Rapid Speciation of Butyltin Compounds in Sediments and Biomaterials by Capillary Gas Chromatography- Microwave-induced Plasma Atomic Emission Spectrometry After Microwave= assisted Leac hing/Digestion Journal of Analytical Atomic Spectrometry JOANNA SZPUNAR VINCENT 0. SCHMITT AND RYSZARD E O B I ~ ~ S K I Laboratoire de Photophysique et Photochimie Moliculaire CNRS URA 348 Universitd Bordeaux I 351 Crs de la Liberation 33 405 Talence France. E-mail Lobinski@frbdxl 1 .cribxl .u-bordeaux.fr JEAN-LOUIS MONOD Laboratoire d'Hydrologie et Molysmologie Aquatique Facultd de Pharmacie Universitd Aix-Marseille 27 Bd Jean Moulin 13 385 Marseille France A rapid procedure for the simultaneous determination of mono- di- and tributyltin in sediments and biological materials is described. The target compounds in sediments were subject to quantitative microwave-assisted leaching with acetic acid.Biomaterials were dissolved in a tetramethylammonium hydroxide solution under the action of a low-power microwave field in such a way that the organotin moiety remained intact. The leaching of the analytes from a sediment as well as the dissolution of a biomaterial in a focused microwave field took only 1-5 min which is much faster than previously reported methods. It was also demonstrated that the leaching efficiency of monobutyltin from certified reference sediments was superior to that of most literature procedures. The butyltins were derivatized with sodium tetraethylborate (NaBEt,) in the aqueous phase and simultaneously extracted into isooctane ( 5 min). The analysis was carried out by capillary gas chroma tograp h y ( GC ) with micr ow ave-induced plasma atomic emission detection.The sample throughput was limited by the duration of the GC run (10 min). The detection limit was 2 ng 8-l for 0.2 g of sample (dry) without preconcentration; it could be improved by a factor 10 if preconcentration was applied. The developed method was applied to a variety of real samples and was validated by analysing three certified reference materials PACS-1 CRM 462 and NIES 11. Keywords Speciation; organotin; microwave-assisted sample preparation; capillary gas chromatography; microwave-induced plasma atomic emission spectrometry; environmental analysis Tributyltin (TBT) which has been used as a biocide in antifouling paints to control the attachment and growth of organisms on the hulls of ships for over 30 years has also been shown to be toxic to non-target b i ~ t a .l - ~ It was found to be responsible for the degradation of edible aquatic resources thus raising economic and ecotoxicological concerns. Legislation has banned TBT on ships smaller than 25 m for over 10 years and has restricted TBT-paints on larger vessels. Some recent studies have shown however unusually high butyltin concentrations in some areas considered clean. Attention has turned to sediments which act as the ultimate sink of organotins. The accumulated species can be released under favourable conditions into the aquatic environment and create an ecotoxicological risk long after the anthropogenic sources have disappeared.For monobutyltin (MBT) and dibu- tyltin (DBT) (products of TBT degradation) no legal environ- mental threshold limits have yet been set but in view of their toxicity and environmental occurrence their simultaneous monitoring is nece~sary.~ The interest in rapid methods for biomaterials has been reinforced apart from ecotoxicological concerns by concerns about their content in foodstuffs. The analytical protocols available are based mostly on the coupling of chromatography and atomic spectrometry Despite the increasing popularity of high-performance liquid chromatography-inductively coupled plasma mass spec- trometry (HPLC-ICP-MS),'-'' the coupling of gas chromatog- raphy (GC) with AS remains the preferred approach owing to good resolution and availability of sensitive detectors. The usual choice for detection is atomic absorption spectrometry (AAS),l1-I4 flame photometric detection ( FPD),15-'' micro- wave-induced plasma atomic emission spectrometry (MIP- AES)19-22 or ICP-MS.23924 These techniques generally show good performance and it is the sample preparation step that determines