Black Carbon-Mediated Destruction of ... - ACS Publications

Jul 20, 2010 - ranges where unexploded ordnance (UXO) renders excavation dangerous. Sulfides generated by biological sulfate reduction in sediments ar...
1 downloads 0 Views 418KB Size
Download PDF
Environ. Sci. Technol. 2010, 44, 6409–6415

Black Carbon-Mediated Destruction of Nitroglycerin and RDX By Hydrogen Sulfide WENQING XU, KATHRYN E. DANA, AND WILLIAM A. MITCH* Department of Chemical Engineering, Environmental Engineering Program, Yale University, New Haven, Connecticut 06520

Received April 22, 2010. Revised manuscript received July 2, 2010. Accepted July 7, 2010.

The in situ remediation of sediments contaminated with explosives, including nitroglycerin and hexahydro-1,3,5-trinitro1,3,5-triazine (RDX), is desirable, particularly at bombing ranges where unexploded ordnance (UXO) renders excavation dangerous. Sulfides generated by biological sulfate reduction in sediments are potent nucleophiles and reductants that may contribute to the destruction of explosives. However, moderately hydrophobic explosives are likely to sorb to black carbons, which can constitute 10-30% of sediment organic carbon. In this study, we evaluated whether the black carbons accelerate these reactions or simply sequester explosives from aqueous phase reactions. Using environmentally-relevant sulfide and black carbon concentrations, our results indicated that black carbons accelerated the destruction of both compounds, yielding relatively harmless products on the time scale of hours. For both compounds, destruction increased with sulfide and graphite concentrations. Using sheet graphite as a model for graphene regions in black carbons, we evaluated whether graphene regions mediated the reduction of explosives by promoting electron transfer from sulfides. Our results demonstrated that the process was more complex. Using an electrochemical cell that enabled electron transfer from sulfides to explosives through graphite, but prevented nucleophilic substitution reactions, we found that nitroglycerin destruction, but not RDX destruction, could be explained by an electron transfer mechanism. Furthermore, surface area-normalized destruction rates for the same explosive varied for different black carbons. While black carbon-mediated destruction of explosives by sulfides is likely to be a significant contributor to their natural attenuation in sediments, a fundamental characterization of the reaction mechanisms is needed to better understand the process.

Introduction In the United States, over 313 million kg of nitrogenous explosives had been released to the environment by 1992 (1). High concentrations in soils and sediments of two of these explosives, nitroglycerin and hexahydro-1,3,5-trinitro1,3,5-triazine (RDX), have resulted from such releases. For example, up to 130 mg/kg nitroglycerin has been detected in soil from a military installation (2); nitroglycerin has a suggested LD50 value of 30-1300 mg/kg for mammals (3). * Corresponding author phone: (203) 432-4386; fax: (203) 4324387; e-mail: [email protected]. 10.1021/es101307n

