Rate and Extent of Aqueous Perchlorate Removal by Iron Surfaces

Department of Civil and Environmental Engineering, University of California, Davis, ...... to determine if the removal rates can be accelerated for us...
0 downloads 0 Views 286KB Size
Environ. Sci. Technol. 2003, 37, 3189-3198

Rate and Extent of Aqueous Perchlorate Removal by Iron Surfaces ANGELA M. MOORE, CORINNE H. DE LEON,† AND THOMAS M. YOUNG* Department of Civil and Environmental Engineering, University of California, Davis, California 95616

The rate and extent of perchlorate reduction on several types of iron metal was studied in batch and column reactors. Mass balances performed on the batch experiments indicate that perchlorate is initially sorbed to the iron surface, followed by a reduction to chloride. Perchlorate removal was proportional to the iron dosage in the batch reactors, with up to 66% removal in 336 h in the highest dosage system (1.25 g mL-1). Surface-normalized reaction rates among three commercial sources of iron filings were similar for acid-washed samples. The most significant perchlorate removal occurred in solutions with slightly acidic or near-neutral initial pH values. Surface mediation of the reaction is supported by the absence of reduction in batch experiments with soluble Fe2+ and also by the similarity in specific reaction rate constants (kSA) determined for three different iron types. Elevated soluble chloride concentrations significantly inhibited perchlorate reduction, and lower removal rates were observed for iron samples with higher amounts of background chloride contamination. Perchlorate reduction was not observed on electrolytic sources of iron or on a mixed-phase oxide (Fe3O4), suggesting that the reactive iron phase is neither pure zerovalent iron nor the mixed oxide alone. A mixed valence iron hydr(oxide) coating or a sorbed Fe2+ surface complex represent the most likely sites for the reaction. The observed reaction rates are too slow for immediate use in remediation system design, but the findings may provide a basis for future development of cost-effective abiotic perchlorate removal techniques.

Introduction Upon the development of a sensitive detection method using ion chromatography, perchlorate contamination has been discovered in groundwaters and surface waters in 13 U.S. states (1). Contamination has been detected in 11% of sampled public water systems in California and in 6.6% of the state’s drinking water sources (2). Contamination levels of up to 400 µg/L have been detected in wells (now offline) used to supply drinking water to residents in Rancho Cordova, CA, and perchlorate levels ranging from 5 µg/L to 1.68 mg/L have been detected in the Colorado River and Lake Mead, which supply irrigation and drinking water to consumers in Arizona, California, and Nevada (3, 4). * Corresponding author e-mail: [email protected]; phone: (530)754-9399; fax: (530)752-7872. † Current address: Psomas, 2295 Gateway Oaks Drive, Suite 250, Sacramento, CA 95833. 10.1021/es026007t CCC: $25.00 Published on Web 06/17/2003

 2003 American Chemical Society

Perchlorate is a health concern because when ingested it can block the uptake of iodine in the thyroid gland, affecting the production of thyroid hormones and possibly causing mental retardation in fetuses and infants (5). The U.S. Environmental Protection Agency (U.S. EPA) formally added perchlorate to the drinking water contaminant candidate list in 1998 and is currently assessing the health and ecological risks resulting from perchlorate exposure. Following the initial detection of contaminated water supplies in northern California, in 1997 the California Department of Health Services (DHS) established an action level of 18 µg/L for perchlorate in drinking water. In January 2002, the DHS lowered the perchlorate action limit to 4 µg/L following the release of the U.S. EPA 2002 draft of the health risk assessment (6). Ammonium perchlorate is commercially manufactured in large quantities for use as a solid propellant for rockets, missiles, and pyrotechnic devices. Perchlorate salts are also used in the manufacture of automotive airbag inflators, as a mordant for fabrics and dyes, and in the production of paints and enamels. Most environmental perchlorate contamination appears to be a result of discharge from rocket fuel manufacturing or from the demilitarization of weaponry (3). The only documented source of naturally occurring perchlorate is found in a few nitrate-rich mineral deposits in Chile (5). Perchlorate salts are highly soluble in water. The perchlorate anion is mobile in the environment and is not significantly sorbed by soil materials. Perchlorate reduction in biological reactions is believed to occur via the sequential reduction ClO4- f ClO3- f ClO2- f Cl- (7, 8). Several microorganisms have been identified that can utilize perchlorate as an electron acceptor, and the biodegradation of perchlorate is an area of active research (9-12). Although biological reduction appears to have the most potential for large-scale perchlorate treatment, development of a chemical remediation technology would be a useful alternative in carbon-limited environments and for the simultaneous treatment of multiple pollutants, such as perchlorate contamination in conjunction with oxidized forms of heavy metals or chlorinated solvents. Dissolved perchlorate is considered to be highly stable and has historically been used as a supporting electrolyte in chemical and electrochemical studies. Thermodynamically, the perchlorate ion is predicted to be unstable and should spontaneously degrade at ambient conditions through the release of molecular oxygen; kinetic barriers to perchlorate reduction are high and lend great stability to the ion (13). Thermodynamics also predicts that perchlorate should readily oxidize aqueous metal ions, but current research illustrates that these reactions do not occur under ambient conditions. Perchlorate reduction has been observed with certain transition metal complexes and metal chelates (3, 13), but the reactions are generally sluggish, and the metals are too expensive to consider in the development of remediation strategies. Perchlorate may be reduced to chlorate or to chloride depending upon the reductant utilized (3). Recent work in electrochemistry has demonstrated the reduction of perchlorate at various metal and metallic electrode surfaces, in contrast to the previous assumption that perchlorate behaved as an inert electrolyte. Perchlorate reduction in acidic media has been observed on aluminum, copper, nickel, and iron metal and on electrodes constructed of aluminum, iridium, platinum, rhenium, rhodium, rubidium, silver, technetium, titanium, and tungsten carbide (14-29). VOL. 37, NO. 14, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3189

