摘要:

Measurement of total boron and 10B concentration and the detection and measurement of elevated 10B levels in biological samples by inductively coupled plasma mass spectrometry using the determination of 10B511B ratios Jennifer A. Moretona and H. Trevor Delvesb aMilestones, Ampleforth, North Yorkshire, UK YO624DA bSAS Trace Element Unit, Chemical Pathology, Southampton General Hospital, Southampton, UK SO16 6YD Received 19th April 1999, Accepted 21st July 1999 Methods were developed to determine total B and 10B content in various biological samples, media and reagents by ICP-MS.Open vessel, wet ash digestion protocols using HNO3 were developed for the biological and media samples. Digestion of blood and plasma also required H2SO4. Despite relatively high blanks, instrumental detection limits using 3s at or near the blank level were, for total boron (total B), <0.1 mmol l-1 (<1 mg l-1), and for 10B, <0.01 mmol l-1 (<0.1 mg l-1). Problems with variable blanks, matrix suppression and peak overlap by 12C on 11B are discussed.The developed methods were used to analyse samples for a collaborative study with Clinical Neurosciences at Southampton General Hospital, related to boron neutron capture therapy of brain tumours, of the uptake of 10B by tumour cells in vitro, using tissue culture, and in vivo, in the rat. Detection limits achieved in cells solubilised in 1 MNaOH and medium were<0.001 mmol ml-1 total B and<0.0001 mmol ml-1 10B and, for blood, 0.002 mmol g-1 total B and 0.0002 mmol g-1 10B,~10-9 Mboronophenylalanine solutions.The analytical precision for duplicate analyses gave mean within-run RSDs of ~7% for 10B concentration and <10% for total B concentration. The determination of 10B511B ratios enabled statistically significant increases in 10B concentration to be detected and confirmed at low B levels where sample digests and digest blanks had similar B contents. Running 10B511B ratios for standard and sample solutions were utilised for novel methods of calculating elevated 10B levels in the samples.Despite achieving low detection limits, elevated 10B levels could not be confirmed without the determination of 10B511B ratios. Boron neutron capture therapy (BNCT) can be used to treat to BNCT, finding a good correlation between determination by ICP-MS and ICP-AES, which has also been found by other brain tumours for which tumour cells are targeted via the incorporation of 10B.Boron has two isotopes, 11B (80.22%) workers.35 ICP methods give the best precision and lower detection limits.34 ICP-MS has the advantages of a sensitivity and 10B (19.78%).1 The neutron capture reaction for 10B is 10B(n,ac)7Li. The boron-containing substances used for BNCT about an order of magnitude higher than ICP-AES and discrimination between the two isotopes. are 10B-enriched. The production of 10B-enriched boron only began in the 1940s.2 Analysis for B was carried out in the Trace Element Unit as part of a collaborative study with Clinical Neurosciences For successful BNCT, suYcient 10B atoms are required in the cancer cells and suYcient neutrons must be delivered to at Southampton General Hospital.The research group of Mr P. D. Lees and Dr. T. Loughlin was working on the uptake the tumour, related to tumour location and depth.3 The most critical and diYcult step is targeting the tumour. A high of 10B by tumour cells in vitro, using tissue culture, and in vivo, in the rat, including the uptake of specific 10B-enriched concentration diVerential is required for 10B between the normal and neoplastic cells.Suitable 10B concentrations in peptides. Methods were developed to determine total B and 10B content by ICP-MS, enabling uptake of 10B by cells to be tumour cells proposed were >10 mg g-1 (ref. 2) and 35–50 mg g-1 (ref. 4). Early experiments with BNCT utilised assessed. Biological samples included blood, plasma, brain tissue, brain homogenate, cells and cells solubilised in 1 M boric acid,2 borax5 and sodium pentaborate.6 Then p-carboxyphenylboronic acid6 and derivatives of polyhedral boranes NaOH.Incubation media and wash media used for cells were also analysed to check on B levels and washout. Other reagents were used.7,8 p-Boronophenylalanine (BPA) gave encouraging results with malignant melanomas.9–16 Compounds now under were also analysed to check for B contamination.Samples of boric acid (H3BO3), BPA [H2BO2C6H4CH2CH(NH2)COOH] evaluation include carborane-containing amino acids,17 boronated thioureas and boron compounds conjugated to and carboranylacetic acid (COOHCH2–C2B10H11) were also analysed, in part to verify that they were enriched with 10B. tumour-seeking macromolecules,18 such as monoclonal antibodies, 19,20 nucleosides, oligonucleotides and polypeptides. Open vessel, wet ash digestion protocols using HNO3 were developed for the biological and media samples.Digestion The ‘prompt gamma’ technique21 was the first available method to measure 10B levels in patient’s blood in 1986, since was diYcult and time consuming and the optimum conditions varied for each type of sample. Digestion of blood and plasma when other methods have been introduced.22–27 By the early 1990s, most NCT teams were using inductively coupled plasma also required H2SO4. Published reports comparing sample dissolution methods for B, including dry ashing, wet digestion atomic emission spectrometry (ICP-AES).28,29 B levels have been determined in biological reference materials by ICP- and microwave dissolution, have produced contradictory results.35 A study by Nyomura et al.35 found no B loss during MS30–35 but have only recently been used by NCT teams.36 Probst et al.36 determined B in mouse tissue for a study related open vessel digestion.J. Anal. At. Spectrom., 1999, 14, 1545–1556 1545Initial experiments at the Trace Element Unit yielded tissue Isotopic shifts therefore occur in terrestrial rocks.30 The 10B511B ratio of terrestrial samples varies from 0.236 to containing 100 mg g-1 10B, but the high cost of the 10Bcontaining compounds was an incentive to work at relatively 0.25537–39 and a ratio of 0.260 has been reported for meteoritic material.30 TIMS has been almost exclusively used in B isotope low concentrations, where problems of contamination and high B digestion blanks became significant.EVorts were made geochemistry40 but one investigation, using ICP-MS, showed that the 10B511B ratio was 0.256 for basalt and 0.239 for to keep background concentrations as low as possible (<5 mg l-1 total B). In order to achieve adequate detection seaweed.38 limits, a sensitivity for total B of >75 000 counts s-1 with a 100 mg l-1 B solution was required. Experimental Despite relatively high background concentrations, instrumental detection limits (DLs) (using three times the standard Instrumental conditions deviation of replicate determinations made at or near the blank level ) were <0.1 mmol l-1 (<1 mg l-1) for total B and Initial investigations were carried out using a VG ( Winsford, Cheshire, UK) PlasmaQuad Mark 1 instrument but in the <0.01 mmol l-1 (<0.1 mg l-1) for 10B.The calculated sample DLs for B (10B, 11B and total B) in biological samples were majority of the analyses a Perkin-Elmer (PE) SCIEX ( Thornhill, ON, Canada) Elan 5000 was used.Details of the as low as 0.001 mmol g-1 (0.01 mg g-1) (10-9M boronophenylalanine solutions), but there were frequent problems with the operating conditions used with the two instruments are given in Table 1. The scan parameters diVered, since the multi- variable digest blanks, background B correction and matrix suppression. These problems have been experienced by other channel scanning mode was used for analysis with the VG PlasmaQuad and the peak hopping mode with the PE SCIEX workers.30,35,36 10B511B ratios were monitored from the beginning of the Elan 5000.Sample preparation procedures were identical for the two instruments. study and it became clear that increases in sample 10B level could be detected using the 10B511B ratio, below the calculated When operating in the multi-channel scanning mode, any overlap of the 12C peak on the 11B peak was visible, necessitat- DL for total B or 10B. 10B511B ratio determination was suYciently precise to enable statistically significant increases ing further digestion of the samples. Peak tailing eVects resulting from overlap of the 12C peak on the 11B peak with in 10B concentration to be detected and confirmed at these low levels. Running 10B511B ratios for standard and sample organic samples have been identified by other workers.33–36,41 For the PE SCIEX Elan 5000, operating in the peak hopping solutions were utilised for novel methods of calculating elevated 10B levels in the samples.The diVerence in m/z between mode, there was less indication of C levels in the samples. An initial 1 min narrow mass range scan (Table 2, boron scan) 10B and 11B is relatively large, therefore micromolar quantities were used for calculations. Despite achieving low DLs, elevated was therefore carried out with the PE SCIEX Elan 5000 in the single continuous scan mode, in order to check C levels 10B levels could not be confirmed without the determination of 10B511B ratios.and also find out whether sample dilution was necessary before analysis. The 10B511B ratio using the normal abundance of the two isotopes of boron is 0.2466.1 For the purpose of this study, For the VG PlasmaQuad a narrow mass range scan of 5–11.5 u for 2 min was used, which included the two B peaks, the background B present was assumed to have the normal abundance for the two isotopes. Boron is known to form 10B and 11B, and one or both of the two internal standards, lithium (7Li) and beryllium (9Be) (Table 2, boron content). volatile compounds and is mobile in the natural setting.Table 1 Operating conditions for ICP-MS instruments VG PlasmaQuad Mark 1— Torch box VG Model 765 Rf generator (ICP) Henry electronics Rf incident power 1370 W Reflected power 5 W (usually 2–3W) Nebuliser Meinhard concentric Type A (TR-30-A3) Spray chamber Scott, double-pass Sample solution uptake rate (Gilson, Minipuls) 0.8 ml min-1 Argon plasma gas flow rate 14.5 l min-1 Argon auxiliary gas flow rate 0.8 l min-1 Argon nebuliser gas flow rate 0.76 l min-1 Sample cone Nickel, aperture 1 mm Skimmer cone Nickel, aperture 0.75 mm Load coil–sampling cone distance 10 mm Electron multiplier Continuous dynode ion multiplier (Channeltron, Galileo) Multichannel analyser TN7200 (Tracor Northern) Data system Compaq 386/20e computer Gas flows and torch position were optimised for maximum sensitivity Perkin-Elmer SCIEX Elan 5000— Rf incident power 1010 W Nebuliser Ryton, crossflow Spray chamber Ryton with flow spoiler Argon plasma gas flow rate 15.0 l min-1 Argon auxiliary gas flow rate 0.8 l min-1 Argon nebuliser gas flow rate 0.9 l min-1 Sample cone Nickel, aperture 1.1 mm Skimmer cone Nickel, aperture 0.9 mm Load coil–sampling cone distance 10 mm Electron multiplier Continuous dynode ion multiplier (Channeltron, Galileo) Data system IBM Personal System/2 Model 70 386 computer All other conditions were standard Gas flows were optimised for maximum sensitivity 1546 J.Anal. At. Spectrom., 1999, 14, 1545–1556Table 2 Scan parameters for boron determination for Perkin-Elmer SCIEX Elan 5000 and VG PlasmaQuad Mark 1 instruments Boron content Boron scan: PE SCIEX Elan VG PlasmaQuad (~1 min narrow (~2 min narrow 10B511B determination: Parameter mass range scan) mass range scan) PE SCIEX Elan PE SCIEX Elan Scanning mode Single continuous Multi-channel Peak hopping Peak hopping scan Mass range (u) 4–13 5–11.5 7, 9, 10 and 11 9, 10 and 11 No.of channels 100 512 4 3 Dwell time/ms 100 0.16 80 80 No. of sweeps 6 1500 75 143 Acquisition/s 62 122 25 36 No. of replicates 5 5 Total acquisition time/s 126 180 Similarly, 2 min scans (Table 2, boron content) in the peak the second internal standard for analysis. Since the degree of ionisation of Be is only 75%,42 a higher concentration was hopping mode using the 10B, 11B and one or both of the two internal standard peaks, 7Li and 9Be, were adequate for used (30 mg l-1 Be).For later analysis where B levels were known to be near the detection limit, Be was used as the determining the relatively high B concentrations at the beginning of the study with the PE SCIEX Elan 5000. Later, for internal standard, since contamination with Li was more likely. samples with very low 10B concentrations, where accurate 10B511B ratio determination was essential, the maximum acqui- Sample digestion/preparation sition time (3 min) was utilised for the sample volume available Two basic protocols were developed for sample digestion: the and only one internal standard peak, 9Be in this case (Table 2, first was for blood and plasma and the second was for other 10B511B ratio determination).biological samples including brain tissue, brain homogenate, The normal resolution mode was used for all analyses cells, cells solubilised in 1 M NaOH, incubation medium and (resolution set at 0.8 u at 10% peak height), since with high wash medium. There were variations with the second method resolution (resolution set at 0.6 u at 10% peak height) not depending on the type of sample, which aVected the ease of only the intensities but also the precision for isotope ratio digestion. A two- or three-stage digestion procedure was determination were reduced. utilised in part to reduce C levels in the final digest.The 12C peak could overlap the 11B peak if high C levels were present.Reagents The reduction of the C content in digests also helped to The reagents used were of the highest purity available unless prevent blockage of the torch and nebuliser and deposition on stated otherwise (Aristar or AnalaR grade from BDH, Poole, the cones, enabling long runs to be utilised. Dorset, UK, or their equivalents from Johnson Matthey, Where appropriate, samples were pipetted into quartz flasks Royston, Hertfordshire, UK).De-ionised water was provided for digestion. Biological samples were weighed accurately by by a Milli-Q system (Millipore, Watford, UK). diVerence either using the container in which the sample Stock mono-elemental standard solutions at 1000 mg l-1 arrived or the quartz flask used for digestion. Samples were boron (BDH, Spectrosol ) and 10 000 mg l-1 lithium and beryl- only analysed in duplicate if suYcient was available, which in lium (Johnson Matthey, Specpure) were used to prepare sub- practice only applied to a limited number of samples of blood, stock solutions at 10 and 100 mg l-1, containing no acid.plasma and media and one batch of cells solubilised in 1 M Working standard solutions of 1 mg l-1 B, 1 mg l-1 Li and NaOH. For blood and plasma, duplicate samples were pre- 5 mg l-1 Be containing no acid were used for final dilutions. pared if the total mass of the sample was >0.5 g and for The normal wash solution used for ICP-MS was 1% v/v media and solubilised cells if the volume of the sample was HNO3–0.1% v/v Triton X-100 solution, which was prepared >1 ml.The biological sample size range, except the one batch by diluting 10 ml of HNO3 and 1 ml of Triton X-100 to 1 l of solubilised cells, was 48–934 mg [Nyomura et al.35 recwith de-ionised water. De-ionised water was also used. ommended a sample size of 0.5 g, particularly for samples A 5%v/v Triton X-100 solution was prepared by diluting with a low B concentration (<3 mg g-1)]. 50 ml of Triton X-100 to 1 l with de-ionised water. At least three blanks were utilised, one with no internal standard, and treated identically with the samples but replacing Methods the sample with the same volume of de-ionised water. A further blank was added if there was a likelihood of the All standard solutions were prepared and final sample dilutions presence of high levels of B requiring sample dilution. were carried out in 30 ml polycarbonate containers fitted with screw caps (Universals, Sterilin, Richmond, Surrey, UK). Where possible, samples were analysed in duplicate.Whole blood and plasma. Approximately 0.5 g of blood or plasma (range 0.17–1.2 g) was weighed accurately by diVerence Initially an internal standard of 10 mg l-1 Li was used. Beryllium was also a satisfactory internal standard and was and placed in a 25 ml quartz conical flask. The internal standard, 100 ml of a 1mg l-1 Li solution or 60 ml of a 5mg l-1 used for blood samples for which lithium heparin had been added as an anticoagulant.The internal standard was added Be solution, was added, then 1–2 ml of de-ionised water, 1 ml of concentrated HNO3 (16M) and 100 ml of concentrated before acid digestion of the samples. For some digested samples, the B concentration was too high (>100 mg l-1) for H2SO4 (36 M). The contents of the flask were then evaporated on a hot-plate (150 °C) until fumes of H2SO4 were produced.analysis without further dilution. In these cases, since an internal standard had already been added before digestion, After cooling for ~10 min, a further 1 ml of concentrated HNO3 was added to each flask and evaporated to dryness the alternative internal standard was added on sample dilution for analysis; to avoid confusion, 10 mg l-1 Li was added as the (with slight charring). After cooling, 0.2 ml of concentrated HNO3 was added to each flask and warmed gently on the hot- first internal standard for digestion and 30 mg l-1 Be as J.Anal. At. Spectrom., 1999, 14, 1545–1556 1547plate, then 2–3 ml of de-ionised water and finally 0.3 ml of The BPA solutions were provided in 1 M NaOH. For the lowest dilution utilised (1+1.7), 100 ml concentrated HNO3 concentrated HNO3 to dissolve the residue. The flasks were removed from the hot-plate and allowed to cool for several were added to a total solution volume of 2.7 ml to neutralise the alkali.hours or overnight. The resulting solutions were then transferred quantitatively into weighed polycarbonate containers Aqueous calibrating standards and made up to 10 g by weight, using 4–6 washings with de-ionised water. Owing to the variety of types of sample analysed, aqueous standards were utilised. Brain tissue, brain homogenate and cells. The sample prep- A series of aqueous calibrating standards for B were pre- aration was as described for the analysis of blood and plasma pared in polycarbonate containers at 0, 5, 10 and 20 mg l-1 but there was no addition of concentrated H2SO4.The sample (occasionally also 40 and 80 mg l-1) containing either 10 mg l-1 size was usually smaller, generally 100 mg (range Li or 30 mg l-1 Be as an internal standard. Working standard 33–460 mg). solutions of 1 mg l-1 B, 1 mg l-1 Li and 5 mg l-1 Be were used. Standards with a total volume of 10 ml were prepared Solubilised cells. Digestion of cells solubilised in 1M NaOH by pipetting 0, 50, 100, 200 ml (400 and 800 ml ) of the 1 mg l-1 was more diYcult than that of non-solubilised cells, requiring B solution into polycarbonate containers.If the internal stana three-stage procedure. dard was Li, then 9.90, 9.85, 9.80, 9.70 ml (9.50 and 9.10 ml ) Samples of solubilised cells of known volume (5 ml ) were of de-ionised water were added to each container, respectively, transferred into 25 ml quartz flasks with three washes of followed by 100 ml of 1mg l-1 Li solution to all the containers.de-ionised water. The internal standard, 100 ml of 1mg l-1 Li If the internal standard was Be, then 9.94, 9.89, 9.84, 9.74 ml solution or 60 ml of 5mgl-1 Be solution, was added, then (9.54 and 9.14 ml ) of de-ionised water was added to each 1 ml of concentrated HNO3, and the contents were carefully container, respectively, followed by 60 ml of 5mg l-1 Be solu- evaporated to dryness on a hot-plate (150 °C), to avoid spitting tion to all the containers.(heavy residue). After cooling, a further 1 ml of concentrated HNO3 was added to each flask and evaporated to dryness Results and discussion (residue slightly reduced). Then, after cooling, a further 1 ml of concentrated HNO3 was added to each flask and evaporated Washout for boron to dryness (residue reduced). After cooling, 0.5 ml of concentrated HNO3 was added to each flask and warmed gently on In general, the wash utilised between samples was 1 min with the normal wash solution (1% v/v HNO3–0.1% v/v Triton the hot-plate, then 2–3 ml of de-ionised water with gentle warming (residue dissolved).The flasks were removed from X-100 solution) followed by 3 min with fresh de-ionised water so an automatic sampler could not be used. Before a run of the hot-plate and allowed to cool for several hours or overnight. The resulting solutions were transferred into weighed samples was initiated, a 3 min wash with the normal wash solution followed by a 6 min wash with fresh de-ionised water polycarbonate containers and made up to 10 g by weight using 4–6 washings with de-ionised water.was used, in order to minimise background B levels. To reduce run times, standard solutions were run in order of ascending concentration utilising a wash of 1 min with de-ionised water Media. Media samples also required a three-stage digestion procedure. between samples. For samples known to contain low B levels from the initial rapid 1 min scan, the wash with 1% v/v A 1 ml volume of medium and 2 ml of de-ionised water were placed in a 25 ml quartz flask.The internal standard, HNO3–0.1% v/v Triton X-100 solution was omitted, since the wash itself contained low levels of B, and replaced by a wash 100 ml of 1mg l-1 Li solution or 60 ml of 5mg l-1 Be solution, was added, then 1 ml of concentrated HNO3, and the contents of 3 min with fresh de-ionised water. Memory eVects are thought to be caused by the release of were carefully evaporated to dryness on a hot-plate (150 °C), to avoid frothing (frothy yellow residue).After cooling, 0.5 ml adhered B from the instrument to subsequent samples and B volatilisation in the spray chamber.35 Since preliminary scans of concentrated HNO3 was added to each flask and evaporated to dryness (residue slightly reduced). After cooling, a further had established the approximate B concentration in each sample, care was taken to run samples in ascending order of 0.5 ml of concentrated HNO3 was added to each flask and evaporated to dryness (fine yellow residue).After cooling, concentration and check on background levels after running high standards, unexpectedly high samples and occasionally 0.2 ml of concentrated HNO3 was added to each flask and warmed gently on the hot-plate, then 2–3 ml of de-ionised during the run. There was a tendency for background B counts s-1 to slowly water with gentle warming (the residue should dissolve).The flasks were removed from the hot-plate and allowed to cool fall during runs. This eVect was more pronounced with the VG PlasmaQuad where the background concentration could for several hours or overnight. The resulting solutions were transferred into weighed polycarbonate containers and made fall by the equivalent of 1 mg l-1 total B per hour. Wash solutions were easily contaminated by B. The utilis- up to 10 g by weight using 4–6 washings with de-ionised water.ation of fresh wash solutions was essential. The Perkin-Elmer AS-90 autosampler did not have a continuous rinse facility Preparation of samples not requiring digestion. Other reagents used by Clinical Neurosciences were analysed to check for B and the wash container was relatively small, ~250 ml, increasing the likelihood of cross-contamination. This was particularly contamination, including saline solution, TRIS buVer, Evan’s Blue solution and lithium heparin solution.The volumes important when samples containing very variable quantities of B were run (range <1–100 mg l-1). The VG PlasmaQuad provided varied from 0.2 to 1.1 ml and were made up to 10 ml with de-ionised water in polycarbonate containers and spiked autosampler was a Gilson Model 221 with a continuous rinse facility, but the PlasmaQuad was only used at the beginning with an internal standard, usually 30 mg l-1 Be, for analysis by ICP-MS. of the study.The 0.5% v/v Triton X-100 solution contained up to The B-containing compounds analysed were provided in solution, including boric acid, o-carboranylacetic acid and p- 1mg l-1 total B and the 2% v/v HNO3 solution contained up to 2.5 mg l-1 total B. Solutions containing ammonia were boronophenylalanine (BPA). Suitable dilutions in polycarbonate containers using de-ionised water were made, spiked with found to contain very high levels of B (up to 100 mg l-1), which precluded the direct dilution of blood solutions for 10 mg l-1 Li or 30 mg l-1 Be as an internal standard. 1548 J. Anal. At. Spectrom., 1999, 14, 1545–1556analysis, since ammonia was present in the normal diluent and to fall during long runs, as shown by the second aqueous calibration curve, possibly related to matrix eVects. wash solutions used for blood.43 Despite the presence of B in HNO3, the high acid content Although Be has consistently been reported to be the best internal standard to use for B determination by of the sample digests (2–5% HNO3) appeared to aid washout and reduce background levels.Alternatively, this could have ICP-MS,33–35,44 several research groups have found that HNO3 suppresses the Be signal more than the B signal.34,35 However, been caused by matrix eVects such as a memory eVect giving rise to matrix suppression.41 this would not explain the non-linearity of the tissue digest calibration graphs near the blank level. (One paper reported The lowest background B level attained was 3.7 mg l-1 total B, which was slightly higher than that reported by other signal enhancement with both B and Be with 2% HNO3.36) A possible solution to the problem would be to utilise a constant workers.33 Vanhoe et al.33 found the background level could not be reduced below 1.7 mg l-1 B, was similar with sub-boiled HNO3 concentration throughout any analytical run, for the sample digests and blanks, calibration standards and blanks 0.14 M HNO3 and water (~2 mg l-1 B), and did not originate from the conventional sample introduction system of boro- and wash solutions, which was used coincidentally in one study.33 silicate glass.They presumed that the background B level originated from the Milli-Q water. Vanhoe et al.33 reported a substantial memory eVect with standards containing >50 mg l-1 B, lasting for ~15 min with 100 mg l-1 B. They therefore ran standard solutions at the end Calibration graphs of the run.Fig. 1 shows the result of an experiment where a set of aqueous calibration standards were run followed by three sets of Variable blanks and matrix suppression standards containing, respectively, HNO3 digests of blood, plasma and brain tissue, then a second set of aqueous stan- As mentioned previously, checks were made on background dards. The graphs were all normalised to Be and subtracted B levels during runs. A set of standard solutions was usually using the lowest aqueous blank.All the graphs were linear run first, but if the background levels fell substantially during between 5 and 80 mg l-1 B. The linear aqueous calibration a run a second calibration curve was constructed at the end graphs had regression equations y=5539x+3432 (r2=0.9996) of the run. A blank or a low standard solution (5 mg l-1 B) and y=5802x+2046 (r2=0.9990), respectively, where y is the was therefore run after every five samples. total B counts s-1 and x is the total B concentration in The use of the internal standard successfully compensated mmol l-1.for the matrix eVect produced by the high Na concentration Digest blanks contained high B concentrations either in the digests of cells solubilised in 1 M NaOH [0.05 M or because washout was not complete after the previous high 1.15 mg ml-1 Na (0.115% Na)]. Both B and Be (the internal standard or possibly due to matrix eVects. The digests con- standard) signals were suppressed by the Na, thus giving the tained 2–5% HNO3.Standard solutions were run in order of impression that the digests contained less B than the digest ascending concentration utilising a wash of 1 min with blanks. Matrix suppression of the Be signal was 69%. de-ionised water between solutions. The general wash method Similarly, the lowest dilutions (1 in 2.7) of the BPA solutions for B samples was only utilised between diVerent sample types. in 1 M NaOH analysed contained 0.37 M NaOH or Since the graphs were normalised, the non-linearity of the 8.5 mg ml-1 Na (0.85%).In this case the internal standard calibration graphs may have been caused by an increasing was Li and the Li signal was suppressed by 48%. matrix eVect at lower B concentrations, not compensated by It is well known that light analytes are more seriously the internal standard. An explanation for the non-linearity of aVected by matrix eVects than heavier analytes.45,46 However, calibration graphs near the blank level, particularly for a light suppression by 10% of the Cu, Zn, Co and Bi signals by element such as B, was suggested by Gregoire in 1990.41 The 1 mgml-1 Na was reported by Gray and Date47 and Douglas molar ratio of B to dissolved solids increases with increasing and Houk.48 Gregoire30 reported suppression of the B signal B concentration so that the matrix eVect can decrease with at >2 mgml-1 Na, with 3% suppression at 2.1 mg ml-1 Na increasing B concentration.and 12% suppression at 4.2 mg ml-1 Na. The suppression Fig. 1 also illustrates the tendency for aqueous blank levels experienced here was substantially higher, presumably because the solubilised cell digests provided a more complex Nacontaining matrix and the BPA solutions had a higher Na content. Probst et al.36 investigated the matrix eVects of single elements on B and Be count rates and found, similarly, considerable suppression with 10 mg ml-1 NaCl or NaNO3.Nyomura et al.,35 when comparing sample preparation methods, found B concentrations in biological reference materials to be slightly higher (although still within an acceptable range) in samples prepared by wet dissolution, presumed to be caused by leakage from borosilicate glass containers. They also found that B concentrations in the blanks tended to be more variable and higher when prepared by wet digestion. Evans and Kra�henbu� hl32 stressed the diYculty of obtaining accurate blank values when determining the B content of biological materials and their importance in determination by isotope dilution.For samples containing low levels of B, the digest blanks contained very similar levels of B to the sample digests. This was true for brain tissue, brain homogenate, blood and plasma digests and even after normalisation to the internal standard, blank subtraction was diYcult or impossible. This was most common with the brain tissue digests and blanks, with which Fig. 1 Boron calibration curves by ICP-MS: aqueous solutions and blood, plasma and brain tissue digests. there was matrix suppression which could not be compensated J. Anal. At. Spectrom., 1999, 14, 1545–1556 1549by the internal standard. The problem did not arise with media digests. Correction for peak overlap by carbon with the VG PlasmaQuad As mentioned previously, the VG PlasmaQuad Mark 1 only operated in the multi-channel scanning mode. For samples containing very high C levels, the 12C peak overlapped the adjacent 11B peak.For these samples it was necessary to correct for the increased counts s-1 on the 11B peak and the count rate was only measured for the part of the peak where there was no overlap, which was normally between 10.5 and 11.3 u. The same area was then measured on the B standard peaks and the mean percentage of counts s-1 for the whole peak, lying between 10.5 and 11.5 u, was calculated. Once the normal percentage of counts s-1 in this area was known, the same percentage could be added to the sample peaks where only part of the peak count rate had been measured.Use of 10B511B ratio determination for the detection and measurement of elevated 10B levels Initial sample total B and 10B concentrations were relatively high: 10B concentrations in blood 14–19 mg g-1, brain tissue 1–13 mg g-1, brain homogenate 2–249 mg g-1 and wash medium 0.007–17.2 mg ml-1. Total B concentrations were slightly higher: blood 14–19 mg g-1, brain tissue 3–19 mg g-1, brain homogenate 9–281 mg g-1 and wash medium 0.008– 18.0 mg ml-1. For these samples the use of an internal standard successfully compensated for matrix eVects.Rapid narrow mass range scans were made of sample digests for comparison. Fig. 2 illustr typical results at the beginning of the study when samples contained very high levels of 10B, whereas Fig. 3 illustrates results later in the study where elevated 10B levels were reduced.Fig. 2 shows an increasing 10B content of tissue digests (brain homogenate) following Fig. 2 10B511B ratios in brain homogenate digests by ICP-MS. initial trial incubation with 10B-enriched boric acid. The 10B concentrations in the original brain homogenate reached a maximum of 249 mg g-1. Fig. 3 illustrates higher C levels in The optimisation of the scan parameters and scan times for 10B511B ratio determination enabled instrumental DLs (using brain tissue digests than for the digests in Fig. 2; a digest blank is included for comparison. All three tissue digests in three times the standard deviation of replicate determinations made at or near the blank level ) for total B of <0.1 mmol l-1 Fig. 3 contained elevated levels of 10B, with concentrations in tissue of 1.2–13 mg g-1 after incubation with BPA. Carbon (<1 mg l-1) and for 10B of <0.01 mmol l-1 (<0.1 mg l-1) to be attained, despite the relatively high background con- levels were not high enough to prevent analysis.The high cost of the 10B-containing compounds was an centrations. The calculated sample DLs for B (10B, 11B and total B) in biological samples were as low as 0.001 mmol g-1 incentive to work at relatively low concentrations, i.e., nanomolar levels, where problems of contamination and high B (0.01 mg g-1) (10-9 M BPA solutions). The DLs attained were related to the amount of sample digest blanks and background concentrations became signifi- cant.For the analysis of samples containing low B levels, the utilised; therefore a lower DL was attained for solubilised cells of 0.0001 mmol ml-1 (0.001 mg ml-1) 10B and for the reagent digest blanks contained very similar B levels to those of the sample digests, where, even after normalisation to the internal solutions analysed 10B concentrations as low as 0.0003 mmol g-1 (0.003 mg g-1) were reported. These sample DLs standard, blank subtraction was diYcult or impossible.This was most common with brain tissue digests but was also found compare favourably with those of other workers.33,36 10B511B ratios were monitored from the beginning of the with brain homogenate, blood and plasma digests. For some samples, the results of analysis were therefore reported without study and it became clear that increases in sample 10B level could be detected using the 10B511B ratio, below the calculated digest blank subtraction or with both ‘running blank’ and digest blank subtraction.The ‘running blank’ was the instru- DL for total B or 10B. 10B511B ratio determination was suYciently precise to enable statistically significant increases mental blank with an aqueous standard blank solution. The power of stable isotope ratio determinations in detecting in 10B concentration to be detected and confirmed at these low B levels. Despite achieving low DLs, elevated 10B levels increases in 10B in situations where no total B concentrations could be accurately determined was illustrated using these could not be confirmed without the accurate determination of 10B511B ratios.Running 10B511B ratios for standard and samples for which normal blank subtraction could not be used. The mean RSD for five replicate determinations of 10B511B sample solutions were utilised for novel methods of calculating elevated 10B levels in the samples. ratio at the blank level was ~2% (range 0.5–5.2%, n=7, three diVerent analytical runs) and similar precision was obtained Signals for the two B isotopes were high enough to permit the determination of 10B511B ratios in all samples, including with low B sample digests.The precision for isotope ratio determination with the aqueous standard solutions was gener- digest blanks. The lowest background B level was ~4 mg l-1 total B, equivalent to a minimum sensitivity for total B of ally <1.5% (range 0.5–2.8%, n=10, three diVerent analytical runs).~2500 counts s-1. 1550 J. Anal. At. Spectrom., 1999, 14, 1545–1556using the mean total B counts s-1, normalised to the internal standard, obtained for the running blank (aqueous blank) and the accurate running 10B511B ratio obtained using the B standards. The lowest running blank total B concentration obtained for any run was 3.7 mg l-1. For this particular run, the running 10B511B ratio using the boron standards was 0.236 (s 0.003; RSD 1.24%; n=5), whereas if the three aqueous blanks had also been used to calculate the running 10B511B ratio, it would have been 0.242 (s 0.009; RSD 3.65%; n=8).During the same run, the running 10B511B ratio for the digest blanks was 0.248 (s 0.006; RSD 2.58%; n=4). The 10B511B ratio using the normal abundance of the two isotopes is 0.2466. Since the bias of the mass spectrometers varied slightly from day to day, a correction factor was used to normalise the running 10B511B ratio sample results, based on the running 10B511B ratio for the B standards.The correction factor assumed that, if no mass bias existed, the standards would have given a 10B511B ratio of 0.2466, since no reference materials of known 10B511B ratio were utilised. The corrected running 10B511B ratio sample results were then reported, allowing comparison between diVerent runs. For digest blank subtraction, the average intensities for the individual 10B and 11B isotopes were utilised in the normal manner. (2) Calculation of elevated 10B levels.Elevated 10B levels were calculated using the running blank subtracted results in micromoles (usually mmol g-1). It was assumed that any 10B present due to contamination or contributions from any source other than a 10B-containing compound would be related to the 11B concentration by the normal abundance of the two isotopes. This ‘blank’ level of 10B could therefore be calculated by multiplying the 11B concentration in the sample by the normal 10B511B ratio of 0.2466.The background 10B level could then be subtracted from the running blank subtracted 10B concentration to give the elevated 10B level in the sample. In this instance, using the running 10B511B ratio for the Fig. 3 10B511B ratios in tissue digests by ICP-MS. standards instead of 0.2466 gave a more accurate result. In summary, the elevated 10B concentration was given by the Calculations following equation: For most analyses by ICP-MS, a convention has developed El[10B]=S[10B]-(S[11B]×RS10B511B) that analysis is carried out with amounts measured in microwhere El[10B]=elevated 10B concentration in sample grams since the detector is more sensitive to elements with (mmol g-1), S[10B]=running blank subtracted 10B con- higher m/z values.At the high end of the m/z range, the centration in sample (mmol g-1), S[11B]=running blank relative diVerences in m/z between isotopes is small, but at the subtracted 11B concentration in sample (mmol g-1) and low end the diVerence is significant.It was therefore convenient RS10B511B =running 10B511B ratio for the standard boron to change to working in micromoles when calculating concensolutions. trations of 10B and 11B, especially at concentrations near to Occasionally it was necessary or more convenient to calcu- the DL. late the elevated 10B concentration using the normalised A new method of calculating elevated 10B levels was develcounts s-1 for B, e.g., when the sample counts s-1 for 11B oped using the determined 10B511B ratios.The method was of were lower than those obtained for the running blank. In these particular value for analysis at the DL, when blanks and cases the elevated counts s-1 for 10B were calculated in a samples contained similar B concentrations and digest blank similar manner and the elevated 10B concentration obtained subtraction could not be utilised. by dividing the elevated counts s-1 for 10B by the X coeYcient for B in counts s-1 per mmol B, using the following equation: (1) Calculation of running 10B511B ratio and running blank 10B and 11B concentrations. The ‘running 10B511B ratio’ was El[10B]= 10Bcs-1-(11Bcs-1×RS10B511B) XB the 10B511B ratio of the standards and samples which did not contain elevated levels of 10B measured during an individual where El[10B]=elevated 10B concentration in sample run.An accurate value of the running 10B511B ratio was (mmol l-1), 10Bcs-1=normalised counts s-1 for 10B (non- required for three reasons: first, for comparison with sample blank subtracted), 11Bcs-1=normalised counts s-1 for 11B 10B511B ratios in order to detect elevated levels of 10B; second, (non-blank subtracted) and XB=X coeYcient for boron to determine the elevated 10B levels in samples; and third, to (counts s-1 per mmol B).determine accurately the 10B and 11B concentrations (count rates) in the running blank. Mass bias interference The most accurate value of the running 10B511B ratio was obtained using the mean 10B511B ratio for the B standards.Assuming that the B standards utilised contained the normal abundance of the two isotopes, the mass discrimination of the The running blank 10B and 11B count rates were then calculated J. Anal. At. Spectrom., 1999, 14, 1545–1556 1551instruments was calculated using the running 10B511B ratio ised running 10B511B ratios were reported (graph j). A more detailed analysis of the data obtained with the final dilutions standard results.For the PE SCIEX Elan 5000, the initial mass bias obtained for 10B511B ratios was -11.3% (range of ~2 mg l-1 10B is given in Fig. 4(B). The eVect of B contamination from NaOH and water was -5.5 to -14.9%, n=6). After optimisation of the mass calibration, the mass bias was reduced to -6.4% (range -1.5 to lower the 10B511B ratio as shown in all three graphs. The lowest ratios were seen when the final B concentrations were to -11.4%, n=10).The mass bias appeared to be generally lower for the VG PlasmaQuad Mark 1 at -5.1% (range -2.3 2 mg l-1 (2×10-7 M). When the final (analysed) B concentrations were adjusted to 10 mg l-1, the 10B511B ratio increased to -7.0%, n=3), where scan conditions were not optimised for isotope ratio determination. Gregoire30 also obtained the as the original concentration of BPA increased, which reflected decreased contamination from NaOH as the dilution for mass bias towards the 11B isotope.Evans and Kra�henbu� hl32 obtained a greater mass bias of 20% with a PE SCIEX Elan analysis increased (graph ii). The solutions containing the highest B concentrations gave ratios lower than the true ratios 5000 when determining 11B510B ratios in biological reference materials. (graph i); for the 10-3M BPA solution diluted to 450 mg l-1 B, this reflected the saturation of the detector with 10B. Gregoire30 found matrix induced mass discrimination with 3 mgml-1 Na.There was only evidence for matrix induced The data obtained with the final dilutions of ~2 mg l-1 10B showed the power of stable isotope measurements by ICP-MS mass discrimination during this study with the lowest dilutions of solutions of boronophenylalanine in NaOH (1 in 2.7), [Fig. 4(B)]. The increase in the normalised 10B511B ratio from the baseline value of 0.247 at 1×10-9 M to 1×10-10M BPA containing 0.37 M NaOH or 8.5 mg ml-1 Na; the mass bias for 10B511B ratios was lower than -2%.(original concentration) to 0.286 at 1×10-8 Mwas statistically significant ( p<0.001). The standard solutions of 0.4–4.0 mg l-1 total B (4×10-8–4×10-7M) gave a mean RSD for five 10B511B isotope ratio determination of boronophenylalanine solutions replicate determinations of 10B511B ratio of 2.1% (range 1.9–2.5%). The precision was similar for the samples. The Solutions of BPA in 1 M NaOH from 10-3 to 10-10 M were data therefore demonstrated, despite high blanks, the possibilfurther diluted for analysis by ICP-MS in order to determine ity of working with concentrations as low as 10-8M BPA, to whether work at 10-9 MBPA was feasible.Fig. 4(A) illustrates study the uptake of 10B in tissues and cells. the resulting 10B511B ratio data. Initially the 10-3 and 10-4 M solutions were diluted for analysis to give 450 (4.5×10-5 M) 10B511B ratio determination in tissue grown with boronoand 50 mg l-1 B, respectively (graph i). The original solutions phenylalanine. A series of tissue digests were analysed.One were then diluted to give~2 mg l-1 10B and re-analysed (graph investigation included tissue samples (mass range 33–69 mg) j); HNO3 was added to neutralise the NaOH with the lowest which had been grown with increasing concentrations of BPA dilutions (10-7 to 10-10 M) as described previously. The from 1×10-8–1×10-3 M. Results for B content and 10B511B original 10-3–10-6M solutions were also diluted to ratio were not digest blank subtracted since the digest blanks ~10 mg l-1 10B for further analysis (graph ii).contained relatively high B levels (Table 3). The 10B511B ratios for the solutions containing 10 mg l-1 The results again illustrated the power of stable isotope were blank subtracted (graphs i and ii), but this was not ratio measurements by ICP-MS. The digest blanks and digests possible for those diluted to ~2 mg l-1 10B where the normal- from tissue grown with 10-5–10-8 M BPA gave normalised running 10B511B ratios at the baseline value of 0.247, but the increase to 0.300 with the tissue grown with 10-4 M BPA was statistically significant.A further increase to a ratio of 1.09 occurred with the tissue grown with 10-3 M BPA. In this particular case the 10B511B ratios calculated using the nonblank subtracted B concentrations for the two isotopes show the same trends. Elevated 10B levels could not have been detected without the accurate determination of 10B511B ratios.Fig. 5 shows narrow mass range scans of some of the tissue digests from the same investigation: the tissue digests grown with 10-3 and 10-4 M BPA (tissue digests 3 and 4) with increased 10B levels and the tissue digests grown with 10-6 and 10-8 M BPA (tissue digests 6 and 8) with normal abundance of the two isotopes. Confounding factors in 10B511B determination Confounding factors all worked against a false detection of an increased 10B511B ratio, with the exception of overloading the detector by running a sample with a very high B concentration, e.g., 450 mg l-1 B, as described previously.The confounding factors included 12C overlap on the 11B peak with sample digests containing high C levels, and matrix suppression. The majority of the medium digests in Table 6 had a slightly lower 10B511B ratio than the blanks due to 12C overlap on the 11B peak. Matrix mass discrimination also tended to decrease the 10B511B ratio. 10B content of boron-containing compounds The 10B content was determined for three diVerent boroncontaining compounds: boric acid, p-boronophenylalanine and Fig. 4 ICP-MS analysis of 10B511B ratios in boronophenylalanine solutions. carboranylacetic acid. 1552 J. Anal. At. Spectrom., 1999, 14, 1545–1556Table 3 10B511B ratio determination in tissue digests by ICP-MS: tissue grown with boronophenylalanine. The blanks used for acid digestion contained relatively high levels of boron.Results were therefore not digest blank subtracted Corrected 10B/ 11B/ Total B/ 10B/ 11B/ Total B/ Final running Sample Sample mg g-1 mg g-1 mg g-1 mmol g-1 mmol g-1 mmol g-1 10B 511B ratioa 10B511B ratio grown with Tissue 8 0.7 3.1 3.8 0.07 0.29 0.35 0.231 0.245 10-8 M BPA Tissue 7 0.4 2.3 2.7 0.04 0.21 0.25 0.198 0.251 10-7 M BPA Tissue 6 0.4 2.0 2.4 0.04 0.18 0.22 0.231 0.245 10-6 M BPA Tissue 5 0.3 1.5 1.8 0.03 0.13 0.16 0.234 0.247 10-5 M BPA Tissue 4 0.5 1.6 2.0 0.05 0.14 0.19 0.323 0.300 10-4 M BPA Tissue 3 1.4 1.0 2.4 0.14 0.09 0.23 1.57 1.09 10-3 M BPA Equivalent values for digest blanks 0.6 2.7 3.2 0.06 0.24 0.30 0.233 0.246 s 0.2 1.0 1.2 0.02 0.09 0.11 0.003 0.002 Running 10B511B ratio for the standard solutionsb 0.2331 s 0.0009 RSD (%) 0.39 aCalculated using boron calculations before rounding up.bThe running 10B511B ratio for the standard solutions indicates the normal abundance of the two boron isotopes as determined by the mass spectrometer during this run.were 1.5 mg l-1 and 2.7 mg l-1, respectively. The estimated level of 10B present in the boric acid was therefore 85%. The data in Fig. 4(A) (specifically graph i, aesults from the 1×10-4 M BPA solution diluted to 50 mg l-1 for analysis) showed that 96% of the B in the BPA sample was 10B. A sample of ~4 mg of carboranylacetic acid (CAA) was dissolved in 1 M NaOH, giving a 1.0 mg ml-1 solution, which was diluted for analysis to give 20 mg l-1 (0.099 M) and 40 mg l-1 (0.198 M) CAA.Since one molecule of CAA contains 10 B atoms, the predicted total B concentrations in the two solutions were 0.99 and 1.98 M, respectively. Analysis by ICP-MS gave total B concentrations slightly higher, 1.23 and 2.37 M. The carboranylacetic acid contained boron with a normal isotopic distribution; the normalised running 10B511B ratio obtained for the CAA solutions was 0.2450±0.0009 (mean±s). The mean running 10B511B ratio for the B standards was 0.2224±0.0009 (mean±s).Analytical precision from duplicates Analytical data for the precision from duplicates were obtained for 12 samples of blood during two runs (runs A and B), six samples of cells solubilised in 1 M NaOH and six samples of medium both analysed in the same run (run C). Table 4 gives the data from runs A, B and C using the conventional methods for calculating 10B and total B concentrations and Tables 5 and 6 give the results from the same analytical runs with the developed calculation methods utilising 10B511B ratio correction.Running blank subtracted results were reported for the blood samples in both cases since the blood digest blanks contained relatively high B levels. Only the blood samples in run B contained significant elevated levels of 10B. The 10B concentration range in the samples was <0.0002 mmol ml-1 in the solubilised cells, <0.007 mmol ml-1 in the media, <0.001 mmol g-1 in the blood samples of run A and 0.007–0.014 mmol g-1 in the blood samples of run B.The equivalent total B concentrations in the samples were Fig. 5 10B511B ratios in tissue digests by ICP-MS: uptake of boron- 0.001 mmol ml-1, <0.04 mmol ml-1, <0.002 mmol g-1 and ophenylalanine. 0.011–0.019 mmol g-1, respectively. Table 4 shows, using the conventional methods for calculation of 10B concentration, that the range and mean of within- The 10B content of a sample of boric acid was determined by measuring the elevated 10B level.The solutions analysed run RSDs for blood (only run B), solubilised cells and medium were very similar, all <17% and with means of ~7%. The contained relatively low levels of boric acid: 10 and 20 mg l-1 (0.162 and 0.323 mmol l-1), equivalent to ~1.7 and mean (range) RSDs were for blood 6.9% (1.2–16.8%), for solubilised cells 7.3% (1.5–16.6%) and for medium 6.9% ~3.4 mg l-1 B, respectively. The normalised running 10B511B ratios obtained for the two solutions were 0.496 and 0.752 (0.5–13.9%).Similarly, using the conventional methods for calculation, and the elevated 10B concentrations found in the two solutions J. Anal. At. Spectrom., 1999, 14, 1545–1556 1553Table 4 Analytical precision from duplicates for total B and 10B concentrations in blood, solubilised cells and medium Total B concentration/mmol g-1 10B concentration/mmol g-1 Type of Mean sample Sample No. 1st assay 2nd assay Mean s RSD (%) 1st assay 2nd assay Mean s RSD (%) reported Blood Run Aa,b— 1l <0.002 <0.002 <0.002 0.003 <0.001 <0.001 <0.001 0.001 2l <0.002 <0.002 <0.002 0.004 <0.001 <0.001 <0.001 0.001 Run Ba— 1W 0.019 0.019 0.019 0.0005 2.7 0.0143 0.0135 0.0139 0.0005 3.7 2W 0.019 0.014 0.016 0.003 19.9 0.0135 0.0106 0.0120 0.0020 16.8 3W 0.016 0.014 0.015 0.001 9.6 0.0099 0.0093 0.0096 0.0004 4.1 4W 0.019 0.017 0.018 0.001 4.9 0.0128 0.0125 0.0127 0.0002 1.2 5W 0.013 0.017 0.015 0.003 19.7 0.0082 0.0104 0.0093 0.0015 6.5 6W 0.014 0.016 0.015 0.002 10.4 0.0088 0.0098 0.0093 0.0007 7.3 1F 0.012 0.013 0.013 0.001 8.9 0.0083 0.0095 0.0089 0.0009 9.9 2F 0.014 0.014 0.014 0.0002 1.2 0.0092 0.0096 0.0094 0.0003 3.5 3F 0.012 0.011 0.012 0.001 10.5 0.0072 0.0068 0.0070 0.0003 4.2 6F 0.011 0.011 0.011 0.0001 0.8 0.0066 0.0068 0.0067 0.0001 1.9 Mean 8.8 6.9 Run Cc— Cells solubilised in 1 M NaOH and medium Cell digests 1 0.0011 0.0011 0.0011 0.00002 2.2 0.00018 0.00019 0.00019 0.000003 1.5 0.0002 2 0.0006 0.0006 0.0006 0.00002 3.0 0.00010 0.00011 0.00010 0.000007 7.0 0.0001 3 0.0007 0.0006 0.0007 0.00007 10.6 0.00013 0.00011 0.00012 0.000019 16.6 0.0001 4 0.0005 0.0005 0.0005 0.00002 3.5 0.00008 0.00008 0.00008 0.000006 7.2 0.0001 5 0.0005 0.0005 0.0005 0.000005 1.1 0.00009 0.00008 0.00009 0.000006 6.6 0.0001 6 0.0006 0.0006 0.0006 0 0.0 0.00010 0.00010 0.00010 0.000003 5.2 0.0001 Mean 3.4 7.3 Medium digests A 0.0261 0.0284 0.0273 0.0016 5.9 0.0044 0.0048 0.0046 0.0003 6 B 0.0238 0.0291 0.0264 0.0038 14.4 0.0041 0.0050 0.0045 0.0006 13.9 C 0.0334 0.0337 0.035 0.0002 0.6 0.0057 0.0058 0.0058 0.00003 0.5 D 0.0357 0.0389 0.0373 0.0023 6.2 0.0061 0.0067 0.0064 0.0004 6.1 E 0.0018 0.0019 0.0019 0.00005 2.6 0.0002 0.0002 0.0002 0.000008 5 F 0.0024 0.0022 0.0023 0.0001 4.3 0.0003 0.0002 0.0003 0.00003 10 Mean 5.7 6.9 aSample masses were 200–600 mg.Detection limits were 0.002–0.005 mmol g-1 for total B and 0.0002–0.0005 mmol g-1 for 10B.bVariation in digestion blank levels of boron increased detection limits for 10B in this run. cSample volumes were 1 ml. Detection limits were <0.001 mmol ml-1 for total B and 0.0001 mmol ml-1 for 10B. 1554 J. Anal. At. Spectrom., 1999, 14, 1545–1556Table 5 Analytical precision from duplicates for 10B and elevated 10B concentrations in blood. Running blank subtracted results with 10B511B ratio concentration 10B concentration/mmol g-1 Elevated 10B concentration/mg g-1 Sample No. 1st assay 2nd assay Mean s RSD (%) 1st assay 2nd assay Mean s RSD (%) Run A— 1l 0.0012 0.0009 .0010 0.0002 18.4 2l 0.0011 0.0011 0.0011 0.00002 2.2 Equivalent value for digest blanksa 0.004 0.002 Run B— 1W 0.0145 0.0138 0.0141 0.0005 3.5 0.130 0.123 0.126 0.005 4.0 2W 0.0137 0.0108 0.0123 0.0020 16.6 0.122 0.098 0.110 0.017 15.6 3W 0.0102 0.0096 0.0099 0.0004 4.3 0.085 0.083 0.084 0.001 2.0 4W 0.0130 0.0128 0.0129 0.0002 1.5 0.113 0.114 0.113 0.00001 0.01 5W 0.0085 0.0106 0.0096 0.0015 15.7 0.071 0.088 0.080 0.012 14.8 6W 0.0090 0.0100 0.0095 0.0007 7.5 0.077 0.083 0.080 0.005 6.1 1F 0.0088 0.0099 0.0094 0.0008 8.2 0.074 0.086 0.080 0.008 9.6 2F 0.0098 0.0101 0.0099 0.0002 1.9 0.081 0.086 0.083 0.003 3.7 3F 0.0076 0.0072 0.0074 0.0003 3.5 0.059 0.058 0.059 0.0007 1.1 6F 0.0069 0.0072 0.0070 0.0002 2.5 0.055 0.056 0.056 0.001 2.7 Equivalent value 0.008 0.003 for digest blanksa aRough estimate only.Table 6 Analytical precision from duplicates for running 10B:11B ratios in blood, solubilised cells and medium Corrected running 10B511B ratio Type of Sample sample No. 1st assay 2nd assay Mean s RSD (%) Blood Run A— 1l 0.257 0.233 0.245 0.017 7.0 2l 0.240 0.247 0.243 0.005 1.9 Equivalent value for digest blanks 0.259 0.007 2.6 Run B— 1W 1.84 1.66 1.75 0.13 7.2 2W 1.72 1.80 1.76 0.06 3.2 3W 1.19 1.40 1.30 0.15 11.6 4W 1.48 1.76 1.62 0.20 12.3 5W 1.16 1.23 1.19 0.05 4.2 6W 1.29 1.17 1.23 0.08 6.7 1F 0.94 1.22 1.08 0.19 17.9 2F 0.88 1.12 1.00 0.17 17.3 3F 0.84 0.87 0.86 0.02 1.8 6F 0.88 0.84 0.86 0.03 3.1 Equivalent value for digest blanks 0.269 0.004 1.4 Solubilised cellsa Run C— 1 0.264 0.263 0.263 0.001 0.4 2 0.265 0.275 0.270 0.007 2.7 3 0.274 0.267 0.270 0.005 0.7 4 0.261 0.264 0.263 0.002 1.8 5 0.275 0.269 0.272 0.005 0.9 6 0.266 0.269 0.268 0.002 0.9 Equivalent value for digest blanks 0.276 0.006 2.3 Medium Run C— A 0.233 0.233 0.233 0.002 0.05 B 0.236 0.234 0.235 0.001 0.5 C 0.236 0.236 0.236 0.0003 0.1 D 0.237 0.237 0.237 0.0001 0.06 E 0.186 0.176 0.181 0.007 3.8 F 0.190 0.180 0.185 0.007 3.8 Equivalent value for digest blanks 0.287 0.002 0.7 aCells were solubilised in 1 M NaOH.J. Anal. At. Spectrom., 1999, 14, 1545–1556 155511 Y. Mishima, M. Ichihashi, M. Tsuji, S. Hatta, M. Ueda, the range and mean of within-run RSDs for total B in blood, C. Honda and T. Suzuki, J. Invest. Dermatol., 1989, 92, 321S. solubilised cells and medium were more variable but all were 12 J.A. Coderre, J. Kalef-Ezra, R. G. Fairchild, P. L. Micca, L. E. <20% with means of 3–9% (Table 4). This was probably a Reinstein and J. D. Glass, Cancer Res., 1988, 48, 6313. result of the variable presence of residual carbon in the digested 13 J. A. Coderre, J. D. Glass, S. Packer, P. L. Micca and samples giving rise to overlap of the 12C peak on the adjacent D. Greenberg, Pigm. Cell Res., 1990, 3, 310. 14 S. Packer, J. A. Coderre, S.Saraf, R. Fairchild, J. Hansrote and 11B peak. For total B, the mean (range) RSDs were for blood H. Perry, Invest. Ophthalmol. Vis. Sci., 1992, 33, 137. 8.8% (1.2–19.9%), for solubilised cells 3.4% (0–10.6%) and 15 K. Z. Matalka, M. Q. Bailey, R. F. Barth, A. E. Staubus, D. M. for medium 5.7% (0.6–14.4%). Adams, A. H. Soloway, S. M. James and J. H. Goodman, in Specific results could not be reported for the blood samples Progress in Neutron Capture Therapy for Cancer, eds.B. Allen, D. in run A using the conventional methods for calculations since Moore and B. Harrington, Plenum Press, New York, 1992, p. 429. the B concentrations were lower than the calculated DLs, but 16 B. J. Allen, J. K. Brown, M. H. Mountford, S. R. Tamat, A. Patwardan, D. E. Moore, M. Ichihashi, Y. Mishima and could be reported using the developed calculation methods S. B. Kahl, Strahlenther. Onkol., 1989, 165, 163. utilising 10B511B ratio correction (Table 5). Table 5 gives the 17 O.Leukhart, M. Caviezel, A. Eberle, E. Escher and A. Tun-khi, running blank subtracted results with 10B5 11B ratio correction Helv. Chim. Acta, 1976, 59, 2184. for the duplicate blood samples and the equivalent concen- 18 M. F. Hawthorne, R. J. Wiersema and M. Takasugi, J. Med. trations for the digest blanks, with the calculated elevated 10B Chem., 1972, 15, 449. concentrations in the blood samples of run B. There was only 19 R. F. Barth, A. H. Soloway, R. G. Fairchild and R.M. Brugger, Cancer, 1992, 70, 2995. a slight diVerence between the recalculated 10B concentrations 20 G. Ko� hler and C. Milstein, Nature (London), 1975, 256, 495. and the original results and the mean (range) precision for 21 R. G. Fairchild, D. Gabel, B. Laster, D. Greenberg, W. Kiszenick duplicate analyses was slightly improved, 6.5% (1.5–16.6%). and P. L. Micca, Med. Phys., 1986, 13, 50. The mean (range) precision (RSD) for the elevated 10B levels 22 W. B.Clarke, M. Koekebakker, R. D. Barr, R. G. Downing and (mg g-1 10B) in the blood samples was 6.0% (0.01–15.6%). R. F. Fleming, Appl. Radiat. Isotop., 1987, 38, 735; Int. J. Radiat. The most useful information obtained using the developed Appl. Instrum., Part A. 23 D. E. Moore, J. Pharm. Biomed. Anal., 1990, 8, 547. calculation methods for identifying increased 10B level was the 24 K. Yoshino, M. Okamoto, H. Kakihana, T. Nakanishi, running 10B511B ratio and the elevated 10B level.Comparisons M. Ichihashi and Y. Mishima, Anal. Chem., 1984, 56, 839. of RSDs for running 10B511B ratios for duplicate samples with 25 F. Salinas, A. M. Pena, J. A. Murillo and J. C. J. Sanchez, Analyst, relevant data are made in Table 6. For the running 10B511B 1987, 112, 913. ratio, the precision was better for the solubilised cells and 26 C. J. Cook, S. V. Dubiel and W. A. Hareland, Anal. Chem., 1985, medium, which showed a normal abundance. The mean (range) 57, 337. 27 W.B. Clarke, C. E. Webber, M. Koekebakker and R. D. Barr, RSDs for duplicates samples were for blood 6.5% (1.8–17.9%), J. Lab. Clin. Med., 1987, 109, 155. for solubilised cells 1.2% (0.4–1.8%) and for medium 1.4% 28 S. R. Tamat, D. E. Moore and B. J. Allen, Anal. Chem., 1987, (0.05–3.8%). The ratios for the medium digests were slightly 59, 2161. lower than those for the digest blanks and standards, owing 29 W. F. Bauer, D. A. Johnson, S. M. Steele, K. Messick, D. L. to the high C levels in the digests, leading to a slight overlap Miller and W. A. Propp, Strahlenther. Onkol., 1989, 165, 176. of the 12C peak on the 11B peak. The data for elevated 10B 30 D. C. Gregoire, Anal. Chem., 1987, 59, 2479. 31 F. G. Smith, D. R. Wiederin, R. S. Houk, C. B. Egan and R. E. concentrations in the blood samples from run B are given in Serfass, Anal. Chim. Acta, 1991, 248, 229. Table 5, since they were the only duplicate samples with 32 S. Evans and U. Kra�henbu� hl, J. Anal. At. Spectrom., 1994, 9, elevated 10B levels, where the RSDs for duplicate samples were 1249. 5.6% (0.01–15.6%). 33 H. Vanhoe, R. Dams, C. Vandecasteele and J. Versieck, Anal. Chim. Acta, 1993, 281, 401. We thank Caroline Dore� for assistance with statistical analysis. 34 S. Evans and U. Kra�henbu� hl, Fresenius’ J. Anal. Chem., 1994, 349, 454. 35 A. M. S. Nyomura, R. N. Sah, P. H. Brown and R. O. Miller, Fresenius’ J. Anal. Chem., 1997, 357, 1185. References 36 T. U. Probst, N. G. Berryman, P. Lemmen, L. Weissfloch, T. Auberger, D. Gabel, J. Carlsson and B. Larsson, J. Anal. At. 1 Handbook of Inductively Coupled Plasma Mass Spectrometry, ed. Spectrom., 1997, 12, 1115. K. E. Jarvis, A. L. Gray and R. S. Houk, Blackie, Glasgow, 37 H. P. Schwartz, E. K. Agyel and C. C. McMullen, Earth Planet. 1992, p. 341. Sci. Lett., 1969, 6, 1. 2 D. N. Slatkin, Brain, 1991, 114, 1609. 38 M. Shima, J. Geophys. Res., 1963, 67, 911. 3 A. Mill, New Sci., 18 Nov., 1989, No. 1691, 56. 39 Handbook of Geochemistry, ed. K. H. Wedepohl, Springer Verlag, 4 M. Javid, G. L. Brownell and W. H. Sweet, J. Clin. Invest., 1952, New York, 1969. 31, 603. 40 A. J. Spivack and J. M. Edmond, Anal. Chem., 1986, 58, 31. 5 H. B. Locksley and L. E. Farr, J. Pharmacol. Exp. Ther., 1955, 41 D. C. Gregoire, J. Anal. At. Spectrom., 1990, 5, 623. 114, 484. 42 R. S. Houk, Anal. Chem., 1986, 58, 97A. 6 W. H. Sweet, A. H. Soloway and G. L. Brownell, Acta Radiol., 43 I. L. Shuttler and H. T. Delves, Analyst, 1986, 111, 651. 1963, 1, 114. 44 F. Vanhaecke, H. Vanhoe, C. Vandecasteele and R. Dams, Anal. 7 A. H. Soloway, H. Hatanaka and M. A. Davis, J. Med. Chem., Chim. Acta, 1991, 244, 115. 1967, 10, 714. 45 D. C. Gregoire, Appl. Spectrosc., 1987, 41, 897. 8 R. L. Jr. Sneath, J. E. Wright, A. H. Soloway, S. M. O’Keefe, 46 D. C. Gregoire, Spectrochim. Acta, Part B, 1987, 42, 895. A. S. Dey and W. D. Smolnycki, J. Med. Chem., 1976, 19, 1200. 47 A. L. Gray and A. R. Date, Analyst, 1983, 108, 1033. 9 M. Tsuji, M. Ichihashi and Y. Ishima, Jpn. J. Dermatol., 1983, 48 D. J. Douglas and R. S. Houk, Prog. Anal. At. Spectrosc., 1985, 93, 773. 8, 1. 10 Y. Mishima, C. Honda, M. Ichihashi, H. Obara, J. Hiratsuka, H. Fukada, H. Karashima, T. Kobayashi, K. Kanda and K. Yoshino, Lancet, 1989, 2, 388. Paper 9/03097B 1556 J. Anal. At. Spectrom., 1999, 14, 154

