ANALYST, JUNE 1991. VOL. 116 585 Electroanalysis for Organotin in Natural Waters Including Sea-water by Cathodic Stripping Voltammetry Constant M. G. van den Berg and Shaukat H. Khan* Oceanography laboratory, University of Liverpool, Liverpool 169 3BX, UK A simple procedure was developed to determine total organotin in natural waters including sea-water. The method entails adsorptive cathodic stripping voltammetry of the 2-hydroxycyclohepta-2r4r6-trienone(tropo- lone) complex of tin for the determination of organotin after its conversion into inorganic tin by ultraviolet irradiation. The method is free from interference by organic surface-active compounds, as these are decomposed during the irradiation stage. The concentration measured has to be corrected for the inorganic tin originally present in the sample by analysis of a sample from which organotin has been removed.Organotins are removed by passing the acidified sample through a CI8 column (Sep-Pak), which adsorbs all organotin from the sample, leaving inorganic tin in solution. The limit of detection for total organotin is approximately 10 pmol dm-3 (with a deposition time of 10 min), but this increases with the amount of inorganic tin present in the sample (typically up to 100 pmol dm-3). Keywords: Cathodic stripping voltammetry; organotin determination; natural waters; sea-water Methyltin compounds can be produced by natural methylation of inorganic tin in sediments,' but those of butyltin are exclusively of industrial origin.2 Natural waters near ports and marinas in coastal waters and estuaries can be contaminated with organotin as a result of the application of antifouling tributyltin preparations.In order to monitor these inputs it should be sufficient to measure the level of organotin as an indicator of the degree of contamination. As no separation is required, a comparatively simple analytical procedure can then be used. Several mcthods exist for determining organotin com- pounds (butyltin) in natural waters. These are based on hydride formation with cold-trapping,3.4 and extraction of the complexes with 2-hydroxycyclohepta-2,4,6-trienone(tropo- lone) ,s both methods involving detection by atomic absorp- tion spectrometry. These methods are not readily automated and are, therefore, not convenient for use in monitoring.Cathodic stripping voltammetry (CSV) is suitable for automa- tion by using such computerized instrumentation as is cur- rently under development in our laboratory and which could provide a better alternative to the methods mentioned above. The proposed method is based on an existing CSV procedure for determining inorganic tin in sea-water.6 Tin forms a complex on the addition of a chelating agent (tropolone) at pH 2. The complex is allowed to adsorb on a hanging mercury drop electrode (HMDE) and is quantified by measuring its reduction current by means of a reducing potential scan, with use of differential-pulse modulation. In this procedure, organotin is converted into inorganic tin by ultraviolet (UV) irradiation prior to the measurement with CSV. In order to calculate the concentration of organotin, the total tin concentration must be corrected for the contribution of inorganic tin.Inorganic tin is determined by the analysis of a sample from which organotin has been removed by pre-treatment on a CIS adsorption column. Experimental Apparatus and Reagents A Princeton Applied Research (PAR) 174A polarograph was connected to a PAR 303 HMDE. The surface area of the HMDE was 2.9 mm2. The solution (10 ml) in the voltammetric * Present address: National Institute of Oceanography, 37-K/6, P.E.C.H.S., Karachi 29, Pakistan. cell was stirred with a polytetrafluoroethylene (PTFE)-coated magnetic stirring bar at a stirring rate of 400 rev min-I. High-density polyethylene (HDPE) containers (Nalgcne) were used to store the water samples.Prior to use these containers were soaked in hot diluted detergent (0.1% at 60 "C for 1 d), rinsed with distilled water, soaked (by immersion) for 1 week in 50% HCI (AnalaR, BDH), then in 2 mol dm-3 HN03 (AnalaR, BDH) (1 week), subsequently filled with dilute acid, 0.01 mol dm-3 HCI (Aristar, BDH), and stored in individual sealed plastic bags. Sea-water (salinity: approximately 