Indium as a Promoter Metal - ACS Publications - American Chemical

Mar 19, 2008 - Reduction of. N-Nitrosodimethylamine: Indium as a Promoter Metal. MATTHEW G. DAVIE, KAIMIN SHIH,. FEDERICO A. PACHECO,. JAMES O...
0 downloads 0 Views 8MB Size
Download PDF
Environ. Sci. Technol. 2008, 42, 3040–3046

Palladium-Indium Catalyzed Reduction of N-Nitrosodimethylamine: Indium as a Promoter Metal MATTHEW G. DAVIE, KAIMIN SHIH, FEDERICO A. PACHECO, JAMES O. LECKIE, AND MARTIN REINHARD* Department of Civil & Environmental Engineering, Stanford University, Stanford, California 94305-4020

Received October 15, 2007. Revised manuscript received January 29, 2008. Accepted January 29, 2008.

An emerging technology for the removal of N-nitrosodimethylamine (NDMA) from drinking and groundwater is reductive destruction using noble metal catalysts and hydrogen gas as a reducing agent. Bimetallic palladium-indium (Pd-In) supported on alumina combines the ability of In to activate NDMA with the hydrogen activating properties of Pd. This study examined the effect of In addition to a commercial 5% Pd by weight on γ-Al2O3 catalyst on the efficacy of NDMA reduction. The pseudo-first-order rate constant increased proportionately to In loading from 0.057 h-1 for 0% In to a maximum of 0.25 h-1 for 1% In and then decreased with additional In loading. Data suggest that hydrogen activation occurred only on Pd surfaces and In activated NDMA 20 times more effectively than Pd on a mass basis. The rate-limiting factor was NDMA activation for In loadings below 1%. The decrease at higher loadings is interpreted as In blocking pore spaces and limiting access to Pd sites, suggesting monatomic hydrogen limitation. The only products detected were dimethylamine and ammonium with carbon and nitrogen balances in excess of 92%, consistent with a mechanism involving reductive N-N bond cleavage. Results from this study serve as a basis for optimizing bimetallic catalysts for treating NDMA contaminated waters.

lesser concern than NDMA (17). Metallic Fe acts as both the catalyst and the electron donor by providing an activating surface for NDMA and reducing water to monatomic hydrogen (H*), which reacts with activated NDMA (NDMA*). Using bimetallic Fe-Ni, Fe supplies H*, a strong reducing agent, and Ni provides a catalytic surface for NDMA activation. Ni surfaces are more effective than Fe for NDMA activation, enhancing the overall transformation rate; Fe alone reduced NDMA with a half-life of 13 ( 2 h compared with 2–20 min for Fe-Ni in the same system (15). Reactivity of the Fe-Ni catalyst was lost over time, possibly because oxide surface coatings inactivated the Fe surface and reduced the supply of H*, as was observed for palladium-iron (Pd-Fe) catalysts in a separate study (18). Nanoscale Fe would be expected to show faster kinetics, as shown for trichloroethylene dechlorination (19), but its long-term reactivity toward NDMA remains to be demonstrated. Pd efficiently transforms nitrite to nitrogen gas using hydrogen as the reductant (21, 22) but does not reduce nitrate to nitrite. It has been shown that the efficacy of Pd for reduction of trace contaminants can be enhanced by adding a promoter metal such as copper (Cu). Bimetallic Pd-Cu reduces nitrate to dinitrogen with a fraction (as low as 5.6%) converted to ammonia (23, 24); Cu transforms nitrate to nitrite, which is then reduced to dinitrogen and ammonia on Pd surfaces. Pd-Cu catalyst (1% Pd and 0.3% Cu by weight) destroyed NDMA 6 times faster than monometallic 1% Pd by weight in a previous study (20); however, this study did not consider other promoter metals nor evaluate the impact of varying the promoter metal loading. In field and laboratory studies examining Pd catalysts in groundwater augmented with hydrogen, growth of hydrogen utilizing and sulfate reducing bacteria led to the formation of hydrogen sulfide, a powerful poison for Pd and other noble metal catalysts (22, 24–27). Controlling the growth of sulfi-

