Anal. Chem. 2005, 77, 511-516
Species-Specific GC/ICP-IDMS for Trimethyllead Determinations in Biological and Environmental Samples Nataliya Poperechna and Klaus G. Heumann*
Institute of Inorganic Chemistry and Analytical Chemistry, Johannes Gutenberg-University Mainz, Duesbergweg 10-14, 55099 Mainz, Germany
An accurate and sensitive species-specific isotope dilution GC/ICPMS method was developed for the determination of trimethyllead (Me3Pb+) in biological and environmental samples. A trimethyllead spike was synthesized from 206Pb-enriched metallic lead by reaction of lead halide with methyllithium and subsequent formation of trimethyllead iodide. The isotopic composition of the spike solution was determined by GC/ICPMS after derivatization with tetraethylborate, and its concentration was determined by reverse isotope dilution analysis. The species-specific GC/ICP-IDMS method was validated by reference material CRM 605 (urban dust) certified for Me3Pb+. The method was also applied to determine the Me3Pb+ content in six biological reference materials (DORM 2, CRM 278, CRM 422, CRM 463, CRM 477, MURST-ISS-A2) and one sediment reference material (CRM 580) for which no certified values of this species exist. The Me3Pb+ concentrations in the biological reference materials vary in the range of 0.3-17 ng g-1 (as Pb) except for the Antarctic Krill (MURST-ISS-A2), where the concentration was less than the detection limit of 0.09 ng g-1, which was also found for the sediment. Up to 20% of total lead was methylated in the biological reference materials, whereas much higher methylation fractions were found for mercury. The method was also applied to seafood samples purchased from a supermarket with Me3Pb+ concentrations in the limited range of 0.3-0.7 ng g-1. On the contrary, the portion of methylated lead in these samples varied over more than 2 orders of magnitude from 0.02 to 7.5%. Tetramethyllead (Me4Pb), tetraethyllead (Et4Pb), and the corresponding mixed methyl-ethyl compounds were used for a long time in Western Europe and the United States as antiknock additives in petrol. However, leaded gasoline is still used in many countries, particularly in Africa and Asia, and this causes a health risk to the local population and ecosystem.1 These volatile compounds undergo degradation processes in the atmosphere, especially by reaction with OH radicals forming inorganic lead but also trialkyllead as a relatively stable species. * To whom correspondence should be addressed. E-mail:
[email protected]. 10.1021/ac048757m CCC: $30.25 Published on Web 12/13/2004
© 2005 American Chemical Society
Besides anthropogenic sources, trimethyllead (Me3Pb+) is also produced by natural biomethylation, for example, in seawater by algae or bacteria.2,3 Similar to monomethylmercury (MeHg+), trimethyllead is expected to be accumulated in marine animals and can therefore also reach the food chain of man, which makes an accurate determination of Me3Pb+ of special importance. Organolead compounds are known to be more toxic than inorganic lead. LD50 values of different alkyllead species in comparison with methylmercury show that trialkyllead species are highly toxic and that the toxicity is in the same range as for methylmercury.1,4 Because of the usually low concentration levels of trimethyllead in environmental and biological samples, sensitive analytical methods must be applied. Analytical methods for alkyllead determinations described in the literature include different sample preparation steps, for example, derivatization, extraction, and chromatographic separation.5,6 Everyone of these analytical steps can affect the accuracy of the analytical result, where transformation is especially risky in the case of elemental speciation.7 Recently, Jitaru et al. published a headspace solid-phase microextraction (SPME) method in combination with multicapillary gas chromatography (MCGC) hyphenated to ICP-TOFMS.8 This method allowed rapid and simultaneous determinations of different alkylated lead and tin compounds and of methylmercury. However, the authors stated that further improvements are expected with respect to the accuracy and the evaluation of possible species transformations by using the species-specific isotope dilution mass spectrometric (IDMS) technique. To validate analytical methods, certified reference materials (CRM) are usually applied. However, there is a lack of CRMs for alkyllead compounds for which only one reference material (urban dust) exists. Doubts on the longterm stability of trimethyllead in an artificial rainwater candidate (1) Yoshinaga, J. In Organometallic Compounds in the Environment; Craig, P. J., Ed.; Wiley & Sons: West Sussex, England, 2003; pp 151-194. (2) Pongratz, R.; Heumann, K. G. Chemosphere 1998, 36, 1935-1946. (3) Pongratz, R.; Heumann, K. G. Chemosphere 1999, 39, 89-102. (4) Hewitt, C. N.; Harrison, R. M. In Organometallic Compounds in the Environment; Craig, P. J., Ed.; Longman: Essex, England, 1986; pp 160197. (5) Quevauviller, P.; Ebdon, L.; Harrison, R. M.; Wang, Y. Analyst 1998, 123, 971-976. (6) Quevauviller, P.; Harrison, R.; Adams, F.; Ebdon, L. Trends Anal. Chem. 2000, 19, 195-199. (7) Demuth, N.; Heumann, K. G. Anal. Chem. 2001, 73, 4020-4027. (8) Jitaru, P.; Goenaga Infante, H.; Adams, F. J. Anal. At. Spectrom. 2004, 19, 867-875.
