Atomic Spectrometry Update: review of advances in elemental speciation

Contents

• 1 Topical reviews

• 2 Sample preparation

• 3 Instrumental techniques and developments
  • 3.1 Development in species separation
  • 3.2 Developments in instrumentation
  • 3.3 Chemical vapour generation
  • 3.4 Solid state speciation

• 4 CRMs and metrology

• 5 Elemental speciation analysis
  • 5.1 Antimony
  • 5.2 Arsenic
  • 5.3 Boron
  • 5.4 Chromium
  • 5.5 Cobalt
  • 5.6 Gadolinium
  • 5.7 Iron
  • 5.8 Halogens
  • 5.9 Mercury
  • 5.10 Phosphorus
  • 5.11 Platinum
  • 5.12 Selenium
  • 5.13 Silicon
  • 5.14 Tin
  • 5.15 Vanadium
  • 5.16 Zinc

• 6 Biomolecular analysis
  • 6.1 Direct macromolecular analysis
  • 6.2 Metalloproteins, metalloproteomics and metallomics
  • 6.3 Tagging and labelling for macromolecular analysis
  • 6.4 Elemental and molecular imaging

• Abbreviations

• References

---

Abstract

This is the seventh Atomic Spectrometry Update (ASU) to focus on advances in elemental speciation and covers a period of approximately 12 months from December 2013. This ASU review deals with all aspects of the analytical 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 and relating to elemental speciation. The review does not specifically deal with fractionation, sometimes termed operationally defined speciation. The focus of the research reviewed includes those methods that incorporate atomic spectrometry as the measurement technique. However, because speciation analysis is inherently focused on the relationship between the metal(loid) atom and the organic moiety it is bound to, or incorporated within, atomic spectrometry alone cannot be the sole analytical approach of interest. For this reason molecular detection techniques, such as mass spectrometry, fluorescence and ultra-violet spectroscopy are also included where they have provided a novel approach to speciation analysis. As in previous years, As and Se speciation continues to dominate the current literature, with many derivative type papers being published, exploring all facets of the established analytical protocols used in these areas. Once again there has been an increase in the number of publications in the area of macromolecular analysis, particularly work concerned with metallomics or metalloproteomics and also tagging or labelling of organic molecules to make them detectable by atomic spectroscopy techniques. The use of atomic and molecular mass spectrometry in combination, either simultaneously on-line or sequentially off-line, continues to offer the best main-stream approach for most conventional applications, particularly when standards for the species under investigation are not available and high mass accuracy MS instrumentation can be used for identification.

---

1 Topical reviews

This latest Update adds to that from last year¹ and complements other reviews^2–6 of analytical techniques in the series of Atomic Spectrometry Updates.

In keeping with the pattern established over the past few years, books covering elemental speciation analysis topics have been published. In ‘Environmental and Low Temperature Geochemistry,’⁷ natural processes at and near the earth's surface, as well as anthropogenic impacts on the natural environment are explained. Topics introduced included the concentration, speciation and reactivity of elements in soils, waters, sediments and air, and attention is given to both thermodynamic and kinetic controls. The roles of microbes in processes such as biomineralization, elemental speciation and reduction–oxidation reactions are also discussed Chemical analysis topics include the role of stable isotopes in environmental analysis, radioactive and radiogenic isotopes as environmental tracers and environmental contaminants, and principles and examples of instrumental analysis in environmental geochemistry. The text concluded with a case study of surface water and groundwater contamination that includes interactions and reactions of naturally-derived inorganic substances and introduced organic compounds (fuels and solvents), and illustrates the importance of interdisciplinary analysis in environmental geochemistry. The important topic of microwave-assisted sample preparation for trace element determinations was the subject of a 416-page, 12-chapter book,⁸ in which contributors described the principles, equipment, and applications involved in sample preparation with microwaves for trace element analysis. The 32-page chapter devoted to sample preparation of elemental speciation analysis summarized the work described in just over 100 original journal articles. Only four elements were considered (As, Hg, Se and Sn), with the bulk of the chapter devoted to As, Hg and Se. There is also a 32-page chapter devoted to ‘omics areas’ that includes a section of 4 pages on metallomics.

A review⁹ of ‘complementary’ As speciation methods, in which the reviewers discussed work in which (a) XAS or (b) ESI-MS has been applied alongside HPLC-ICP-MS, is of considerable interest. The reviewers point out a number of limitations of the HPLC-ICP-MS approach, including the serious problem of extraction of the analytes from solid samples without changing the speciation. The reviewers gave numerous examples of ‘discrepancies,’ in which the As species in the plant was shown by XANES to be As^III bound to sulfur, whereas after extraction and separation by HPLC only inorganic As^III and As^V were found. A summary table contains material taken from 18 articles. The reviewers considered that the most effective use of ESI-MS as a complement to ICP-MS is when the two detection modes are used simultaneously for the same HPLC eluent. The problems of selecting a suitable mobile phase that is compatible with both types of instrument was discussed, but the very significant differences in detection capabilities were only mentioned in passing. Several examples are provided of instances when previously unknown As compounds had been identified, in matrices as diverse as sheep urine and giant clam extract, even when the compounds were not resolved by the HPLC separation. A summary table contains material taken from 24 articles. There would appear to be only one study so far in which all three techniques have been applied to the elucidation of the species (As-peptides in the roots of black-eyed Susan vine, Thunbergia alata). Overall, 129 articles were included, and the titles were provided.

Classifying reviews is something of a challenge, but four devoted to microextraction procedures may be readily identified. Rutkowska et al.¹⁰ cited 84 references (with titles) in an overview of microextraction, both liquid- and solid-phase, of the organic compounds of Hg and Sn (mostly alkyltin compounds). The introduction is a well-written, nicely illustrated account of the various procedures identified as part of the microextraction family. The three main microextraction subgroups include: single-drop, DLLME, and hollow-fibre liquid-phase. The single-drop group is further subdivided into headspace, direct-immersion, three-phase, liquid–liquid–liquid, and continuous flow. This 26-reference introduction is followed by a table of applications to a variety of samples that summarizes the contents of about 30 research articles. The next section is entitled ‘Analytical Procedures Using Microextraction Techniques for the Stationary Phase,’ which appears to be about SPME, though of course, its use with both liquid and gas chromatography is discussed. Zaitsev and Zui,¹¹ confining themselves to the solid-phase, entitle their review “Preconcentration by SPME” and cite 157 references, though the titles of these are not given. Elemental speciation is really only featured as such in a short section of 8 not very recent articles, entitled ‘coupling of SPME with atomic spectrometry.’ The bulk of the review is devoted to the chemistry of the extractant materials and the devices in and on which they can be deployed. In surveying applications of microextraction to the analysis of foods Vinas et al.¹² confine themselves to DLLME, but still come up with 226 references (with titles). However the promise of a considerable coverage of elemental speciation analysis suggested by the abstract, is not fulfilled: only three articles deal with inorganic speciation of As or Se. Strategies for coupling SPME with MS have been reviewed by Deng et al.¹³ The 82 references (with titles) cover five types of MS: electron-impact, ICP, laser-desorption/ionization, atmospheric-pressure ionization, and ambient. The ICP section is short: five articles describing thermal-desorption interfacing are featured together with one article describing a solvent–desorption interface. The speciation capability would have been embodied in the selectivity of the extraction stage and any prior derivatization. A considerable portion (32 articles) of the review is devoted to the emerging area of ambient ionization techniques such as electrospray, desorption electrospray, desorption corona beam, and DART (direct analysis in real time). The reviewers concluded that SPME coupled with ambient MS will be a hot research topic for the next several years, combining isolation, enrichment, and analysis into one step under ambient, open-air conditions with minimal sample pre-treatment and without chromatographic separation. The researchers considered that the combination has the potential for highly sensitive and specific detection of trace analytes in complex matrices. It is to be hoped that such a technique would find applications to elemental speciation analysis, for which SPME procedures already have been developed for elements such as As, Hg, Pb, Se and Sn, though there is clearly some way to go.

A sample pre-treatment procedure with the potential for applications in elemental speciation analysis is cloud point extraction (CPE), which is the subject of a survey article,¹⁴ in which the writers summarized mainly the results of trace metal preconcentrations from a variety of samples. The relevant literature since 2006 was reviewed (120 references, no titles). The short section on the mechanisms of CPE is not easy to follow, but citations are provided to relevant primary sources. Although there is a section entitled ‘speciation,’ with subsections devoted to As, Cr, Fe, Hg, Mn, Sb, Se, Sn, and Tl, it is difficult to see the extent to which CPE methods have been applied. For example, the section on As contains a summary of just one paper, devoted to the determination of As^V, but a table entitled ‘cloud point preconcentration of representative elements’ contains an entry for As^III in which two articles are mentioned. There is very little in the way of critical evaluation of the methods described, though the authors emphasize the ‘green’ aspects of the methodology, which they describe as ‘sustainable.’

Three reviews of separations by CZE have appeared in this period. The first¹⁵ covers the broad areas of environmental chemistry, toxicology, and the biomedical sciences (45 references with titles). Some emphasis was placed on As, Cr, Hg, and Se speciation and the reviewers included a useful detailed discussion of species stability for each of these elements, pointing out that chemical reactions between species, interaction with the container material, microbial activity, temperature, pH, and light activation can all cause changes in speciation. The researchers summarized the best storage conditions of pH, temperature and lighting in a table. The reviewers concluded that CE has a future in the speciation field, not just for the determination of trace species (1 μg L⁻¹ concentrations are possible with ICP-MS detection) but also as a technique for studying biological and physicochemical interactions in the environment. The researchers also considered preconcentration before the CE separation to be important and discuss several strategies for sample stacking. The second review¹⁶ is confined to the elemental speciation analysis of environmental samples and 65 relevant articles (with titles) are cited. Methods for slightly broader range of elements (As, Cr, Fe, Hg, Pb and Sn) are highlighted. The review was not confined to those methods based on atomic spectrometric detection. Although the reviewers considered that the coupling of CE with ICP-MS has been successful, they pointed out limitations imposed by the small numbers of CRMs and stressed the importance of interlaboratory tests, comments that might be applied to all speciation analyses. The third review is confined to applications of CE for studying the speciation of actinides with an emphasis on CE coupled with ICP-MS. As well as studies of the species encountered in the environment, the review (47 references with titles) also deals with the determination of binding constants with a number of ligands, including humic acid. Measurements involving the latter ligand are discussed in some detail for both natural humic acids and model systems. The reviewers concluded that whatever the future holds for nuclear power in the twenty-first century, analytical chemistry will be intimately involved in a range of topics related to protecting the public and the environment.

Grotti et al.¹⁷ have reviewed speciation analysis by small-bore HPLC coupled with ICP-MS (77 references with titles). Six classes of column are identified based on internal diameters that range from 25 μm to 5 mm. The introduction sets out what advantages might accrue from the use of decreased dimension LC separations. There is also a comprehensive discussion of the various interface devices (nebulizers, spray chambers and desolvators), but the reviewers did not draw any conclusions as to what the best approach might be. It would seem that direct injection devices offer some benefits in terms of limiting the extra-column peak broadening. As might be expected, there is considerable focus on As and Se, though Cr, Hg and Pb also get a mention. The topic of column recovery is mentioned but the reason why some analyte molecules are permanently retained in the column and others are not is not explained, and there is no discussion of compound-dependent responses, though there is a short section entitled ‘effect of the mobile phase on the ICP-MS signal,’ which is a shame, as the reviewers have closely examined many publications.

In a review of the introduction of gaseous samples, Meng et al.¹⁸ examined the current status and future trends of microplasma-based detectors for GC. Apparently MIP detectors were only included if they had dimensions less than 1 mm and so only three of the 79 references (with titles) describe work with these kinds of devices. Almost all the applications of the other devices (capacitively coupled, glow-discharge, and micro-hollow-cathode) discussed, were concerned with the determination of non-metallic compounds. Disappointingly, the important topic of compound-dependent responses was not discussed. Only one of the cited articles included the word ‘speciation’ in the title, although clearly these approaches would be applicable to such studies.

Grafe et al.¹⁹ have reviewed the application of XAS to the speciation of metals and metalloids in environmental samples. The review included a tutorial introduction to the theory and practice of the techniques, in which it was pointed out that XAS is the only speciation tool from which quantitative speciation data can be obtained for an environmental sample regardless of its physical state and crystallinity, provided sample integrity can be maintained from sampling through to analysis. This may not be so difficult for minerals, but is problematic for samples that contain biological material. In addition XAS provides information about speciation at environmentally relevant solid–liquid interfaces as well as in a variety of matrices including plant tissues, the rhizosphere, soils, sediments, ores, and mineral process tailings. Data processing methods, such as abstract factor analysis and linear combination fitting, are discussed in detail. The need for spectra of reference materials, collected at the same beam line and the same time as the samples are processed is emphasized. The reviewers also discussed the advances in optics, detectors, and data processing that have led to the development of XRF microprobes, capable of focusing X-rays to micro- and nanometer spot sizes, which have been adapted for μ-XANES imaging, a new combinatorial approach of XRF spectrometry and XANES spectroscopy at the micrometer scale capable of imaging of chemical states. The review is based on 145 references (with titles) and is highly recommended.

The use of tandem mass spectrometry for metallomic studies has been reviewed.²⁰ It appears from the entries in the summary tables that much of the work selected for inclusion in the review features analyses by ICP-MS as well as by MS/MS. The reviewers devote space in the text to sample preparation topics and to the use of multiple chromatographic separations (GE is included in this section). There is also a section devoted to metal tagging and labelling techniques. A separate table is devoted to ‘ICP-MS selected applications in the discussed tandem MS reports’ that prominently features the determination of As and Se compounds. The review, which is based on 106 articles (but no titles are provided), would have benefited from some desk editing to enhance readability.

Procedures involving HPLC-ICP-MS also feature prominently in a critical overview of Cr speciation analysis.²¹ The review covers material in 91 articles (no titles are provided). None-the-less, this article is recommended reading. The introduction is a good survey of the environmental chemistry of Cr^III and Cr^VI, and the section on ICP-MS detection contains a thorough discussion of how to deal with the isobaric interferences on the commonly determined isotopes at m/z 52 and 53 by the use of high resolution spectrometers or reactions in collision cells, including high energy collisions with helium. There is a comprehensive section on the preparation and the use of isotopically enriched Cr^III and Cr^VI solutions for spiking experiments, which is followed by an equally comprehensive discussion of the potential of speciated ID-ICP-MS in the analysis of a variety of materials. Previous reports of the determination of Cr^VI in plants, soils, bread and tea infusions are roundly criticized for erroneously concluding (based on supposed selective extraction) that the species extracted was Cr^VI. It was pointed out that results not supported by speciation analysis may be wrong, as a subsequent repeat of the tea analysis has shown. With regard to Cr^III, it was concluded that due to the widespread use of Cr^III in nutritional supplements, there is an increasing need for the development of speciation procedures for selective determination of Cr^III. It was considered that HPLC-ICP-MS in combination with tandem organic MS provides a potential tool for speciation in nutritional supplements, foodstuffs and environmental samples.

The particular difficulties of sample preparation for As speciation in terrestrial plants, alluded to earlier, has been the subject of a separate review.²² A breakdown of the literature from 2000 to 2012 is provided based on searching the Web of Science data-base, from which just over 37% of the articles are concerned with rice. Nearly 54% of articles (almost certainly a different set) are concerned with roots and only 18% with grain. The review contains 103 references, but the titles are not given. The reviewers also highlighted the difficulty of preserving phytochelatin species (in which the As is bound to sulfur), and pointed out that despite numerous studies, there is still much to understand about this detoxification mechanism. The introduction provides a useful overview of the uptake and metabolism of As compounds by plants and there is discussion of studies of As species stability following the extraction step, from which the reviewers concluded that it is important that the analysis be carried out as soon as possible after extraction. The reviewers predict that HPLC-ICP-MS will be the instrumentation of choice for some time to come, but the role of ESI-MS is also highlighted.

The larger field of speciation and detection of As in aqueous samples has also been reviewed.²³ Methods involving optical atomic spectrometry were not included because they were considered to be sufficiently mature and have been reviewed recently with other reviews available. Despite the omission of many of the techniques regularly featured in the pages of JAAS, this review is highly recommended: it is well written and genuinely critical. A summary table of previous reviews, dating back to 1989, contains 80 entries covering a variety of sample materials other than waters, and a variety of topics in addition to chemical measurement. The analytical methods reviewed are taken from the literature dating from 2005 to 2013. The spectrometry topics included are colorimetry (visible absorbable/reflectance), luminescence (mostly chemiluminescence), LIBS, surface enhanced Raman spectrometry, TXRF, attenuated total reflectance-FTIR, and surface plasmon spectrometry. The reviewers also discussed electrochemical methods (mostly stripping voltammetries). There is a lengthy section devoted to ICP-MS and a short section on neutron activation analysis. Overall, 203 article are cited (no titles are provided). There is plenty of critical commentary, such as electrodes are ‘finicky,’ detection limits of some procedures are ‘barely adequate,’ or of ‘marginal utility,’ some researchers' conclusions and claims are described, unflatteringly, as ‘remarkable’ or ‘implausible.’ One apparatus is described as ‘not well designed.’ The reviewers threw their considerable weight behind the possibilities for gas-phase chemiluminescence spectrometry for the speciation and detection of As in aqueous samples.

The speciation analysis of Se-containing compounds has been reviewed. Latorre et al.²⁴ surveyed SPE procedures for the speciation and preconcentration of iSe in water samples. The reviewers classified the procedures for the two analytes, Se^IV and Se^VI on the basis of the retention and elution strategy for the Se species as anions and the retention of one or other of the species as chelates. Some 79 references (no titles) are cited, of which the first 30 or so are in support of the general introduction that covers Se toxicity and other preconcentration procedures, such as precipitation/redissolution and DLLME. As many procedures are based on the retention of the Se anions, much of the review is devoted to a discussion of the nature of the solid phase materials. The writers do offer critical evaluations, pointing out the drawbacks of various procedures. It was concluded that the best strategies are those involving (a) the retention of both iSe species on the same sorbent followed by the separate elution of Se^IV and Se^VI or (b) preferential chelate formation with Se^IV. The reviewers considered that these approaches are rapid, avoid excessive sample manipulation, do not require prior reduction or oxidation, allowing the determination of both species (at concentrations down to single-digit ng L⁻¹ in real water samples) in only a few steps. Furthermore, the chelation procedure increased selectivity due to the specificity of the ligand used. The reviewers pointed out that the ease of regeneration and the number of times a material can be reused, two important method characteristics, were often not provided in the account of the performance. Despite the large body of work, it was considered that further studies are needed, particularly directed towards the development of materials based on functionalized nano-sized carbon and on molecularly imprinted polymers. The equally active field of the determination of selenoamino acids in biological samples has been reviewed (82 references, but no titles) with particular emphasis on chiral separations.²⁵ It was pointed out that chiral speciation, including that of selenoamino acids is needed not only for a more efficient evaluation of the bio-function of Se supplements, but also for a better understanding of the bioavailability of Se in living organisms, in which the majority of amino acids incorporated in proteins are the L-forms; the corresponding D-forms that coexist in supplements as racemates or even as impurities may be toxic. The reviewers divided the field into two major strategies: indirect and direct approaches. The former is based on derivatization with an optically pure chiral compound to form diastereoisomers that are separated by conventional chromatographic or electrophoresis methods. The latter is based on using a chiral column in chromatography or a chiral additive in electrophoresis to directly separate the enantiomers. For each of these two strategies, methods involving HPLC, GC, or CE were discussed. In addition, different detectors, sample preparation, and applications to (a) dietary supplements and (b) biological fluids were discussed. It was concluded that further work is needed and that, although HPLC-ICP-MS procedures have been the most widely used, there is still a need for the development of columns and better sample preparation procedures. It was speculated that separations by capillary electrochromatography may have the potential for application in this area.

2 Sample preparation

Very few articles describing advances in sample preparation have been published during the current review period. Sample preparation topics have been covered in several of the review articles and books discussed in the previous section.

An investigation of As species in plants stored under different conditions for 12 months has been reported by Amaral et al.²⁶ Plant samples (Brachiaria brizantha StapE cv. Marandu) grown in As-containing soil were harvested and divided into four portions. The first portion was freeze-dried, milled in the presence of liquid N₂ and stored at room temperature; the second was lyophilized and stored at 4 °C; and the remaining portions were kept at two storage temperatures (−18 and −80 °C) without any prior procedure. Aliquots of samples stored under different conditions were extracted every 2 months for a 12 month period. The total As content was 230 ± 0.27 mg kg⁻¹ and the As species were determined by HPLC-ICP-MS. The extraction was more efficient in sample aliquots that were lyophilized and ground (87–90%) compared to those only stored under different temperatures (53–66%). Furthermore, DMA was extracted only for lyophilized and ground samples. Storage time under the two tested temperatures had no significant effect. It was concluded that lyophilization and cryogenic grinding strategies were the most suitable sample pre-treatments for As speciation in plant tissue. Stability tests²⁷ for As species found in commercially available edible alga Hijiki (Hizikia fusiformis), have been performed for both the dry sample and its water extracts. Samples were stored in amber glass and polystyrene containers at +18 and +4 °C in the dark. Extractions were carried out with deionized water and MAE, at a temperature of 90 °C and three extraction steps of 5 min each. Arsenic speciation analysis was performed by anion exchange HPLC-photo-oxidation-HG-AFS. The results obtained for the dry alga showed that the As species present (As^V, DMA and the arsenosugars glycerol, phosphate, sulfonate and sulfate) were stable for at least 12 months when the sample was stored in polystyrene containers at +18 °C in the dark. The water extracts were best stored in polystyrene containers at a temperature of +4 °C, for a maximum storage time of seven days.

3 Instrumental techniques and developments

3.1 Development in species separation

Several reports of analyte separation by CZE have appeared in the current review period. For the determination of Br species in bread, a procedure involving field-amplified sample injection followed by separation by microchip electrophoresis and detection by ICP-MS has been developed.²⁸ The researchers achieved separation of bromide and bromate within 35 s in a 1.5 cm separation channel under an electric field of approximately 260 V cm⁻¹ in a 65 mM sodium acetate buffer at pH 8.0. The sensitivities were improved by a factor of 13 for bromide and 12 for bromate by preparing the standards or samples in 3 mM sodium acetate instead of the background electrolyte. The LOD values were 0.2 μg L⁻¹ for both species, corresponding to about 10 μg kg⁻¹ in the bread. This was not low enough to detect bromide, but bromate was quantified in all but one of the samples. Recoveries of spikes (corresponding to 2.5 mg kg⁻¹) in six bread samples were between 97 and 105% with no significant difference between the results by the new method and those obtained by a method involving HPLC-ICP-MS. The viability of a new flow-focusing nebulizer interface was evaluated by application to Cr speciation.²⁹ In addition to the novelty of the use of this kind of low-flow nebulizer, the flow of liquid in the CE system was driven by the application of argon gas pressure to the carrier buffer reservoir rather than by electroosmosis. The sample was injected via a 6-port rotary valve operated in timed-injection mode so that a sharp rear sample zone boundary was obtained. Even so, the resolution between the two peaks for the Cr species was poor. Better separation was achieved with shorter columns and higher voltages, with 1 m and 15 kV being the optimum conditions. When the applied voltage was turned off, one rather broad peak (half width approximately 20 s) corresponding to total Cr was obtained. The system was coupled with both ICP-OES and ICP-MS giving LOD values of between 0.03 and 0.1 μg L⁻¹ for the latter detector, the sensitivity of which was about three orders of magnitude greater that the emission instrument used. The CE run time was about 2 min. A CZE procedure has been developed for the speciation of Cr in gelatin capsules.³⁰ The interface is not new having been described in an earlier report. The Cr species in the capsules were extracted into a mixture of the running buffer (20 mM NaH₂PO₄ plus 5 mM Na₂B₄O₇ at pH 6.0) and 5 mM derivatizing agent 1,2-cyclohexanediaminetetraacetic acid and then separated at 15 kV on a fused-silica capillary column (30 cm × 75 μm). The method was validated by spike recoveries and accurate analysis of a cabbage CRM (GBW10014). The LOD values were 12 and 9 ng g⁻¹ for Cr^III and Cr^VI, respectively and the latter was not low enough to detect Cr^VI in the real samples. Thirty capsules were analyzed and these contained between 0.24 and 72 μg g⁻¹ of Cr^III only. A similar problem of inability to detect some species in real samples was encountered by Chen et al.³¹ in the speciation analysis of Pb in marine animals. Again the interface was not new, having been described in an earlier article. Much of the novelty of the work concerns the development and evaluation of a two-step MAE procedure to solubilise Pb^II, trimethyllead and triethyllead from oyster and clam tissue. The method was appropriately validated by the analysis of a shrimp CRM (GBW 10050) and spike recoveries from this and some real oyster and clam samples. Solution LOD values of between 0.012 and 0.084 ng mL⁻¹ were reported, but not enough information was given to calculate the corresponding values for the dried samples. They detected Pb^II in clam, and Pb^II and trimethyllead in oyster. In a study of Al speciation by both CZE-ESI-MS and CZE-ICP-MS, Nakamoto and Tanaka³² did report, in Japanese, on the design of the interfaces. Their instrumentation was applied to the speciation of aluminium fluoride complexes, which they found to be partly dissociated during the CE separation process.

A new analytical interface combining CE coupled on-line with ETAAS has been developed and used for the determination of Se species.³³ A separation buffer solution consisting of 20 mmol L⁻¹ Na₂HPO₄, 0.2 mmol L⁻¹ NaH₂PO₄, 0.2 mmol L⁻¹ CTAB, and 10% CH₃OH (pH 10.0), and a separation voltage of 21 kV were selected as optimal conditions for achieving Se species separation in less than 10 minutes. The LOD values were determined to be 1.7, 2.2, 0.89, and 0.97 μg L⁻¹, for SeMet, SeCys₂, Se^IV, and Se^VI, respectively. The recoveries ranged from 95.6% to 104%. The applicability of the CE-ETAAS technique was demonstrated by analyzing enzymatic extracts of a reference ginger containing native Se at trace levels (μg g⁻¹) and ginger plants sprayed with a mixture containing Se^IV and Se^VI. Results indicated the presence of SeMet, SeCys₂, Se^IV, and Se^VI in the ginger, while SeCys₂ was not detected.

A centrifugal microfluidic platform with integrated monolithic capillary columns has been developed and applied for Cr speciation.³⁴ This is a SPE procedure in which, by adjusting the pH to between 2 and 4, Cr^III was not retained but Cr^VI was. The Cr^VI was eluted with 0.1 M ammonia solution. The device consisted of eight columns (530 μm id × 2 cm long) positioned radially in a device mounted on a rotor spun at 1600 rpm. The columns contained the product of the polymerization of N-(β-aminoethyl)-γ-aminopropyltriethoxysilane and tetraethoxysilane. Following elution, the Cr was determined by FI-ICP-MS. Total Cr was determined directly, and hence Cr^III was determined by difference. The procedure was validated by the accurate analysis of an environmental water CRM (GBW08607) certified for total Cr, a spiked bottled water, and a sewage water sample. The LOD values were 1.2 and 0.9 μg L⁻¹ for Cr^VI and Cr^III, respectively, which were not low enough to detect either species in the drinking water (though both were measured in the sewage water), and allowed only Cr^III to be detected in the CRM. The parallel processing of the extraction stage increased the throughput so that eight solutions could be processed in 10 min. The researchers suggested, possibly somewhat optimistically, that the device could be used for field sampling and pre-treatment.

An SPE procedure for the speciation of Hg in surface waters based on selective retention of Hg^II on polymeric ion-imprinted nanoparticles has been developed.³⁵ Much of the report is devoted to descriptions of the synthesis and characterization of the extractant which was prepared by co-polymerization of methacrylic acid (functional monomer), trimethylolpropane trimethacrylate (cross-linking agent), 2,2′-azo-bis-isobutyronitrile (initiator) and Hg^II complexes with 1-pyrrolidinedithiocarboxylic acid (chelating agent). The analytical utility was demonstrated by the analysis of a CRM (BCR 505, estuarine water), but there it is not clear whether the value found of 0.62 nmol kg⁻¹ is significantly different from the certificate value of 0.69 nmol kg⁻¹. Total Hg was determined after any OMCs were degraded by a procedure described as “MW digested according to EPA” (with no citation). The difference between the total Hg value and that for iHg was interpreted as due to MeHg. Recoveries of 92–96% were obtained for MeHg spiked into both the CRM and various waters (sea, river and mineral), but no details were given. The AAS LOQ for both species was 0.02 μg L⁻¹. Details were not provided concerning the operation of the spectrometer, whose QT-atomizer was operated at room temperature for iHg and in an air/acetylene flame for MeHg.

Also of interest is an evaluation of hydrodynamic chromatography for sizing and quantifying colloids in environmental media.³⁶ It was shown that for two commercial columns, flow rate and eluent composition had little influence on the size resolution and that even weak agglomerates could be measured. The researchers measured Ag nanoparticles in synthetic surface waters and titania and zironia nanoparticles in sunscreens. When ICP-MS detection was combined with UV-visible absorption and molecular fluorescence spectrometries, it was possible to distinguish between organic molecules and inorganic colloids. Wacker and Suebert³⁷ determined stability constants of the complexes of the trivalent ions of Cr, Fe, Ga, In and Sc with N,N,N′,N′-ethylenediaminetetraacetic acid, with a competitive procedure in which species were separated by anion or cation exchange chromatography and quantified by ICP-OES. As the results obtained were the ratios of the stability constants of the two elements, the values were independent of the pH and ionic strength. It was found that Fe species were not stable towards sunlight and that storage in the dark was necessary.

A few procedures for simultaneous speciation analysis of more than one element have appeared. Wu and Pichler³⁸ determined the inorganic anions of As, Sb and Se by SF-ICP-MS after separation on a Hamilton PRX-X100 anion-exchange HPLC column. The mobile phase was a gradient from 5 mM to 30 mM of EDTA (pH 4.7 with formic acid) and 3% methanol. The LOD values ranged from 0.02 μg L⁻¹ (As^III, Sb^V) to 0.4 μg L⁻¹ (Se^VI). The method was validated by the analysis of spiked hot spring water samples and of two CRM (NIST 1643e, trace elements in water, and High Purity Standards CRM-SW, simulated seawater in 2% nitric acid). The analysis time was 11 min at a flow rate of 1.5 mL min⁻¹, though no information was provided about the time needed to re-equilibrate the column. The transformations of inorganic As^V and Hg^II ions by ‘metal-resistant’ oral bacteria have been studied by imaging TOF-SIMS.³⁹ Bacterial strains, isolated from the oral cavities of healthy volunteers, were grown in the presence of Hg^II (1–10 mg L⁻¹) or As^V at concentrations of (0.1–1.0%). Imaging and depth profiling of biofilms of air-dried bacteria on aluminium foil surfaces allowed subcellular nm scale depth resolution with μm lateral resolution. The researchers found that MeHg was formed, preferentially in the periplasmic space between the inner and outer membranes that constitute the cell wall of Gram negative bacteria, and that As^V was reduced to As^III inside the cells, close to the cell membrane. The results were discussed in the light of current concepts in bacterial resistance to metals and antibiotics. The determination of Pb and Sn species by GC-AFS is described in the following section and of Cu and Se by X-ray methods in Section 3.3.