32) used for the experi- ments was collected from the Menai Straits and stored in 60 I HDPE containers. This water was filtered (0.45 pm Nucle- pore) prior to use. Distilled water, used for reagent dilution and rinsing, was produced by a double-distillation system (made of silica). An aqueous stock solution of 0.1 mol dm-3 tropolone (Aldrich) was prepared weekly in distilled watcr and used without further purification. Stock standards of monobutyltin trichloride (Strem Chemicals), dibutyltin di- chloride, tributyltin chloride and tetrabutyltin (Fluka) were stored in a freezer, and were diluted immediately prior to use with methanol (HPLC grade) in glass calibrated flasks; these solutions contained 1 x 10-4 mol dm-3 organotin. Working organotin standards were prepared in glass containers by dilution with methanol.Procedure for Determining Organotin Dissolved organic materials were decomposed by UV pho- tolysis of the acidified [to pH 2 with HC1 (Aristar)] sample in a capped 40 ml silica tube using either a 100 W low-pressure or a 1 kW medium-pressure mercury vapour lamp (Hanovia), housed in a laboratory-built aluminium container fitted with a cooling fan.Preliminary tests showed that organotin was converted quantitatively by this treatment. Total dissolved tin (including organotin and inorganic tin) was then determined by CSV. The CSV procedure used to determine combined tin (inorganic tin and organotin) was adapted from that of van den Berg et af.6 Briefly, this procedure consisted of the addition of 40 pmol dm-3 of tropolone to 10 ml of the acidified, UV-irradiated sample. The solution was de-aerated by purging with nitrogen (6 min). Deposition was carried out at a potential of -0.8 V for 1-5 min, depending o n the expected concentration of tin. The re-oxidation potential was -0.4 V, and the potential scan was in the negative direction,586 ANALYST, JUNE 1991, VOL. 116 using differential-pulse modulation.The re-oxidation time was 20 s. The concentration of inorganic tin was measured after the removal of organotin from the sample by passage of an acidified (pH 2) aliquot (15 ml) through a Sep-Pak CIS column (Waters; column volume approximately 1 ml) at a flow rate of approximately 2 ml min-1. The filtrate was UV irradiated, and dissolved inorganic tin was determined by CSV. The organotin concentration was calculated from the difference between the concentrations of total dissolved tin and total dissolved inorganic tin. Results and Discussion Determination of Organotin by CSV Previous experiments involving the use of polarography with a dropping mercury electrode have shown that organotin compounds can be reduced at potentials more negative than those required for inorganic tin, v i z ., -0.8 V (monobutyltin) and -0.9 V (dibutyltin).’ In preliminary experiments an attempt was made to determine organotin directly by CSV using a deposition potential of <-0.8 V. No peak was obtained for organotin (monobutyltin, dibutyltin and tributyl- tin) even with deposition potentials as negative as -1.6 V. It is not clear why the dissolved organotin could not be reduced and plated on to the mercury electrode under these condi- tions, but organotin could not be determined directly by CSV without prior conversion into inorganic tin by other means. Ultraviolet irradiation is commonly used to photolyse dissolved organic material prior to voltammetric analysis.8 Determinations of tin (between 1 and 100 nmol dm-3) in sea-water by CSV, before and after UV treatment (with use of either the 100 W or the 1 kW system), showed that the organotin was quantitatively converted into inorganic tin by -0.4 -0.6 -0.8 PotentialN Fig.1 Determination of organotin in sea-water using CSV. (a) 20 nmol dm-3 organotin recovered from a CI8 column using 0.5 mol dm-3 HC1; deposition time, 30 s. (b) 1 nmol dm-3 organotin in sea-water after UV irradiation treatment; deposition time, 3 min. (c) 14 nmol dm-3 organotin eluted with methanol from a CIS column; deposition time, 30 s irradiation for 3 h. Thereafter, the concentration of converted tin was measured by CSV. Comparative experiments invol- ving irradiation at neutral and low pH showed that the irradiation had to be carried out after acidification of the sample to pH 2 in order