Electrochemical Reduction of Hexahydro-1, 3, 5-trinitro-1, 3, 5-triazine

38, 5, 1595-1599. Abstract. Electrochemical reduction of RDX, hexahydro-1,3,5-trinitro-1,3,5-triazine, ... Environmental Science & Technology 0 (proof...
1 downloads 0 Views 82KB Size
Environ. Sci. Technol. 2004, 38, 1595-1599

Electrochemical Reduction of Hexahydro-1,3,5-trinitro-1,3,5-triazine in Aqueous Solutions PASCALE M. L. BONIN,† DORIN BEJAN,† LEAH SCHUTT,† JALAL HAWARI,‡ AND N I G E L J . B U N C E * ,† Department of Chemistry and Biochemistry, University of Guelph, Guelph, Ontario, Canada N1G 2W1, and Biotechnology Research Institute, National Research Council of Canada, 6100 Royalmount Avenue, Montreal, Quebec, Canada H4P 2R2

Electrochemical reduction of RDX, hexahydro-1,3,5trinitro-1,3,5-triazine, a commercial and military explosive, was examined as a possible remediation technology for treating RDX-contaminated groundwater. A cascade of divided flow-through cells was used, with reticulated vitreous carbon cathodes and IrO2/Ti dimensionally stable anodes, initially using acetonitrile/water solutions to increase the solubility of RDX. The major degradation pathway involved reduction of RDX to the corresponding mononitroso compound, followed by ring cleavage to yield formaldehyde and methylenedinitramine. The reaction intermediates underwent further reduction and/or hydrolysis, the net result being the complete transformation of RDX to small molecules. The rate of degradation increased with current density, but the current efficiency was highest at low current densities. The technique was extended successfully both to 100% aqueous solutions of RDX and to an undivided electrochemical cell.

Introduction Hexahydro-1,3,5-trinitro-1,3,5-triazine (RDX, structure 1) is a widely used commercial and military explosive (1). During its manufacture, process wastewater (“pink water”) may be

discharged to the environment at concentrations up to 12 mg/L RDX (2). Plumes of RDX have developed from pink water lagoons (3), and from military testing, training, waste disposal, and demilitarization operations (4, 5). RDX and its degradation products are reported to be toxic, mutagenic, and carcinogenic (6, 7); as a result there is a need for remediation of contaminated sites, as well as treatment of process wastewater prior to discharge. * Corresponding author phone: (519) 824-4120, ext 53962; fax: (519) 766-1499; e-mail: [email protected]. † University of Guelph. ‡ National Research Council of Canada. 10.1021/es0305611 CCC: $27.50 Published on Web 01/28/2004

 2004 American Chemical Society

The decomposition of RDX can be initiated chemically (8, 9), photochemically (10, 11), biologically (12-17), or electrochemically (18, 19). Although mechanisms of decomposition are not known with certainty, postulated sites of attack generally involve one of the nitro groups of RDX. McCormick et al. (12) proposed that the degradation of RDX in municipal anaerobic sludge occurred via successive reduction of the nitro groups to give hexahydro-1-nitroso3,5-dinitro-1,3,5-triazine (MNX), hexahydro-1,3-dinitroso5-nitro-1,3,5-triazine (DNX), and hexahydro-1,3,5-trinitroso1,3,5-triazine (TNX). Subsequent fragmentation of the ring was proposed as the route to the end products hydrazine, 1,1-dimethylhydrazine, 1,2-dimethylhydrazine, formaldehyde, and methanol. More recently, Hawari et al. (13) reported the identification of an intermediate ring cleavage product, methylenedinitramine [MDNA, CH2(NHNO2)2], which was detected in almost stoichiometric amounts under anaerobic conditions. This suggested a different degradation pathway, in which RDX underwent ring cleavage via MNX or via direct enzymatic attack (14), in place of successive reductions of the nitro groups. Nitrous oxide and formaldehyde, which further biotransformed to CO2, were identified as end products. Electrolysis is a developing technology for the remediation of industrial wastes, and has successfully been applied to the remediation of explosives (18-20). Rodgers et al. (20) demonstrated electrochemical reduction of 2,4,6-trinitrotoluene (TNT) at a reticulated vitreous carbon (RVC) cathode in high current efficiency and high chemical yield. A subsequent oxidative treatment was needed to remove the reduction products, anilines, by precipitation (20). Electrochemical oxidation (18) and reduction (19) of RDX have been reported, but in both cases, degradation of RDX was based only on the loss of a chromatographic response and the reaction products were not identified. The aim of the present investigation was to develop an electrochemical method for the remediation of RDXcontaminated water under reductive conditions. Specific goals included finding conditions to optimize the disappearance of RDX and its potentially toxic reaction products, product identification, and the determination of chemical and current yields.