3.2 Developments in instrumentation

A number of speciation procedures based on developments in detection following GC separation have been described. Zeng et al.⁴⁰ employed AFS detection for the speciation analysis of Se and the simultaneous determination of Pb and Sn species. The spectrometer was a commercial two-channel AFS instrument (SK-2003A, Beijing Jinsuokun Technology Developing Co. Ltd.) for which a new quartz hydrogen-diffusion flame atomizer had been constructed. The burner supported three concentric gas flows: the argon mobile phase formed the central channel of the burner, the argon/hydrogen fuel mixture was delivered in the first annular region and an argon sheath flow formed the second annular region. The whole device was heated by electrical heating tape. Detailed optimization experiments for the burner design were reported. The exit from the GC oven was positioned sufficiently close to the spectrometer that no heated transfer line was needed. The system was applied to the separation and detection of dimethylselenide and dimethyldiselenide (burner temperature 120 °C) and, in a separate experiment, to the simultaneous detection of tetraethyllead and tetramethyltin (burner temperature 260 °C). The Se species were separated in 30 s; the retention time for the tetramethyltin was 20 s and that for the tetraethyllead was 90 s. The LODs ranged from 4 pg for dimethyldiselenide, to 4 ng for tetraethyllead, based on measurement of standards dissolved in methanol. For a 1 μL sample injection volume, the corresponding concentrations were 4 μg L⁻¹ and 4 mg L⁻¹.

The novel use of different detectors for GC separations has been reported. A low-temperature dielectric-barrier discharge (DBD) device for the detection of carbon-containing compounds separated by GC has been constructed.⁴¹ The DBD device was housed in a box heated to 300 °C and the C atomic emission line at 193 nm was monitored by a miniature CCD spectrometer. The device required a separate flow of discharge gas (argon was selected). Details of the optimization of the performance were provided in ESI. Although compound-dependent responses were not discussed, and no values for the sensitivities of the calibrations for the various compounds were given, the range of LOD values of 0.12–0.28 ng on a sample volume of 1 μL, suggests that the responses were compound-dependent. The instrumentation was used for the determination of four compounds in wine and the results compared to those obtained with a flame ionization detector. The DBD device could not detect ethyl acetate or methanol, whereas the flame ionization detector could not detect formaldehyde. The researchers point out the potential further applications when coupled with a multichannel spectrometer capable of monitoring both atomic and molecular emissions. The simultaneous use of EI and ICP-MS has been implemented for GC detection.⁴² The GC eluent was split 75 : 25 by a ‘simple’ Y-device fabricated from silica capillary tubing (custom designed and made by Agilent Technologies). The instrumentation performance was demonstrated by the determination of MeHg, DBT and TBT, and SeMet in DOLT-4 (dogfish liver, NRCC), PACS-2 (marine sediment), and SELM-1 (Se-enriched yeast) CRMs, respectively. The analyses were all performed by derivatization (propylation for Hg, ethylation for Sn, and reaction with methylchloroformate for Se). Accurate results were obtained, though for Se, the researchers reported that ‘SeMet in SELM-1 was just qualitatively determined.’ It was pointed out that the specificity of both fragmentation and isotopic patterns plays a pivotal role in the characterization of the organometals and that with this instrument, species can be both identified by EI-MS and quantified using ICP-MS.

Preisler and co-workers⁴³ have further developed their diode laser thermal vaporization technique, this time as a means of coupling TLC with ICP-MS. The introduction to the article includes a survey of previous reports of coupling TLC with ICP detection as well as citations to their own work in which species were vaporized by this kind of laser from a paper surface that had been overprinted with black ink from a regular ink-jet printer to absorb the laser light. The overprinted cellulose TLC plate was cut into 3 mm wide strips containing the entire separation lane, which were inserted into a 17 cm long × 3.8 mm i.d tubular vaporization cell. The center of the overprinted area was scanned by a 1.2 W 808 nm continuous-wave diode laser (RLDH808-1200-5, Roithner LaserTechnik, Austria) mounted on a syringe pump that served as the movable stage. No details of this arrangement were given. The procedure was applied to detection of four cobalamins (hydroxo-, adenosyl-, cyano-, and methyl-). It was found that the plastic-backed plates gave a background signal at m/z 59, which was interpreted as Co, that was about 300× that from the aluminum-backed plates, which were therefore used in the final version of the method.

3.3 Chemical vapour generation

For the speciation of Hg, Yu et al.⁴⁴ developed a miniature long-optical path AA spectrometer with dielectric barrier discharge as atomizer for Hg^II and MeHg. The original article is in Chinese, but contains a lengthy abstract in English. The Hg⁰ and MeHg vapor were generated in a sequential injection system, and after passage through a gas–liquid separator, a glass wool moisture-removal microcolumn and the DBD atomizer, were delivered into the long optical-path cell for quantitative analysis by AAS. The signal for Hg⁰ was obtained when the atomizer was off, while the signal for both Hg⁰ and MeHg was measured when the atomizer was on. The LOD was 0.4 μg L⁻¹ for both species for a sample volume of 1 mL. The reliability of the system was demonstrated by the analysis of CRMs and real samples, but no details were given in the abstract.

3.4 Solid state speciation

The role of solid-state speciation in illuminating changes induced by sample preparation for determination by liquid-phase techniques was highlighted in a review of As speciation discussed earlier.⁹ Weekley et al.⁴⁵ studied Se and Cu speciation and distribution in the kidneys of selenite-supplemented rats by XAS and XRF-microscopy. Selenium K-edge X-ray absorption spectra of the rat tissues were recorded at the Stanford Synchrotron Radiation Lightsource, CA on beamline 9-3 and XRF elemental distribution maps were collected on beamline 2-ID-E at the Advanced Photon Source Argonne National Laboratory, USA and on the XRF microprobe beamline at the Australian Synchrotron, Australia. Somewhat contradictory results were obtained. Analysis by XRF-microscopy revealed a strong correlation between the distribution of Se and the distribution of Cu in the kidney, a phenomenon that had previously been observed by the researchers in cell culture. However, the XAS results indicated that most of the Se in the kidney was in the form of Se–Se species, rather than Cu-bound, and that most of the Cu was bound to S and N, (which was interpreted as amino acid residues in proteins). The article features an in-depth discussion of these findings, together with the role of SOD-1 (containing Cu and Zn), whose expression did not change in response to the high Se diet.

In a study of the subcellular distribution, cytotoxicity and speciation of environmental Mn compounds, Carmona et al.⁴⁶ combined micro-SR-XRF for imaging with micro-XANES to determine the Mn oxidation state at the single cell level. The experiments were conducted at European Synchrotron Radiation Facility ID21 beamline, where the scanning X-ray microscope end-station of ID21 allows the complementary use of micro-SR-XRF and micro-XANES in the 2–9 keV energy range with spatial resolution at the μm level. A vibration-free cryo-stage, cooled by liquid nitrogen, allowed the analysis and preservation of frozen hydrated samples, which is essential for biological applications and, in particular for metal speciation in the near native state. For many compounds, Mn^II was observed located mainly in the Golgi apparatus of the cell, probably for detoxification purposes via exocytosis. It was found that organic compounds including methylcyclopentadienyl manganese tricarbonyl and manganese dithiocarbamate, (a foliate fungicide sold as maneb) were degraded and behaved similar to the soluble Mn ^II inorganic compounds. To investigate the nature of the Mn enrichments in sediments from the Holocene Thermal Maximum (approx. 8000–4000 cal. year BP) at a site in the northern Gotland Basin, Lenz et al.⁴⁷ used high resolution synchrotron EXAFS and XANES combined with LA-ICP-MS and micro-XRF. Measurements were carried out at the DUBBLE beamline at the European Synchrotron Radiation Facility (BM26A). The researchers conclude that the Mn-containing materials can by identified as (a) rhodochrosite, (b) a Mn aluminosilicate, and (c) Mn co-precipitated with iron sulfides.

4 CRMs and metrology

Details of two CRMs, both certified for MeHg content, have been published during the period covered by this review. The first of these, Dourada-1, is a sterilised, freeze dried Dourada (Brachyplatystoma Flavicans) fish, with all of the fish apart from the head and tail being homogenised for analysis.⁴⁸ The total Hg content in the homogenised fish material was determined by IDMS, with the ²⁰²Hg spike added prior to a hot nitric acid digestion procedure. For the MeHg content the method is less clear as it is only partially reported. However, it can be surmised from the method used to assess sample homogeneity that the candidate CRM was subjected to a 6 mol L⁻¹ HCl leach procedure, followed by retention of the Hg in the leachate on an anion exchange column (Dowex 1 × 8 cm, 100–200 mesh), UV oxidation of the MeHg in the eluent to Hg^II and measurement by FI-CV-AAS. The procedure was validated by the analysis of two NRCC CRMs, namely DORM-2 and DOLT-2. The certified values in the Dourada-1 CRM are 0.271 ± 0.059 and 0.245 ± 0.053 μg g⁻¹ for total Hg and MeHg respectively. The extraction method is assumed to be selective for MeHg although no evidence is given for this. Therefore, the MeHg values obtained for each CRM should really be designated “organic mercury” as no effort was made to determine the individual Hg species in the sample. Expanded uncertainties (k = 2) are reported, and take into account sample stability and homogeneity as well as the uncertainty associated with the analytical result. The second CRM to be reported this year, IAEA 452, was prepared from scallop (Pecten maximus) adductor muscle and gonads.⁴⁹ The certified MeHg mass fraction in IAEA 452 CRM, of 0.11 ± 0.02 μmol kg⁻¹ was determined by an interlaboratory comparison certification process and further details of this can be found in section 8.1.4 of the certification report available at http://www-pub.iaea.org/MTCD/Publications/PDF/IAEA-AQ-23_web.pdf. This CRM is also certified for Cd, Cu, Hg, Pb and Zn. As part of the certification process reference procedures were developed, by the IAEA laboratory in Monaco, for the total metal mass fractions by IDMS and the MeHg mass fraction, by ssIDMS using a ²⁰¹Hg enriched MeHg spike material. A MAE procedure, 80 W for 30 min, and 15% HCL was used to extract the organomercury in the scallop homogenate, to which the MeHg spike had been previously added. After centrifugation, the supernatant was passed through an anion exchange column, packed with Bio-Rad AG1X-8 500 mesh resin, to remove any co-extracted iHg and the eluent collected. Finally, the ²⁰⁰Hg : ²⁰¹Hg isotope amount ratio was measured in the column eluent by a single collector SF-ICP-MS instrument. The mass bias correction factor was determined by the measurement of the ²⁰⁰Hg : ²⁰¹Hg isotope amount ratio using the bracketing method. The reference value for MeHg of 0.109 ± 0.0049 μmol kg⁻¹, obtained by this procedure, was in agreement with the certified MeHg mass fraction of 0.11 ± 0.02 μmol kg⁻¹, established by the interlaboratory comparison exercise.

5 Elemental speciation analysis

5.1 Antimony

There have been few reports on developments in Sb speciation during this review period. The speciation, methodology and influence of temperature, time and anticoagulants on the distribution of Sb^IIIand Sb^Vin human erythrocytes (plasma and cytoplasm) have been reported.⁵⁰ The method involved purification of the sample by salting out the proteins followed by cleaning the supernatant by elution through a C₁₈ SPE cartridge using 3 mL of EDTA. Finally, a chromatographic separation, using anion exchange PRP X-100 (100 × 4.1 mm, 10 μm) and EDTA 20 mmol L⁻¹ as mobile phase, was used prior to detection by HG-AFS. The method was optimised by experimental design with a recovery of 90% for Sb^V and 55–75% for Sb^III. The results showed that both Sb^V and Sb^III were capable of entering the red blood cells but after 90 minutes excreted to below 30% within the cell. An increase in the culture temperature increased the capacity of Sb^V and Sb^III to penetrate the membrane barrier and reach the cytoplasm. In order to preserve the original distribution of Sb in blood, heparin was the best anticoagulant for sample preservation.

An analytical method for the determination of total Sb, Sb^Vand Sb^IIIin road dust and airborne particulate matter from Valparaiso, Chile was developed using HPLC-HG-AFS.⁵¹ For total Sb determination, samples were digested with 6 mL of HNO₃ and 2 mL of HBF₄ in closed microwave digestion system at 200 °C for 30 min. Two Sb species, Sb^V and Sb^III, were extracted using oxalic acid 100 mmol L⁻¹ in 1% w/v ascorbic acid at 70 °C in a closed vessel microwave extraction procedure. The analytical method was applied to airborne particulate matter (<10 μm), and road dust (<37 μm) samples. A CRM, NIST 1648a, was also used. Up to 70% of total Sb extracted from road samples was Sb^III indicating an anthropogenic origin of this element.

5.2 Arsenic

The speciation of As continues to be a lively area for research, with advances being made in many of the fields covered in last year's review.¹ The effect of different As species on the ICP-MS signal when working at a low liquid flow rate has been investigated by Grotti et al.⁵² The influence of the analytical concentration, the matrix, and various sample introduction systems was considered in the study. A significant decrease (up to 65%) in the relative sensitivity of As^III compared to As^V was found, while MMA, DMA and AB gave the same response (within 6%) as As^V, with flow rates in the 20–1000 μL min⁻¹ range. The effect was independent of the analytical concentration over the range 1–100 μg L⁻¹ as As, and was ascribed to processes related to both the sample introduction system and the ion generation and transport. Ion defocussing due to dissimilar kinetic energy of the As ions generated from As^III and As^V were ruled out following measurement of the As⁺ signal as a function of ion lens voltage. Results obtained by various micronebulizer/spray chamber configurations including a PFA-ST micronebuliser (Elemental Scientific) or a HEN (Meinhard) coupled to either a cyclonic or Cinnabar spray chamber, showed that the temperature of the spray chamber was also relevant in determining the relative responses of the As species: heating the spray chamber at 60 °C caused a decrease in relative sensitivity of As^III and DMA compared to As^V, while the As^III-to-As^V signal ratio was improved by cooling at 4 °C. The relative response of As^III and As^V was also significantly influenced by the presence of ammonium phosphate, which mitigated the difference between the species using conventional sample introduction devices. The influence of the chemical species on the ICP-MS signal was also investigated for Hg, Se and Sn species and significant differences in sensitivity when working at a low liquid flow rate were found.

Once again, there have been a number of reports focusing on the extraction and preconcentration of As from aqueous systems. Sol–gel based amine-functionalized SPME fibres (PDMS-weak anion exchanger) have been prepared and used for the direct extraction of DMA, MMA, and As^V from aqueous solutions followed by determination using HPLC-ICP-MS.⁵³ Electrospinning was used for the preparation of sol–gel based SPME fibers and was found to be superior in terms of As extraction, coating homogeneity, accessibility of amine groups on the surface, and preparation time for a single fibre when compared to dip coating. The optimum extraction conditions were determined as pH 5.0, extraction time 30 min, agitation speed 700 rpm and extraction temperature 20 °C. The extraction ability of the coating decreased with the addition of NaCl as a consequence of the competition between anionic As species and Cl ions for active sites on the weak anion exchanger. Vibrational spectroscopy revealed alignment of PDMS chains by elongational force during the electrospinning process. Titanium dioxide nanotubes have been used as a solid phase adsorbent for on-line separation and preconcentration of As^III and As^V prior to determination by ICP-MS.⁵⁴ In the pH range 3.0–6.0 both species were quantitatively retained, whereas within the pH range 6.0–10.0 only As^III was quantitatively adsorbed with As^V passing directly through the column without retention. Under the optimised conditions, the LOD of this method was 0.0019 ng mL⁻¹ for As^III with an enrichment factor of 75, with a RSD of 2.5% for As^III (n = 9, 1.0 ng mL⁻¹). The method was applied to the determination of iAs species in spiked environmental water samples with recoveries in the range of 95.5–102%. In order to verify the accuracy of the method, a CRM water sample was analyzed with satisfactory results. Preconcentration of DMA from water samples using a strong cation exchange disk functionalized with sulfonic groups, before analysis using WDXRF has been reported.⁵⁵ The authors suggested that the adsorption of DMA occurred preferentially on the surface of the SCX disk, regardless of pH, due to interactions with the sulfonic functional groups. However, no other As species, such as As^V, As^III, or MMA, were retained. The SCX-WDXRF method produced a linear calibration curve (r = 0.9996) with a LOD of 0.218 μg L⁻¹ when a 1 L water sample was used for preconcentration. The intensity of the As signal was sensitive to the Pb content retained on the SCX disk owing to the proximity of the As-K-alpha and Pb-L-alpha lines. To compensate for this interference, a correction factor was developed by considering the calibration slope ratio between the X-ray intensity measured at a Bragg angle of 48.781° and the Pb content of the SCX disks. The results of spike tests for As^V, As^III, MMA, and DMA with and without the addition of Pb in synthetic landfill leachate gave recoveries of between 98–105% after the spectral adjustment for the Pb interference. A thiol-modified sand⁵⁶ has been synthesized to rapidly and selectively remove As^III from aqueous matrices in situ without absorbing As^V. The thiol-modified sand was placed in a disposable cartridge and used to separate As^V (37–970 μg L⁻¹) and As^III (LOD to 488 μg L⁻¹) in 23 groundwater samples collected in areas with naturally occurring As. The results in the field were consistent with those obtained later using HPLC-AFS. Furthermore, the approach was applicable to a wide variety of matrices, including groundwater, leachate from contaminated soil and in vitro gastrointestinal extractions. A urine CRM (GBW09115) was also analysed and good agreement was found with the certified values. A series of thiol- and amine-bifunctionalized mesoporous silicas have been synthesized via one-pot co-condensation of tetraethylorthosilicate, 3-mercaptopropyltrimethoxysilane and N-(2-aminoethyl)-3-aminopropyltriethoxysilane.⁵⁷ The mesoporous materials were characterized by XRD, SEM, TEM, nitrogen gas adsorption, IR spectroscopy, TGA and elemental analysis. The inorganic species As^V and As^III were effectively adsorbed by amine and thiol groups on the functionalized silica, respectively, through electrostatic interaction and chelation. Adsorption isotherms and kinetic uptake profiles of As^V and As^III onto these adsorbents were investigated by batch adsorption experiments. In addition, the material was used for As speciation in water using a home-made syringe-based SPE device. Both species were effectively separated and preconcentrated in a single run through a sequential elution strategy, in which 0.1 M HNO₃ was first used to selectively elute As^V, and then 1 M HNO₃ with 0.01 M KIO₃ was used to elute As^III.

Arsenic speciation following absorption onto functionalized carbon nanotubes continues to attract attention. 3-(2-Aminoethylamino) propyltrimethoxysilane has been used with microcolumn SPE-ICP-MS for the determination of As and other elements in environmental waters.⁵⁸ The As^V could be selectively adsorbed onto the packed microcolumn at pH 2.2, whilst As^III was not retained at this pH and passed through. Total iAs was determined after oxidation of As^III with 10.0 μmol L⁻¹ KMnO₄, and the As^III calculated by difference. Once optimised, the LOD for As^V was 15 ng L⁻¹ with an RSD of 7.4%. The method was validated by using four Chinese environmental water based CRMs, rain, river and lake water. In a similar application for drinking and environmental waters, a sequential SPME system consisting of two monolithic capillary columns has been developed for the simultaneous separation and preconcentration of iAs followed by ICP-MS detection.⁵⁹ An N-(beta-aminoethyl)-gamma-aminopropyltriethoxysilane incorporated organic–inorganic hybrid monolithic column was prepared in situ by sol–gel technology in a fused capillary and employed as the extraction medium, utilising the amino active sites on the synthesized monolith which possess a high adsorption selectivity for As^V. Using an on-line design of dual columns and an oxidation coil, As^V was quantitatively extracted by the first column, and As^III in the effluent quantitatively extracted by the second column after oxidization to As^V with a KMnO₄ solution. The retained species were then sequentially eluted by diluted HNO₃ prior to determination by ICP-MS. On-line SPME of 1 mL sample solutions gave an impressive signal enhancement factor of 60 for both As^V and As^III using the system and the RSD for six replicate measurements of 1 μg L⁻¹ As^V and As^III were 3.8% and 3.2%, respectively. The LOD for As^V and As^III were 0.005 μg L⁻¹. The thermoacidophilic iron-oxidizing archaeon Acidianus brierleyi is a single-celled microorganism that could be used for the removal of iAs from wastewater, because it simultaneously oxidizes As^III and Fe^II to As^V and Fe^III in an acidic culture medium, resulting in the immobilization of As^V as FeAsO₄. The successive determination of As^III, As^V, and total As in A. brierleyi and its culture medium via a method based on ICP-OES with a FI pre-treatment system using a mini-column packed with an anion-exchange resin has been reported.⁶⁰ The FI pre-treatment system consisted of a syringe pump, a selection valve and a switching valve, controlled by a PC. Sample at pH 5.0 flowed into the mini-column where the As^V was retained, whereas As^III was not retained and flowed to the ICP-OES. An acidic solution (1 M HNO₃) was then pumped into the mini-column to elute As^V prior to measurement by ICP-OES. The total As was determined by ICP-OES without the sample passing through the mini-column. The calibration curves showed good linearity with values for the LOD of 158, 86, and 211 μg L⁻¹ for As^III, As^V, and total As, respectively. The results suggested that the oxidation of As^III was influenced by the presence of Fe^II in the culture medium, i.e., Fe^II enhanced the oxidation of As^III in A. brierleyi. In addition, the authors found that no soluble As species was present in the cell pellets and more than 60% of the As^III in the culture medium was oxidized by A. brierleyi after a 6 day incubation.

A number of reports have been published that describe techniques based on sample introduction by HG approaches. A method has been developed for the determination of As^III and As^V in natural water samples based on the generation of arsine (AsH₃) from the reaction between the As species in the injected solution and tetrahydroborate immobilized on a strong anion-exchange resin (Amberlite IRA-400) and then detection by AFS.⁶¹ Speciation was based on two different measurement conditions: (i) acidification to 0.7 M with HCl and (ii) acidification to 0.1 M with HCl in the presence of 0.5% L-cysteine, which produced two calibration equations with different sensitivities for each species. The LOD for a 0.5 mL sample volume was 13 ng L⁻¹ and 15 ng L⁻¹ for As^III and As^V respectively. The precision, expressed as % RSD of the measurement of 0.5 μg L⁻¹ As was 4.3% and 4.1% for determination of As^III and As^V, respectively, in 0.7 M HCl; and 3.8% and 3.6%, for the determination in 0.1 M HCl and 0.5% L-cysteine. Interferences from transition metals (Fe, Mn and Zn) and hydride-forming elements were eliminated by the addition of L-cysteine. The method was evaluated by the analysis of spiked natural waters. The recoveries for 0.5 and 1 μg L⁻¹ As^III were 92–108% and 88–112%, respectively; the recoveries for 0.5 and 1 μg L⁻¹ As^V were 94–111% and 95–112%, respectively. This method was also validated for total As by the analysis of a seawater CRM, NASS-6, which contains 1.43 ± 0.12 μg L⁻¹ (total As). The method was applied to the analysis of a number of real water samples. The time required per measurement was less than 4 min and the procedure consumed about 100× less HCl than the conventional continuous-flow procedure. A method for selective HG-cryotrapping (HG-CT) coupled to an in-house assembled and designed AFS instrument for determination of toxicologically important As species has been reported.⁶² A HG-CT system was interfaced to an advanced flame-in-gas-shield atomizer (FIGS) and its performance compared to a standard miniature diffusion flame (MDF) detector. A significant improvement over the MDF atomiser was found for both sensitivity and baseline noise and this was reflected in improved LOD values: 0.44, 0.74, 0.15, 0.17 and 0.67 ng L⁻¹ for As^III, iAs, MMA, DMA and TMAO respectively. The sensitivities with FIGS and MDF were equal for all As species, allowing for the possibility of single species standardization with As^V standard for accurate quantification of all other As species. The accuracy of HG-CT-AFS with FIGS was verified by speciation analysis in two samples of spiked bottled drinking water and the NRCC CRMs, CASS-5 (nearshore seawater) and SLRS-5 (river water) that contain traces of methylated As species. The sums of all quantified species corresponded with the certified total As. The feasibility of HG-CT-AFS with FIGS was also demonstrated by the speciation analysis in microsamples of exfoliated bladder epithelial cells isolated from human urine collected from residents of Chihuahua with a high chronic exposure to iAs from drinking water. The results for the sums of trivalent and pentavalent As species corresponded well with the results obtained by HG-CT-ICP-MS. On-line electrokinetic extraction and electrochemical HG coupled with AFS has been used for iAs speciation in water samples.⁶³ The system employed an H-type cell which consisted of four PTFE components; a working chamber, an auxiliary chamber, a connecting tube between the two (to give the ‘H’ configuration) and four Nafion membranes placed at the top of both chambers and at each end of the connecting tube. The As^V ions in the sample solution were firstly extracted into the working chamber of the H-type integration cell and reduced to As^III. Subsequently, the integrated cell was converted to an electrochemical HG unit to generate arsine by changing power supply and the direction of the electrical field. Finally, the arsine generated in the working chamber was separated using the gas–liquid separator and detected by AFS. Factors which may affect the extraction and HG such as pH and conductivity, composition of solutions (buffers and acids), effect of voltage and time, and interferences were investigated in detail. The precision (RSD, n = 10) ranged from 2.3–3.5% for peak area response for As^V at the 2 μg L⁻¹ level. A LOD (3σ) of 0.020 μg L⁻¹ As^V was achieved. The recoveries of three samples ranged from 98 to 104%. The use of the method to determine As^V in the CRM (BW3210) gave results which agreed well with the certified value, and the proposed method was successfully applied to preconcentration of As^V species in natural water samples, with As^III being calculated by difference. A method for the analysis of As species in aqueous sulfide samples has been presented.⁶⁴ The method uses IC-HG-AFS to determine a range of As species including: As^III, As^V, thioarsenite, monothioarsenate, dithioarsenate, trithioarsenate and tetrathioarsenate. The peak identification and retention times were determined based on standard analysis of the various As compounds, although only using a single approach. The LOD was 1–3 μg L⁻¹, depending on the variability of the baseline. The method was applied for on-site determination of As species in natural sulfidic waters, in which seven species were identified.

Prior to the widespread adoption of chromatography coupled to ICP-MS detection for As speciation studies, differential HG in which the conditions were altered to facilitate the detection of different As species was a popular technique. Arsenic speciation, without chromatography using selective HG has seen a return this year as a screening method for large numbers of samples.⁶⁵ The HG is used as a selective step to determine iAs in the gaseous phase while organically bound As remains in the solution. Mechanistically, iAs forms volatile arsine species with high efficiency when treated with NaBH₄ at acidic conditions, whereas most other organoarsenic compounds do not form any or only less volatile arsines. Additionally, using high concentrations of HCl further reduces the production of the less volatile arsines and arsine from iAs is almost exclusively formed, therefore enabling the measurement of iAs without a prior step of species separation using chromatography. In this study, a commercially available HG system was coupled to an ICP-MS and optimised for determination of iAs in rice and samples of marine origin (edible seaweeds) using different acid concentrations, wet and dry plasma conditions, and different reaction gas modes for HG. Comparing this method to HPLC-ICP-MS, no statistical difference in iAs concentration were reported and comparable LOD values were achieved in less analysis time. The efficiency of chemical generation of arsanes from iAs, MMA and DMA, to arsane, AsH₃, monomethylarsane, CH₃AsH₂ and dimethylarsane, (CH₃)₂AsH has been investigated in different reaction media (with and without additives such as L-cysteine and HClO₄), with the aim to better elucidate the mechanisms controlling the HG process.⁶⁶ The experimental conditions for non-chromatographic As speciation based on the selective determination of some As species was also investigated. Studies were performed by CF-HG-AS, using different reductants such as borane-ammonia, borane-tert-butylamine, and sodium tetra-hydridoborate in HCl and HClO₄ media, in the presence or absence of L-cysteine. The efficiency of the HG process was mainly controlled by the reactivity of the substrates with the borane, which could be strongly influenced by the formation of ion couples. The protonation of arsane did not play a significant role in the employed reaction system. By taking advantage of the different reactivity pattern of As species in selected generation conditions, DMA and MMA could be selectively determined in 0.5 and 10 M HClO₄ solutions, respectively, in the presence of L-cysteine, with borane-ammonia as the reducing agent, although clearly when using such high concentrations safety concerns also need to be considered. The presence of L-cysteine as a masking agent and the reducing properties of borane-ammonia ensured good control of interferences from Co^II, Cu^II, Fe^II, Fe^III and Ni^II. The selective determination of DMA with borane-ammonia/L-cysteine in HClO₄ was optimised and validated with CRMs of natural waters CASS-4, SLRS-4, and NASS-4 (NRCC).

A range of clinical applications of As speciation have been reported. A study to determine urinary As speciation reference values in 95 non-occupationally exposed volunteers based in the UK has been reported to help aid interpretation of As speciation results and better assess exposure.⁶⁷ Five species of As (AB, As^III, As^V, MMA and DMA) were determined using μHPLC-ICP-MS. Separation was achieved using an anion exchange column. The results presented give the 95th percentile of concentrations, both uncorrected for creatinine (μg L⁻¹) and creatinine corrected (nmol mmol⁻¹) in urine for the 95 volunteers. Statistical analysis was performed on the dataset using a Bayesian model to determine and quantify effects of gender, smoking and diet. Because 63% of As^V measurements were below the LOQ, the median As^V concentration could not be quantified. The 95th percentiles for As^III and As^V were 0.54 mg L⁻¹ (0.99 nmol mmol⁻¹ creatinine) and 0.23 mg L⁻¹ (0.35 nmol mmol⁻¹ creatinine), respectively. The 95th percentiles for MMA, DMA and AB gave values of 2.37 mg L⁻¹ (3.08 nmol mmol⁻¹ creatinine), 12.68 mg L⁻¹ (16.08 nmol mmol⁻¹ creatinine) and 126.7 mg L⁻¹ (174.7 nmol mmol⁻¹ creatinine), respectively. The statistical results show that the consumption of fish, shellfish and red wine has a significant elevating effect on AB, DMA and MMA urinary concentrations. However, no significant effect was observed for smoking. The regression model results indicate that creatinine correction was effective for arsenic species As^III, MMA, DMA and AB.

Melarsoprol, an organoarsenic compound, is widely used for the treatment of trypanosomiasis (human African sleeping sickness), although very little is known about its fate in the human body, its active metabolites passing through the blood–brain barrier and the mode of action. Previous pharmacological studies based on the determination of Melarsoprol by HPLC-UV or by a bioassay method have produced different results. Raber et al.⁶⁸ have reported a HPLC-ICP-MS method suitable for determining Melarsoprol and its metabolites in biological fluids. The major product reported was melarsen (4-((4,6-diamino-1,3,5-triazin-2-yl)amino)phenyl)arsonic acid identified by HPLC-ESI-MS, and by accurate mass measurements. Investigations into the stability of Melarsoprol in serum showed that within 30 h about 10% is converted to melarsen. In blood, however, most of the Melarsoprol was bound to proteins and only 1% was converted after 30 hours. The LOD for Melarsoprol and its conversion products were in the range of 1 μg As L⁻¹ (13 nmol As L⁻¹) based on a S/N ratio of 3 with a 10 μL injection volume allowing direct determination of the compounds in blood and serum after protein precipitation, at therapeutically realistic concentrations. In a recent study by Yehiayan et al.⁶⁹ dimethylarsinothioyl glutathione (DMTA^V(GS)) has been reported as a metabolite in human multiple myeloma cell lines exposed to Darinaparsin (a small-molecule organic arsenical with potential antineoplastic activity). Having ICP-MS as the detector indicated the presence of S along with As in an unidentified molecule in the chromatograms of cell lines treated with DMA^III(GS). Analysis of the unknown peak by LC-ESI-MS in the MS and tandem MS modes revealed molecular ion peaks at m/z 443.9 and 466.0, corresponding to (DMTA^V(GS) + H)⁺ and (DMTA(V) (GS) + Na)⁺, as well as peaks at m/z 314.8 for the loss of glutamic acid and m/z 231.1 for the loss of glycine. In addition, peaks were observed at m/z 176.9 corresponding to cysteine and glycine adducts and at m/z 137.1 for the (C₂H₆AsS)⁺ ion. An increase in the peak area of the unidentified peak was observed upon spiking the cell extracts with a standard of DMTA^V(GS). The DMTA^v(GS) was found to be stable in cell extracts at both acidic and neutral pH conditions.