Introduction N-Nitrosodimethylamine (NDMA) is frequently found as a contaminant in reclaimed and groundwater disinfected with chloramines, with concentrations ranging from 0.01 to 400 µg L-1 (1, 2). In groundwater, NDMA stems from spilled 1,1dimethylhydrazine, a rocket fuel containing NDMA as an impurity (2). In reclaimed and drinking water, NDMA is a byproduct of disinfection with monochloramine and to a lesser degree chlorine (3–6). Its presence is increasingly troublesome in drinking, reclaimed, and groundwater (7, 8) because action levels are as low as 10 ng L-1 in California (1). Traditional physicochemical treatment technologies are relatively inefficient (8–11), and ultraviolet (UV) treatment is an effective but costly alternative (12–14). Recently, zerovalent iron (Fe) and nickel-enhanced iron (Fe-Ni) were found to transform NDMA to dimethylamine (DMA) and ammonium (NH4+) (15, 16), both products of * Corresponding author phone: (+1) 650-723-0308; fax: (+1) 650723-7058; email: [email protected]. 3040

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 8, 2008

FIGURE 1. NDMA reduction profiles for (a) selected Pd-In catalysts to show the influence of In loading and (b) other catalysts for comparison of Pd-In with other Pd-based catalysts. All experiments were run with 100 mg L-1 catalyst. Dashed lines represent pseudo-first-order model fits. 10.1021/es7023115 CCC: $40.75

 2008 American Chemical Society

Published on Web 03/19/2008

TABLE 1. Observed Pseudo-First-Order Rate Constants k′obs for the Reduction of NDMA in Solutions Containing 100 mg L-1 Catalyst As a Function of Pd, In, or Cu Composition, the Increase in k′obs Attributed to Promoter Metal, and the Metal-Specific Second-Rate Constant k ′obs/{Me}b (L gme-1 h-1) catalyst

k ′obs (h-1) overall k ′obs increasea (h-1) Pd

c

Pd1 Pd5 Pd1-Cu0.3c Pd5-In0.5 Pd5-In0.7 Pd5-In1 Pd5-In1.2 Pd5-In1.5 In5

0.012 ( 0.001 0.057 ( 0.008 0.067 ( 0.007 0.13 ( 0.02 0.14 ( 0.02 0.25 ( 0.02 0.19 ( 0.02 0.16 ( 0.01 0d

In

11.5 11.5 0.055 (Cu) 0.073 0.083 0.193 0.133 0.103

222 (Cu) 250 195 251 160 104

Increase in k ′obs relative to respective Pd-only base catalyst (Pd1 or Pd5). b Metal-specific second-order rate constant; rate obtained by normalizing k ′obs with respect to the measured metal content from Table 2. c Data taken from previous study (20). d No NDMA reduction observed with In-only catalyst. a

dogenic bacteria requires periodic oxidative treatment with bleach (25, 27) or hydrogen peroxide (28). In a field demonstration treating anoxic water for TCE removal, Pd catalyst withstood multiple cycles of poisoning and bleaching for extended periods (25, 29, 30) and single batches of Pd catalyst have been effective in laboratory and field demonstrations for periods exceeding five years (26–28). By contrast, tests with bimetallic Pd-Cu catalysts showed that while Cu significantly increased the rate of NDMA removal (20), it was oxidatively stripped from the catalyst surface during bleach regeneration (24). This precludes Cu or any other metal that cannot withstand oxidative conditions for field applications where regeneration is necessary. Indium (In) was considered as a potential promoter metal for NDMA reduction because In and Cu showed similar promoting effects for nitrate reduction when added to Pd catalysts (24), and In withstood oxidative regeneration (31). To extend bimetallic catalysis to field applications, the kinetics must be optimized while maintaining efficacy after multiple regeneration cycles. The objectives of this study were to (i) determine the In loading that maximizes catalyst activity with Pd-based catalyst and (ii) develop a mechanistic understanding of the role of In during NDMA transformation with Pd-In catalysts. This study demonstrates that Pd-In catalysts are potentially applicable for the reductive destruction of NDMA in ground and drinking water. Kinetic and spectroscopic data are interpreted in terms of a possible reduction mechanism.

FIGURE 2. Catalyst mass-normalized pseudo-first-order rate constants plotted against In loading (n > 4), indicating a bell-shaped curve with maximum NDMA reduction capacity with 1% In by wt.

