JEQ Journal of Natural Resources and Life Sciences Education
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Published online 9 August 2005
Published in J Environ Qual 34:1539-1546 (2005)
DOI: 10.2134/jeq2005.0002
© 2005 American Society of Agronomy, Crop Science Society of America, and Soil Science Society of America
677 S. Segoe Rd., Madison, WI 53711 USA
This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Related articles in JEQ
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Matocha, C. J.
Right arrow Articles by Wagner, K. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Matocha, C. J.
Right arrow Articles by Wagner, K. M.
Agricola
Right arrow Articles by Matocha, C. J.
Right arrow Articles by Wagner, K. M.
Related Collections
Right arrow Redox Processes
Right arrow Heavy Metals

TECHNICAL REPORTS

Heavy Metals in the Environment

Reduction of Copper(II) by Iron(II)

C. J. Matocha*, A. D. Karathanasis, S. Rakshit and K. M. Wagner

Department of Agronomy, University of Kentucky, N-122 Agricultural Science Center-North, Lexington, KY 40546-0091

* Corresponding author (cjmato2{at}uky.edu)

Received for publication January 4, 2005.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Laboratory and field investigations have clearly demonstrated the important role of reduced iron (Fe(II)) in reductive transformations of first-row transition metal species. However, interactions of Fe(II) and copper (Cu) are not clearly understood. This study examined the reduction of Cu(II) by Fe(II) in stirred-batch experiments at pH 5.2 and 5.5 as influenced by chloride (Cl–) concentration (0.002–0.1 M), initial metal concentration (0.1–9.1 mM), and reaction time (1–60 min) under anoxic conditions. Reduction of Cu(II) to Cu(I) by dissolved Fe(II) was rapid under all experimental conditions and the stability of the products explains the driving force for the redox reaction. Under conditions of low [Cl–] and high initial metal concentration, >40% of total Cu and Fe were removed from solution after 1 min, which accompanied formation of a brownish-red precipitate. X-ray diffraction (XRD) patterns of the precipitates revealed the presence of cuprite (Cu2O), a Cu(I) mineral, based on d-spacings located at 0.248, 0.215, 0.151, and 0.129 nm. Fourier transform infrared (FTIR) spectroscopy corroborated XRD data for the presence of Cu2O, with features located at 518, 625, and 698 cm–1. Increasing [Cl–] stabilized the dissolved Cu(I) product against Cu2O precipitation and resulted in more Fe precipitated from solution (relative to Cu) that appears to be present as poorly crystalline lepidocrocite ({gamma}-FeOOH). This process may be important in anoxic soil environments, where dissolved Fe(II) levels can accumulate.

Abbreviations: FTIR, Fourier transform infrared • T (subscript), total • XRD, X-ray diffraction


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
COPPER BEHAVIOR in soils and natural waters is influenced by electron transfer, precipitation, complexation, and sorption processes. Cupric (Cu(II)) and cuprous (Cu(I)) oxidation states can exist in aqueous solution, but Cu(I) is thought to be relatively insignificant because it disproportionates to Cu(II) and Cu(0) (Shriver et al., 1994). However, field studies have shown that dissolved Cu(I) comprises a significant fraction of total Cu in various freshwater (Xue et al., 1991; Glazewski and Morrison, 1996), saltwater (Moffett and Zika, 1988; Voelker et al., 2000), and rainwater environments (Kieber et al., 2004). Chloride (Cl–) is an important ligand in soil solutions that can stabilize Cu(I) with respect to disproportionation (Kamau and Jordan, 2001) and may explain the stability of dissolved Cu(I) in natural waters. In fact, Moffett and Zika (1983) showed that the half-life of dissolved Cu(I) in water drastically increased with an increase in Cl– concentration.

Direct reductants of Cu(II) include hydrogen peroxide and sulfide. Kinetics of Cu(II) reduction by hydrogen peroxide were followed by measuring Cu(I) formation and it was found that rates increased with an increase in pH and Cl– (Millero et al., 1991). Mixing Cu(II) with sodium bisulfide or sodium sulfide rapidly forms covellite, a solid CuS mineral in which Cu is present as Cu(I) (Silvester et al., 1991; Pattrick et al., 1997). Covellite was proposed to control dissolved Cu solubility in flue gas sludge (Rai et al., 1994) and it has been recently identified in near surface sediments enriched in Cu (Martin et al., 2003).