The determination of As species in biological samples remains of interest. An analysis of the biological response of mouse liver (Mus musculus) exposed to As₂O₃ based on integrated-omics approaches has been reported.⁷⁰ The toxicological effects of As^III after oral administration (3 mg kg⁻¹ body weight and per day) were investigated over a period of 7 days using metallomics, metabonomics and proteomics approaches. The combination of SEC-ICP-MS with AEC was used to characterize the biological response of the exposed mice. Direct infusion ESI-QTOF-MS of polar and lipophilic extracts using +ve and −ve ion modes, provided information about time-dependent changes in endogenous metabolites identified by Partial Least Squares Discriminant Analysis. Finally, the study evaluated the regulation of enzymes related to oxidative stress such as SOD, glutathione reductase, catalase and peroxidases in connection with metal toxicity issues. The results show that the iAs methylation in the liver may reach saturation point upon chronic exposure to the element. The SEC-ICP-MS coupling provided information about metal containing-proteins and metabolites related to As exposure which has been correlated with the changes in metabolism. The study showed that As causes biochemical pathway alterations, such as energy metabolism (e.g. glycolysis, Krebs cycle), amino acid metabolism, choline metabolism and degradation of membrane phospholipids. The lack of relevant standards limits investigations of the quality of the measured isotopologue distributions. To meet that need, Millard et al.⁷¹ have developed a theoretical and experimental framework for the biological production of metabolites with fully controlled and predictable labelling patterns. The strategy is valid for different isotopes and different types of metabolism and organisms, and was applied to two model microorganisms, Pichia augusta and Escherichia coli, cultivated on ¹³C-labeled methanol and acetate as sole carbon source, respectively. The isotopic composition of the substrates was designed to obtain samples in which the isotopologue distribution of all the metabolites should give the binomial coefficients. The strategy was validated using LC-MS-MS by quantifying the complete isotopologue distributions of different intracellular metabolites, which were in close agreement with predictions. This strategy can be used to evaluate entire experimental workflows, from sampling to data processing, or different analytical platforms in the context of isotope labelling experiments. One challenging area in the analysis of As-glutathione/phytochelatin complexes is their extraction from small amounts of biological material and the maintenance of their stability during HPLC separation. Focused sonication has been used to extract these complexes from the single-celled green algae Chlorella vulgaris and the integrity of the complexes determined by HPLC (C₁₈ column) online with simultaneous HR-ICP-MS and ES-MS/MS detection.⁷² Water soluble As species were extracted with an improved efficiency of 71% and much reduced extraction times (30 s) allowing for the determination of unstable As phytochelatin and glutathione species in small biomass samples making the method particularly well-suited for cell cultures. In one of the more esoteric applications reported this year, As exposure of Pre-Columbian populations that inhabited the Tarapaci mid river valley in the Atacama Desert in Chile (AD 500–1450) were assessed using SR-XRF mapping, XAS, XRD and FTIR spectromicroscopy measurements on ancient human hair.⁷³ High concentration of As, mainly in the form of inorganic As^III and As^V were detected suggesting chronic arsenicism through ingestion of As-polluted water rather than external contamination by the deposition of heavy metals due to metallophilic soil microbes or diffusion of As from the soil, although it is not clear from the paper how post-burial changes were evaluated.

Arsenic speciation in marine based products continues to be the most popular area of research. Arsenolipids are the major As species present in fish oils and several chemical structures have been elucidated in recent years demonstrating the complexity of this metal(loid) in the marine environment. In a recent study⁷⁴ commercial oils from 8 species of fish and oil from the Greenland seal, were analyzed for arsenolipids using RP HPLC-ICP-MS. The total As concentrations ranged from 1.6 to 12.5 mg kg⁻¹ oil, the arsenolipids were quantified using three different As-containing calibration standards; DMA, triphenylarsinoxide (Ph₃AsO) and a synthesized As-containing hydrocarbon (AsHC) (dimethylarsinoyl nonadecane; C₂₁H₄₃AsO). The observed variation in signal intensity for As during the gradient elution profile in RP HPLC was compensated for by determining the time-resolved response factors for the arsenolipids. Isotopes of germanium (⁷⁴Ge) and indium (¹¹⁵In) were used as post column internal standards, added via a T-piece using a peristaltic pump. The most suitable calibration standard for the quantification of arsenolipids was shown to be DMA, with recoveries between 91% and 104% compared to total As measured in the same extract. The species C₁₇H₃₈AsO, C₁₉H₄₂AsO and C₂₃H₃₈AsO, were identified as the major arsenolipids in the extracts of all oils by HPLC coupled with QTOF-MS. Minor amounts of two As-containing fatty acids (C₂₃H₃₈AsO₃ and C₂₄H₃₈AsO₃) were also detected. The sum of the arsenolipids and the As-containing fatty acids determined in the study accounted for 17–42% of the total As in the oils. Amayo et al.⁷⁵ have reported the identification of sixteen arsenolipids in cod-liver oil. The fish oil was fractionated on a Varian ionosphere C silica gel column (150 × 4.6 mm) and the fraction enriched with As analysed using RP-HPLC on a Eclipse, XBD, C₁₈ column (150 × 4.6 mm) online with ICP-MS and ES-Orbitrap-MS. Among the arsenolipids identified six saturated arsenolipid compounds (C₁₆H₃₄O₃As, C₁₈H₃₈O₃As, C₂₀H₄₂O₃As, C₂₃H₄₈O₃As, C₂₅H₅₂O₃As, C₂₇H₅₆O₃As) and three unsaturated arsenolipids (C₁₉H₃₆O₃As, C₂₆H₅₂O₃As and C₂₁H₄₄OAs) have not been reported before. Structural assignment was achieved by using the As signal from ICP-MS, retention time behaviour and accurate mass determination of fragment and molecular peaks. In addition, the unknown degradation products of arsenolipids eluting in the void volume were investigated using fraction collection, CEC and accurate mass determination, and were found to contain predominantly DMA with trace amounts of MMA, dimethylarsenopropanoic acid (DMAP) and dimethylarsenobutanoic acid (DMAB). This finding is reported to be useful in epidemiologic studies where urinary DMA and other arsenic metabolites have been used as biomarker in accessing human exposure to arsenic, since the degradation products identified here may contribute to total urinary As concentrations and lead to over estimates of exposure. The determination and identification of hydrophilic and hydrophobic As species in methanol extracts of fresh cod liver by RP-HPLC with simultaneous ICP-MS and ESI-QTOF-MS detection has also been reported.⁷⁶ The total concentration of As in the fresh cod liver was determined by ICP-MS (1.53 ± 0.02 mg As kg⁻¹ w/w). The extraction recovery was ca. 100% and the column recovery >93%. Besides polar inorganic and methylated As species (>70%) hydrophobic As-containing fatty acids and hydrocarbons were found. Based on the MS data, proposals for molecular structures were elaborated for 20 of the organic As species including 10 As-containing fatty acids (AsFA) and an AsHC mentioned for the first time in fresh cod liver. Arsenobetaine was found as the main water-soluble As compound in cod liver. The same research group have used a similar approach for Capelin oil (Mallotus villosus).⁷⁷ Twelve arsenolipids were identified in the fish oil including three AsFAs and seven AsHCs. Among the AsHCs, four that were identified had protonated molecular masses of 305, 331, 347, and 359 and have not been reported before. In addition, the compounds with molecular formulas C₂₀H₄₄AsO⁺ and C₂₄H₄₄AsO⁺ were found in low concentrations and showed chromatographic properties and MS data consistent with cationic trimethylarsenio fatty alcohols. Derivatization by acetylation and thiolation coupled with MS was successfully used to establish the occurrence of this new class of arsenolipids as cationic trimethylarsenio fatty alcohols. Glabonjat et al.⁷⁸ have described a method based on HPLC-ICP-MS/ES-MS to quantitatively measure seven of the major arsenolipids present in a seaweed (Hijiki) CRM 7405-a from the National Metrology Institute of Japan (NMIJ), which is a rich source of arsenolipids. Sample preparation involved extraction with dichloromethane and methanol, a clean-up step with silica, and conversion of the (oxo)arsenolipids originally present to thio analogues by brief treatment with H₂S. Compared to their oxo analogues, the thioarsenolipids showed much sharper peaks using RP HPLC, which facilitated their resolution and quantification. The concentrations of two As-containing hydrocarbons and five arsenosugar phospholipids are reported in the Hijiki seweed CRM (NMIJ). Thio-methylated As species in marine organisms where the oxygen bonded to As is replaced by an S group and thioarsenate species in sulfide rich water environments have been studied by Maher et al.⁷⁹ using HPLC-ICP-MS. Thio-methylated As species were separated using an Alantis C₁₈ RP column and elution with an aqueous 20 mM phosphate buffer (pH 3.0). Thioarsenate species were separated using an Ion Pac, AS16 anion exchange column with a sodium hydroxide gradient (20–100 mM) and the use of an anionic self-regenerating suppressor to remove sodium ions before the ICP-MS spray chamber. The thio-methylated arsenic species in marine biota were stable to freeze drying and microwave extraction with methanol–water (1 : 1 v/v) at 70 °C. Rotational mixing at 25 °C for long periods caused the loss of species whilst freeze–thawing of extracts resulted in oxidation of species. The thio-methylated As species in the extracts were stable if the solvent was removed and the residues stored dry in a dessicator. The thioarsenate (mono, di, tri and tetra) species in water samples were also stable during analysis if manipulated under a N₂ atmosphere and if during chromatography mobile phases were degassed and precautions taken to exclude air from samples.

Many extraction methods reported for seafood samples are not particularly novel and consequently are excluded from this review. An ID LC-MS-MS method has been developed and applied to the determination of AB from an oyster based candidate CRM.⁸⁰ The exact matching isotope dilution approach was adopted for accurate determination of AB using ¹³C-labelled AB as an internal standard. Efficiencies of different AB extraction methods were evaluated using a codfish RM and a sonication method was selected for the certification of the oyster candidate CRM. HILIC combined with ESI-MS-MS was used for quantification of AB in the samples. The certified value of AB was determined as 6.60 mg kg⁻¹ ± 0.31 mg kg⁻¹ and showed excellent between-bottle homogeneity between subsamples of less than 0.42%. The major source of uncertainty was the certified value of the AB standard solution. The technique of INAA has been used for validation of the As mass fraction determined by LC-ICP-MS in extracts of both candidate and certified fish based RMs.⁸¹ Various methods for the extraction of water-soluble As species were evaluated and the best results were acquired after methanol/acetone/water extraction yielding 93% extractable As in a tuna RM. This procedure was then used for the LC-ICP-MS studies. The results demonstrated that INAA could account for 100% of the distribution of As species in analytical processes and complemented LC-ICP-MS for the validation of the characterization of As species in the development of new fish RMs, including a tuna tissue and robalo (Centropomus undecimalis) liver tissue. Good agreement with CRM values were obtained.

A method for the detection of AB, AC, As^III, As^V, MMA, DMA in fish using CE-ICP-MS analysis has been reported.⁸² The results showed that the six species of As were separated within 20 min under the optimised conditions. Good linearity was observed in the range from 2–50 μg L⁻¹ with a linear correlation r > 0.996, LODs were 0.10–1.08 μg L⁻¹ and peak area precision (n = 5) < 7% RSD. The method was successfully used for the determination of the As species in Japanese Spanish mackerel, Scomberomorus niphonius. Recoveries were between 93% and 98%, and AB was the main species in the sample. Arsenic in mussels has been determined using HPLC-ES-MS with parallel determination of total As by AAS.⁸³ The proposed method was evaluated using a range of figures of merit (mean recoveries, repeatability, specificity, LOQ and LOD) to ensure that it was fit for purpose. Total As concentrations ranged from 1.38–12.79 mg kg⁻¹. Arsenobetaine and DMA were detected in all samples (AB: 0.72 to 10.36 mg kg⁻¹; DMA: 0.28 to 1.08 mg kg⁻¹), and concentrations of AC and MMA ranged from 0.20–1.53 mg kg⁻¹. The method was used to carry out the first survey of the concentrations of AB, AC, MMA, and DMA in seafood from southern Italy. A method utilising CE-ICP-MS has also been developed for the detection of trace As^III, As^V, MMA, DMA in groundwaters.⁸⁴ The results showed that the four species of As could be baseline separated within 9 min under optimised conditions with LODs between 0.2–0.5 μg L⁻¹. The recoveries were between 93%–106%. It was concluded that As^III and As^V were the main As species in the groundwater of Inner Mongolia.

Arsenic speciation in rice continues to be of significant interest reflecting the international importance of this food crop, both as a staple in its own right and as an ingredient within other products. One of the more novel approaches to disseminating information on this topic is a web seminar by co-author of this review Julian Tyson. In the seminar Prof Tyson explains how to develop a reliable method for As speciation and he addresses questions raised during the original web seminar.⁸⁵ A recording of the web seminar titled “Speciation Analysis: A Critical Look at Methods Involving HPLC with ICP-MS Detection, with a Focus on Rice” is available for free at http://www.spectroscopyonline.com/ArsenicRiceWebinar2013. Arsenic species have been determined in rice using HPLC-ICP-MS employing a pentafluorophenyl column.⁸⁶ Five As species were separated with a Discovery HS F5 column and a simple, volatile, and isocratic mobile phase of 0.1% (v/v) formic acid and 1% (v/v) methanol. The Discovery HS F5 column with a pentafluorophenyl (PFP) stationary phase gave sharp peaks and full separation of the arsenic species in 5 min; other PFP columns delivered a lower performance. The extraction of As from rice samples was performed using 0.15 M nitric acid. The methodology was validated by use of a white rice floor CRM 7503-a (NMIJ) and rice flour SRM 1568a (NIST). Fourteen different extracting solvents for As species present in rice flour samples have been evaluated using heat-assisted extraction and detection by HPLC-ICP-MS.⁸⁷ The extraction efficiencies for total As species and As^III, As^V and DMA found in the samples were investigated, although the concentration of DMA did not vary with treatment conditions. However, the concentrations of extracted total As and those of As^III, As^V were dependent on the extracting solvents used. When an extracting solvent was highly acidic, the concentrations of extracted total As found were in good agreement with the total As concentration determined by ICP-MS after microwave digestion, although a part of the As^V was reduced to As^III during the extraction process. Extraction under neutral conditions increased the extraction efficiency for As^V, but extracted total As was decreased because part of the As^III was not extracted. A heat block extraction technique using 0.05 mol L⁻¹ HClO₄ and Ag-containing 0.15 mol L⁻¹ HNO₃ was also developed and proved suitable for total iAs in rice flour. A study to evaluate the As species in Korean and USA rice grains using MAE and HPLC-ICP-MS has shown that As^III and DMA were the major species present.⁸⁸ The percentage of iAs ranged from 54.5–87.9% for Korean rice and 52.9–72.9% for USA samples. The order and percentage of As species observed in Korean and USA rice were As^III (70%) > DMA (24%) > As^V (5%) > MMA (1%), and As^III (64%) > DMA (28%) > As^V (5%) > MMA (3%), respectively. The use of HPLC-HG-AFS has been reported in a study of the As content of Chinese rice.⁸⁹ The As species As^III, As^V, DMA, and MMA, were extracted by methanol–water (50 : 50, v/v) containing 0.02 mol L⁻¹ nitric acid in a MAE procedure. The results showed that the method offered good recoveries (>90%) for rice, with LOQ values of 8.0, 20, 12, and 12 ng g⁻¹ for As^III, As^V, DMA, and MMA, respectively.

Arsenic speciation in other food and drink products has also been reported. An online coupled system based on GC-HG-AFS for the analysis of As species in food seasoning (soy sauce and vinegar) has been reported.⁹⁰ Using absorbent cotton as separation medium, the volatile As species released from the samples were cold trapped online by liquid nitrogen. Baseline separation of the volatile As species was achieved at room temperature and the As species detected by AFS. The experimental conditions such as flow rates of carrier gas (He), acid type and concentration, reducing reagent concentration and reaction time were optimised to give LOD values for As^III, As^V, MMA and DMA of 0.2, 0.2, 0.3 and 0.8 μg L⁻¹, respectively. A methodology for the non-chromatographic screening of the As species present in 29 edible oils of different origin has been reported.⁹¹ Reverse dispersive LLE was used to extract water soluble As compounds (iAs, MMA, DMA and AB) from the oils into a slightly acidic aqueous medium. The total As content was measured in the extracts by ETAAS using Pd as the chemical modifier. By repeating the measurement using Ce instead of Pd, the sum of iAs and MMA was obtained. The LOD value was 0.03 ng As g⁻¹ oil. The determination of iAs at low levels in cereal-based food continues to be of interest.⁹² Thus, IC-ICP-MS was used to determine iAs, MMA and DMA in a range of 29 cereal based products including bread, biscuits, breakfast cereals, wheat flour, corn snacks, pasta and infant cereals. The total and iAs levels found ranged between 3.7–35.6 and 3.1–26.0 μg As kg⁻¹, respectively. A method for the simultaneous determination of arsanilic, nitarsone and roxarsone residues in foods of animal origin has been developed using ASE and measurement by HPLC-AFS.⁹³ Ultrasound centrifugation extraction and ASE were compared following optimisation. The calibration curves for arsanilic, nitarsone and roxarsone were linear up to 2.0 mg L⁻¹ with correlation coefficients of 0. 9992–0.9998. The LOD values were 2.4, 7.4 and 4.1 μg L⁻¹ for arsanilic, nitarsone and roxarsone respectively, and spiked recoveries were better than 87.1%.

The spatial distribution and speciation of As using X-ray techniques has been used to study minerotrophic peatland in Gala di Lago, Switzerland.⁹⁴ Thus μXRFS and μXAS were used to examine thin sections of undisturbed peat from depths of 0–37 cm and 200–249 cm. It was found that As in the near-surface peat was mainly concentrated in 10–50 μm sized hotspots, identified by μXAS as realgar (α-As₄S₄). In the deep peat layer samples, however, As was more diffusely distributed and mostly associated with particulate natural organic matter of varying decomposition stages. The natural organic matter associated As was present as As^III bound by sulfhydryl groups. Arsenopyrite (FeAsS) and arsenian pyrite (FeAs_xS_2−x) of <25 μm size, which have escaped detection by bulk As and Fe K-edge XAS, were found as minor As species in the peat. X-ray resonant Raman scattering spectroscopy has been reported for As speciation.⁹⁵ Mineral samples containing As in different oxidation states (As^III and As^V) and two biological forms of As (MMA and DMA) were analysed. The technique is clearly in the early stages of development and only qualitative assessments were possible, although the authors claimed that the residuals of compounds could be used to identify the oxidation state of the elements under study. Arsenic metabolism in non-hyperaccumulator plants remains of research interest, and has been studied with respect to Macrophyte Ceratophyllum demersum.⁹⁶ Concentrations and tissue-dependent speciation and distribution were assessed using HPLC-ICP-ES-MS in whole-plant extracts and by confocal μXANES in intact shock-frozen hydrated leaves. The cellular element distribution was also determined using μXRF. Chromatography revealed up to 20 species binding more than 60% of the accumulated As, of which eight were identified as thiol-containing species including, phytochelatins, glutathione, and cysteine. Three newly identified complexes were among the species present: Cys-As^III-reduced glutathione₂, Cys-As-(GS)₂, and GS-As^IIIdesgly-reduced glutathione₂. Confocal μXANES identified As^V, As^III, As-(GS)₃, and As-phytochelatin species with varying ratios in various tissues.

5.3 Boron

A species-specific response for B by ICP-OES has been observed.⁹⁷ Significant bias between the results obtained from two different instruments arose during the laboratory validation of a new ICP-OES spectrometer. Upon investigation, it was found that B was present in at least two chemical forms in acidic (HF/HNO₃) digests of metallurgical grade Si. One species was identified by HPLC-ICP-MS, utilising a porous graphite carbon column (100 × 4.6 mm, 10 μm) and a gradient elution from 0 to 100% of 1 mol L⁻¹ formic acid, as boric acid whilst a second B species was not identified. Experiments using ¹⁰B enriched boric acid showed the species was not formed during the acidic sample digestion procedure and that the B : Si stoichiometry was 1 : 10. Further attempts to identify the unknown B species, using ES-MS, NMR, XRD, XPS and EPR were also unsuccessful. It was tentatively concluded that the unknown B species, which was sensitive to oxidation by H₂O₂ was a B–Si cluster or polymer and that it was responsible for B fractionation in the PTFE/PFA ICP-OES sample introduction system, which caused signal instability and necessitated prolonged washout times and led to significant underreporting of B concentrations when compared to calibration standards prepared from different B species. The addition of H₂O₂ was required to improve measurement reproducibility and the authors conclude that this should be considered for acidic digests of Si rich matrices when B is the target analyte. The reported fractionation presumably occurred due to retention of the unknown B species on part of the sample introduction system of one ICP-OES instrument, which was fitted with a V-groove nebuliser and a Sturmann Masters type spray chamber. However, the polymer type of these components is not stated. This would be useful information to other workers as this type of spray chamber is available in at least two polymer types. It would also have been beneficial to swap the nebuliser/spray chamber assemblies between instruments to isolate these components as the probable source of the observed fractionation.

5.4 Chromium

Most of the methods for Cr speciation reported this year are for water and/or effluent samples. Two different approaches for estimating the uncertainties associated with an HPCL-ICP-MS measurement procedure, for Cr species present in drinking water, have been compared.⁹⁸ The Cr species were extracted and separated from drinking water samples using RP ion pair HPLC. The mobile phase comprised 0.8 mmol L⁻¹ TBAH and 0.6 mmol L⁻¹ EDTA at pH 6.9 flowing isocratically at 1.2 mL min⁻¹ through a 33 mm length C₈ column. The use of commonly employed organic mobile phase components (e.g. MeOH) was avoided to maintain instrument sensitivity and minimise the formation of C based interferences on Cr. The HPLC eluent was directly coupled to an ICP-MS instrument, which was operated in DRC mode with NH₃ as the reaction gas to minimise polyatomic interferences arising from the plasma gases and reagents used. Water samples were diluted 3 : 1 with the mobile phase and allowed to stand for at least one hour, to allow Cr–EDTA complexes to form, prior to analysis. Using these HPLC conditions, Cr^III and Cr^VI were separated in under three minutes, with sensitivity greatest for the latter species, presumably due to a greater sample transport and/or ionisation efficiency for the Cr₂O₇²⁻ ion than for the Cr–EDTA complex. The LOD was 0.1 μg L⁻¹ for both species. This implies that baseline noise increased after elution of the Cr–EDTA complex as the sensitivity (signal to analyte amount) increased for the Cr₂O₇²⁻ ion. The method repeatability and intermediate precision values were 1.5 and 3.4% for Cr^III and 1.6 and 3.5% for Cr^VI respectively. Method trueness was assessed by spiking experiments of up to 2 μg L⁻¹ added Cr, as no suitable species specific CRM was available, and spike recoveries ranged from 93 to 115%. Finally, the uncertainties associated with the method were estimated. For the classical ‘bottom up’ or modelling approach the expanded uncertainty was reported to be <5% relative for both Cr species. This value rose to <8% relative when the ‘single-laboratory validation’ approach was used. This increase in the expanded uncertainty was ascribed to an increase in the variability of each of the individual components, such as precision and bias, used to calculate the expanded uncertainty. The ‘bottom up’ approach, although it is more time consuming to undertake, generally produces lower method uncertainty estimates than those obtained by the ‘single-laboratory validation’ approach. This latter uncertainty estimate is easier to produce though, as routinely collected QC data can be used, and the adoption of these principles is to be encouraged in research laboratories. Finally, both uncertainty estimates and the LOD values presented allow the determination of Cr in potable waters at less than the WHO guideline value.

Additional reports of Cr speciation in aqueous based samples include a redox speciation and preconcentration study of Cr^III and Cr^VI.⁹⁹ To achieve this a FI system, with dual mini-columns prepared from cross-linked polymers-poly(methacrylic acid) and polyvinylimidazole, was operated online by coupling to a FAAS instrument. Samples containing Cr^III and/or Cr^VI at pH 4.0 were loaded onto the dual mini-columns in parallel at a flow rate of 3.0 mL min⁻¹. The Cr^III was selectively retained on the poly(methacrylic acid) stationary phase whilst Cr^VI was retained on the polyvinylimidazole phase. Subsequently, the retained species on each column was sequentially eluted with 2.5 mol L⁻¹ HNO₃ and the Cr species quantified by FAAS. The LODs were found to be 0.84 and 1.58 μg L⁻¹ for Cr^III and Cr^VI respectively. The preconcentration factors were reported as 47.3 and 8.6 for Cr^III and Cr^VI, respectively. The method was applied to the speciation of Cr in different types of water sample. Spike recovery values, for additions of up to 40 μg L⁻¹ of Cr^III and Cr^VI, ranged from 90 to 108%. A surfactant assisted DLLME method has been developed for the sequential and simultaneous preconcentration of Cr^III and Cr^VI from tap, river, ground and wastewaters.¹⁰⁰ The procedure is based on selective complexation of Cr^III with sodium dodecylbenzenesulfonate (SDBS) and, separately, the complexation Cr^VI with diphenylcarbazide (DPC). The total Cr present in the samples was extracted into 1-undecanol after adding both complexing reagents simultaneously. Subsequently, the Cr complexes formed by each procedure were extracted into 1-undecanol, which was then solidified by chilling, removed from the water sample and redissolved in MeOH prior to Cr quantification by ETAAS. The LOD of the method was 1 pg mL⁻¹ for Cr^III and 4 pg mL⁻¹ for Cr^VI, and spiking experiments, in the range of 100–20 000 pg mL⁻¹ added Cr, gave recoveries of between 92 and 98% which were independent of the spiked concentration of Cr. The sequential extraction method was applied to three CRMs, BCR 713 (effluent wastewater) and 714 (influent wastewater) and NIST 1643c (trace elements in water), and four water samples. For the CRMs the sum of Cr^III and Cr^VI measured in each was stated to be in good agreement with the certified total Cr value, although no statistical basis was given for this statement. For the water samples, Cr^III concentrations ranged between 0.1 and 18 ng mL⁻¹ and 0.06 to 7.5 ng mL⁻¹ for Cr^VI. The authors stated that the method meets the current US EPA requirements for Cr determinations in water samples. Online SPE-FAAS has also been used to extract and quantify Cr^III and Cr^VI from water samples.¹⁰¹ Two mini-columns, one packed with poly-2-(5-methylisoxazole)methacrylamide-co-2-acrylamido-2-methyl-1-propanesulfonic acid-co-divinylbenzene for extraction of Cr^III and the second packed with Dowex 21K to extract Cr^VI, were fitted in parallel to a FI manifold. The manifold also incorporated two switching valves and two peristaltic pumps controlled by a timer module, with the eluent line from each column being directly coupled to a FAAS instrument. All solutions were adjusted to pH 3.0 prior to analyte extraction, the Cr species were eluted from the columns with 3 mol L⁻¹ HNO₃ and each Cr species was determined sequentially. After a hot oxidation procedure total Cr was extracted using the Dowex 21K filled column only. The enrichment factors for Cr^III and Cr^VI were 48 and 30 and the LOD values were found to be 0.05 and 0.3 μg L⁻¹ respectively. For the CRM analysed (CWW-TM-D, waste water) the sum of the Cr^III and Cr^VI species value of 0.99 mg L⁻¹ and the total value obtained, after hot oxidation of the Cr species to Cr^VI, of 1.02 mg L⁻¹ was in agreement with the certified value of 1.00 mg L⁻¹. Finally, a method for Cr speciation based on anion-exchange HPLC has been developed.¹⁰² To achieve this, Cr^III present in the samples was complexed with EDTA, 2 mmol L⁻¹, at pH 7.0 and 70 °C to form an anionic species. The HPLC mobile phase consisted of 60 mmol L⁻¹ NH₄NO₃ at pH 9.3, with an isocratic elution separating the two Cr species in <8 min, was introduced directly to the ICP-MS instrument, which was operated in CC mode with He as the collision gas. After careful optimisation of the separation conditions the reported LOD values were 0.051 and 0.078 μg L⁻¹ for Cr^III and Cr^VI respectively. No CRM was used to validate the method, however, spike recoveries were between 95 and 109% for both Cr species. The measured concentrations in various water samples ranged between 0.19 and 0.27 μg L⁻¹ for Cr^III and 0.72 and 0.88 μg L⁻¹ for Cr^VI.

Three papers report methods for Cr speciation in solid samples. In a comprehensive study, species specific IDMS was applied to soil samples, with spikes of ⁵⁰Cr^III and ⁵³Cr^VI, and MAE.¹⁰³ Two different extraction solutions were used; for Cr^VI, 0.5 mol L⁻¹ NaOH, 0.28 mol L⁻¹ Na₂CO₃, a pH 7.0 phosphate buffer and 2 mg L⁻¹ MgCl₂ at 95 °C for 1 h and for Cr^III and Cr^VI a 50 mmol L⁻¹ solution of EDTA at pH 10.0. For this latter extraction, the MAE procedure was 90 °C for 5 min followed by 110 °C for 5 min with cooling to 25 °C between the two cycles. After both procedures, the extractant mixture was centrifuged and the supernatant reserved for analysis by AEC-ICP-MS with an isocratic elution with 2 mmol L⁻¹ EDTA, adjusted to pH 10.0, as the mobile phase. The solid extraction residues were then subjected to a HNO₃/HF digestion procedure and the resulting digests analysed for total Cr to allow a mass balance to be calculated. Mass bias correction factors were determined by analysing naturally abundant Cr by either direct aspiration or, for ssIDMS, AEC-ICP-MS of a mixed Cr^III and Cr^VI standard at the beginning, middle and end of each analytical sequence. The ICP-MS instrument was operated in CC mode, with He as the cell gas, to minimize polyatomic interferences on the three Cr isotopes monitored (m/z 50, 52 and 53). Three soil CRMs, SRM 2709a, SRM 2711a (both NIST) and SQC012 (Sigma Aldrich) and two candidate CRMs were subjected to the developed analytical procedures. In all cases, for both extraction solutions, and with corrections for species inter-conversions made using the IDMS methodology employed, agreement was obtained between the found and certified values. The same ssIDMS methodology was also applied to the determination of soluble Cr species in 24 dietary supplements, in which the found Cr^VI mass fraction ranged from below the LOD, of 2.8 ng g⁻¹, up to 120 μg g⁻¹.¹⁰³ In addition, the Cr^VI values obtained for four Cr raw material samples by ssIDMS were compared with those obtained by conventional external calibration IC-ICP-MS and EPA Method 7196A. For each sample, the results obtained by the EPA method were significantly lower, by e.g. 100 to 3000%, than those of the other two methods, which were in better agreement in some cases. An analytical method for the determination of soluble and insoluble Cr^III and Cr^VI species in welding fume workplace air has been developed.¹⁰⁴ Soluble Cr^III and Cr^VI were extracted from aerosol filters using a solution of 0.05 mol L⁻¹ HNO₃ at pH 4.0 and UAE for 30 min. For Cr^VI, the extractant solution was 2% KOH 3% Na₂CO₃ at pH 9.5 with UAE at 70 °C for 30 min. The AEC-ICP-MS procedure was the same as that cited earlier⁹⁸ except for minor differences in the eluent pH and flow rate. The total leachable Cr^VI value obtained for BCR 545 (welding dust), of 40.2 g kg⁻¹ agreed favourably with the certified value of 40.7 g kg⁻¹. In addition, a mass balance approach gave recoveries of >97% for the sum of the Cr species compared with the total Cr content, of various spiked blank filters and collected workplace air samples.