Materials and Methods The batch experiments in this study were performed using the same materials and methods reported in a previous study (20) except as noted below. Chemicals. All chemicals in the batch experiments were the same as previously reported (20). For NH4+ analysis, ammonium cyanurate was purchased in preweighed packets (Hach Analytical; Loveland, CO) and sodium hydroxide (NaOH) was purchased as 2 N solution (Sigma Aldrich; St. Louis, MO). All chemicals were used as received without further purification. Catalysts. Two catalysts from the previous study (20) were used: 1% Pd by weight on γ-Al2O3 (Pd1) and 1% Pd 0.3% Cu by weight on γ-Al2O3 (Pd1-Cu0.3). Other catalysts were prepared by Shapley and collaborators using the incipient wetness method (24, 38). Starting with prereduced 5% Pd by weight (Pd5) on γ-Al2O3 (Sigma Aldrich), In was added to five separate catalyst batches, yielding final nominal catalyst loadings: Pd5-In0.5, Pd5-In0.7, Pd5-In1, Pd5-In1.2, and Pd5-In1.5, where the number after each metal represents its respective nominal weight loading on the γ-Al2O3 support (e.g., Pd5-In0.5 ) 5% Pd, 0.5% In by weight). A batch of Pd5 was reserved and In5 (no Pd) was prepared using γ-Al2O3 powder. The final metal content of each catalyst was measured by inductively coupled plasma-mass spectrometry (ICP-MS). Catalyst Characterization Methods. Catalysts were analyzed using scanning electron microscopy (SEM) and transmission electron microscopy (TEM) in order to evaluate the impact of In addition to the Pd base catalyst. The particle cross-sections for SEM and energy dispersive spectroscopy (EDAX) analyses were prepared by first suspending the particles in 150 °C heated CrystalboundTM 509 adhesive liquid (Ted Pella; Redding, CA). The adhesive was then

TABLE 2. Metal Loading, Metal Ratio Measured at the Particle Surface, and Surface Area of Catalysts Used catalyst Pd1a Pd1-Cu0.3a Pd5 Pd5-In0.5 Pd5-In0.7 Pd5-In1 Pd5-In1.2 Pd5-In1.5 In5

nominal loading [% (w w-1)] 1% 1% 5% 5% 5% 5% 5% 5% 5%

Pd Pd-0.3% Cu Pd Pd-0.5% In Pd-0.7% In Pd-1% In Pd-1.2% In Pd-1.5% In In

measured loadingb [% (w w-1)]

5.19% 5.89% 5.47% 5.52% 4.44% 5.79%

Pd Pd-0.52% Pd-0.68% Pd-1.03% Pd-1.40% Pd-1.57%

particle surface ratioc [In:Pd]

In In In In In

0.00 1.52 ( 0.05 1.31 ( 0.03 2.00 ( 0.12 1.40 ( 0.05 1.32 ( 0.08

catalyst specific surface aread (m2 gcat.-1) 153.3 140.8 133.2 93.4 90.3 88.9 88.2 83.7 71.3

a Data taken from previous study (20). b Measured by ICP-MS. c Ratio of metal masses on external catalyst particle surface, by XPS analysis, 95% confidence intervals shown (n > 4). d Measured by nitrogen adsorption using a Coulter SA 3100.

VOL. 42, NO. 8, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3041

FIGURE 3. SEM images of Pd5-In1 (5% Pd, 1.5% In by weight) with scale bar representing 10 µm: (a) surface of one catalyst particle, approximately 20 µm in diameter; and (b) cross-section of the catalyst particle showing relatively uniform γ-Al2O3 base structure. TEM images of Pd5 (5% Pd by weight) with scale bar representing 5 µm: (c) cross-section of the catalyst particle showing pore structure of alumina support with internal Pd deposition; and (d) edge of the catalyst surface, showing Pd clusters ranging from 8 to 100 nm in diameter (measured from multiple TEM images; data not shown). allowed to cool to room temperature for hardening. Crosssections were created by cutting through the adhesive block with diamond cutting disk and then polishing with 30 to 0.1 µm diamond lapping films to enhance image quality. All samples were cleaned by sonication in water for 5 min prior to SEM and EDAX analyses. For TEM analysis, catalyst particles were embedded with EMbed 812 (Electron Microscopy Sciences; Hatfield, PA) in BEEM capsules (Ted Pella) and cured overnight at 48 °C. Thin sections were prepared using an Ultracut S microtome (Leica; Wetzlar, Germany) and mounted onto carbon/Formvar coated copper TEM grids. Images were obtained with a TEM system consisting of a Jeol TEM1230 microscope and a Gatan 967 cooled CCD camera at an accelerating voltage of 80 kV. Catalyst specific surface area was measured using a Coulter SA 3100 surface area and pore size analyzer (Beckman Coulter; Fullerton, CA) and represents the total surface area, not the reactive metal surface areas (e.g., Pd, In).