the duration efficiency precision and accuracy of the over-all analytical procedure. Indeed the procedures reported so far are not only time- consuming but also usually inefficient in terms of analyte recovery and unreliable.As shown by Zhang et only three out of ten sample preparation methods described in the literature for the analysis of sediments were able to recover more than 90% of TBT from a sediment sample whereas none of them was able to recover MBT in a non-erratic and reproducible manner.A high scatter of results caused by leaching problems also prevented certification of MBT in the Community Bureau of Reference (BCR) Certified Reference Material (CRM) 462 Sediment.26 Recently several papers have appeared on supercritical fluid extraction which might be expected to provide a solution to these In addition to the high equipment cost however the extraction step still required 10-50min and the recoveries of di- and especially monosubsituted compounds (even when added as spikes) were far from being quantitative. Microwave-assisted processes have been gaining in popular- ity in analytical and environmental c h e m i ~ t r y . ~ ~ ~ ~ ~ Digestion with a mixture of concentrated acids in a pressurized vessel at temperatures of 200-300°C is commonly used to achieve complete sample dissolution which for biomaterials is associ- ated with the destruction of carbon-containing The potential of microwave heating to enhance the extraction of various compounds from emulsions biomaterials and soils has been reviewed.34 Studies of selective leaching of organic ana- lytes from an inorganic matrix have mainly been confined to Journal of Analytical Atomic Spectrometry March 1996 Vol.11 (1 93-1 99) 193polyaromatic hydrocarbons and pesticides in sediments and A low power microwave field was recently shown to accelerate and enhance leaching of organotin species from a sediment without affecting the C-Sn bonds.37 The poor com- patibility of the determination procedure (GC-FPD) with the leachate required additional steps and resulted in an over-all time for the procedure of more than 1 h which makes it unsuitable for routine analysis.18 The present study was aimed at developing a faster procedure internal standard solution was prepared by diluting the Pr3SnC1 stock solution with methanol to give a concentration of 1 pg m1-l.Two sediments with certified contents of butyltin species were used uiz. PACS- 1 from the National Research Council of Canada (NRCC) and CRM 462 from BCR. A fish tissue NIES 11 from the National Institute of Environmental Studies of Japan with a certified content for TBT was also used. adapted to routine work by optimizing the transfer of organo- tins from a sediment to the organic phase and the application Procedures of a tin-selective detector. Particular attention is given to the Analysis of sediments recovery of MBT the accurate determination of which is still beyond the capability of most procedures published so far.This work also attempts to give a first approach to the microwave-assisted solubilization of tissue samples without the destruction of the Sn-C bond in order to accelerate speciation analysis for organotin in biomaterials. EXPERIMENTAL Instrumentation Organotin compounds were extracted in a 50ml open vessel with a condenser made of borosilicate glass using a Microdigest Model A301 (2.45 GHz maximum power 200 W) microwave digester (Prolabo Briare France) equipped with a TX32 programmer which allows the applied energy to be varied from 20 to 200 W in steps of 10 W.The time of exposure up to 99 min can be set in steps of 1 min. The ethylated species were separated on a DB-210 (J&W) column (30 m x 0.32 mm x 0.25 pm) using an HP Model 5890 Series I1 gas chromatograph (Hewlett-Packard Avondale PA USA) equipped with a split/splitless injection port. Detection was achieved with an HP Model 5921A atomic emission detector. Injections were made by means of an HP Model 7673A automatic sampler. Data were handled using an HP Model 5895A ChemStation. A piece of HP-1 GC column (Hewlett- Packard) (0.32 mm x 0.17 pm) served as transfer line. Reagents Analytical-reagent grade chemicals (Merck Darmstadt Germany) and water de-ionized