Elevated concentrations of dissolved Fe(II) occur in anaerobic environments where microorganisms couple oxidation of organic carbon to reduction of solid Fe(III) minerals in soil (Lovley, 1987). Laboratory and field investigations have clearly demonstrated the important role of reduced iron (Fe(II)) in reductive transformations of first row transition metal species (Eary and Rai, 1988; Wehrli, 1990; White and Peterson, 1996; Buerge and Hug, 1997). Dissolved Fe(II) can reduce solid Mn(III,IV) (hydr) oxides, an important secondary reaction to consider in some natural environments (Postma, 1985; Wehrli, 1990). The reduction of hexavalent chromium (Cr(VI)) to Cr(III) by Fe(II) occurs very rapidly. The products, Cr(III) and Fe(III), form insoluble brown precipitates that produced no X-ray lines but were thought to be mixed Cr(III)–Fe(III) hydroxide minerals (Eary and Rai, 1988; Buerge and Hug, 1997).

It is well known that addition of trace quantities of Cu(II) can catalyze the oxidation of dissolved Fe(II) and structural Fe(II) by O2 (Stumm and Lee, 1961; Sayin, 1982), but there is evidence to suggest that Cu(II) can serve as a direct oxidant of Fe(II) in the absence of O2. Under anoxic conditions, addition of Cu(II)Cl2 solutions to magnetite (Fe3O4) and ilmenite (FeTiO3) containing structural Fe(II) resulted in loss of Cu(II) and production of reduced Cu species (Cu(I) or Cu(0)) based on X-ray photoelectron spectra (White and Peterson, 1996). Green rust, a mixed Fe(II)–Fe(III) mineral identified in reducing soil environments (Trolard et al., 1997), was shown to reduce Cu(II) to Cu0 (O'Loughlin et al., 2003). Dissolved Cu(I) accounted for greater than 50% of the total Cu in a river water that contained low oxygen concentrations (Glazewski and Morrison, 1996). It was speculated that cycling of Fe(II) to Fe(III) may be involved in Cu(I) production. This conclusion was not based on direct experimental evidence. The objective of this study was to investigate the reactivity of dissolved Cu(II) with Fe(II) by performing experiments that bear on questions of kinetics, stoichiometry, and product characterization.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Materials
We used ACS reagent-grade CuCl, bathocuproine (2,9-dimethyl-4,7-diphenyl-1,10-phenanthroline), and Na4EDTA to prepare Cu(I)–bathocuproine standards, and ACS reagent-grade CuCl2 and FeCl2 to prepare deoxygenated 0.1 M CuCl2 and FeCl2 stock solutions for reactivity experiments and standards. 2-(N-morpholino)ethanesulfonic acid monohydrate (MES) was used to buffer pH. All deionized water was deoxygenated by purging with Ar for 3 h before transferring into the oxygen-free anaerobic chamber (95% Ar to 5% H2, Pd catalyst; Coy Laboratory Products, Grass Lake, MI) where the experiments were performed. Dissolved oxygen concentrations of deoxygenated water were determined with a Clark-type polarographic electrode (Warner Instruments Incorporated, Hamden, CT) and found to be <1 µM, verifying anoxic conditions.