ISSN:0267-9477    

DOI:10.1039/a903097b

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

Speciation of cadmium in plant tissues by size-exclusion chromatography with ICP-MS Detection Ve�ronique Vacchina, Katarzyna Po�ec�† and Joanna Szpunar* CNRS EP132, He�lioparc, 2, av. Pr. Angot, 64053 Pau-Pyre�ne�es, France. E-mail: joanna.szpunar@univ-pau.fr Received 17th June 1999, Accepted 12th August 1999 A systematic approach based on the coupling of high resolution size-exclusion chromatography (SEC) with ICP-MS detection was developed for the speciation of Cd complexes with oligopeptides known to be biosynthesized by plants exposed to metal stress (phytochelatins, PCs).The separation conditions were optimized using a series of standards prepared by incubation of cadmium with purified (GluCys)2Gly (PC2), (GluCys)3Gly (PC3) and (GluCys)4Gly (PC4) ligands. The formation of artefact species was investigated and conditions allowing the reproducible chromatography and quantitative determination of Cd species in the range 0.01–15 mg ml-1 were established.The method developed was applied to investigate the speciation of Cd in cytosols of plant tissues and plant cell cultures. In addition to Cd complexes with PC2, PC3 and PC4 that eluted at the times corresponding to standards, two other Cd species, one excluded from the column and the other with an apparent molecular mass of 6800 Da, were observed in all the samples investigated. The latter compound, isolated and analysed by electrospray MS-MS, turned out to be a mixed Cd complex with diVerent PC ligands. The species eluted in the void volume could not be identified.The method was validated by comparing the results obtained by SEC-ICP-MS with those obtained by reversed-phase HPLC with post-column derivatization of phytochelatin. ICP-AES or ICP-MS12 were used to determine the metal Introduction concentration oV-line in the eluted fractions. A faster, more Aquatic and terrestrial plants are the primary entry points of reliable and more elegant analysis turned out to be possible essential micronutrients and toxic metals (such as cadmium) by the coupling of SEC with ICP-MS and this has become the into the food chain leading to animals and humans.The most widely used technique to detect stable metallodetermination of the total concentration of heavy metals is a compounds in plants.3,13 The applications included a multiroutine method to monitor the exposure of a plant to environ- elemental screening for elemental species in a particular matrix, mental metal pollution but there is increasing evidence that a such as tea leaves,14 fruit and vegetable extracts15 or plant more suitable approach to investigate the ecotoxicity of heavy extracts,16 or the investigation of a particular class of commetals, their pathways in the ecosystem and their metabolism pounds, e.g., metalloproteases of bacterial origin,17 polyby living organisms is the identification, characterization and saccharide complexes of lead in wine,18 boron complexes in determination of the particular metal species involved.1–6 plant extracts19,20 or platinum metabolites in grass.6 Plants have developed a number of internal mechanisms to Biochemical studies of metal speciation by SEC-ICP-MS control the homeostasis of essential elements and to cope with have suVered from a number of limitations and pitfalls.21–24 the stress induced by toxic elements.The most frequent case The most pertinent include (i) the non-availability of standards is the biosynthesis of a ligand, such as a phenolic compound, which means that the only information to be gained is an organic acid or oligo- or polypeptide, that would be able to approximate molecular mass of the analyte species estimated complex the excess of the toxic element into a compound on the basis of the elution volume of the complex, (ii) unconinnocuous to the organism.2,4,5 In particular, a class of oligope- trolled recovery of the metallo-complex because of its sorption tides having the general formula (EC)nG (n=2–11), called or dissociation following the interaction with the column phytochelatins (PCs), was found to play a crucial role in the packing, (iii) formation of artefacts on the column because of detoxification and homeostasis of heavy metals in plants.4,5 the exchange of the analyte metal ion or of the ligand with The complexation of metals occurs through the cysteine sulfur metal ions or ligands adsorbed earlier on the column and atom, leading to a number of relatively poorly characterized (iv) dependence of the complex composition on the chemical metal complexes.The understanding of mechanisms con- conditions of the mobile phase. The prerequisite for developing trolling the detoxification is possible only with the availability a valid approach to speciation analysis of metallo-complexes of analytical data on the species formed. by SEC-ICP-MS is therefore the synthesis of standards (that Various analytical approaches have been proposed to study are usually not commercially available) and a careful evalumetal speciation in plants.The classical approaches were based ation of the chromatographic separation conditions. This on the use of an ultra- and diafiltration technique7,8 and size- approach becomes indispensable when the complex is fairly exclusion chromatography (SEC)9,10 for the separation of the labile, which is often the case of metal complexes with oligoindividual metal species, whereas atomic absorption spec- and polypeptides.25 trometry,7–9 total reflection XRF,10 c-ray spectrometry11 and Whereas a lot of research has been devoted to the characterization of ligands bioinduced as a response of a plant to the metal stress,4,5 few data are available on the actual †On leave from the Department of Chemistry, Warsaw University of Technology, ul.Noakowskiego 3, 00-664 Warsaw, Poland. complexes formed by these ligands with heavy metals.J. Anal. At. Spectrom., 1999, 14, 1557–1566 1557Recently, SEC-ICP-MS allowed the detection of a 13 kDa where.26 The fractions with the individual PC (PC2, PC3 and PC4) peaks were heart-cut. The acetonitrile was rotavaporated compound containing Cd and Cu in which PC2 and desglycine-PC were detected by ESI-MS.3,16 Since the stan- and the fractions were lyophilized. The heart-cut fractions were repurified in the same way to produce compounds used dards were not available, the exact nature of this compound was not elucidated.as standards below. The objectives of this research were to investigate the Preparation of the PC–Cd complexes. A 1 ml volume of the behaviour of oligopeptide–Cd complexes in SEC at the trace solution (100 mg ml-1) of an individual PC in 30 mM and ultratrace levels, to evaluate the potential of SEC-ICP-MS TRIS–HCl buVer (pH 7.5) was mixed with an equimolar coupling for speciation of cadmium in plant extracts and to amount of a neutral solution of Cd2+.develop a method for the quantification of the individual Cd species found. Analytical procedures Experimental Sample preparation. Cells were vacuum filtered and washed. Plants roots were cut from the rest of the plant and washed Instrumentation (in both cases only with water). From here on, both cells and plants were treated in the same way. They were frozen in Chromatographic separations were performed using a Model liquid nitrogen to break the cell walls, ground with a mortar 1100 HPLC pump (Hewlett-Packard, Wilmington, DE, USA) and pestle and extracted with water or 10 mM TRIS–HCl as the sample delivery system.Injections were made using a buVer (pH 8). They were centrifuged (30 min, 10 000 g, 4°C), Model 7725 injection valve with a 100 ml injection loop filtered and lyophilized. (Rheodyne, Cotati, CA, USA). All the connections were made A sample of 10 mg of the freeze-dried material was dissolved of PEEK tubing (0.17 mm id).An ELAN 6000 ICP mass in 1 ml of 30 mM TRIS–HCl buVer (pH 7.5). The solution spectrometer (Perkin-Elmer SCIEX, Thornhill, ON, Canada) was ultracentrifuged for 30 min at 50 000 rpm at 4 °C. Prior was used as an element-specific detector in HPLC. The column to SEC-ICP-MS analysis, the supernatant was diluted 20-fold eluate was introduced into the ICP via a cross-flowor Silene cucubalus analysis, 10-fold for Agrostis tenuis analy- fitted in a Ryton spray chamber.For total analyses samples sis and not diluted for maize and Rauwolfia serpentina analysis. were fed by means of a Minipuls 3 peristaltic pump (Gilson, No dilution was used when SEC was used prior to ESI- Villiers-le-Bel, France) that also served for draining the spray MS/MS analysis. chamber. Chromatographic data were processed using Turbochrom4 software (Perkin-Elmer, Norwalk, CT, USA). SEC and ICP-MS conditions. Three types of SEC columns All signal quantifications were performed in the peak area (300×10 mm id) (Pharmacia Biotech, Uppsala, Sweden) were mode.ESI-MS/MS measurements were performed using a investigated, namely Superdex Peptide HR 10/30 designed for Perkin-Elmer SCIEX API 300 electrospray triple-quadrupole the separation of peptides (optimum separation range 100– mass spectrometer. A Hitachi (Tokyo, Japan) Model 7000 kDa), Superdex 75 HR 10/30 with an exclusion limit of CS120GX refrigerated ultracentrifuge was used for the separa- 100 kDa and an eVective separation range between 3 and tion of the supernatant after leaching of Cd species from plant 70 kDa and Superdex 200 with an exclusion limit of 1300 kDa tissues and cell cultures.and an optimum separation range of 10–600 kDa. A TSK PWXL guard column (40×3 mm id) (Tosoh, Tokyo, Japan) Reagents, solutions and materials was always used. The columns were calibrated (UV detection Analytical-reagent grade reagents purchased from Sigma- was used) with the following standards: glutathione (Mr 307), Aldrich (St.Quentin Fallavier, France) were used throughout PC2 (Mr 539), PC3 (Mr 771), rabbit liver metallothionein–Cd unless specified otherwise. Water (18 MV) prepared with a complex (Mr 6918), cytochrom c (Mr 12 384), bovine albumin Milli-Q system (Millipore, Bedford, MA, USA) was used (Mr 66 000) and thyroglobulin (Mr 660 000). throughout. The columns were conditioned between the injections with Cell cultures of the plants Silene cucubalus, Agrostis tenuis a 2mM b-mercaptoethanol solution in 30 mM TRIS buVer and Rauwolfia serpentina were investigated.They were grown (pH 7.5) for 30 min followed by removal of the chelating for 4 d in a 300 mM Cd2+ solution. Seeds of maize (Zea mays reagent by conditioning the column with the mobile phase for L.) were sown in sand followed by the addition of Hoagland another 30 min. Care should be taken to avoid traces of metals nutrient solution.When the plant was large enough (10 d after in the conditioning mobile phase solution. The column congermination), 300 mM Cd2 + solution was added. The maize dition should be controlled prior to injection of a sample by was harvested 1 week after Cd addition. injecting a 2 mM b-mercaptoethanol solution. Typical peak height values observed for Cd and Cu were lower than 200 Standards and 500 counts s-1, respectively. Preparation of metal-free phytochelatins.The synthesis of PC–metal complexes were eluted with 30 mM TRIS–HCl phytochelatins was carried out according to Grill et al.26 In buVer (pH 8.5) at a flow rate of 0.75 ml min-1. A 100 ml brief, 2000–10 000 pkat of PC synthase27 were incubated at aliquot of the sample solution was injected. The mobile phase 25 °C with 1 mM GSH and 0.8 mM Cd(NO3)2 in 1120 ml of was degassed ultrasonically. ICP-MS measurement conditions the buVer solution (pH 8.0). NaN3 (0.02%) was added to (nebulizer gas flow, rf power and lens voltage) were optimized retard bacterial growth.daily using a standard built-in software procedure. The iso- Proteins were precipitated from the resulting solution by topes monitored were 114Cd and 65Cu. The SEC-ICP-MS the addition of (NH4)2SO4 to 85% followed by centrifugation optimized instrumental conditions were optimized daily using at 8000 g for 30 min. PCs were precipitated from the super- the standard built-in optimization algorithm.Calibration was natant by the addition of 20 ml of 1 M Cd(NO3)2 solution. performed using standard graphs in the range 0.01–15 mg ml-1 The precipitated mixture of PCs was centrifuged and washed (as PC) established with the Cd–PC2–4 standards prepared as twice with water. The washed precipitate was stable for several described above. months when stored at -20 °C. The precipitated PCs were dissolved in 3.5M HCl and ESI-MS-MS analysis. ESI-MS/MS was applied to identify Cd species for which the elution volume in an SEC-ICP-MS separated by semi-preparative HPLC by elution with a concentration gradient of acetonitrile in water as described else- chromatogram could not be matched by one of the standards 1558 J.Anal. At. Spectrom., 1999, 14, 1557–1566prepared as described above. The fraction containing such a peak was collected and lyophilized. It was dissolved in 500 ml of water containing 0.1% of trifluoroacetic acid (TFA) (pH 2.3), filtered and desalted by reversed-phase chromatography by elution with 0.1% TFA in water for 5 min followed by gradient elution with 0.1% TFA in acetonitrile–water (50+50 v/v) for a further 40 min at 0.75 ml min-1.The eluate of the first 5 min was discarded. To the rest of the eluate 5 mM (dithiothreitol ) (DTT) was added, acetonitrile was removed by rotavaporation and the remaining solution was freeze-dried. The dried material obtained was dissolved in 200 ml of 0.06M acetic acid in 30% methanol and analysed by ESI-MS.ESI-mass spectra were acquired (10 scans) in the range Fig. 1 EVect of the excess of cadmium on the synthesis of PC–Cd 50–1100 u using a 10 ms dwell time and a 0.5 u step size. The complex. The determination was performed by SEC-ICP-MS. The orifice potential was 40 V, the ionspray voltage was 4100 V intensity corresponds to the area of the peak corresponding to a and the ion multiplier potential was 2400 V. The signals in the particular phytochelatin. 1, (GluCys)2Gly (PC2); 2, (GluCys)3Gly mass spectrum were identified by MS/MS using collision- (PC3); 3, (GluCys)4Gly (PC4). induced dissociation and the product ion scan mode. The collision gas was nitrogen and the collision energy was varied from 25 to 50 eV to obtain optimum fragmentation. Purity of the PC–Cd standards. The purity of the PC–Cd complexes obtained was monitored by SEC-ICP-MS. The chromatograms of the Cd complexes with the individual Results and discussion phytochelatins are shown in Fig. 2(a)–(c). Fig. 2(d) shows a chromatogram corresponding to the complexes formed by Preparation of standard solutions of Cd complexes with incubating cadmium with a mixture of apo-PCs prior to phytochelatins separation of the individual PCs by reversed-phase chromatography. Phytochelatins (oligopeptides with up to 11 amino acids) can be readily obtained in large amounts in the reaction of Chromatograms of the PC2 and PC3 standard solutions show that in each of the cases one major complex is formed polymerization of glutathione in the presence of an enzyme, phytochelatin synthase.26,27 The reaction produces a mixture that elutes as a well defined peak.Values of the molecular weight can be assigned to these peaks by calibrating the of oligopeptides that can be isolated by precipitation with Cd from the supernatant remaining after the precipitation of column with a series of oligopeptide standards. The molecular masses calculated in this way for the PC2–Cd and PC3–Cd larger proteins with ammonium sulfate.The precipitate (soluble in fairly concentrated HCl) can be desalted by reversed- complexes with cadmium (pH 8.5) are 2300±500 and 3400±500 Da, respectively. These results agree with an earlier phase chromatography under acidic conditions under which the mixture of PCs is recovered in the metal-free (apo) form.26 observation that phytochelatin peptides form complexes with Cd2+ of Mr 2500 and 3600.29 Individual PCs (PC2, PC3 and PC4) (apoforms) can be isolated by heart-cutting of the peaks in the reversed-phase In the case of the PC4–Cd complex, the chromatogram is more complex. In addition to the PC4–Cd signal with an chromatogram and characterized by ESI-MS/MS as described elsewhere.28 The species obtained are metal-free and cannot estimated molecular mass of 5100±500 (at pH 8.5) two other intense peaks are present.One of them corresponds to the be used as calibration standards for SEC-ICP-MS.The preparation of metallated standards requires the reconstitution of elution volume of the PC3–Cd complex. This may be explained by a PC3 impurity in the standard that is also indicated by an PC–Cd complexes under optimized reaction conditions. The metal-free PCs can apparently be reassociated with ESI mass spectrum (not shown) of the apo-PC4 standard solution. The intensity of this peak in the SEC-ICP-MS heavy metal ions simply by mixing a metal ion with an apophytochelatin.Particular care was taken to avoid oxidation of chromatogram is nevertheless higher than one would expect from the ESI mass spectrum. The other peak corresponds to the sulfhydryl groups during reconstitution.26 Having this in mind, we investigated two procedures for the preparation of a compound with an apparent molecular mass of 6800 u. The identity of this compound is unknown but it is likely to PC–Cd complexes. In the first, NaBH4 was added to the reaction mixture during reconstitution to ensure in situ reduc- correspond to a PC–Cd species of a diVerent stoichiometry.A trace of species at the same elution volume is also observed ing conditions according to Grill et al.26 The alternative was to deoxygenate the reaction solutions by sparging with helium, in the SEC-ICP-MS chromatogram of the PC3 standard [Fig. 2(b)]. which was expected to be suYcient to ensure a non-oxidizing environment. The chromatogram of a solution prepared by the incubation of a crude (not subjected to reversed-phase HPLC) mixture It was found that the SEC-ICP-MS chromatograms of the PC–Cd complexes obtained by the two procedures were ident- of apo-PCs with Cd2+ shown in Fig. 2(d) contains three baseline separated peaks with elution volumes corresponding ical. Therefore, the more straightforward procedure based on mixing an apo-PC solution with that of Cd2+ ions in TRIS to the PC2, PC3 and PC4 standards. Neither the small impurities observed in the PC3 chromatogram nor the intense signal buVer was used throughout.The excess of Cd to be added was determined by adding increasing amounts of Cd to the at Mr 6800 can be seen. The reason may be some degradation (e.g., oxidation of PCs during the reversed-phase chromato- solution of an apo-PC standard. The increase in the peak height as a function of increasing Cd concentration is shown graphic purification and freeze-drying procedures of the individual PCs).The mixture of PCs used to obtain the in Fig. 1. A PC-to-Cd ratio of 151 was applied for the reconstruction of PC–Cd complexes. Despite it being higher chromatogram in Fig. 2(d) was not subjected to any chromatographic or lyophilization steps (see Analytical procedures). than the stoichiometric ratio, it was chosen in view of the possible interactions of the complex with the column stationary phase. Note that the excess of Cd in the solution is anyway Stability of PC–Cd standard solutions with time.Fig. 3 shows the overlapped chromatograms of a standard solution freshly retained on the column. J. Anal. At. Spectrom., 1999, 14, 1557–1566 1559Fig. 3 Stability of standard solution of PC–Cd complexes with time. Chromatograms were obtained by SEC-ICP-MS. Solid line, freshly prepared solution of PC–Cd complexes; dashed line, solution 12 d old. Peaks: 1=unidentified; 2=PC4; 3=PC3; 4=PC2. The likely reasons for such a behaviour are the oxidation of PC and the fact that the complex formed with the oxidized form, if any, is much weaker than that with the reduced form.The formation of a diVerent complex is evident judging from the shift of the elution volume of PC2. There is also evidence for the formation of an aggregate with an apparent molecular mass of 6800 Da as seen in Fig. 2(c). This suggests that PCs may aggregate under some conditions and there is a risk of misidentification of chromatographic signals if judged from the elution volume only.Optimization of the separation conditions Because of their strongly hydrophilic anionic character, PC–metal complexes cannot be retained on a reversed-phase column even in organic media (methanol, acetonitrile). The use of strong anion exchangers leads to the need for concentrated buVers which would adversely aVect ICP-MS or ESIMS/ MS detection. The size-exclusion mode was therefore selected as the chromatographic separation mechanism.An earlier study showed the possibility of eluting Cd and Cu species in plant cytosol extracts from a size-exclusion column (Eurogel GFC, 300×7.5 mm id) as a fairly broad single peak corresponding to a compound with an apparent molecular mass of ca. 13 kDa,16 but no optimization of the separation conditions was carried out because of the non-availability of standards.16 In this work, the synthesis and purification of individual Cd–PC species allowed the systematic optimization of the SEC-ICP-MS conditions and investigation of the potential of this technique for the determination of Cd complexes with oligopeptides in plant cell extracts.Fig. 2 SEC-ICP-MS chromatograms obtained for the PC–Cd stan- Choice of the column. The SEC columns examined in this dards synthesized from individual purified PC preparations. work contained the same type (from the chemical point of (a) (GluCys)2Gly (PC2); (b) (GluCys)3Gly (PC3); (c) (GluCys)4Gly view) of stationary phase (cross-linked agarose–dextran) but (PC4); (d) mixture of PC2, PC3 and PC4 prior to separation of the diVered in terms of the exclusion limit and the optimum individual PCs equilibrated with Cd.The dashed lines correspond to the molecular masses found by the calibration of the columns as separation range. Superdex-75 oVers an eVective separation described under Analytical procedure. Peaks: 1=unidentified; 2= range (for globular proteins) of 0.5–80 kDa which has been PC4; 3=PC3; 4=PC2.judged the most suitable for the separation of metal complexes with metallothioneins so far.24 Since PC peptides are much smaller than metallothioneins, another column (Superdex prepared from a mixture of apo-PCs and one stored for 12 d at 4 °C. It is evident that solutions of standards and sample Peptide HR), designed for the separation of amino acids and oligopeptides, was investigated in this work, despite the lack extracts should be prepared fresh since storage adversely aVects the intensity of Cd–PC signals, especially for PC2 and PC3.of reports of its application to the separation of metal com- 1560 J. Anal. At. Spectrom., 1999, 14, 1557–1566plexes. A guard column filled with a packing with an exclusion limit of 80 kDa was used. The separation eYciency was investigated for a mixture of the PC2–4–Cd complexes prepared as described in the under Analytical procedures. No diVerence in terms of the separation eYciency between the Superdex 75 and Superdex Peptide columns was observed. The duration of the run was slightly shorter (by ca. 2 min) using a Peptide HR column, so this was chosen for further work. EVect of pH. Fig. 4 shows that the pH of the mobile phase aVects not only the separation eYciency but also the morphology of the chromatogram. The eVect was tested in the pH range 6.0–10.0 adjusted with 30 mM TRIS–HCl or TRIS– ammonia buVer. Lower pH values were not investigated because the PC–Cd complexes are known to be unstable in acidic media.4 At pH 6.0 the signal from PC2 is already suppressed, probably because of the dissociation of the complex and sorption of the Cd2+ on the column stationary phase.Chromatograms at pH 7.5 and 8.5 are almost identical. The separation between PC2 and PC3 is slightly better at pH 7.5, but on the other hand the run at pH 8.5 is faster by 2 min. Both pH values were examined further for the chromatography of plant cell extracts. At pH 10.0 all the cadmium elutes as a single compound in the void volume of the column.The total cadmium eluted (calculated from the area under the peak) is similar to that at pH 7.5 and 8.5. It is therefore concluded that another compound is formed with PC2, PC3 and PC4 as ligands. This corroborates the observation of Grill et al.26 who predicted, under some conditions, the formation of Cd–thiol complexes consisting of a mixture of the PC peptides of various lengths with small amounts of cysteine and glutathione.EVect of buVer concentration. Fig. 5 shows that the concentration of the buVer in the mobile phase does not play an important role provided that it is not lower than 10 mM. In the absence of salt the elution eYciency is fairly low, and all the cadmium elutes as a single peak in the void volume of the column. This is not the eVect of the pH since at the same pH in the presence of buVer a ‘normal’ elution pattern was observed [cf., Fig. 4(a)].The chromatogram in Fig. 5(a) suggests that in the absence of salt aggregated species are formed. They are apparently less stable than the complexes and the exchange of the cadmium with the column stationary phase occurs more readily. At elevated buVer concentrations (70 mM) the baseline is higher, probably because of the column bleeding with Cd2+. Figures of merit The precision of the method was evaluated by fivefold injection of a mixture of Cd–PC2, –PC3 and –PC4 complexes at the 1 mg ml-1 level (as PC) and the determination of the peak area. The relative standard deviations were 4.7, 4.6 and 6.9% for Cd–PC2, –PC3 and –PC4, respectively.Fig. 4 EVect of pH on the separation of PC–Cd complexes by SEC. The linearity of the PC peak area as a function of concen- ICP-MS detection of cadmium. Amount injected, 1 mg ml-1 (as PC). tration was investigated in the range 0.01–15 mg ml-1 of PC pH: (a) 6.0; (b) 7.5; (c) 8.5; (d) 10. Peaks: 1=unidentified (void volume of the column); 2=PC4; 3=PC3; 4=PC2.in the chromatographed solution. Calibration curves are shown in Fig. 6. Two linearity ranges can be distinguished; a well marked break point is observed around concentrations of 2–3 mg ml-1 for each of the analysed PCs. The detection limits cadmium from the injected PC–Cd complex by other metals present on the column cation-exchange sites was investigated. (calculated as three times the baseline noise) are at the 10 ng ml-1 level for each of the phytochelatins.The approach undertaken consisted of injecting a metal-free PC solution and monitoring the chromatographic eluate for The reason for such an anomalous behaviour of the calibration curve seems to be due either to the formation of the presence of metal complexes by ICP-MS. It was found that Cd and Cu were present in the eluate, diVerent Cd complexes as a function of the PC concentration, or to a reaction taking place on the column. Since the giving rise to signals matching the elution volumes corresponding to those of the PC–Cd complexes.The intensities of these morphology of the chromatogram was found to be independent of the PC concentration injected, the second reason seems signals were random and depended on the history of the column. When a PC–Cd standard was injected, intense signals to be more likely. The possibility of the displacement of J. Anal. At. Spectrom., 1999, 14, 1557–1566 1561Fig. 6 Calibration curves for the PC–Cd complexes.Calibration curves for the lower concentration range are shown in the inset. (a) (GluCys)2Gly (PC2); (b) (GluCys)3Gly (PC3); (c) (GluCys)4Gly Fig. 5 EVect of the TRIS–HCl buVer concentration on SEC of PC–Cd (PC4). complexes. ICP-MS detection. Amount injected, 0.1 mg ml-1. BuVer concentration: (a) none (water at pH 6.0); (b) 10 mM; (c) 30 mM; (d) 70 mM. Peaks: 1=unidentified (void volume of the column); 2= intensity) and (iii) Cu2+ adsorbed on the column exchanges PC4; 3=PC3; 4=PC2.with Cd from a PC–Cd complex leading to the destruction of the latter. It was therefore decided to develop a protocol for column conditioning in order to be able to obtain reproducible were observed on the Cu channel, sometimes in addition to chromatograms. the Cd signals but often instead of Cd signals. When a Cu2+ solution was injected on to the column the following chromato- Column conditioning gram of a PC–Cd standard showed a distorted morphology.It was concluded that (i) metal ions (Cd, Cu and to some The most straightforward solution to the problem of artefact formation consists of preventing the free Cu2+ and Cd2+ ions extent Zn) accumulate on the column until an equilibrium is reached, (ii) these metals may be complexed by free ligands from accumulating at the head of the column. A similar problem was solved in the case of metal complexes with present in the injected solution giving rise to ghost signals (or to signals from analyte compounds but of uncontrollable metallothioneins by adding b-mercaptoethanol to the injected 1562 J. Anal.At. Spectrom., 1999, 14, 1557–1566solution.30 This chelating compound complexed Cu2+ and TRIS–HCl buVer (pH 7.5) and 30 mMacetate buVer (pH 7.0). No diVerence in terms of the morphology of the SEC-ICP-MS Cd2+ ions without aVecting the complexation equilibrium of the analyte species (MT–Cd), which allowed the elution of chromatograms of these extracts was observed.The buVer was used for extraction of metal–peptide complexes in order to the free metal ions as a peak in the total volume of the column, after the peak of the analyte compound. In the case of PC ensure reproducible extraction conditions in terms of pH. Chromatograms obtained under the optimized analytical complexes it turned out, however, despite some reports claiming the stability of PC–metal complexes in the presence of conditions (Peptide HR 10/30 column) for the above samples are shown in Fig. 8. The intensity of the signals was found to b-mercaptoethanol,16 that b-mercaptoethanol competes with the PC ligands, leading to the destruction of the PC–Cd decrease even after 1 d of storage of the extract but no changes in the morphology of the chromatogram were observed. complexes. An alternative is the cleaning of the column by repetitive Chromatography at pH 7.5 was verified and was found to give very similar chromatograms (not shown). injections of b-mercaptoethanol to remove the metals accumulated in the previous run.The advantages are the ease of The extract of Silene cucubalus contains Cd complexes of PC2, PC3 and PC4 as confirmed by the matching elution control of the eluted metal (peak intensity monitored in real time by ICP-MS) and the lack of need to change the mobile volumes of the appropriate standards. PC2, PC3 and a trace of PC4 were found in Rauwolfia serpentina extract. PC2 and phase between sample injections.However, whereas two injections of the reagent solution are suYcient to remove Cd from PC3 signals were also identified in Agrostis tenuis extract. The intensity of Cd complexes in the maize samples is much lower. the column, a large amount of Cu2+ stays on the column even after the fourth injection of b-mercaptoethanol (Fig. 7). The Nevertheless, a distinct PC2 signal can be identified. Note that Silene and Agrostis extracts were diluted for the analysis.only possibility therefore seems to be the introduction of a regular washing step between the injections. The most promising results in terms of reproducibility (ca. 5% as indicated above) were obtained by conditioning the column between the injections with a 2 mMb-mercaptoethanol solution in 30 mM TRIS buVer (pH 7.5) for 30 min. The removal of the chelating agent prior to next injection was accomplished by conditioning the column with the mobile phase for another 30 min.Care should be taken to avoid traces of metals in the conditioning solutions, otherwise trace metals will accumulate at the head of the SEC column during the conditioning phase. A sorption column removing trace metals (especially Cu) at the exit from the pump is recommended. 31 The state of the column should be controlled prior to introduction of a sample by injection of a b-mercaptoethanol solution. Typical peak height values were lower that 200 counts s-1 for Cd and 500 counts s-1 for Cu.The cleanliness of the column aVects the slope of the calibration curve in the lower concentration range and the position of the break point on the latter. Analysis of plant extracts Detection of metal–peptide complexes in plant extracts. The method developed was applied to speciation of cadmium in extracts of a plant, maize (Zea mays L.), and of three diVerent plant cell cultures of Silene cucubalus, Agrostis tenuis and Rauwolfia serpentina, known to biosynthesize PCs when exposed to Cd2+.The procedures investigated to extract the Cd complexes included extraction with water, 30 mM Fig. 8 SEC-ICP-MS chromatograms of plant extracts prepared according to the Analytical procedure. (a) Silene cucubalus; (b) Agrostis tenuis; (c) Rauwolfia serpentina; (d) maize (Zea mays L.). Fig. 7 EYciency of the post-run removal of Cd (&) and Cu (#) from Peaks: 1=unidentified (void volume of the column); 2=unidentified (Mr 6800); 3=PC4; 3=PC3; 4=PC2. The dashed vertical lines corre- the column by repetitive injections of b-mercaptoethanol. Mobile phase, water (open symbols) and 30 mM TRIS buVer (pH 8.5) spond to the molecular masses found by the calibration of the columns as described under Analytical procedure.(closed symbols). J. Anal. At. Spectrom., 1999, 14, 1557–1566 1563In addition to the above discussed signals for which identifi- the careful column conditioning but the regular washing of the column makes it occur in a reproducible way.It should cation was possible on the basis of their elution volumes matching those of the previously synthesized and characterized also be noted that the amounts of Cd2+ and that of free PC ligands in the injected solutions are actually unknown. It is PC2–4–Cd standards, at least one (sometimes two) intense Cd signals are observed in the SEC-ICP-MS chromatograms of therefore impossible to evaluate whether this apparently low recovery is due to the decomposition of the complex on the the extracts investigated. All the extracts contain a compound with an apparent molecular mass of 6800 u similar to that column or simply because only 50% of Cd in the injected solution is actually complexed.The second possibility is cor- already observed in the isolated PC4 standard [Fig. 2(c)]. Moreover, the extracts show the presence of a Cd species roborated by the fact that chromatography is reproducible, the baseline is low and stable and the chromatographic peaks excluded from the column.The chromatograms in Fig. 8 exhibit a diVerent morphology do not tail. in comparison with the only chromatogram reported so far, to our knowledge, in the literature and obtained by a similar Quantification of PC–Cd complexes in plant extracts. The technique. Leopold et al.,3,16 analysing an extract from Silene above data indicate that Cd occurs in the extracts of plants in vulgaris by SEC-ICP-MS, showed that cadmium eluted as a a number of fairly labile complexes.This makes the approfairly broad single peak at an elution volume corresponding priate calibration a critical issue. Attempts to use the method to an apparent molecular mass of ca. 13 kDa. Since no of standard additions instead of the calibration curve failed calibration standards were available, this compound was ident- since the addition of a standard modified the equilibrium in ified as a Cd, Cu–PC2 complex on the basis of the detection the system, leading to changes in the mutual ratio of the of PC2 and desglycyl-PC by ESI-MS.3 In order to obtain a diVerent Cd–PC complexes present.External calibration using deeper insight into the identity of the unknown signals in the a calibration graph established as in Fig. 6 was therefore used. chromatograms in Fig. 8, two approaches were used in this The results obtained for the diVerent samples analysed are work. The extracts were chromatographed on size-exclusion summarized in Table 2.Note that, in contrast to the injection columns with a higher exclusion limit to determine more of the standard solution, the average precision of the determiprecisely the apparent molecular mass of the early eluted nation of PC–Cd complexes in real samples is ca. 20%, which compounds. Further, the fractions containing the unknown is probably a consequence of a number of dynamic equilibria peaks were collected and characterized by ESI-MS-MS. occurring in the system.Signal identification in SEC-ICP-MS chromatograms of plant Validation of the method developed. Since reference materials extracts. Chromatograms (not shown) obtained using the with known PC concentrations in plant samples are not Superdex 75 column (exclusion limit 100 kDa) and Superdex available, the only way to validate the method developed was 200 (exclusion limit 1300 kDa) for the plant extracts analysed to compare the results obtained with those obtained by an in Fig. 8 always show a signal excluded from the column that independent analytical method. No other method is available, corresponds, judging from the morphology of the chromatog- however, for the determination of Cd–PC species. Therefore, rams, to the first eluting peak in Fig. 8. The fraction containing the only possibility to validate the method developed was to this peak collected and analysed by ESI-MS did not show any compare the concentrations of the individual PC ligands found trace of PCs.A possible hypothesis of the identity of this peak by SEC-ICP-MS with those obtained by reversed-phase HPLC is that it contains a complex of cadmium with polymers of with post-column derivatization with 5,5¾-dithiobiscell membranes which were not completely removed by ultra- (2-nitrobenzoic acid) and spectrophotometric detection. The centrifugation. latter measurements were carried out in a diVerent laboratory The second of the unknown signals in Fig. 8 behaves on the by a diVerent operator. The agreement is satisfactory but, in Superdex 75 and Superdex 200 columns like a compound with general, the results obtained by SEC-ICP-MS are 20–30% an apparent molecular mass of ca. 10 kDa. The fraction lower. This may be due to the fact that the PCs present in the containing this peak collected, desalted on a reversed-phase peak 2 in Fig. 8 are not taken into account in the SECcolumn as described in the under Analytical procedure and ICP-MS procedure.Note that the ratio of Cd to PC for peak analysed by ESI-MS produced the spectrum shown in Fig. 9. 