Environ. Sci. Technol. 2004, 38, 1746-1752
Nonbiological Removal of cis-Dichloroethylene and 1,1-Dichloroethylene in Aquifer Sediment Containing Magnetite M A R K L . F E R R E Y , * ,† RICHARD T. WILKIN,‡ ROBERT G. FORD,‡ AND JOHN T. WILSON‡ Minnesota Pollution Control Agency, 520 Lafayette Road, St. Paul, Minnesota 55155-4194, National Risk Management Research Laboratory, Office of Research and Development, U.S. Environmental Protection Agency, 919 Kerr Research Drive, Ada, Oklahoma 74820
The U.S. EPA Technical Protocol for Evaluating Natural Attenuation of Chlorinated Solvents in Groundwater emphasizes biological reductive dechlorination as the primary mechanism for destruction of chlorinated solvents. However, biological reductive dechlorination could not explain the removal of cis-dichloroethylene (cis-DCE) and 1,1DCE from a plume of contaminated groundwater in Minnesota. Several recent laboratory studies have demonstrated that common iron minerals such as magnetite can also transform chlorinated alkenes. Laboratory microcosms were constructed with sediment from three depth intervals in the aquifer near the source of the plume. The microcosms were autoclaved to prevent biological transformations. In these autoclaved sediments, the rates of removal of cis-DCE in samples from the shallow, intermediate, and deeper depth intervals in the aquifer were 0.58 ( 0.09, 2.29 ( 0.26, and 0.31 ( 0.08 per year at 95% confidence. The rate of removal of 1,1-DCE in sediment from the shallow interval was 1.37 ( 0.50 per year. The rates of removal in the microcosms are similar to the rates of attenuation observed in the field. Magnetite was identified in the sediment by X-ray diffraction and optical microscopy. Published rates of transformation of cisDCE by magnetite are consistent with the rates of removal in the microcosm study.
Introduction Groundwater contamination with chlorinated solvents is one of our most persistent environmental problems. Despite aggressive remediation efforts, the concentrations of these compounds or their transformation products often remain above regulatory limits for decades. Monitored natural attenuation (MNA) is an important alternative for managing chlorinated solvents in groundwater (1). Currently, the U.S. Environmental Protection Agency (U.S. EPA) protocol for evaluation of MNA focuses on natural attenuation achieved through natural biotransformation processes (2). The familiar pattern of microbial reductive * Corresponding author phone: (651)296-7775; e-mail:
[email protected]. † Minnesota Pollution Control Agency. ‡ U.S. Environmental Protection Agency. 1746
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 6, 2004
dechlorination of trichloroethylene (TCE) to cis-dichloroethylene (cis-DCE), then to vinyl chloride, and then to ethylene or ethane (3, 4) is often cited as the primary mechanism leading to the degradation of chlorinated ethylenes in groundwater. The current U.S. EPA protocol and U.S. EPA screening model used to evaluate the prospects for monitored natural attenuation of chlorinated solvents in groundwater presume that the most important mechanism for contaminant destruction is biological reductive dechlorination (2, 5). With the exception of the hydrolysis of 1,1,1trichloroethane (TCA) to acetate and an elimination reaction of TCA to form 1,1-dichloroethene (1,1-DCE), nonbiological transformations are ignored. At the Twin Cities Army Ammunition Plant (TCAAP) in Minnesota, disposal of TCE and 1,1,1-TCA produced a plume of contaminated groundwater that was over 8 km long. The development of such a large plume was consistent with the high seepage velocity of groundwater in the aquifer system and the prevailing redox status of the aquifer, which is manganese-reducing or iron-reducing. Sulfidogenic or methanogenic conditions have not been observed. Appreciable reductive dechlorination of chlorinated aliphatic contaminants is not expected to occur in groundwater poised at these redox levels (2). Nonetheless, fate and transport modeling indicated that contaminant mass was being lost from the groundwater over time and distance. A first-order rate constant for removal of chlorinated ethylenes of at least 0.17 per year was necessary to account for the size of the plume and the distribution of contaminants within the plume in 1998 (6). This rate of attenuation for TCE was within the range of rates reported for other sites where microbial reductive dechlorination is occurring (7, 8). Reductive dechlorination could account for removal of TCE; the groundwater plume consistently contains cis-DCE and 1,1-DCE, attributed to the reductive dechlorination of TCE and the elimination reaction of 1,1,1-TCA, respectively. However, the concentrations of cis-DCE and 1,1-DCE did not increase as the concentrations of TCE and 1,1,1-TCA declined. Potential transformation products of cis-DCE and 1,1-DCE include vinyl chloride and ethene, but neither product was measured in the groundwater at concentrations above 1 µg/L when selected monitoring wells on site were sampled by U.S. EPA in 1996, 1997, and 1998 (6). Long-term monitoring beginning in 1989 indicated that the concentrations of the reductive dechlorination daughter products have never exceeded 10% of the concentration of TCE (Tables S1-S4 in the Supporting Information). Thus, microbial reductive dechlorination alone cannot explain the behavior of cis-DCE and 1,1-DCE in the plume. DCE can be biologically oxidized to carbon dioxide under iron-reducing and manganese-reducing conditions (9, 10). Recent reports in the literature demonstrate that a variety of iron and sulfur minerals in aquifer material can chemically transform chlorinated alkenes or alkanes (11-17). To determine the role of anaerobic biological degradation and nonbiological transformations of cis-DCE and 1,1-DCE in groundwater at the TCAAP, microcosms were constructed with aquifer sediment collected at the site. We found that the rates of removal in the sediment as collected and in autoclaved sediment were equivalent, demonstrating that a nonbiological process is primarily responsible for removal of the dichloroethenes. To our knowledge, this is the first study to show that nonbiological degradation is an important mechanism for the natural attenuation of cis-DCE and 1,1-DCE in a plume of contaminated groundwater. 10.1021/es0305609 CCC: $27.50
2004 American Chemical Society Published on Web 02/03/2004
FIGURE 1. Simplified depiction of the plume of chlorinated ethylenes in groundwater at the Twin Cities Army Ammunition Plant in Minnesota (TCAAP). The black circle is the location from which core samples were acquired for the laboratory studies. The other open circles are the locations of monitoring wells that describe the behavior of chlorinated ethylenes in the plume (see ref 6 and the Supporting Information). The dashed line is the location of a geological cross section presented in Figure S1 in the Supporting Information. The down-gradient portion of the plume is not depicted.