 2010 American Chemical Society

Published on Web 07/20/2010

Nitroglycerin and RDX are listed as class C possible human carcinogens (4), and are on the U.S. Environmental Protection Agency’s Contaminant Candidate List 3 (5). Marine sediment contamination is of concern from both direct contamination at marine bombing ranges and discharges from contaminated rivers. Pore waters can contain up to 5 mM hydrogen sulfide due to microbial sulfate reduction (6). As both potent nucleophiles and reductants (7-9), sulfides may affect the fate of nitroglycerin and RDX. However, because these explosives are moderately hydrophobic (e.g., log Kow ) 1.6 for nitroglycerin (10) and 0.9 for RDX (11)), they likely will sorb to black carbons, which constitute 10-30% of the total organic carbon in marine sediments (12, 13). Black carbons include chars from forest fires and soots from combustion processes. The importance of black carbons for sequestering hydrophobic contaminants in marine sediments has been noted (14, 15). A critical unresolved issue is whether the sorption of explosives to black carbons hinders their destruction by sequestering them from reactions with sulfides in the aqueous phase, or enhances their destruction by enabling black carbonmediated reactions with sulfides. Kemper et al. (16) reported that RDX was destroyed within 2 h in the presence of hydrogen sulfide and activated carbon, an anthropogenic model black carbon, at concentrations relevant to marine sediments. Products included nitrite and formaldehyde, indicating cleavage of the RDX ring, and no nitrosated reduction products were observed. Indeed, destruction rates slowed significantly for nitrosated analogues of RDX. No RDX decay was observed in the absence of either sulfides or activated carbon. Although slightly slower, similar destruction was observed with other black carbons, including graphite powder and pine and oak chars. Oh and Chiu (17) extended these results by demonstrating destruction of RDX and 2,4-dinitrotoluene by dithiothreitol, an organosulfide, in the presence of sheet graphite or soot particles. Although both groups claimed that the reaction involved mediation by black carbons of sulfide reduction reactions, rather than nucleophilic substitution reactions, neither group provided direct evidence. Oh and Chiu (17) suggested that the conductive properties of sheet graphite promoted reduction by sulfides based upon analogy to their previous studies employing elemental iron (18). Kemper et al. (16) observed that no RDX destruction occurred below 1.2 mM sulfides. This threshold suggested an adjustment of the solution reduction potential; thresholds should not be observed with nucleophilic substitution reactions. In addition to conductive graphene regions, black carbons contain oxygenated functional groups, including quinones, which may serve as electron shuttles between sulfides and explosives (19). Alternatively, sulfide reactions with surface functional groups, including quinones, may form organosulfides (20), which may serve as potent nucleophiles. Direct evidence for distinguishing reductive vs nucleophilic substitution reactions, and the structural requirements for the explosives and the black carbons remain important questions. One goal of this work was to extend the range of explosive structures susceptible to black carbon-mediated destruction by sulfides from the cyclic N-nitramine, RDX (16, 17), and the nitroaromatic, 2,4-DNT (17), to nitroglycerin, an aliphatic organonitrate. Previous research has demonstrated graphite mediation of nitroglycerin reduction by cast iron (21). For black carbon-mediated sulfide destruction of nitroglycerin, we evaluated the time scale of the reaction, concentration requirements for black carbon and sulfides, the effect of sulfide speciation (i.e., H2S/HS-), the black carbon type VOL. 44, NO. 16, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6409

dependence, and product formation over sulfide and black carbon concentration ranges relevant to marine sediments. Our results suggest that, like RDX, nitroglycerin is rapidly degraded to low toxicity products in the presence of sulfides and black carbon. A second goal was to further evaluate the reaction mechanism. Upon analogy to previous work involving the graphite-mediated reduction of RDX (18) and nitroglycerin (21) by elemental iron, we hypothesized that black carbonmediated destruction of both explosives by sulfides could be attributed to graphene regions within the black carbons (22). Accordingly, sheet graphite was employed as a model black carbon. Based upon the dialysis cell design of Chiu et al. (18), we developed an electrochemical cell which restricted observable solid phase-catalyzed reactions to those involving electron transfer from sulfides to nitroglycerin via the sheet graphite. Nucleophilic substitution reactions between nitroglycerin and aqueous sulfides or surface sulfides formed, for example, by bisulfide incorporation into quinone functional groups on the sheet graphite (20), would be prevented by the physical separation of both the aqueous phases and the sheet graphites to which the sulfides and nitroglycerin were exposed. Therefore, we were able to isolate reactions occurring through the solid black carbon phase, and to tune anodic and cathodic conditions separately by using the electrochemical cell design. Our results demonstrated significant differences in the reaction pathways associated with different explosives, and varying reactivities for different black carbons. These results indicate that mediation of sulfide reduction reactions by graphene regions within black carbons does not capture the system complexity, and demonstrate a need for a more fundamental understanding of the black carbon-mediated reactions.

Materials and Methods Material sources and nanotube purity analysis are provided in the Supporting Information. Hydrogen sulfide stock solutions were made fresh daily by washing the surfaces of sodium sulfide solids with deionized water, and then dissolving them in deoxygenated deionized water (8, 9). Sulfide concentrations, representing the sum of all hydrogen sulfide species ([H2S] + [HS-] + [S2-]), were determined iodometrically (23). Batch Reactor Experiments. Deionized water solutions buffered at pH 7.0 with phosphate buffer were purged with nitrogen gas for 30 min prior to spiking explosives. The deaerated solutions were decanted into vials containing either nitroglycerin or RDX and a weighed amount of black carbon and equilibrated for 12 h. After spiking sulfides, samples were capped without headspace with Teflon-lined septa, placed on a gently rotating bed at room temperature, and covered with aluminum foil to prevent photolytic reactions. Electrochemical Cell Experiments. Two 24-mL vials were connected through Teflon-lined septa by insulated copper wires attached via graphite-based electrical tape (NEM tape, Nisshin EMCO Ltd.) to graphite sheets serving as electrodes (Figure SI-1). The electrical circuit was completed using a salt bridge constructed of Teflon tubing filled with an agarose gel containing 1 M potassium chloride, and passed through the Teflon caps. Each cell was filled with deaerated deionized water buffered at pH 7.0 to a level permitting solution contact with the salt bridge tubing and the graphite sheet, but not the wire. The headspace was purged with nitrogen gas, and the vials were capped. The cathodic cell was spiked with an explosive, and gently mixed for 12 h on a rotating bed to permit sorption equilibrium. The anodic cell was spiked with the sulfide stock solution, the electrochemical cell was returned to the gently rotating bed, and shielded from light by an aluminum foil sheet. With both cells containing 6410

