Environ. Sci. Technol. 2004, 38, 4106-4112
Biotransformation of Tributyltin to Tin in Freshwater River-Bed Sediments Contaminated by an Organotin Release J A M E S E . L A N D M E Y E R , * ,† TERRY L. TANNER,‡ AND BRUCE E. WATT§ U.S. Geological Survey, 720 Gracern Road, Suite 129, Columbia, South Carolina 29210, U.S. Environmental Protection Agency, 61 Forsythe Avenue, Atlanta, Georgia 30303, and Data Resources, Inc., 3005 Broad River Road, Columbia, South Carolina 29210
The largest documented release of organotin compounds to a freshwater river system in the United States occurred in early 2000 in central South Carolina. The release consisted of an unknown volume of various organotin compounds such tetrabutyltin (TTBT), tributyltin (TBT), tetraoctyltin (TTOT), and trioctyl tin (TOT) and resulted in a massive fish kill and the permanent closures of a municipal wastewater treatment plant and a local city’s only drinkingwater intake. Initial sampling events in 2000 and 2001 indicated that concentrations of the ecologically toxic TTBT and TBT were each greater than 10 000 µg/kg in surfacewater bed sediments in depositional areas, such as lakes and beaver ponds downstream of the release. Bedsediment samples collected between 2001 and 2003, however, revealed a substantial decrease in bed-sediment organotin concentrations and an increase in concentrations of degradation intermediate compounds. For example, in bed sediments of a representative beaver pond located about 1.6 km downstream of the release, total organotin concentrations [the sum of TTBT, TBT, and the TBT degradation intermediates dibutyltin (DBT) and monobutyltin (MBT)] decreased from 38 670 to 298 µg/kg. In Crystal Lake, a large lake about 0.4 km downstream from the beaver pond, total organotin concentrations decreased from 28 300 to less than 5 µg/kg during the same time period. Moreover, bed-sediment inorganic tin concentrations increased from pre-release levels of less than 800 to 32 700 µg/kg during this time. These field data suggest that the released organotin compounds, such as TBT, are being transformed into inorganic tin by bed-sediment microbial processes. Microcosms were created in the laboratory that contained bed sediment from the two sites and were amended with tributyltin (as tributyltin chloride) under an ambient air headspace and sacrificially analyzed periodically for TBT, the biodegradation intermediates DBT and MBT, and tin. TBT concentrations decreased faster [half-life (t1/2) ) 28 d] in the organic-rich sediments (21.5%) that characterized the beaver pond as compared to the slower * Corresponding author phone: (803)750-6128; fax: (803)750-6183; e-mail:
[email protected]. † U.S. Geological Survey. ‡ U.S. Environmental Protection Agency. § Data Resources, Inc. 4106
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 15, 2004
(t1/2 ) 78 d) degradation rate in the sandy, organic-poor, sediments (0.43%) of Crystal Lake. Moreover, the concentration of inorganic tin increased in microcosms containing bed sediments from both locations. These field and laboratory results suggest that biotransformation of the released organotins, in particular the ecologically detrimental TBT, does occur in this fresh surface-water system impacted with high concentrations of neat organotin compounds.
Introduction Organotin compounds, such as tetrabutyltin (TTBT) and tributyltin (TBT), were first used in the early 1920s as mothproofing agents. Today, these organometallic compounds are widely used as light and heat stabilizers in plastics, plastic plumbing, and vinyl home siding; as coatings for glass containers; as curing agents for silicone adhesives and caulks; as pesticides for wood treatment and preservation; and as antifungal agents for industrial water-cooling towers. TBT has been employed most extensively since the 1960s, however, as an antifouling agent on the hulls of ocean-going vessels due to the toxicity of TBT to barnacles, see grass, and other organisms that preferentially attach to the bottoms of ships. The Organotin Antifouling Paint Control Act of 1988 controls the amount and rate of TBT compounds that can leach from marine paints in contact with water to not exceed 4 µg of organotin cm-2 d-1). Such properties of organotin compounds also render TBT an ecological risk to nontarget organisms exposed to TBT in aquatic environments, such as in harbors and areas that service boats (1-3). Dissolved-phase concentrations of TBT less than 100 µg/L can cause growth and reproductive problems in mussels and oysters, and exposure to aquatic invertebrates is believed to cause endocrine disruption, such as imposex and intersex in gastropods (4, 5). The toxic effects of TBT on microbes in municipal wastewater treatment plants (WWTP) have also been reported at parts per trillion concentrations (6). In the aquatic environment, TBT is quickly removed from the water column and adheres to bed sediments because TBT has a high specific gravity (1.2), low solubility (less than 10 mg/L at 20 °C and pH 7.0), and log Kow values near 3.2 (6). Because the adsorption of TBT to sediments is reversible, contaminated sediments can act as a long-term source of dissolved-phase contamination to the overlying water column (7). The affinity of organotins for adsorption to sediments is positively correlated to the extent of organo-substitution on the tin, such that increasing adsorption is seen for monobutyltin (MBT) < dibutyltin (DBT) < TBT , TTBT (1). As