, di and tri thioarsenics concentrations in the presence or absence of Al-hydroxide were determined by LC-AFS. In most of the samples As^V–S^II formed at pH 5 to 7, the percentage of thioarsenic showed an increase and then decrease during the reaction process, indicating that thioarsenics may act as intermediates in this reaction. In high S^II conditions, di and tri thioAs^V were identified which may be convert to amorphous As₂S₃. At pH 9, thioarsenic intermediates were not observed, indicating a direct reduction of As^V to As^III which was also the case when the initial As concentration was low. The pH, S^II : As^V molar ratio as well as the initial As^V concentration were thus the major factors controlling the reaction rate. The Al hydroxide acted as an acid-catalyst enhancing the generation of thioarsenic and then increasing the reaction rate at slightly acidic to neutral pH.

The use of honey bees as potential biomarkers for As contamination has been reported by Zaric et al.⁵¹ Four different procedures were evaluated in triplicate for As species extraction from the honey bees, H₂O at room temperature and 90 °C, 20% v/v MeOH and 1% v/v formic acid. For each extraction, 2.00 mL of extraction solvent was added to approximately 100 mg of sample followed by UAE for 15 min. For As speciation analysis, H₂O₂ (10% v/v) was added to an aliquot of each filtered sample. These mixtures were shaken and put in an oven for one hour at 45 °C to convert any labile trivalent or thiolated As species in the sample to As^V and speciation was by HPLC-ICP-MS. Two CRM materials were also evaluated (NCR BOVM-1 bovine muscle and NIST SRM 1640a, trace elements in natural waters). It was found that the procedure using water and UAE at 90 °C for 1 h offered the best extraction efficiency (>90%). The study found that the highest As concentrations were determined in the vicinity of coal fired thermal power plants (367 μg kg⁻¹), followed by an urban region (213 μg kg⁻¹), an industrial city (28.8 μg kg⁻¹) and rural areas (41 μg kg⁻¹). The iAs accounted for 95% of As species in bees from most locations, the exception being the industrial city sites, where it represented only 80% of As species with 15% present as DMA.

The speciation of As in other biological matrices has also been reported on this year. A method for separating As species using a 5 μL sample loop combined with a capillary IE column and ICP-MS has been developed for analysing small volume biological samples.⁵² Although the chromatographic resolution was not high enough to separate As^III, As^V, DMA, MMA, the capillary column could be used to separately quantitate total iAs (As^III + As^V) after oxidation of As^III to As^V with H₂O₂ prior to IC. The LOD was 0.13 μg kg⁻¹ for iAs dissolved in water, inferior to conventional columns with a 100 μL sample loop in IC-ICP-MS analysis (0.033 μg kg⁻¹). However, in terms of mass, the LOD of the capillary column method was found to be at the sub-picogram level, better than that of the conventional column method. The developed method was then successfully validated using a water SRM (NIST SRM 1643f, Trace Elements in Water) and demonstrated good recovery efficiency (101%). Finally, the developed method was used to determine iAS in a human urine SRM (NIST SRM 2669, Arsenic Species in Frozen Human Urine) and the iAs and organoarsenic species were well separated in the human urine. The measured concentration of total iAs was 4.69 ± 0.47 μg kg⁻¹, within the experimental uncertainties of the certified value, 3.88 ± 0.40 μg kg⁻¹. The speciation of As (As^III, As^V, DMA, MMA, AB and AC) in biological matrices (crayfish, laver and kelp) using ES-HG-MPT-MS has been reported.⁵³ The novel ionisation source of MPT combined with HG used an in-house made sample introduction device. The device resulted in unique mass spectrum behaviour for As and the authors reported improvements in sensitivity and matrix tolerance. The LOD values for As^III, As^V, MMA, DMA, AB and AC were 0.01, 0.02, 0.08, 0.08, 0.05 and 0.02 μg kg⁻¹ respectively. A spike recovery experiment was also conducted, showing a recovery of 94 ± 11%, with an RSD of <10%. The As hyperaccumulator fern Pteris vittata has been widely studied as it is an ideal plant for the phytoremediation of As-contaminated soil. The micro-distribution pattern of As in spores from Pteris vittata have been investigated using SR-XRF microprobe analysis and micro-speciation with SR-XAS.⁵⁴ The distribution of As was then compared to that of Ca, Fe, K, S and Zn and it was found that the distribution of As, Ca and S were similar indicating some form of interaction. The micro speciation analysis indicated that the main species of As in spores and sporangium was As^III. Considering that As^III has higher mobility and toxicity than As^V in most organisms, the authors suggest that the results indicated a high tolerance to As by P. vittata throughout its life cycle. The speciation of atmospheric As adsorbed on pollen (Vachellia caven) and other aerobiological samples before and during the COVID-19 pandemic period has been reported.⁵⁵ The As species adsorbed on pollen samples were extracted utilising 1% acetone and UAE for 15 minutes followed by separation and detection of the As species by AEC-ICP-MS. The As^III and As^V concentrations ranged from 0.08 to 0.62 μg g⁻¹, and 0.33 to 0.89 μg g⁻¹ respectively. A comparison of the data collected during 2019 and 2021 in the city (San Luis, Argentina), showed a marked decrease at times of lower traffic intensity. The method LOD and LOQ were 0.01 and 0.04 μg g⁻¹ for As^III; and 0.01 and 0.06 μg g⁻¹ for As^V, respectively for 0.05 g of pollen. The precision RSD was 4.1% (n = 10).

A number of studies based on clinical applications of As speciation have been published. A large scale study to determine the levels and distribution of As species (As^III, As^V, MMA, DMA, AB, and AC) in urine and blood, as well as to investigate the methylation efficiency and related factors in the Korean population has been reported by Choi et al.⁵⁶ The study used a total of 2025 urine samples and 598 blood samples obtained by the Korean Ministry of Food and Drug Safety. The 6 As species were determined using UHPLC-ICP-MS. Multiple linear regression models were used to examine the relationship between As species (concentrations and proportions) and covariates. The most prevalent species in urine and blood was AB. The relative composition of iAs, MMA, DMA, and AC in urine and blood differed significantly. Consumption of blue-backed fish was linked to higher levels of AB in urine and blood. Drinking water and multigrain rice consumption were associated with increased iAs concentration in urine. Except for iAs, every species had correlations in urine and blood in both univariate and multivariate analyses. Adolescents and smokers presented a lower methylation efficiency (higher % MMA and lower % DMA in urine) and females presented a higher methylation efficiency (lower % iAs, % MMA, and higher % DMA in urine). The study concluded that due to the correlations and differences in As species distribution between urine and blood, each medium may have different role as an As biomarker. Arsenic methylation capacity has also been studied by Jiang et al.⁵⁷ These workers measured urinary concentrations of iAs, MMA, and DMA using HPLC-HG-AFS and calculated the primary methylation capacity index and secondary methylation capacity index in 209 university students in Hefei, China, a non-As endemic area. Volunteers were given a standardised questionnaire asking about their sociodemographic characteristics. Bayesian kernel machine regression analysis was used to estimate the methylation indices. The median concentrations of iAs, MMA, and DMA were 1.22, 0.92, and 12.17 μg L⁻¹, respectively and the proportions of iAs, MMA, and DMA were 8.76%, 6.13%, and 84.8%, respectively. It was found that most of the students had been exposed to As although the levels of iAs, MMA, DMA, and total As differed by sex. The females seemed to have stronger As methylation capacity than males. The application of nanoscale SIMS analysis to define the subcellular pharmacokinetics of the cytotoxic chemotherapy drug As trioxide has been reported.⁵⁸ The aim of the study was to gain insights into the mechanisms underlying ATO efficacy. It was found that ATO-treated cells revealed As accumulated in the nucleolus. After prolonged ATO exposure, 40 nm As- and S rich protein aggregates appeared in the cell nucleolus, nucleus, and membrane-free compartments in the cytoplasm. The study suggested that the partitioning of nanoscale aggregates could be relevant to cell survival. An asymmetric serpentine microfluidic device with high sampling efficiency and good focusing performance has been developed for single-cell focusing.⁵⁹ The device coupled with ICP-MS enabled single-cell assay to provide information on for As^III uptake by HepG2 cells. The study investigated the heterogeneity of cellular As distribution in order to elucidate the As elimination behaviours in single HepG2 cells. The metabolism and transformation of As^III in HepG2 cells was also tracked using CE-ICP-MS. The results for single-cell analysis and As elimination kinetics showed that the half-life of As elimination is 0.9 ± 0.04 h with the elimination constant of 0.77 ± 0.03, i.e., 77% of accumulated As in HepG2 cells may be eliminated per hour. The study also found that AB was the main metabolite and biotransformation species of As in HepG2 cells. A study to investigate the association of As levels with blood pressure levels and prevalence of hypertension among general non-Hispanic Asians has been presented by Tang et al.⁶⁰ The study used the 2011–2018 US National Health and Nutrition and Examination Survey data. Participants were aged 20 years and older were tested for total As and DMA in blood and urine samples. The As species As^III, As^V, DMA, MMA, AB and AC were determined by HPLC-ICP-MS with LOD values of: As^III 0.12 μg L⁻¹; As^V 0.79 μg L⁻¹; AB 1.16 μg L⁻¹; AC 0.11 μg L⁻¹; DMA, 1.91 μg L⁻¹ and MMA 0.20 μg L⁻¹. Systolic and diastolic blood pressure levels were examined through a standardised protocol. Censored normal regression model and logistic regression model were employed to explore the associations of As and other metals. The study found that urinary As was associated with increased diastolic blood pressure levels. The potential use of toenails as a biomarker for As related disease has been reported.⁶¹ The primary objectives of the study were to (1) establish an analytical method for As speciation analysis in toenails, (2) describe preliminary As speciation profiles of toenail samples from individuals with skin, lung, bladder, and kidney cancer, type II diabetes, and no known disease, and (3) determine if these speciation patterns differ between disease groups to inform the feasibility of subsequent research. A total of 60 toenail samples and baseline questionnaire data from the US Atlantic Partnership for Tomorrow's Health (Atlantic PATH) study were used. Arsenic speciation profiles were determined using HPLC-ICP-MS. No differences in total As were found between groups although As speciation profiles were significantly different between certain cancer groups and the reference group with no known disease. The percentage of MMA was found to be significantly higher in the toenails of individuals with lung cancer and kidney cancer, compared to healthy individuals with similar total As exposure. The authors conclude that toenail As speciation analysis is feasible and could potentially have important implications for research on As-related diseases.

The assessment of As speciation in Chinese medicines continues to be popular. Realgar, a traditional Chinese medicine containing As₂S₂, has been used for hundreds of years. Today, NiuHuangJieDu tablets (NHJDT) are one of the most commonly prescribed Realgar-containing preparations and are used for the treatment of sore throat, swelling, and aching of gums. Wu et al.⁶² have assessed the health risk associated with NHJDT in a study involving healthy volunteers following single and multiple dose oral administration. Blood, plasma, and urine samples were collected and total As and As speciation (As^III, As^V, DMA, and MMA) determined by HG-AFS and HPLC-HG-AFS, respectively. No significant differences in total As were observed in human blood, and no traces of As were found in human plasma. After dosing, DMA was found to be the predominated As species in human urine. It was suggested that a single dose of NHJDT may be relatively safe but long-term medication may still pose health risks due to the accumulation of As in blood and its slow rate of excretion. Medicinal earthworms are consumed widely in China, but the potential health risks from As are unknown. Li et al.⁶³ have reported on work to investigate the total concentration, bioaccessibility, and speciation of As in earthworms by ICP-MS and HPLC-ICP-MS in order to evaluate the potential health risks to humans. Arsenic was found in all earthworms at concentrations ranging from 0.4 to 53.6 mg kg⁻¹. The bioaccessibility of As varied significantly and ranged from 12 to 69%, with As^III and As^V as the predominant species. Small amounts of AB were also found. The estimated daily intake dose, hazard quotient, and carcinogenic risk of As in most of the samples exceeded the safe threshold level. Results from this study indicated that the potential health risks by the consumption of earthworms may not be negligible and the authors suggest regulatory limits on their use.

3.3 Cadmium

A paper reports on the capability of the hyperaccumulator plant Noccaea caerulescens to phytoextract Cd in Cd-contaminated solids.⁶⁴ In this broad and detailed study 29 lines of N. caerulescens were employed and grown in soils contaminated from the discharge of a chemical industry, a steel plant and sewage irrigation. Plants were grown in greenhouse under controlled conditions of light and temperature. After 6 months of growing, the leaves of the plants were selected, dried, and analysed by FAAS for total Cd concentration and X-ray absorption spectrometry (XAS) for Cd speciation and elemental distribution. Significant differences were found among the different plant lines evaluated. The EXAFS analysis identified the presence of Cd-thiol and Cd-carboxyl complexes in the leaves of high Cd-accumulating plant lines whereas Cd-thiol and Cd-phytate/phosphate complexes were detected in leaves of low Cd accumulating lines. Cd-thiol complexes were the most dominant Cd-species in the plants accounting for 60–70% of total cadmium content.

3.4 Chromium

There has been a significant decrease in the number of papers published in the current review period describing Cr speciation analysis compared with that of the previous review. About two-thirds of the papers in the current review have emanated from just two research groups. The presence of Cr^VI in foods is still being reported, despite the compelling evidence to the contrary discussed in last year's ASU.¹

The majority of Cr speciation procedures described in the past year have featured HPLC-ICP-MS. The group of Milacic and Scancar has reported on speciation in wine and beer,⁶⁵ speciation and bioimaging in dandelions,⁶⁶ the simultaneous speciation of arsenate, chromate, and molybdate in lysimeter water,⁶⁷ and the reduction of Cr^VI by bacteria in tannery effluent.⁶⁸ In each of these studies, Cr speciation was carried out by an HPLC-ICP-MS procedure, originally described in 2012, in which analytes were separated on a SAX fast-protein column (Mono Q HR 5/5) that had been conditioned with 1.0 mol L⁻¹ NaCl. Gradient elution from 100% water to 100% 0.7 mol L⁻¹ NaCl was used in all studies, except those featuring the dandelions, for which the gradient was 100% water to 100% of 1.0 mol L⁻¹ NH₄Cl. The report of the wine and beer study⁶⁵ started with a critical evaluation of the presence of Cr^VI in both matrices reported by other workers that the authors of the present study considered to be erroneous. They cited several of their own previous reports of investigations in which enriched stable isotopic tracers of Cr were deployed in speciation analysis by HPLC-ICP-MS; a procedure that significantly contributes to the validity of the analytical data. The behaviour of Cr^III and Cr^VI was followed by adding ⁵³Cr^III and ⁵⁰Cr^VI stable isotope tracers to the wine and beer samples, and the researchers concluded that none of the samples (3 red wines, 3 white wines, 2 alcoholic beers, 2 non-alcoholic beers and 2 radlers, purchased on the Slovenian market) contained Cr^VI, as the concentrations were below the LOD of 0.06 ng mL⁻¹. Total Cr concentrations, determined by ICP-MS following MAE with HNO₃ and H₂O₂, ranged from 1.8 to 22.1 ng mL⁻¹. For Cr speciation in dandelions,⁶⁶ the researchers synthesised a number of complexes with low molecular weight acids (aconitate, citrate, malate, oxalate and quinate) and then identified the species in the plant extracts (water at 70 °C for 6 h) by retention time matching and from the mass spectra obtained (off line) with high resolution ES-MS. Species were quantified by both internal standardisation with Rh and with species-unspecific IDMS with ⁵³Cr^III, both of which were added post-column. Total Cr was determined by ICP-MS following MAE with 4 mL of HNO₃, 1 mL of HCl, and 2 mL of HF. After cooling, 12.5 mL of 4% H₃BO₃ were added to dissolve the insoluble CaF₂ and to complex the excess HF. The procedure was validated by the analyses of CRM SPS-SW1 (surface water), BCR CRM 320R (trace elements in river sediment), and NIST SRM 1573a (tomato leaves). The sum of species found (LOD values 0.3–0.4 ng mL⁻¹) was 60–76% of the total Cr concentrations for plants grown in soils with a variety of Cr^III and Cr^VI amendments. Additionally, spatially resolved studies of leaves with LA-ICP-MS showed that Cr was localised mainly at the apices, indicating that Cr species were relatively mobile during the transport within the leaf. For speciation of As, Cr and Mo in lysimeter waters,⁶⁷ in support of a study of the leaching of these elements from 3 geotechnical composites made of recycled waste in both compacted and uncompacted (20 times less dense) forms, the same HPLC column was used, though this time the eluent was a gradient from 0 to 0.7 mol L⁻¹ over 10 min at a flow rate of 1.5 mL min⁻¹. The samples were adjusted to pH 12 and 100 μL was injected. Without any further optimisation, almost baseline resolution of the peaks for arsenate, chromate and molybdate was obtained. The spectrometer was operated in He collision cell mode and Ge, In, Rh, Sc were added to the mobile phase at concentrations of 100 ng mL⁻¹ as internal standards. Total element concentrations were also measured by ICP-MS, the procedure for which was validated by the analysis of SPS Spectrapure Standards AS (Oslo, Norway) SPS-SW1 (quality control material for surface water analysis). The determination of chromate was validated by the analysis of Merck CRM Cr Standard Solution (0.050 ± 0.002 mg L⁻¹ as K₂CrO₄ in H₂O) and by spike recovery, which was also used for the other two species. The LOD values were 0.16, 0.16, and 0.31 ng mL⁻¹ for arsenate chromate, and molybdate respectively. The researchers found that both the degree of compaction and the composition of the various composites studied both have significant effects on the leaching of the 3 species and much of the report is given over to a discussion of the relevant chemistry. Essentially the same speciation procedure was used to monitor Cr^VI in a study of the reduction of Cr^VI by bacteria isolated from tannery effluent.⁶⁸ Total Cr was also determined, and the same procedures were used for validation of the methods. The bulk of the report was devoted to (a) the isolation and growth of the microbial communities and (b) the speed and extent to which Cr^VI was reduced by them. The researchers concluded that the ability of the enriched microbial community and the isolated bacterial strains to reduce Cr^VI indicated potential for the rapid bioremediation of contaminated wastewaters. It should perhaps be noted that the gradient elution procedure required almost the same time for regeneration as for analysis (about 10 minutes), and that in the case of the plant samples, the column was cleaned after every 10 injections by a procedure that took well over an hour.

Davis and Wise discussed the possible role of HPLC-ICP-MS to analyse leather following the lowering of the limit for Cr^VI in leather to 1 mg kg⁻¹, as suggested in a consultation document on skin sensitising substances from the European Chemicals Agency.⁶⁹ They demonstrated that the current limit of 3 mg kg⁻¹ is a challenge for both the colorimetric (part 1) and chromatographic with spectrophotometric detection (part 2) methods described in BS EN ISO 17075. They suggested that the IC separation already set out in BS EN ISO 17075-2 should be interfaced with ICP-MS detection to enhance detection limits. The same researchers, together with Italian collaborators, have devised a new HPLC-ICP-MS procedure⁷⁰ with separation on a Hamilton PRP-X100 (25 cm × 4 mm i.d.) anionic column for the determination of Cr^VI in particulates rich in Cr^III collected as the inhalable fraction in the breathing zone of selected workers in the leather industry. A key feature of the work was the use of speciated isotope dilution (⁵³Cr^III and ⁵⁰Cr^VI), by which they showed that during the analysis of particulates according to the NIOSH 7600 protocol, significant oxidation of Cr^III to Cr^VI occurred. Their new procedure, a 48 hour extraction at room temperature with a pH 8 phosphate buffer, produced no interconversions provided a redox buffer was added. In addition to this novelty, they also dealt with the problem of detecting Cr^VI in the presence of a large excess of Cr^III by precipitating the Cr^III as Cr(OH)₃ and trapping it on the head of the column. Best results were obtained with a mobile phase of 20 mmol L⁻¹ NH₄NO₃ at pH 9.5. Interestingly, no column clogging was observed over hundreds of injections because, as the researchers demonstrated, the phosphate buffer, present in all injected solutions, gradually dissolved the Cr(OH)₃, probably as CrHPO₄⁺. They measured the oxidation–reduction potential (ORP) in the personal sampling extracts and observed a significant reduction of Cr^VI when the ORP was < 100 mV. To prevent this cause of low bias, an Fe^II/Fe^III redox buffer was added to the extractant (2.28% K₂HPO₄·2H₂O adjusted to pH 8.0 with concentrated H₃PO₄) thereby raising the ORP to 220 mV. They also noted that it was not necessary to degas the solution, but that the redox buffer constituents interfered with the DPC colorimetric determination of Cr^VI. The MS was operated in KED mode, with He as the collision gas, and V as the internal standard. The solution LOD was 0.51 μg kg⁻¹, corresponding to an LOD of 0.013 μg m⁻³ in air (sample volume 400 L and 10 mL of extractant).

A preconcentration procedure, based on tandem electromembrane extractions has been devised for the determination of Cr^VI in food (milk powder, fish tissue and basil) by ETAAS.⁷¹ Samples (1.0 g) were first dry-ashed (550 °C for 4–5 h), dissolved in 100 mL of dilute HCl and the pH adjusted to 5.0. Milk powder was also dispersed in water. An 8 mL subsample was taken for the first electromembrane procedure, in which species migrated across a supported liquid membrane (1-octanol containing aliquat-336 as carrier) into 40 μL of acceptor solution that acted, following acidification, as the donor solution for the second procedure, in which species migrated across 1 μL of the same organic solvent, acting as a free liquid membrane, into 10 μL of acceptor solution. Separation from Cr^III was assured as the two Cr species bore opposite charges in the separation stages. The LOD was 0.003 μg L⁻¹, and the method was validated by the analysis of CRM NIST SRM 2700 (hexavalent chromium in contaminated soil), which contains 14.9 ± 1.2 mg kg⁻¹ and for which a relative measurement error of 11% was reported that is not significant, and by spike recoveries, though it appears that these were to the sample digest solutions and not to the original samples. The concentration in the basil was below the method LOD, which is presumably 0.3 μg kg⁻¹, though this was not calculated by the authors. The concentrations of Cr^VI in the other samples were reported to be 0.9 (water extraction) and 1.3 (dry ashing) ng g⁻¹ in milk and for the fish (tilapia) 4.8 ng g⁻¹. Neither Cr^III nor total Cr were determined.

A procedure for the separation of Cr species in water samples into two aqueous phases has been devised for determination by FAAS.⁷² The two phases were created by adding to the acidified sample sufficient amounts of poly(ethylene oxide), average molar mass 1500 g mol⁻¹ (PEO1500), and Na₂SO₄ such that after centrifuging, sitting, and warming (25 °C, 20 min), two phases formed consisting of a top phase containing 43.41% (m/m) PEO1500 and 1.46% (m/m) Na₂SO₄ and a bottom phase containing 1.03% (m/m) PEO1500 and 21.71% (m/m) Na₂SO₄. Preconcentration was achieved for Cr^VI the species predominantly extracted into the top phase, as the volume of top phase was considerably smaller than the original sample volume when analysing tap water, local wastewater treatment plant effluent, and Piracicaba River water. For the analysis of wastewater from electroplating plants, a much smaller volume of sample was taken. The Cr concentration in each phase was determined by FAAS. The researchers showed that species did not interconvert during the procedure by spiking with each species first simultaneously and then sequentially; the extraction efficiencies measured in each case were not significantly different. The concentrations of species in the tap, waste and river waters were below the LOD of 5 μg kg⁻¹, but the electroplating eluent contained 1.66 g t⁻¹ Cr^III and 8.41 g t⁻¹ Cr^VI. The abbreviation “t” was not defined, so it is not clear whether this refers to a ton (long or short) or a tonne.

Pechancova et al. have reported on a detailed study of Cr species released from failed hip or knee CoCrMo implants, by a number of techniques including HPLC-ICP-MS and blue native polyacrylamide gel electrophoresis (BN-PAGE).⁷³ Sample preparation involved adding isotopically enriched spike of ⁵³Cr^III (0.5 μg g⁻¹) and 10 mmol L⁻¹ EDTA solution (pH 10.5) to 80 mg of homogenised periprosthetic tissue and shaking for 60 min. After centrifugation, the supernatant was heated (70 °C, 90 min), and filtered (0.20 μm PTFE). For patients with extensive metallosis, an extra 3 kDa ultrafiltration step was inserted to remove NPs (3–200 nm) and protein complexes prior to the HPLC-ICP-MS determinations. The chromatographic procedure, published in 2020, involved isocratic elution of a 20 μL subsample with 30 mmol L⁻¹ NH₄NO₃ through a Hamilton PRP-X100 PEEK column (150 × 2.1 mm, 5 μm) maintained at 21 °C. The spectrometer contained an octopole reaction/collision cell operated in helium mode. The procedure was validated in the earlier study by the analysis of two CRMs, NRCC, LUTS-1 and TORT-2 (lobster hepatopancreas) and the LOD was 0.03 μg g⁻¹. Total Cr in all the relevant fractions was also determined by ICP-MS. The protein complexes, separated by PAGE and visualised by staining with Coomassie Brilliant Blue, were cut from the gel and the Cr quantified by ICP-MS. It is possible that Sc was also used as an internal standard but it is not clear at what stages of the various procedures this was added. The authors devoted a considerable amount of space to the discussion of the results, which they summarised as in agreement with the results of several prior Cr speciation studies analysing ex vivo human periprosthetic tissues surrounding metal-on-metal and metal-on-polyethylene implants. They found elevated Cr in the periprosthetic tissues caused by the presence of metallic Cr, CrPO₄, and Cr₂O₃ particles, whereas Cr^VI was found only in ventricular, hepatic, and splenic tissues of diabetic patients collected post-mortem. They also observed, for the first time, the binding of Cr and other implant-based metals (Al, Co, Mo, Nb, Ti and V) to albumin in periprosthetic tissues for patients both with and without metallosis.

3.5 Copper

Whilst this section deals with the elemental speciation of Cu, all of the publications reviewed involve the macromolecular forms of Cu in various biological systems or the binding of Cu within biomolecules. Recent advances in biochemical Cu analysis by inorganic MS⁷⁴ have been reviewed (33 references, over the period 2000 to 2020). The instrumental scope of the review covered the basic areas of speciation, imaging and the newer method of single cell analysis. The utility of these techniques is illustrated with a particular focus on tissue extracts, fresh tissue samples and cells. Whilst giving some basic background to Cu speciation, the most interesting part of the review highlighted the miniaturisation of HPLC separations in metalloprotein analysis and the advantages that it offers when using low flow, low volume, small diameter microbore LC columns coupled to ICP-MS detection. These include: the reduction in sample dispersion within the column; their applicability to the analysis of samples with a limited volume (injection volume 1–5 μL), including extracts from tissue biopsies; and the introduction of a low flow rate of eluent (40 μL⁻¹) with no loss in sensitivity compared to conventionally sized HPLC columns. The use of narrow, micro- and capillary bore columns in metallomics is an overlooked area of development, particularly as the newest ICP-MS/MS instrumentation offers increased sensitivity, so are suited to low volume applications. The review also highlighted the use of SIMS and LA-ICP-MS for Cu-imaging of tissues and scanning XFM for the analysis of Cu, Fe and Zn in fibroblast cells from ATOX1 (a Cu transporter) knockout mice. The newer technique of single cell analysis, which is essentially based on single particle ICP-MS, whereby a cell suspension is introduced to the ICP-MS through a customised nebuliser and spray chamber operating in fast time resolved mode (1 ms signal integration time), was also covered. Whilst not a true speciation method, the single cell analysis approach can be used in conjunction with speciation analysis to provide useful information on the metabolism of metals. In this instance it was used for measurement of Cu in red blood cells, cells from cancer cell-lines and to evaluate the effect of Cu-based algaecides on algae.

The use of speciation methods to investigation the role of Cu in human disease and biomedical sciences has been reported at the molecular level, for the study of Alzheimer's disease, but also transplant and cancer patients. The evidence for the involvement of Cu²⁺ in Alzheimer's disease is equivocal, even though it has been reported to influence the aggregation of the amyloid-beta 1–42 peptide (Aβ_1–42). The direct assessment of metal–Aβ complexes requires an analytical method with a minimal disruption of the complexes involved. Duroux and Hagege⁷⁵ have developed an analytical approach to assess the distribution of Cu²⁺ among the different Aβ_1–42-species, which are known to be involved in the disease. However, the small size of the peptides and oligomers and the lability of the complexes, makes the analysis of Cu²⁺ complexes of Aβ_1–42 a difficult proposition. In this work CE-ICP-MS was used to study the influence of Cu²⁺ on Aβ_1–42 aggregation and evaluate the Cu content of oligomers during the aggregation process. The CE separation was hyphenated to an ICP-MS instrument via an in-house made sheath-flow interface, to a micro-uptake glass concentric nebuliser. The capillary interface went through the first inlet of a micro-cross piece toward the tip of the nebuliser and the electrical connection was achieved by the ground electrode positioned through the third inlet. The sheath liquid used (Tris 5 mmol L⁻¹, pH 7.4) was introduced by self-aspiration to the fourth inlet and also contained 10 μmol L⁻¹ EDTA to reduce Cu absorption on ICP-MS glass parts. The sheath liquid solution height was adjusted to both eliminate the suction effect from the nebuliser and avoid any flow reversal into the capillary. The optimal flow was estimated to be around 450 μL min⁻¹ by measuring the sheath flow consumption over a 30 min period. Samples were introduced into the capillary by hydrodynamic injection under 1 psi for 10 s (around 90 nL). Electrophoretic separations were performed at 25 °C at −7 kV assisted by a pressure of 0.3 psi. The fused silica capillary (id 75 μm, od 375 μm, total length 64 cm) was coated with hydroxypropyl cellulose to reduce the interaction between the silanol groups of the capillary surface and the metal ions. The developed CE-ICP-MS method was used to monitor the distribution of Cu between soluble species of Aβ_1–42 in various in vitro peptide-metal mixtures and confirmed that Cu exhibits an affinity for both monomeric and oligomeric Aβ_1–42 species and its presence lead to the formation of anionic species. However, the utility of the method was impaired by poor resolution, resulting from the use of mild separation conditions which were required to avoid dissociation of the complexes and loss of Cu. As total separation of the species was not achieved, an accurate quantitation of Cu in the different species was not possible. Markovi et al.⁷⁶ have developed a speciation method for Cu in human serum using conjoint liquid chromatography on short-bed monolithic disks with UV and post-column ID-ICP-MS for detection and quantitation. The method was used to investigate the Cu bound to low molecular mass species (LMM), ceruloplasmin (Cp) and HSA, which form the main Cu species in human serum. Two immunoaffinity CIMmic™ albumin depletion (alpha-HSA) disks and one CIMmic™ weak AEC diethylaminoethyl (DEAE) disk were assembled in a single housing, forming a CLC monolithic column. By applying isocratic elution with a 50 mmol L⁻¹ MOPS buffer (pH 7.4) in the first 3 min at a flow rate of 0.3 mL min⁻¹, followed by gradient elution with 1 mol L⁻¹ NH₄Cl (pH 7.4) in the next 9 min, HSA was retained by the alpha-HSA disk, allowing subsequent separation of the LMM from Cp on the DEAE disk. Further elution with 500 mmol L⁻¹ acetic acid in the next 4 min rinsed the HSA from the alpha-HSA disk. The separated Cu species were quantified using post-column ID-ICP-MS, monitoring Cu isotopes at m/z 63 and 65. The isotopically enriched ⁶⁵Cu was delivered with a peristaltic pump via a T-piece after the separation. The mass flow of Cu was plotted versus time during the chromatographic separation and the concentrations of Cu species were calculated by means of equations for the post-column SUIDMS analysis. The accuracy of the speciation method was not directly addressed due to the lack of speciated CRM materials available, however the column recovery data was determined from the sum of the Cu for each species in comparison to the total Cu in the sample, which gave quantitative column recoveries, with values in the range 97–105%. The repeatability (n = 6 replicates) was between 1.0–9.0% RSD and the LOD values for the Cu–Cp, Cu–HSA and Cu–LMM species were 6.1, 5.3 and 3.3 μg L⁻¹ as Cu, respectively. The technique was successfully applied to the determination of Cu–Cp, Cu–HSA and a fraction that most probably corresponds to the Cu–LMM species in the human serum of healthy individuals, kidney transplant patients and cancer patients. It would be interesting to see this approach used for the investigation of Wilson disease patients, where the total serum Cu content is much lower than the aforementioned patient cohorts.

3.6 Gadolinium

In a similar vein as discussed in this review last year¹the investigation of Gd-based contrast agents (GBCAs) in the environment has focused on the aquasphere.⁷⁷ To demonstrate the species-dependent differences in the interaction of GBCAs with humic acids as potential binding partners, model experiments with gadodiamide and gadobutrol, a linear and a macrocyclic GBCA, respectively, were conducted. Humic substances are an important part of soil and aquatic organic matter and because of their diverse, macromolecular structure, with many carboxylic and phenolic groups and polyelectrolytic characteristics, they represent suitable binding partners for metal ions. Solutions of the GBCAs and humic acid were prepared in triplicate and incubated for 24 h at room temperature. Afterwards, 450 μL was filtered using centrifugal filters (Amicon® Ultra, 3 kDa cut off), centrifuged for 30 min at 14 000g. The low molecular weight Gd-species passed through the filter, whilst the macromolecular Gd–humic acid adduct was retained. The Gd concentration in the different weight fractions was determined and a different reaction behaviour for the examined GBCAs was observed, with 73% of the total Gd amount present in the macromolecular fraction of the linear GBCA compared to 0.41% in case of the macrocyclic GBCA. The Gd-species present were determined using SEC-UV-ICP-MS analysis. An analytical TSKgel UP-SW2000 column (4.6 × 150 mm, particle size 2 μm) was used at a column temperature of 30 °C with gradient elution. Ammonium formate 5 mmol L⁻¹, pH 5.8 was held for 6 min at 100%, after which a salt gradient from 0 to 100% ammonium formate 150 mmol L⁻¹, pH 5.8 was applied, at a flow rate of 0.4 mL min⁻¹ and 5 μL injection volume. A triple quadrupole ICP-MS/MS instrument was used with O₂ as reaction gas and monitoring of the ¹⁵⁸Gd⁺ and ¹⁵⁸Gd¹⁶O⁺ ions, humic acids were identified using a UV detector at 300 nm. Whilst the peaks obtained for the Gd–humic acid complexes were excessively broad in nature, the LOD and LOQ obtained via speciation analysis were 0.7 and 2.2 nmol L⁻¹ for gadodiamide and 0.5 and 1.8 nmol L⁻¹ for gadobutrol, respectively. The speciation of the macromolecular fractions by SEC-UV-ICP-MS confirmed that Gd–humic acid adducts were formed with the linear gadodiamide, but not with the macrocyclic gadobutrol. The potential environmental consequences of these findings were unclear, but establishment of the potential toxicity of Gd and also these individual compounds to sensitive organisms is clearly required, with a view to determining their ultimate fate in the environment.

Continuing on their work into the biomedical fate of Gd-containing MRI contrast agents in the rat brain, the group from Pau in France, have published a third paper⁷⁸ in a series investigating the differences between macrocyclic and linear forms of these contrast agents. In this paper they examined the comprehensive speciation analysis of residual Gd in deep cerebellar nuclei (DCN) in rats repeatedly administered either macrocyclic gadoterate meglumine or linear gadodiamide structured GBCAs via 4 daily intravenous administrations of 2.5 mmol kg⁻¹ of drug for 5 weeks, which corresponds to 80-fold the clinical dose if adjusted for patients. The animals were sacrificed 4 months after the last injection and the DCN were dissected and stored at −80 °C. To provide a suitable amount of tissue for sample preparation and further analysis using multiple techniques, DCN from each group of 6 rats were pooled. Sample preparation involved 2 consecutive steps using water followed by urea to extract the Gd-species. The total Gd concentrations were determined by ICP-MS and the extracted Gd-species were analysed by SEC-ICP-MS. Speciation analysis of the Gd-complexes used isocratic elution from a Superdex-200 column (300 × 10 mm) with ammonium acetate (100 mmol L⁻¹ pH 7.4) over 45 minutes at a flow rate of 0.7 mL min⁻¹ and a 100 μL injection volume. The LOQ was determined using standard GBCA solutions as 0.3 pmol mL⁻¹ for both GBCAs with the criteria that the signal-to-noise ratio should be at least 5. Elemental analysis of the insoluble Gd species used sNP-ICP-MS, nanoscale SIMS, and STEM-EDX. The total Gd concentrations in the pooled DCN from animals treated with gadoterate or gadodiamide were 0.25 and 24.3 nmol g⁻¹, respectively. For gadoterate, the highest amount of Gd was found in the water-soluble fractions, being present exclusively as low-molecular-weight compounds, most likely as the intact GBCA form. In the case of gadodiamide, the water-soluble fraction of DCN was composed of high-molecular-weight Gd species of approximately 440 kDa and contained a small amount (less than 1%) of intact gadodiamide. The column recovery calculated for this fraction was incomplete, which suggested presence of labile complexes of dissociated Gd³⁺ with endogenous molecules. The highest amount of Gd was detected in the insoluble residue, which was demonstrated, by sNP-ICP-MS, to be a particulate form of Gd. The two imaging techniques, NanoSIMS and STEM-EDX, allowed further characterisation of these Gd species as amorphous, spheroid structures of approximately 100–200 nm of sea urchin-like shape. Furthermore, Gd was consistently co-localised with Ca, O, and P, strongly suggesting the presence of structures composed of mixed Gd/Ca phosphates. No or occasional co-localisation with Fe and S was observed. The results showed that for the macrocyclic gadoterate the main fraction of retained Gd was present in the soluble fractions as the intact GBCA form. In contrast the linear gadodiamide, had less than 10% of Gd solubilised and characterised using SEC-ICP-MS. The main Gd species detected in the soluble fractions were macromolecules of 440 kDa, one of which was speculated to be a Gd complex with the iron-binding protein ferritin. However, the major fraction of residual Gd was present as insoluble particulate species, very likely composed of mixed Gd/Ca phosphates. This comprehensive Gd speciation study provided important evidence for the de-chelation of linear GBCAs and offered an insight into the mechanisms of Gd deposition in the brain.

