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


     

Published online 6 July 2006
Published in J Environ Qual 35:1193-1203 (2006)
DOI: 10.2134/jeq2005.0463
© 2006 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 Similar articles in this journal
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 Google Scholar
Google Scholar
Right arrow Articles by McLean, J. E.
Right arrow Articles by Sorensen, D. L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by McLean, J. E.
Right arrow Articles by Sorensen, D. L.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by McLean, J. E.
Right arrow Articles by Sorensen, D. L.
Related Collections
Right arrow Biogeochemical Processes
Right arrow Bioremediation and Biodegradation
Right arrow Water Pollution

TECHNICAL REPORTS

Bioremediation and Biodegradation

Iron and Arsenic Release from Aquifer Solids in Response to Biostimulation

J. E. McLeana,*, R. R. Dupontb and D. L. Sorensena

a Department of Biological and Irrigation Engineering
b Department of Civil and Environmental Engineering, Utah Water Research Laboratory, Utah State University, Logan, UT 84322-8200

* Corresponding author (jmcle{at}cc.usu.edu)

Received for publication December 15, 2005.

   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Biostimulation has been used at various contaminated sites to promote the reductive dechlorination of trichloroethylene (TCE), but the addition of carbon and energy donor also stimulates bacteria that use Fe(III) as the terminal electron acceptor (TEA) in potential competition with dechlorination processes. Microcosm studies were conducted to determine the influence of various carbon donors on the extent of reductive dissolution of aquifer solids containing Fe(III) and arsenic. Glucose, a fermentable and respirable carbon donor, led to the production of 1500 mg Fe(II) kg–1, or 24% of the total Fe in the aquifer sediment being reduced to Fe(II), whereas the same concentration of carbon as acetate resulted in only 300 mg Fe(II) kg–1 being produced. The biogenic Fe(II) produced with acetate was exclusively associated with the solid phase whereas with fermentable carbon donors as whey and glucose, 22 and 54% of the Fe(II) was in solution. With fermentation, some of the metabolites appear to be electron shuttling chemicals and chelating agents that facilitate the reductive dissolution of even crystalline Fe(III) oxides. Without the presence of electron shuttling chemicals, only surficial Fe in direct contact with the bacteria was bioavailable, as illustrated when acetate was used. Regardless of carbon donor type and concentration, As concentrations in the water exceeded drinking water standards. The As dissolution appears to have been the result of the direct use of As as an electron acceptor by dissimilatory arsenic reducing bacteria. Our findings indicate that selection of the carbon and energy donor for biostimulation for remediation of chlorinated solvent impacted aquifers may greatly influence the extent of the reductive dissolution of iron minerals in direct competition with dechlorination processes. Biostimulation may also result in a significant release of As to the solution phase, contributing to further contamination of the aquifer.

Abbreviations: AQDS, anthraquinone-2,6-disulfonate • DIRB, dissimilatory iron reducing bacteria • HAFB, Hill Air Force Base • HA-HCl, hydroxylamine hydrochloride • HRC, hydrogen releasing compound • OU, operable unit • TCE, trichloroethylene • TEA, terminal electron acceptor


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
THE ADDITION of electron donor material to many aquifers contaminated with chlorinated solvents, such as trichloroethylene (TCE), stimulates the microbial reductive dechlorination of these solvents (Smidt and de Vos, 2004). The stimulated microbial community, however, will use a variety of terminal electron acceptors (TEAs), potentially in competition with the dechlorination process. Studies on halorespiratory activity as a function of hydrogen concentration indicate that tetrachloroethene (PCE) and TCE reduction resulting in dehalogenation (<0.3–2 nM H2) will occur in the same hydrogen concentration range as sulfate (1–1.5 nM H2) and Fe(III) reduction (0.1–0.8 nM H2) (Loffler et al., 1999; Yang and McCarty, 1998; Lovley and Goodwin, 1988). Dechlorinating activity of dehalorespiring bacteria was repressed in the presence of nitrate and sulfate (Gerritse et al., 1997, 1999) and iron (Dupont et al., 2003; Heidorn et al., 2003; He and Sanford, 2003).