such, organotins can readily bioaccumulate, especially in fish (8). Moreover, numerous investigations that confirm the deleterious effect of TBT on nontarget aquatic organisms have led The International Maritime Organization to call for a global treaty that bans the application of TBT-based paints starting January 1, 2003, and total prohibition by January 1, 2008 (9). Due to the toxicity of TBT on nontarget organisms, the fate of these compounds in the water column and bedsediment environments has been extensively studied, in particular for marine or estuarine surface-water systems that are most likely to be exposed to such contaminants (10-14). Biotransformation of TBT by bacteria, algae, and fungi does occur and is believed to proceed by successive debutylation reactions from TBT f DBT f MBT f inorganic tin [as Sn(IV)] via β-hydroxylation with appropriate dioxygenases (15). The rate-limiting step of this successive biotransfor10.1021/es030697z CCC: $27.50
2004 American Chemical Society Published on Web 06/24/2004
FIGURE 1. Study area, Red Bank Creek, Lexington County, SC, showing the sampling locations of the beaver pond and Crystal Lake. mation is the transformation of DBT f MBT (16). Each successive breakdown intermediate, however, is less toxic than TBT. Half-lives (t1/2) are faster in the water column as compared to bed sediments, with calculated half-lives for TBT in the range of days to months (17-19). In bed sediments, the half-life increases to between 1 and 3 months under oxic conditions and is on the order of years in sediments depleted of oxygen. Comparatively less information is known about the fate of organotins in fresh surface-water systems because of the much lower need to use anti-fouling paints in inland, freshwater bodies (20). Even less is known about the fate of TBT following an acute, point-source discharge of high concentrations of organotins, in either fresh or saline surfacewater systems. The investigation reported here was performed to determine the role that microbially mediated biotransformation processes played in the observed decrease in TBT concentrations in organotin-contaminated bed sediments of a freshwater river in South Carolina that received one of the world’s largest documented releases of organotin compounds.
Study Area In February 2000, an industry that used organotin compounds such as TTBT, TBT, TTOT, and TOT and was permitted to discharge partially treated wastes into a sewage system in Lexington County, SC, had an unpermitted discharge of an unknown volume and composition of organotin compounds into this sewage system. Upon reaching the local WWTP, the plant’s biological treatment was significantly reduced or killed off. This organotin-rich, untreated effluent left the WWTP and discharged through an underground pipe into Red Bank Creek (Figure 1). Residents along the creek reported seeing dead fish shortly thereafter. The WWTP had to close in late February 2000, and the downstream City of Cayce had to shut down a drinking water intake located on Congaree Creek. Four years after the release, neither the WWTP nor the drinking-water intake have been reopened. Moreover, the SC Department of Health and Environmental Control (DHEC) has proposed to include organotin compounds in its updates to the Hazardous Waste Management Regulations of South Carolina. In response to the release, a remedial investigation (RI) was initiated by the U.S. Environmental Protection Agency (EPA) in mid-2000 and determined that although the majority of the organotin release was flushed to downstream areas by surface-water flow, some contaminants were retained in the bed sediments. Some of these contaminated bed sediments may have also moved downstream, which may explain the observation of higher organotin concentrations detected in bed sediments at the mouth of lakes and ponds relative to river bed sediments (21). Red Bank Creek flows to the east
about 3.2 km through rural and residential areas, through a series of beaver ponds, until it empties into a 0.26 km2 manmade lake (Crystal Lake), continues downstream to the smaller Durham Pond, and then converges 1.6 km downstream with Congaree Creek (Figure 1). This creek flows for about 14.5 km to ultimately converge with the Congaree River. As part of the EPA’s RI, the composition of the bed sediments in the Red Bank Creek study area was determined by the U.S. Army Research and Development Center (USARDC) using standard grain-size analysis and a laser diffraction particle sizer, and the data were input into the Hydraulic Process Analysis System (HyPAS) for gradation curve generation. The grain sizes ranged from fine to coarse sands; to sands and some silts in depositional areas such as lakes and ponds; and to organic-rich mucks and partially decomposed plant material in swamps, marshes, and beaver ponds (22). To determine the potential for sediment erosion, a particle entrainment simulator was used on undisturbed bed-sediment samples collected from Crystal Lake. As such, critical shear stress values ranged from 0.013 and 0.106, and erosion constant values ranged from 0.006 to 0.205 (22). In general, the surface-water quality of the study area is typical of many “blackwater” streams and rivers that drain this area of central South Carolina, with water being characterized by acidic pH due to the lack of a significant geologic source of buffering capacity and the presence of high concentrations of organic acids from decomposing plant material.
