Atomic spectrometry update: review of advances in elemental speciation

Contents

• 1 Introduction

• 2 Books and topical reviews

• 3 CRMs and metrology

• 4 Elemental speciation analysis
  • 4.1 Actinides
  • 4.2 Antimony
  • 4.3 Arsenic
  • 4.4 Caesium
  • 4.5 Chromium
  • 4.6 Cobalt
  • 4.7 Gadolinium
  • 4.8 Gold
  • 4.9 Halogens
  • 4.10 Lead
  • 4.11 Mercury
  • 4.12 Nickel
  • 4.13 Phosphorus
  • 4.14 Platinum
  • 4.15 Selenium
  • 4.16 Tin
  • 4.17 Titanium
  • 4.18 Vanadium
  • 4.19 Zinc

• 5 Biomolecular speciation analysis

• 6 Abbreviations

• Conflicts of interest

• Data availability

• References

---

Abstract

This is the 18th Atomic Spectrometry Update (ASU) to focus on advances in elemental speciation and covers a period of approximately 12 months from January 2025. This ASU review deals with all aspects of analytical atomic spectrometry speciation methods developed for: the determination of oxidation states; organometallic compounds; coordination compounds; metal- and heteroatom-containing biomolecules, including metalloproteins, proteins, peptides and amino acids; and the use of metal tagging to facilitate detection via atomic spectrometry. As with all ASU reviews, the focus of the research reviewed includes those methods that incorporate atomic spectrometry as the measurement technique, although molecular detection techniques are also included where they have provided a complementary approach to speciation analysis. The number of publications covered this year has fallen to the lowest number since this ASU has been published. However, there is still a good breadth of elements covered, with the most popular elements still being As, Hg and Se, and the focus of many papers is on elucidating mechanisms and transformation pathways, particularly in plants and animals/humans, rather than concentrating on the speciation method. A number of papers covering speciation analyses of elements related to nuclear power generation have been published, which reflects the increasing demand for this power source in the global efforts to decarbonise. There has also been a report on the use of XRF spectroscopy as a detector for LC and three papers of note that describe comprehensive methods for metallobiomolecules. Once again, the quality of the abstract for many of the papers is poor, with detail on the methodology, key results with data included, conclusions and implications thereof missing. This is likely to lead to fewer researchers reading the article or considering it for reviews such as those produced by the ASU.

---

1 Introduction

This latest update adds to that from last year¹ and complements the five other annual Atomic Spectrometry Updates, advances in environmental analysis,² advances in the analysis of clinical and biological materials, foods and beverages,³ advances in atomic spectrometry and related techniques,⁴ advances in X-ray fluorescence spectrometry and its special applications⁵ and advances in the analysis of metals, chemicals and functional materials.⁶

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

This year’s speciation ASU is the shortest produced, reviewing less than 100 papers. The number of papers covered in the first 10 years of this ASU, 2009–2018, averaged at 250 and since then there has been a gradual decline to the present low point. This suggests that research funding is being sought for other areas of interest, with single NP and cell analysis being likely candidates as these research topics have shown a rapid growth in publications since 2018.

2 Books and topical reviews

No textbook devoted entirely to elemental speciation has appeared in the current review period; however, several books have been published in which elemental speciation topics are featured. Two of these cover the instrumental techniques commonly applied. In “Elemental Analysis” (five chapters, 470 pages) four well-known practitioners, including one current and one former ASU writer, dealt with all aspects of AAS, AFS, ICP-MS, ICP-OES and XRF spectrometry.⁷ Although speciation topics were mentioned in the chapter on ICP-MS, they did not feature to any great extent in the X-ray chapter, as neither X-ray absorption spectrometry nor synchrotron sources were discussed. In “Atomic Spectrometric Methods of Analysis”, the first volume in a new series from The Royal Society of Chemistry described as practical and technical guides for laboratory-based chemists, Andrew Fisher discussed (in nine chapters, 206 pages) AAS, AFS, ICP-MS, ICP-OES, LIBS, VG and XRF spectrometry.⁸ The book started with chapters on sample collection and preparation and finished with a chapter devoted to “Which technique to use”. At the more specialised level, in “Macromolecular Separation Science”, elemental speciation topics were mentioned several times throughout the book (12 chapters, 665 pages).⁹ Various chromatographic separation techniques were discussed, including hydrodynamic chromatography, polymer interaction chromatography and SEC, together with several modes of FFF. The author also covered high-temperature and multi-dimensional separations and, in the final chapter, several specialised techniques, including SFC and magnetic chromatography. The applications of 3D printing in analytical chemistry have now reached the stage of being the subject of a book. In “3D Printing in Analytical Chemistry: Sample Preparation, Separation, and Sensing” (18 chapters, 473 pages), 56 authors contributed to this comprehensive coverage.¹⁰ In addition to the three chapters in the section devoted to 3D printing in analytical separation techniques (fabrication of stationary phases, electrophoretic techniques, and mass spectrometry), there were also chapters on SPE (including SPME), membrane separation, and microfluidic devices. Elemental speciation topics were featured prominently in “Organometallic Compounds in the Marine Environment: Analytical and Ecological Aspects” (six chapters, 177 pages).¹¹ As may be imagined, compounds of Hg, Pb and Sn were the subject of the bulk of the topics discussed. One chapter of 25 pages was devoted entirely to chemical analysis topics. The availability of information about speciation underpins many of the other topics discussed, including stability, chemical reactions, biotransformations and toxicology (both eco and human).

Two reviews of very different instrumental techniques have appeared. Chandrakasan et al. have reviewed (143 references) recent advances in SR-based analytical techniques for investigating the occurrence, transformation and environmental fate of organic contaminants.¹² As such, the review focusses on non-metallic elements such as C, F and S. In addition to recent advances, the review contains a significant amount of tutorial material, in which the advantages and limitations of the various techniques were explained. The reviewers showed how multi-scale analyses, combining spectroscopic and imaging approaches, have enabled precise tracking of contaminants, including microplastics, pharmaceuticals, hydrocarbons, persistent organic pollutants and PTEs, within complex environmental matrices (soil, sediments, water and air) with high spatial resolution, chemical specificity and good detection capability. The authors considered that their cross-disciplinary perspective provided a practical guide for environmental scientists seeking to apply the analytical capabilities of synchrotron facilities. Fadillah et al. have reviewed (125 references) advanced electrodialysis techniques for analytical separations.¹³ Part of the review was a tutorial introduction explaining how the presence of an electric field to drive ions across ion-exchange membranes enables the isolation and enrichment of target ionic species and matrix removal. The review covered how advanced membranes, such as monovalent cation-permselective membranes, chelate-type membranes, and chiral-imprinted membranes, have further enhanced the selectivity, resolution, and stability of electrodialysis systems, expanding their application in fields such as environmental monitoring, biomedical analysis, and industrial process control. Elemental speciation was covered in a separate section of the review (34 references) with an emphasis on Cr^III/Cr^VI, though procedures for As and Hg were also mentioned. The reviewers pointed out the advantages of electrodialysis in providing clean baselines, low reagent consumption, improved LOD values and extended instrument lifespans by eliminating interfering species prior to analysis. They also noted that developments in flow-based, microfluidic systems have created new possibilities for automated and miniaturised analytical platforms.

3 CRMs and metrology

The development of a new CRM, INM-0391-1 striped catfish (Pseudoplatystoma fasciatum), certified for total Hg and with an informative MeHg content has been reported.¹⁴ The total Hg in the freeze-dried fish muscle was extracted with a 3 : 2 mixture of HNO₃ : H₂O₂ and MAE. The MeHg was extracted with TMAH in 25% KOH at 80 °C for 8 h, followed by derivatisation with 2% NaBEt₄ and extraction of the product into a 9 : 1 mixture of toluene and isooctane. Total Hg was determined by CV-AAS with a bracketing calibration and ICP-MS with a gravimetric calibration. The derivatised MeHg was quantified using GC-MS fitted with a 5% diphenyl/95% dimethylpolysiloxane column with He as the carrier gas and a linear temperature program. The total Hg mass fraction is given as 3.94 ± 0.27 mg kg⁻¹ whilst the informative MeHg mass fraction was found to be 3.79 ± 0.31 mg kg⁻¹. The major uncertainty components, characterisation, homogeneity and long-term stability were 25 and 47%, 24 and 20% and 48 and 37% for Hg and MeHg, respectively. Method validation was by the analysis of DORM-4 and ERM-CE464 CRMs.

The assignment of metrological traceable values for QC materials that can be used for specific metalloproteins has the potential to make significant improvements to routinely used biomedical assays. The reliable quantification of protein biomarkers is limited by the lack of specific and traceable QC and RM for use in routine clinical laboratories. In a recent report using HPLC coupled to ICP-MS/MS, sulfur ssIDA has been used to assign an SI traceable value to human ferritin light chain (FTL), a relatively low-abundance serum protein biomarker at supraphysiological concentrations in a WHO standard.¹⁵ The reported methodology used proteins prepared recombinantly and then isotopically enriched for the SI traceable quantification of this biomarker for the diagnosis of Fe-related disorders such as hemochromatosis. The separation of species in standards and samples was achieved using a SAX column (MonoQ 5/50 GL) and elution with an optimised gradient of 0.5 mol L⁻¹ NHOAc in Tris–HCl pH 7.40. The ICP-MS/MS instrument was operated in O₂ mode and S isotopes were monitored with a mass shift of m/z +16 to resolve the interferences. Natural and ³⁴S isotopically enriched human FTL standards were recombinantly expressed using the HEK-293 mammalian cell system. Both standards were fully characterised for their total S content, isotopic abundance, species distribution and primary sequence before being used for IDA. Confirmation of structural identity was also undertaken by HR LC-MS/MS. Treatment of serum spiked with FTL was optimised to selectively remove high abundant serum proteins without significantly affecting the integrity of the analyte being measured. This was achieved using methanol protein precipitation in combination with heat treatment and ultra-filtration. Complementary asymmetrical FFF and AEC coupled to ICP-MS detection, were used to monitor the impact of sample treatment on protein removal and FTL recovery. Both natural and enriched standards, were used with the optimal sample treatment in a double ssIDMS workflow for the determination of FTL in the WHO 4th International Standard for Ferritin (NIBSC 19/118). The SI traceable value for FTL using the developed procedure was 11.0 ± 1.0 µg mL⁻¹ representing a relative combined expanded uncertainty of 9.4% (k = 2). This was in good agreement with the assigned value and its associated uncertainty (10.5 ± 0.3 µg ampoule made up in 1 mL; k = 2.23). For the first time S ssIDA was used to assign an SI traceable value to a relatively low-abundance serum protein biomarker in a WHO standard. The developed methodology will be invaluable for metrology institutes and quality control material producers for adding SI traceability to QC materials and for the certification of new biomarker RMs.

4 Elemental speciation analysis

4.1 Actinides

Lohmann et al. investigated the carbonate complexation of Am^III, Np^V, Th^IV and U^VI by CE-ICP-MS.¹⁶ The work constitutes a contribution to the safety analysis of a high-level nuclear waste repository, which requires an understanding of the diffusion and sorption behaviour of long-lived actinides under environmentally relevant conditions (trace concentrations of the elements and appropriate concentrations of ligands, such as hydroxide and carbonate). They determined the formation constants for the three successive actinide carbonate complexes in a medium of ionic strength 0.33 mol L⁻¹ containing different alkali metal (Li, K and Na) chlorides. Separation occurred in a fused silica capillary (50 µm id, 50 cm long) into which the 15 nL sample was injected hydrodynamically at 100 mbar for 5 s, then 10 kV and 60 mbar were applied at constant temperature (25 °C). A large number of background electrolytes were used with pHs ranging from 1.6 to 11. To aid aerosol formation, the eluent was merged with a make-up electrolyte consisting of 1.25% HNO₃, 10% EtOH and 5 µg L^{−1 7}Li, ²⁴Mg, ⁵⁹Co, ⁸⁹Y, ¹⁴⁰Ce, and ²⁰⁵Tl internal standards, delivered at a flow rate of 15 µL min⁻¹. 2-Bromopropane, added to the samples, allowed the electroosmotic flow to be detected via ⁷⁹Br. In the case of Th^IV, they found that, in contrast to the behaviour of the other actinides investigated, well-defined peaks for Th were produced only at pH < 2.6. Above this value, peaks in the electropherograms showed considerable tailing or were so ill defined that no peak could be discerned. Even though samples were measured directly after Th^IV addition, there were clear signs of precipitation and colloid formation. They deduced that the significant amount of tailing indicated retention on the fused silica of the capillary via surface complexation and therefore the electrophoretic mobilities determined from the front of the signal were not sufficiently reliable for the determination of complex formation constants. They considered that fused silica capillaries were not suitable and cited recent work by other researchers who studied the Ca–U^VI–CO₃ system by CE in PEEK capillaries with detection by ICP-MS. For U^VI, they found significant discrepancies between the stability constant values they calculated and those already in the literature. In the case of UO₂(CO₃)₃⁴⁻, the difference of three orders of magnitude was explained by the fact that previous studies had been made at concentrations of around 3× 10⁻⁵ mol L⁻¹, at which concentrations of polynuclear complexes make a significant contribution to the speciation. However, at the concentrations used in this study (10⁻⁷ mol L⁻¹), polynuclear complexes were considered unlikely. The researchers also pointed out that the results of previous studies based on solubility measurements were highly dependent on the characterisation of the relevant solid phases, whereas complex formation can be studied directly in the liquid phase by CE-ICP-MS. They also showed that alkali metal cations had very little influence on the first two U^VI–CO₃ complexes, but that in the case of the 1 : 3 complex, the formation of ternary complexes was needed to accurately model the limiting mobility of U^VI. They determined the complex formation constants of XUO₂(CO₃)₃³⁻ and X₂UO₂(CO₃)₃²⁻ for the alkali metal cations, where X⁺ is one of Li⁺, Na⁺, K⁺, Rb⁺ and Cs⁺. For Am, they found that the electropherograms did not show reliable peaks between pH 5 and 8 (where Am(CO₃)²⁻ predominates), which they ascribed to strong interaction or sorption of Am^III on the fused silica capillary by analogy with known similar behaviour of Cm. At pH > 8, clear peaks for Am were observed, indicating that the formation of Am(CO₃)³⁻ decreases the sorption on the fused silica. They were able to calculate formation constants for this species as well as those for Am(HCO₃)²⁺ and Am(CO₃)⁺. With regard to the role of alkali metal cations, they found that, in general for highly negatively charged actinide-carbonate complexes, there was a significant effect on electrophoretic mobility, though they were unable to estimate the formation constants of any of the ion-association constants. In the case of Np^V, the free species is considered to be NpO₂⁺, and thus the carbonate complexes are NpO₂(CO₃)⁻, NpO₂(CO₃)₂³⁻, and NpO₂(CO₃)₃⁵⁻, which they considered only formed to 30%, and that for pH > 11, the predominant species was NpO₂(CO₃)₂OH.⁴ The formation constants determined were in agreement with previously determined values by other researchers. Experiments to determine the alkali metal association were complicated by the fact that Np^V carbonate complexation does not reach a point where a limiting carbonate complex is the sole species present: there are always at least two carbonate species present in significant proportions. They deduced from the impact of the presence of K ions on the calculated formation constants that the highly negatively charged Np^V complexes (carbonate or hydroxy carbonate) also form associated complexes with the alkali cations. They were able to calculate complex formation constants for KNpO₂(CO₃)₂²⁻, Li₂NpO₂(CO₃)₂⁻ and Na₂NpO₂(CO₃)₂⁻.

Lohman et al. have also studied the Th^IV-acetate system by CE-ICP-MS with essentially the same instrumental set-up.¹⁷ They explained how kinetically stable species were characterised by evaluating the peak areas, while labile species were characterised through variations in electrophoretic mobility. They discovered that previous work in which it was generally accepted that complexes with one to five acetate ligands are formed, was not entirely correct. They did not detect the Th(acetate)₅⁻ species, instead they discovered two hydrolysed species: Th(OH)(acetate)₃ and Th(OH)(acetate)₄⁻. They determined formation constants for all species, from which they constructed a speciation plot showing that above pH 6 only the two hydrolysed species were present and that the Th(OH)(acetate)₄⁻ species was dominant with a relative concentration of about 96%.

To investigate the distribution and speciation of Th in a titania slag (the Rio Tinto chloride slag, RTCS, from the Rio Tinto Fer et Titane operation in Quebec, Canada), Chadirji-Martinez et al. applied several analytical techniques.¹⁸ The slag, which is about 90% titania, contained about 170 mg kg⁻¹ Th and 16 mg kg⁻¹ U and was slightly radioactive, raising some environmental concerns. Based on the results from powder XRD, ICP-MS, LA-ICP-MS, EPMA, SEM, Raman spectroscopy, µSXRFM, synchrotron Laue XRD and synchrotron XAS, they showed that about 99.4% of Th in the RTCS occurred in a chevkinite-like Th-REE-Ti aluminosilicate containing an average of 8.05 ± 0.64% ThO₂, which occurred as acicular or tabular crystals in association with a Th-bearing aluminosilicate glass (0.41 ± 0.35% ThO₂). They observed that crystals occurred as infillings either in interstitials or along the fractures of the main Ti–Fe oxides of the sassite-ferropseudobrookite solid-solution series. From LA-ICP-MS analyses and µSXRF mapping, they showed that the main Ti–Fe oxides in the RTCS contained an average of only 0.32 ± 0.60 mg kg⁻¹ Th. They concluded that future pyrometallurgy operations with Th- and U-bearing minerals must consider the environmental effects and mitigate radioactivity. On the upside, they proposed that preferential acid dissolution of the Th-REE-Ti aluminosilicate in RTCS and other titania slags could recover Th and REE, with both environmental and economic benefits.

