Environ. Sci. Technol. 2004, 38, 4510-4518
Sorption of the Herbicide Dichlobenil and the Metabolite 2,6-Dichlorobenzamide on Soils and Aquifer Sediments LISELOTTE CLAUSEN,* FLEMMING LARSEN, AND HANS-JØRGEN ALBRECHTSEN Environment & Resources DTU, Technical University of Denmark, Bygningstorvet, Building 115, DK-2800 Kgs. Lyngby, Denmark
The worldwide used herbicide dichlobenil (2,6-dichlorobenzonitrile) has resulted in widespread presence of its metabolite 2,6-dichlorobenzamide (BAM) in pore- and groundwater. To evaluate the transport of these compounds we studied the sorption of dichlobenil and BAM in 22 sediment samples of clayey till, sand, and limestone including sediments exhibiting varying oxidation states. Dichlobenil sorbed to all investigated sediments, with a high sorption in topsoils (Kd ) 7.4-17.4 L kg-1) and clayey till sediments (Kd ) 2.7-126 L kg-1). The sorption of the polar metabolite BAM was much lower than the sorption of dichlobenil but followed the same tendency with the highest sorption in the topsoils (Kd ) 0.24-0.66 L kg-1) and in the clayey till sediments (Kd ) 0.10-0.93 L kg-1). The sorption of both compounds was significantly higher (2-47 times) in the unoxidized (reduced) clayey till than in the weathered (oxidized) clayey till. Such a difference in sorption capacity could neither be explained by a higher organic carbon content, sorption to clay minerals, differences in clay mineralogy, nor by blocking of reactive surface sites on clay minerals by iron oxides. However, by removing an average of 81% of the organic carbon from the reduced clayey till with H2O2, the sorption decreased on average 50%. Therefore, most of the sorption capacity in the reduced clayey till was related to organic carbon, which indicates that sorption processes are affected by changes in organic compound composition due to weathering.
Introduction In recent years, herbicides and their degradation products have been detected in an increasing number of aquifers all over the world, including Europe (1) and the United States (2). One of the problematic compounds is the herbicide dichlobenil (2,6-dichlorobenzonitrile), since its discharge into groundwater can result in widespread presence of its metabolite 2,6-dichlorobenzamide (BAM) (Table 1). Dichlobenil is used worldwide for total weed protection in orchards, ornamental plants, and on nonagricultural sites (courtyards, driveways, parking areas, railroads, etc.). In U.S dichlobenil is still being used today, and e.g. approximately 110-161 tons of dichlobenil was used during 1993-1995 (5). In Denmark 556 tons of dichlobenil was sold from 1970 to 1996. However, since 1997 dichlobenil has been banned in * Corresponding author phone: +45 45 25 15 96; fax: +45 45 93 28 50; e-mail:
[email protected]. 4510
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 17, 2004
TABLE 1. Structural Formulas and Relevant Properties of Dichlobenil and BAM (3) dichlobenil
BAM
172.01 88 (20 °C) 18 (20 °C) 2.70
190.03 not found not found 0.77a
structural formulas
molecular weight (g mol-1) vapor pressure (mPa) solubility in H2O (mg L-1) Log Kow a
Reference 4.
Denmark because BAM has been detected in 21.4% of the investigated groundwater wells during 1992-2002 and at higher concentrations than the European Union drinking water guideline (0.1 µg L-1) allows in 6.6% of the wells (6). The high number of detections in Danish groundwater indicates that other countries, which have used dichlobenil intensively, would find their groundwaters contaminated by BAM, if this metabolite was included in their monitoring programs. To predict the mobility of dichlobenil and BAM in soils, aquitards, and aquifers it is crucial to understand the sorption of these compounds thoroughly. There are, however, only a few published studies on the sorption of dichlobenil (7-9) and of BAM (10, 11). Dichlobenil was only studied in topsoils with a high organic carbon content and with concentrations in milligrams per liter range. In these soils the sorption correlated well with the organic carbon content (TOC) (7, 8), which is also generally accepted as the main factor controlling sorption of nonionic pesticides (12-15). However, in subsoils and aquifers with a low organic carbon content, the sorption may not only be controlled by TOC, since other factors including mineralogy, cation exchange capacity, pH, and ionic strength may have a role to play (e.g. refs 15-21). Especially clay mineralogy and content may be important for nonionic pesticides (17, 19-21). Sorption of BAM has only been investigated in two aerobic sandy sediments from the same aquifer (Vejen, Denmark) with low organic carbon and clay content, and in these sediments the sorption of BAM was limited (Kd ) 0.00 and 0.08 L kg-1, respectively) (10). This was in accordance with the insignificant retardation of BAM observed in a field injection experiment preformed in the Vejen aquifer (11). However, the sorption of BAM has not been investigated in clayey sediments or in topsoils. Although clayey aquitards may