Experimental Section Chemicals. RDX was synthesized in our laboratory following a procedure derived from Hale (21). Its identity was confirmed by chromatographic comparison to an analytical standard (g98% pure) from Chem Service, Inc. and by melting point analyses (mp 205.5 °C and mmp 208 °C). The synthesized RDX had a purity of 98% by HPLC. MNX and TNX were provided by Defense Research and Development Canada (DRDC), Valcartier, Canada. Methylenedinitramine (MDNA) was obtained from the Rare Chemical Library of Aldrich, Canada. All other chemicals used were of reagent grade. HPLC Analysis. The equipment comprised a Waters model 600 pump (flow rate 1.0 mL/min), a Rheodyne injector containing a 20 µL sample loop, and a Waters model 486 tunable absorbance detector operated at 240 nm. The detector output was processed with Waters Millennium version 3.2 software. Separation of RDX, MNX, TNX, and MDNA was achieved on a Genesis C18 (25 cm × 4.6 mm) stainless steel column, using a 65:35 volumetric solution of acetonitrile (ACN)/H2O as the mobile phase. All solvents used were Fisher Scientific HPLC grade. Analyses for formaldehyde involved mixing the aqueous solution with an equal volume of Nash reagent (22), resulting in the development of a yellow VOL. 38, NO. 5, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

1595

color when positive, and then analyzing by UV-vis spectrophotometry at 412 nm ( ) 81.6 L mol-1 cm-1) with a Shimadzu 160 UV-vis recording spectrophotometer. Voltammetry. For cyclic voltammetry experiments, an EG&G model 273A potentiostat was employed with a conventional three-electrode glass cell. The working electrode was a glassy carbon electrode (3 mm diameter), and the counter electrode was a platinum foil. The reference electrode was a “RED ROD” Ag/AgCl electrode from Radiometer Analytical. Electrolysis. Electrolyses were carried out under galvanostatic conditions using a PAR model 363 potentiostat/ galvanostat. The plug flow reactor was a flow-through, sandwich-type Plexiglas cell with two inner compartments of 58 mm × 15 mm × 4.5 mm each, separated by a DuPont Nafion 424 cation exchange membrane (CEM), constructed in-house. For some experiments the membrane was removed, creating an undivided cell. The working electrode was a rectangular (55 mm × 15 mm × 4.5 mm) Duocel 100 pores per inch (PPI) RVC with a platinum feeder. RVC was chosen as cathode because of its high surface area, allowing relatively high currents to be used while maintaining low current density. Using the volume and porosity of the RVC electrode, and a graph provided by the manufacturer, the area of the electrode was estimated as 231 cm2. The counter electrode was an ELTECH IrO2/Ti dimensionally stable anode (DSA) (55 mm × 15 mm × 1.5 mm) with a stainless steel feeder. An IrO2/Ti DSA was chosen because it is well suited for oxygen evolution. The anolyte consisted of an aqueous solution of 20.0 mM H2SO4. The catholyte consisted of a 1:1 ACN/H2O volumetric solution of 10.0 mM RDX at pH 6.0, with 10 mM Na2SO4 as supporting electrolyte. ACN was used to increase the solubility of RDX to facilitate the study of reaction products. To mimic contaminated-water scenarios, other experiments employed aqueous solutions of 0.22 mM (48 mg/L) RDX (the solubility limit is ∼60 mg/L (23)). The catholyte and anolyte solutions were delivered separately, at equal flow rates, with a Masterflex C/L peristaltic pump to their respective compartments; Teflon tubing was used for connections. Two types of electrolytic experiments were performed. The first was designed to replicate a cascade of reactors. A given volume (30 mL) of catholyte was passed through the reactor, collected, and then reintroduced into the reactor for a second pass. This procedure was repeated as many times as necessary to mimic a series of “n” reactors. A sample of the catholyte was collected for HPLC and UV analyses after each pass. The second type of electrolytic experiment involved a single reactor, operating in recirculation mode, with the catholyte exiting the electrolytic cell returned to the storage reservoir. Samples were taken at time intervals equivalent to integral numbers of passes through the cell, based on the flow rate and the volume in the reservoir. For all the experiments, the catholyte volume was corrected for the aliquots removed to calculate current efficiencies.

