Critical review of analytical methods for determination of inorganic mercury and methylmercury compounds.код для вставкиСкачать
APPLIED ORGANOMETALLIC CHEMISTRY, VOL. 8, 293-302 (1994) REVIEW Critical Review of Analytical Methods for Determination of Inorganic Mercury and Methylmercury Compounds Richard Puk and James H. Weber* University of New Hampshire, Chemistry Department, Parsons Hall, Durham, NH 03824, USA This review describes determinations of mercury compounds under three categories: total mercury; separate determinations of inorganic mercury(I1) and organomercury compounds by selective reduction; and speciation of inorganic mercury(II), monomethylmercury cation, and dimethylmercury. Topics described for each category include sample treatment, separation, detection, and limit of detection. Finally, we note that most methods would not detect dimethylmercury if it were present. Keywords: Analytical methods, mercury, monomethylmercury, dimethylmercury ETAA EXT GC GFAA HCL Hg" Hg(I1) Hg,,, HOM HPLC ICP LOD MeHg Me,Hg MIP ox NOTATl0N AA AF APDC Carbo trap cc Conc Cry-PC CVAA CVAF DER DIG DME DMF ECD Et,Hg Atomic absorption Atomic fluorescence Ammonium pyrrolidine dithiocarbamate Graphitized carbon-black column Column chromatography Concentrated Cryogenic packed column Cold vapor atomic absorption Cold vapor atomic fluorescence Derivatization Digestion Dropping mercury electrode Dimethylformamide Electron capture detector Diet h ylmercury ' Author to whom correspondence should be addressed. CCC 02hX-2605/94/040293- 10 0 1994 by John Wiley & Sons. Ltd PRE RED RHg RSD SAMP STOR Electrothermally heated atomic absorption Extraction Gas chromatography Graphite furnace atomic absor;tior. Hollow-cathode lamp Elemental mercury Mercury(I1) Total mercury Homogenization High-performance liquid chromatography Inductively coupled plasma Limit of detection Monomethylmercury cation Dimethylmercury Microwave-induced plasma detector Oxidation Pretreatment Reduction Organomercury (R = M e or Me,) Relative standard deviation Sample Sample storage GENERAL INTRODUCTION Long-term and increasing interest in the speciation of inorganic mercury [ Hg(ll)], monomethylmercury cation (MeHg), and occasionally dimethylmercury (MezHg) in environmental and biological samples has resulted in a large number of published analytical methods. A recent interlaboratory study on the determination of methylmercury (MeHg) in fish and mussels' clearly shows the importance of a methods review. The Receioed 16 Fehruury I994 Accepied I April 1994 R. PUK AND J. H. WEBER 294 Table 1 Scheme for the determination of mercury compounds. ~ Sample treat men t Homogenization Digestion Oxidation Extraction Reduction Derivatization DETERMINATION OF TOTAL MERCURY Introduction Separation Detection Gold GC HPLC Cry-PC Carbo trap CVAA WAF ECD GFAA MIP ETAA DME cc results were a relative standard deviation between laboratories in the range of 20-25%, indicating a lack of reproducibility of existing methods and the urgent need for improvements in MeHg determinations. It is clearly impossible for us in a short review to describe hundreds of publications using dozens of analytical methods and their variations. We will summarize only major themes and organize existing methods into three major