2 seems to be much higher than that for the PC2, PC3 and It shows at least three peaks, at m/z 540, 772 and 1004, which PC4 complexes. may correspond to the protonated molecular ions of PC ligands. Indeed, the product ion spectra of the three peaks (not shown) confirmed the presence of amino acid sequences Conclusions of PC2, PC3 and PC4. The data shown in Fig. 9 were obtained SEC with ICP-MS detection is a valuable tool for screening with Silene cucubalus. Similar spectra were obtained for the for metal complexes in biological tissue cytosols, but its proper other samples investigated, except maize, for which the ESI use requires a systematic approach to each problem to be mass spectrum was hardly diVerent from that of the blank, solved in terms of standardization and verification of the probably because of the 100-times lower concentration of the stability of the species studied.In contrast to organometallic species of interest. compounds, metal coordination complexes are likely to undergo metal (ligand) exchange either with the metal or Recovery. Recovery of Cd from the column was evaluated ligand contaminants retained at the head of the column or by comparing the amount of the metal injected with that with the column active sites themselves. Nevertheless, when a eluted. The results are given in Table 1.The average recovery number of precautions are taken, the SEC-ICP-MS method of Cd oscillates around 50% with a trend of being lower for allows not only the qualitative analysis of heavy metal species more dilute solutions. This corresponds well with the lower in plants but also their quantitative determination. slope of the calibration curve in the low concentration range (cf., Fig. 6). One of the reasons for such a low recovery may be the exchange of the Cd from the PC–Cd complex with Acknowledgements Cu2+ present on the column.Indeed, the results shown in Table 1 demonstrate that a significant amount of copper is The authors thank Professor Dr. M. Zenk and Matjaz Oven (University of Munich) for valuable discussions. Dr. released from the column which can only occur by complexation with a PC ligand. This phenomenon occurs despite R. Lobinski is acknowledged for critical reading of the manu- 1564 J. Anal. At. Spectrom., 1999, 14, 1557–1566Fig. 9 ESI tandem mass spectra of peak 2 in Fig. 8. Data are shown for Silene cucubalis. Peaks: 1=PC2; 2=PC3; 3=PC4. Table 1 Recovery of cadmium and copper in SEC for standards and sample extracts Metal injecteda/ Metal recovered/ Recovery (%) ng ng Sample analysed Cd Cu Cd Cu Cd Cu Apophytochelatin (a-PC), 1 mg ml-1 <0.1 <0.1 <0.1 0.5 — — a-PC+0.1 mg Cd 3 0.1 0.6 0.5 20 500 a-PC+0.25 mg Cd 7 0.2 2 0.9 29 450 a-PC+0.5 mg Cd 14 0.2 5 0.2 36 100 a-PC+1 mg Cd 25 0.1 10 0.4 40 400 Silene vulgaris 11 0.1 5.5 1.1 51 1100 Silene cucubalis 26 0.1 10 1.1 39 1100 Agrostis tenuis 30 0.1 18 1.0 60 1000 Rauwolfia serpentina 71 0.4 30 1.6 43 400 Maize 8.5 0.2 2.7 1.7 32 850 aDetermined by ICP-MS in the injected solution.script. The work was financed by grant No. 130201 Z0025 (Region Aquitaine–FEDER). Table 2 Validation of the results obtained by SEC-ICP-MS in lyophilized plant cell extracts by an independent procedure Concentrationa/mg g-1 References Sample Phytochelatin SEC-ICP-MS HPLCb 1 J.Szpunar and R. Lobinski, in Heavy Metal Stress in Plants— From Molecules to Ecosystem, ed. M. N. V. Prasad and Silene cucubalis J. Hagemeyer, Springer, Heidelberg, 1999, pp. 349–370. PC2 3.2±0.8 3.6 2 M. N. V. Prasad, Analusis, 1998, 26, M25. PC3 7.0±1.2 10.5 3 I. Leopold, D. Guenther and D. Neumann, Analusis, 1998, 26, PC4 3.4±1.2 3.6 M28. Agrostis tenuis 4 M. H. Zenk, Gene, 1996, 179, 21. PC2 1.2±0.7 0.9 5 W. E. Rauser, Annu.Rev. Biochem., 1990, 59, 61. PC3 2.1±0.6 2.7 6 D. Klueppel, N. Jakubowski, J. Messerschmidt, D. Stuewer and PC4 0.3 0.16 D. Klockow, J. Anal. At. Spectrom., 1998, 13, 255. Rauwolfia serpentina 7 K. Lange-Hesse, L. Dunemann and G. Schwedt, Fresenius’ PC2 0.17±0.02 0.3 J. Anal. Chem., 1991, 339, 240. PC3 2.3±0.3 2.7 8 K. Lange-Hesse, L. Dunemann and G. Schwedt, Fresenius’ PC4 0.8±0.3 1.3 J. Anal. Chem., 1994, 340, 460. Maize (Zea mays L.) 9 K.Gu� nther and H. Waldner, Anal. Chim. Acta, 1992, 259, 165. PC2 0.03±0.01 0.03 10 K. Gu�nther and A. von Bohlen, Spectrochim. Acta, Part B, 1991, PC3 <0.01 <0.01 46, 1413. PC4 <0.01 <0.01 11 A. V. Harms and J. T. van Elteren, J. Radioanal. Nucl. Chem., 1998, 228, 139. aIn the lyophilized plant extract. bObtained by HPLC with post- 12 H. Fingerova� and R. Koplik, Fresenius’ J. Anal. Chem., 1999, column derivatization with 5,5¾-dithiobis(2-nitrobenzoic acid) and 363, 545. spectrophotometric detection. 13 A. Makarov and J. Szpunar, Analusis, 1998, 26, M44. J. Anal. At. Spectrom., 1999, 14, 1557–1566 156514 K. E. Oedegard and W. Lund, J. Anal. At. Spectrom., 1997, 12, 25 K. De Cremer, J. De Kimpe and R. Cornelis, Fresenius’ J. Anal. 403. Chem., 1999, 363, 519. 15 J. Szpunar, P. Pellerin, A. Makarov, T. Doco, P. Williams and 26 E. Grill, E. L. Winnacker and M. H. Zenk, Methods Enzymol., R. £obin� ski, J. Anal. At. Spectrom., 1999, 14, 639. 1991, 205, 333. 16 I. Leopold and D. Gu� nther, Fresenius’ J. Anal. Chem., 1997, 359, 27 E. Grill, S. Lo�Zer, E. L. Winnacker and M. H. Zenk, Proc. Natl. 364. Acad. Sci. USA, 1989, 86, 6838. 17 I. Leopold and B. Fricke, Anal. Biochem., 1997, 252, 277. 28 V. Vacchina, H. Chassaigne, M. Oven, M. Zenk and R. Lobinski, 18 J. Szpunar, P. Pellerin, A. Makarov, T. Doco, P. Williams, Analyst, in the press. B. Medina and R. £obin� ski, J. Anal. At. Spectrom., 1998, 13, 749. 29 E. Grill, E. L. Winnacker andM. H. Zenk, Science, 1985, 230, 674. 19 T. Matsunaga, T. Ishii and H.Watanabe, Anal. Sci., 1996, 12, 673. 30 K. A. High, B. A. Methven, J. W. McLaren, K. W. M. Siu, 20 T. Matsunaga and T. Nagata, Anal. Sci., 1995, 11, 889. J. Wang, J. F. Klaverkamp and J. S. Blais, Fresenius’ J. Anal. 21 L. M. W. Owen, H. M. Crews, R. C. Hutton and A. Walsh, Chem., 1995, 351, 393. Analyst, 1992, 117, 649. 31 G. F. Van Landeghem, P. C. D’Haese, L. V. Lamberts and M. E. 22 R. Cornelis, J. De Kimpe and X. Zhang, Spectrochim. Acta, Part De Broe, Anal. Chem., 1994, 66, 216. B, 1998, 53, 187. 23 A. Sanz-Medel, Spectrochim. Acta, Part B, 1998, 53, 197. 24 J. Szpunar and R. Lobinski, Pure Appl. Chem., 1999, 71, 899. Paper 9/04845F 1566 J. Anal. At. Spectrom., 1999, 14, 1

ISSN:0267-9477    

DOI:10.1039/a904845f

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

Quality assurance of arsenic, lead, tin and zinc in copper alloys using axial inductively coupled plasma time-of-flight mass spectrometry (ICP-TOF-MS) Ha°kan Emteborg, Xiaodan Tian and Freddy C. Adams* Micro and Trace Analysis Center (MiTAC), Department of Chemistry, University of Antwerp (UIA), Universiteitsplein 1, B-2610 Antwerp, Belgium Received 25th May 1999, Accepted 12th July 1999 Results for As, Pb, Sn and Zn are reported for five copper alloys. Following dissolution in a refluxing HCl–HNO3 mixture and subsequent dilution, the elements were determined by inductively coupled plasma time-of-flight mass spectrometry (ICP-TOF-MS) using In as an internal standard.The following isotopes: 64Zn, 66Zn, 67Zn, 68Zn, 70Zn, 75As, 206Pb, 207Pb 208Pb, 115In, 116Sn, 118Sn and 120Sn were monitored and the intensity ratios to 115In as an internal standard were used. The analytical results were compared with results obtained by flame atomic absorption spectrometry (FAAS) for assessment of accuracy.The results agreed fairly well, except for Zn in some compositions, which was due to background interference eVects, and for Sn, where dissolution problems occurred. The relative standard deviations, % RSDs, for six replicate measurements of the elements in each copper alloy were in the range 2.0–6.1% for As, 3.8–11.0% for Pb, 1.4–9.4% for Sn and 3.6–12.9% for Zn. Uncertainty budgets for Sn and Zn associated with these determinations are described.The detection limits (3s criterion) in the solid materials with a 1.0 g sample intake and appropriate dilutions were 0.7 mg g-1 for Pb, 2.5 mg g-1 for Sn, 11 mg g-1 for As and 15 mg g-1 for Zn. trometer was commercialised by LECO, St. Joseph, MI, USA, Introduction and the performance of this instrument is described here and ICP-MS combines parts per trillion detection limits, a linear compared with results obtained independently by flame atomic range of 6–7 orders of magnitude, isotopic measurement absorption spectrometry (FAAS).capability and limited spectral interferences with a high sample Principle of operation of ICP-TOF-MS throughput and almost complete elemental coverage.1–4 Since the introduction of commercially available ICPs in 1983, most The ions are formed in the plasma and extracted into the mass instruments incorporate quadrupole mass filters as these com- spectrometer through the sampler and skimmer cones.7 The bine the necessary resolution with a reasonable cost.4 With ions are accelerated to similar kinetic energy before entering the introduction of high-resolution ICP-MS the sensitivity was the field-free flight-tube and the arrival time of an ion at the increased and isobaric interferences such as the overlaps detector is proportional to the square root of its m/z since from 40Ar+, 56ArO+ and 80ArAr+ on Ca, Fe and Se could be ions of diVerent mass acquire a diVerent velocity.To increase eliminated.5 resolution, a reflectron is used to diminish small variations of A less attractive feature of quadrupole ICP-MS is the the kinetic energy acquired by ions of the same mass, doubling moderate signal stability over a reasonably short time, typically also the flight path.The continuous production of ions in the 1–5% RSD.3 The origin of instability has several sources such ICP requires modulation since a new bundle of ions can only as changes in nebulization eYciency and variations in the be allowed to enter the flight tube when the heaviest ion has plasma tail where the ions are sampled.The signal stability reached the detector. The measurement over the whole mass can be increased significantly by measuring intensity ratios to range is essentially simultaneous and>20 000 full mass spectra an internal standard provided that sequential measurements per second can be acquired. The repetition rate of a TOF-MS of analyte and internal standard are performed within a short instrument is determined by the flight time of the slowest time span.1,3 The quadrupole mass spectrometer is only cap- (heaviest) ion.Owing to the large data flow some trimming able of such data acquisition speed for 5–7 isotopes especially of the raw data is necessary. Present systems can measure up in very short and rapid sample pulses. High-resolution mass to 65 isotopes quasi-simultaneously under default conditions. spectrometers normally do not allow this scanning speed Obviously, a very fast detector and recording system is mandaexcept for a short mass range.6 Another important limitation tory.The selected integration time and the number of replicate of scanning mass spectrometers is that the total time for a measurements performed in each sample determine the total measurement is directly proportional to the number of isotopes time of analysis. measured. Another important feature of the axial ICP-TOF-MS system There are considerable advantages in the use of a faster is the transverse rejection ion pulse (TRIP). This feature is mass spectrometer coupled to ICP, for example, time-of-flight implemented to deflect matrix ions such as O+, OH+, NO+, mass spectrometry (TOF-MS) as outlined by Hieftje et al.1,3 Ar+, ArO+ and Ar2+ which otherwise would over-saturate In this case, the plasma can be placed either orthogonal or or destroy the detector since ions of all masses theoretically axial to the time-of-flight mass spectrometer.During 1998, an enter the mass spectrometer. The deflection windows have reproducible side eVects on neighbouring isotopes and aVect axial inductively coupled plasma time-of-flight mass spec- J. Anal. At. Spectrom., 1999, 14, 1567–1572 1567roughly 5 u seen over the whole mass range. For lower m/z Prior to further dilution and analysis all samples from C, D and E were placed in an ultrasonic bath for 15 min and when (40) the narrowest TRIP setting aVects 1–3 u.The main advantage of ICP-TOF-MS is thus speed, leading these samples were to be diluted further they were shaken vigorously to suspend all of the white precipitate. to a much higher sample throughput compared with quadrupole ICP-MS. A full elemental coverage in minute volumes of Instrumentation and quantification sample and very rapid sample transients from chromatographic or laser ablation systems is also feasible using ICP-TOF-MS. An axial ICP-TOF-MS instrument (Renaissance, LECO) was Simultaneous ion extraction and high speed also enhances used for the determination of As, Zn, Sn and Pb measuring precision of isotope ratio measurements.1,3 Finally, a rapid the isotopes: 64Zn, 66Zn, 67Zn, 68Zn, 70Zn, 75As, 206Pb, 207Pb mass spectrometer allows collection of more information 208Pb, 115In, 116Sn, 118Sn and 120Sn.The signals were measured before, for example, cone maintenance is necessary when simultaneously in two diVerent detection modes, the analog nebulizing solutions with a high content of total dissolved (at high ion intensity) and the ion counting mode.The ion solids. counting mode was used for As and Zn while Pb and Sn were This paper describes the first results obtained in our labora- measured in the analog mode. Indium-115 was used as an tory using axial ICP-TOF-MS. It was decided to test the internal standard and all isotopes were also measured as ratios capability of ICP-TOF-MS by analysing five copper alloys to 115In.It is important to add the internal standard at a also independently analysed with FAAS, as this allows a direct concentration that gives a reliable signal in both detection comparison with an independent analytical technique for a modes and 20 ng ml-1 In was added to the samples and critical evaluation of its performance. The purpose was to calibrants. All samples and calibrants were acidified to 2% obtain concentrations of As, Pb, Sn and Zn in five diVerent with HNO3 (Merck suprapur grade, v/v).From the calibration homogeneous copper alloys. The copper alloys are named graphs it was clear that the contribution from 115Sn (natural A–E for the five materials. For each composition, six diVerent abundance 0.36%) on the internal standard was negligible and samples were supplied. Five replicate measurements of 13 samples containing considerable amounts of Sn were diluted diVerent isotopes were performed with a 10-s integration time to a degree where the contribution from 115Sn had no eVect for each sample and standard.For each set of five replicates on the ratio. Single element stock solutions were obtained approximately 50 s is consequently necessary using ICP- from Z-TEK (Amsterdam, The Netherlands) containing TOF-MS, while for quadrupole ICP-MS, more than 10 min 1000±3 mg ml-1 of each element (traceable to NIST SRM would be required. 3128–790704). The standards were compared with another set The measurements summarised here were performed on of single element stock solutions obtained from Merck also three diVerent days with freshly prepared calibration standards containing approximately 1000 mg ml-1 of the elements.at each occasion. Except for Sn, where a diVerence of 4.4% between both standards was measured, negligible diVerences, (below 1%) were observed between the commercial stock solutions. Equal Experimental volumes of the four Z-TEK single element stock solutions containing As, Pb, Sn and Zn were mixed to give a multi- Sample intake, cleaning and digestion element stock solution from which further dilutions were then Pieces of approximately 1 g were cut from each of 30 copper made.Calibration graphs were obtained by the analysis of 12 discs that were prepared from ultra-pure powders of the mixed standard solutions prepared to contain 0, 0.2, 0.5, 1.0, required elements. The powders were thoroughly mixed and 2.5, 5.0, 10, 25, 50, 100, 250 and 500 ng ml-1 of As, Sn, Zn then fused using hot isostatic pressing to obtain copper alloy and Pb.Samples and standards were analysed with a 10-s rods from which smaller discs could be taken. The samples integration time (n=5). For the final calculations only the (five compositions, six discs) were supplied by the Institute for ratios of 208Pb, 68Zn 75As and 120Sn to 115In were used. The Reference Materials and Measurements, Geel, Belgium, Joint operating parameters for ICP-TOF-MS are reported in Research Centre of the European Commission.The samples Table 1. For the FAAS measurements (Perkin-Elmer AAnalyst were decontaminated by washing in 10 ml of acetone (this and 300), external calibration was used and the operating paramall further products were of pro analysi grade, Merck, eters are listed in Table 2. Darmstadt, Germany) in an ultrasonic bath for 15 min. After washing, the samples were etched with 1 mol l-1 HCl (Merck) Results and discussion for 10 min using ultrasound.Following etching, the samples were washed with Milli-Q water and dried in an oven at Analytical results 100 °C. This treatment caused a black tint on some samples As can be seen from Table 3 the results obtained using ICP- and others became stained with white spots. The pieces were TOF-MS agree fairly well with those obtained by FAAS. weighed and placed in an Erlenmeyer flask and 5 ml of 50% Exceptions are Zn in compositions D and E.Reasons for this HCl (Merck) and 10 ml of 50% HNO3 (Merck) were added. are discussed below in more detail. A full spectral scan of The solution was refluxed for 20–30 min to allow complete copper alloy A is shown in Fig 1–3. In Fig. 1 the mass region dissolution of the solid. After the solution had cooled and the for Zn and As is displayed and in Fig. 2 and 3 the mass cooler had been rinsed with Milli-Q water, the resulting regions for Sn and Pb are shown.solution was transferred into a poly(ethylene) vessel and made up to approximately 100 g. The blanks were processed in the Uncertainty budgets—error propagation same way without sample present. Five procedural and reagent blanks were made i.e. one blank for each alloy composition. The uncertainty budget for Sn in alloy E and the uncertainty budgets for Zn in alloys B (high in Zn) and C (low in Zn) are For sample compositions C, D and E a white precipitate was formed rapidly after dilution with water.No further attempts included to illustrate the overall uncertainties associated with these measurements. The uncertainty budgets are presented in were made to dissolve chemically the precipitate, which was likely to be meta-stannic acid8 since X-ray fluorescence analy- Tables 4–6 and summarised in Table 7 and were made according to the guidelines issued by Eurachem.9 ses (XRF) of the precipitate mainly showed the presence of Sn. The loss of Sn during dissolution and subsequent dilution Because of the complexity of a multi-step analytical procedure where all sub-steps are associated with an uncertainty, of these solutions might aVect the analytical results for analytical techniques that require dissolution, as discussed below. it is necessary to simplify reality when an uncertainty budget is 1568 J. Anal.At. Spectrom., 1999, 14, 1567–1572Table 1 Operating parameters for axial ICP-TOF-MS Category Parameter Day 1a Day 2a Day 3a ICP conditions: Forward power/kW 1.29 1.30 1.31 Plasma flow/l min-1 16.1 14.5 14.5 Auxiliary flow/l min-1 1.00 0.804 0.782 Nebulizer flow/l min-1 0.730 0.731 0.751 Frequency/MHz 40.68 Ion deflection: O+, 16–19 u Defl. 1 start/ms 0.966 Defl. 1 width/ms 0.150 O2+, 32–36 u Defl. 2 start/ms 1.356 Defl. 2 width/ms 0.104 Ar+, 40 u Defl. 3 start/ms 1.504 Defl. 3 width/ms 0.100 Ar–Ar+, 80 u Defl. 4 start/ms 2.076 Defl. 4 width/ms 0.030 Ion focusing: Ion lens 1/V -412 Ion lens 2/V -294 Mass calibration (flight time): 7Li/ns 6746 24Mg/ns 12 138 59Co/ns 18 800 89Y/ns 22 996 115In/ns 26 084 138Ba/ns 28 536 140Ce/ns 28 740 208Pb/ns 34 942 209Bi/ns 35 024 Mass spectrometer: Flight tube/V -1470 Reflectron low/V 199 Reflectron high/V 1540 Detector/V -2490 Y steering/V -1640 Einzel lens 1/V -1280 Einzel lens 2/V -768 X steering/V -1470 aFor the categories Ion deflection, Ion focusing, Mass calibration and Mass spectrometer the settings were set identically between days.Table 2 Operating parameters for FAAS (Perkin-Elmer AAnalyst 300) Parameter As Pb Sn Zn Wavelength/nm 193.7 283.3 286.3 213.9 Slit-width/nm 0.7 0.7 0.7 0.7 Read time/s 5 5 5 5 Read delay/s 5 5 5 5 Read frequency 5 3 3 3 Flame type Air–C2H2 Air–C2H2 N2O–C2H2 Air–C2H2 made. It is the responsibility of the analyst to make sound judgements on which uncertainty components should be included and which can be overlooked. Simplifications are also necessary to break down or reduce mathematical expressions to simpler forms.The general mathematical expression to obtain the final concentration in the bronze components is: Fig. 1 Full spectral scan of mass range 62–76 from the analog channel for copper alloy A. Prominent peaks are 63Cu, 64Zn, 65Cu, 66Zn and Y={[(S/k)×Vinit×Df1×Df2]/w} (1) 68Zn. Minor peaks are derived from 67Zn, 70Zn and 75As. Note that where Y is the concentration in the sample, S is the ratio the ion counting mode was used for quantification of As and Zn.between the analytical signal for the analyte and the internal standard and k is the slope of the calibration graph. Vinit is the density-corrected weight (i.e. volume) of the initial dilution diVerent calibrants was first assessed separately. It was found that subsequent dilutions on the balance resulted in low of the dissolved sample. Df1 and Df2 are dilution factors used and w is the weight of the solid sample. The expression uncertainties as described for steps 3 and 4 given in Table 4.The two major sources of uncertainty in the calibrants thus contains only products or quotients and thus the uncertainties associated with each term can be summed up as squared arise from the initial concentration in the commercial stock solutions (i.e. 0.173% RSD) and if volumes below 0.25 ml are RSDs.9 Uncertainties associated with the concentration of the J. Anal. At. Spectrom., 1999, 14, 1567–1572 1569Table 3 Results for As, Pb, Sn and Zn in the alloys. Values are reported as % in the solid materials.Spread is given as ± one standard deviation (n=6) Copper alloy Analytical technique As Pb Sn Zn A ICP-TOF-MS 0.201±0.007 7.55±0.62 7.24±0.19 5.87±0.26 FAAS 0.184±0.003 7.93±0.59 7.26±0.09 5.96±0.14 B ICP-TOF-MS 0.106±0.006 0.397±0.015 2.10±0.04 15.09±0.54 FAAS 0.088±0.003 0.382±0.004 2.13±0.08 15.03±0.36 C ICP-TOF-MS 4.56±0.16 0.182±0.009 0.150±0.014 0.049±0.006 FAAS 4.53±0.09 0.172±0.009 0.032±0.017 0.054±0.005 D ICP-TOF-MS 0.263±0.016 9.01±1.00 8.03±0.50 0.101±0.006 FAAS 0.262±0.007 8.89±1.06 7.60±0.20 0.146±0.007 E ICP-TOF-MS 0.183±0.011 0.207±0.017 6.23±0.09 0.123±0.005 FAAS 0.181±0.007 0.196±0.003 6.30±0.40 0.153±0.003 Fig. 3 Full spectral scan of mass range 202–210 from the analog Fig. 2 Full spectral scan of mass range 110–125 from the analog channel for copper alloy A. Prominent peaks are 206Pb, 207Pb and channel for copper alloy A. Prominent peaks are 115In (internal 208Pb.Minor peak is derived from 204Pb. standard+trace amounts of 115Sn), 116Sn, 117Sn, 118Sn, 119Sn, 120Sn, 122Sn and 124Sn. Minor peaks are derived from 112Sn, 113In, 114Sn, 121Sb and 123Sb. bration graph would be the best way to account for the uncertainties in the calibrants as included in step 6. The expanded combined uncertainties and final results for used for further dilution. In the worst case, the uncertainty associated with the lowest concentration (0.2 ng ml-1) of the three examples reported in Tables 4–6 are summarised in Table 7.It can be seen that the calculated combined uncertaint- either As, Pb, Sn or Zn has been calculated to 1.02% relative uncertainty. For the 500 ng ml-1 standard solution the corre- ies are somewhat larger than the standard uncertainty from the concentrations in the six replicate samples with one excep- sponding uncertainty was only 0.201% relative. Since In was added as an internal standard to all calibrants and samples tion.This is an indication that no major source of uncertainty has been neglected and that the uncertainty budgets are likely (concentration #20 ng ml-1), the uncertainty associated with this addition to the samples must also be taken into account to be reasonably accurate. On the other hand, the lower calculated combined uncertainty for Zn in copper alloy B and is included in step 7. The addition of In is, in fact, a greater source of uncertainty alone than that associated with indicates that more sources of uncertainty possibly need to be identified and quantified.Other possible sources are (i) incom- the concentration of As, Pb, Sn and Zn in the calibrants. It was concluded that the uncertainty of the slope of the cali- plete digestion eYciency; (ii) carry-over or losses from the Table 4 Uncertainty budget for the determination of Sn in copper alloy E using ICP-TOF-MS Standard Relative Contribution to Step Description and unit Parameter Value uncertainty, u uncertainty, u (%) total uc (%) 1 Weighing of solid sample/g Ws 1.000 0.0010 0.1 2.9 2 Make up to #100 g on balancea/g D1 100.000 0.0010 0.001 0.03 3 Second dilutionb Df48 48.0 0.0010 (v1,v2) 0.1 2.9 4 Third dilutionb Df3500 50.0 0.0010 (v1,v2) 0.143 4.1 5 Measurement on ICP-TOF-MSc (ratio 120Sn/115In) MICP 1.1611 0.00191 0.16 4.7 6 Uncertainty of slope of calibration graph Uslope 0.0137 0.000225 1.64 47.8 7 Conc.of internal standard in sample/ng ml-1 SIS 20.0 0.20 1.0 29.1 8 Correction for densitya/g ml-1 r 1.0405 0.00298 0.286 8.3 aThe resulting solution following dissolution in an acid mixture was made up to approximately 100 g with Milli-Q water.The density of this solution was slightly higher than 1.00 g ml-1. In order to simplify the final calculations, a density of 1.04 was used for the initial solution for all samples. The uncertainty associated with the density has been determined to 0.286% relative.bThe uncertainties for the second and third dilution factors are calculated as Df=Ó(uv1/v1)2+(uv2/v2)2. Note that all dilutions were performed on the balance and that the uncertainties introduced in these steps are relatively small. cMeasurements on the ICP-TOF-MS instrument were undertaken in the analog mode at m/z 120Sn and m/z 115In using an integration time of 10 s and n=5. The ratio of m/z 120Sn to m/z 115In was used for calibration and quantification. 1570 J. Anal. At.Spectrom., 1999, 14, 1567–1572Table 5 Uncertainty budget for the determination of Zn in copper alloy B (high in Zn) using ICP-TOF-MS Standard Relative Contribution to Step Description and unit Parameter Value uncertainty, u uncertainty, u (%) total uc (%) 1 Weighing of solid sample/g Ws 1.000 0.0010 0.1 1.8 2 Make up to # 100 g on balancea/g D1 100.000 0.0010 0.001 0.02 3 Second dilutionb Df48 48.0 0.0010 (v1,v2) 0.1 1.8 4 Third dilutionb Df6000 50.0 0.0010 (v1,v2) 0.25 4.6 5 Measurement on ICP-TOF-MSc (ratio 68Zn/115In) MICP 0.2401 0.00413 1.72 32.0 6 Uncertainty of slope of calibration graph Uslope 0.00131 2.52 E-5 1.93 35.9 7 Conc.of internal standard in sample/ng ml-1 SIS 20.0 0.20 1.0 18.6 8 Correction for densitya/g ml-1 r 1.0405 0.00298 0.286 5.3 abAs in Table 4. cMeasurements on the ICP-TOF-MS instrument were undertaken in the ion counting mode at m/z 68Zn and m/z 115In using an integration time of 10 s and n=5. The ratio of m/z 68Zn to m/z 115In was used for calibration and quantification. Table 6 Uncertainty budget for the determination of Zn in copper alloy C (low in Zn) using ICP-TOF-MS Standard Relative Contribution to Step Description and unit Parameter Value uncertainty, u uncertainty, u (%) total uc (%) 1 Weighing of solid sample/g Ws 1.000 0.0010 0.1 0.6 2 Make up to #100 g on balancea/g D1 100.000 0.0010 0.001 0.006 3 Second dilutionb Df48 48.0 0.0010 (v1,v2) 0.1 0.6 4 Third dilutionb Df2500 50.0 0.0010 (v1,v2) 0.1 0.6 5 Measurement on ICP-TOF-MSc (ratio 68Zn/115In) MICP 0.0032 0.0004 12.5! 78.1 6 Uncertainty of slope of calibration graph Uslope 0.00131 2.52 E-5 1.93 12.1 7 Conc. of internal standard in sample/ng ml-1 SIS 20.0 0.20 1.0 6.2 8 Correction for densitya/g ml-1 r 1.0405 0.00298 0.286 1.8 abAs in Table 4.cA high degree of uncertainty is associated with step 5 and comes from the large signal fluctuation due to high background caused by the close proximity of intense signals from 65Cu and 63Cu.The background was subtracted and the ratio to 115In was then calculated manually. Otherwise as stated in Table 5. Table 7 Summary of results from uncertainty budgets given in Tables 4–6 and measurements given in Table 3 Concentration (%)±expanded Calculated combined Standard uncertainty from combined uncertaintyb (Cx±U) Element/sample uncertainty, uc a measurements (Table 3) using a coverage factor of 2 Sn in E 0.020 0.014 6.33±0.27 Zn in B 0.028 0.036 14.92±0.82 Zn in C 0.127 0.127 0.047±0.012 aTotal combined uncertainty of u in Tables 4–6.bU is calculated according to ref. 9. A coverage factor of 2 gives an interval containing approximately 95% of the distribution of values. sample digestion vessels; and (iii) memory eVects in the sample signal obtained in the ion counting mode for the 0.2 ng ml-1 standard divided by the slope of the calibration graph for each introduction system. The latter sources of uncertainty are element and then multiplied by the dilution factor.In this diYcult to quantify and they represent type B uncertainties example, samples of approximately 1 g of material were diluted i.e. an educated estimation by the analyst would have to approximately 45 000 times (1 g to 100 ml and then diluted suYce.9 450 times). In most cases the samples were diluted further (2500–6000 times) and the detection limits under these con- Calibration, detection limits and precision ditions are obviously somewhat higher.For the five procedural Equations for the calibration graphs and their correlation blanks, the four elements were below the detection limits in coeYcients are given in Table 8 along with detection limits for the solution obtained following the second dilution. The As, Pb, Sn and Zn in the solid materials. The detection limits detection limits without including the dilution factors were: 15 ng l-1 for Pb, 55 ng l-1 for Sn, 280 ng l-1 for As and were calculated as three times the standard deviation of the Table 8 Figures of merit for the ICP-TOF-MS system for determination of As, Pb, Sn and Zn in copper alloys Element Equationa (slope) r2 Detection limitb/mg g-1 Comments As 0.0020 0.9995 11 Ion counting mode used for analytical measurements Pb 0.0336 0.9998 0.7 Analog mode used for analytical measurements Sn 0.0137 0.9963 2.5 Analog mode used for analytical measurements Zn 0.0013 0.9995 15 Ion counting mode used for analytical measurements aCalibration equations for ratio of given element to 115In, with a forced zero intercept.bNote that the ion counting mode was used to calculate detection limits. J. Anal. At. Spectrom., 1999, 14, 1567–1572 1571320 ng l-1 for Zn. The detection limits reported above can be Consequences of precipitation of tin in compositions C, D and E improved to between 4 and 50 ng l-1 for these elements by For discs C, D and E a white precipitate was formed following further optimisation of several parameters.These detection dissolution and dilution of the solid material. As mentioned limits are roughly a factor of 10 worse than for quadrupole previously, Sn is a major component of the precipitates as ICP-MS. This diVerence can be explained by the modulation qualitatively assessed by XRF. For compositions A and B no of the ion source, leading to a lower sensitivity which results precipitate was observed and the results are consequently of a in higher detection limits.The RSDs for six replicate measurehigher quality. The reason for the observed diVerence between ments of the four elements in each copper alloy were in the the alloys following dissolution could be the high amount of range 2.0–6.1% for As, 3.8–11.0% for Pb, 1.4–9.4% for Sn Zn present in compositions A and B which is more easily and 3.6 – 12.9% for Zn. oxidised than Sn when partially using nitric acid for dissolution. The low content of Zn in C, D and E thus possibly Isobaric interference facilitates formation of meta-stannic acid, which is notoriously The axial ICP-TOF-MS instrument is diVerent from a conven- diYcult to dissolve as described in ref. 8. Since both ICPtional quadrupole ICP-MS instrument as all ions theoretically TOF-MS and FAAS require dissolved samples, the results reach the detector, measured intentionally or not. The obtained for Sn in C, D and E should be interpreted with implementation of the TRIP is necessary in order to avoid some caution. One major discrepancy is also obvious from the detector overload.This is achieved by removing matrix ions result for Sn in composition C as shown in Table 3. from the plasma or the sample by pulsing a high voltage perpendicular to the ion beam at regular intervals. The TRIP Conclusions settings in this particular case are displayed in Table 1. Since copper is the matrix element in the samples (see Fig. 1 and 4), Axial ICP-TOF-MS gives reasonable precision and accuracy Zn is measured close to over-saturating copper signals.The for As, Pb, Sn and Zn in most copper alloys, quantitative data background level in the ion counting mode is therefore chang- becoming available in a short time. Following initial screening, ing rapidly for masses 64, 66, 67, 68 and 71 as shown in Fig. 4. considerable saving of time can be expected compared with To avoid an over-estimation of the Zn content, the raw signals quadrupole ICP-MS.It should be noted that the four elements were treated separately in an Excel spreadsheet and the determined have diVerent masses (64–208 u) and ionization excessive background was subtracted. Nevertheless, the final energies, (As 9.81, Zn 9.39, Pb 7.41 and Sn 7.34 eV). This result for Zn in copper alloys D and E is low compared with demonstrates the wide elemental coverage, which is associated FAAS as can be seen in Table 3. The mean values obtained with ICP-MS in general and ICP-TOF-MS in particular.The are well outside±two standard deviations of the mean of the nature of the spectral interference must, however, be attributed other method of analysis. A possible reason may be that the to an inherent ICP-TOF-MS feature, namely, that all ions detector to a certain degree is blinded following the intense reach the detector unless they have been intentionally deflected. signals on m/z 63 and m/z 65. In this mass range the signals In this particular case, the very high concentration of copper at m/z 63–64, 65–66 and 65–68 are only 150 and 450 ns apart, caused interference in the determination of low concentrations respectively, and the detector does not seem to recover com- of Zn.Use of alternative plasma gases and/or the implemenpletely. This results in apparent concentrations that are too tation of a collision cell might be a way to reduce such low. Note that the observed eVect is not likely to be due to problems in the future.low abundance sensitivities as the manufacturer reports an abundance sensitivity of 105 for m/z 24 (m/m-1). It is also Acknowledgements feasible to deflect masses 63 and 65. This would aVect the Zn isotopes in a reproducible way. It was nevertheless decided A visiting post-doctoral fellowship from ‘Fonds voor not to use any deflection of the copper ions because the Wetenschappelijk Onderzoek-Vlaanderen, FWO,’ is gratefully intensities of the Zn signals are already low and adding a acknowledged by H.E. The authors are also indebted to Mr. deflection would not increase the S/N. The narrowest deflection W. Van Mol, University of Antwerp, UIA, for the FAAS window aVects roughly 5 u in this mass region. Normally, the measurements. Dr. M. Adriaens, UIA, is acknowledged for background level was low and stable i.e. around 2–5 counts the information concerning the preparation of the copper for As, Pb and Sn. alloys and Mr. P. Lemberge, UIA, is acknowledged for the XRF analyses. References 1 G. M. Hieftje, D. P. Myers, G. Li, P. P. Mahoney, S. J. Burgoyne, S. J. Ray and J. P. Guzowski, J. Anal. At. Spectrom., 1997, 12, 287. 2 Inductively Coupled PlasmaMass Spectrometry, ed. A. R. Date and A. L. Gray, Blackie, London, 1989. 3 P. P.Mahoney, S. J. Ray and G. M. Hieftje, Appl. Spectrosc., 1997, 51, 16A. 4 R. S. Houk, Anal. Chem., 1986, 58, 97A. 5 L. Moens, F. Vanhaecke, J. Riondato and R. Dams, J. Anal. At. Spectrom., 1995, 10, 267. 6 F. Vanhaecke, L. Moens, R. Dams and P. Taylor, Anal. Chem., 1996, 68, 567. 7 H. Niu and R. S. Houk, Spectrochim. Acta, Part B, 1996, 51, 779. 8 Merck Index, Merck, Rahway, NJ, 100th edn., 1989, entry 9376, Fig. 4 Zn measured in bronze component E. Note that the signals for tin. the copper isotopes (63Cu and 65Cu) over-saturate the ion-counting 9 Quantifying Uncertainty in Analytical Measurement, Eurachem detector resulting in a high and rapidly changing background. Since Guide, Laboratory of the Government Chemist, Teddington, UK, this mode had to be used for quantification, the increased background 1995. will lead to an overestimation of Zn unless it is corrected for as described in the text. The detector is, however, partially blinded, which results in an apparent concentration that is too low in the final Paper 9/04208C corrected result. 1572 J. Anal. At. Spectrom., 1999, 14, 1567–1572