TABLE 1. Summary of Iron Type, Preparation, and Specific Surface Areas (A) iron type

source

nominal size

preparation (this study)

A (m2 g-1)

range of reported values for A (m2 g-1)a

Fisher Fisher AWc Electrolytic Electrolytic AW Connelly Connelly AW Peerless Peerless AW

Fisher Scientific Fisher Scientific Fisher Scientific Fisher Scientific Connelly GPM Connelly GPM Peerless Metals and Abrasives Peerless Metals and Abrasives

-40 mesh filings, degreased -40 mesh filings, degreased -100 mesh powder -100 mesh powder -20/+50 mesh filings -20/+50 mesh filings -20/+50 mesh filings -20/+50 mesh filings

untreated acid washed untreated acid washed untreated acid washed untreated acid washed

5.69b 2.50d 0.08d 0.43d 1.54b 1.69b 1.09b 1.24b

0.0005-2.43 0.7-5.76 0.05-0.611 0.3-4.17 1.68-1.9 1.58-4.36 0.08-2.53 0.87-1.54

a Data compiled from various studies as reported by Alowitz and Scherer (33). Literature values for iron filings pretreated by acid washing or sonication were compared with acid-washed iron in this study. b Analyses performed by Micromeritics Instrument Corporation (Norcross, GA). c AW indicates acid-washed sample. d Analyses performed by Dr. Eugene LeBoeuf, Vanderbilt University.

Permeable reactive barriers (PRBs) can be an effective means for removing certain heavy metals and organic compounds in contaminated aquifers. Use of barriers filled with reduced metals is an emerging technology, and removal mechanisms include reduction to a nontoxic species, adsorption, precipitation, and biologically mediated transformations (30). Reduced iron is the most common PRB medium used for the dechlorination of halogenated hydrocarbons as well as for the reduction and precipitation of heavy metals and radionuclides (31). Iron media have also been investigated for remediation of nitrate in contaminated groundwater (32, 33). Possible mechanisms considered for contaminant reduction on iron media include (i) direct reduction at the metal surface, possibly via pitting in the oxide surface, (ii) reduction by dissolved Fe2+, (iii) reduction by elemental hydrogen, and (iv) reduction by sorbed Fe2+ species, (32, 34-39). A limited amount of research has been performed on the potential reduction of perchlorate by iron surfaces. An evaluation of the use of zerovalent metals in PRBs demonstrated no reduction of perchlorate upon contact with iron filings (40). Gurol and Kim (41) examined the feasibility of perchlorate reduction on metallic iron surfaces in potable water sources; perchlorate reduction was demonstrated, and it was shown that UV light accelerated the reaction rate. Batch experiments with iron provided evidence of perchlorate reduction to chloride in strongly acidic media, although the rate of transformation was not reported (29). Electrochemical studies of corrosion and pitting of iron electrodes in perchlorate media have also been published. Prinz and Strehblow (42) concluded that the perchlorate ion decomposes on the surface of iron electrodes during the pitting process, and the reduction to chloride may be a key factor in the corrosion process. However, in another corrosion study of the perchlorate/iron system, the authors attributed pitting to the nonuniform dissolution of iron rather than to the perchlorate reduction reaction (43). Perchlorate reduction on iron surfaces has been observed but to date has not been systematically evaluated under environmental conditions relevant to contaminated groundwater. The mechanism for the removal of aqueous perchlorate by solid iron phases has not been characterized and is complicated by the potential importance of iron oxide surfaces. This paper reports the results of experiments designed to (i) determine the rate and extent of aqueous perchlorate removal by iron surfaces, (ii) delineate conditions under which the reaction is favored or retarded, and (iii) provide initial information about reaction mechanisms to guide future research.

