JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY AUGUST 199 1 VOL. 6 385 Determination of Tributyltin and Inorganic Tin in Sea-water by Solvent Extraction and Hydride Generation Electrothermal Atomic Absorption Spectrometry Ni Zhe-ming Hang Heng-bin Li Ang He Bin and Xu Fu-zheng Research Centre for Eco-Environmental Sciences Academia Sinica P. 0. Box 934 Beijing China An ultrasensitive method for the determination of tributyltin and inorganic tin has been developed by in situ concentration of tin hydrides on a zirconium coated graphite tube with subsequent detection by electrothermal atomic absorption spectrometry (ETAAS). Characteristic mass values of 20 and 14 pg were obtained for tributyltin and inorganic tin respectively where characteristic mass is defined as that mass of analyte which provides a defined peak absorbance of 0.0044 A.The relative standard deviations for ten replicate measurements were 5.6% for tributyltin and 3.4% for inorganic tin. For speciation studies tributyltin was extracted in dichloromethane and determined by ETAAS. The sensitivity of tin in organic solution was about the same as that for inorganic tin in aqueous solution. The method proposed was applied to the analysis of water and satisfactory results were obtained. Keywords Electrothermal atomic absorption spectrometry with in situ Concentration; solvent extraction; zirconium coated graphite tube; stannane and tributyltin hydride generation; sea-water The occurrence of organotin compounds in the environ- ment and their impact on the biota have urged the need for the development of sensitive rapid and reliable analytical methods for their determination.The chemistry of tin from an analytical viewpoint has been comprehensively re- viewed.'q2 A variety of techniques have been proposed for the determination of alkyltin species in environmental samples including extraction of tributyltin followed by electrothermal atomic absorption spectrometry (ETAAS),3-6 hydride generation followed by cryogenic trapping and atomic absorption detection using a quartz cuvette at~mizer,~-l coupling of gas chromatographic or liquid chromatographic separation with electron capture,'* flame photometric,13 mass spectr~metricl~ or atomic ab- sorption spectrometric detectors. 15-17 This paper describes a sensitive and rapid method for the determination of tributyltin and inorganic tin in sea-water by solvent extraction and in situ concentration of the tin hydrides on a zirconium coated graphite tube with subsequent atomiza- tion and detection by AAS.The present method has the advantage of avoiding multiple manipulations which might introduce contamination or lead to loss of the analyte. Characteristic mass values of 20 and 14 pg were obtained for tributyltin and inorganic tin respectively where charac- teristic mass is defined as that mass of analyte which provides a defined peak absorbance of 0.0044 A. Experimental Apparatus All the experiments were carried out on a Perkin-Elmer 4000 atomic absorption spectrometer with a deuterium arc background corrector and equipped with an HGA-400 graphite furnace.The operating parameters for direct injection were wavelength 224.4 nm; dry at 100 "C (40 s); char at 500 "C (30 s); and atomize at 2400 "C (4 s). The gas flow was interrupted during atomization and the maximum power heating mode was used. Peak height measurements were used throughout. A laboratory-built hydride generator HG- 100 (Research Centre for Eco-Environmental Sciences Beijing China) was used; the construction and function are similar to those described previously. Hydride generation was accom- plished by using two channels of a peristaltic pump to deliver the sample solution and potassium tetrahydroborate solution. The liberated hydrides were stripped from the solution by the argon gas and were introduced onto the zirconium coated graphite tube via the tip of a quartz tube. When the adsorption of the tin hydrides was complete the quartz tube was removed from the furnace and the tin absorbance was recorded at the atomization temperature of 2400 "C.Insertion of the quartz tube into the graphite tube and its removal were performed