categories: (1) determination of total mercury (Hgt,,,), (2) separate determination of Hg(I1) and organomercury compounds (RHg) by selective reduction; and (3) speciation of Hg(II), MeHg, and MezHg. Within each category we emphasize originators of methods and newest applications. Most environmental samples require a sequence of three analytical steps (Table 1) namely sample treatment, separation and detection '.' Each step is futher divided into one or more specific techniques depending on the mercury compound(s) of interest. Tables 2-4 list analytical methods for determining Hg(I1) and methylmercury compounds. They include sample type and treatment, separation method, detection method, and limit of detection (LOD). LOD is the lowest concentration that is statistically different from the blank." ' The tables list LOD based on the mass of mercury per sample weight (fresh weight or dry weight are not always mentioned) or volume. Missing information in some papers made it impossible to calculate LOD as the absolute mass of mercury in samples. Thc first entry in Table 2 exemplifies the interpretation of Tables 2-4. Hatch and Ott' determined Hg,,,, in metal, rock, or soil by a sample treatment of digestion and reduction followed by detection by cold vapor atomic absorption (CVAA). Their method has an LOD of 1 ng Hg g - ' . Determination of Hg,,,, (Table 2) requires conversion of all forms of mercury to Hg(I1). Since samples may contain Hg(I1) bound to other molecules such as proteins o r humic matter, it must be liberated from ligands present. In addition, sample treatment must convert bound or free MeHg and Me2Hg to free Hg(I1). Researchers usually convert all mercury compounds to free Hg(1I) by an acidic, oxidative digestion (Eqns [l] and ). DIG/OX Bound Hg(I1)-Free RED Hg(1I)-Hg" DIG/OX MeHg(or Me2Hg)-Free [ 11 RED H g ( I I ) - - H g "  The final two steps are reduction of Hg(I1) to Hg" by Sn(I1) or NaBH, and detection of Hg", typically by atomic absorption (AA) or atomic fluorescence (AF). Sample treatment Researchers have digested and clxidized samples in various ways to release mercury compounds from binding substances, or have sometimes omitted this ~ t e p . Bricker9 ~.~ oniitted the digestion and storage steps by using a field method for reduction and volatilization of mcrcury and purging Hg" onto a gold column. Match and Ott" digested samples with the oxidizing acid HNO,. Other researchers used acid digestion and oxidation."'-'h with or without heating. The digestion procedure (if any) is followed by reduction of Hg(I1) with tin(II)'? (Eqn ). '' Hg(I1) + Sn(II)-tHg''+ Sn(lV) [31 Hatch and Ott,h Rezende et ul.," Bloom and Crecelius"' I t and Robinson and !$humani5added hydroxlyamine (NH20H) as a preliminary reducing agent before adding the tin(I1) solution. NaBH, also reduces Hg(1I) to Hg" (Eqn ). Hg(I1) + 2NaBH, + 6 H 1 0 - +Hg"+ 7H2 + 2H,BO, + 2Na' 141 Bricker" and Tsalev et ul.'" usecl NaBH, alone. Two group^^.