Experimental Methods Chemicals. Deionized Milli-Q (Millipore Corp.) water was used in all experiments. Argon gas (99.995%) and nitrogen gas (99.997%) were obtained from Puritan-Bennett. Am3190

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 14, 2003

monium perchlorate (99.999%), sodium chlorate (>99%), and sodium chloride (99.999%) were obtained from Sigma Aldrich (St. Louis, MO). Sodium chlorite (80% unstabilized, technical grade) was obtained from Acros Chemical (Pittsburgh, PA). Certified ACS grade iron(II) perchlorate, iron(II) sulfate, and iron(II) chloride were purchased from Fisher Scientific (Pittsburgh, PA). All acids were trace metal grade and were purchased from Fisher Scientific. Iron Metal Preparation and Characterization. Iron types used in this study included Fisher iron filings (-40 mesh, Fisher Scientific), electrolytic iron (certified and uncertified powder, -100 mesh, Fisher Scientific), Fe3O4 (laboratory grade, Fisher Scientific), iron aggregate filings from ConnellyGPM Inc. (Chicago, IL), and iron aggregate filings from Peerless Metal Powders and Abrasive Co., Inc. (Detroit, MI). The Connelly-GPM and Peerless samples were sieved, reserving the -20/+50 mesh particles for use in this study. The Fisher 40 mesh filings, Peerless, and Connelly samples were tested as received and also tested after acid washing. Iron types, preparative procedures, and abbreviations used in this study are summarized in Table 1. Initially, the Fisher 40 mesh iron filings were washed using deaerated 1.0 N hydrochloric acid (HCl) and rinsed 15 times with deaerated Milli-Q water. This procedure produced soluble chloride levels greater than 7 mM in the batch studies and prevented quantification of chloride potentially produced by perchlorate reduction. Hydrochloric acid was subsequently replaced by deaerated 1.0 N hydrofluoric acid (HF) in the acid-washing procedure. Iron samples (500 g) were rinsed with 500 mL of HF for 30 min, followed by 5 rapid rinses of 250 mL of deaerated Milli-Q water. The iron was packed into a column and further rinsed with deaerated Milli-Q water using a peristaltic pump; rinsing continued until the effluent chloride levels were below 500 µg/L and effluent fluoride levels were sufficiently low so that they did not interfere with subsequent chloride analyses. The iron was dried at 100 °C under an N2 atmosphere for a minimum of 12 h and stored in amber bottles under N2 until use. Acid-washed iron samples used in this study are identified by the supplier name followed by the suffix AW. Upon receipt, the iron filings varied in the level of visible surface oxidation. The Peerless filings were a uniform dark gray to black color while the Connelly filings and some of the Fisher filings had a more rusty appearance. Presumably, the orange deposits visible on the surface of the unwashed Fisher and Connelly iron filings were iron(III) oxides, and the black layers evident on the Peerless, Connelly, and Fisher filings were a more reduced iron(II) or mixed iron(II)/iron(III) oxide. The orange deposits were no longer visible following the acid-washing procedure, and all surfaces were a similar dark gray to black color. The electrolytic powder (certified and uncertified) was a shiny metallic silver as received, but the color dulled to a steel gray after extensive rinsing with deaerated Milli-Q during the acid-washing procedure.