Reduction Experiments
Stirred-batch kinetic experiments were initiated by adding a measured volume of Cu(II) stock solution to well-mixed deoxygenated Fe(II) solutions at 23°C in 125-mL Nalgene bottles (Nalge Nunc International, Rochester, NY). Initial concentrations of Cu(II) and Fe(II) were varied between 0.1 and 9.0 mM and experiments were conducted in duplicate or triplicate. The reduction reactions were followed as a function of chloride concentration (2 mM–0.10 M) in both deoxygenated water and MES buffer to control pH. Reduction experiments performed with 1 mM Cu(II) and Fe(II) at pH 5.5 in 0.1 M MES buffer were comparable with those using a pH-stat, indicating that MES had a negligible effect. The pH values of the experimental solutions were 5.2 and 5.5 measured using a Metrohm (Herisau, Switzerland) 744 pH meter with a combination glass electrode. Higher concentrations of 9.0 mM Cu(II) and Fe(II) were used in some experiments to cause precipitation of sufficient solid to allow for structural characterization. The resulting precipitates in these experiments were formed at pH 5.2 (±0.2) and a Metrohm Model 716 automatic titrator in pH-stat mode was used to maintain constant pH by addition of standardized 1 M NaOH. Control experiments were performed with 9.0 mM Cu(II) and Fe(II) at pH 5.2 in separate solutions and the change in dissolved Cu(II) and Fe(II) was negligible over the 60 min that the reaction was monitored.

Suspension aliquots were removed at increasing time intervals with an automatic pipette and the reaction between Cu(II) and Fe(II) was stopped by filtration with 0.2-µm pore size membrane filters. The filtrate was immediately divided into two portions to determine dissolved Cu(I) and Fe(II). One aliquot was added to a bathocuproine–EDTA reagent mixture to complex Cu(I) while preventing Cu(II) interference (Moffett et al., 1985). The dissolved Cu(I)–bathocuproine complex exhibits a maximum absorbance at a wavelength ({lambda}max) of 483 nm, with a molar extinction coefficient ({epsilon}) of 8.5 x 103 M–1 cm–1. Dissolved Fe(II) was determined in a separate filtrate after complexation with ferrozine [3-(2-pyridyl)-5,6-diphenyl-1,2,4-triazine-4',4''-disulfonic acid monosodium salt] at 562 nm (Stookey, 1970). Absorbance spectra were recorded in quartz cuvettes with a 1-cm optical pathlength at 24°C on a double-beam Shimadzu (Kyoto, Japan) UV-3101PC spectrophotometer equipped with a temperature-controlled carrier (TCC) Peltier temperature controller. Copper(I) also forms a brownish-colored complex with ferrozine at 470 nm (Kundra et al., 1974). Both Cu(I) and Fe(II) concentrations (CCu and CFe) were determined using the principle of additive absorbances and solving Eq. [1] and [2] simultaneously:

[1]

[2]

The {epsilon} values at the respective wavelengths were empirically determined to be 3.47 x 103, 9.84 x 103, 1.87 x 103, and 2.75 x 104 M–1 cm–1 for {epsilon}Cu470, {epsilon}Fe470, {epsilon}Cu562, and {epsilon}Fe562, respectively. These values reasonably agree (±16%) with Kundra et al. (1974). The results for dissolved Cu(I)–ferrozine and Cu(I)–bathocuproine methods compared favorably with one another (±20%), considered within experimental error given the rapid kinetics of the redox process and subsequent precipitation of products (see results below). Dissolved Cu(I)–bathocuproine data were presented rather than Cu(I) determined using Eq. [1] and [2] because this method was more sensitive. In some experiments, Cl– concentrations in filtrates were quantified using ion chromatography (Metrohm).

Total Fe (FeT) and Cu (CuT) concentrations in ferrozine solutions were measured using flame atomic absorption spectrometry (Shimadzu AA-6800) against appropriate standards. It was assumed that the difference between CuT and Cu(I) represented dissolved Cu(II) because there was no evidence of Cu0 production. The difference between FeT and Fe(II) was used to estimate dissolved Fe(III). Copper and iron speciation were evaluated using constants in the MINEQL+ database (Schecher, 1998).

Solid-Phase Analysis
X-ray diffraction (XRD) was performed with a PW 1840 Phillips (Eindhoven, the Netherlands) diffractometer, equipped with CoK{alpha} radiation (40 kV, 30 mA), on wet mounts dried under Ar. Scans were made from 10 to 90°2{theta} at increments of 0.02°2{theta}.

Fourier transform infrared spectroscopy (FTIR) was conducted on a Nicolet 20 SXC (Thermo Electron, Waltham, MA) on approximately 3 mg of sample that was homogenized in a ball mill with 600 mg of KBr and pressed into pellets. The spectrometer chamber was allowed to purge 5 min under N2 before and during spectral acquisition, which consisted of a collection of 512 scans co-added together. Spectral processing was performed using GRAMS 32AI.

