Environ. Sci. Technol. 2004, 38, 457-467
Can Degradation Products Be Used as Documentation for Natural Attenuation of Phenoxy Acids in Groundwater? LOTTE A. REITZEL,* NINA TUXEN, ANNA LEDIN, AND POUL L. BJERG Environment & Resources DTU, Technical University of Denmark, Building 115, Bygningstorvet, DK-2800 Kgs. Lyngby, Denmark
In situ indicators of degradation are important tools in the demonstration of natural attenuation. A literature survey on the production history of phenoxy acids and degradation pathways has shown that metabolites of phenoxy acid herbicides also are impurities in the herbicide products, making the bare presence of these compounds useless as in situ indicators. These impurities can make up more than 30% of the herbicides. Degradation of phenoxy acids was demonstrated in microcosm experiments using groundwater and sediment contaminated with MCPP, dichlorprop, and related compounds such as other phenoxypropionic acids and chlorophenols. Field observations at two phenoxy acid-contaminated sites showed the occurrence of several impurities including metabolites in the groundwater. Neither the microcosm experiments nor the field observations verified that metabolites were actually produced or accumulated in situ. However, it was demonstrated that the impurity/parent herbicide ratios can be useful in situ indicators of degradation.
Introduction Herbicides as phenoxy acids have frequently been detected in groundwater worldwide (1, 2). This groundwater contamination can be a result of large-scale agriculture but may also be due to point sources caused by, for instance, landfills, machine pools, or market gardens. Natural attenuation as a remedial process for phenoxy acids from point sources has been suggested because of their degradability under aerobic conditions (3, 4). During natural attenuation as a remedy for contamination plumes in groundwater, the degradation of contaminants must be documented. A detailed mapping of the plume can demonstrate mass reduction of dissolved contaminants; however, degradation is not always easily distinguished from other processes such as dispersion and sorption. Consumption of electron acceptors and production of reduced species in the plume can be used to indicate biodegradation (5), although this can only be done if the contaminant in question is the primary substrate (i.e., the contaminant constitutes a major part of the biodegradable dissolved organic carbon). This requirement is fulfilled by petroleum hydrocarbons from gasoline stations but not for specific contaminants in landfill leachate plumes (6). More direct evidence of degradation could be the production of degradation products, the changes * Corresponding author phone: +45 4525 1591; fax: +45 4593 2850; e-mail:
[email protected]. 10.1021/es030039e CCC: $27.50 Published on Web 12/05/2003
2004 American Chemical Society
in the isotopic signature along a flow path (7), or in the case of chiral compounds such as phenoxy propionic acids, even the changes in the enantiomeric ratio (8, 9). The presence of lower chlorinated compounds (e.g., ethene, mono- and dichloroethene from PCE and TCE) are examples of degradation products that have previously been used as documentation of biodegradation (10). Metabolites can only be used to document degradation of their parent compounds if they are sufficiently specific. Beller (11) defined an ideal indicator of in situ biodegradation as a compound that could definitively be related to the metabolism of a specific compound (i.e., it should not be naturally occurring or commercially produced). The compound should be stable enough to make detection probable, but it should also be an intermediate of mineralization rather than a dead-end product. The most commonly reported metabolites of phenoxy acid degradation are the corresponding chlorophenols. The identification of these chlorophenols at phenoxy acid sites may therefore be indicative of phenoxy acid degradation. However, this interpretation could be complicated by the presence of chlorophenols as well as other phenoxy acids as impurities in the herbicides (12-14) due to incomplete synthesis reactions and insufficient cleanup procedures in herbicide production. This point deserves more attention because detection of such impurities is sometimes mistakenly interpreted as being metabolites. For instance, Grass et al. (15) and Kelly et al. (16) suggested 4,6-dichloro-o-cresol and other phenolic compounds with repositioned chloro and methyl groups as metabolites, without considering whether the compounds were initially present as impurities. The aim of this study is to investigate if metabolites can be used as indicators of phenoxy acid degradation in the field, and how in situ degradation can be quantified. The required information was obtained using four approaches: (i) studying the history of production (i.e., the content of impurities in phenoxy acid herbicides at different times); (ii) reviewing the fate of phenoxy acids and related chlorophenols by providing an overview of known and proposed degradation pathways; (iii) laboratory microcosm experiments to study degradation of phenoxy acid herbicides and related compounds in contaminated groundwater; and (iv) field observations in phenoxy acid herbicide-contaminated groundwater.
Materials and Methods Field Sites. The Bornholm site (Figure 1a) has operated as a machine pool (common supply of machinery for field work) since 1973, but the site may have been contaminated by pesticides earlier on, from the handling of spraying equipment. The aquifer is shallow and unconfined, with layers of moraine clay and silt underlain by sandstone. A leaky manure tank has resulted in elevated concentrations of, for example, NH4+ and nonvolatile organic carbon (NVOC). The tank is also a major source of herbicides (although not the only source) since over the past few years it has been used to collect excessive water from the washing of pesticide-spraying equipment. The transformation of NH4+ and NVOC has led to nitrate- and manganese-reducing conditions at the site. The groundwater and aquifer material for the microcosm experiments was sampled close to the manure tank. Monitoring wells exist at three different depths: 2, 7, and 14 m below the surface. Sjoelund is an old landfill located in a former gravel pit (Figure 1b). The landfill was in use from 1965 to 1975 and has since been covered with a layer of clay. It has no liners VOL. 38, NO. 2, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
457
FIGURE 1. Map of field sites. Bornholm is a former machine pool, and Sjoelund is an old landfill. The level of the water table is shown as heads (meters above sea level).
TABLE 1. Hydrochemistry in the Microcosmsa microcosm
dichlorprop (µg/L)
MCPP (µg/L)
O2 (mg/L)
NO3--N (mg/L)
Mn2+ (mg/L)
Fe2+ (mg/L)
SO42--S (mg/L)
NH4+-N (mg/L)
NVOC (mg/L)
redox conditions during experiment
Bornholm 1 (B1) Bornholm 2 (B2) Sjoelund 1 (S1) Sjoelund 2 (S2) Sjoelund 3 (S3) Sjoelund 4 (S4)
5300 15 70 5