^ used a transition-metal catalyst with NaBH,. DETERMINATION OF INORGANIC MERCURY AND METHYLMERCURY 295 Table 2 Method summary for the determination of total mercury (Hg,,,J Sample typc and treatment Separation Detection LOD Reference 6 SAMP: DIG: RED: Metal, rock, soil 7 M HNO1/4.S M HZSO, NH,OH/NaCI/Sn(lI) Air purge None CVAA 1 ngg-' SAMP: DIG: RED: Urine None N:IBH,/CLI(II)/~H6.5 buffer Ar purge None ETAA 1-2 ng mt SAMP: River water, rainwater, pond water, sewage effluent None NaBH, (in the field) Air purge Gold Helium-dc plasma emission SAMP: DIG: RED: Water, waste water None NaBH,IFe(III)/S M HCI Ar purge Gold CVAA SAMP: DIG: Fish Hot cone HNO;/H,SO, (reflux) 0.2 M BrCl Sn(1l) He purge Gold CVAF 1.0ngg Urine, river, lake water. rain water Microwave (SO-90 "C) KBrO,/KBr/HCI NaBH,/O. I M HCI Ar purge Gold or without Gold CVAA 0.0 1 ng ml I (with gold) 0.2 ng ml Seawater. sediment, sewage effluent 2% HNO, Conc HNO,/H,SO, (sediment only) 0.2 M BrCl NH,OH/Sn(lI) N, purge Gold CVAA Stream water. river water HNO,/K,Cr,O,/cysteine Hot (60°C) HISOJHNO? KMnO,/ K&O, NH,OH/Sn(lI) N ? purge Gold CVAA 0.02 ng ml Fish HNOJHTSO, (4.5 M) KMnO, NH,OH.HCI/Sn(II) in IS'%, HCI Air purge None CVAA 2Sngg DIG: RED: ox: RED: SAMP: DIG/OX: RED: SAMP: STOR: DIG: ox: RED: SAMP: STOR: DIG: ox: RED: SAMP: DIG: ox: RED: ~ 8 ' 13. 14 (without gold) 10. I I I I 1s 12 R . PUK A N D J. H. WEBER 296 Table 3 Method summary lor the determination of (Hg(I1) and RHg by selective reduction - Sample typc and treiitment Fish, niisc. biological samples 35'% NaOH/ I '%cysteind 20%) NaCI/ 100 "C Separation Detection LOD Relerence None CVAA 1 0 ng mt None CVAA ~).003-0.0~)5 ng rid None CVAA 25ngg 22 I Sn(II)/X M H?SO,/cysteine/NaCI Air purge (NaOH added) KH s RED: Sn(ll)-Cd(ll)/X M HISO,/cystcine/NaCI Air purge (NaOlI added) H g ( I I ) + R H g= Hg,,,, SAMP: DIG: OX: Txp water, tuna, hair, urine 10 M KOH (90 "C) 0.24 M HNO,/O.Ol'% K,CrlO,/ 1 '%NaCl ' Sn( II)/HNO:/K,Cr20, N? purge R I i(: RED: NsBH,/HNOJ K?Cr@, N? purge lish water/4.5 M H:SO,/KBr ' wiiter phase after extraction with CHCli/NaHH, Air purge ,is McHgBr i n t o RED: CHCI, NaBH,/DMF/HNO, Air purge Hg,,,,- Iig(lI) = RHg SAMP: SI'OR: I i S ("1 RED: Sn(ll)/O.OS M HCL NaBH,: mi ' 0. 192 iig 21 DETERMINATION OF INORGANIC MERCURY AND METHYLMERCURY 291 Table 3 continued ~~~~~ Sample type and treatment ~ Separation Detection LOD Reference standards, tap water, river water 0.16 M HN03/KMn04 Sn(II)/O.12 M HCI Air purge None CVAA 0.001 ng mt-' 2s seawater, rain water, estuarine water, river water None 0.06 M HNO, Sn(I1) N, purge Gold CVAA 0.042 ng m1-I 24 Hg(1t)~' SAMP: ox: RED: H g (1)' SAMP: DIG: STOR: RED: Hg,,,,- Hg(I1) = MeHg: this Hg(I1) method is used with a Hg,,,, method in Table 1 to determine MeHg by difference. Separation/concentration The high 1.2 x lo-' mm Hg vapor pressure of Hgo at 20°C simplifies its detection by AA. Hgo formed during the reduction step is often concentrated on a gold ~ o I u m n ~ - I I ,and ~ ~ -then ' ~ thermally desorbed prior to detection. Water, volatile organic compounds, sulfides or CI, can sometimes interfere with the amalgamation of Hg".''~" Concentration of Hg' on the gold column can decrease LOD by 20-fold. For example, Tsalev et a/." had an LOD of 0.2ngml-I without a gold column and of 0.01 ng ml-l with it. Detection M~~~ groups6. 8 . 10. 12. 15. 