Sun et al. also measured formation constants, of CaUO₂(CO₃)⁻ and Ca₂UO₂(CO₃)⁻ complexes, by CE-ICP-MS, but were able to largely overcome the limitations of silica capillaries (interaction with the analytes and dissolution at high pH values) by using PEEK tubing.¹⁹ The separations were conducted using a commercial CE system with capillaries of inner diameters 50, 75 or 100 µm, outer diameter 360 µm, and 105 cm long. The background electrolyte was 0.1 mol L⁻¹ NaCl with 25 mmol L⁻¹ Tris buffer. The samples were injected at 50 mbar for 7 s, and the electrophoresis was carried out at the desired voltage between −5 and +5 kV, under a pressure of 50 mbar. The CE system eluent was merged with a make-up solution (90% background electrolyte + 10% EtOH, w/w) at about 10 µL min⁻¹ and introduced to the single-quadrupole ICP-MS instrument via a MiraMist CE nebuliser. The electroosmotic flow was measured directly by ICP-MS by monitoring the Ga signal of a stable macrocyclic neutral complex with 2,2′,2″-(1,4,7-triazacyclononane-1,4,7-triyl)triacetic acid. The impact of Joule heating was evaluated and found not to be problematic. However, on prolonged use, an interaction between the analyte species and the capillary was observed. As all the U could be retrieved by the electrolyte flow, the researchers assumed that the interaction was always against the direction of this flow, and they were able to compensate for the effect by performing electrophoresis twice: first with a positive voltage and then with a negative voltage. Calculations of formation constants from the results of this two-mode strategy were in agreement with literature values. The researchers concluded that their results verified the thermodynamic data available for these species, and validated PEEK-based CE for the measurement of thermodynamic constants of inorganic complexes in alkaline media.

4.2 Antimony

As Sb species can exhibit isotopic variability, with isotopic fractionation occurring during species transformations, the ability to determine isotopic compositions of Sb species in environmental samples can provide valuable insights into the biogeochemical cycling these species and their impact on the environment. Chen et al. have developed a method for the determination of the isotopic composition of inorganic Sb^III and Sb^V species by HPLC-HG-ICP-MC-ICP-MS.²⁰ They separated the species by AEC (Hamilton PRP X100 column) with a mobile phase of 10 mmol L⁻¹ EDTA and 1 mmol L⁻¹ KHP. To overcome the mismatch in flow rates between the HPLC eluent and the nebuliser uptake rate, increase sensitivity and prevent the uptake of mobile phase salts, they converted the eluting analyte species to SbH₃ by merging the eluent first with HCl and then with BH. Despite the additional broadening, the peaks were baseline separated for a total analysis time of 10 min. They observed that the combination of HPLC separation and HG caused notable isotopic fractionation in the chromatographic signals. They optimised a number of parameters affecting precision, including the use of In as an IS, the HG conditions, transient signal processing and peak integration windows, achieving external 2 SD precisions better than 0.05 parts per thousand for both species. They applied their method to the analysis of some standard solutions (including NIST SRM 3102a) and two IGGE RMs GSD-11 and GSD-12 (stream sediments) obtaining δ¹²³Sb values that were characterised as “in excellent agreement with previously reported data”. They also studied isotopic fractionation during the reduction of Sb^V to Sb^III, a commonly used procedure in the determination of Sb by HG, as the sensitivity for Sb^V in HG procedures, especially those with the limited reaction times in flowing streams, is significantly lower than that for Sb^III. For reduction by KI, they found preferential reduction of the light Sb isotope (¹²¹Sb) with consequent enrichment of the remaining unreacted Sb^V with respect to ¹²³Sb, leading to an increase in δ¹²³Sb of up to 1.18%. They also noted that their method required smaller samples than other methods, needing only 5 ng for Sb^V and 0.5 ng for Sb^III. When total Sb was determined in FI mode, just 0.25 ng was needed.

4.3 Arsenic

The speciation of As is now routine in many areas of study with robust methodologies available to determine a wide range of As species in a variety of matrices. The number of publications discussing novel applications for As speciation drops year by year, and reports of both instrumental and method developments are now few in number. However, there is continued interest in As speciation in some specific areas of study, and these continue to attract attention.

The speciation of As in water samples, particularly potable water, has received much attention over the years, and many research papers have been published in this area. The subject has also been covered by many reviews, some more authoritative than others. This year, a review of this topic has been published by Otiwa et al.²¹ The review considers spectroscopic detection methods for As in water samples and cites 161 references, although over half of these are over 10 years old. The review is rather focused on more traditional techniques such as HG-AAS, and the text is a little dismissive of developments in the chromatographic separation of As species from water samples. The review could also make better use of Tables. However, for those interested in an overview of traditional methods to detect As species, principally As^III and As^V, this review cites a lot of the key publications in this area. A common problem encountered in As speciation studies of waters is the possibility of post-sampling changes, such as oxidation during storage, treatment, and analysis, hindering the accurate quantification of As^III and As^V. A flow-type electrodialytic ion-transfer device (ITD) capable of the immediate on-site separation of As species and particularly targeting the selective isolation of As^III has been reported for use with HG-AAS.²² The ITD comprised donor, acceptor, and isolator channels, and it enabled efficient separation of As^III and As^V by applying 100 V across the system. The As^III was selectively transferred through the donor channel, while As^V migrated into the acceptor channel along with strongly acidic anions, such as chloride and sulfate. Using this approach, it was found that the As^III : As^V ratio in a hot spring water sample significantly differed between immediate on-site separation and separation after 1 month storage, demonstrating the risk of As^III oxidation. The trace As species in hot spring water were quantified by HG-AAS, while coupling the ITD with ion-exclusion chromatography facilitated the separation of As^III and silicate in brine, with desalination effectively achieved by the ITD. Selective sampling and in situ detection of As^III and As^V in groundwater using diffusive gradients in thin-films have again been reported this year.²³ In this study, two selective adsorption matrices, thiol-functionalised mesoporous silicon-modified polyacrylamide gel (MCM-41-SH-PAM) and polyethyleneimine functionalised polyacrylamide gel (PEI-PAM) were developed as the binding phases. The MCM-41-SH-PAM achieved 99.5% adsorption of As^III within 12 h with only 0.6% cross-adsorption of As^V, whilst the PEI-PAM adsorbed over 99.0% of the As^V within 1 h with only 3.8% adsorption of As^III even after 12 h. The adsorbed As^III and As^V were then eluted using 0.5 mol L⁻¹ HNO₃ and H₂O₂ (99 : 1 v/v) and 0.5 mol L⁻¹ HNO₃ with recovery rates of 80% and 94.5%, respectively. Using ICP-MS detection, the method achieved LOD values of 0.04 µg L⁻¹ for As^III and 0.02 µg L⁻¹ for As^V. A study to examine As contamination in the Caplina Basin, Tacna, Peru, focusing on As speciation in surface waters used for drinking, has been published by Sosa et al.²⁴ Using ICP-MS, total As levels ranging from 0.0304 mg L⁻¹ to 0.0730 mg L⁻¹ were determined, exceeding the WHO limit of 0.01 mg L⁻¹ for drinking water. Speciation by HPLC-ICP-MS found As^III levels reached up to 0.0154 mg L⁻¹ at one sampling point, characterised by acidic pH (3.82) and high conductivity (1456 µS cm⁻¹) whilst As^V peaked at another more alkaline site (pH 8.42) at 0.0260 mg L⁻¹. Organic species such as DMA and MMA were also detected, indicating active methylation processes and anthropogenic influences. The risk assessment revealed carcinogenic risk values ranging from 2.56 × 10⁻⁵ to 6.28 × 10⁻⁵, below the US EPA threshold (1 × 10⁻⁴), but significant for prolonged exposure. Non-carcinogenic hazard quotient (HQ) values ranged from 0.09 to 0.21. The findings highlighted severe public health implications and emphasised the urgent need for comprehensive strategies, including continuous monitoring, targeted treatment technologies, and community education, in order to guarantee water quality in the most vulnerable areas covered in the study.

The speciation of As in foodstuffs continues to attract attention, but the focus of many studies has shifted to the health implications of As in the diet rather than the elucidation of new As species. A method for multi-matrix As speciation analysis by HPLC-ICP-MS has been developed by Ghaffour et al. for use in the third French total diet study.²⁵ To optimise the sample preparation process, microwave and heat-assisted extraction techniques were evaluated and compared. Separation of four As species (As^III, As^V, DMA, and MMA) was achieved with baseline resolution within 10 min, utilising an AEC with mobile phase gradient of 0.5 and 5 mmol L⁻¹ ammonium carbonate with 3% (v/v) MeOH at a pH of 9.3. The method validation followed the accuracy profile approach, employing six different food matrices analysed in duplicate over six separate days within a 6 weeks period. The method demonstrated good intermediate precision, with coefficients of variation ranging from 4.7% to 5.5%, while the bias was <3%. The LOQ for the four species was 6.25 µg kg⁻¹ (dw), and the LOD was 1.88 µg kg⁻¹ (dw).

Several groups have reported studies evaluating As speciation in rice. A method aimed at reducing the overall retention time to less than 4 minutes while enhancing peak separation when using HPLC-ICP-MS for the determination of As^III, As^V, DMA, and MMA in 30 white and brown rice samples from Malaysia has been reported.²⁶ The method incorporates chromatographic modifiers and ion-pairing agents (2 mmol L⁻¹ 1-octanesulfonic acid sodium salt, 2 mmol L⁻¹ malonic acid, 4 mmol L⁻¹ TMAH) in the mobile phase. The method was successfully validated using the NIST SRM 1568b Rice Flour. Rice samples, obtained from a supermarket in Malaysia, were finely ground using a high-speed blender, then sifted through a 60-mesh sieve (<250 µm) and stored in 50 mL centrifuge tubes. The powered samples were dried at 60 °C over a period of two days in a forced convection oven to constant weight. No significant association was found between As levels and place/country of origin, but some samples, both brown and white rice, exceeded the European Commission’s limit for iAs. The highest levels were in samples labelled ‘organically cultivated’. Health risk assessments showed all rice samples had a target hazard quotient above 1, potentially indicating non-carcinogenic adverse health risks. Additionally, estimated cancer risks exceeded the 1 in 1000 lifetime risk threshold under the revised cancer slope factor (CSF) value. A similar study of rice from Croatian Supermarkets has determined the concentrations of 29 essential and toxic elements and four As species (As^III, As^V, DMA, MMA) in 58 rice samples.²⁷ In addition, the influence of rice variety, cultivation methods, and origin on the composition of trace elements was assessed. The elemental concentrations were determined using ICP-MS, and As species were quantified with HPLC-ICP-MS. Organic brown rice had higher concentrations of essential trace elements (Co, Cr, Cu, Fe, Mn, Se and Zn) than white rice, with organic brown rice containing more essential elements than conventionally grown rice. The average total As concentration in all samples was 142 ± 57 µg kg⁻¹ (ranging from 36.7 to 259 µg kg⁻¹ dry weight) with brown conventionally grown rice containing a higher amount. Arsenite was the predominant As species and regional differences in As concentrations were observed. A study to significantly reduce the extraction time for nine As species in rice using high-pressure MAE has been reported by Zhao et al.²⁸ The proposed extraction method took 3 min and increased the extraction efficiency by up to 2.5 times when compared to UAE, heat-assisted, and MAE methods under the same conditions (solvent, temperature, and time). The nine As species extracted were: As^III, As^V, AB, AC, DMA, MMA, ROX, 4-nitrophenylarsonic acid (NIT), p-hydroxyphenylarsonic acid (NAPP). After demonstrating the improvement in extraction efficiency, other factors including extraction solvent, temperature and time were also optimised. For a 0.50 g rice sample, optimal extraction conditions were found with 5 mL of HCl (0.023 mol L⁻¹), 45 bar, 90 °C and 3 min for the nine As species. For rice, the LOD and LOQ values were from 0.041 to 0.12 µg kg⁻¹ and from 0.12 to 0.36 µg kg⁻¹, respectively. Method validation was by spiking, at three levels, two randomly selected rice samples and the recoveries ranged from 88.6 to 103.1% with RSDs between 1.54% to 4.55%.

The speciation of As in mushrooms is known to be diverse. A profile with unusual As compounds has been identified in the edible mushroom Sparassis crispa by Steiner et al.²⁹ The study analysed four different fruiting bodies of S. crispa originating from Austria and Czechia. The As species were separated on an UHPLC system comprising a binary pump, an autosampler, and a heated column compartment at 40 °C. Both cation-exchange (Zorbax 300-SCX) and anion-exchange (Hamilton PRP-X100) chromatographic columns were used with mobile phases of 30 mmol L⁻¹ NH₄HCO₂, pH 2.3 and 20 mmol L⁻¹ (NH₄)₂HPO₄, pH 6.0 respectively. High-resolution mass spectra were obtained using ES-TOF. Besides the previously known As species AC, trimethylarsine oxide, trimethyl(2-carboxyethyl)arsonium, arsenobetaine amide, and the tetramethylarsonium ion, the researchers identified the uncommon alpha-glycerophosphorylarsenocholine as a major As species. This compound has previously only been identified as a minor As species in marine animals. Furthermore, a novel As compound, beta-methyl arsenocholine, was identified as a naturally occurring compound in all investigated samples of S. crispa.

The determination of As in seafood continues to attract attention. Liu et al.³⁰ have reported on the determination and monitoring of six As species (As^III, As^V, AB, AC, DMA, and MMA) in various seafood samples. Species separation and detection was by HPLC-ICP-MS with an AEC analytical column and the chromatographic and sample preparation conditions optimised. A relatively low buffer-salt concentration (50 mmol L⁻¹) was selected. The method was validated using CRMs (Tuna Fish Tissue BCR-627, and Lobster Hepatopancreas TORT-3, NRCC). Baseline separation of the As species was achieved within 12 min, and oven heat extraction was selected for sample preparation. The method gave a linear calibration for the As species from 0.1 to 100 µg L⁻¹, with recoveries ranging from 92 to 123%, intra-day precision of 0.9–5.0%, and inter-day precision from 1.1 to 19.4%. The developed method was used to determine the As species in 109 molluscs, 195 crustaceans, and 195 marine fish samples from Guangdong, China. Distinctive distribution patterns for the As species were observed among different seafood categories, in-line with other studies. The predominant species detected across all samples was AB, with a detection frequency higher than 97.9%, and it contributed to more than 90% of the total As. The other As species showed marked variations in their occurrence and concentrations. In another study, As^III, As^V, AB and MMA have been determined in anchovy, haddock, mussels and prawns using HPLC-ICP-MS after UAE sample preparation.³¹ The separation of the As species was achieved using 50 mmol L⁻¹ (NH₄)₂CO₃ and 0.50 mmol L⁻¹ EDTA in 1% methanol at pH 9.50 as the mobile phase. The LOD values of the method were 0.12, 0.03, 0.07and 0.06 ng mL⁻¹ for As^III, As^V, AB and MMA, respectively. The RSD calculated for the lowest calibration standards ranged between 1.01 and 7.88% and the accuracy, assessed using spike recovery experiments, was 85–117%. A CRM was also analysed using the method for total As, and gave good agreement with the certified values, although the choice of matrix, NIST-1573A Tomato leaves, was surprising for this study. Finally, the method was applied for the qualitative and quantitative determination of As^III, As^V, AB and MMA in the seafood samples with the four species giving from 6.0 to 5700 ng g⁻¹ As and accounting for between 78 and 97% of the total As in the samples. The carcinogenic and non-carcinogenic risk for haddock and anchovies was low, whereas the carcinogenic risk for prawns was relatively high but below the US EPS threshold value. A method for the quantification of As^III, As^V, AB and MMA in dried seafood via pressure tank-assisted extraction with HPLC-AFS has been presented.³² The key parameters, which included extraction temperature, extraction time, and extraction liquid concentration, were optimised using orthogonal testing at three levels. The optimal extraction conditions were identified as 1% HNO₃, an extraction temperature of 150 °C and an extraction time of 75 min. The results demonstrated that the four As species could be completely separated within 6 min. Good linear relationships were observed in the concentration range from 1.00 to 100.0 µg L⁻¹ for all species, with LOD values ranging from 0.21 to 0.85 µg L⁻¹. Spike recoveries were between 88.6 and 103%, with RSDs ranging from 1.4 to 4.8%. The method was applied to determine the concentrations of As species in a range of dried seafood products (dried shrimp meat and dried mussel, clam meat, dried squid, whitebait and cowfish). The total As concentration was also quantified using AFS after wet digestion. The total concentration of As determined in the samples was between 4.22 mg kg⁻¹ and 24.4 mg kg⁻¹, and the extraction efficiency, calculated as the percentage of the total As species content extracted by the method to the total As concentration, ranged from 1.35 to 13.5% (AB, AC and any arsenosugars were not determined in this study). Although the results indicated that the total As content in the dried seafood products was high, the iAs content in the tested samples did not exceed the regulatory limit.

