Electrochemical Oxidation of N-Nitrosodimethylamine with Boron

Sep 23, 2009 - N-nitrosodimethylamine (NDMA) has been found in the environment in air, water, and soil (1); in food products (2); and more recently in...
0 downloads 0 Views 393KB Size
Environ. Sci. Technol. 2009, 43, 8302–8307

Electrochemical Oxidation of N-Nitrosodimethylamine with Boron-doped Diamond Film Electrodes BRIAN P. CHAPLIN,* GLENN SCHRADER, AND JAMES FARRELL Department of Chemical and Environmental Engineering, University of Arizona, Tucson, Arizona 85721

Received May 29, 2009. Revised manuscript received September 4, 2009. Accepted September 9, 2009.

This research investigated NDMA oxidation by boron-doped diamond (BDD) film electrodes. Oxidation rates were measured as a function of electrode potential, current density, and temperature using rotating disk and flow-through reactors. Final NDMA reaction products were carbon dioxide, ammonium, and nitrate, with dimethylamine and methylamine as intermediate products. Reaction rates were first-order with respect to NDMA concentration and surface area normalized oxidation rates as high as 850 ( 50 L/m2-hr were observed at a current density of 10 mA/cm2. The flow-through reactor yielded mass transfer limited reaction rates that were first-order in NDMA concentration, with a half-life of 2.1 ( 0.1 min. Experimental evidence indicates that NDMA oxidation proceeds via a direct electron transfer at potentials >1.8 V/SHE with a measured apparent activation energy of 3.1 ( 0.5 kJ/mol at a potential of 2.5 V/SHE. Density functional theory calculations indicate that a direct two-electron transfer can produce a stable NDMA(+2) species that is stabilized by forming an adduct with water. The transfer of two electrons from NDMA to the electrode allows an activation-less attack of hydroxyl radicals on the NDMA(+2) water adduct. At higher overpotentials the oxidation of NDMA occurs by a combination of direct electron transfer and hydroxyl radicals produced via water electrolysis.

Introduction N-nitrosodimethylamine (NDMA) has been found in the environment in air, water, and soil (1); in food products (2); and more recently in drinking water (3). It is classified by the U.S. Environmental Protection Agency as a probable human carcinogen (4) and the California Department of Health Services has set an action level of 10 ng/L for NDMA in drinking water (5). The main sources for NDMA in the environment are from the release and subsequent oxidation of the rocket fuel, unsymmetrical dimethylhydrazine (6), and its formation as a byproduct during disinfection of drinking and wastewater (7). Traditional treatment methods, such as carbon adsorption and ion exchange, are not cost-effective for removing NDMA from contaminated waters (3). Recent research has investigated both oxidative and reductive methods for destructive removal of NDMA from water. These methods include reduction by zerovalent iron (8, 9) and Ni- and Pd-based * Corresponding author phone: 217-369-5529; fax: 520-621-6048; e-mail: [email protected]. 8302 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 21, 2009

catalysts (8, 10-12), advanced oxidation processes (AOPs) that generate hydroxyl radicals (e.g., ozone/H2O2 (13), UV/ H2O2 (14), and photocatalysis (15)), and direct UV photolysis (16). There are limitations to each of the above methods. Catalysts are prone to fouling by species found in natural waters (17-19), and zerovalent iron deactivates due to oxide formation (20). The use of AOPs to generate hydroxyl radicals is effective at degrading NDMA, but their capacity for oxidation is limited in natural waters due to hydroxyl radical scavenging by dissolved organic carbon and bicarbonate (13). The use of direct UV photolysis is expensive since water readily absorbs UV light, requiring a UV dose approximately 10 times higher than that required for virus inactivation (3). Electrochemical oxidation using boron-doped diamond (BDD) film electrodes has the potential for overcoming the limitations of other treatment methods. Many studies have shown that BDD electrodes can oxidize compounds by a combination of direct electron transfer (21, 22) and by hydroxyl radicals produced from water oxidation (23, 24); the former mitigates problems associated with hydroxyl radical scavenging. BDD electrodes have a high overpotential for water oxidation, and the potential at the electrode surface can be more oxidizing than that of hydroxyl radicals generated by other AOPs. This feature allows BDD electrodes to oxidize very recalcitrant compounds like perfluorooctane sulfonate by direct electron transfer reactions (21). BDD electrodes are also resistant to fouling by constituents in water and have a long service-life compared to other electrode materials (25, 26). Several recent studies have documented the ability of BDD electrodes to oxidize various waste streams (27-29). Our study investigated the ability of BDD-film electrodes on p-silicon supports to oxidize NDMA. Batch experiments were used to quantify reaction rates as a function of current density, electrode potential, and temperature. Experimental results were compared to density functional theory calculations to elucidate the reaction mechanisms of NDMA oxidation.

