Transport of Atomic Hydrogen through Graphite and Its Reaction with

Graphite is a major non-iron component in commercial iron granules that are typically used for groundwater remediation. Recent studies suggest graphit...
0 downloads 0 Views 126KB Size
Download PDF
Environ. Sci. Technol. 2006, 40, 3959-3964

Transport of Atomic Hydrogen through Graphite and Its Reaction with Azoaromatic Compounds JIANCHANG YE AND PEI C. CHIU* Department of Civil and Environmental Engineering, University of Delaware, Newark, Delaware 19716

Graphite is a major non-iron component in commercial iron granules that are typically used for groundwater remediation. Recent studies suggest graphite inclusions in commercial iron may serve as both adsorption and reaction sites for nitrogenous pollutants such as nitroaromatics, nitrate esters, and heterocyclic nitramines. In this study, we investigated graphite-mediated reduction of azoaromatic compounds with elemental iron in dialysis cells, where azo compounds and iron were physically separated by graphite foil. Both the nonpolar azobenzene and the water-soluble orange G were reduced to aniline, suggesting that exposed graphite in granular iron may mediate reduction of both polar and nonpolar compounds. Orange G reduction was zero-order and commenced after a long initial lag. Both the lag time and the zero-order rate constant varied with graphite thickness, consistent with the explanation that orange G reduction was limited by atomic hydrogen, which was formed via anaerobic iron corrosion and spilled over to graphite. Involvement of atomic hydrogen was confirmed by detection of deuterated aniline when iron was placed in a D2O-based buffer. Our results indicate that atomic hydrogen is mobile in graphite at room temperature, is reactive toward azoaromatic compounds, and may be consumed during transport in graphite.

Introduction Elemental iron has been used as a reactive material in permeable reactive barriers (PRBs) for subsurface remediation for the past decade. Over a hundred PRBs have been installed to date and most of them are in the United States (1). Many PRBs contain elemental iron to treat groundwater contaminants including chlorinated aliphatics (2), metals and radionuclides (3, 4), and nutrients (3, 5). In addition, elemental iron has been shown to be potentially useful for water and wastewater treatment. Several iron-based technologies were recently proposed for removing arsenic and viruses from water (6, 7) and enhancing degradation of nitroaromatic and nitramine explosives and azo dyes in wastewater (8, 9). For field applications, cast iron granules are typically used, which are relatively low-cost and contain impurities including oxides, silica, carbon, and sulfur (10). Reductive transformation of contaminants by elemental iron has been widely investigated. With a few exceptions (11), redox reactions in iron-water systems are believed to occur at the iron-solution interface (12). Although elemental * Corresponding author phone: 302-831-3104; fax: 302-831-3640; e-mail: [email protected]. 10.1021/es060038x CCC: $33.50 Published on Web 05/19/2006

 2006 American Chemical Society

iron is the ultimate source of electrons, surface iron oxides and oxyhydroxides play important roles in contaminant reduction. Surface oxide films are (semi)conducting and permit transfer of electrons from Fe(0) to oxidants such as chlorinated solvents (13, 14). Fe(II)-bearing minerals; e.g., green rust and magnetite, can reduce chlorinated (15) and nitroaromatic (16) compounds, Cr(VI) (17), and nitrate (18). Furthermore, iron oxides can adsorb Fe(II) and, in the presence of aqueous Fe(II), exhibit high reactivity (19, 20) toward polyhalogenated methanes (21), nitroaromatics (20, 22), nitrate esters (23), nitrite (23), and a heterocyclic nitramine (24, 25). In addition to iron oxides, a non-iron component in cast iron that has been suggested to be involved in contaminant reduction is graphite. Cast iron contains >2% carbon by weight (>6 vol %) and is considered to be an iron-carbon alloy (26). A large portion of this carbon can be graphite (26, 27). Exposed graphite in cast iron was shown by Burris and co-workers to lower apparent reduction rates of chlorinated ethenes via hydrophobic adsorption (28, 29). In addition, Oh et al. (23, 24, 30) showed that graphite in cast iron could mediate reduction of nitrogenous compounds such as nitroaromatics, nitrate esters, nitrite, and heterocyclic nitramines by serving as reaction sites. On the basis of the different intermediate distributions of 2,4-dinitrotoluene (DNT) observed in pure iron and cast iron systems (30), Jafarpour et al. (31) showed through numerical modeling that most DNT was reduced on the graphite surface. Although this is only the first attempt to quantify the significance of graphite in contaminant reduction by cast iron, the study suggests that graphite may be potentially important because of its unique ability to both adsorb hydrophobic compounds and conduct electron. In a study to demonstrate graphite’s ability to mediate organic reduction by elemental iron, Oh et al. (30) reported observations that suggested a reductant other than electron was conducted by graphite and responsible for the observed reaction. These authors proposed that this reductant was atomic hydrogen, produced through reduction of proton by elemental iron. Although iron is not normally considered catalytic, it has been suggested that surface defects or secondary phases in iron may catalyze hydrogenation reactions (32). There is also growing evidence that hydrogen atom may be an important reductant involved in the reduction of chlorinated ethenes (33, 34) and N-nitrosodimethylamine (35) with iron. However, the potential role of graphite as a conduit for atomic hydrogen and the possible involvement of hydrogen atom in graphite-mediated reduction have not been investigated. This study was prompted to address these questions. The primary focus of the study was to understand the nature of the reductant(s) involved in graphite-mediated reduction. Azo-aromatic compounds of different hydrophobic properties were used to differentiate removal by adsorption and reduction. Experiments were conducted with a hydrogen isotope (deuterium) to obtain direct evidence for the involvement of atomic hydrogen.

