![[logo] US EPA](https://webarchive.library.unt.edu/eot2008/20081106080455im_/http://www.epa.gov/epafiles/images/logo_epaseal.gif){kind=logo}
2004 Progress Report: Kinetic and Mechanistic Framework for Remediation Using Zerovalent Iron (SEERII)
EPA Grant Number: R829422E03Title: Kinetic and Mechanistic Framework for Remediation Using Zerovalent Iron (SEERII)
Investigators: Zhang, Tian C. , Shea, Patrick J.
Institution: University of Nebraska at Lincoln
EPA Project Officer: Winner, Darrell
Project Period: August 5, 2002 through August 4, 2004 (Extended to August 4, 2005)
Project Period Covered by this Report: August 5, 2003 through August 4, 2004
Project Amount: $215,061
RFA: EPSCoR (Experimental Program to Stimulate Competitive Research) (2001)
Research Category: EPSCoR (The Experimental Program to Stimulate Competitive Research)
Description:
Objective:The objectives of this Science and Engineering Environmental Research (SEER) II project are to: (1) elucidate the kinetics and mechanisms of zerovalent iron (Fe0) treatment processes; (2) develop new approaches to enhance Fe0 performance; and (3) implement a successful cleanup of a contaminated field site.
Progress Summary:Kinetics and Mechanisms of Fe0 Performance
The effect of low pH (2-4.5) on nitrate reduction in an iron/nitrate/water system was investigated in a pH-stat. The results showed rapid nitrate reduction to ammonium at pH 2-4.5. A black coating, consisting of Fe(II) and Fe(III) formed on the iron grains as a corrosion product. X-ray diffractometry (XRD) indicated the black coating was poorly crystalline and its spectrum did not match commonly known iron oxides/hydroxides/oxide hydroxides or green rust I/II. The black coating did not inhibit Fe0 reactivity (at least at pH < 3), but was unstable and evolved with time into other oxides. A kinetic model incorporating the effects of pH on nitrate reduction and Langmuir adsorption of nitrate was proposed, and parameters estimated by nonlinear curve fitting. This model indicated two major effects of pH on the nitrate reduction kinetics: (1) H+ ions directly participate in the redox reaction of nitrate reduction following first-order kinetics and (2) H+ ions affect nitrate adsorption onto reactive sites.
The effects of three Good’s pH buffers on performance of a Fe0/nitrate/water system were evaluated. The buffer itself did not reduce nitrate directly. Nitrate reduction by iron powder at near-neutral pH was negligible in an unbuffered system, but it was greatly enhanced in the presence of the buffer. A significant amount of aqueous Fe2+ (or Fe3+) was released after adding the buffer to the Fe0/water system with or without nitrate. The pH of the buffered solution generally increased from the initial pH of about 4.6-5.3 (depending on buffer pKa) to near-neutral. After the initial increase, pH was more or less stable for about 5-10 hours, usually concurrent with a fairly stable aqueous Fe2+ concentration. The pH then drifted to approximately 7.1 to 8.6, depending on the initial concentration and pKa of the buffer, and the consumption of Fe2+ concurrent with nitrate reduction. Although a common assumption made by researchers is that Good’s pH buffers do not directly participate in reaction processes involved in contaminant remediation, this study shows that the Good’s pH buffer may react with iron powder.
A conventional tracer study using Li+ and Cl- was conducted on four Fe0-packed column reactors for nitrate removal. Both Li+ and Cl- showed strong adsorption onto iron media, and thus were not ideal for the study. Tests using an impulse loading of nitrate were then innovated to investigate nitrate transport and reduction in the reactors. The impulse loading was superimposed on a continuous constant feeding of nitrate, which generated a steady effluent baseline. A multi-variable model incorporating hydraulic dispersion, adsorption/desorption, and reduction of nitrate was developed and numerically solved. Both Langmuir adsorption and linear adsorption isotherms were separately applied to describe nitrate adsorption on the reactive surface. The parameters of the model were estimated by fitting the model with the response curves from the impulse loading tests. These estimated parameters were consistent with previous studies. Specifically, the modeling results suggest a significant adsorption of nitrate by the iron media, causing an evident retardation effect. This research may lead to new methods for studying the fate of contaminants in porous reactive environments.