ISSN:0267-9477    

DOI:10.1039/a904208c

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

Direct isotope ratio measurement of ultra-trace lead in waters by double focusing inductively coupled plasma mass spectrometry with an ultrasonic nebuliser and a desolvation unit Franck Poitrasson*a and Siv Hjorth Dundasb aLaboratoire de Ge�ochimie, UMR 5563 ‘Me�canismes de Transfert en Ge�ologie’, CNRS–Universite� Paul Sabatier, 38 rue des 36 Ponts, 31400 Toulouse, France. E-mail: franck.poitrasson@cict.fr bNorges Geologiske Undersøkelse, Leiv Eirikssons vei 39, Pb 3006-Lade, 7002 Trondheim, Norway Received 26th April 1999, Accepted 22nd July 1999 A double focusing inductively coupled plasma mass spectrometer coupled with an ultrasonic nebuliser and a desolvation unit was used to measure 206Pb/207Pb, 206Pb/208Pb and 207Pb/208Pb isotope ratios in Pb standard SRM981, diluted to total Pb concentrations ranging from 1 to 1000 ng l-1.The synthetic water standard SRM1654d, diluted by a factor of 100, and the natural riverine water standard SLRS-3 were also analysed for their Pb isotope ratios.It was found that lead isotope data with good accuracy could be obtained down to 10 ng l-1. At 1 ng l-1, the blanks became too problematic. The internal precisions found appear to be well correlated with counting statistics, and range from ca. 3 to 0.2% RSD, depending on the Pb concentration and the instrument sensitivity. The long term reproducibility is even better and may reach values down to 0.02% RSD at concentrations above 100 ng l-1 total Pb.Lead isotope values are proposed for the SRM1654d and SLRS-3 water standards. It appears that the reproducibility is aVected by matrix eVects in the latter, natural water. In future work aimed at lead isotope measurements at the sub-ng l-1 level with this approach, it will be necessary to solve serious instrument contamination and memory eVects for Pb. to implement, as it always requires an initial chemical separa- Introduction tion step carried out under very strict clean laboratory con- Lead isotopes are widely used in geological and biological ditions to avoid contamination.sciences as tracers and there is a need to measure their isotopic Hence the isotope capability of the more recent inductively composition in increasingly diverse types of samples. Thermal coupled plasma mass spectrometry (ICP-MS) was rapidly ionisation mass spectrometry (TIMS) was until recently the introduced for Pb isotope ratio determination. Quadrupole only method available to obtain precise isotope ratios.This ICP-MS proved to be much faster than TIMS for this type of technique oVers several strengths, including the ability to analysis, but it remains significantly less precise,5–8 even when produce intense and highly stable ion beams, and the thermal a refined approach is used.9 This results mostly from the less ionisation source is highly eYcient, as 5% of the Pb atoms stable nature of the ICP-MS quadrupole filter compared with loaded on a single filament with silica gel can typically be the TIMS magnetic sector.The plasma torch of the ICP also detected.1 This is a decisive advantage for very small samples. shows more short-term fluctuations than a thermal ionisation Analytical developments are still under way, and ion yields source, which is detrimental for isotopic analysis using single above 10% have recently been reported.2 Nevertheless, a major collector instruments, such as quadrupole ICP-MS.On the limitation of this technique is the diYculty of obtaining very other hand, ICP ionisation systems do not necessarily require accurate Pb isotope ratios for natural samples. With the chemical purification of the sample if the total load in solution exception of 204Pb, the other lead isotopes occurring in nature can be handled by the nebulisation system used. It also allows (206Pb, 207Pb and 208Pb) are present in variable abundance. the direct in situ measurement of lead isotopes in solids via Hence the time-dependent mass fractionation that occurs laser ablation.10–12 during the thermal ionisation process has to be corrected for More recently, magnetic sector ICP-MS showed much better by repeated analyses of a standard with a well known Pb capabilities than quadrupole instruments, with accuracy isotopic composition throughout every analytical session.and precision approaching13–15 or even as good as those of However, as shown recently by Woodhead et al.,3 this TIMS for multi-collection, magnetic sector ICP-MS.16,17 approach still results in a bias between the standards and the Nevertheless, and with the exception of a preliminary study,18 unknown samples.These latter are typically obtained after a the results reported so far with these new techniques were chemical purification from a sometimes complex matrix, and carried out using solutions with Pb concentrations ranging any remaining impurity is likely to induce specific mass typically from about 10 mg l-1 13,19 for double focusing (DF) fractionation behaviour during the thermal ionisation process.ICP-MS to about 1 mg l-1 for magnetic sector, multi-collection This matrix eVect can be successfully corrected with the ICP-MS.16,17 In contrast to other ICP-MS instruments, the simultaneous use of the ‘double-spike’ method and multi- results reported with multi-collection instruments to date were collection TIMS.3 Although this approach leads to significant acquired with Faraday cups, thus requiring intense ionic improvements in accuracy and long-term reproducibility,4 beams.Solutions of 50 mg l-1 were still required when high TIMS measurements remain time consuming, thus hindering nebulisation eYciency devices were used with multi-collection ICP-MS.20 the use of this technique for large surveys. It is also not easy J. Anal. At. Spectrom., 1999, 14, 1573–1577 1573Some geological and biological samples may display Pb water obtained from a Milli-Q system (Millipore, Bedford, MA, USA) (18 MV cm) and nitric acid (sub-boiling, quartz concentrations much lower than the mg l-1 level (e.g., spring water, snow, blood), thus requiring preconcentration tech- doubly distilled grade).These solutions were acidified to 2% HNO3 and also doped with Tl (Spectrascan certified standard niques prior to ICP-MS analysis. However, these samples may only be available in small amounts, and any sample handling solution, Teknolab, Krøbat, Norway, 1000±3 mg ml-1) at concentrations matching those of lead in the various Pb is likely to induce significant contamination for samples with lead concentrations at the sub-mg l-1 level unless extremely standard solutions analysed, in order to obtain equivalent counts for the Pb and Tl isotopes of interest.Both synthetic strict laboratory precautions are followed. We have therefore evaluated the capability of a DF-ICP-MS instrument with (NIST SRM 1643d) and natural (SLRS-3, Conseil National de Recherches, Ottawa, Canada) water standards were ana- ultrasonic nebulisation (USN) and a desolvation unit for the direct measurement of lead isotope ratios in solutions with Pb lysed in the course of this study.These standards were acidified to 2% HNO3 and their Tl concentrations were adjusted to concentrations ranging from the ng l-1 to the mg l-1 level. We show that with this approach, it is possible to obtain isotope match the Pb count rates. The SRM 1643d standard was diluted by a factor of 100 in order to obtain Pb concentrations ratios for the major lead isotopes (i.e., 206Pb, 207Pb and 208Pb) with precision, accuracy and reproducibility suitable for closer to the figures commonly found in spring waters.22–24 All containers used were cleaned using ultra-pure HNO3 environmental studies.and water. Experimental Data acquisition Instrumentation Masses 203Tl, 205Tl, 206Pb, 207Pb and 208Pb were measured using the electric scanning capability of the instrument, the The DF-ICP-MS instrument was an Element from Finnigan MAT (Bremen, Germany), the basic characteristics of which magnet being set at a fixed mass.Only the central 20% of the peaks was scanned in order to stay at the flattest part of the have been described elsewhere.21 Briefly, it consists of a reverse Nier–Johnson mass spectrometerhe magnetic sector peak tops, which should result in improved accuracy and precision for isotopic analysis.1,25 The 204Pb mass was not placed before the electrostatic field on the ion beam path.This instrument has a single electron multiplier, with discrete measured as it would increase the lowest Pb concentration measurable since it is significantly less abundant (1.4%) than dynodes, as signal collector. Although the instrument used is capable of medium (m/Dm=3000; 10% valley definition) and the other isotopes (24.1, 22.1 and 52.4% for 206Pb, 207Pb and 208Pb, respectively) in natural common lead.Further, its high mass resolution (m/Dm=7500), only the low mass resolution mode (m/Dm=300) was used since no interference determination is not critical for many environmental studies. The detector dead time used was 25 ns, a value close to the problems occur for the lead masses, and this permitted the best ion transmission to be obtained. figures experimentally determined with the Element for isotopic analysis.14,15,26 Additional details on the data acquisition In order to improve the instrument sensitivity, the Meinhard concentric nebuliser fitted on this instrument in its standard parameters used during this study are reported in Table 2 and will be discussed in more detail below.It is important to stress configuration was replaced by an ultrasonic nebuliser attached to a desolvation unit (USN 6000AT+, Cetac, Omaha, NE, that all the uncertainties reported in this paper are expressed as standard deviation (s) or relative standard deviation (RSD), USA).This improved the instrument sensitivity by two orders of magnitude (reaching 5×109–10×109 counts s-1 per mg l-1 and not as standard error (SE=s/Ón, where n is the number of measurements), as is often the case in isotopic analysis.27 208Pb) without significantly increasing the instrumental background (well below 1 count s-1), but reduced the signal stability by nearly an order of magnitude. The ICP-MS was Results and discussion not fitted with a shielded torch (CD-1 system) for this study.Further details on the instrument settings are reported in Precision Table 1. DiVerent acquisition parameters were tested in order to find the best settings for the Pb isotope measurements while keeping Reagents and standards a reasonable total analysis time: the dwell time was varied from 1 to 5 ms, the number of measurement points per mass The National Institute of Standards and Technology (NIST) Standard Reference Material (SRM) 981 for Pb isotopes was peak from 1 to 50 and the number of sweeps up to 10 000.It diluted to 1, 10, 100 and 1000 ng l-1 total lead with ultra-pure Table 2 Data acquisition parameters Table 1 Details of the ICP-MS and ultrasonic nebuliser settings Take-up time 90 s Wash time 30 s Detection mode Counting ICP-MS— Forward power 1350 W Detector dead time 25 ns Scan type E-scan Coolant flow rate 13.5 l min-1 Auxiliary flow rate 1.33 l min-1 Magnet mass 202.972 u Mass scanning window 20% Nebuliser flow rate 1.17 l min-1, adjusted for signal intensity Mass range for— 203Tl 202.905–203.040 u Ion lens settings Adjusted for signal intensity Sampling cone Nickel, 1.1 mm aperture 205Tl 204.906–205.043 u 206Pb 205.906–206.043 u Skimmer cone Nickel, 0.8 mm aperture Resolution setting 300 (m/Dm) 207Pb 206.907–207.045 u 208Pb 207.907–208.046 u Sample uptake rate 2.5 ml min-1 Ultrasonic nebuliser— Measurement points per mass peak 50 Dwell time per measurement point 0.002 s Heater temperature 140 °C Condenser temperature 3 °C Settling time between masses 0.001 s Total time per sweep 0.505 s Desolvation unit— Heater temperature 160 °C Number of sweeps 750 Total run time per sample 8 min 18 s Sweep gas flow rate 2.39 l min-1 1574 J.Anal. At. Spectrom., 1999, 14, 1573–1577appeared that whereas the best precision on each mass count rate was obtained if a long dwell time and/or a large number of points per peak were used, it was necessary to scan the various masses of interest at a suYciently high speed to improve the precision of isotope ratio measurements.The compromise found (Table 2) was similar to the settings used by Vanhaecke et al.13 It appeared that the absolute values for the dwell time or the number of points per peak are not critical. Rather, the total time spent on each mass per sweep appeared to be significant, whether adjusted through the dwell time or the number of points per peak.Similarly, the way the data reduction is carried out is important. For example, taking the standard deviation of three blocks of 250 measurements dramatically improves the precision compared with the simple calculation of the standard deviation of 750 ratios. In the present case, the data reduction used followed more or less the procedures commonly used in TIMS27 for comparison purposes. The ratios outside two standard deviations calculated on the 750 ratio were rejected. Fig. 1 Comparison of the uncertainty obtained on 206Pb/207Pb ratio This statistical treatment led to the rejection of about 5% of for a single analysis of SRM 981 Pb isotopic standard at various the measured ratios, but it did not significantly aVect the concentrations with the theoretical best uncertainty achievable accordmean. Then, the mean and standard deviation of 10 blocks of ing to Poisson counting statistics, defined as RSD (%)=100 (1/N1+1/N2)D, where N1 and N2 are the total number of counts for 75 ratios (or less if some of the corresponding data were 206Pb and 207Pb, respectively.Observed data are from Table 3. rejected ) were calculated. The final lead isotope ratios were eventually corrected for mass discrimination with the final thallium isotope ratios and the uncertainties of the Pb ratios times more concentrated solutions with the standard instruwere calculated accordingly by error propagation. Owing to ment configuration,14,15 and this makes this adaptation of the the final level of precision found for the individual ratios USN to DF-ICP-MS worthwhile.(Table 3), the type of mass discrimination law used was Another limitation of the USN is its behaviour with soluunimportant. Therefore, the simplest, linear law27 was adopted. tions characterised by a high concentration of dissolved Similarly, the level of precision is such that an additional solids.29 For instance, the data shown in Table 3 were acquired correction based on the dependence of the mass discrimination after several hours of analysis of spring waters with a total on mass, such as suggested by Hirata,17 was not necessary for load of ca. 0.4 g l-1, during which the instrument sensitivity the present calculations. It is also noteworthy that whereas decreased by a factor of 10. This signal decrease is greater the mass discrimination correction significantly improves the than what should be expected with this instrument if this was precision of the TIMS data, it is much less the case with simply related to a reduced performance of cones and injector DF-ICP-MS data.This certainly results from the progressive tube. Another set of SRM 981 standards analysed on a mass fractionation (from positive to negative) trend observed diVerent day, not after samples with a high load, gave 10 times throughout TIMS runs, which gives large standard deviations higher signals.All RSDs for solutions ranging from 1 to before correction, whereas the isotope ratios determined by 1000 ng l-1 were in that case between 0.24 and 0.63% for all plasma spectrometry are only aVected by the time-independent lead isotope ratios. mass discrimination, which is typically less than 1% u-1.5,17 The within-run uncertainties vary greatly with the lead Accuracy concentration (Table 3), as expected from counting statistics. It appears that the RSDs are between three and ten times The very high sensitivity of the instrument with the USN makes the blank problem really challenging. On the first day higher than those predicted by Poisson counting statistics (Fig. 1). This is higher than the typical factor of three reported of work, the acid blank yielded more than 105 counts s-1 on 208Pb. The 206Pb/207Pb, 206Pb/208Pb and 207Pb/208Pb ratios for in previous isotope measurements with this type of instrument in its standard configuration.14,15,28 This certainly results from the first blank measured that day were respectively 1.1953±0.0154 (1 s), 0.5025±0.0155 and 0.4213±0.0085, thus the USN, which decreases the signal stability of the instrument, as noted above.Hence part of the improved precision for low well above the SRM 981 values, at least for the first two isotope ratios. Thereafter, for the first set of SRM 981 samples concentrations gained with the USN, because of better counting statistics, is lost as a result of increased instability.analysed, a blank correction was required to find the recommended values for all concentrations, except at 1000 ng l-1. Nevertheless, this ratio between observed and theoretical precision remains not far from what is observed for about 1000 As it appeared in the course of this study that instrumental Table 3 Precision of Pb isotope ratios obtained by DF-ICP-MS on SRM 981. Each concentration corresponds to a single analysis consisting of 750 measured ratios and corrected for mass discrimination using Tl according to the procedure described in the text 206Pb/207Pb 206Pb/208Pb 207Pb/208Pb Total Pb concentration/ng l-1 Mean s RSD (%) Mean s RSD (%) Mean s RSD (%) 1 1.1517 0.0368 3.19 0.4782 0.0112 2.35 0.4203 0.0093 2.21 10 1.1088 0.0157 1.42 0.4697 0.0050 1.06 0.4241 0.0066 1.54 100 1.0946 0.0068 0.62 0.4607 0.0041 0.89 0.4221 0.0024 0.57 1000 1.0968 0.0051 0.46 0.4630 0.0032 0.68 0.4224 0.0020 0.47 Reference valuesa 1.0933 <0.0001 0.004 0.4616 0.0002 0.038 0.4222 0.0002 0.038 aValues of Hirata17 obtained with 2 mg l-1 total Pb solutions using multi-collector ICP-MS.J. Anal. At. Spectrom., 1999, 14, 1573–1577 1575contamination was a significant limitation, special attention values than the rather large internal uncertainties might suggest (Table 3). Only at 10 ng l-1 is the uncertainty associated with was paid to the instrument cleanliness before the analytical sessions, and blank analyses for Pb were run periodically to the reproducibility larger than the internal precision, but this certainly reflects the scatter induced by the blank correction, monitor the background level.It also became rapidly apparent that only doubly distilled HNO3 should be used for the blanks which becomes significant at such a concentration. A reproducibility of better than 1% RSD for all isotope ratios at this and to clean the instrument during the analytical sessions, sometimes alternately with ultra-pure water, if needed.Hence, concentration may be obtained if no blank correction is applied, but this may be detrimental to the data accuracy if as our procedures improved, a blank correction was no longer necessary for 10, 100 and 1000 ng l-1 total lead concentration the laboratory techniques are not clean enough (see above). Only the reproducibility found with 1000 ng l-1 solutions (Table 3). Only the 1 ng l-1 standard still yielded wrong values and its count rate remained close to the blank values.Hence (Table 4) approaches the best values reported in the literature for 206Pb/207Pb, 206Pb/208Pb and 207Pb/208Pb isotope ratios further work regarding the acid purity and the instrument cleanliness is required for isotopic work at such very low with this type of DF-ICP-MS.13 However, in the present study, the data were obtained on Pb solutions at much lower concen- concentration levels. It should be noted that the reference values taken in Table 3 for SRM 981 are those of Hirata,17 trations, with additional instabilities generated by the USN, after blank correction and mass bias correction with Tl.Also, determined using multi-collector ICP-MS. These figures are more precise than the NIST determinations carried out more this reproducibility represents analyses made over more than 3 months, sometimes after an analytical sequence of waters than 30 years ago by TIMS, and also probably more accurate, although marginally diVerent, because of a careful within-run which decreased the instrument sensitivity by a factor of 10 (see above).It is therefore believed that the accuracy and Tl correction for mass discrimination.17 reproducibility reported here represent conservative estimates of what may be expected from practical analyses of solutions Reproducibility with lead at ultra-trace levels using the analytical methodology The reproducibility of lead isotope measurements was esti- described in this paper.mated on the basis of analyses made on three diVerent days spanning more than 3 months (Table 4). All the data were Results for water standards blank corrected, using blank solutions prepared in the same way as the samples, although only the first-day data were Both a synthetic (SRM 1643d) and a natural (SLRS-3) standard water were analysed for their 206Pb/207Pb, 206Pb/208Pb significantly diVerent from the reference values prior to this correction if the uncertainties were taken into account. This and 207Pb/208Pb isotope ratios (Table 5).The latter displays a total lead concentration of 68 ng l-1 whereas the former was produced isotope ratios with good accuracy down to 10 ng l-1 total lead owing to the uncertainties (Table 4). It has been diluted by a factor of 100 in order to obtain a concentration of 182 ng l-1, which is in the range of concentration of interest reported before that, in contrast to what is commonly observed with many analytical techniques, including TIMS, the internal for this study.Three replicate analyses were carried out on the same dates as the three separate analyses of SRM 981 at precision is often close to the external reproducibility by ICP-MS.14 The present study is no exception as the reproduc- various concentrations (Table 4), that is, over a time span of more than 3 months. The data for SRM 1643d confirm the ibility shown, at least at 100 and 1000 ng l-1, is often better than the internal precision (compare Table 4 with Table 3).results obtained for SRM 981 in terms of internal precision according to its total Pb concentration, slightly above This suggests that the internal uncertainties of the measurements are overestimated at these concentrations. This view is 100 ng l-1 (compare Tables 3 and 5). The reproducibility of the values is excellent, even when compared with the supported by the accuracy of individual analyses, which show that the measured values are much closer to the reference 1000 ng l-1 SRM 981 standard (compare Tables 4 and 5).Table 4 Reproducibility of Pb isotope ratios obtained by DF-ICP-MS on SRM 981. Each ratio corresponds to the average of three analyses made on 16 and 17 October 1997 and 30 January 1998. See text for details 206Pb/207Pb 206Pb/208Pb 207Pb/208Pb Total Pb concentration/ng l-1 Mean SD RSD (%) Mean SD RSD (%) Mean SD RSD (%) 10 1.0766 0.0191 1.77 0.4583 0.0085 1.85 0.4211 0.0046 1.09 100 1.0951 0.0053 0.48 0.4642 0.0054 1.17 0.4225 0.0020 0.48 1000 1.0952 0.0020 0.19 0.4626 0.0007 0.15 0.4213 0.0017 0.40 Reference valuesa 1.0933 <0.0001 0.004 0.4616 0.0002 0.038 0.4222 0.0002 0.038 aValues of Hirata17 obtained with 2 mg l-1 total Pb solutions using multi-collector ICP-MS.Table 5 Lead isotope ratios obtained by DF-ICP-MS on a synthetic (SRM 1643d) and a natural water standard (SLRS-3). See text for details 206Pb/207Pb 206Pb/208Pb 207Pb/208Pb Analysis Sample No.Mean SD RSD (%) Mean SD RSD (%) Mean SD RSD (%) SRM 1643d 1 1.1507 0.0026 0.23 0.4702 0.0019 0.40 0.4082 0.0010 0.23 (182 ng l-1) 2 1.1499 0.0055 0.48 0.4703 0.0063 1.33 0.4099 0.0029 0.71 3 1.1491 0.0049 0.43 0.4702 0.0027 0.57 0.4098 0.0017 0.41 Average 1.1499 0.0008 0.07 0.4702 0.0001 0.02 0.4093 0.0010 0.24 SLRS-3 1 1.1827 0.0029 0.25 0.4856 0.0020 0.41 0.4093 0.0011 0.26 (68 ng l-1) 2 1.1833 0.0148 1.25 0.4891 0.0072 1.47 0.4127 0.0030 0.72 3 1.1745 0.0104 0.88 0.4826 0.0040 0.83 0.4111 0.0028 0.68 Average 1.1802 0.0049 0.41 0.4858 0.0033 0.68 0.4110 0.0017 0.42 1576 J.Anal. At. Spectrom., 1999, 14, 1573–1577This may result, at least in part, from the fact that no blank Magne Ødega°rd are thanked for critical reading of the manuscript. correction was applied to the SRM 1643d data (or to SLRS-3). For SLRS-3, the reproducibility is not as good as that for SRM 1643d. Although having a Pb concentration three times References lower, counting statistics alone can account for only less than 1 K.Habfast, in Modern Isotope Ratio Mass Spectrometry, ed. I. T. a factor of two diVerence between the uncertainties at these Platzner, John Wiley & Sons, Chichester, 1997, pp. 11–82. two concentrations (see the theoretical curve in Fig. 1). It is 2 H. Gerstenberger and G. Haase, Chem. Geol., 1997, 136, 309. therefore suspected that the more complex matrix of the 3 J. D. Woodhead, F. Volker and M.T. McCulloch, Analyst, 1995, SLRS-3 natural water also played a significant role in the 120, 35. 4 J. D. Woodhead and J. M. Hergt, Chem. Geol., 1997, 138, 311. diVerence in observed reproducibility, especially for the 5 K. E. Jarvis, A. L. Gray and R. S. Houk, Handbook of Inductively 206Pb/207Pb and 206Pb/208Pb ratios (Table 5). Nevertheless, Coupled Plasma Mass Spectrometry, Blackie, Glasgow, 1992. the precision of individual analyses is still in good agreement 6 M. E. Ketterer, J.Anal. At. Spectrom., 1992, 7, 1125. with that for the more concentrated 100 ng l-1 SRM 981 Pb 7 A. M. Ghazi, Appl. Geochem., 1994, 9, 627. standard (compare Tables 3 and 5), and its reproducibility 8 A. Cocherie, P. Ne�grel, S. Roy and C. Guerrot, J. Anal. At. compares favourably with this standard (compare Tables 4 Spectrom., 1998, 13, 1069. 9 I. S. Begley and B. L. Sharp, J. Anal. At. Spectrom., 1997, 12, 395. and 5). 10 B. J. Fryer, S. E. Jackson and H. P. Longerich, Chem.Geol., 1993, It can be seen that, as for SRM 981 isotopic analyses, the 109, 1. external reproducibility is often better than the internal pre- 11 R. Feng, N. Machado and J. Ludden, Geochim. Cosmochim. Acta, cision (Table 5). It is therefore concluded that, in agreement 1993, 57, 3479. with previous ICP-MS work on Pb isotopes at higher concen- 12 A. J. Walder, I. D. Abell, I. Platzner and P. A. Freedman, trations,14 the external reproducibility is certainly the best Spectrochim.Acta, Part B, 1993, 48, 397. 13 F. Vanhaecke, L. Moens and R. Dams, Anal. Chem., 1996, 68, 567. estimate of the uncertainties attached to each unknown sample 14 R. Gwiazda, D. Woolard and D. Smith, J. Anal. At. Spectrom., since the internal precision may easily be overestimated using 1998, 13, 1233. standard data processing schemes. 15 A. T. Townsend, Z. Yu, P.McGoldrick and J. A. Hutton, J. Anal. To the authors’ knowledge, no lead isotopic data have At. Spectrom., 1998, 13, 809.previously been published on these water standards. The 16 A. J. Walder and P. A. Freedman, J. Anal. At. Spectrom., 1992, 7, 571. assessment of the accuracy of the results obtained on these 17 T. Hirata, Analyst, 1996, 121, 1407. two waters is therefore not possible yet, but the data presented 18 T. Do� ring, M. Schwikowski and H. W. Ga�ggeler, Fresenius’ here will certainly be useful for future studies on the topic. J. Anal. Chem., 1997, 359, 382. 19 D. Woolard, R.Franks and D. R. Smith, J. Anal. At. Spectrom., 1998, 13, 1015. Conclusions 20 A. J. Walder, D. Koller, N. M. Reed, R. C. Hutton and P. A. Freedman, J. Anal. At. Spectrom., 1993, 8, 1037. The use of a DF-ICP-MS coupled with USN and a desolvation 21 U. Gießmann and U. Greb, Fresenius’ J. Anal. Chem., 1994, 350, unit allows the measurement of 206Pb/207Pb, 206Pb/208Pb and 186. 22 A. Criaud and C. Fouillac, Geochim. Cosmochim. Acta, 1986, 50, 207Pb/208Pb isotope ratios in waters at concentrations down to 1573. the 10 ng l-1 level with precision, accuracy and reproducibility 23 F. Monna, D. Ben Othman and J. M. Luck, Sci. Total Environ., good enough for many environmental studies. With the analyt- 1995, 166, 19. ical conditions used in the course of this work (Tables 1 and 24 F. Poitrasson, S. H. Dundas, J. P. Toutain, M. Munoz and 2), this corresponds to total lead samples as small as 200 pg. A. Rigo, Earth Planet. Sci. Lett., 1999, 169, 269. 25 A. J.Walder, in Modern Isotope Ratio Mass Spectrometry, ed. I. T. Similar studies at the ng l-1 level and lower will require Platzner, John Wiley & Sons, Chichester, 1997, pp. 83–108. further work to reduce the procedural blanks. Better blanks 26 C. Latkoczy, T. Prohaska, G. Stingeder and M. Teschler-Nicola, for reagents and laboratory conditions may be attained if the J. Anal. At. Spectrom., 1998, 13, 561. very strict conditions used, for example, in U–Pb geochronol- 27 Modern Isotope Ratio Mass Spectrometry, ed. I. T. Platzner, John ogy on accessory minerals, or glaciology, are followed. In Wiley & Sons, Chichester, 1997. contrast, improvements required with regard to the instrument 28 S. Stu�rup, M. Hansen and C. Mølgaard, J. Anal. At. Spectrom., 1997, 12, 919. cleanliness will probably be more challenging. 29 M. Thompson and J. N. Walsh, Handbook of Inductively Coupled Plasma Spectrometry, Blackie, Glasgow, 1989. This work benefited from a Leonardo da Vinci travel grant from the European Commission to F.P. Bernard Bingen and Paper 9/03300I J. Anal. At. Spectrom., 1999, 14, 1573–1