to prevent adsorption of tin on the wall of the silica tube; losses of between 20 and 50% of the dissolved tin were observed if irradiation was carried out at neutral pH.A CSV determination of 1 nmol dm-3 of organotin (tributyltin) in sea-water is shown in Fig. 1. In ‘clean’ sea-water (sea-water of oceanic origin) it is possible to determine tin without prior UV treatment.6 The concentration of surface-active organic material in this water is sufficiently low to allow detection of even very low (10 pmol dm-3) levels of tin ( i . e . , labile tin, organotin being non-labile) without interference. The organotin concentration in such water can, therefore, be measured by the difference between the labile and total (after UV treatment) concentra- tions.Organic material in waters of estuarine and coastal origin interferes with the determination of labile tin (which is normally low at <0.1 nmol dm-3 and requires a deposition time of several minutes), causing poor sensitivity. Inorganic tin in such waters can, therefore, be determined only after UV treatment, which also converts organotin into inorganic tin. In practice, therefore, it is not possible to determine organotin separately from inorganic tin in natural waters without some other treatment to remove either inorganic tin or organotin from the sample. Specific Removal of Organotin from Solution The efficiency of CIS coated Sep-Pak for the adsorption of organotin from sea-water was investigated. This adsorbent came pre-packed in small (1 ml) columns that were connected to the spout of a 250 ml extraction funnel (HDPE) fitted with a PTFE tap (Nalgene) via a short length of silicone tubing.The sea-water was passed through the adsorption columns at a low pressure stream of nitrogen (approximately 0.2 bar) and at a flow rate of about 2 ml min-1. The adsorption efficiency was tested by determining residual tin in the filtrate after UV irradiation. In addition, the recoveries of organotin by elution with various organic solvents were compared. Adsorption experiments with 10 and 100 nmol dm-3 tributyltin added to sea-water and distilled water (acidified and unacidified), also containing 0.2 nmol dm-3 inorganic tin, showed that the organotin adsorbed quantitatively (100%) onto the CIS column, leaving non-adsorbed inorganic tin in the filtrate.Separate experiments with inorganic tin at low (50 pmol dm-3) and elevated (0.2 nmol dm-3) levels showed no adsorption on the CI8 column from acidified or unacidified sea-water and distilled water, indicating that tributyltin was adsorbed specifically by this material. The organotin concen- tration in a natural water sample can, therefore, be calculated from the total dissolved (combined organotin and inorganic tin) concentrations by correcting these for the inorganic tin concentration. Elution of Adsorbed Organotin With Solvents Attempts were made to recover the organotin adsorbed on the Sep-Pak column in order to develop a procedure by which organotin could be determined specifically. Various solvents and acids were assessed and recoveries were evaluated by CSV analysis after the conversion of the organotin into inorganic tin by UV irradiation.As the UV treatment prior to the CSV analysis required that the solution containing the eluate was mainly aqueous, the solvent used for elution had to be soluble in water or had to be evaporated to dryness prior to the dissolution in water. The recovery was tested by adsorption of a known amount (0.2-2 nmol) of organotin on a Sep-Pak column followed by elution, and analysis of the eluate after UV treatment. TestsANALYST, JUNE 1991, VOL. 116 587 with toluene, pentane and dichloromethane showed that a considerable fraction of the organotin evaporated with the solvent when it was dried by blowing down with nitrogen at room temperature (approximately 20 "C).For instance, the recovery of 50 nmol dm-3 organotin in pentane was only 3.8 nmol dm-3 (8%) and for 100 nmol dm-3 organotin in dichloromethane the recovery was between 7 and 18 nmol dm-3 (7-18%); residues of toluene also interfered with the subsequent analysis of tin after UV treatment of the eluate. Attempts to carry out partial blow-downs in order to diminish the evaporation of organotin were not successful