Dupont et al. (2003) conducted a laboratory microcosm experiment to identify an electron donor for use at a TCE contaminated site (OU5) at Hill Air Force Base (HAFB), Utah. The selection of donors to be tested, low solubility oils (coconut oil, low melting point oil, and high melting point oil) and soluble donors [whey, lactic acid, and hydrogen releasing compound (HRC) (Regenesis, San Clemente, CA)], was based on the use of these donors at field sites across the United States for enhanced anaerobic bioremediation (Air Force Center for Environmental Excellence, 2004). Even with the addition of 1000 mg C L–1, only minimal TCE dechlorination was observed after more than 300 d incubation. The reducing conditions produced from organic matter addition resulted in the complete removal of nitrate and sulfate, the reductive dissolution of Fe(III) containing minerals with the resulting release of Fe(II) and As to solution, along with production of methane. Likewise pilot and field studies conducted at Dover Air Force Base, Delaware, over 200 to 500 d of lactate (Ellis et al., 2000) or acetate, lactate, sugar, alcohol, or molasses (Lee et al., 2000; Heidorn et al., 2003) addition showed only partial dechlorination of PCE to cis-DCE. There was, however, a significant increase in the dissimilatory iron reducing bacteria (DIRB) population (Heidorn et al., 2003), and increased dissolved iron concentrations in field wells from background concentrations of 0.01 up to 60 mg L–1 after treatment (Ellis et al., 2000). The lack of complete dechlorination of PCE/TCE in these studies may be due to (i) dechlorinating bacteria not being present in the aquifer solids; (ii) dechlorinating bacteria being present but the reductive dissolution of iron minerals created conditions that were unfavorable for their growth or activity; (iii) competition by non-dehalogenating bacteria that use alternative TEAs and effectively outcompete dehalorespiring bacteria for electron donor; and/or (iv) dechlorinating bacteria being present and utilizing iron as their preferred TEA over chlorinated alkenes. At both Air Force sites the dehalorespiring bacteria Dehalococcoides has been detected using standard polymerase chain reaction (PCR) procedures with 16s rDNA probes. In addition, biostimulation with guar gum used during construction of an air sparge trench led to transient dechlorination of TCE at HAFB. These observations are evidence that the first two explanations for the lack of TCE dechlorination in the field and the HAFB microcosm studies are not likely.

Iron(III) is typically the most abundant electron acceptor in anaerobic subsurface environments. The Fe(III) in amorphous oxides is bioavailable due to the high surface area of these minerals, but numerous soil bacteria have been identified that will also use Fe(III) in crystalline oxides such as goethite and hematite as summarized by Roden and Zachara (1996). More recently, bacteria in the genus Desulfitobacterium (D. frappiere G2 and D. metallireducens), Shewanella putrefaciens CN32, S. oneidensis, and Geobacter sp. have been shown to utilize iron associated with clay minerals (Kostka et al., 1999; Dong et al., 2003; Ernstsen et al., 1998; Stucki et al., 1987; Shelobolina et al., 2003). Many of these organisms will utilize a variety of terminal electron acceptors. For example, Shelobolina et al. (2003) showed that D. frappieri G2 was capable of reducing poorly crystalline Fe(III) oxide and structural Fe(III) in smectite as well as the chlorinated solvents, TCE and PCE, and Finneran et al. (2002) described a D. metallireducens strain that reduces Fe(III), PCE, and TCE.