Results and Discussion Characterization of RDX and Its Products by HPLC. Analytical standards for MNX, TNX, and MDNA were prepared by dissolving weighed amounts of the pure material in HPLC grade ACN. Under the aforementioned HPLC conditions, these standards had retention times as follows: MDNA (3.0 ( 0.1 min), TNX (3.9 ( 0.1 min), MNX (4.2 ( 0.1 min), and RDX (4.3 ( 0.1 min). MNX and RDX were incompletely separated. Attempts to improve the separation of the MNX and RDX peaks interfered with the resolution of the MDNA peak. Na2SO4 and NaNO2 gave responses at 2.1 ( 0.1 and 2.3 ( 0.1 min, respectively. MDNA [(O2NNH)2CH2] hydrolyzes rapidly in aqueous solutions of pH 3-8, resulting in the formation of nitramide 1596

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 5, 2004

FIGURE 1. Cyclic voltammograms (scan rate 20 mV/s) in deaerated ACN/water solutions of 10.0 mM RDX (1) and 5.0 mM MNX (2) and in the presence of the supporting electrolyte only (10.0 mM Na2SO4, 3). (NH2NO2) and formaldehyde (HCHO); nitramide in turn hydrolyzes to nitrous oxide (N2O) (14, 24). Solutions of MDNA in acetonitrile/water were sampled over time to monitor the progress of hydrolysis. The concentration of MDNA was found to decrease within a few hours (decrease of the intensity of the HPLC peak at 3.0 min), while formation of a peak attributed to nitramide was observed at 2.8 ( 0.1 min. Formaldehyde was also present in all the samples, as confirmed by UV analysis. RDX Stability. To dissociate the contributions of hydrolysis and electrolysis reactions to the overall degradation of RDX, the stability of RDX solutions was studied. Under normal laboratory illumination, nondegassed aqueous solutions of RDX at pH 6 were stable for several months, but in basic solutions, RDX concentrations declined due to hydrolysis (8, 9, 16). For example, when the pH of a nondegassed ACN/water solution of 10.0 mM RDX was adjusted to 10 with NaOH, 43% of the initial RDX hydrolyzed over 24 h, during which the pH decreased to 8. After an additional 24 h, a further 3% of the RDX concentration had hydrolyzed and the final pH was 7. These results are consistent with the findings of Hoffsommer et al. (8). In our experiments, the two main peaks seen by HPLC were at 2.6 and 3.8 min. The first peak corresponds to formamide (NH2CHO), which, under the operating conditions of our HPLC, was found to have a retention time of 2.6 ( 0.1 min. The second peak, at 3.8 min, is tentatively attributed to the ring cleavage product 4-nitro2,4-diazabutanal (4-NDAB, O2NNHCH2NHCHO), which has been identified as the main product of the alkaline hydrolysis of RDX by Balakrishnan et al. (9). Traces of MDNA and nitrite were also detected by HPLC, while the presence of formaldehyde was confirmed by UV. Voltammetry. Cyclic voltammograms were recorded at a scan rate of 20 mV/s in deaerated ACN/water solution of 10.0 mM Na2SO4 and 10.0 mM RDX or 5.0 mM MNX (Figure 1). The reduction peak of RDX was found to be at -1.32 V vs the Ag/AgCl electrode, and the reaction was irreversible since there is no evidence of an oxidation peak (curve 1). The reduction of MNX occurred in the same potential range as that of RDX, suggesting that MNX is readily reduced upon its formation (curve 2). Electrolytic Experiments. RDX was readily reduced under all the electrolysis conditions investigated. Using a cascade of divided flow-through reactors at a flow rate of 1.25 mL/ min, the rate of degradation of a nondegassed ACN/water solution of 10.0 mM RDX increased with current intensity (Table 1). For example, with a current of 50 mA (0.22 mA/ cm2), the RDX concentration had decreased to 5.42 mM after the first reactor and there was 0.06 mM RDX left after the equivalent of six reactors. At a current of 200 mA (0.86 mA/