16 used CVAA for the detection of Hg". CVAA avoids problems with nebulization and atomization that occur in classical flame AA." Cold vapor atomic fluorescence (CVAF) detection for Hg" often has improved the LOD relative to A A detection.13 Other types of detection include a electrothermally heated atomic absorption cell (ETAA)' and helium-dc plasma emission.' DETERMINATION OF Hg(ll) AND RHg BY SELECTIVE REDUCTION Introduction Separate determinations of Hg(I1) and RHg (Table 3 ) combine two reactions that are carried out sequentially on one or more sample aliquots. The first reaction is typically reduction by the mild reducing agent tin(I1) that reduces free Hg(1I) to Hg* (Eqn [ 3 ] ) , but not C-Hg bonds in RHg. After complete purging of the resulting Hgo, the same aliquot is treated under acid and oxidizing conditions to break C-Hg bonds in RHg and form Hg(I1) which is reduced to Hg" (Eqn ). The separation and detection steps are similar to those described above. A disadvantage of selective reduction techniques is the impossibility of confirming the identity of RHg. For example, the selective reduction method does not permit researchers to distinguish MeHg from Me2Hg. The typical assumption is that RHg is predominant MeHg since it is the nearly exclusive form of RHg in fish.I4 Researchers have used selective reduction on samples such as fish,'2.20-'2 hair and urine," animals,, and blood.23 Sample treatment for mercury(l1) The sample treatment used to determine Hg(I1) should separate free Hg(I1) from any chemical that binds it without breaking C-Hg bonds. Digestion with NaOH and cysteine" or KOH" leaves the C-Hg bond intact but frees any bound Hg(I1). Determination of Hg(I1) in aquatic matrices involves sample treatment with tin(I1) in aqueous dilute hydrochloric acid (HCI) to reduce Hg(I1) to Hg".24,25HCI prevents adsorption of mercury compounds on the sam le container Pused NaBH4 prior to reduction. Rezende ef u1.Irather than tin(I1) with air purging to reduce the Hg(I1) in the aqueous phase to Hg". R. PUK AND J. H. WEBER 298 Table4 Method summary for speciating Hg(I1). M eHg and Me,Hg ~ ~~~ ~ Sample type and treatment Separation Detection LOD Reference MeHg SAMP: DIG: HOM: EXT: Fish, eggs, meat, liver None WaterIHCI Benzene, cysteine, HCI, benzene GC ECD 70-400 ng g -I 27, 28 MeHg SAMP: DIG: DER: Fish, biological tissue Conc H2S04 Iodoacetic acid GC MIP 20 ng g-' 36 Fish Water16 M HCV Celite 545 Elute with CC14, Na2S203(to eluate) Air or N2 purge HPLC HeatICVAA or DME 0.37-0.6 ng g-' 33 Fish KOHlMeOH (70 "C)/neutralize NaBEt,/pH 4.5 acetate buffer N, or air purge to Carbo trap He purge to Cry-PC Carbo trap/ Cry-PC CVAF MeHg: 0.5 rig g - ' Me2Hg:0.1 ng g- ' 13, 14 MeHg SAMP: DIG: HOM: EXT: Fish, eggs, meat, liver None WaterIHCI Benzene, cysteine, HCI, benzene GC ECD 70-400 ng g 27,28 Hg(ll), MeHg SAMP: HOM: Fish, biological materials WaterlNaClIl M HCI GFAA 3.0 ng m L - ' 34 NaBH,ICVAA (on eluted MeHg: 0.5 ng ml-' Hg(I1): 0.015 ng ml-' 43 MeHg SAMP: HOM: EXT: MeHg, Me,Hg SAMP: DIG: RED: Hg(W DERIEXT: I Me,Sn/MeOH (100 "C) Benzene/Na2SZ03/ benzene MeHg EXT: Benzene, Na2S203/Cu(II)Ibenzene H g ( l l ) , MeHg SAMP: DER: Standards, tap water Hg-APDC complex formed on RED: column Elute with APDC Eluent with NaBH, HPLC sample) DETERMINATION O F INORGANIC MERCURY AND METHYLMERCURY ~ - ~~~ 299 ~~ Table 4 continued Sample type and treatment Separation Hg(ll), MeHg SAMP: River water, tap water cc STOR: 0.08 M HNO, Hg(lI) Pre: Pass through sulfhydryl cotton column with 0.01 M HCI ox: KBr/KBrO, RED: Eluate with Sn(II)/HCI MeHg PRE: RED: Elute column with 3 M HCI KBr/KBrO, Sn(II)/HCI Me2Hg SAMP: Seawater ox: H g ( I l ) , M e H g , MerHg SAMP: Fish, seawater HOM: KOH/MeOH RED: NaBEt,/pH 4.5 acetate buffer He purge M e H g , Me,Hg SAMP: Sediment EXT: Acetic acid DER: NaBH,/acetic acid (pH 3.5) He purge LOD Reference CVAF 6 ng g-' 35 Carbo trap/Cry-PC HeatICVAF GoldlCarbo trap/ Cry-PC Heat/CVAF 13 Seawater: 0.2 ng I- I (Hg(lI), 0.003 ng I-' (MeHg and MeZHg) 13, 14 Fish: 0.5 ngg-' (MeHg), 0.1 ngg-l (Me&) Cry-PC Hg(Il), MeHg SAMP: Fish, lobster Cry-PC DIG: KOH/MeOH DER: NaBEtJpH 4.5 acetate buffer He purge H g ( I I ) , M e H g , MerHg SAMP: S . alterniflora, eelgrass EXT: 0.1 M HCI, MeOH DERIRED: NaBHJ0.01 M HCI He purge Detection Cry-PC Hz/Oz/ETAA 39 ETAA MeHg: 4 ng g-' Hg(I1): 75 ng g-' 44 ETAA Hg(I1): 0.11 ngg-' 42 MeHg: 0.05 ng g- Sample treatment for RHg Separation and detection Under most conditions, NaBH, does not reduce MeHg. However, NaBH, and air purge in the presence of dimethylformamide,'2 NaBH, and nitric acid," or tin(1I) plus metal-ion c a t a l y d 2 reduce RHg left in the sample aliquot after the removal of Hg( 11). A separation using a gold trap is generally unnecessary since both Hg(I1) and RHg are separately reduced to volatile Hg".". 'I." However, Gill and Bruland'h and Gill and F i t ~ g e r a l ddid ~ ~use a gold trap. The usual detectors for H ~ I I are CVAAI?. 21, ??. 24.25 and CVAF.2h R. PUK AND J. H. WEBER 300 SPECIATION OF Hg(ll), MeHg AND Introduction Speciation of Hg(II), MeHg, and Me,Hg (Table 4) generally requires extraction and derivatization (not all methods), separation and concentration (not all methods), and detection. This speciation section differs from the selective reduction section (Table 3) by identifying RHg compounds and confirming that the organomercury compound most commonly observed in the environment is MeHg. Sample treatment Extraction Hundreds of papers on the determination of MeHg in the environment have appeared during the past 25 years. Most researchers use a variation of the original extraction method reported by Westooj.?l.?X Researchers have applied the method to a wide variety of sample types including fish, food, seawater, sediment, blood and urine. One goal of these extraction methods is to separate Hg(I1) from MeHg. A second goal is to concentrate MeHg and separate it from chemicals such as proteins and humic matter to avoid interferences during the detection step. Extraction methods are commonly used before separation and detection by gas chromatography (GC).2X.2y In the W e ~ t o o ~ ’ method .~’ samples are typically homogenized in water, acidified with hydrochloric acid and treated with benzene to extract MeHgCl from the aqueous phase into benzene. After extraction of MeHgCl from the benzene phase with aqueous cysteine, the intitial benzene layer is discarded. The aqueous solution is acidified to break up the cysteine-MeHg complex and MeHg is again extracted into benzene. Many determinations of MeHg in environmental samples are performed by variations of the Westoo extraction method in which bound Hg(I1) and MeHg are converted to free forms using HCI or HBr .30. 