Experimental Section Rotating Disk Electrode Experiments. Reaction rates for NDMA were measured under both constant current and constant potential conditions. Currents and electrode potentials were controlled using a Princeton Applied Research (PAR) model 273A potentiostat. Experiments were performed over a temperature range of 10-40 °C using a circulating water bath. A 1.1 cm diameter BDD film on a p-silicon substrate was used as the working electrode (Adamant Technologies, Neutchatel, Switzerland). The working electrode was mounted in a PAR model 316 rotating disk electrode (RDE) assembly and rotated at 3000 rpm to eliminate mass transfer limitations. The counter electrode was either a 20 cm long by 3.2 cm diameter titanium rod (constant current experiments) or a 12 cm long by 0.3 mm diameter platinum wire (constant potential experiments). The anode and cathode chambers were separated by a Nafion membrane (Fuel Cell Scientific, Stoneham, MA) in order to prevent reduction of NDMA oxidation products. A PAR Hg/Hg2SO4 reference electrode saturated with K2SO4 was used. Potentials were adjusted for uncompensated solution resistance and are reported versus the standard hydrogen electrode (SHE). Constant current and constant potential experiments were conducted in 10 mM NaClO4 and 1 M HClO4 background electrolytes, respectively. Before each experiment the BDD electrode was preconditioned in a blank electrolyte solution at a current density of 20 mA/cm2 for 10 min. This 10.1021/es901582q

Not subject to U.S. Copyright. Publ. 2009 Am. Chem. Soc.

Published on Web 09/23/2009

preconditioning was performed to prevent the accumulation of adsorbed organic compounds on the electrode surface. Initial Reaction Rate. Two methods were used to calculate initial reaction rates as a function of electrode potential. In one method (current analysis) the BDD electrode was polarized at a fixed potential in a blank electrolyte solution and the current increase (∆I) following addition of 1.35 mM NDMA was recorded. The current increase was converted to an initial reaction rate (r (mol hr-1)) by using Faraday’s law, r)

∆I nF

(1)

where n are the number of electrons transferred and F is Faraday’s constant. Analytically determined initial reaction rates were also calculated by measuring the disappearance of NDMA from solution up to approximately 10-20% NDMA conversion. Linear regression of the concentration vs time profiles was used to obtain r. Current Efficiency. The current efficiency of NDMA oxidation was calculated by dividing the measured rate of NDMA disappearance by the rate calculated from eq 1, replacing ∆I with the total current observed in each experiment. A background current was observed in each experiment before NDMA addition, and is attributed to water oxidation (see Table 1). Cyclic Voltammetry. Cyclic voltammetry (CV) experiments were conducted using the same experimental setup as described in the constant potential RDE experiments, except the electrode was stationary. The potential was swept from the open circuit potential to 3.64 V and back to the open circuit potential. The scan rate was varied from 0.010 to 16 V/s. Flow-through Reactor. The flow-through reactor contained one bipolar and two monopolar BDD films on p-silicon substrates that were 5 cm long and 2.5 cm wide (Mini-DiaCell, Adamant Technologies). The bipolar electrode was situated between the two monopolar electrodes with an interelectrode gap of 3 mm on both sides. The monopolar electrodes were connected to a Protek (Stayton, OR) model 3005B galvanostatic power supply operated without a reference electrode. The total anode and cathode surface areas were 25 cm2 each, and the total reactor solution volume was 15 mL, yielding a specific surface area of 1.67 cm-1. The flow-through reactor was operated as a closed-loop at a temperature of 20 °C. A 0.5 L solution of 1.35 mM NDMA and 10 mM NaClO4 background electrolyte was circulated at a rate of 100 mL/ min using a peristaltic pump. The reservoir was connected to a pressure gauge containing a 0.3 mm diameter platinum wire mesh (Aesar, Ward Hill, MA) to catalyze the reaction between H2 and O2 produced from water hydrolysis. NDMA removal and product distributions were monitored with time. To account for the time that the fluid spent in the reactor, the elapsed electrolysis times (te) were calculated from