and further purified in a Milli-Q system (Millipore Milford MA USA) were used throughout unless otherwise stated.The glassware used was cleaned with a common detergent thoroughly rinsed with tap water soaked for 12 h in a 10% nitric acid solution and finally rinsed with de-ionized water just before use. A sample of 0.1-0.2g of dry sediment (or 1-2g of wet sediment) 100 pl of the TPrT solution and 10 ml of acetic acid solution (1 + 1) were placed in an extraction tube and exposed to microwaves at a power of 60 W for 3 min. The supernatant solution was transferred by means of a Pasteur pipette into a narrow-neck 20 ml extraction tube (similar to the design described by Witte et aL3*). Volumes of 10 ml of buffer 5 ml of ammonia 1 ml of NaBEt solution and 1 ml of isooctane containing 20ngml-' of tetrabutyltin were added to the supernatant. The mixture was shaken for 5 min.After separa- tion of the phases (several seconds) sufficient water was added to force the organic phase into the narrow neck to facilitate its recovery. The extract was analysed by GC-AED. Analysis of biomaterials A sample of 0.1-0.2 g of lyophilized tissue (or 1-2g of wet tissue) and 5 ml of 25% aqueous tetramethylammonium hydroxide (TMAH) solution were placed in an extraction tube and exposed to microwaves at a power of 60 W for 3 min. The solution was diluted with 15 ml of water brought to a pH of about 5 by the addition of concentrated acetic acid and buffered with 5 ml of the buffer solution. Volumes of 1 ml of the NaBEt solution and 1 ml of the extracting solvent were added and the tube was shaken for 5 min. The emulsion was broken up by subjecting the extraction tube to the microwave field for 2 min at 20 W.Thereafter the organic phase was recovered and analysed by GC-AED. GC-AED conditions The optimum parameters used for GC-AED are listed in Table 1. RESULTS AND DISCUSSION acetic acid. Standards Individual stock solutions (0.5 mg ml-' as Sn) of BuSnCl (MBT) Bu,SnCl (DBT) Bu3SnC1 (TBT) and Pr3SnC1 (TPrT) (Aldrich St. Quentin Fallavier France) were prepared in methanol. Mixed working solutions were prepared daily by dilution of the stock solutions with methanol. Tetrabutyltin used as internal standard was prepared in the same way. A multi-compound working solution was prepared at 0.25 pg ml-' and diluted with methanol as required. The and was chosen in the present work. The primary concern was focused on the quantitative separation of butyltins from sedi- ment and tissue matrices using a medium suitable for the reaction of the analytes with NaBEt,.Speciation of Butyltins in Sediments Separation of butyltins from the sediment matrix The classical approach is based on the extraction of butyltins from a sediment as their chelate complexes (typically with tropolone) into a water-immiscible solvent.6 Recent studies by 194 Journal of Analytical Atomic Spectrometry March 1996 Vol. 11Table 1 Optimum GC-AED parameters GC parameters Injection port Injection port temperature Injection volume Column head pressure Oven programme Initial temperature Ramp rate Final temperature Interface parameters Transfer line Transfer line temperature AED parameters Wavelength Helium make-up flow Scavenger gases H2 pressure 0 pressure Spectrometer purge flow Solvent vent-off time Column-detector coupling Cavity temperature Splitless 200 "C 1 Yl 130 kPa helium 60 "C (1 min) 20 "C min-l 200 "C (1.5 min) HP-1 250 "C 303.419 nm 240 ml min-'* 50 psi 20 psi 2 1 min-' nitrogen 1.5 min Column-to-cavi ty 250 "C * Measured at the cavity vent.Chau et a1." and Ceulemans and Adams41 showed that the presence of a polar solvent in the extractant mixture is essential for the quantitative recovery of MBT. This makes the analytical procedure a priori time-consuming because of the need for elimination of the polar solvent prior to the Grignard derivatiz- ation which is carried out on the organic phase. For this reason another approach i.e.leaching of native organotin compounds with water or a water-miscible polar solvent was investigated in detail in this work. Although organotin compounds are not involved in mineralogical pro- cesses the presence of acid is essential to destroy surface carbonates and