Our previous work demonstrated that Al2(SO4)3 promoted Fe0-mediated dechlorination of a model chloroacetamide herbicide (metolachlor), with faster kinetics when FeSO4 was added with the Al2(SO4)3. The appearance of green rust (GR(II); Fe6(OH)12SO4) prompted determination of the oxides formed and their influence on metolachlor dechlorination. Peerless annealed Fe0 contains a thin layer of magnetite (Fe3O4), maghemite (γ-Fe2O3), and wüstite (FeO), and XRD indicated akaganeite (β-FeOOH) formation during treatment. Goethite (α-FeOOH) and some lepidocrocite (γ-FeOOH) formed when Al2(SO4)3 was present, although goethite and magnetite (Fe3O4) were identified in Fe0 treatments containing FeSO4. GR(II) formed in the presence of FeSO4 and facilitated Fe0-mediated dechlorination of metolachlor. Adsorption but no dechlorination was observed when GR(II) was synthesized in the presence of metolachlor and Eh/pH changed to favor Fe(III) oxyhydroxide or magnetite formation. In contrast, dechlorination occurred in a batch system containing magnetite or natural goethite + FeSO4 at pH 8 and continued when additional Fe(II) was provided. Whereas metolachlor was not dechlorinated by GR(II) during a 48-hour incubation, the GR(II) provided a source of Fe(II) and produced magnetite (and other oxide surfaces) that coordinated Fe(II), which then facilitated metolachlor dechlorination. Aside from dechlorination through reaction with Fe0, results indicate the potential of surface-bound Fe(II) to reduce chloroacetamides. Metolachlor dechlorination may be promoted by creating conditions resulting in the formation of oxides with a high surface density of reactive surface sites and under which sufficient Fe2+ is available. The appearance of green rust is a good indicator of these conditions.
Develop New Approaches to Enhance Fe0 Performance
Tests were conducted in Fe0-packed columns to investigate the effects of adding selected cations on nitrate removal by Fe0. Because of rapid passivation of the Fe0, only negligible nitrate was reduced in the columns without adding the selected cation. Adding certain cations (Fe2+, Fe3+, or Al3+) to the feed solution, however, significantly enhanced nitrate reduction. Extending hydraulic retention time (HRT) increased nitrate removal by the columns, but the increase was not linearly proportional to HRT. Decreases in column hydraulic conductivity (K) were monitored during an 8-month operating period. A modest decrease in K was recorded in the upper and the middle section of the media bed, whereas a significant decrease in K occurred in the inlet section. XRD indicates that magnetite was the dominant corrosion product in the entire height of the column media under anoxic and other test conditions. In the inlet section, however, lepidocrocite and goethite also were identified. Cementation only occurred in the inlet section, suggesting that lepidocrocite and goethite, rather than magnetite, might be responsible for the cementation thereby causing the hydraulic clogging observed in permeable reactive iron barriers. The magnetite coating, however, would not necessarily cause clogging of the media and may prolong the lifetime of the reactive barrier.
Batch tests were conducted in Fe0 systems to investigate oxygen consumption and the effect of dissolved oxygen (DO) on the composition of iron corrosion products, nitrate reduction, the reactivity of Fe0, the role Fe2+ (aq), and the fate of Fe2+ in the system. Results indicate that without augmenting Fe2+ (aq), neither nitrate nor DO could be removed efficiently by Fe0. In the presence of Fe2+ (aq), nitrate and DO could be reduced concomitantly with limited cross-interference. Unlike nitrate reduction, DO removal by Fe0 didn’t consume Fe2+ (aq). A two-layer structure, with an inner layer of magnetite and an outer layer of lepidocrocite, may be formed in the presence of DO. When DO depleted, the outer lepidocrocite layer transformed to magnetite. The inner layer of magnetite, even in a substantial thickness, might not impede the Fe0 reactivity as much as the thin interfacial layer between the oxide coating and bulk solution. Surface-bound Fe2+ may greatly enhance electron transfer from the Fe0 core to the solid-liquid interface and thus improve the performance of the Fe0 process.