3.7 Halogens

The geochemical cycling of atmospheric I is of interest because of its role in climate change where it can react with atmospheric matter to form secondary particulates affecting the surface irradiation flux. Additionally, because ¹²⁹I plays an important role in the safety assessment of nuclear facilities, understanding the environmental geochemical cycling and speciation transformations of radioactive ¹²⁹I is of interest. Iodine in the air exists mainly in particulate, gaseous inorganic (e.g., I₂, HI, and HOI), and gaseous organic forms (e.g., CH₃I) and the speciation of I is difficult, because of the low concentrations encountered and the species interchangeability. The atmospheric cycling of I species was studied after developing a method for the speciation analysis of ¹²⁹I and ¹²⁷I by sequential collection of particulate, gaseous inorganic and gaseous organic forms of I, followed by chemical separation and measurement of ¹²⁷I by ICP-MS and ¹²⁹I by AMS.⁷⁹ A specially designed cascade sampler was divided into three sequential sections, the first section contained a glass fibre filter for sampling particulate I, the second section contained a glass fibre filter impregnated with alkaline solution for sampling gaseous inorganic I, whilst the final section comprised a nylon cartridge (90 od × 15 mm) filled with activated carbon (50 g, 30–60 mesh) for sampling gaseous organic I. A vacuum pump was used with a control system that allowed adjustment of the sampling flow rate. Air samples were collected at a flow rate of 100 mL min⁻¹ for 72 h, which amounted to a total air volume sample of 340 m³. Particulate I from the first fibre filter was isolated from other particulates by combustion using an optimised temperature programme and trapping of the released I₂ in a solution containing NaOH (0.5 mol L⁻¹) and NaHSO₃ (0.02 mol L⁻¹), a small aliquot was then analysed for ¹²⁷I by ICP-MS. The I₂ in the remaining solution was coprecipitated as AgI–AgCl for the AMS measurement of ¹²⁹I. A similar approach was taken for the second impregnated fibre filter, except a different temperature programme and trapping solution was used. The activated carbon from the last filter cartridge was placed in a quartz boat, spiked with ¹²⁵I which acted as a chemical yield tracer and heated in the presence of N₂ and O₂ inside a tube furnace. The released I₂ was trapped in the same alkaline solution as used for the other fractions, but recovery from that solution used a solvent extraction method, rather than coprecipitation. The I₂ was extracted into CCl₄ and back extracted with H₂O with the addition of a few drops of NaHSO₃ (0.01 mol L⁻¹) followed by addition of AgNO₃ to form a AgI precipitate for analysis. The LOD values for the three species (particulate, gaseous inorganic, and gaseous organic) were between 0.30 to 2.21 ng m⁻³ for ¹²⁷I and 0.05 to 0.22 × 10⁵ atoms m⁻³ for ¹²⁹I. The method was applied to the determination of these species of ¹²⁹I and ¹²⁷I in ambient air from Xi'an, China, between May and August 2020. Gaseous organic I was found to be the dominant species of ¹²⁷I and ¹²⁹I, accounting for about half of the total I, and gaseous inorganic I and particulate I accounted for one-quarter each during the sampling period.

Iodine in the ocean mainly exists in the form of iodide (I⁻) and iodate (IO³⁻) and dissolved organic iodine (DOI). The anthropogenic production of ¹²⁹I has been significantly greater than its natural formation, which has lead to an imbalance of its distribution. This is reflected in the ¹²⁹I : ¹²⁷I atom ratio, having a difference of up to 7 orders of magnitude, depending on the geographical location and the I species. Although some water samples with high ¹²⁹I concentration can be determined by ICP-MS, for most seawater samples, because of the ultralow concentration of ¹²⁹I and the high NaCl concentration, the measurement is not amenable by this method. At present, the most widely used analytical approach for ¹²⁹I with the best accuracy is AMS. This technique requires the sample to be converted into AgI for determination and therefore, the separation and pretreatment steps to generate Ag¹²⁹I are very important in its analysis. A method⁸⁰ for the speciation analysis of ¹²⁹I for both inorganic and organic I in seawater used a complex coprecipitation workflow and developed a novel SPE analysis method using Bond Elut PPL cartridges (bed mass, 5 g; volume, 60 mL; particle size, 125 μm), for the direct observation of DO¹²⁹I in seawater by AMS. By modifying the I⁻ and IO³⁻ separation process by developing an improved coprecipitation method, a significant reduction in the cross-contamination of I⁻ and IO₃⁻, from about 3% to less than 0.05%, was achieved. Clearly this is very important for samples with a large difference between the concentration of ¹²⁹I⁻ and ¹²⁹IO₃⁻. The speciation analysis of inorganic ¹²⁹I, provided separation efficiencies of about 95% and 93% for I⁻ and IO₃⁻, respectively. The direct method for the measurement of the DO¹²⁹I using SPE was claimed by the authors to be the first direct observation of the DOI ¹²⁹I : ¹²⁷I ratio in seawater (Tokyo Bay).

3.8 Iron

Analytical methods for Fe speciation continue to be developed and reported due the important biological role of this element. In plants Fe is usually taken up via the secretion of siderophores and then transported in the form of low-molecular mass citrate and malate complexes. These complexes have been determined qualitatively using e.g. LC-MS/MS but are rarely quantified and this latter aspect has been addressed in a comprehensive paper from researchers in Pau.⁸¹ In this work coconut water was spiked with Fe enriched with either ⁵⁷Fe or ⁵⁸Fe which was allowed to equilibrate with the naturally occurring citrate and malate ligands present in the sample for 12 hours. Subsequently, the spiked sample was diluted and separated on a HILIC column (150 × 2.1 mm, 2.6 μm Ø) with a gradient elution using 5 mmol L⁻¹ ammonium acetate in water and 5 mmol L⁻¹ ammonium acetate in 95% acetonitrile, each at a pH of 5.5, flowing at 0.5 mL min⁻¹ for 23 minutes. The eluent was analysed by either ES-MS or an ICP-MS, the former for species identification and the latter for species quantification which was by means of the calculated ⁵⁷Fe : ⁵⁶Fe isotope ratio for each Fe/citrate/malate complex. The method allowed this quantification without the need to prepare individual standards of each Fe complex, the structure of which may well be unknown, and the authors suggest that the methodology can be applied to labile complexes of other metals containing at least two non-interfered stable isotopes, which would also require that the complexing ligand is naturally in excess compared to metal under study. Two papers report on the use of XANES for Fe speciation studies. The first of these investigated the composition of Fe aerosol samples collected at the Palmer Station in the West Antarctic Peninsula with the aim of deducing the source of this material.⁸² The Fe minerals in the aerosol particles were found to be predominantly hematite and biotite, whilst lesser fractions of pyrite and ilmenite were also detected observed as well. The Fe oxidation state varied seasonally, with Fe^II accounting for 71% of the total Fe in the summer falling to 60% in winter and the authors suggest that this change in speciation was controlled by variations in wind speed, relative humidity and solar irradiation. The results also suggested that region's crustal emission is the primary source of aerosol Fe and that total aerosol emissions are correlated with snow depth, which has implications for future emissions due to global heating. Some people suffer hypersensitivity reactions of the skin from the use of tattoo inks and potential insights as to the causes have been investigated using XRF, XANES and Raman spectroscopy.⁸³ Analysis of skin biopsies of affected patients showed the presence of a suite of metals, Cr, Cu Fe, Mn, Ni, Ti and Zn, which could have tattoo ink origins and also the pigment carbon black. XANES analysis identified that the Fe species were present as oxyhydroxides, and that Fe^III+ was present in an octahedral environment, suggesting that the Fe present was not of endogenous origin. Raman spectroscopy also identified that the TiO₂ was present as anatase.

3.9 Mercury

Two reviews covering Hg speciation in the environment have been published this year. The first of these, with 99 references, is mainly focused on Hg geochemistry in soils and sediments including a section on methylation/demethylation processes.⁸⁴ Also included are sections on sampling approaches including collection, pre-treatment and storage and the analysis of soils and sediments for MeHg and tHg with the authors noting that the commonly used distillation process for MeHg extraction can lead to artefact formation and that for tHg care needs to be taken to prevent volatile Hg species loss. The second review reported on here (121 references from 2014–2021) is concerned with recent applications and novel strategies for the measurement of Hg in environmental samples using microextraction-based approaches.⁸⁵ The review is based on the premise that established methods from organisations such as the US EPA, ISO and CEN, were developed and validated when there was not a focus on ‘green’ chemistry and suggests that these need updating to encompass techniques such as solid and liquid phase microextractions. The key working principles of these techniques were discussed, and various studies compared in terms of analytical performance and environmental impact, as well as the limits, challenges and the future outlook in this research area. The concluding remarks and future directions section is a comprehensive summary of the work presented and is well worth a read. Some highlights of that section are that methods which are described as new strategies are really minor modifications to existing procedures, and that there is an overall lack of rigorous method validation in much of the literature covered, with few performance statistics presented which makes method comparisons difficult.

Three papers report on Hg measurements in soils. The first of these is covered in Section 2 as it concerns the certification of a soil CRM.²² The second paper describes the procedures used to distinguish between deposited and in situ formed MeHg in soils in a Swiss valley.⁸⁶ Two separate procedures, iHg spiking and double-spike (isotopically enriched iHg and MeHg) ID-HPLC-ICP-MS were used to determine and correct for artificial Hg methylation in a MeHg-selective acid-leaching/organic solvent extraction procedure (6 mol L⁻¹ HCl, DCM, L-cysteine). The HPLC separations were performed on a C18 column with a mobile phase consisting of 0.1% (w/v) L-cysteine (98% v/v) and methanol (2% v/v) flowing at 1 mL min⁻¹ for 10 minutes. Subsequently, the methodology was combined with PCR amplification of hgcA genes which allowed the authors to suggest that MeHg was deposited and not formed in situ in two out of three studied locations. The paper gives a wealth of detail on the whole approach and is worth reading by practitioners in this field. Species specific ID-GC-ICP-MS has been used to assess Hg methylation potentials in sediments of an ancient cypress wetland.⁸⁷ Sediment cores from the site were spiked with CH₃¹⁹⁹Hg⁺ and ²⁰⁰Hg⁺ at 1 cm intervals and left to incubate on site for 2 hours before further processing. The Hg species were extracted based on US EPA distillation methods and the ethylated extracts analysed by GC-ICP-MS. Using the data Hg methylation and demethylation rates were calculated and both were found to be higher in summer than winter, correlating with water temperature but not with most other water quality parameters.

The measurement of Hg and its species in natural waters can be challenging due to the very low concentrations usually encountered and thus a pre-concentration step is often required; two papers reported on this year. The first method involves the adsorption of Hg species into a 3D printed thiol functionalised metal scavenger followed by sequential elution with acidic thiourea solutions before extract analysis by ICP-MS.⁸⁸ After optimisation of preconcentration factors such as the sample matrix, effect of the flow rate on adsorption, eluent composition, and elution mode the LOD values were 0.05 ng L⁻¹ and 0.08 ng L⁻¹, with preconcentration factors of 42 and 93, for MeHg and iHg, respectively. The method was validated for tHg with a groundwater CRM, ERM-CA615 (certified tHg concentration 37 ± 4 ng L⁻¹ found 41.2 ± 0.5 ng L⁻¹). The determined MeHg concentration in the CRM was below the LOD and this was validated by spike recovery at 5 ng L⁻¹ concentration with recoveries of 88–97% obtained. The procedure was then applied to two lake water samples where MeHg and iHg concentrations ranged from 0.18 to 0.24 ng L⁻¹ and 0.50 to 0.62 ng L⁻¹, respectively. The preparation of the 3D printed metal scavenger is described in a previously published paper (cited) and this paper gives full details of the validation of its use for natural freshwater samples. The second paper describes an automated method for measuring dissolved Hg⁰ and dimethylHg (DMHg) in seawaters. In this approach the Hg species in the seawater are equilibrated into an air-filled headspace and then swept by Ar to a Carbotrap® (B, 20–40 mesh, Supelco) for a suitable time period followed by thermal desorption to a GC-AFS system. After optimisation and validation, the system could sample seawater at a 30 minute frequency, with an RSD of 5%, and was found to be stable for two weeks before the traps needed replacing. Estimated LOD values for DMHg in the atmosphere and in surface seawater were found to be 10 pg m⁻³ and 0.2 fM, respectively. The paper gives full detail on development and deployment and also cautions workers on the dangers of working with DMHg, which is highly toxic, adsorbed by the skin and can pass through most protective laboratory gloves. The system was deployed in Long Island sound and the results obtained suggest that this region has lower DMHg levels than other coastal or open ocean regions for which literature data is available.

The measurement of Hg species in biota, either for the amounts present or isotopic composition for source apportionment, continues to be a popular area of research. Suarez-Criado et al. have compared GC-ICP-MS, GC-EI-MS and GC-EI-MS/MS for the determination of MeHg, EtHg and iHg in a suite of biological CRMs, IAEA-085 and -086 human hair, NRCC DOLT-4 dogfish liver and NIST 955c caprine blood, by ssIDMS.⁸⁹ The Hg species in the CRMs were solubilised after spiking, using ¹⁹⁸Hg-enriched EtHg, ¹⁹⁹Hg-enriched Hg^II and ²⁰¹Hg-enriched MeHg, with 25% TMAH and MAE (35 W for 4.5 min) followed by in situ derivatisation with sodium tetra-n-propylborate in 1 mol L⁻¹ acetate buffer at pH 4 and extraction into hexane. The full results of the work are too complex to be cited here but can be summarised by the LOD values for GC-ICP-MS (0.12 ng Hg g⁻¹) were three times lower for Hg^II and MeHg than those obtained with GC-EI-MS and between 4 to 10 times lower for those calculated for GC-EI-MS/MS. For GC-ICP-MS the recoveries of Hg^II and MeHg agreed with the certified values for all of the CRMs analysed whilst this was only the case for the hair and dogfish CRMs when the samples were analysed by GC-EI-MS. Although GC-EI-MS/MS offers higher selectivity in SRM mode this did not result in an improved analytical performance compared SIM mode, due to the lower sensitivity obtainable, and the authors do not recommend this technique for routine use. Overall GC-ICP-MS gave the best performance in terms of accuracy, precision and LOD values. The paper and the ESI contain a wealth of detail and are recommended reading for workers in this field. Some interesting work has been reported on concerning Hg speciation in 117 liver samples from Scottish raptors.⁹⁰ Following lyophilisation, MeHg was extracted with 10 mmol L⁻¹ APDC in 80% MeOH for 30 minutes at 60 °C followed by UAE for 15 minutes and filtration. The extracts were analysed using a previously published LC-PVG-AFS method, with 25% acetic acid and UV used to reduce Hg species to Hg⁰. Total Hg and Se measurements were also made on the livers, after an hot H₂O₂/HNO₃ extraction procedure, by CV- or HG-AFS, respectively, whilst sNP-ICP-MS was used to detect HgSe NPs, LA-ICP-MS to map the Hg and Se in a Golden Eagle liver section and δ¹⁵N, δ¹³C, and δ³⁴S in Golden Eagle livers were made with an elemental analyser. Validation of the MeHg, tHg and t-Se determinations was achieved with NRCC DOLT-4 CRM, with recoveries ranging from 96 to 109%, and a suite of IAEA and USGS CRMs were used to validate the stable isotope ratio measurements with all results agreeing with the certified δ values. The LOD values were 0.18, 0.006 and 0.012 ng g⁻¹ for MeHg, tHg and tSe, respectively and the measurement uncertainties were estimated according to the JCGM Guide “To The Expression Of Uncertainty In Measurement”. Again, the data set is too complex to fully report on here, and interested readers are recommend to obtain the paper, but in summary the median MeHg mass fractions ranged from 0.25 mg kg⁻¹ in Goshawk liver to 3.77 mg kg⁻¹ in White Tailed Sea Eagle liver which presumably reflects the mainly fish based diet of these birds whilst the MeHg/tHg fraction was lowest in Golden Eagle (79%) and highest in Hen Harrier (104%). The authors postulate that the low value for this fraction in Golden Eagle could be due to the formation of HgSE NPs in the liver via MeHg demethylation processes or that these NPs could have been taken up through their diet as they periodically have access to cetacean carcass (due to stranding) and that these NPs have previously been observed in cetaceans.

Two papers report on the use of AEC to selectively extract MeHg from biota. The first of these presents a method combining acidic leaching and batch AEC for selectively extracting MeHg from fish and aquatic invertebrates prior to Hg isotope ratio measurements.⁹¹ The first step involves 30% HNO₃ at 60 °C for 12 hours which is followed by a two stage extraction, each of 2 hours duration, with Bio-Rad AG 1-X4 resin and HCl, to adsorb negatively charged iHg species, e.g., HgCl₃⁻ and HgCl₄²⁻, with neutral MeHg species, e.g. MeHgCl, remaining in solution for oxidation with BrCl followed by reduction of the Hg²⁺ formed to Hg⁰ and then a purge and trap step prior to analysis with MC-ICP-MS. The method was applied to four NRCC CRMs, DOLT-2 and DOLT-5 dogfish liver, DORM-3 dogfish muscle and TORT-2 lobster hepatopancreas. The recovery of MeHg, quantified after each step in the procedure, was 93.4 ± 2.9% (1 SD, n = 28) for the CRMs and natural biota samples and 96.9 ± 1.8% (1 SD, n = 5) for aqueous MeHgCl standards, with the average MeHg purity being 97.8 ± 4.3% (1 SD, n = 28) across all CRMs and natural biota samples after the final separation step. The measured MeHg isotopic compositions of the CRMs agreed with literature values and the average precision (2 sd) for the CRMs was 0.10‰ for δ²⁰²MeHg and 0.04‰ for Δ¹⁹⁹MeHg. The authors conclude that the developed method is simpler than other widely used methods, such as distillation, and uses less toxic materials than DCM or toluene based extractions. The method in the second paper utilised 5 mol L⁻¹ HCL and UAE for 15 minutes to extract Hg species from a range of CRMs, tuna fish muscle (IRMM BCR-463 and ERM-CE464), plankton (IRMM BCR 414) dogfish liver (NRCC DOLT-5) and cephalopod, fish and plankton samples from the Gulf of Djibouti.⁹² For the Hg measurements the extracts were diluted 10-fold and passed through in-house fabricated column (ID 2.5 mm, length 20 mm, packed with AmberChrom® 1 × 2 chloride form to retain anionic iHg species) inserted between the peristaltic pump and nebuliser of an ICP-MS instrument with a sample uptake rate of 1 mL min⁻¹. A 1.44% thiourea solution was added to the sample flow post-column to minimise retention on the instrument's spray chamber whilst Lu was added, presumably to the samples as the paper states it was not retained on the column, as an internal standard and measurements appear to have been made after a 100 s uptake time. Using this setup recoveries for all of the CRMs analysed were within the certified ranges and the MeHg content in the marine organism samples ranged from <LOD (7.8 ng kg⁻¹ in solution, 1.6 μg kg⁻¹ in a solid sample) to 2.9 mg kg⁻¹ with precision ranging from 6 to 20% RSD depending on the MeHg mass fraction in the original sample.