ISSN:0267-9477    

DOI:10.1039/a903300i

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

Direct REE determination in fresh waters using ultrasonic nebulization ICP-MS Ludwik Halicz, Irina Segal and Olga YoVe Geological Survey of Israel, 30 Malkhe Israel Street, 95501 Jerusalem, Israel. E-mail: Ludwik@mail.gsi.gov.il Received 29th October 1998, Accepted 7th January 1999 Fifteen samples of fresh spring water from northern Israel were analyzed for extremely low REE concentrations. The concentration of LREE was between 0.02 and 13 ng l-1 and that of HREE was between 0.01 and 3 ng l-1.The limit of detection was between 0.005 and 0.05 ng l-1 depending on the element. The eYciency of ultrasonic nebulization (USN) is dependent on the total dissolved solids in the samples and therefore the depression of the signal varies from sample to sample as a function of their chemical composition. This problem was solved by using rhenium as an internal standard. The recoveries (between 85 and 120%) found for the determination of REE in spiked natural samples together with the good agreement with known data for SRM SLRS-3 riverine water indicated the accuracy of the direct determination using the developed USN-ICP-MS procedure.The chemistry of REE makes them particularly useful in Instrumentation studies of water and marine geochemistry. The REE concen- The measurements were performed with a Perkin-Elmer trations in terrestrial waters, such as groundwaters, surface SCIEX (Concord, Ontario, Canada) ELAN 6000 ICP-MS waters and lakes, have been examined as potential tracers of system.This instrument has been described in detail by processes aVecting their geochemical evolution and of rock– Denoyer14 and Tanner15 and was equipped with a CETAC water interactions.1,2 However, REE analysis of these waters 6000AT ultrasonic nebulizer (CETAC Technologies, Omaha, by ICP-MS is still diYcult. The major restrictions are the NE, USA). The operation was fully controlled by a computer extremely low concentrations of the REE, which in many cases with a Windows-NT driven dedicated software package.The are below the instrumental detection limits. To eliminate the ICP-MS and USN instrumental operating conditions and the above problems, the determination of REE in waters by mass spectrometer acquisition settings are summarized in ICP-MS or other techniques is commonly preceded by oV- or Table 1. on-line preconcentration and matrix separation.2–13 These established methods have two disadvantages: (a) time consum- Reagents and standard solutions ing procedures (even in the case of on-line procedures a single A blank solution of 0.1 M HNO3 was prepared from J.T. repetition takes about 10 min); and (b) the high likelihood of Baker (Phillipsburg, NJ, USA) Instra-Analyzed Reagent nitric contamination of samples with concentrations in the pg l-1 acid and doubly de-ionized water (DDW) obtained by passage range. These disadvantages [in the case of fresh water samples of purified water through a NanoPure water purification containing up to a few hundred mg l-1 total dissolved salts system (Barnstead, Dubuque, IA, USA). An REE stock (TDS)] can be overcome by using an enhanced method of standard solution (High-Purity Standards, Charleston, SC, sample introduction such as ultrasonic nebulization (USN), USA; ICP-MS Multielement Standard B) was diluted with which results in an order of magnitude improvement in the 0.1 M HNO3 to 100 ng l-1.LODs of the REE.The problem of matrix eVects due to relatively high (for USN) TDS contents in natural samples can be resolved by using Re as an internal standard. The good recovery values found for spiked natural samples indicate the Table 1 Operating conditions and ICP-MS and USN settings for the accuracy of direct determination using the USN-ICP-MS direct determination of REE in fresh waters procedure. ICP-MS operating conditions— Rf power 1050 W Nebulizer gas flow rate 0.98 l min-1 Auxiliary gas flow rate 0.8 l min-1 Experimental Plasma gas flow rate 15 l min-1 Lens setting AutoLens Sample collection and processing Interface cones Nickel Mass spectrometer acquisition settings— The water samples used for this study were collected in Galilee, Dwell time 70 ms the northern part of Israel, in early June 1998.The samples Number of sweeps 11 were filtered in the field through Millipore 0.45 mm filters and Number of readings 1 immediately acidified with nitric acid (to pH about 1) and Number of replicates 5 stored in the acid pre-cleaned, low density polyethylene bottles. Scan mode Peak hopping Rhenium was added as an internal standard (to the blank, MCA channels per peak 1 USN operating conditions— standard and samples), resulting in a final Re concentration Desolvating temperature: of 1.0 ng ml-1. The TDS of the water samples ranged from Cooling -5 °C 0.012 to 0.053%, which is relatively high for USN with Heating 120 °C continuous mode aspiration. J.Anal. At. Spectrom., 1999, 14, 1579–1581 1579Table 2 Selected REE isotopes and their relative abundances, blanks, sensitivities and detection limits Abundance Blank/ Sensitivity/ Detection limit/ Analyte Mass (%) cps Mcps ppm-1 ng l-1 La 139 99.9 6 485 0.04 Ce 140 88.5 11 498 0.06 Pr 141 100 3 663 0.015 Nd 143 12.2 2 81 0.1 Sm 147 15 3 103 0.06 Eu 151 47.8 2 344 0.02 Tb 159 100 2 730 0.01 Gd 160 21.9 2 169 0.04 Dy 163 25 1 181 0.04 Ho 165 100 4 717 0.01 Er 166 36.4 1 243 0.03 Tm 169 100 1 734 0.005 Yb 174 31.8 2 242 0.03 Lu 175 97.4 2 720 0.01 Table 3 REE range in Galilee fresh water samples and results for selected spring samples and SRM SLRS-3 riverine water Range of SLRS-3/ng l-1 analytical results/ Dan spring/ Shamir spring/ Analyte ng l-1 ng l-1 ng l-1 This work Brenner et al.17 La 0.4–13 5.62 4.61 250 210 Ce 0.7–12 3.55 10.7 293 250 Pr 0.09–3.0 1.27 1.05 61 53 Nd 0.3–13 5.79 4.36 239 200 Sm 0.09–2.5 1.09 0.82 43 39 Eua 0.02–0.7 0.26 0.21 6.5 6.6 Tb 0.03–0.5 0.20 0.11 4.5 3.6 Gd 0.08–3.3 1.28 0.75 39 28 Dy 0.08–3.6 1.29 0.60 22 19.8 Ho 0.015–0.8 0.28 0.13 4.9 3.8 Er 0.06–2.5 0.86 0.35 14 11 Tm 0.01–0.3 0.12 0.06 1.6 1.5 Yb 0.04–2.0 0.70 0.35 12 9.4 Lu 0.01–0.3 0.10 0.06 1.6 1.4 aAfter correction for molecular interference of 135BaO.Results and discussion Blanks, sensitivities and detection limits Selected REE isotopes and their relative abundances, the absolute blanks [in counts s-1 (cps)], sensitivities in 106 cps ppm-1 (Mcps ppm-1) and detection limits are given in Table 2.The detection limits are based on calculation of the uncertainty involved in measurement of the blank, using 3.29 times the standard deviation,16 and are at least an order of magnitude better than those obtained with a conventional nebulizer. High blanks for La and Ce resulted in relatively poor detection limits for these elements. Matrix eVects and internal standard To test the overall recovery and matrix eVects, five water samples (with diVerent TDS) were spiked with 1, 5 and 10 ng l-1 of REE.The recovery was TDS dependent and varied from 60 to 85%. After recalculation using an internal standard, the recovery was fairly good and varied for a 1 ng l-1 spike from 85 to 120% and for a 10 ng l-1 spike from 97 to 107%. Results of water analysis The range of analytical results is given in Table 3. The precision of analysis (RSD) in general varied as a function Fig. 1 REE patterns of representative water samples and associated of concentration and at the 1–10 ng l-1 level was about 5% rocks: (a) Shamir spring and basaltic rocks;17 (b) Dan spring and limestone rocks.16 or better. Since no suitable standard reference materials of 1580 J. Anal. At. Spectrom., 1999, 14, 1579–1581fresh water exist for this level of REE, the accuracy of the References results was validated by (a) comparing our data for SRM 1 P. Smedley, Geochim. Cosmochim.Acta, 1991, 55, 2767. SLRS-3 riverine water with those obtained by Brenner et al.17 2 G. E. M. Hall, J. E. Vaive and J. W. McConnell, Chem. Geol., (Table 3) and (b) normalizing our data to North American 1995, 120, 91. Shale Composite (NASC). The latter is a convenient pro- 3 H. Elderfield and M. J. Greaves, Nature (London), 1982, 296, 214. cedure conducted in geochemistry to characterize an REE 4 H. J. W. DeBaar, P. G. Brewer and M. P. Bacon, Geochim. pattern.18 This pattern is an important tool for understanding Cosmochim.Acta, 1985, 49, 1943. 5 H. J. W. DeBaar, M. P. Bacon and P. G. Brewer, Geochim. the geochemical processes and also to detect anomalous data Cosmochim. Acta, 1985, 49, 1961. that can be due to natural processes, contamination (anthro- 6 A. Masuda and Y. Ikeuchi, Geochem. J., 1979, 13, 19. pogenic in field or laboratory) or analytical error. We found 7 D. G. Piepgras and G. J. Wasserburg, Earth Planet. Sci. Lett., two types of pattern which are represented by the Dan and 1980, 50, 128.Shamir springs (Table 3, Fig. 1). Water from the Dan spring 8 G. Glinkhammer, H. Elderfield and A. Hudson, Nature (London), gives the typical sedimentary carbonate pattern,19 whereas 1983, 305, 185. 9 M. B. Shabani, T. Akagi, H. Shimizu and A. Masuda, Anal. water from the Shamir spring corresponds to that obtained Chem., 1990, 62, 2709. from rocks found in the basaltic province on Golan Heights20 10 M. B. Shabani, T.Akagi and A. Masuda, Anal. Chem., 1992, which lie to the east of the Galilee. All samples give patterns 64, 737. corresponding to one of these two types, which suggests good 11 J. R. Jezorek and J. Freiser, Anal. Chem., 1979, 51, 366. analytical accuracy. 12 B. K. Esser, A. Volpe, J. M. Kenneally and K. Smith, Anal. Chem., 1994, 66, 1736. 13 L. Halicz, I. Gavrieli and E. Dorfman, J. Anal. At. Spectrom., 1996, 11, 811. 14 E. R. Denoyer, Int. Lab., 1995, 8. Conclusion 15 S. D. Tanner, J. Anal. At. Spectrom., 1995, 10, 905. It has been demonstrated that USN-ICP-MS with Re as an 16 D. W. Medley, R. L. Kathren and A. G. Miller, Health Phys., 1994, 67, 122. internal standard is a rapid and accurate method for determin- 17 I. B. Brenner, M. Liezers, J. Godfrey, S. Nelms and J. Cantle, ing REE in fresh water at the sub-ng l-1 level. Using this Spectrochim. Acta, Part B, 1998, 53, 1087. approach, it is possible to overcome serious problems of 18 H. R. Rolinson, in Using Geochemical Data: Evaluation, contamination and the time-consuming procedures connected Presentation, Interpretation,Wiley, New York, 1993, pp. 133–142. with preconcentration and separation processes. The REE 19 A. Bellanca, D. Masett and R. Neri, Chem. Geol., 1997, 141, 141. patterns of the water samples reflected underlying water–rock 20 Y. S. Wenstein, PhD Thesis, Institute of Earth Sciences, Hebrew University, Jerusalem, 1998. interactions. The proposed method is an excellent tool for the geochemical interpretation of REE data even at concentrations at sub-ng l-1 levels. Paper 8/08387H J. Anal. At. Spectrom., 1999, 14, 1579–1581 1581

ISSN:0267-9477    

DOI:10.1039/a808387h

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

EDTA as the modifier for the determination of Cd, Hg and Pb in fish by slurry sampling electrothermal vaporization inductively coupled plasma mass spectrometry Hsien-Chung Liao and Shiuh-Jen Jiang* Department of Chemistry, National Sun Yat-Sen University, Kaohsiung, 804, Taiwan. E-mail: sjjiang@mail.nsysu.edu.tw; Fax: 886-7-5253908 Received 1st July 1999, Accepted 11th August 1999 Ultrasonic slurry sampling electrothermal vaporization inductively coupled plasma mass spectrometry (USS-ETVICP- MS) has been applied to the determination of Cd, Hg and Pb in several fish samples.The influences of the instrument operating conditions and slurry preparation on the ion signals were reported. In this study, a relatively low vaporization temperature was used which separated the analyte from the major matrix components and improved the ion signals significantly. EDTA was used as the modifier to enhance the ion signals. Since the sensitivities of Cd, Hg and Pb in various fish slurries and aqueous solution were quite diVerent, the standard addition method and isotope dilution method were used for the determination of Cd, Hg and Pb in these fish samples.This method has been applied to the determination of Cd, Hg and Pb in dogfish muscle reference material (DORM-2) and a swordfish sample purchased from the market. The analysis results agreed with the certified values. The precision between sample replicates was better than 14% with the USS-ETV-ICP-MS method.Detection limits estimated from standard addition curves were about 3–6, 3–9 and 6–7 ng g-1 for Cd, Hg and Pb, respectively, in diVerent samples. applied to the determination of various trace elements in high Introduction salt content samples in several previous studies.11–14 Certain The majority of analyses by ICP-MS are carried out on organic additions promote atomization of the analyte prior to solutions using a conventional pneumatic nebulizer. However, vaporization of the bulk matrix, thus facilitating the temporal the type of analytical tasks that can be solved by ICP-MS can separation of the analyte and the matrix and leading to less be extended using a number of other sample introduction interference eVects.For example, EDTA was used as the techniques that can be easily adapted to ICP-MS. chemical modifier to improve the signals of Cd and Pb in Electrothermal vaporization (ETV) is one of the sample previous ETAAS applications.15–17 introduction techniques currently employed in ICP-MS.1–7 Most of the methods which exist for determining the This alternative technique to solution nebulization presents concentrations of elements in fish samples involve the digestion several advantages, including improved sensitivity, small of the sample with acid before analysis.18 The volatile elements sample size requirements and the capability for solids analysis.may be lost during the sample pretreatment step.Furthermore, Perhaps the most notable benefit of ETV-ICP-MS is the the sample dissolution step might increase the risk of contamipossibility to perform direct solids analysis.2–7 nation. In this study, ultrasonic slurry sampling ETV-ICP-MS Ultrasonic slurry sampling is one of the methods for direct was used to determine the concentrations of Cd, Hg and Pb solid sample introduction, which has been successfully used in in several fish samples directly. EDTA was used as the modifier.electrothermal atomic absorption spectrometry (ETAAS).8–10 A relatively low vaporization temperature was used which separated the analyte from the major matrix components and More recently, this approach has been extended to ETVimproved the ion signals significantly. The influences of the ICP-MS.2–4 Compared with traditional sample preparation instrument operating conditions and slurry preparation on the methods, such as acid digestion and dry ashing, slurry sampling ion signals were investigated.This method was used for the oVers several benefits including reduced sample preparation determination of Cd, Hg and Pb in dogfish muscle reference time, reduced possibility of sample contamination and material (DORM-2) and a swordfish sample purchased from decreased possibility of analyte loss before analysis. the market. Furthermore, slurry sampling combines the benefits of solid and liquid sampling and permits the use of conventional liquid sample handling apparatus such as an autosampler.2–4,8–10 Experimental Although slurry sampling ETV-ICP-MS has been applied to Apparatus and conditions several types of sample analysis, it still suVers from heterogeneity problems, transport interferences due to the matrix, residue A Perkin-Elmer Sciex (Thornhill, Ontario, Canada) ELAN accumulation in the furnace and poor reproducibility prob- 5000 ICP-MS equipped with an HGA-600MS electrothermal lems.In ETV-ICP-MS analysis, in order to alleviate non- vaporizer was used.Pyrolytic graphite coated tubes and matespectroscopic interference, a relatively low vaporization tem- rial platforms were used throughout. The transfer line consisted perature can be used which separates the volatile analyte from of 80 cm long, 6 mm id PTFE tubing. The sample introduction the major matrix components and alleviates the non- system included a Model AS-60 autosampler equipped with a USS-100 ultrasonic slurry sampler.Teflon autosampler cups spectroscopic interference significantly. This strategy has been J. Anal. At. Spectrom., 1999, 14, 1583–1588 1583Table 1 Equipment and operating conditions Slurry preparation The applicability of the method to real samples was ICP mass spectrometer Outer gas flow rate/l min-1 15 demonstrated by the analysis of dogfish muscle reference Intermediate gas flow rate/l min-1 0.74 material DORM-2 (National Research Council of Canada, Carrier gas flow rate/l min-1 1.1 Ottawa, Canada) and a swordfish sample purchased from the Rf power/kw 1.10 local market.The fresh swordfish sample obtained from the Sampler/skimmer Nickel market was cut into small pieces and dried in an oven at Data acquisition 60 °C. The dried fish sample was then ground at room tempera- Dwell time/ms 10 ture for 30 min by a Retsch MM2000 mixer mill and sieved Scan mode Peak hopping by a Retsch VE1000 sieving machine (Retsch, Haan, Sweeps per reading 1 Germany).The powder with a particle size below 100 mm was Readings per replicate 400 Signal measurement mode Integrated collected for the experiments. Before weighing for analysis, Isotopes monitored 111Cd, 112Cd, 201Hg, the fish samples were dried to constant weight by drying at 202Hg, 204Pb, 208Pb reduced pressure at room temperature in a vacuum desiccator over Mg(ClO4)2 for 24 h. Various fish samples were determined by the isotope dilution method and standard addition method.The slurry was prewere used. The USS-100 was set at 20W (50% power), and a pared as follows. A 0.1 g portion of the powder material was 10 s mixing time was used to mix the slurries before injection transferred into a 5 ml flask. A suitable amount of EDTA was of 20 ml sample aliquots for analysis. added to make a final solution containing 1% m/v EDTA. The ICP conditions were selected to maximize Cd, Hg and Suitable amounts of enriched isotope (to make the isotope Pb ion signals while a solution containing 10 ng ml-1 of these ratios of the elements studied in the slurry mixture close to elements in 1% HNO3 was continuously introduced with a unity) or various amounts of Cd, Hg and Pb element standard conventional nebulizer.The sensitivity of the instrument might solutions were added, and the slurries were diluted to the vary slightly from day to day. Since an ETV sampling device mark with pure water. For standard addition analysis, the was used and the signal obtained by ICP-MS was a transient slurries were spiked with various amounts of Cd (0, 0.5, 1, 2, signal, a 10 ms dwell time was used in the following experi- 4 and 8 ng ml-1 in the final solutions), Hg (0, 10, 20, 50, 100 ments.The peak area of the transient signal was used for data and 250 ng ml-1 in the final solutions) and Pb (0, 2, 10, 20, handling. Since the concentration of Pb in the swordfish 30 and 40 ng ml-1 in the final solutions) element standards. sample was quite high, an OmniRange setting of 5 was used The slurries were then sonicated for 10 min in an ultrasonic for the measurement of Pb.This was performed to eVectively bath, and 1 ml aliquots were removed as needed for analysis reduce the sensitivity of the ICP-MS. The ICP-MS and ETV with the use of a pipette while the slurry was being mixed with operating conditions used throughout this work are a vortex mixer. These aliquots were then deposited in Teflon summarized in Tables 1 and 2, respectively.autosampler cups for analysis. A blank containing all reagents used for slurry preparation was taken through the procedure, Reagents as outlined above, to correct for any analyte in the reagent used for sample preparation. There was no significant amount Trace metal grade HCl (35% m/m) and HNO3 (70% m/m) of the analyte in the blank. The analyte concentration in the and HPLC grade methanol were obtained from Fisher (Fair sample was then calculated by the equation described in Lawn, NJ, USA).NH4OH, EDTA and H2O2 were obtained previous papers4,21 and/or from the standard addition cali- from Merck (Darmstadt, Germany). EDTA stock solution bration curves. In order to study the eVect of ETV conditions was passed through a Chelex-100 column three times to remove and slurry preparations on the ion signals, a swordfish slurry the trace analyte in the solution.19 Element standard solutions sample was prepared as described above.were obtained from Spex (Edison, NJ, USA). Triton X-100 was obtained from Sigma Chemicals (St. Louis, MO, USA). Thioacetamide (TAC) was obtained from Tokyo Chemical Results and discussion Industry (Tokyo, Japan). Enriched isotopes of 111CdO and 204Pb(NO3)2 were Selection of modifier purchased from the Oak Ridge National Laboratory (Oak Ridge, TN, USA) and 201HgO was obtained from Cambridge Fig. 1 shows the temporal behaviour of Cd, Hg and Pb in fish slurry. The drying temperatures were set at 80 °C for 30 s and Isotope Laboratories (Andover, MA, USA).Stock solutions of each element were prepared by dissolution of an accurately 120 °C for 10 s; the pyrolysis temperature was set at 200 °C and the vaporization temperature was set at 1200 °C. The weighed quantity of the material in HNO3 or HCl and dilution to volume. The concentrations of the spike solutions were swordfish slurry was injected into the furnace without any further pretreatment. As shown in Fig. 1, all the elements verified by reversed spike isotope dilution ICP-MS.20 Owing to the mass discrimination eVect, the intensities obtained studied evaporated at 1200 °C, although a double vaporization peak was observed for Pb and significant ion signals of Hg during isotope ratio determination of each solution were used to calculate the isotopic abundance of each element. and Pb were detected in the condition stage when no modifier Table 2 HGA-600MS temperature programme; sample volume 20 ml Step Drying Pyrolysis Vaporization Cooling Condition Cooling Temperature/°C 80 120 180 1000 20 1500 2500 20 Ramp time/s 30 10 10 1 1 1 1 1 Hold time/s 10 10 20 13 5 10 10 10 Signal acquisition — — — ON — — — — 1584 J. Anal.At. Spectrom., 1999, 14, 1583–1588Fig. 1 ETV-ICP-MS signals of Cd, Hg and Pb. Drying temperatures were set at 80 °C and 120 °C, the pyrolysis temperature was set at 200 °C and the vaporization temperature was set at 1200 °C.No modifier was used. The slurry solution contained 2% m/v swordfish sample. An OmniRange setting of 5 was used for Pb measurement. was used. Modifiers are commonly used in ETV-ICP-MS in order to reduce losses of analyte due to condensation on diVerent parts of the ETV cell or the transfer line that connects the ETV to the ICP-MS.1–3 Although most eVects of the Fig. 2 EVect of various modifiers on the ion signals: (a) with 1% modifiers used in ETV-ICP-MS analysis are believed to be EDTA as modifier; (b) with 1% EDTA and 1% HCl as modifier; (c) physical, it is likely that chemical eVects could also be involved with 1% EDTA and 1% NH4OH as modifier.Temperature programme in specific instances.3 The use of a modifier changes the was the same as that in Fig. 1. The composition of the slurry was the chemical or physical characteristics of the sample and/or the same as that in Fig. 1, except for the addition of various modifiers. atomizer surface in order to improve quantification.Certain additions promote vaporization of the analyte before vaporiz- study. EDTA was also used as extracting reagent for the ation of the bulk matrix, thus facilitating the temporal separa- extraction of metal ions from biological material in a previous tion of the analyte and background signals.11,12 In this study, application.22 Fig. 3 shows the eVect of the amount of EDTA several modifiers, including EDTA and several EDTA contain- on the ion signals.As shown, the signals of the elements ing mixture modifiers, were tested to obtain the best signals studied decreased with increasing EDTA concentration when of the elements studied. Fig. 2(a) shows the eVects of the the EDTA concentration was greater than 1% m/v. However, EDTA modifier on the ion signals. As shown, the Pb vaporiz- as shown in Fig. 2(a), a more symmetrical sharp peak was ation peak was improved significantly when 1% m/v EDTA obtained when 1% EDTA was used as the modifier.After was used as the modifier. In this experiment, no significant evaluation, 1% m/v of EDTA was selected in the subsequent ion signal of Cd, Hg or Pb was observed at the condition ETV-ICP-MS analysis. stages which means that all the Cd, Hg and Pb evaporated at 1200 °C. The condition temperatures were set at 1500 and EVect of slurry preparation on ion signal 2500 °C. Fig. 2 also shows the eVects of other modifiers on the ion signals of the elements studied.As shown in Figs. 2(b) ETAAS has been successfully applied to the analysis of slurries.8–10 Certain factors, such as particle size, analyte and 2(c), the addition of other reagents suppressed the ion signals. Table 3 shows the eVects of various modifiers on the partitioning, maximum slurry concentration and slurry homogeneity, are important for the success of the analysis of slurries integrated peak area of the elements studied. Although the ion signal of Pb increased when 10% v/v CH3 OH was added by ETAAS.8,9 The eVects of several parameters of the slurry preparation on the ion signals were investigated in the and the Cd signal increased slightly when 0.1% TAC was added, for the simultaneous determination of these elements, experiments.An important factor in the slurry technique is the slurry after evaluation, EDTA was selected as the modifier in this Table 3 EVect of various modifiers on ion signalsa (n=7) Peak area/counts Modifier 111Cd 202Hg 208Pb 1% EDTA 18100±1800 77000±4200 96300±2700 1% EDTA+0.1% TAC 20900±2100 55900±3480 95400±2330 1% EDTA+1% HCl 11300±1160 53800±3360 42900±3740 1% EDTA+1% HNO3 10600±2260 67100±6640 41400±2860 1% EDTA+1% NH4OH 17700±1100 65700±5870 74900±5910 1% EDTA+1% H2O2 15800±807 67900±1980 102000±7860 1% EDTA+10% CH3OH 15600±953 71800±3030 118000±9280 1% EDTA+0.1% Triton X-100 16100±2220 61400±6130 84900±8420 aValues are means of seven measurements±standard deviation.J. Anal. At. Spectrom., 1999, 14, 1583–1588 1585Fig. 3 EVect of EDTA concentration on the ion signals. Slurry Fig. 4 EVect of the pyrolysis temperature on the ion signal. The slurry solution contained 2% m/v swordfish sample and was spiked with solution contained 2% m/v swordfish sample and 1% m/v EDTA. The various amounts of EDTA. Each data point represents the mean of vaporization temperature was set at 1200 °C. All data are relative to five measurements. All data are relative to the first point.the first point. concentration. However, dilution of the slurry can only be carried out within a limited range. In order to balance sample homogeneity, analyte signals and complete vaporization of the introduced sample, a dilution factor of 50 was used in the experiments.23 The concentration of acid in the slurry solution can aVect the rate of extraction of metal ions and the precision of signal measurement.10 If a large percentage of the analyte is extracted into the liquid phase, the precision will approach that obtainable with a conventional liquid digest.Also, as described by Gregoire et al.,1 the presence of mineral acid in ETV-ICP-MS analysis can aVect the ion signals. They found that the analyte signals were enhanced by as much as a factor of two in the presence of 1% v/v HNO3. The eVect of HNO3 and HCl in the slurry sample on the ion signals was studied in this work. However, as shown in Fig. 2 and Table 3, the addition of acid did not improve the ion signals in the current study.This may be because the addition of acid may deteriorate the completion of EDTA complex formation. For the best ion signals, no acid was used in the slurry preparation in the experiments. Furthermore, as shown in Table 3, the addition of surfactant suppressed the ion signals slightly. In the experiments, no surfactant was used in the slurry preparation. Selection of pyrolysis and vaporization temperatures Fig. 5 EVect of the vaporization temperature on (a) the ion signal Fig. 4 shows the eVect of the pyrolysis temperature on the ion and (b) the signal-to-background ratio. The slurry solution contained signals. The ion signals of the elements studied increased with 2% m/v swordfish sample and 1% m/v EDTA. The pyrolysis increasing pyrolysis temperature at temperatures less than temperature was set at 180 °C. All data are relative to the first point. 180 °C. This may be due to the removal of the more volatile matrix during the pyrolysis stage and alleviation of the non- peak was obtained when a higher vaporization temperature spectroscopic interference at the vaporization stage.However, was used. In the experiments, 1000 °C was selected as the the signals of the elements studied decreased rapidly at pyrol- vaporization temperature. A summary of the ETV operating ysis temperatures above 180 °C. This may be due to vaporiz- conditions is given in Table 2. ation of the EDTA complexes.For the simultaneous determination of the elements studied, the pyrolysis tempera- Comparison of proposed method and conventional method ture was set at 180 °C. Fig. 5 shows the eVect of the vaporization temperature on Table 4 lists the ion signals obtained with diVerent temperature programmes. The temperature programme used for the ‘con- the ion signals. As shown in Fig. 5(a), the signals of the elements studied decreased gradually with increasing vaporiz- ventional method’ was reported in ref. 24.As shown in Table 4, the ion signals obtained with the proposed method were 8.7, ation temperature. An increase in vaporization temperature increased the amount of matrix vaporized with the analyte, 9.7 and 14.1 times greater than those obtained using the conventional method for Cd, Hg and Pb, respectively. The which caused a gradual decrease in the integrated peak area. Furthermore, as shown in Fig. 5(b), a higher signal-to- precision of the peak area of seven consecutive injections was better than 7% with the current method, which was also better background ratio was obtained when the vaporization temperature was set at 1000 °C.Moreover, a more symmetrical than that obtained with the conventional method. 1586 J. Anal. At. Spectrom., 1999, 14, 1583–1588Table 4 Comparison of ion signals for diVerent temperature Determination of Cd, Hg and Pb in fish by USS-ETV-ICP-MS programmesa (n=7) In order to validate the USS-ETV-ICP-MS method, the Peak area/counts concentrations of Cd, Hg and Pb were determined in two fish samples.In order to evaluate the possibility of using the Methodb Cd Hg Pb external calibration method for the determination of trace elements in fish samples by USS-ETV-ICP-MS, the vaporiz- Proposed method 5380±380 14900±410 17300±1000 ation behaviour of Cd, Hg and Pb in diVerent matrices was Conventional method 620±60 1530±240 1230±170 carefully studied. Fig. 6 shows the vaporization curves of the Relative signal 8.7 9.7 14.1 elements studied when the analytes were in diVerent matrices.aValues are means of seven measurements±standard deviation. All elements appeared at approximately the same time when bProposed method: signal obtained with low vaporization temperature. Conventional method: temperature programming was as in ref. 24. the analytes were in diVerent matrices. This may be due to the Relative signal: peak area count obtained with conventional method high volatility of the EDTA complexes of these elements.assigned as unity. However, the signals of the elements studied were much higher in aqueous solution than in the slurry solution. This may be due to the diVerent transport eYciency and non-spectroscopic Table 5 EVect of Mo concentration on Cd isotope ratio determinationa interference between the slurries and aqueous solution. (n=7) Therefore, the external calibration method could not be used Peak area/counts for the quantification of these elements in these samples.The isotope dilution method and standard addition method were Sample 111Cd 113Cd 113Cd/111Cd used for the determination of Cd, Hg and Pb in these samples. The analysis results are shown in Table 6. The determined DORM-2 fish slurry 1420±110 1380±100 0.97±0.04 concentrations in the dogfish muscle reference sample by the DORM-2 fish slurry 1480±150 1450±140 0.98±0.03 +0.1 mg ml-1 Mo aThe fish slurry solution contained 2% m/v DORM-2 fish muscle and 1% m/v EDTA.Furthermore, since no Pd modifier was used in this study, no thermal pretreatment of the Pd modifier was necessary. Thus, the analysis time and reagent blank were reduced. Interference study For ICP-MS analysis, the components of the sample could produce several metal oxide or metal hydroxide ions which might interfere with the determination of the analyte. For example, 111Cd experiences interference from 94Zr16OH+ and 95Mo16O+, 202Hg from 186W16O+, etc.In order to evaluate the significance of these interferences, experiments were performed to check the interferences caused by Zr, Mo and W in the fish samples. A stock slurry sample was prepared as described previously. The prepared fish slurry sample was spiked with 0.1 mg ml-1 of Mo. This concentration is about 4–8 times that of Mo in the prepared slurry sample. As shown in Table 5, although the ion signals changed slightly, the isotope ratio of Cd was not aVected by Mo at these concentrations.The precision of isotope ratio determination was better than 4% with the ETV sample introduction device. Although not shown, in other experiments we found that Zr and W in the fish sample did not aVect Cd and Hg Fig. 6 ETV-ICP-MS vaporization curves of Cd, Hg and Pb in (a) determinations. These experiments demonstrated that the slurry solution and (b) aqueous solution. Slurry solution contained concentrations of Cd and Hg in the fish samples can be 2% m/v DORM-2 sample and 1% m/v EDTA.Aqueous solution determined directly by USS-ETV-ICP-MS without significant contained same concentrations of the analytes as in (a) in 1% EDTA solution. spectroscopic interference. Table 6 Determination of Cd, Hg and Pb in fish samples by USS-ETV-ICP-MSa (n=3) Element Sample Methodb Cd/mg g-1 Hg/mg g-1 Pb/mg g-1 DORM-2 Method 1 0.039±0.003 4.53±0.15 0.062±0.004 Method 2 0.040±0.005 4.50±0.22 0.057±0.008 Certified 0.043±0.008 4.64±0.26 0.065±0.007 Detection limit 0.003 0.009 0.007 Swordfish Method 1 0.211±0.014 0.840±0.068 2.18±0.18 Method 2 0.217±0.018 0.831±0.073 2.21±0.11 Detection limit 0.006 0.003 0.006 aResults are means of three measurements±standard deviation.bMethod 1: isotope dilution method. Method 2: standard addition method. Certified: NRCC certified values. Detection limit: estimated from standard addition calibration curve. J. Anal. At. Spectrom., 1999, 14, 1583–1588 15873 R.D. Ediger and S. A. Beres, Spectrochim. Acta, Part B, 1992, standard addition method and isotope dilution method were 47, 907. in good agreement with the certified values. The results for 4 M.-J. Liaw and S.-J. Jiang, J. Anal. At. Spectrom., 1996,11, 555. the swordfish sample, for which no reference values were 5 J.Wang, J. M. Carey and J. A. Caruso, Spectrochim. Acta, Part B, available, were also found to be in good agreement between 1994, 49, 193. the isotope dilution method and standard addition method. 6 F.Vanhaecke, S. Boonen, L. Moens and R. Dams, J. Anal. At. Spectrom., 1995, 10, 81. The precision between sample replicates was better than 14% 7 C. J. Park, J. C. Van Loon, P. Arrowsmith and J. B. French, Anal. for all determinations. This experiment indicated that Cd, Hg Chem., 1987, 59, 2191. and Pb in fish could be readily quantified by the proposed 8 C. Bendicho and M. T. C. de Loos-Vollebregt, J. Anal. At. USS-ETV-ICP-MS method. Spectrom., 1991, 6, 353.The detection limits of the elements studied in various 9 N. J. Miller-Ihli, Anal. Chem., 1992, 64, 964A. samples (Table 6) were determined from the standard addition 10 N. J. Miller-Ihli, J. Anal. At. Spectrom., 1994, 9, 1129. 11 R. J. Bowins and R. H. McNutt, J. Anal. At. Spectrom., 1994, curves of each element in the diVerent samples. The detection 9, 1233. limit was based on the usual definition as the concentration 12 K.-H. Lee, S.-H. Liu and S.-J. Jiang, Analyst, 1998, 123, 1557. of the analyte yielding a signal equivalent to three times the 13 K.-H. Lee, S.-J. Jiang and H.-W. Liu, J. Anal. At. Spectrom., 1998, standard deviation of the reagent blank signal (n=7). The 13, 1227. detection limits estimated from the standard addition curves 14 J. W. H. Lam, R. E. Sturgeon and J. W. McLaren, Spectrochim. were in the range 3–6, 3–9 and 6–7 ng g-1 for Cd, Hg and Acta, Part B, 1999, 54, 443. 15 F. Dolinsek and J. Stupar, Analyst, 1973, 98, 841. Pb, respectively, in diVerent samples. Better detection limits 16 R. Guevremont, Anal. Chem., 1980, 52, 1574. are to be expected with better purified reagents. 17 R. Guevremont, R. E. Sturgeon and S. S. Berman, Anal. Chim. Acta, 1980, 115, 163. Acknowledgement 18 M. Navarro, M. C. Lopez, H. Lopez and M. Sanchez, Anal. Chim. Acta, 1992, 257, 155. This research was supported by a grant from the National 19 S.-C. Pai and T.-H. Fang, Mar. Chem., 1990, 29, 295. Science Council of the Republic of China under Contract 20 C.-C. Chang and S.-J. Jiang, J. Anal. At. Spectrom., 1997, 12, 75. 21 T.-J. Hwang and S.-J. Jiang, J. Anal. At. Spectrom., 1996, 11, 353. NSC 88–2113-M-110–017. 22 T. Uchida, H. Isoyama, K. Yamada, K. Oguchi, G. Nakagawa, H. Sugie and C. Iida, Anal. Chim. Acta, 1992, 256, 277. References 23 Y.-C. Li and S.-J. Jiang, Anal. Chim. Acta, 1998, 359, 205. 24 M.-J. Liaw and S.-J. Jiang, Spectrochim. Acta, Part B, 1997, 52, 1 D. C. Gregoire, D. M. Goltz, M. M. Lamoureux and 779. C. L. Chakrabarti, J. Anal. At. Spectrom., 1994, 9, 919. 2 D. C. Gregoire, N. J. Miller-Ihli and R. E. Sturgeon, J. Anal. At. Spectrom., 1994, 9, 605. Paper 9/05328J 1588 J. Anal. At. Spectrom., 1999, 14, 1583–1588