Nitrate reduction in an iron/nitrate/water system with or without an organic buffer was investigated using multiple batch reactors under strict anoxic/anaerobic conditions. Nitrate reduction was very limited (< 10%) at near-neutral pH in the absence of an organic buffer. Nitrate reduction would be greatly enhanced, however, if the system (1) had a low initial pH (~2-3), (2) was primed with adequate aqueous Fe2+, or (3) was in the presence of the organic buffer. In cases (1) and (3), nitrate reduction usually involved three stages. The first stage was very fast and was associated with acidic corrosion of the iron grains. The second stage was very slow as a result of the formation of amorphous oxides on the surface of the iron grains, and the third stage was characterized by a rapid nitrate reduction concurrent with disappearance of aqueous Fe2+. Acetate (at pH = 4.1) was used to extract various Fe(II) species that were loosely adsorbed on the surface of iron grains. Once nitrate was exhausted in the system, no more Fe2+ would be consumed. In the presence of nitrate, however, surface-complexed Fe(II) might be oxidized and become structural Fe(III), resulting in a steadily increasing ratio of Fe(III)/Fe(II) in the oxides formed. Sustained adsorption and hydrolysis of Fe2+ was found to shift nonstoichiometric amorphous iron oxides into crystalline magnetite that might trigger (or be a result of) rapid nitrate removal.
Differences in solubility and reactivity among high explosives complicate the development of remediation treatments for munitions-contaminated soils. When Fe0 was used to treat soils containing multiple energetics, high destruction rates were observed for RDX and TNT but not HMX (as indicated in the 2003 annual report). The objective was to enhance HMX destruction, particularly in the presence of solid-phase material, using cationic surfactants to increase HMX solubility and availability at the Fe0 surface. In subsequent studies, HMX destruction by surfactant-coated Fe0 also was evaluated. Batch experiments were conducted with Fe0 and 0 to 4 percent (w/v) didecyldimethylammonium bromide (didecyl) to determine optimum surfactant concentration for HMX removal in aqueous solutions (with and without solid-phase HMX) and in soil slurries. Results showed that Fe0 plus 2 percent didecyl was most effective in removing HMX, RDX, and TNT from a munitions-contaminated soil (Los Alamos National Laboratory, NM). When soil slurries were treated with 2 percent didecyl, equilibrium concentrations were slightly above the critical micelle concentration (CMC). XRD analysis indicated that didecyl affected Fe0 corrosion, with more oxides present (most notably goethite) in the Fe0 plus didecyl treatment compared to Fe0 alone. Pretreating Fe0 with didecyl was superior to using Fe0 alone or mixing Fe0 and didecyl together, in removing HMX from aqueous solution, but not as effective as Fe0 plus didecyl when solid-phase HMX was present. Reseeding experiments in which HMX was added intermittently to the batch reactor to mimic dissolution of solid phase HMX showed that Fe0 pretreated with didecyl was highly reactive, but reaction rates were not sustained. It was concluded that didecyl surfactant can enhance HMX destruction by Fe0 and is optimized when concentrations are below the CMC so that only patchy or mono surfactant layers form on the iron surface.
In situ treatment of aquifers by dithionite to produce Fe(II)-bearing minerals and surface-bound Fe(II) (i.e., redox barrier) also has been proposed as an effective remediation treatment. Initial experiments indicated that dithionite rapidly dechlorinates chloroacetanilide herbicides in water, with the following order of reactivity: propachlor > alachlor > acetochlor > metolachlor. Mass balance showed stoichiometric release of chloride. In batch tests, propachlor (0.76 mM) was exposed to dithionite-treated aquifer sand (91% sand, 3% silt, and 6% clay) and dechlorination rates were determined. Dechlorination kinetics increased as the amount of reduced aquifer material increased (25 to 125 g aquifer sand L-1 treated with 18.8 mM dithionite) or as the dithionite concentration increased from 9.4 to 188 mM (50 g aquifer sand L-1). Washing the reduced sediments removes Fe(II) and thus hinders herbicide transformation. In contrast, Sharpsburg soil, rich in clay and iron, effectively degraded alachlor even after being washed. Dechlorination also occurred when the washed, dithionite-treated sediments were amended with Fe(II) (as FeSO4) at pH 8.5 and continued as long as additional Fe(II) was provided. The dechlorination of chloroacetanilide herbicides by dithionite and dithionite-treated sediments and soil indicates a remedial option that could be employed in natural environments where iron-bearing minerals are abundant in soils and sediments and can be reduced to Fe(II). Because of the low toxicity of dithionite and its reaction products, dithionite can be used as an environmentally suitable reductant for abiotic degradation of chloroacetanilides.