Results and Discussion Kinetics of Surface Reaction. NDMA reduction followed pseudo-first-order kinetics (k ′obs) in all cases, as shown in Figure 1 and summarized in Table 1. Rate constants depended on Pd loading as well as the type and loading of promoter metal. Pd5 performed 5 times better than Pd1 (Figure 1b), confirming the hypothesis from a previous study (20) that the number of active sites and thus catalyst activity increased proportionately to Pd mass in the loading range analyzed. NDMA transformation occurs over time scales of days, relatively slowly when compared with the formation of H*, which is mobile on Pd (32). Therefore, the reaction rate is 3042

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 8, 2008

assumed to be independent of hydrogen pressure as long as it exceeds a threshold value and sufficient Pd sites are accessible to provide an excess of H* (25). Taken together, the pseudo-first-order transformation of NDMA and the linear dependence on Pd mass indicate a second-order process, first-order in NDMA and metal surface. Pd1-Cu0.3, included in this study for comparison with Pd-In catalysts, removed NDMA in a pseudo-first-order process approximately equal in time scale to Pd5 (Figure 1b), indicating a strong promoting effect of Cu-i.e., 0.3% Cu by weight was as effective as an additional 4% Pd. Effect of In Loading on NDMA Reduction Rate. Adding In to Pd5 increased catalyst efficacy for NDMA reduction proportionately to In content up to approximately 1% In by weight (Pd5-In1), above which the trend reversed (Table 1 and Figure 2). NDMA reduction was not observed using an In-only catalyst (In5) over a period of 3 weeks, indicating that In cannot produce H* from dihydrogen, as was found previously for Cu (20). Adding 0.5% In by weight to Pd5 approximately doubled the rate constant, from 0.057 to 0.13 h-1. Further addition of In increased the rate in proportion to In loading up to 0.25 h-1 at 1% In by weight (i.e., Pd5-In1). The rate decrease above 1% In is interpreted as In blocking pore spaces within the catalyst, thereby limiting the availability of Pd surfaces and, therefore, H*. Figure 2 displays the k ′obs normalized by total catalyst mass in the system (100 mgcat L-1), to provide a visual depiction of the promoting effect of In. The curve shows a bell shape with a maximum at approximately 1% In by weight.

FIGURE 4. Composition of a particle of Pd5-In1 (5% Pd, 1.5% In by weight) catalyst analyzed by SEM and EDAX: (a) SEM of catalyst particle, approximately 20 µm in diameter; (b) EDAX analysis of the same catalyst particle selecting for Al content, revealing uniform Al structure; (c) EDAX analysis of Pd content, showing uniform distribution of Pd; and (d) EDAX analysis of In content, where distribution of In appears relatively even throughout the catalyst particle. The Pd and In images do not have as much contrast as the Al image because the signal was much weaker, so obtaining a significant signal-to-noise ratio was more difficult. A dashed line was added as a visual guide to show the outline of the catalyst particle and make it easier to see the even distribution of each metal throughout the particle. The promoting effect of In and Cu were evaluated from the kinetic data summarized in Table 1. The observed pseudofirst-order rate constant, defined as d[NDMA] ) k′obs[NDMA] dt

(1)

was calculated by regression from the kinetic data shown in Figure 1 at catalyst concentrations of 100 mg L-1. The massspecific second-order rate constants for the promoter metals, In or Cu, were obtained by first calculating the increase in k′obs attributed to the effect of reactions at the In or Cu surfaces: ′

′

′

k In,obs ) k obs - k Pd

(2)

where k′Pd represents the pseudo-first-order rate constant for the corresponding Pd-only catalyst (Pd5 for Pd-In catalysts tested). The mass-specific second-order rate constants for In were obtained from k′In ) k′In,obs ⁄ {In}