The reduction of structural Fe(III) from oxides and clays may not be observed in nutrient poor natural environments (Glasauer et al., 2003; Lovley et al., 2004), but biostimulation with organic materials provides an environment similar to the nutrient rich conditions used in laboratory studies that have shown the reduction of structural Fe(III) (Kostka et al., 1999; Dong et al., 2003; Ernstsen et al., 1998; Stucki et al., 1987; Shelobolina et al., 2003). Structural Fe(III) may provide a large pool of bioavailable iron under biostimulation conditions that is a source of Fe not usually accessible to DIRB under natural environmental conditions.

Bacteria may use direct attachment mechanisms to reduce Fe(III) associated with crystalline minerals. The use of Fe(III) deep within crystalline minerals implies that some organisms may be using mechanisms for accessing this iron other than direct attachment, although the role of pili as electron carrying wire in Geobacter spp. is just being described (Reguera et al., 2005). In natural systems, some species of microbes including Geothrix fermentans (Nevin and Lovley, 2002a), S. alga BrY (Nevin and Lovley, 2002b; Caccavo and Das, 2002), and S. oneidensis (Dong et al., 2003; Kostka et al., 1999) produce electron shuttling compounds to transfer electrons to otherwise inaccessible Fe(III). An important genus of DIRB, Geobacter, however, does not produce its own electron shuttle compounds (Nevin and Lovley, 2000; Straub and Schink, 2003) but can utilize various chemicals in the environment as electron shuttles. A number of studies have shown that the addition of anthraquinone-2,6-disulfonate (AQDS), a humic substance analogue, to microcosms containing various DIRB and ferrihydrite (Hacherl et al., 2001), illite (Dong et al., 2003), poorly crystalline Fe(III) oxide imbedded in alginate (Nevin and Lovley, 2000), hematite (Rosso et al., 2003), and ferrihydrite in agar (Straub and Schink, 2003) enhanced microbially mediated reductive dissolution of these minerals, as AQDS serves as an electron shuttle between the organism and inaccessible Fe(III). The Fe(III) in the above studies was not bioaccessible without the addition of AQDS. Humic substances, which contain quinone moieties, are suspected of acting as electron shuttles for electron flow in anaerobic environments. Commercially available humic acids as well as humic substances associated with leaf extracts, freshwater aquatic sediments, and ground water are utilized as electron shuttles by DIRB (Nevin and Lovley, 2002b). Other redox active compounds such as cysteine have also been shown to transfer electrons from bacteria to Fe(III) oxides (Doong and Schink, 2002). Studies of the role of electron shuttles have been limited to defined organic chemicals such as AQDS or humic materials.

Chelating agents may also play a role in the reductive dissolution of Fe(III) minerals by either complexing biogenic Fe(II), preventing the passivation of the Fe(III) mineral surface with sorbed Fe(II) or by directly complexing Fe(III). The first mechanism was demonstrated with the addition of the synthetic chelating agent, nitiloacetic acid (NTA), to goethite in the presence of S. alga BrY (Urrutia et al., 1999). No Fe(III) was detected in solution indicating that the enhanced reductive dissolution that was observed was due to removal of Fe(II) from the oxide surface, preventing passivation. In contrast, Nevin and Lovley (2002a) interpreted the high concentrations of Fe(III) in solution, in studies using microbially produced chelating agents with G. fermentans and a non-crystalline Fe(III) oxide, as evidence of the direct removal of Fe(III) from the oxide surface by the chelating agent. As with electron shuttling compounds, this mechanism eliminates the need for the organism to make direct contact with the oxide surfaces. Reductive dissolution of Fe(III) in smectite clays in the presence of S. putrefaciens MR-1 was increased by a factor of two with the addition of NTA, with only minor enhancement of dissolution with the addition of naturally occurring complexing agents (oxalate or malate) (Kostka et al., 1999). The addition of soil humic acid increased the extent of biotic dissolution of goethite in the presence of S. alga BrY (Roden and Urrutia, 1999). These authors acknowledged that the added humic acid could act as both a chelator and an electron shuttle.