( )

te ) t

VR VL

(2)

where t is the elapsed time and VR and VL are the reactor and solution volumes, respectively. Analytical Methods. Concentrations of NDMA were measured using high pressure liquid chromatography with ultraviolet absorption detection at a wavelength of 226 nm. Nitrate and formate concentrations were determined by ion chromatography (Dionex ICS-3000). Ammonium, dimethylamine (DMA), and methylamine (MA) concentrations were measured by ion chromatography (Dionex DX 500). Total organic carbon (TOC) measurements were made on a Shimadzu model VCSH total organic carbon analyzer. Quantum Mechanics Simulations. Density functional theory (DFT) simulations were performed to investigate possible NDMA reaction mechanisms. All DFT calculations were performed using the DMol3 (30, 31) package in the Accelrys Materials Studio (32) modeling suite. Unrestricted spin, all-electron calculations were performed using doublenumeric with polarization (DNP) basis sets (33) and the gradient corrected Perdew-Burke-Ernzerhof (PBE) functional for exchange and correlation (34). Implicit solvation was incorporated using the COSMO-ibs (35) polarized continuum model. Further details of DFT methods are provided in the Supporting Information (SI).

Results and Discussion Constant Current Experiments. Log-concentration versus time profiles for NDMA removal at current densities ranging from -10 to 20 mA/cm2 at 20 °C are shown in Figure 1. All rates were first-order with respect to NDMA concentration. Reaction rates increased between 1 and 10 mA/cm2, and then decreased between 10 and 20 mA/cm2. The decrease in rate between 10 and 20 mA/cm2 was attributed to formation of oxygen bubbles on the electrode surface that physically blocked NDMA access to the electrode. The first observable products of NDMA oxidation were NO3- and DMA, which are formed by cleavage of the NsN bond (15). DMA was oxidized to form MA (H3C-NH3+), which reacted on the anode to form CO2 and NH4+ via cleavage of the CsN bond (13). Trace levels of formate were also observed and were always less than 2% of the NDMA removed, indicating it may be an intermediate to CO2 formation. Significant concentrations of DMA, MA, and NH4+ were observed in the cathode chamber as a result of diffusion through the Nafion membrane. This finding was confirmed by control experiments with each compound in the absence of an applied current. Similar control experiments also showed that NDMA did not diffuse through the Nafion membrane. A reduction experiment conducted with NDMA showed final products of DMA and NH4+, indicating that the cationic amines passing through the Nafion membrane were not reactive at the cathode. Measurements of TOC in solution

TABLE 1. Summary of Initial Rates for Constant Potential Experiments potential (V)

initial rate (µmol/h)a

background current (mA)

current increase (mA)b

rate from current increase (µmol/h)

% direct oxidationc

current efficiency (%)d

2.00 2.14 2.39 2.64 3.14 4.14

1.01 3.50 10.8 32.4 41.7 37.4

0.06 0.14 0.23 0.33 1.51 4.94

0.05 0.17 0.50 1.26 1.22 0.85

0.98 3.20 9.39 23.5 22.8 15.9

97 91 87 73 55 42

50 64 77 99 77 35

a Rate calculated by NDMA disappearance from solution. b Current increase calculated after the addition of 1.35 mM NDMA. c Calculated by the ratio of the rate determined from the observed current increase and the measured initial rate. d Current efficiency calculated based on comparison of the initial rate to the total rate of electrons transferred.

VOL. 43, NO. 21, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

8303

FIGURE 1. Removal of NDMA as a function of applied current density. Solid lines represent regressions of duplicate experiments. First order rate constants of 0.012 ( 0.002, 0.25 ( 0.01, 0.34 ( 0.02, 0.18 ( 0.02, and 0.12 ( 0.01 h-1 were found for applied current densities of 1, 5, 10, 20, and -10 mA/cm2, respectively. Experiments were conducted in 250 mL of 10 mM NaClO4 background electrolyte, at 20 °C using a 1 cm2 BDD electrode. Reported errors represent 95% confidence intervals.