Analytical Chemistry, Vol. 77, No. 2, January 15, 2005 511
reference material did not allow its certification.5 IDMS with its accepted high accuracy and its special properties to validate analytical speciation procedures is an important alternative to the use of CRMs. A species-specific HPLC/ICP-IDMS method therefore had been developed to determine trimethyllead in the mentioned artificial rainwater candidate reference material.9,10 A detection limit of 3 ng g-1 was reached with this method.10 Because trimethyllead concentrations below the nanogram per gram range must be expected in many environmental and biological samples, an improvement of the detection limit by also producing reliable results was necessary. GC/ICPMS coupling is known for its excellent sensitivity, which has been demonstrated for the determination of a couple of organometallic compounds.11 In addition, descriptions of GC/ICP-IDMS methods also exist, e.g., in ref 12 which have been exclusively focused on the speciation of mercury, selenium, and tin. There is lack of a corresponding method for the determination of trimethyllead. A sensitive GC/ ICP-IDMS method was therefore developed in this work for accurate and precise determinations of trimethyllead in environmental and biological samples that can be used down to the subnanogram per gram level. EXPERIMENTAL SECTION Chemicals. Me3PbCl (98%) was obtained from ABCR (Karlsruhe, Germany). A stock solution was prepared by dissolving this compound in 0.5% (by weight) HNO3 and stored in the dark at 4 °C. Diluted solutions in MQ water, adapted to the necessary concentration for analysis, were prepared prior to each series of measurements. (Safety note: Alkyllead compounds are extremely toxic. Gloves, for example, Silvershield, must be used and good ventilation in the working area is absolutely necessary.) Metallic lead, enriched in 206Pb, was obtained from Euriso-top (Saint Aubin Cedex, France). Methyllithium (1.6 mol L-1 solution in diethyl ether) and iodomethane (99%) for synthesis of the trimethyllead spike were obtained from Acros Organics (Geel, Belgium). Derivatization was carried out with sodium tetraethylborate (Galab Technologies, Geesthacht, Germany). Tetramethylammonium hydroxide (TMAH) solution in water (25 wt %) and hexane (purris. g99.5%) were purchased from Fluka (Buchs, Switzerland). Sodium acetate and acetic acid (Suprapure, Merck, Darmstadt, Germany) were mixed to receive a 4 mol L-1 acetate buffer at pH 5. Ultrapure water was obtained from a Milli-Q system (Millipore, Eschborn, Germany). HNO3 (p.a., Acros Organics) was purified by distillation under sub-boiling conditions in a quartz apparatus. Samples. The only reference material certified for trimethyllead at the moment is an urban dust (CRM 605) with a certified value of 7.9 ( 1.2 µg kg-1 (as Me3Pb+). Other reference materials, which are not certified for trimethyllead, were analyzed for this lead species by GC/ICP-IDMS (Table 1). In addition, mussels, large peeled prawns, fillets of a tuna fish, a plaice, and a pollock were bought in a supermarket. These samples were freeze-dried directly after purchase, then milled, and stored in polyethylene bottles in the dark. The water content of these samples, deter(9) Brown, A. A.; Ebdon, L.; Hill, S. J. Anal. Chim. Acta 1994, 286, 391-399. (10) Ebdon, L.; Hill, S. J.; Rivas, C. Spectrochim. Acta B 1998, 53, 289-297. (11) Prange, A.; Jantzen, E. J. Anal. At. Spectrom. 1995, 10, 105-109. (12) Garcia Alonso, J. I.; Ruiz Encinar, J. In Handbook of Elemental Speciation: Techniques and Methodology; Cornelis, R., Caruso, J., Crews, H., Heumann, K. G., Eds.; Wiley: London, 2003; pp 163-200.