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 16, 2010

deaerated 25 mM phosphate buffer at pH 7 and with 3 mM sulfides in the anodic cell, the potential difference between cathodic and anodic cells was 475 mV for 1 µmole nitroglycerin and 395 mV for 1 µmole RDX, respectively, measured using a multichannel potentiostat VSP (Bio-Logic Instruments). Analyses. Periodically, samples were sacrificed for analysis. The aqueous phase was extracted by shaking for 5 min into 5 mL of dichloromethane and the solid phase was extracted by 5 mL of methyl tert-butyl ether. Extracts were evaporated to dryness under a nitrogen gas stream, resuspended in 250 µL of acetonitrile, and analyzed by HPLC-UV. However, glycerol was analyzed by GC/MS in the chemical ionization mode, following an extraction procedure described below. Details for the HPLC and GC/MS analyses are provided in the Supporting Information. Nitrite was analyzed by either ion chromatography with conductivity detection or colorimetrically (23).

Results Nitroglycerin Destruction and Product Formation. The decay of nitroglycerin and product formation were monitored over time using batch reactors. Nitroglycerin (1.0 µmole) was spiked into 16.4-mL vials containing 12 g/L preweighed sheet graphite, yielding an aqueous concentration of 61 µM. An experiment in the absence of sulfides was conducted to study the sorption kinetics of nitroglycerin with sheet graphite. Aqueous nitroglycerin concentrations declined over 12 h, with no further changes observed through 24 h. Over an introduced nitroglycerin mass range of 0.3-1.5 µmoles, the Freundlich adsorption isotherm was approximately linear (qe ) 0.036Ce1.10 (Figure SI-2A)). For subsequent experiments, 1.0 µmole nitroglycerin was spiked into each vial 12 h prior to sulfide introduction, resulting in 42.3% of the nitroglycerin adsorbed to the sheet graphite. Because both the solid and liquid phases were analyzed for nitroglycerin and the decay products, the results are presented as the total mass retrieved from both phases. No significant nitroglycerin decay was observed in controls lacking both sulfides and graphite or in the presence of 12 g/L sheet graphite alone (Figure 1A). In samples containing only nitroglycerin and 3 mM sulfides, nitroglycerin decayed slowly by ∼64% over 24 h, likely due to a nucleophilic substitution reaction (24). However, nitroglycerin decay was complete within 4 h in the presence of both 12 g/L sheet graphite and 3 mM sulfides, indicating that black carbon promoted nitroglycerin destruction by sulfides. In the presence of sulfides and black carbon, nitrite was the major nitrogenous product, accounting for 90% of the nitrogen mass balance after 24 h (Figure 1B). Similar to previous biodegradation studies with nitroglycerin (25), 1,3dinitroglycerin and 1,2-dinitroglycerin were detected as transient intermediates, and a strong regioselectivity was observed, with 1,3-dinitroglycerin formation exceeding that of 1,2-dinitroglycerin by a factor of 3. Sheet graphite greatly enhanced the transformation of these intermediates (Figure 1C and D); both the formation and decay of 1,2- and 1,3dinitroglycerin occurred within 24 h in the presence of sheet graphite. However, formation, but not decay, of dinitroglycerins was observed by sulfides in the absence of sheet graphite. Due to low extraction efficiencies for 1- and 2-nitroglycerin and glycerol by our standard analytical procedure, we conducted separate experiments to verify their formation. We spiked 0.44 µmole 1,3-dinitroglycerin into vials under the same conditions as previous experiments. After 5 h, the aqueous phase was extracted 3 times with 5 mL of ethyl acetate. The extracts were combined, blown down to dryness under nitrogen gas, and resuspended in 250 µL of acetonitrile. Although sensitivity issues prevented quantification, we

FIGURE 1. Decay of 1.0 µmole nitroglycerin (NG) and product formation over time in 25 mM phosphate buffer at pH 7.0 and 25 °C. (A) NG decay, (B) nitrite formation, (C) 1,2-dinitroglycerin (1,2-DNG) formation, (D) 1,3-dinitroglycerin (1,3-DNG) formation: 9 ) 3 mM sulfides and 12 g/L sheet graphite; 0 ) 3 mM sulfides only; 1 ) 12 g/L sheet graphite only; 2 ) no graphite or sulfides. Error bars represent 1 standard deviation of experimental duplicates.