The determination of As species in seaweeds continues to be reported. The use of IC-ICP-MS to determine As^III, As^V, AB, DMA and MMA in seaweed has been reported by Thang et al.³³ An optimised extraction procedure was established, involving a 20 mL MeOH : H₂O (1 : 1, v/v) solution at 50 °C for 10 min. Chromatographic separation of all the species was achieved in less than 7 min in gradient elution mode using AEC, and the mobile phase consisted of 300 mmol L⁻¹ (NH₄)₂CO₃ and 3% MeOH at pH 10.3. The linear range for each As species was from 0.1 to 100 µg L⁻¹, and the LOD values for As^III, As^V, AB, DMA and MMA were 4.29, 2.28, 21.2, 4.76 and 4.29 µg kg⁻¹, respectively. The LOQ values for the same species were 14.3, 7.62, 70.63, 15.86 and 9.27 µg kg⁻¹, respectively. Analysis of 15 seaweed samples revealed the presence of As^III in some samples, while As^V was not detected. The acquisition of reliable speciation data for dietary exposure assessment for seaweeds is challenged by the lack of calibration standards for arsenosugars, the predominant As species in algae. A study to investigate the retention and elution behaviour of anionic arsenosugars (PO₄-Sug, SO₃-Sug and SO₄Sug) using various dispersive solid-phase (d-SPE) sorbents – SAX, primary secondary amine (PSA) and HILIC – has been reported by Morales-Rodríguez et al.³⁴ The aim was to assess their suitability for isolating and preconcentrating these compounds from aqueous algae extracts, not for quantification purposes, but as a strategy for subsequent standard preparation. Key parameters included extract pH, adsorption time, ionic strength, and elution solvent. The d-SPE eluates were analysed by ICP-MS for total As and IC-ICP-MS for As speciation. Among the sorbents tested, SAX was the most effective, achieving approximately 90% retention of arsenosugars. Two consecutive elution steps with 0.5% ammonium formate enabled recoveries ranging from 75 to 88% of the retained compounds in the studied algae samples. A method for As speciation in the edible seaweeds Ascophyllum nodosum and Fucus vesiculosus L., and the edible prokaryote Arthrospira platensis, has been reported using enzyme-assisted extraction and LC-ICP-MS.³⁵ The total As content and extraction efficiencies were evaluated using both conventional and enzyme-assisted extraction methods. Of the 11 methods evaluated, enzymatic extractions were generally more effective than non-enzymatic in preserving individual As species. Among the tested enzymes, Viscozyme™ at 37 °C for 24 h provided the most consistent recovery of both organic and inorganic species, while SDS (serine dehydratase) at 37 °C for 24 h achieved the highest total As yield for brown algae (Ascophyllum nodosum and Fucus vesiculosus), though with poorer preservation of the speciation. Overall, the extraction efficiency for total As in this study ranged from 35 to 41% for the most effective protocols. While these values may seem low, they are typical for algae matrices, where As is often bound to complex structures preventing complete recovery without more harsh treatments, which may change the integrity of the species. Despite this limitation, the optimised procedure gave repeatable and reliable identification of the major As species, and enabled the identification of As^III, As^V, DMA, MMA and AB in all tested samples. The results indicated significant interspecies variation, with the highest As concentration found in brown algae (62.1 µg g⁻¹ dry weight) and the lowest in Arthrospira platensis (0.682 µg g⁻¹). Overall, enzyme-assisted extraction consistently outperformed traditional methods in terms of both yield and species preservation.

There have been few reports of As speciation in clinical studies published this year, although As in urine continues to be of interest. The determination of As species in urine is now routine and few new developments are reported. However, occasionally work is published that serves as a timely reminder to carefully consider all aspects of the experimental design, particularly sample collection. Heitland et al.³⁶ have reported finding an unexpected As compound in the routine determination of As species in urine samples using LC-ICP-MS. In addition to the chromatographic peaks of As^III, As^V AB, DMA, and MMA, which occur regularly, they also observed an additional As peak overlapping with that of As^V. Further investigations identified this unknown As compound as p-arsanilic acid, which had contaminated the urine samples during the sampling process. The contamination originated from the application of a test strip for the rapid determination of other clinical parameters before shipment of the samples to the laboratory. The authors note that this important aspect of sampling is not always considered in occupational biomonitoring. The simultaneous determination of As and Hg species in human urine has also been reported in a study aimed at reducing costs, time, and consumption of samples.³⁷ The As and Hg species were separated and analysed using HPLC-ICP-MS, achieved using a C18 (250 × 4.6 mm × 5 µm) column, a mobile phase composed of 0.1% L-cysteine, 5 mmol L⁻¹ NH₄H₂PO₄, and 4 mmol L⁻¹ TBAH flowing at 1 mL min⁻¹ in 15 minutes. The samples were diluted 5 fold with water prior to injection onto the column and polyatomic interferences were reduced using KED mode. Calibration curves for four As species (As^V, AB, DMA and MMA) and three Hg species (iHg, MeHg and EtHg) exhibited good linearity within the range from 1 to 20 µg L⁻¹, with correlation coefficients above 0.999. The LOD values were in the range from 0.030 to 0.086 µg L⁻¹, while the LOQ ranged from 0.10 to 0.29 µg L⁻¹. The spiking recoveries for all species varied from 87 to 110%, and the intra- and inter-day RSDs were determined to be 1.1–6.0%, and 0.8–9.2%, respectively.

4.4 Caesium

The speciation and distribution of Cs in radiocaesium-bearing microparticles emitted during the Fukushima nuclear accident was determined using XANES in fluorescence mode and µXRF with a transition edge sensor detector (TES).³⁸ The application of TES as a high-energy-resolution X-ray detector involves significant technical challenges, which are discussed in detail by the authors, including technological development, interdisciplinary collaboration, and instrumental adaptations, particularly for hard X-ray spectrometry. The TES system employed consisted of a 240-pixel detector with a detection area of 80 mm² and a counting rate of 2400 counts. The results provided insight into the sequence of processes occurring during the accident. The Cs was mainly found in two forms: (i) incorporated into phyllosilicates and (ii) embedded within silica microparticles. More specifically, Cs was present as (i) Cs dissolved in silicate glass and (ii) CsCl located on the surface of voids within the microparticles. The presence of CsCl suggests that Cs may be readily released into aqueous environments, contributing to its environmental mobility. This study represents a multidisciplinary effort involving approximately 20 research centres across Japan.

4.5 Chromium

For the determination of available Cr^VI in soil, Fu et al. devised a method in which the analyte was quantified by pXRFS after its retention on an ion-exchange resin.³⁹ They investigated the effect of a number of parameters, including ratio of soil mass to extractant volume, extraction time, resin dosage, enrichment time, soil particle size, resin drying, and soil-solution separation methods. They did not vary the extractant, which was 0.01 mol L⁻¹ CaCl₂ in all cases nor the ion-exchange resin, which was a Cr^VI specific resin (A21-S Tulsion, a strongly basic anion-exchange resin featuring quaternary ammonium functional groups on a gel polystyrene matrix). Details of the optimised method were not collected in one section and were scattered throughout the paper. Sample mass was not given, only that the optimum sample to extractant ratio was 1 : 5 with extraction for 30 min. Soil and extractant were separated by centrifugation. The resin mass was 0.1 g and the adsorption time allowed was 20 min. Resin and extractant were separated by filtration. It appeared as though soil particle size and drying of the resin had minimal effects on the results. The procedure was applied to several soils that had been spiked with Cr^III and Cr^VI, and results of the pXRFS method were compared with those of a spectrophotometric procedure based on reaction with 1,5-diphenylcarbazide (which is misprinted as “diphenyl carbamide” and not listed among the reagents). The LOD in a soil sample was not given, only a value for soil solution of 0.07 mg kg⁻¹. The researchers concluded that, in contrast to traditional methods, their pXRFS method minimised secondary environmental contamination, was cost-effective, did not require complex soil pretreatment, and consequently is highly suitable for rapid, on-site soil analyses.

Increasing interest in the possibilities of using microalgae as alternatives to synthetic dietary supplements raises questions about the speciation of both the potentially beneficial elements and the PTEs they contain. Szczesniak et al. developed several LC-MS procedures to examine the speciation of Cr and Zn in two species of microalgae: diatoms (Phaeodactylum tricornutum) and spirulina (Arthrospira platensis).⁴⁰ They examined diatoms grown in Poland, Portugal and Israel, and spirulina grown on Cr, Zn and Zn–Cr media. From preliminary experiments with SEC-ICP-MS, they found that the diatom material from Israel contained the highest concentrations of Cr and Zn, and this material was selected for the development of the extraction procedure. They investigated seven different extraction reagents: (1) 10 mmol L⁻¹ ammonium acetate pH 7.4; (2) 2% SDS; (3) 30 mmol L⁻¹ Tris–HCl pH 7.3; (4) methanol + water (1 : 1 v/v); (5) 7% Viscozyme; (6) 1.5% Alcalase; (7) 0.1% Viscozyme. One of these must be an error, as the list of reagents included Flavourzyne (a mixture of a peptidases prepared from the mould Aspergillus oryzae), which was in fact selected. The procedure consisted of vortexing 0.5 g of ground dried powder with 5 mL of Flavourzyne (concentration presumably either 7% or 0.1%) at 50 °C for 24 h, followed by centrifuging (30 min, 12 000 rpm) and filtering (0.45 µm). They separated Cr compounds by high-performance AEC (15 cm × 4.6 mm column) with isocratic elution (14 mmol L⁻¹ NH₄NO₃ + 0.5 mmol L⁻¹ EDTA, pH 7.2), obtaining two peaks in the ICP-MS/MS chromatograms. By comparison with the chromatograms of standards, they showed that one was due to Cr^III and that the other was not due to Cr^VI. They collected appropriate fractions and attempted to characterise the species by ES-MS/MS; however, they were unsuccessful. On the other hand, they were able to characterise a number of Zn compounds, and these results are discussed in Section 4.20.

To better understand the reasons for implant failure, Liu et al. studied the imaging and speciation information available regarding intracellular metallic implant debris with µSXRF.⁴¹ Debris generated from total hip replacement (THA) components made from metal alloys (most commonly CoCrMo and TiAlV alloys) can cause inflammatory cell responses that lead to adverse local tissue reactions and implant failure. The lack of information on intracellular chemical alterations of metal debris has hindered the understanding of the pathogenesis of such local tissue reactions. They studied the periprosthetic tissues of two implants: one consisting of a metal-on-metal articulation and the other a metal-on-polyethylene articulation both of which showed corrosion of the metal femoral head. Experiments were carried out at the Advanced Photon Source of the Argonne National Laboratory. All tissue thin sections were first screened at Beamline 8-BM with a spatial resolution of about 30 µm, and the areas of interest selected were further analysed at other beamlines, including the Bionanoprobe and the microprobes at 2-ID-D/E, providing a spatial resolution of 80 to 600 nm. They found that the metal-on-metal implant debris showed different oxidation states of both Cr- and Co-containing debris, consistent with three different species: Cr₂O₃, CrPO₄, and an CoCrMo-oxide mixture. In addition, they found that Cr₂O₃ formed in the tribological interfaces of the implant, while CrPO₄ is a by-product of the phagocytosis (uptake by macrophages – white blood cells) processing of debris containing a Co alloy. For the metal-on-polyethene case, they found no intracellular particulate debris associated with macrophages, from which they deduced that the adverse local tissue reaction may be entirely triggered by metal ionic species.

4.6 Cobalt

A method for the determination of vitamin B12 in nutritional products by HPLC-ICP-MS has been developed by Wolle et al.⁴² following an extensive evaluation of extraction and SPE preconcentration and clean-up procedures. Eight extraction procedures were investigated by analysing four CRMs, NIST SRM 1869 (adult nutritional formula), SRM 3252 (protein drink mix) and SRM 3233 (breakfast cereal), together with ERM-BD600 (whole milk powder). The optimised method consisted of the addition of 10 mL of 0.25 mol L⁻¹ NaOAc to between 0.2 and 0.5 g of solid sample or 5.0 g of liquid (or ready to feed/drink) samples. Samples containing starch were incubated with α-amylase at 40 °C for 30 min. Then 250 µL of 1% (w/v) NaCN was added (to stabilise all forms of cobalamin) and the mixture heated for 30 min at 95 °C. Following cooling, centrifuging and filtration, the extract was subjected to SPE. Four SPE materials were evaluated, of which Oasis HLB sorbent was selected. After conditioning the column, 15 mL of sample was loaded, and the analyte eluted with 3.5 mL of MeOH, which was evaporated to dryness (vacuum concentrator at 55 °C) and redissolved in 0.5 mL of 0.1% (v/v) formic acid for subsequent separation by isocratic HPLC at room temperature on a C-8 column with a mobile phase of 0.1% (v/v) formic acid in MeOH + water (30 + 70). The ICP-MS was operated in O₂/Ar option gas mode to oxidise the 30% MeOH in the mobile phase (flowing at 0.4 mL min⁻¹) and prevent soot buildup on the cones. The collision cell gas was He to negate interferences from polyatomic ions at m/z 59. An IS solution, 5 ng g⁻¹ Co, was injected post-column via a switching valve. For the determination of total Co the IS was Ge (20 ng g⁻¹ in 5% v/v HNO₃, 0.5% v/v HCl, and 4% v/v isopropanol) added via a T-piece prior to the nebuliser. For total Co determination, samples were subjected to MAE according to the method described in the US FDA Elemental Analysis Manual Section 4.7. In addition to the four CRMs mentioned above, the validation also included the analyses of NIST SRM 1846 (infant formula), 1549a (whole milk powder), 8260 (infant nutritional formula), and 8261 (adult nutritional formula). They applied the method to 38 samples (infant formula, toddler nutritive drink, adult nutritive drink, milk protein powder/drink, and breakfast cereal) purchased from local stores and measured spike recoveries from each at three different concentrations. They obtained recoveries of both spikes and of the mass fractions in the CRMs ranging from 80 to 120%, except for NIST SRM 3252 (protein drink). The LOD values were 0.004 µg per 100 g (for liquid and ready-to-feed/drink samples) and 0.06 µg per 100 g (for solids), expressed as cyanocobalamin, which was detected in all samples except one, a pea and rice ready-to-drink protein drink. The low values obtained for SRM 3252 were corroborated by analyses conducted by two other laboratories, and this finding was communicated to NIST, who retracted the reference value. The researchers considered that their method is simpler, less time-consuming and with broader application and better analytical performance than most official and other published methods.

To better understand the reasons for implant failure, Liu et al. studied the imaging and speciation information available regarding intracellular metallic implant debris with SXRFM.⁴¹ The study rationale and method details are given in the final paragraph of Section 4.5. They found that the metal-on-metal implant debris showed different oxidation states of both Cr- and Co-containing debris, consistent with three different species: Cr₂O₃, CrPO₄, and a CoCrMo-oxide mixture. They also deduced from Co-specific XANES spectra of the same particles that Co was present in a mixed state between the alloy and Co oxide. For the metal-on-polythene case, they found no intracellular particulate debris associated with macrophages, from which they deduced that the adverse local tissue reaction may be entirely triggered by metal ionic species.

4.7 Gadolinium

The speciation of Gd due to the use of Gd-containing Magnetic Resonance Imaging (MRI) contrast agents continues to be of interest. These reports can be divided into two categories: the analysis of the contrast agents themselves and their effect and fate in the human body and the investigation of their environmental fate and degradation. In the former category a UHPLC-ICP-MS method has been developed to separate and quantify a tetrameric macrocyclic contrast agent, gadoquatrane, from by-products produced in its synthesis.⁴³ The separations were performed on a RP BEH phenyl column (100 mm, 3.0 mm, 1.7 µm) using a gradient elution of (A) 0.1% formic acid (FA), 0.1% NH₄HCO₂ and 0.3% ACN in water and (B) an aqueous solution of 0.1% FA and ACN (2 : 3, v/v), flowing at 0.4 mL min⁻¹ with a total runtime of 47 minutes, which includes 10 minutes for column re-equilibration, and a column temperature of 35 °C. An internal standard, 1 ng ml⁻¹ Tb in 2% HNO₃, was added post-column via a T-piece to account for changes in sensitivity due to the gradient elution profile and gadobutrol was used to prepare the calibration standards. The paper gives a detailed discussion on the complexities of compound independent calibration for gradient HPLC, and the method development required to be able to undertake this. The method revealed that 7 by-products were present from the production process before purification, and these were identified by ES-MS and HR-ES-MS. After full optimisation of the method the following LOQ values were achieved, 38 nmol L⁻¹ for monomeric, 19 nmol L⁻¹ for dimeric, 13 nmol L⁻¹ for trimeric and 10 nmol L⁻¹ for tetrameric Gd complexes.