Argon dried samples were deposited on carbon tape attached to Al holders that were sputter coated with Au and Pd to reduce sample charging in the beam. A Hitachi (Tokyo, Japan) S-3200 scanning electron microscope equipped with a Noran Voyager energy dispersive X-ray system (Thermo Electron) was used to image precipitates at an excitation voltage of 20 keV, a spectrum acquisition time of 60s, and a working distance of 15 mm.


   RESULTS AND DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
CONCLUSIONS
 REFERENCES
 
Solution Concentrations
Under all experimental conditions, reduction of Cu(II) by Fe(II) was rapid, generally complete within 1 to 2 min based on immediate production of Cu(I). The overall redox reaction between Cu(II) and Fe(II) was initially assumed to be a net one-electron transfer, resulting in one mole of Fe(II) consumed and one mole of Cu(I) produced, as shown in Eq. [3]:

[3]

At equimolar concentrations of Cu(II) and Fe(II) (1 mM), initial [Cl–] of 2 mM, and pH 5.5, the molar ratio of Fe(II) consumed to Cu(I) produced ({Delta}[Fe(II)]/[Cu(I)]) after 1 min was 0.14 (Fig. 1a) , much less than that predicted from Eq. [3]. This value increased to 0.52 as [Cl–] was raised to 0.1 M under similar chemical conditions (Fig. 1c). Additional experiments conducted with 0.1 mM initial [Cu(II)] and [Fe(II)] resulted in {Delta}[Fe(II)]/[Cu(I)] = 0.05 (data not shown). Increasing the initial metal concentrations to 9.1 mM Cu(II) and Fe(II) resulted in {Delta}[Fe(II)]/[Cu(I)] ratios of 2.3 and 1.2 at 0.04 and 0.1 M [Cl–] after 1 min (Fig. 2) . Precipitation was more rapid in these experiments following the initial electron transfer step in Eq. [3]. For example, greater than 40% of CuT and FeT were removed from solution after 1 min and approximately 95% after 60 min, which accompanied formation of a brownish-red precipitate (Fig. 2a). Control experiments in separate solutions ruled out precipitation of incipient Cu(II) and Fe(II) solids (due to supersaturated conditions in the initial solutions) where the change in Cu(II) and Fe(II) was negligible.



View larger version (30K):
[in this window]
[in a new window]
  		Fig. 1. Dissolved Cu and Fe concentrations over time under anoxic conditions with initial metal concentrations of 1.0 mM and (a) 2 mM [Cl–], (b) 4 mM [Cl–], and (c) 0.1 M [Cl–]. Reactions contained 0.1 M pH 5.5 MES buffer.

 


View larger version (24K):
[in this window]
[in a new window]
  		Fig. 2. Dissolved Cu and Fe concentrations over time under anoxic conditions with initial metal concentrations of 9.1 mM and (a) 0.04 M [Cl–] and (b) 0.1 M [Cl–]. Experiments were performed in water and an automatic titrator was used to maintain a pH of 5.2 ± 0.2.

 
Precipitation of Cu and Fe solids is the most likely explanation for nonstoichiometric reduction, particularly at higher initial metal concentrations (Fig. 2), where loss of Cu(I), CuT, Fe(II), and FeT from solution accompanied formation of solid-phase precipitates (see Solid Characterization, below). Thermodynamic modeling indicated that solutions in the 0.04 and 0.1 M [Cl–] treatments for equimolar (9.1 mM) metal levels were supersaturated for several Cu(I) and Cu(II) minerals (cuprite, nantokite, atacamite), numerous Fe(III) oxide mineral phases (akaganeite, hematite, maghemite, goethite, lepidocrocite, ferrihydrite), and mixed Fe–Cu minerals (cupric ferrite and cuprous ferrite) based on saturation index (SI) values of >0 (Table 1).