The reducing environment induced by the addition of the various electron donors to the OU5 microcosm studies also resulted in the release of arsenic in concentrations that exceeded drinking water limits (10 µg As L–1) (Dupont et al., 2003). Researchers have noted the bacterially induced, simultaneous release of As(V) and Fe(II) to solution in laboratory experiments using a Fe-As mineral (scorodite), As saturated sediments (Cummings et al., 1999), and As sorbed to ferrihydrite and goethite (Jones et al., 2000; Langner and Inskeep, 2000). These observations support the theory that As(V) is released to solution with the microbial mediated dissolution of Fe(III) minerals and, once in solution, As(V) can be reduced to As(III) by either microbial or chemical processes. Arsenic, however, can also be directly mobilized from aquatic sediments through microbial respiration of arsenic (Ahmann et al., 1997).

High levels of bioavailable iron in the aquifer solids at OU5 may be inhibiting TCE degradation (Dupont et al., 2003). Also, the health risk associated with the concurrent release of As may limit the application of this remediation technique at this site. Therefore the present study was designed to identify the forms of iron, as defined by chemical extraction procedures, which are available to the native microbial community with the addition of various carbon and energy sources. The bioavailability of iron depends on iron mineralogy, the community of bacteria present under given site conditions, and the overall microbial ecology of the system. Carbon source type will influence the ecosystem including metabolic product formation, pH changes, and the rate of bacterial growth. The formation of electron shuttles and/or chelating compounds during carbon source metabolism is a critical controller on the nature and extent of Fe(III) transformation in natural soil environments, yet has not been reported on in the literature. This lead to the following objectives of this study: (i) to assess the extent of release of Fe(II) from aquifer solids with the addition of various carbon donors, focusing specifically on the role that fermentation products play in enhancing iron bioavailability as they function as electron shuttles and/or chelating agents in aquifer systems; (ii) to identify the iron minerals that are bioavailable under various organic compound loading conditions using chemical extraction procedures; and (iii) to evaluate the relationship between iron and arsenic release in response to various carbon loading conditions.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Hill Air Force Base (HAFB) is located in northern Utah, approximately 40 km north of Salt Lake City. Operable units (OUs) have been identified for site investigation and remediation at HAFB. These OUs are defined based on geology and type and source of contamination. Aquifer solids used in this study were from OU5 that is located on the west side of the Base and has ground water contamination associated with historical open pit disposal of chlorinated solvents. Aquifer solids were collected from the smear zone, which ranged from 0.6 m above to 1.5 m below the water table; 3 to 5 m below ground surface. Ground water was collected from an extraction system used to control ground water migration over a small area of the OU5 plume. The aquifer solids were 760 g kg–1 sand and 240 g kg–1 clay. The EC of the solids was 0.6 dS m–1, the cation exchange capacity was 6.5 cmol kg–1, the total S concentration was 90 mg kg–1, and the organic carbon content was 3 mg kg–1. Selected metals concentrations were (mg kg–1) As, 2.55; Ca, 32000; Mg, 5500; and Fe, 6100. The ground water pH was 7.8 and the electrical conductivity was 0.9 dS m–1. Ground water concentrations of selected chemical species were (mg L–1) organic carbon, 2; NO3––N, 0.5; SO42–, 60; Fe, <0.02; and alkalinity (as CaCO3), 300. The arsenic concentration was 2.0 µg L–1.