(3)

where {In} represents the In content in suspension. The constant for Cu was calculated similarly. Indicating the content of In in the suspension in gIn L-1 results in k′′In with units of (L gIn-1 h-1), it is evident that, up to a content of 1% In by weight, the In-specific second-order rate constant is remarkably constant, varying only between 195 and 250 L gIn-1 h-1. This value is approximately 20 times larger than the Pd-specific second-order rate constant (11.5

L gPd-1 h-1) and nearly equal to that of Cu (222 L gPd-1 h-1). Increasing the In content in the catalyst above 1% by weight decreased both the overall activity of the catalyst (Figure 2) and the In-specific second-order rate constant (Table 1), presumably for the reasons discussed above. Assuming that Pd provides a constant contribution to k′obs, eq 4 represents the overall rate for catalysts loaded with In, which we verified up to 1% In by weight. It was further assumed that dissolved hydrogen was available to produce excess H*: -

d[NDMA] ) (c + k′In{In})[NDMA] dt

(4)

where the constant c accounts for NDMA reduction at Pd sites and {In} corresponds to the mass of In present per liter of suspension (gIn L-1). For monometallic Pd5, eq 1 reduces to the first term, c[NDMA], and c is equal to the massnormalized pseudo-first-order rate constant shown in Table 1, 11.5 L gPd-1 h-1. The efficacy of similarly prepared Pd-In catalysts may be predicted using c and kIn as long as H* is present in excess. Also, on the basis of the values of c and kIn, it can be estimated that In generates NDMA* approximately 20 times more effectively per weight In than Pd. Catalyst Characterization. As seen in the SEM picture of Figure 3a, the catalyst particles were approximately 20 µm in diameter and highly porous, as specified by the manufacturer and further evidenced by the fragmented surface. Figure 3b shows an SEM cross-section, revealing the uniform VOL. 42, NO. 8, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3043

FIGURE 5. NDMA destruction and product formation. Using 100 mg L-1 Pd-In catalyst (5% Pd, 1% In by weight on γ-Al2O3): (a) normalized by initial carbon concentration, carbon balance > 97%; (b) normalized by initial nitrogen concentration, nitrogen balance > 92%. Dashed lines are visual guides and do not represent model fits. alumina structure of the catalyst particle. Figure 3c shows the same cross-section under TEM, highlighting deposition of Pd within the macro pore structure. A TEM image of the surface edge in Figure 3d reveals cluster sizes approximately 8–100 nm in diameter. Figure 4a shows an SEM image of a Pd5-In1 catalyst particle analyzed in Figure 4b-d using EDAX to determine metal distribution. This 20 µm diameter particle was outlined with a dashed line to provide a visual guide for the subsequent EDAX elemental maps of Pd and In (Figure 4c,d) because the metals are barely detected. Figure 4b shows the EDAX results when selecting for Al. As expected, there is a large concentration of Al within the catalyst structure, as it is the primary metal in the Al2O3 support material. Parts c and d of Figure 4 show the EDAX results for Pd and In, respectively. A relatively even distribution is apparent within the catalyst particle for both Pd and In, consistent with the assumption that both metals penetrated the catalyst pore structure. The XPS data shown in Table 2 analyze external catalyst particle surfaces. The Pd:In ratio varies within a relatively narrow range and does not track the bulk metal composition measured by ICP-MS, as would be expected. The fact that the highest rate agrees with the highest In:Pd ratio measured at the external particle surface is difficult to explain because the reaction is presumed to occur at internal surfaces where the surface ratios are presumed to correspond to the measured metal contents in the catalyst. The even deposition of Pd shown in the SEM/EDAX images agrees with the finding that NDMA reduction rate increased proportionally to Pd mass loading-i.e., most of the Pd loaded to the alumina support was available for NDMA reduction as small clusters within the macro pore structure. The surface area measurements of Table 2 show that In loading using the incipient wetness method reduced the specific metal surface approximately proportionately to the In mass added. This implies In formed larger clusters that may have blocked some of the pores, making sites deeper in the pore structure unavailable for reaction by limiting mass transfer. Highresolution SEM and TEM, plus dynamic chemisorptions, 3044