to facilitate the penetration of the leaching agent. Acetic acid was preferred to HCl for this purpose because of the danger of a nucleophilic attack on the organotin compounds by HCl resulting in a possible cleavage of side groups under the microwave field. It was found that in a microwave field acetic acid (1 + 1) alone is able to leach quantitatively all three butyltin species within 3min. The presence of methanol as advocated in a previous was found not to affect the leaching process.However the presence of methanol adversely affected the efficiency of the derivatization/extraction process thereby requiring a large dilution of the leachate prior to extraction." Note also that TBT released from a sediment on sonication with methanolic HCl was reported to resorb on the sediment immediatel~.,~ The use of acetic acid in the present work apparently makes this sorption more difficult. Extraction of butyltins with NaBEt Initially attempts were made to extract the organotins directly from the leachate containing the suspended sediment after pH adjustment and addition of NaBEt and hexane. A literature report showed the possibility of 70-90% recovery of butyltin species from a dilute (0.5-270 dry mass/volume) sediment suspension by using a 30min extraction with fairly concen- trated (0.5%) NaBEt into a large (10 ml) volume of he~ane.~ We found that this procedure failed when the volume of hexane or the extraction time was reduced which was necessary to avoid the lengthy evaporation step and to keep the over-all analysis time short.Further it was observed that the presence of sediment during 5 min of shaking with 1 ml of the NaBEt solution and 1 ml of hexane reduced recoveries to 10-20% for DBT and TBT whereas virtually none of the MBT was recovered. The same result was observed for the extraction with tropolone and subsequent Grignard derivatization. This phenomenon was observed irrespective of whether the sample had been treated in the microwave field prior to extraction or not. It was ascribed to the fact that in the absence of a polar solvent organotin compounds could sorb on the sediment during shaking.It was observed however that when extraction was per- formed from a supernatant free of the suspended sediment spike recoveries approached 100% after a 5 min extraction. Isooctane was preferred to hexane because it allows a higher GC injection port temperature which makes the GC run shorter. GC-AED conditions Most of the work on capillary GC of tetraalkyltin compounds has been carried out using columns with non-polar phases (cross-linked 100% polydimethylsiloxane).6 These columns ensure excellent resolution sharp peaks and good sensitivity in peak height mode. The disadvantage is the fairly long retention times of polyaromatic hydrocarbons (abundant in many sediments) and hence the need to heat the column to 280-300°C. This increases the analysis time up to 15 min and prolonged cooling of the oven is required.In the present work a weakly polar column was chosen. The peaks were broader (the half-width was increased from 0.016 to 0.025 min) and the sensitivity was slightly poorer but the chromatogram was completed at 180-200 "C. No problems were observed in about 200 chromatograms of real sample extracts (including tissues without clean-up) run in this work. Apart from causing the build-up of a black deposit on the injector liner which required periodic cleaning introduction of the crude extracts did not appear to detract from column performance. Another contribution to shortening the analysis time was the use of isooctane in place of the commonly used hexane.The starting temperature of 60 "C not only allowed advantage to be taken of the solvent effect but also decreased the oven cooling time. By using the oven programme in Table 1 four runs per hour (including the data handling) were achieved. Sediment samples tend to contain considerable amounts of sulfur either native (e.g. as s6) or organically bound. Fig. 1 shows a chromatogram for the PACS-1 sediment on the S 181 nm channel which demonstrates the presence of an intense sulfur background and of sulfur compounds at considerable concentrations. In real-life sediments these peaks may occur at random retention times sometimes influencing the analysis. 