Methods and Materials Field Investigation. The RI provided data for the organotin concentrations of more than 50 bed-sediment samples collected along the length of the impacted surface-water system. These bed-sediment samples were collected from between 0 and 8 cm below the bed-sediment/surface-water interface using an auger (21). The highest concentrations of total organotin compounds (sum of TTBT, TBT, DBT, and MBT) detected in bed sediments during the RI were found downgradient of the release in two depositional areas, a beaver pond and Crystal Lake (Figure 1), at 38 670 and 28 300 µg/kg, respectively (21). The highest bed-sediment concentrations of TBT, the organotin compound typically used to determine risks to human health and the ecosystem, were also located in the beaver pond and Crystal Lake (21). As such, these two locations provided the opportunity to observe and monitor the fate of TBT over time as well as to document the potential for the biotransformation of organotin to inorganic tin. Between March 2002 and January 2003, additional bedsediment samples were collected at these locations by the EPA and U.S. Geological Survey (USGS), with samples being collected in the same locations as had been collected during the earlier RI. The bed-sediment samples were collected using VOL. 38, NO. 15, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4107
a 0.7-m-long, stainless steel core device (AMS, Inc.) with a check valve at the bottom to prohibit the loss of sample during retrieval. The cores retrieved were from the top 4-10 cm of bed sediment and were contained in the core device in acetate liners. Ambient surface water/porewater was collected with each bed-sediment core. The cores were capped and transferred to the lab in a cooler. In the lab, multiple cores of bed-sediment material collected at each site were homogenized separately in 2-L polycarbonate containers; polycarbonate is known to resist adsorption or absorption of organotin compounds (18). These containers were stored in the dark at room temperature. The samples collected in Crystal Lake were taken from the northeastern shore in about 0.3 m of water. The samples consisted of sandy material with very little organic matter (0.43%, as determined by loss on ignition and expressed as a percentage of sediment dry weight). The samples collected in the beaver pond were collected near a dam, in about 1 m of water, and consisted of mud with abundant organic matter (21.5%, as determined by loss on ignition). Laboratory Microcosm Investigation. Because the potential exists for changes in the hydraulic conditions of Red Bank Creek to affect organotin concentrations over time through sediment transport (22), a laboratory-based study was performed to evaluate organotin fate under static conditions. Microcosms were created in the laboratory by placing about 30 g of a sediment-water slurry of either beaver pond or Crystal Lake bed sediments into 60-mL polycarbonate microcosms (Fisher Scientific, Inc.). Because these sediments were collected almost 2 yr after the initial release, organotin concentrations in these sediments were lower than that measured in 2000, suggesting that significant degradation had already occurred. Hence, the bed sediments were amended with TBT, as tributyltin chloride (TBT-Cl). A working solution was created by adding 0.339 g of a >97% TBT-Cl (Fluka Chemicals) stock into a glass vial containing 100 mL of autoclaved, deionized water, for a working solution concentration of 3390 mg/L TBT-Cl. To maintain this concentration, a stirrer bar and hot plate were used. Each microcosm received 2.5 mL of the working solution, for a final TBT-Cl concentration of 565 mg/L (565 mg/kg wet weight, assuming porosity near 0.50) per microcosm. Each microcosm was then capped with a polycarbonate screw-lid to create a 30-mL headspace of ambient air. Sterilized controls were prepared in the same manner and autoclaved for 1 h at 15 psi at 121 °C. All microcosm experiments were incubated in the dark at room temperature. At specified time points, 30-g sediment-water slurry samples were sacrificed and analyzed for the presence of TBT, the degradation byproducts DBT and MBT, and inorganic tin. Organotin Analytical Method. Organotin compounds in sediment from the field and laboratory microcosm experiments were determined by solvent extraction, derivatization, separation, and detection following a method given by Peven and Uhler (23). Each 30-g sediment bed-sediment sample (either collected in the field or from a sacrificed microcosm) was spiked with 25 µL of a surrogate standard solution consisting of 4000 mg/L tri-n-propyltin chloride and subsequently extracted with 50 mL of “extraction hexane”, a chromatographic grade hexane prepared with 0.05% tropolone. The extract was then dried by passing through a column of anhydrous sodium sulfate and placed into an inert environment chamber (anoxic glovebox) for derivatization following a method by Peven and Uhler (23). The derivatization step converts the organotin cations to more volatile compounds. Approximately 1 mL of 2 M n-pentylmagnesium bromide was added to the extract, and the derivatization was allowed to proceed for 15 min. The derivatized extract was then removed from the inert environment chamber and passed through a cleanup column of 10 g of Florisil/5 g of 4108
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 15, 2004
silica gel to remove simultaneously extracted, nontarget lipids. The cleanup column was rinsed with 60 mL of chromatographic grade hexane and combined with the initial extract. The extract was then concentrated to 1 mL, to which was added a 10-µL internal standard. The internal standard consisted of a 4000 mg/L solution of 1,4-dichlorobenzened4, naphthalene-d8, acenaphthalene-d10, phenanthrene-d10, chrysene-d12, and perylene-d12. The derivatized extract containing the internal standards was introduced to a Hewlett-Packard (HP) (Palo Alto, CA) 5980 gas chromatograph attached to an HP 5971 mass selective detector using an HP 7673A autosampler. The chromatographic capillary column was a Restek Rtx-5, 30 m × 0.25 mm i.d., 0.25 µm film thickness. The column flow was 1 milliliter per minute (mL/min) of helium. A solvent delay of 3.75 min with an initial oven temperature of 35 °C was held for 3 min. The oven temperature was ramped at a rate of 10 °C/min to a temperature of 200 °C with a hold time of