Two different chromatographic approaches for Hg species determinations have also been published this year. It is unusual to see TLC used in conjunction with atomic spectrometry but this has been undertaken by Haraguchi et al. to quantify MeHg in hair.⁹³ The MeHg was extracted using 2.2 mol L⁻¹ HCl and heating at 100 °C for five minutes, neutralisation with 2 mol L⁻¹ NaOH and back extraction into 0.01% (w/v) dithizone in toluene. The toluene layer was removed and added to 2 mL of 1 mol L⁻¹ NH₃, the mixture shaken and then centrifuged, the toluene layer removed and evaporated and the residue redissolved in acetone and this was applied to a TLC plate with n-hexane : acetone (9 : 1, v/v) used as the developing solvent. The MeHg–dithizonate spot was removed from the TLC plate and decomposed at 850 °C followed by Au trapping and thermal desorption to a CV-AAS instrument. The LOD obtained using this method was 0.18 ng MeHg (as Hg), which corresponds to 0.018 mg MeHg per kg of hair for a 10 mg sample. Recoveries from IAEA-086 and NIMD-01 human hair CRMs were 102 and 99%, respectively and the results for real samples were in agreement with those obtained using an acidic leaching/toluene extraction/GC-ECD method. Finally in this section an in-line SPE-LC-AFS method is reported on.⁹⁴ To achieve this a C18 mini-column (10 mm × 4.6 mm, 60 μm), modified with diethyldithiocarbamate, was fitted in place of the sample loop on a 6 port valve and 10 mL of sample passed through the column. The retained Hg species were then eluted with 0.5 mmol L⁻¹ (m/v) L-cysteine to a second 6 port valve attached to an HPLC system fitted with a 150 mm × 4.6 mm, 5 μm C₁₈ column, with a mobile phase of 1 mmol L⁻¹ (m/v) L-cysteine, 60 mmol L⁻¹ (m/v) ammonium acetate and 5% (v/v) acetonitrile flowing at 1 mL min⁻¹. The eluted Hg species were oxidised post-HPLC column with K₂S₂O₈ and then reacted with KBH₄ to form hydrides before detection by AFS. After optimisation of the system LOD values of 0.3 ng L⁻¹ and 0.2 ng L⁻¹ were obtained for Hg²⁺ and MeHg, respectively with an enrichment factor of 100 for a 10 mL sample loaded over three minutes. The method was validated using three rice CRMs, GBW(E)100348 and GBW(E)100361 for tHg and META-DJTZK-024 for MeHg with recoveries of >97% after extraction of the HG species with 5 mol L⁻¹ HCL and UAE for 60 minutes with periodical shaking. Finally, the method was applied to 9 rice samples and the iHg and MeHg mass fractions varied from <LOD to 19 and 16 ng g⁻¹, respectively. 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, tHg, S species	CV-AFS, XANES	Peatland soil	MeHg: US EPA 1630	MeHg: distillation, aqueous ethylation	tHg: 0.03 ng g^−1	Spike recovery, results not given	95
			tHg: US EPA 1631		MeHg: 0.006 ng g^−1
			S: dried under inert gas, cryohomogenised
MeHg	CV-AAS	Freshwater fish	MeHg: 0.3–0.5 g sample, 10 mL HBr, stirred, 20 mL toluene, shaken, centrifuged, back extraction into 1% L-cysteine	Selective extraction	MeHg: 0.007 μg g^−1 (LOQ 0.023 μg g^−1)	DORM-2, recovery 99.0 ± 3.7%	96
MeHg, tHg	ID-GC-ICP-MS, ID-ICP-MS	Soils	MeHg: distillation with H2SO4, KCl, CuSO4	Selective extraction	MeHg: 0.004 ng g^−1	MeHg: IAEE-158, recovery 104%	97
			tHg: hot HNO3, BrCl oxidation, SnCl2 reduction		tHg: 0.2 ng g^−1	tHg: MESS-4 recovery 84%
iHg, MeHg	CV-DBD-AAS	PM10	MeHg: UAE 2.0 mol L^−1	Selective extraction	MeHg: 6 ng	Spike recovery: MeHg : Hg mixtures, MeHg recovery > 91%	98
			tHg: HG with NaBH4		tHg: 9 ng
iHg, tHg, organoHg by difference	CV-AFS	Waters	Hg species extracted with dithiol coated magnetic NPs	tHg: organoHg oxidised with KBr–KBrO3	0.55 ng L^−1	Spike recoveries: iHg 94–106%, MeHg 85–91%, EtHg 83–94%, PhHg 84–95%	99
				iHg: SnCL2 reduction
MeHg, tHg	ID-GC-ICP-MS	Serum	Cited method	Cited method	MeHg: 0.03 μg L^−1	Not given	100
					tHg: 0.2 μg L^−1
EtHg, MeHg, iHg	ID-GC-ICP-MS	Blood	Cited method	Cited method	0.064–0.26 μg L^−1	Not given	60

---

3.10 Molybdenum

For the HPLC-ICP-MS speciation of Mo (and As, and Cr) in lysimeter waters⁶⁷ in support of a study of the leaching of these elements from three geotechnical composites, made of recycled waste, in both compacted and uncompacted (20 times less dense) forms, a SAX fast-protein column (Mono Q HR 5/5) HPLC column was used, with gradient elution from 0 to 0.7 mol L⁻¹ NaCl over 10 min at a flow rate of 1.5 mL min⁻¹. The samples were adjusted to pH 12, and 100 μL were injected. Without any further optimisation, almost baseline resolution of the peaks for arsenate, chromate and molybdate was obtained. The spectrometer was operated in collision mode with He as the bath gas and 4 ISs, ⁷²Ge, ¹¹⁵In, ¹⁰³Rh, ⁴⁵Sc, were added to the mobile phase at concentrations of 100 ng mL⁻¹. Total element concentrations were also measured by ICP-MS, the procedure for which was validated by the analysis of SPS Spectrapure Standards AS (Oslo, Norway) SPS-SW1 (quality control material for surface water analysis). The determination of molybdate (LOD 0.31 ng mL⁻¹) was validated by spike recovery. The researchers found that both the degree of compaction and the composition of the various composites studied both had significant effects on the leaching of the three species and much of the report is given over to a discussion of the relevant chemistry.

3.11 Phosphorus

A comparison of two methods for the measurement of phosphonic acid (H₃PO₃), a fungicide used in the wine industry, in the presence of phosphate has used IC coupled to ICP-MS or a conductivity detector (CD).¹⁰¹ A range of samples were tested including conventional and organic white and red wine, vine leaf, stem, roots, white and red grapes, grape cane and soil/sediment. The liquid samples were filtered <0.45 μm, the plant materials were first freeze-dried and then milled to a particle size <0.5 mm, and the soil/sediment samples wet milled, dried and then sieved <2.0 mm. The liquid samples were diluted prior to analysis whereas the solid samples were extracted with water for 4 h, followed by centrifugation and removal of the supernatant for analysis after a suitable dilution. A Metrosep A Supp 10 (250 mm × 2.0 mm id, 4.6 μm), AEC was used for all samples and the column was protected by two sequentially placed guard columns, a Metrosep A Supp 10 and a Metrosep RP2. The composition of the aqueous mobile phase was held isocratic and contained Na₂CO₃ (5 mmol L⁻¹), NaHCO₃ (5 mmol L⁻¹), and 5 μmol HClO₄ (5 μmol L⁻¹). For CD detection suppression of the baseline conductivity of the eluent was performed with 0.1 mol L⁻¹ aqueous H₂SO₄ (0.1 mol L⁻¹) and the suppressor was regenerated with purified water. The same Na based eluent was used with the ICP-MS detector, whereas a volatile buffer such as an ammonium salt would be preferred, so as to reduce the build-up of salt deposits on the interface. The ICP-MS and IC module were connected through a remote cable that enabled synchronisation of both instruments regarding chromatographic runtime alignment and simultaneous detection with CD and ICP-MS. Detection using CD identified a number of compounds by retention time matching including acetate, chloride, phosphonate, phosphate, malate, tartrate and sulfate. Coupling the IC to ICP-MS achieved greater specificity than CD detection as it eliminated peaks due to non-P-containing coeluting substances. The ICP-MS conditions used interference removal by a dynamic reaction cell containing O₂ and measurement of the PO⁺ formed at m/z 31. Unsurprisingly, the ICP-MS instrument provided improved sensitivity compared to CD, giving LOQs ranging in liquid and solid samples from 3.8 to 16.8 μg kg⁻¹ and 0.08 to 2.41 μg kg⁻¹ compared to the liquid and solid samples measured using CD detection of 39.9 to 593.7 μg kg⁻¹ and 3.51 to 58.7 μg kg⁻¹, respectively. Extraction recoveries ranged from 95.1 to 99.3%, but no CRM or RM was analysed to demonstrate the accuracy and traceability of the methodologies. Data on an anonymised selection of 100 conventionally and 30 organically produced wines were briefly presented to demonstrate the suitability of the method.

3.12 Platinum

Two papers related to the analysis of Pt-containing chemotherapeutic agents have been reported, each dealing with a different aspect of metallodrug analysis, including the drug delivery system and impurities in the drug preparations. Wroblewska et al.¹⁰² has focused on the use of a CE-ICP-MS/MS method for the investigation of liposome–cisplatin nanosystems and their interactions with transferrin. Drug targeting can reduce the systemic toxicity of drugs such as cisplatin and improve treatment efficacy, by improving tumour uptake. Liposomes are artificial vesicles with at least one lipid bilayer, in this case formed from a phospholipid, which can encapsulate the drug for delivery. The CE separation method chosen for speciation analysis is not easy to use because of the difficulty in coupling to ICP-MS/MS, but the authors reported it to work particularly well with charged molecules and it is possibility of using a background electrolyte adjusted to reflect physiological conditions. Conventional HPLC separations are more problematic because of the interaction between the liposome and the stationary phase, leading to poor recoveries and leakage from the liposome. In this work a self-aspirating nebuliser interface was used for sample introduction to a low volume spray-chamber, with a cross-piece to merge the sheath liquid flow and a grounded Pt wire to maintain electrical connectivity. Recovery of liposome from the capillary was reported to be 80% based on the ³¹P¹⁶O⁺, signal. A triple quadrupole ICP-MS/MS instrument was used in reaction cell mode with O₂ as the cell gas and the signals for ³¹P¹⁶O⁺, ³²S¹⁶O⁺ and ¹⁹⁵Pt were monitored to quantify (in retention time order) free cisplatin, hydrolysed cisplatin, cisplatin bound to transferrin and the encapsulated liposome–cisplatin complex. The drug–protein species was suggested to be a transferrin adduct with free drug originating from the non-encapsulated cisplatin. The analytical figures of merit for the liposome complex were based on the ³¹P¹⁶O⁺ signal and gave an LOD and LOQ of 1.06 and 3.19 μmol L⁻¹, respectively, with intermediate precision <5%. The method was applied to the solutions resulting from experiments on the preparation of liposome–cisplatin complexes. As the use of drug delivery systems become more wide spread then suitable methods will be needed that can fulfil the testing requirements of regulatory authorities and for these to be used with the variety of metallodrugs that are available, speciation methods such as this will be required. The second paper deals with the analysis by HPLC-ICP-MS of the residual Pt-containing raw materials occurring at trace levels in the final drug preparations of cisplatin, carboplatin and oxaliplatin.¹⁰³ The main challenge was to overcome the stark difference in concentration between the final Pt-drug product and the much lower concentration of impurity. The aim being to ensure that the detector is not exposed to very high concentrations of Pt, which would affect the blank signal and have the potential for carry-over effects in the sample introduction system. It would not be possible to dilute the samples sufficiently to reduce the concentration of the Pt-drug because of the high drug: impurity concentration ratio, as dilution would reduce the concentration of impurity to a level below the LOD/LOQ for the method. To overcome this issue, the authors used a simple switching valve approach to divert the flow from the HPLC column in such a way that the very high concentration of the Pt-containing pharmaceutical drug did not reach the ICP-MS detector. Thus, the flow from the HPLC separation would be diverted to waste during the periods covering the retention times of the drugs being tested and only connected to the ICP-MS detector when the specific impurity eluted. In this way the impurity concentration could be measured without the drug reaching the detector. Separation of inorganic platinum–chloride complexes was accomplished in less than 3 min on an AEC with a mobile phase containing 2 mmol L⁻¹ (v/v) HCl. The LOD values were 0.2 μg L⁻¹ and 0.3 μg L⁻¹ for potassium hexachloroplatinate (K₂PtCl₆) and potassium chloroplatinate (K₂PtCl₄), respectively. Method precision and accuracy were satisfactory at <3% RSD for 10 μg L⁻¹ of K₂PtCl₆ and K₂PtCl₄, and >90% recovery for real samples. The method has the advantages of simplicity, rapidity and sensitivity, and provides a basis for quality control of trace impurities in the production process of Pt-based drugs.

3.13 Selenium

Two reviews have appeared this year on Se speciation focused on different applications. The preparation, purification, bioactivities, and application of Se-enriched proteins have been extensively reviewed by Wen et al.¹⁰⁴ The paper is divided into several sections: (1) preparation and purification of Se-enriched proteins obtained from different resources (microorganisms, plants, and animals). In this section the performance of different methods of extracting Se-proteins are critically evaluated such as nature of the solvents (water, salt, ethanol, alkali solution extraction) and extraction techniques (ultrasound assisted and pulsed electric field assisted extraction); (2) bioactivity of Se-enriched proteins and their hydrolysates. This part describes antioxidant, immunoregulation, neuroprotective and hyperglycaemic activities among others and (3) applications of Se-proteins as stabilisers, dietary supplements, and sauces. The review contains useful information for researchers working with Se and it includes new perspectives for the application of Se-enriched proteins in the development of functional and nutritional foods. Of special relevance for the readers are the tables that compile relevant information on extraction, separation and purification techniques and the bioactive properties of the proteins and their hydrolysates. The review includes more than 80 references. The focus of the second review is the current state of VG-AFS methodology for Se speciation.¹⁰⁵ The review discusses different methodologies for performing the pre-reduction process of Se^VI to Se^IV such as conventional heating, microwave irradiation, UV-irradiation as well as different reagents (KI, NaBr, HCl, HNO, KBrO₃ and organic acids). Other less conventional procedure such as the use of nano-TIO₂ as catalyst with UV radiation are also evaluated. Finally, the review describes recent advances in the application of automated and flow-injection methods involving chromatographic and non-chromatographic systems with pre-reduction steps using MW or UV irradiation.

Elucidation of Se species and metabolites in Se supplements and Se-fortified food is still a topic of research. An interesting paper¹⁰⁶ reports on the development of a zero-interfacing approach for coupling microbore HPLC with ICP-MS. The system includes a home-made column-nebuliser assembly (COL-NEB) that allowed authors to adjust the capillary monolithic column and nebuliser just before the base of the ICP. This interface eliminates the dead-volume resulting when applied other nebulisers such as DIN and DIHEN thus maintaining the separation resolution of the compounds and the LOD values of ICP-MS as well. The proposed method was applied to separate Se-containing peptides in trypsin lysate selenised yeast (CRM SELM-1) by means of a homemade 690 μm OD × 330 μm ID methacrylated-C18 hybrid silica-fused capillary monolithic column. The proposed COL-NEB assembly was first tested with a SLUG peptide by measuring the full width at half maximum (FWHM) of the resulting SLUG chromatographic peak and comparing the FWHM values obtained with those provided by a Meinhard concentric nebuliser, a HEN working with a 15 mL cyclonic spray chamber and a self-made DIN of 67 μL dead volume. The COL-NEB was selected as the best interface for coupling μHPLC-ICP-MS considering the narrowest FWHM values and the highest ICP-MS signal obtained. Under optimal conditions a LOD value for Se of 1.44 ng L⁻¹ corresponding to an absolute 0.72 fg was obtained. Up to 32 Se-peptides were quantified within 10 minutes chromatographic run whereas when using the self-made DIN nebuliser only 12-Se peptides were determined. The authors highlighted the broad applicability of the zero-interfacing design for separating and quantifying targeted analytes different from Se. One paper reports on the stability of Se^IV, Se^VI, SeMet, SeCy, MeSeCy and SeMetO in aqueous solution and aqueous extracts of tablets of three types of dietary supplements (Se-yeast, Se organic and Se containing vitamin E).¹⁰⁷ Samples were prepared in different solutions (water, HCl, aqueous solution pH = 2 and NH₄Ac pH = 7), mixed for 1 hour, filtered and stored in glass vials under different conditions. The extracts were subsequently analysed by HILIC coupled to MS for Se speciation and ICP-MS for total Se content. The Se species were less stable in aqueous solution than in aqueous supplements extracts confirming the great influence of the sample matrix on Se stability. It was found that acidification of the solution increases Se stability except for SeMet in yeast samples. Light did not affect Se-species stability whereas the effect of temperature was highly dependent on the Se compounds and sample matrix. According to the data obtained in this study, the authors recommend reaching a compromise between extraction efficiency and stability. For instance, NH₄Ac provided the highest extraction efficiency, however SeMet was rather unstable in this extraction medium. A complete and detailed study on Se speciation in eggs by HPLC-ICP-MS measurements has been presented by Zhao et al.¹⁰⁸ Four types of chromatographic columns for performing separation were evaluated (RPC18 columns and a Hamilton PRPX-100 column). An Agilent SB-Ag (4.6 mm × 250 mm, 5 μm) was selected for further studies as provided the highest peak intensities and the best resolution for the five Se-species tested (SeCys2, MeSeCys, Se^IV, Se^VI and SeMet). Special attention was paid to Se-species extraction by evaluating five extraction procedures (water extraction, acid extraction (0.1 mol L⁻¹ HCl), basic extraction (0.1 mol L⁻¹ NaOH), buffer extraction (25 mmol L⁻¹ NH₄Ac containing 5% MeOH) and enzymatic extraction). The best extraction efficiency was obtained when performing enzymatic hydrolysis. Based on that, a detailed optimisation of the enzymatic hydrolysis was carried out by including different extraction conditions (ultrapure water, Tris–HCl, PBS, NaHCO₃ and NH₄HCO₃) different proteases (pepsin, trypsin, protease XIV and pronase) and the combined use of lipase. The Se species in eggs were quantitatively released by using a pre-treatment with 100 mmol L⁻¹ Tris–HCl and UAE for 30 minutes followed by a treatment with 75 mg of pronase and 75 mg of lipase at 37 °C for 18 h. Spike recoveries of the Se species at the 0.1–0.4 μg g⁻¹ spike levels all exceeded 80%. The method was applied to determine Se species in different brands of regular eggs and Se-enriched eggs and cooked eggs and SeCys2 and SeMet were the dominant Se-species in fresh eggs whereas Se-species distribution in cooked eggs was affected by the temperature applied during cooking. The results obtained are of interest for researchers, however the work lacks a proper and unambiguous identification Se-species by using molecular mass spectrometry. The use high resolution mass spectrometry techniques in speciation analysis is to be encouraged to ensure the correct characterisation of known and unknown species containing Se.

The essentiality of Se for humans is related to its presence as SeCys in three major selenoproteins which are selenoprotein P (i), glutathione peroxidase (GPx) and selenoalbumine (SelAl). Moreover, several studies indicate the important role of selenoproteins against the toxicity of other elements such as Pt, Au, and Cd. An interesting paper deals on the reactivity of SelP with two metalloproteins: aurofin and cisplatin.¹⁰⁹ The SelP was first purified from human serum by sequential affinity chromatography by using two columns: an immobilised Co²⁺ affinity (IMAC) and a heparin affinity column. The SelP was well separated from other Se-containing proteins (GPx and SeAl) by the IMAC column. Separation was achieved in two hours producing a 50 mL fraction containing 35% of total Se present in the sample. The SelP from the IMAC column was collected and retained on the heparin column and subsequently eluted as a very narrow peak completely separated from matrix components. The SelP-fraction from heparin column was then preconcentrated and digested with trypsin and analysed by nanoHPLC-ES-MS/MS to identify the produced peptides and to build the SelP sequence. A full coverage of the protein sequence corresponding to its full-length (381 amino acids) was achieved. The reactivity of SelP with metallodrugs was subsequently investigated by SEC-ICP-MS with simultaneous detection of Se and metals. This approach allowed authors to trace the intact SelP, SelP-adduct and free metallodrugs. The results evidenced the coelution of the SelP form with Au and Pt. In order to ensure the direct involvement of Se as other binding sites (Cys and His) exist in the binding sequence, the digest was analysed by 2D-nanoHPLC-ESI-MS/MS. It was found that cisplatin reacts preferentially with Cys and His residues whereas auranofin is bound to SeCys and Cys residues. The combined use of SEC-ICP-MS with nanoHPLC-MS/MS of the tryptic digest of the proteins offers a great potential to evaluate metallodrugs binding sites. The effect of SeMet in preventing Cd toxicity in HepG2 cells, through the modulation of selenoproteins, was evaluated by Ramirez-Acosta et al.¹¹⁰ For this purpose, a widely implemented procedure consisting of the use of two chromatographic affinity columns (Heparin Sepharose, HEP, and Blue Sepharose, BLUE) coupled to ICP-MS/MS was employed. In brief, SelP is selectively retained on HEP column whereas SelP and SeAl are strongly bound in BLUE column. In contrast, selenoprotein GPx is not retained in any of the affinity columns tested. Cells were cultivated in the presence of 100 μM SeMet and 5 μM, 15 μM and 25 μM CdCl₂ and in combination of both species for 24 hours. The results obtained evidenced that the co-exposure of SeMet and Cd slightly increases HepG2 viability, with the highest enhancement at high levels of Cd where SeMet increased the cell survive from 35% (25 μM Cd level) to 46% (25 μM Cd + SeMet). The presence of Cd decreased the concentration of selenometabolites (Se-species unretained) and it had an important effect on GPx and Selp concentration suggesting modifications in selenoproteins synthesis in the presence of cadmium. The results obtained in this paper can be considered as a first step in the way of elucidating the mechanisms involved in Cd–Se interaction. A method based on the use of HPLC-ICP-MS was developed for determining selenoaminoacids in selenoproteins from Lactococcus lactis (L. lactis) N2900.¹¹¹ Bacteria cells were treated with Na₂Se₂O₃ to induce the expression of the recombinant GPx proteins. The incubated cells were subsequently broken in ice by UAE and centrifuged. Proteins were then precipitated with (NH₄)₂SO₄ and subsequently purified in a Ni–NTA chromatographic column. Selenoaminoacids were released from selenoproteins by using a UAE protease hydrolysis and analysed by HPLC-ICP-MS by using a Zorbax SB-Aq C18 column. Separation was performed under gradient elution by using a mobile phase of A: 20 mmol L⁻¹ citric acid and 5 mmol L⁻¹ sodium hexane sulfonate (pH 2.3, 0.1% TFA) and B: 20 mmol L⁻¹ citric acid and 5 mmol L⁻¹ sodium hexane sulfonate (pH 5.5.3, 0.1% TFA). Different Se species were detected (SeMet, SeCys2, SeCys and MeSeCy) however the method lacks an unambiguous identification of the species by applying high resolution molecular mass spectrometry.