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 8, 2008

would be necessary to effectively evaluate the pore blocking. Pd and In dispersion and their spatial relationship at the nanoscale merits further exploration but was outside the scope of this study. NDMA Reduction Products. Figure 5 shows NDMA disappearance and product formation using Pd5-In1; similar results were obtained for all catalysts tested (data not shown). The only products detected were DMA and NH4+, which appeared to be stable. DMA accounted for nearly all of the carbon balance (>97%) as was the case in other studies (15, 16, 20). As postulated by Gui et al. (15) and verified in this study, the nitrosyl group of NDMA was reduced to NH4+; combined with DMA, the nitrogen balance was >92%. DMA is only a weak NDMA precursor in waters that are subsequently chloraminated or chlorinated, with yields less than 1% after 4 h contact with 0.01 mM monochloramine (39). With NDMA concentrations in drinking water typically below 10 ng L-1 (1), the amount of ammonium formed is expected to be immaterial. NH4+ formation may be problematic at sites with very high levels of NDMA, but even heavily contaminated industrial sites are below 1 mg L-1 (2, 8). Under the experimental conditions of this study, NDMA was at such low concentrations that the pH was not measurably impacted. Proposed NDMA Reduction Mechanism. Mechanistic details for the transformation of NDMA on Pd were hypothesized in a previous study (20). Using Pd-based catalysts, Pd forms H* from dihydrogen which can then “spillover” to sites elsewhere on the catalyst surface. Hydrogen spillover has been verified for multiple catalysts, reactions and conditions (32, 33, 35–37) and is postulated here for the reaction of NDMA* adsorbed at In surfaces of the Pd-In catalyst. Following reductive cleavage of the N-N bond, DMA and NH4+ are formed. On the basis of the fact that nitric oxide is an intermediate in nitrite reduction on Pd catalysts (40), it is also possible that the nitrosyl group migrates to Pd before reduction to NH4+. The following simplified reaction scheme depicts NDMA reduction by Pd-In with hydrogen as the reductant: activation of dihydrogen to H* (eq 5); activation of NDMA to NDMA* by adsorption to the Pd or In surface (eqs 6a and 6b); and NDMA* reduction by H* (eqs 7a and 7b). Pd

H2 98 2H/ Pd

NDMA 98 NDMA/ In

NDMA 98 NDMA/ Pd

NDMA/ + 4H/ 98 DMA + NH4+ + OHIn

NDMA/ + 4H/ 98 DMA + NH4+ + OH-

(5)

(6a)

(6b)

(7a)

(7b)

Reaction 5 indicates H* formation from dihydrogen on Pd (32, 33), which is formed with an activation energy close to zero (34). Therefore, in the presence of sufficient Pd and excess dihydrogen, H* is not expected to be rate-limiting. Mass transfer of dihydrogen and NDMA from solution to the surface is relatively fast (20) compared to the overall transformation rate, which occurs on a time scale of hours. Hence, the rate-limiting steps are presumed to be activation of NDMA (reaction 6a or 6b) or the reductive transformation (reaction 7a or 7b). There is insufficient data to distinguish between these two hypotheses. Improvement of Pd-In catalysts should focus on increasing the number of In sites

while maintaining a sufficient H* supply by maintaining the Pd area sufficiently large. This study found that a bimetallic Pd-In loading of 5% Pd-1% In by weight optimized NDMA reduction in laboratory batch experiments starting with 5% Pd by weight. Similar Pd-In catalysts were found robust under simulated field conditions that may experience occasional sulfide poisoning and require aggressive oxidative regeneration with hypochlorite (31). The Pd-In catalyst evaluated here provides a significant improvement over previously studied Pd-Cu catalysts (20, 24), which would be destroyed at the first sulfide poisoning event. To increase the reduction rate for field application, both Pd and In loading should be increased in the ratio of approximately 5:1 as high as the physical structure can support mass transport. This would feasibly maintain excess H* while increasing the NDMA reduction rate per mass catalyst.

Acknowledgments The authors would like to thank Brian Chaplin at Charles Werth at the University of Illinois, Urbana–Champaign, for preparing the Pd-In catalysts and providing XPS characterization data. Partial support for this research came from the U.S. Environmental Protection Agency’s STAR Fellowship to M.G.D. (Grant FP916417); additional support included the National Science Foundation’s Science and Technology Center program (Grant CTS-0120978).

(12)

(13)

(14)

(15)

(16)

(17)

(18)

(19)

Supporting Information Available The kinetic rate expression in eq 4 was derived from the Langmuir–Hinshelwood model. This material is available free of charge via the Internet at http://pubs.acs.org.