FIGURE 2. NDMA and products as a function of time in the flow-through reactor in a 10 mM NaClO4 background electrolyte at a current density of 10 mA/cm2. Error bars represent 95% confidence intervals on triplicate experiments. In some cases the error bar is contained within the data point. after ∼24 h of reaction agreed well with the measured carbon species in the anode chamber. However, a total mass balance of carbon and nitrogen could not be achieved, because the products of NDMA oxidation (DMA, MA, and NH4+) that diffused through the Nafion membrane were volatile at the elevated pH values (pH ∼10–11) in the cathode chamber. A closed system flow-through experiment was used to close the mass balance, and is discussed later. The surface area normalized rate constants for NDMA disappearance in the batch experiments were as high as 850 ( 50 L/m2 hr for the 10 mA/cm2 oxidation experiment and 300 ( 30 L/m2 hr for the -10 mA/cm2 reduction experiment. These rates are 2 orders of magnitude greater than those measured for catalytic reduction of NDMA by H2 using Ni and Pd-Cu catalysts (0.678 to 5.53 L/m2-hr) (10-12). Final products of catalytic NDMA reduction were DMA and either NH4+ or N2, depending on the catalyst structure (10-12). The photocatalytic oxidation of NDMA was investigated with modified TiO2, and the highest rate was approximately 0.06 L/m2 hr for Nafion-coated TiO2, with final products of NO2-, NO3-, DMA, and MA (15). Flow-Through Experiments. Concentration versus time profiles for NDMA and its products are shown in Figure 2 for a current density of 10 mA/cm2. NDMA removal was first-order with respect to NDMA concentration. Comparison of rates at current densities of 2, 5, and 10 mA/cm2 indicate that NDMA removal in the flow-through reactor was mass-transfer limited at a current density of 10 mA/cm2 (SI Figure S-1). The first-order reaction rate constant of 0.33 ( 0.01 min-1 corresponds to an NDMA destruction half-life of 2.1 ( 0.1 min. The products observed in the flow-through reactor were similar to the RDE experiments, except formate was not 8304

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 21, 2009

FIGURE 3. Cyclic voltammetry scan in the presence and absence of 1.35 mM NDMA. Scan rate ) 100 mV/s, 1 M HClO4 background electrolyte, 1 cm2 BDD electrode. detected in the flow-through experiments. The total N balance was between 90 ( 9% and 101 ( 7% with a final mass balance of 94 ( 4%. This result indicates minimal volatilization loss of nitrogen species from solution. Nitrate and DMA were the dominant products, indicating that oxidation of NDMA at the anode dominated over NDMA reduction at the cathode. For example, NO3- concentrations after 15 min of electrolysis were 92.6 ( 10.6% of the NDMA degraded. Since batch experiments determined that the rate of NDMA oxidation was roughly double that of reduction, preferential transport of NDMA to the anode over the cathode may have occurred in the flow-through reactor. This observation can be explained by the fact that the oxygen in the nitroso group of NDMA is easily polarized, thus giving it a net negative charge (15). The BDD anode possesses a positive charge that would contribute to electrophoretic transport of NDMA to its surface. The total carbon balance tracked the measured TOC concentration. The final TOC concentration and total carbon balance were 0.90 ( 0.16 and 1.19 ( 0.20 mM C, respectively, indicating that all major aqueous carbon species were detected and the methyl groups of NDMA were oxidized to CO2. Rate-Determining Step of NDMA Oxidation. Several lines of evidence indicate that at low overpotentials NDMA oxidation occurs primarily via direct electron transfer, while at higher overpotentials NDMA oxidation occurs via a combination of direct electron transfer and reaction with hydroxyl radicals. The CV scans in Figure 3 show that measurable rates of NDMA oxidation occur at lower potentials than those required for measurable rates of water oxidation. The currents in the electrolyte solution containing 1.35 mM NDMA were greater than those in the blank electrolyte over the potential range of 1.8-2.9 V. The greater currents in the presence of NDMA and the current peak at 2.6 V can be attributed to direct oxidation of NDMA adsorbed to the electrode surface. The absence of a reduction peak in the reverse scan is indicative of irreversible electron transfer, or a reversible electron transfer followed by a fast chemical reaction. Scan rates as high as 16 V/s did not detect a reduction peak in the reverse scan, and the oxidation peak was shifted to more positive potentials at faster scan rates, both of which suggest an irreversible direct electron transfer reaction (36). The products of NDMA oxidation also support a direct electron transfer mechanism. The first detectable NDMA oxidation products in the batch and flow-through experiments were NO3- and DMA. The proposed half reaction may be expressed as