In urine, TMSe in one of the LMW Se-species most often detected and analysis for TMSe is usually performed by HPLC-ICP-MS. However, HPLC-HG-AFS has been also used as it also provides an adequate LOD to determine TMSe in biological fluids. Based on that, Slejkovec et al.¹¹² report a detailed study on the conditions affecting on-line conversion of TMSe by varying the mobile phase composition (pyridine, phosphate and acetate) testing different reaction coils and several concentrations of reagents for hydride generation. The optimised HPLC-HG-AFS system consisted of a Dionex Ion Pac AS7anion exchange column with 20 mmol L⁻¹ KH₂PO₄ at pH 4.65 as the mobile phase at a flow-rate of 1 mL min⁻¹. After separation the eluent was on-line mixed with NaBH₄ (0.7%) and 1 mol L⁻¹ HCl to form volatile compounds. Under these conditions TMSe eluted at 4.4 min and was baseline separated from SeMe, SeCys2 and MeSeCy which eluted at the void volume. The method provides a LOD for TMSe of 0.2 ng mL⁻¹. The technique was subsequently implemented for the determination of TMSe in 64 urine samples from 16 healthy volunteers. The system was able to distinguish between TMSe producers (hINMt genotype GA) containing 2.5 ± 1.7 ng mL⁻¹ of TMSe and non-TMSe producers (hINMT genotype GG).

The determination of Se in environmental samples such as waters focused the interest of different research groups. A method based on dual cold point extraction (CPE) combined with ICP-MS determination is presented for the isolation and speciation of SeNPs, Se^IV and Se^VI in water samples.¹¹³ The SeNPs were synthesised by reducing Na₂Se₂O₃ with ascorbic acid with different capping agents: alginate (Al-SeNPs), carboxymethylcellulose (CMC-SeNPs), citrate (Cit-SeNPs) and cysteine (Cys-SeNPs). The synthesised SeNPs were spherical with a diameter in the range of 3–96 nm. The paper includes a detailed evaluation of the parameters influencing extraction efficiency such as: pH, type and concentration of non-ionic surfactants (TX-114 and TX-45), addition of salts (NaCl and CaCl₂), extraction time, incubation time and presence of natural organic matter (NOM). Due to the superficial charge of nanoparticles, the extraction yield was highly influenced by pH. The extraction efficiency gradually increased with a reduction in pH with a maximum at pH 1.0. Consequently, Se species in water samples adjusted to pH 1.0 and containing 0.2 mL of 1 mol L⁻¹ NaCl and 0.2 mL of 10% TX-45 were efficiently separated by using two CPE. In the first CPE SeNPs were adsorbed in the lower TX-45 rich phase while Se^IV and Se^VI remains in the upper aqueous phase. In the second CPE the upper aqueous phase resulting from the first CPE was treated with 2% diethyldithiocarbamate (DDTC), 0.2 mL of 10% TX-45 at pH 7.0. The complex Se^IV–DDTC was retained in the TX-45 rich phase whereas Se^VI remained in the upper aqueous phase, thus separating these species. Transmission electron microscopy (TEM) measurements showed that the SeNPs morphology did not change significantly after CPE. Similar extraction efficiencies were obtained regardless the capping agent used to stabilise SeNPs and in the presence of NOM. This analytical approach offers a LOD of 0.03 μg L⁻¹. The method was applied to waters (well water, river water from a Se-enriched region and river water collected from areas nearby Beijing). Recovery values between 65 and 113% were obtained in the different samples. In another paper a deep eutectic solvent (DES)-based LME has been proposed to preconcentrate Se^IV present in ground waters of coal mining areas.¹¹⁴ The DES employed was synthesised by heating ZnCl₂ : CH₃CONH₂ in a molar ratio of 1 : 2 and 1 : 3. The Se^IV was first complexed with a chelating agent (0.1–0.5% APDC) and mixed with DES for extraction. The resulting dispersion was subsequently shaken at different intervals (5–20 s) on a vortex mixer to enhance dispersion. Finally, the DES-enriched Se^IV was separated and back-extracted into a dilute HCl solution to decrease the matrix effect prior to ETAAS measurements. Total Se was determined after a reduction step of Se^VI to Se^IV with 2 mol L⁻¹ HCl and heating at 80 °C. The LOD for Se^IV was found to be 25.2 ng L⁻¹ with an enrichment factor of 64. The accuracy of the proposed methodology was evaluated by spiking known standards of both Se^IV and Se^VI in ground water samples with recoveries >98%. The concentration of Se^IV was lower than Se^VI in all water samples analysed. It was found that the concentration of Se increases with the depth of the ground water. The authors suggest DES as a cost-effective and green extractant to be used instead of ionic liquids for speciation purposes. A passive sampling technique based on DGT has been developed for simultaneous and in situ measurements of As^III, Sb^III, Se^IV species.¹¹⁵ The gel binding substrate was made of aminopropyl and mercaptopropyl-bifunctionalised mesoporous silica spheres (AMBS). Different techniques were applied: X-ray photoelectron spectra to analyse the chemical states of the elements and confirm thiol and amino groups functionalisation, HPLC-ICP-MS to determine As^III, Sb^III, Se^IV in river water samples resulting from a field deployment trial and LA-ICP-MS for micromapping As^III, Sb^III, Se^IV lability in rice roots. The AMBS-DGT provided stable measurements of As^III, Sb^III and Se^IV through a wide range of conditions (pH 4–9 for As^III, Sb^III; pH 5–7.5 for Se^IV; ionic strength 0.01–200 mmol L⁻¹ NO³⁻). The system was applied to measure the time-averaged concentration of the selected analytes from a field deployment trial in an inland river. Samples were collected after a deployment time of 72 h. The target analytes were eluted from the AMBS-DGT gel using a mixture of 3 mL of 1 mol L⁻¹ HNO₃ and 1% KIO₃ (m/v). A PRP-X100 column (10 μm, Hamilton, UK) was used to separate different As, Sb, and Se species. The mobile phase for Se speciation was 5 mmol L⁻¹ ammonium citrate at pH 4.3 containing 2% MeOH at a flow rate at 1.0 mL min⁻¹. The results obtained agreed with those provided by high frequency grab sampling. It is worth noting the application of the AMBS gel disc combined to LA-ICP-MS to visualise the spatial patterns of As^III, Sb^III, Se^IV around rice roots. The results showed a depletion of As^III and Sb^III and a significant mobilisation of Se^IV within the rice rhizosphere. The developed AMBS-DGT gel disc seems to be a quite promising tool to for risk assessment and As, Se and Sb biogeochemical behaviour at critical interphase zones. Table 2 shows examples of other applications of Se speciation presented in the literature during the time period covered by this ASU.

Table 2 Applications of speciation analysis: Se. Note that the for these citations the identity of the organoSe species detected was not confirmed by molecular MS

Analyte species	Technique	Matrix	Sample treatment	Separation	LOD	Validation	Reference
Se^IV, Se^VI, tSe	HG-AAS	Tap water	HCl added to reduce Se^VI to Se^VI	None, speciation by difference	tSe: 0.56 μg L^−1	tSe: NIST 1643e, 96% recovery	116
					Se^IV: 0.72 μg L^−1	Se^IV: spike recovery 92–116%
Se^IV, Se^VI, tSe	ICP-MS	Natural waters	Adjusted to pH 5	Se^IV, SiO2 with PDDG, eluted with DMPS at 50 °C	Se^VI: 0.75 ng L^−1	tSe: GSB Z50029-94, 99% recovery. Spike recoveries of 97–104% for both species	117
				Se^VI, SiO2 with MPhS, eluted with 1 mol L^−1 HNO3	Se^IV: 1.25 ng L^−1
Se^IV, Se^VI, SeMet, SeCys, MeSeCys, SeMet	HPLC-ICP-MS	Wheat grains	Extraction with protease XIV in Tris–HCl, pH 7.5, shaking at 37 °C for 22 h	Hamilton PRPX-100, ammonium citrate 5 mmol L^−1 and 2% MeOH at 1 mL min^−1	Not given	NIST SRM 1570a (spinach), SRM 1568a (rice flour) recoveries of 90–110% and 94–116% respectively	118
Se^IV, Se^VI, SeMet, SeCys2, MeSeCys and SeEt	HPLC-ICP-MS/MS	Water, salt, tea	Water, direct analysis	Hamilton PRPX-100, 25 mmol L^−1 sodium citrate, 2% MeOH (pH = 4.0) at 1 mL min^−1	Ranged from 0.04 to 0.15 ng mL^−1	Spiking with Se species. Recoveries 93–105%	119
			Salt and tea samples, dried, dissolved in water
Se^IV, Se^VI, SeMet, SeCys2, MeSeCys	HPLC-ICP-MS/MS	Rice	Enzymatic hydrolysis, protease XIV, pH 6.8, 45 °C, 6 h	C18 column, 5 mmol L^−1 1-butanesulfonic acid, 2 mmol L^−1 malonic acid, 5% MeOH, pH 5.0 at 1 mL min^−1	Ranged from 0.04 to 0.12 ng g^−1	Spiking with Se species. Recoveries 96 to 103% for all analytes except SeCys (77%)	120

---

3.14 Silver

Most papers covering Ag speciation involve the separation of ionic Ag species from Ag NPs and a report on the use of humic-acid-modified silica (HAMS) for this purpose has been published this year.¹²¹ The HAMS adsorbant was prepared by dissolving humic acids in 0.1 mol L⁻¹ NaOH and 0.02 g L⁻¹ aminopropyl silica, pH adjustment to 7.5–8.0, stirring at room temperature for 20 h, separation of the product by centrifugation and drying at 120 °C for 5 h. Starch and citrate stabilised Ag NPs were also prepared, using cited procedures, for use in extraction/adsorption studies. Ionic Ag and Ag NPs (10 mg L⁻¹) were spiked into an aqueous suspension of the HAMS (5 mg L⁻¹) adjusted to pH 3 with dilute HNO₃, followed by shaking for 1 hour and centrifugation to extract the HAMS. Under these conditions the added Ag⁺ was retained on the adsorbant whilst the Ag NPs remained suspended in solution. After elution from the HAMS with dilute HNO₃ the Ag⁺ content of the eluate was determined by ICP-OES. The methodology was then applied to diluted suspensions of two commercially available anti-bacterial products spiked with Ag⁺ and Ag NPs⁺. Recoveries of 81 and 88% were obtained for the added Ag⁺ whilst the Ag NP content in the spiked samples was determined as 86 and 94% of that originally added.

3.15 Tellurium

The relevance of Te from the technological and economical point of view has increased in the last years and consequently its release to the environment. However, the real impact of Te is not clear yet due to the lack of reliable data and many of the reported methodologies do not reach sufficient LOD values (ng L⁻¹ range) for environmental samples. Additionally, Te has a relatively high 1^st ionisation potential and most of its stable isotopes are affected by isobaric interferences although the development of ICP-MS/MS negates some of these problems. Garcia-Figueroa et al. explored the capabilities of ICP-MS/MS to reduce the influence of spectral interferences on the most natural abundant Te isotopes.¹²² For this purpose, FI-HG was interfaced to ICP-MS/MS. Most of the paper is devoted to the optimisation of the experimental conditions affecting HG-ICP-MS/MS. The combination of O₂ and H₂ in the reaction cell was used to overcome Xe isobaric interferences, allowing measurements with the most abundant Te isotopes: ¹²⁸Te and ¹³⁰Te. Subsequently, Te^IV was determined by selective generation of TeH₂ using 4 mol L⁻¹ HCl and 2% (w/v) NaBH₄ whereas total Te was quantified after pre-reduction of Te^VI with a TiCl₃ solution. Different reduction rates were found depending on the concentration of TiCl₃ (1, 2 and 5 mmol L⁻¹) and the nature of the sample stabilisation medium: HCl (0.2 or 2%), HNO₃ (0.2 or 2%) and EDTA (2 mmol L⁻¹) with 5 mmol⁻¹ TiCl₃ being chosen as best option for the reduction of Te^IV. Both 0.2% HNO₃ and 0.2–2% HCl solutions were considered as adequate stabilisation media whereas EDTA impaired the reduction by TiCl₃. Under optimal conditions, the LOD was 0.07 ng L⁻¹ for both Te^IV and iTe with a RSD of 5.2%. The method was validated by spike recoveries from two CRMs, SLRS-Q river water and CASS-6 seawater to represent natural samples and then applied to the analysis of Te^IV and Te^VI (by difference) in bottled and lake water samples. A second paper presents an interesting approach in which FI-PVG-ICP-MS/MS is evaluated.¹²³ The mechanisms of PVG are based on photodecomposition of LMW organic acids by UV radiation and the subsequent formation of radical species such as H., COO. and R. These radicals interact with the elements of interest to form hydrides, carbonyl or alkyl compounds. Recently, the introduction of different metals as photocatalyst has expanded the application of this methodology to other analytes. In this work the PVG of Te^IV to form dimethyl telluride was carried out by means of a reactor consisting of a PTFE reaction coil wrapped around a Hg lamp. Relevant PVG conditions such as composition of the reaction medium, irradiation time and metal sensitisers were tested. The PVG efficiency was investigated by adding various metal ion sensitisers (Cd²⁺, Co²⁺, Cu²⁺, Fe²⁺, Mn²⁺, Ni²⁺ and W⁶⁺) The presence of Mn²⁺ and Fe²⁺ led to the greatest signal enhancement: 2.3 and 1.97 fold, respectively whereas no positive or negative effects were observed in the presence of the other metal ions. The negative effects of Cd²⁺ and Cu²⁺ were attributed to the formation of CdTe or to the decomposition of a volatile Te compound, respectively. The best results were obtained by using a mixture of 5 mmol L⁻¹ HAc and 3.5 mmol L⁻¹ formic acid containing 250 mg L⁻¹ of Mn²⁺ and 15 mg L⁻¹ Fe²⁺. A pre-reduction step in heated 6 mol L⁻¹ HCl was required for the determination of total Te. The analytical approach offered a LOD of 1.3 ng L⁻¹ and a RSD of 0.9%. The method was applied to Te speciation in water samples (fresh water, well water, seawater and contaminated water). Method accuracy was by spike recoveries for Te^IV and total Te in water samples and also by total Te determination in a fresh-water CRM NIST 1643f with recoveries in the range of 100–107%. The authors discuss in detail the difficulties found in validating results due to the lack of CRMs certified for Te content of environmental significance. For instance, in NIST 1643f the certified value is 977 ± 84 ng L⁻¹ in comparison to the expected levels of Te in natural or slightly contaminated water samples of 10 s of ng L⁻¹. Moreover, the presence of HNO₃ in CRMs as a stabiliser hampers the determination of Te^IV for PVG as NO₃⁻ is recognised as an interference when this methodology is employed.

3.16 Vanadium

In contrast to the situation described in lasts year's ASU, when all V speciation methods reported were based on HPLC-ICP-MS, most methods reported in the past year were based on LLE. Researchers are still devoting considerable efforts to optimisation by the single-cycle alternating variable as opposed to a multivariate approach and it is disappointing to see such reports still being approved by journal referees.

Chen and coworkers at Wuhan have had a productive year with at least six articles published on inorganic speciation analysis by non-chromatographic methods and ETAAS. For the speciation of V^IV and V^V in water samples, they developed two methods based on direct immersion drop microextraction.^124,125 In the first method, two different organic solvent drops on the tips of microsyringe needles were immersed in the stirred sample solution maintained at 40 °C to which 1-(2-pyridylazo)-2-naphthol (PAN) had been added. Both species form complexes with PAN, but that of V^IV was selectively extracted into an undecanol drop, whereas that with V^V was extracted into a chloroform drop. Each drop was diluted to 100 μL with ethanol and 20 μL taken for ET-AAS determination. The researchers showed that at the optimum pH of 6.5, only 5% of the other species was extracted into each solvent but were not able to account for the selectivity other than by speculating on the roles of hydrophobicity (of the V complexes) and polarity (of the solvents). The LOD values were 2.4 ng L⁻¹ and 1.5 ng L⁻¹, for V^IV and V^V, respectively. They validated the method by the analysis of a CRM GSBZ 50029-94 (environmental water) finding V^IV and V^V concentrations whose sum was not significantly different from the certified value of 0.500 ± 0.022 μg mL⁻¹. As the concentration of V^IV “found” was 5% of the total V concentration, it is possible that this result is inaccurate, being an artifact of the 95% selectivity of the method. They found both species in 3 real samples, mineral, lake and river water. Spike recoveries (of both species) from these samples at 2 and 5 ng mL⁻¹ ranged from 90 to 110%; and spike recoveries (again of both species) at 0.1, 0.5 and 1.0 μg mL⁻¹ from the CRM ranged from 88 to 113%. In their second report,¹²⁵ they described a procedure for the analysis of apple juice, green tea, cola, lemon-lime beverage, and soda in which each V species was selectively extracted sequentially. First, V^V was extracted from 20 mL (stirring for 20 min at 40 °C) as the thenoyltrifluoroacetone complex into a 15 μL suspended chloroform drop at pH 2.5, and then the V^IV complex was extracted at pH 4.5. In each case, the chelating agent was contained in the chloroform drop (at 1.5 mmol L⁻¹) and not added to the sample solution. The drops were diluted to 100 μL and 20 μL taken for ETAAS analysis. Samples were degassed by vortexing (2 min), diluted (5-times), filtered (0.45 μm) and stored at 5 °C. The LOD values were 3.1 ng L⁻¹ and 2.6 ng L⁻¹, for V^IV and V^V, respectively. They validated the method as described in the earlier paper and found V^V in all beverage samples, but V^IV was detected only in the apple juice. They claimed that both methods had an enrichment factor of 300, but this is misleading. In the first study, the ratio of sample volume (20 mL) to final volume (100 μL) gives a value of 200. In the second study, as the samples were diluted 5-times prior to analysis, the enrichment factor was 40. The extent to which one V species was extracted under the conditions selected for the other species was not reported.