512 Analytical Chemistry, Vol. 77, No. 2, January 15, 2005
Table 1. Reference Materials, Not Certified for Trimethyllead, for Which Me3Pb+ Is Determined by Species-Specific GC/ICP-IDMS material code
matrix
DORM 2
dogfish muscle
CRM 463
tuna fish
CRM 422 CRM 477 CRM 278 MURST-ISS-A2 CRM 580
cod muscle mussel tissue mussel tissue Antarctic krill sediment
certified compound trace elements and methylmercury total mercury and methylmercury trace elements butyltin compds trace elements trace elements total mercury and methylmercury
agencya NRCC BCR BCR BCR BCR ISS BCR
a NRCC, National Research Council, Canada; BCR, Community Bureau of Reference, Brussels; ISS, National Institute of Health, Italy.
mined as mass difference before and after the freeze-drying procedure, was in the range of 75-84%. The humidity of the reference materials was also determined by drying 100-mg portions at 105 °C for 24-30 h until a constancy in the weight. The water content in the reference materials varied from 2.3% in CRM 605 (urban dust) to 12.5% in CRM 463 (tuna fish). The results of alkylated metals in this work are always related to the dry sample material. Instrumentation. A gas chromatograph (model HP 6890, Agilent Technologies, Wilmington, DE), fitted with a split/splitless injector and a HP-1 capillary column (cross-linked methyl siloxane; 21-m length, 0.32-mm i.d., and 0.17-µm film thickness), was used for separation of the volatile alkyllead compounds after derivatization. The gas chromatograph was coupled to an ICPMS (model HP 4500, Agilent Technologies) via a homemade transfer line. The transfer line is made of a Silcosteel-coated stainless steel tube (Resteck; 0.51-mm i.d., 1.59-mm o.d.) directly heated by an electrical transformer (type 3234D, Statron, Fu¨rstenwalde, Germany) in order to avoid broadening of separated peaks. For safety reasons, the transfer line is covered with a PTFE tube. All connections are sealed by metal or graphite ferrules. The transfer line was coupled with the plasma torch of the ICPMS via a flexible glass interface with a spherical ground joint described by Schwarz and Heumann.13 The ICPMS conditions were optimized in order to obtain maximum sensitivity for the lead isotope ratio measurement of the trimethyllead peak under the dry plasma conditions. Parameters such as radio frequency power, the flow rate of carrier gas, and the torch position have significant influences on the sensitivity. A maximum of the signal intensity for lead isotope ratio measurements of the trimethyllead peak was observed at a rf power of 1200-1400 W (Figure 1). In addition, the signal intensity increased by a factor of 7-8 when the argon carrier flow rate was increased from 1.0 to 1.25 L min-1 at 1300-W rf power. Oxygen additions to the plasma gas can be used, in principle, to avoid carbon depositions on the ICPMS cones, but the signal of trimethyllead decreases dramatically under this condition (Figure 2), so that no oxygen was applied. The optimized operating conditions for the GC/ICPMS system are summarized in Table 2. Lead Isotope Ratio Measurements. The isotope ratio 206Pb/208Pb with 206Pb as the spike and 208Pb as the reference (13) Schwarz, A.; Heumann, K. G. Anal. Bioanal. Chem. 2002, 374, 212-219.
Figure 1. Effect of the rf power (a) and of the argon carrier flow rate (b) on the signal intensity of lead isotopes measured in the trimethyllead peak (10 pg of trimethyllead of natural isotopic composition; measurement at conditions listed in Table 2 if parameter is not varied). Table 2. Operating Conditions of the GC/ICPMS System ICPMS Parameters radio frequency power 1300 W plasma gas flow rate 15 L min-1 carrier gas flow rate 1.25 L min-1 auxiliary gas flow rate 1.20 L min-1 torch position (“sampling depth”) 6 mm integration time per point 0.1 s column carrier gas/flow rate oven program
Figure 2. Influence of oxygen as auxiliary gas on the signal intensity of lead isotope measurements in the trimethyllead peak.
injection mode injection volume injection temperature transfer line
GC Conditions HP-1 (methyl siloxane) He/2 mL min-1 40 °C (2.5 min) f 60 °C (10 °C/min) f 260 °C (30 °C/min, 5.0 min) split/splitless 1 µL 250 °C Silicosteel tube, 180-cm length, 0.51-mm i.d.