Materials and Methods Materials. Azobenzene (99%), orange G (93%), dichloromethane (99.9%), aniline (99.9%), and iron powder (99.5%) were purchased from Aldrich (Milwaukee, MI). HEPES (N-[2-hydroxyethyl]piperazine-N′-ethanesulfonic acid) was obtained from Sigma (St. Louis, MO). Acetronitrile (HPLC grade) was acquired from Fisher Scientific (Pittsburgh, PA). VOL. 40, NO. 12, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3959

Perdeuterated (d7) aniline, deuterium oxide (D2O, 99.9% D), and deuterium chloride (DCl, 37 wt % in 99.5% D2O) were obtained from Aldrich. The aniline-d7 had a deuterium content of 99.6% for the ring hydrogen and 99.2% for the amino hydrogen, as determined by Aldrich through D and H NMR analyses. All chemicals were used as received. Graphite foils were purchased from SGL Carbon (>98%, Wiesbaden, Germany) and Alfa Aesar (99.8%, Ward Hill, MA). The thicknesses of the graphite foils were measured to be 0.14, 0.22, and 0.38 mm, corresponding to nominal thicknesses of 0.0057, 0.0087, and 0.0157 in., respectively. Dialysis cells, each having a 10-mL capacity, were purchased from Bell-Art Product (Pequannock, NJ). Dialysis Cell Experiments. The setup and procedures for dialysis cell experiments with graphite foil were similar to those described by Oh et al. (30). Briefly, the two 5-mL compartments of a dialysis cell were separated by graphite foil, and the cell was assembled using two-sided tape and stainless steel screws (Figure S1 of the Supporting Information). An opening at the top of each compartment permitted transfer of solution and sampling. One compartment was filled with approximately 32 g of high-purity elemental iron powder in 0.1 M deoxygenated HEPES buffer solution (pH 7.4). The other compartment contained azobenzene or orange G solution in deoxygenated deionized water. Multiple replicate dialysis cells were prepared and shaken on an orbital shaker at 60 rpm in an anaerobic glovebox (N2/H2 ) 95/5, Coy Laboratory, Grass Lake, MI). Cells were sacrificed at different times for analyses. A 3-mL aliquot was withdrawn from the azo compartment and extracted with 1 mL of CH2Cl2 to measure aqueous masses. Graphite foil was extracted twice, each with 4 mL of acetonitrile, to obtain adsorbed masses. For orange G reduction experiments, aqueous orange G concentration was measured before these extractions. For experiments with deuterated solution, the buffer in the iron compartment was prepared with HEPES base (sodium salt) and DCl in D2O, giving a solution deuterium content of >99.5%. A solution of azobenzene or orange G for the other compartment was still prepared in deoxygenated deionized H2O. Control Experiments. Controls were run in parallel to test possible alternative data explanations: (1) Dialysis cells were set up in the same fashion, except the “azo” compartment contained only deoxygenated water (no azoaromatic compound), which was analyzed over time for Fe(II) and pH. (2) Iron powder was omitted to test the possibility of azo reduction by graphite, H2, or their combination. (3) HEPES buffer solution for the iron compartment was prepared in D2O, and aniline (instead of an azo compound) in H2O was added to the azo compartment to test whether deuterated aniline could form through nonredox processes. In addition, a hydrogen-deuterium exchange experiment was performed using 5-mL amber borosilicate vials to demonstrate, and assess the extent of, proton exchange between water and aniline. In each replicate vial, 4 µL of aniline in acetone (50.9 mM) was added to 2 mL of D2O (pH ≈ 9, deuterium content g 99.85%). After different elapsed times, aniline was extracted with dichloromethane. Analytical Methods. Dichloromethane and acetonitrile extracts were analyzed for azobenzene and aniline using an Agilent (Palo Alto, CA) 6890 GC-5973 MSD equipped with a J&W DB-1701 capillary column (30 m, 0.25 mm i.d., 0.25 µm film thickness). Identification and quantification of these analytes in nondeuterated experiments were based on the retention times (6.34 min for aniline and 10.51 min for azobenzene), quantification ions (m/z ) 93 for aniline and 182 for azobenzene), and confirmation ions (m/z ) 39 and 66 for aniline and 77 and 105 for azobenzene). The detection limit for both compounds was 1 ng/µL. The ionization mode 3960