Three papers report on the detection of Gd compounds in waters. In the first of these hospital waste waters were analysed for the presence of MRI contrast agents using HPLC-ICP-MS.⁴⁴ The developed method was also able to simultaneously measure I- and Pt-containing pharmaceuticals at trace levels (ng L⁻¹), showing the versatility of ICP-MS as a multi-elemental speciation detector when coupled with HPLC. The methodology used HILIC to separate and quantify ten drug residues, including: three I-, three Pt-, and four Gd-complexes. Three HILIC stationary phases, including two zwitterionic, i.e. phosphorylcholine and ammonium–sulfonic acid phases, and one silica charged phase, were compared using different compositions of mobile phase, including organic solvent percentage, salt concentration and pH. The phosphorylcholine stationary phase, with an acetonitrile gradient from 75 to 50% (v/v) in 10 mmol L⁻¹ NH₄HCO₂ aqueous solution, achieved the separation of the 10 tested complexes in 10 minutes. The LOD values for the various complexes were between 1 and 3 ng L⁻¹ Gd, 46 and 63 ng L⁻¹ I and 3 and 17 ng L⁻¹ Pt. The developed method was applied to urban and hospital wastewaters. Two I-containing compounds, iohexol and iopamidol, and one Gd, gadoterate, were quantified while no Pt-drug compounds or metabolites were detected in the samples studied. The other two papers covered here both used AEC to separate and quantify the Gd compounds for detection by ICP-MS. Dalla Costa et al. developed a tertiary gradient elution profile that was able to separate four compounds, Gd-BT-DO3A, Gd-BOPTA, Gd-DOTA and Gd-DTPA, with good baseline separation in 15 minutes.⁴⁵ The three components of the mobile phase were (A) 100 mmol L⁻¹ NH₄NO₃ at pH 7, (B) high-purity water and (C) MeOH with a flow rate of 0.45 mL min⁻¹. The LOD and LOQ values obtained were 2 and 4 ng L⁻¹ for Gd-BT-DO3A and Gd-DOTA, 5 and 30 ng L⁻¹ for Gd-BOPTA and 4 and 8 ng L⁻¹ for Gd-DTPA, respectively. The method was applied to river water samples from various locations in France, the results of which ranged from 42 to 279 ng L⁻¹ depending on species and location. The second paper to use AEC utilised a gradient elution with a binary mobile phase of (A) 1.5 mmol L⁻¹ NH₄NO₃ at pH 10.5 and (B) 150 mmol L⁻¹ NH₄NO₃ at pH 9.2 to separate 6 Gd species in 180 s, although two pairs of them, gadobutrol and gadoterate and gadopentetate and gadobenate, were not quite baseline resolved.⁴⁶ The LOD and LOQ values ranged from 0.8 to 5.8 and 2.6 to 19.5 pmol L⁻¹, respectively. The method was then applied to water samples from 4 German rivers, in which 3 Gd compounds, gadobutrol, gadoterate and gadoteridol, were detected at concentrations ranging from 0.03 to 164 nmol L⁻¹, 0.112 to 128 nmol L⁻¹ and <LOD to 0.149 nmol L⁻¹, respectively. All three papers give a detailed account of the method development and optimisation with good discussions on the significance of the results obtained.

4.8 Gold

In Au nanomaterial synthesis, the speciation of the Au precursor dictates nucleation kinetics and growth mechanisms, directly controlling nanoparticle size, dispersion and method repeatability. However, the speciation of these precursors is currently unclear. Thus, Volchek et al. utilised HPLC-ICP-OES and ES-MS to investigate the speciation of a group of fluorinated Au benzoethiolate precursors, H[Au(SR)₂], where R = C₆F₅, C₆F₄H, C₆F₄CF₃ or C₆F₄Cl, which were synthesised using cited methods.⁴⁷ Species separation was on a RP column using a gradient elution with a mobile phase of (A) 0.1% HCOOH and (B) ACN + 0.1% HCOOH flowing at 0.25 mL min⁻¹ for 18 minutes. For analyses by ICP-OES, the eluent was diluted with H₂O to reduce the effects of organic carbon on the plasma whilst for ES-MS analyses the eluent was introduced undiluted to the instrument used. Using both techniques, these separations resolved three peaks for each gold complex, which were assigned as an organodisulfide (RSSR) and two Au species: H[Au^I(SR)₂] and H[Au^III(SR₄)], which was confirmed using scanning UV-vis spectrometry. The stoichiometric Au : S ratios of 1 : 2 and 1 : 4 for the second and third peaks were calculated from the HPLC-ICP-OES spectra. The optimisation of the method and data interpretation is covered in detail in the paper, and the authors highlight the benefits of elemental speciation over bulk analysis for complex systems and conclude that the study enables the rational design of Au–thiolate complexes to achieve desired functional characteristics.

4.9 Halogens

It is exciting to see some ground-breaking research work being undertaken into the types of instrumentation used for speciation analysis– in particular, the use of new chromatographic detectors that are not often associated with speciation analysis, such as XRF. But also, applications using instrumentation, such as ICP-OES, where new instrumental developments make it feasible for coupling to LC separations. Coincidently, both of these new developments have been used for the speciation of halogen-containing compounds, where once again the main focus of recent applications is related to concerns regarding their environmental impact.

The use of ICP-OES as an HPLC detector rather than ICP-MS is unusual, both because of the better sensitivity of ICP-MS but also because the ICP-OES software is often not suitable to acquire a time resolved signal. However, the use of ICP-OES as a chromatographic detector is an attractive analytical approach as it is a truly simultaneous detection system, unlike Q-ICP-MS, which acquires the signal sequentially. Other advantages are: that there is no sampling cone interface, which is susceptible to blockage when using mobiles phases containing less volatile buffers; the use of organic solvents does not require the addition of oxygen to reduce interface clogging due to the formation of carbon; and depending on the application, it can be less affected by the interferences encountered in ICP-MS when measuring elements with an m/z below 50. A study on the quality of pool and spa water has evaluated the hyphenation of IC with ICP-OES for the simultaneous quantification of the inorganic oxy-halide species formed during oxidative water treatments.⁴⁸ These hazardous chemical by-products of this process are monitored to ensure the concentrations present do not exceed the regulatory limits in place. The conventional methods used are based on IC separation with conductivity detection, which lacks the specificity needed to discriminate between closely eluting compounds containing different halogens. For this reason, a method based on ICP-OES detection was developed to measure five prevalent Br and Cl species: Br⁻, BrO₃⁻, Cl⁻, ClO₂⁻, and ClO₃⁻, at trace concentrations ranging from 2.5 to 100 µmol L⁻¹ for Br species and 5 to 100 µmol L⁻¹ for Cl species. The study showed that ICP-OES detection in the vacuum UV region below 170 nm achieved higher sensitivity for Br species, with BrO₃⁻ showing a significantly stronger signal response than Br⁻. The developed method enabled the selective quantification of co-eluting oxy-halides and showed linearity (R² ≥ 0.999) for all analytes. The LOD values achieved were below 100 µg L⁻¹ and mean recoveries of 98 ± 4% were achieved for BrO₃⁻ and ClO₃⁻ in model solutions mimicking high-salinity samples. For the first time, Br and Cl speciation was successfully achieved by using IC-ICP-OES.

The second paper highlighted this year details the development of a novel HPLC detector, using XRF detection, which was described as being the ideal LC detector because of its characteristic of having a universal response.⁴⁹ In this respect the authors were making comparison to the widely used UV and fluorescence detection (FLD) systems commonly used as HPLC detectors. While widely used, the response of UV and FLD remains highly molecule specific, always necessitating authentic molecular standards for quantification, whereas the use of XRF meant that calibration of the detector could be undertaken using simple and highly traceable, elemental standards. Another useful characteristic of XRF as a detector is that it is non-destructive, allowing for a flow-through design, which can be coupled to other detectors to combine the advantages of each in a single setup. Therefore, HPLC-XRF has the potential of universal quantification based on elemental instead of molecule-specific standards, provides elemental information, and is non-destructive allowing further hyphenation with other instruments. The detector itself was a µXRF instrument based on a customised version of the X-Beam Superflux MP excitation source creating monochromatic radiation at the energy of the Mo Kα line (17.4 keV). The emitted fluorescence X-rays were detected under 90° relative to the incoming X-ray beam by a SiriusSD Silicon Drift Detector. Monte Carlo simulations were used to aid in the creation of the HPLC-XRF setup and study the influence of the parameters involved. A proof of principle was established for the analysis of three brominated organic molecules: tribromophenol, tribromomethoxyphenol, and tribromohydroxybenzoicacid of environmental concern, and calibration curves were created, which show LOD values below 0.3 mmol L⁻¹ for all tested compounds and good linearity between 0.2 and 20 mmol L⁻¹ (R² ≥ 0.99). The precision for the peak areas varied between 10.5 and 1.06%. Sensitivity was a limitation but could be further improved through multidetector setups, improved acquisition hardware and software, and techniques such as stop-flow chromatography and preconcentration. While simulations predicted a minor influence of mobile phase composition on the response, this effect proved to be negligible under the experimental conditions. Whilst this instrumental set-up might have limited application to elements that are not easily detected by ICP-MS, such as the halogens, it does not appear to have many advantages over inorganic MS. Contrary to the authors opinion that HPLC-ICP-MS suffers numerous issues, e.g. cost, complexity and polyatomic interferences, that limit its usability, the considerable number of applications of HPLC-ICP-MS as the benchmark speciation instrumentation would prove this outlook to be wildly incorrect. One wonders how the referees for the manuscript did not draw this to the authors attention.

The speciation of I-containing pharmaceutical drugs in environmental waste waters has been studied using HPLC coupled to ICP-MS.⁴⁴ The developed method was also able to simultaneously measure the Pt- and Gd-containing pharmaceuticals at trace levels (ng L⁻¹), showing the versatility of ICP-MS as a multi-elemental speciation detector when coupled with HPLC.

The methodology used hydrophilic interaction chromatography (HILIC) to separate and quantify 10 drug residues, including: three I-, three Pt-, and four Gd-complexes. Three HILIC stationary phases, including two zwitterionic, i.e. phosphorylcholine and ammonium-sulfonic acid phases, and one silica charged phase, were compared using different compositions of mobile phase, including organic solvent percentage, salt concentration and pH. The phosphorylcholine stationary phase with an acetonitrile gradient from 75 to 50% (v/v) in 10 mmol L⁻¹ NH₄HCO₂ aqueous solution, achieved the separation of the 10 tested complexes. The LOD values for the various complexes were between 1 and 3 ng L⁻¹ Gd, 46 and 63 ng L⁻¹ I and 3 and 17 ng L⁻¹ Pt. The developed method was applied to urban and hospital wastewaters. Two I-containing compounds, iohexol and iopamidol, and one Gd compound, gadoterate, were quantified while no Pt-drug compounds or metabolites were detected in the samples studied.

4.10 Lead

The determination of Pb in seafood presents a number of analytical challenges, including the need for an extraction method applicable to a variety of samples and a rapid, robust separation prior to quantification. Both of these challenges were addressed by Xu et al. who devised a method consisting of an extraction/derivatisation followed by short-column, cation-exchange HPLC with ICP-MS detection.⁵⁰ For the development of the extraction procedure, they selected three materials representative of the variety of seafoods: a shellfish (Morulauva), a seaweed (kelp: Laminaria japonica) and a fish (Ditrema temmincki Bleeker). They found that the extraction could be performed solely by EDTA, and that the compounds formed could be separated by CEC without any further treatment. The optimised procedure consisted of a “pre-soaking” step (0.10 g freeze-dried, ground sample, 6.0 mL of 12 mmol L⁻¹ EDTA, 6 h at room temperature), followed by MAE (10 min ramp to 120 °C held for 30 min), which was repeated with a further 6 mL of EDTA. The two extracts were combined, filtered (0.45 µm) and diluted (H₂O) as needed before injection into the HPLC system. No information was given concerning the microwave apparatus other than the reaction vessel was described as a “Teflon jar”. They confirmed, both by spiking and taking standards through the extraction procedure, that neither trimethylPb nor triethylPb were degraded. The extracted species were separated on two guard columns of Zorbax-300SCX (4.6 × 12.5 mm) connected in series by isocratic elution with 1.0 mol L⁻¹ EDTA + 0.25 mmol L⁻¹ TBAH (pH 3.0) at a flow rate of 1.0 mL min⁻¹ and column temperature 30 °C. The elution order was Pb²⁺, trimethylPb, triethylPb for which a total time of about 9 min was required. However, as none of the six samples examined contained triethylPb, the separation time could be around 6.5 min. Visual inspection of the chromatograms of standards, suggests that Pb²⁺ was not retained and eluted with the solvent front. The slopes of the calibration plots show evidence of compound-dependent responses. The validation included the analysis of a CRM, GBW08521 (laver), with an inorganic Pb concentration of 0.81 ± 0.03 µg g⁻¹, the result for which was not significantly different from the certificate value, and spike recoveries from 6 samples that ranged from 94 to 105%. The LOD values ranged from 0.60–0.72 ng g⁻¹ (dry weight), and they found that seaweed and shellfish contained only Pb²⁺, whereas the fish samples contained Pb²⁺ and trimethylPb.

4.11 Mercury

The high toxicity of Hg species to biota is the one of the main drivers for their determination and the continual analytical development in this field. Iwai et al. have reported the development of an HPLC-ICP-MS method for the determination of Hg species in breast milk.⁵¹ Extraction of the Hg species present in the samples was achieved by adding 30 µL of 8% TMAH to 200 µL of milk followed by UAE for 10 minutes, vortex mixing and ultrafiltration 21,300 g for 15 min. The ultrafiltered sample was then injected (10 µL) onto a C18 column with a mobile phase of 60 mmol L⁻¹ NH₄HCO₃, 0.05% (m/v) Na 2,3-dimercaptopropanesulfonate monohydrate, 5% MeOH and 0.05% TMAH neutralised with HCL flowing at 0.22 mL min⁻¹. These conditions gave sharp peaks with good baseline separation and a runtime of 350 s. Detection of the eluted species was by ICP-MS and analysis of NIST SRM 1953, breast milk certified for tHg gave a recovery of 99% for the sum of the two Hg species (iHg and MeHg) detected with a method LOD of 1.9 and 1.4 pg g⁻¹, respectively. The iHg content in fore-and hindmilk was found to be around the 0.1 ng g⁻¹ level whilst MeHg was higher ranging from 0.11 to 0.31 ng g⁻¹. The paper contains a detailed discussion on the portioning of Hg species in breast milk. The same research group then applied the method to a larger number of breast milk samples, and the detailed results of this work have been published in a separate paper.⁵² Pacer et al. have developed a rapid method for the quantification of Hg species in whole blood by HPLC-CV-ICP-MS.⁵³ The Hg species were extracted from 250 µL of blood by the addition of 4.75 mL of 0.1% TMAH in H₂O and UAE for 15 minutes followed by centrifugation of 5 minutes at 400 g and filtration. The HPLC conditions were a C8 column with a mobile phase of 0.06 mol L⁻¹ ammonium acetate, 0.05% (v/v) mercaptoethanol, 0.4% (m/v) L-cysteine, 5% (v/v) methanol and a pH of ∼6.7 flowing at 1 mL min⁻¹. This gave a runtime for iHg and MeHg of four minutes or if EtHg was also required, 8 minutes. The CV generation was by the use of 0.5% (m/v) NaBH₄ in 0.1% (m/v) NaOH as the reductant and 0.01 mol L⁻¹ HCl, with Ar as the carrier gas. Method LOD values with and without CV were 0.45 and 0.22, 0.42 and 0.23 and 0.84 and 0.33 µg L⁻¹ for iHg, MeHg and EtHg, respectively. Method validation was by the use of the NIST 955d, frozen human blood, and NIST 955c, caprine blood, which have certified and reference values for iHg, MeHg and EtHg, respectively. These samples were analysed by both HPCL-ICP-MS and HPLC-CV-ICP-MS and all values obtained, n = 33, were in statistical agreement with the certified or reference values. The authors conclude that the developed VG method gives lower LOD values than GC-ID-ICP-MS methods and due to its rapidity, 100 samples in ∼7 hours if only iHg and MeHg are required and 50 samples if EtHg is added in, is competitive with this alternative approach and that the HPLC column used is capable of 1000 analyses before needing replacement. The paper also contains a good discussion of the optimisation process and is well worth a read. A combination of LA-ICP-MS and HPLC-ICP-MS has been used to study Hg species in rat and brain tissues after exposure to iHg or MeHg.⁵⁴ The Hg species distribution was mapped in tissue section and then areas of interest excised of approximately 0.5 mm³ using the same laser. The Hg species in both whole brain and the excised sections were extracted with 100 or 50 mmol L⁻¹ Tris–HCl at pH 7.4, respectively, UAE and then ultracentrifuged at 100 000 g. The extracts were then injected onto a SEC column and eluted with a mobile phase of 50 mmol L⁻¹ Tris–HCl at pH 7.4 flowing at 40 µL min⁻¹. It was found that iHg exposure led to the induction of MTs, resulting in elevated levels of Cu and Zn in the renal cortex and outer medulla whilst MeHg was primarily bound to Hb, highlighting its metabolic pathway and potential for bioaccumulation. Finally, the simultaneous determination of As and Hg species in human urine has been reported in a study aimed at reducing costs, time, and consumption of samples.³⁷ The As and Hg species were separated and analysed using HPLC-ICP-MS, achieved using a C18 (250 × 4.6 mm × 5 µm) column, a mobile phase composed of 0.1% L-cysteine, 5 mmol L⁻¹ NH₄H₂PO₄, and 4 mmol L⁻¹ TBAH flowing at 1 mL min⁻¹, in 15 minutes. The samples were diluted 5-fold with water prior to injection onto the column and polyatomic interferences were reduced using KED mode. Calibration curves for four As species (As^V, AB, DMA and MMA) and three Hg species (iHg, MeHg and EtHg) exhibited good linearity within the range from 1 to 20 µg L⁻¹, with correlation coefficients above 0.999. The LOD were in the range from 0.030 to 0.086 µg L⁻¹, while the LOQ ranged from 0.10 to 0.29 µg L⁻¹. The spiking recoveries for all species varied from 87 to 110%, and the intra- and inter-day RSD were determined to be 1.1–6.0%, and 0.8–9.2%, respectively.