Specific surface area determinations for the treated and untreated iron samples are listed in Table 1. Specific surface area (A, m2 g-1) was determined from a Brunauer-EmmettTeller (BET) krypton gas adsorption isotherm, measured using a Micromeritics ASAP 2010 (Fisher AW, electrolytic powder) or a Micromeritics ASAP 2405 (remaining samples). With the exception of the unwashed Fisher filings, the specific surface areas determined for the iron types in this study were similar to values reported in the literature as shown in Table 1, from data summarized by Alowitz and Scherer (33). The lot of unwashed Fisher filings sent for analysis was more highly oxidized than other lots received, which may account for the anomalous value. As observed by others, the specific surface area of the iron generally increased after acid washing, presumably because removal of surface oxidation exposed previously inaccessible pore structure. The Fisher AW iron filings and the electrolytic iron were analyzed by X-ray diffraction (XRD) to determine the crystalline mineralogy of the iron surfaces. XRD analysis was performed with a Scintag XDS-2000 diffractometer utilizing Cu KR radiation. The surface mineralogy of the Fisher AW samples corresponded to magnetite, a mixed iron oxide. No crystalline oxide phases were detected on the electrolytic iron powder. Batch Reactors. Reactors consisting of 25-mL amber glass vials with Teflon-lined septa and Teflon-taped threaded caps were used to conduct batch perchlorate experiments. All solutions used in the batch and column experiments were deaerated by vigorous sparging with nitrogen or argon gas. All solutions were introduced in a glovebag under an argon atmosphere, and each vial was filled to capacity, leaving no headspace. The reactors were placed into a custom-built enclosed rotational tumbler operated at 20 rpm, and vials were removed for analysis at specified times. Each iron experiment was performed in triplicate and included a set of control vials containing the introduced solution but no iron and a set of blank vials containing only the iron and deaerated Milli-Q water. Most of the batch experiments were performed with a nominal initial concentration of 0.1 mM ammonium perchlorate, with the exception of two experiments performed with Fisher AW filings with initial concentrations of 0.001 mM and 1 M ammonium perchlorate. The initial pH of the ammonium perchlorate solutions was adjusted with 0.1 N perchloric acid, 1 N HNO3, or 0.1 N NaOH. Columns. Two columns were packed with Fisher AW filings for a long-term column study. The borosilicate columns (Kontes) were 24 cm long with a 2.5 cm i.d. with polyethylene frits and caps at the ends. The two columns were filled with 299 and 305 g of iron, respectively; purged with N2 gas; and quickly filled with deaerated Milli-Q water using a kdScientific model 200 syringe pump and 100-mL SGE glass syringes with Teflon plungers. After the initial rapid filling, the flow rate from the syringe pump was maintained at 0.75 mL/h. Tracer tests were performed using a 1.15 µM solution of LiBr during the first 2 weeks of operation. After flushing for 1 additional week with deaerated Milli-Q water, a 0.1 mM ammonium perchlorate solution was introduced as the influent. The influent solution was continuously sparged with argon, and solution pH was not controlled. The columns were covered with aluminum foil to minimize possible photochemical transformations during the experiment. Sampling and Analytical Methods. Perchlorate, chlorate, chlorite, and chloride were analyzed on a Dionex 600 ion chromatograph with an EG40 eluent generator. A 4-mm Dionex AS-16 anion-exchange analytical column and guard column were used for anion separation, with a conductance detector for analysis. The potassium hydroxide eluent flow was 1.25 mL min-1 with a concentration gradient of 10 mM for 5 min increasing to 55 mM at a rate of 9 mM min-1, held constant at 55 mM for 6 min, and reduced to 10 mM for 2 min to allow for equilibration before the next sample

FIGURE 1. Aqueous perchlorate removal in the presence of acidwashed iron filings. Initial concentration 0.1 mM ClO4-, initial solution pH ∼7.0. Experiments were performed in triplicate; error bars indicate one standard deviation about the mean. injection. For all experiments with a nominal initial perchlorate concentration of 0.1 mM, a 25-µL injection loop was used to analyze the samples; in one experiment with a lower initial concentration, a 1000-µL injection loop was used. Method detection limits determined for this system using the 25-µL loop were 0.21 µM for perchlorate, 0.49 µM for chlorate, 1.45 µM for chlorite, and 0.37 µM for chloride. Samples for anion and oxyanion analyses were collected from the batch reactors with 12-mL disposable syringes. Effluent samples from the columns were collected in glass vials over a period of several hours to obtain enough liquid to analyze. Both types of samples were passed through syringe filters to remove particulates (99.995 98.16

9.03 6.19 2.78 81.57 >99.995 >99.995

t ) 100 d Ga ) 1000 m2 L-1 Ga ) 3500 m2 L-1 59.38 42.20 21.50 >99.995 >99.995 >99.995

95.73 85.32 57.14 >99.995 >99.995 >99.995

The kSA values listed for chlorinated hydrocarbons are from Johnson et al. (46).

FIGURE 11. Perchlorate removal in columns packed with Fisher AW filings, effluent concentration/influent concentration (Ceff/Cin). Influent solution 0.1 mM ClO4-1, flow rate 0.75 mL h-1. Column 1: retention time 3.65 d, linear flow velocity 0.28 cm h-1, porosity 0.54. Column 2: retention time 3.8 d, linear flow velocity 0.26 cm h-1, porosity 0.59. Transport parameters were determined from tracer data using the CXTFIT model (60). used in other bench and field studies of iron media (58, 59); retention of the bromide in this study indicates anion sorption is occurring on the surface of the iron filings. As shown in Figure 11, the extent of perchlorate removal was about 90% upon introduction of perchlorate to the columns, but performance began to degrade almost immediately. After 50 d, the extent of perchlorate removal stabilized and subsequently fluctuated near 40% removal of perchlorate in both of the columns. The specific reaction rate constants (kSA) for perchlorate removal in the columns were calculated using parameters derived from the tracer test and were based upon the 40% steady-state removal observed after 50 d of operation. The kSA values for columns 1 and 2 were 4.95 E-07 and 5.35 E-07 L h-1 m-2, respectively, and were similar to values reported for the batch experiments in Table 2. Surficial Oxides. The nature of oxides and other mineral phases on the surface of the iron filings is obviously an important factor to consider in identifying the reactive surface. Amorphous iron (hydr)oxides, goethite, magnetite, akageneite, lepidocrocite, and green rusts have been identified on iron filings used in field and laboratory studies of PRB materials (31). An XRD analysis of the Fisher AW filings used in this study indicated that magnetite was the dominant phase on the filing surface; however, batch studies with a mixed oxide powder (Fe3O4) did not demonstrate any perchlorate removal (data not shown). This result does not rule out magnetite as the reactive surface, however, because the reactivity of a conducting oxide film on a reduced metal substrate probably differs from that of a pure iron oxide phase. 3196