Characterization of Iron in the Sediment
To determine the association of iron with the solid phases a six-step sequential extraction was performed on the original OU5 aquifer solids using a modification of the procedure of Amacher (1998). A solution of 0.025 M calcium chloride was used as the first extractant as recommended by McLaren and Crawford (1973). The solids (1 g) were serially extracted in 50-mL centrifuge tubes with (i) 20 mL of 0.025 M calcium chloride; (ii) 25 mL of 1 M acetate buffer; (iii) 35 mL of 0.1 M sodium pyrophosphate; (iv) 25 mL of 0.01 M hydroxylamine hydrochloride (pH 2); (v) 25 mL of 0.25 M hydroxylamine hydrochloride + 0.25M HCl; and (vi) 10 mL of 0.2 M ammonium oxalate + 0.2 M oxalic acid + 5 mL of 0.1 M ascorbic acid. The fractions removed with these extractants are operationally defined as metals associated with (i) surface exchangeable sites; (ii) carbonates; (iii) organic matter; (iv) Mn oxides; (v) non-crystalline Fe oxides; and (vi) crystalline Fe oxides, respectively. The difference between total Fe and the sum of these fractions is defined as the residual fraction. After each extraction step the extracting solution was decanted and the supernatant filtered through a 0.45-µm filter. The centrifuge tubes were weighed between each extraction step to determine the mass of extracting solution left on the solids. The mass of iron extracted was corrected for the Fe associated with this residual solution from the previous extractant. Analysis of Fe was performed using flame atomic absorption spectroscopy (Aanlyst 800; PerkinElmer, Wellesley, MA).

A separate 1-g portion of the solid was extracted with 0.5 M HCl. This extract was analyzed for Fe(II) using the ferrozine procedure (Lovley and Phillips, 1986). Another separate 1-g sample of the sediment was extracted with 25 mL of 0.25 M hydroxylamine hydrochloride + 0.25 M HCl. Microbially reducible Fe(III), as defined by Lovley and Phillips (1987), was determined by subtracting the Fe(II) extracted with 0.5 M HCl from the Fe extracted with the hydroxylamine hydrochloride + HCl solution (HA-HCl minus HCl).

Microcosm Preparation
Microcosms used for the evaluation of Fe reduction were prepared using 15-mL glass culture tubes with rubber stoppers. Each culture tube contained 3 g dry weight of OU5 aquifer material and 9 mL of site ground water. Three soluble carbon donors [whey, hydrogen releasing compound (HRC) (Regenesis, San Clemente, CA), and lactic acid] and three low solubility carbon donors (low melting point vegetable oil, high melting point vegetable oil, and coconut oil) were evaluated. These are the same carbon donors and dosing rates used in the study by Dupont et al. (2003). The soluble donors were added directly into the microcosms at a nominal dose of 100 or 1000 mg C L–1 liquid volume. To minimize the formation of a separate oil phase in the reactors, the low solubility vegetable oils were added by first coating a washed, fine sand with the individual donors at a concentration of 1 mg donor g–1 inert media and then adding the coated sand to the microcosms to yield a nominal dose of 150 mg C L–1 liquid volume. Each microcosm was amended with 50 mg L–1 yeast extract. Both biotic (aquifer solids with no carbon addition) and abiotic (aquifer solids with carbon addition to reactors containing autoclaved aquifer solids and 0.2-µm filtered sterilized OU5 ground water) controls were carried through the microcosm study. All microcosms were constructed and stored statically at 15°C in an anaerobic glove bag (Coy, Grass Lake, MI) that was constantly purged (1.5 volumes d–1) with a pure nitrogen gas stream.

Sampling and Analytical Methods
Triplicate microcosms of each treatment were sacrificed for analysis at various time intervals from 0 to 140 d. The solution phase was filtered (0.45 µm) in the glove bag and immediately analyzed for Fe(II) using the ferrozine procedure (Lovley and Phillips, 1986). The pH was also recorded. The remaining solution was split between two vials, one containing HCl and the other containing HNO3. The solutions in HCl were immediately (within 24 h) analyzed for As(III) using hydride generation atomic absorption spectroscopy (Aanlyst 800). The solution with nitric acid was analyzed for total As [As(V) + As(III)] using graphite furnace atomic absorption spectroscopy.