2 3 4 5 6 7 8 9 10 Retention time/min Fig. 1 Chromatogram obtained for the PACS-1 sediment using the procedure optimized for the butyltins (Table 1) on the sulfur (emission wavelength = 181 nm) channel Journal of Analytical Atomic Spectrometry March 1996 Vol.11 195A desulfurization step is required which makes the analysis time longer; moreover such a procedure does not affect organosulfur compounds that can be co-determined.I8 The advantage of using a tin-selective detector is that it allows the over-all procedure to be shortened. The low selectivity of FPD against sulfur creates a danger of overlap of sulfur peaks with analytical peaks which are impossible to detect unless a truly tin-selective detector is used. Internal standardization Because of the complexity of real-life aquatic sediments the extraction efficiencies may vary with the type of sediment despite the use of the same or a similar extraction pr~cedure.~ An internal standard is necessary.Tripropyltin was chosen. Its main role is to correct for volume changes and spray losses during the microwave heating and the separation of the supernatant. A poor recovery of the internal standard can be considered as an early warning of the malfunction of the procedure. However although the internal standard can contribute sig- nificantly to the correction of errors caused by sorption/ desorption processes on different sediments and of those resulting from the suppression of the extraction caused by the release of an unexpected contaminant from a sediment it should not be used for corrections exceeding 10-15%. Should such a need exist a standard additions run is highly advisable especially when a new batch of sediments is to be analysed.Another internal standard Bu4Sn (TeBT) was added to the extracting solvent (isooctane) to correct for the precision of the injection. Analytical characteristics Typical chromatograms obtained for the reference sediments are shown in Fig. 2. Values of precision and spike recovery are listed in Table 2. Results of the recoveries of standard additions are also included. The method is fast. The throughput is limited by the duration of the chromatographic run and is 4 samples h - l . The detection limit is 2ngg-' for a 0.2g sample size. Evaporation of the extract ten times resulted in a decrease in the detection limit; no additional matrix effects were observed. High volume injection (as described elsewhere6) is thus feasible.Validation of the method Table 3 shows the results for the determination of MBT DBT and TBT in the CRMs. Good agreement with the certified values for DBT and TBT for both sediments is observed. The value for MBT is much higher than the certified value for the PACS-1 sediment and is one of the highest ever reported. This deserves further examination. The highly scattered values for MBT recoveries in the PACS-1 and CRM 462 sediments reported in the literature are summarized in Table 4. It is surprising that none of the values published matches the mean by closer than 20% and only three out of 12 results published hitherto fall within the certified range. Fewer results have been published for CRM 462; three of them (out of five) lie closely together and are close to the certification mean.The most probable reason for this is that MBT is not completely recovered from the sediment by most of the reported procedures. The most reliable assessment of recovery can be made by the method of standard additions the accuracy of which is dependent on whether the analyte and the spike behave similarly. This similar behaviour is apparently not the case for MBT in sediments. Indeed for two intrinsically different procedures recoveries of the MBT spike varied by about 20% (84.7 and 62.3%) but the calculated concentration in the sediment differed by a factor of 2.5 (1.03 196 Journal of Analvtical Atomic SDectrometrv. March 1996. 1 2 l 3 60 50 40 h u) c .- c % 30 .