isotope was selected for the isotope dilution technique. The isotope ratios of the isotope diluted samples were always determined by an evaluation of the corresponding isotopes in the Me3Pb+ peak of the gas chromatogram. Because of better precisions of the isotope ratio determination, the peak areas and not the peak height were used. Integration of the chromatographic peaks was performed by the ICPMS software. Mass discrimination effects in the ICPMS were compensated by always using the measured isotope abundances or isotope ratios of the spike, the sample, and the isotope-diluted sample for trimethyllead calculations by the IDMS equation.14 In the case of lead, natural isotopic variations can occur by radiogenic production. The isotope abundances of all different samples were therefore determined independently, and the corresponding values were used for IDMS calculations instead of the IUPAC data. Synthesis of 206Pb-Enriched Trimethyllead Spike. A 110mg sample of 206Pb-enriched metallic lead was dissolved under heating in an oil bath (150 °C) with 7 mL of 47% HBr. After evaporation of the acid, the precipitate was washed once with ethanol and twice with diethyl ether and then dried under vacuum conditions. Under argon atmosphere, the corresponding PbBr2 product was suspended in 10 mL of ether, which included an excess of iodomethane (MeI) compared to the stoichiometric reaction (0.4 g). (Safety note: Iodomethane is highly toxic because it may cause cancer. It is fatal if this substance is inhaled, swallowed, or absorbed by the skin. Therefore, safety glasses,
gloves, and a good ventilation at the working place are necessary.) A 2.4-mL aliquot of 1.6 mol L-1 methyllithium (MeLi) solution in ether was then dropwise added under intensive stirring and icewater cooling. After an additional 5 h of stirring at room temperature, the ether phase was washed with water and then dried over anhydrous sodium sulfate. The ether phase was placed in a dry ice-acetone bath (-60 °C) and 0.13 g of I2, dissolved in ether, was slowly added under stirring within 5 h. After evaporation of ether, the precipitate was air-dried under cooling. Drying under vacuum conditions should be avoided because this causes degradation of trimethyllead to lead iodide. The yield was 185 mg of 206Pb-enriched product, which was dissolved in 250 mL of 0.5% HNO3. This stock spike solution was diluted prior to analyses in a way that the corresponding concentration fits the optimum range of spike addition for the isotope dilution technique.15 The isotopic analysis of the spike solution was performed by GC/ICPMS after ethylation with NaBEt4. Figure 3 shows the GC/ ICPMS chromatograms obtained for 206Pb and 208Pb for a Me3Pb+ standard solution of natural isotopic composition and of the spike, respectively. Besides the Me3EtPb peak, caused by ethylation of trimethyllead, a second small peak was observed in the spike chromatogram that corresponds to Pb2+ derivatized by NaBEt4 to Et4Pb. The isotope abundances of trimethyllead in the 206Pb-enriched spike solution measured by GC/ICPMS are as follows: 204Pb, 0.03 ( 0.02%; 206Pb, 99.28 ( 0.04%; 207Pb, 0.41 (
(14) Heumann, K. G.; Gallus S. M.; Ra¨dlinger G.; Vogl J. J. Anal. At. Spectrom. 1998, 13, 1001-1008.
(15) Heumann, K. G. In Inorganic Mass Spectrometry; Adams, F., Gijbels, R., van Grieken, R., Eds.; Wiley: London, 1988; pp 301-376.
Analytical Chemistry, Vol. 77, No. 2, January 15, 2005
513
Figure 4. Schematic diagram of the sample preparation procedure for trimethyllead determinations in environmental and biological samples by species-specific GC/ICP-IDMS.
Figure 3. GC/ICPMS chromatograms of 206Pb and 208Pb (a) of a Me3PbCl standard solution of natural isotopic composition and (b) of the synthesized spike solution.