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 14, 2003

Significant reduction of nitrobenzene by magnetite was found to occur only when sorbed Fe2+ concentrations were increased (35); however, this possible effect was not evaluated in the present study. One or more iron phases on the surface of the filings may be responsible for the perchlorate sorption and subsequent reduction, but these films were probably too thin to detect using XRD. Fisher AW iron filings in a 0.1 mM perchlorate batch study were sonicated after 48 h of reaction time to determine if the surface reactivity could be regenerated by removing surficial deposits, but removal rates were no different from the unsonicated control for subsequent samples (data not shown). It is possible that the sonication procedure did not successfully remove surficial oxides. Implications for Field Application. Specific reaction rate constants calculated for the batch reactors can be used to estimate the performance of an iron PRB for perchlorate remediation. Typical iron metal barriers are 50-150 cm thick (61); for the following estimate, a midrange value of 1 m was selected. Groundwater flow velocities of 1 and 30 cm d-1 were both evaluated, corresponding to retentions times of 100 and 3.3 d, respectively. Reasonable field values for Fa in PRBs range from 1000 (62) to 3500 m2 L-1 (47). To evaluate the performance of the proposed PRB, eq 2 was rearranged:

Ce ) e-kSAFat Ci

(3)

where Ci is the influent perchlorate concentration (mM), Ce is the effluent perchlorate concentration (mM), and t is the residence time (h). Perchlorate removal rates are presented in Table 3, and calculations for selected chlorinated hydrocarbons have been included for comparison. Longer retention times would be required to remove perchlorate from contaminated groundwater using ironbased PRBs, as compared with percent removal for the listed organic halocarbons. A retention time of 100 d is not a reasonable design for compounds with rapid degradation rates but could be feasible in an aquifer with low groundwater velocities. The column studies demonstrated a higher percent removal than the above calculations as a result of a higher Fa value (approximately 1E+04 in the columns), likely due to the smaller and more uniform particle size of the Fisher AW filings as compared to field material. The calculations show some limited promise for perchlorate remediation by iron media, but anion concentrations present in groundwater will likely decrease perchlorate removal efficiency; the production of chloride from the reduction of chlorinated hydrocarbons in a mixed contaminant plume may also be a concern. Further study is required to determine if the removal rates can be accelerated for use in the field. Possible Reaction Mechanisms. In biological and chemical studies, it has been shown that, in the sequential reduction

of perchlorate to chloride, the rate-determining step is the initial reduction of perchlorate to chlorate (13, 63, 64). An iron batch experiment performed with 10-g Fisher AW filings, and 0.06 mM initial sodium chlorate exhibited greater than 98% transformation to chloride within 3 h (data not shown), a much faster rate of removal than observed for perchlorate. The lack of detectable intermediate product accumulation in the perchlorate batch experiments also suggests that the reduction to chlorate is the rate-determining step in the iron system. An alternative explanation for the failure to detect anticipated intermediates is that reduction to chloride occurs entirely in the adsorbed phase with no release of products to the bulk solution, but based on the observed chlorate removal rate, the initial step is still likely rate determining. It is important to develop an understanding of the perchlorate removal mechanism at the reactive surface in order to investigate strategies for improving the removal efficiency in engineered systems. Although the present work does not definitively establish the mechanism responsible for perchlorate reduction, the results implicate sorption and subsequent reduction at iron hydr(oxide) surface layers. The various lines of evidence supporting this conclusion are outlined below. It has been shown that certain contaminants can be reduced indirectly by atomic hydrogen generated from the reduction of water by the iron metal (32, 65, 66). This reaction may require the presence of a catalyst (32), and contaminant reactions with adsorbed hydrogen may involve the formation of hydride complexes on the iron surface (65). In electrochemical studies, perchlorate reduction on platinum, ruthenium, iridium, and tungsten carbide electrodes has been attributed to reaction with adsorbed hydrogen atoms (16, 21, 42, 55). This type of reaction is fast on these metals because of their low hydrogen overpotentials (53) but slow on metals such as iron that have high hydrogen overpotentials. Although not definitively ruled out by this study, indirect perchlorate reduction by atomic hydrogen is not supported by experimental work at low pH values in this study or by the results of other researchers (29, 42, 45). The extent of perchlorate removal was strongly dependent on the amount of iron surface area in the batch reactors, and similar surface area-normalized perchlorate removal rates were calculated for the batch and long-term column studies. These observations and calculations thus support the hypothesis that perchlorate removal occurs at the iron surface. Suppression of perchlorate removal by added chloride indicates that pitting of the oxide surface and subsequent exposure of bare iron is not responsible for perchlorate removal. In addition, the observation that differences in iron reactivity among the various sources could be rationalized by considering the amount of exchangeable chloride suggest that the reactive iron phase is not zerovalent but is instead a combination of the elemental metal coated by a mixed valence iron hydr(oxide) phase. The absence of detectable perchlorate removal or reduction by the electrolytic iron powders provides further evidence for the involvement of iron hydr(oxide) coatings. Understanding the molecular interactions between perchlorate and the iron metal or oxide surface is a key factor in determining the reaction mechanism. The presence of oxidized surface films or sites has been suggested as a requirement for the reduction of perchlorate on aluminum, rhenium, rhodium, tin, and titanium electrodes (14, 20, 23, 24, 49). The perchlorate ion cannot easily accept an additional electron because it has no low-lying unfilled electronic orbitals. Therefore, the transfer of an oxygen atom is required for a reduction reaction (67). In studies with titanium electrodes, oxygen transfer only occurred when the oxygen was stabilized by another species through the existence of an “oxygen vacancy” (49). The surface of the titanium