automatically. A high intensity tin lamp (General Research Institute of Non-Ferrous Metals Beijing China) was used as the line source. Pyrolytic graphite coated graphite tubes (Union Spectroscopic Technology Investment Beijing China) were used. Coating of the tube surface was achieved by soaking the tube overnight in a solution of 5% ZrOC12-8H20. The coated tube was dried at about 200 "C for 3-4 h. The tube was again treated in the furnace by injection of 100 pi of a zirconium solution onto it with subsequent drying at 100 "C and surface activation at 2000 "C.The treatment was repeated five times. The tube had a lifetime of 150-200 firings. Reagents A stock solution of tin 1 pg ml-' was prepared by dissolving 0.100 g of tin metal powder (99.999%) in 30 ml of 12 mol dm-3 hydrochloric acid and diluting to 100 ml with 5% citric acid in de-ionized water. The stock solution was diluted with 5Oh citric acid in 3 mol dm-3 hydrochloric acid to a concentration of 10 pg ml-' of tin. The final standard solutions were obtained by diluting the above solution with 0.01 mol dm-3 hydrochloric acid to the ng ml-l concentration range. These solutions were prepared just prior to use. Bis(tributy1tin) oxide (Beijing Chemical Industries Beijing China) containing 0.100 g of tin was dissolved in 50 ml of ethanol and diluted with 3 mol dm-' hydrochloric acid containing 5% citric acid to 100 ml to obtain 1 mg ml-l of tin.The organotin solutions were diluted to 10 pg ml-' with 5% citric acid in 3 mol dm-3 hydrochloric acid. The final standard solutions in ng ml-l concentration ranges were prepared just prior to use. Potassium tetrahydroborate solution (4% m/v) was pre- pared daily by dissolving KBH4 (Shanghai Chemical Reagent Factory Shanghai China) in de-ionized water and was used without filtration and without the use of a stabilizing agent. All chemicals were of analytical-reagent grade.386 JOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY AUGUST 1991 VOL. 6 Procedure Hydride generation ETAAS The furnace programme and procedures for hydride genera- tion collection and atomization are briefly listed in Table l.The sequence of operations was as follows. After the initiation of the programme the tip of the quartz tube from the outlet of the hydride generator was inserted through the sample introduction port at the centre of the zirconium coated graphite tube and held in contact with the opposite interior wall. The furnace was heated to 500 "C and the peristaltic pump was then started. The samples in 0.01 mol dm-3 hydrochloric acid and 4% KBH4 solutions were delivered to the generator at a rate of 3 ml min-l for 20 s. The generated hydrides were swept with argon and hydro- gen into the furnace and adsorbed onto the zirconium coated graphite tube. The argon flow rate was 0.3 1 min'l.A period of 100 s was used to collect the tin hydrides at a temperature of 500 "C. The quartz tube was then automati- cally removed from the introduction port of the graphite tube and the atomic absorbance was measured at 2400 "C. Solvent extraction-E TAAS A 10-20 ml volume of diluted sea-water (1 + 1) containing tin compounds was transferred into a 125 ml glass separat- ing funnel. The solution contained 0.001 mol dm-3 hydro- chloric acid. A 1.0 ml volume of 4% m/v NaBH solution and 1 .O ml of dichloromethane were added. The funnel was capped with a glass stopper and shaken by hand for 1 min vented and then shaken for 5 min. After a 10 min settling period the lower organic layer was removed into a microsampling cup with a cap. A 10 p1 volume of the organic solution was injected into the zirconium coated graphite tube using a microsyringe and the absorbance of tin was measured.Results and Discussion Trapping of Tin Hydrides on the Zirconium Coated Graphite Tube The use of a palladium coated graphite tube greatly improved the trapping efficiency of metallic hydrides as compared with the uncoated graphite tube.18J9 The sensi- tivities for the determination of arsenic selenium tellu- rium bismuth antimony and germanium by hydride generation combined with ETAAS were sufficiently high for the analysis of natural water and sea-water. In situ concen- tration of stannane in a graphite furnace has been re- ported.20 The efficiency of generation trapping and atomi- zation for tributyltin relative to inorganic tin hydrides was 46%.The addition of palladium to the furnace produced the same efficiency for the trapping of stannane.20 Our experi- ments showed that in comparison with the standard graphite tube both palladium and zirconium coated graph- ite tubes were found to be 2-4 times more efficient for trapping the hydride derivatives of inorganic tin and organotin. However the zirconium coated graphite tube was preferred because the zirconium remained on the Table 1 Furnace programme Temperature/ Ramp/ Hold/ Step "C S S Procedure - 1 100 5 2 2 500 5 5 Insert quartz tube 100 Hydride generation and collection 8 Remove quartz tube 3 2400 0 4 Atomization graphite tube and the enhanced trapping effect for tin lasted for more than 100 firings. The calibration graphs for stannane and tributyltin are linear over the concentration range 0-2.0 ng ml-l of tin and the sensitivity of tributyltin is only 7Ooh relative to that of stannane.A probable reason for the result was the lower volatility of the tributyltin hydride (b.p. 280 "C) as compared with stannane (b.p. - 52 "C). Characteristic mass values obtained using peak height measurement of 14 and 20 pg were obtained for inorganic tin and tributyltin respectively which is better than that reported previously.20 The relative standard deviations for ten replicate measurements were 3.4Oh for inorganic tin and 5.6Oh for tributyltin. The detection limits for inorganic tin and tributyltin were 58 and 78 pg respectively. The reagent blank was found to contain 0.340 2 0.003 ng of tin. Optimization of Signal The conversion of tin into stannane prior to determination by AAS has been well d~cumented.~g*~-~~ Studies on the conditions required for optimization of the signal such as NaBH concentration and pH value for the generation of tin hydride have shown that the highest absorbance resulted from acidification of the sample solution to pH 2 followed by the addition of 1.5 ml of 4% NaBH s~lution.~ Our experimental work confirmed the above results.There- fore a sample solution containing 0.01 mol dm-3 hydro- chloric acid and 1.0 ml of 4% KBH4 which gave stable absorption signals for the hydrides of both inorganic tin and organotin was used. The dependence of the recovery of tin on the sorption temperature of the furnace is shown in Fig. 1. As can be seen the recoveries of stannane increased on raising the temperature from 100 to 400 "C and then levelled off between 400 and 600 "C while that of tributyltin was nearly constant over the temperature range of 100-700 "C.The optimum response as a compromise for both inorganic tin and organotin can be obtained at 500 "C which was used for further studies. The stripping of the hydrides of tin out of solution increased with purge time using argon at a flow rate of 0.3 1 min-l. As shown in Fig. 2 a period of 80-90 s was- required for the complete recovery of the tin compounds examined. Separation and Determination of Tributyltin Since the contribution of tributyltin hydride to the absorp- tion signal was lower than that of inorganic tin hydride generation ETAAS was not adequate for total tin determi- nation. For quantitative measurements of tributyltin a separation procedure using solvent extraction should be carried out.The inorganic tin concentration can be calcu- lated by subtraction of the absorbance due to tributyltin from the total absorbance obtained by hydride generation ETAAS. I 1 I I 1 300 500 700 Temperature/"C Fig. 1 Dependence of tin recovery on the sorption temperature of the furnace 1 ng of Sn in A stannane; and B tributyltinJOURNAL OF ANALYTICAL ATOMIC SPECTROMETRY AUGUST 199 1 VOL. 6 387 ~~~~~~~ ~ Table 2 Analysis of sea-water Sea-water Sn sample addedhg ml-' 1 1 .o 2.0 5.0 2 6.0 5.0 2.0 * Mean of three determinations. Sn added as tributyltidng ml-1 1 .o 0.50 1 .o 3.0 4.0 6.0 Sn found*/ng ml-I 0.94 2.2 5.2 5.5 5.4 1.8 Sn found as tributyltidng ml-1 1.2 0.59 1 .o 3.3 3.6 6.0 I * 30 50 70 90 Time/s Fig.2 Effect of sorption time on recovery of tin 1 ng of Sn in A