Two papers reviewed here cover Hg speciation in waters, where concentrations can be at the ng L⁻¹ level or lower, so a pre-concentration step is often required. Shi et al. have addressed this by synthesising a magnetic covalent organic framework (COF) based on silica coated Fe₃O₄ NPs to which thianthrene was bonded, with full details of the procedure and characterisation given, for extracting iHg, MeHg and EtHg from various types of water, surface, ground, sea and tap.⁵⁵ The COF was added to 300 mL of sample at a concentration of 0.02 mg L⁻¹ and the mixture was agitated for 10 minutes on a shaking table. Subsequently, the COF was extracted magnetically and the adsorbed Hg species eluted with a solution of 2% m/v thiourea in 1.5% v/v HCl with the elution being repeated for a second time to recover all of the Hg present on the COF. The Hg species in the eluted sample were then separated by HPLC in 800 s, using a C18 column (15.0 cm × 6.0 mm × 5.0 µm) and a mobile phase of 5% (v/v) methanol, L-cysteine (1.0 g L⁻¹) and ammonium acetate (60 mmol L⁻¹) flowing at 1 mL min⁻¹ with detection by ICP-MS. Although baseline resolved, this separation resulted in significant tailing for two of the Hg peaks in the chromatogram, but they are not labelled in the paper. The method gave LOD values for Hg²⁺, MeHg and EtHg of 0.50, 0.19 and 0.67 ng L⁻¹ with enrichment factors, based on 300 mL of sample, of 338, 356 and 329, respectively. The method was validated using NRCC DORM-4, fish protein, after a cited MAE procedure solubilised the Hg species present, with recoveries of 98% achieved for both MeHg and total Hg and spike recoveries of the four waters analysed that ranged from 94 to 104% for iHg, 95 to 103% for MeHg and 95 to 102% for EtHg. The COF was found to be stable in the presence of either acidic or alkali solutions, in the 0.1 to 1.0 mol L⁻¹ range, and at least 35 consecutive adsorption–desorption cycles. The second paper related to Hg species in waters investigated the formation of nanoparticulate HgS in the presence of DOM.⁵⁶ The model Hg^II-DOM-S^II system contained 20 µg L⁻¹ Hg²⁺, 10 mg L⁻¹ DOM, and 10 µmol L⁻¹ Na₂S in N₂-deoxygenated water, adjusted to pH 5.0–9.0 with 10 mmol L⁻¹ Na₂HPO₄–NaH₂PO₄. The Hg species were separated using an amino-modified SEC column with a mobile phase of 2% v/v FL-70, a mixture of non- and anionic surfactants, and 2 mmol L⁻¹ Na₂S₂O₃ flowing at 0.5 mL min⁻¹ with detection by ICP-MS. The formed HgS NPs were characterised by TEM after ultrafiltration. The authors suggest that nano-HgS is bioavailable to methylation bacteria and that, in some cases, nano-HgS can be more effectively internalised and methylated than Hg^II and that high-RMM DOM may stabilise nanoparticles via steric hindrance, whereas low-RMM DOM, such as fulvic acids, could enhance dissolution or aggregation of the NPs.

For the extraction of Hg species from lithogenic materials, acidic or alkaline reagents are usually used. Lei et al. have compared two extraction methods one acidic and one alkaline for the extraction of MeHg from marine sediment, soil and rock samples.⁵⁷ The acidic method utilised 18% KBr in 0.9 mol L⁻¹ H₂SO₄, 1 mol L⁻¹ CuSO₄ and DCM, shaking at 25 °C for 1 hour, centrifugation and back extraction from the DCM into H₂O (USGS Method 5), whilst the alkaline method involved vortex mixing and shaking the sample in 4.5 mol L⁻¹ KOH in 25% MeOH followed by heating at 75 °C for 4 hours then dilution with H₂O and centrifugation. Subsequently, the MeHg in the extracts from both approaches were ethylated using NaBEt₄. The ethylated samples were then analysed using two different measurement techniques, an automated MeHg analyser and purge and trap GC-CV-AFS, which are briefly described in the paper. Method validation was by the use of ERM CC580, estuarine sediment, for which the recovery was 101 to 111% for the alkaline method compared with 76 to 92% for the acidic extractions. Spike recoveries, which are of dubious use with solid materials as the spike is only mixed with the sample and not incorporated into the sample matrix, were 87 to 123% for the alkaline method compared with 70 to 95% for the acidic extractions. The authors suggest that the lower extraction efficiency for the acidic method was due to readsorption of MeHg to the sample and the inability of this method to fully release all of the MeHg bound to thiol groups on the particles, which is accomplished with the alkaline method, and the possible breakdown of MeHg under acidic conditions due to the presence of Fe^III generated free radicals. It was also found that the alkaline method gave superior precision, 5% RSD, compared to 10% RSD for the acidic extraction.

4.12 Nickel

A novel methodology based on the combined use of a paper-based plate as a sample treatment approach, coupled with direct analysis by LA-ICP-MS, was developed for the separation of Ni metalloporphyrins in light, medium, and heavy crude oil samples.⁵⁸ Cellulose-based paper plates were prepared from filter paper (0.4 mm thickness) with dimensions of 5 × 5 cm. Crude oil samples were initially treated with heptane to separate asphaltenes (solid fraction) from maltenes (supernatant). The asphaltene fraction was subsequently dissolved in toluene, deposited onto the base of the paper plate, and separated by gradient elution using solvents of increasing polarity applied every 1 cm along the plate. The maltene fraction was directly applied onto the paper plate and subjected to fractionation prior to porphyrin speciation. In this case, maltenes were first separated into subfractions on one side of the plate; after rotating the plate by 90°, porphyrin speciation was achieved using solvents of increasing polarity. Following separation, the resulting spots were analysed by LA-ICP-MS to detect Ni associated with metalloporphyrins. Octaethylporphyrin-Ni was identified in the most polar fractions, such as resins and asphaltenes, and its structure was confirmed by HR-MS. Notably, only 1 µL of crude oil and 4 mL of solvent (hexane, toluene, methanol, and dichloromethane) were required to complete the separation in 5 minutes. This method enables the direct identification of metalloporphyrins in crude oil matrices with minimal sample preparation, low solvent consumption, and the potential for field-deployable analysis.

4.13 Phosphorus

The environmental transformation and degradation of P-containing compounds, including inorganic and organic species, has been studied using IC coupled to ICP-MS. Phosphorus is an important and often limiting element in terrestrial and aquatic ecosystems and an insufficient understanding of the environmental chemistry of P-containing compounds limits the development of controls and mitigation of their environmental impacts. Two papers have used similar analytical work-flows to investigate the environmental transport, stability and fate of these important and widely used chemicals. An IC-ICP-MS method was developed to provide suitable sensitivity for the investigation of the transformation and dephosphorylation of specific P-containing compounds in model environmental systems.⁵⁹ The separation of eight P compounds was facilitated using AEC (4 mm × 250 mm) at 25 °C. The gradient and flow rate were optimised, and the best peak resolution was achieved using water at pH 7.0 as eluent A and 100 mmol L⁻¹ NaOH in water pH 13.5 as eluent B at a flow rate of 1.0 mL min⁻¹. The use of different ICP-MS cell modes including O₂, He, H₂ and no-gas were investigated and surprisingly, as O₂ is usually used as the reaction cell gas for the interference free measurement of P, no-cell gas mode was shown to provide suitable sensitivity and robustness to interferences. The calibration curves, determined by peak area, exhibited a linear relationship over a broad concentration range of 1 to 10 000 µg P L⁻¹ (R² > 0.999), thus clearly indicating a wide dynamic range. Environmentally relevant LOD values were achieved, ranging from 1.2 µg P L⁻¹ (phosphite) to 4.2 µg P L⁻¹ (triphosphate). The method was compatible with the measurement of a range of both inorganic (phosphite, phosphate, pyrophosphate, and triphosphate) and organic compounds (mono-, di-, and tri-adenosine phosphates and phytic acid). Detecting both inorganic and organic P-containing compounds during the transformation process often requires combining multiple analytical techniques due to their restricted detection sensitivity or applicability. However, the high sensitivity and broad applicability of IC-ICP-MS to both compound types enabled it to simultaneously track the dynamics of all potential P products predicted during the transformation, saving time and effort. The IC-ICP-MS method efficiently and comprehensively tracked both inorganic and organic hydrolysis products of model P compounds in water containing reactive CeO₂ surfaces, on which triphosphate and adenosine triphosphate were quickly dephosphorylated.

Another recently developed robust P speciation method to investigate the reactivity of organic and inorganic P-containing species, such as polyphosphonates, in terrestrial and aquatic ecosystems also utilised IC-ICP-MS/MS.⁶⁰ The focus of this work was to develop a comprehensive, fast, robust and sensitive quantification method for the most commonly used aminopolyphosphonates and their potential transformation products, as well as relevant monophosphonate compounds, including diethylenetriamine penta(methylenephosphonate) (DTPMP), which can be transformed to glyphosate, a non-selective herbicide. An automated sample dilution, chromatographic separation, and addition of a post-column internal standard was developed using a prepFAST AEC system equipped with two 500 µL sample loops. Unlike the previous study on P-compounds in environmental samples, to overcome interferences and reduce background noise, triple-quadrupole mode with O₂ as a reaction gas was selected and P was detected as ³¹P¹⁶O⁺. One advantage of using an elemental detector such as ICP-MS/MS is that it allows for the quantification of unidentified transformation products with a species-unspecific calibration approach, which can help to complete the P mass balance. Instrumental LOD values ranged from 0.06 to 0.73 µg L⁻¹ P for each species, but for the later eluting species, the LOD values increased due to the increasing background signal resulting from the NH₄NO₃ gradient, which was attributed to changed plasma conditions. Chromatographic repeatability was good with peak area and retention time RSDs <2.5% and <0.6% and the run was complete within 3.5 min, with good peak shapes, indicating a rapid analytical workflow. The applicability of the developed method was demonstrated by the analysis of a transformation experiment of DTPMP under environmental oxidative conditions over several time points, to gain insight into the mechanisms of degradation in the environment, both into known and unidentified transformation products.

The investigation of inositol phosphate (InsP) and metabolites, which play a crucial role in various signal transduction pathways and are responsible for many biological and cellular functions, has used HPLC-ICP-MS/MS with detection of the ³¹P¹⁶O⁺ reaction cell product ion.⁶¹ There are 63 possible isomers of phosphate monoester-substituted myo-inositol, and discrete signalling functions are assigned to only a small number of these species. Their analysis in tissues is plagued by multiple difficulties of sensitivity, regioisomer resolution and the need for radiolabelling with metabolic precursors. The developed method reported in this work, used a short, steep linear gradient of HCl to resolve all known naturally occurring inositol species in avian tissue, plant material and soil used in the controlled experimental investigation of these compounds. Two gradient elution profiles were used for the separation: either a linear or exponential gradient, of the solvents A (water) and B (0.8 mol L⁻¹ HCl) at a flow rate of 0.4 mL min⁻¹ and delivered by a metal free ion-chromatography LC system. Upwards of 20 compounds, were well resolved with good peak shape, on a 250 × 3 mm id column. The use of HPLC-ICP-MS tolerates complex matrices and, by powerful chromatography, resolves in a single run multiple non-enantiomeric myo-inositol tetrakisphosphates, myo-inositol pentakisphosphates and all inositol hexakisphosphates known in nature.

In a departure from the obvious trend in P speciation methodology shown by the papers reviewed above, the use of GC coupled to ICP-MS/MS detection for the analysis of organophosphorus chemical warfare agents has been compared to GC-FPD.⁶² Organophosphorus chemical warfare agents such as sarin (GB), soman (GD), and cyclosarin (GF) rank among the most toxic substances known, making trace-level detection critical for public and military safety. Diluted samples of the aforementioned agents were prepared under controlled laboratory conditions and then analysed by both methods. The same GC column, a Rtx-OPPesticides capillary separation column, was used with both detectors. The GC was coupled to the ICP-MS instrument via a commercially available interface that uses a heated transfer line and a commercially available GC-FPD instrument was used for the comparison. The analytical figures of merit were established, and it was shown that GC-ICP-MS enabled LOD values of 0.12–0.14 ng mL⁻¹ to be achieved, whereas GC-FPD gave LOD values of approximately 0.36–0.43 ng mL⁻¹. Although GC-ICP-MS was only 10× more sensitive than GC-FPD, the authors point out that the two methods can be used in a complementary fashion, as GC-FPD is more portable and could be used as a screening method, whereas GC-ICP-MS would be applied in situations where the greatest sensitivity was required, such in post-conflict forensic and judicial investigations. Repeatability (RSD) was better than 5% for all compounds in both methods, indicating the high stability and reliability of the measurements. It was concluded that the greater sensitivity of GC-ICP-MS makes it suitable as a technique for ultra-sensitive detection of trace amounts of chemical warfare agents in environmental samples and in confirmatory testing for compliance with the Chemical Warfare Convention, while simultaneously employing GC-FPD for rapid preliminary monitoring.

4.14 Platinum

The measurement of Pt species in environmental samples has been the subject of two recent studies, which have demonstrated the breadth of the types of Pt species of environmental concern. Due to its varied uses, including in automobile catalysts and as a cancer treatment, and the discharge of Pt into the aquatic environment, a renewed interest in its possible environmental mobility and toxicity has instigated the development of new speciation methods.

To evaluate the ecotoxicity of inorganic Pt in water the inter-conversion of two Pt oxidation states and their interaction with organic material was investigated to determine which species are present in freshwater, under different pH and organic matter conditions.⁶³ Because of the paucity of thermodynamic constants for Pt species the authors developed experimental approaches to establish Pt speciation under simulated environmental conditions. Analytically, they used SEC chromatography coupled to ICP-MS detection to separate Pt^II, Pt^IV and investigate the binding of these oxidation states to natural organic matter (NOM). This heterogeneous mixture of functionalised organic macromolecules, derived from living organisms is operationally classified into different fractions, the main ones being humic acids (HA) and fulvic acids (FA). The former tends to be composed of larger aromatic molecules, mainly from a terrestrial origin, whereas the latter fraction often contains smaller more functionalised molecules. Separation using SEC allowed for the determination of the range of MWs of Pt bound with NOM, as well as calculation of the average and median MWs of the species present. Based on a comparison of retention times, Na polystyrene sulfonate standards (PSS) were used to determine the size range of the Pt-containing NOM, as well as the inorganic Pt species present. The SEC column contained highly cross-linked agarose with covalently bound dextran chains. This was used to minimise conformational changes during the separation and the mobile phase was the same as the sample matrix for the experiment carried out. Compounds with a range of MWs, between 0.2 and 80 kDa were shown to be present in the samples. However, apart from the inorganic Pt oxidation states, the structural identity of the species was not elucidated and the majority of the report was qualitative in nature.

The speciation of organometallic Pt-containing pharmaceutical drugs in environmental waste waters has been studied using HPLC coupled to ICP-MS.⁴⁴ The developed method was also able to simultaneously measure I- and Gd-containing pharmaceuticals present at trace levels (ng L⁻¹), showing the versatility of ICP-MS as a multi-elemental speciation detector when coupled with LC. The methodology used HILIC to separate and quantify ten drug residues, including: three I-, three Pt-, and four Gd-complexes. Three HILIC stationary phases, including two zwitterionic, i.e. phosphorylcholine and ammonium-sulfonic acid phases, and one silica charged phase, were compared using different compositions of mobile phase, including organic solvent percentage, salt concentration and pH. The phosphorylcholine stationary phase with an acetonitrile gradient from 75 to 50% (v/v) in 10 mmol L⁻¹ ammonium formate aqueous solution, achieved the separation of the 10 tested complexes. The LOD values for the various complexes were between 1 and 3 ng L⁻¹ Gd, 3 and 17 ng L⁻¹ Pt and 46 and 63 ng L⁻¹ I. The developed method was applied to urban and hospital wastewaters. Two I-containing compounds iohexol and iopamidol and one Gd compound, gadoterate, were quantified while no Pt-drug compounds or metabolites were detected in the samples studied.

4.15 Selenium

Two reviews on Se speciation have been published this year. One review focused on the use of nanostructured sorbents for the separation and preconcentration of iSe species in environmental waters using SPE.⁶⁴ The paper compiled 70 recent references and includes tables and schematic illustrations describing experimental procedures for the determination of tSe, Se^IV, and Se^VI using different nanostructured materials. Attention is given to nanoscale zero-valent iron, layered double hydroxides (LDHs), and metal–organic frameworks (MOFs), which are highlighted as emerging materials for Se accumulation and extraction from environmental samples. The review also discussed surface modification strategies, such as functionalisation with polymers or metal oxides, aimed at enhancing sorption efficiency. The authors emphasise the need for methodologies based on nanosorbents capable of accumulating multiple Se species followed by selective elution, as an alternative to the commonly applied approach based on selective sorption of Se^VI and indirect determination of Se^IV by difference, which remains relatively costly and time-consuming. The second review covered here provides an overview of analytical methodologies for the determination of Se species in staple crops.⁶⁵ The article includes 147 references and critically examines well-established analytical approaches, outlining their applications, advantages, and limitations. A section is devoted to sample preparation strategies for Se speciation. Although the review may be particularly useful for researchers entering the field, the presentation of analytical techniques and sample treatment procedures lacks clear systematic organisation. Moreover, some techniques, such as HPLC-ICP-MS and GC-ICP-MS, are described as recent advances in speciation, despite being widely established methods. Notably, the review also presents several examples of the application of LIBS for Se speciation.