TABLE 1. Concentration of RDX, MDNA, and HCHO (mM) and Current Efficiency (CE; %) in the Effluent of Each Equivalent Reactor for a Current of 50, 100, and 200 mAa no. of reactors

RDX

1 2 3 4 5 6

5.42 2.14 0.95 0.35 0.13 0.06

a

50 mA MDNA HCHO 3.48 3.04 1.95 0.80 0.35 0.23

5.15 4.64 4.44 3.72 3.20 3.00

CE

RDX

73.6 52.8 19.1 9.6 3.6 1.2

4.96 1.87 0.55 0.14 0.02 0.00

100 mA MDNA HCHO 2.65 1.39 1.07 0.64 0.34 0.20

3.02 2.63 2.02 1.66 1.07 0.74

CE

RDX

40.5 24.8 10.7 3.3 1.0 0.2

2.32 0.42 0.06 0.00 0.00 0.00

200 mA MDNA HCHO 2.19 0.86 0.62 0.36 0.12 0.11

2.53 1.62 1.17 0.96 0.38 0.26

CE 30.9 7.6 1.5 0.2

The starting solution was a nondegassed ACN/water solution of 10.0 mM RDX.

FIGURE 2. HPLC chromatograms of RDX and its major electrolysis products in an experiment carried out at 100 mA using a cascade of reactors at a flow rate of 1.25 mL/min after the equivalent of (A) two reactors, (B) four reactors, and (C) six reactors. The starting solution was a nondegassed ACN/water solution of 10.0 mM RDX. Retention times: nitrite, 2.3 min; nitramide, 2.8 min; MDNA, 3.0 min; MNX, 4.2 min; RDX, 4.3 min. cm2), the RDX concentration had decreased to 2.32 mM after the first reactor and only four reactors were needed to remove all the RDX. These experiments were replicated with a single reactor operating in recirculation mode; the rate of degradation was identical to that observed with a cascade of reactors (data not shown). Experiments were also carried out at different flow rates (0.75 and 1.75 mL/min). As expected, the proportion of RDX degraded during each pass through the reactor decreased as the flow rate increased, due to shorter contact times. For example, after one pass, the concentration of RDX had decreased by 62% with a flow rate of 0.75 mL/ min, while decreases of 46% and 22% were observed with flow rates of 1.25 and 1.75 mL/min, respectively. In all these experiments, the solution pH was found to be unchanged at 6 after each pass in the reactor. A similar distribution of products was observed at applied currents of 50, 100, and 200 mA. Figure 2 shows the chromatographic response for RDX and its major electrolysis products (except formaldehyde) as a function of the number of equivalent reactors (two, four, and six) in an experiment carried out at 100 mA. The degradation of RDX resulted in the formation of MDNA (3.0 min), nitramide (2.8 min), and nitrite (2.3 min). The chromatogram also shows two small unidentified peaks at 3.2 and 3.6 min, and a peak at 2.1 min