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 12, 2006

FIGURE 1. Aqueous and surface masses of (a) azobenzene and (b) orange G during their reduction in dialysis cells with iron powder and 0.22-mm graphite foil. Aniline masses in part a are presented as equivalent azobenzene masses (i.e., divided by 2). Arrow indicates commencement of orange G reduction. was electron impact (70 eV) and the ion source temperature was 250 °C. The GC temperature program used was as follows: isothermal at 40 °C for 1 min, 20 °C/min to 230 °C, and isothermal at 230 °C for 1 min. Orange G concentration was measured by a Hach (Loveland, CO) DR/2010 portable spectrophotometer at 475 nm. Soluble iron concentration was measured using inductively coupled plasma (model EOP, Spectro Analytical Instruments, Kleve, Germany) and was taken to be dissolved Fe(II) concentration (detection limit ) 5 ppb).

Results and Discussion Azobenzene was reduced to aniline in dialysis cells with elemental iron and graphite foil. Figure 1a shows the masses of aqueous and adsorbed azobenzene and aniline at different times. Note that aniline mass is expressed as equivalent azobenzene mass (divided by 2). Azobenzene was removed from the aqueous phase due to both reduction and adsorption to graphite. The adsorbed azobenzene was reduced to aniline over several hours. Compared to the hydrophobic azobenzene (Kow ) 3.82), aniline was more soluble and adsorbed less strongly to graphite. About 89% of the azobenzene was reduced to aniline in 16 h and the final mass recovery was 98 ( 5.7%. The high mass balance shows that azobenzene and aniline were confined to the azo compartment and that the graphite foil was impermeable to these compounds. The lower mass balance (as low as 74.8%) in early times was presumably due to accumulation of intermediates not measured in this study, such as hydrazobenzene. In control cells where iron was omitted, no azobenzene reduction (only adsorption) occurred under H2 over the same period, indicating that H2 and graphite could not have reduced azobenzene and that Fe(0) was necessary for the reduction.

Figure 1a shows that azoaromatic compounds can be reduced by elemental iron that is in contact with graphite but not the azo compound. An implication of this result is that graphite inclusions in cast iron may be reaction sites for hydrophobic azo-aromatic compounds. That is, graphite may be a conduit through which reductants derived from elemental iron are transferred to adsorbed oxidants. This is a possible mechanism for all nitrogenous compounds examined to date, including azo- and nitro-aromatics (30), nitrate esters (23), and heterocyclic nitramines (24). We also conducted a parallel experiment with orange G, for two reasons. First, in contrast to azobenzene, which adsorbs rapidly and reversibly to graphite (Figure S2 of the Supporting Information), orange G is negatively charged and highly water-soluble and thus adsorbs only minimally to graphite. This would allow us to test whether hydrophobic adsorption is necessary for graphite-mediated reduction with iron. Second, without adsorption as a potentially confounding process, more detailed reduction kinetic data might be obtained. Orange G was reduced to aniline and 1-amino-2-hydroxynaphthalene-6,8-disulfonate (not measured) over a time scale similar to that for azobenzene, as shown in Figure 1b. The aqueous orange G mass was essentially the total mass, as the fraction adsorbed to graphite never exceeded 3% of initial mass in any experiment. The aniline produced accounted for approximately 85% of the total orange G. This low mass recovery relative to azobenzene may be attributed to the low purity (93%) of orange G. No orange G reduction was observed in controls without elemental iron. Figure 1b indicates that strong adsorption is not necessary for graphitemediated reduction with iron. Thus, it is possible for exposed graphite in cast iron to mediate reduction of both polar and nonpolar compounds, although this mechanism is probably more important for hydrophobic compounds as they would preferentially accumulate on exposed graphite. Interestingly, orange G reduction exhibited two distinct features that were not observed with azobenzene. First, there appeared to be an initial lag period of about 1 h before orange G reduction commenced (Figure 1b). Second, the reduction kinetics was zero-order, independent of orange G concentration. These features were presumably masked in the case of azobenzene due to its rapid and substantial adsorption to graphite (Figure S2). We believed the lag and zero-order kinetics were mechanistically significant and decided to investigate them further. The lag time suggests that orange G reduction would not occur until the reactant involved, most likely a reductant derived from corroding elemental iron, traveled through the graphite foil and reached the orange G compartment. In addition, the considerable lag time (∼1 h) indicates that transport of this reductant through graphite was slow. If the initial lag indeed reflects the time required for the reductant responsible for orange G reduction to pass through graphite, then one would expect the lag time to vary with graphite foil thickness. Figure 2 shows the effect of graphite thickness on orange G reduction in dialysis cells. For all three foil thicknesses tested, both an initial lag and zero-order kinetics were observed. As expected, the lag time increased with graphite foil thickness: approximately 0.5, 1.0, and 2.5 h for 0.14, 0.22, and 0.38 mm, respectively. The result supports the hypothesis that orange G reduction was controlled by a reductant that was formed in the iron compartment and conducted slowly through graphite. The zero-order rate constant for orange G reduction following the initial lag also varied with graphite thickness, decreasing from 43.6 µM/h for 0.14 mm to 7.6 µM/ h for 0.38 mm. This kinetic effect will be discussed later.