View this table:
[in this window]
[in a new window]
  		Table 1. Saturation indices (SI), calculated as log (IAP/Kso), where IAP is ion activity product, for predicted mineral solids computed using MINEQL+ after 1 min of reaction time. Reactions contained 9.1 mM Cu(II) and Fe(II) and were performed at pH 5.2 ± 0.2 in water at two chloride levels.

 
At lower initial metal concentrations, where precipitation was slower (Fig. 1), competing processes may be invoked to explain the deviation from Eq. [3]. The self-exchange electron transfer reaction between Fe(II) and Fe(III) is a very rapid process (Jolley et al., 1990) and the products of this reaction are equivalent to the reactants (Rosso et al., 2004). This may explain why it appears that negligible Fe(II) reacted despite the production of nearly 0.5 mM dissolved Cu(I) (Fig. 1a, 1b). It turns out that Cu+ and Cu2+ also undergo self-exchange, but not as readily because of the reorganization energies involved in converting between different coordination geometries, four-coordinate (tetrahedral) for the d10 Cu(I) and six-coordinate (Jahn–Teller tetragonally distorted octahedral) for d9 Cu(II) (Rorabacher, 2004). Oxidation of dissolved Cu(I) by Fe(III) has been reported (Parker and Espenson, 1969; Rorabacher, 2004). This represents another process that may be important in lowering the {Delta}[Fe(II)]/[Cu(I)] value.

Increasing Cl– concentration resulted in large changes in Cu behavior following the initial electron transfer step (Fig. 1). In dilute Cl– solutions (2 mM), roughly 65% of the total Cu (CuT) was removed from solution after 60 min, with a significant fraction of CuT present as dissolved Cu(I). In contrast, only 15% of CuT was removed from solution at 0.1 M Cl– and the fraction as Cu(I) increased from 0.54 to 0.90 as the reaction proceeded (Fig. 1c). More FeT and Fe(II) were lost from solution relative to Cu at 0.1 M Cl– (Fig. 1c). These same general trends were reflected at higher initial metal concentrations, where dissolved Cu(I) was stabilized over time in solutions containing 0.1 M Cl–, when compared with 0.04 M Cl– solutions (Fig. 2).

The pronounced stabilization effect of Cl– on Cu(I) can be understood when one considers that Cl– complexes Cu(I) strongly as revealed from the successive formation constants taken from the MINEQL+ database (Schecher, 1998):

[4]

[5]

[6]

The total concentration of Cu(I), Cu(I)T, is the sum of free and complexed forms. Speciation calculations using MINEQL+ and data presented in Fig. 1 predict that the proportion of free Cu+ decreases from 28% at 2 mM to <1% at 0.1 M [Cl–] while the dichloro complex increases in importance from 18% at 2 mM to 93% at 0.1 M. The remainder of the Cu(I)T exists as CuCl0. Data from Fig. 2 are similar in that there is sufficient Cl– to complex >95% of the Cu(I)T, with the CuCl2– complex predicted to account for 85 to 92%. The CuCl2– species has been shown to be less reactive toward oxidation by O2 than CuCl0 and free Cu+ (Sharma and Millero, 1988). During Cu(II) reduction by hydrogen peroxide to form Cu(I), Millero et al. (1991) reported that the rates of Cu(I) production increased with increasing Cl– concentration. They attributed this finding to the greater stability of the CuCl2– product, suggesting that CuCl20 was the reactive species. Xu and Jordan (1990) reported a rate acceleration due to Cl– for anaerobic Cu(II) reduction to Cu(I) by ascorbic acid. The impact of Cl– on Cu(I) stability has been reported elsewhere (Moffett and Zika, 1983; Kamau and Jordan, 2001). It is unclear whether the reduction rate of dissolved Cu(II) by Fe(II) increased with greater Cl– because initial reaction rates were faster than we could measure using the methods employed. Our results do indicate that high Cl– stabilized Cu(I) in solution, slowing removal by precipitation.