due to sulfate ions in the supporting electrolyte. MNX was detected as a small shoulder on the RDX peak. No TNX was detected and no chromatographic peak was observed in the retention time window for DNX, which appears between TNX and MNX (3, 25). An important observation is that no intermediate electrolysis products accumulated in the solution upon electrolysis, but were further reduced and/or hydrolyzed to small molecules. The presence of formaldehyde was confirmed by UV analysis; as for the other reaction products, their concentration decreased with subsequent passes in the reactor (Table 1). Each experiment was repeated at least twice. No significant difference in terms of degradation rate ((5%) or distribution of reaction products was observed. Aerobic vs Anaerobic Conditions. The preceding experiments were performed in nondegassed solutions. Doppalapudi et al. (19) have reported a slower rate of electrochemical reduction under air, but we observed that the rate of degradation was identical when the electrolysis was conducted under argon, suggesting that, under our conditions, any oxygen originally present was rapidly reduced at the cathode. The reaction products were also unchanged, except that the peaks at 3.2 and 3.6 min were not detected under argon. Mass Balance and Degradation Pathway. Table 1 provides quantitative data from which the mass balance may be deduced. The highest concentrations of intermediate products were obtained after one pass at 50 mA; when 4.58 mM RDX had reacted, 5.15 mM HCHO, 3.48 mM MDNA, and 0.40 mM MNX were found. The concentration of MNX was based on the area of the MNX shoulder in the RDX peak. Knowing that RDX and MNX each contain three atoms of carbon per molecule, and MDNA and HCHO each contain one, these data translate to a carbon mass balance of 72% (38% as HCHO, 25% as MDNA, and 9% as MNX). Factors contributing to the low mass balance for carbon include electrochemical reduction of formaldehyde to methanol (which was not detected under our conditions) and vaporization losses of both formaldehyde and methanol. At higher currents (100 and 200 mA after one pass), the chemical yields were lower (37% and 20%, respectively) due to the faster transformation of RDX to small molecules. After six passes through the reactor at a current of 50 mA, only 11% of the original mass could be accounted for (1% as MDNA and 10% as formaldehyde); this ratio decreased to 3% and 1% with currents of 100 and 200 mA, respectively (Table 1). Chemical yields of MNX were 1-2%, except after one pass at 50 mA in which a yield of 9% was observed. As reported by Halasz et al. (14), detection of MNX in only trace amounts is expected, because it can be further reduced to a hydroxylamine derivative that is unstable in water (26). This is consistent with the results of our voltammetry experiments indicating that MNX readily reduces. The yield of MDNA was 76% based on RDX consumed; considering the instability of aqueous solutions of MDNA, we conclude that the pathway for electrochemical degradation of RDX must overwhelmingly pass through MDNA. VOL. 38, NO. 5, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

1597

SCHEME 1. Reaction Pathway for the Electrochemical Reduction of RDXa

a

The hydroxylamine derivative was not detected. Adapted from ref

14.

Scheme 1, adapted from ref 14, represents a plausible reaction pathway for the electrochemical reduction of RDX. According to this scheme, RDX is reduced to MNX, which is further reduced to a hydroxylamine derivative. The latter must hydrolyze without significant competing reduction of a second nitro group, as evidenced by the high yield of MDNA and the absence of DNX or TNX. Our proposal for this transformation involves the release of 2 mol of HCHO and two inorganic nitrogen fragments, followed by the hydrolysis of MDNA to form nitramide and formaldehyde. These are further hydrolyzed or reduced, leaving only small molecules such as formaldehyde, methanol, nitrite, and N2O. Our electrolytic studies thus appear to follow pathways parallel to those of previous microbial studies under anaerobic conditions (12, 13) in which formaldehyde and methanol, nitrite, and nitrous oxide (which can be generated anaerobically from nitrite reduction (27)) have been reported as carbon-containing and nitrogen-containing end products, respectively. A second reduction route has been recently identified during the anaerobic biodegradation of RDX, in which initial denitration of RDX leads to the formation of 4-NDAB (28), the compound observed as the main product of the alkaline hydrolysis of RDX. However, we found no evidence for formation of 4-NDAB during the electrochemical reduction of RDX at pH 6. Current Efficiency. The electrochemical reduction of RDX involves two successive transfers of 2 electrons and 2 protons, yielding successively MNX and the hydroxylamine derivative, for an overall 4-electron reduction. Current efficiencies were calculated using eq 1 for applied currents of 50, 100, and 200

current efficiency ) (cnF/I)v

(1)

mA (Table 1). c is the concentration of RDX degraded (mol/ L), n is the number of electrons transferred (4), F is Faraday’s constant (96485 C/mol), I is the current (A ) C/s), and v is the flow rate (L/s). In each experiment, performed under amperostatic conditions, the highest current efficiency was observed after the first pass. The efficiency declined with subsequent passes, due to the decrease in the concentration of RDX, and hence more competition from reduction of water. After the first pass, a current efficiency of 74% was observed with a current of 50 mA, while efficiencies of only 41% and 31% were obtained after the first pass at 100 and 200 mA, respectively. Higher currents, which gave higher rates of degradation, give lower current efficiencies due to mass-transfer limitations, as reported by Doppalapudi et al. (19). Undivided Cell. On the laboratory scale, divided cells generally offer the advantage of higher overall current efficiency; they also avoid back-reactions, which occur when reaction products migrate to the counter electrode. Back1598