- c 1 s2 34 t IS1 I 2 281 I I 1 I I I I I 0 1 2 3 4 5 6 7 8 9 Retention time/rnin Fig.2 Chromatograms obtained by the optimized procedure for the certified reference sediments. (a) PACS-1; (b) CRM 462; 1 MBT; 2 DBT; 3 TBT; IS1 TPrT; IS2 TeBuT Table2 sample (five experiments) Results of the spike recovery experiments for a sediment Compound Addedlng as Sn MBT - 20 40 DBT - 20 40 TBT - 20 40 Foundlng as Sn 34.0 54.7 72.5 24.5 44.9 63.8 12.2 31.5 49.9 s,* Recovery (Yo) (Yo) 8.5 - 7.3 101 6.8 98.0 5.1 - 6.3 101 4.9 98.9 11.5 - 7.3 102 6.6 97.9 * s = Relative standard deviation. Table3 certified reference sediments (five experiments) Results for the determination of butyltin compounds in PACS-lIpg g-' as Sn CRM 462Jngg-' as Sn Compound Certified Determined Certified Determined DBT 1.16f0.18 1.01+0.06 128k16 122f6 TBT 1.27f0.22 1.19k0.08 70+14 61 +7 MBT 0.28f0.17 0.76k0.05 (12-244)* 172f 15 * Literature data.26 and 0.41 respectively).22 This discrepancy suggests that the spike is not recovered in the same way as the compound in the sediment itself. This is corroborated by the study of Siu et ~ l .~ ~ who showed an 86% recovery of the spike but were not able to detect MBT in PACS-1 despite a sufficiently low experimental detection limit. The value obtained in this work is one of the highest ever VOl. 11Table 4 Literature values for the recovery of MBT from PACS-1 and CRM 462 Procedure* Certified value SFE in the presence of DDTC with CO doped with 5% methanol SFE with C02 doped with 10% methanol Extraction with aqueous NaBEt into hexane Extraction with tropolone into hexane Acidification with HC1 extraction with tropolone into hexane-ethyl Extraction with tropolone into toluene Leaching with concentrated acetic acid acetate Leaching with concentrated acetic acid Microwave-assisted leaching with 0.5 mol 1-' acetic acid in methanol Microwave-assisted leaching with 8.5 mol I-' acetic acid MBT concentration in PACS-l/pg g-' 0.28 f 0.17 0.025 & 0.06 0.41 k 0.04 0.49 f 0.09 0.52 & 0.1 5 0.36 & 0.17 0.94 & 0.06 1.03 50.01 0.55 k 0.05 0.72k0.16 0.59 k 0.06 0.37 & 0.01 0.76 0.05 MBT concentration in CRM 462/ng g-' ( 13-244)t 9.5 & 3.8 NAS NA 102 & 38 126+ 16 NA 8 7 f 4 NA 28 & 4 172k 15 Ref.21 22 23 24 41 22 9 43,44 18 This work * SFE = Supercritical fluid extraction; DDTC = diethyldithiocarbamate.7 The range reported for the certification study.26 The mean value obtained by the laboratories participating in the campaign is 148 f 64. It is 1 NA =Not analysed. neither a certified nor an indicative value. reported. It is slightly lower than the results of Chau et a1.22 and Ceulemans and Adams,,' who have pointed out that their methods extracted three times more MBT from the PACS-1 sediment than the certified value. Our value is also higher than that previously reported using the microwave-assisted pro- cedure.18 This is probably due to the absence of methanol which was found to hamper derivatization and extraction of MBT with NaBEt,. It should be pointed out that it is unlikely that any MBT may be generated by degradation of TBT and/or DBT in a microwave field; the values obtained for the last two compounds match perfectly the certified values.Speciation of Butyltins in Biological Materials In contrast to sediments organotin compounds tend to be incorporated into tissue. Despite some success with leaching reported by several groups,45 the digestion (complete solubiliz- ation) approach was preferred. This was justified because the reports of successful leaching were mostly based on the spike recovery experiments. Intrinsically bound organotin may behave differently. Microwave-assisted dissolution of tissue A recent comparison study showed that hydrolysis with TMAH is superior to acid and enzymic hydrolysis in terms of time and efficacy.46 Biological tissues can be dissolved during 1 h at 40-60°C.46 It was observed in this study that a microwave field can shorten this time to 1-5 min depending on the applied power (100-20 W respectively).Despite rapid (1 min) dissolu- tion at 80-100 W spray losses were considerable and precision was poor. It was preferable to dissolve samples at 60 W under which conditions a transparent solution was obtained after 3 min from the fish oyster and mussel samples investigated. In the last two instances a fine silica-like suspension was sometimes observed at the end. Its presence did not however affect the recoveries of the butyltins. Extraction