The fortification of plants or mushrooms with Se is the subject of two papers. A mixed-mode (RP/AEC) chromatographic method coupled to ICP-MS was applied for the rapid determination of Se species in enzymatic extracts of Pleurotus mushrooms supplemented with Se^IV and Se^VI.⁶⁶ Separation of the Se compounds was achieved using a Newcrom AH column (4.6 × 150 mm, 5 µm) and an isocratic mobile phase consisting of 3 mmol L⁻¹ H₃PO₄ and 10% acetonitrile (pH 1.8). The study mainly focused on the optimisation of sample preparation and chromatographic conditions. Under optimised conditions, Se^IV, Se^VI, SeMet, MeSeCys, and SeCys₂ were baseline separated in less than 7 min, representing a faster analysis than most separation approaches reported in the literature. Column recoveries for the different Se species ranged from 92 to 103%. Several extraction procedures, both with and without enzymatic assistance (protease, pancreatin, pepsin, and driselase) were evaluated. The use of pepsin and protease XIV in Tris–HCl with UAE was identified as the most efficient approach, providing extraction recoveries between 67 and 102%. Application of the method revealed partial transformation of inorganic Se to SeMet, and higher concentrations of inorganic Se were detected in mushrooms cultivated in the presence of Se^VI. The analytical procedure provided limits of detection as low as 0.5 µg L⁻¹ for the Se species studied. Method accuracy was assessed through spiking experiments, yielding recoveries between 92 and 109%. Another study investigated Se speciation in Chuju plants (Chrysanthemum) grown in soils fertilised with Se^IV and Se^VI under field conditions.⁶⁷ This approach is relatively uncommon, as most Se biofortification studies are conducted under hydroponic conditions. In addition, the influence of geographical origin on Se speciation was evaluated. Chuju plants and rhizosphere soils were collected from different regions in China and subsequently transplanted to a common site to perform Se-enrichment experiments under comparable conditions. After plant growth and harvesting, total Se was determined by acid digestion followed by ICP-MS, whereas Se speciation was investigated after enzymatic hydrolysis using HPLC-ICP-MS. However, the absence of details regarding the chromatographic column and separation conditions limits the analytical evaluation of the methodology. The results indicated that in Se^IV-enriched roots most Se remained as Se^VI, suggesting a limited conversion to organic Se species, with a substantial fraction of Se being translocated to the aerial parts of the plant as Se^VI. In contrast, roots exposed to Se^IV contained both Se^IV and SeMet, and Se was mainly transported to the aerial tissues in organic forms. The predominant Se species in all plant organs following Se^IV treatment was identified as SeMet. Nevertheless, the overall Se uptake and enrichment capacity of plant tissues was significantly higher in plants exposed to Se^VI than in those treated with Se^IV, likely because Se^IV tends to be more strongly adsorbed and immobilised in soils, reducing its mobility and bioavailability for plant uptake.

In recent years, selenoneine has attracted increasing attention. Selenoneine has been proposed to play a key biological role as a potent free-radical scavenger and as a protective agent against Hg-induced toxicity. However, its determination remains analytically challenging due to its limited stability, as it readily oxidises to a dimeric form. Consequently, in-house prepared standards are commonly employed. Several stabilisation strategies have been proposed, including oxidation with H₂O₂ or acetonitrile to produce oxidised selenoneine, or reduction with dithiothreitol (DTT) to maintain the reduced form, allowing its quantification in a single defined species. Following this approach, an analytical method based on the on-line coupling of HPLC-UV-thermal treatment-HG-AFS was developed for the simultaneous determination of selenoneine and other Se species (SeMet and MeSeCys) in aqueous extracts of seabird liver.⁶⁸ Both reduced and oxidised selenoneine were used as standards. Separation was achieved on a C8 column (250 × 4.6 mm, 5 µm) using 2% MeOH in 0.1% HCOOH as mobile phase under isocratic conditions (0.7 mL min⁻¹). The column effluent was mixed with a reducing solution (5% KBr in 50% HCl) and subjected to UV irradiation followed by thermal treatment to quantitatively convert Se species to Se^IV. The resulting Se^IV was subsequently introduced into the HG-AFS system for hydride generation using 0.7% NaBH₄ in 1% NaOH. Structural confirmation of reduced and oxidised selenoneine in the standards was performed by HPLC-ES-MS/MS, while total Se content in an in-house standard was determined by HG-AFS and ICP-MS. The study demonstrated that the combined UV and thermal treatment was required for quantitative conversion of selenoneine to Se^IV. Under optimised conditions, selenoneine eluted at approximately 7 min, and the Se species (SeCys₂, MeSeCys, selenoneine and SeMet) were separated within 12 min. However, complete baseline separation between selenoneine and MeSeCys was not achieved. The method provided a LOD of 0.5 µg Se L⁻¹ and a relative standard deviation as low as 1.6%. In seabird liver extracts, selenoneine was the only Se compound detected in the water-soluble fraction, and its presence was further confirmed by HPLC-ES-MS/MS. In a separate study, selenoneine was extracted from several seafood and seafood-derived products and determined by SEC-ICP-MS.⁶⁹ Several thiol-reactive reagents dithioreitol (DTT), glutathione (GSH) and tris(2-carboxyethyl)phosphine hydrochloride (TCPE) as well as protease treatment to enhance extraction efficiency were evaluated. Selenoneine standards were obtained from a genetically modified fission yeast strain (FY25320) cultured in the presence of 10 µmol L⁻¹ Se^IV, followed by extensive purification. Optimal extraction of selenoneine from samples was achieved using 50 mmol L⁻¹ DTT combined with UAE for 1 h and incubation at 37 °C for 24 h. Extracts were analysed using SEC column with a mobile phase consisting of 0.1 mmol L⁻¹ CH₃COONH₄ and 0.1% IGEPAL®. A Se-containing peak detected at 10 min was identified as selenoneine. The method reported LOD and LOQ values of 0.020–0.030 mg kg⁻¹ and 0.067–0.099 mg kg⁻¹, respectively. Method accuracy was evaluated by spiking experiments, yielding recoveries between 94 and 106%. Selenoneine was detected in all seafood samples analysed, with the highest levels found in bluefin tuna, 7.82 ± 0.13 mg kg⁻¹, corresponding to approximately 90% of total Se. Lower concentrations were found in yellowfin tuna, Pacific mackerel, swordfish and skipjack tuna (0.17–0.26 mg kg⁻¹), representing about 20% of total Se. Interestingly, selenoneine was also detected in processed products such as canned fish, dried fish flakes and Japanese soups, indicating the stability of this compound during thermal processing at 100 °C. Due to its high water solubility, approximately 70% of selenoneine was found in the liquid fraction of soups prepared from processed fish products.

Methods for the production and detection of SeNPs continue to be reported. Kieliszek et al. investigated the biogenic synthesis and characterisation of SeNPs produced by two yeast strains (Saccharomyces cerevisiae ATCC 7090 and Rhodotorula glutinis CCY 20-2-26) grown in the presence of Se^IV (1000 mg Se per L).⁷⁰ The total Se in yeast biomass was determined after MAE by ICP-MS. Selenium speciation was investigated in aqueous extracts treated with DTT and iodoacetamide to reduce disulfide bridges, followed by enzymatic digestion with protease. Chromatographic separation of Se species was achieved using a C8 column (4.6 × 250 mm, 5 µm) with formic acid and MeOH under gradient elution, while structural identification was performed by UHPLC-ES-Orbitrap-MS. Isolation of SeNPs from yeast pellets involved a sequential enzymatic extraction consisting of water extraction, driselase digestion, protease digestion and SDS extraction. The final supernatant was analysed by ICP-MS and spICP-MS to determine total Se and SeNPs, respectively. Results showed that S. cerevisiae exhibited a higher capacity for Se accumulation (1559 µg Se per g) compared with R. glutinis (1202 µg Se per g), with most Se associated with the insoluble fraction. Speciation analysis revealed the presence of SeMet and SeCys, as well as several known and potentially new Se derivatives such as glutathione-2,3-DHP-selenocysteine and methylthioselenoglutathione. Measurements by sNP-ICP-MS revealed differences in NP_size distribution between yeast strains, ranging from 60 to 180 nm for S. cerevisiae and from 40 to 260 nm for R. glutinis. The authors proposed both yeasts as promising producers of Se-enriched dietary supplements and nanomaterials with potential biotechnological applications. Investigations of SeNPs as protective agents against cisplatin-induced nephrotoxicity in cell culture models have also been conducted.⁷¹ Two cell lines (RPTEC/TERT1 and HeLa) were exposed to cisplatin, modified chitosan-Se nanoparticles (Ch-SeNPs) and SeMet, either individually or in combination. Several analytical techniques were applied, including MTT assays for cell viability, TEM for NP localisation, and both SEC and AEC coupled to ICP-MS or UV-vis detection to evaluate Se transformations during cellular uptake. The results indicated distinct behaviours for the two Se species. It was found that SeMet reduced intracellular levels of cisplatin and Pt-DNA adducts, whereas SeNPs did not significantly affect cisplatin uptake or its binding to nuclear DNA. However, co-administration of SeNPs produced a synergistic enhancement of the antitumor activity of cisplatin, suggesting their potential use as adjuvants to mitigate nephrotoxicity without compromising anticancer efficacy. Cytosolic extracts from cell culture were also submitted to SEC and AEC analysis coupled with ICP-MS. The separation of Se species by SEC was achieved with 50 mmol L⁻¹ NH₄HCO₃ (pH 8.4) as the mobile phase, whereas AEC analysis was performed by means of with a mobile phase composed of 10 mmol L⁻¹ ammonium citrate in 2% MeOH (pH 5.0). The SEC chromatograms revealed that in cells exposed to Ch-SeNPs or SeMet, an increase in Se-containing peaks corresponding to MW between 66 and 79.5 kDa was observed, which the authors attributed to selenoproteins. Comparative analysis suggested that Ch-SeNPs stimulated selenoprotein synthesis to a greater extent than SeMet. Co-incubation with cisplatin resulted in a decrease in the intensity of these peaks. The AEC-ICP-MS analyses further indicated transformation of both SeMet and SeNPs into SeCys₂ and MeSeCys. These findings were used by authors to support the assignment of the SEC peaks to selenoproteins, although further confirmation using molecular MS techniques would be required.

The metabolism of Se in animals has been the subject of three papers. Gut bacteria have been shown to produce DMSe, MMSe and DMDSE from Se^IV, SeMet and MeSeCys. However, the nutritional availability of these volatile methylated Se compounds remains unclear. In this context the content of two serum proteins (GPX3 and selenoprotein P) and TMSe and selenosugars were measured in blood serum and urine of Whistar rats supplemented with MeSeCys, DMDSe and DMSe.⁷² The Se speciation in serum was carried out by its direct injection onto a SEC column with 50 mmol L⁻¹ Tris–HCl (pH = 7.4) at a flow rate of 0.6 mL min⁻¹ as the mobile phase. Selenium speciation in urine was performed by means of SEC column and 50 mmol L⁻¹ NH₄Ac (pH = 6.5) flowing at 0.6 mL min⁻¹ as the mobile phase. The Se species were detected by ICP-MS/MS in the O₂ mass shift mode. Both MeSeCys and DMDSe were efficiently employed for the biosynthesis of selenoproteins and the production of urinary metabolites, whereas DMSe is neither incorporated into serum selenoproteins nor metabolised into urinary Se-sugars. The incorporation of DMDSe into selenoproteins suggests its demethylation to selenide. By contrast, DMSe, the dimethylated compound, was primarily methylated into TMSe, a non-bioactive form of Se, which is readily excreted in urine, facilitating the elimination of Se from the body. The authors conclude that gut microbiota contributes to Se metabolisms through methylation processes. An interesting study with rats evidenced that SeMet administration in vivo decreases the oxidative stress induced during irrigation-reperfusion injury of ablated organs, so decreasing proteolysis and consequently favouring post-mortem organ preservation.⁷³ Oral supplementation of SeMet to rats was at a dose rate of 150 mg kg⁻¹ per day for 72 days. After this period, liver, kidney and heart were ablated and proteins were extracted at different post-mortem intervals. Total protein analysis by the Bradford method showed a decrease of proteolysis in kidney and heart but not in liver. Proteins and proteolysis products profile (PPs) changes were analysed by a SEC-UV with 100 mmol L⁻¹ phosphate buffer and 0.025% NaN₃ as the mobile phase, whereas GPx1 was determined by affinity chromatography-ICP-MS by using a heparine-sepharose column. The SeMet-administered rats showed lower PPs (<29.5 kDa) concentration than controls, showing that SeMet decreased proteolysis in the organs tested. The SeMet administration decreased GPx1 degradation in the liver, kidney, and heart, with the greatest effect being in the liver. The evaluation of the proteolysis velocity of GPx1 at different post-mortem intervals showed a faster decay in the kidney followed by the liver. In the heart, the proteolysis velocity of GPx1 was constant and lower. In a different study, the metabolisms of selenotrisulfides (STS, R-S-Se-S-R) were evaluated in vivo by X-ray fluorescence and high-resolution fluorescence-XAS.⁷⁴ The STS compounds are considered as the major metabolic intermediates in the bioconversion of iSe to Se organic compounds and are very reactive with thiol-containing proteins facilitating Se transport and subsequent utilisation within the body. In the current work, STS compounds were in-house synthesised by reacting of L-penicillamine with selenious acid to produce L-penicillamine selenotrisulfide (PenSSeSPen), whereas selenenyl penicillamine bound to rat serum albumin (RSASSeSPen) was synthesised from PenSSeSPen and albumin from rat serum (via thiol-exchange reaction). In both cases, purity was assessed by Se content measurement by ICP-MS. Rats were administered with 1 mg Se per kg b.w, of STS compounds via intraperitoneal injection. The ICP-MS measurements confirmed a significant increase in Se content in liver, kidney and blood after 48 h treatment. Interestingly, XRF revealed Se and Cu co-localisation in the Se-treated rodent, especially in the renal cortex and to a lesser extent in renal medulla, liver and blood. The use of SeKα1 HERFD-XAS combined with PCA indicated the presence of CuSe and FeSe. The elevated proportion of CuSe found compared to other Se species of biological relevance was quite unexpected, with the authors suggesting further investigation by improving experimental design as well as the use of complementary techniques such as HPLC-ICP-MS. The STS compounds were identified in the spectra of all tissues highlighting the role of these compounds in transporting Se through the bloodstream and organs. The occurrence of selenosugars binding to proteins in aqueous liver extracts from turkey and male rats fed with Se-containing diets has been demonstrated using LA-ICP-MS.⁷⁵ Extracts were subjected to SDS-PAGE and Native-PAGE with and without pretreatment with β-mercaptoethanol. The separated proteins were electrophoretically transferred to membranes, which were then analysed by LA-ICP-MS. The results showed that without applying β-mercaptoethanol, Se appeared distributed across the protein molecular weight profile, whereas β-mercaptoethanol dramatically reduced Se binding, releasing selenosugar-decorating proteins and leaving just selenoproteins. The use of LA-ICP-MS demonstrated that selenosugars are present in the HMW fraction.

Finally, Se speciation methodologies have also been applied to monitor the removal of Se from electrolyte solutions used in the production of high-purity Ni.⁷⁶ The presence of Se impurities during the electrodeposition process can adversely affect the quality of Ni-based alloys. Therefore, efficient removal of Se is required, although this remains challenging due to the complexity of industrial electrolyte matrices. In this study, electrolyte solutions and Ni deposits were characterised by HPLC-ICP-MS and XPS, respectively. Nickel deposits were prepared by adding controlled amounts of Se to industrial Ni electrolyte solutions. The XPS analysis revealed that Se was present in the deposits mainly as NiSe, NiSe₂, Ni₃Se₂, and SeO₂. These results suggest that Se is initially reduced to elemental Se during the electrodeposition process, which subsequently reacts with Ni to form predominantly NiSe and Ni₃Se₂. After systematic optimisation, reduction with NaBH₄ was identified as the most effective approach for Se removal from electrolyte solutions containing 20 mg L⁻¹ of Se^IV/Se^VI. To minimise the interference caused by the high concentration of Ni during chromatographic analysis, the electrolyte solution was pretreated with methanol to precipitate NiSO₄. Selenium species were then separated using a AEC with a NaHCO₃ mobile phase at pH 8.4. Under optimised conditions (25 °C, pH 2, and 1000 mg L⁻¹ NaBH₄), removal efficiencies of 99% for Se^IV and 84% for Se^VI were achieved. Overall, this work illustrates an interesting and less conventional application of Se speciation analysis in the field of metallurgical and metal-processing industries.

4.16 Tin

One review on Sn speciation has been published by Song et al.,⁷⁷ which comprehensively overviews the literature since 2012. The review has 152 references, with some predating 2012. The authors discuss the occurrence, toxicity, pretreatment and detection methods available for OTCs for different environmental matrices. The review includes a useful table of the methods that have been used to determine a range of OTCs in a wide range of matrices. The advantages and disadvantages of the various approaches, including different pretreatment techniques, are discussed.