FIGURE 4. Initial measured and calculated NDMA reaction rates (r) in the RDE reactor at different electrode potentials. Error bars represent 95% confidence intervals on duplicate experiments. For current analysis error bars are contained within the data point. These results differ from prior studies involving oxidation of NDMA by hydroxyl radicals, where the initial amino product was MA (13, 37), and was attributed to a reaction mechanism that involves H-atom abstraction from the methyl groups of NDMA (13, 37). The difference in the initial amino product of NDMA oxidation observed here suggests that H-atom abstraction was not the rate-determining step at low overpotentials. Comparison between the two methods used to calculate initial reaction rates of NDMA oxidation also support direct electron transfer at low overpotentials. Figure 4 shows a plot of the initial rates of NDMA oxidation calculated by current analysis (eq 1) assuming a 2e- oxidation, as shown in eq 3, and the analytically measured initial rates of NDMA disappearance from solution. At potentials below 1.84 V no current increase was observed upon NDMA addition, which is consistent with results from the CV scans. For electrode potentials between 2.00 and 2.39 V, the rates calculated by the two methods were nearly identical, as shown in Table 1. The agreement between rates calculated from a current analysis and the analytically measured reaction rates supports a mechanism involving direct electron transfer, with both electrons being transferred from NDMA at approximately the same rate. With increasing electrode potentials, the rates calculated by current analysis fall increasingly below the analytically measured rates (Figure 4 and Table 1). This result indicates that at higher overpotentials there is a second mechanism accounting for NDMA removal from solution, possibly hydroxyl radical attack on the methyl groups of NDMA (13, 37). The apparent activation barrier can also provide insight into the rate-determining step for NDMA oxidation by comparing it to activation barriers calculated by DFT for different reaction mechanisms. As shown in SI Figure S-4, apparent activation barriers of 3.1 ( 0.5 and 9.5 ( 0.8 kJ/mol were measured at potentials of 2.5 and 3.14 V, respectively. Activation barriers this low are often attributed to unactivated processes (38), such as the effect of temperature on the composition and thickness of the electrical double layer at the electrode surface, or the relative adsorption strengths of water and NDMA on the electrode surface. Density Functional Theory Modeling. The activation barrier for oxidation of NDMA via a direct electron transfer mechanism was determined by calculating the activation barrier associated with the loss of one electron from NDMA by a vertical electron transfer reaction using the method of Anderson and Kang (39). DFT simulations were used to calculate the optimized geometry for NDMA and its cation radical oxidation product, NDMA•+. As shown in SI Figure S-5, the largest structural change between NDMA and NDMA•+ is the shortening of the NsN bond. This indicates that the NsN bond length can be used as a reaction coordinate in calculating the activation barrier for direct electron transfer. Figure 5a shows the energies of the reactants

FIGURE 5. (a) Energy profiles as a function of the NsN bond length reaction coordinate at an electrode potential of 1.3 V/ SHE for the reactant and products shown in eq 4. (b) Activation barrier calculations as a function of electrode potential for direct oxidation based on the calculations in (a). and products for reaction 4 as a function of the NsN bond length for an electrode potential of 1.3 V.

The reactant energies were calculated by varying the length of the NsN bond from its minimum energy length of 1.07 Å, followed by geometric optimization of the structure. The product energies were calculated using the atomic positions determined from the optimized reactant structures, followed by self-consistent field optimization of the electronic configurations. Electron energies from the vacuum scale were converted to the SHE scale by subtracting 4.6 eV (39). Product energies as a function of electrode potential were determined by shifting the energy profile of the product species downward by 96.5 kJ/mol (i.e., 1.0 eV) to increase the electrode potential by 1.0 V and upward by 96.5 kJ/mol to decrease the electrode potential by 1.0 V (39). Intersection of the product and reactant energy profiles yields the bond length of the transition state and the activation energy for the reaction, as illustrated in Figure 5a. The higher the electrode potential, the smaller the NsN bond distortion required to reach the transition state. By shifting the products energy profile up and down, activation energies as a function of electrode potential were calculated, as shown in Figure 5b. Figure 5b shows that the activation barrier for direct electron transfer decreases from 108 kJ/mol at 1.3 V to 0 kJ/mol at 1.8 V. The activation-less barrier at electrode potentials >1.8 V indicates that electronic energy states in the electrode are lower than those in NDMA. The activationless electron transfer at potentials >1.8 V is consistent with the minimum potential necessary to observe a measurable current increase upon NDMA addition to solution in both VOL. 43, NO. 21, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