of butyltin compounds The solution obtained is highly alkaline and must be brought to pH 5 to permit derivatization and extraction. During acidi- fication a white precipitate probably of free fatty acids was formed especially with fish tissue hydrolysates.The precipitate was found to affect precision and its formation should be limited. This was achieved by diluting the hydrolysate 4-fold with water and carrying out the extraction from dilute solu- tions. During shaking a foam was formed which prevented the separation of the phases. Instead of centrifugation exposure to a microwave field at 20 W for 1-3 min was found to break up the emulsion and allowed for clear separation of the phases. Isooctane is required not only for making the GC run shorter but also because hexane evaporates faster during microwave treatment. Injection of 1-2 yl on a routine basis did not require a clean-up when a polar column was used. When a non-polar column was used a clean-up procedure e.g.on an alumina column was indispensable. Analytical characteristics A typical chromatogram of the NIES 11 reference tissue extract is shown in Fig. 3. Table 5 illustrates results from the spike recovery experiments on an uncontaminated oyster tissue homogenate and the NIES 11 tissue. Good precision and linearity can be seen. Table 6 shows the results obtained for the determination of MBT DBT and TBT. Good agreement with the certified value was found for TBT. For MBT and DBT no certified values are available. The pattern is similar to that found by Ceulemans et The method was also validated by a survey analysis (single determinations) of some fish and mussel samples from the Marseille harbour area analysed 6 months earlier using a sample preparation pro- cedure similar to that described by Pannier et ~ 1 .~ ~ The I 46 48 I x 44 c .- 2 42 a .r 40 C .o 38 vl vl 'E 36 W 34 32 30 w 3 I S 4 ms 28 I 1 I 1 I I I I I 0 1 2 3 4 5 6 7 8 9 Retention tirnelmin Fig. 3 Chromatogram obtained for NIES 11 Fish Tissue 1 MBT; 2 DBT; 3 TBT; 4 TPhT; IS TeBuT Journal of Analytical Atomic Spectrometry March 1996 Vol. 11 197Table 5 Results of the spike recovery experiments for biological tissue (five determinations) Sample Compound Fish tissue MBT Southern France DBT TBT NIES 11 MBT DBT TBT Added/ng as Sn - 20 40 20 40 20 40 40 40 40 - - - - - Found/ng as Sn NDt 13.2 21.3 ND 19.5 40.9 ND 17.0 37.6 1.7 39.9 10.9 50.1 88.7 121 s,* (Yo) - 7.9 6.8 5.3 5.1 6.6 6.2 6.5 7.1 5.9 5.7 4.9 - - 10 Recovery (Yo) 66.2 53.2 97.2 - - 102 - 85.2 94.1 95.7 98.4 94.0 - - - * s = Relative standard deviation. ND = Not determined (below the detection limit).Table 6 Results for the determination of butyltin compounds in NIES 11 Fish Tissue (five determinations) Certified/yg g-’ Determinedlpg g-’ Compound as chloride as chloride MBT NA* 0.02 f 0.002 DBT NA 0.14 f 0.01 TBT 1.3f0.1 1.22 f 0.07 * NA = Not available. 0- 1 2 3 4 5 6 7 Concentration (Sn) found by procedure Npg g-’ Fig. 4 Results of an inter-method comparison of TBT determination in mussels and fish (Marseille harbour area). Procedure A this method. Procedure B method based on sample preparation according to ref. 47 followed by GC-AED agreement shown in Fig. 4 can be considered to be good taking into account that an independent set of calibrants was used.CONCLUSIONS This study is the first to present a method for the simultaneous speciation of butyltins in sediments and tissues in which the sample throughput is controlled by the duration of the GC run and not by the sample preparation step. Time-consuming (several hours) sonication and operations such as evaporation and solvent replacement are avoided. Application of micro- waves to leaching of butyltins from sediments not only reduces the time of sample preparation but also enhances the recovery of MBT. Microwave-assisted digestion also reduces the time required for the dissolution of biomaterials 20 times without affecting Sn-C bonds. The proposed method appears promis- ing for routine agricultural and biological speciation analysis.We thank D. Mathe (Prolabo) for the gift of the microwave digester and Dr. Olivier F. X. Donard for valuable discussions. J. S . acknowledges a long-term fellowship from the European Environmental Research Organization (EERO). REFERENCES 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 Maguire R. J. Appl. Organomet. Chem. 1987 1 475. Huggett R. J. Unger M. A. Seligman P. F. and Valkirs A. O. Environ. Sci. Technol. 