For the determination of V^V and total V in foods by ICP-OES, Mortada et al. devised a CPE method with back extraction in which samples (0.5–1.0 g) were subject to MAE with HNO₃ and H₂O₂, and the digest made up to 25 mL with pH adjustment to 7.0. A 5 mL subsample was mixed with 500 μL of 1 mmol L⁻¹ of bis(3,4-dihydroxybenzylidene)isophthalohydrazide (DHBIP), 500 μL of 2.5% (v/v) Triton X-114 and 2.0 mL of hexamine buffer (pH 7.0) and diluted to 50 mL. The mixture was incubated at 45 °C for 15 min, and the phases were separated by centrifuging (4000 rpm, 3 min) and cooling in an ice-bath (10 min). The aqueous phase was removed, and after warming to room temperature, the V in the surfactant-rich phase was back-extracted into 1.0 mL of 1 mol L⁻¹ HNO₃ by incubation at 45 °C for 15 min and subsequent phase separation by centrifugation (4000 rpm for 5 min). The V in the upper aqueous phase was determined by ICP-OES (all the relevant details are in the ESI). Following the addition of H₂O₂ to oxidise V^IV to V^V, the procedure was repeated to determine total V and hence V^IV by difference. The LOD was 0.12 μg L⁻¹ (with an enrichment factor of 50) and the method was validated by the analysis of two CRMs, NIST 1643f (water) and 1570a (spinach leaves), with relative measurement errors that were not significant. Both species were found in samples of water (mineral and river), and food (cabbage, carrots, mint and tomato) and spike recoveries of both V^IV and V^V to all samples were between 95 and 100%. It is difficult to imagine that such a method could find application for routine analysis: the chelating agent is not commercially available and was synthesised in-house and it takes almost an hour to make one measurement, not counting any preparation and digestion of the solid samples, though this part of the procedure would be needed no matter how the digests were subsequently analysed. In addition, CPE (even without a back-extraction step) is difficult to automate, so parallel processing of samples is not an option.

To determine V species in V-enriched apples (obtained from Andong Solnaeum Agriculture, Inc., in South Korea), Nam et al. devised an HPLC-ICP-OES procedure in which the species were fixed by reaction with EDTA.¹²⁶ Apples were divided into peel and flesh and dried for a week prior to grinding to a powder (no details of temperature or particle size were given). Total V was determined following a two-step extraction of 1.0 g with firstly, 10 mL of 70% HNO₃ (at 90 °C). Once the volume had decreased to 2 mL, a small volume (not specified) of 30% H₂O₂ was added and the digest filtered (0.2 μm). For speciation, 1.0 g was subject to repeated UAE (2 min, 10 times) with 10 mL of 3 mmol L⁻¹ EDTA, the supernatant was obtained after centrifugation (no details given) and the extract filtered (0.2 μm). A 100 μL sub-sample was injected into the mobile phase (3 mmol L⁻¹ EDTA, 0.05 mmol L⁻¹ tetrabutylammonium phosphate in 5% methanol at pH 4.66) for separation on a RP C8 column (150 mm × 4.6 mm) with isocratic elution at 0.2 mL min⁻¹. Peaks for V^IV and V^V were baseline separated with a total run time of about 6 min. No citations to previous reports of this separation were given, so it is possible that this separation is novel. The LOD was 0.02 mg kg⁻¹, and it was found that all the V (as V^IV) was contained in the peel at an average concentration of 0.77 mg kg⁻¹. The researchers showed that the oxidation that can occur at pH 4.4 was prevented by the addition of EDTA, otherwise no validation of the speciation methodology was reported. They did show that the concentration of total V found in the acid/peroxide extraction was 95% of that found after MAE with HNO₃ alone.

Brito and coworkers have published a further study in their on-going work on characterising the reactions that occur in the interior of graphite furnace atomisers by XPS.¹²⁷ They indicated that many previous studies have been based on techniques, such as XRD or electron microscopy that require both unrealistic amounts of analyte (maybe 1000-fold over what would be typical of the analytical situation) and excessively long heating times. On the other hand, XPS is capable of probing materials at the concentrations appropriate for analytical measurements. The focus of their work was on the identification of carbides formed by reaction with the surface of pyrolytic graphite platforms. The researchers acknowledged that drawbacks of XPS are the need for operating in ultrahigh vacuum and at room temperature, conditions which are, of course, quite different from those of ETAAS. Consequently, samples from ETAAS must be quenched by fast cooling and the possible alteration or contamination of samples, when transferring from the AAS spectrometer to the XPS system, due to contact with the atmosphere has to be considered. The researchers confirmed that V carbide, V₈C₇, usually denoted as VC in the literature, is the phase formed at ashing temperatures of ≥1400 °C, while the thermodynamically stable phase under these conditions (V₂C) was not observed, suggesting kinetic control in the carbide formation reaction. They suspected that some V oxides detected were artifacts of the XPS analysis technique. In particular, VO (a V^II species) produced at the excessively low temperature of 120 °C, probably originated from reduction induced by irradiation with X-rays, and that V₂O₅, observed after ashing at 1400 °C, was due to reoxidation of the lower V oxides during transfer from the ETAAS equipment to the UHV system. They concluded that their results confirmed the presence of the “VC” intermediate in the ashing and atomisation steps as suggested previously.

3.17 Zinc

The speciation of As, Ag, Cu, Pb and Zn in four natural plants growing in a mining area in Peru has been studied.¹²⁸ Analysis were performed on the water extractable fraction (25 mmol L⁻¹ NH₄Ac) by using SEC and HILIC separation modes combined with ICP-MS and ES-Orbitrap-MS detection. Information of metal accumulation and translocation from root to aerial parts was achieved by acid digestion followed by ICP-MS measurements. Translocation factors were found to be exceptionally high for Zn and Cu where their concentrations in leaves exceeded toxic concentrations (100–400 mg kg⁻¹). With the aim of determining the ligand involved in metal translocation the water-soluble fraction was investigated by using HILIC with a parallel (not simultaneous) detection by ICP-MS and ES-Orbitrap-MS. For this purpose, a TSKgel-Amide (20 μm, 2 × 150 mm) column was employed with a mobile phase composed of A: NH₄Ac at pH 7.0 and B: ACN under gradient elution at a flow-rate of 0.2 mL min⁻¹. To get more information an ultrahigh performance (4.6 × 150 mm, 1.7 μm) SEC column was also used with 25 mmol L⁻¹ NH₄Ac at pH 7.0 as the mobile phase at a flow-rate of 0.3 mL min⁻¹ and with post-column ACN addition. Data obtained revealed the presence of nicotianamine (NA) and deoxymugineic acid (DMA) Cu and Zn complexes. Interestingly, an unreported compound, dihydroxy-nicotianamine was identified as the most abundant ligand in the plant Hipericum laricifolium. The results seem to corroborate the hypothesis that, in contrast of hyperaccumulators plants which transport metals using weakly binding ligands, strong ligands such as NA or DMA are required in non-hyperaccumulating plants. The presence of arsenobetaine and arsenosugars were confirmed by ES-MS whereas Ag and Pb were found as HMW compounds. In another paper the use of natural deep eutectic solvents (NADES) with SEC-ICP-MS/MS is evaluated for the fractionation of metal (Mn, Cu, Zn, Co, Mo) complexes with bioligands from young barley.¹²⁹ In comparison with other extractants NADES show lower toxicity and higher biodegradability and they are considered as green solvents. Different mixtures of hydrogen bond acceptor and hydrogen bond donors were tested which efficiency was dependent on the type of fraction (LMW, intermediate MW and HMW) to be extracted. The extracts were further analysed by SEC-ICP-MS/MS by using a SEC Superdex 200 10/300 GL column with 10 mmol L⁻¹ NH₄Ac (pH 7.4) as mobile phase flowing at 7 mL min⁻¹. The use of NADES based on choline chloride solution with citric acid or glucose was efficient for extracting intermediate MW compounds. In contrast, those solvents based on β-alanine with citric acid shown a better ability to extracts the fraction of LMW and HMW compounds. NADES based on betaine were efficient in extracting intermediate MW compounds. The metals tested were found to be bound to macromolecular compounds with molecular weights ranged between 17 to 44 kDa. These compounds were efficiently extracted by NADES based on choline chloride with citric acid and betaine with citric acid. Consequently, the use of different NADES permits the selective extraction of metal–bioligands complexes of different molecular weights.

4 Biomolecular speciation analysis

Studies dealing with the practical method-based aspects of elemental speciation and which are applicable to macromolecules are a welcome addition to the literature, as they enhance the methodologies available and broaden the potential application areas. Two recent papers have the potential to make a significant impact in metallomics using ICP-MS, in particular non-targeted studies which should be a strength of this technique, but which have so far not been fully developed. Following on from previous work on new eluents for HPLC-ICP-MS the group from Graz, Austria, have investigated the separation of lipophilic compounds, using elution with 1,2-hexanediol, to enable the use of RP separations with ICP-MS.¹³⁰ A number of different fatty acids, lipids and medicinal drugs, including: non-steroidal anti-inflammatory drugs (NIASD); antibiotics; proton-pump inhibitors; and steroids, were tested and detected via the different heteroatoms (As, Cl or S) they contain. The advantage of this solvent is that it overcomes the need to use a chilled spraychamber and/or the addition of O₂ post-nebulisation to reduce the organic load on the plasma, an approach which is time consuming, requires the use of expensive Pt-tipped cones and which can decrease detection sensitivity by up to 100-fold compared to the conventional non-organic configuration. Historically, elemental speciation via HPLC-ICP-MS has focused on hydrophilic, aqueous soluble, polar compounds, mainly because the separation of lipophilic, non-polar species is restricted due to the mobile phases that are required for separation, which are not readily compatible with the ICP. However, researchers are now looking further into the speciation of elements so as to investigate non-polar species and hence the applicability of unexplored solvents, which can aid such separations are of great interest. In addition, of all the modes of HPLC separations that have been developed, RP chromatography using organic eluents are by far the largest in number, indicating the wide applicability of this mode of separation particularly in the pharmaceutical industry. The results showed that unlike other organic eluents, 1,2-hexanediol was remarkably compatible with ICP-MS detection, even at high flow rates of 1.5 mL min⁻¹ and concentrations of >30% v/v, under the standard instrumental setup normally used with 100% aqueous media. Use of 1,2-hexanediol at concentrations within the range of 1.0 to 25% v/v could replace methanol, acetonitrile, and isopropanol within the concentration ranges of 20 to 95%, 10 to 85%, and 5.0 to 60%, respectively. Sensitivity for all tested elements (As, Br, Cl, P, S, and Se) was enhanced with 10% v/v 1,2-hexanediol relative to that of 100% aqueous media by 1.5 to 7.0-fold depending on the element. Concentrations of 1,2-hexanediol at ≤30% v/v were superior in elution strength to concentrations at >90% v/v of the common organic phases such as methanol, acetonitrile and isopropanol, which greatly decreases the amount of C required to elute highly hydrophobic compounds such as lipids and steroids, enabling detection at ultra-trace levels. The proposed approach was applied to detect As-containing fatty acids in spiked human urine, and LOD values of <0.01 μg As per L were achieved, which is >100-fold lower than those previously reported using the organic ICP-MS mode. Another potential advantage of using mobile phases with higher elution strength is in the area of non-targeted speciation analysis using HPLC-ICP-MS, where more novel compounds normally missed using conventional eluents might be identified. For example, the non-targeted speciation analysis in Allium sativum revealed the presence of a large number of hydrophobic S-containing metabolomic features at trace levels. Overall, 1,2-hexanediol appeared to be less toxic than acetonitrile and methanol, and in light of the far lower concentrations of 1,2-hexanediol needed for elution, it might be considered as a greener alternative as a general eluent in RP LC with and without ICP-MS detection. The only drawback of concern was its high viscosity, which required the use of column temperatures >45 °C to ensure compatibility with standard 4.00 × 10⁷ Pa (400 bar) HPLC pumps at a chromatographic column length of 250 mm.

The second paper of use from a practical aspect detailed new “greener” extractants termed natural deep eutectic solvents (NADES) for the fractionation of compounds of selected elements from plant materials, in this case young barley.¹²⁹ The study used SEC-ICP-MS to highlight similarities in the molecular weight of different metal-containing compounds present in barley plants compared to a Rye Grass CRM (ERM CD2810). The authors tested a combination of chemicals with specific H bond donor and H bond acceptor properties in a two component mixture to determine the different molecular weight metal-containing species that were extracted. These different mixtures included: choline chloride with citric acid; betaine with citric acid; β-alanine with citric acid; choline chloride with glycerol; and choline chloride with glucose. They were used to extract different molecular weight species in the range 1.35 to 670 kDa, of six elements including: Cd, Co, Cu, Mn, Mo and Zn. The extracted metal species were eluted from a SEC Superdex 200 10/300 GL column with 10 mmol L⁻¹ ammonium acetate buffer (pH 7.4) as a mobile phase, in isocratic mode, and a reported flow rate of 7 mL min⁻¹. The column was “size” calibrated with a mixture of thyroglobulin (670 kDa), γ-globulin (158 kDa), ovalbumin (44 kDa), myoglobin (17 kDa) and vitamin B12 (1.35 kDa). The paper did not report quantitative results, which seems a missed opportunity as it would have been straight-forward to determine the extraction efficiency of the methods investigated by comparing the sum of the species to the total metal concentration in the accepted way, which is one of the strengths of elemental speciation. However, the report did highlight new extraction conditions which would be of value to explore in tissues other than plant materials.

New methods for the absolute quantification of proteins using elemental standards, which provide advances in method traceability, of particular importance for applications in the clinical and biomedical areas, have been reported. One of the significant strengths of using ICP-MS/MS as a detector in macromolecular speciation analysis is its ability to directly measure P and S, which are key heteroatoms in many important macromolecules, e.g. DNA and proteins, respectively. This feature of the technique allows for quantification of macromolecules containing these heteroatoms using elemental rather than molecular standards. This is a significant advantage of elemental detection, both in terms of their wide applicability, but also because they can provide better traceability compared to molecular detectors, which require an authentic standard for quantitation. The provision of authentic standards for calibration in the clinical and biomedical application areas is not always possible and current immunochemical based methods can suffer from poor traceability, requiring comparisons based on provision of matrix-based RMs, used as analytical standards. This leads to difficulty when changes in the RMs are required, often resulting in a lack of interlaboratory and temporal comparability in clinical results and analyte reference ranges. Any work on the provision of methods and materials with enhanced traceability in this area is a significant advance and worth implementing in real-world studies. A practical approach¹³¹ to the calibration of protein standards with low measurement uncertainty using SUIDMS, which whilst not true elemental speciation, is of wide applicability to quantitative proteomics using HPLC-ICP-MS/MS, has been developed. The comparability of testing in clinical and biomedical proteomic studies can only be achieved by traceability to the International System of Units (SI). This requires the availability of well-characterised and SI-traceable protein standards. However, there is a pronounced lack of sufficiently characterised and quantified protein standards and only very few metrologically traceable certified reference proteins available. In theory, the gravimetric determination of the protein concentration in a solution prepared from commercially available lyophilised protein is possible, however in practice it would require information on the water and salt content, as well as the other impurities, because these are usually not provided by the manufacturer and can easily sum up to a significant fraction of the material. There are methods for protein quantitation which can be used including amino acid analysis (AAA), whereby the protein is hydrolysed into its amino acids, which are subsequently quantified by molecular MS using an amino acid calibrator. Although AAA is considered the gold standard, it requires an isotopically-labelled amino acid standard and the optimisation of hydrolysis conditions is critical and can strongly influence the accuracy of the result. The developed workflow utilised SUIDMS to determine the S content of protein standards of known stoichiometry, including a procedure to correct for S-containing impurities and non-protein-bound S-containing compounds. As no spike specific for the protein measurand needs to be produced, the method is easily applicable to every protein of known S composition. The protein quantification method was developed and tested on the NIST CRM SRM 927e BSA and a commercially available avidin glycoprotein. The protein solutions to be measured were diluted to a desired target concentrations between 0.3 to 1 mg kg⁻¹ as S by weighing and then spiked close to a 1 : 1 ratio with a ³⁴S-enriched inorganic S spike. Samples and sample–spike blends were digested after the addition of concentrated HNO₃ (0.5 to 2.0 mL, 65%) with small-vessel MW digestion. The total S content of the protein solution to be characterised was determined by IDMS after blank correction and then the protein fraction was isolated by molecular weight using a membrane filtration column. The S concentration of the unbound fraction from gel-filtration (i.e. the unretained fraction) was measured, also by IDMS. The protein concentration was determined from the total blank corrected S-content corrected for the non-protein-bound S. To work effectively the protein must be free from any other protein impurities, which need to be verified by applying molecular MS or SDS-PAGE. In practice the small molecular weight S species were separated from the protein fraction by membrane filtration using PD-10 gel filtration columns containing Sephadex™ G-25 resin with a molecular cut-off of about 5 kDa. Elution volumes of different proteins were tested and were adapted accordingly and NaCl (25 mmol L⁻¹), was used for column equilibration, as mobile phase, and for elution. The sample solution (200–220 mg) was added dropwise to the resin and eluted by stepwise addition of NaCl (25 mmol L⁻¹). Protein fractions (3.2 mL for BSA, 2.6 mL for avidin) and a low molecular mass fraction (3.8 mL for BSA, 4.4 mL for avidin) were collected. Samples were weighed into digestion vessels, blended with the ³⁴S-enriched spike, followed by MW digestion and matrix removal, which was carried out using a self-packed AE column (AG 1-X8 resin, BioRad). Isotope ratio measurements for ³²S and ³⁴S were performed using a single-collector SF-ICP-MS. As measurement sequences were up to 14 h long, instrument drift was corrected by the bracketing method using repeated measurements of a S standard throughout the whole sequence. The quantified mass fraction of NIST SRM 927e agreed very well with the certified value and showed similar uncertainties (3.6%) as established methods while requiring less sample preparation and no species-specific standards.

An overview of the approaches to the absolute quantitation and provision of SI traceable results for the determination of proteins using the measurement of S by ICP-MS/MS has been reported. The group from Oviedo, who for many years have proposed and developed this area, have published an overview of the implementation of ICP-MS/MS in quantitative proteomic workflows.¹³¹ The paper discusses the use of ICP-MS/MS quantitation in comparison to ESI-MS/MS for proteomic workflows. The discussion makes the point that in theory, by measuring the concentration of S present in cysteine and methionine, the quantitation of proteins using ICP-MS/MS can be undertaken at the intact protein level as well as via the peptides generated in tryptic digestion. This means it is possible to undertake both top-down and bottom-up proteomics. However, to undertake the analysis of proteins via the peptides generated from enzymatic digestions, it is necessary to be able to chromatically resolve all the peptides present. This is possible using LC-ES-MS/MS because of the second resolution level available via MS/MS, making the identification of co-eluting peptides possible. Unfortunately, detection via ICP-MS/MS cannot discriminate between S atoms from different co-eluting peptides. So, in reality, quantitative proteomics methods using ICP-MS/MS are limited to a bottom-up approach, whereby intact proteins are separated and measured after chromatographic separation. Another useful feature discussed relating to protein analysis is the possibility of using ICP-MS/MS to obtain an indication of the degree of post-translational modification of the proteome, in this case phosphorylation, by using the measurement of S as well as P.

Recently published direct methods for the measurement of vitamin B₁₂, metalloproteins, and oligo-nucleic acids using HPLC coupled to different elemental detectors, including FAAS, ICP-OES or ICP-MS, illustrate the versatility and important analytical potential for the combination of coupling chromatography with different elemental detection systems, for the analysis of a wide range of macromolecules. In all three of these applications the chromatographic separation was based on SEC, which has the advantage of being able to separate metalloproteins and other biomolecules under physiological conditions and without affecting the binding of the metal. However, the resolving power of SEC is intrinsically poor in comparison to other LC modes, so much so that it is often not possible to separate proteins even those which have a large difference in molecular weight. This is particularly the case when the sample matrix is complex such as with human plasma, so it would be an advantage to also use SEC in combination with a second separation method such as IE, to provide better peak resolution and therefore greater certainty as to the identity of the proteins being measured. Unfortunately, to date very few studies using 2D-LC systems have been explored in this regard, but this may change now that commercially available analytical systems are available.