(20)

(21)

Literature Cited (1) CA DHS. A brief history of NDMA findings in drinking water supplies and related activities; California Department of Health Services, 2006; available at http://www.dhs.ca.gov/ps/ddwem/ chemicals/ndma/history/default.htm. (2) MacDonald, A. Perchlorate and NDMA contamination in the Sacramento area. Presentation at the Fourth Symposium in the Series on Groundwater Contaminants: Perchlorate and NDMA in Groundwater: Occurrence, Analysis and Treatment, 2002. (3) Choi, J.; Duirk, S. E.; Valentine, R. L. Mechanistic studies of N-nitrosodimethylamine (NDMA) formation in chlorinated drinking water. J. Environ. Monit. 2002, 4 (2), 249. (4) Schreiber, I. M.; Mitch, W. A. The influence of the order of reagent addition on NDMA formation during chloramination. Environ. Sci. Technol. 2005, 39, 3811. (5) Pehlivanoglu-Mantas, E.; Hawley, E. L.; Deeb, R. A.; Sedlak, D. L. Formation of N-nitrosodimethylamine (NDMA) during chlorine disinfection of wastewater effluents prior to use in irrigation systems. Water Res. 2006, 40 (2), 341. (6) Mitch, W. A.; Sedlak, D. L. Formation of N-nitrosodimethylamine (NDMA) from dimethylamine during chlorination. Environ. Sci. Technol. 2002, 36, 588. (7) OCWD. Orange County Water District takes a proactive stance on newly regulated compound N-nitrosodimethylamine: OCWD recommends taking two drinking water wells out of service, Press Release; Orange County Water District, 2000. (8) 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. (9) Liang, S. Photolysis and advanced oxidation processes for NDMA removal from drinking water; Presentation at the Fourth Symposoum in the Series on Groundwater Contaminants: Perchlorate and NDMA in Groundwater: Occurrence, Analysis, and Treatment, 2002. (10) Plumlee, M. H.; Lopez-Mesas, M.; Heidlberger, A.; Ishida, K. P.; Reinhard, M. Occurrence and removal of N-nitrosodimethylamine (NDMA) in an advanced wastewater purification facility by LC-MS/MS. Water Res., submitted for publication. (11) Steinle-Darling, E.; Zedda, M.; Plumlee, M. H.; Ridgway, H.; Reinhard, M. Evaluating the impacts of membrane type, coating, fouling, chemical properties and water chemistry on reverse

(22)

(23)

(24)

(25)

(26)

(27)

(28)

(29)

(30)

(31)