FIGURE 2. Dependence on reagent concentrations of observed pseudo-first-order decay rates (kobs) at pH 7.0 for 1.0 µmole nitroglycerin in 25 mM phosphate buffer and for 0.108 µmoles RDX in 20 mM phosphate buffer. (A) kobs for nitroglycerin as a function of sulfide concentrations in the presence (9) or absence (0) of 12 g/L graphite. (B) kobs for RDX as a function of sulfide concentrations in the presence (b) or absence (O) of 4 g/L graphite. (C) kobs for nitroglycerin as a function of graphite concentrations in the presence (9) or absence (0) of 3 mM sulfides. (D) kobs for RDX as a function of graphite concentrations in the presence (b) or absence (O) of 3.5 mM sulfides. Error bars represent one standard deviation of experimental duplicates. detected both 1-nitroglycerin and glycerol using HPLC and the pseudo-first-order observed rate constants (kobs) for GC/MS, respectively. To quantify glycerol yields, we added nitroglycerin decay were first order in sulfides over 0.36-6.26 270 µM nitroglycerin to 3 mM sulfides and 12 g/L sheet mM (Figure 2A). As the 95% confidence interval of the linear graphite. After 24 h, samples were extracted three times with regression model between sulfides and kobs suggested (Table SI-1), there was no threshold in sulfide concentrations below 10 mL of ethyl acetate and the extracts were combined and which nitroglycerin was not destroyed. blown down to 100 µL for GC/MS analysis. The molar yield Previously, destruction of the cyclic N-nitramine, RDX, of glycerol was 78% ((2.9%), suggesting a sequential deniin the presence of activated carbon was reported to be trification during graphite-mediated nitroglycerin degradapseudo-first-order in sulfides, but only above a threshold tion by sulfides. concentration (16); based on the linear regression between Reagent Requirements for Nitroglycerin and RDX Desulfides and kobs, the 95% confidence interval for this sulfide cay. Both nitroglycerin and RDX decay followed first order threshold was 1.2 ((0.1) mM. RDX destruction in the presence kinetics (Figure SI-3). In the presence of 12 g/L sheet graphite, VOL. 44, NO. 16, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6411

FIGURE 3. (A) 1.0 µmole nitroglycerin decay by 3 mM sulfides after 4 h in the presence or absence of 3 g/L sheet graphite using batch reactors or electrochemical cells. (B) Nitrite formation from 1.0 µmole nitroglycerin destruction at various pH after 4 h in electrochemical cells with 3 g/L sheet graphite serving as electrodes. Cathodic cell ) nitroglycerin at pH 7.0; anodic cell )3 mM sulfides at various pH. (C) 0.18 µmole RDX decay by 3 mM sulfides after 24 h in the presence or absence of 3 g/L sheet graphite at pH 7.0 using batch reactors or electrochemical cells. Samples were buffered in 25 mM phosphate at pH 7.0. Error bars represent one standard deviation of experimental duplicates. of 4 g/L sheet graphite was evaluated as a function of sulfide concentrations (Figure 2B), but the regression model indicated no threshold (Table SI-1). The dependence of nitroglycerin decay on sheet graphite concentration (Figure 2C) was investigated. Between 3 and 36 g/L sheet graphite, nitroglycerin decay by 3 mM sulfides leveled off. As calculated from the Freundlich adsorption isotherm, the sorbed nitroglycerin percentage likewise leveled off at higher graphite concentrations (Figure SI-2B). For RDX, kobs was linear through 4 g/L sheet graphite (Figure 2D), within the linear sorption range (Figure SI-2C). The results for both nitroglycerin and RDX decay indicate that adsorbed species undergo faster decay kinetics than aqueous species. Electrochemical Cell Experiments. Within batch reactors, sulfide-mediated nitroglycerin decay proceeded by concurrent aqueous phase- and solid phase-catalyzed pathways. To separate these pathways and better characterize the solid phase-catalyzed pathway, nitroglycerin decay was evaluated in three systems: (1) the batch reactor, which combined aqueous and solid phase reactions; (2) batch reactor controls without sheet graphite, which isolated the aqueous phase nitroglycerin reaction with sulfides; and (3) electrochemical cells, which isolated solid phase-catalyzed reactions by physically separating the aqueous phases containing sulfides and nitroglycerin. In batch reactors containing 3 g/L sheet graphite and 3 mM sulfides, 0.86 ((0.05 standard deviation; n ) 2) µmoles of the 1.0 µmole nitroglycerin was destroyed after 4 h. Over 4 h, 0.53 ((0.02) µmoles nitroglycerin was destroyed in the electrochemical cells, providing direct evidence for nitroglycerin reduction via electron transfer from sulfides through the sheet graphite. In batch reactors without sheet graphite, 0.38 ((0.11) µmoles nitroglycerin decayed after 4 h. Demonstrating the utility of this system for isolating the multiple reaction pathways occurring in batch reactors, the sum of the nitroglycerin decay observed for the isolated solid phasecatalyzed system using the electrochemical cells and the aqueous phase reactions using the batch reactor sheet graphite-free controls (0.91 ( 0.11 µmoles) was not significantly different from the decay observed in batch reactors containing sheet graphite and sulfides (Figure 3A). The electrochemical cell design also enabled separate tuning of the conditions in each cell. To evaluate pH dependence, the cell containing nitroglycerin was maintained at pH 7.0 to ensure a common percentage of sorbed nitroglycerin. The pH of the cell containing sulfides was varied from 5.7 to 8.5 to evaluate the importance of sulfide speciation. Nitrite formation in the nitroglycerin-containing cell increased with pH (Figure 3B), suggesting that bisulfide, rather than hydrogen sulfide, was responsible for the sheet graphite-mediated reaction. 6412