8305

FIGURE 6. (a) Initial reactants, (b) transition state, and (c) final products for hydroxyl radical attack on the nitroso group of NDMA. Atom key: C ) gray; N ) blue; O ) red; H ) white. CV and NDMA oxidation experiments and is consistent with the low measured activation barrier at 2.5 V of 3.1 ( 0.5 kJ/mol. Since experimental results in Figure 4 suggest that two electrons were transferred from NDMA at approximately the same rate, the activation barrier for hydroxyl radical attack at NDMA(+2) was also calculated using DFT. Simulations showed that NDMA(+2) was a stable species that formed a water adduct (SI Figure S-5). DFT calculations indicate that hydroxyl radical attack on NDMA(+2) and the NDMA(+2) water adduct proceed without activation barriers. In both instances, DMA and HNO2 were produced. The absence of an activation barrier suggests that this reaction is important in NDMA oxidation by BDD electrodes. At higher overpotentials considerable concentrations of hydroxyl radicals are produced and the direct reaction between NDMA and hydroxyl radicals steadily increases. For example, at 2.0 and 3.14 V direct oxidation was estimated to account for 97 and 55% of NDMA oxidation, respectively (Table 1), whereas the balance was attributed to oxidation by hydroxyl radicals. However, at potentials between 3.14 and 4.14 V, NDMA oxidation rates were similar and direct oxidation estimates decreased only marginally (55 and 42%, respectively). At potentials above 3.0 V, it was previously estimated that the BDD surface is saturated with hydroxyl radicals (40). Thus, at potentials where hydroxyl radicals are readily produced, the direct oxidation of NDMA on the electrode surface is still occurring. In order to elucidate the mechanism of hydroxyl radical attack, DFT simulations were used to calculate the activation barriers for hydroxyl radical attack at the nitroso group and methyl group of NDMA, as shown in eqs 5 and 6, respectively. The reaction pathway shown in eq 5 results in the production

of DMA and NO3-, which was previously observed during photocatalytic oxidation of NDMA in a UV-A/TiO2 system (15). Figure 6 shows the initial reactants, transition state, and final products for hydroxyl radical attack at the nitroso group. The overall Gibbs energy change for this reaction is 34 kJ/mol and the activation barrier is 42 kJ/mol. The positive overall reaction energy and the high activation barrier compared to the experimental values indicate that hydroxyl radical attack at the nitroso group of NDMA was not an important reaction in NDMA oxidation. The increase in the measured activation barrier from 3.1 to 9.5 kJ/mol for a potential increase from 2.5 to 3.14 V, suggests that a small fraction of NDMA oxidation may be occurring by hydroxyl radical attack at the nitroso group of NDMA at higher overpotentials. The reaction pathway shown in eq 6 has previously been reported to form MA and NO3- during the homogeneous 8306

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 43, NO. 21, 2009

reaction between NDMA and hydroxyl radicals (13, 37). DFT simulations indicate that hydroxyl radical attack at the methyl group proceeds without an activation barrier and this pathway is likely an important contributor to NDMA oxidation at higher overpotentials. The experimental and DFT results indicate that the ratedetermining step of NDMA oxidation on BDD electrodes proceeds by a combination of direct electron transfer and hydroxyl radical oxidation. Direct oxidation proceeds readily at potentials >1.8 V, and is the dominant reaction pathway at low overpotentials. The pathway involves a two electron transfer from NDMA to the BDD anode which allows an activation-less reaction between a hydroxyl radical and the NDMA(+2) water adduct. At higher overpotentials, where hydroxyl radicals are produced in significant quantities, both direct electron transfer and the reaction between NDMA and hydroxyl radicals contribute approximately equally to the rate of NDMA oxidation.

Acknowledgments We thank the Donors of the American Chemical Society Petroleum Research Fund (PRF 43535-AC5) and the Technology Research Infrastructure Fund at the University of Arizona, funded by Proposition 301 by the state of Arizona, for support of this work. We also thank Mr. Craig Duncan for assistance in conducting oxidation experiments and sample analysis.

Supporting Information Available Methods for DFT simulations, the effect of current density on reaction rates in the flow-through reactor, the calculation of activation barriers for NDMA, and figures of molecular structures of NDMA, NDMA(+1), and the NDMA(+2)/H2O adduct. This material is available free of charge via the Internet at http://pubs.acs.org.