31 Free MeHg halide is extracted into an organic solvent such as benzene,2X di~hloromethane’~,” or CC14.33MeHg halide is extracted from the organic phase by aqueous or cysteine.2X.2y Other sample thiosulfate treatments requiring separation have been accomplished on microcolumns of sulfhydryl cotton,”” which binds MeHg but not Hg(I1). The eluted Hg(I1) is treated with an oxidizing agent and reduced to Hg” with tin(I1). MeHg is eluted from the column with 3 M HCI, an oxidizing agent is added to the solution, and Hgi:II) is reduced to Hg” with tin(I1) before its detection. Derivatization Some researchers have emphasized methods of volatizing MeHg to avoid extractions. For example, Lansens and B a e y e n ~anti ~ ~ Decadt et a!.’’ separated MeHg from biological tissue by treatment with concentrated sulfuric acid in a closed vial and conversion of MeHg into volatile MeHgI by addition of iodoacetic acid. Bloom‘3 developed based on ethylation of Hg(I1) and MeHg to produce volatile compounds. The digested sample is reacted with sodium tetraethylborate (NaBEt,) to convert MeHg to methylethylmercury (MeEtHg) (Eqn [ 5 ] ) and Hg(I1) into diethylmercury (Et,Hg) (Eqn [61). MeHg’ Hg(I1)’’ + NaBEt,+ + ‘BEt,’ + Na’ EtzHg+ 2 BEt,’ + 2Na’ MeEtHg + 2NaBEt,+   In both equations ‘BEt,’ represents unstable BEt, which reacts with air and water. More recently two g r o ~ p s ’ ~ -reported ~’ a hydride derivatization method in which NaBH, converts MeHg into volatile MeHgH (Eqn ). + NaBH, + 3HzO+ h4eHgH + 3H2 + H,B03 + Na’ [71 MeHg’ NaBH, also reduces Hg(1I) to Hg” (Eqn ) and Me,Hg is purged unchanged. Quevauviller et af.” used this hydride generation method for detection of MezHg in sediment samples. Puk and Weber42 further developed the method fclr determinations of Hg(II), MeHg, MezHg and EtzHg. Separation by GC, HPLC or cryogenic packed column (Cry-PC) GC2X.29.34 or derivatization followed by GC” is often used for separation of MeHg after extraction. HPLC has an advantage over G C in that formation of volatile derivatives is not necessary .33. 43 The volatile products formed by reactions with NaBEt, or NaBH, can be separated in several ways. H i ’ , Me2Hg, MeEtHg (Eqn IS]) and EtzHg (Eqn ) are separated sequentially with a gra- DETERMINATION OF INORGANIC MERCURY A N D METHYLMERCURY phite carbon column, gold column and Cry-PC"." or Cry-PC alone.14 Hg" (Eqn ), MeHgH (Eqn ), and Me,Hg from the NaBH, reaction can be separated on a Cry-PC.""' Detection MeHg from an extracted sample can be separated by G C and detected by an electron capture detector (ECD)?~. ZL).4i.46 , helium microwave induced plasma emission spe~trometry,~".".'~ mass spectrometry.'" inductively coupled plasma-mass spectometry5"or graphite furnace atomic absorption (GFAA).jJ Hgl can be detected after Cry-PC separation by atomic spectrometry either by CVAFl3 or by ETAA."." CVAA is a sensitive enough method for determining MeHg in environmental samples, provided it is converted into Hg" before detection. The eluate from a high-performance liquid chromatography (HPLC) separatiod3.'" must be reduced by N a B H P or atomized by thermal decomposition33for determination of MeHg and/ or Me,Hg as Hg". Hg" is detected after HPLC by atomic spectrometry using CVAA,33.U atomic emission or CVAF after separation on a microcolumn .35 Critique of the possible presence of Me& Recently Baldi er al." reported formation