5.5 Cobalt

The speciation of Co compounds in nutritional supplements has been performed using HPLC-ICP-MS.¹⁰⁵ Cobalt species were solubilised from locally obtained supplement tablets using 0.5% v/v HNO₃ and MAE at 90 °C for 10 min. The extract was subsequently centrifuged and the supernatant collected and diluted with the HPLC mobile phase (22% v/v MeOH, 8 mmol L⁻¹ NH₄OAc). After filtration (0.2 μm PVDF membrane) the Co species were separated isocratically using a RP C₈ column (30 × 3 mm, 5 μm). The total HPLC eluent flow was interfaced with the ICP-MS instrument via a commercially available membrane desolvator to reduce solvent loading on the plasma and to minimise the ⁴³Ca¹⁶O⁺ polyatomic interference on ⁵⁹Co⁺. Using the optimised HPLC-ICP-MS system, three Co species, Co^II, hydroxycobalamine (OH-Cbl) and cyanocobalamine (CN-Cbl), were baseline resolved and separated in six minutes. Subsequently, these three species were identified, by both retention time matching and HPLC-ES-MS analysis, in two of the nutritional supplements whilst only Co^II was detected in a third supplement type. A mass balance approach, which compared the sum of the Co species detected with the total Co found in each supplement by a concentrated HNO₃ MAE digestion procedure, gave ‘recoveries’ of 98% or greater. The LOD values were 8, 13 and 14 ng L⁻¹, as Co, for Co^II, OH-Cbl and CN-Cbl respectively. The authors concluded that the developed method would also be suitable for the determination of cobalamin species in food. This, of course, will be dependent on a suitable extraction procedure for this sample type.

5.6 Gadolinium

Due to the use of Gd-based MRI contrast agents, the small but steady growth of the body of papers covering methods for Gd speciation, has continued. A method combining HILIC, membrane desolvation and ICP-MS has been developed for the determination of Gd species in surface waters that receive treated waste water effluent.¹⁰⁶ Separation of four Gd species, Gd-BOPTA, Gd-DTPA, Gd-BT-DO3A and Gd-DOTA was achieved in under 15 min, using a 10 × 2.1 mm HILIC column and an isocratic elution with a mobile phase of 50 mmol L⁻¹ ammonium formate at pH 3.75. The LOD, calculated from a signal-to-noise ratio of 10, was reported as 0.09 nmol L⁻¹ for Gd-DTPA, 0.08 nmol L⁻¹ for Gd-BT-DO3A and 0.10 nmol L⁻¹ for Gd-DOTA. A LOD value for Gd-BOPTA was not reported and the chromatographic peak for this species exhibited a ‘shoulder’ that was attributed to the existence of several isomers. The different LOD values obtained for each species, which the authors claim to be the lowest yet reported for HILIC-ICP-MS methods without any sample enrichment, was assigned to the differing peak widths obtained for each species. In the surface water samples the three contrast agents Gd-DTPA, Gd-DOTA and Gd-BT-DO3A were identified by retention time matching and quantified in constant ratios in those samples. The concentrations were found in a range from 0.59 nmol L⁻¹ for Gd-DOTA up to 3.55 nmol L⁻¹ for Gd-BT-DO3A. The contrast agent concentration was found to account for 74–89% of total Gd concentration in the samples collected, indicating that other Gd species may have been present but not detected. The authors conclude that further work is required to assess the ecotoxicological effects of Gd-based MRI contrast agents on aquatic biota.

5.7 Iron

A method for determining the redox speciation of Fe, based on in-column complexation with dipicolinic acid (DPA), has been developed.¹⁰⁷ For this work, a metal-free HPLC system was directly coupled to an ICP-MS instrument, operated in collision cell mode with He as the cell gas to minimise polyatomic interferences on the Fe isotopes being monitored. Due to the lability and reactivity of both Fe^II and Fe^III species, a thorough investigation was undertaken into the HPLC separation including column type, mobile phase composition and how these affected the potential for species inter-conversion(s). Isotopically modified standards, enriched in either ⁵⁴Fe or ⁵⁷Fe, allowed the amount of carryover between samples and standards, or vice versa, to be calculated by IDMS, one of the first uses of IDMS in this way. Dithiothreitol was added to all Fe^II solutions to prevent oxidation to Fe^III. Investigation using the enriched Fe species and IDMS showed that significant amounts, >20%, of Fe^II and Fe^III were carried over from sample to sample. The source of this carryover was identified as the HPLC column and possible interactions of the Fe species with residual dissolved oxygen leading to Fe precipitation. Thus, the shortest column possible was used to minimise these effects. Under the optimal HPLC conditions using a PEEK 50 × 4 mm anion-exchange column and a mobile phase of 40 mmol L⁻¹ DPA, 20 mmol L⁻¹ ammonium nitrate at pH 4.3, the two Fe species were separated in 11 minutes. The LOD for both Fe species was 2 ng g⁻¹. The method was then used with mixed, synthetic, Fe^II and Fe^III standards and also for the determination of the Fe content of a beverage sample. Whilst no loss of Fe, due to precipitation, or carryover was observed, up to 15% of the Fe^II was oxidised to Fe^III during the analytical procedure, thus leading to biased results. Unfortunately for such a comprehensive study, there is no mention of mass bias correction and so the isotope amount ratios and concentration values quoted may be biased. There are few reports of isotope ratio measurements using ICP-MS with CCT mode, which by design increases ion beam collisions with an inert and/or reactive gas, potentially resulting in significant mass bias. Further bias can be introduced if H₂ is used as the cell gas as a consequence of hydride generation. For this reason, it would have been useful for other workers in the field of Fe isotope ratio measurements, if the mass bias data obtained was presented and discussed.

5.8 Halogens

In response to the need for the detection of brominated flame retardants (BFRs), several papers detailing the use of approaches based on GC coupled to atomic spectroscopy, have been published in this ASU period. An analytical procedure based on GC-ICP-MS was developed for determining six BFRs and in particular, polybrominated diphenyl ether (PBDE) congeners in environmental water samples.¹⁰⁸ The PBDEs were extracted from water using iso-octane and Tris–citrate buffer with mechanical shaking for 6 hours. Chromatographic separation was performed using a DB-5MS capillary column (15 m × 0.25 mm, 0.25 μm film thickness) coated with 5% phenyl methyl polysiloxane. Extraction efficiencies of the various PBDEs from spiked water samples were close to 100% and an overall LOD of 0.109 ng L⁻¹ was reported which is low enough to satisfy the EU Water Framework Directive (WFD) requirements for ∑SPBDE. The method was successfully applied to detect PBDE in sea and river water. A method using GC coupled to plasma-assisted reaction chemical ionization mass spectrometry (PAR-CI-MS) has been employed with the aim of overcoming the limitations of ICP-MS for the ionization of halogens¹⁰⁹ when measuring Br containing organic compounds. Separation by GC was performed on a DB-5MS column (30 m × 0.25 mm, 0.25 μm film thickness) with a fused silica capillary transfer line to the PAR-CI-MS. The eluting Br-containing compounds were converted to HBr in a low-pressure He MIP with trace amounts of hydrogen added as a reaction gas. Two critical parameters: reaction gas (H₂) and ionization gas (N₂) were optimised. The results obtained suggest in-plasma reactions followed by electron capture as the main mechanisms for producing Br⁻ from organobromines. Interestingly, a compound-independent response factor was attained which allowed authors to quantify organobromines without the need of using compound-specific standards. A LOD of 29 fg was obtained, which is 4-fold lower than values reported in the literature when using ICP-MS. In another paper¹¹⁰ up to 12 PBDE congeners were determined in airborne particulate matter (PM_2.5) by MAE followed by gel permeation chromatography (GPC) clean-up and determination by a programmed temperature vaporizer (PTV)-GC-MS/MS. Parameters affecting PTV injection and MAE extraction were optimised by means of a central composite experimental design. The optimum MAE conditions for extracting PM_2.5-bound PBDEs were achieved by using 50 mL of hexane : acetone (1 : 1, v/v) as solvent at 75 °C for 2 min. Sample clean-up was undertaken by injecting 700 μL of the extract into the GPC system and further elution with methylene chloride. Under optimised conditions, the method LOD values ranged from 0.063 to 0.210 pg m⁻³ when collecting air volumes of 723 m³. A derivatization protocol consisting of heptafluorobutyric anhydride under triethylamine amine was developed for the simultaneous determination of alkylphenol ethoxylates and BFRs using GC-MS analysis.¹¹¹ The LOD values were estimated to be 0.01–0.20 μg L⁻¹. The improved derivatization protocol allowed authors to successfully detect the target compounds in environmental water samples.

There have been few reports of any note concerning the use of HPLC coupled atomic spectroscopy techniques for the detection of BFRs. Fu et al.¹¹² employed HPLC-ICP-MS for measuring three BFRs decabromodiphenyl ether (BDE209), decabromobiphenyl (BB209) and hexabromobenzene (HBB) in polymers. Samples were extracted with toluene, using UAE, and then precipitated using isooctane and methanol. Chromatographic separation was achieved by RP-LC using a C₁₈ column (250 × 4.6 mm, 5 μm) with isopropanol : methanol : water (60 : 30 : 10) as mobile phase. The effect of polymer type on signal intensity was also evaluated. Data revealed that polystyrene (PS), acrylonitrile-butadiene-styrene copolymer (ABS) and polyethylene (PE) provided signal intensity reduction of 25%, 10% and 3%, respectively. The LOD values were within the range of 0.097–0.1 mg L⁻¹.

A method based on SPE and GC-MS for the determination of 7 emerging halogenated flame retardants (HFR) and PDBE compounds in human serum, have been reported.¹¹³ The HFR compounds tested were: hexabromobenzene (HBB), Dechlorane Plus® (DPs, syn- and anti-isomers), DBDPE, BTBPE, hexachlorocyclopentenyl-dibromocyclooctane (HCDBCO), dechlorane 602 (Dec602) and dechlorane 603 (Dec603). Analytes from spiked human serum samples were collected on Oasis®HLB columns (500 mg) and then eluted with 8 mL of dichloromethane and 1 mL of heptane. Separation was performed on a DB5-MS column (15 m × 0.25 mm, 0.1 μm film thickness). The LOD values were between 0.3 and 5.4 pg mL⁻¹ and spike recoveries ranged between 88–102%. Because of the lipophilic nature of these compounds, the content of triglycerides, phospholipids, total cholesterol and total lipid content was determined in each of the serum samples tested. The method was applied to 10 Norwegian serum samples. Five serum samples were collected from residents in Oslo (Group A) and the other 5 serum samples were collected from residents around Lake Mjøsa (Group B), a lake previously known to be contaminated with PBDEs. Data from Group B provided higher concentrations of PBDEs than Group A, which is likely to be related to consumption of fish from the polluted lake. For the first time, emerging HFRs: hexabromobenzene, 1,2-bis[2,4,6-tribromophenoxy]ethane, Dechlorane Plus®, Dechlorane 602 and 603 were detected in human serum.

The detection of chlorophenols (CPs) in different matrices has been the subject of investigation. Determination of three CPs (2,4-dichlorophenol, 2,4,6-trichlorophenol and pentachlorophenol) in toilet paper was reported by Zhang et al.¹¹⁴ CPs from paper toilet were extracted by UAE using acetone following simultaneous DLLME employing chlorobenzene as extracting solvent. Cholorphenol-containing extracts were further analyzed by GC-ECD on a DB-5MS capillary column (30 m × 0.25 mm, 1 μm film thickness). Under optimal conditions LOD values of 0.25–1.5 μg kg⁻¹ were obtained. Three toilet paper samples spiked at three concentration levels were used for method validation with recoveries ranging from 70.9–118.5%. Pentachlorophenol was the only chlorophenol found in toilet paper and measured at a concentration of 1.2 μg kg⁻¹. The DLLME¹¹⁵ was also employed in another study for extracting CPs in water followed by GC-MS. A funnel shaker was developed to shake centrifuge tubes up and down facilitating emulsification of the water samples during extraction and therefore avoiding the need for using a disperser solvent. Chorophenols from 5 mL of water were extracted and derivatised in less than 1 min by using 1-heptanol (12 μL) as the extraction solvent and acetic anhydride (50 μL) as the derivatization reagent. Separation was performed on a fused silica capillary column (30 m × 0.25 mm, 0.25 μm film thickness). The LODs ranged from 0.02 to 0.1 μg L⁻¹, which are lower than the standard of 0.5 μg L⁻¹ set by European Community legislation. The use of Turboflow C₁₈ column for trapping CPs from human urine samples was described by Guo et al.¹¹⁶ Turboflow columns are novel columns packed with large particles (60 μm) providing an efficient separation of matrix components (e.g. proteins) from smaller molecules when using turbulent flow. Chlorophenols were first trapped on-line in the Turbo flow column, then eluted in back-flush mode and finally determined by LC-UV. The experimental set up allowed the direct introduction of a large volume of sample (1 mL) thus improving sensitivity. The LOD values achieved ranged from 0.5–2.0 μg L⁻¹, as measured by a UV detector and were comparable to those provided by LC-MS/MS, GC-MS, GC-FID and GC-ECD. The method was successfully applied to urine samples collected from 20 volunteers from a hospital, an urban and a rural area.

Directly suspended droplet microextraction (DSDME) combined with GC-MS was evaluated for determining organochlorine pesticide residues in rice.¹¹⁷ Pesticide residues were extracted from rice by using 30% (v/v) acetonitrile and 10 μL toluene as a free suspended droplet. The droplet was retracted into the micro-syringe and injected directly into the GC-MS and separated using a HP-5MS (30 m × 0.32 mm, 0.25 μm film thickness) column. The enrichment factor was reported to be 221–550. The LOD values were 0.0005–0.033 mg kg⁻¹ with an RSD of 2.0–14.0%. The method was applied to spiked urine samples with acceptable recovery values ranged from 72.0–89.6%.

Three papers describing iodinated compound determination by applying different analytical tools have been published in this review period. In the first¹¹⁸ iodine speciation in human urine by RP HPLC-ICP-MS is described. Seven I containing species including: IO₃⁻; I⁻; monoiodotyrosine (MIT); di-iodotyrosine (DIT); tri-iodothyronine (T₃); reversed tri-iodothyronine (rT₃); and thyroxine (T₄) were separated using a C₁₈ column (12.5 × 4.6 mm; 5 μm) with a ternary mobile phase comprising: 4 mmol L⁻¹ TBAH at pH 8.5 containing 0.5 mmol L⁻¹ NH₄Cl, used for eluting IO₃⁻, I; 4 mmol L⁻¹ TBAH at pH 8.5 containing 0.5 mmol L⁻¹L-phenylalanine, for eluting MIT and DIT; and 4 mmol L⁻¹ TBAH at pH 8.5 containing 20 mmol L⁻¹ deoxycholic acid for eluting T₃, rT₃ and T₄. Under these conditions separation of the seven I containing species was achieved in less than 7 minutes. The reported LOD values were in the range 0.046–0.094 mg L⁻¹. The method was validated by performing a spike recovery test using certified reference materials of human urine (GBW09108i, GBW09109g and GBW09110n) for total I with recovery values of 94–106%. The method was then applied for I speciation analysis in urine samples from four volunteers after 2-fold dilution. As expected, I⁻ was found to be the dominant species, showing 96.6–98.4% of total I in the urine samples. In addition, low concentrations of rT₃, T₃ and T₄ were also detected in human urine samples. In the second paper¹¹⁹ a study to detect I-containing proteins in the water insoluble residue of edible seaweed (Nori) by 2D isoelectric focusing SDS-PAGE-LA-ICP-MS has been described. Four reagents: 0.1 M NaOH, 0.1% SDS, 0.2% TritonX-100 and 8 M urea were evaluated for extracting water-insoluble iodinated proteins. Once extracted with urea, I-containing proteins were digested with trypsin and protease XIV and analyzed by SEC-ICP-MS. Data indicated the presence of the iodinated amino acids MIT and DIT suggesting the proteinaceous nature of I in marine algae samples. In a second approach, I-containing proteins were separated by GE and then directly analysed at room temperature by LA-ICP-MS but in this case proteins were extracted with the detergent 3-((3-cholamidopropyl)dimethylammonio)-1-propanesulfonate, followed by LLE with phenol. It was found that I was associated with alkaline proteins with RMM of 10, 20, 28, 40 and 110 kDa in the alkaline (pH 8.0–9.0) isoelectric point (pI) range. Finally, dehalogenation of iodinated aromatic compounds when performing HPLC-ES-MS/MS measurement in both +ve and −ve modes has been reported by Hvattum et al.¹²⁰ The iodinated aromatic compounds were analyzed on a YMC-Pack ODS AM column (150 × 4.6 mm, 5 μm) using gradient elution from 100% water to 50% acetonitrile. The effect of mobile phase additive on the deiodination reaction was evaluated by using formic acid, acetic acid, trifluoroacetic acid, ammonium formate and ammonium acetate as modifiers. Formic acid was the only additive that resulted in the generation of deiodinated compounds, HI and CO₂ as reaction products. The deiodination reaction took place in the ESI capillary and its extent increased with increasing capillary voltage. The authors concluded that formic acid should be used with care when analyzing iodinated aromatic compounds with LC-MS/MS because changes in chemical structure of the compounds may take place.

Methods for the determination of fluorinated compounds have been developed during the current review period. Arvaniti et al.¹²¹ determined 10 perfluoroalkyl carboxylate acids (PFCAs), 5 perfluoroalkyl sulfonates (PFASs), and 3 perfluoroalkyl sulfonamides (PFSAs) in dissolved and particulate phases of wastewater (raw and treated) and in dried sludge by solid–liquid extraction with LC-ES-MS/MS detection. Dried sludge (100 mg) or a filter containing the particulate matter were treated with 1% v/v acetic acid and MeOH for extracting perfluorinated compounds by UAE. Chromatographic separation was performed using a RP C₁₈ (100 × 2.1 mm, 3.5 μm) column and a gradient elution with 5 mM ammonium formate aqueous solution (solvent A) and MeOH (solvent B) as a binary mobile phase. The gradient elution started with 30% (v/v) MeOH and increased linearly to 75% MeOH in 1.5 min, and then to 100% MeOH in 12.0 min which was held for 5.0 min and finally reverted to 30% for a total run time of 30 min. The LOD values ranged from 0.29 to 3.0 ng g⁻¹ and from 0.15 to 1.5 ng g⁻¹ (dw) for dissolved phase, and particulate matter and sludge samples, respectively. It was reported that 11 out of 18 target PFCs were detected either in wastewater (dissolved/particulate phase) or in sludge samples.

Several papers have been published covering the determination of haloacetic acids (HAAS) in drinking water. The first¹²² describes the development of a 2D IC separation with conductimetric detection for determining nine chlorinated and brominated HAAS: monochloroacetic acid (MCAA), dichloroacetic acid (DCAA), trichloroacetic acid (TCAA), monobromoacetic acid (MBAA), dibromoacetic acid (DBAA), bromochloroacetic acid (BCAA), bromodichloroacetic acid (BDCAA), dibromochloroacetic acid (DBCAA), and tribromoacetic acid (TBA). After separation in the first system using a guard column AG24 (50 × 4 mm) and an analytical column AS24 (250 × 4 mm) at 10 °C, fractions of interest were collected and injected into the second system, which used the same columns and conditions as the first. The method allowed separation of the 9 HAAs from each other and from 10 interfering anions. Following optimization, LOD values ranged from 0.3 to 1.7 μg L⁻¹ and recoveries for spiked water samples were in the range of 88 to 119%. The method presents the advantage of simplicity, fast sample preparation and LOD values comparable to GC-MS procedures. Cardador et al.¹²³ have undertaken the determination of chlorinated, brominated and iodinated HAAs (13 species) in water by micro liquid–liquid extraction (MLLE) combined with large volume injection (LVI) coupled to programmed temperature vaporisation (PTV)-GC-MS technique. Three derivatization procedures: aqueous derivatization with DMS and TBA–HSO₄, Fischer esterification, and acidic methanol derivatization in aqueous medium were compared for the methylation of the HAA. The best results were obtained by using the acidic methanol derivatization in aqueous medium following extraction of the resulting methyl esters in MTBE (250 μL). The use of microwave irradiation resulted in a drastically reduced reaction time (2 min versus 2 hours). The reported LOD was 0.05 μg L⁻¹. The developed method was then used for analysing 13 HAAs in treated (chlorinated and chloraminated water) and untreated water. Up to 8 HAAs were found at detectable levels in the treated water. The most remarkable finding was the detection of low levels of mono-iodoacetic acid in chloraminated-treated water. In conjunction with the results the authors concluded that the developed method represents a simple, fast and sensitive alternative to EPA method 552.2 (Determination of Haloacetic Acids in Drinking Water by LLE-GC-ECD) and minimized the generation of hazardous residues in accordance with the principles of “Green Chemistry”. Continuing the interest in developing eco-friendly sample treatment procedures, a vortex-assisted emulsification microextraction (VAEME) procedure followed by in syringe ultrasound-assisted back-microextraction and HPLC-DAD was employed for extracting nine haloacetic acids from water, before HPLC-DAD.¹²⁴ The method utilized 5 mL of water sample containing Na₂SO₄ (45%, w/v) and 600 μL of isopropyl ether as extractant followed by 5 min of vortex to ensure the efficient formation of droplets. Subsequently the isopropyl ether droplet containing the extracted HAAs was collected with a Hamilton syringe and mixed with a low volume (50 μL) of an aqueous solution of (NH₄)₂SO₄ (0.2 M). Finally the syringe was subjected to ultrasound for 5 min and the aqueous solution containing the HAAs was directly injected into the chromatograph. The method LOD values ranged from 1.02–60.1 μg L⁻¹. The method did not provide acceptable recovery values (77.7–89.0%) when applied to spiked water samples and a further development may be needed. In another paper¹²⁵ HAA were selected along with halogenated aromatic compounds to evaluate the performance of LC-IRMS for analysing highly halogenated molecules. Correct isotope values δ¹³C were obtained for mono- and dichlorinated, fluorinated, and tribrominated acetic acids and also for aniline, phenol, benzene, bromobenzene, chlorobenzene, 1,2-dichlorobenzene, 2,4,6-trichlorophenol, pentafluorophenol, and nitrobenzene. In contrast, trichloroacetic acid and trifluororacetic acid lead to an isotopic shift compared to values obtained when using an elemental analyzer, suggesting that these compounds are resistant to persulfate oxidation.

The determination of dehalogenated impurities in halogen-containing pharmaceuticals in a single chromatographic run has been explored.¹²⁶ The method employed a pentafluorophenyl column (2.1 × 50 mm, 1.9 μm) and gradient elution using a binary mobile phase consisting of water, 2 mM ammonium formate, pH 3.5 as solvent A and acetonitrile, 2 mM ammonium formate, pH 3.5 as solvent B. The separated compounds were analysed by HPLC-DAD-ESI-MS. Among the 22 pharmaceuticals tested, only three: the fluorine-containing antibiotic drugs, linezolid, and paroxetine, and the chlorine-containing anticonvulsant drug, lamotrigine, exhibited chromatographic profiles containing dehalogenated impurities.

The isomeric differentiation of halogenated phenylmethylidene hydrazinecarbodithioates (X = F, Cl, Br) was achieved by Ni(II) complexation followed by ES/MS analysis.¹²⁷ The ortho-, meta-, and para-positional isomers were differentiated based on the characteristic product. It is was found in MS³ experiments, that the complex [M_X + SCH₃ + Ni]⁺ ion, resulting from dissociation of the ES-generated complex [2M_X − H + Ni]⁺ ion, underwent ligand-exchange reactions with residual gas molecules, such as water, acetonitrile, and nitrogen in the ion trap while the o-isomers [Mo − X + SCH₃ + Ni]⁺ were found to undergo HX elimination reactions to produce several unique ions. Each set of three isomers [M_X + SCH₃ + Ni]⁺ showed significantly different reactivity which makes the method useful for differentiating isomers. However, the method is not straightforward to implement since costly instrumentation and complicated theoretical calculations are required.

5.9 Mercury

It is evident from the papers covered in this year's ASU review, and those in recent years, that Hg speciation analysis using alkylation and GC as the separation technique, coupled with either AFS or ICP-MS as the detector, can be considered as a mature field with few novel improvements reported. However, for separations involving HPLC the optimisation of various method parameters, including the choice of stationary phase, column dimensions and mobile phase composition and pH continue to be reported. Whether this is due to new workers entering the field and starting method development from scratch, to the nature of Hg speciation by HPLC, or arising from the use of the wide variety of HPLC columns available, or a combination of all three is difficult to discern. This suggests a need for a comprehensive critical review of the topic that gives details of columns and mobile phases used, species separated and retention times and method validation. It is also apparent, from the number of papers recently published on the topic, that IDA applied to Hg speciation studies, once the preserve of national measurement institutes, is now in widespread use in research laboratories worldwide. The next step will be for the wider community to fully embrace the concept of measurement uncertainty, by the use of full uncertainty budgets, for which the technique is ideally suited.

Two research groups have reported on the determination of iHg, MeHg, and EtHg in blood samples by triple spiking head space (HS) SPME-GC-ICP-MS. In the first paper covered here whole blood samples were spiked with enriched spikes of ¹⁹⁹HgCl₂, CH₃²⁰⁰HgCl, and C₂H₅²⁰¹HgCl and Hg species derivatisation was undertaken using NaBPr₄.¹²⁸ The Hg species were solubilised with a 25% TMAH solution in MeOH, added to 100 μL of blood which had been previously spiked with 100 μL of the mixed enriched Hg spike solution, followed by heating in a convection oven at 80 °C for one hour. This was followed by the transfer of 200 μL of the solubilised sample into a 20 mL glass SPME analysis vial and the addition of 7.7 mL of 0.1 M sodium acetate buffer, at pH 4.75, to bring the final pH to between 5 and 6, and 250 μL of 0.2% w/v of NaBPr₄. A commercially available robotic sample processing unit, fitted with two SPME heads, was used to load the volatile Hg species onto the SPME fibres and for subsequent injections onto a diphenyl (5%)-polydimethyl siloxane (95%) capillary GC column (30 m × 0.25 mm, 0.25 μm film thickness). The GC column outlet was inserted, via a heated transfer line at 250 °C, into the ICP torch injector so that the transfer line end was 0.5 to 1 mm from the injector tip. Under these conditions the calculated LOD values for iHg, MeHg, and EtHg were 0.27, 0.12, and 0.16 μg L⁻¹ and the method expanded uncertainties (k = 2) were ± 1.6, ± 0.6, and ± 0.6 μg L⁻¹, respectively. The method was validated by analysing NIST SRM 955c (Caprine blood) level 3 CRM and the results for each Hg species were stated to “fall within the allowable certified limits”. In addition, 28 Centre de Toxicologie du Québec PT scheme human blood samples were analysed, which contain iHg and MeHg, and the results were not statistically different (p value of 1 for a difference plot) from the target values. In an unusual approach, the effect of mass bias correction on the calculated concentrations of Hg species was evaluated by comparing the theoretical and measured Hg isotopic patterns, and not isotope ratios, in the samples analysed. This difference was statistically determined to not differ from zero. As the effect of mass bias correction on the final concentrations was found to be less than 1% it was not corrected for. The paper also contains ESI giving further details of the method optimisation steps undertaken, including solubilisation temperature, spiking levels, SPME sampling and the sample and/or spike stability throughout the analytical procedure. In the second paper the major difference is the extraction protocol.¹²⁹ In this case the Hg species in the blood samples were solubilised, after spiking with a mixed isotopically modified spike solution, with HNO₃ and an MAE protocol. Each blood sample was thawed at room temperature and 7.2 mL of 2 mol L⁻¹ HNO₃ and ‘an appropriate’ amount of each spike solution added to 0.5 g of blood before irradiation at 100 °C for 10 min. After cooling and centrifugation, 2 mL of supernatant was transferred to a glass vial, the pH adjusted to 5.2 with acetate buffer and 0.1 mL of 1% w/v aqueous NaBPr₄ added. After stirring, the volatile Hg species formed were extracted onto the SPME fibre, with the vial in a water bath at 65 °C, for 10 min. Isotope ratio measurements of the Hg species, after GC separation, using the same type of column as detailed in the first paper, were again performed using ICP-MS. In this case, mass bias was conventionally measured and corrected for by the analysis of natural isotopic composition analogues of the three Hg species that underwent the solubilisation and/or derivatisation procedure. In addition, the total Hg content of the blood samples was determined, by IDMS and a mineral acid digestion procedure, for mass balance purposes. A detailed description and discussion of the complex IDMS equations used, as they allow for the correction of any species interconversions caused by the analytical procedure, are also given. The found mass fractions of the total Hg, iHg and MeHg in NIST SRM 996 Level 2 bovine blood were in good agreement with the certified values with a recovery of 100% for the sum of the species detected. The LOD values were as low as 30 pg g⁻¹. For the human blood samples, the results obtained, in which MeHg ranged from not detectable to 25 ng g⁻¹ and the iHg ranged from 0.5 to 31 ng g⁻¹, matched the dietary intake of the study subjects, with, as would be expected, the highest MeHg values being obtained from those who ate fish and/or other seafood products. For all samples analysed, there was statistical agreement, at the 95% confidence level, between the total Hg and sum of species concentrations.