4.17 Titanium

To better understand the reasons for implant failure, Liu et al. studied the imaging and speciation information available regarding intracellular metallic implant debris with SXRFM.⁴¹ The study rationale and method details are given in the final paragraph of Section 4.5. Most of the speciation results were for Co and Cr (discussed in Sections 4.6 and 4.5, respectively), whereas they found that Ti debris occurred in a mixed crystalline/amorphous oxide state. The researchers were unclear as to whether this chemical state resulted from the tribochemical processes at the implant surface or from intracellular alterations.

4.18 Vanadium

The presence of asphaltenes in crude oil poses significant challenges regarding extraction, transportation and storage due to precipitation, largely due to porphyrinic species containing transition metals, primarily Ni and V, which also deactivate catalysts during refining. Characterising these compounds is, therefore, essential for understanding the mechanisms of asphaltene aggregation and optimising the refining processes. Thus, Machado et al. have devised a fractionation-speciation method based on 2D paper chromatography and LA-ICP-MS.⁵⁸ They first separated the crude oil samples into asphaltene and maltene fractions by adding 1 mL of heptane to 100 mg of oil to precipitate the asphaltenes, which were redissolved in 1 mL of toluene and 1 µL applied the chromatographic paper plate. The supernatant containing the maltenes was evaporated to dryness and the residue taken up in 1 mL of toluene. The asphaltenes were fractioned by successive elution with solvents of increasing polarity with drying between each elution: the plate was first eluted up to 1 cm with hexane, then up to 2 cm with toluene, next up to 3 cm with dichloromethane and, finally, up to 4 cm with MeOH. The maltene fraction (1 µL) was similarly fractioned with three solvents, hexane, hexane + dichloromethane (70 + 30 v/v), and toluene + MeOH (50 + 50 v/v). The plate was rotated 90° and a further fractionation carried out with hexane, toluene, dichloromethane and MeOH. Between each solvent, the plate was removed from the solvent chamber and dried before inserting into the next solvent chamber. Signals were acquired for both ⁶⁰Ni and ⁵¹V during ablation with radiation from a Nd:YAG laser operating at 213 nm, 20 Hz, spot size of 110 µm and a scan speed of 110 µm s⁻¹. Identification was aided by running standards through the same procedures, including vanadyl tetraphenoxyphthalocyanine, vanadyl phthalocyanine, Ni octaethylporphyrin, Ni phthalocyanine, and Ni tetrasulfonated phthalocyanine. To confirm the identities of the species, HR-MS was applied to the contents of paper squares (1 cm × 1 cm) dissolved by UAE in 1.5 mL toluene : methanol, 50 : 50 v/v, + 0.01% formic acid. One Ni compound (octaethylporphyrin) and five oxygenated-V porphyrins were identified in the more polar fractions. They used one heavy crude oil for the development of the method, which was applied to another six crude oils ranging from light to heavy. The researchers envisioned that the high sample throughput of 20 min per sample together with minimal reagent consumption could lead to routine applications via either the establishment of ICP-MS facilities near refineries, allowing near real-time monitoring of Ni and V, or by centralised laboratories acting as analytical hubs for multiple refineries.

Thakur et al. have devised a method for the selective determination of V^V in a wide variety of materials by SCF followed by ICP-MS. They used a commercial SCF apparatus with CO₂ as the extracting medium to which they added a solution of a chelating agent, 2,8,14,20-tetramethyl-11,23-dinitro-calix[4]resorcinarene-5,17-dihydroxamic acid (NMC4RADHA) in octanol.⁷⁸ As the V complex has a large molar absorptivity (1.8 × 10⁴ L mol⁻¹ cm⁻¹), they also quantified V by solution spectrophotometry. All samples were subjected to MAE with concentrated HCl and HNO₃, but no details were given apart from the final volume (100 mL). They presented results for the analysis of 16 foods, tobacco, blood, urine, hair, nails, river water, seawater, effluent and phosphate rock, all of which were given in units of µg mL⁻¹ They also analysed a number of CRM metal alloys containing percent concentrations of V. Some of the results were obtained by ICP-OES, but it was not clear which those were. The LOD for the ICP-MS method was given as 0.005 ng mL⁻¹ in the text, but this is contradicted by the numbers in the abstract. They indicated that the introduction of the SCF effluent into the plasma spectrometer resulted in a 40-fold increase in sensitivity, but there is some doubt as to whether the extraction was directly coupled to the spectrometer although this is what was shown in the schematic diagram of the apparatus, as the diagram also showed the spectrometer located prior to a back-pressure regulator. The only speciation studies reported were performed with standard solutions (at single and low double-digit ng mL⁻¹ concentrations) in a batch LLE experiment. Following extraction of the V^V complex (by NMC4RADHA in octanol) from 6.5 mol L⁻¹ HCl, the remaining V^IV was oxidised to V^V by 0.1 mol L⁻¹ KMnO₄ solution and the extraction repeated.

4.19 Zinc

Microalgae have attracted significant attention due to their potential to address nutritional deficiencies as alternatives to synthetic supplements and a study investigating Zn speciation in microalgae using HPLC-ICP-MS and HPLC-ES-MS/MS has been reported.⁴⁰ Several extraction reagents were evaluated for isolating Zn from dried ground microalgae, including 100 mmol L⁻¹ NH₄Ac (pH 7.4), 2% SDS, 30 mmol L⁻¹ Tris–HCl, methanol : water (1 : 1), 7% Flavourzyme, 0.1% Viscozyme, and 1.5% Alcalase. The extraction temperature and time were optimised for each condition. The selection of the extracting solution was based on chromatographic profiles obtained by SEC-ICP-MS. After optimisation, Flavourzyme was selected as the most effective extractant, enabling the recovery of multiple Zn species as well as LMW compounds not obtained with other reagents. The Zn-containing compounds in the extracts were separated using a ZIC-HILIC column (150 × 2.1 mm, 100 Å) with a mobile phase consisting of 100 mmol L⁻¹ NH₄Ac and acetonitrile under gradient conditions. Several Zn species were identified, including Zn–nicotianamine in diatoms cultivated in Poland, and Zn–citrate, Zn–nicotianamine, and a Zn complex with 2′-deoxymugineic acid in Spirulina grown in Zn-enriched media.

5 Biomolecular speciation analysis

An important requirement for the measurement of intact proteins and other metallobiomolecules in complex biological systems are methods with the chromatographic power to resolve the species of interest prior to detection. In the current review period, three papers have been published that investigate improvements to the chromatography used in the separation phase. One of the papers proposes a generic purification for use in metalloproteomics whilst the other two focus on specific applications using SEC in combination with other chromatographic modalities.

A comprehensive and technically challenging study has been reported for the development of a generic method, termed the Metalloproteome Factory for the isolation, purification and characterisation of metalloproteins in plants, but also with applicability to other cellular systems.⁷⁹ The main technical challenges for metalloproteomics include: achieving a sufficient separation of the many thousands of metalloproteins in an organism to identify the actual target proteins, and simultaneously achieving sensitivity in metal detection and protein identification suitable even for low-abundance proteins, which can play a key regulatory role in all organisms. To achieve these requirements the authors describe the use of different HPLC separation modalities in different combinations, for the investigation of metalloproteomics in various cell-based samples. Three different chromatographic separation approaches were used to investigate the metal-containing fractions and used on-line coupling in parallel to two detectors, SF-ICP-MS in medium or high-resolution mode and DAD, to identify metal-containing peaks for isolation using fraction collection. The different separation chromatographies used included: AEC followed by either chromatofocusing (CF) or hydrophobic interaction chromatography (HIC), and finally, SEC. Briefly, CF is a variant of IEC in which a pH gradient elutes proteins mostly depending on their isoelectric point and is similar to isoelectric focusing, a technique used in gel electrophoresis. As the elution by a pH gradient leads to a different elution sequence than elution by a salt gradient at constant pH, CF can separate proteins that co-elute on AEC, thus making it an attractive option as a second dimension after AEC. For optimal compatibility with ICP-MS, all the buffers used were based on volatile ammonia salts that do not lead to clogging of the torch injector. Metal-containing peaks of interest were isolated during the chromatographic run using fraction collection and then further structural characterisation undertaken using planar gel-electrophoresis and analysis by trapped ion mobility mass spectrometry. The sensitivity and S/N ratio reported were sufficient to identify proteins at concentrations <1 nmol L⁻¹ and total amounts <1 pmol in the final SEC. The project developed a method for purifying, on an analytical scale, a very wide range of metalloproteins without requiring specific adaptations for a particular protein. This was possible by making two previously ICP-MS-incompatible types of chromatography, HIC and CF, fully compatible with on-line coupling to SF-ICP-MS. To check how far 3D performance improves compared to the previously best 1D-metalloproteomics, a myoglobin standard was spiked into a soybean root extract. The mixture was then run either only over SEC-ICP-MS or the new 3D method. This showed that with the 1D method, most of the myoglobin in the mix would have been missed due to oligomerisation, which led to a different elution time in the mix compared to the pure standard, whereas the developed method was able to account for this. However, only a single known protein was used to validate the power of the method for the identification of unknown proteins, which seems rather limited and further work should focus on addressing how robust the method is to the identification of low-abundance proteins in complex extracts.

The use of SEC hyphenated in parallel to SF-ICP-MS and ES-MS for the analysis of metalloprotein profiles in human samples shows that simultaneous assessment of elemental and molecular information is feasible, if specific aspects are considered.⁸⁰ This proof of concept experimental work with a limited sample set, has investigated and compared the use of different types of SEC columns for the separation of metalloproteins in 3 subjects in matched serum and CSF fluid samples. The early detection and determination of the treatment response for neurodegenerative diseases is challenging due to the limited number of specific biomarkers that can be used to quantify disease state and progression. Typically, either elemental or molecular signatures are targeted, but simultaneous detection of both could yield further information and thus diagnostic potential. An approach to study elements associated with biomolecules that has the potential to identify anomalous elemental signatures linked to alterations in metalloprotein or phosphorylation patterns in human body fluids has been reported. Optimisations and measurements were performed using UHPLC coupled in parallel to a SF-ICP-MS instrument equipped with a desolvating nebuliser system and a Q Exactive Orbitrap MS equipped with a heated ES source. A DAD detector was also used and placed in series with the ES-MS instrument. Parallel coupling of the two MS instruments was achieved via a micro-splitter valve, with flows set to a 2 : 1 split using a liquid flowmeter. Interestingly, dilution of the eluent flow (3 : 1) with water was required to maintain a stable ion beam for the ES-MS. The SEC columns from four different suppliers were compared: AdvanceBio SEC 200A, PEEK, 1.9 µm, 2.1 × 150 mm; Inertsil WP300 Diol, PEEK, 5 µm, 4.6 × 250 mm; TSKgel BioAssist G3SWxl, PEEK, 5 µm, 7.8 × 300 mm; and bioZen SEC-3, titanium, 1.8 µm, 4.6 × 300 mm. All column solid phases tested were resins composed of ultrapure spherical silica covered with a diol phase. Eluents consisting of either NH₄HCO₂ or CH₃COONH₄ at different concentrations in water and a varying organic modifier content (0–18% v/v) were evaluated. The final experimental setup used composed of two serially installed Inertsil WP300 Diol (4.6 × 250 mm, 600–10 kDa) columns for high molecular weight separations and in the low molecular weight region, a separation was achieved with a ProteoSil 100-SEC column (4.6 × 250 mm, <5 kDa). Afterwards, the flow was split 2 : 1 and directed in parallel to the ICP-MS and DAD/ES-MS instruments. Variation in the split flow rates was usually found to be within ± 5% so that a constant splitting ratio and no clogging on the ICP-MS side of the system was observed. Clogging of the ES needle could be correlated to increased backpressure of the HPLC and was observed during the optimisation of the setup with undiluted serum samples. Using the developed system, the ICP-MS LOD values were: Al 0.32, Ca 24.3, Cd 0.18, Co 0.37, Cu 0.84, Fe 3.00, Hg 5.28, I 96.6, Mg 29.7, Mn 4.48, Ni 0.81, P 161, Pb 0.07, S 410, Se 91.3, Sr 3.08 and Zn 0.22, all nmol L⁻¹. The method was applied to the analysis of three matched CSF and serum samples. The results suggest that the simultaneous acquisition of molecular and elemental mass spectra yields additional information for data interpretation and biomarker identification. However, one major limitation of this approach is the inherent lower resolution and separation capability of SEC in comparison to other techniques like RPLC, IEC or CE. This trade-off was justified as necessary to ensure best stability of labile protein-metal complexes, and compatibility with both ionisation sources so that information can be obtained from both mass spectrometers. However, many applications using AEC have been successfully used to characterise metalloproteins in human biofluids.

An SEC-ICP-MS platform has been developed recently for the characterisation of metallobiomolecules in human serum, with a range of applications, including: understanding metal homeostasis, metal biomarker discovery, and toxicity assessment.⁸¹ The method enables the simultaneous detection and quantitation of ten metals and metalloids including Ca, Co, Cu, Fe, Hg, Mg, Mn, Pb, Se, and Zn within a single sample. The SEC column used was a Discovery BIO GFC 150 (30 cm × 4.6 mm), and an eluent containing 100 mmol L⁻¹ ammonium acetate pH 7.2 at a flow rate of 0.7 mL min⁻¹. The overall approach used 4 injections at specific time points during the run, which included: an initial on-column injection of the blood serum sample (0 s); diluted blood serum injection (200 s) for total element concentration determination; sensitivity control standard injection (300 s), for instrumental stability monitoring; and lastly on-column injection (550 s) of an EDTA solution (8 mmol L⁻¹), and the metal–EDTA complexes formed in the column are eluted, after the end of the chromatographic analysis (1400–1600 s). By using this complex sample introduction programme integration of the post-column FI peak allowed for element calibration using ionic standards containing EDTA, total element determination following the injection of an acid-diluted blood serum sample, and instrument sensitivity monitoring and correction. Method validation was performed using a human RM (Seronorm Trace Elements Level 2). Element recoveries exceeded 80% for most analytes, both following total element determination and column elution, confirming the robustness and accuracy of the approach. An on-column EDTA injection strategy effectively mitigated metal/metalloid interactions with the stationary phase, enhancing column recoveries for Co and Zn while preventing cross-contamination between samples. The developed analytical platform expands the number of elements that can be simultaneously monitored and quantified, and maximises the information obtained for the analysis of each sample. The analysis of the Seronorm RM provided insights into the element distribution across different relative molecular mass ranges. The detected metallobiomolecule bands aligned with biomolecules known to be present in human serum and previous studies of this RM. It was unfortunate that important elements, such as Zn, were not stable in the chromatographic conditions used and were lost from the binding protein and retained on the column to be eluted by EDTA in the final step. Interestingly the authors explain their use of serum rather than whole blood as the sample type to use, because it does not require pre-treatment. However, serum is an isolate from whole blood, which requires pre-treatment in the pre-analytical phase and as a result, may not be representative. Some elements, such as Pb, are significantly different in serum and often analysed in whole blood samples. Lead in particular is determined in whole blood as this is required to conform with legislative requirements. The use of the Seronorm RM, which is a nutritional QC in most clinical laboratories, for the investigation of toxic heavy metals also seems misplaced as it will have very low concentrations of these toxic metals present, which are not representative of true toxicological situations. That said, the on-line approach to measuring total and biomolecular bound metals in a single run is a useful new addition to the methods in this area.