Complex formation of Cu(I) by Cl– strongly influences reduction potentials (Table 2). The greater stability of Cu(I)–chloro complexes has the effect of raising the standard reduction potential (Eo) of the Cu2+–Cu+ couple, by as much as 0.414 V (Moffett and Zika, 1987). As a result, Cu(II) is a better oxidant. If it is assumed that Fe(II) is oxidized directly to freshly precipitated Fe(OH)3(s), the Nernst factor can be used with chemical conditions pertinent to this study to calculate an effective reduction potential (E'). Assuming that CuCl20 is the reactive species, it can be seen that the stability of the Cu products (CuCl2– and Cu2O(s)) and Fe(OH)3(s) is what makes the reaction possible thermodynamically (Table 2).


View this table:
[in this window]
[in a new window]
  		Table 2. Reduction potentials of Cu(II)–chloro complexes relative to Fe(III) oxides. Values for solid-phase minerals and water were assumed equal to unity. The estimated {Delta}Gf (Gibbs free energy of formation) for lepidocrocite was taken from Roden (2003). Solubility data for Cu2O and E0 values for CuCl20 were used to estimate CuCl20–Cu2O couple.

 
Solid Characterization
Insights were sought on the chemical nature of the brownish-red colored precipitate produced in the reactions involving low (0.04 M) and high (0.1 M) [Cl–] from Fig. 2. X-ray diffraction (XRD) patterns for the precipitates and corresponding plots of Fe(II) in solution and hydroxyl consumed are shown in Fig. 3 . In the 0.04 M Cl– treatment, peaks appeared corresponding to d-spacings located at 0.248, 0.215, 0.151, and 0.129 nm. All the peaks produced by the precipitate identified it as cuprite (Cu2O), cuprous oxide (Fig. 3a). Cuprite has a cubic structure and crystallizes in the space group Pn3 (Berry and Mason, 1959). There was good agreement of relative intensities with those reported in JCPDS Card 5-0667 (Joint Committee on Powder Diffraction Standards, 1980). Most of the oxidized Fe had apparently precipitated as poorly crystalline Fe(OH)3 or mixed Cu–Fe minerals not detected by the XRD analysis. Poorly crystalline ferrihydrite (Fe(OH)3) gives very weak and broad XRD lines, in the region between 20 and 40°2{theta} (Fig. 3a), making it difficult to detect (Schwertmann et al., 1982). Rapid Fe(II) oxidation by HCrO4– resulted in production of Fe(III) and Cr(III), which coprecipitated together in a poorly crystalline Fe(III)–Cr(III) precipitate (Eary and Rai, 1988; Buerge and Hug, 1997). Chromium(III) is more structurally compatible to coprecipitate with Fe(III) oxides because of the similarity in its ionic radius (0.62 Å) compared with Fe(III) (0.64 Å), whereas Cu(II) is larger (0.72 Å in sixfold coordination) and will not coprecipitate as readily (McBride, 1994). In addition, the tetragonally distorted octahedral geometry of Cu(II) due to the Jahn–Teller effect (Shriver et al., 1994) may further limit coprecipitation and substitution with Fe(III) (Singh et al., 2000).



View larger version (29K):
[in this window]
[in a new window]
  		Fig. 3. (a) X-ray diffraction patterns of the solid-phase precipitates collected after 60 min of reaction time for 0.04 M and 0.10 M [Cl–] treatments, and (b) corresponding plots of base added during the precipitation reactions. Reactions contained 9.1 mM Cu(II) and Fe(II) and were performed at pH 5.2 ± 0.2 in water.

 
Formation of Cu2O was suppressed in the 0.1 M Cl– treatment based on XRD analysis (Fig. 3a). The absence of Cu2O agrees with solution results that showed dissolved Cu(I) was stabilized in high Cl– media with respect to precipitation (Fig. 1 and 2). The exact mechanism of this suppression is unclear. These observations agree with Kear et al. (2004), who reported that Cu2O stability was inversely related to free Cl– concentration. The percentage of free Cl– increased from 88.4% at 0.04 M [Cl–] to 92% in the 0.1 M treatment based on MINEQL+ calculations.

The FTIR spectra of the precipitates were compared with reported scans for Cu2O and other potential Cu and Fe minerals. The band at 625 cm–1 in the 0.04 M Cl– treatment agreed with the most intense peak for Cu2O, located at 615 cm–1 (Nyquist and Kagel, 1971). Additional shoulders located at 698 and 518 cm–1 corresponded to published values for Cu2O at 690 and 510 cm–1 (Fig. 4) . This corroborated the XRD data for the presence of Cu2O.