FIGURE 2. Orange G reduction in dialysis cells with graphite foil of different thicknesses. Initial orange G concentration was 237.2 µM. Arrows indicate lag times, and fitted lines show zero-order reduction kinetics. Data were corrected against control (without iron) for slight adsorption of orange G to graphite. To further test the hypothesis of transport-controlled reduction, we decided to give the reductant a head start. In a separate set of experiments, the azo compartment was initially filled with deoxygenated water and a concentrated orange G solution was introduced at different times after a HEPES solution was added to the iron compartment to initiate corrosion. The result is shown in Figure S3. When orange G was added at the same time as the buffer (no delay), an orange G reduction curve similar to that in Figure 2 was obtained, with a lag time of 2.4 h and a zero-order rate constant of 5.8 µM/h. When orange G was introduced 2 h after buffer addition (i.e., after iron had been corroding for 2 h), the lag time was shortened to only 40 min. Finally, when orange G was introduced at a 3-h delay, the lag disappeared completely and orange G reduction occurred immediately upon its addition. These results show that formation of the reductant and its transport through graphite started at the onset of iron corrosion (i.e., upon buffer addition), regardless of whether orange G was present in the other compartment or not. During the lag time, orange G was “waiting” for the reductant and its reduction commenced immediately upon arrival of this reductant. The long lag times suggest that the reductant responsible for the orange G reduction was not electron, whose passage through graphite would occur instantly. Electron transfer across graphite foil was shown to take place only in the beginning of the reaction and diminished quickly (30). This is because the dialysis cell resembles an electrochemical cell without a salt bridge and thus electron transfer would result in charge separation across graphite foil, that is, accumulation of cations (e.g., Fe2+, eq 1a) in the iron compartment and anions (e.g., OH-, eq 2a) in the azo compartment. As a result, the pH of azo compound solution would increase and a counter potential would develop across graphite foil, impeding further electron flow. This reductant also could not have been H2 or graphite, which were ruled out by controls as noted earlier. Fe(II) was also ruled out by another set of control cells, which showed that the graphite foil was impervious to ions including Fe(II) and proton over the same period. Therefore, the only probable reductant to explain the observed azo-aromatic reduction appears to be atomic hydrogen, as proposed by Oh et al. (30). Proposed reactions involving atomic hydrogen in the iron and azo compartments are shown in eqs 1b and 2b, respectively. Note that passage of atomic hydrogen through graphite would not be affected by the counter potential across graphite foil. VOL. 40, NO. 12, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3961

Elemental Iron Compartment 2Fe(0) f 2Fe2+ + 4e- (exported to azo compartment) (1a) 2Fe(0) + 4H+ f 2Fe2+ + 4H• (exported to azo compartment) (1b) 2Fe(0) + 4D+ f 2Fe2+ + 4D• (exported to azo compartment) (1c) Azo Compound Compartment Ar-NdN-Ar′ + 4e(imported from iron compartment) + 4H2O f

Ar-NH2 + Ar′-NH2 + 4OH- (2a)

Ar-NdN-Ar′ + 4 H•(graphite) (imported from iron compartment) f Ar-NH2 + Ar′NH2 (2b) Ar-NdN-Ar′ + 4D•(graphite) (imported from iron compartment) f Ar-ND2 + Ar′ND2 (2c) Ar-ND2 + 2H+ ) Ar-NHD + D+ + H+ )

Ar-NH2 + 2D+ (2d)