as the organic solvents continued to interfere with the CSV analysis even after UV treatment. This is a drawback of the CSV technique as blow-down to dryness would not be essential for an alternative detection technique such as atomic absorption spectrometry. Organotin (20 nmol dm-3) added to sea-water or distilled water also containing 20% methanol was recovered fully by the UV treatment (carried out at pH 2) with subsequent CSV analysis.However, the recovery of the organotin by UV treatment of the methanolic and ethanolic Sep-Pak eluates was poor and variable. Variable recoveries ranging from 50 to 100% were achieved by increasing the volume of the eluent (methanol and ethanol). It appeared that the eluent contained an unknown interferent that lowered the sensitivity of the analysis for tin and sometimes produced a peak that over- lapped and interfered with the peak for tin. A scan for organotin eluted with methanol is shown in Fig. l(c); the magnitude of the small peak next to that of tin varied and the compound producing it probably originated from the methanol or the C18 column. The elution with HCI was incomplete, but at least no interference was introduced: the recovery was 15% with 0.2 rnol dm-3 HCI, 23% with 0.5 rnol dm-3 HCI and 19% with 1 rnol dm-3 HCI (5 ml aliquots); 66% recovery was achieved with a mixture of 50% methanol and 1 mol dm-3 HCI (5 ml).These recoveries were considered to be too low to provide an accurate and reproducible determination of organotin. A scan for organotin eluted with 0.5 rnol dm-3 HCl is shown in Fig. l(a). Effective recovery of the adsorbed organotin from the Sep-Pak column was, therefore, not achieved. Sample Storage The effects of sample storage on the concentration of dissolved tin were determined from the investigation of contamination arising from leaching of the bottle, and adsorption of organotin on to the bottle.Acidified (pH 2 ) sea-water, with and without added organotin, was therefore stored in HDPE and glass containers. The HDPE containers cleaned by soaking in 2 rnol dm-3 HCI (for 2 weeks) were compared with containers that had been soaked in 50% HCl (for 1 week) followed by a soak in 1 rnol dm-3 HN03. It was found that tin was released from HDPE containers cleaned with 2 rnol dm-3 HCI, but not from those cleaned with 50% HCl (Table 1). Further, organotin was found to adsorb on both HDPE and glass containers (Table 2), even if the water had been acidified to pH 1 with HN03 (Table 3). Adsorption was minimized if a large sample volume was used (reducing the surface : volume ratio of the container), but the adsorption rate was still approximately 10% per day from 2.5 1 of acidified sea-water containing 100 nmol dm-3 organotin.Samples should, therefore, be analysed as quickly as possible in order to minimize adsorption, or the samples should be stored in the silica tubes in which the UV irradiation is to be carried out. Interferences Potential interfering species include inorganic tin, other metals that form an adsorptive complex with tropolone, Table 1 Effect of storage on the concentration of inorganic tin in acidified (pH 2) sea-water in HDPE containers Inorganic tin/nmol dm-3 Immediate After 3 d After 5 d Bottle 1" 0.58 2.00 2.6 Bottle 21- 0.58 0.60 0.56 * Cleaned in 2 rnol dm-3 HCl only. f Soaked in 50% HCI and 1 mol dm-3 HN03. Table 2 Effect of storage on the organotin concentration added to acidified (pH 2) sea-water in HDPE containers Organotidnmol dm-3 lmmediate After 1 week Loss (YO) 250 ml HDPE container 15.4 3.1 80 104.8 19.6 81 container 17.7 4.4 75 102.4 19.3 81 11 HDPE Table 3 Storage of 100 nmol dm-3 organotin added to sea-water acidified with 4 ml of concentrated HN03 per litre of sea-water (pH about 1) in a brown glass container of 2.5 1 capacity Storagc Organotid time/d nmol dm-3 Loss (Yo) 0 100.7 0 1 100.5 0.2 2 80.2 20 4 72.2 28 6 64.0 36 10 45.9 54 surface-active organic substances that adsorb competitively on the HMDE and organic complexes of tin.The major interferent is inorganic tin as this is a co-determinant with organotin. Inorganic tin must be determined separately after removal of organotin by passing the sample through a CIS column. Organic complexes of tin in natural waters could interfere if these