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 16, 2010

RDX was evaluated similarly as nitroglycerin (Figure 3C). After 24 h, no RDX decay was observed in batch reactor controls omitting either sheet graphite or sulfides. Meanwhile, 98% of 0.18 µmole RDX was destroyed in batch reactors with 3 mM sulfides and 3 g/L sheet graphite at pH 7.0, forming 1 mol equiv of nitrite. In contrast, neither RDX decay nor nitrite formation was observed in electrochemical cells after 24 h or even 6 d. The results suggested that RDX destruction observed in the batch reactors was not associated with electron transfer from sulfides to RDX via the sheet graphite as previous researchers had suggested (16, 17). Black Carbon Type Dependence. Previous research demonstrated differences in RDX destruction rates by sulfides in the presence of different black carbons, including activated carbon, graphite powder, chars, and soot (16, 17). The decay of nitroglycerin by 3 mM sulfides was investigated in batch reactors with a variety of black carbons, including activated carbon, sheet graphite, multiwall carbon nanotubes (MWCNTs), single wall carbon nanotubes (SWCNTs), and diesel soot. Nitroglycerin decay experiments with sulfides and 3 g/L activated carbon were conducted as for sheet graphite. After 3 h, 78% of nitroglycerin decayed in the presence of sulfides and activated carbon, but only 26% with sulfides alone. For MWCNTs, SWCNTs, and diesel soot, initial sorption experiments with 1.0 µmole nitroglycerin and 8-16 mg/L of these materials without sulfides indicated that aqueous concentrations no longer declined after 72 h, indicating that sorption equilibrium was attained. Attempts to recover nitroglycerin from the solid phase MWCNTs and SWCNTs were unsuccessful. In subsequent experiments, samples were pre-equilibrated with nitroglycerin for 72 h before sulfide introduction. We employed significantly lower concentrations (8-16 mg/L) of MWCNTs, SWCNTs, and diesel soot than for sheet graphite and activated carbon, and quantified only the aqueous phase concentrations (Figure SI-4). MWCNTs and SWCNTs coagulated and settled from solution, facilitating phase separation. Diesel soot remained suspended, and was removed from the aqueous phase by filtration through 0.2-µm filters. Controls under these conditions where sulfides were omitted suggested that >90% of the nitroglycerin was recovered from the aqueous phase after 3 h. No nitrite or organic transformation intermediates were detected in the aqueous phase, indicating that the other ∼10% of the nitroglycerin was sorbed to the black carbons, rather than transformed. The high percentage of nitroglycerin in the aqueous phase at equilibrium resulting from the low black carbon concentrations enabled monitoring of nitroglycerin decay via the aqueous phase. In the presence of 3 mM sulfides and 11 mg/L MWCNTs or 8 mg/L SWCNTs, more than 60% of nitroglycerin was

FIGURE 4. Nitroglycerin decay by 3 mM sulfides in 25 mM phosphate buffer at pH 7.0 and 25 °C after 3 h in the presence of various black carbons (12 g/L graphite sheet, 3 g/L granular activated carbon, 16 mg/L diesel soot particulate matter, and 11 mg/L single wall carbon nanotubes or 8 mg/L multiwall carbon nanotubes). The observed reaction rates were normalized by surface areas of black carbons. Error bars represent one standard deviation of experimental duplicates.