stannane; and B tributyltin Tributyltin in sea-water has been extracted by using toluene.6 However this procedure when applied to fresh- water samples was hindered by emulsification of the toluene in the aqueous layer. Recoveries were poor and strong interferences occurred in the graphite furnace. In order to reduce the emulsified layer sodium chloride and methanol should be added. Butyltin hydrides have been extracted by using dichloromethane followed by gas chro- matography-flame photometric detection.25 It was found that the tributyltin hydrides in dichloromethane can be directly determined by ETAAS. The organic layer was easily separated from the aqueous layer and remained clear after shaking for 5 min.The dependence of absorbance on concentration of tin is found to be similar for tributyltin in dichloromethane and inorganic tin in aqueous solution giving a linear calibration with absorbance increasing from 0 to 0.20 for 0-6.0 ng of tin. Since inorganic tin was not extracted into the organic layer the method provided a useful procedure for the separation of tributyltin from inorganic tin prior to its determination. Analysis of Sea-water The sea-water samples received which had previously been filtered contained no measurable levels of tin compounds. In order to test the applicability of the proposed method mixtures of varying proportions of inorganic tin and tributyltin were added to the sea-water samples and the concentrations were determined using the procedures de- scribed.The results shown in Table 2 indicate that recoveries of inorganic tin and tributyltin are 90-1 10 and 90- 120% respectively. The suppression effect of the matrix on tin hydride generation can be minimized by diluting the sea-water. Conclusion In situ concentration of inorganic and tributyltin hydrides onto a zirconium coated graphite tube with subsequent detection by ETAAS provided a sensitive and rapid method for their determination. The method can be applied to other hydride forming elements such as lead and to volatile organometallic compounds. This work was supported by the National Natural Science Foundation of China under Grant No. 2870321. 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 References Weber G.Fresenius Z. Anal. Chem. 1985 321 217. Thompson J. A. J. Sheffer M. G. Pierce R. C. Cooney J. J. and Maguire R. J. Organotin Compounds in the Aquatic Environment NRCCKNRC 22494 Ottawa Canada 1985. Chamsaz M. and Winefordner J. D. J. Anal. At. Spectrom. 1988 3 119. Sullivan J. J. Torkelson J. D. Wekell M. M. Hollingworth T. A. Saxton W. L. Miller G. A. Panaro K. W. and Uhler A D. Anal. Chem. 1988 60 626. Stephenson M. D. and Smith D. R. Anal. Chem. 1988 60 696. Donaghy C. Harriott M. and Thorburn Bums D. Anal. Proc. 1989 26 260. Donard 0. F. X. Rapsomanikis S. and Weber J. H. Anal. Chem. 1986 58 772. Chau Y. K. Wong P. T. S. and Benqart G. A. Anal Chem. 1982 54 246. Randall L. Donard 0. F. X.and Weber J. H. Anal. Chim. Acta 1986 184 197. Balls P. W. Anal. Chim. Acta 1987 197 309. Hodge V. F. Seidel S. L. and Goldberg E. D. Anal. Chem. 1979,51 1256. Sodcquist C. J. and Crosby D. G. Anal. Chem. 1978 50 1435. Muller M. D. Anal. Chem. 1987 59 617. Meinema H. A. Burger-Wiersma T. Versluis-de Haan G. and Geners E. Ch. Environ. Sci. Technol. 1978 12 288. Ebdon L. Hill S. J. and Jones P. Analyst 1985 110 515. Brinkman F. E. Blair W. R. Jewett K. L. and Iverson W. P. J Chromatogr. Sci. 1977 15 493. Jewett K. J. and Brinkman F. E. J. Chromatogr. Sci. 1981 19 583. Zhango L. Ni Z.-m. and Shan X.-q. Spectrochim. Acta Part B 1989,44 339. Sturgeon R. E. Willie S. N. Sproule G. I. Robinson P. T. and Berman S. S. Spectrochim. Acta Part B 1989 44 667. Sturgeon R. E. Willie S. N. and Berman S. S. Anal. Chem. 1987 59 2441. Subramanian K. S. and Sastri V. S. Talanta 1980,27 469. De Doncker K. Dumarey R. Dams R. and Hoste J. Anal. Chim. Acta 1986 187 163. Andreae M. O. and Byrd J. T. Anal. Chim. Acta 1984,156 147. Evans W. H. Jackson F. J. and Dellar D. Analyst 1979 104 16. Matthias C. L. Olson G. J. Brinkman F. E. and Bellama J. M. Environ. Sci. Technol. 1986 20 609. Paper 0/02866E Received June 26th 1990 Accepted March 7th 1991