The potential of selected graphitic/activated carbon materials to improve the Fe0 treatment process by simultaneous action as adsorbents and redox mediators was explored. In preliminary tests, RDX destruction by Peerless cast iron aggregates was faster than HMX but the addition of graphite increased the HMX destruction rate. A similar increase was observed when zinc metal was used with the graphite in place of Fe0. All reaction rates were greater under anaerobic conditions than under aerobic conditions. Graphitic carbon is believed to provide sites for adsorption and chemical reduction by serving as a conduit for electrons and hydrogen from Fe0 oxidation. These preliminary observations and related research indicate that graphite may have the potential to increase the efficiency of chemical reduction treatment processes.
Implement a Successful Cleanup of a Contaminated Field Site
The 2003 annual report described the initial results of a field study in which pesticide-contaminated soil from a Nebraska agricultural cooperative was treated with Fe0 and aluminum sulfate. The soil contained average concentrations of 653 mg atrazine, 177 mg metolachlor, and 4,220 mg nitrate-N per kg. As reported, the treatment decreased metolachlor, atrazine, and nitrate-N concentrations by greater than 99 percent, 80 percent, and 40 percent. Although remediation of metolachlor contamination was complete after treatment with Fe0 plus Al2(SO4)3, secondary treatment was necessary to further reduce concentrations of atrazine and nitrate-N. Table sugar (sucrose) was added as a supplemental carbon source by incorporating 128 25-lb paper bags of sugar into the soil with the high speed, soil-mixing implement. The addition of readily available carbon (table sugar) while maintaining anaerobic conditions resulted in a greater than 95 percent decrease in atrazine and greater than 88 percent decrease in nitrate-N concentrations.
Future Activities:In Year 3 of this SEER II project, we will complete studies of the mechanisms of reaction with Fe0 and the role Fe(II) associated with iron oxide surfaces (Objective 1). Research with dithionite will be completed and we will complete recently initiated studies with graphite to enhance the kinetics of Fe0–promoted remediation in water and soil (Objective 2). A field study previously was completed and we anticipate that Dr. Comfort (SEER I project R829422E01 co-principal investigator) will initiate a major field demonstration project next year as a result of new funding (Objective 3). One Ph.D. and two M.S. students associated with the project are expected to complete their degree programs during the next year. Additional publications and presentations are anticipated. Funding will be sought to continue this research.
Journal Articles on this Report: 4 Displayed | Download in RIS Format
| Other project views: | All 34 publications | 14 publications in selected types | All 14 journal articles |
| Type | Citation | ||
|---|---|---|---|
|
|
Gibb C, Satapanajaru T, Comfort SD, Shea PJ. Remediating dicamba-contaminated water with zerovalent iron. Chemosphere 2004;54(7):841-848. |
R829422E03 (2004) |
not available |
|
|
Huang YH, Zhang TC. Effects of low pH on nitrate reduction by iron powder. Water Research. 2004;38(11):2631-2642. |
R829422E03 (2004) |
not available |
|
|
Park J, Comfort SD, Shea PJ, Machacek TA. Remediating munitions-contaminated soil with zerovalent iron and cationic surfactants. Journal of Environmental Quality 2004;33(4):1305-1313. |
R829422E03 (2004) |
not available |
|
|
Satapanajaru T, Shea PJ, Comfort SD, Roh Y. Green rust and iron oxide formation influences metolachlor dechlorination during zerovalent iron treatment. Environmental Science and Technology 2003;37(22):5219-5227. |
R829422E03 (2003) R829422E03 (2004) |
not available |
cleanup, sediments, restoration, engineering, environmental
chemistry, remediation technologies, nitrate, iron, Science and Engineering
Environmental Research, SEER,
,
POLLUTANTS/TOXICS, Water, Geographic Area, Scientific Discipline, Waste, Remediation, Water Pollutants, Contaminated Sediments, Groundwater remediation, Ecology and Ecosystems, State, water quality, dehalogenation, verticle attachment, fate and transport, predictive understanding, reductive treatment, ecology assessment models, kinetic studies, contaminated aquifers, contaminated sediment, remediation technologies, contaminated soil, groundwater contamination, hazardous waste, contaminated groundwater, sediment treatment, environmental engineering, Nebraska (NE), nitrate, zero valent iron
Progress and Final Reports:
2003 Progress Report
Original Abstract
Final Report