destroyed and the transformation intermediates (1,2- and 1,3-dinitroglycerin) were detected after 3 h. In contrast, in the presence of 3 mM sulfides and 16 mg/L diesel soot, nitroglycerin decay after 3 h was comparable to the 26% decay observed in controls containing 3 mM sulfides without soot. Nitrite and 1,2- and 1,3-dinitroglycerin were observed as products with or without soot. To enhance the importance of any surface-catalyzed reaction compared to the aqueous phase sulfide-nitroglycerin reaction, we also evaluated 200 mg/L diesel soot; again, nitroglycerin decay after 3 h was no different than that in soot-free controls. Attempts to use higher soot concentrations were unsuccessful due to clogging of filters used for phase separation. To compare the ability of black carbons to mediate these reactions, nitroglycerin decay rates were normalized by black carbon surface areas (Figure 4; see Figure SI-5 for decay rates normalized by black carbon mass). The order of reactivities was MWCNTs > SWCNTs > sheet graphite > activated carbon. For diesel soot, any surface-catalyzed nitroglycerin destruction reactions were less important than aqueous phase sulfide-nitroglycerin reactions, preventing the exploration of any solid phase-mediated reactions. Based upon the level of nitroglycerin destruction observed for the soot-free control, at 200 mg/L soot, any surface-catalyzed nitroglycerin destruction reactions must have been hexahydro-1-nitroso-3,5dinitro-1,3,5-triazine (MNX) > RDX (26), the opposite of that observed for RDX destruction by sulfides in the presence of activated carbon (16). One alternative is that black carbon-mediated RDX destruction involves a nucleophilic substitution reaction. As observed previously for natural organic matter functional groups (20), sulfide reactions with surface functional groups on the black carbons may have formed organosulfides. Organosulfides may serve as potent nucleophiles (20) that can react directly with RDX. Note that the 1 mol equiv of nitrite and 2 mol equiv of formaldehyde that formed from RDX in the presence of sulfides and activated carbon (16) are similar to the 1 mol equiv of nitrite and 1 mol equiv of formaldehyde observed during alkaline hydrolysis (27), a form of nucleophilic substitution reaction. In contrast, during reduction reactions mediated by Fe(II), products included nitrosated triazine reduction products (28), and 3 mol equiv of formaldehyde but no nitrite (26); however, in the latter case, it is possible that any nitrite that formed was rapidly reduced to the ammonia that was observed. The lack of direct contact between RDX and sulfide-modified sheet graphite would have prevented such a reaction within the electrochemical cell. However, the nature of the black carbon-mediated RDX reaction remains unproven. Second, black carbons differed significantly in their ability to mediate the destruction of the same compound, nitroglycerin, by sulfides. The surface-area normalized pseudo-first-order nitroglycerin decay rate constants for various black carbon surfaces followed the order MWCNTs > SWCNTs > sheet graphite > GAC (Figure 4). Note that the high reactivities of MWCNTs and SWCNTs were not due to their large specific surface areas since the decay rate constants were already normalized by surface areas. Although the nanotubes were not purified to remove trace metals, metals may not have been responsible for their high reactivities. Previous research with natural organic matter mediation of the destruction of substituted nitrobenzenes by sulfides indicated the importance of quinone functional groups, rather than trace metals (7). The variation in reactivity among black carbons may arise from differences in conductivity, functional groups, and possibly unidentified reactive sites. Studies of glassy carbon electrodes indicated that oxygenated functional groups, including carbonyl, hydroxyl, and carboxyl groups, could VOL. 44, NO. 16, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6413