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 5, 2004

FIGURE 3. Concentration of RDX as a function of the number of equivalent reactors in a divided cell at a current of 50 mA (full squares), in an undivided cell at 50 mA (empty squares), and in an undivided cell at 200 mA (empty triangles). The starting solution was a nondegassed ACN/water solution of 10.0 mM RDX.

FIGURE 4. Experiment in 100% water: concentrations of RDX (full squares), MDNA (empty triangles), and HCHO (empty circles) as a function of the number of equivalent reactors. The starting solution was a nondegassed aqueous solution of 0.22 mM RDX. reactions were not a problem in our case, because the reduction reactions were electrochemically irreversible, and the rapid decomposition of the hydroxylamine precluded reoxidation. On a large scale, undivided cells are preferable because they are less complex to construct and avoid the use of membranes, which are both costly and susceptible to fouling. ACN/water solutions of 10.0 mM RDX were treated in a cascade of undivided reactors. For a given current (50 mA), the degradation of RDX was less efficient than in a divided cell (Figure 3). Nevertheless, full degradation, giving identical products, was achieved with higher currents and/ or a larger number of equivalent reactors. Fully Aqueous Solutions. We used 100% water as the solvent to simulate actual environmental conditions more closely. A cascade of divided reactors was used for these experiments. Starting with 0.22 mM (i.e., 48 mg/L) RDX in water, the main products observed were MDNA and HCHO. Even using a lower current of 10 mA (0.04 mA/cm2), the degradation of RDX was very rapid, and after a single pass through the reactor, the RDX concentration had decreased to 0.07 mM, and there was no RDX left after the equivalent of three reactors (Figure 4). After one pass the carbon mass balance was 55% (5% as MDNA and 50% as HCHO). Although the current efficiency (12% after one pass) was low, on account of the smaller initial RDX concentration and the competing water reduction, the degradation pathway in water appeared to be identical to that observed in ACN/water. In conclusion, this study confirmed that RDX is electrochemically active and easily reducible. Using a divided flowthrough cell, an RVC cathode, and a DSA, it was found that the rate of degradation of RDX increased with current density, and that total degradation was achieved with a small cascade of reactors. As the reaction proceeded, the intermediate reaction products did not accumulate. A plausible reaction pathway involves successive reduction of RDX to a mono-

nitroso compound and then a monohydroxylamino compound, followed by ring cleavage to MDNA, which further hydrolyzes, as well as formaldehyde and inorganic nitrogen compounds. These results offer proof of concept only; whether a practical technology for the treatment of RDXcontaminated groundwater is possible will depend on a suitable engineering design.

Acknowledgments We thank Dr. Annamaria Halasz and Louise Paquet of the Biotechnology Research Institute of the National Research Council of Canada for technical help. Financial support was provided by the Natural Sciences and Engineering Research Council of Canada through a postdoctoral fellowship to P.M.L.B. and a research grant to N.J.B., and by the U.S. Strategic Environmental Research and Development Program (DOE, DOD, and EPA).