of Me,Hg from MeHg by sulfate-reducing Desuffouibrio desuffulricans strains. This result suggests that Me,Hg may be more common in the aquatic environment that was generally believed. Despite this possibility, only Mason and Fitzgerald," Quevauviller et af.,'" Puk and Weber" and Bloom13 have observed Me,Hg in environmental samples. Virtually all known procedures for speciating RH, in environmental and biological matrices have emphasized MeHg because MeHg was considered the sole RHg synthesized by bacteria in the aquatic environment. In this section we will attempt to convince the reader that with many common analytical methods for speciation of mercury compounds, Me,Hg would not be observed even if it were present. There for several reasons for non-observance of Me,Hg in environmental samples: (1) Volatilization of Me,Hg may occur during 30 I storage, homogenization, hot digestion or other means of sample preparation. (2) Me,Hg may be lost during extraction of toluene or benene phases with aqueous cysteine or S20:-. We confirmed (unpublished results) that Me,Hg sometimes remains in the original organic phase which is usually discarded. (3) Me,Hg may not be detected when GC-ECD is used. The ECD is very sensitive to compounds like MeHgCl that contain at least one electronegative element, but it is not very sensitive to Me'Hg. In addition, the absence of a retention time from Me,Hg standards would accentuate the difficulties of identifying it. (4) Me,Hg may not be seen because of its high density and low solubility in H 2 0 . We sometimes observed significant amounts of Me,Hg after extraction of plant material with 0.1 M HCl, and sometimes saw none in the same extract. The reason was that Me,Hg was insoluble and at the bottom of the water layer. Addition of MeOH (1 : 1) to the 0.1 M HCI solubilized Me,Hg and allowed its determinati~n.~' ( 5 ) Determination of Me2Hg during extractions requires sufficient acidity to free it from ligands in the matrix, but not enough to convert it to MeHg or Hg(I1). Papers have estimated the maximum concentration of acid that leaves C-Hg bonds intactl3.53".S but the results are inconsistent. Our unpublished work tested the stability of Me,Hg in aqueous HCI solutions while sonicating 1 h at 40°C. Under these conditions Me'Hg is stable in 0.1 M HCI, partially decomposed in 1 M and 3 M HCl, and unstable in 6~ HCI. Many researchers used sufficiently strong acidic solutions for extractions, etc., to decompose Me,Hg. REFERENCES I . P . Ouevauviller. I . Drahaek. H. Muntau. B . Griepink, Appl. Organomer. Cliern. 7. 413 (1993). 2. 0. F. X . Donard and F. M. Martin, Trmds Atid. Chrm. 11- 17 (1992). 3. D. E. Wclls, Mikrochim. Ac/u 109. 13 (1992). 4. L. H. Kcith. W. Crummctt. J . Dcegan Jr. R. A . Lihhy. J . K . Taylor and G . Wcntlcr, A n d . Chem. 55. 2210 (1983). 302 R. PUK. A N D J . H. WEBER 5. Anon. AriuIv.st (London) 112, 199 (1987). 6. R, Hatch and W. L. Ott. Anul. Chem. 40,2085 (1968). 7. J. Toffaletti and J. Savory, Anal. Chem. 47 , 2091 (1975). 8. B. Welz and M. Schubert-Jacobs, Fres. Z . Anal. Chem. 331. 324 (1988). 9 , J. L. Bricker, Anal. Chem. 52, 492 (1980). 10. N. S . Bloom and E. A . Crecelius, Mar. Chem. 14, 49 (1983). I I . N. S. Bloom and E. A . Crecelius, Mar. Chem. 21, 377 (1987). 