The speciation of Hg in waters has again been reported this year. A method for Hg speciation analysis by on-line preconcentration, using an AE guard column, HPLC-ICP-MS has been developed.¹³⁰ The analytes, Hg⁺, MeHg, EtHg and Hg²⁺ were absorbed onto the AE column, which had been preconditioned with sodium 3-mercapto-1-propanesulfonate, and then rapidly eluted with 2-mercaptoethanol (3% v/v) onto a C₁₈ RP column of 50 mm in length. The Hg species were then eluted with a mobile phase of 0.5% v/v 2-mercaptoethanol and 5 ng mL⁻¹ Bi as internal standard. The HPLC eluent was then directly coupled to the nebuliser of the ICP-MS instrument used. The four Hg species were separated in <5 minutes, in the order Hg⁺, MeHg, EtHg and Hg²⁺ but with degrading peak shape as retention time increased. All solution flows were controlled using three manually operated six port switching valves. Enrichment factors, of 1025 for Hg⁺, 1084 for MeHg, 1108 for EtHg and 1046 for Hg²⁺, were obtained using 6 mL of sample in a 1.5 min enrichment procedure giving a total analytical time of <7 minutes. The LOD values for Hg⁺, MeHg, EtHg and Hg²⁺ were 0.015, 0.010, 0.009 and 0.016 ng L⁻¹ respectively whilst peak height and area RSDs (at 5 ng L⁻¹ for each Hg species) were all <5%. The method was validated using two CRMs GBW(E) 080041 (freshwater) and 080042 (seawater) and the results, expressed as the sum of the species detected, were in good agreement with the certified total Hg value of 1 ng mL⁻¹ for each CRM. The individual Hg species concentrations found also agreed with those reported in the literature. Mercury speciation in three freshwater, two drinking water and two seawater samples was then undertaken using the developed method. MeHg and Hg²⁺ concentrations down to 0.14 and 0.56 ng L⁻¹ were detected in the drinking waters. Somewhat surprisingly, as Hg²⁺ is the usual species present in oxygenated waters, Hg⁺ was detected in a river water sample, at 0.18 ng L⁻¹, and a seawater sample at 1.4 ng L⁻¹. The method shows potential for high throughput analysis if an autosampler and an automated switching valve and pump system were to be used. Workers in Finland determined MeHg in humic-rich water samples by N₂ distillation IDA and online purge and trap GC-ICP-MS.¹³¹ The analytical procedure comprised the equilibration of a CH₃²⁰¹HgCl spike into 50 g of unfiltered sample, preserved with 0.4% HCl, followed by N₂ assisted distillation at 145 °C using an in-house manufactured system capable of distilling 20 samples simultaneously. Subsequently, MeHg was purged from the distillate onto Tenax traps, desorbed at 200 °C and swept with He gas to the diphenyl (5%)-polydimethyl siloxane (95%) capillary GC column (15 × 0.32 mm ID, 0.25 μm film thickness) via a transfer line heated to 60 °C. The GC column outlet was inserted directly into the torch injector with, unusually, no heated transfer line being used. To check for possible methylation during the analytical procedure water samples containing 20, 35 and 50 mg L⁻¹ of DOC were spiked with ²⁰¹HgCl₂ prior to the distillation procedure. The ²⁰¹Hg : ²⁰²Hg isotope ratio for the MeHg peak in the resultant GC chromatogram of 0.45 was comparable with the natural ratio of 0.44. The authors thus state that no significant methylation of iHg occurred. The LOD value obtained for MeHg was 0.05 ng L⁻¹ with single-day precision of <10% RSD. The results obtained for a number of water samples by GC-ID-ICP-MS were also compared with those obtained from parallel samples analysed by GC-AFS (the method details are not given in the paper). The results were found to be in good agreement (n = 28, 95% confidence level) although visual analysis of the data plot suggests this is not the case when samples above a MeHg concentration of 2 ng L⁻¹ are excluded. Two different modifiers, diethyldithiocarbamate and 2-mercaptoethanol, have been compared for the preconcentration of Hg species in water by SPE using a C₁₈ column.¹³² Mercury species in the SPE eluents were determined by HPLC-AFS with minimal detail given about this aspect of the work. The eluent type, pH, chloride ion concentration, humic acid concentration, and storage time were evaluated to compare the preconcentration efficiency. L-Cysteine was employed to elute the mercury compounds. It was found that less eluent was needed for the 2-mercaptoethanol modified SPE compared to the diethyldithiocarbamate version, at an L-cysteine concentration of 0.12%. However, the diethyldithiocarbamate modified SPE could be used over a wider pH range and higher humic acid concentrations, whereas the 2-mercaptoethanol version was less affected by the chloride concentration of the sample waters. Both modified SPE systems were used to store Hg species for 5 days, but the diethyldithiocarbamate modified material could be stored longer. Diethyldithiocarbamate SPE provided LOD values of 3.5, 2.5, and 4 ng L⁻¹ and average recoveries of 91%, 97%, and 85% for iHg, MeHg and EtHg, respectively with RSDs of <6.5%. For the 2-mercaptoethanol modified SPE cartridges, the LOD values were 1.4, 1, and 1.6 ng L⁻¹ and recoveries of 88%, 87% and 91% for iHg, MeHg and EtHg respectively. In this latter case RSDs were higher at 10%. It was concluded that water samples should be examined for their physical and chemical characteristics before Hg preconcentration is attempted so as to choose the most suitable SPE modification method.

The use of diffusive thin film (DFT) devices for the extraction of MeHg from waters has been investigated in two separate studies that reached opposing conclusions. Each study evaluated the use of in-house prepared agarose (AG) and polyacrylamide (PA) gels for the extraction of MeHg from waters. In the first study,¹³³ twelve of each type of gel were separately placed in vessels containing 40 ng L⁻¹ of MeHg (presumably in high purity water but this is not stated), and the DFT devices were withdrawn at regular intervals, along with liquid samples, until 48 h had elapsed. The DFT discs were also subjected to further performance tests including the effect of the diffusion layer thickness and, separately, the effect of ionic strength on the capacity to adsorb MeHg. To extract MeHg from the two DFT gels, two separate reagents, thiourea, at concentrations between 0.5 to 50 mmol L⁻¹, and a solution of 5% H₂SO₄, 18% KBr and 40 mmol L⁻¹ CuSO₄, were employed. The extracted MeHg was then ethylated and quantified using GC-AFS, with full method details given in a cited reference. For the two extraction reagents used it was found that either MeHg was not extracted with thiourea or the ethylation process was interfered with, or both. The diffusion coefficient for the AG gel was 5.1 × 10⁻⁶ cm² s⁻¹, and it was found that this gel was suitable for use at high and low ionic strength, in the range 1 mol L⁻¹ to 0.1 μmol L⁻¹ NaNO₃, and a pH range of 4.0–6.0. It was also found that the PA gel had a much greater affinity for MeHg than the AG gel did whereas it was indicated that PA gel has a high affinity for iHg via a cited reference. Somewhat strangely, the authors then recommended the AG gel for Hg speciation studies. This seems rather a perverse conclusion, unless the aim is to collect only one Hg species and no evidence was given that this was the case for the AG gel. In the second paper¹³⁴ both PA and AG gels were evaluated for MeHg uptake, at a MeHg concentration of 1 μg L⁻¹ in 0.01 M NaCl at pH 7.0, in the presence or absence of natural organic matter (NOM). The authors observed no problems with MeHg ethylation after extraction from the PA and AG gels with 1.3 mmol L⁻¹ thiourea in 0.1 mol L⁻¹ HCl. Once again, quantification of MeHg was undertaken using GC-AFS, with full method details given in cited references. In this case the MeHg diffusion coefficients were found to be higher for the AG gel than for the PG gel in high NOM waters (2.68 × 10⁻⁶ cm² s⁻¹ and 1.69 × 10⁻⁶ cm² s⁻¹, respectively) and in low NOM water (3.15 × 10⁻⁶ cm² s⁻¹ and 2.49 × 10⁻⁶ cm² s⁻¹, respectively). The PG gel was also found to take up a greater amount of MeHg compared with that of the AG gel, 9.2 and 8.5 ng in 24 hours respectively. Therefore, due to this higher uptake rate, the PG gel was recommended for use for monitoring MeHg, particularly in eutrophic waters where biofouling might occur.

Methods for the determination of Hg species in fish tissues have been described in a number of papers. Two papers on this topic, one with Hg species separation using AEC¹³⁵ whilst the second details the use of CEC for the same purpose¹³⁶ have been published. In each case, the column eluent was directly coupled to an ICP-MS instrument for Hg quantification. The AEC separations were achieved using a 12.5 × 4.6 mm column and an isocratic elution using a mobile phase of 1.0 mmol L⁻¹ sodium 3-mercapto-1-propanesulfonate (MPS), at pH 7.0, and a flow rate of 1.5 mL min⁻¹. Four mercury species, Hg²⁺, MeHg, EtHg and PhHg, were separated in this order in under five minutes using these conditions. However, the peak width rapidly degraded with elution time such that the EtHg and Hg²⁺ peak widths were over one minute at a concentration, as Hg, of 5 μg L⁻¹ for each species. This is evident in the LOD values reported, 0.008, 0.024, 0.029 and 0.034 μg L⁻¹ for Hg²⁺, MeHg and total Hg, respectively, which increase with elution time and hence peak width. The method was then applied to the measurement of Hg species in various locally bought fish samples and one CRM following extraction with 5 mol L⁻¹ HCl and 10 mmol L⁻¹ MPS with UAE assistance at 40 °C for 15 min. Concentrations in the CRM, GBW 10029 (which is a fish tissue of an unstated species), were reported as 0.020 ± 0.002, 0.82 ± 0.03, and 0.84 ± 0.03 mg kg⁻¹ for Hg²⁺, MeHg and total Hg, respectively which were comparable to the certified values of 0.84 ± 0.03 mg kg⁻¹ for MeHg and 0.85 ± 0.03 mg kg⁻¹ for total Hg. For the fish sample analysis, the iHg content varied between not detectable to 36 μg kg⁻¹, MeHg was in the range 3–750 μg kg⁻¹ whilst EtHg and PhHg were not detected. The CEC method described by the same group¹³⁶ used 2 SEC columns (12.5 × 4.6 mm) coupled in series and an isocratic elution using a mobile phase of 2 mmol L⁻¹L-cysteine at pH 2.0 and a flow rate of 1.5 mL min⁻¹. This setup separated the same four Hg species, Hg²⁺, MeHg, EtHg and PhHg in that elution order, in under three minutes and with much sharper peaks than the AEC method. Nevertheless, the detection limits reported, 0.019, 0.027, 0.031 and 0.022 μg L⁻¹ for Hg²⁺, MeHg, EtHg and PhHg, respectively, were similar to those for the AEC method. The method was validated by the analysis of CRM GBW 10029, for which the found and certified values were again in good agreement. The method was then applied to locally bought fish samples with similar ranges of Hg²⁺ and MeHg detected as cited above. In addition, five seawater samples were directly analysed without pre-treatment but no Hg species were detectable. Mercury speciation in Cuban fish samples, using double IDMS and HPLC-ICP-MS, has also been reported this year.¹³⁷ The Hg species were extracted from fish muscle, after spiking with CH₃²⁰¹HgCl and ²⁰⁰Hg²⁺, with 0.1% v/v 2-mercaptoethanol (2-ME), 0.1% v/v HCl and 0.05% m/v L-cysteine and UAE for 30 min. After sonication, the samples were centrifuged and the supernatant retained for analysis. The total Hg in the fish samples was determined by ICP-MS after a MAE with HNO₃ and H₂O₂. The optimal HPLC mobile phase was a 0.05% v/v 2 ME, 0.075% m/v L-cysteine and 0.06 mol L⁻¹ ammonium acetate at a flow rate of 0.5 mL min⁻¹, using a RP C₈ column (100 × 4.6 mm) that separated MeHg and Hg²⁺ in 10 min. The LOD values obtained were 18 and 26 ng g⁻¹ for Hg²⁺ and MeHg, respectively. The method was validated by the analysis the CRMs, DOLT-2 and DORM-2 (NRCC Dogfish liver and muscle respectively), for which recoveries of both the total Hg and MeHg content were in good agreement with the certified values. The inter-conversion of MeHg to Hg, and vice versa, was calculated from isotope amount ratio data to be <1.2% for each species in each CRM. For the 13 fish samples analysed, the MeHg content ranged between 0.04 and 0.3 mg kg⁻¹ whilst for iHg the range was from not detectable to 0.062 mg kg⁻¹.

Continuing on the theme of Hg speciation in fish tissues, the performance of ID-GC-MS and ID-GC-ICP-MS has been compared for the measurement of iHg and MeHg in a bivalve CRM, BCR 710 oyster tissue.¹³⁸ The Hg species were extracted from the CRM (0.1 g) using 5 mL of TMAH and MAE at 70 °C for 4 min. After centrifugation and collection of the supernatant, 5 mL of a 0.1 mol L⁻¹ acetate buffer at pH 5.0, and the ¹⁹⁹Hg and CH₃²⁰¹Hg isotopically enriched spike materials suitable to obtain a spike : reference isotope amount ratio of close to unity, were added to 0.1 mL of the CRM extract, the pH was readjusted to 5.0 followed by the addition of 1 mL of hexane and 1 mL of either 1% NaBPr₄ or 1% NaBEt₄. After 5 minutes of vigorous shaking, the hexane phase was collected and stored frozen until analysis. A cross-linked diphenyl (5%)-dimethyl siloxane (95%) (30 m × 0.25 mm, 0.25 μm film thickness) GC column was used for all separations. For detection of the Hg species by ICP-MS the GC column outlet was connected to the ICP torch via a commercially available heated transfer line. Full details of the instrumental conditions and representative chromatograms are provided as ESI. The LOD values obtained for ethylated Hg species were 0.366 and 0.008 pg for MeHg and 0.168 and 0.017 pg for iHg, for GC-MS and GC-ICP-MS methods respectively. For propylated species, the LOD values were 0.079 and 0.005 pg for MeHg and 0.041 and 0.013 pg for iHg, for GC-MS and GC-ICP-MS methods, respectively. Thus, species derivatisation by propylation followed by analysis by GC-ICP-MS gave lower LOD values and, from the RSD values cited, greater precision by a factor of approximately four than GC-MS determinations. However, as the authors state, both methods exhibited overall good performance for the analysis of the CRM. Recoveries for ethylated MeHg were 115 and 92% for GC-MS and GC-ICP-MS, respectively. The method was also used in the determination of Sn species, for which ethylation was preferable. The use of methanesulfonic acid (MSA), which is often used to solubilise amino acids, as an extractant for MeHg from DOLT-4 CRM, has also been reported.¹³⁹ The sample (0.5 g) was spiked with ¹⁹⁸Hg enriched MeHg to give a spike : reference isotope ratio close to unity was placed in a glass flask and 8 mL MSA and 16 mL water were added. The mixture was then refluxed for 16 h, cooled, centrifuged and the supernatant stored at 4 °C until analysis. Subsequently, the MeHg was derivatised with 1% NaBPr₄ and collected onto a PDMS fibre prior to separation and detection by ID-GC-ICP-MS. Once again, the column used was a cross-linked diphenyl (5%)-dimethyl siloxane (95%) (30 m × 0.25 mm, 0.25 μm film thickness) connected to the ICP torch via an in-house fabricated transfer line (further details of this arrangement were given in a cited reference). Mass bias was determined by the analysis of a natural MeHg standard every fourth run as mass bias drift was found to be insignificant. Desorption of the MeHg from the PDMS fibre took place in the GC inlet port heated to 220 °C. The concentration of MeHg found in DOLT-2 was 1.37 ± 0.060 μg g⁻¹, which was in good agreement (t-test, P = 0.05) with the certified value of 1.33 ± 0.12 μg g⁻¹.

The final paper reviewed in this section reports on the determination of Hg species in foodstuffs by HPLC-ICP-MS.¹⁴⁰ A relatively short C₈ RP column, 74 × 4 mm with particles of 3 μm diameter, a mobile phase of 0.02 mol L⁻¹ CH₃COONH₄, 0.2% v/v 2-ME and 1% v/v CH₃OH, at a flow rate of 0.8 mL min⁻¹, was used to separate MeHg, iHg and EtHg. However, the separation of all three analytes took 30 min, which is considerably longer than some other HPLC methods, e.g. those using IEC, described in this section. This increase in separation time can be ascribed to the use of a C₈ stationary phase, which was chosen to allow a lower proportion of MeOH in the mobile phase, compared with that used in some methods using C₁₈ columns, hence removing the need to add O₂ to the ICP-MS plasma gases. The HPLC eluent was connected to a second injection valve to allow post-column calibration standards to be flow injected if required, which, in turn, was connected to the ICP-MS nebuliser via a T-piece that allowed the addition of a Rh internal standard (50 ng L⁻¹ in 0.15 mol L⁻¹ HNO₃ at 0.4 mL min⁻¹). The authors concluded that a 10 mg L⁻¹ iHg solution in dilute HNO₃ was unstable over a period of three weeks when stored in plastic bottles because an additional peak appeared in the chromatograms. This extra peak was found to be Hg⁰, possibly formed from reduction of Hg²⁺ by inorganic or organic impurities leaching from the plastic container. This effect (and/or the loss of Hg through the walls of plastic containers) is well known, but those new to the subject are reminded that pre-cleaned glass containers should be used for storage of solutions of Hg, which is what the researchers subsequently did. Two different approaches to calibration were compared: (1) conventional injection onto the HPLC column and (2) flow injection of increasing amounts, by volume (hence in absolute mass terms), of an MeHg standard via the second injection valve fitted to the system. For this latter approach to work, there needs to be no difference in the instrumental response, given by peak areas, for different Hg species. This was found to be the case for MeHg and iHg by comparison of the slopes of calibration curves from the conventional calibration approach. The advantage of the FI-ICP-MS approach was that a complete calibration could be completed in <20 min. The LOQ values reported, presumably for the FI-ICP-MS approach but this is not stated, for MeHg and iHg were 0.012 ng mL⁻¹ (as Hg) and 0.08 ng mL⁻¹, respectively. The method was validated by the analysis of five CRMS, NRCC DORM-2 dogfish muscle, BCR CRM 422 cod muscle, NIST SRMs 2976 mussel tissue and 1570a spinach leaves and BCR 185 bovine liver. The CRMs, and subsequent locally bought food samples, were subjected to an extraction procedure with 0.2% v/v 2-ME and 1 mol L⁻¹ HCl with shaking at an oscillation frequency of 800 min⁻¹. This procedure was then repeated on the solid residue obtained after centrifugation and the supernatants combined. Finally, the pH of sample extract was adjusted (the final value is not stated) with 2 mol L⁻¹ CH₃COONH₄ immediately prior to analysis. For all of the CRMs except BCR 185 bovine liver the MeHg, iHg and total Hg levels found, given by the sum of the species, mass fractions were in good agreement with the certified values. However, for BCR 185 bovine liver, only 65% of the Hg content was extracted. Subsequently, prior to the above extraction method, this CRM was subjected to acid hydrolysis with 5 mol L⁻¹ HCl at 90 °C for 1 h. This double extraction step resulted in a measured value of 48 ± 6 ng g⁻¹ Hg comparable to the certified value of 44 ± 3 ng g⁻¹ Hg. The method was then applied to various food types, with MeHg, comprising the main Hg content in fish of between 78 and 98%, concentrations ranging between 6 and 250 ng g⁻¹. For cereal and vegetable samples, the total Hg content was generally lower than 15 ng g⁻¹, with the dominant species being iHg except for some spinach samples in which trace amounts of MeHg were also detected. Surprisingly, full details of the results obtained from the foodstuffs are not given in the paper.

5.10 Phosphorus

Several studies to measure organophosphorus pesticides (OPPs) have been reported during this year. Five OPPs in samples of beans were extracted by LLE using dichloromethane as solvent and then determined by GC-FPD on a SPB 35 capillary column (30 m × 0.53 mm, 0.5 μm film thickness).¹⁴¹ The LOD values ranged from 0.24 to 0.85 μg g⁻¹ and recoveries from the spiked samples were of the order of 100% for diazinon, ethion, malathion, methyl parathion and methyl pirimiphos. The method was further applied to determine OPPs in beans samples from different sources. No OPPs were detected in any of the samples analyzed suggesting degradation of OPPs between their use and harvesting of the beans. In an interesting investigation, Nedaei et al.¹⁴² applied for the first time magnetic multi-walled carbon nanotubes (MWCNTs) for extracting OPPs (thionazin, sulfotep, phorate, disulfoton, methyl parathion, malathion, parathion, o,o,o-triethyl phosphorothioate, ethion, famophos, dimethoate, and fenitrothion) from water samples prior to subsequent separation and detection by GC-MS. For optimal extraction efficiency, 10 mg of MWCNTs were added to 10 mL of water sample and then the adsorbent particles were brought to the bottom of the test tube using a magnet. Finally, the analytes were desorbed from the MWCNTs with acetonitrile prior to GC-MS analysis. The method was quite simple to perform providing LOD values in the range of 3–15 ng L⁻¹. The method was validated with environmental water samples and the spiked recoveries were in the range 74–103%. The study clearly shows the potential of nanomaterials as sorbents for determining pesticides. A sample treatment for detecting nine OPPs (ethoprophos, fenitrothion, malathion, chlorpyrifos, isocarbophos, methidathion, fenamiphos, profenofos, and triazophos) in beverage samples has been reported.¹⁴³ The method combines ultrasonication and vortexing for liquid–liquid microextraction prior analysis by GC-FPD. The optimised conditions selected were: 5.0 mL of sample solution, 15 μL of chlorobenzene as the extraction solvent, 3 min of ultrasonication, 3 min of vortexing and 10.0 min of centrifugation. Chromatographic separation was performed on a phenyl (5%)-methylpolysiloxane (95%) capillary GC column (30 m × 0.25 mm, 0.25 μm film thickness). The LOD values obtained were between 0.01–0.05 μg L⁻¹, making this a simple and rapid method capable of determining OPPs in mineralized drinking water, soda water and carbonated beverages. The enrichment factors for the nine OPPs were between 224 and 339.

Methodologies based on HPLC have been used to measure OPP, for example Kazui et al.,¹⁴⁴ carried out simultaneous determination of glyphosate and glufosinato and their hydrolysis products, aminomethylphosphonic acid and 3-methylphosphinicoacetic acid in urine and serum samples by HPLC-ICP-MS. Compounds were reported to be baseline separated in 40 minutes using AEC (250 × 4.6 mm, 9 μm) with a mobile phase consisting of 95% solvent A (0.3 mM NaOH) and 5% solvent B (0.3 mM NaOH, 1 mM Na₂CO₃) in gradient mode. Detection was performed by monitoring PO⁺ (m/z 47) using oxygen as a reaction gas. The method gave LOD values in the range of 0.1–0.7 μg mL⁻¹ and 0.2–1.6 μg mL⁻¹ for serum and urine samples, respectively. The recoveries of spiked targets were satisfactory (91–104%) for both serum and urine samples. The article is of interest as the applicability of HPLC-ICP-MS methodology for determining OPPs is extended to biological and forensic analysis. Lastly, a rapid LC-MS/MS method has been developed¹⁴⁵ for the determination of chlorpyrifos, azinphos methyl and their oxygen analogs chlorpyrifos-oxon and azinphos methyl-oxon on XAD-2 resin and polyurethane foam (PUF), used as active air sampling matrices. Compounds were first extracted by UAE with an acetonitrile solution containing stable-isotope labelled internal standards and then analyzed using LC-MS/MS in multiple reaction monitoring, using a RP C₁₈ column (150 × 2.0 mm, 3 μm). The LOD values ranged from 0.15–1.1 ng for all the compounds on the sampling matrices The spike recovery of the compounds were between 78–113% from XAD-2 active air sampling tubes and 71–108% from PUF tubes. Storage stability provided recoveries ranging from 73–98% after storage-time periods of 2 to 10 months. The method was validated by an interlaboratory study conducted between the California Department of Pesticide Regulation and the University of Washington. Data from the study revealed that the proposed method provided similar results when it was compared to the traditional GC-MS method.

A methodology¹⁴⁶ based on AEC (250 × 4.0 mm, 10 μm) and ICP-MS for simultaneous measurement of the P-containing compounds, orthophosphate and myo-inositol hexakisphosphate (IP6), in plant and soil-related-samples has been published. The developed method offered baseline separation of orthophosphate, P containing impurities, and IP6 within 12 min. Phosphorus compounds were extracted from soils by using two sample extraction procedures. Procedure 1 consisted of placing 2 g soil in 0.25 mol L⁻¹ NaOH and 0.005 mol L⁻¹ EDTA for 2 h at a soil–solution ratio of 1 : 10 and procedure 2 comprised treating 10 g of soil with a 1 mmol L⁻¹ citrate solution over seven days. The HPLC-ICP-MS analysis of the extracts revealed detectable amounts of IP6 when NaOH/EDTA was used whereas for the citrate extracts the IP6 concentration was below the LOD (0.3 μmol L⁻¹). The IP6 was detected in root and shoot acid extracts of Brassica napus, ranging from 10 to 90 mmol kg⁻¹ fresh weight. The use of SPE combined with HPLC-ICP-MS has been reported for the determination of five phosphonates (1-hydroxyethane(1,1-diphosphonic acid)) (nitrilotris(methylene phosphonic acid), ethylenediaminetetra(methylene phosphonic acid)), hexamethylenediaminetetra(methylene phosphonic acid) and diethylenetriaminepenta(methylene phosphonic acid) in natural waters.¹⁴⁷ The main novelty of the article is the transformation of metal complexes of the phosphonates into free phosphonic acids by treating water samples with a strong acid cation exchanger following preconcentration by SPE onto a weak AE IonPac AS16 column (4.0 × 250 mm). The HPLC-ICP-MS method gave LOD values between 0.01 and 0.02 μg L⁻¹ and was applied to environmental water samples and the concentration levels were determined to be within the 0.1–0.2 μg L⁻¹ range. The develop method is superior to others used for phosphonate analysis, especially with respect to sensitivity.

Two reports cover the metabolomic study of P-containing metabolites in biological samples. In the first paper,¹⁴⁸ a study was conducted by AEC coupled to ICP-MS to determine sugar phosphates in yeast cell extracts. A CarboPAC PA1 column (250 × 2.0 mm, 10 μm) was used with gradient elution based on the following mobile phases: solvent A was 50 mM NaOH and solvent B contained 400 mM Na₂CO₃. Once separated, the sugar phosphates entered an anion self-regenerating suppressor (ASRS-ultra II, Dionex) before analysis by ICP-MS. Oxygen was used as a reaction gas for monitoring PO⁺ at m/z 47. Using these conditions, the method was capable of baseline separation for most sugar phosphates and inorganic phosphorus within 50 min. Post-column injection of inorganic P was used for the purity assessment of commercially available sugar phosphate standards. This approach was also successfully applied to the quantification of sugar phosphates in yeast samples (Pichia pastoris). Thus 17 phosphorylated compounds were detected in the cellular ethanolic extracts without the need of using a preconcentration or derivatization step. The method was fully validated by comparing the results of the proposed method with those obtained by an in-house developed LC × LC-ES-MS/MS method utilizing a fully ¹³C-labelled standard for IDA. The comparison revealed a good agreement between the results provided by both procedures which makes the proposed method a valuable analytical tool for the quantification of sugar phosphates in biological samples. In the second paper¹⁴⁹ FIA-ICP-MS and HPLC-ICP-MS were used to statistically differentiate between tumour cell lines derived from breast, lung and colon cancer cells based solely on metabolites that contain P or S. Metabolomic profiling was achieved via RP-LC using a C₁₈ column (100 × 2.1 mm) in gradient mode by using 0.1% (v/v) formic acid as solvent A and MeOH containing 0.1% (v/v) formic acid as solvent B. The total P analysis was more discriminating than the corresponding S analysis. The study is a useful contribution to the on-going work in this area since the proposed ICP-MS based method significantly simplified metabonomic studies as only the target compounds are monitored instead of the complicated profiles obtained when using more conventional metabonomic analytical platforms such as LC-MS/MS.

5.11 Platinum

The nephrotoxicity of cisplatin platinum has always been a serious drawback for its clinical use and this was the main driver for the development of other platinum containing drugs with less serious side effects. In an effort to better understand the mechanisms involved in this process, two papers from a Spanish group^150,151 have investigated the binding of cisplatin, oxaliplatin and carboplatin to proteins present in kidney tissue by developing different gel-based separation systems and LC-MS/MS. A proteomics approach based on 2D GE in both reducing and native forms was developed to separate the three drugs bound to different common proteins. Excision of the bands was then carried out prior to Pt measurement by ICP-MS after acid mineralization or in-gel digestion with trypsin for the identification of the protein using nLC-ESI-LTQ-MS/MS. It was not possible to maintain the stability of the carboplatin–protein conjugates using either form of GE, but a reducing separation was able to maintain the integrity of the oxaliplatin–proteins. The method was based on SDS-PAGE under conditions provided by the thiol-free reducing agent tris(2-carboxyethyl) phosphine, which allowed the separation of oxaliplatin-binding proteins in narrow bands with almost quantitative recoveries. Although off-line in nature, the approach offers some interesting advantages, no less the possibility for development of a generic metallomic approach, mainly because the system maintains the stability of the drug–protein conjugates under investigation. However, it's clear from the analysis of Pt in the bands from the gel that a high Pt background hinders the identification of the protein bands binding Pt; presumably the background results from unbound Pt present in the cellular extracts. In a second paper¹⁵¹ cisplatin–protein complexes were investigated and an intermediate stage developed which allowed further purification of the complex prior to MS analysis. The combined purification-digestion step used a filter-aided separation procedure (FASP), involving a commercially available ultrafiltration device to remove detergent, exchange buffer and as a reactor for protein digestion with trypsin. The integrity of the cisplatin–protein complexes was maintained and the purified peptide–cisplatin complexes were recovered by centrifugation. This methodology was applied to the separation of platinum-enriched protein fractions obtained by SEC-ICP-MS in a kidney tissue extract from a rat treated with cisplatin, followed by further identification by nLC-ESI-LTQ-MS/MS after FASP tryptic digestion of selected platinum-containing liquid fractions.

Analytical studies into the interaction of platinum-containing drugs with proteins in the human circulatory system continue to be undertaken, providing new approaches to a much investigated area, although with little actual clinical relevance in the applications reported. The development¹⁵² of a rapid 2D chromatographic separation of unbound Pt-drugs and complexes with proteins in human serum using monolithic polymer disks, offers a utilitarian approach to Pt-speciation in the clinical setting. Conjoint liquid chromatography using monolithic convective interaction media disks based on affinity and IE chromatography were coupled on-line to UV and ICP-MS detection. The developed system was applied, as is too often the case, to in vitro studies of the interaction of cisplatin, carboplatin and oxaliplatin, with serum proteins and the distribution of Pt in spiked human serum. Presumably subsequent papers will describe the application of this approach to actual patient samples. A second paper in this area¹⁵³ investigated the derivatization of the Pt species with diethyldithiocarbamate prior to analysis by LC-MS/MS. Whilst it is encouraging to see novelty in an oft-visited clinical area, it is difficult to understand how the justification for another method can be that ICP-MS is inferior to ESI-MS/MS because of the potential for polyatomic interference from Hg and the oxides of rare earth elements on Pt, when no evidence is given that these would hamper patient sample analysis. In all GLP compliant laboratories the instrument specification will be such that oxides and polyatomics would be insignificant. One wonders with papers of this nature, which have a good idea but do not include the current state-of-the-art methods for comparison, that the reviewing process did not question the evidence for statements such as ICP-MS is more prone to interference in biological samples than ESI. In reality excellent methods involving ICP-MS are available and provide a significantly better LOD compared to that on offer in this work.

The speciation of inorganic Pt compounds, particularly in relation to the precious metal refining and recycling industry is not often reported on, presumably because of intellectual property issues involved. Speciation of the inorganic Pt(IV) anions, [PtCl₆-nBr_n]²⁻ (n = 0–6) and their corresponding mono-aquated [PtCl₅-nBr_n(H₂O)]⁻ (n = 0–5) anions in dilute solutions has been reported.¹⁵⁴ An ion-pairing separation on a RP column was used to separate the species and ESI-QTOF-MS for their detection. Sufficient chromatographic resolution for the series of Pt^IV complexes was achieved using the (n-butyl)₃NH⁺ ion generated in a formic acid/water/methanol (pH 3.5) mobile phase as the ion-pairing reagent. This mobile phase composition facilitated a low-background for optimal ESI-QTOF-MS detection with enhanced sensitivity and identification of the species from the characteristic m/z pattern of the fragment ions, generated by ‘reductive conversion’ in the ESI source.