View larger version (26K):
[in this window]
[in a new window]
  		Fig. 4. Fourier transform infrared (FTIR) spectra of the solid-phase precipitates collected after 60 min of reaction time for 0.04 M and 0.10 M [Cl–] treatments. Reactions contained 9.1 mM Cu(II) and Fe(II) and were performed at pH 5.2 ± 0.2 in water.

 
Precipitates collected from the 0.1 M Cl– treatment were characterized by a broad peak centered at 468 cm–1, and a weak shoulder at 560 cm–1 (Fig. 4). These peaks are similar to those of atacamite (Cu2(OH)3Cl) at 475 and 572 cm–1, assigned as Cu(II)–O stretching vibrations (Braithwaite et al., 2004). Atacamite is a solid Cu(II) mineral that has been shown to precipitate from CuCl2 solutions (Pollard et al., 1989) and has been identified in close association with Cu2O in copper mineral deposits (Hawthorne et al., 2002). Appearance of a subtle XRD peak located at 16.1°2{theta} (0.549 nm; Fig. 3a) agrees with the most intense peak reported for atacamite (Parise and Hyde, 1986). Although no XRD lines were observed for any Fe(III) oxide minerals, a pronounced feature at 1019 cm–1 and appearance of weaker bands at 742 and 1155 cm–1 agreed with IR spectra for lepidocrocite, {gamma}-FeOOH (Weckler and Lutz, 1998). The most intense peak for {gamma}-FeOOH occurs at 1024 cm–1, consistent with a pronounced peak in the 0.1 M Cl precipitates at 1019 cm–1. High [Cl–] favors {gamma}-FeOOH precipitation from slow oxidation of FeCl2 solutions by O2 (Schwertmann and Cornell, 1991). In our experimental solutions, precipitation was quite rapid (Fig. 1 and 2), likely explaining the poorly crystalline nature of the precipitates.

The scanning electron microscope observations of precipitates from low (0.04 M) and high (0.1 M) Cl– experiments at pH 5.2 shed more light on associations of Fe and Cu. Although Cu2O crystals are characterized by cubic morphologies, no clear cubic faces were observed at 0.04 M [Cl–]. However, where more Fe was precipitated relative to Cu in the 0.1 M [Cl–] experiments, distinct particles were identified that contained a greater proportion of Fe relative to Cu. These particles exhibited needle-like morphology, generally <0.5 µm in width and variable in length (Fig. 5) . This morphology is consistent with published results for poorly crystalline {gamma}-FeOOH. Larsen and Postma (2001) prepared five different {gamma}-FeOOH minerals of varying crystallinity by changing Cl to Fe ratios and pH. The poorly crystalline {gamma}-FeOOH was prepared at high Cl to Fe ratios and adopted needle-shaped habits with a surface area of 112 m2 g–1. Results of energy dispersive spectroscopy analysis on multiple needles indicated an elemental ratio for Fe to Cu of 1.8:1 (Fig. 5). The Cu could be present as underlying Cu not associated with the needles.



View larger version (48K):
[in this window]
[in a new window]
  		Fig. 5. Representative scanning electron microscope image of precipitates after 60 min of reaction time for the 0.1 M [Cl–] treatment and corresponding energy dispersive X-ray spectra shown below. Reactions contained 9.1 mM Cu(II) and Fe(II) and were performed at pH 5.2 ± 0.2 in water. Scale bar equals 2 µm.

 
These results allow two possible overall reactions, Eq. [7] and [8], to be hypothesized for experiments at 0.04 and 0.1 M [Cl–]:

[7]

[8]

During the precipitation reactions, protons were produced and continuously neutralized with base to maintain constant pH (5.2 ± 0.2) during the reactions. For the 0.1 M [Cl–] treatment, 3.7 protons were produced per Fe(II) consumed (taken from slope value of 0.27) as shown in Fig. 3b. This is in good agreement with that predicted by Eq. [8]. For the 0.04 M Cl– level, there was a discrepancy between measured protons produced per Fe(II) consumed (6.6 to 1) and that predicted by Eq. [7] (4 to 1). The discrepancy between measured and predicted protons may involve reactions during the precipitation of Cu2O. Kear et al. (2004) reported nantokite (CuCl(s)) to be an intermediate in the corrosion of Cu0. Nantokite transforms in water to Cu2O resulting in proton production:

[9]

There was a slightly positive saturation index (SI) value for CuCl(s) precipitation (Table 1). Nantokite may form first in the initial stages of the reaction and convert to Cu2O.