The analysis of Fe(II) in solution provides information on soluble Fe(II) produced during incubation of the sediment with various electron donors. Biogenic Fe(II), however, may also be associated with sediment solid phases either as an adsorbed species or as a precipitate. The amount of biogenic Fe(II) associated with solids was determined using the 0.5 M HCl extraction with analysis of Fe(II) using ferrozine, as described above. Post incubation amorphous Fe oxides and crystalline Fe oxides were determined on a separate 1 g of solid using serial extractions with hydroxylamine hydrochloride and oxalate/oxalic acid as described above. The distribution of Fe between the solid phases was compared with the original sediment. Aquifer solid extractions were performed after 42 d for the high dosing of soluble carbon donor, and 138 d for the low dosing of the soluble and low solubility donors since these donors stimulated slower reactivity.

Estimating Maximum Iron(II) and Total Arsenic Release from the Aquifer Solids
To determine the maximum amount of Fe(II) that could be released from the aquifer solids, glucose was used as the C and energy source. The addition of glucose (2000 or 20000 mg C L–1) provides an environment for the rapid growth of fermenting and respiring microbes that produced a mixture of metabolites, some of which may act as electron shuttling compounds, maximizing the reductive dissolution of Fe(III) minerals. The microcosms were set up as described above, except with the addition of 200, 2000, and 20000 mg C kg–1 as glucose. Samples were collected in triplicate at 7-d intervals for 21 d. Aqueous samples were analyzed for Fe(II), pH, and As species as described above. Aquifer solids were extracted on Day 21. A single dosing of acetate (2000 mg C L–1 as calcium acetate) was included in this study since, as a product of fermentation, acetate is likely to be an important electron donor under Fe(III)-reducing conditions (Lovley, 2000; Coates et al., 1996). These microcosms were only evaluated for end of study (36 d) biogenic Fe(II) and extractable Fe.

Statistical Analysis
All measurements were analyzed with Statview 4.01 using the general linear model for analysis of variance (Abacus Concepts, 1992). When the factors or other interactions in the ANOVA were determined to be significant ({alpha} = 0.05), multiple comparison analyses were performed using the Bonferroni–Dunn procedure ({alpha} = 0.05/m, where m = number of comparisons). This procedure was chosen since the model is conservative for false positive comparisons.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS
 REFERENCES
 
Iron(II) Release from Aquifer Solids Incubated with Carbon Donors
The abiotic microcosms contained background concentrations of Fe(II) (<0.02 mg L–1) and arsenic (2.4 ± 0.21 µg As L–1) for all electron donors over all time intervals. The observed reactivity of iron and arsenic was therefore due to biological activity and not abiotic, chemical reactions.

For the biotic reactors, the release of Fe(II) to solution followed a similar pattern with time regardless of soluble electron donor type and concentration, as displayed for the soluble donors at 1000 mg C L–1 (Fig. 1). The pattern displayed was (i) a delay in onset of Fe(II) release to solution that lasted 10 to 15 d; (ii) Fe(II) release reached a maximum between 50 and 77 d; and (iii) a decrease in Fe(II) in solution due to either sorption and/or precipitation of the Fe(II). For the low solubility electron donor (150 mg L–1), again there was a delay in onset of Fe(II) to solution followed by a rapid rise in solution concentration. A decrease in Fe(II) in solution was not observed since the concentration of Fe(II) remained relatively low (Table 1).


Figure 1
View larger version (17K):
[in this window]
[in a new window]
  		Fig. 1. Iron(II) release from aquifer solids incubated with soluble carbon donor (1000 mg C L–1) and no carbon addition. Least significant difference (lsd) by the Bonferroni–Dunn procedure.

 

View this table:
[in this window]
[in a new window]
  		Table 1. Maximum release of Fe(II) to solution (average ± 95% CI).{dagger}

 
With dosing at 1000 mg C L–1, the pH dropped to a low of 6.2 for HRC and whey but increased over time to the original pH of 7.8. The rise in soluble Fe(II) coincided with the lower pH values. With 1000 mg C L–1 lactic acid and the low dosing of the soluble C donors, the same pattern of increasing Fe(II) solution concentration with time was observed, yet the pH of these solutions remained between 7.5 and 7.8 over the time course of this study. Changes in pH alone do not explain the observed increase and decrease in Fe(II) solution concentration.