To obtain direct evidence for the involvement of atomic hydrogen in azoaromatic reduction, we conducted additional experiments in which the HEPES buffer for the iron compartment was prepared using D2O and DCl. Azobenzene and orange G solutions were still prepared in deoxygenated DI H2O. Anoxic iron corrosion in D2O would generate atomic deuterium (eq 1c) and, if D• could indeed pass through graphite foil to react with azo-aromatic compounds, deuterated aniline would be formed, as shown in eq 2c. In contrast, no deuterated aniline would be produced if the reduction occurred through electron transfer and the hydrogens in the amino group of aniline were derived from H2O in the azo compartment. The mass spectrum of regular (nondeuterated) aniline is shown in Figure S4 of the Supporting Information as a reference. The base peak was the molecular ion (m/z ) 93). The small amounts of 94 (7%) and 95 (0.5%) ions were due to the naturally occurring isotopes 13C (1.1%) and, to a lesser extent, 15N (0.37%). The mass spectra of aniline formed from azobenzene and orange G in dialysis cells containing iron in deuterated solution are respectively shown in parts a and b of Figure 3. In contrast to Figure S4, markedly greater amounts of ions 94 and 95 were detected, with abundances (relative to the base peak, m/z 93) of 24% and 20% for azobenzene and 20% and 7% for orange G, respectively. Higher abundances of 94 and 95 ions were consistently observed for aniline samples extracted at different reaction times and from both graphite and aqueous solution. The higher abundances of these ions can be attributed to monodeuterated (Ar-NHD) and dideuterated (Ar-ND2) aniline. The result shows that a portion of aniline formed via reduction of azobenzene and orange G was deuterated, supporting the hypothesis that the azo-aromatic compounds were reduced by atomic hydrogen (deuterium) produced in the iron compartment and transported through graphite. The fact that ions 94 and 95 were less abundant than ion 93 appears to suggest that only small portions of these azo compounds (about 21% of azobenzene and 11% of orange G) were reduced by hydrogen atom. However, the low deuterium contents were due to the fact that the hydrogens of the amino group are labile; i.e., they would undergo proton 3962

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 12, 2006

FIGURE 3. Mass spectra of aniline from reduction of (a) azobenzene and (b) orange G in dialysis cells containing elemental iron in deuterated solution. Iron powder was placed in HEPES buffer prepared with D2O and DCl, and azobenzene or orange G solution was prepared in deoxygenated deionized H2O. The graphite foil thickness was 0.22 mm. exchange with H2O, according to eq 2d. This H-D exchange occurred both in the azo compartment and during GC/MS analysis. The former is illustrated in Figure S5 of the Supporting Information: When regular aniline (Figure S4) was added to D2O and extracted back after 1 h, a large portion of the amino hydrogen had been replaced by deuterium. The H-D exchange during GC/MS analysis was demonstrated through injection of perdeuterated aniline. A molecular ion of 98 (d5) instead of 100 (d7) was observed (data not shown), indicating that, while the ring deuterium was intact, most of the deuterium in the amino group was lost through H-D exchange with moisture in the GC/MS system. On the basis of these results, atomic hydrogen probably made up a much higher percentage of the total reducing equivalent than the deuterium contents in Figure 3 reflect. Although the contribution of atomic hydrogen could not be determined due to the uncertainty associated with the rate and extent of H-D exchange, the lack of orange G reaction during the long initial lags suggests that orange G reduction was limited almost entirely by availability of atomic hydrogen. Thus, atomic hydrogen probably accounted for most of the reducing equivalent involved in the observed azo-aromatic reduction. Interaction between atomic hydrogen and graphite has been studied in recent years, often in the context of hydrogen storage in graphitic carbons. It was shown through molecular simulation that hydrogen atoms chemisorbed to graphite were bound strongly, with a calculated activation energy of 0.67 eV (65 kJ/mol) (36, 37). The relatively high enthalpy was due

FIGURE 4. Zero-order rate constants for orange G reduction with different graphite thicknesses. to formation of a partial covalent bond between hydrogen and an sp2-hybridized carbon in graphite, which needs to pucker out of the graphene plane and adopt a partial sp3 configuration to form a partial C-H bond. In contrast, other modeling studies (38, 39) showed that atomic hydrogen physically adsorbed to graphite was mobile along a graphene plane due to the low binding energy of 8 meV (0.77 kJ/mol). This low energy likely involves van der Waals interactions and suggests that hydrogen atom can diffuse along graphene planes at room temperature (40). Indeed, modeling results (41) showed that diffusion of a hydrogen atom intercalated in graphite could be initiated at temperature as low as 100 K, but diffusion through the graphite basal plane was energetically difficult (42). Our results provide experimental evidence that atomic hydrogen is mobile in graphite at ambient temperature, lending direct support to the above modeling studies. Transfer of atomic hydrogen from one surface to another has been observed and is often referred to as “spillover” (43). It was reported that hydrogen atoms could be transferred from a metal to a carbon surface and subsequently hydrogenate adsorbed organic compounds (44, 45). The spillover distance from Pt to silica, for example, could be as far as a centimeter (46). These observations are consistent with our hypothesis that the azo compounds were reduced by atomic hydrogen spilled over from corroding iron to the adjacent graphite. The transport of spilled-over hydrogen atom was slow (on the order of hours per millimeter, as the lag times suggest) and was presumably driven by its concentration gradient across graphite. As shown in Figures 1b, 2, and S3, reduction of orange G after an initial lag was zero-order. This means that orange G reduction in dialysis cells was limited by another reactant involved in the rate-limiting step. As discussed earlier, reduction of orange G was most likely limited by atomic hydrogen. In many zero-order reactions, such as surfaceand enzyme-catalyzed reactions where all reactive sites are utilized, zero-order rate constants are proportional to the concentration of the limiting agent (i.e., surface site or enzyme). Similarly, it is plausible that the observed zeroorder kinetics reflects that orange G reduction was limited by, and proportional to, the concentration of hydrogen atom at the graphite surface in contact with orange G solution. On the basis of this postulate, orange G reaction rate would be first-order with respect to surface hydrogen atom concentration, as shown in eq 3.