Cuprite occurs as an oxidation product in copper ore deposits, often closely associated with other Cu minerals (Berry and Mason, 1959; Hawthorne et al., 2002). It has been shown to form as a product of Cu0 corrosion by direct oxidation with dissolved O2 or H2O (North and Pryor, 1970; Kear et al., 2004). The results of this work show that homogeneous reduction of Cu(II) by dissolved Fe(II) under anoxic conditions is possible, leading to Cu(I) stability in solution or precipitation as Cu2O depending on chloride concentration. Although the majority of experiments were performed at pH 5.2 and 5.5, instantaneous precipitation occurred at pH 6.0 and 6.5 (data not shown), suggesting more rapid reaction rates at higher pH as predicted from Eq. [7] and [8].

An example of a possible environment where this process may be observed would be in Fe(III)-reducing soil environments (Lovley, 1987), where dissolved Fe(II) accumulates and would react with Cu(II) introduced to solution by weathering processes or waste applications. Recent research identified Cu2O in a Cu-contaminated soil taken from the 0- to 20-cm depth based on X-ray absorption near edge structure spectroscopy (Liu and Wang, 2004). These results may assist in predicting Cu mobility in field soils, where models typically include sorption and complexation processes of Cu(II), but do not account for reduction of Cu(II) to Cu(I). Copper(II) reduction to Cu(I) has been invoked to explain model underestimations in dissolved Cu concentration (Hesterberg et al., 1993). A significant fraction of soil Cu(II) is typically complexed with naturally occurring organic ligands (McBride, 1994), which may affect its reactivity with dissolved Fe(II).


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
REFERENCES
 
Our studies demonstrate that Cu(II) can be reduced by Fe(II) to Cu(I) on extremely short timescales in anoxic environments and depending on Cl– concentration, can precipitate as Cu2O or remain stabilized in solution as the dichloro complex . The stability of the products (CuCl2–, Cu2O, and Fe(OH)3) explains the driving force for the redox reaction. In soils, the process of Cu(II) reduction by Fe(II) may be operative in Fe(III)-reducing environments, where dissolved Fe(II) levels can accumulate and react with Cu(II).


   ACKNOWLEDGMENTS
 
We are grateful for the support and collaboration of the Department of the Interior, U.S. Geological Survey, and the University of Kentucky Research Foundation, Grant Agreement no. 01HQGR0133. This project was supported by National Research Initiative Competitive Grant no. 2002-35107-12214 from the USDA Cooperative State Research, Education, and Extension Service. We thank Larry Rice at ASTECC for assistance in scanning electron microscope–energy dispersive spectroscopy analysis and Darrell Taulbee at the CAER for help and access to the FTIR spectrometer.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS AND DISCUSSION
 CONCLUSIONS
 REFERENCES
 


Related articles in JEQ:

This Issue in Journal of Environmental Quality

JEQ 2005 34: v. [Full Text]  




This Article
Right arrow Abstract Freely available
Right arrow Figures Only
Right arrow Full Text (PDF) Free
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Related articles in JEQ
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via ISI Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Matocha, C. J.
Right arrow Articles by Wagner, K. M.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Matocha, C. J.
Right arrow Articles by Wagner, K. M.
Agricola
Right arrow Articles by Matocha, C. J.
Right arrow Articles by Wagner, K. M.
Related Collections
Right arrow Redox Processes
Right arrow Heavy Metals

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
The SCI Journals Agronomy Journal Crop Science
Vadose Zone Journal Journal of Plant Registrations
Journal of Natural Resources
and Life Sciences Education		 Soil Science Society of America Journal