Materials and Methods Sediment Collection. The Twin Cities Army Ammunition Plant is located 10 km north of St. Paul, MN (Figure 1). Groundwater flows to the Mississippi River approximately 9 km to the southwest. The aquifer containing the water table at TCAAP is the Hillside Sand, an unconsolidated sandy aquifer that varies in thickness from 8 to 137 m. Below the Hillside Sand is an aquifer of fractured dolomite, the Prairie du Chien Group (Figure S1 in the Supporting Information). Disposal of large quantities of spent solvents from the 1940s through the 1960s contaminated groundwater in the Hillside Sand and the underlying bedrock aquifers. Sediment samples at the TCAAP site were collected from a Rotosonic drill core recovered from 50 to 56 m below the ground surface, penetrating the top 5 m of the unconsolidated sandy aquifer. The sediment core was divided into three samples corresponding to depths of 1.2-1.5, 3.0-3.6, and 4.6-6.1 m below the water table in the aquifer, subsequently referred to as shallow, intermediate, and deep sediments (see Figure 1 for a map showing the location of the samples and Figure S1 in the Supporting Information for a geological cross section). To minimize exposure to air in the field and in the laboratory, the sediment samples were sealed in glass Mason jars immediately after collection. The jars contained no headspace. The samples were refrigerated until they were opened inside an anaerobic glovebox for chemical analysis or to construct microcosms. Microcosms. To protect anaerobic microorganisms from oxygen in the atmosphere, all manipulations in the preparation of microcosms were carried out in an anaerobic glovebox containing 2-5% (v/v) hydrogen and less than 1 ppmv oxygen. Microcosms were prepared in 25 mL glass serum bottles. Groundwater from the site was added to the wet sediment to make a slurry (10% groundwater, 90% sediment (v/v)). A 45 g amount of this slurry was transferred to the serum vials. Each microcosm contained 38.5 ( 1.7 g of dry sediment and 6.5 ( 0.3 mL of water. Half of the microcosms were autoclaved overnight to provide killed controls. Living and autoclaved microcosms were prepared from material
collected from the shallow, intermediate, and deep sediments, for a total of six experimental treatments containing sediment. A seventh treatment contained autoclaved reverse-osmosis water and no sediment. All microcosms received 1.0 mL of a stock solution containing cis-DCE and 1,1-DCE. The different treatments were spiked at different initial concentrations. The initial concentrations of cis-DCE varied between 500 and 2000 µg/L. The initial concentrations of 1,1-DCE varied between 10 and 50 µg/L. Every microcosm in a particular treatment was spiked to the same initial concentration. The microcosms were sealed with a Teflon-faced gray butyl rubber septum and a crimp cap and then were stored, inverted, in the same anaerobic chamber at room temperature (20-22 °C). Container controls were prepared by spiking sterile water in serum bottles with the stock DCE solutions. Analytical. (a) Chlorinated Aliphatic Hydrocarbons in Water. Microcosm sediments were resuspended prior to sampling by agitation with a vortex mixer. The solids were allowed to settle. The crimp cap and septa were removed, and approximately 1 mL of supernatant was removed and transferred to a 20 mL glass serum tube containing 14 mL of distilled water and 0.15 g of Na3PO4‚12H2O. The serum tubes were sealed with a Teflon-faced septum and a crimp cap and stored at 4 °C until analyzed. The precise volume of water removed from the microcosms was determined gravimetrically. Samples were analyzed for chlorinated aliphatic hydrocarbons by headspace sampling and GC/MS analysis (modified EPA Method 524.2, 18). (b) Sediments. To prepare sediments for mineralogical characterization, composite samples were dried to constant mass under a 97:3 (v/v) N2:H2 gas mixture in an anaerobic glovebox. Dried sediment composites were sieved to isolate the