Literature Cited (1) Fine, D. H.; Rounbehler, D. P.; Rounbehler, A.; Silvergleid, A.; Sawicki, E.; Krost, K.; Demarrais, G. A. Determination of dimethylnitrosamine in air, water, and soil by thermal-energy analysis - measurements in Baltimore, MD. Environ. Sci. Technol. 1977, 11 (6), 581–584. (2) Lijinsky, W. N-nitroso compounds in the diet. Mutat. Res., Genet. Toxicol. Environ. Mutagen. 1999, 443 (1-2), 129–138. (3) Mitch, W. A.; Sharp, J. O.; Trussell, R. R.; Valentine, R. L.; AlvarezCohen, L.; Sedlak, D. L. N-nitrosodimethylamine (NDMA) as a drinking water contaminant: A review. Environ. Eng. Sci. 2003, 20 (5), 389–404. (4) O’Neill, I. K.; Borstel, R. C. V.; Miller, C. T.; Long, J.; Bartsch, H. IARC Scientific Publication 57; Oxford University Press: Lyon, 1984. (5) California Department of Health Services. California drinking water: NDMA-related activities. http://www.cdph.ca.gov/certlic/ drinkingwater/pages/ndma.aspx (2002). (6) Greene, B.; McClure, M. B.; Johnson, H. T. Destruction and decomposition of hypergolic chemicals in a liquid propellant testing laboratory. Chem. Health Saf. 2004, 6–13, Jan./Feb. . (7) Mitch, W. A.; Sedlak, D. L. Formation of N-nitrosodimethylamine (NDMA) from dimethylamine during chlorination. Environ. Sci. Technol. 2002, 36 (4), 588–595. (8) Gui, L.; Gillham, R. W.; Odziemkowski, M. S. Reduction of N-nitrosodimethylamine with granular iron and nickel-enhanced iron. 1. Pathways and kinetics. Environ. Sci. Technol. 2000, 34 (16), 3489–3494. (9) 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 (16), 3495–3500. (10) Davie, M. G.; Reinhard, M.; Shapley, J. R. Metal-catalyzed reduction of N-nitrosodimethylamine with hydrogen in water. Environ. Sci. Technol. 2006, 40 (23), 7329–7335. (11) Frierdich, A. J.; Shapley, J. R.; Strathmann, T. J. Rapid reduction of N-nitrosamine disinfection byproducts in water with hydrogen and porous nickel catalysts. Environ. Sci. Technol. 2008, 42 (1), 262–269.

(12) Frierdich, A. J.; Joseph, C. E.; Strathmann, T. J. Catalytic reduction of N-nitrosodimethylamine with nanophase nickel-boron. Appl. Catal., B 2009, 90 (1-2), 175–183. (13) Lee, C.; Yoon, J.; Von Gunten, U. Oxidative degradation of N-nitrosodimethylamine by conventional ozonation and the advanced oxidation process ozone/hydrogen peroxide. Water Res. 2007, 41 (3), 581–590. (14) Kruithof, J. C.; Kamp, P. C.; Martijn, B. J. UV/H2O2 treatment: A practical solution for organic contaminant control and primary disinfection. Ozone: Sci. Eng. 2007, 29 (4), 273–280. (15) Lee, J.; Choi, W. Y.; Yoon, J. Photocatalytic degradation of N-nitrosodimethylamine: Mechanism, product distribution, and TiO2 surface modification. Environ. Sci. Technol. 2005, 39 (17), 6800–6807. (16) Lee, C.; Choi, W.; Kim, Y. G.; Yoon, J. UV photolytic mechanism of N-nitrosodimethylamine in water: Dual pathways to methylamine versus dimethylamine. Environ. Sci. Technol. 2005, 39 (7), 2101–2106. (17) Lowry, G. V.; Reinhard, M. Pd-catalyzed TCE dechlorination in groundwater: Solute effects, biological control, and oxidative catalyst regeneration. Environ. Sci. Technol. 2000, 34 (15), 3217– 3223. (18) Chaplin, B. P.; Roundy, E.; Guy, K. A.; Shapley, J. R.; Werth, C. J. Effects of natural water ions and humic acid on catalytic nitrate reduction kinetics using an alumina supported Pd-Cu catalyst. Environ. Sci. Technol. 2006, 40 (9), 3075–3081. (19) Chaplin, B. P.; Shapley, J. R.; Werth, C. J. Regeneration of sulfurfouled bimetallic Pd-based catalysts. Environ. Sci. Technol. 2007, 41 (15), 5491–5497. (20) Farrell, J.; Kason, M.; Melitas, N.; Li, T. Investigation of the longterm performance of zero-valent iron for reductive dechlorination of trichloroethylene. Environ. Sci. Technol. 2000, 34 (3), 514–521. (21) Carter, K. E.; Farrell, J. Oxidative destruction of perfluorooctane sulfonate using boron-doped diamond film electrodes. Environ. Sci. Technol. 2008, 42 (16), 6111–6115. (22) Zhi, J. F.; Wang, H. B.; Nakashima, T.; Rao, T. N.; Fujishima, A. Electrochemical incineration of organic pollutants on borondoped diamond electrode: Evidence for direct electrochemical oxidation pathway. J. Phys. Chem. B 2003, 107 (48), 13389– 13395. (23) Gyorgy, F. G., D.; Comninellis, C.; Perret, A.; Haenni, W. Oxidation of organics by intermediates of water discharge on IrO2 and synthetic diamond anodes. Electrochem. Solid-State 1999, 2 (5), 228–230. (24) Marselli, B.; Garcia-Gomez, J.; Michaud, P. A.; Rodrigo, M. A.; Comninellis, C. Electrogeneration of hydroxyl radicals on boron-