12. M. R. Rezende, R. C. Compos and A. J. Curtius, J . Anal. Atom. Sprctrom. 8 , 247 (1993). 13. N. S. Bloom. Cun. J . Fish. Ayuat. Sci. 46, 1131 (1989). 14. N. S. Bloom, Cun. J . Fish. Ayuat. Sci. 49, 1010 (1992). 15. K . G. Robinson and M. S. Shuman, In[. J . Enuiron. Anal. Chem. 36. I 1 I (1989). 16. D . L. Tsalev. M . Sperling and B. Welz, Analysf (London) 117. 1729 (1092). 17. M. H . Bothner and D . E. Robertson, Anal. Chem. 47, 592 (1975). 18. W . F. Fitzgerald and G . A. Gill. 51, 1714 (1979). 19. A . M. Ure. Anal. Chim. Acta76, l(1975). 20. K. Fukushi. S . N. Willie and R. E. Sturgeon, A n d . Lelt. 26. 325 (1993). 21. C . E. Oda and J . D . Ingle Jr, Anal. Chem. 53, 2305 (1981). 22. I-. Magos. Analyst (London) 96, 847 (1971). 23. I-. Magos and T. W. Clarkson. J . Assoc. Off. Anal. Chem. 55. 966 (1972). 24. G . A . Gill and W. F. Fitzgerald, Mar. Chem. 20, 227 (1987). 25. J. E. Hawley and J . D. Ingle Jr. Anal. Chem. 47, 719 (1975). 26. G . A . Gill and K. W. Bruland, Enoiron. Sci. Technol. 24, 1392 (1990). 27. G . Wcstoii. Actu Chem. Scand. 20. 2131 (1966). 28. G . Westiiii. Actu Chern. Scund. 21. 1790 (1967). 29. M . Uorvat. A . R . Byrne and K. May, Talantu 37. 207 ( 199(1). 30. C . Mculeman, C. C. Lairio, P. Lansens and W. Baeyens, Wuter Res. 27. 1431 (1993). 31. G. Cerrati, M. Bernhard and J. H. Weber, Appl. Orgunornet. Chem. 6 , 587 (1992). 32. Y. Thibaud and D. Cossa. Appl. Organomet. Chem. 3, 257 (1989). W . Holak, Analyst (London) 107, 1457 (1982). M. Filippelli, Anal. Chem. 59, 1 I6 (1987). W. Jian and C . W . McLeod, Talunra 39, 1537 (1992). P. Lansens and W. Baeyens, Anar. Chim. Acta 228, 93 ( 1990). 37. G . Decadt, W. Baeyens, D . Bradley and L. Goeyens, Anal. Chem. 57, 2788 (1985). 38. M. Filippelli, F. Baldi, F. E. Brinckman and G . J. Olson, Enoiron. Sci. Technol. 26, 1457 (1942). 39. P. Quevauviller, 0. F. X . Donard, J . C . Wasserman, F. M. Martin and J. Schneider, Appl. Organomet Chem. 6 , 221 (1992). 40. P. J . Craig, D. Mennie, N. Ostah, 0. F. X. Donard and F. Martin, Analysl (London) 117, 823 :l992). 41. P. J. Craig, D. Mennie, M. Needham. N. Ostah. 0. F. X. Donard and F. Martin, J . Orgonomet. Chem. 447. 5 (1993). 42. R. Puk and J. H. Weber, Anal. Chim. Acta (1994). in press. 43. C. Sarzanni, G . Sacchero, M. Aceto, 0. Abollino and E. Mentasti, J . Chromatogr. 1992, 626, 151 (1992). 44. R . Fischer. S. Rapsomanikis and M. 0. Andreae. Anal. Chem. 65, 763 (1993). 45. C. J . Cappon and J. C . Smith, Anul. Chem. 49, 365 (1977). 46. V. Zelenko and L. Kosta, Talantu 21). I15 (1973). 47. Y. Talmi, Anal. Chim. Acla 74, I07 (1975). 48. K. Chiba, K. Yoshida, K. Tanabe, 11. Haraguchi and K. Fuwa, Anal. Chem. 55, 450 (1983). 49. B. Johansson, R. Ryhage and G. \Yestiiii. Acta Chem. Scand. 24, 2340 ( 1970). 50. D. Beauchemin, K. W. M. Siu and <. S . Bcrman. Anul. Chem. 60, 2587 (1988). 51. M. Fujita and E. Takabatake, Anul. Chem. 55, 454 (1983). 52. F. E. Brinckman, W. R. Blair, K. 1.. Jewett and W . P. Iverson, J . Chromatogr. Sci. 15, 493 (1977). 53. F. Baldi, M. Pepi and M. Filippclli, Appl. Etiuirori. Microbiol. 59, 2479 ( 1993). 54. R. P. Mason and W . F. Fitzgerald, I?/a/. Air Soil Pollut. 56, 779 (1991). 55. N. Imura, E. Sukewaga, S. K. Pan, K . Nagao. J. K. Kim, T . Kwan and T. Ukita. Science 172. !23K (1972). 33. 34. 35. 36.