An improved sample introduction system for coupling HPLC to FAAS has used a concentric nebuliser combined with a T-shaped slotted quartz tube (T-SQT) in order to introduce the HPLC eluate to the FAAS system and also improve the sensitivity of the method by confining the ground state atoms in the light path for longer.¹³² Clearly the use of FAAS has a number of draw backs, such as poor sensitivity and only offering single element detection. However the instrumentation is relatively low cost and easy to use and this paper shows that it has its uses if a suitable application can be found, in this case the analysis of vitamin B₁₂ tablets. One of the main challenges in the hyphenation of HPLC to FAAS is the incompatibility of the nebuliser flow rate (7–8 mL min⁻¹) and that of the mobile phase (1.0 mL min⁻¹). To address this, the eluent was introduced to the flame region via a lab-made glass nebuliser along with air pressure, which balanced out the flow-rates so that the flame was not starved of liquid. The poor LOD of FAAS is partly due to low nebulisation efficiency and the short dwell time of the atoms in the HCL beam, compared to other atomic spectroscopic methods. This was addressed in the current work by use of the T-SQT to increase the dwell time of the atoms in the beam. Vitamin B₁₂ was selected as a model analyte to validate the HPLC-T-SQT-FAAS system and after conducting analytical performance studies, two different brands of vitamin tablet were analysed. After optimising some parameters such as mobile phase flow rate and pH, nebuliser gas flow rate, T-SQT height and injection volume, the linear range was determined to be between 4.7 and 92 mg kg⁻¹ as Co and the LOD and LOQ were calculated to be 1.6 and 5.3 mg kg⁻¹ as Co, respectively. Recovery studies were also conducted to verify the accuracy and applicability of the developed method for vitamin tablets and excellent percent recovery results (similar to 100%) with low SD values were obtained.

A group from the University of Calgary have used SEC-ICP-OES to determine the Fe, Cu and Zn-containing metalloproteins in human plasma samples from individuals with multiple sclerosis, acute ischemic stroke and healthy controls.¹³³ A Superdex™ 200 Increase 10/300 GL (300 × 10 mm id, 8 μm particle size) SEC column was used with a PBS eluent at pH 7.4 at a flow rate of 0.75 mL min⁻¹ and a column temperature of 22 °C. To estimate the molecular weight of the detected metalloprotein peaks the SEC column was calibrated by injecting a size calibration standard protein mixture and measurement of the carbon emission line at a wavelength of 193.091 nm, to monitor elution of the proteins. All the elements of interest were simultaneously detected by means of a high-dispersion, radial-view ICP-OES instrument, including: Cu at 324.754 nm, Fe at 259.940 nm and Zn at 213.856 nm. It is unusual for ICP-OES to be used as a chromatographic detector, mainly because the software used does not usually offer the option for the time-resolved data acquisition mode required. This is unfortunate, because in some application areas, such as for human sample analysis, the metal concentration levels are not particularly low so as to require the ultimate LOD values possible with e.g. ICP-MS instruments and also OES can be, depending on the instrument used, a truly simultaneous detection method unlike Q-ICP-MS. In this study the chromatographic raw data for the time-resolved emission lines of Cu, Fe and Zn were imported into a separate software programme for data analysis including, data smoothing using the bi-square algorithm and determination of the retention times and peak areas. With regard to Cu, the area of a single Cu peak which corresponded to Cp was reported, while for Fe the peak areas corresponding to an α2-sialoglycoprotein haptoglobin–haemoglobin complex and transferrin, whereas for Zn, only the peak area for the first eluting Zn peak, which corresponded to α2-macroglobulin was reported. As a measure of the column integrity on the recovery of Fe and Zn, a rabbit plasma stock was injected with and without a column and the emission lines of Fe and Zn were simultaneously monitored. The subtraction of the areas that were obtained for Fe and Zn with a column from the area counts that were observed without a column allowed the metal recovery to be calculated, which was 37–87% for Fe and 48–107% for Zn (no recovery for Cu was reported). Although the work is of some interest, the recoveries for the limited number of metalloproteins studied, were inconsistent between individuals and not quantitative. It is also clear that there were other unresolved proteins present which remained unidentified, which represents the main difficulty when relying solely on SEC, with its inherently poor peak resolution it is often not possible to separate the proteins sufficiently for quantitation or identification.

The simultaneous quantification of oligonucleic acids and a ferritin nanocage by SEC-ICP-MS has been reported for use in the development of drug delivery systems. Nucleic acid drugs, such as small interfering RNA (siRNA) and micro RNA (miRNA) are non-coding forms of RNA which are increasingly being used to treat intractable genetic conditions.¹³⁴ However, they are susceptible to enzymatic degradation when administered and because of their inherent negative charge state have limited cellular uptake. To overcome these drawbacks different drug-delivery systems have been investigated, in the current case, the Fe-storage protein ferritin which has an inner cavity capable of encapsulating these small lengths of RNA. To be able to optimise the encapsulation process and determine the purity of the formulation, requires measurement of the drug carrier complex (ferritin-RNA), the carrier (ferritin) and the drug (RNA). The use of SEC coupled to ICP-MS/MS overcomes the issues around quantitation, multiply charged ions and intolerance of complex samples that are encountered in LC-ES-MS/MS and with the use of a triple quadrupole ICP-MS, the oligonucleic acid and protein can be measured via their PO⁺ and SO⁺ signals. The chromatographic columns used in this work included a Superdex 200 Increase 3.2/300 and a 41 Superdex 200 increase at a flow rate of 0.10 mL min⁻¹. Isocratic separations were performed using ammonium citrate buffer (200 mmol L⁻¹, pH 7.0) as the mobile phase, and 10 °L of each sample was injected into the column, which was maintained at a temperature of 25 °C. The column was interfaced to a micro-mist nebuliser by a short length of PEEK tubing. Measurement of the oligonucleic acids and ferritin was carried out using the ³¹P¹⁶O⁺ and ³²S¹⁶O⁺ ions generated by reacting O₂ gas in the reaction cell of the ICP-MS/MS instrument with the analytes. Baseline drift was assessed by adding Se into the mobile phase and monitoring the isotope at m/z 78. The calibration curve (0.25–20 mg L⁻¹) for P and S showed good linearity (R², 0.9999). Interestingly, both the PO⁺ and SO⁺ signals had the same signal intensities, indicating that they have similar efficiencies not only for ionisation by ICP but also for reactivity with the O₂ cell gas used. The LOD, calculated as three times the signal-to-noise ratio (3s), for P and S were 12 and 6 μg L⁻¹, respectively. The concentrations of siRNA and ferritin in the range of 0.25 to 5.0 μmol L⁻¹ gave recovery values of 100 ± 10% and 100 ± 15% in the range 0.25 to 5.0 μmol L⁻¹ compared to theoretical values. These results indicated that no siRNA was adsorbed or that there were any significant losses in the system. In this application of SEC the resolution between ferritin, ferritin aggregates and the siRNA-ferritin complex was acceptable and the gentle nature of the separation preserved the drug delivery complex. Also, no differences in the elution time of ferritin and its aggregates were observed between the samples with or without siRNA, indicating that the encapsulation process did not affect ferritin size. Additionally, no free siRNA was observed, indicating that the analytical conditions did not release the siRNA encapsulated in ferritin during the whole process.

A recently published indirect method for the analysis of Pb containing proteins has utilised labelling strategies for the analysis of biological fluids from bone marrow.¹³⁵ The method for the separation and detection of proteins by using SEC-ICP-MS has used a commercially available High Matrix Introduction (HMI) system to overcome perceived issues around eluents containing high levels of buffer-salts. The HMI technology uses an aerosol dilution approach with clean, dry argon gas to reduce the loading on the plasma, which is reported to allow the plasma to tolerate much higher total dissolved solids levels than the 0.1 to 0.2% that is typical for ICP-MS. The study looked at the binding of Pb to proteins extracted from the bone marrow of mice exposed to Pb by gavage treatment with lead acetate (0, 1, 5 and 10 mg kg⁻¹, respectively) every day for 28 d. The neutrophil proteins were extracted from bone marrow using a protein lysis solution and purified using ultrafiltered with 3 kDa ultrafiltration tubes. These proteins were reacted with I and then measured using SEC-ICP-MS. In order to optimise the detection conditions, the influence of different plasma modes, including standard mode and several HMI modes on Pb detection were investigated. Using the HMI system the LOD values for Pb increased with the degree of dilution gas as follows: low mode 0.0089 μg L⁻¹, medium mode 0.030 μg L⁻¹and high mode 0.063 μg L⁻¹. The low gas mode had an LOD 3 times lower than the LOD of 0.0277 μg L⁻¹ for the standard mode, but not as low as the LOD for Pb in DI water of 0.0013 μg L⁻¹. The results suggest that HMI mode under low dilution gas conditions can improve the sensitivity of separations using high buffer sale concentrations by eliminating the adverse effect of salt-rich matrices on detection. The SEC conditions were optimised for the simultaneous separation of metalloproteins with different molecular weights by using 4 iodine labelled proteins, including BSA (66.0 kDa), ovalbumin (44.0 kDa), carbonic anhydrase (29.0 kDa), ribonuclease A (13.7 kDa). The 4 standard proteins were labelled with I by incubation with a KI solution (50 mmol L⁻¹) using a protein : KI ratio 9 : 1, at room temperature. The mixture was purified using a 3 kDa ultrapure tube (no details given) with 0.1 mol L⁻¹ Tris–HCl (pH = 7.5) and centrifugation for 10 min (10 000g) at 4 °C. The authors are unclear on why labelling with I, an element that does not ionise well in ICP-MS, was required when it would have been possible to detect the proteins via their S signal. The separation of the metalloproteins was performed using a TSK-GEL G3000 SWXL column (7.8 mm × 30 cm, 5 μm particle size, 250 Å pore size) and a guard column TSK Guard SWXL (6.0 × 40 mm). An isocratic elution using HEPES and anhydrous Na₂SO₄ as the mobile phase was used at a flow rate of 0.5 mL min⁻¹. The relationship between the retention time and log molecular weight of the I-labelled proteins was established to characterize the Pb-containing proteins with unknown molecular weight, but for some reason the size of the 3 Pb-containing proteins was not described and there was no further information on their identity given.

Abbreviations

AAS	atomic absorption spectrometry
AB	arsenobetaine
AC	arsenocholine
AE	atomic emission
AEC	anion-exchange chromatography
AFS	atomic fluorescence spectrometry
AMS	accelerator mass spectrometry
APDC	ammonium pyrrolidine dithiocarbamate
ASU	Atomic Spectrometry Update
ATO	arsenic trioxide
BCR	Community Bureau of Reference
BS	British Standard
BSA	bovine serum albumin
CD	conductivity detector
CE	capillary electrophoresis
CEN	European Committee for Standardisation
CPE	cloud point extraction
CRM	certified reference material
CV	cold vapour
Cys	cysteine
CZE	capillary zone electrophoresis
Da	discriminant analysis
DBD	dielectric barrier detector
DCM	dichloromethane
DCN	deep cerebellar nuclei
DDTC	diethyldithiocarbamate
DES	deep eutectic solvent
DGT	diffusive gradient in thin film
DHBIP	bis(3,4-dihydroxybenzylidene)isophthalohydrazide
DIHEN	direct injection high efficiency nebuliser
DIN	Deutches Institut für Normung
DMA	dimethylarsenic (include oxidation state if known)
DMHg	dimethylmercury
DMMTA	dimethylmonothioarsinic acid (include oxidation state if known)
DMP	2,3-dimercapto-1-propanesulfonic acid
DNA	deoxyribonucleic acid
EDI	estimated daily intake
EDTA	ethylenediaminetetraacetic acid
ED	energy dispersive
EDX	energy-dispersive X-ray
EELS	electron energy loss spectrometry
EI	electron ionisation
ELNES	energy loss near edge structure
EN	European Committee for Standardisation
EPA	Environmental Protection Agency
ERM	European reference material
ES	electrospray
ETAAS	electrothermal atomic absorption spectrometry
EtHg	ethylmercury
EtOH	ethanol
EXAFS	extended X-ray absorption fine structure
FAAS	flame atomic absorption spectrometry
FDA	Food and Drug Administration
FI	flow injection
FTIR	Fourier transform infrared
FWHM	full width at half maximum height
GC	gas chromatography
GBCA	Gd-based contrast agents
GPx	glutathione peroxidase
Hac	acetic acid
HAS	human albumin serum
HCL	hollow cathode lamp
HG	hydride generation
HILIC	hydrophilic interaction liquid chromatography
HMI	high matrix introduction
HMW	high molecular weight
HPLC	high performance liquid chromatography
HR	high resolution
IAEA	International Atomic Energy Agency
iAs	inorganic arsenic
IC	ion chromatography
ICP	inductively coupled plasma
ID	internal diameter
IDMS	isotope dilution mass spectrometry
IE	ion exchange
iHg	inorganic mercury
IMAC	immobilised metal affinity chromatography
IRMM	Institute for Reference Materials and Measurements
ISO	International Organization for Standardisation
JCGM	Joint Committee for Guides in Metrology
LA	laser ablation
LC	liquid chromatography
LLE	liquid–liquid extraction
LME	liquid microextraction
LMW	low molecular weight
LOD	limit of detection
LOQ	limit of quantification
MAE	microwave-assisted extraction
MeHg	methyl mercury
MeOH	methanol
MeSeCys	methylselenocysteine
MMA	monomethylarsenic
MOPS	3-(N-morpholino)propanesulfonic acid
MPT	microwave plasma torch
MS	mass spectrometry
MW	molecular weight
NADES	natural deep eutectic solvents
NIST	National Institute of Standards and Technology
NMIJ	National Measurement Institute of Japan
NOM	natural organic matter
NP	nanoparticle
NRCC	National Research Council of Canada
NTA	nitrilotriacetic acid
OC	organic carbon
OES	optical emission spectrometry
PAGE	polyacrylamide gel electrophoresis
PAN	1-(2-pyridylazo)-2-naphthol
PBS	phosphate buffered saline
PCA	principal component analysis
PDDG	poly(4,9-dioxododecane-1,12-guanidine)
PEEK	polyetheretherketone
PGE	platinum group element
PTFE	poly(tetrafluoroethylene)
PVG	photochemical vapour generation
Q	quadrupole
REE	rare earth element
RM	reference material
RNA	ribonucleic acid
RP	reversed phase
RSD	relative standard deviation
SAX	strong anion exchange
SDS	sodium dodecylsulfate
SEC	size exclusion chromatography
SeCys	selenocysteine
SeCys2	selenocystine
SelAl	selenoalbumin
SelP	selenoprotein
SEM-EDS	scanning electron microscopy-energy dispersive (X-ray) spectrometry
SeMet	selenomethionine
SF	sector field
SIM	single ion monitoring
SIMS	secondary ion mass spectrometry
sNP	single nanoparticle
SPE	solid phase extraction
SR	synchrotron radiation
SRM	standard reference material
SR	synchrotron radiation
IS	internal standard
ssIDMS	species specific isotope dilution mass spectrometry
STEM	scanning transmission electron microscope
SUIDMS	species unspecific isotope dilution mass spectrometry
TCE	technology critical element
TEM	transmission electron microscopy
TFA	trifluoroacetic acid
tHg	total mercury
TLC	thin layer chromatography
TMAH	tetramethylammonium hydroxide
TMAO	trimethylarsine oxide
TMSe	trimethylselenium
TOF	time-of-flight
TRIS	tris(hydroxymethyl)aminomethane
TXRF	total reflection X-ray fluorescence
UAE	ultrasound-assisted extraction
UHPLC	ultra high performance liquid chromatography
UHV	ultrahigh vacuum
USGS	United States Geological Survey
UV	ultraviolet
VG	vapour generation
XANES	X-ray absorption near-edge structure
XAS	X-ray absorption spectroscopy
XFM	X-ray fluorescence microscopy
XPS	X-ray photoelectron spectroscopy
XRF	X-ray fluorescence
XRD	X-ray diffraction

Conflicts of interest

There are no conflicts to declare.

