Environ. Sci. Technol. 2006, 40, 2206-2212
Nitroaromatic Reduction Kinetics as a Function of Dominant Terminal Electron Acceptor Processes in Natural Sediments L I S A A . H O F E R K A M P * ,† A N D ERIC J. WEBER‡ University of Alaska Southeast, 11120 Glacier Highway, Juneau, Alaska 99801, and National Exposure Research Laboratory, U.S. Environmental Protection Agency, 960 College Station Road, Athens, Georgia 30605-2700
The reductive transformation of p-cyanonitrobenzene (pCNB) was investigated in laboratory batch slurries exhibiting dominant terminal electron accepting processes (TEAPs). Pseudo-first-order rate constants (kobs) were measured for the reduction of pCNB in nitrate-reducing, iron-reducing, sulfate-reducing, and methanogenic sediment slurries. Reduction was extremely slow in nitrate-reducing slurries but increased in slurries exhibiting TEAPs with significant concentrations of solution phase Fe(II). As the reduction of pCNB progressed in the Fe(II) rich systems, significant but nonstoichiometric decreases in aqueous Fe(II) concentration were measured. Normalization of kobs to initial aqueous Fe(II) concentrations (kobs/[Fe(II)]t)0) gave values ranging from 0.0040 to 0.0052 d-1 µM-1 for nitratereducing, iron-reducing, and methanogenic sediment slurries as well as sulfate-reducing sediment slurries in which lactate served as a source of organic carbon. The kobs/ [Fe(II)]t)0 ratios were 1-fold greater for sulfate-reducing batch slurries amended with acetate and iron-reducing slurries equilibrated with a 3% H2 atmosphere indicating that the electron source and system parameters such as pH play a determinant role in the reaction kinetics. Although these data demonstrate that aqueous phase Fe(II) must be present for significant reduction to occur, a limited role for aqueous phase Fe(II) as a quantitative indicator of reactivity is suggested.
Introduction The nitroaromatic functional group is common among agrochemicals, textile dyes, munitions, and other industrial chemicals, thus elucidation of the reaction processes controlling the transformation of nitroaromatic chemicals (NACs) in ecosystems is essential. The production, deployment, and improper disposal of NACs has resulted in environmental accumulations of these chemicals, and several different approaches for the remediation of contaminated sites have been summarized (1, 2). Several investigators have demonstrated the utility of monosubstituted nitrobenzenes as probe chemicals for elucidating reaction pathways and kinetics for electron transfer of NACs in both model and natural systems (3, 4). * Corresponding author phone: (907)796-6538; e-mail: jflh@ uas.alaska.edu. † University of Alaska Southeast. ‡ U.S. Environmental Protection Agency. 2206
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 7, 2006
Studies of model and natural systems have shown the reductive transformation of NACs under anaerobic conditions occurs primarily through abiotic processes as opposed to biotransformation (3, 5-10). Model systems designed to mimic iron- and sulfate-reducing conditions have provided insight into the possible processes controlling reduction under these conditions (3, 10, 11). The general conclusions drawn include the following: (1) the importance of surface associated ferrous iron as a chemical reductant for the reductive transformation of NACs, (2) precursor complex formation as the initial step in the mechanism for electron transfer, and (3) formation of Fe(II) surface sites as potential rate-determining steps in the overall reduction of NACs. Extrapolation to natural systems is limited given the small number of studies designed to elucidate pathways for electron transfer in these systems. One such study compared reductive rate constants for a series of NACs under both field (field injections, in situ microcosms) and laboratory (batch experiments) conditions and concluded that surface-associated Fe(II) was largely responsible for reduction in the anaerobic portion of a landfill leachate plume (5). Furthermore, high aqueous Fe(II) concentrations that remained invariant throughout reduction of the NACs suggested adsorption of aqueous Fe(II) to mineral surfaces was largely responsible for the regeneration of reactive sites. In this laboratory the reductive transformation of a model compound, p-cyanonitrobenzene (pCNB), was investigated in a sediment column that had been characterized with respect to redox zonation (12). Characterization of the redox zones was assessed by the measurement of the aqueous concentrations of NO2-, NO3-, Mn(II), Fe(II), and SO42-. pCNB was reduced in the nitrate-reducing oxic zone of the column to p-cyano-N-hydroxylaniline (pCNH), which was reduced further to p-cyanoaniline (pCNA) in the iron-reducing zone:
This study further explores the relationship between reduction kinetics and the dominant terminal electron accepting processes (TEAPs) in sediment slurries. Toward this goal, the reductive transformation of pCNB was investigated in laboratory batch slurries exhibiting specific TEAPs. The results were used to assess the role of aqueous Fe(II) as a chemical indicator of reactivity.