ISSN:0267-9477    

DOI:10.1039/a905328j

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

Determination of copper, zinc, cadmium and lead in a fish otolith certified reference material by isotope dilution inductively coupled plasma mass spectrometry using oV-line solvent extraction Jun Yoshinaga,*†a Masatoshi Moritaa and John S. Edmonds‡b aNational Institute for Environmental Studies, 16-2 Onogawa, Tsukuba, Ibaraki 305-0053, Japan bWestern Australian Marine Research Laboratories, P.O. Box 20, North Beach, WA 6020, Australia Received 8th June 1999, Accepted 29th July 1999 Copper, Zn, Cd and Pb concentrations in a fish otolith (ear stone) certified reference material (CRM) were determined by isotope dilution inductively coupled plasma mass spectrometry (ID-ICP-MS).The material consists of almost pure calcium carbonate, and contained 39% calcium and less than 1 ppm of these heavy metals. An oV-line solvent extraction was employed to separate the metals from the sample matrix. Contamination from reagents and from the extraction procedure was minimal and precise results were obtained for these four metals.The mean concentration (mg kg-1 dry wt) and standard deviation (1s) were 0.742±0.007 for Cu, 0.471±0.002 for Zn, 0.0028±0.0002 for Cd and 0.0234±0.0003 for Pb (n=5). An ICP-MS determination by external calibration after solvent extraction was used as a back-up method and the results obtained (Cu, 0.73; Zn, 0.43; Cd, 0.0028; Pb, 0.026; n=3 for each) were in good agreement with the ID-ICP-MS results. The present ID-ICP-MS results are used as reference values of the fish otolith CRM.populations of Atlantic cod Gadus morhua. The facility to Introduction include information on heavy metals, in addition to that on Otoliths consisting of aragonite are contained in the inner ear alkali metals and alkaline earths, is likely to extend the of all bony fish (teleosts) and are thought to have various usefulness of otolith chemistry for studying fish migrations functions, including the maintenance of body balance.Element and for fish population discrimination. concentrations in sagittal otoliths (usually the largest of the The analysis reported here aimed at providing the CRM three pairs present) are known to vary with physiological with reference values for selected heavy metals, viz. Cu, Zn, factors, e.g. age and growth rate,1,2 as well as environmental Cd and Pb. These heavy metals were determined by isotope factors, e.g. temperature, salinity and water chemistry.3,4 This dilution inductively coupled plasma mass spectrometry variation has attracted the interest of marine ecologists and (ID-ICP-MS) with solvent extraction using ammonium pyrrol- fisheries scientists as a tool for tracking the migration of idinedithiocarbamate (APDC) or 8-hydroxyquinoline (HOQ) individual fish5,6 and for the delineation of fish populations.7–9 as a chelate.Although solvent extraction is an established The National Institute for Environmental Studies (NIES), procedure for the separation and concentration of heavy Japan Environment Agency, in collaboration with Western metals from a variety of samples, oV-line solvent extraction Australian Marine Research Laboratories (WAMRL), has not often been used for pretreatment in ICP-MS analysis.recently prepared a fish otolith certified reference material Although an on-line system using a chelating resin has been (CRM) to meet a growing demand for the quality control of gaining wide acceptance for the separation and concentration otolith element analysis.The CRM was certified for concen- of metals,11 Batterham et al.12 recently demonstrated the utility trations of Na, Mg, K, Ca, Sr and Ba based on a collaborative of oV-line dithiocarbamate solvent extraction in the ICP-MS analysis involving NIES and five other laboratories. These determination of metals in sea water. The present study alkali metals and alkaline earths were the analytes of interest demonstrates the successful application of an oV-line solvent in most previous research on otolith chemistry.The details of extraction for the separation of the four selected heavy metals the preparation, homogeneity assessment and certification will from the otolith carbonate matrix. be published elsewhere.10 Fish otoliths consist of almost pure calcium carbonate and contain very low concentrations of heavy metals. The accurate determination of these heavy metals at trace levels in the Materials and methods carbonate matrix is a considerable analytical challenge.This Materials may be the reason why the analysis of heavy metals has usually been omitted in previous research. However, Campana et al.9 The fish otolith CRM was prepared by pulverizing and homosuccessfully determined the concentrations of six elements, genizing about 1.4 kg of sagittal otoliths extracted from red including Zn and Pb, in otoliths for the discrimination of emperor (Lutjanus sebae) collected from the northwest coast of Western Australia.Each vial contains approximately 3 g of †Present address: Institute of Environmental Studies, The University prepared powdered otoliths. The moisture content of the vials of Tokyo, Hongo 7-3-1, Bunkyo, Tokyo 113-0033, Japan. used in the present analysis was measured according to a ‡Present address: Department of Chemistry, De Montfort University, The Gateway, Leicester, UK LE1 9BH. specified drying method.10 J. Anal.At. Spectrom., 1999, 14, 1589–1592 1589Reagents in the back-extracted sample solution was approximately 0.5 M. ‘Ultrapure’ grade nitric acid for sample digestion (Kanto Cadmium and Pb were determined directly from the back- Chemicals, Tokyo, Japan) was used without further purifi- extracted sample solutions. Zinc was determined after a fivecation. A 2% (w/v) aqueous solution of APDC [atomic absorp- fold dilution of the back-extracted solution. Copper was tion spectrophotometry (AAS) grade (Kanto Chemicals)] extracted with HOQ from the back-extracted solution at pH 4 was prepared daily and purified by twice extracting with after the Zn, Cd and Pb determinations.AAS grade 2,6-dimethyl-4-heptanone (DIBK) (Wako Pure Sample digestion and subsequent extraction and back- Chemicals, Osaka, Japan), which had, in turn, been purified extraction were carried out in a class 1000 clean room. by extraction with 2 M HNO3 (2×) and with Milli-Q (MQ) water (2×) prior to use.A 1% solution of HOQ (Merck, Instrument Darmstadt, Germany) was prepared daily by dissolving 0.5 g The ICP-MS used in this study was an HP-4500 of Yokogawa in 2.5 mL of acetic acid (semiconductor grade, Kanto Analytical Systems (Tokyo, Japan). Typical operational con- Chemicals) and diluted to 50 mL with MQ water. Acetate ditions were as follows: rf power, 1200 W; reflected power, buVer (1 M, pH 5.2) was prepared by mixing appropriate <10 W; plasma Ar flow rate, 15 L min-1; auxiliary Ar flow amounts of ammonium hydroxide and acetic acid (AA-100 rate, 1.25 L min-1; nebulizer Ar flow rate, 1.10 L min-1.The grade, Tama Chemicals, Kawasaki, Japan). settings of the ion lens systems and other parameters were Standard solutions (1000 mg kg-1) of Cu, Zn and Cd were tuned daily to obtain maximum stability and sensitivity by prepared by dissolving high purity metal in dilute HNO3. The using a tuning solution containing 10 ppb each of Co, Ga, Y, standard solution of Pb was SRM 981 purchased from the In, Ce, Tl and Bi.National Institute of Standards and Technology (NIST, MD, USA). These solutions were mixed and diluted to prepare a Isotope ratio measurements working multi-element standard solution (20 ppb of each) in 0.5 M or 0.14M HNO3 for the mass discrimination correction In ID analyses, the isotope ratios, i.e. 63Cu/65Cu, 68Zn/66Zn, in isotope ratio measurements. 112Cd/114Cd and 208Pb/206Pb, were measured in the back- Stable isotope spikes, 65Cu, 66Zn and 112Cd, were purchased extracted solution.Ion counts at each mass, 3 points acquifrom Oak Ridge National Laboratory (TN, USA) in metallic sition mass-1, were integrated for 3 s by peak jumping. The form. They were dissolved in dilute HNO3 and stored individu- number of scans was set at 100; thus the dwell time was 30 ms ally in clean Teflon bottles. An accurate concentration of the mass-1 or 10 ms point-1. One sample run consisted of five spike was determined by a reverse isotope dilution technique replicate measurements. Typical within-run precision of the using the standard solutions mentioned above.The 206Pb spike isotope ratio measurements at the 20 ppb level was 0.3–0.5%. used was SRM 911 purchased from NIST. The four stable Mass discrimination was corrected by periodic measurements isotope solutions were mixed to prepare a mixed-spike solution of a standard solution prepared from high purity metal and (0.14 M HNO3).NIST SRM 981 for Pb. The following known spectroscopic interferences on the selected isotopes were monitored and corrected mathemat- Digestion ically. Interference from 137Ba2+ on 68Zn was corrected by A portion of the CRM (1 g) was accurately weighed into a measuring the Ba2+ production rate (2–3% of the ion counts Teflon beaker of 100 or 200 mL capacity. An appropriate of 138Ba) daily and monitoring the ion count at m/z=138 in amount of the mixed-spike solution was then added and all of the samples.Isobaric interference from Sn to Cd isotopes accurately weighed. In the case of a natural abundance was corrected mathematically by monitoring the ion count at measurement, no stable isotope spike was added.Water (5 mL) m/z=118 in all of the samples. was added to suspend the powder and the beaker was covered A typical measurement sequence was as follows: HNO3 with a Teflon watch-glass. Nitric acid (5 mL) was slowly added blank, 20 ppb standard solution, procedural blank, sample 1, and, after 30 min standing at room temperature, when the procedural blank, sample 2, HNO3 blank, standard solution.violent reaction had ceased, the beaker was placed on a hot The blank counts, measured just prior to any samples and plate and heated at 100 °C for a further 30 min. The tempera- standards, were subtracted from the ion counts of the subture was raised to 140 °C and the sample was heated for sequent sample or standard.When >0.5% bias was found for another 4 h. The watch-glass was then removed and the a standard solution, a mass calibration was performed and solution was evaporated to dryness at 180 °C. The residue the sample measurements based on the biased standard were was dissolved in 0.14 M HNO3 (20 g) and transferred to a remeasured or corrected using an appropriate factor. screw-capped perfluroalkoxy (PFA) centrifuge tube (Nalgene, The measured isotope ratios were converted to element NY, USA). concentrations in the sample according to the ID equation.13 For the determination of Co, Ni, Cu, Zn, Cd and Pb by external calibration, 0.25 g of the CRM was digested, without Results and discussion the stable isotope spike, as described above.Contamination and extraction eYciency Extraction and back-extraction A procedural blank was prepared for each series of sample digestion–extraction–back-extraction. Means and standard A preliminary experiment showed that the addition of 5 mL of 1 M acetate buVer to the 20 g digest (in 0.14M HNO3) deviations of the heavy metal concentrations in the procedural blanks were 6.2±1.4 ng for Ni, 0.87±0.08 ng for Cu, brought the solution to pH 4.The purified 2% APDC solution (2 mL) and purified DIBK (4 g or 5 mL) were then added. 16±10 ng for Zn and 0.22±0.07 ng for Pb (n=3–5 depending on the metal ). The levels were similar to those of the reagent The mixture was manually shaken for 3 min.The DIBK layer was transferred to another screw-capped PFA vial with a blanks (5.4 for Ni, 0.52 for Cu and 0.24 for Pb), indicating that reagent impurity was the primary source of the procedural quartz Pasteur pipette. 2 M HNO3 (1 mL) was added and the mixture was manually shaken for 3 min. After centrifugation, blank. However, the procedural blank for Zn was significantly larger than the reagent blank (6 ng) and it was variable. This the HNO3 layer (~1 g) was pipetted into a 7 mL capacity screw-capped vial (Savillex, Gasukuro Kogyo, Japan) and indicated that contamination from another source(s) was significant for Zn.Procedural blanks were not detected for Co diluted to 4 g with MQ water. Thus the HNO3 concentration 1590 J. Anal. At. Spectrom., 1999, 14, 1589–1592and Cd (absolute detection limits of 0.008 and 0.02 ng, respect- Isotope ratios in an unspiked otolith CRM sample ively). The blank levels will be discussed in relation to the The ratio of the selected isotopes in the sample is considered metal concentrations in the fish otolith CRM in a later section. a sensitive indicator of spectroscopic interference.Table 1 Examination of blanks also revealed, importantly, that the shows the ratios of selected isotopes in the unspiked CRM blank concentrations of Cu determined at m/z=65 were higher along with their natural abundance.14 For Zn and Cd, the by one order of magnitude than those determined at m/z=63.isotopic ratios were in agreement with the natural abundance This result indicated that there was interference from sulfur, indicating the absence of spectroscopic interference. Barium 32S16O2H+. One of the blanks contained 84 mg L-1 sulfur as was not extracted by the present procedure and no interference determined by ICP atomic emission spectrometry. The source on 68Zn was expected. The relatively low Sn content of the of the sulfur in the back-extracted solution was likely to have CRM allowed Cd isotope ratio measurement without inter- been APDC.Some APDC might be decomposed in the process ference, although some Sn was extracted by the extraction of extraction and/or back-extraction. This result demonstrates procedure used. However, as expected from the blank evaluthat APDC extraction–back-extraction is not a suitable ation mentioned earlier, 63Cu/65Cu in samples extracted with method for the determination of Cu by ID-ICP-MS. APDC was not consistent with their natural abundance The extraction eYciency was examined by a spike recovery because of molecular interference (mainly 32S16O2H+) in the test.One nanogram each of Co, Ni, Cd and Pb was added to back-extracted sample solution. Leaching of sulfur from the a 0.25 g sample prior to digestion and the recovery through organic layer was evident from an analysis by ICP atomic the whole procedure was measured. When the recovery from emission spectrometry of major element concentrations in the a standard solution (without sample) was corrected, the mean back-extracted sample solution (Table 2).Although part of recovery (n=3) of the elements added to the sample was in the interference might have been cancelled by blank subtracthe range 96–110%, depending on the metal, indicating that tion, APDC extraction was not considered suitable for the ID the present digestion–extraction–back-extraction procedure analysis of Cu. Therefore, sulfur was removed by re-extracting was quantitative.Although the extraction eYciency does not the Cu with HOQ after APDC extraction and back extraction. aVect the accuracy of the isotope dilution analysis, quantitative The 63Cu/65Cu value in Table 1 was the ratio after HOQ extraction is desirable to obtain maximum ion counting, which extraction and it was consistent with the IUPAC value within gives a higher precision to the isotope ratio measurements. ±0.5% deviation. Higher concentrations of Cu and Zn were indicated in a Table 2 also shows that the concentrations of selected preliminary measurement; therefore, the recoveries of these elements, including Na and Si which can give molecular two elements were not evaluated.interference on 63Cu (40Ar23Na+) and 68Zn (40Ar28Si+), are insignificant and no spectroscopic interference was expected in the ID-ICP-MS determination when the selected masses Results of extraction–back-extraction–ICP-MS with external were used for measurement.calibration The naturally variable isotopic composition of Pb did not allow the evaluation of spectroscopic interference by natural An ICP-MS determination of Co, Ni, Cu, Zn, Cd and Pb with abundance analysis, but no interference has been reported in external calibration after APDC extraction–back-extraction the m/z region for Pb. was performed. There were three objectives for this analysis. The analytical results could be used as back-up data for the Determination of Cu, Zn, Cd and Pb by ID-ICP-MS subsequent ID-ICP-MS analysis because the analytical principle was diVerent.Estimation of the suitable amount of stable Results of the ID-ICP-MS determinations of Cu, Zn, Cd and isotope spikes to add for ID-ICP-MS analysis was facilitated. Pb concentrations in the fish otolith CRM are shown in The sample solutions prepared could be used to check for Table 3 along with the results from the back-up method. The spectral interference(s) by natural abundance measurements.ID-ICP-MS results were precise, i.e. relative standard devi- Concentrations of Ni, Cu (m/z=63), Zn, Cd and Pb ations of five independent analyses were 0.4% for Zn, 0.9% were determined to be 0.015±0.002, 0.73±0.02, 0.43±0.02, for Cu, 1.3% for Pb and 5.8% for Cd. The moderate precision 0.0028±0.0002 and 0.026±0.001 mg kg-1 dry wt, respectively. Cobalt was not detected (detection limit, 0.0001 mg kg-1). Table 1 Isotopic ratios in unspiked fish otolith CRM after digestion– Copper determined at m/z=65 gave 0.72±0.02 mg kg-1, extraction–back-extraction (n=3) which was comparable with the value at m/z=63.The agree- Unspiked sample IUPAC14 ment of the Cu values at m/z=63 and 65, despite the spectroscopic interference at 65, was attributable to the compensation 63Cu/65Cu 2.252±0.002a 2.244 of the interference by blank subtraction. Using these heavy 66Zn/68Zn 0.677±0.001 0.674 metal concentrations, the percentage contributions of the mean 114Cd/112Cd 1.185±0.023 1.191 procedural blank to the amount of the element in 1 g of the 208Pb/206Pb 2.147±0.002 — fish otolith CRM were calculated to be as follows: 29% for aDetermined after HOQ extraction–back-extraction of APDC Ni, 0.1% for Cu, 3.6% for Zn, <0.1% for Cd and 0.8% for Pb.extracted–back-extracted sample. The procedural blank levels were acceptable for the accurate determination of the analytes by ID-ICP-MS, except for Ni which was judged to be too large and thus was excluded from Table 2 Major element concentrations in APDC extracted–2 MHNO3 back-extracted sample solution (n=6) the ID-ICP-MS analysis.Based on this result, the amount of stable isotope spike Concentration (ppm) which would be added to 1 g of the CRM was determined as follows: 800 ng 65Cu to make the 63Cu/65Cu value of the Na <0.04 spiked sample 0.5, 200 ng 66Zn (66Zn/68Zn=0.25), 3.6 ng 112Cd Mg <0.04 (114Cd/112Cd=0.2) and 21 ng 206Pb (208Pb/206Pb=0.5).The Si <0.2 P <0.2 target isotope ratios for the ID analysis were selected taking S 148±30 both the theoretical optimum ratio (square root of the sample K <0.2 and spike abundance) and optimum ion counting in ICP-MS Ca <0.04 into consideration.13 J. Anal. At. Spectrom., 1999, 14, 1589–1592 1591Table 3 Heavy metal concentrations in the fish otolith CRM (mg kg-1 of the Cd result could be attributed to the lower concentration dry wt) determined by ID-ICP-MS and by ICP-MS with external involved; Cd in the back-extracted sample solution gave as calibration (back-up method) low as 5000 cps for 112Cd and 800 cps for 114Cd.The results of the back-up method were consistent with the ID-ICP-MS (n=5) External calibration (n=3) ID-ICP-MS values. The consistency of the results obtained Co — <0.0001 from ID-ICP-MS and the back-up method indicated the Ni — 0.015±0.002 validity of both of the results. The trueness of the Zn result Cu 0.742±0.007 0.73±0.02 was further confirmed by the consistent Zn result determined Zn 0.471±0.002 0.43±0.02 by neutron activation analysis (0.46±0.02 mg kg-1, n=5),15 Cd 0.00278±0.00016 0.0028±0.0002 which was reported from a laboratory which participated in Pb 0.0234±0.0003 0.026±0.001 the collaborative analysis for the certification.The trueness of the ID-ICP-MS results can also be assumed from the fact that no spectroscopic interference was expected in the present to be negligible for the four heavy metals (Ni was an exception) analysis.Spectroscopic interference is the primary factor that in comparison with the concentrations in the CRM. The produces biased results in ID-ICP-MS. precise results obtained (0.4–5.8% RSD, n=5) were evaluated Thus the present ID-ICP-MS results are considered accurate in terms of their trueness by comparing them with the values enough to be used as reference values. Reference values were obtained by other analytical methods.Reference values for determined to be 0.74 for Cu, 0.47 for Zn, 0.0028 for Cd and these four heavy metals were determined based on the present 0.023 for Pb (mg kg-1 dry wt). analysis. The fish otolith CRM will be of value in developing a routine analytical method, and for quality assurance of Comparison of the heavy metal concentrations in the otolith existing analytical methods for heavy metals in fish otoliths CRM with reported values for marine aragonites and other marine biogenic aragonites.The otolith CRM is available from NIES, Japan. Address One of the few reliable determinations of heavy metals in fish correspondence to M. Morita. otoliths was performed by Campana et al.9 using ID-ICP-MS. They reported the concentrations of Zn and Pb in a large number of cod otoliths sampled from six sites in Canada. The Acknowledgements mean concentrations for the six sites ranged from 0.31 to The authors thank M. Takano and R.Kumada, Environment 1.75 mg kg-1 for Zn and from 0.009 to 0.0155 mg kg-1 for Pb. Research Center, for the ICP atomic emission spectrometry Heavy metals have been measured in other biogenic marine measurements. Thanks are also due to K. Takata and aragonites. Corals have been frequently analysed for the A. Nakama, NIES, for assistance. reconstruction of marine pollution chronologies. Dodge and Gilbert16 reported Pb concentrations in coral to be 0.087– 0.395 mg kg-1. Shen and Boyle17 reported 0.01–0.1 mg kg-1 References Pb, 0.0004–0.008 mg kg-1 Cd and 0.02–0.07 mg kg-1 Zn. 1 J.M. Kalish, J. Exp. Mar. Biol. Ecol., 1989, 132, 151. Thus the concentrations of heavy metals in corals are similar 2 Y. Sadovy and K. P. Severin, Can. J. Fish Aquat. Sci., 1994, to those in the otolith CRM, and this CRM is useful not only 51, 133. for trace heavy metal analysis of fish otoliths but also for 3 G.R.HoV and L. A. Fuiman, Bull. Mar. Sci., 1995, 54, 578. 4 K. D. Friedland, D.G. Reddin, N. Shimizu, R. H. Haas and A. F. other biogenic marine aragonites such as coral. Youngson, Can. J. Fish Aquat. Sci., 1998, 55, 1158. 5 K. E. Limburg, Mar. Ecol. Prog. Ser., 1995, 119, 25. Conclusions 6 S. R. Thorrold, C. M. Jones and S. E. Campana, Limnol. Oceanogr., 1997, 42, 102. Copper, Zn, Cd and Pb concentrations in a fish otolith CRM 7 J. S. Edmonds, M. J. Moran, N. Caputi and M. Morita, Can. were measured by ID-ICP-MS after oV-line solvent extraction J. Fish Aquat. Sci., 1989, 46, 50. 8 J. S. Edmonds, R. C. J. Lenanton, N. Caputi and M. Morita, Fish. and back extraction to remove the Ca matrix and spectroscopic Res., 1992, 13, 39. interferences. Although solvent extraction is a classical 9 S. E. Campana, J. A. Gagne� and J. W. McLaren, Mar. Ecol. Prog. approach for the separation of heavy metals from sample Ser., 1995, 122, 115. matrices, it is still of practical value even for ICP-MS. It is 10 J. Yoshinaga, A. Nakama, M. Morita and J. S. Edmonds, submitwell known that ICP-MS suVers from matrix-related spectro- ted for publication. scopic and non-spectroscopic problems and separation from 11 D. Beauchemin and S. S. Berman, J. Anal. At. Spectrom., 1989, 61, 1857. the matrix is often essential for the analysis of real samples. 12 G. J. Batterham, N. C. Munksgaard and D. L. Parry, J. Anal. At. Solvent extraction is believed to be time consuming and liable Spectrom., 1997, 12, 1277. to procedural contamination, however, the present APDC 13 J. D. Fassett and P. J. Paulsen, Anal. Chem., 1989, 61, 643A. extraction–back-extraction procede took 2 h for the prep- 14 IUPAC, Pure Appl. Chem., 1998, 70, 217. aration of a dozen samples and thus was not unacceptable. 15 S. Hirai, S. Suzuki and S. Okada, unpublished results. Procedural blanks, mainly arising from reagent blanks in the 16 R. E. Dodge and T. R. Gilbert, Mar. Biol., 1984, 82, 9. 17 G. T. Shen and E. A. Boyle, Chem. Geol., 1988, 67, 47. present analysis, do exist, but the limitation by the blank is dependent on the analyte concentration in the sample and is not always unacceptable. In fact, the blank levels were found Paper 9/04572D 1592 J. Anal. At. Spectrom., 1999, 14, 1589–1592

ISSN:0267-9477    

DOI:10.1039/a904572d

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

Simultaneous multi-element determination of hydride-forming elements by ‘‘in-atomiser trapping’’ electrothermal atomic absorption spectrometry on an iridium-coated graphite tube James Murphy,a Gerhard Schlemmer,b Ian L. Shuttler,b Phil Jonesa and Steve J. Hill*a aDepartment of Environmental Science, University of Plymouth, Drake Circus, Plymouth, Devon, UK PL4 8AA bBodenseewerk Perkin-Elmer GmbH., Postfach 107161, D-88647 U� berlingen, Germany Received 4th June 1999, Accepted 4th August 1999 A simultaneous multi-element approach utilising ‘‘in-atomiser trapping’’ electrothermal atomic absorption spectrometry (ETAAS) for As, Bi, Sb and Se was developed.The approach uses flow injection methodology and hydride formation with sodium tetrahydroborate to sequestrate the hydrides of the elements of interest on an Ir precoated graphite tube. Since the eYciency of the hydride formation depends on the oxidation state of the analyte, an oV-line reduction process was included to ensure that the analyte to be determined was in the most sensitive and favourable oxidation state.Initially five elements, As, Bi, Sb, Se and Te, were considered for simultaneous ‘‘inatomiser trapping’’. The elements were split into two groups reflecting the nature of the reducing agent required by each of the elements. Group A consisted of As, Bi and Sb and used L-cysteine as the reducing agent, whilst Group B consisted of Bi, Se and Te and used concentrated HCl as the reducing agent.However, Te was later removed from Group B due to problems in identifying a set of compromise conditions which enabled all three elements to be determined simultaneously. Bismuth featured in both groups as it did not require a reduction step. Various tube coatings were considered and Ir and Zr were evaluated. Iridium was found to be well suited to this application. The characteristic masses obtained using this method were 177, 91, 107 and 90 pg for As, Bi, Sb and Se, respectively, yielding detection limits (500 ml sample loop) of 0.82, 0.04, 0.26 and 0.29 mg l-1.Precision for analytes at the 5 mg l-1 level was typically better than 3.5% RSD. The method was validated by the analysis of two Certified Reference Materials and good agreement was found with the certified values. thus oVering characteristic concentrations in the low ng l-1 Introduction range;5 (c) a decrease or elimination of the interference eVects Hydride generation atomic absorption spectrometry in both the liquid and gas phases;5 and (d) easy integration (HGAAS) is a useful and sensitive technique for the determi- with flow injection methodology, thereby facilitating increased nation of those elements which form volatile hydrides.use of automation and allowing for a higher sample throughput Traditionally, HGAAS has been performed by passing the than for batch HGAAS methods.4 volatile hydrides, once formed, into an atomiser cell, where As stated above, the ‘‘in-atomiser trapping’’ technique utildecomposition of the hydride takes place.The atomiser cell ises the graphite tube as both the preconcentrating medium can either be a flame or more commonly a heated quartz tube. and the atomisation cell. The preconcentrating surface can be When coupled with flow injection (FI ) technology, the whole either the uncoated tube6 or a tube pre-treated with a precious process, i.e. sample introduction, control of the hydride gener- metal or a carbide-forming element.7,8 As the analytes are ation stage and the final measurement step, may be easily trapped on the graphite tube, this allows the carrier gas (Ar) automated.However, more recently the potential of using a and the hydrogen gas (reaction by-product) to escape before graphite electrothermal atomiser as both the hydride trapping the atomisation step, whereas in a quartz tube atomiser, these cell and the atomisation cell (‘‘in-atomiser trapping’’) has been gases would dilute the analyte during the atomisation stage.reported by a number of workers and recently reviewed by This extra gas production can also alter the hydride pro- Matusiewicz and Sturgeon.1 We have previously reported the duction/transfer rates which in a HG-quartz tube atomiser potential of ‘‘in-atomiser trapping’’ to allow simultaneous method can influence the atomisation signal. The in situ multi-element determination of Bi and Se.2 The work described preconcentration step is expected to eliminate the possible here extends this study to consider both the possibilities and influence of the hydride generation kinetics on the signal shape limitations of the simultaneous determination of five hydride- which in quartz tube atomisers leads to extended peak widths.9 forming elements, i.e.As, Bi, Sb, Se and Te. Performing the atomisation in the graphite tube atomiser The ‘‘in-atomiser trapping’’ technique has a number of means that conventional ETAAS peaks are obtained with peak advantages for the atomisation of hydride-forming elements widths of 1–5 s instead of 10–15 s (typical for quartz tube over the use of conventional heated quartz cells.These advan- atomiser) for a similar mass of analyte,2 and the careful tages can be summarised as: (a) the high absolute sensitivity control of the atomisation parameters, i.e. temperature, time intrinsic to electrothermal atomisation,3 where the character- and heating rates, is possible, leading to improved reproducistic masses are generally lower than for direct transfer ibility of the atomisation process and can reduce interferences.HGAAS with quartz tube atomisers;4 (b) the high relative A key requirement in achieving the simultaneous multielement determination of hydride-forming elements with sensitivity due to in situ preconcentration from larger samples, J. Anal. At. Spectrom., 1999, 14, 1593–1600 1593ETAAS is the selection of the instrumental parameters and and tetravalent Se.Based on a consideration of the chemistries involved the selected elements were divided into groups the hydride generation chemistries. The selection of the instrumental parameters is relatively straightforward, i.e. easily reflecting the optimum oxidation state required for hydride generation. Group A contained As, Sb and Bi, and used L- achieved by taking the lowest trapping temperature and the highest atomisation temperature from the optimum instrumen- cysteine as the reducing agent and Group B contained Bi, Se and Te and employed elevated temperatures with concentrated tal conditions for the group of elements to be determined.However, with respect to the hydride generation chemistries, HCl as the reducing agent. Bismuth was grouped with As and Sb due to it also being a Group Vb element; however, it can these have to be a compromise of the optimum single element conditions and these may show considerable variation. Clearly, also be determined with Se as shown in our earlier study.2 therefore, the sensitivity for a single element being determined will depend on the oxidation state of the analyte in the sample.Experimental To achieve the pre-reduction step required, potassium iodide10 or a mixture of potassium iodide and ascorbic acid11 Instrumentation has been used to reduce Asv to AsIII, but recently, L-cysteine A simultaneous electrothermal atomic absorption spectrometer has found favour12,13 since it is compatible with FI systems (Model SIMAA 6000, Bodenseewerk Perkin-Elmer, and uses reagents at lower concentrations.Other advantages U� berlingen, Germany) with a transversely heated graphite of L-cysteine include more eYcient reduction of As/SbV to the atomiser (THGA) with longitudinal Zeeman-eVect back- trivalent oxidation state, and the stabilising eVect on the ground correction equipped with a Model AS-72 autosampler analyte solutions.12,14 was employed for this study.Standard THGA tubes with For Se and Te, the highest oxidation state is hardly reduced integrated platforms and Perkin-Elmer EDL ‘‘System 2’’ elec- at all by tetrahydroborate,15 owing to the extremely slow trodeless discharge lamps wused. The instrumental param- kinetics of the reduction. Therefore, if total Se and Te conceneters used are shown in Table 1 and the THGA programme trations are required, a pre-reduction step must be used is shown in Table 2.implemented to reduce the hexavalent state to the tetravalent Flow injection (FI ) hydride generation was performed using state. For the reduction of SeVI to SeIV, the use of concentrated a commercial FI system (Perkin-Elmer FIAS 400) equipped hydrochloric acid is considered an eYcient and successful with a Model AS-91 autosampler. The control programme for method, with the chloride ion being the reductant.15,16 the FI system is shown in Table 3.Tygon tubing was used for For an ‘‘ideal’’ multi-element analysis, a simple preall reagent lines. The sample and acid carrier lines (id 1.52 mm) treatment step should be used, which converts all the elements had flow rates of 12 and 6 ml min-1, respectively, while the to their optimum oxidation state for hydride generation. Lreductant line (id 1.14 mm) had a flow rate of 4 ml min-1. Cysteine will reduce the As/SbV to the lower trivalent oxidation The PTFE capillary of the furnace autosampler was discon- state, but it will also reduce SeVI or SeIV to the elemental state, nected from the autosampler arm and replaced by a quartz while concentrated hydrochloric acid will reduce Se/TeVI to capillary (id 1.0 mm) and PTFE transfer line (id 2.0 mm) to the tetravalent state, but it has no eVect on reducing the convey the hydrides from the gas–liquid separator.The quartz As/SbV to the trivalent state.16 The ideal reductant that will capillary was carefully adjusted so as to inject the hydrides meet all the criteria discussed above (i.e.reduce the hydrideover the heated platform of the THGA without touching the forming elements to their most sensitive oxidation state) has platform or the tube walls. The optimum distance between the yet to be identified. However, various diVerent approaches quartz tip and the heated platform surface was 1.0 mm. have been tried to find the ‘‘ideal’’ reducing agent for simul- The graphite tube and platform were pre-treated with Ir taneous multi-element hydride generation determinations.after the new tube had been thermally conditioned. Using a Uggerud and Lund17 used a reducing agent consisting of two standard micropipette, a 50 ml volume of IrCl3 (1000 mg l-1) separate reagents (concentrated HCl and thiourea) to deterstock standard (Inorganic Ventures, Lakewood, NJ, USA) mine As, Bi, Sb, Se and Te by HG-ICP-AES. First, the was injected manually onto the platform and then the THGA concentrated HCl was added to the sample oV-line, to reduce was run using the programme previously described in the the Se and Te to the lower oxidation state and then the literature.8 This complete sequence was performed twice, and thiourea was added again oV-line to reduce any As/Sbv to the then the THGA tube was ready for use and no further coating lower+3 oxidation state. This combination of reducing agents or conditioning was necessary during the lifetime of the tube.works well, but the thiourea also slowly reduces the Se/TeIV The pre-treatment was suYcient to last the lifetime of the to the elemental state; therefore, once the thiourea has been tube, provided that the clean-out temperature was not excessive added, the analysis must be performed within a short period (>2300 °C). It was also possible to prepare batches of tubes of time for precise and reproducible results. Bowman et al.18 in this way and store them under dust-free conditions for took this idea of the two-component reducing agent one step future use.The system was controlled using AA Winlab further when they determined As, Sb and Se in water samples software (Perkin-Elmer Version Beta 2.38). by HG-ICP-MS. The concentrated HCl at an elevated tempera- The combined FI procedure and synchronisation of the ture of 80 °C was still added in an oV-line manner, but the atomiser time/temperature programme was controlled by the thiourea was added on-line to eliminate the time restraint. sequences shown in Table 4.This operation took 2 min and Stroh and Vollkopf19 used a diVerent approach to determine consisted of six individual steps, which started with the sample As, Bi, Hg, Sb, Se and Te in water and sea-water samples at loop being filled with fresh sample, sequestration of the hydride ultra-trace levels by FI-vapour generation-ICP-MS. The on the tube and finally the atomisation and measurement step.sample was split into two sub-samples, with one sample being The atomiser ran for 90 s (which included a 60 s trapping used to determine As and Sb by using a mixed reducing agent time) and the FI stage lasted for 75 s. Before the first replicate (5% m/v potassium iodide–5% m/v ascorbic acid), while the of any new sample, fresh sample was pumped into the sample second sub-sample was used to determine Se and Te using loop expelling any previous sample left in the tubing to waste.concentrated HCl at an elevated temperature as the reducing agent. Bi and Hg were found not to require any reducing agent; Reagents therefore, they could be determined in either sub-sample. The object of the study presented here was to assess the All reagents used were of pro analysi grade (Merck, Darmstadt, Germany) unless otherwise stated. Dilutions were made using performance of a multi-element approach where elements are present in their optimum oxidation states, i.e.trivalent As/Sb de-ionised water (Nanopure, Barnstead, Boston, MA, USA). 1594 J. Anal. At. Spectrom., 1999, 14, 1593–1600Table 1 Instrumental parameters Parameter Arsenic Bismuth Antimony Selenium Wavelength/nm 193.7 223.1 217.6 196.0 Lamp type As EDL Bi EDL Bi EDL Se EDL Lamp current/mA 400 380 380 280 Read time/s 5 5 5 5 Read delay/s 0 0 0 0 Carrier Group A 1% (v/v) hydrochloric acid Carrier Group B 10% (v/v) hydrochloric acid Reductant 0.5% (m/v) sodium tetrahydroborate in 0.5% (m/v) sodium hydroxide Reducing agent Group A 1% (m/v) L-cysteine in 0.1 mol l-1 hydrochloric acid Reducing agent Group B 50% (v/v) hydrochloric acid Signal measurement Integrated absorbance Signal type Background-corrected AA The reductant solution, 0.5% m/v NaBH4 in 0.5% m/v Table 2 Furnace programme NaOH, was prepared by dissolving 5.0 g of sodium hydroxide pellets followed by 5.0 g of sodium tetrahydroborate (95% Temperature/ Ramp time/ Hold time/ Gas flow/ assay, Riedel-de Ha�en, Seelze, Germany) and diluting to Step °C s s Read ml min-1 1000 ml.This solution was prepared daily and stored in a plastic Nalgene bottle. The carrier solution was either 1 or 1 a 1 60 50 2 a 1 20 250 10% v/v HCl depending on the group of analytes being deter- 3 2200 0 5 Yes 0 mined. The solutions were prepared by diluting either 10 or 4 2300 1 3 250 100 ml of concentrated hydrochloric acid (Suprapur, Merck) to 1000 ml. This solution was prepared when needed.aGroup A 300 °C and Group B 250 °C. The working metal standards were made by serial dilutions with either (a) 1% v/v HCl and 1% m/v L-cysteine for As, Bi and Sb or (b) 10% v/v HCl for Bi and Se from the original Table 3 Optimised programme for flow injection systema stock solutions, 1000 mg l-1 as SeVI, 1000 mg l-1 as SbIII, 1000 mg l-1 as BiIII (Merck) and 1000 mg l-1 as AsIII (Fixanal, Time/ Pump No. 1/ Pump No. 2/ s rpm rpm Position Riedel-de Ha�en). Research-grade argon was used for both the FI system and the internal and external gas streams for the Prefill 20 100 0 Fill THGA system. 1 10 100 0 Fill The reducing agent for Group A elements was L-cysteine. 2 5 100 80 Fill A stock solution of 10% L-cysteine in 0.5 mol l-1 HCl was 3 40 0 80 Inject made by adding 10 g of L-cysteine (Biochemistry, Merck) to aSample loop 500 ml; argon gas flow 150 ml min-1. 4.5 ml of concentrated HCl (Suprapur, Merck) and diluting to 100 ml and kept for 1 week.From this solution, enough Lcysteine was added to make the final concentration in the standard or sample 1% m/v. The reducing agent for Group B Table 4 Sequence control programme for ETAAS and FIAS elements was concentrated hydrochloric acid (apur, Merck) and suYcient acid was added to each sample to give Step Operation a final concentration of 50% v/v HCl. Sample solution pumped from the autosampler vessel to fill A Zirconium trapping solution: 0.02 mol l-1 ZrOCl2 8H2O the 500 ml sample loop, excess sample passes to waste.B Furnace Step 1, the tube is pre-heated to 250 or 300 °C, i.e. A 0.02 mol l-1 solution of ZrOCl2 8H2O was prepared by the trapping temperature. In parallel, the flow injection dissolving 0.06445 g of the salt in 10 ml of water. A total of system pumps fresh carrier and reductant solutions into 100 ml was injected onto the platform in two aliquots of 50 ml the manifold, and the argon gas sweeps the gas–liquid followed by the temperature programme shown in Table 5.separator clean. Sample loading is completed. The coating lasted for the lifetime of the tube.20 C The autosampler arm moves the quartz pipette tip from the standby position into the graphite tube. The sample is mixed with the carrier stream followed by the Results and discussion tetrahydroborate solution and argon gas and then passed through a reaction coil. The mixture enters the gas–liquid Blank contamination problem separator where the hydride vapour is swept into the graphite tube by the argon gas flow, and the waste is During the initial experiments, elevated blank values for As pumped out.A 0.45 mm membrane filter is positioned and to a lesser extent for Sb were observed. It was confirmed between the gas–liquid separator and the delivery tube to prevent excessive moisture from entering the graphite tube. Table 5 Furnace programme for graphite tube pre-treatment The tube is maintained at the trapping temperature.D The autosampler arm moves back to its standby position. Temperature/ Ramp time/ Hold time/ Gas flow/ E Pumps on the flow injection system are stopped to reduce Step °C s s Read ml min-1 reagent and sample consumption. F The atomiser time/temperature programme runs to 1 110 10 50 250 completion. A drying step to removes excess hydrogen and 2 130 30 50 250 any water vapour before the atomisation step and then a 3 1200 30 20 250 clean-out step cleans the tube before the next sample is 4 a 1 3 Yes 250 introduced. A maximum temperature of 2300 °C is used to prevent the iridium coating from being removed.aFor Ir 2000 °C and for Zr 2600 °C. J. Anal. At. Spectrom., 1999, 14, 1593–1600 1595experimentally that the contamination was due to the sodium experiments, Bi was also monitored, and it was found that, with prolonged heating times, both the accuracy and reproduc- tetrahydroborate as noted in the literature.20 DiVerent batches of sodium tetrahydroborate were therefore assessed, giving ibility of the results decreased.Based on these findings, it became clear that a single set of each batch tested gave slightly diVerent blank values, and thus aVected the detection limits of the method. The ‘‘cleaning-up’’ conditions that would reduce Te and Se completely, but without losses of either Se or Bi, would be diYcult to identify of the sodium tetrahydroborate was therefore considered. Activated alumina has previously been used to preconcentrate and that these three elements were incompatible in terms of the proposed simultaneous procedure.Consequently, it was oxyanions21 and it was assumed that, since the sodium tetrahydroborate was in alkaline solution, this approach may be decided to remove Te from Group B. Overall, the optimised conditions for the Se reduction step were found to be 70 °C used. Initially, an on-line column approach appeared to be the best option since no special treatment would then be for 120 min in 50% v/v HCl; under these conditions reproducible results were obtained for Bi.needed and the reductant would be cleaned-up immediately prior to use, keeping the method as simple as possible. Comparison between Ir and Zr trapping materials In preliminary experiments, it became clear that although the on-line activated alumina column was eVective in reducing Whilst many diVerent trapping materials have been proposed, the contamination levels, the limited column capacity (approxi- Ir and Zr are the most favoured, because these coatings only mately 30 replicates) was a serious disadvantage. The next need to be applied once during the lifetime of the tube.Both step was to clean the tetrahydroborate oV-line. The contami- of these materials have been applied successfully to single nated tetrahydroborate was passed through an activated alumelement analysis.7,8,20 In this study, a comparison was made ina column and collected in a fluorinated ethylene propylene to assess the suitability of both materials for simultaneous (FEP) bottle prior to the start of analysis.A large column multi-element analysis. The results are shown in Fig. 1. (150×10 mm id) was made from plastic tubing and a plastic Using the Zr-coated tube, it was not possible to find a restrictor was placed in one end of the tube, with a glass wool temperature where As, Bi and Sb could be successfully trapped plug. Approximately 2 g of activated alumina were slurried together [Fig. 1(a)–(c)]. Similar results were found for Bi and with a small volume of water and carefully poured into the Se. The findings are in line with the work of Haug and Liao,20 column. The alumina was allowed to settle and a glass wool who concluded that in terms of the optimum trapping tempera- plug was inserted above the alumina. The column was activated tures, As, Bi and Sb are all mutually exclusive. Thus, it was by passing 1% v/v HCl through it for 5 min followed by concluded that Zr was not an eVective trapping material for distilled water for 5 min and finally the alkaline reductant was this study.Using an Ir coating, As, Bi and Sb could be then passed through the column for a further 5 min or until successfully trapped at 300 °C [Fig. 1(d)] while Se and Bi the waste solution was alkaline. The column was left in the required a trapping temperature of 250 °C. To the best of our alkaline solution when not in use.This procedure removed knowledge, this is the first reported comparison of the applicaover 80% of the As contamination and 30% of the Sb bility of Ir- and Zr-coated graphite tubes to simultaneous contamination. multi-element hydride trapping. This oV-line approach gave similar blank values to the on-line method (As 0.005 s-1, and Sb 0.003 s-1) but the Long-term stability tests useful lifetime of the procedure was greatly increased, as 500 ml of ‘‘contaminated’’ reductant could be cleaned-up in In order to determine the lifetime of the Ir-coated tubes, tests were run overnight to check the stability of the analyte signal 2 h by slowly passing the solution through the column using a pump rate of 5 ml min-1.This volume of reductant was in terms of integrated absorbance and the precision of the signal (RSD %) for ten replicates. A solution of 5 mg l-1 As, suYcient for 160 analytical cycles and a fresh batch of reductant was prepared daily following re-activation of the column Bi and Sb, Group A, and Bi and Se, Group B, was used and all other instrumental parameters were as stated in Tables 1–3.as previously described. The results obtained are shown in Fig. 2. For Group A, initially for all three elements [Fig. 2(a)–(c)], Choice of reducing agents the integrated absorbance was high with good precision, but the mean peak absorbance then dropped to a constant level Although the instrumentation used in this study would facilitate the simultaneous determination of As, Bi, Sb, Se and Te for the next 120+ samples.This is as expected as the instrument will be most sensitive just after the graphite tube has as described above, the analytes were split into two groups, the criteria being those analytes which could be determined been freshly treated with the trapping material. Overall, however, for the first 5 h, i.e. 150 cycles, the results are steady and with the same reducing agent and conditions.Grouping the elements into two distinct sets in this way ensured that the reproducible, but then for As and Sb, but not for Bi, the precision is degraded and the average RSD exceeds 5%. This elements would be in their most favourable states prior to the hydride generation reaction. is thought to be because both As and Sb require L-cysteine as the reducing agent but L-cysteine also acts as a complexing During the initial experiments, Te was included in Group B along with Bi and Se.Tellurium requires the same reducing agent for these two analytes.13 When using a 1% v/v HCl carrier solution, the L-cysteine is not stable for long periods agent as Se, i.e. 50% v/v HCl, but the optimum temperature for reduction is 100 °C for 20 min. This temperature is 20 °C of time, as lower acid carrier concentrations (<0.1 mol l-1 HCl) are recommended with 1% m/v L-cysteine.12 higher than the optimum Se temperature. It is known that Se can be lost from solution as the chloride,16 and hence the It was also found in these studies that Sb was more stable (i.e.gave more reproducible results) in 10% v/v HCl carrier temperature of the reduction vessel should be kept as low as possible in order to eliminate any loss of Se. The results solution than in 1% v/v HCl carrier solution. AsIII could not be determined satisfactorily with a 10% v/v HCl carrier solu- obtained in this study showed that for an oV-line reduction, if the temperature was increased to 100 °C for 3 h to facilitate tion; therefore, to allow the simultaneous determination of AsIII and SbIII, a compromise was reached which used a carrier the reduction of Te, significant losses of Se were observed. The 3 h reduction time was required to produce the best results solution of 1% v/v HCl.These issues, with respect to the HCl carrier concentration and the stability of L-cysteine in the for Te. If the temperature was reduced to 70 °C, optimum for the Se reduction, then the results for Te were irreproducible, solutions, explain the degradation in precision noted after 5 h of operation. It has been observed by Welz and Sucmanova12 indicating incomplete reduction. During the course of these 1596 J.Anal. At. Spectrom., 1999, 14, 1593–1600Fig. 1 Simultaneous signals for As, Bi and Sb (Group A) at the 5 mg l-1 level; atomisation temperature 2200 °C, sample volume 500 ml. (a)–(c) Zr-coated tube with trapping temperature of: (a) 300 °C; (b) 600 °C; (c) 800 °C.(d) Ir-coated tube with a trapping temperature of 300 °C. that the integrated absorbance of an AsIII standard (10 mg l-1) interferent (0.01, 0.1, 1.0, 10, 100 mg l-1) were then added in 1 mol l-1 HCl with no reductant decreased almost linearly and the new response (n=3) was recorded. An interference with time to 50% of its original value within 3–6 h. This was defined as any signal which gave a deviation of ±10% highlights the instability of the AsIII standard (which is oxidised from the standard response.The levels at which such deviations to AsV) which must be correctly stabilised to prevent poor were observed are shown in Table 6. precision. Bi does not appear to be aVected by the acid Of the four analytes, SeIV was the most aVected with two concentration of the L-cysteine and both the precision and the out of six metals (CoII and PbII) giving an interference at the integrated absorbance values remain steady during the full 9 h 0.01 mg l-1 level, i.e.only a 2-fold increase above the original of the experiment. Se concentration. Two other metals, FeIII and NiII, gave an For the Group B elements [Fig. 2(d) and (e)], Bi showed a interference at the 0.1 mg l-1 level, i.e. a 20-fold increase. The slight increase in integrated absorbance over the 9 h period worst interference on AsIII was from four elements at the (i.e. 270 runs) with a precision of approximately 2% for the 0.1 mg l-1 level, FeIII, NiII, PbII and AgI.However, As can be whole run. determined free from interference in the presence of up to Overall, these results indicate that Ir as the trapping reagent 10 mg l-1 CoII and CuII. SbIII was the least aVected analyte, for simultaneous multi-element analysis is reliable, shows good and could tolerate up to 10 mg l-1 CuII, NiII, AgI and PbII precision and is stable for reasonable working periods, i.e. 5 h and 100 mg l-1 FeIII and CoII without interference eVects.The for As and Sb and up to 9 h for Bi and Se. The reduced interference eVects on Bi varied between Groups A and B. In analysis time for As and Sb is due to instability in the chemistry Group A, both FeIII and AgI gave an interference at the over prolonged periods of time rather than degradation of the 0.1 mg l-1 level, while only Ag gave the same level of intertrapping material used. ference in Group B. In Group A, there was no interference from CoII up to 100 mg l-1 although under Group B con- Interference studies ditions, CoII interfered at the 1 mg l-1 level.The results of this brief interference study show that Se is Gas phase interferences are caused by the presence of other the element that is potentially aVected the most by interferents, volatile elements and result from the interferent removing the with four elements giving an interference at the 0.1 mg l-1 hydrogen radicals required by the analyte hydride for atomislevel, i.e.a 2-fold increase in concentration of the interferent ation.22 This process happens in conventional hydride generover the original Se concentration. At the 0.1 mg l-1 inter- ation within the quartz cell,23 but since ‘‘in-atomiser trapping’’ ference level, i.e. a 20-fold increase in interferent concentration utilises higher atomisation temperatures and the atomisation over the analyte concentration, it is possible that only FeIII mechanism is diVerent to that in a quartz tube atomiser, this may interfere with the Group A elements (As and Bi) and Se type of interference may be eliminated.The mutual interference in Group B, but all the other interferent concentrations in the among the hydride-forming elements is much lower than in samples should be below this interferent level. Therefore, when externally heated quartz tube atomisers.24 This interference analysing natural water samples, care must be taken either to should therefore be expected to depend on the total mass of reduce the potential interferences whenever possible, by adding interferent trapped in the furnace rather than the interferent a complexing agent (e.g., EDTA or a greater concentration of concentration in the sample.The literature identifies six metals L-cysteine) or confirm the absence of any potential interferent. (AgI, CoII, CuII, FeIII, NiII and PbII) which may be considered Overall, Group A elements appear to be more resistant to as potential interferences.A 5 mg l-1 multi-element standard the interfering elements than Group B. This is probably due of these elements was analysed (n=3) in the absence of interferent to give a base response. Increasing amounts of the to the presence of the reducing agent (L-cysteine) used in J. Anal. At. Spectrom., 1999, 14, 1593–1600 1597Fig. 2 Long-term stability as indicated by integrated absorbance (s-1) and precision (%RSD, n=10). Concentration for all elements 5 mg l-1, sample volume 500 ml, trapping time 40 s, atomisation temperature 2200 °C.(a)–(c) Group A, trapping temperature 300 °C, (a) As; (b) Bi; (c) Sb; (d) and (e) Group B, trapping temperature 250 °C, (d) Bi; (e) Se. Table 6 Results from interference study. Level (mg l-1) at which the Analysis of Certified Reference Materials interferent produces a deviation of 10% or more from the 5 mg l-1 Two reference materials (RMs) were analysed: High Purity metal standard Standard (HPS) Certified Reference Material No. 490915 AsIII BiIII SbIII BiIII SeIV Trace Metals in Drinking Water and National Institute of Standards and Technology (NIST) Standard Reference CuII 10 1 10 10 1 Material 1643c Trace Elements inWater. As the concentrations FeIII 0.1 0.1 >100 1 0.1 of the RMs were outside the calibration range for all the CoII 10 >100 >100 1 0.01 analytes of interest, appropriate dilutions were made during NiII 0.1 10 10 1 0.1 the oV-line digestion step.AgI 0.1 0.1 10 0.1 0.1 PbII 0.1 1 10 100 0.01 For Group A, two separate dilutions had to be made. The As required a 20-fold dilution to bring it within range, while Sb and Bi only required a 2.5-fold dilution. During the dilution step, the appropriate amount of 10% m/v L-cysteine was added to each calibrated flask. After the 30 min reaction period, the analysis Group A. This reducing agent also has a releasing eVect on was performed. For Group B, a 2-fold dilution was used.The the potential liquid phase interferences.9 Many workers have solution was heated for 120 min at 70 °C, and, once cool, used used the dual nature of L-cysteine as a reducing agent and to for analysis. All solutions were analysed under the conditions minimise or eliminate the liquid phase interferences in HGAAS.13,24 stated in Tables 1–3, and the results are shown in Table 7. 1598 J. Anal. At. Spectrom., 1999, 14, 1593–1600Table 7 Results obtained for Certified Reference Materials HPS CRM SRM 1646c Certified value/mg l-1 Experimental/mg l-1 Certified value/mg l-1 Experimental/mg l-1 As 80.00±0.40 80.8±2.1 82.10±1.2 80.9±0.5 Bi 10.00±0.05 9.7±0.4 12 12.6±0.2 Sb 10.00±0.05 11.1±1.2 NDa ND Bi 10.00±0.05 9.2±0.0 12 11.7±0.0 Se 10.00±0.05 9.7±0.0 12.7±0.7 12.1±0.0 aND=Not determined.The HPS RM was certified for all four elements of interest ETAAS is possible using conditions derived from single element analysis. in this study, and good agreement was achieved between the experimental and the certified values as shown in Table 7. The Advantages of simultaneous multi-element analysis over single element analysis include speed with respect to the overall NIST Trace Elements in Water SRM was certified for As and Se, with a recommended but not certified value for Bi and no analysis time (the time reduction being proportional to the number of elements run together), and reduced consumption value was provided for Sb. Again, good agreement between the found and certified values was achieved, with all results of reagents per sample.The disadvantage is reflected by the higher detection limits obtained for multi-element analysis falling within the certified range of two standard deviations (Table 7). when compared with single element analysis; this is because in the latter the conditions can be fully optimised with respect to the instrumental and chemical parameters to attain the best Analytical figures of merit possible detection limits for each element. However, while a The characteristic mass (as determined for integrated flow injection approach with 500 ml sample volumes was absorbance) and instrument and method detection limits for utilised in this study, improvements in the detection limit can the method are shown in Table 8.The instrumental detection be obtained by using larger sample volumes by either increasing limit (pg) is given as equal to the mean of the blank signal the size of the sample loop, by using multiple trapping steps (Yblank) for ten replicates plus three times the standard devi- with a small sample loop (as the trapping step is independent ation of the blank (sblank) for ten replicates (IDL= of the atomisation step), or by applying continuous-flow Yblank+3×sblank).The method detection limit (MDL) is based sampling.27 on the instrument detection limit but takes into account the The optimised reducing conditions used for Group B line of regression from the calibration graph such that MDL elements in this study were determined by the need for Se to (mg l-1)=(IDL-intercept)/gradient.be in the tetravalent state for the hydrides to be formed, but The detection limit for As is dependent on the background for the Group A elements, e.g. As and Sb, the hydrides can levels from the sodium tetrahydroborate. Consequently, the be formed from either the trivalent or the pentavalent state.detection limit for As could be improved further if this source This paper describes the determination of hydride-forming of contamination could be eliminated. elements in the lowest oxidation state by using L-cysteine as When the detection limits for this simultaneous multi- the reducing agent. The use of L-cysteine also appears to element method are compared against those of single element reduce the interference eVects on the analytes of interest. methods,7,20,24,25 the single element method detection limits The ‘‘in-atomiser trapping’’ technique reported here can are lower by a factor of between 2 and 6 depending on the easily be expanded from single element analysis to multielement.This is to be expected as the single element methods element analysis provided that an Ir-coated tube is used. The use fully optimised parameters for each element, while the analysis may be set up and left to operate unaided with only simultaneous multi-element method uses a set of compromise occasional re-calibration for up to 5 h with good precision conditions based on the suite of elements being determined.and accuracy. Iridium was the only coating material found to However, when the absolute detection limits of the simul- be applicable to this multi-element procedure, as a universal taneous multi-element method (As 107 pg, Se 73 pg) are trapping temperature could not be found for a Zr-coated tube. compared with those of a previously published simultaneous The conditions developed were shown to produce acceptable method26 (As 92 pg, Se 120 pg), the absolute detection limits results for the analysis of two reference materials.are very similar. Acknowledgements Conclusion The authors thank Professor Dimiter Tsalev (University of The simultaneous multi-element determination of combi- Sofia, Bulgaria), Dr. Michael Sperling and Miss Michaela nations of hydride-forming elements by ‘‘in-atomiser trapping’’ Feuerstein (Bodenseewerk Perkin-Elmer GmbH.) for their helpful discussions and advice.Table 8 Analytical figures of merit References Group A Group B 1 H. Matusiewicz and R. E. Sturgeon, Spectrochim. Acta, Part B, 1996, 51, 377. AsIII BiIII SbIII BiIII SeIV 2 J. Murphy, G. Schlemmer, I. L. Shuttler, P. Jones and S. J. Hill, Anal. Commun., 1997, 34, 359. Characteristic mass/pg 175 136 107 91 90 3 H. O. Haug and Y. Liuo, J. Anal. At. Spectrom., 1995, 10, 1069. Instrumental detection 107 28 27 43 73 4 D. L. Tsalev, A. D’Ulivo, L. Lampugnani, M. Di Marco and limit/pg R. Zamboni, J. Anal. At. Spectrom., 1995, 10, 1003. Method detection 0.82 0.04 0.26 0.12 0.29 5 J. Dedina and D. L. Tsalev, Hydride Generation Atomic Absorption limit/mg l-1 Spectrometry,Wiley, Chichester, 1995. J. Anal. At. Spectrom., 1999, 14, 1593–1600 15996 R. E. Sturgeon, S. N. Willie, G. I. Sproule and S. S. Berman, 19 A. Stroh and U. Vollkopf, J. Anal. At. Spectrom., 1993, 8, 35. J. Anal. At. Spectrom., 1987, 2, 719. 20 H. O. Haug and Y-P. Liao, Fresenius’ J. Anal. Chem., 1996, 356, 7 X.-p. Yan and Z.-m. Ni, J. Anal. At. Spectrom., 1991, 6, 483. 435. 8 I. L. Shuttler, M. Feuerstein and G. Schlemmer, J. Anal. At. 21 M. Sperling, S. Xu and B.Welz, Anal. Chem., 1992, 64, 3101. Spectrom., 1992, 7, 1299. 22 P. Barth, V. Krivan and R. Hausbeck, Anal. Chim. Acta, 1992, 9 X-p. Yan and Z-m. Ni, Anal. Chim. Acta, 1994, 291, 89. 263, 111. 10 J. F. Tyson, S. G. OZey, N. J. Seare, H. A. B. Kibble and 23 B. Welz and P. Strauss, Spectrochim. Acta, Part B, 1993, 48, 951. C. Fellows, J. Anal. At. Spectrom., 1992, 7, 315. 24 Y. An, S. N. Willie and R. E. Sturgeon, Spectrochim. Acta, Part 11 J. Dedina and B. Welz, J. Anal. At. Spectrom., 1992, 7, 307. B, 1992, 47, 1403. 12 B. Welz and M. Sucmanova, Analyst, 1993, 118, 1417. 25 Z.-m. Ni, B. He and H.-b. Han, J. Anal. At. Spectrom., 1993, 13 D. L. Tsalev, A. D’Ulivo, L. Lampugnani, G. Pellegrini and 8, 995. R. Zamboni, J. Anal. At. Spectrom., 1996, 11, 989. 26 S. Garbos, M. Walcerz, E. Bulska and A. Hulanicki, Spectrochim. 14 B. Welz and M. Sucmanova, Analyst, 1993, 118, 1425. Acta, Part B, 1995, 50, 1669. 15 R. Bye and W. Lund, Fresenius’ Z. Anal. Chem., 1988, 332, 242. 27 H.-W. Sinemus, J. Kleiner, H.-H. Stabel and B. Radziuk, J. Anal. 16 R. Bye, Talanta, 1990, 37, 1029. At. Spectrom., 1992, 7, 433. 17 H. Uggerud and W. Lund, J. Anal. At. Spectrom., 1995, 10, 405. 18 J. Bowman, B. Fairman and Catterick, J. Anal. At. Spectrom., 1997, 12, 313. Paper 9/04468J 1600 J. Anal. At. Spectrom., 1999, 14, 1593–1600