electrodes was covered with a monolayer of adsorbed titanium (hydr)oxide species, and adsorbed Ti(II) was identified as the species likely responsible for stabilization of transferred oxygen; a similar mechanism of oxygen stabilization and transfer has been proposed for rhodium electrodes (24). In the case of contaminant reduction at iron surfaces, the iron oxide can behave as a coordinating surface, and adsorbed Fe2+ on the oxide surface may serve as the reductant (34). Under conditions where the initial concentration of aqueous Fe2+ is lower than the concentration of surface sites, adsorption of Fe2+ onto ferric (hydr)oxide surfaces takes place as the formation of the surface complexes ≡FeOFe+ and ≡FeOFeOH0 (68). Sorption of Fe2+ on magnetite primarily occurs between pH 6 and pH 8 (35, 69), coinciding with initial pH values in the more reactive batch experiments in this study. This mechanism may be responsible for the perchlorate sorption and subsequent reduction observed on the iron filings in this study, and the importance of sorbed ferrous iron is the focus of ongoing research. This work represents the first systematic effort to quantify the rate and extent of perchlorate removal by iron surfaces. The capacity of iron filings to remove aqueous perchlorate has been shown under various conditions. Rapid decline of aqueous perchlorate concentrations in the early stages does not coincide with a rapid increase in chloride concentrations, indicating that sorption may be responsible for this high initial rate of removal. Chloride concentrations do increase steadily throughout the experiment and are significantly greater than chloride concentrations in the blank vials, indicating that perchlorate reduction does occur following sorption. Removal of perchlorate from solution proceeds most rapidly under neutral pH conditions similar to those found in many groundwater systems, but the slow overall rate of reduction and inhibition by chloride prevents immediate application in subsurface PRBs. Further research to elucidate the perchlorate reduction mechanism may produce a feasible remediation technology designed to exploit the initial rapid rate of perchlorate sorption and subsequent reduction on iron surfaces or might potentially be used in conjunction with in situ bioremediation strategies.

Acknowledgments The authors thank Dr. Peter Green and Dr. W. Ron Fawcett for their advice and assistance and also David Crawford and Nicki Giese for their efforts in the lab. We also thank Dr. Eugene LeBoeuf of Vanderbilt University for performing the BET analyses, and the donations of iron filings from Peerless Metals and Abrasives and Connelly GPM were greatly appreciated. We also thank the anonymous reviewers, whose comments and insights significantly improved the manuscript. This project was supported by Grant 5 P42 ES0469916 from the National Institute of Environmental Health Sciences, NIH, with funding provided by the U.S. EPA. The contents are solely the responsibility of the authors and do not necessarily represent the official views of the NIEHS, NIH, or EPA.

Literature Cited (1) Jackson, P. E.; Laikhtman, M.; Rohrer, J. S. J. Chromatogr. A 1999, 850, 131-135. (2) CA Department of Health Services. http://www.dhs.ca.gov/ps/ ddwem/chemicals/perchl/perchlindex.htm (accessed 2/19/03). (3) Urbansky, E. T. Biorem. J. 1998, 2, 81-95. (4) U.S. EPA. Perchlorate Environmental Contamination: Toxicological Review and Risk Characterization, External Review Draft; NCEA-1-0503; National Center for Environmental Assessment: January 16, 2003. (5) Renner, R. J. Environ. Monit. 1999, 1, 37-38. (6) CA Department of Health Services. http://www.dhs.ca.gov/ps/ ddwem/chemicals/perchl/actionlevel.htm (accessed 2/19/03). VOL. 37, NO. 14, 2003 / ENVIRONMENTAL SCIENCE & TECHNOLOGY