were adsorbed by the CIS column.The interference from such organically complexed tin is minimized by carrying out the adsorption at low pH (pH 2 ) , when all the metal complexes are expected to be dissociated. Adsorption of inorganic tin on 'uncapped' silica groups of the CI8 cartridge is also prevented at this pH (such adsorption has been shown to occur for several trace metals when adsorption was carried out at neutral pH) .9 Organic surfactants occurring in natural waters are known to diminish the sensitivity of CSV as a result of competitive adsorption .S Electroactive organic compounds could adsorb on the electrode and produce an interfering reduction wave.8 These interfering species are removed effectively by the UV pre-treatment of the sample .8 Metals other than tin can interfere by forming a complex with the added tropolone and by forming an electroactive adsorbed complex.A number of metals have been studied previously and only molybdenum was found to present serious interference in sea-water as its concentration is comparatively high (approximately 100 nmol dm-3) and it produces a reduction peak close to that of tin.6 Interference by this element is eliminated by using a very negative deposition potential (-0.8 V), where complexed molybdate ions are reduced and desorbed from the electrode. Tin is amalgamated during the deposition step as the deposition potential is more negative than the reduction potential of the tropolone complex with tin. The tin is re-oxidized and re-adsorbed, during the re-oxidation step at -0.4 V, from the unstirredANALYST, JUNE 1991, VOL.116 solution during a period of 20 s prior to the CSV scan. Hence, only the amalgamated elements such as lead or cadmium can interfere with the analysis for tin. The reduction potential of lead is more positive than that of tin and the response is poor, hence this element was found not to interfere. The reduction potential for cadmium is very close to that of tin, and a high concentration of this element (100-fold higher than that of tin) can interfere to the extent that 0.5 nmol dm-3 of tin is masked by 50 nmol dm-3 of cadmium. Normally, cadmium concentra- tions in sea-water are fairly low (0.5 nmol dm-3 or less), hence no interference is to be expected in unpolluted waters.Limit of Detection The limit of detection for organotin in the proposed procedure is determined by two factors: the limit of detection of CSV for tin and the background level of inorganic tin. The sensitivity of the measurement is not affected by surface-active organics because the CSV analysis is carried out after UV irradiation of the sample, hence the limit of detection is equal to that for inorganic tin at approximately 5 pmol dm-3 when a deposition time of 10 min is used.6 The background level of inorganic tin is low at about 50-100 pmol dm-3 in river water, and it is only 5-10 pmol dm-3 in unpolluted sea-water.6 The organotin concentration should be twice that of inorganic tin in order to give an accurate result, as the organotin concentra- tion is corrected for the contribution of inorganic tin to the total dissolved tin concentration. Therefore, the limit of detection for organotin is of the order of 20-200 pmol dm-3, depending on the actual concentration of inorganic tin. This work has been carried out with the support of the Procurement Executive of the Ministry of Defence. References Gilmour, C. C., Tuttle, J. G., and Means, J. C., in Marine and Estuarine Geochemistry, eds. Sigleo, A. C., and Hattori, A., Lewis Chelsey, MI, 1985, pp. 239-258. Francois, R., and Weber, J. H., Mar. Chem., 1988, 25, 279. Andreae, M. O., in Trace Metals in Sea Water, eds. Wong, C. S . , Boyle, E., Bruland, K. W., Burton, J. D., andGoldberg, E. D., Plenum, London, 1983, pp. 1-19. Balls, P. W., Anal. Chirn. Acta, 1987, 197, 309. Chapman, A. H., and Samuel, A., Appl. Organomet. Chem., 1988,2, 73. van den Berg, C. M. G., Khan, S. H., and Riley, J. P.. Anal. Chim. Acta, 1989, 222,43. Weber, G., Anal. Chim. Acta, 1986, 186, 49. van den Berg, C. M. G., in Chemical Oceanography, ed. Riley, J. P., Academic Press, London, 1989, vol. 9, pp. 197-245. Mackey, D. J., Mar. Chem., 1985, 16, 105. Paper 0104785F Received October 24th, 1990 Accepted December 20th, 1990