Literature Cited (1) Urbanski, T. Chemistry and Technology of Explosives; Pergamon Press: Oxford, U.K., 1983; Vol. 3, pp 77-117. (2) Jackson, M.; Green, J. M.; Hash, R. L.; Lindsten, D. C.; Tatyrek, A. F. Nitramine (RDX-HMX) Wastewater Treatment at the Holston Army Ammunition Plant; Report ARLCD-77013; U.S. Army Armament Research and Development Command: Dover, NJ, 1978. (3) Beller, H. R.; Tiemeier, K. Environ. Sci. Technol. 2002, 36, 20602066. (4) Haas, R.; Schreiber, E.; von Low, E.; Stork, G. Fresenius’ J. Anal. Chem. 1990, 338, 41-45. (5) Myler, C. A.; Sisk, W. In Environmental Biotechnology for Waste Treatment; Sayler, G. S., Fox, R., Blackburn, J. W., Eds.; Plenum Press: New York, 1991; pp 137-146. (6) Yinon, J. Toxicity and Metabolism of Explosives; CRC Press: Boca Raton, FL, 1990. (7) Talmage, S. S.; Opresko, D. M.; Maxwel, C. J.; Welsh, C. J. E.; Cretella, F. M.; Reno, P. H.; Daniel, F. B. Rev. Environ. Contam. Toxicol. 1999, 161, 1-156. (8) Hoffsommer, J. C.; Kubose, D. A.; Glover, D. J. J. Phys. Chem. 1977, 81, 380-385. (9) Balakrishnan, V. K.; Halasz, A.; Hawari, J. Environ. Sci. Technol. 2003, 37, 1838-1843. (10) Glover, D. J.; Hoffsommer, J. C. Photolysis of RDX. Identification and Reaction of Products; Technical Report NSWC TR-79-349;

Naval Surface Weapons Center: Silver Spring, MD, 1979. (11) Hawari, J.; Halasz, A.; Groom, C.; Deschamps, S.; Paquet, L.; Beaulieu, C.; Corriveau, A. Environ. Sci. Technol. 2002, 36, 51175123. (12) McCormick, N. G.; Cornell, J. H.; Kaplan, A. M. Appl. Environ. Microbiol. 1981, 42, 817-823. (13) Hawari, J.; Halasz, A.; Sheremata, T.; Beaudet, S.; Groom, C.; Paquet, L.; Rhofir, C.; Ampleman, G.; Thiboutot, S. Appl. Environ. Microbiol. 2000, 66, 2652-2657. (14) Halasz, A.; Spain, J.; Paquet, L.; Beaulieu, C.; Corriveau, A.; Hawari, J. Environ. Sci. Technol. 2002, 36, 633-638. (15) Bhushan, B.; Halasz, A.; Spain, J.; Thiboutot, S.; Ampleman, G.; Hawari, J. Environ. Sci. Technol. 2002, 36, 3104-3108. (16) Fournier, D.; Halasz, A.; Spain, J.; Fiurasek, P.; Hawari, J. Appl. Environ. Microbiol. 2002, 68, 166-172. (17) Hawari, J.; Halasz, A. The Encyclopedia of Environmental Microbiology; John Wiley & Sons Ltd.: New York, 2002; pp 19791993. (18) Martins, A. M.; Ferreira, M.; Tremiliosi-Filho, G. Proceedings of the 200th Electrochemical Society Meeting, San Francisco, CA, 2001; Electrochemical Society: Pennington, NJ, 2001; Abstract 789. (19) Doppalapudi, R. B.; Sorial, G. A.; Maloney, S. W. Environ. Eng. Sci. 2002, 19, 115-130. (20) Rodgers, J. D.; Bunce, N. J. Environ. Sci. Technol. 2001, 35, 406410. (21) Hale, G. C. J. Chem. Soc. 1925, 47, 2754-2763. (22) Nash, T. Biochem. J. 1953, 55, 416-421. (23) Dannenfelser, R. M.; Yalkowski, S. H. Sci. Total Environ. 1991, 109-110, 625-628. (24) Lamberton, A. H.; Lindley, C.; Speakman, J. C. J. Chem. Soc. 1949, 1650-1656. (25) Groom, C. A.; Beaudet, S.; Halasz, A.; Paquet, L.; Hawari, J. J. Chromatogr., A 2001, 909, 53-60. (26) Corbett, M. D.; Corbett, B. R. In Biodegradation of Nitroaromatic Compounds; Spain, J. C., Ed.; Plenum Press: New York, 1995; p 151. (27) St- Jones, R. T.; Hollocher, T. C. J. Biol. Chem. 1977, 252, 212218. (28) Zhao, J.-S., Paquet, L., Halasz, A., Hawari, J. Appl. Microbiol. Biotechnol. 2003, 63, 187-193.

Received for review July 29, 2003. Accepted December 14, 2003. ES0305611

VOL. 38, NO. 5, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

1599