9

3197

(7) Rikken, G. B.; Kroon, A. G. M.; van Ginkel, C. G. Appl. Microbiol. Biotechnol. 1996, 45, 420-426. (8) Coates, J. D.; Bruce, R. A.; Haddock, J. D. Nature 1998, 396, 730. (9) Logan, B. Biorem. J. 1998, 2, 69-79. (10) Coates, J. D.; Michaelidou, U.; O’Connor, S. M.; Bruce, R. A.; Achenbach, L. A. In Perchlorate in the Environment; Urbansky, E. T., Ed.; Kluwer Academic/Plenum Publishers: New York, 2000. (11) Giblin, T. L.; Herman, D. C.; Frankenberger, W. T. J. In Perchlorate in the Environment; Urbansky, E. T., Ed.; Kluwer Academic/ Plenum Publishers: New York, 2000. (12) Tipton, D. K.; Rolston, D. E.; Scow, K. M. J. Environ. Qual. 2003, 32, 40-46. (13) Espenson, J. H. In Perchlorate in the Environment; Urbansky, E. T., Ed.; Kluwer Academic/Plenum Publishers: New York, 2000. (14) Almeida, C. M. V. B.; Giannetti, B. F.; Rabockai, T. J. Electroanal. Chem. 1997, 422, 185-189. (15) Ahmadi, A.; Evans, R. W.; Attard, G. J. Electroanal. Chem. 1993, 350, 279-295. (16) Sanchez Cruz, M.; Gonza´lez-Tejera, M. J.; Villamana, M. C. Electrochim. Acta 1985, 30, 252. (17) Vasina, S. Y.; Petrii, O. A. Soviet Electrochem. 1970, 6, 231. (18) Clavilier, J.; Wasberg, M.; Petit, M.; Klein, L. H. J. Electroanal. Chem. 1994, 374, 123. (19) Hora´nyi, G.; Bakos, I. J. Appl. Electrochem. 1992, 23, 547-552. (20) Painot, J.; Augustynski, J. Electrochim. Acta 1975, 20, 747. (21) Colom, F.; Gonza´lez-Tejera, M. J. J. Electroanal. Chem. 1985, 190, 243-255. (22) Rhee, C. K.; Wasberg, M.; Hora´nyi, G.; Wieckowski, A. J. Electroanal. Chem. 1990, 291, 281-287. (23) Hora´nyi, G.; Bakos, I. J. Electroanal. Chem. 1994, 378, 143-148. (24) Wasberg, M.; Hora´nyi, G. J. Electroanal. Chem. 1995, 385, 6370. (25) Wasberg, M.; Hora´nyi, G. J. Electroanal. Chem. 1995, 381, 151158. (26) Bakos, I.; Hora´nyi, G.; Szabo´, S.; Rizmayer, E. M. J. Electroanal. Chem. 1993, 359, 241-252. (27) Ujva´ri, M.; La´ng, G.; Hora´nyi, G. J. Appl. Electrochem. 2001, 31, 1171-1173. (28) La´ng, G. G.; Hora´nyi, G. Electroanal. Chem. In press [DOI: 10.1016/S0022-0728(02)01302-5]. (29) Reference deleted during revision. (30) Blowes, D. W.; Ptacek, C. J.; Benner, S. G.; McRae, C. W. T.; Bennett, T. A.; Puls, R. W. J. Contam. Hydrol. 2000, 45, 123-137. (31) Roh, Y.; Lee, S. Y.; Elless, M. P. Environ. Geol. 2000, 40, 184-194. (32) Matheson, L. J.; Tratnyek, P. G. Environ. Sci. Technol. 1994, 28, 2045-2053. (33) Alowitz, M. J.; Scherer, M. M. Environ. Sci. Technol. 2002, 36, 299-306. (34) Scherer, M. M.; Balko, B.; Gallagher, D. A.; Tratnyek, P. G. In Mineral-Water Interfacial Reactions: Kinetics and Mechanisms; Sparks, D., Grundl, T., Eds.; American Chemical Society: Washington, DC, 1998. (35) Klausen, J.; Trober, S. P.; Haderlein, S. B.; Schwarzenback, R. P. Environ. Sci. Technol. 1995, 29, 2396-2404. (36) Cui, D.; Eriksen, T. E. Environ. Sci. Technol. 1996, 30, 22592262. (37) Ernstsen, V. Clays Clay Miner. 1996, 44, 599-608. (38) Charlet, L.; Liger, E.; Gerasimo, P. J. Environ. Eng. 1998, 124, 25-30. (39) Charlet, L.; Bosbach, D.; Peretyashko, T. Chem. Geol. 2002, 190, 303-319. (40) Yarmoff, J. A.; Amrhein, C. Fundamental Studies of The Removal of Contaminants from Ground and Waste Waters Via Reduction By Zero-Valent Metals; DOE/ER/14707; U.S. Department of Energy: Washington, DC, 2002. (41) Gurol, M. D.; Kim, K. In Perchlorate in the Environment; Urbansky, E. T., Ed.; Kluwer Academic/Plenum Publishers: New York, 2000. (42) Prinz, H.; Strehblow, H.-H. Corros. Sci. 1998, 40, 1671-1683. (43) Ma, H.; Li, G.; Chen, S.; Zhao, S.; Cheng, X. Corros. Sci. 2002, 44, 1177-1191.