d[OG] ) k0[OG]0 ) k0 ) k1[H•](S) dt

-

(3)

where [OG] is orange G concentration (µM) at time t, k0 is zero-order rate constant (µM/h), [H•](S) is atomic hydrogen concentration (µmol/cm2) at the graphite-orange G solution interface, and k1 is the reaction rate constant (cm2/L/h).

The fitted k0 values from Figure 2 and from additional orange G reduction experiments are shown in Figure 4. While only three graphite thicknesses were available for this study, it is clear that k0 decreased with increasing thickness. Assuming k0 was proportional to [H•](S) (eq 3), the data suggest that the concentration of atomic hydrogen that had passed through graphite and become available to react with orange G decreased with increasing graphite thickness. That is, the longer the distance atomic hydrogen had to travel, the lower its concentration became. This suggests that atomic hydrogen was “lost” during transport in graphite. In fact, Figure 4 may be viewed as a profile of (relative) atomic hydrogen concentration in graphite versus distance from the iron compartment. Possible mechanisms for the loss may include coupling of H• to form H2 (42, 47) and reaction with edge/ corner carbon atoms in graphite. Results of this study can help us understand the mechanism for contaminant transformation in iron treatment systems. First, graphite inclusions in cast iron may conduct electron and atomic hydrogen and serve as redox reaction sites for nonpolar and polar/ionic nitrogenous pollutants. Second, of all the nitrogenous compounds examined in dialysis cells to date, the only one that exhibited limited reduction (and only in early times) was nitrite (23). Because orange G is also negatively charged and was reduced largely by atomic hydrogen, the comparison suggests that nitrite was reduced by electron rather than atomic hydrogen. Third, graphite inclusions in cast iron may contain high concentrations of atomic hydrogen. Because graphite flakes in cast iron are much thinner than the foil used in this study, the steady-state concentration of hydrogen atom in graphite inclusions is probably much higher, as suggested by Figure 4. Results of this study, however, shall not be extrapolated directly to iron treatment systems. As noted earlier, a dialysis cell is an electrochemical half-cell that differs substantially from an iron PRB. The role of electron was suppressed because its flow across graphite foil was impeded by a counter potential. On the other hand, by suppressing electron flow, a dialysis cell is a useful system to study transport of atomic hydrogen and its reactivity toward different contaminants.

Acknowledgments This material is based upon work supported by the National Science Foundation under Grant No. 9984669.

Supporting Information Available (a) Schematic of a dialysis cell and reduction of orange G, (b) adsorption of azobenzene to graphite in controls without iron, (c) result of orange G reduction with delayed introduction of orange G, (d) mass spectrum of nondeuterated aniline, and (e) mass spectrum of D2O-treated aniline. This material is available free of charge via the Internet at http:// pubs.acs.org.

Literature Cited (1) EnviroMetal Technologies, Inc. http://www.eti.ca (accessed January, 2006). (2) Orth, W. S.; Gillham, R. W. Dechlorination of trichloroethene in aqueous solution using Fe0. Environ. Sci. Technol. 1996, 30, 66-71. (3) Blowes, D. W.; Ptacek, C. J.; Benner, S. G.; McRae, C. W. T.; Bennett, T. A.; Puls, R. W. Treatment of inorganic contaminants using permeable reactive barriers. J. Contam. Hydrol. 2000, 45, 123-137. (4) Gu, B.; Liang, L.; Dickey, M. J.; Yin, X.; Dai, S. Reductive precipitation of uranium(VI) by zerovalent iron. Environ. Sci. Technol. 1998, 32, 3366-3373. (5) Su, C. M.; Puls, R. W. Nitrate reduction by zerovalent iron: Effects of formate, oxalate, citrate, chloride, sulfate, borate, and phosphate. Environ. Sci. Technol. 2004, 38, 2715-2720. VOL. 40, NO. 12, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3963