Experimental Section Chemicals. p-Cyanonitrobenzene (pCNB), p-cyanoaniline (pCNA), sodium acetate (NaC2H3O2‚3H2O), sodium lactate (NaC3H5O3, 60% w/w), potassium nitrate (KNO3), and sodium sulfate (Na2SO4) were purchased from Aldrich Chemical Corporation and used as received. Sediment. Sediment was collected from the Oconee River (OR), Athens, GA at a water depth of 38 cm. The sediment was processed for use in batch experiments and analyzed for mineral content as described in ref 12. Results of the sediment analysis are presented in the Supporting Information. Analysis of the sediment-associated pore water after col10.1021/es051780k CCC: $33.50
2006 American Chemical Society Published on Web 02/22/2006
lection showed the presence of nitrate and sulfate but no reduced iron or manganese. Typical pH values of the pore water were between 6.5 and 6.7. Batch Experiments with Oconee River Sediment and Water. Batch experiments with OR sediment slurries were performed on the benchtop or in a COY anaerobic chamber (3% H2). Carbon sources included sodium acetate and sodium lactate. Potassium nitrate or sodium sulfate was employed as a nitrate or sulfate source. Stock solutions of pCNB were prepared in argon-purged, Nanopure water with acetonitrile (10% maximum) added when necessary to enhance solubility. A typical batch experiment consisted of a 65-mL serum bottle charged with the appropriate carbon source (acetate or lactate) and the appropriate amending reagent(s), added such that the final concentration would be 10 mM. To this was added 10 mL of sediment slurry and sufficient argon-purged site water to give a final volume of 60 mL. The serum bottles were crimp-sealed with Teflon-faced butyl rubber septa and placed on a shaker at room temperature. Equilibration periods ranged from 10 min (unamended) to 12 weeks (methanogenic). Specific equilibration periods are provided in the Supporting Information. After the equilibration period, a volume of pCNB stock solution was added to provide a concentration of approximately 0.1 mM. This addition marked t0 for the kinetic experiments. Two 1-mL samples were withdrawn (gastight syringe) at appropriate time intervals. The first sample was filtered through a 0.2 µm PTFE filter and analyzed for inorganic anions, pCNB, and any reduction products. The second sample was filtered through a 0.2 µm PTFE filter, acidified with 12 M HCl for a final pH ) 1, and analyzed for Mn(II) and Fe(II). Rate constants for the reduction of pCNB were calculated based on pseudofirst-order kinetics. Iron and Manganese Analysis. A 50 µL aliquot of the acidified sample of the sediment-associated water was analyzed for Fe(II) and Mn(II) by isocratic ion chromatography on a cation-exchange column (IonPac CS5A, Dionex) with 20/80 pyridine-2,6-dicarboxylic acid (PDCA)/H2O as eluent. Separated metal ions were complexed with 4-(2pyridylazo)resorcinol (PAR), and the absorbance was monitored at 530 nm. Peaks were identified and quantified by comparison of retention times and peak area with standards. Anion Analysis. A 100-µL aliquot of the neutral sample of sediment-associated water was analyzed chromatographically for nitrate, nitrite, sulfate, and phosphate concentrations as described in ref 12. Sulfide Analysis. Sulfide levels in the sediment-associated water were determined spectrometrically according to the method of Cline (13). Hydrogen Analysis. Headspace samples were measured for hydrogen using a Trace Analytical model RGA3 gas chromatograph (Menlo Park, CA) with a 1-mL loop and a reduction gas detector (RGD2). Aqueous-phase H2 concentration was calculated by assuming equilibrium as follows: [H2]aq ) (LP)/(RT), where [H2]aq is the concentration of dissolved H2 in molarity; L is the Ostwald coefficient for H2 () 0.02) (14); R is the universal gas constant; P is pressure (atm); and T is the temperature (K). Methane Analysis. Headspace samples were analyzed for methane using a HP 5890 GC equipped with an FID; carrier gas: He; carrier flow: 13.5 mL/min; packed column: Porapak N, 80/100 mesh, 6 feet × 1/8 in. × 0.085 in. stainless steel (Alltech, Deerfield, IL); oven temp: 50 °C. The methane peak was identified and quantified by comparison to methane standards. pCNB and Reduction Product Analysis. A 200-µL aliquot of the neutral sample of sediment-associated water was analyzed by liquid chromatography with UV/vis detection as described in ref 12.
FIGURE 1. Reduction kinetics of pCNB in an OR sediment slurry amended with 10 mM acetate as a function of redox conditions and time. Sediment slurry treated with acetate and pCNB at t ) 0 min: (a) changes in concentrations of the cationic redox indicators, [Mn(II)] (0), [Fe(II)] (O), and (b) pseudo-first-order reduction kinetics of pCNB (4).