0.03%; 208Pb, 0.28 ( 0.04%. Reverse isotope dilution analyses, using trimethyllead and Pb2+ standard solutions of natural isotopic abundances, were carried out to determine the concentration of trimethyllead and inorganic lead in the spike solution. Three independent analyses resulted in concentrations of 309.4 ( 2.2 µg g-1 for Me3Pb+ and 63.4 ( 2.2 µg g-1 for Pb2+ (both as Pb). Taking into account that the spike contains some inorganic lead iodide, the total yield for the synthesis of the 206Pb-enriched trimethyllead iodide was 75%. The peak areas of the two lead species in Figure 3 do not represent the determined concentration ratio because the inorganic lead peak is much too small. This is due to an incomplete derivatization of Pb2+ by NaBEt4. However, this has no influence on the accuracy of the species-specific GC/ ICP-IDMS determination because only the isotope ratio of the corresponding isotope-diluted species represents the exact amount of a species. Only the precision is little affected by the low signal intensities of the Pb2+ peak. The long-term stability of a diluted portion of the stock spike solution (480 ng g-1) was investigated. No significant change in its concentration and isotopic composition was found within 10 months of storage at 4 °C in the dark. Sample Treatment. The sample treatment procedure for the determination of trimethyllead in road dust was adapted from the work of Witte et al.,16 which was optimized for the isotope dilution technique. The recommended minimum sample intake of 1 g was weighed together with 0.3-0.4 g of a diluted stock spike solution (usually in the range of∼10 ng g-1, but exactly determined) into (16) Witte, C.; Szpunar-Lobinska, J.; Lobinski, R.; Adams, F. C. Appl. Organomet. Chem. 1994, 8, 621-627.
514 Analytical Chemistry, Vol. 77, No. 2, January 15, 2005
a 7-mL vial with screw cap. A 3-mL aliquot of an extraction reagent, which consists of a citrate buffer (pH 7-8), EDTA, and sodium diethyldithiocarbamate, was added. The mixture was shaken for 2 h. After addition of 500 µL of hexane and 500 µL of 2% aqueous NaBEt4 solution, the mixture was shaken again for 10 min and then centrifuged (4000 min-1) to facilitate phase separation. The yellow hexane phase was collected with a micropipet and transferred to a 2-mL glass vial. After injection of 1 µL into the GC (split ratio 10:1 or 20:1), the GC chromatograms of 206Pb and 208Pb were detected by ICPMS. In the case of biological samples, 0.2-1 g of the sample and 0.1-0.3 g of the diluted stock spike solution were also weighed into a 7-mL vial and mixed with 3-4 mL of TMAH (25% in water). The mixture was shaken for 2-3 h at room temperature in order to dissolve the organic matrix. Sample decomposition can also be carried out by TMAH in a few minutes using microwaveassisted treatment.17 However, in this case, the single-used 7-mL vials cannot be applied. A 0.6-0.8-mL sample of concentrated HNO3 and 0.5 mL of 4 mol L-1 acetate buffer were then added to obtain pH 5-6. Ethylation and the GC/ICPMS measurement were similar to the dust sample treatment except that a split ratio of 1:1 or a splitless mode was applied due to the lower Me3Pb+ concentrations in these samples. The schematic sample preparation procedure for the determination of trimethyllead in environmental as well as in biological samplessor more general, in inorganic and organic matrixessby species-specific GC/ICP-IDMS is presented in Figure 4. The total lead content in biological samples was determined by ICP-IDMS after microwave-assisted treatment. A 0.2-0.3-g aliquot of the sample and 0.1-0.6 g of a diluted 206Pb-enriched Pb2+ spike solution were accurately weighed into the Teflon microwave vessel and treated with 5 mL of 65% HNO3 and 2 mL of 30% H2O2. After digestion (2.3 min at 75 °C, 5.3 min at 130 °C, 24 min at 210 °C), the Teflon vessel was opened and the solution was diluted with Milli-Q water (1:5 or 1:100 depending on the lead concentration) and measured by ICPMS using a PFA microflow nebulizer. This method was validated by the reference material DORM 2 (dogfish muscle), for which the total lead content of 61 ( 6 ng g-1 by ICP-IDMS was in good agreement with the certified value of 65 ( 7 ng g-1. (17) Heumann, K. G. Anal. Bioanal. Chem. 2004, 378, 318-329.
Table 3. Trimethyllead Concentrations of Reference Materials Determined by Species-Specific GC/ICP-IDMS Compared with Total Lead, Methylmercury, and Total Mercury Concentrations methylated species reference material DORM 2 CRM 463 CRM 422 CRM 477 CRM 278 MURST-ISS-A2 CRM 580 a
ng
MeHg+, g-1 (as Hg)
4470 ( 320a 2830 ( 150a 400 ( 20d ndc 130 ( 50d nd 70.2 ( 3.4a
total Pb+,
Me3 ng g-1 (as Pb)
Hg, ng g-1
Pb, ng g-1
6.41 ( 0.29b 4.39 ( 0.16b 16.74 ( 0.19b 0.30 ( 0.02b 2.72 ( 0.06b