5.12 Selenium

Studies on the conversions and transformations of Se in plants by HPLC-ICP-MS and HPLC-ESI-MS/MS have been of interest this year. The speciation of SeMet, MeSeCys and several unknown species¹⁵⁵ in selenate-enriched pakchoi sprouts (Brassica chinensis Jusl var parachinensis) was determined using HPLC-ICP-MS. The optimum separation conditions were an eluent containing 1-butanesulfonic acid (8 mM) and trifluoroacetic acid (4 mM) at pH 4.5 using a flow rate of 1.2 mL min⁻¹ and a RP C₁₈ column (250 × 4.6 mm). The method LOD and LOQ values obtained were lower than 5 and 16 μg Se L⁻¹ respectively for all species. The group of Lobinski¹⁵⁶ continues to investigate Se metabolism in plants by applying a multistep approach combining three chromatographic clean-up steps, ion exchange, reversed phase and ion pairing, and ES-TOF-MS, ES-Orbitrap-MS and ICP-MS detection. This approach was used to unambiguously identify MeSeCys and γ-Glu-MeSeCys in green beans (Phaseolus vulgaris) harvested in a natural seleniferous region of China and the Se-species were quantified by standard addition. Quantification of SeMet and Se^IV was performed by AEC using a PRP-X100 column (250 × 4.1 mm, 10 μm) and detection by ICP-MS. Validation of the results was performed by mass balance and showed that the cumulated amount of Se from the four quantified species represented 93% of the enzymatically extracted Se. This corresponded to only a 72% total recovery of Se, similar to that obtained by other workers. A similar approach was used by Nemeth et al.¹⁵⁷ for determining the selenometabolome of Lecythidaceae seeds (Lecythis minor and Bertholletia excels). Low molecular weight seleno-metabolites from SEC-ICP-MS were collected and analyzed by AE and RP-ICP-MS and RP-ES-QTOF-MS. Species identification was by using different database search strategies, which allowed the identification of new selenocompounds mainly selenohomocysteine (SeHCy) derivatives which open new insights into Se metabolic pathways in plants. The incorporation of Se into rice (Oryza sativa) proteins grown naturally on seleniferous soils was evaluated by a multi-technique approach combining isoelectric focusing SDS-PAGE and LA-ICP-MS for Se-containing protein mapping, and capillary HPLC-ES-Orbitrap-MS/MS for protein identification.¹⁵⁸ The results indicated that 19 kDa globulin and 66 kDa granule-bound starch synthase were the most abundant Se-containing proteins present in the water-insoluble fraction, while only a 19 kDa globulin was detected in the water soluble protein fraction. The most prominent components of the proteins were SeMet and SeCys residues. Ni et al.¹⁵⁹ performed Se speciation by HPLC-ICP-MS and applied this to bamboo shoot extracts grown in Se-rich soil. Isocratic elution using a citric acid mobile phase at pH 4.7 and AEC using a PRP-X100 column (250 × 4.1 mm, 10 μm) were used for Se species separation. Interestingly, trypsin, pepsin, and papain enzymes were used instead of protease for extracting Se. The use of pepsin provided recovery values greater than 80% after 8 h extraction time, whilst maintaining species integrity. The amount of total Se ranged from 45.9 to 216 μg kg⁻¹ and no iSe was detected, with SeMet as the pre-eminent organoselenium species in the extracts. The production of Se-metabolites by garlic (Allium Sativum L) cultivated in a selenite-enriched medium was explored by HPLC-ICP-MS.¹⁶⁰ Optimal separation conditions were achieved using a RP separation C₁₈ column (250 × 4.6 mm, 5 μm) and a buffer of 20 mmol L⁻¹ ammonium acetate containing 5% (v/v) methanol as the mobile phase. Four extractant solutions: 0.1 mol L⁻¹ HCl, 0.1 mol L⁻¹ NaOH, 20 mmol L⁻¹ ammonium acetate–5% methanol and protease XIV were compared and the most suitable results were obtained using 0.1 mol L⁻¹ HCl. As might be expected the main Se species in garlic was γ-Glu-MeSeCys representing 65% of total Se content. The concentration of total Se and selenoamino acids was evaluated in mushrooms (Agaricus bisporus) cultivated in compost irrigated with Se^IV.¹⁶¹ Determination of selenoamino acids was achieved by RP LC-ES-MS on a C₁₈ column (2.1 × 50 mm, 1.8 μm) before enzymatic hydrolysis with protease and derivatization of the resulting amino acids with aminoquinolyl-N-hydroxysuccinimidyl carbamate in order to facilitate Se species retention on the RP column. The maximum level of selenoamino acids was found in caps and stalks with values of 9.65 mg g⁻¹ for SeCys, 0.58 mg g⁻¹ for SeMet, and 0.10 mg g⁻¹ for MeSeCys. The most notable result obtained was the much higher levels of SeCys (detected as SeCys₂), accumulated by A. bisporus, compared to SeMet and MeSeCys which is in discrepancy with the results reported in the literature. The authors attributed these differences to the yield of the different methods used for extracting selenoamino acids.

The investigation of Se speciation in edible oils is of particular interest because of the difficulties relating to the low level of Se being measured. Extraction of the species by SPE using an XAD resin followed by HPLC-ES-MS/MS was used to determine the selenoamino acids, SeMet, MeSeCys and SeCys.¹⁶² Compounds containing Se were first extracted from olive oil using a water : methanol (80 : 20 v/v) solution and then preconcentrated on an XAD resin, followed by elution using 0.5 M formic acid. The resulting eluates were analyzed using a RP C₈ column (2.1 × 50 mm, 1.7 μm) coupled to ES-MS/MS. With a sample volume of 10 mL an enrichment factor of 60 was derived to give an LOD of 0.01 μg kg⁻¹. MeSeCys was the only Se-species detected in olive oil samples with a measured concentration from 2.0 to 8.3 μg kg⁻¹. The second paper¹⁶³ presents a non-chromatographic method for iSe and total Se determination in edible oils (olive, sunflower, avocado, macadamia, fish oil, sesame oil and peanut oil) based on DLLME followed by ETAAS determination. The optimum DLLME conditions for extracting iSe, SeMet, SeCys₂ and SeCM in 10 g of oil were achieved by using 300 μL of a mixture containing diluted nitric acid and isopropyl alcohol as aqueous phase. Meanwhile, for selectively extracting the Se organic compounds the acidified aqueous solution were replaced by an ionic liquid (C12min)(Tf2N). The enrichment factor was reported to be 140 and the LOD was as low as 0.03 μg kg⁻¹ in the original oil sample.

A 2D chromatography approach has been applied to the investigation of Hg–Se antagonism in Water Hyacinth (Eichhornia crassipes).¹⁶⁴ Proteins were screened using a TSK gel G3000_SW SEC column (300 × 7.5 mm, 10 μm) and GE Mono-Q, AEC column (100 × 10.0 mm, 10 μm). In roots, stems and leaves, Se was mainly distributed in two molecular mass fractions of 670 kDa and 40 kDa. However, the proportion between these two fractions was inverted when Hg was co-administered. The AEC HPLC-ICP-MS data revealed that Hg and Se were mostly not associated with the same entity. Specific proteins associated with Hg and Se were identified through trypsin enzymatic digestion, followed by RP nanoLC-ESI/MS and a Mascot database search.

Selenium speciation in biological samples using coupled techniques has been reported. Bay scallop (Argopecten irradians) tissues have been investigated by utilizing HPLC-ICP-MS.¹⁶⁵ Selenium was completely released from the tissues by applying a new enzyme combination consisting of Flavourzyme®500 L, carboxypeptidase Y and trypsin (3 + 2 + 1) after pre-hydrolysis of the tissues by papain in a 1 mol L⁻¹ NaHCO₃ solution at 30–40 °C for 24 h. Selenium species were further separated on an AE hydroxide selective IonPac® AS11 column (4.0 × 250 mm) operating at 30 °C and using 10 mmol L⁻¹ NaHCO₃ with 2% acetonitrile as mobile phase. MeSeCys, SeCys₂, SeMet and Se^IV were detected in the adductor muscle, the mantle and the visceral mass, all three of the anatomical parts tested. The method was validated using SELM-1 a selenized yeast CRM, certified in SeMet. However suitable validation was only achieved when the method incorporated matrix matching with the CRM. In another paper,¹⁶⁶ incorporation of Se in probiotic Lactobacillus reuteri (Lb2 BM-DSM 16143) was evaluated by a multi-technique approach. This combined SDS-PAGE with LA-ICP-MS for mapping the Se-containing proteins, and capillary-HPLC coupled to ESI-Orbitrap-MS/MS for the protein identification. Results demonstrated that all the Se in the bacteria was present as SeCys. One remarkable aspect of this study concerns the investigation on SeCys location in proteins. The SeCys was found in two glycolytic enzymes (glyceraldehyde 3-phosphate dehydrogenase (GAPDH) and pyruvate kinase), two pentose phosphate pathway enzymes (phosphoketolase and 6-phosphogluconate dehydrogenase), two arginine deiminase pathway enzymes (arginine deiminase and ornithine carbamoyltransferase) and a ribonucleoside hydrolase RihC. In some of these enzymes all the cysteine residues reported in the sequence were modified into SeCys while in GAPDH and ADI two out three Cys residues were replaced by SeCys.

The determination of Se species in food and food supplements is once again of interest. Multi-vitamin tablets and two different brands of Se containing tablets purchased in local markets were investigated using RP C₁₈ separation (250 × 4.6 mm, 5 μm) coupled to DRC-ICP-MS.¹⁶⁷ Chromatographic separation was performed in a gradient elution mode using a binary mobile phase consisting of 2 mmol L⁻¹ sodium 1-pentanesulfonate and 5 mmol L⁻¹ citric acid in 3% methanol (pH 2.70) as mobile phase A, and 5 mmol L⁻¹ NaH₂PO₄ and 5 mmol L⁻¹ citric acid in 3% methanol (pH 2.77) as mobile phase B. Selenium compounds were solubilised with a mixture of Protease XIV and mobile phase solution with MAE at 37 °C for 30 min. The species Se^IV, SeMet and γ-Glu-MeSeCys were identified by ES-MS/MS and found in the multi-vitamin sample and in one brand of Se containing tablet, whereas only traces of Se^IV were detected in the other tablet brand. The LOD values for the various Se species were in the range of 0.04–0.07 μg Se L⁻¹. The accuracy of the method was established by comparing the sum of the concentration of individual species obtained by the present procedure with the total concentration of the element. The recovery values were acceptable and ranged from 93–105%. A method to determine Se species in food supplements using HPLC and ES-MS/MS detection has been reported.¹⁶⁸ Chromatographic separation of SeMet, SeCys₂, Se^IV, Se^VI and MeSeCys was achieved under conditions of RP-HPLC on a TSK-Gel ODS-100V column, using a mixture of 5 mM ammonium acetate and methanol as mobile phase. This approach provided LOD values in the range 0.1–3 μg Se L⁻¹. Three sample treatments were tested: UAE with hot water; enzymatic hydrolysis using the non-specific enzyme Protease XIV; and sequential extraction combining water extraction and enzymatic hydrolysis. The highest efficiency was obtained when using sequential extraction. Palomo et al.¹⁶⁹ employed a multi-technique metallomics approach to characterise the Se-containing proteins in Se-enriched yogurt. For this purpose, Se-enriched yogurt samples were first defatted and lyophilised. Proteins were then extracted and separated using asymmetrical flow field-flow fractionation (AF4) and SDS-PAGE utilizing ICP-MS detection. The AF4 was used for protein pre-fractionation before SDS-PAGE with the aim of improving the 1D gel separation of proteins from Se-enriched yoghurt. Protein identification after the post-tryptic digestion of the Se-containing bands was performed by nano-HPLC ES-LT-QMS. The detection of chaperones in the controls, but not in the Se-enriched yogurt was the most remarkable finding. Chaperones are heat-shock proteins expressed in response to elevated temperature or other cellular stresses suggesting that Se could play a role in chaperone expression and therefore in decreasing the stress factor affecting the Lactobacillus. For the amino acids analysis, SeCys was the primary Se-containing species.

Methods for the determination of Se in various body fluids and cultured cells have also been developed in the period covered by this Update. Two papers from Garcia-Sevillano et al. cover simultaneous determination of selenometabolites and selenoproteins in human serum and mice plasma by multi-dimensional chromatography coupled on line to ICP-MS. In the first paper¹⁷⁰ separation of the analytes in plasma from laboratory mice (Mus musculus) and in human serum CRM BCR-637 was performed by in-series stacking of two 5 mL HiTrap®Desalting columns to separate Se^VI, Se^IV, Br⁻ and Cl⁻ from selenoprotein P (SeP), selenoalbumin (SeAlb) and extracellular glutathione peroxidase (eGPx) proteins. This was connected to a dual affinity column arrangement comprising a 1 mL heparin-sepharose column, which selectively retained SeP and a 1 mL blue-sepharose column, which retained both SeP and SelAlb. In the second study¹⁷¹ a new chromatographic dimension using an AE PRP-X100 column (250 × 4.1 mm, 10 μm) was added to the previous experimental set-up to separate Se^VI and Se^IV. The proposed methods allowed quantitative speciation of SeP, eGPx, SeAlb, Se^VI and Se^IV in human serum using suIDMS. The LOD values obtained ranged from 0.2–1.3 ng g⁻¹ for all the detected species. The determination of selenosugars, TMSe, DMSe and DMDSe in urine by using RP C₁₈ column (150 × 4.6 mm) with ICP-MS detection has been reported.¹⁷² An instrumental set-up based on the use of a flow-splitter installed between the HPLC and the ICP-MS combined with a methanol gradient separation was optimised with the aim of reducing the concentration of methanol required to separate DMSe and DMDSe. The method allowed separation of selenosugars and TMSe in presence of DMSe and DMDSe without compromising the stability of the plasma. Plasma stability was tested by repeated measurements (40 repeat injections over 12 hours) of urine samples spiked with 3 selenosugars and TMSe. The retention time and signal intensity precision were reported to be <0.5% and 5% RSD, respectively. The LOQ values reported for each species, 0.25 μg Se L⁻¹ were suitable for application of the method to baseline separate of the species present in urine samples. A study¹⁷³ has been conducted to evaluate transport of the Se species Se^IV, Se^VI SeMet and MeSeCys by using the human colon adenocarcinoma cell line, Caco-2. The main novelty of the study was the use of a Ussing chamber apparatus (used for the measurement of epithelial membrane properties) which consisted of a donor compartment (luminal/apical side) and the receiver compartment (basal/serosal side) each filled with 4.5 mL pre-warmed Hank's Balance Salt Solution, kept at 37 °C, and aerated with carbogen (95% O₂ and 5% CO₂). Cell monolayers were incubated with the four Se compounds at concentrations of 100 and 400 μg L⁻¹. Samples were collected after 40, 80, and 120 min of incubation and analyzed by using an ion exchange IonPac AS7 column (250 × 4.0 mm, 10 μm) coupled to ICP-MS, using 50 mM nitric acid, 1% (v/v) methanol, pH 1.1 as mobile phase. Transport efficiencies were dependent on the chemical form of Se and type of application (separate or simultaneous) with the highest values for SeMet (9.1%) and Se^VI (7.9%) when adding Se chemical forms separately or simultaneously, respectively. In other study¹⁷⁴ Se-containing proteins were isolated from HEK 293 kidney cells incubated with 100 μM SeMet for 24 h. Proteins were screened by using a 2D chromatography method involving SEC (300 × 7.5 mm, 10 μm) and a RP C₄ column (150 × 0.50 mm, 5 μm). Selenium-containing RP fractions were digested with trypsin for identification using HPLC-Chip-ES-IT-MS. Numerous proteins were found to contain SeMet, suggesting SeMet is not specifically incorporated into proteins. In another study,¹⁷⁵ a method based on the direct detection of selenoproteins using iso-electrofocusing (IEF) electrophoretic strips and LA-ICP-MS was developed. The method was applied to different cultured human cells (Hek293 (kidney), HepG2 (liver), HaCaT (skin) and LNCaP (prostate)). The best performance for the separation of proteins by IEF was achieved using 500 μg of protein and 3–10 pI strips as well as a methanol/chloroform procedure for protein precipitation. Subsequently, Se-containing proteins in the IEF strips were digested with trypsin for analysis by ICP-MS and ES-Orbitrap MS/MS and Mascot database search. The LOD for protein detection was 1.0 μg Se L⁻¹. The principal proteins found were selenoprotein 15 (SeP15), glutathione peroxidise 1 (GPx1) and glutathione peroxidase 4 (GPx4), thioredoxin reductase 1 (TRxR1) and thioredoxin reductase 2 (TRxR2).

The number of papers focused on Se speciation in environmental samples has been scarce this year. Zhang et al.¹⁷⁶ described a novel method for the speciation of iSe in water based on DLLME followed by ETV-ICP-MS. Selective extraction of Se^IV by using 500 μL ethanol as the dispersion solvent, containing 70 μL chloroform as the extraction solvent and 0.2 g L⁻¹ Bismuthiol II as the chelating reagent, at pH 2.0. Additionally, the use of Bismuthiol II favoured Se vaporization in the instrumental set up. Total Se was determined after the reduction of Se^VI to Se ^IV by treatment with 5 mol L⁻¹ HCl for 50 min in a water bath. The Se^VI concentration was obtained by subtracting Se^IV from total Se concentration. This approach provided an LOD for Se of 0.047 μg L⁻¹ and an enrichment factor of 64.8 for 5 mL of water sample. The method was validated using a CRM of environmental water GBW(E)080395 and good agreement was shown between the reported value of 1.0357 ± 0.015 μg L⁻¹ Se^IV and the certified value 1.0007 ± 0.09 μg L⁻¹ Se^IV. The developed method was applied to Se speciation in spring and lake waters achieving recovery values close to 100%. In another study¹⁷⁷ SF-ICP-MS combined with a laboratory made continuous FI-HG-AAS allowed authors to determine inorganic Se in natural waters at low levels. The precision of Se determination in river water and seawater was better than 4%, and the spiked recovery ranged from 97% to 103%. The applicability of solid phase extraction (SPE) for the speciation and preconcentration of iSe in water matrices has been described.²⁴ The paper summarized recently published data on the use of different SPE strategies and sorbents. The authors highlighted the use of new materials such as nano-sized carbon derivatives, nano-tubes, nanoparticles and graphene for iSe speciation. A good demonstration of the usefulness of nanomaterials as sorbents for iSe preconcentration was given by Kim et al.¹⁷⁸ Thus Se^IV was selectively extracted from natural waters by using TiO₂@SiO₂/Fe₃O₄ nanoparticles as sorbents. The synthesized nanoparticles had two noticeable capabilities: photocatalytic reduction of Se cations to Se⁰ atoms by the TiO₂ shell with UV irradiation and an attractive superparamagnetic force for particle collection by the Fe₃O₄ core. The separation efficiency was estimated by determining the remaining Se in the sample solution and the Se on the nanoparticles. It was found that the photocatalytic reduction was dependent on the power of the UV radiation source and the reaction time. Based on that finding, Se^IV was separated from Se^VI and organic Se by photocatalytic reaction with TiO₂ in presence of UV at a low radiation power of 6 watts for 3 hours. The capability of TiO₂ NPs to induce photocatalytic reduction of iSe into gaseous products was also explored by Shih et al.¹⁷⁹ A microfluidic-based VG system was employed in conjunction with HPLC-ICP-MS. High optical grade poly(methyl methacrylate) (PMMA) substrate was used to fabricate the TiO₂-coated microfluidic-based photocatalyst-assisted reduction device with dimensions 619 μm (W) × 784 μm (D) × 154 cm (L) and a void volume of 317 μL. Both Se^IV and Se^VI were first separated by AE using a PRP-X100 column (250 × 4.1 mm, 10 μm) and then vaporized in the microfluidic-based VG system and detected in the ICP-MS. Both Se^IV and Se^VI were efficiently vaporised within 15 s at pH 3.0 in the presence of 25 mmol L⁻¹ HCOOH. The LOD values were reported to be 0.043 and 0.042 μg L⁻¹ for Se^IV and Se^VI, respectively. The applicability of the method was examined by determining iSe species in the CRM NIST 1643e, artificial saline solution and fortified irrigation water with recovery values close to 100%. The Se species that naturally existed in the irrigation water were also clearly identified with the developed system even though the concentration of iSe was below the μg L⁻¹ level. This approach was also used for determining Se^IV, Se^VI and SeMet in water, urine and yeast extract samples.¹⁸⁰ Using 1 g of nano-TiO₂ per litre, at pH 3.0, and illuminating for 80 s, Se species were all quantitatively converted into volatile Se products. The LOD values of 3.9, 7.2 and 8.3 ng L⁻¹ were obtained for Se^IV, Se^VI and SeMet, respectively. Acceptable recovery values (85–105%) were obtained for NIST SRM 1643e (trace elements in water), yeast extracts and urine samples.

In the context of developing new methodologies for Se speciation it is worth noting the appearance of three papers dealing with chiral speciation. The state of the art of chiral speciation of selenoamino acids has been reviewed.²⁵ A brief description of the use of indirect and direct enantioseparation by using HPLC, GC and CE coupled on-line with element specific detectors (mainly ICP-MS) was provided in the first section. The second part of the review describes the applicability of the technique to biological samples such as food supplements, biological fluids and bioavailability studies. The final section contains a perspective on future developments such as chiral LC columns and CE for the simultaneous separation of various selenoamino acid enantiomers. The determination of enantiomeric forms of SeMet by RP HPLC using an indirect procedure was examined in the second of these papers.¹⁸¹ For this purpose the synthesis of a new chiral derivatizing reagent, phthalimidyl-(S)-naproxen ester for producing SeMet diastereomes via microwave irradiation or vortexing was undertaken. Diastereomers were successfully resolved by RP HPLC using a C₁₈ column (250 × 4.6 mm, 5 μm) coupled to a UV detector (231 nm) and by using a mobile phase consisting of aqueous triethylammonium phosphate (10 mM, pH 3.4) and acetonitrile in a linear gradient of 25 to 65%. Under these conditions analytes were separated in 35 minutes with D-SeMet eluting earlier than the corresponding L-isomer. This approach offered a LOD of 0.10 pmol mL⁻¹ and a linear range up to 14 pmol mL⁻¹. The final paper¹⁸² contained a report on enantioresolution of DL-SeMet. For this purpose, different experimental set-ups including: TLC separation of DL-SeMet using (−)quinine as chiral detector; and RP TLC and RP HPLC separation using three difluorodinitrobenzene based reagents having L-amino acids (L-phenylalanine, S-benzyl-L-cysteine and L-methionine) for synthesizing chiral derivatizing reagents. The LODs were at the level of ng L⁻¹, μg L⁻¹ and mg L⁻¹, for RP-HPLC, RP TLC and direct TLC, respectively. The best performance in terms of chromatographic resolution was attained by RP HPLC.

5.13 Silicon

A GC-ICP-MS method has been described for Si speciation in light petroleum products.¹⁸³ Silicon compounds can severely degrade catalysts used in the processing of this oil fraction. In part, Si species in oils arise from the breakdown of polydimethylsiloxane, which is used as an agent to increase oil recovery or as an anti-foaming agent during refining stages. A DB-5 GC column and multi-stage temperature program were used to separate Si species during the optimisation and sample analysis steps in the study. The GC carrier gas was He and the column was connected to the ICP-MS instrument via an 80 cm transfer line heated at 250 °C. The column was inserted into the central channel of the ICP torch and the flow combined with a make-up gas flow of Ar, O₂ and Xe. The final position of the transfer line outlet in the torch was optimised by monitoring the ¹²⁸Xe signal arising from the make-up gas. To minimise the effects of polyatomic interferences (e.g.¹²C¹⁶O and ¹⁴N₂ on ²⁸Si) H₂ was used as a reaction gas, at 2 mL min⁻¹, in a DRC. Hydrides were formed with the interfering ions but Si was unaffected as it is unreactive towards H₂. Under the optimised GC and DRC conditions, 19 different Si compounds were baseline separated. The LOD values, based on six standard deviations of the background, ranged from 24 to 136 μg kg⁻¹ as Si. These values are higher than those reported for GC-GC-TOF-MS analysis, of 5–33 μg kg⁻¹ as Si, which the authors attributed to the use of a quartz torch. The method was linear from 50–1000 μg kg⁻¹ of Si compounds in acetone with compound-dependent response factors, as shown by the slope of the calibration curves, also being observed. The possible causes of the compound-dependent responses, such as the partial loss of polar compounds to the transfer line, were discussed and compared with other observations reported in the literature. It was recommended that the transfer line temperature was raised to 300 °C as a result of this investigation. Analysis of thermally degraded PDMS by GC-ICP-MS revealed the occurrence of a wide variety of previously undetected compounds. Using a combination of retention time matching, GC-TOF-MS, SIM-GC-MS and ES-FT-ICR-MS, more than 10 new Si species were characterised. The new species contained various oxygen-based functional groups, which would easily react with the alumina catalyst supports, and the study has therefore provided an insight into catalytic degradation processes which need to be addressed.

5.14 Tin

There have been a small number of publications within this review period focusing on Sn speciation. A method for the determination of MBT, DBT, TBT, MPhT and DPhT, in seawater using SPME coupled with GC-MS/MS after ethylation with NaBEt₄ has been reported.¹⁸⁴ An internal standard (TPrT) was added to the samples and the pH was adjusted to 5.0 using a buffer. The seawater samples were then stirred with a magnetic agitator and headspace SPME performed with a 75 μm polydimethylsiloxane/carboxen (PDMS/CAR) fibre for 20 min. After extraction, the SPME fibre was directly injected into the GC inlet for GC-MS/MS analysis. The method offered LOD values for the five OTC in the range of 0.10–0.20 ng L⁻¹ (Sn). The recoveries were in the range of 80.2–93.6%, and the RSD values were all below 13% for the five OTC at the spike level of 20 ng L⁻¹ (Sn). The proposed method was applied to 96 seawater samples collected from mariculture areas along the Bohai Bay in Yellow Sea NE China. Coscolla et al.¹⁸⁵ have also used headspace SPME with GC-MS/MS for the determination of MBT, DBT, TBT, MPhT, DPhT, and TPhT in a range of surface and marine water samples following statistical design experiments. The main experimental parameters influencing the extraction efficiency were pre-incubation time, incubation temperature, agitator speed, extraction time, desorption temperature, buffer (pH, concentration and volume), headspace volume, sample salinity, preparation of standards, ultrasonic time and desorption time in the injector. The main factors affecting the GC-IT-MS/MS response included, excitation voltage, excitation time, ion source temperature, isolation time and electron energy and were also optimised using the same experimental design. The proposed method offered reasonable linearity (correlation coefficient r > 0.99) and somewhat variable repeatability (1–25%) for the compounds under study. Percentage recovery for the compounds in spiked surface and marine waters was >70% for all compounds studied. Headspace SPME followed by GC-FPD analysis has also been used for the simultaneous determination of 11 OTC, including methyl-, butyl-, phenyl- and octyltin derivatives, in human urine.¹⁸⁶ Samples were spiked at two concentration levels, 8–100 ng Sn L⁻¹ and 15–200 ng Sn L⁻¹. An experimental design approach was used to model the storage stability of OTC in human urine, demonstrating that the organotins were highly degraded in this medium, although their stability was satisfactory during the first 4 days of storage at 4 °C and pH 4.0. Finally, this methodology was applied to urine samples collected from harbour workers exposed to antifouling paints and methyl- and butyltins were detected.

A number of different approaches have been applied to Sn speciation in water samples. The use of DLLME in combination with an in situ derivatization has been reported for butyltin compounds in water samples.¹⁸⁷ The derivatization was carried out with sodium tetraethylborate at pH 4.5. The effects of extraction and dispersion solvent type, volume, extraction time and ionic strength of the solution on the extraction efficiency were investigated. Tetrachloromethane containing n-hexadecane as an internal standard was used as an extracting solvent and methanol was used as the dispersion solvent. Calibration graphs were linear from 2.8, 4.2 and 9.8 ng L⁻¹ up to 10 μg L⁻¹ for MBT, DBT and TBT, respectively (correlation coefficients were 0.996–0.999). The LOD values were 1.7, 2.5 and 5.9 ng L⁻¹ for MBT, DBT and TBT, respectively. The application of LC-MS-IT-TOF for the speciation of TPrT, TBT, TPhT and tripentyltin (TPeT) has been evaluated for aqueous solutions.¹⁸⁸ Various chromatographic parameters such as choice of chromatographic column, mobile phase, flow rate and gradient program were optimised, and a RP C₁₈ bonded-silica column with a mobile phase of water, methanol and acetic acid at a flow rate of 0.4 mL min⁻¹ was selected. The LOD values for the above species ranged between 13–45 pg as Sn with RSDs between 4.3 and 9.9%. The ICP-MS and ESI-MS methods have been used as complementary approaches to identify Sn–pentaacetic acid complex formation.¹⁸⁹ Thus ESI-MS was used to initially confirm the formation of (Sn(DTPA))³⁻ and (Sn(DTPA))¹⁻ and their MS spectra suggested that these Sn complexes were stable in solution. On-column complexation of Sn with DTPA and the separation of (Sn(DTPA))³⁻ and (Sn(DTPA))¹⁻ were performed with AEC with a mobile phase containing 20 mM NH₄NO₃ and 3 mM DTPA at pH 6.0, and the subsequent detection of (Sn(DTPA))³⁻ and (Sn(DTPA))¹⁻ was achieved by ICP-MS. Linear plots were obtained in a concentration range of 1.0–1000 μg L⁻¹ with LOD figures of merit ranging from 0.1 to 0.3 μg L⁻¹. Simultaneous determination of (Sn(DTPA))³⁻ and (Sn(DTPA))¹⁻ was possible in less than 5 min with a RSD between 2.1 and 2.7%. The recoveries of (Sn(DTPA))³⁻ and (Sn(DTPA))¹⁻ were found to be 96.8 and 99.4%, respectively, when the sample was spiked with 20 μg L⁻¹ standard. The proposed procedure was used for the determination of Sn species in contaminated water.

The use of isotopically enriched Sn tracers to follow the transformation of OTC in landfill leachate over a six month period has been reported.¹⁹⁰ The degradation and biomethylation processes were followed using isotopically enriched Sn tracers: ¹¹⁷Sn-enriched tributyltin (TBT), ¹¹⁹Sn enriched DBT, ¹¹⁷Sn-enriched SnCl₂, ¹¹⁷Sn-enriched SnCl₄ and a ¹¹⁹Sn-enriched butyltin mix containing TBT, DBT and MBT. Transformation of OTCs in the spiked leachates was followed at m/z of the enriched spikes and at m/z 120, which allowed simultaneous observation of the transformation of OTCs in the leachate itself and of the added spike. In parallel, these processes were also monitored in a non-spiked leachate sample at m/z 120. Quantification of OTCs was performed by GC-ICP-MS. To discriminate between the biotic and abiotic transformations of OTCs and iSn species, sterilization of leachate was also performed and data compared to non-sterilized samples. During the course of the experiment the microbial degradation of TBT was clearly manifested in Sn-enriched spiked leachate samples, while abiotic pathway of degradation was observed for DBT. Biomethylation process were also observed in the leachate spiked with Sn-enriched Sn²⁺ or Sn⁴⁺, in concentrations close to those found for total Sn in landfill leachates. MMT was formed first and then stepwise alkylation resulted in DMT and TMT species formation. Hydrolysis of Sn²⁺ and Sn⁴⁺ species were found to be a limiting factor which controlled the extent of methyltin formation. The same research group have also reported the micro-scale synthesis of ¹¹⁷Sn-enriched TBT chloride and its characterization by GC-ICP-MS and NMR techniques.¹⁹¹ The synthetic pathway started with bromination of metallic Sn, followed by butylation with butyl lithium. The TBT formed was transformed to TBT chloride using concentrated HCI with a yield of around 60%. The purity of the synthesized TBT was verified using GC-ICP-MS and NMR spectroscopy. The results showed that TBT had a purity of more than 97% (the remaining 3% corresponded to DBT). The TBT was quantified by reverse ID GC-ICP-MS. The advantage of this procedure over those previously reported is its application on a micro-scale, starting with 10 mg of enriched metallic Sn.

Five OTC (DMT, DBT, DPhT, TPhT and TBT) have been determined in wines from China using MAE and HPLC-ICP-MS.¹⁹² The extraction, using n-hexane with 0.02% tropolone, was optimised with respect to the extraction temperature and extraction time by an orthogonal array experimental design. The procedure offered LOD values between 0.029 and 0.049 μg L⁻¹. The linearity was in the range 0.5–100 μg L⁻¹ and the precision was below 9.43% RSD. The method detected the OTC at levels of 0.053–1.14 μg L⁻¹ in the wine, with confirmation by HPLC-MS/MS.