(25) (26) (27)

(28) (29) (30) (31) (32) (33) (34) (35) (36) (37) (38) (39)

(40)

doped diamond electrodes. J. Electrochem. Soc. 2003, 150 (3), D79–D83. Ferro, S. Synthesis of diamond. J. Mater. Chem. 2002, 12 (10), 2843–2855. Chen, X. M.; Chen, G. H.; Gao, F. R.; Yue, P. L. High-performance Ti/BDD electrodes for pollutant oxidation. Environ. Sci. Technol. 2003, 37 (21), 5021–5026. Zhu, X. P.; Shi, S. Y.; Wei, J. J.; Lv, F. X.; Zhao, H. Z.; Kong, J. T.; He, Q.; Ni, J. R. Electrochemical oxidation characteristics of p-substituted phenols using a boron-doped diamond electrode. Environ. Sci. Technol. 2007, 41 (18), 6541–6546. Liu, L.; Zhao, G. H.; Wu, M. F.; Lei, Y. Z.; Geng, R. Electrochemical degradation of chlorobenzene on boron-doped diamond and platinum electrodes. J. Hazard. Mater. 2009, 168 (1), 179–186. Anglada, A.; Urtiaga, A.; Ortiz, I. Pilot scale performance of the electro-oxidation of landfill leachate at boron-doped diamond anodes. Environ. Sci. Technol. 2009, 43 (6), 2035–2040. Delley, B. An all-electron numerical-method for solving the local density functional for polyatomic-molecules. J. Chem. Phys. 1990, 92, 508–517. Delley, B. From molecules to solids with the DMOL3 approach. J. Chem. Phys. 2000, 113, 7756–7764. Materials Studio, v.4.2; Accelrys Corporation: San Diego, CA. Delley, B. Fast calculation of electrostatics in crystals and large molecules. J. Phys. Chem. 1996, 100, 6107–6110. Perdew, J.; Burke, K.; Ernzerhof, M. Generalized gradient approximation made simple. Phys. Rev. Lett. 1996, 77 (18), 3865– 3868. Delley, B. The conductor-like screening model for polymers and surfaces. Mol. Simul. 2006, 32, 117–123. Wang, J., Analytical Electrochemistry, 3rd ed.; John Wiley and Sons, Inc.: Hoboken, NJ, 2006; p 250. Mezyk, S. P.; Cooper, W. J.; Madden, K. P.; Bartels, D. M. Free radical destruction of N-nitrosodimethylamine in water. Environ. Sci. Technol. 2004, 38 (11), 3161–3167. Smith, J. M., Chemical Engineering Kinetics; McGraw-Hill: New York., 1980. Anderson, A. B.; Kang, D. B. Quantum chemical approach to redox reactions including potential dependence: Application to a model for hydrogen evolution from diamond. J. Phys. Chem. A 1998, 102 (29), 5993–5996. Farrell, J.; Martin, F. J.; Martin, H. B.; O’Grady, W. E.; Natishan, P. Anodically generated short-lived species on boron-doped diamond film electrodes. J. Electrochem. Soc. 2005, 152 (1), E14–E17.

ES901582Q

VOL. 43, NO. 21, 2009 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

8307