play a crucial role for the transport of nonionic pesticides to the aquifers, no information exists on sorption of dichlobenil and BAM to such sediments. In clayey aquitards varying oxidation states can exist due to weathering effects, which may change the functional groups of the sediment available for sorption due to changes in mineralogy or organic carbon composition. To evaluate the risk of groundwater contamination sorption data on both oxidized and reduced aquitard sediment are necessary. Therefore, the purpose of this study was to provide sorption data for dichlobenil and BAM, at environmentally relevant low concentration levels (micrograms per liter range) for a range of sediments by investigating the sorption in topsoils, clayey subsoils overlying the groundwaters (aquitard sediments) with varying oxidation states, and aquifer sediments (sand and limestone). Furthermore, the main purpose is to improve the understanding of controlling parameters for sorption of both polar (BAM) and nonpolar (dichlobenil) 10.1021/es035263i CCC: $27.50
2004 American Chemical Society Published on Web 07/22/2004
profiles from the 6 sites were evaluated from the matrix sediment color and from the water chemistry (Table 2), and from all sites both oxidized and reduced sediment were sampled (Table 4). The aquifer at Staurbyskov consists of glacial sand deposits until 15 m below surface (mbs), where the aquifer is bounded by a clay aquitard. The water table was at 5.25 mbs, with aerobic conditions until 13 mbs. The Eskærhøj site had glacial sand deposits and minor clay lenses below the topsoil (∼0.75 m thick). This aquifer is bounded by clayey till at 16 mbs, with the water table at 5 mbs, and aerobic conditions until 7 mbs The sediments from the eastern part of Denmark are clayey till deposits (Weichselian glaciation) covering limestone-chalk aquifers (Hvidovre and Avedøre) or glacial sand aquifers (Kirke Syv and Strøby Egede). Due to weathering the upper 2.5-6 m of the investigated clayey tills were visible oxidized, and below these depths the clayey tills transition to reduced gray material. The sediment cores were collected by a Geoprobe macro corer (Staurbyskov and Eskærhøj), Shelby tubes (Hvidovre and Strøby Egede), or hand-drilling equipment (Avedøre and Kirke Syv). Groundwater was collected using a drive-point piezometer, where the water samples were forced to the surface using N2 gas displacement (Staurbyskov and Eskærhøj) or collected from groundwater abstraction wells close to the drill hole (Hvidovre). Pore water used in experiments with sediments from the unsaturated zones was collected using suction cups (Hvidovre) or, due to the presence of perched water tables, taken directly from the drill holes with a bailer-system (Strøby Egede, Avedøre, and Kirke Syv). At Staurbyskov and Eskærhøj pore water could not be collected from the unsaturated zone, and water for the experiments was sampled just below the groundwater table. Sediments Characterization. The outermost 1 cm of sediment from each end of a core section was discarded. The remaining sediment from each depth interval was thoroughly mixed and freeze-dried, where the high vacuum and lowtemperature prevent oxidation of samples, thus preserving sediment components. Grain size was characterized by sieving and particle distribution (Micrometrics Sedigraph 5000 ET). The specific surface area was measured by Multipoint N2-BET analysis (Micromeritics, Gemini III 2375 surface area analyzer), after being outgassed (Micromeritics, FlowPrep 060 Degasser) for 24 h at room temperature, as previously recommended (22). Total organic carbon content (TOC) was measured using a total elemental carbon analyzer (LECO CS-225) after removal of carbonates using 6% sulfurous acid (H2SO3).
FIGURE 1. Map of Denmark showing sampling sites. nonionic compounds by (1) determining the contribution to sorption from common aquifer minerals by investigating sorption to quartz, calcite, clay minerals (kaolinite, montmorillonite), and iron oxides (goethite, lepidocrocite); (2) determining the contribution to sorption from organic carbon by correlating sorption and content of organic carbon in the sediments and by investigating the effect on sorption by removing organic carbon from reduced clayey till sediments by hydrogen peroxide treatment; and (3) determining the influence of redox status on sorption by investigating sediment samples from both oxidized and reduced clayey till aquitards.
Materials and Methods Sediment and Water Sampling. A total of 22 sets of water and sediment samples were collected from six locations in Denmark (Figure 1). The redox status of the sediments in the
TABLE 2. General Chemistry of Pore and Ground Water Used in Sorption Experiments Strøby Egede
Kirke Syv
Avedøre
depth (mbs)
6i
Staurbyskov 8i
14i
6i
Eskærhøj 15i
3i
Hvidovre 12-50i
1.5-5.0i
1.5-2.2i
1.4-3.4i
pH EC (µS cm-1) DOCa (mg L-1) O2b (mg L-1) cations (mg L-1)c Na+ K+ Ca2+ Mg2+ Fe2+ anions (mg L-1)d ClNO3SO42TAL (mequiv L-1)e
7.0 556 8.1 5.6
7.7 648 3.4 0.7
7.4 621 4.1