5.15 Vanadium

In a comprehensive study, vanadium speciation in human liver cells, after in vitro exposure to bis(maltolato)oxovanadium(IV) (BMOV), a potential anti-diabetic drug, has been reported.¹⁹³ Cells (HepG2) were grown in Eagle's essential minimum medium (EMEM), supplemented with 10% foetal bovine serum (FBS), or Hank's Balanced Salt Solution both with and without BMOV. Subsequently, after cell lysis and ultrafiltration total V concentrations were determined in the lysate fractions by ICP-MS using He as the collision cell gas. In addition, the cell lysates were size fractionated on a gel permeation column, with 10 mmol L⁻¹ ammonium acetate as the eluent, and the eluent analysed online and in parallel by ICP-MS and ES-MS. Two internal standards, Rh and Y, were added online to the flow stream for ICP-MS whilst the flow stream for ES-MS was split further. This configuration resulted in a slight offset for the retention time, of no more than 0.2 minutes, for the measurements by ES-MS. For on-line quantification of V species by ICP-MS, V standards were introduced along with the internal standard solution and concentrations of the individual species calculated from the resultant mass flow chromatograms. On line accurate mass measurements were performed using SEC-ES-QTOF and ES-MS-MS. In the latter case the identification of BMOV was achieved using a synthesised ⁵⁰V enriched standard. The obtained MS/MS data for the main V compound detected was consistent with the presence of a divanadate-phosphate derivate that included a carbohydrate or polycarboxylic acid substructure.

5.16 Zinc

A complete analytical protocol for Zn speciation in leaf tissues of Plantago lanceolata, collected from a mining waste site, has been reported.¹⁹⁴ In order to separate the Zn compounds, both SEC and IC were used in direct sequential and reverse sequential modes followed by detection of Zn species by ICP-MS. In the direct sequential mode, the entire tissue extract underwent SEC separation, with 0.025 mol L⁻¹ sodium acetate at a flow rate of 0.35 mL min⁻¹ as the mobile phase, and individual fractions, as detected by UV, were collected and then further separated with IC. In the reverse sequential mode, an aliquot of the entire tissue extract was injected onto the IC column. Species were eluted with an initial mobile phase, of 0.005 mol L⁻¹ acetate buffer at pH 6, which changed to 0.1 mol L⁻¹ acetate buffer at pH 4.7. Individual fractions collected were then injected onto the size-exclusion column. For the subsequent analyses of the fractions collected, a HPLC-UV-ICP-MS system was used, with the eluent from the UV detector coupled directly to the nebuliser of an ICP-MS instrument. Zinc species were extracted from shredded leaf tissue using a combination of sodium acetate, sodium bicarbonate and Tris–HCl buffers and mechanical shaking for 1 h at 4 °C followed by centrifugation and supernatant collection. Using this procedure, four Zn species were detected in the high molecular mass fraction, and three in the low molecular mass fractions. The authors noted that the number of compounds detected was strongly dependent on the extractant pH and that further work would be needed to identify the Zn species by molecular mass spectrometry.

6 Biomolecular analysis

6.1 Direct macromolecular analysis

The measurement of RNA is widely performed in genetic testing, microbial identification and disease biomarker studies, usually by employing molecular biology techniques. However, the separation and quantification of RNA using SEC-ICP-MS has been investigated.¹⁹⁵ This approach was applied to the analysis of commercially available RNA ladder marker solutions and a CRM containing single-strand 533 and 1033 base pair RNA fragments (NMIJ CRM 6204-a). A silica based diol SEC column derivatized with 1,2-dihydroxypropane was employed for the RNA separations, using a Tris–HCl buffer at pH 8 and 40 °C, conditions under which the silica will be significantly unstable. The RNA was quantified by measuring 32P⁺ using a collision cell ICP-MS instrument, although the cell gas and conditions are unfortunately not reported. Inorganic phosphate was used as the P calibration standard with Se added as an internal standard. Interestingly the phosphate eluted from the SEC column with a retention time of 17 min and Se at 18 min, whereas the 500 and 1000 base pair RNA had a retention time of 10 min which presumably was the exclusion volume for this SEC column. The LOD and absolute detection limit of P were 0.33 μg kg⁻¹ and 3.3 pg (equivalent to 37 pg of RNA) respectively and RNA was quantified with a repeatability of 2.7% RSD and the accuracy, as shown by measurement of RNA in the CRM, was in agreement with the certified value. Analysis of the RNA ladder solution showed good separation of fragments in the range 50 to 1000 base pairs.

6.2 Metalloproteins, metalloproteomics and metallomics

The separation of metalloproteins by HPLC-ICP-MS and quantification using different calibration methods has been the focus of recent work by a number of groups. This provides an in-sight into the advantages and disadvantages of different calibration approaches, including, external calibration, suIDMS and ssIDMS. A SEC-ICP-MS method¹⁹⁶ for profiling the metal-binding (Cu, Fe and Zn) proteins in primary cultures of neurons and astrocytes, used a simple external calibration approach. Ferritin was used as the metalloprotein standard to quantify the Fe containing proteins, whilst for those containing Cu and Zn, superoxide dismutase-1 (SOD-1) was used as the calibrator. Standards in the range 6–150 ng of Fe and 0.6–15 ng Cu and Zn produced calibration plots after linear regression analysis with correlation coefficients >0.999. The limits of detection (3σ) were 0.825 ng for Fe, 130 ng for Cu and 13.6 ng for Zn. The repeatability of responses for a 20 ng Fe (as ferritin) and 2 ng Cu and Zn (both as SOD-1) produced coefficients of variation of 1.3%, 7.3% and 2.4% for 4 injections. The column was size calibrated mn using a range of metalloproteins including: ferritin (450 kDa), ceruloplasmin (151 kDa), conalbumin (75 kDa), SOD-1 (32 kDa (dimer)) and metallothionein IIa (6.5 kDa). This approach produced a very linear inverse relationship, which was used to estimate the size of unknown proteins in the samples analysed. The methodology is really aimed at high through-put, rapid screening (<15 min) of metalloproteins, using a similar sample size (20–150 μg protein) as that used in the conventional Western blotting procedure. Clearly the main drawback of the method is the poor resolving power provided by SEC, which limits the facility to identify the proteins present, particularly low abundance metalloproteins, which may be of most interest. The lack of any attempt to identify the peaks in the elemental chromatograms by molecular mass spectrometry or any evidence relating to the purity of the peaks in the chromatograms presented, would seem to be a significant omission. It would be fairly straight-forward to determine the purity of the peaks by fraction collection and then GE separations, which would provide greater resolving power than SEC. Researchers at the National Institute of Metrology in Beijing have reported methods^197,198 for the high accuracy measurement of transferrin (Tf) and albumin (Alb) in human serum, using AEC, which provides greater resolution than SEC. Post-column addition of the enriched isotopes ³⁴S and ⁵⁴Fe was employed to allow IDMS measurement of the proteins in species-unspecific mode. The methodology involved pre-saturation of Tf with naturally abundant Fe prior to anion exchange separation from Alb and then detection using SF-ICP-MS in the medium resolution mode (m/Δm = 4000). To allow for the significant difference in concentration of the two proteins and to achieve a suitable isotope ratio, the flow rate of the spike solution was increased from 0.02 to 0.12 mL min⁻¹ after 15 min to coincide with the elution of the much more abundant Alb. The methodology was validated using a human serum protein CRM (ERM-DA470/IFCC), achieving good agreement with the certificate value for both Tf and Alb, with the results for Tf using S and Fe in close agreement. Clearly this approach has significant applications in providing primary methods for measurement of these proteins in clinical chemistry. The use of S as the measurand could potentially be applied to other proteins with a suitable number of cysteine residues. Ordonez et al.¹⁹⁹ have revisited the measurement of Tf and carbohydrate-deficient Tf (a biomarker for congenital disorders of glycosylation) and chronic alcohol consumption, in clinical samples. The use of a Dionex ProPac SAX-10 (250 × 4.0 mm) anion exchange column showed some improvement in separation of the five different sialoforms of Tf. Each form was measured in two CRMs including human serum protein certified for individual proteins (ERM-DA470/IFCC) and frozen human serum certified for total protein (SRM 909c). After Fe saturation of the solution both suIDMS and ssIDMS procedures using SF-ICP-MS detection, generated accurate results in agreement with the certified values for the ERM material.

Three studies by different groups with an interest in the speciation of metal-containing proteins in seeds and nuts illustrate the range of analytical approaches available for metallomic studies have been reported. In the first of these²⁰⁰ a combination of SEC-ICP-MS, tryptic digestion and MALDI-TOF-MS were used to look at the metalloproteome of Cu in brazil nuts. The authors make the point that the mass spectrum for any metal containing peptide will be identifiable because the isotopic composition of the metal, (with the exception obviously, of monoisotopic elements), will affect the isotopic pattern of the peptide. Whilst this is true, in practice the metal centre in the protein must be stable under the enzymatic digestion conditions used but also the laser desorption process. A Cu-containing protein was identified from an apo-peptide present at m/z 1375.21, with a corresponding metallated form found at m/z 1438.12. The mass difference of 62.9 Da between the two signals corresponded to the mass of the most abundant (62.9%) Cu isotope. Thus a peptide containing CuI was identified, although it is not clear if Cu is present in the protein in this oxidation state. The identification of the oxidation state appears premature without further confirmation of the actual protein binding site, which was not presented. The detailed mass spectrum for the metallated cluster ions showed the presence of both Cu isotopes, but this was not discussed in the paper. It is clear from this report that such high accuracy MALDI-TOF-MS work will have some application in metallomics. However because of the need for protein digestion and the unknown stability of the metalloprotein in the laser desorption process it is clearly not a panacea. It is difficult therefore to see how the methodology could be used as a generic approach, as it is in conventional proteomics. In what appears to be a good basis for a generic metalloproteomic protocol,²⁰¹ the second approach uses immobilized metal affinity chromatography (IMEC) to “fish-out” proteins with a Cu binding motif. These are then subjected to high resolution separation using 2D GE, followed by in-gel enzymatic digestion using essentially the same conditions described in the first paper. Obviously because the proteins are isolated prior to digestion by their affinity to a Cu containing resin, the loss of the metal centre during the digestion step does not render the method incompatible with a full metalloproteomic approach. The protocol was used for a systematic screen for Cu-binding proteins in soybean seeds and yielded a total of 32 protein spots displaying Cu-binding ability, which were unambiguously identified by MALDI-TOF-MS analysis. About 78% of the identified proteins contained the possible copper-binding motifs, namely, H–(X)_n–H (n^1/4 0–5, 7, and 12), H–(X)₃–C, H–(X)₆–M, M–(X)₇–H, and C–(X)_n–C (n^1/4 2–4). In the third study,²⁰² which really sits between the first two in terms of specificity, a 3D separation system was used to enhance the separation of metalloproteins in soybean seeds. Separations using SEC-ICP-MS facilitated the identification of three metal fractions containing Co, Cu, Fe, Mg, Mn and Zn. The second separation of the collected fractions was by AEC and the resultant sub-fractions were lyophilized and subjected to a third dimension of separation using SDS-PAGE. After the separation the bands were digested and in addition to others, the following proteins, previously associated with metals, were identified: 3-lipoxygenase A chain (soybean) complex with 13(S)-hydroperoxy-9(Z), 11(E)octadecadienoic acid, beta-amylase [Glycine max], seed lipoxygenase-1, lipoxygenase [Glycine max], seed lipoxygenase-2 (Pisum sativum) and beta-conglycinin. For a full identification process it would be appropriate to look further at the metal binding motifs in these proteins.

6.3 Tagging and labelling for macromolecular analysis

Critical reviews^203–205 describing the detection of proteins by hyphenated techniques using endogenous heteroatom tags and chemical labelling with metals highlight different approaches to the application of ICP-MS detection. Certainly whilst naturally occurring heteroatoms such as P, S and Se incorporated in some macromolecules allow for their detection, the poor ICP-MS characteristics of these elements makes labelling with true metals a more attractive approach, particularly for the measurement of low abundance proteins. Clearly the incorporation of elements which have a good signature in ICP-MS and are not widely distributed in biology, such as the rare earth elements, make for ideal labels as they will intrinsically have a good S/N. As is often the case with review articles strong evidence supporting the authors views is presented, in this case In the first review²⁰³ the authors present strongly held views that the use of chemical derivatization followed by elemental tagging is far better than current biomedical methods based on immunochemistry, without seeming to appreciate that in the real world this technique is the mainstay of routine analysis. The use of chemical derivatization requires extensive separation of the analytes from the biological matrix as the tagging method is not specific to the molecules being measured, as the side groups being labelled are present in a wide range of other macromolecules present in the sample. In fact it is difficult to envisage immunochemistry, which is by nature a highly specific labelling method, not being the route for the further implementation of ICP-MS detection into this important area. Many new techniques are being developed, which replicate the use of the well-established ELISA or dot-blot methodologies with ICP-MS detection. In another critical review²⁰⁴ a more balanced assessment of the current direction in the tagging of proteins, makes the point that although some popular tagging reagents, such Hg and I are effective, they are undesirable. This is because of their toxicity, poor ionisation and in the case of these 2 elements the possibility of carry over effects in the plasma. The reviewers consider the rare earth elements to offer significant advantages in terms of sensitivity. The review covers the use of Hg, I and ferrocene tagging reagents, the metal-coded affinity tag technique and the use of isotopic-tagging which allows for the absolute quantitation of proteins, rather than relative quantitation. It was conclude that rather than adopting tagging reagents used in other areas, the concept is mature enough that specific reagents should be developed for use with ICP-MS. The final tutorial review²⁰⁵ summarises the use of covalently bound metal tags, most notably 1,4,7,10-tetraazacyclododecane N,N′,N′′,N′′-tetraacetic acid (DOTA) chelate complexes carrying lanthanides as the metal core. This review covers the scope and limitations of peptide and protein quantification and makes the point that the metal tags not only provide low LODs, but also because of the large number of different lanthanides and lanthanide isotopes, allow for multiplexing capabilities. The tutorial illustrates very nicely the sorts of analytical work-flows that have been developed and includes the calibration methods that can be used to allow for accurate and high accuracy IDMS methods to be used.

A number of new tagging reagents for use with different biomolecules have been reported. A novel²⁰⁶ element-labelled activity-based photo-cleavable biotinylated chemical tagging reagent, termed a “Hub” by the authors, was designed and synthesized to combine ICP-MS and ESI-MS workflows. The developed reagent combined the following four parts: (1) a sulfonyl fluoride as an example of a binding group, with specificity towards the hydroxyl group of the serine residue at the active site of serine proteases; (2) a Eu-loaded DOTA to produce ICP-MS signals for protein quantification; (3) a conjugated biotin for capture by streptavidin-coated beads; and (4) a photo-cleavable o-nitrobenzyl ether linker to release the captured serine protease upon UV irradiation for ESI-IT-MS identification. A detailed and timely characterization of the metal-tagged antibodies used for quantitative bioanalysis using ICP-MS has been reported.²⁰⁷ The common modification employing bifunctional ligands containing maleimide residues as a reactive group was investigated in detail via SEC-ICP-MS and HPLC-TOF-MS to determine the preservation of the antibody structure after tagging. Mouse monoclonal IgG modified with metal-coded tags was used as a model system and several antibody fragments were identified carrying different numbers of metal tags. There are some merits to this approach as the “hub” can be easily extended to the quantification and identification of other targeted proteases such as cysteine, threonine and aspartic proteases. Other binding groups could be added with specificity to other targets and containing different elements or isotopes, allowing the targeting of other molecules. The ability to isolate specific proteins from a complex matrix via streptavidin-coated beads, prior to LC-MS/MS analysis is also a significant advantage, effectively overcoming the differences in sensitivity between inorganic and organic MS of large molecules. A sensitive and selective method²⁰⁸ using magnetic separation coupled with ICP-MS detection has been developed for simultaneous determination of multiple glycoproteins, such as aptoglobin (HP), hemopexin (HPX) and ovalbumin (OVA). These species were selected by lectin conjugated magnetic particles via their glycol-structure and then immunoreacted with antibodies labelled with Cd, Hg, Pb through poly(acrylic acid). The metal ions, corresponding to the concentration of glycoproteins, were released from the antibody with an acid-dissolution step and subjected to subsequent ICP-MS detection. Using a 50 mL sample volume which seems quite large, the LODs achieved with the method were 0.032, 0.027 and 0.13 ng mL⁻¹, respectively for HP, HPX and OVA. The response of the magnetic immuno-ICP-MS assay for HP, HPX and OVA was linear over a dynamic range of 0.1–100, 0.1–100 and 0.5–100 ng mL⁻¹, respectively. The recoveries for HP, HPX and OVA in spiked human serum samples were in the range 97.6–105%. Benda et al.²⁰⁹ have extended the application of lanthanide-ion-based tags from quantification using ICP-MS into the quantification of labelled intact proteins using ESI-MS and ESI-MS/MS. The proteins were labelled with the metal chelate tag MeCAT-iodoacetamide (IA) (1,4,7,10-tetraazacyclododecane N,N′,N′′,N′′′-tetra acetic acid with a IA reactive site) and separated using a C3 RP HPLC column interfaced to ESI-MS. Even large proteins were completely labelled at all available cysteine residues using MeCAT-IA with only a small excess of reagent. Fragmentation of labelled proteins either using infrared multiphoton dissociation in FT-ICR-MS or higher-energy collision dissociation with an Orbitrap gave characteristic fragments, which were used to quantify several intact proteins avoiding digestion. To demonstrate the applicability HSA was quantified in blood serum and HPLC-ICP-MS used to verify the results. Undoubtedly, because the metal within the tag may be any of the lanthanides, multiplexing capabilities are inherent in the design.

6.4 Elemental and molecular imaging

Elemental imaging involves the use of approaches which can show the spatial distribution of elements, usually in tissue or other biological materials. The molecular version includes the analysis of gels from GE separations to identify bands containing elements either intrinsically present in the molecule or tagged on to make the compound “visible” using atomic spectroscopy.

Researchers investigating the binding of metals to proteins in plankton and seawater have continued their studies reported in the last speciation ASU, to develop the use of LA-ICP-MS for the spatial analysis of PAGE gels, this time in 2D rather than 1D.²¹⁰ After extraction and precipitation of the proteins using different trichloroacetic acid–acetone and methanol steps, electrophoretic 2D PAGE separation yielded a complex array of protein bands, which were then interrogated using LA-ICP-MS to reveal the presence of Cd, Cr, Cu and Zn in proteins of low molecular weight and variable pI. Unfortunately the identity of the metalloproteins was limited to size and isoelectric point, but using this approach it would be possible to sequence the proteins by LC-MS/MS, which will presumably be the focus of further work.

The limiting step in applying LA-ICP-MS to detect elements in metalloproteomic studies has been recognized as the lack of reliable methods to separate the proteome with high resolution, while retaining bound metal ions, particularly in the final electrophoresis preceding metal analysis. To this end the development of non-denaturing gel-separation systems that can be used for high resolution separations without loss of metal have been initiated. An investigation into the impact of changing the conditions of SDS-PAGE on the quality of protein separation and retention of functional properties has resulted in the development of a high resolution PAGE method providing for the separation of native proteins.²¹¹ Removal of SDS and EDTA from the sample buffer together with omission of a heating step had no effect on the results of PAGE separations. Reduction of SDS in the running buffer from 0.1% to 0.0375% together with removal of the EDTA also made little impact on the quality of the electrophoretograms of fractions of pig kidney cell proteome in comparison with that achieved with the standard SDS-PAGE approach. Retention of Zn^II bound in proteomic samples increased from 26 to 98% upon shifting from standard to modified conditions. Moreover, seven of nine model enzymes, including four Zn^II proteins separated using the developed method retained their enzymatic activity. All nine proteins investigated were active in BN-PAGE, whereas all underwent denaturation during SDS-PAGE. Retention of the metal within the protein after electrophoresis was additionally confirmed using LA-ICP-MS and in-gel Zn–protein staining using the fluorophore 6-methoxy-8-p-toluenesulfonamidoquinoline (TSQ). A non-denaturing 2D GE protocol for screening in vivo non-covalent uranium–protein complexes in fresh water crayfish (Procambarus clarkii) with LA-ICP MS followed by protein identification by HPLC-Orbitrap MS has been developed.^212,213 The approach was based on non-denaturing 1D and 2D PAGE GE, using isoelectric focusing in the first step, in conjunction with LA-ICP MS for the detection of U-containing proteins. The proteins were then identified by microbore LC-MS/MS using an Orbitrap MS. The method was applied to the analysis of the cytosol of the hepatopancreas of crayfish exposed to U. The imaging of U in 2D gels revealed the presence of 11 U-containing protein spots, six protein candidates (ferritin, glyceraldehyde-3-phosphate dehydrogenase, triose-phosphate isomerase, cytosolic manganese superoxide dismutase (Mn-SOD), glutathione S transferase D1 and H3 histone family protein) were then identified by matching with the MS data base for a similar crustacea.

The use of laser desorption coupled to inorganic and molecular MS detectors to provide both elemental and molecular information has been developed. Certainly the coupling of LA sample introduction to both elemental and molecular MS has numerous applications in the area of bioimaging. Such an approach²¹⁴ has been developed for the simultaneous spatial analysis of thin tissue sections by LA coupled both to elemental (ICP-MS) and molecular MS. The experimental setup was based on a LA system with a 213 nm laser coupled in parallel via flow-split transfer lines both with an ICP ionization source and an APCI source on a Orbitrap MS. Simultaneous elemental and molecular bioimaging with high lateral resolution down to 25 μm was presented for the staining agents eosin Y and haematoxylin as well as for the chemotherapy drug cisplatin, in thin (3 μm) kidney sections from mice exposed to cisplatin. The paper goes some way towards developing an important analytical methodology, for which applications in the histopathology laboratory have yet to be fully established. Perhaps more interesting is the combined use of ICP-MS, MALDI-MS and SALDI-MS for imaging CE separations off-line. Tomolova et al.^215,216 have developed an off-line coupling interface for CE separations which utilized a liquid junction and sub-atmospheric deposition chamber to collect fractions of CE effluent on to disposable PETG targets plated with a thin gold film, making them usable for both MALDI and SALD ICP MS analyses after covering with a solution of MALDI matrix. Only a small amount of material from the individual dried droplets was consumed by the MALDI-MS and the separation record was consequently subjected to SALD ICP MS, thus providing quantitative elemental analysis. The developed system was evaluated using the analysis of rabbit liver metallothioneins (MTs) as a model system, to investigate the overall performance of the off-line coupling. The report details excellent results for a painstakingly difficult application area, with good resolution of the MT isoforms and good signal-to-noise evident for the MALDI mass spectra. The two main issues to contend with in this area are the difficulty in maintaining the non-covalent metal–protein interaction during the MALDI analysis, which requires specific matrices and pH conditions, and the very low sample volumes (nL) which make quantitative analysis of low abundant proteins by SALD ICP-MS extremely difficult and probably unlikely in real samples. Whilst quantitation using inorganic Cd standards was excellent (r = 0.9997), the standard concentrations had to be in the mg L⁻¹ concentration range to provide an appropriate LOD for the application.

Abbreviations

1D	One dimensional
2D	Two dimensional
AAS	Atomic absorption spectrometry
AB	Arsenobetaine
AC	Arsenocholine
AE	Anion exchange
AEC	Anion exchange chromatography
AFS	Atomic fluorescence spectrometry
APCI	Atmospheric pressure chemical ionisation
APDC	Ammonium pyrrolidine dithiocarbamate
APL	Acute promyelocytic leukemia
ASE	Accelerated solvent extraction
ASU	Atomic spectrometry update
BA	Bacillus anthracis
BH	Borohydride
CE	Capillary electrophoresis
CEC	Cation exchange chromatography
CP	Cisplatin
CPE	Cloud point extraction
CRC	Colorectal cancer
CRM	Certified reference material
CSF	Cerebrospinal fluid
CT	Computer tomography
CV-AFS	Cold vapour atomic fluorescence spectrometry
Cys	Cysteine
CZE	Capillary zone electrophoresis
DDTC	Diethyldithiocarbamate
DESI	Desorption electrospray ionisation mass spectrometry
DGT	Diffusive gradient in thin film
DLLME	Dispersive liquid–liquid microextraction
DMA	Dimethylarsenic (include oxidation state if known)
DMAE	Dimethylarsenoethanol
DMMTA	Dimethylmonothioarsinic acid
DNA	Deoxyribonucleic acid
DOTA	1,4,7,10-Tetraazacyclododecane-1.4.7.10-tetraacetic acid
DPAA	Diphenylarsinic acid
DPP	Differential pulse polarography
EDTA	Ethylenediaminetetraacetic acid
EI	Electron ionisation
ESI	Electrospray ionisation
ETAAS	Electrothermal atomic absorption spectrometry
EtHg	Ethylmercury
EXAFS	Extended X-ray absorption fine structure
FAAS	Flame atomic absorption spectrometry
FBS	Foetal bovine serum
FTIR	Fourier transform infrared
GC	Gas chromatography
Gd-BOPTA	Gadobenate
Gd-BT-DO3A	Gadobutrol
Gd-DOTA	Gadoterate
Gd-DTPA	Gadopentetate
Gd-DTPA-BMA	Gadodiamide
GE	Gas electrophoresis
GF-AAS	Graphite furnace atomic absorption spectrometry
GPx	Glutathione peroxidase
GSH	Glutathione
HAS	Human serum albumin
HAV	Hepatitis A virus
HBV	Hepatitis B virus
HG	Hydride generation
HIDA	Hybridisation isotope dilution analysis
HILIC	Hydrophilic interaction liquid
HIV	Human immunodeficiency virus
HPCN	High performance concentric nebuliser
HPLC	High performance liquid chromatography
iAs	Inorganic arsenic
IC	Ion chromatography
ICP	Inductively coupled plasma
ICP-MS	Inductively coupled plasma mass spectrometry
ICP-OES	Inductively coupled plasma optical emission spectrometry
ICR	Ion cyclotron resonance
Id	Internal diameter
IDA	Isotope dilution analysis
ID-ICP-MS	Isotope dilution inductively coupled plasma mass spectrometry
ID-MS	Isotope dilution mass spectrometry
IEC	Ion exchange chromatography
iHg	Inorganic mercury
IR	Infrared
iSn	Inorganic selenium
LA	Laser ablation
LA-MC-ICP-MS	Laser ablation multicollector inductively coupled plasma mass spectrometry
LC	Liquid chromatography
LLE	Liquid–liquid extraction
LLLME	Liquid–liquid–liquid microextraction
LMM	Low-molecular mass
LOD	Limit of detection
LOQ	Limit of quantification
MAE	Microwave assisted extraction
MALDI	Matrix-assisted laser desorption ionisation
MC-ICP-MS	Multicollector inductively coupled plasma mass spectrometry
MeHg	Methyl mercury
MeSeCys	Methylselenocysteine
Met	Methionine
MIP	Microwave induced plasma
MMA	Monomethylarsenic
MMMTA	Monomethylmonothioarsonic acid
MMP	Magnetic microparticles
MMTA	Monomethylthioarsonic acid
MRI	Magnetic resonance imaging
MRT	Molecular recognition technology
MS	Mass spectrometry
m/z	Mass to charge ratio
MW	Molecular weight
NIST	National Institute of Standards and Technology
NMG	N-Methylglucamine
NRCC	National Research Council of Canada
OMC	Organomercury compounds
OTC	Organotin compounds
PAGE	Polyacrylamide gel electrophoresis
PCR	Polymerase chain reaction
PDMS	Polydimethylsiloxane
PEEK	Polyetheretherketone
PEG	Polyethylene glycol
PFA	Perfluoroalkyl
PhHg	Phenylmercury
PHWE	Pressurised hot water extraction
PLS	Partial least squares
PM	Particulate matter
PTFE	Poly(tetrafluoroethylene)
Q	Quadrupole
QC	Quality control
RP	Reversed phase
RSD	Relative standard deviation
SARS	Severe acute respiratory syndrome
SAX	Strong anion exchange
SCX	Strong cation exchange
SDS	Sodium dodecylsulfate
SeAlb	Selenoalbumin
SEC	Size exclusion chromatography
SeCys	Selenocysteine
SeCys2	Selenocystine
SeEt	Selenoethionine
SeHLan	4,4′-Selenobis[2-aminobutanoic acid]
SEM	Scanning electron microscopy
SeMeG	Methylselenoglutathione
SeMet	Selenomethionein
SeMetSeCys	Selenomethioneinselonocysteine
(SePh)2	Dyphenyl diselenide
SePP	Seleno protein P
SeSug1	Methyl-2-acetamido-2-deoxy-1-seleno-b-D-galactopyranoside
SeSug3	Methyl-2-amino-2-deoxy-1-seleno-b-D-galactopyranoside
SF-ICP-MS	Sector field inductively coupled plasma mass spectrometry
SIMS	Secondary ion mass spectrometry
SO	2-Hydroxy-4-methylselenobutanoic acid
SOD	Superoxide dismutase
SPAX	Synchrotron radiation excited X-ray fluorescence
SPE	Solid phase extraction
SPME	Solid phase microextraction
SRM	Standard reference material
SR-XAS	Synchrotron radiation X-ray absorption spectroscopy
SR-XRF	Synchrotron radiation X-ray fluorescence
ssIDMS	Species specific isotope dilution mass spectrometry
suIDMS	Species unspecific isotope dilution mass spectrometry
TLC	Thin layer chromatography
TMAH	Tetramethylammonium hydroxide
TMAO	Trimethylarsine oxide
TMSb	Trimethylantimony
TMSe	Trimethylselenium
TOF-MS	Time-of-flight mass spectrometry
TRIS	Tris(hydroxymethyl)aminomethane
TrxR	Thioredoxin reductase
TXRF	Total reflection X-ray fluorescence
UAE	Ultrasound-assisted extraction
UPLC	Ultra high performance liquid chromatography
USAEME	Ultrasound-assisted emulsification-microextraction
UV	Ultraviolet
WDXRF	Wavelength dispersive X-ray fluorescence
WHO	World Health Organisation
XANES	X-ray absorption near-edge structure
XAS	X-ray absorption spectroscopy
XAFS	X-ray absorption fine structure spectrometry
XRD	X-ray diffraction