ISSN:0267-9477    

DOI:10.1039/a904468j

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

Tungsten–rhodium permanent chemical modifier for lead determination in digests of biological materials and sediments by electrothermal atomic absorption spectrometry E� der C. Lima,* Fernando Barbosa Jr., Francisco J. Krug and Uelinton Guaita Centro de Energia Nuclear na Agricultura, Universidade de Sa�o Paulo, Postal Box 96, 13400-970, Piracicaba-SP. Brazil. E-mail: ederlima@cena.usp.br Received 26th April 1999, Accepted 20th July 1999 A tungsten carbide–rhodium coating on the integrated platform of a transversely heated graphite atomizer was used as a permanent chemical modifier for the determination of Pb in digests of biological materials and sediments by electrothermal atomic absorption spectrometry.Coating with 250 mgW+200 mg Rh was as eYcient as a Pd+Mg(NO3)2 conventional modifier for obtaining good Pb recoveries (95.2–102.3%). TheW–Rh permanent modifier remained stable for approximately 350 firings when 20 ml of digested sample were delivered into the atomizer.In addition, the permanent modifier increased the tube lifetime by 50–115% with respect to untreated integrated platforms. Also, there was less degradation of sensitivity during the atomizer lifetime when compared with the conventional modifier, resulting in a decreased need for re-calibration during routine analysis. The W–Rh permanent modifier withstood acid concentrations up to 5.0% v/v HNO3 without changes in the coating lifetime as well as in the analytical signal.The detection limit, based on integrated absorbance, was 15.5 and 124 ng g-1 Pb for biological materials and sediments, respectively. The RSDs after 1140 and 1250 consecutive measurements of 10 ml of digested Plankton reference material and 10 ml of River Sediment were, respectively, 3.6 and 3.3%. Results for the determination of Pb in the samples were in agreement with those obtained with digested solutions by using Pd+Mg(NO3)2, since no statistical diVerences were found after applying a paired t-test at the 99% confidence level.for Cd determination in fish by slurry-sampling.5 It was Introduction demonstrated that a similar selectivity to conventional modi- With the introduction of the concept of the stabilised tempera- fiers for environmental and biological samples was attained ture platform furnace (STPF), the use of modifiers has become and that the total tube lifetime was increased to 1700 analytical an essential part of electrothermal atomic absorption spec- firings.4 In addition, the detection limit for Cd determination trometry (ETAAS).1 Since first proposed by Ediger,2 the main in fish slurries was improved 8-fold when compared with a purpose of chemical modifiers is either to stabilise the analyte 15 mg Pd+9 mg Mg(NO3)2 modifier.5 thermally or increase the volatility of the matrix.Hence, the Permanent modifiers present attractive advantages such as bulk of the matrix may be removed, by volatilization or (i) simpler and faster heating programmes for ETAAS determidecomposition, during the pyrolysis step, prior to atomization nation, (ii) elimination of volatile impurities of the modifier of the analyte.during its coating process on a graphite surface, which allows Chemical modifiers are generally used by introducing an a decrease in detection limits, minimising the analytical costs aliquot of the modifier solution into the atomizer together by increasing tube lifetime, and (iii) improvement in hydride with the sample solution. Alternatively, the modifier is intro- trapping.4–7 However, some potential drawbacks of permanent duced first onto the graphite surface, followed by drying and modifiers, such as multiple peaks, overstabilisation of some pyrolysis, and then the sample solution is delivered into the analytes, and limitation of the maximum applicable temperaatomizer.Both procedures are repeated at each firing and have ture in the heating programme in order to avoid elimination their advantages and limitations.The former procedure of the permanent modifier from the graphite surface, were requires high-purity chemical modifier reagents in order to also reported.4–7 avoid high blank values. In the latter procedure, the modifier Although permanent chemical modifiers provide various is reduced over the graphite surface, often allowing better advantages, they have not been widely employed for direct analyte stabilisation.3 Additionally, this procedure eliminates introduction of liquid and slurry samples into the atomizer volatile impurities in the modifier solution during the pre- for ETAAS,4–13 as well as for electrothermal vaporization treatment cycle, allowing a decrease in the detection limits.inductively coupled plasma mass spectrometry (ETV-ICP-MS) On the other hand, the total heating programme time is greatly determinations.14 Also, they have scarcely been applied to the increased, becoming unsuitable when analysis of a large analysis of real samples.4,5,9,10,12,14 In this context, there is a number of samples is required.These drawbacks of using a necessity for a more extensive investigation of permanent chemical modifier might be solved if it were present in the chemical modifiers applied to the analysis of samples in tube as a ‘permanent chemical modifier’.4 ETAAS and ETV-ICP-MS. Recently, a tungsten carbide–rhodium coating on the inte- The aim of this work was to propose the use of an easily grated platform of a transversely heated graphite atomizer prepared W–Rh permanent chemical modifier, coated on (THGA) was successfully employed as a permanent chemical the integrated platform of a THGA, for the determination of Pb in digests of biological materials and sediments.modifier for the determination of Cd, Pb and Se in water4 and J. Anal. At. Spectrom., 1999, 14, 1601–1605 1601Improvement in tube lifetime, decrease of blank values, shorter A 1000 mg l-1 Pb stock solution was prepared from Pb(NO3)2 (Johnson Matthey, Royston, Hertfordshire, UK) heating programmes and simplification of the previously described ETAAS procedures that employ conventional by dissolving 0.7992 g of the reagent in 0.5 l of 1% v/v HNO3. Analytical reference solutions were prepared in 1.0% v/v HNO3 modifiers were the main goals.in the range 1.50–35.00 ng ml-1 Pb by suitable serial dilution of the stock solution.The calibration was checked periodically Experimental after every 24 measurements with 20.00 ng ml-1 Pb solution. Apparatus Samples A Perkin-Elmer 4100ZL atomic absorption spectrometer (U� berlingen, Germany) with a longitudinal Zeeman-eVect The following reference materials were used for checking the accuracy of the proposed method: Coppepod Homogenate background correction system, furnished with end-capped THGA tubes (Part No. B300–0653), was employed. THGA (MA-A-1), Fish Flesh Homogenate (MA-A-2), Lake Sediment (SL-1), and Marine Sediment (SD-M-2/TM) from tubes with integrated platforms were used, either without any previous treatment (referred to as the pyrolytic carbon plat- the International Atomic Energy Agency (IAEA, Vienna, Austria); Bovine Liver (SRM 1577a), and BuValo River form), or after pre-treatment first with W and then with Rh (referred to as the W–Rh treated platform).Sediment (SRM 2704) from the National Institute of Standards and Technology (NIST, Gaithersburg, USA); Pig All measurements were based on integrated absorbance and performed at 283.3 nm (slit 0.7 nm) by using a Perkin-Elmer Kidney (CRM 186), Cod Muscle (CRM 422), Rye Grass (CRM 281), Plankton (CRM 414), Brown Bread (CRM 191), EDLII electrodeless discharge lamp (EDL) system.Working solution aliquots of 10–50 ml were taken from and River Sediment (CRM 320) from the Community Bureau of Reference (BCR, Brussels, Belgium); and Chlorella (NIES poly(propylene) cups and delivered into the tube by means of an AS-71 autosampler from the same manufacturer. 3), and Pond Sediment (NIES 2) from the National Institute for Environmental Studies (NIES, Ibaraki, Japan). The heating programmes for Pb determination using pyrolytic carbon platforms with the conventional modifier and the Biological materials and sediments were digested in a MD microwave oven (Graz, Austria). The samples were W–Rh permanent modifier are shown in Table 1.A preheating of the tube at 100 °C prior to sample delivery and a decomposed in digestion vessels with a suitable mixture of acids, following the manufacturer’s procedure for each sample higher drying temperature were employed in order to reduce the total heating programme time. Although this is not usually type.16 The volume was made up with water (1.0% v/v HNO3 final acidity) and the resulting solution was analysed by recommended by the manufacturer,15 no significant deterioration in the analytical performance for Pb determination was ETAAS by delivering 20.0 ml of the digest into the atomizer.The heating programme of Table 1 was used. When the observed under the chosen conditions (1.0% v/v HNO3, 20.0 ml sample solution, end-capped THGA tubes). conventional modifier was employed, after delivering the sample solution, 10.0 ml of chemical modifier were subsequently All measurements were made with at least three replicates. Argon (AGA, Campinas, Brazil ) was used as protective gas delivered into the atomizer, and the heating programme of Table 1 was also employed. throughout.Materials, reagents and solutions Results and discussion High-purity de-ionized water (resistivity 18.2MV cm) obtained The Pb characteristic mass employing a standard THGA and from a Milli-Q water purification system (Millipore, Bedford, Pd+Mg(NO3)2 chemical modifier is 30 pg.15 End-capped MA, USA) was used throughout. Analytical-reagent grade THGA tubes provide an increase of 50% on the analyte HNO3 and HCl (Merck, Rio de Janeiro, Brazil ) were distilled integrated absorbance when compared with a standard THGA, in quartz sub-boiling stills (Ku� rner, Rosenheim, Germany).as previously reported.17 Lead contents present in digests of All solutions were stored in high-density poly(propylene) the materials chosen in this work range from 1.7 to 40 ng ml-1 bottles. Plastic bottles, autosampler cups and glassware mate- in biological materials (sample volume to sample mass ratio rials were cleaned by soaking in 20% v/v HNO3 for 24 h and kept at 100 ml g-1; samples were decomposed and diluted by rinsing five times with Milli-Q water and were dried and stored a factor of 100 ml g-1, which means 0.01 g of sample per in a Class-100 laminar flow hood. All operations were per- millilitre of final solution) and from 20 to 40 ng ml-1 in formed in a clean bench.sediments (sample volume to sample mass ratio kept at The W–Rh permanent chemical modifier as well as its 5000 ml g-1).In order to attain measurable analytical signals, solutions were prepared as described elsewhere.4 end-capped THGA tubes were used throughout. The conventional chemical modifier (added at each firing), used for analyte stabilisation with untreated pyrolytic Chemical modifier carbon platforms, was 0.05% m/v Pd+0.03% m/v Mg(NO3)2, which was prepared from 10.0 g l-1 Pd (Suprapur, Merck, Lead is a volatile element that is lost from the graphite atomizer at temperatures higher than 500 °C in the absence of Darmstadt, Germany) and 10.0 g l-1 Mg(NO3)2 solution (Suprapur, Merck).a chemical modifier.18 Use of a chemical modifier is required Table 1 Heating programmes for the determination of Pb in digests of biological materials and dissolved sediments Step Temperature/°C Ramp/s Hold/s Ar flow rate/ml min-1 1 150 2 20a, 30b 250 2 200 5 10a, 20b 250 3 950a, 1000b 10 25 250 4 1700a, 1800b 0 5 0 5 2200 1 4 250 Injection temperature 100 °C a250 mg W+200 mg Rh permanent modifier.b5 mg Pd+3 mg Mg(NO3)2. 1602 J. Anal. At. Spectrom., 1999, 14, 1601–1605Table 2 Recoveries (%) of Pb in digests of biological materials and to increase the analyte thermal stability, thereby decreasing sediments. The value given is the mean±standard deviation (n=5). the matrix eVects and the background signal, and allowing its Final acidity 1.0% v/v HNO3; Pd+Mg: 5 mg Pd+3 mg Mg(NO3 )2; determination in real samples. W–Rh: 250 mg W+200 mg Rh (permanent modifier).The heating In this work, W–Rh coatings were evaluated as a permanent programmes of Table 1 were employed chemical modifier for the stabilisation of Pb in digests of Sample/modifier Recovery (%) biological materials and sediments. For comparison purposes, the recommended mixture of Pd+Mg(NO3)219 was employed Pb spike/pg as a conventional chemical modifier. Several coatings of the graphite platform were performed Chorella, NIES 3: 100 200 300 with W–Rh4 for Pb determination in digests of biological Pd+Mg 94.8±2.9 97.3±2.5 98.7±3.5 materials (digest of Rye Grass—CRM 281—sample volume W–Rh 95.2±2.1 100.1±1.1 102.1±1.5 Brown Bread, CRM 191: to sample mass ratio 100 ml g-1; Pb content 476±22 pg per Pd+Mg 94.3±3.0 95.9±3.5 97.7±4.2 injection; see Fig. 1). Better analytical signals and recoveries W–Rh 95.4±2.1 96.2±2.2 99.5±3.0 for digested samples were obtained with the following Pig Kidney, CRM 186: sequence: 250 mg W+200 mg Rh (100%) >250 mg W+120 mg Pd+Mg 104.2±2.5 99.8±3.0 103.5±2.7 Rh (95%) >250 mg W+80 mg Rh (89%) >250 mg W+ W–Rh 102.3±1.2 100.9±1.7 99.2±2.0 40 mg Rh (86%) #250 mg W+250 mg Rh (85%) >250 mg Coppepod Homogenate, MA-A-1: Pd+Mg 95.6±3.8 94.2±4.1 102.1±3.7 W+300 mg Rh (79%) >250 mg W (75%) (Fig. 1). Decreases W–Rh 97.2±2.1 101.2±1.9 98.8±1.8 in sensitivity for Rh masses higher than 200 mg were observed.Rye Grass, CRM 281: This depressive eVect with increasing amount of noble metal Pd+Mg 93.7±3.5 94.2±4.0 96.5±4.5 chemical modifier has been observed previously.4,6 The maxi- W–Rh 96.2±2.8 97.2±3.0 101.9±2.5 mum attainable pyrolysis temperature for all the permanent Pond Sediment, NIES 2: modifiers tested was 950 °C, independent of the analytical Pd+Mg 92.5±3.5 98.2±2.9 103.2±2.5 W–Rh 94.5±2.5 99.5±2.1 98.2±1.9 signal obtained. For the permanent modifiers that did not Marine Sediment, SD-M-2/TM: achieve a quantitative recovery (normalised absorbance= Pd+Mg 95.2±5.0 92.4±4.8 93.5±4.5 1.00), analyte losses probably occur during the atomization W–Rh 96.2±3.0 95.8±2.8 94.6±3.5 stage of the heating programme, since there is no dependence BuValo River Sediment, SRM 2704: on pyrolysis temperature, as previously observed.4 From these Pd+Mg 94.5±3.8 98.2±3.2 104.2±3.1 results, it can be inferred that Rh plays an important role in W–Rh 102.2±2.5 100.7±2.7 100.4±2.8 Pb thermal stabilisation in sample digests.A similar performance of the permanent chemical modifiers, related to analytical sensitivities, was also observed for sedi- of the robustness of the proposed method. In addition, there ment digests, with the maximum pyrolysis temperature not was a linear relationship between absorbance and increasing exceeding 950 °C and an atomization temperature of 1700 °C. amount of Pb in the digests. For the remainder of this work, a 250 mg W+200 mg Rh With the 250 mg W+200 mg Rh permanent modifier, slightly permanent modifier and a 5 mg Pd+3 mg Mg(NO3)2 conven- better recoveries and lower standard deviations were achieved, tional modifier were chosen.as compared with the conventional modifier. The graphite coating process probably produces a very thin layer of the Recovery modifier on the graphite surface. This may allow faster, and more eYcient, release of the analyte compared with the An extensive interference study, with isolated inorganic species situation where it has to migrate out of relatively large Pd within a large concentration range, on Pb atomization droplets.3 employing the 250 mg W+200 mg Rh permanent modifier has been reported previously.4 In order to ascertain whether there Analytical characteristics was any interference from the sample matrix on Pb atomization, under the established conditions describeabove, recov- A linear range up to 35.0 ng ml-1 Pb was obtained.The Pb ery experiments (Table 2) were carried out by spiking known characteristic masses for the 250 mg W+200 mg Rh permanent Pb amounts to the digested samples. modifier, and for 5 mg Pd+3 mg Mg(NO3 )2, based on inte- No significant interference from the digests of biological grated absorbance, were 19±1 and 20±2 pg, respectively materials and sediments on Pb atomization was observed for (uncertainty based on ten average results obtained on diVerent the chemical modifiers employed, which is a good indication days).The experiments were performed in a 1.0% v/v HNO3 medium, although concentrations as high as 5.0% v/v could be tolerated by the permanent modifier without significant variations in the tube lifetime, as observed previously.4,5 Detection limits, calculated from 20 consecutive measurements of the blank solution (1.0% v/v HNO3) according to IUPAC,20 for the samples employing the 250 mg W+200 mg Rh permanent modifier were 15.5 ng g-1 (biological materials) and 124 ng g-1 (sediments), and for 5 mg Pd+3 mg Mg(NO3)2 25.0 ng g-1 (biological materials) and 200 ng g-1 (sediments).The poorer detection limits obtained with 5 mg Pd+3 mg Mg(NO3)2 may be attributed to high blanks in the Pd solution. As described previously,4 W–Rh coatings increase tube lifetimes when compared with untreated pyrolytic carbon platforms for typical 20 ml direct sample solution introduction. It is important to point out that each platform treatment Fig. 1 Lead normalised absorbance of digested Rye Grass CRM 281; lasted for about 300–350 analytical firings when working with sample volume to sample mass ratio 100 ml g-1 (Pb content digests of biological materials and sediments [Fig. 2(A) and 476±22 pg per injection) for diVerent permanent modifiers. Heating programmes of Table 1 were employed. (B), respectively]. The used treated platform could be recon- J. Anal. At. Spectrom., 1999, 14, 1601–1605 1603day of 8–10 h of instrument operation (i.e., 300–350 firings).It is important to mention that significant variations in the Pb peak profile were not observed during the several coatings that were applied during the total tube lifetime. Long-term stability curves for a pyrolytic carbon platform with Pd+Mg(NO3)2 and a W–Rh treated platform for Pb determination in digests of Plankton, CRM 414 (BCR), and digests of River Sediment, CRM 320 (BCR), are presented in Fig. 2. For the biological material, and with the permanent modifier, the tube lifetime was 95% longer than that of the untreated platforms with use of 5 mg Pd+3 mg Mg(NO3 )2, when reconditioning platform treatments were performed during four working days.For the dissolved sediment, the increase of tube lifetime with the permanent modifier was 112% in relation to 5 mg Pd+3 mg Mg(NO3)2, for four reconditioning platform treatments. Variations of sensitivity are lower for the permanent modifier (digest of biological material, RSD 3.6%, n=1140; dissolved sediment, RSD 3.3%, n=1250) as compared with untreated platforms with 5 mg Pd+3 mg Mg(NO3)2 (digest of biological material, RSD 7.3%, n=585; dissolved sediment, RSD 7.1%, n=590).Hence, there is less need for re-calibration during the routine analysis of large numbers of samples, compared with the situation when Pd+Mg(NO3)2 is employed. Analysis of samples The accuracy of the proposed method employing 250 mg W+200 mg Rh as a permanent modifier and also using the 5 mg Pd+3 mg Mg(NO3 )2 conventional modifier was assessed with nine reference materials of biological samples Fig. 2 Long-term stability curves for untreated and W–Rh treated and five reference materials of sediments (Table 3). Results platforms. Each point represents an average of ten measurements after injection of 10 ml of: (A) digest of Plankton, CRM 414 (Pb content for all modifiers employed were in agreement with the certified 794±38 pg per injection); (B) dissolved River Sediment, CRM 320 values (Table 3).After applying a t-test between each pair of (Pb content 705±27 pg per injection). Arrows indicate a new coating data related to diVerent modifiers, no diVerences were found withWfollowed by Rh.+, 250 mgW+200 mg Rh permanent modifier; within each pair of results at the 1% probability level, which #, 5 mg Pd+3 mg Mg(NO3)2. is another indication of the accuracy of the proposed method. The performance and accuracy of the method employing W–Rh platform treatment with diVerent coatings and tubes ditioned by performing 4–5 firings at 2400 °C for 5 s to eVect cleaning, and subsequently carrying out the treatment pro- for Pb determination in the chosen samples are better than that obtained with conventional modifiers.It should be pointed cedure described elsewhere.4 The new re-coated platform again lasted for 300–350 firings, with sensitivity not significantly out that the standard deviations of results of all samples using the 250 mg W+200 mg Rh permanent modifier were always diVerent (<2%) from that of the first coating.The total time for treatment of the graphite surface with W–Rh is about lower than those related to the conventional modifier. The lower blank values and longer signal stability as a function of 35–40 min, and this procedure could be carried out during the required warm-up of the EDLs, without increasing the total time when the proposed permanent modifier is employed account for this achievement.analysis time. A coating would typically last for a working Table 3 Lead determination in biological materials and sediments by ETAAS (n=3). Final acidity 1.0% v/v HNO3. Standard deviations are given in parentheses. Permanent modifier: 250 mg W+200 mg Rh; conventional modifier: 5 mg Pd+3 mg Mg(NO3)2 Sample Pb/mg g-1 Certified value Permanent modifier Conventional modifier Coppepod Homogenate, MA-A-1 2.10±0.30 2.11 (0.09) 2.07 (0.15) Fish Flesh Homogenate, MA-A-2 0.58±0.07 0.56 (0.02) 0.59 (0.04) Pig Kidney, CRM 186 0.306±0.011 0.310 (0.03) 0.312 (0.05) Plankton, CRM 414 3.97±0.19 3.98 (0.05) 3.95 (0.10) Cod Muscle, CRM 422 0.085±0.015 0.082 (0.006) 0.95 (0.011) Rye Grass, CRM 281 2.38±0.11 2.37 (0.08) 2.36 (0.11) Brown Bread, CRM 191 0.187±0.014 0.180 (0.006) 0.194 (0.009) Bovine Liver, SRM 1577a 0.135±0.015 0.139 (0.008) 0.142 (0.010) Chlorella, NIES 3a 0.60 0.58 (0.01) 0.59 (0.03) BuValo River Sediment, SRM 2704 161±17 162 (4) 163 (6) Lake Sediment, SL-1 37.7±7.4 37.3 (3.2) 36.8 (5.5) Marine Sediment, SD-M-2/TMb 22.8 23.1 (1.8) 20.7 (2.0) River Sediment, CRM 320 42.3±1.6 42.9 (0.5) 41.2 (0.7) Pond Sediment, NIES 2 105±6 102.8 (2.0) 103.8 (2.5) aNot a certified value.bConfidence interval 20.1–25.6 mg g-1. 1604 J. Anal. At. Spectrom., 1999, 14, 1601–16053 H. Qiao and K. W. Jackson, Spectrochim. Acta, Part B, 1991, Conclusion 46, 1841. 4 E. C.Lima, F. J. Krug and K. W. Jackson, Spectrochim. Acta, In general, treatment of graphite with W followed by Rh Part B, 1998, 53, 1791. presented a better analytical performance relative to a conven- 5 E. C. Lima, F. J. Krug, A. T. Ferreira and F. Barbosa, Jr., J. Anal. tional chemical modifier for Pb determination in biological At. Spectrom., 1999, 14, 269. materials and sediments. The detection limit obtained with the 6 D. L. Tsalev, A. D’Ulivo, L. Lampugnani, M. D. Marco and permanent modifier was 1.61 times better than that obtained R.Zamboni, J. Anal. At. Spectrom., 1995, 10, 1003. 7 C. J. Rademeyer, B. Radziuk, N. Romanova, N. P. Skaugset, with Pd+Mg(NO3)2 mixtures. The tube lifetime was increased A. Skogstad and Y. Thomassen, J. Anal. At. Spectrom., 1995, by 50–115% in relation to a pyrolytic carbon platform, leading 10, 739. to a marked decrease in the variable analytical costs.21 The 8 E. Bulska and W. Jedral, J. Anal. At. Spectrom., 1995, 10, 49.better long-term stability achieved with the 250 mg W+200 mg 9 E. Bulska, W. Kandler and A. Hulanicki, Spectrochim. Acta, Part Rh permanent modifier (digest of biological material, RSD B, 1996, 51, 1263. 3.6%, n=1140; dissolved sediment, RSD 3.3%, n=1250) when 10 C. J. Rademeyer, B. Radziuk, N. Romanova, Y. Thomassen and P. Tittarelli, J. Anal. At. Spectrom., 1997, 12, 81. compared with untreated platforms with 5 mg Pd+3 mg 11 E. Bulska and K. Pyrynska, Spectrochim. Acta, Part B, 1997, Mg(NO3)2 (digest of biological material, RSD 7.3%, n=585; 52, 1283.dissolved sediment, RSD 7.1%, n=590) allows a decrease of 12 J. B. B. da Silva, M. B. O. Giacomelli, I. G. de Souza and the re-calibration during routine analysis; hence, a larger A. J. Curtius, Microchem. J., 1998, 60, 249. number of samples can be analysed in a working day. 13 E. Bulska, K. Liebert-Ilkowska and A. Hulanicki, Spectrochim. Furthermore, at least the same accuracy and analyte thermal Acta, Part B, 1998, 53, 1057. 14 D. Pozebon, V. L. Dressler and A. J. Curtius, J. Anal. At. stabilisation provided by the conventional modifier was Spectrom., 1998, 13, 7. attained by using the 250 mg W+200 mg Rh permanent 15 The THGA Graphite Furnace: Techniques and Recommended modifier. Conditions, Perkin-Elmer, U� berlingen, 1992. 16 Pressurized Microwave Decomposition System, Instruction Manual, Paar, Graz, 1992. Acknowledgements 17 N. Hadgu and W. Frech, Spectrochim. Acta, Part B, 1994, 49, 445. 18 E. C. Lima, F. J. Krug and M. A. Z. Arruda, Spectrochim. Acta, The authors thank A. T. Ferreira (CENA-USP) for technical Part B, 1998, 53, 601. support, and are grateful to Fundac�a� o de Amparo a` Pesquisa 19 B. Welz, G. Schlemmer and J. R. Mudakavi, J. Anal. At. do Estado de Sa�o Paulo (FAPESP), Financiadora de Spectrom., 1992, 7, 1257. Estudos e Projetos (PRONEX) and Conselho Nacional de 20 Commission on Spectrochemical and Other Optical Procedures for Analysis. Nomenclature, Symbols, Units and their Usage in Desenvolvimento Cientý�fico e Tecnolo�gico (CNPq) for finan- Spectrochemical Analysis—II. Data Interpretation, Spectrochim. cial support and fellowships. Acta, Part B, 1978, 33, 241. 21 D. L. Massart, B. G. M. Vandeginste, S. N. Deming, Y. Michotte and L. Kaufman, Costs in Data Handling in Science and References Technology—Chemometrics—A Textbook, Elsevier, Amsterdam, 1988, vol. 2, pp. 137–147. 1 W. Slavin, D. C. Manning and G. R. Carnrick, At. Spectrosc., 1981, 2, 137. 2 R. D. Ediger, At. Absorpt. Newsl., 1975, 14, 127. Paper 9/03270C J. Anal. At. Spectrom., 1999, 14, 1601&nda