osmosis rejection of seven nitrosoalklyamines, including NDMA. Water Res., submitted for publication. Stefan, M. I.; Bolton, J. R. UV direct photolysis of N-nitrosodimethylamine (NDMA): kinetic and product study. Helv. Chim. Acta 2002, 85 (5), 1416. Sharpless, C. M.; Linden, K. G. Experimental and model comparisons of low- and medium-pressure Hg lamps for the direct and H2O2 assisted UV photodegradation of N-nitrosodimethylamine in simulated drinking water. Environ, Sci. Technol. 2003, 37, 1933. 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, 2101. 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, 3489. 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. U.S. EPA. Integrated Risk Information System; Office of Research and Development (ORD), National Center for Environmental Assessment, U.S. Environmental Protection Agency: 2006; available at http://www.epa.gov/iris/subst/0045.htm. Muftikian, R.; Nebesny, K.; Fernando, Q.; Korte, N. X-ray photoelectron spectra of the palladium-iron bimetallic surface used for the rapid dechlorination of chlorinated organic environmental contaminants. Environmental Science & Technology 1996, 30 (12), 3593. Liu, Y.; Majetich, S. A.; Tilton, R. D.; Sholl, D. S.; Lowry, G. V. TCE dechlorination rates, pathways, and efficiency of nanoscale iron particles with different properties. Environmental Science & Technology 2005, 39 (5), 1338. Davie, M. G.; Reinhard, M.; Shapley, J. R. Metal-catalyzed reduction of N-nitrosodimethylamine with hydrogen in water. Environmental Science & Technology 2006, 40 (23), 7329. Pintar, A.; Batista, J.; Levec, J. Potential of mono- and bimetallic catalysts for liquid-phase hydrogenation of aqueous nitrite solutions. Water Science & Technology 1997, 37 (8), 177. Pintar, A.; Bercic, G.; Levec, J. Catalytic liquid-phase nitrite reduction: Kinetics and catalyst deactivation. AIChE J. 1998, 44 (10), 2280. Pintar, A.; Batista, J.; Muševicˇ, I. Palladium-copper and palladium-tin catalysts in the liquid phase nitrate hydrogenation in a batch-recycle reactor. Applied Catalysis B: Environmental 2004, 52 (1), 49. 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. Environmental Science & Technology 2006, 40 (9), 3075. Lowry, G. V.; Reinhard, M. Pd-catalyzed TCE dechlorination in groundwater: solute effects, biological control, and oxidative catalyst regeneration. Environmental Science & Technology 2000, 34 (15), 3217. McNab, W. W.; Ruiz, R.; Reinhard, M. In-situ destruction of chlorinated hydrocarbons in groundwater using catalytic reductive dehalogenation in a reactive well: Testing and operational experiences. Environmental Science & Technology 2000, 34 (1), 149. Munakata, N.; Reinhard, M. Palladium-catalyzed hydrodehalogenation for groundwater remediation: sulfide deactivation model, regenerant comparison and chemical surface changes. Applied Catalysis B: Environmental 2007, 75, 1. Schüth, C.; Kummer, N. A.; Weidenthaler, C.; Schad, H. Field application of a tailored catalyst for hydrodechlorinating chlorinated hydrocarbon contaminants in groundwater. Applied Catalysis B: Environmental 2004, 52 (3), 197. Lowry, G. V.; Reinhard, M. Pd-catalyzed TCE dechlorination in water: effect of [H2](aq) and H2-utilizing competitive solutes on the TCE dechlorination rate and product distribution. Environmental Science & Technology 2001, 35 (4), 696. Munakata, N.; Roberts, P. V.; Reinhard, M.; McNab, W. W. Catalytic dechlorination of halogenated hydrocarbon compounds using supported palladium: A preliminary assessment of matrix effects. IAHS Publication (International Association of Hydrological Sciences) 1998, (250), 491. Chaplin, B. P.; Shapley, J. R.; Werth, C. J. Regeneration of sulfurfouled bimetallic Pd-based catalysts. Environmental Science & Technology 2007, 41 (15), 5491.

VOL. 42, NO. 8, 2008 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3045

(32) Ponec, V.; Bond, G. C. Catalysis by metals and alloys; Elsevier: Amsterdam: 1995. (33) Boudart, M.; Wang, H. S. Solubility of hydrogen in small particles of palladium. J. Catal. 1975, 39 (1), 44. (34) Pintar, A.; Batista, J.; Levec, J.; Kajiuchi, T. Kinetics of the catalytic liquid-phase hydrogenation of aqueous nitrate solutions. Applied Catalysis B: Environmental 1996, 11 (1), 81. (35) Khoobiar, S. Particle to particle migration of hydrogen atoms on platinum-alumina catalysts from particle to neighboring particles. J. Phys. Chem. 1964, 68 (2), 411. (36) Boudart, M.; Djéga-Mariadassou, G. Kinetics of heterogeneous catalytic reactions; Princeton University Press: 1984. (37) Nutt, M. O.; Hughes, J. B.; Wong, M. S. Designing Pd-on-Au bimetallic nanoparticle catalysts for trichloroethene hydrodechlorination. Environmental Science & Technology 2005, 39 (5), 1346.

3046

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 42, NO. 8, 2008

(38) Pintar, A.; Batista, J.; Areon, I.; Kodre, A. Characterization of γ-Al2O3-supported Pd-Cu bimetallic catalysts by EXAFS, AES and kinetic measurements; In Delmon, B.; Jacobs, P. A.; Maggi, R.; Martens, J. A.; Grange, P.; Poncelet, G. Eds.; Preparation of Catalysts; Elsevier Science: Amsterdam: 1998. (39) Mitch, W. A.; Sedlak, D. L. Characterization and fate of N-nitrosodimethylamine precursors in municipal wastewater treatment plants. Environmental Science & Technology 2004, 38 (5), 1445. (40) Prusse, U.; Vorlop, K. D. Supported bimetallic palladium catalysts for water-phase nitrate reduction. Journal of Molecular Catalysis A: Chemical 2001, 173 (1), 313.

ES7023115