affect electron transfer kinetics (22). Other groups have suggested that edge-planes on graphite or surface defects on carbon nanotubes are reactive sites (29, 30). The variations in sulfide concentration thresholds needed for RDX decay among different black carbon types may reflect differences in the prevalence of surface functional groups that scavenge sulfides, competing with other groups that mediate sulfide destruction of RDX. The catalytic role of graphitic carbon has drawn attention for fuel cell applications (29, 30). However, its potential role in mediating reactions in sediments has not been a research focus. Our results indicate that the catalytic role of black carbons can not be attributed simply to electron transfer mediation by graphene regions. Different reaction mechanisms may apply to different contaminants, and surface functional groups may play a significant role. To develop a more fundamental understanding of the nature of black carbon-mediated reactions, we are currently evaluating the roles played by specific functional groups. Environmental Relevance. Regardless of the mechanisms, our results suggest that abiotic destruction of nitroglycerin and RDX by sulfides and black carbons could be an important contributor to their natural attenuation in marine sediments. Rather than hindering degradation of explosives by sequestering them from aqueous reactions, sorption to black carbons accelerated their destruction. Black carbons contribute 10-30% of total organic carbon in soils and sediments (12, 13), which corresponds to black carbon concentrations of 12.5-37.5 g/L, comparable to the concentrations we employed. Sulfide concentrations up to 5 mM occur in sediment pore waters (6). Using sheet graphite as a model black carbon, both RDX and nitroglycerin were transformed to relatively harmless products within hours. Compared to the week-long biodegradation of these compounds (25, 31), black carbon-mediated sulfide destruction may contribute significantly to their degradation. Natural attenuation of explosives may be particularly important for marine bombing ranges, where unexploded ordnance excavation is dangerous. Where lower concentrations of native black carbons are found, sediments could be supplemented with black carbons to enhance degradation rates, just as activated carbon has been added to sediments to sequester PCBs (15). However, our results indicated that the black carbon-mediated destruction rates depend significantly on black carbon varieties. To understand the importance of black carbon-mediated destruction, the functional groups in native black carbons must be characterized, and the dependence of destruction rates on the prevalence of functional groups must be understood.

Acknowledgments The authors would like to acknowledge the assistance of Drs. Marcelo Carmo and André Taylor of Yale University Department of Chemical and Environmental Engineering for performing the potentiostat measurements. This research was supported by a grant from the National Science Foundation (CBET-0747735).

Supporting Information Available Additional information including 5 figures and 4 tables. This material is available free of charge via the Internet at http:// pubs.acs.org.

Literature Cited (1) Spain, J. C.; Hughes, J. B.; Knackmuss, H. Biodegradation of Nitroaromatic Compounds and Explosives; Lewis Publishers: New York, 2000. 6414