References

1. R. Clough, C. F. Harrington, S. J. Hill, Y. Madrid and J. F. Tyson, J. Anal. At. Spectrom., 2014, 29(7), 1158–1196.
2. O. T. Butler, W. R. L. Cairns, J. M. Cook and C. M. Davidson, J. Anal. At. Spectrom., 2015, 30(1), 21–63.
3. A. Taylor, M. P. Day, S. Hill, J. Marshall, M. Patriarca and M. White, J. Anal. At. Spectrom., 2014, 29(3), 386–426.
4. E. H. Evans, J. Pisonero, C. M. M. Smith and R. N. Taylor, J. Anal. At. Spectrom., 2014, 29(5), 773–794.
5. M. West, A. T. Ellis, P. J. Potts, C. Streli, C. Vanhoof and P. Wobrauschek, J. Anal. At. Spectrom., 2014, 29(9), 1516–1563.
6. S. Carter, A. S. Fisher, M. W. Hinds, S. Lancaster and J. Marshall, J. Anal. At. Spectrom., 2013, 28(12), 1814–1869.
7. P. Ryan, Environmental and Low Temperature Geochemistry, John Wiley and Sons, New York, 2014.
8. Microwave-Assisted Sample Preparation for Trace Element Determination, ed. E. M. de Moraes Flores, Elsevier, Amsterdam, 2014.
9. M. M. Nearing, I. Koch and K. J. Reimer, Spectrochim. Acta, Part B, 2014, 99, 150–162.
10. M. Rutkowska, K. Dubalska, P. Konieczka and J. Namiesnik, Molecules, 2014, 19(6), 7581–7609.
11. V. N. Zaitsev and M. F. Zui, J. Anal. Chem., 2014, 69(8), 715–727.
12. P. Vinas, N. Campillo, I. Lopez-Garcia and M. Hernandez-Cordoba, Anal. Bioanal. Chem., 2014, 406(8), 2067–2099.
13. J. W. Deng, Y. Y. Yang, X. W. Wang and T. G. Luan, TrAC, Trends Anal. Chem., 2014, 55, 55–67.
14. P. Samaddar and K. Sen, J. Ind. Eng. Chem., 2014, 20(4), 1209–1219.
15. J. Harvanova and L. Bloom, J. Liq. Chromatogr. Relat. Technol., 2015, 38(3), 371–380.
16. J. Muse, V. Tripodi and S. Lucangioli, Curr. Anal. Chem., 2014, 10(2), 225–230.
17. M. Grotti, A. Terol and J. L. Todoli, TrAC, Trends Anal. Chem., 2014, 61, 92–106.
18. F. Y. Meng, X. Yuan, X. M. Li, Y. Liu and Y. X. Duan, Appl. Spectrosc. Rev., 2014, 49(7), 533–549.
19. M. Grafe, E. Donner, R. N. Collins and E. Lombi, Anal. Chim. Acta, 2014, 822, 1–22.
20. C. G. Vogiatzis and G. A. Zachariadis, Anal. Chim. Acta, 2014, 819, 1–14.
21. J. Scancar and R. Milacic, J. Anal. At. Spectrom., 2014, 29(3), 427–443.
22. C. D. B. Amaral, J. A. Nobrega and A. R. A. Nogueira, Talanta, 2013, 115, 291–299.
23. J. Ma, M. K. Sengupta, D. X. Yuan and P. K. Dasgupta, Anal. Chim. Acta, 2014, 831, 1–23.
24. C. H. Latorre, J. B. Garcia, S. G. Martin and R. M. P. Crecente, Anal. Chim. Acta, 2013, 804, 37–49.
25. B. B. Chen, M. He, C. Zhong and B. Hu, J. Chromatogr. A, 2014, 1363, 62–70.
26. C. D. B. Amaral, J. A. Nobrega and A. R. A. Nogueira, Microchem. J., 2014, 117, 122–126.
27. S. Garcia-Salgado and M. A. Quijano, Talanta, 2014, 128, 83–91.
28. H. Y. Cheng, C. Han, Z. G. Xu, J. H. Liu and Y. C. Wang, Food Analytical Methods, 2014, 7(10), 2153–2162.
29. N. Kovachev, M. A. Aguirre, M. Hidalgo, K. Simitchiev, V. Stefanova, V. Kmetov and A. Canals, Microchem. J., 2014, 117, 27–33.
30. L. Lin, J. P. Zheng and J. H. Sun, Anal. Lett., 2014, 47(14), 2406–2416.
31. Y. Q. Chen, L. M. Huang, W. H. Wu, Y. J. Ruan, Z. J. Wu, Z. M. Xue and F. F. Fu, Electrophoresis, 2014, 35(9), 1346–1352.
32. D. Nakamoto and M. Tanaka, Bunseki Kagaku, 2014, 63(5), 383–390.
33. B. Y. Deng, C. Y. Shen, X. D. Qin, S. J. Liang and Y. Y. Liang, J. Anal. At. Spectrom., 2014, 29(10), 1889–1896.
34. P. Li, Y. J. Chen, H. Z. Lian and X. Hu, J. Anal. At. Spectrom., 2014, 29(10), 1785–1790.
35. T. Yordanova, I. Dakova, K. Balashev and I. Karadjova, Microchem. J., 2014, 113, 42–47.
36. A. Philippe and G. E. Schaumann, PLoS One, 2014, 9(2), 9.
37. M. Wacker and A. Seubert, J. Anal. At. Spectrom., 2014, 29(4), 707–714.
38. D. B. Wu and T. Pichler, Anal. Methods, 2014, 6(14), 5112–5119.
39. H. Nygren, G. Dahlen and P. Malmberg, Basic Clin. Pharmacol. Toxicol., 2014, 115(1), 129–133.
40. Y. Zeng, K. L. Xu, X. D. Hou and X. M. Jiang, Microchem. J., 2014, 114, 16–21.
41. B. J. Han, X. M. Jiang, X. D. Hou and C. B. Zheng, Anal. Chem., 2014, 86(1), 936–942.
42. L. D'Ulivo, L. Yang, Y. L. Feng, J. Murimboh and Z. Mester, J. Anal. At. Spectrom., 2014, 29(6), 1132–1137.
43. A. Bednarik, I. Tornalova, V. Kanicky and J. Preisler, J. Chromatogr. A, 2014, 1364, 271–275.
44. Y. L. Yu, F. Gao, M. L. Chen and J. H. Wang, Acta Chim. Sin., 2013, 71(8), 1121–1124.
45. C. M. Weekley, A. Shanu, J. B. Aitken, S. Vogt, P. K. Witting and H. H. Harris, Metallomics, 2014, 6(9), 1602–1615.
46. A. Carmona, S. Roudeau, L. Perrin, G. Veronesi and R. Ortega, Metallomics, 2014, 6(4), 822–832.
47. C. Lenz, T. Behrends, T. Jilbert, M. Silveira and C. P. Slomp, Chem. Geol., 2014, 370, 49–57.
48. J. C. Ulrich and J. E. S. Sarkis, Accredit. Qual. Assur., 2013, 18(6), 511–516.
49. E. Vassileva, I. Wysocka and M. Betti, Microchem. J., 2014, 116, 197–205.
50. W. Quiroz, L. Aguilar, M. Barria, J. Veneciano, D. Martinez, M. Bravo, M. G. Lobos and L. Mercado, Talanta, 2013, 115, 902–910.
51. W. Quiroz, M. Cortes, F. Astudillo, M. Bravo, F. Cereceda, V. Vidal and M. G. Lobos, Microchem. J., 2013, 110, 266–272.
52. M. Grotti, F. Ardini, A. Terol, E. Magi and J. L. Todoli, J. Anal. At. Spectrom., 2013, 28(11), 1718–1724.
53. E. Boyaci, N. Horzum, A. Cagir, M. M. Demir and A. E. Eroglu, RSC Adv., 2013, 3(44), 22261–22268.
54. S. Z. Chen, X. L. Guo, Y. Y. He and D. B. Lu, At. Spectrosc., 2013, 34(4), 113–118.
55. J. An, K. H. Kim, J. A. Kim, H. Jung, H. O. Yoon and J. Seo, J. Hazard. Mater., 2013, 260, 24–31.
56. J. J. Du, D. S. Che, J. F. Zhang and C. Y. Jing, Environ. Toxicol. Chem., 2014, 33(8), 1692–1696.
57. P. Li, X. Q. Zhang, Y. J. Chen, T. Y. Bai, H. Z. Lian and X. Hu, RSC Adv., 2014, 4(90), 49421–49428.
58. H. Y. Peng, N. Zhang, M. He, B. B. Chen and B. Hu, Talanta, 2015, 131, 266–272.
59. P. Li, X. Q. Zhang, Y. J. Chen, H. Z. Lian and X. Hu, Anal. Methods, 2014, 6(12), 4205–4211.
60. N. Higashidani, T. Kaneta, N. Takeyasu, S. Motomizu, N. Okibe and K. Sasaki, Talanta, 2014, 122, 240–245.
61. N. Wang and J. Tyson, J. Anal. At. Spectrom., 2014, 29(4), 665–673.
62. S. Musil, T. Matousek, J. M. Currier, M. Styblo and J. Dedina, Anal. Chem., 2014, 86(20), 10422–10428.
63. M. Huang, W. E. Gan and S. S. Xie, Anal. Methods, 2014, 6(6), 1796–1801.
64. N. S. Keller, A. Stefasson and B. Sigfuson, Talanta, 2014, 128, 466–472.
65. S. Musil, A. H. Petursdottir, A. Raab, H. Gunnlaugsdottir, E. Knipp and J. Feldmann, Anal. Chem., 2014, 86(2), 993–999.
66. E. Pitzalis, M. Onor, M. C. Mascherpa, G. Pacchi, Z. Mester and A. D'Ulivo, Anal. Chem., 2014, 86(3), 1599–1607.
67. E. Leese, J. Morton, E. Tan, P. H. E. Gardiner and V. A. Carolan, J. Anal. Toxicol., 2014, 38(1), 24–30.
68. G. Raber, T. Raber, R. Raml, M. Murko, C. Magnes and K. A. Francesconi, Talanta, 2013, 116, 876–881.
69. L. Yehiayan, S. Stice, G. L. Liu, S. Matulis, L. H. Boise and Y. Cai, Chem. Res. Toxicol., 2014, 27(5), 754–764.
70. M. A. Garcia-Sevillano, T. Garcia-Barrera, F. Navarro and J. L. Gomez-Ariza, Metallomics, 2013, 5(12), 1644–1655.
71. P. Millard, S. Massou, J. C. Portais and F. Letisse, Anal. Chem., 2014, 86(20), 10288–10295.
72. L. P. Munoz, D. Purchase, H. Jones, J. Feldmann and H. Garelick, Anal. Methods, 2014, 6(3), 791–797.
73. I. Kakoulli, S. V. Prikhodko, C. Fischer, M. Cilluffo, M. Uribe, H. A. Bechtel, S. C. Fakra and M. A. Marcus, Anal. Chem., 2014, 86(1), 521–526.
74. V. Sele, J. J. Sloth, B. Holmelid, S. Valdersnes, K. Skov and H. Amlund, Talanta, 2014, 121, 89–96.
75. K. O. Amayo, A. Raab, E. M. Krupp and J. Feldmann, Talanta, 2014, 118, 217–223.
76. U. Arroyo-Abad, S. Lischka, C. Piechotta, J. Mattusch and T. Reemtsma, Food Chem., 2013, 141(3), 3093–3102.
77. K. O. Amayo, A. Raab, E. M. Krupp, H. Gunnlaugsdottir and J. Feldmann, Anal. Chem., 2013, 85(19), 9321–9327.
78. R. A. Glabonjat, G. Raber, K. B. Jensen, J. Ehgartner and K. A. Francesconi, Anal. Chem., 2014, 86(20), 10282–10287.
79. W. A. Maher, S. Foster, F. Krikowa, E. Duncan, A. St John, K. Hug and J. W. Moreau, Microchem. J., 2013, 111, 82–90.
80. W. Y. Lee, Y. H. Yim, E. Hwang, Y. Lim, T. K. Kim and K. S. Lee, Bull. Korean Chem. Soc., 2014, 35(3), 821–827.
81. V. M. O. Carioni, C. S. Nomura, L. L. Yu and R. Zeisler, J. Radioanal. Nucl. Chem., 2014, 299(1), 241–248.
82. F. R. Chen, L. Zheng, Z. G. Wang, J. Sun, L. H. Han and X. R. Wang, Spectrosc. Spectral Anal., 2014, 34(6), 1675–1678.
83. F. P. Serpe, R. Russo, P. Gallo and L. Severino, J. Food Prot., 2013, 76(7), 1293–1299.
84. M. Han, G. X. Zhao, S. Z. Li, Z. H. Sun, Y. Li and N. Jia, Chin. J. Anal. Chem., 2013, 41(11), 1780–1781.
85. J. Tyson, Spectroscopy, 2013, 28(10), 40–41.
86. K. Baba, T. Arao, N. Yamaguchi, E. Watanabe, H. Eun and M. Ishizaka, J. Chromatogr. A, 2014, 1354, 109–116.
87. T. Narukawa, T. Suzuki, K. Inagaki and A. Hioki, Talanta, 2014, 130, 213–220.
88. J. Y. Kim, W. I. Kim, A. Kunhikrishnan, D. W. Kang, D. H. Kim, Y. J. Lee, Y. J. Kim and C. T. Kim, Food Sci. Biotechnol., 2013, 22(6), 1509–1513.
89. H. Yang, S. H. Dai, M. Wang, X. F. Mao, Y. T. Huang and F. H. Wang, Anal. Lett., 2014, 47(15), 2601–2612.
90. C. G. Yuan, W. P. Jiang, T. Zhu, B. Yuan and X. W. Song, Spectrosc. Spectral Anal., 2014, 34(8), 2259–2263.
91. I. Lopez-Garcia, M. Briceno, Y. Vicente-Martinez and M. Hernandez-Cordoba, Food Chem., 2015, 167, 396–401.
92. T. Llorente-Mirandes, J. Calderon, F. Centrich, R. Rubio and J. F. Lopez-Sanchez, Food Chem., 2014, 147, 377–385.
93. Y. B. Xiao, M. Zhang and H. W. Wen, Spectrosc. Spectral Anal., 2014, 34(4), 1100–1103.
94. P. Langner, C. Mikutta, E. Suess, M. A. Marcus and R. Kretzschmar, Environ. Sci. Technol., 2013, 47(17), 9706–9714.
95. H. J. Sanchez, J. J. Leani, C. A. Perez and R. D. Perez, J. Appl. Spectrosc., 2014, 80(6), 912–916.
96. S. Mishra, G. Wellenreuther, J. Mattusch, H. J. Stark and H. Kupper, Plant Physiol., 2013, 163(3), 1396–1408.
97. P. Galler, A. Raab, S. Freitag, K. Blandhol and J. Feldmann, J. Anal. At. Spectrom., 2014, 29(4), 614–622.
98. D. Baralkiewicz, B. Pikosz, M. Belter and M. Marcinkowska, Accredit. Qual. Assur., 2013, 18(5), 391–401.
99. M. Z. Corazza, E. S. Ribeiro, M. G. Segatelli and C. R. T. Tarley, Microchem. J., 2014, 117, 18–26.
100. N. N. Meeravali, R. Manjusha and S. J. Kumar, J. Anal. At. Spectrom., 2014, 29(11), 2168–2175.
101. S. Sahan, S. Sacmaci, S. Kartal, M. Saccmaci, U. Sahin and A. Ulgen, Talanta, 2014, 120, 391–397.
102. J. Sun, L. Ma, Z. G. Yang and L. Wang, J. Sep. Sci., 2014, 37(15), 1944–1950.
103. M. M. Wolle, G. M. M. Rahman, H. M. S. Kingston and M. Pamuku, J. Anal. At. Spectrom., 2014, 29(9), 1640–1647.
104. M. Stanislawska, B. Janasik and W. Wasowicz, Talanta, 2013, 117, 14–19.
105. F. Y. Yang, S. J. Jiang and A. C. Sahayam, Food Chem., 2014, 147, 215–219.
106. M. Birka, C. A. Wehe, L. Telgmann, M. Sperling and U. Karst, J. Chromatogr. A, 2013, 1308, 125–131.
107. M. M. Wolle, T. Fahrenholz, G. M. M. Rahman, M. Pamuku, H. M. Kingston and D. Browne, J. Chromatogr. A, 2014, 1347, 96–103.
108. P. Novak, T. Zuliani, R. Milacic and J. Scancar, Anal. Chim. Acta, 2014, 827, 64–73.
109. N. H. Lin, H. P. Wang, K. Kahen, H. Badiei and K. Jorabchi, Anal. Chem., 2014, 86(15), 7954–7961.
110. M. I. Beser, J. Beltran and V. Yusa, J. Chromatogr. A, 2014, 1323, 1–10.
111. T. B. Chokwe, J. O. Okonkwo, L. L. Sibali and E. J. Ncube, Water Sci. Technol., 2014, 69(12), 2389–2396.
112. H. Fu, X. Q. Li, C. Wei, M. W. Shao, Q. H. Zhang and Q. Li, Chin. J. Anal. Chem., 2013, 41(12), 1881–1886.
113. E. Cequier, R. M. Marce, G. Becher and C. Thomsen, J. Chromatogr. A, 2013, 1310, 126–132.
114. M. Zhang, J. Cheng, M. Wu, T. Du, X. H. Wang and M. Cheng, Anal. Methods, 2014, 6(1), 207–214.
115. K. D. Wang, P. S. Chen and S. D. Huang, Anal. Bioanal. Chem., 2014, 406(8), 2123–2131.
116. F. Guo, Q. Liu, J. B. Shi, F. S. Wei and G. B. Jiang, Talanta, 2014, 119, 396–400.
117. L. J. Wu, Y. W. Zhu, X. H. Hao, S. F. Ye, Q. Q. Li and S. G. Min, J. Anal. Chem., 2014, 69(3), 262–269.
118. C. Han, J. N. Sun, H. Y. Cheng, J. H. Liu and Z. G. Xu, Anal. Methods, 2014, 6(14), 5369–5375.
119. V. Romaris-Hortas, J. Bianga, A. Moreda-Pineiro, P. Bermejo-Barrera and J. Szpunar, Talanta, 2014, 127, 175–180.
120. E. Hvattum and H. Priebe, Rapid Commun. Mass Spectrom., 2013, 27(21), 2504–2514.
121. O. S. Arvaniti, A. G. Asimakopoulos, M. E. Dasenaki, E. I. Ventouri, A. S. Stasinakis and N. S. Thomaidis, Anal. Methods, 2014, 6(5), 1341–1349.
122. D. Verrey, M. V. Louyer, O. Thomas and E. Baures, Microchem. J., 2013, 110, 608–613.
123. M. J. Cardador and M. Gallego, J. Chromatogr. A, 2014, 1340, 15–23.
124. P. Rocio-Bautista, V. Pino, B. Delgado, A. M. Afonso and J. H. Ayala, Anal. Methods, 2014, 6(12), 4115–4123.
125. T. Gilevska, M. Gehre and H. H. Richnow, Anal. Chem., 2014, 86(15), 7252–7257.
126. E. L. Regalado, R. K. Dermenjian, L. A. Joyce and C. J. Welch, J. Pharm. Biomed. Anal., 2014, 92, 1–5.
127. Y. Q. Wu, C. Guo, N. W. Zhang, G. F. Bian and K. Z. Jiang, Rapid Commun. Mass Spectrom., 2014, 28(19), 2111–2120.
128. Y. L. Sommer, C. P. Verdon, M. R. Fresquez, C. D. Ward, E. B. Wood, Y. Pan, K. L. Caldwell and R. L. Jones, Anal. Bioanal. Chem., 2014, 406(20), 5039–5047.
129. G. M. M. Rahman, M. M. Wolle, T. Fahrenholz, H. M. Kingston and M. Pamuku, Anal. Chem., 2014, 86(12), 6130–6137.
130. H. Y. Cheng, C. L. Wu, L. H. Shen, J. H. Liu and Z. G. Xu, Anal. Chim. Acta, 2014, 828, 9–16.
131. H. Pietila, P. Peramaki, J. Piispanen, L. Majuri, M. Starr, T. Nieminen, M. Kantola and L. Ukonmaanaho, Microchem. J., 2014, 112, 113–118.
132. X. Xiong, X. T. Qi, J. T. Liu, J. Wang and C. X. Wu, Anal. Lett., 2014, 47(14), 2417–2430.
133. Y. Gao, S. De Craemer and W. Baeyens, Talanta, 2014, 120, 470–474.
134. C. Fernandez-Gomez, J. M. Bayona and S. Diez, Talanta, 2014, 129, 486–490.
135. X. P. Chen, C. Han, H. Y. Cheng, J. H. Liu, Z. G. Xu and X. F. Yin, Anal. Chim. Acta, 2013, 796, 7–13.
136. X. P. Chen, C. Han, H. Y. Cheng, Y. C. Wang, J. H. Liu, Z. G. Xu and L. Hu, J. Chromatogr. A, 2013, 1314, 86–93.
137. A. Montero-Alvarez, M. D. F. de la Campa and A. Sanz-Medel, Int. J. Environ. Anal. Chem., 2014, 94(1), 36–47.
138. J. Cavalheiro, H. Preud'homme, D. Amouroux, E. Tessier and M. Monperrus, Anal. Bioanal. Chem., 2014, 406(4), 1253–1258.
139. L. D'Ulivo, L. Yang, Y. L. Feng and Z. Mester, Anal. Methods, 2013, 5(24), 7127–7131.
140. R. Koplik, I. Klimesova, K. Malisova and O. Mestek, Czech J. Food Sci., 2014, 32(3), 249–259.
141. K. B. Borges and E. F. Freire, Acta Sci., Technol., 2014, 36(3), 531–535.
142. M. Nedaei, A. R. Salehpour, S. Mozaffari, S. M. Yousefi and S. R. Yousefi, J. Sep. Sci., 2014, 37(17), 2372–2379.
143. T. Wu, W. T. Zhao, Z. H. Yang, H. X. Gao and Z. Q. Zhou, J. Sep. Sci., 2013, 36(24), 3918–3925.
144. Y. Kazui, Y. Seto and H. Inoue, Forensic Toxicol., 2014, 32(2), 317–322.
145. J. L. Armstrong, R. L. Dills, J. B. Yu, M. G. Yost and R. A. Fenske, J. Environ. Sci. Health, Part B, 2014, 49(2), 102–108.
146. A. Rugova, M. Puschenreiter, J. Santner, L. Fischer, S. Neubauer, G. Koellensperger and S. Hann, J. Sep. Sci., 2014, 37(14), 1711–1719.
147. C. K. Schmidt, B. Raue, H. J. Brauch and F. Sacher, Int. J. Environ. Anal. Chem., 2014, 94(4), 385–398.
148. D. B. Chu, K. Klavins, G. Koellensperger and S. Hann, J. Anal. At. Spectrom., 2014, 29(5), 915–925.
149. D. F. Thompson, F. Michopoulos, C. J. Smith, C. J. Duckett, R. W. Wilkinson, P. Jarvis and I. D. Wilson, Rapid Commun. Mass Spectrom., 2013, 27(22), 2539–2545.
150. M. L. Mena, E. Moreno-Gordaliz and M. M. Gomez-Gomez, Talanta, 2013, 116, 581–592.
151. I. Moraleja, E. Moreno-Gordaliza, M. L. Mena and M. M. Gomez-Gomez, Talanta, 2014, 120, 433–442.
152. A. Martincic, M. Cemazar, G. Sersa, V. Kovac, R. Milacic and J. Scancar, Talanta, 2013, 116, 141–148.
153. C. M. Tang, C. Li, C. X. Tang, W. Zhan, H. Zheng and X. Z. Peng, Anal. Methods, 2013, 5(24), 7117–7126.
154. P. H. van Wyk, J. B. van Dyk, W. J. Gerber, M. Stander, A. de Villiers and K. R. Koch, Rapid Commun. Mass Spectrom., 2014, 28(5), 505–519.
155. W. Thosaikham, K. Jitmanee, R. Sittipout, S. Maneetong, A. Chantiratikul and P. Chantiratikul, Food Chem., 2014, 145, 736–742.
156. S. X. Shao, X. B. Mi, L. Ouerdane, R. Lobinski, J. F. Garcia-Reyes, A. Molina-Diaz, A. Vass and M. Dernovics, Food Analytical Methods, 2014, 7(5), 1147–1157.
157. A. Nemeth, J. F. G. Reyes, J. Kosary and M. Dernovics, Metallomics, 2013, 5(12), 1663–1673.
158. P. Cheajesadagul, J. Bianga, C. Arnaudguilhem, R. Lobinski and J. Szpunar, Metallomics, 2014, 6(3), 646–653.
159. Z. L. Ni, Y. H. Liu and M. H. Qu, Food Sci. Biotechnol., 2014, 23(4), 1049–1053.
160. N. Zhong, L. Zhong, L. H. Hao, C. Z. Luan and X. R. Li, Anal. Lett., 2015, 48(1), 180–187.
161. A. Galitsopoulou, D. Georgantelis and M. G. Kontominas, Food Addit. Contam., Part A, 2013, 30(11), 1900–1908.
162. S. Torres, S. Cerutti, J. Raba, P. Pacheco and M. F. Silva, Food Chem., 2014, 159, 407–413.
163. I. Lopez-Garcia, Y. Vicente-Martinez and M. Hernandez-Cordoba, J. Agric. Food Chem., 2013, 61(39), 9356–9361.
164. P. Pacheco, T. Hanley and J. A. L. Figueroa, Metallomics, 2014, 6(3), 560–571.
165. Q. H. Zhang and G. P. Yang, J. Chromatogr. A, 2014, 1325, 83–91.
166. E. Galano, E. Mangiapane, J. Bianga, A. Palmese, E. Pessione, J. Szpunar, R. Lobinski and A. Amoresano, Mol. Cell. Proteomics, 2013, 12(8), 2196–2204.
167. Y. J. Hsieh and S. J. Jiang, Microchem. J., 2013, 110, 1–7.
168. J. Zembrzuska, H. Matusiewicz, H. Polkowska-Motrenko and E. Chajduk, Food Chem., 2014, 142, 178–187.
169. M. Palomo, A. M. Gutierrez, M. C. Perez-Conde, C. Camara and Y. Madrid, Food Chem., 2014, 164, 371–379.
170. M. A. Garcia-Sevillano, T. Garcia-Barrera and J. L. Gomez-Ariza, Anal. Bioanal. Chem., 2014, 406(11), 2719–2725.
171. M. A. Garcia-Sevillano, T. Garcia-Barrera and J. L. Gomez-Ariza, J. Chromatogr. A, 2013, 1318, 171–179.
172. S. Kokarnig, N. Kroepfl, D. Kuehnelt and K. A. Francesconi, Anal. Methods, 2014, 6(5), 1603–1607.
173. D. Pick, C. Degen, M. Leiterer, G. Jahreis and J. W. Einax, Microchem. J., 2013, 110, 8–14.
174. K. R. Chitta, J. A. Landero-Figueroa, P. Kodali, J. A. Caruso and E. J. Merino, Talanta, 2013, 114, 25–31.
175. J. Bianga, Z. Touat-Hamici, K. Bierla, S. Mounicou, J. Szpunar, L. Chatiatte and R. Lobinski, J. Proteomics, 2014, 108, 316–324.
176. Y. Zhang, J. K. Duan, M. He, B. B. Chen and B. Hu, Talanta, 2013, 115, 730–736.
177. Y. Chang, J. G. Qu, R. F. Zhang and J. Zhang, Chin. J. Anal. Chem., 2014, 42(5), 753–758.
178. J. Kim and H. B. Lim, Bull. Korean Chem. Soc., 2013, 34(11), 3362–3366.
179. T. T. Shih, C. H. Lin, I. H. Hsu, J. Y. Chen and Y. C. Sun, Anal. Chem., 2013, 85(21), 10091–10098.
180. Y. N. Tsai, C. H. Lin, I. H. Hsu and Y. C. Sun, Anal. Chim. Acta, 2014, 806, 165–171.
181. R. Bhushan and H. Nagar, Biomed. Chromatogr., 2014, 28(1), 106–111.
182. H. Nagar and R. Bhushan, Anal. Methods, 2014, 6(12), 4188–4198.
183. F. Chainet, C. P. Lienemann, J. Ponthus, C. Pecheyran, J. Castro, E. Tessier and O. F. X. Donard, Spectrochim. Acta, Part B, 2014, 97, 49–56.
184. Z. Y. Cui, N. Ge, Y. Z. Cao, Y. M. Liu, J. Li and Y. P. Wu, Chin. J. Anal. Chem., 2013, 41(12), 1887–1892.
185. C. Coscolla, S. Navarro-Olivares, P. Marti and V. Yusa, Talanta, 2014, 119, 544–552.
186. A. Valenzuela, G. Lespes, W. Quiroz, L. F. Aguilar and M. A. Bravo, Talanta, 2014, 125, 196–203.
187. V. Smitiene, I. Baskirova and V. Vickackaite, Chemija, 2013, 24(3), 210–216.
188. N. M. Tzollas, G. A. Zachariadis and E. Rosenberg, Curr. Anal. Chem., 2013, 9(2), 279–287.
189. L. Huang, D. Yang, X. Q. Guo and Z. L. Chen, J. Chromatogr. A, 2014, 1368, 217–221.
190. K. Peeters, T. Zuliani, J. Scancar and R. Milacic, Water Res., 2014, 53, 297–309.
191. K. Peeters, J. Iskra, T. Zuliani, J. Scancar and R. Milacic, Chemosphere, 2014, 107, 386–392.
192. Y. X. Liu, Y. Q. Wan and L. Guo, Anal. Lett., 2014, 47(2), 343–355.
193. V. Nischwitz, J. T. Davies, D. Marshall, M. Gonzalez, J. L. G. Ariza and H. Goenaga-Infante, Metallomics, 2013, 5(12), 1685–1697.
194. J. Karasinski, W. Cegielkowska, M. Wojciechowski, M. Wierzbicka and E. Bulska, Chem. Pap., 2014, 68(3), 291–299.
195. S. Fujii, K. Inagaki, S. Miyashita, A. S. Groombridge, K. Nagasawa, K. Chiba and A. Takatsu, Electrophoresis, 2014, 35(9), 1315–1318.
196. D. J. Hare, A. Grubman, T. M. Ryan, A. Lothian, J. R. Liddell, R. Grimm, T. Matsuda, P. A. Doble, R. A. Cherny, A. I. Bush, A. R. White, C. L. Masters and B. R. Roberts, Metallomics, 2013, 5(12), 1656–1662.
197. L. X. Feng, D. Zhang, J. Wang and H. M. Li, Anal. Methods, 2014, 6(19), 7655–7662.
198. D. Zhang, L. X. Feng, J. Wang and J. P. Xiong, Chin. J. Anal. Chem., 2014, 42(2), 179–185.
199. Y. N. Ordonez, R. F. Anton and W. C. Davis, Anal. Methods, 2014, 6(12), 3967–3974.
200. S. B. Jayasinghe and J. A. Caruso, Int. J. Mass Spectrom., 2013, 356, 33–36.
201. Y. Wang, H. Li, Y. F. Qiu, N. Li, W. H. Sun and Z. H. Shan, Plant Biosystems, 2014, 148(1), 88–95.
202. L. R. V. Mataveli and M. A. Z. Arruda, J. Proteomics, 2014, 104, 94–103.
203. B. Campanella and E. Bramanti, Analyst, 2014, 139(17), 4124–4153.
204. Y. He, Y. Y. Zhang, C. Wei, C. Y. Li, Y. Gao and R. Liu, Appl. Spectrosc. Rev., 2014, 49(6), 492–512.
205. G. Schwarz, L. Mueller, S. Beck and M. W. Linscheid, J. Anal. At. Spectrom., 2014, 29(2), 221–233.
206. X. W. Yan, Z. X. Li, Y. Liang, L. M. Yang, B. Zhang and Q. Q. Wang, Chem. Commun., 2014, 50(50), 6578–6581.
207. L. Mueller, T. Mairinger, G. Hermann, G. Koellensperger and S. Hann, Anal. Bioanal. Chem., 2014, 406(1), 163–169.
208. H. Y. Peng, Y. Jiao, X. Xiao, B. B. Chen, M. He, Z. R. Liu, X. Zhang and B. Hu, J. Anal. At. Spectrom., 2014, 29(6), 1112–1119.
209. D. Benda, G. Schwarz, S. Beck and M. W. Linscheid, J. Mass Spectrom., 2014, 49(1), 13–18.
210. N. Garcia-Otero, M. C. Barciela-Alonso, P. Bermejo-Barrera, A. Moreda-Pineiro, M. S. Jimenez, M. T. Gomez and J. R. Castillo, Anal. Chim. Acta, 2013, 804, 59–65.
211. A. B. Nowakowski, W. J. Wobig and D. H. Petering, Metallomics, 2014, 6(5), 1068–1078.
212. M. Xu, S. Frelon, O. Simon, R. Lobinski and S. Mounicou, Anal. Bioanal. Chem., 2014, 406(4), 1063–1072.
213. M. Xu, S. Frelon, O. Simon, R. Lobinski and S. Mounicou, Talanta, 2014, 128, 187–195.
214. C. Herdering, C. A. Wehe, O. Reifschneider, I. Raj, G. Ciarimboli, K. Diebold, C. Becker, M. Sperling and U. Karst, Rapid Commun. Mass Spectrom., 2013, 27(23), 2588–2594.
215. I. Tomalova, P. Foltynova, V. Kanicky and J. Preisler, Anal. Chem., 2014, 86(1), 647–654.
216. I. Tomalova, P. Foltynova, V. Kanicky and J. Preisler, Chem. Listy, 2013, 107, S329–S330.

---