(6) Melitas, N.; Conklin, M.; Farrell, J. Electrochemical study of arsenate and water reduction on iron media used for arsenic removal from potable water. Environ. Sci. Technol. 2002, 36, 3188-3193. (7) You, Y.; Han, J.; Chiu, P. C.; Jin, Y. Removal and inactivation of waterborne viruses using zerovalent iron. Environ. Sci. Technol. 2005, 39, 9263-9269. (8) Oh, S.-Y.; Chiu, P. C.; Kim, B.-Y.; Cha, D. K. Enhancing Fenton oxidation of TNT and RDX through pretreatment with zerovalent iron. Water Res. 2003, 37, 4275-4283. (9) Perey, J. R.; Chiu, P. C.; Huang, C. P.; Cha, D. K. Zero-valent iron pretreatment for enhancing biodegradability of azo dyes. Water Environ. Res. 2002, 74, 221-225. (10) Tamara, M. L.; Butler, E. C. Effects of iron purity and groundwater characteristics on rates and products in the degradation of carbon tetrachloride by iron metal. Environ. Sci. Technol. 2004, 38, 1866-1876. (11) Buerge, I. J.; Hug, S. J. Kinetics and pH dependence of chromium(VI) reduction by iron(II). Environ. Sci. Technol. 1997, 31, 14261432. (12) Weber, E. J. Iron-mediated reductive transformation: Investigation of reaction mechanism. Environ. Sci. Technol. 1996, 30, 716-719. (13) Odziemkowski, M. S.; Schuhmacher, T. T.; Gillham, R. W.; Reardon, E. J. Mechanism of oxide film formation on iron in simulating groundwater solutions: Raman spectroscopic studies. Corro. Sci. 1998, 40, 371-389. (14) Balko, B. A.; Tratnyek, P. G. Photoeffects on the reduction of carbon tetrachloride by zerovalent iron. J. Phys. Chem. B 1998, 102, 1459-1465. (15) Erbs, M.; Bruun Hansen, H. C.; Olsen, C. E. Reductive dechlorination of carbon tetrachloride using iron(II) iron(III) hydroxide sulfate (green rust). Environ. Sci. Technol. 1999, 33, 307-311. (16) Heijman, C. G.; Holliger, C.; Glaus, M. A.; Schwarzenbach, R. P.; Zeyer, J. Abiotic reduction of 4-chloronitrobenzene to 4-chloroaniline in a dissimilatory iron-reducing enrichment culture. Appl. Environ. Microbiol. 1993, 59, 4350-4353. (17) Legrand, L.; El Figuigui, A.; Mercier, F.; Chausse, A. Reduction of aqueous chromate by Fe(II)/Fe(III) carbonate green rust: Kinetic and mechanistic studies. Environ. Sci. Technol. 2004, 38, 4587-4595. (18) Hansen, H. B.; Koch, C. B.; Nancke-Krogh, H.; Borggaard, O. K.; Sorensen, J. Abiotic nitrate reduction to ammonium: Key role of green rust. Environ. Sci. Technol. 1996, 30, 2053-2056. (19) Williams, A. G. B.; Scherer, M. M. Spectroscopic evidence for Fe(II)-Fe(III) electron transfer at the iron oxide-water interface. Environ. Sci. Technol. 2004, 38, 4782-4790. (20) Elsner, M.; Schwarzenbach, R. P.; Haderlein, S. B. Reactivity of Fe(II)-Bearing Minerals toward reductive transformation of organic contaminants. Environ. Sci. Technol. 2004, 38, 799807. (21) Pecher, K.; Haderlein, S. B.; Schwarzenbach, R. P. Reduction of polyhalogenated methanes by surface-bound Fe(II) in aqueous suspensions of iron oxides. Environ. Sci. Technol. 2002, 36, 17341741 (22) Hofstetter, T. B.; Heijman, C. G.; Haderlein, S. B.; Holliger, C.; Schwarzenbach, R. P. Complete reduction of TNT and other (poly)nitroaromatic compounds under iron-reducing subsurface conditions. Environ. Sci. Technol. 1999, 33, 1479-1487. (23) 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. (24) Oh, S.-Y.; Cha, D. K.; Kim, B. J.; Chiu, P. C. Reductive transformation of hexahydro-1,3,5-trinitro-1,3,5-triazine, octahydro-1,3,5,7-tetranitro-1,3,5,7-tetrazocine, and methylenedinitramine with elemental iron. Environ. Toxicol. Chem. 2005, 24, 2812-2819. (25) Gregory, K. B.; Larese-Casanova, P.; Parkin, G. F.; Scherer, M. M. Abiotic transformation of hexahydro-1,3,5-trinitro-1,3,5triazine by FeII bound to magnetite. Environ. Sci. Technol. 2004, 38, 1408-1414. (26) Walton, C. F.; Opar, T. J. In Iron Castings Handbook Covering Data on Gray, Malleable, Ductile, White, Alloy and Compacted Graphite Irons; Walton, C. F., Opar, T. J., Eds.; Iron Castings Society, Inc.: Rocky River, OH, 1981.