Results and Discussion Kinetic Studies under Changing Redox Conditions. Initial experiments utilized batch systems prepared with unamended sediment slurries and sediment slurries amended with acetate at t ) 0. The slurries were charged with pCNB and sealed, and the concentrations of pCNB, the anionic redox indicators (NO3-, NO2-, and SO42-), and the cationic redox indicators (Mn(II) and Fe(II)) were monitored over time. Both unamended and amended slurries showed similar changes in redox conditions; however, changes in the batch systems amended with acetate occurred over shorter time periods. Results for acetate amended slurries are shown in Figure 1. Aqueous Mn(II) and Fe(II) were detected shortly after sealing the serum bottles and showed a slow but gradual increase over the course of the experiment (Figure 1a). Anion analyses (data not shown) indicated quick reduction of nitrate and then nitrite. Sulfate reduction (nominal sulfate concentration of 90 µM) did not occur over the time period of the experiment nor were any significant changes in the pH of the sediment associated water observed. Plots of time elapsed versus the ln([pCNB]t /[pCNB]0) suggest an increase in the disappearance rate for pCNB at ∼30 d (Figure 1b), which is nearly coincident with increases in the aqueous [Fe(II)] at 28 d (Figure 1a). Analysis of Figure 1b from 0 to 23 d and from 28 to 56 d shows a doubling in the rate at which pCNB degrades (0.008 d-1 vs 0.015 d-1). Attempts to fit the kinetic data to second-order kinetics (i.e., including [Fe(II)]aq in the rate equation) were not successful for this data set or subsequent data sets. Sediment Incubation and Characterization of Dominant TEAPS. Subsequent experiments measured the reduction of pCNB in sediment slurries in which a dominant TEAP had VOL. 40, NO. 7, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
2207
TABLE 1. Reduction of PCNB in Oconee River Sediment Exhibiting Various TEAPsa TEAP
[Fe2+]aq (µM)
[S(II)] (µM)
[CH4] (µM)
[H2]aq (nM)
kobs (d-1)b (r2)
kobs/[Fe2+] (µM d)-1
mole balance
nitrate red.c iron red. iron red. iron red.g iron red.g sulfate red.h sulfate red.h sulfate red.i sulfate red.i methanogenic methanogenic methanogenic
0.0049 0.0041 0.0048 0.0195 0.0272 0.0206 0.0220 0.0049 0.0040 0.0052 0.0070 0.0041
8 54 77 100 100 87 88 89 98 83 81 98
a Redox indicators were measured prior to addition of pCNB. See Table 1S for equilibration periods. b First-order rate constant for the reduction of pCNB with initial [pCNB] ) 100 µM. r2 is shown following kobs in parentheses. c Represents the average of two batch slurry experiments (% R.D. < 10%). d nd ) not determined; sulfate levels g 90 µM. e Below detection limit. f Lack of detectable CH4 suggests iron-reducing conditions. g Experiment performed in anaerobic chamber (3.0% H2). h Soluble carbon source was acetate. i Soluble carbon source was lactate.
been established prior to the addition of pCNB. Slurries exhibiting nitrate-reduction, iron-reduction, sulfate-reduction, and methanogenesis were prepared (Table 1). Measurements of redox indicators including the solution phase concentrations of Fe(II), S(II), and H2 and headspace CH4 levels were used to characterize the dominant TEAP (Table 1). The incubation periods required (see the Supporting Information) for the formation of the dominant TEAPs were determined by the time required to reach fairly constant concentrations of the soluble redox indicators (i.e., 20 µM. While this [H2]aq is higher than typically associated with iron reduction (0.2-0.6 nM), a lack of evidence for methanogenesis and sulfate reduction suggested iron reduction was the dominant TEAP in these slurries (15). Acetate-amended slurries with [H2]aq > 0.9 nM, [Fe(II)]aq > 20 µM, and with detectable CH4(g) in the headspace were considered methanogenic. The porewater pH values of the iron-reducing and methanogenic slurries increased slightly (6.8-7.1) over the course of the incubation periods and throughout reduction of pCNB. Slurries to which an external source of sulfate and organic carbon (lactate or acetate) were added became sulfate-reducing within 3.5 weeks and were distinguished by significant levels of aqueous S(II) (0.5-3.95 µM), dissolved hydrogen (6-11 nM), black coloration, and slightly increased pH levels (7.6 for lactate-amended, 8.0 for acetate-amended). Higher levels of aqueous Fe(II) were found in the sediment slurries amended with lactate. The origin of the higher Fe(II) concentrations in the lactate-amended sulfate-reducing slurries was not further explored from a microbial aspect. The elevated Fe(II) concentrations and pH in these slurries, however, may be in part due to reductive dissolution of iron oxide by sulfide (eq 1) (16):
former is a much slower process than the oxidation of sulfides presented in reaction 1 (19). The black precipitate present in all sulfate-reducing sediment slurries was attributed to iron sulfides and elemental sulfur. Scanning Electron Microscopy (SEM) and sequential extraction data (see the Supporting Information) confirmed the presence of iron sulfides. The presence of polysulfides, was not verified.
8Fe2O3 + 8H2S + 32H+ f 16Fe(II) + S8 + 24H2O
Very limited transformation of pCNB was observed over the course of 10 weeks in nitrate-reducing sediment slurries. These slurries exhibited low, but detectable levels of aqueous Fe(II) (