References

1. R. Clough, C. F. Harrington, S. J. Hill, Y. Madrid and J. F. Tyson, J. Anal. At. Spectrom., 2022, 37(7), 1387–1430.
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., 2023, 38(1), 10–56.
3. M. Patriarca, N. Barlow, A. Cross, S. Hill, A. Robson, A. Taylor and J. Tyson, J. Anal. At. Spectrom., 2022, 37(3), 410–473.
4. E. H. Evans, J. Pisonero, C. M. M. Smith and R. N. Taylor, J. Anal. At. Spectrom., 2022, 37(5), 942–965.
5. C. Vanhoof, J. R. Bacon, U. E. A. Fittschen and L. Vincze, J. Anal. At. Spectrom., 2022, 37(9), 1761–1775.
6. S. Carter, R. Clough, A. Fisher, B. Gibson and B. Russell, J. Anal. At. Spectrom., 2022, 37(11), 2207–2281.
7. M. A. Z. Arruda and J. R. de Jesus, ICP-MS and Trace Element Analysis as Tools for Better Understanding Medical Conditions, Elsevier, Amsterdam, 2022.
8. V. K. Singh, J. Kawai and D. K. TripathiMagma Redox Geochemistry, Wiley, New Jersey, 2021.
9. R. Saran, Analytical Techniques for Trace Elements in Geochemical Exploration, The Royal Society of Chemistry, 2022.
10. M. L. Bishop, E. P. Fody, C. V. Siclen, J. M. Mistler and M. Moy, Clinical Chemistry: Principles, Techniques, and Correlations, 9th edn, 2023.
11. P. K. Srivastava, Arsenic in Plants Uptake, Consequences, and Remediation Techniques, 1st edn, Wiley, Hoboken, NJ, 2022.
12. A. Malik, M. K. Kidwai and V. K. Garg, Bioremediation of Toxic Metal(loid)s, 1st edn, CRC Press, Taylor & Francis Group, Boca Raton, 2023.
13. V. D. Rajput, Toxicity of nanoparticles in plants: An evaluation of Cyto/Morpho-physiological, biochemical and molecular responses, Academic Press is an imprint of Elsevier, London, United Kingdom, 2022.
14. S. Mandal and S. Lahiri, Microchem. J., 2022, 175, 107150.
15. M. E. I. Badawy, M. A. M. El-Nouby, P. K. Kimani, L. W. Lim and E. I. Rabea, Anal. Sci., 2022, 38(12), 1457–1487.
16. Y. Zhong, M. Ji, Y. Hu, G. Li and X. Xiao, J. Chromatogr. A, 2022, 1681, 463458.
17. D. Clases and R. Gonzalez de Vega, Anal. Bioanal. Chem., 2022, 414(25), 7337–7361.
18. D. Clases and R. Gonzalez de Vega, Anal. Bioanal. Chem., 2022, 414(25), 7363–7386.
19. D. Bazin, S. Reguer, D. Vantelon, J. P. Haymann, E. Letavernier, V. Frochotd, M. Daudon, E. Esteve and H. Colboc, C. R. Chim., 2022, 25, 189–208.
20. M. R. Beccia, G. Creff, C. Den Auwer, C. Di Giorgio, A. Jeanson and H. Michel, ChemPlusChem, 2022, 87(8), e202200108.
21. G. E. Batley and P. G. C. Campbell, Environ. Chem., 2022, 19(1), 23–40.
22. A. A. Krata, J. Sep. Sci., 2022, 45(18), 3624–3634.
23. S. Chen, J. Liu, J. Yan, C. Wang and D. Lu, Food Addit. Contam.,Part A, 2022, 39(3), 499–507.
24. L. A. Portugal, E. Palacio, V. Cerda, J. H. Santos-Neto, L. Ferrer and S. L. C. Ferreira, Chemosensors, 2022, 10(4), 139.
25. W. Fang, Y. Yang, P. N. Williams, H. Sun, H. Chen, D. Yang, X. Shi, R. Fu and J. Luo, Anal. Chem., 2022, 94(11), 4576–4583.
26. H. Z. Chen, W. Zhong and C. Y. Jing, J. Anal. At. Spectrom., 2022, 37(7), 1578–1586.
27. H. Li, K. Gong, X. Jin, G. Owens and Z. Chen, Chemosphere, 2022, 307(Pt 1), 135778.
28. C. D. Quarles, P. Sullivan, N. Bohlim and N. Saetveit, J. Anal. At. Spectrom., 2022, 37(6), 1240–1246.
29. B. Lajin, J. Feldmann and W. Goessler, Anal. Chem., 2022, 94(24), 8802–8810.
30. B. Krasnodebska-Ostrega, A. Drwal, E. Biadun and J. Kowalska, J. Anal. At. Spectrom., 2022, 37(2), 229–232.
31. F. Aslan and A. Tor, Chemosphere, 2022, 307(Pt 1), 135661.
32. E. P. Jahrman, L. L. Yu, W. P. Krekelberg, D. A. Sheen, T. C. Allison and J. L. Molloy, J. Anal. At. Spectrom., 2022, 37(6), 1247–1258.
33. M. M. Thosago, A. Botha, A. A. Ambushe and T. W. Godeto, Anal. Lett., 2022, 55(18), 2860–2875.
34. A. Raab, K. Kubachka, M. Strohmaier, M. Preihs and J. Feldmann, Environ. Chem., 2022, 9, 74–82.
35. J. Dai, Z. Tang, A. X. Gao, B. Planer-Friedrich, P. M. Kopittke, F. J. Zhao and P. Wang, Environ. Sci. Technol., 2022, 56(6), 3575–3586.
36. X. Li, Q. Ma, C. Wei, W. Cai, H. H. Chen, R. Xing and P. S. Song, Separations, 2022, 9(5), 105,  DOI:10.3390/separations9050105.
37. A. Spanu, I. Langasco, F. Barracu, M. A. Deroma, J. F. Lopez-Sanchez, A. Mara, P. Meloni, M. I. Pilo, A. S. Estrugo, N. Spano and G. Sanna, J. Environ. Manage., 2022, 321, 115984.
38. T. Sarwar, S. Khan, J. Nawab, S. Muhammad, S. Amin, J. Khan, A. Sarwar, I. Haider and Q. Huang, Exposure Health, 2022, 15, 299–313.
39. A. J. D. Perera, M. Carey, P. M. C. S. De Silva, C. Meharg and A. A. Meharg, Exposure Health, 2023, 15(1), 133–144.
40. A. Nawrocka, M. Durkalec, M. Michalski and A. Posyniak, Food Chem., 2022, 379, 132045.
41. W. O. Matos, F. L. F. da Silva, S. Sinaviwat, E. J. Menzies, A. Raab, E. M. Krupp and J. Feldmann, J. Trace Elem. Med. Biol., 2022, 71, 126968.
42. H. Cao, Z. Wang, J. Meng, M. Du, Y. Pan, Y. Zhao and H. Liu, Food Chem., 2022, 393, 133345.
43. G. Subramaniam, J. Bakar, N. Surugau and K. Muhammad, Food Anal. Methods, 2022, 15(10), 2858–2878.
44. Z. Huang, R. Bi, S. Musil, A. H. Petursdottir, B. Luo, P. Zhao, X. Tan and Y. Jia, Sci. Total Environ., 2022, 847, 157429.
45. A. Morales-Rodriguez, M. Perez-Lopez, E. Puigpelat, A. Sahuquillo, D. Barron and J. F. Lopez-Sanchez, J. Chromatogr. A, 2022, 1684, 463549.
46. J. Tibon, H. Amlund, A. I. Gomez-Delgado, M. H. G. Berntssen, M. S. Silva, M. Wiech, J. J. Sloth and V. Sele, Chemosphere, 2022, 302, 134906.
47. C. B. Miller, M. B. Parsons, H. E. Jamieson, O. H. Ardakani, R. T. Patterson and J. M. Galloway, Environ. Earth Sci., 2022, 81(4), 137.
48. A. Sharma, J. Guinness, A. Muyskens, M. L. Polizzotto, M. Fuentes and D. Hesterberg, Geoderma, 2022, 411, 115697.
49. P. W. Zhang, S. F. Wang, D. N. Zhang, Y. M. Wang, Y. Song and Y. F. Jia, Appl. Geochem., 2022, 142, 105344.
50. Y. Wang, P. Zhang, S. Wang, Y. Song, F. Xiao, Y. Wang, D. Zhang and Y. Jia, Chemosphere, 2022, 307(Pt 4), 135971.
51. N. M. Zaric, S. Braeuer and W. Goessler, J. Hazard. Mater., 2022, 432, 128614.
52. S. J. Yang, Y. Lee and S. H. Nam, J. Anal. Sci. Technol., 2022, 13(1), 7.
53. D. B. Wu, D. A. Li, L. L. Dong, G. L. Li, L. Wang, Z. Y. Tang, M. M. Rahman and S. P. Yang, J. Anal. At. Spectrom., 2022, 37(10), 2103–2110.
54. X. M. Wan, W. B. Zeng, M. Lei and T. B. Chen, Spectrosc. Spectral Anal., 2022, 42(2), 478–482.
55. D. R. Escudero, A. C. Isaguirre, M. F. Moyano, P. E. Hasuoka, P. Pacheco and M. M. Moglia, Int. J. Environ. Anal. Chem., 2022, 12.
56. J. W. Choi, Y. C. Song, N. Y. Cheong, K. Lee, S. Kim, K. M. Lee, K. Ji, M. Y. Shin and S. Kim, Environ. Res., 2022, 214(Pt 2), 113846.
57. R. Jiang, Q. Zhang, D. Ji, T. Jiang, Y. Hu, S. He, L. Tao, J. Shen, W. Zhang, Y. Song, Y. Ma, S. Tong, F. Tao, Y. Yao and C. Liang, Environ. Sci. Pollut. Res., 2022, 29(19), 28714–28724.
58. S. Huang, K. Chen, J. K. Leung, P. Guagliardo, W. Chen, W. Song, P. Clode, J. Xu, S. G. Young and H. Jiang, Anal. Chem., 2022, 94(40), 13889–13896.
59. X. Men, C. X. Wu, X. Zhang, X. Wei, W. Q. Ye, M. L. Chen, T. Yang, Z. R. Xu and J. H. Wang, Anal. Chim. Acta, 2022, 1226, 340268.
60. J. Tang, Q. Zhu, Y. Xu, Y. Zhou, L. Zhu, L. Jin, W. Wang, L. Gao, G. Chen and H. Zhao, Ecotoxicol. Environ. Saf., 2022, 241, 113776.
61. N. K. Smith, E. Keltie, E. Sweeney, S. Weerasinghe, K. MacPherson and J. S. Kim, Ecotoxicol. Environ. Saf., 2022, 232, 113269.
62. X. Wu, R. Yan, R. Guan, Y. Du, Y. Liu, S. Wu, S. Zhu, M. Song and T. Hang, Front. Pharmacol., 2022, 12, 761801.
63. Y. Li, H. Li, K. Zan, Y. Wang, T. Zuo, H. Jin, B. Zhang and S. Ma, Front. Pharmacol., 2022, 13, 795530.
64. J. L. Yan, Z. Tang, M. Fischel, P. Wang, M. G. Siebecker, M. G. M. Aarts, D. L. Sparks and F. J. Zhao, Plant Soil, 2022, 475(1–2), 379–394.
65. S. Markovic, M. Gabric, M. I. Razborsek, R. Milacic and J. Scancar, J. Food Compos. Anal., 2022, 108, 104422.
66. S. Markovic, L. Levstek, D. Zigon, J. Scancar and R. Milacic, Front. Chem., 2022, 10, 863387.
67. M. Djuric, L. Levstek, P. Oprckal, A. Mladenovic, A. M. Pranjic, J. Scancar and R. Milacic, Sci. Rep., 2022, 12(1), 15186.
68. E. Plestenjak, B. Kraigher, S. Leskovec, I. M. Mulec, S. Markovic, J. Scancar and R. Milacic, Sci. Rep., 2022, 12(1), 11.
69. S. J. Davis and W. R. Wise, J. Soc. Leather Technol. Chem., 2022, 106(4), 191–195.
70. A. Spinazze, D. Spanu, P. Della Bella, C. Corti, F. Borghi, G. Fanti, A. Cattaneo, W. R. Wise, S. J. Davis, D. M. Cavallo and S. Recchia, Int. J. Environ. Res. Public Health, 2022, 19(19), 12111.
71. L. Goodarzi, M. R. Bayatloo, S. Chalavi, S. Nojavan, T. Rahmani and S. B. Azimi, Food Chem., 2022, 373(Pt A), 131442.
72. P. R. Patricio, J. C. Quintao, L. H. M. da Silva and M. C. Hespanhol, J. Braz. Chem. Soc., 2022, 33(8), 978–986.
73. R. Pechancova, J. Gallo, D. Baron, D. Milde, P. Antal, Z. Slobodova, K. Lemr and T. Pluhacek, J. Biomed. Mater. Res., Part B, 2023, 111(2), 271–283.
74. Y. Ogra, Y. K. Tanaka and N. Suzuki, J. Clin. Biochem. Nutr., 2022, 71(1), 2–6.
75. C. Duroux and A. Hagege, Metallomics, 2022, 14(1), 9.
76. K. Markovi, M. Cemazar, G. Sersa, R. Milacic and J. Scancar, J. Anal. At. Spectrom., 2022, 37(8), 1675–1686.
77. K. Sommer, M. Sperling and U. Karst, Chemosphere, 2022, 300, 134528.
78. I. Strzeminska, C. Factor, J. Jimenez-Lamana, S. Lacomme, M. A. Subirana, P. Le Coustumer, D. Schaumloffel, P. Robert, J. Szpunar, C. Corot and R. Lobinski, Invest. Radiol., 2022, 57(5), 283–292.
79. L. Y. Zhang, X. L. Hou, T. Zhang, M. Fang, H. Kim, H. Jiang, N. Chen and Q. Liu, Anal. Chem., 2022, 94(27), 9835–9843.
80. Y. Z. Qi and H. Matsuzaki, Anal. Methods, 2022, 14(37), 3623–3631.
81. K. Kinska, G. Alchoubassi, L. Maknun, K. Bierla, R. Lobinski and J. Szpunar, J. Anal. At. Spectrom., 2022, 37(10), 2155–2164.
82. S. Fan, Y. Gao, B. Lai, E. J. Elzinga and S. Yu, Sci. Total Environ., 2022, 824, 153890.
83. H. Colboc, D. Bazin, S. Reguer, I. T. Lucas, P. Moguelet, R. Amode, C. Jouanneau, A. Soria, F. Chasset, E. Amsler, C. Pecquet, S. Aractingi, L. Bellot-Gurlet, L. Deschamps, V. Descamps and N. Kluger, J. Synchrotron Radiat., 2022, 29, 1436–1445.
84. G. Ignatavicius, M. H. Unsal, P. E. Busher, S. Wolkowicz, J. Satkunas, G. Sulijiene and V. Valskys, AIMS Environ. Sci., 2022, 9(3), 277–281.
85. D. Amico, A. Tassone, N. Pirrone, F. Sprovieri and A. Naccarato, J. Hazard. Mater., 2022, 433, 128823.
86. L. Gfeller, J. N. Caplette, A. Frossard and A. Mestrot, Environ. Pollut., 2022, 311, 119854.
87. D. D. Bussan, C. Douvris and J. V. Cizdziel, Molecules, 2022, 27(15), 4911.
88. S. Kulomaki, E. Lahtinen, S. Peramaki and A. Vaisanen, Talanta, 2022, 240, 123163.
89. L. Suarez-Criado, S. Queipo-Abad, A. Rodriguez-Cea, P. Rodriguez-Gonzalez and J. I. G. Alonso, J. Anal. At. Spectrom., 2022, 37(7), 1462–1470.
90. S. T. Lancaster, G. Peniche, A. Alzahrani, M. Blanz, J. Newton, M. A. Taggart, W. T. Corns, E. M. Krupp and J. Feldmann, Sci. Total Environ., 2022, 829, 154557.
91. E. R. Crowther, J. D. Demers, J. D. Blum, S. C. Brooks and M. W. Johnson, Anal. Bioanal. Chem., 2023, 415(5), 759–774.
92. D. Spanu, L. Butti, G. Boldrocchi, R. Bettinetti, S. Recchia and D. Monticelli, Anal. Chim. Acta, 2022, 1206, 339553.
93. K. Haraguchi, H. Akagi and A. Matsuyama, Anal. Sci., 2022, 38(1), 215–221.
94. Z. Z. Yao, J. X. Liu, X. F. Mao, G. Y. Chen, Z. H. Ma and B. R. Li, Food Chem., 2022, 379, 132116.
95. C. E. Pierce, O. S. Furman, S. L. Nicholas, J. C. Wasik, C. M. Gionfriddo, A. M. Wymore, S. D. Sebestyen, R. K. Kolka, C. P. J. Mitchell, N. A. Griffiths, D. A. Elias, E. A. Nater and B. M. Toner, Environ. Sci. Technol., 2022, 56(2), 1433–1444.
96. A. Cruz-Esquivel and J. Marrugo-Negrete, Acta Biol. Colomb., 2022, 27(1), 28–35.
97. R. J. Strickman, S. Larson, H. Huang, E. Kakouros, M. Marvin-DiPasquale, C. P. J. Mitchell and R. B. Neumann, Ecotoxicol. Environ. Saf., 2022, 230, 113143.
98. G. D. S. Coelho Junior, K. B. Fontana, T. A. Maranhao and D. L. G. Borges, Anal. Methods, 2022, 14(13), 1371–1377.
99. F. A. D. Cunha, M. J. de Oliveira, P. P. Florez-Rodriguez and J. C. C. Santos, Spectrochim. Acta, Part B, 2022, 192, 9.
100. L. Bjorkman, F. Musial, T. Alraek, E. L. Werner and H. J. Hamre, Acta Odontol. Scand., 2023, 81(4), 298–310.
101. B. May, S. Otto and R. Schweiggert, J. Agric. Food Chem., 2022, 70(33), 10349–10358.
102. A. M. Wroblewska, J. Samsonowicz-Gorski, E. Kaminska, M. Drozd and M. Matczuk, J. Anal. At. Spectrom., 2022, 37(7), 1442–1449.
103. Z. Z. Yao, B. R. Li, C. Li, B. H. Wang, M. Zhao and Z. H. Ma, J. Anal. At. Spectrom., 2022, 37(8), 1652–1657.
104. C. T. Wen, X. D. He, J. X. Zhang, G. Y. Liu and X. Xu, Food Funct., 2022, 13(10), 5498–5514.
105. S. L. C. Ferreira, V. Cerda, L. A. Portugal, L. B. Goncalves, J. H. S. Neto, J. B. Pereira and E. Palacio, TrAC, Trends Anal. Chem., 2022, 152, 116617.
106. Y. Zhou, X. R. Song, X. W. Yan, L. M. Yang, S. Chen and Q. Q. Wang, Anal. Chem., 5, 16975–16979.
107. A. Sentkowska and K. Pyrzynska, J. Pharm. Biomed. Anal., 2022, 214, 114714.
108. Y. Zhao, M. Wang, M. R. Yang, J. Zhou and T. T. Wang, Separations, 2022, 9(2), 21.
109. J. Lamarche, K. Bierla, L. Ouerdane, J. Szpunar, L. Ronga and R. Lobinski, J. Anal. At. Spectrom., 2022, 37(5), 1010–1022.
110. S. Ramirez-Acosta, R. Uhlirova, F. Navarro, J. L. Gomez-Ariza and T. Garcia-Barrera, Front. Chem., 2022, 10, 8.
111. J. J. Peng, Y. Liu, F. T. Yu, H. L. Fan, S. Y. Yue, Y. H. Fang, X. L. Liu and C. H. Wang, J. Chromatogr. A, 2022, 1685, 463590.
112. Z. Slejkovec, A. Stajnko, D. Mazej, M. J. Hudobivnik, O. Mestek, B. Lajin, W. Goessler, J. T. van Elteren and I. Falnoga, Anal. Bioanal. Chem., 2023, 415(2), 317–326.
113. R. Yang, Q. C. Li, W. J. Zhou, S. J. Yu and J. F. Liu, Anal. Chem., 9, 16328–16336.
114. A. A. Lashari, T. G. Kazi, J. A. Baig, H. I. Afridi and A. A. Memon, Water, Air, Soil Pollut., 2022, 233(11), 11.
115. W. Fang, Y. Yang, P. N. Williams, H. T. Sun, H. Y. Chen, D. X. Yang, X. Y. Shi, R. B. Fu and J. Luo, Anal. Chem., 2022, 94(11), 4576–4583.
116. M. Atasoy, D. Yildiz and I. Kula, J. Indian Chem. Soc., 2022, 99(7), 100511.
117. V. Losev, S. Didukh-Shadrina, A. Orobyeva, E. Borodina, E. Elsuf'ev, S. Metelitsa and U. Ondar, Anal. Methods, 2022, 14(28), 2771–2781.
118. M. Yeasmin, D. Lamb, G. Choppala and M. M. Rahman, J. Soil Sci. Plant Nutr., 2022, 22(3), 3243–3253.
119. Y. Luo, G. Chen, X. Q. Deng, H. Q. Cai, X. H. Fu, F. J. Xu, X. N. Xiao, Y. M. Huo and J. Luo, Separations, 2022, 9(9), 242.
120. Q. Ma, Q. H. Zhang, X. Li, Y. Gao, C. Wei, H. M. Li and H. Jiao, J. Chromatogr. A, 2022, 1674, 463134.
121. P. Anekthirakun, N. Bhawawet and A. Imyim, Chem. Pap., 2022, 76(12), 7333–7342.
122. A. Garcia-Figueroa, S. Musil and T. Matousek, Anal. Chem., 2022, 94(40), 13995–14003.
123. E. Jenikova, E. Novakova, J. Hranicek and S. Musil, Anal. Chim. Acta, 2022, 1201, 339634.
124. J. T. Yan, C. H. Zhang, C. L. Wang, D. B. Lu and S. Z. Chen, Microchem. J., 2022, 182, 107927.
125. S. Z. Chen, J. T. Yan, C. H. Zhang, C. L. Wang and D. B. Lu, Anal. Lett., 2023, 56(9), 1488–1501.
126. S. H. Nam, N. Y. Kim, E. S. Park and Y. Lee, Bull. Korean Chem. Soc., 2022, 43(3), 402–406.
127. F. Ruiz, Z. Benzo, A. Garaboto, J. Salas and J. L. Brito, J. Anal. At. Spectrom., 2022, 37(3), 668–676.
128. K. Kinska, E. Cruzado-Tafur, M. Parailloux, L. Torro, R. Lobinski and J. Szpunar, Sci. Total Environ., 2022, 809, 151090.
129. L. Ruzik and A. Dyoniziak, Molecules, 2022, 27(3), 1063.
130. B. Lajin, J. Feldmann and W. Goessler, Anal. Chem., 2022, 94(24), 8802–8810.
131. A. J. Nosti, L. C. Barrio, F. C. Celis, A. Soldado and J. R. Encinar, J. Proteomics, 2022, 256, 104499.
132. S. Erarpat, S. Bodur, O. T. Guenkara and S. Bakirdere, J. Pharm. Biomed. Anal., 2022, 217, 114855.
133. S. Sarpong-Kumankomah, K. B. Knox, M. E. Kelly, G. Hunter, B. Popescu, H. Nichol, K. Kopciuk, H. Ntanda and J. Gailer, PLoS One, 2022, 17(1), e0262160.
134. J. Yamazaki, I. Inoue, A. Arakawa, S. Karakawa, K. Takahashi and A. Nakayama, Anal. Methods, 2022, 14(22), 2219–2226.
135. Y. Y. Tang, L. H. Liu, Q. Y. Nong, H. Guo, Q. F. Zhou, D. Y. Wang, Y. G. Yin, J. B. Shi, B. He, L. G. Hu and G. B. Jiang, J. Chromatogr. A, 2022, 1677, 463303.

---