1992 26 232. Tsangaris J. M. and Williams D. R. Appl. Organomet. Chem. 1992 6 3. Stab J. A. Brinkman U. A. Th. and Cofino W. P. Appl. Organomet. Chem. 1994 8 577. Donard O. and Ritsema R. in Techniques for Environmental Analysis ed. Barcelo D. Elsevier Amsterdam 1993 pp. 549-606. Dirkx W. M. R. tobinski R. and Adams F. C. Anal. Chim.Acta 1994 286 309. Lobinski R. and Marczenko Z. Spectrochemical Trace Analysis for Metals and Metalloids Elsevier Amsterdam 1995. MacLaren J. W. Siu K. W. M. Willie S. N. Maxwell P. S. Palepu A. Koether M. and Berman S . S. Fresenius’ J. Anal. Chern. 1990 337 721. Dauchy X. Cottier R. Batel A. Borsier M. Astruc A. and Astruc M. Environ. Technol. 1994 15 569. Inoue Y. Kawabata K. and Suzuki Y. J . Anal. At. Spectrom. 1995 10 363. Dirkx W. M. R. tobinski R. and Adams F. Anal. Sci. 1993 9 273. Forsyth D. S. and Hayward S. Fresenius’ J. Anal. Chem. 1995 351 403. Martin F. M. and Donard 0. F. X. Fresenius’ J. Anal. Chem. 1995,351,230. Sarradin P. M. Leguille F. Astruc A. Pinel R. and Astruc M. Analyst 1995 120 79. Tolosa I. Bayona J. M. AlbaigCs J. Alencastro L. F.and Tarradellas J. Fresenius’ J. Anal. Chem. 1991 339 646. Cai Y. and Bayona J. M. J. Chromatogr. Sci. 1995 33 89. Gomez Ariza J. L. Beltran R. Morales E. Giraldez I. and Ruiz Benitez M. Appl. Organomet. Chem. 1995 9 51. Lalere B. Szpunar J. Budzinski H. Garrigues Ph. and Donard 0. F. X. Analyst 1995 120 2665. Scott B. F. Chau Y. K. and Rais-Firouz A. Appl. Organomet. Chem. 1991 5 151. Lobinski R. Dirkx W. M. R. Ceulemans M. and Adams F. C. Anal. Chem. 1992 64 159. Liu Y. Lopez-Avila V. Alcarez M. and Beckert W. F. Anal. Chem. 1994 66 3788. Chau Y. K. Yang F. and Brown M. Anal. Chim. Acta 1995 304 85. Prange A. and Jantzen E. J. Anal. At. Spectrom. 1995 10 105. Kuballa J. Wilken R. D. Jantzen E. Kwan K. K. and Chau Y. K. Analyst 1995 120 667. Zhang S. Chau Y. K.Li W. C. and Chau A. S. Y. Appl. Organomet. Chem. 1991 5,431. Quevauviller P. Astruc M. Ebdon L. Desauziers V. Sarradin 198 Journal of Analytical Atomic Spectrometry March 1996 KiE. 11P. M. Astruc A. Kramer G. N. and Griepink B. Appl. Organomet. Chem. 1994 8 629. 27 Liu Y. Lopez-Avila V. Alcaraz M. and Beckert W. F. J. High Resolut. Chromatogr. 1993 16 106. 28 Cai Y. Alzaga R. and Bayona J. M. Anal. Chem. 1994,66 1161. 29 Bayona J. M. and Cai Y. Trends Anal. Chem. 1994 13 327. 30 Introduction to Microwave Sample Preparation eds. Kingston H. M. and Jassie L. B. American Chemical Society Washington DC 1988. 31 Zlotorzynski A CRC Crit. Rev. Anal. Chem. 1995,25 43. 32 Matusiewicz H. and Sturgeon R. E. Prog. Anal. Spectrosc. 1989 4 14. 33 KuD H . M. Fresenius’ J. Anal. Chem. 1992 343 788. 34 Pare J. R. J. Belanger J. M. R. and Stafford S . S. Trends Anal. Chem. 1994 13 176. 35 Onuska F. I. and Terry K. A. Chromatographia 1993 36 191. 36 Lopez-Avila V. and Young R. Anal. Chem. 1994 66 1097. 37 Donard 0. F. X. Lalbre B. Martin F. and Lobinski R. Anal. Chem. 1995 67,4250. 38 Witte C. Szpunar J. Lobinski R. and Adams F. C. Appl. Organomet. Chem. 1994 8 621. 39 40 41 42 43 44 45 46 47 Quevauviller P. Donard 0. F. X. Maier E. A. and Griepink B. Mikrochim. Acta 1992 109 169. Rapsomanikis S. Analyst 1994 119 1429. Ceulemans M. and Adams F. C. Anal. Chim. Acta in the press. Siu K. W. M. Maxwell P. S. and Berman S. S. J. Chromatogr. 1989,475 373. Pannier F. Dauchy X. Potin-Gautier M. Astruc A. and Astruc M. Appl. Organomet. Chem. 1993 7 231. Astruc A. Dauchy X. Pannier F. Potin-Gautier M. and Astruc M. Analusis 1994 22 257. Dirkx W. M. R. Lobinski R. and Adams F. in Quality Assurance in Environmental Analysis eds. Quevauviller Ph. Maier E. and Griepink B. Elsevier Amsterdam 1994. Ceulemans M. Witte C. Lobinski R. and Adams F. C. Appl. Organomet. Chem. 1994 8 451. Pannier F. Astruc A. and Astruc M. Appl. Organomet. Chem. 1994 8 595. Paper 5/06320& Received September 26 1995 Accepted November 1 1995 Journal of Analytical Atomic Spectrometry March 1996 VoE. 11 199