3964

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 12, 2006

(27) Deng, B. L.; Campbell, T. J.; Burris, D. R. Hydrocarbon formation in metallic iron/water system. Environ. Sci. Technol. 1997, 31, 1185-1190. (28) Burris, D. R.; Allen-King, R. M.; Manoranjan, V. S.; Campbell, T. J.; Loraine, G. A.; Deng, B. Chlorinated ethene reduction by cast iron: Sorption and mass transfer. J. Environ. Eng. 1998, 124, 1012-1019. (29) Campbell, T. J.; Burris, D. R.; Roberts, A. L.; Wells, J. R. Trichloroethylene and tetrachloroethylene reduction in a metallic iron-water-vapor batch system. Environ. Toxicol. Chem. 1997, 16, 625-630. (30) 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. (31) Jafarpour, B. Y.; Imhoff, P. T.; Chiu, P. C. Quantification and modeling of 2,4-dinitrotoluene reduction with high-purity and cast iron. J. Contam. Hydrol. 2005, 76, 87-107. (32) Gotpagar, J.; Lyuksyutov, S.; Cohn, R.; Grulke, E.; Bhattacharyya, D. Reductive dehalogenation of trichloroethylene with zerovalent iron: Surface profiling microscopy and rate enhancement studies. Langmuir 1999, 15, 8412-8420. (33) Li, T.; Farrell, J. Electrochemical investigation of the rate-limiting mechanisms for trichloroethylene and carbon tetrachloride reduction at iron surfaces. Environ. Sci. Technol. 2000, 35, 35603565. (34) Wang, J. K.; Farrell, J. Investigating the role of atomic hydrogen on chloroethene reactions with iron using Tafel analysis and electrochemical impedance spectroscopy. Environ. Sci. Technol. 2003, 37, 3891-3896. (35) Odziemkowski, M. S.; Gui, L.; Gillham, R. W. Reduction of N-nitrosodimethylamine with granular iron and nickel-enhanced iron. 2. Mechanistic studies. Environ. Sci. Technol. 2000, 34, 3495-3500. (36) Zecho, T.; Gu ¨ ttler, A.; Sha, X.; Jackson, B.; Ku ¨ ppers, J. Adsorption of hydrogen and deuterium atoms on the (0001) graphite surface. J. Chem. Phys. 2002, 117, 8486-8492. (37) Miura, Y.; Kasai, H.; Din ˜ o, W. A.; Nakanishi, H.; Sugimoto, T. Effective pathway for hydrogen atom adsorption on graphene. J. Phys. Soc. Jpn. 2003, 72, 995-997. (38) Jeloaica, L.; Sidis, V. DFT investigation of the adsorption of atomic hydrogen on a cluster-model graphite surface. Chem. Phys. Lett. 1999, 300, 157-162. (39) Sha, X. W.; Jackson, B. First-principles study of the structural and energetic properties of H atoms on a graphite (0001) surface. Surf. Sci. 2002, 496, 318-330. (40) Majer, G.; Eberle, U.; Kimmerle, F.; Stanik, E.; Orimo, S. Hydrogen diffusion in metallic and nanostructured materials. Phys. B (Amsterdam, Neth.) 2003, 328, 81-89. (41) Shimizu, A.; Tachikawa, H. Thermal behavior of hydrogen atom intercalated between two layers of C150H30 graphite plane: MD simulation. J. Phys. Chem. Solids 2003, 64, 419-423. (42) Ferro, Y.; Marinelli, F.; Allouche, A. Density functional theory investigation of H adsorption and H2 recombination on the basal plane and in the bulk of graphite: Connection between slab and cluster model. J. Chem. Phys. 2002, 116, 8124-8131. (43) Conner, W. C.; Falconer, J. L. Spillover in heterogeneous catalysis. Chem. Rev. 1995, 95, 759-788. (44) Srinivas, S. T.; Rao, P. K. Direct observation of hydrogen spillover on carbon-supported platinum and its influence on the hydrogenation of benzene. J. Catal. 1994, 148, 470-477. (45) Robell, A. J.; Ballou, E. V.; Boudart, M. Surface diffusion of hydrogen on carbon. J. Phys. Chem. 1964, 68, 2748-2753. (46) Candau, J. F. Cevallos; Conner, W. C. Spillover of hydrogen onto silica. IV. The use of scanning FTIR to follow spillover from a point source. J. Catal. 1987, 106, 378-85. (47) Denisov, E. A.; Kompaniets, T. N.; Kostyushkin, K. L.; Makarenko, I. V.; Titkov, A. N. Retention of hydrogen in well-ordered pyrolytic graphite as a result of its interaction with atomic hydrogen. Phys. Scr. T 2003, T103, 43-46.

Received for review January 8, 2006. Revised manuscript received March 23, 2006. Accepted April 19, 2006. ES060038X