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 16, 2010

(2) Pennington, J. C.; Jenkins, T. F. Distribution and Fate of Energetics on DoD Test and Training Ranges: Interim Report 1; Technical Report ERDCTR-01-13; U.S. Army Engineer Research and Development Center: Vicksburg, MS, 2001. (3) Wendt, T. M.; Cornell, J. H.; Kaplan, A. M. Microbial degradation of glycerol nitrates. Appl. Environ. Microbiol. 1978, 36, 693–699. (4) Integrated Risk Information System. United States Environmental Protection Agency; retrieved April 1, 2010; http:// www.epa.gov/iriswebp/iris/index.html. (5) Drinking Water Contaminant Candidate List 3. United States Environmental Protection Agency; retrieved April 1, 2010; http:// www.epa.gov/safewater/ccl/ccl3.html. (6) Rickard, D.; Luther, G. W. Chemistry of iron sulfides. Chem. Rev. 2007, 107, 514–562. (7) Dunnivant, F. M.; Schwarzenbach, R. P. Reduction of substituted nitrobenzenes in aqueous solutions containing natural organic matter. Environ. Sci. Technol. 1992, 26, 2133–2141. (8) Roberts, A. L.; Sanborn, P. N.; Gschwend, P. M. Nucleophilic substitution reactions of dihalomethanes with hydrogen sulfide species. Environ. Sci. Technol. 1992, 26, 2263–2274. (9) Lippa, K. A.; Roberts, A. L. Nucleophilic aromatic substitution reactions of chloroazines with bisulfide (HS-) and polysulfides (Sn2-). Environ. Sci. Technol. 2002, 36, 2008–2018. (10) Lyman, W. J.; Reehl, W. F.; Rosenblatt, D. H. Handbook of Chemical Property Estimation Methods; McGraw-Hill: New York, 1985. (11) Monteil-Rivera, F.; Paquet, L.; Deschamps, S.; Balakrishnan, V. K.; Beaulieu, C.; Hawari, J. Physico-chemical measurements of CL-20 for environmental applications s Comparison with RDX and HMX. J. Chromatogr., A 2004, 1025, 125–132. (12) Gustafsson, O.; Gschwend, P. M. The flux of black carbon to surface sediments on the New England continental shelf. Geochim. Cosmochim. Acta 1998, 62, 465–472. (13) Middelburg, J. J.; Nieuwenhuize, J.; Van Breugel, P. Black carbon in marine sediments. Mar. Chem. 1999, 65, 245–252. (14) Gustafsson, O.; Haghseta, F.; Chan, C.; MacFarlane, J.; Gschwend, P. M. Quantification of the dilute sedimentary soot phase: Implications for PAH speciation and bioavailability. Environ. Sci. Technol. 1997, 31, 203–209. (15) Cho, Y. M.; Ghosh, U.; Kennedy, A. J.; Grossman, A.; Ray, G.; Tomaszewski, J. E.; Smithenry, D. W.; Bridges, T. S.; Luthy, R. G. Field application of activated carbon amendment for in-situ stabilization of polychlorinated biphenyls in marine sediment. Environ. Sci. Technol. 2009, 43, 3815–3823. (16) Kemper, J. M.; Ammar, E.; Mitch, W. A. Abiotic degradation of hexahydro-1,3,5-trinitro-1,3,5-triazine in the presence of hydrogen sulfide and black carbon. Environ. Sci. Technol. 2008, 42, 2118–2123. (17) Oh, S. Y.; Chiu, P. C. Graphite- and soot-mediated reduction of 2,4-dinitrotoluene and hexahydro-1,3,5-trinitro-1,3,5-triazine. Environ. Sci. Technol. 2009, 43, 6983–6988. (18) Oh, S. Y.; Cha, D. K.; Chiu, P. C. Graphite-mediated reduction of 2,4-dinitrotoluene with elemental iron. Environ. Sci. Technol. 2002, 36, 2178–2184. (19) Schwarzenbach, R. P.; Stierli, R.; Lanz, K.; Zeyer, J. Quinone and iron porphyrin mediated reduction of nitroaromatic compounds in homogeneous aqueous-solution. Environ. Sci. Technol. 1990, 24, 1566–1574. (20) Perlinger, J. A.; Kalluri, V. M.; Venkatapathy, R.; Angst, W. Addition of hydrogen sulfide to juglone. Environ. Sci. Technol. 2002, 36, 2663–2669. (21) Oh, S. Y.; Cha, D. K.; Kim, B. J.; Chiu, P. C. Reduction of nitroglycerin with elemental iron: Pathway, kinetics, and mechanisms. Environ. Sci. Technol. 2004, 38, 3723–3730. (22) Chen, P.; McCreery, R. L. Control of electron transfer kinetics at glassy carbon electrodes by specific surface modification. Anal. Chem. 1996, 68, 3958–3965. (23) Standard Methods for the Examination of Water and Wastewater, 20th ed.; Eaton, A. D., Clesceri, L. S., Greenberg, A. E., Eds.; American Public Health Association, American Water Works Association, Environmental Federation Publishers: Washington, DC, 1998. (24) Merrow, R. T.; Cristol, S. J.; Van Dolah, R. W. The reaction of n-butyl nitrate with alkaline hydrosulfides. J. Am. Chem. Soc. 1953, 75, 4259–4265. (25) Bhaumik, S.; Christodoulatos, C.; Korfiatis, G. P.; Brodman, B. W. Aerobic and anaerobic biodegradation of nitroglycerin in batch and packed bed bioreactors. Water Sci. Technol. 1997, 36, 139–146. (26) Kim, D.; Strathmann, T. J. Role of organically complexed iron(II) species in the reductive transformation of RDX in anoxic environments. Environ. Sci. Technol. 2007, 41, 1257–1264.

(27) Balakrishnan, V. K.; Halasz, A.; Hawari, J. Alkaline hydrolysis of the cyclic nitramine explosives RDX, HMX, and CL-20: New insights into degradation pathways obtained by the observation of novel intermediates. Environ. Sci. Technol. 2003, 37, 1838–1843. (28) Gregory, K. B.; Larese-Casanova, P.; Parkin, G. F.; Scherer, M. M. Abiotic transformation of hexahydro-1,3,5-trinitro-1,3,5-triazine by FeII bound to magnetite. Environ. Sci. Technol. 2004, 38, 1408–1414. (29) Banks, C. E.; Davies, T. J.; Wildgoose, G. G.; Compton, R. G. Electrocatalysis at graphite and carbon nanotube modified electrodes: edge-plane sites and tube ends are the reactive sites. Chem. Commun. 2005, 7, 829–841.

(30) Zhang, J.; Liu, X.; Blume, R.; Zhang, A.; Schlo¨gl, R.; Su, D. S. Surface-modified carbon nanotubes catalyze oxidative dehydrogenation of n-butane. Science 2008, 322, 73–77. (31) Bhatt, M.; Zhao, J. S.; Halasz, A.; Hawari, J. Biodegradation of hexahydro-1,3,5-trinitro-1,3,5-triazine by novel fungi isolated from unexploded ordnance contaminated marine sediment. J. Ind. Microbiol. Biotechnol. 2006, 33, 850–858.

ES101307N

VOL. 44, NO. 16, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

6415