3198

9

ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 37, NO. 14, 2003

(44) Eaton, A. D.; Clesceri, L. S.; Greenberg, A. E. Standard Methods for the Examination of Water and Wastewater; American Public Health Association, American Water Works Association, Water Environment Federation: Washington, DC, 1995. (45) Kim, K. M.S. Thesis, San Diego State University, 1999. (46) Johnson, T. L.; Scherer, M. M.; Tratnyek, P. G. Environ. Sci. Technol. 1996, 30, 2634-2640. (47) Tratnyek, P. G.; Johnson, T. L.; Scherer, M. M.; Eykholt, G. R. Ground Water Monit. Rem. 1997, Fall, 108-114. (48) Cox, E. E.; Edwards, E.; Neville, S. In Perchlorate in the Environment; Urbansky, E. T., Ed.; Kluwer Academic/Plenum Publishers: New York, 2000. (49) Brown, G. M. J. Electroanal. Chem. 1986, 319-330. (50) Hora´nyi, G.; Bakos, I. ACH-Models Chem. 1994, 131, 25-41. (51) McBride, M. B. Environmental Chemistry of Soils; Oxford University Press: Oxford, 1994. (52) Jolivet, J.-P. Metal Oxide Chemistry and Synthesis; John Wiley & Sons Ltd.: West Sussex, 2000. (53) Johnson, T. L.; Fish, W.; Gorby, Y. A.; Tratnyek, P. G. J. Contam. Hydrol. 1998, 29, 379-398. (54) Aramaki, K.; Tomihari, M.; Furuya, S.; Yamaguchi, M.; Nishihara, H. Corros. Sci. 1994, 36, 1133-1141. (55) Wesolowski, D. J.; Machesky, M. L.; Palmer, D. A.; Anovitz, L. M. Chem. Geol. 2000, 167, 193-229. (56) Dzombak, D. A.; Morel, F. M. M. Surface Complexation Modeling: Hydrous Ferric Oxide; John Wiley & Sons: New York, 1990. (57) Eykholt, G. R.; Baghel, S. S.; Sivavec, T. M.; Mackenzie, P. D.; Haitko, D.; Horney, D. Proceedings of the 209th National Meeting of the American Chemical Society, Anaheim, CA, 1995; American Chemical Society: Washington, DC, 1995. (58) Gupta, N.; Sass, B. M.; Gavaskar, A. R.; Sminchak, J. R.; Fox, T. C.; Snyder, F. A. In Designing and Applying Treatment Technologies: Proceedings of the First International Conference on Remediation of Chlorinated and Recalcitrant Compounds, May 18-21, 1998, Monterey CA; Wickramanayake, G. B., Hinchee, R. E., Eds.; Battelle Press: Columbus, OH, 1998. (59) Cumming, L.; Sass, B. M.; Gavaskar, A. R.; Drescher, E.; Williamson, T.; Drescher, M. In Designing and Applying Treatment Technologies: Proceedings of the First International Conference on Remediation of Chlorinated and Recalcitrant Compounds, May 18-21, 1998, Monterey, CA; Wickramanayake, G. B., Hinchee, R. E., Eds.; Battelle Press: Columbus, OH, 1998. (60) Toride, N.; Leije, F. J.; van Genuchten, M. Th. The CXTFIT code for estimating transport parameters from laboratory or field tracer experiments, version 2.1; 137; U.S. Salinity Laboratory, U.S. Department of Agriculture: Riverside, CA, 1995. (61) Scherer, M. M.; Richter, S.; Valentine, R. L.; Alvarez, P. J. J. Crit Rev. Environ. Sci. Technol. 2000, 30, 363-411. (62) Gillham, R. W.; O’Hannesin, S. F. Ground Water 1994, 32, 958967. (63) Urbansky, E. T.; Schock, M. R. J. Environ. Manage. 1999, 57, 79-95. (64) Miller, J. P.; Logan, B. E. Environ. Sci. Technol. 2000, 34, 30183022. (65) Brewster, J. H. J. Am. Chem. Soc. 1954, 76, 6361-6363. (66) Li, T.; Farrell, J. Environ. Sci. Technol. 2001, 35, 3560-3565. (67) Taube, H. Mechanistic Aspects of Inorganic Reactions; ACS Symposium Series 198; American Chemical Society: Washington, DC, 1982. (68) Liger, E.; Charlet, L.; Van Cappellen, P. Geochim. Cosmochim. Acta 1999, 20, 2939-2956. (69) Charlet, L.; Silvester, E. J.; Liger, E. Chem. Geol. 1998, 151, 8593.

Received for review July 29, 2002. Revised manuscript received February 21, 2003. Accepted April 23, 2003. ES026007T