6 Abbreviations

1D	one dimensional
2D	two dimensional
3D	three dimensional
AAS	atomic absorption spectrometry
AB	arsenobetaine
AC	alternating current
ACN	acetonitrile
AEC	anion-exchange chromatography
AFS	ammonium pyrrolidine dithiocarbamate
ASU	angular resolved
CE	capillary electrophoresis
CF	chromatofocusing
Ch-Se	chitosan-Se
COF	covalent organic framework
CRM	certified reference material
CSF	cerebral spinal fluid
CV	coefficient of variation OR cold vapour
DAD	diode array detector
DCM	dichloromethane
DMA	dimethylarsenic
DMDSe	dimethyldiselenide
DMSe	dimethylselenium
DNA	deoxyribonucleic acid
DOM	dissolved organic matter
DTPMP	diethylenetriamine penta(methylenephosphonate)
DTT	dithiothreitol
EDTA	ethylenediaminetetraacetic acid
EPA	Environmental Protection Agency
EPMA	electron probe microanalysis
ERM	European reference material
ES	electrospray
EtHg	ethylmercury
EtOH	ethanol
FFF	field flow fractionation
FLD	fluorescence detection
FPD	flame photometric detector
FTL	ferritin light chain
GC	gas chromatography
GPX3	glutathione peroxidase 3
GSH	glutathione
Hb	haemoglobin
HERFD	high-energy fluorescence detection
HG	hydride generation
HIC	hydrophobic interaction chromatography
HILIC	hydrophilic interaction liquid chromatography
HMW	high molecular weight
HPLC	high performance liquid chromatography
HR	high resolution
iAs	inorganic arsenic
IC	ion chromatography
ICP	inductively coupled plasma
ID	internal diameter
iHg	inorganic mercury
IS	internal standard
iSe	ion selective electrode
ITD	ion-transfer device
KED	kinetic energy discrimination
LA	laser ablation
LC	liquid chromatography
LDH	layered double hydroxide
LIBS	laser-induced breakdown spectroscopy
LMW	low molecular weight
LOD	limit of detection
LOQ	limit of quantification
MAE	microwave-assisted extraction
MC	multicollector
MeHg	methyl mercury
MeOH	methanol
MeSeCys	methylselenocysteine
MMA	monomethylarsenic
MOF	metal organic framework
MRI	magnetic resonance imaging
MS	mass spectrometry
MTT	3-[4,5-dimethylthiazol-2-yl]-2,5 diphenyltetrazolium bromide
Nd:YAG	neodymium-doped: yttrium aluminium garnet
NIST	National Institute of Standards and Technology
NOM	natural organic matter
NP	nanoparticle
NRCC	National Research Council of Canada
OES	optical emission spectrometry
OTC	organotin compounds
PAGE	polyacrylamide gel electrophoresis
PCA	principal component analysis
PEEK	polyetheretherketone
PSA	primary secondary amine
PTE	potentially toxic element
pXRFS	portable X-ray fluorescence spectrometry
Q	quadrupole
QC	quality control
RM	reference material
RMM	relative molecular mass
ROX	roxarsone
RP	reversed phase
RSD	relative standard deviation
RTCS	Rio Tinto chloride slag
S/N	signal-to-noise ratio
SAX	strong anion exchange
SCF	supercritical fluid
SCX	strong cation exchange
SD	standard deviation
SDS	sodium dodecylsulfate
SEC	size exclusion chromatography
SeCys	selenocysteine
SEM	scanning electron microscopy
SeMet	selenomethionine
SF	sector field
SFC	supercritical fluid chromatography
SI	Système International d’unités
SPE	solid phase extraction
SPME	solid phase microextraction
SR	synchrotron radiation
SRM	standard reference material
ssIDMS	species specific isotope dilution mass spectrometry
STS	selenotrisulfides
TBAH	tetrabutyl ammonium hydroxide
TCPE	tris(2-carboxyethyl)phosphine hydrochloride
TEM	transmission electron microscopy
TES	transition edge sensor
TMAH	tetramethylammonium hydroxide
TMSe	trimethylselenium
Tris	tris(hydroxymethyl)aminomethane
UAE	ultrasound-assisted extraction
UHPLC	ultra-high-performance liquid chromatography
US FDA	United States Food and Drug Administration
UV	ultraviolet
VG	vapour generation
Vis	visible
WHO	World Health Organisation
XANES	X-ray absorption near-edge structure
XAS	X-ray absorption spectroscopy
XPS	X-ray photoelectron spectroscopy
XRD	X-ray diffraction
XRF	X-ray fluorescence

Conflicts of interest

There are no conflicts to declare.

Data availability

There is no additional data associated with this article.

References

1. R. Clough, C. F. Harrington, S. J. Hill, Y. Madrid and J. F. Tyson, J. Anal. At. Spectrom., 2025, 40(7), 1615–1644.
2. W. R. L. Cairns, E. C. Braysher, O. T. Butler, O. Cavoura, C. M. Davidson, J. L. T. Torro and M. von der Au, J. Anal. At. Spectrom., 2025, 55.
3. M. Patriarca, N. Barlow, A. Cross, S. Hill, D. Milde and J. Tyson, J. Anal. At. Spectrom., 2025, 40(3), 541–664.
4. E. H. Evans, J. Pisonero, C. M. M. Smith and R. N. Taylor, J. Anal. At. Spectrom., 2025, 40(5), 1136–1157.
5. C. Vanhoof, A. Cross, U. E. A. Fittschen and L. Vincze, J. Anal. At. Spectrom., 2025, 40(9), 2275–2289.
6. A. Agarwal, E. Bolea-Fernandez, R. Clough, A. Fisher, B. Gibson and S. Hill, J. Anal. At. Spectrom., 2025, 40(11), 2982–3022.
7. F. Vanhaecke, J. L. Todolí, G. Schlemmer and M. W. Hinds, Elemental Analysis, De Gruyter, Berlin, Boston, 2025.
8. A. Fisher, Atomic Spectrometric Methods of Analysis, Royal Society of Chemistry, 2025.
9. A. M. Striegel, Macromolecular Separation Science, Springer Nature Switzerland, Cham, 2025.
10. J. M. Herrero-Martínez, M. Miró, E. J. Carrasco-Correa and M. Vergara-Barberán, 3D Printing in Analytical Chemistry, in Sample Preparation, Separation, and Sensing, Elsevier, 2025.
11. T. O. Said and G. M. El Zokm, Organometallic Compounds in the Marine Environment: Analytical and Ecological Aspects, Springer Nature Switzerland, Cham, 2025.
12. G. Chandrakasan, S. Pongpiachan, G. Marcus and G. M. S. Zarazua, TrAC, Trends Anal. Chem., 2026, 195(16), 118589.
13. G. Fadillah, R. Hidayat, A. Saputra, H. S. Wasisto, R. Rani, K. Saeki, K. Toda and S. I. Ohira, Anal. Chim. Acta, 2025, 1380(19), 344637.
14. D. A. Garzón, D. A. Ahumada, C. Paredes and E. Castillo, Anal. Bioanal. Chem., 2025, 417(12), 2717–2726.
15. C. L. Ward-Deitrich, S. Nehete, M. E. D. Busto, S. Hill, F. P. Robertson, L. Luckau, H. Hannemann, B. Fox and H. Goenaga-Infante, Anal. Chem., 2025, 97(46), 25643–25654.
16. J. Lohmann, S. I. Demel, J. C. Sander and T. Reich, Inorg. Chem., 2025, 64(42), 20915–20926.
17. J. Lohmann, C. Tamain, P. Moisy, T. Reich and J. Aupiais, Inorg. Chem., 2025, 64(46), 22674–22682.
18. K. Chadirji-Martinez, G. Hudon, R. Chernikov, E. Heredia, R. F. Feng, A. Crawford and Y. M. Pan, Mineral. Mag., 2025, 89(1), 113–126.
19. R. Sun, E. Dupuis, J. Aupiais and P. E. Reiller, Dalton Trans., 2025, 54(2), 719–727.
20. L. J. Chen, Y. Y. Du, S. Y. Li, X. Liu, J. Shen and Z. L. Zhu, Anal. Chem., 2025, 97(4), 2264–2272.
21. S.O. Otiwa, F. S. Mani and S. Prasad, Appl. Spectrosc. Rev., 2025, 1–33.
22. H. Imaiida, O. Nogawa, M. Mori, K. Toda, S. I. Ohira and M. Mori, Talanta, 2026, 300(8), 129246.
23. H. C. Che, C. Y. Yang, Y. L. Nie, C. Yang, X. K. Tian and Y. Hwu, Water Res., 2026, 288(12), 124684.
24. L. Sosa, D. U. M. Cabrera, W. De Leon, E. A. H. Ramos and A. Y. T. Ventura, Sustainability, 2025, 17(2), 611.
25. D. Ghaffour, A. Leufroy and P. Jitaru, Anal. Bioanal. Chem., 2025, 417(8), 1519–1530.
26. R. Navaretnam, A. Z. Aris, M. F. A. Shukor, M. N. F. Norrrahim, V. F. Knight, T. T. Chin, N. M. Isa and L. J. Looi, Environ. Geochem. Health, 2025, 47(10), 411.
27. I. R. Samarin, A. Sulimanec, T. Orct, A. Benutic, B. Marcius, K. Tomljanovic and J. Jurasovic, Foods, 2025, 14(13), 2261.
28. F. Zhao, R. Dong, D. W. Xu, H. S. Zhang, W. J. Yu, D. Y. Lv, Y. M. Liu and M. Sun, Food Chem., 2026, 500(8), 147436.
29. L. Steiner, A. Raab, B. Lajin, J. Borovicka, J. Truschner and W. Goessler, Anal. Bioanal. Chem., 2025, 8.
30. H. C. Liu, Y. F. Cheng, H. W. Zhou, X. H. Pan, R. F. Peng, Z. J. Bai, J. T. Li and L. Tan, Anal. Methods, 2025, 17(34), 6800–6809.
31. S. Sel, B. Aydogan and I. Koyuncu, Talanta, 2026, 296(9), 128518.
32. C. J. Ma, P. P. Ouyang, Q. Q. Zheng and F. B. Qiu, Anal. Methods, 2025, 17(27), 5726–5735.
33. N. Q. Thang, H. K. Trung, L. V. Tan, N. T. K. Phuong, N. T. K. Ngan and T. Q. Hieu, J. Anal. Chem., 2025, 80(5), 877–886.
34. A. Morales-Rodríguez, B. Vilarrúbies, Y. Badar, P. Eiriz, A. Barquin, J. F. Lopez-Sánchez, D. Barron and A. Sahuquillo, J. Chromatogr., 2026, 1765(7), 466528.
35. A. Kresa, I. Sowik and L. Ruzik, Anal. Lett., 2025, 14.
36. P. Heitland, A. Raab and S. Krell, J. Trace Elem. Med. Biol., 2025, 92(3), 127745.
37. C. Y. Zheng, J. Tang, X. D. Pan, H. T. Shen, Z. Y. Hu, J. S. Zhang, L. X. Wang, P. G. Wu and Y. Tan, Chemosensors, 2025, 13(3), 78.
38. Y. Takahashi, H. Miura, S. Yamadad, O. Sekizawa, K. Nitta, T. Hashimoto, T. Yomogida, A. Yamaguchi, S. Okada, T. Itai, H. Sugaj, T. Kashiwabara, K. Sakata, M. Tanaka, M. Kurisu, M. Nagasawa, T. Kawai, A. Takemoto, S. Yoshida, T. Ohno, H. Tatsuno, R. Hayakawa, H. Suda, Y. Ishisaki, T. Okumura, Y. Ichinohe, T. Hayashi, Y. Imai, H. Noda, T. Tamagawa, T. Isobe, T. Azuma, W. B. Doriese, J. N. Ullom, D. S. Swetz, M. Durkin, G. C. O’Neil, S. Kohjiro, S. Higaki, D. Tsumune, M. Yamamoto, T. Uruga and Y. Kurihara, J. Hazard. Mater., 2025, 495(19), 139031.
39. T. G. Fu, W. Liu, X. L. Fu, Y. H. Huang, H. H. Lou, J. Song and G. L. Li, Agriculture, 2025, 15(5), 545.
40. B. Szczesniak, I. Sowik, S. Oszwaldowski and L. Ruzik, Microchem. J., 2025, 210(8), 113044.
41. S. Y. Liu, S. Chen, B. Lai, O. Antipova, Y. Q. Luo, D. J. Hall, Q. L. Jin, E. Maxey, J. J. Jacobs and R. Pourzal, Sci. Rep., 2025, 15(1), 12467.
42. M. M. Wolle, J. Escavage and P. J. Gray, Food Anal. Meth., 2025, 15.
43. S. Weishaupt, W. Holkenjans, S. Balzer, A. Jeremias, M. Sperling, M. Vogel and U. Karst, J. Chromatogr., 2025, 1746(7), 465808.
44. J. Côme, M. Bueno, P. Karamanis, F. Pannier and S. Mounicou, Talanta, 2026, 298(11), 128974.
45. M. Dalla Costa, D. Dumoulin, M. Lenski, A. Garat and G. Billon, J. Anal. At. Spectrom., 2025, 40(8), 2150–2161.
46. M. Athmer, L. Marotz and U. Karst, J. Anal. At. Spectrom., 2025, 40(8), 2138–2149.
47. V. V. Volchek, D. G. Sheven, P. A. Tkachenko, A. M. Maksimov, M. N. Sokolov and P. A. Abramov, Talanta, 2026, 298(8), 129025.
48. P. C. Gillemot, A. Kretschmer, R. Grimmig, U. E. A. Fittschen and S. Witzleben, Anal. Chim. Acta, 2025, 1362(11), 344185.
49. G. Spileers, P. Tack, L. Vincze and F. Lynen, Anal. Chem., 2025, 97(43), 23905–23913.
50. H. C. Xu, Y. W. Zhu, Y. Peng, C. C. Sun, F. F. Fu and Y. Lin, J. Anal. At. Spectrom., 2025, 40(5), 1364–1372.
51. K. Iwai, M. Iwai-Shimada, N. Tatsuta, Y. Kobayashi, K. Asato, M. Nishimoto, K. Nakai and S. F. Nakayama, Microchem. J., 2025, 211(9), 113142.
52. N. Tatsuta, M. Iwai-Shimada, K. Asato, K. Iwai and K. Nakai, Environ. Res., 2025, 285(7), 122736.
53. E. J. Pacer, C. D. Palmer and P. J. Parsons, Anal. Methods, 2025, 17(8), 1840–1849.
54. M. Iwase, Y. K. Tanaka, Y. Fukumoto, N. Suzuki and Y. Ogra, Anal. Chem., 2025, 97(9), 5306–5312.
55. J. H. Shi, Z. H. Wang, J. B. Chao, L. G. Hu, B. Li and C. F. Dai, J. Chromatogr., 2025, 1755(12), 466049.
56. Y. Chen, P. J. Yang, Y. Y. Guo, Y. W. Liu, T. Jiang, M. Y. Song, J. B. Shi, L. G. Hu, Y. G. Yin, Y. Cai and G. B. Jiang, J. Environ. Sci., 2026, 161, 445–453.
57. P. Lei, J. Zhang, Y. Q. Kong, H. He and H. Zhong, Bull. Environ. Contam. Toxicol., 2025, 114(1), 9–12.
58. E. Q. Machado, B. B. E. Salles, L. W. Hantao, C. J. F. de Oliveira and M. A. Z. Arruda, Talanta, 2026, 298(9), 128988.
59. T. Y. Li, A. Henke, B. L. Wu, Z. Zhao, A. Farsad, M. Serpa, P. Herckes and P. Westerhoff, Water Res., 2026, 288(11), 124554.
60. M. Athmer, A. M. Röhnelt, T. J. Maas, S. B. Haderlein and U. Karst, J. Chromatogr., 2025, 1748(10), 465843.
61. C. Sprigg, H. L. Whitfield, P. T. Leftwich, H. F. Kuo, T. J. Chiou, A. Saiardi, M. L. Shipton, A. M. Riley, B. V. L. Potter, D. Scholey, E. Burton, M. R. Bedford and C. A. Brearley, Biochem. J., 2025, 482(21), 1627–1644.
62. M. Wiktorko, P. Kot, A. Puchala, P. Bryczek-Wróbel, K. I. Rzadkowska and B. Wiaderek, Molecules, 2025, 30(20), 4086.
63. O. Hourtané, D. S. Smith and C. Fortin, Environ. Chem., 2025, 22(5), En25014.
64. K. Pyrzynska, Separations, 2025, 12(3), 64.
65. A. T. Mlangenia, F. Jagota, S. Namaumboa, N. J. Kapitoa, D. C. Tsukuluzaa, L. Bothaa, P. Ndoviea and P. Kumambalab, Chin. J. Anal. Chem., 2025, 53(4), 100511.
66. A. P. de Oliveira, J. Naozuka and J. A. Landero-Figueroa, J. Food Compos. Anal., 2025, 142(10), 107500.
67. Y. Li, N. Li, D. Hu, Z. Chen, G. Yu, J. Liu, X. Xue and L. Zhang, Russ. J. Plant Physiol., 2025, 72(3), 89.
68. C. Marchán-Moreno, C. L. Ward-Deitrich, J. Allen, M. Bueno, D. Amouroux, Z. P. Zayas, H. Goenaga-Infante and W. T. Corns, Spectrochim. Acta, Part B, 2025, 232(8), 107261.
69. E. Matsumoto, T. Seko, Y. Yamashita and M. Yamashita, J. AOAC Int., 2025, 108(5), 779–785.
70. M. Kieliszek, W. Sek, A. Angaïts and J. Szpunar, Biol. Trace Elem. Res., 2025, 14.
71. A. Iglesias-Jiménez, G. Artiaga, E. Moreno-Gordaliza and M. M. Gómez-Gómez, Eur. J. Pharm. Biopharm., 2025, 212(13), 114737.
72. K. Takahashi, S. Iijima and Y. Ogra, Biol. Pharm. Bull., 2025, 48(4), 410–414.
73. P. E. Hasuoka, F. Tonelli, L. Mariño-Repizo and P. Pacheco, Free Radic. Res., 2025, 10.
74. A. T. Baker, C. J. Kidman, L. I. Vogt, G. N. George, D. Sokaras, D. L. Howard, V. F. S. Pape, K. Kulik, S. Yoshida and H. H. Harris, Inorg. Chem., 2025, 64(20), 9961–9983.
75. K. Bierla, J. Szpunar, R. Lobinski and R. A. Sunde, Metallomics, 2025, 17(2), mfaf002.
76. H. Li, Y. T. Xu, Y. Wang, D. Y. Zhang and T. D. Xia, Korean J. Chem. Eng., 2025, 14.
77. Y. Song, Y. Bian, Y. Zhang, J. L. Wang, Z. W. He and X. S. Feng, J. Environ. Chem. Eng., 2025, 13(5), 117625.
78. B. Thakur, N. Pandya, V. Shukla and Y. Agrawal, Bull. Chem. Soc. Jpn., 2025, 98(5), uoaf037.
79. H. Küpper, S. N. H. Bokhari, P. Konik, F. Dycka, I. J. Flores-Sanchez, E. Andresen, N. Jaime-Pérez, A. Colussi and F. Morina, Anal. Chem., 2025, 97(32), 17641–17649.
80. N. Helle, M. A. Hobert, W. Maetzler and M. Gledhill, Microchem. J., 2025, 219(13), 115790.
81. G. Panagou, N. Lydakis-Simantiris and S. A. Pergantis, J. Anal. At. Spectrom., 2025, 40(9), 2582–2598.

---