ISSN:0267-9477    

DOI:10.1039/a903270c

出版商:RSC    

年代:1999

数据来源: RSC

摘要:

Rapid and quantitative release, separation and determination of inorganic arsenic [As(III )+As(V )] in seafood products by microwaveassisted distillation and hydride generation atomic absorption spectrometry Ociel Mun�oz,a Dinoraz Ve� lez,a Marý�a Luisa Cerverab and Rosa Montoro*a aInstituto de Agroquý�mica y Tecnologý�a de Alimentos (CSIC), Apdo. Correos 73, 46100 Burjassot, Valencia, Spain. E-mail: Rmontoro@iata.csic.es bDepartamento de Quý�mica Analý�tica, Universidad de Valencia, Dr.Moliner 50, 46100 Burjassot, Valencia, Spain. E-mail: M.Luisa.Cervera@uv.es Received 22nd June 1999, Accepted 29th July 1999 A precise, simple and rapid method is described for the determination of inorganic arsenic [As(III)+As(V)] in seafood products. The inorganic species were isolated from the matrix by microwave-assisted distillation and determined by hydride generation atomic absorption spectrometry (HGAAS). The microwave and chemical parameters were optimized in order to obtain quantitative inorganic arsenic recoveries.The analytical features of the method are as follows: detection limit 10 ng g-1 (dry mass) or 2 ng g-1 (fresh mass); precision (RSD) 4%; recoveries 106±3% for As(III ) and 113±4% for As(V). Under the optimized conditions, arsenobetaine, arsenocholine and tetramethylarsonium ion added to samples of seafood were not distilled; however, minor species were distilled and were detected in various percentages: 109% monomethylarsonic acid; 11% dimethylarsinic acid; 0.2% trimethylarsine oxide.The concentrations expressed as arsenic contributed by these species would imply a mean overestimate of inorganic arsenic of about 24 ng g-1 dry mass (dm). The concentrations of inorganic arsenic found in natural seafood samples ranged between 0.053 and 1.145 mg g-1 (dm) (mean moisture content 78%). The procedure was compared with an alternative methodology in which acid digestion-solvent extraction-HGAAS was used for the determination of inorganic arsenic.A Student’s t-test for paired samples of the data obtained by the two methodologies showed no significant diVerences (P-value 0.66). In fact, most of the studies that quantify inorganic arsenic in Introduction biological matrices after extraction by means of organic sol- The concentrations of arsenic found in seafood products vary vents do not provide recovery values for these arsenic over a wide range [values expressed in mg g-1, fresh mass species,5–7 or else the recoveries are low for the two species,8 (fm)]: fish 0.06–45; lamellibranchs 0.1–14.4; cephalopods especially for As(III).9 This is because the organic solvent is 0.1–49; gasteropods 1.0–40; crustaceans <0.04–91.1 This total unable to break the bond of As(III ) to thiol groups in proteins, arsenic is contributed by diVerent forms of arsenic which diVer and to release the As(III).For this reason it is necessary to considerably in their toxicological connotations.For example, use an attack with a non-oxidizing acid (HCl), which the most toxic species are As(III ) and As(V), the sum of which completely denatures the proteins in the sample without alterconstitutes the inorganic arsenic, whereas no harmful eVects ing the chemical structure of the organic arsenic species.9 have been attributed to arsenobetaine (AB).2 For the methyl- Procedures for determining inorganic arsenic based on attackated species such as monomethylarsonic acid (MMA), ing the sample with acid have previously been reported in the dimethylarsinic acid (DMA) and trimethylarsine oxide literature.10–12 Our laboratory has recently established a meth- (TMAO), with the exception of tetramethylarsonium ion odology based on solubilization with HCl, reduction and (TMA+), the toxicity diminishes as the degree of methylation subsequent extraction with chloroform, back-extraction into increases. The International Agency for Research on Cancer dilute HCl, dry-ashing and quantification by HGAAS (IARC) has evaluated arsenic ingested or inhaled as a Group which allows quantitative recovery of inorganic arsenic.9 I carcinogen to humans.3 The scientific community attributes Other organoarsenic species [DMA, AB, arsenocholine (AC), the majority of cancerous pathologies to inorganic arsenic.4 TMA+] are not co-extracted or degraded.MMA is quantitat- This situation creates a need for methodologies capable of ively recovered and TMAO is recovered to a small extent quantifying inorganic arsenic in order to make a real evaluation (3–10%).of the health risk involved in the consumption of seafood As an alternative to the extraction procedures cited above, products that contain arsenic. procedures based on distillation appear in the literature. For The procedures for determining inorganic arsenic reported example, Lunde13 developed a methodology that obtained in the literature include some that are performed by means of quantitative recoveries for As(III) and As(V), using conextraction with organic solvents.5–8 However, in a previous ventional distillation that combined release and separation paper we showed that extraction of inorganic arsenic by means in a single stage employing HCl and a reducer [Fe(II)]. Subsequently, Flanjak14 modified this methodology, using HBr of organic solvents does not provide quantitative recoveries.9 J.Anal. At. Spectrom., 1999, 14, 1607–1613 1607as the reducing agent, and obtained inorganic arsenic recover- with a Perkin-Elmer Model 5000 atomic absorption spectrometer (Perkin-Elmer, Norwalk, CT, USA) equipped with ies of the order of 94–106%.None of the procedures mentioned included a study of co-distillation or degradation of the organic a Perkin-Elmer FIAS 400 system operating as a hydride generator in continuous flow mode. An electrothermally heated arsenic species present in seafood products.In order to simplify and speed up the determination of inorganic arsenic proposed quartz cell was employed. Other equipment used included a lyophilizer equipped with by Lunde,13 our laboratory has developed a methodology that employs microwave-assisted distillation as an alternative to a microprocessor controlling the lyophilization process (FTS Systems, New York, USA) and connected to a computer; a conventional distillation, with subsequent determination of inorganic arsenic in the distillate by HGAAS.This method- PL 5125 sand bath (Raypa, Scharlau, Barcelona, Spain); a K 1253 muZe furnace equipped with a Eurotherm Controls 902 ology reduces the distillation process, improves the reduction of As(V) to As(III ) and proves the non-degradability of AB.15 control program (Heraeus, Madrid, Spain); a KS 125 Basic mechanical shaker (IKA Labortechnik, Merck Farma y In the present study, starting from the methodology of Lo�pez et al.,15 we have redesigned the system of the distillation Quý�mica, Barcelona, Spain) and an Eppendorf 5810 centrifuge (Merck Farma y Quý�mica, Valencia, Spain).collector flask, optimized the physical and chemical parameters, replaced HGAAS batch determination of arsenic by Reagents HGAAS determination in continuous flow mode, and studied the possible interference that could be caused by organoarsenic De-ionized water, of 18 MV cm resistivity, obtained with a species frequently found in fish (MMA, DMA, AC, TMAO Milli-Q water purification system (Millipore, Millipore Ibe�rica, and TMA+).The methodology developed improves on the Madrid, Spain), was used for the preparation of reagents and methodology of Lo�pez et al. in terms of speed and oVers standards. All chemicals including standards and reagents better analytical characteristics and absence of interference. It were of pro analysi quality or better. was applied to natural seafood samples and the results were The stock standard solution of As(III ) (1000 mg l-1) was compared with those obtained by acid digestion-solvent prepared by dissolving 1.320 g of arsenic trioxide (Riedel extraction-HGAAS.9 de-Hae�n, Hannover, Germany) in 25 ml of 20% m/v KOH solution, neutralizing with 20% v/v H2SO4, and diluting to 1 l Experimental with 1% v/v H2SO4.The standarmg l-1) was prepared by dilution of the Titrisol Instrumental standard (Merck, Darmstadt, Germany).The standard solutions (1000 mg l-1) of MMA, DMA, TMAO and TMA+ The microwave-assisted distillation was performed with a were prepared by dissolving appropriate amounts of domestic Moulinex Micro-Chef 700 microwave oven CH3AsO(ONa)2·6H2O (Carlo Erba, Milan, Italy), (Moulinex, Valencia, Spain). Microwave emission was (CH3)2AsO2Na·3H2O (Fluka, Madrid, Spain), (CH3)3AsO unpulsed, using maximum power (700 W) and an operating (Hot Chemical Co., Tokyo, Japan) and (CH3)4As+I- frequency of 2450 MHz.The sample was placed inside a (Hot Chemical Co.), respectively, in water. Solutions of poly(tetrafluoroethylene) (PTFE) vessel with an internal AB [(CH3)3As+CH2COO-] (973 mg l-1) and AC volume of 120 ml (4.0 cm id, height 9.5 cm) and 10 mm wall [(CH3)3As+CH2CH2OH] (1000 mg l-1) were obtained from thickness and with a tight-fitting screw-cap lid designed in our the Service Central d’Analyse du CNRS (Vernaisson, France). laboratory. The lid was provided with an outlet so that the Ashing aid suspension was prepared by stirring 20 g of distillate could be emptied from the vessel by means of a Mg(NO3)2·6H2O and 2 g of MgO in 100 ml of water until PTFE tube (1.0 mm id).The tube was passed through the homogeneous. The following reducing solutions were used: vent holes of the oven. The distillate was collected in a 63 ml 5% m/v KI+5% m/v ascorbic acid, aqueous solution; 1 ml of gas collection flask (2.0 cm id, height 20.0 cm) which had a 1.5% m/v hydrazine sulfate aqueous solution+2 ml (48%) glass tube in the centre (0.3 cm id, length 35.0 cm) connected HBr; 1 ml of 1.0% m/v hydrazine sulfate aqueous solution+ to the PTFE tube through which the distillate from the PTFE 1 ml (48%) HBr.vessel passed. A porous glass plate was cemented to the funnel- As reducing solution for hydride generation, sodium tetra- shaped lower end of the glass tube to allow uniform diVusion hydroborate(III) solution (1% m/v) was prepared by dissolving of the distillate in the collecting liquid (Fig. 1). NaBH4 powder in 0.7% m/v NaOH solution and filtering Determination of inorganic and total arsenic was performed through Whatman No. 42 paper. Fresh NaBH4 solution was prepared daily. All glassware was treated with 10% v/v HNO3 for 24 h, and rinsed three times with de-ionized water before being used. The reference materials employed, DORM-1 and DORM-2 (Dogfish Muscle, Squalus acanthias) and TORT-2 (Lobster Hepatopancreas), were purchased from the National Research Council Canada (Ottawa, Ontario, Canada).Sample preparation Inorganic arsenic was determined in various fresh, frozen and canned products purchased at local retail outlets. The fresh samples were prepared as they are normally eaten. The brine or sauce in the canned seafood products was removed by the method for determining the drained mass of canned foods.16 The drained samples of fresh and canned products were cut into pieces and frozen at -20 °C and afterwards freeze-dried for 48 h at a chamber pressure of 0.225 Torr.Sublimation heat Fig. 1 Diagram of microwave-assisted distillation system. A: Collector was supplied by conduction from heating plates at 20 °C. The gas tube (2.0 cm id, height 20.0 cm); B: PTFE tube (1.0 mm id); C: lyophilized samples were crushed and homogenized to a fine PTFE vessel (4.0 cm id, height 9.5 cm); D: security load; E: microwave oven; and F: poly(propylene) container. powder in a mill.The resulting powder was stored in previously 1608 J. Anal. At. Spectrom., 1999, 14, 1607–1613decontaminated twist-oV flasks and kept in the refrigerator at inorganic arsenic in the chloroform phase by agitating for 3 min with 10 ml of 1 mol l-1 HCl. Separate the phases by 4 °C until analysis. centrifugation, aspirate the aqueous phase and pour it into a beaker; repeat this stage again and combine the back- Determination of total arsenic extraction phases obtained.Weigh 0.25±0.01 g of dry sample in a 250 ml beaker. Add The determination of inorganic arsenic is performed by 1 ml of ashing aid suspension and 5 ml of 50% v/v HNO3. means of the following procedure: Add 2.5 ml of ashing aid After evaporation to total dryness on a sand-bath, mineralize suspension and 10 ml of concentrated HNO3 to the combined in accordance with the programme described in a previous back-extraction phases. Evaporate and treat in the same way paper.17 Dissolve the ash from the mineralized samples in 5 ml as for total arsenic.of 50% v/v HCl, and pre-reduce the extract with 5 ml of With samples that, in the back-extraction phase, generate KI+ascorbic acid. After 30 min, dilute to volume with water emulsions that cannot be broken up by centrifugation at over and filter through Whatman No. 1 filter-paper into a 25 ml 2000 rpm, transfer the emulsion into a beaker. After adding calibrated flask. The instrumental conditions used for the the ashing aid suspension and HNO3 and heating gently in determination of arsenic by HGAAS in continuous-flow mode the sand-bath for not more than 30 s, break up the emulsion are shown in Table 1.and remove the chloroform phase created by aspiration. Determination of inorganic arsenic Results and discussion Microwave-assisted distillation (proposed method). Weigh Microwave oven parameters 0.50±0.01 g of lyophilized sample in the PTFE vessel. Add 2 ml of H2O and 1 ml of reducing solution (KI+ascorbic For optimization of the analytical method a domestic microacid). Agitate, wait 5 min until the sample is completely moist, wave oven was used, with the rotating platform removed.The and add 12 ml of 8.25 mol l-1 HCl. Cover the PTFE vessel. PTFE vessel was placed inside a poly(propylene) container in In samples of products conserved in oil perform the wetting the lid of which a hole was made for the PTFE tube from the process for 10 min.Add 8 ml of H2O to the collector flask. PTFE vessel. This container protects the microwave oven from Connect the PTFE vessel to the collector flask and radiate at acid corrosion if there is leakage during distillation. The maximum power (700 W) until bubbling stops in the collector diVerent positions in the microwave cavity are not identical flask (4–6 min). Dilute the distillate to 50 ml, take an aliquot from the microwave radiation point of view. Therefore, the of 10 ml and add 1 ml of KI+ascorbic acid.Leave to react methodology of Yba�n� ez et al.17 was used to determine the for 30 min and determine inorganic arsenic by HGAAS, area of maximum radiation in which to place the PTFE vessel. applying the operating conditions described in Table 1. In This position was 13 cm from the door and 13 cm order to avoid cross-contamination between samples, clean from the magnetron. Similarly, the security load required to the inside of the PTFE tube after each distillation by con- avoid damage to the magnetron through overheating was necting it to a peristaltic pump and passing 20 ml of de-ionized established (25 ml ).water through it in the opposite direction to that of the distillation. Distillation parameters This study was aimed, in the first instance, at determining Solvent extraction.9 Weigh the lyophilized sample As(III ) and As(V) separately in seafood samples. The following (0.50±0.01 g) into a 50 ml screw-top centrifuge tube.Add strategy was proposed: distillation of As(III) during a first 4.1 ml of water and agitate until completely moistened. Then stage, and addition of a reducer and distillation of As(V) in a add 18.4 ml of concentrated HCl and agitate again for 1 h. second stage. With this in mind, the acidity conditions and Leave to stand for 12–15 h (overnight). collection of distillate proposed by Lo�pez et al.15 were tested Add the reducing agent (1 ml of 1.5% m/v hydrazine sulfate with each of the standards [As(III ), As(V), MMA or DMA], solution+2 ml of HBr) and agitate for 30 s.Add 10 ml of without a reducing agent, and the reagent volumes were CHCl3 and agitate for 3 min. Separate the phases by centrifuadjusted so that the volume of distillate was in keeping with gation at 2000 rpm for 5 min. Separe chloroform phase the capacity of the collector flask. Accordingly, 1 ml of by aspiration and pour it into another tube. Repeat the 1 mg ml-1 of each standard solution (expressed as As), 4 ml extraction process two more times.Combine the chloroform of H2O and 10 ml of 9.9 mol l-1 HCl (final concentration phases and centrifuge again. Eliminate the remnants of the 6.6 mol l-1 HCl) were placed in the PTFE vessel. The distil- acid phase completely by aspiration (acid phase remnants in lates were collected in 15 ml of water. The analytical determi- the chloroform phase cause substantial overestimates of nation by HGAAS of the arsenic present in the distillates was inorganic arsenic).Eliminate possible remnants of organic performed immediately after the process had ended. The material in the chloroform phase by passing it through a recoveries obtained for the various species determined were: Whatman GD/X syringe filter with a 25 mm PTFE membrane As(III ) 104%, As(V) 0–26%, MMA 0% and DMA 0%. The (Merck Farma y Quý�mica, Barcelona, Spain). Back-extract the variability in the distillation of As(V) suggests the possible non-quantitative formation of arsenic pentachloride (AsCl5).Table 1 Operating conditions for HGAAS This species is unstable and is completely dissociated to AsCl3 Hydride generation continuous-flow mode— and Cl2.18 Cell temperature 900 °C A study was made of whether the distillation of As(III ) took Sample solution 1 ml min-1 flow rate place throughout radiation or during a shorter period. It was Reducing agent 1.0% m/v NaBH4 in 0.7% m/v found that the entire volume in the PTFE vessel had to be NaOH; 1.0 ml min-1 flow rate distilled in order to obtain quantitative recoveries.The process HCl solution 1.5 mol l-1; 2.5 ml min-1 flow rate takes place within a single stage, providing a considerable Carrier gas Argon; 45 ml min-1 flow rate reduction in distillation time (4–6 min) by comparison with Atomic absorption spectrometer— the methodology of Lo�pez et al.15 Wavelength 193.7 nm The variation in the distillation percentage of the As(V) Spectral bandpass 0.7 nm standard calls into question the viability of the working Lamp power As EDL system 2; 400 mA conditions employed previously for the independent J.Anal. At. Spectrom., 1999, 14, 1607–1613 1609determination of As(III) and As(V). However, despite the HGAAS. Greater concentrations of L-cysteine cause interference in the hydride generation, reducing the sensitivity results obtained with standards, it was considered useful to study the overall eVect of the matrix on the distillation of drastically.The optimized conditions for standards were then tested As(III ) and As(V). Accordingly, applying the procedure described earlier, recovery tests of As(III), As(V) and with natural samples. Lyophilized sardine (1 and 0.5 g) was spiked with 1 ml of 1 mg ml-1 (as As) of standard [As(III), [As(III)+As(V)] were made on natural seafood samples (sardines and small squid). In each assay 1 g of lyophilized sample As(V), MMA or DMA], 5 ml of 0.1% m/v L-cysteine and 10 ml of 9.9 mol l-1 HCl.Quantitative recoveries were was placed in the PTFE vessel together with 1 ml of 1 mg ml-1 of each standard solution (expressed as As), 4 ml of water and obtained for As(III ), As(V) and MMA, and recoveries of 78% for DMA, showing that the methylated species are also pre- 10 ml of 9.9 mol l-1 HCl. The distillate obtained was quantified in the same way as for the standards. It was shown that the reduced to their trivalent state, forming organosulfur derivatives. 21 Taking as a basis the mean concentrations of MMA As(III ) added to the two samples was recovered quantitatively (sardine 118%; small squid 113%), as was the As(V) (sardine and DMA found by us in a wide range of seafood products [MMA: 6 ng g-1 dm; DMA: 167 ng g-1 dm, both expressed 103%; small squid 87%). Combined addition of the two standards also provided quantitative recoveries (sardine 97%; as As (dm=dry mass)],22 we evaluated the significance of the recoveries obtained for MMA and DMA on the level of small squid 100%).Comparison of these results with those obtained in the distillation of the standards [As(III ) 104%; inorganic arsenic detected when this reducer is used. The concentration of DMA, higher than that of MMA, causes the As(V) 0–26%] revealed a diVerence in the behaviour of As(V), which was not distilled quantitatively without the matrix but high recovery percentage in the distillate to lead to a substantial mean overestimate of the inorganic arsenic (136 ng g-1 dm), was distilled quantitatively when added to a natural sample.This suggests the existence of reducing agents in the samples which rules out the use of this reducer. Also, although the use of L-cysteine at concentrations of 0.1% does not cause inter- tested, which transform the As(V) into As(III) and make its distillation possible. A similar phenomenon was reported by ference in the determination of standards, in natural samples it triggers the formation of volatile sulfurized compounds us in a previous paper dealing with the existence of reducing agents in a natural sample which was able to reduce As(V) which interfere in hydride generation.during the extraction stage of inorganic arsenic.9 Consequently, Use of the reducing mixture HBr–hydrazine sulfate it was decided to add to the sample a reducer which, irrespective of the type of sample, would ensure quantitative reduction In view of the previous use of HBr as a reducing agent in of As(V) to As(III).Under these conditions it is only possible conventional distillation of inorganic arsenic,14 and the use of to make a quantification of the two species combined. the reducing mixture HBr–hydrazine sulfate for the reduction of arsenate,9,23 it was decided to test the validity of this Use of the reducer KI reducing mixture under the following conditions: 1 ml of concentrated HBr, 1 ml of 1.0% m/v hydrazine sulfate and First, a study was made of the individual behaviour of each of the arsenic species [As(III), As(V), MMA or DMA] in the 12 ml of 8.25 mol l-1 HCl (final concentration 6.6 mol l-1 HCl) were added to 1 ml of 1 mg ml-1 (as As) of standard presence of KI as reducer.Accordingly, 1 g of sample was placed inside the PTFE vessel together with 1 ml of 1 mg ml-1 [As(III ), As(V), MMA or DMA]. The distillate was collected in 8 ml of water and an aliquot of the distillate (10 ml ) was (as As) of each standard [As(III ), As(V), MMA, DMA], 5 ml of 30% m/v KI, and 10 ml of 9.9 mol l-1 HCl.In the collector then reduced with 1 ml of KI+ascorbic acid and quantified by HGAAS. The distillation of this mixture showed that the flask 15 ml of a 0.25% m/v solution of hydroxylamine hydrochloride were placed, this being the solution used previously arsenic species As(III), As(V) and MMA are recovered quantitatively, whereas the recovery of DMA is 81%.The recovery by Lo�pez et al.15 to maintain reducing conditions in the collecting flask. It was observed that As(III ) and As(V) distilled of these species suggests that the As(V) in the methylated species has been reduced to the corresponding species of quantitatively, whereas MMA and DMA were determined non-quantitatively (56 and 10%, respectively). The distillation As(III ), and that the bromides of these methylated species of As(III ) co-distil with the inorganic arsenic, as for reduction of the methylated species shows that the As(V) in these species has been reduced to the corresponding species of As(III), and with KI described above.Use of the HBr–hydrazine sulfate mixture was rejected the iodides of these methylated species of As(III) co-distil with inorganic arsenic. The existence of these species of As(III) because of the high recovery percentage of DMA, since, as with L-cysteine, it would cause a substantial overestimate of (MMAI2 and DMAI) was described by Suzuki et al.19 The use of 30% m/v KI as a reducing agent generates a the inorganic arsenic. considerable amount of I3- in the distillate, which cannot be Use of the reducing mixture KI–ascorbic acid completely reduced by the hydroxylamine hydrochloride. Contly, the presence of I3- creates interference in the The reducing mixture KI+ascorbic acid has been extensively subsequent hydride generation, making it necessary to use the used for reduction of arsenic.24 The reducing agent (I-) is method of additions or high dilutions which increase the limits constantly regenerated each time the As(V) is reduced to of detection (LODs).Given the low levels at which inorganic As(III ), since the I3- produced in the reaction is reduced by arsenic is found in some seafood products, the search for the ascorbic acid. This increases the eYciency of the iodide as methodologies with low LODs makes it advisable to test a reducer during distillation and eliminates interference due alternative reduction conditions.to excess of I3-. This reducing mixture was first tested on arsenic standards. Use of the reducer L-cysteine In the PTFE vessel 1 ml of water, 1 ml of the reducing mixture (KI+ascorbic acid) and 12 ml of 8.25 mol l-1 HCl (final The use of cysteine in hot acidic medium reduces As(V) quantitatively to As(III ).20 Therefore, it was decided to test concentration 6.6 mol l-1 HCl) were added to 1 ml of 1 mg ml-1 (as As) of each standard [As(III), As(V), MMA, this substance as a reducing agent.The optimum conditions required for quantitative distillation of 1 ml of 1 mg ml-1 of DMA, AB, AC, TMAO or TMA+]. The distillate was collected in 8 ml of water and made up to 50 ml with water. As(III ) and As(V) were: 10 ml of 9.9 mol l-1 HCl and 5 ml of 0.1% m/v L-cysteine (final concentration 6.2 mol l-1 HCl). The An aliquot of 10 ml of the distillate was pre-reduced with KI+ascorbic acid and quantified by HGAAS.The recoveries distillate was collected in water and immediately quantified by 1610 J. Anal. At. Spectrom., 1999, 14, 1607–1613Table 3 Analytical characteristics of the method Table 2 Recovery of distilled standards of arsenic species. Quantification by HGAAS or dry-ashing HGAAS LOD/ng g1 Asa 10 (dm)&2 (fm) Precision (RSD %)b Sardine 4% [104 (dm)&25 (fm)] Species Arsenic recovery (%) Cockle 4% [678 (dm)&163 (fm)] Small squid 3% [93 (dm)&22 (fm)] Distillation, Distillation, HGAAS dry-ashing HGAAS Mean 4% Recovery (%)c As(III ) Sardine 103¡À4% (25, 48) Cockle 109¡À5% (163, 168) As(III ) 108 107 As(V) 108 101 Mean 106¡À3% As(V) Sardine 109¡À2% (25, 48) MMA 96 96 DMA 6 65 Cockle 116¡À5% (163, 168) Mean 113¡À4% AB 0 0 AC 0 0 MMA Sardine 106¡À2% (25, 48) Cockle 112¡À3% (163, 168) TMAO 0.2 12 TMA+ 0 0 Mean 109¡À3% DMA Sardine 8¡À2% (25, 48) Cockle 13¡À3% (163, 168) Mean 11¡À3% were quantitative for As(III) (98%), As(V) (99%) and MMA aNine reagent blanks were employed; (dm) dry mass; (fm) fresh mass.(101%). For the remaining organoarsenic species the recoveries bRelative standard deviation from six independent analyses. Values in obtained ranged from 0 to 6%: AB (0%), DMA (6%), AC square brackets are the mean inorganic arsenic concentrations for the (0%), TMAO (0.2%) and TMA+ (0%) (Table 2). In order to samples analysed, expressed in ng g1 As dm and fm. cPercentage recoveries from three independent analyses expressed as verify whether the recovery percentages obtained previously mean¡Àstandard deviation.Values in parentheses are the mean inor- were due to distillations of the various species or to the ganic arsenic concentrations of the unspiked samples (first value) and numerous factors that aVect hydride generation (type and concentration added (as arsenic) of each arsenic species (second value) concentration of reagents, reduction temperature, oxidation in ng g1 As (fm).state, chemical form, sample matrix),25 the following experiment was performed: an aliquot of 25 ml of distillate was mineralized by dry-ashing and the arsenic was quantified by HGAAS (Table 2). It was seen that for As(III ), As(V) and MMA the results obtained by direct reading of the distillate (106¡À3%) and As(V) (113¡À4%) demonstrate the validity of and readings after dry-ashing coincided. The results obtained this methodology. These analytical characteristics are an by dry-ashing HGAAS show that the distillation of DMA was improvement on the LOD (68 ng g1 dm) and precision (9%) 64%, but in the distillate read directly only 6% was detected.obtained in the methodology proposed by Lo¡äpez et al.15 The behaviour of TMAO was similar: the distillation was 12%, although in the direct reading by HGAAS only 0.2% Application to reference materials was detected. The other arsenic species (AB, AC and TMA+) were not detected in the direct reading of the distillate, as was The proposed methodology was applied to reference materials with a certified total arsenic content: DORM-1, DORM-2, to be expected, since they are not hydride-generating species, nor were they detected in the mineralized distillate, since they TORT-2. In these samples total arsenic was quantified by dryashing HGAAS.For all the samples the results obtained were are not volatile species. The distillation conditions were then applied to natural within the certified concentration ranges: value found/value certified (mg g1, dm): DORM-1: 16.2¡À0.4/17.7¡À2.1; samples of sardine and cockle.The use of this reducer does not cause negative interference during instrumental reading; DORM-2: 17.9¡À0.5/18.0¡À1.1; TORT-2: 22.3¡À0.2/21.6¡À1.8. As there are no certified values for the concentrations of under these conditions it does not reduce analytical sensitivity and it is possible to use samples of 0.5 g.In the PTFE vessel inorganic arsenic in DORM-1, DORM-2 and TORT-2, the results obtained were compared with those provided by other 1 ml of 1 mg ml1 (as As) of each standard [As(III), As(V), MMA, DMA, AB, AC, TMAO or TMA+], 1 ml of water, workers. For the matrices mentioned, Table 4 shows the concentrations of As(III)+As(V)+MMA, and also As(III)+As(V) 1 ml of the reducing mixture (KI+ascorbic acid) and 12 ml of 8.25 mol l1 HCl were added to 0.5 g of sample. As in the and MMA found by various workers,5,9,26 together with the concentrations of As(III )+As(V)+MMA+11% DMA found tests with standards, the distillation percentages were quantitative for As(III), As(V) and MMA and remained at 0.2% for in the present work.DORM-1 is the sample that has been most analyzed in TMAO, and there was no distillation of AB, AC and TMA+. However, for DMA the mean distillation percentage in the studies of arsenic speciation. Its mean concentration of DMA, obtained from data in the literature,27,28 is 0.500 mg g1(dm), samples increased from 6 to 11%.Given the mean values of MMA and DMA in seafood products reported previously,22 of which 11% (0.055 mg g1, dm) is detected with the methodology developed in this work. If we subtract from the concen- the distillation of these methylated species would lead to a mean overestimate of the concentration of inorganic arsenic tration of As(III)+As(V)+MMA+11% DMA obtained in this work (0.196 mg g1, dm) the arsenic contributed by the of 6 ng g1 by co-distillation of MMA (100% of 6 ng g1 dm) and of 18 ng g1 by co-distillation of DMA (11% of percentage of DMA quantified with the present methodology (0.055 mg g1, dm), it is possible to compare the 167 ng g1 dm).In the light of the results obtained we consider that the most suitable reducer is the mixture of KI+ascorbic mean value resulting (0.141 mg g1) with the values of As(III )+As(V)+MMA reported by other workers. Those acid, given the insignificant mean overestimate of inorganic arsenic that it produces (24 ng g1 dm) and the fact that it values (0.180 mg g1 Beauchemin et al.,5 0.165 mg g1 S¡¦ lejkovec et al.,26 0.124 mg g1 Muno¡ä z et al.9) are in reasonable does not cause interference under the working conditions employed. agreement with the result obtained in this work (0.141 mg g1).The concentrations of inorganic arsenic obtained for the In order to determine whether the methodology developed is suitable for the determination of inorganic arsenic in seafood other CRMs with the proposed methodology (DORM-2: 0.102¡À0.029 mg g1; TORT-2: 0.506¡À0.031 mg g1) are products, its analytical characteristics were evaluated (Table 3).The values obtained for LOD (10 ng g1 dm) and close to those reported by Munoz et al.9 (DORM-2: 0.145¡À0.011 mg g1; TORT-2: 0.581¡À0.055 mg g1). precision (4%) and the satisfactory recoveries of As(III) J. Anal. At. Spectrom., 1999, 14, 1607�C1613 1611Table 4 Concentrations of inorganic arsenic in certified reference materials CRM Arsenic species/mg g.1 (dm) Reference DORM 1 As(III)+As(V)+MMA 0.180¡¾0.040 Beauchemin et al.5 As(III)+As(V) 0.129¡¾0.003 S¢§ lejkovec et al.26 MMA 0.036¡¾0.010 S¢§ lejkovec et al.26 As(III)+As(V)+MMA 0.124¡¾0.004 Mun.oz et al.9 As(III)+As(V)+MMA+11% DMA 0.196¡¾0.004 This work DORM-2 As(III)+As(V)+MMA 0.145¡¾0.011 Mun.oz et al.9 As(III)+As(V)+MMA+11% DMA 0.102¡¾0.029 This work TORT-2 As(III)+As(V)+MMA 0.581¡¾0.055 Mun.oz et al.9 As(III)+As(V)+MMA+11% DMA 0.506¡¾0.031 This work Table 5 Total arsenic and moisture content in seafood.Inorganic arsenic content of samples obtained by solvent extraction and microwave distillation Sample Moisture (%) Total Asa/ Inorganic Asb/mg g.1 (dm) mg g.1 (dm) Solvent extraction Microwave distillation Fresh. Anchovy 76 18.78¡¾0.45 0.176¡¾0.005 0.123¡¾0.006 Clam 86 16.12¡¾0.56 1.009¡¾0.090 1.145¡¾0.012 Cockle 78 20.71¡¾0.43 0.688¡¾0.021 0.678¡¾0.029 Mussel 83 11.52¡¾0.03 0.223¡¾0.004 0.265¡¾0.009 Mussel 82 11.37¡¾0.11 0.220¡¾0.022 0.202¡¾0.069 Mussel 79 12.03¡¾0.18 0.335¡¾0.009 0.270¡¾0.013 Mussel 84 9.14¡¾0.56 0.212¡¾0.007 0.211¡¾0.051 Mussel 75 9.42¡¾0.08 0.245¡¾0.007 0.280¡¾0.001 Mussel 83 14.79¡¾0.09 0.238¡¾0.005 0.285¡¾0.033 Mussel 82 17.48¡¾0.10 0.400¡¾0.019 0.305¡¾0.030 Mussel 83 11.83¡¾0.24 0.224¡¾0.003 0.277¡¾0.047 Mussel 84 11.42¡¾0.04 0.196¡¾0.007 0.151¡¾0.012 Sardine 76 16.72¡¾0.41 0.137¡¾0.008 0.099¡¾0.004 Small squid 80 9.85¡¾0.30 0.063¡¾0.002 0.053¡¾0.005 Squid 86 1.90¡¾0.09 0.074¡¾0.003 0.104¡¾0.006 Canned.Cockle 77 9.09¡¾0.01 0.749¡¾0.031 0.894¡¾0.064 Langostillo 72 6.97¡¾0.09 0.477¡¾0.025 0.385¡¾0.027 Octopus 69 9.47¡¾0.23 0.319¡¾0.007 0.313¡¾0.008 Razor clam 69 1.98¡¾0.14 0.201¡¾0.019 0.249¡¾0.023 Shrimp 73 3.18¡¾0.02 0.247¡¾0.013 0.246¡¾0.029 Frozen. Shrimp 82 6.00¡¾0.16 0.384¡¾0.028 0.396¡¾0.042 aTotal arsenic determined in duplicate (mean¡¾standard deviation).bInorganic arsenic determined in triplicate (mean¡¾standard deviation). Application to natural samples; comparison of methodologies methodology described here oVers the performance necessary for use in routine analyses with the instrumentation available The optimized methodology was applied to 21 natural seafood in many control laboratories. samples. Table 5 shows the concentrations of total arsenic and inorganic arsenic and the moisture content for each sample.It also shows the values of inorganic arsenic obtained for the Acknowledgements same samples by applying the methodology of extraction with The authors gratefully acknowledge the financial support organic solvents developed by Mun.oz et al.9 The Student¡�s tof the Comisio¢¥n Interministerial de Cienc©¥¢¥a y Tecnolog©¥¢¥a test for paired samples was used to compare the values of (CICyT), Project ALI96.0511, for which they are deeply inorganic arsenic found by the two methodologies.The Pindebted. O. Mun.oz received a Research Personnel Training value obtained (0.66) indicates that there are no significant Grant from the Instituto de Cooperacio¢¥n Iberoamericana diVerences between the two methodologies with a confidence (ICI). level of 95%. The concentrations of inorganic arsenic for the samples analysed by the proposed method were in the range References 0.053.1.145 mg g.1 (dm). The percentage of inorganic arsenic in the samples varied between 0.5.12.6%, with a mean value 1 M. L.Cervera, A. Navarro, R.Montoro and J. Gomez, Fresenius¡� of 3.9¡¾3.3%. The percentages of inorganic arsenic do not J. Anal. Chem., 1993, 347, 58. 2 P.Mitchell and D. Barr, Environ. Geochem. Health, 1995, 17, 57. diVer substantially from those given in the literature for 3 T. Tsuda, A. Babazono, T. Ogawa, H. Hamada, Y. Mino, H. seafood products 0.1.6.5%.9,14,29 Aoyama, N. Kurumatani, T. Nagira, N. Hotta, M. Harada and S. Inomata, Appl.Organomet. Chem., 1992, 6, 309. 4 D. J. Hewitt, G. C. Millner, A. C. Nye and H. F. Simmons, Conclusion Environ. Res., 1995, 68, 73. The proposed method allows simple and rapid determination 5 D. Beauchemin, M. E. Bednas, S. S. Berman, J. W. McLaren, K. W. M. Siu and R. E. Sturgeon, Anal. Chem., 1988, 60, 2209. of inorganic arsenic in seafood products. Moreover, the 1612 J. Anal. At. Spectrom., 1999, 14, 1607.16136 E. H. Larsen, G. Pritzl and S. H. Hansen, J. Anal. At. Spectrom., 19 N. Suzuki, K. Satoh, H. Shoji and H. Imura, Anal. Chim. Acta, 1993, 8, 1075. 1986, 185, 239. 7 S. Branch, L. Ebdon and P. O’Neill, J. Anal. At. Spectrom., 1994, 20 M. F. Huang, S. J. Jiang and C. J. Hwang, J. Anal. At. Spectrom., 9, 33. 1995, 10, 31. 8 H. Helgensen and E. H. Larsen, Analyst, 1998, 123, 791. 21 D. L. Tsalev, J. Anal. At. Spectrom., 1999, 14, 147. 9 O.Mun� oz, D. Ve� lez and R. Montoro, Analyst, 1999, 124, 601. 22 D. Ve� lez, N. Yba�n� ez and R. Montoro, J. Agric. Food Chem., 1996, 10 H. C. Beard and L. A. Lyerly, Anal. Chem., 1961, 33, 1781. 44, 859. 11 A. Yasui, C. Tsutsumi and S. Toda, Agric. Biol. Chem., 1978, 23 W. Holak and J. J. Specchio, At. Spectrosc., 1991, 12, 105. 42, 2139. 24 W. A. Maher, At. Spectrosc., 1987, 8, 88. 12 H. Mu�nz and W. Lorenzen, Fresenius’ Z. Anal. Chem., 1984, 319, 25 D. Beauchemin, J. Anal. At. Spectrom., 1998, 13, 1. 395. 26 Z. a� lejkovec, A. Byrne, B. Smodio� and M. Rossbach, Fresenius’ 13 G. Lunde, J. Sci. Food Agric., 1973, 24, 1021. J. Anal. Chem., 1996, 354, 592. 14 J. Flanjak, J. Sci. Food Agric., 1982, 33, 579. 27 W. Goessler, D. Kuehnelt, C. Schlagenhaufen, Z. a� lejkovec and 15 J. C. Lo� pez, C. Reija, R. Montoro, M. L. Cervera and M. de la K. J. Irgolic, J. Anal. At. Spectrom., 1998, 13, 183. Guardia, J. Anal. At. Spectrom., 1994, 9, 651. 28 D. Ve� lez, N. Yban� ez and R. Montoro, J. Anal. At. Spectrom., 16 Presidencia de Gobierno, orden 21 de Noviembre de 1984 por la 1996, 11, 271. que se aprueban las normas de calidad para conservas, BOE, 30/11, 29 P. J. Brooke and W. H. Evans, Analyst, 1981, 106, 514. 1–3/12/1984, 287, 34574. 17 N. Yba�n� ez, M. L. Cervera, R. Montoro and M. de la Guardia, J. Anal. At. Spectrom., 1991, 6, 379. 18 E. S. Gould, in Inorganic Reactions and Structure, ed. H. Holt and Paper 9/04999A Co., New York, 1958, ch. 16, p. 301. J. Anal. At. Spectro

ISSN:0267-9477    

DOI:10.1039/a904999a

出版商:RSC    

年代:1999

数据来源: RSC