The maximum amount of Fe(II) release was dependent on donor type, as is illustrated for the 1000 mg C L–1 dosing with the soluble donors (Fig. 1). Hydrogen releasing compound (1000 mg C L–1) sponsored the most Fe(II) release to solution, with the maximum concentration of Fe(II) being the same for whey and lactic acid treatments (Table 1). The concentration of soluble donor also affected Fe(II) release, with 16 times (for HRC addition) and 30 times (for whey or lactic acid addition) less Fe(II) in solution with the low dosing of 100 mg C L–1 compared to the 1000 mg C L–1 dose level (Table 1). The maximum amount of Fe(II) released with low solubility oils (150 mg C L–1) and the low dosing (100 mg C L–1) of whey or lactic acid was not statistically different from the no carbon treatment.

Arsenic Release from Aquifer Solids Incubated with Carbon Donors
The release of total As [As(V) + As(III)] to solution followed a similar pattern with time regardless of carbon donor type and concentration as displayed for the soluble donors at 1000 mg C L–1 (Fig. 2). Unlike Fe(II), there was no delay for As release to solution. After varying time intervals dependent on the electron donor, the concentration of As in solution was invariant with time (p > 0.05), not showing the decline that was displayed for Fe(II) (Fig. 1 and 2). Once solution equilibrium conditions for As release were reached, the concentration of As in solution was not dependent on the type of donor at the lower carbon additions (Table 2). The addition of 1000 mg C L–1 of the soluble donors however promoted higher As release than with 100 mg C L–1, and whey at this dose produced more As than lactic acid and HRC. Of the As released to solution, 100% was as As(III) for all treatments with the exception of that using 1000 mg C L–1 whey.


Figure 2
View larger version (19K):
[in this window]
[in a new window]
  		Fig. 2. Total As release from aquifer solids incubated with soluble carbon donors (1000 mg C L–1) and no carbon addition. Least significant difference (lsd) by the Bonferroni–Dunn procedure.

 

View this table:
[in this window]
[in a new window]
  		Table 2. Total As [As(V) + As(III)] and As(III) averaged over time once equilibrium conditions were reached (average ± 95% CI).{dagger}

 
Differences in the patterns of release of Fe(II) and As (Fig. 1 and 2) in this study indicate the independent reactivity of As. To further investigate the apparent lag time for Fe(II) release not observed for As, the release of Fe(II) and As to solution using the high dosing of the soluble donors was plotted over the first 8 d of the study (Fig. 3). Since Fe(II) concentration is in mg L–1 and As is in µg L–1, the data were normalized by dividing Ct, the concentration of Fe(II) or As at time equals t, by Cmax, the maximum concentration of Fe(II) or As released to solution with each treatment (Tables 1 and 2). The release of As to solution over the initial 8 d of the study was extensive, accounting for 20 to 35% of the total amount of As released over the course of the study. Iron(II) solution concentration remained low over the 8-d period, with less than 4% of the maximum release of Fe(II) occurring over this time period (Fig. 3).


Figure 3
View larger version (23K):
[in this window]
[in a new window]
  		Fig. 3. Iron(II) and As release from aquifer solids incubated over the first 8 d of the study. Concentrations at each sampling time have been normalized to the maximum Fe(II) or As release to solution during the experiment (Tables 1 and 2). Error bars are standard deviation of triplicate microcosms.

 
There was also As release from biologically active aquifer sediments with no added carbon (14.7 ± 5.3 µg L–1). The release of As to solution with no added carbon cannot be attributed to chemical processes since the abiotic controls remained at the background concentration of 2.4 ± 0.21 µg As L–1 throughout the study. The amount of As released without added carbon donor was statistically lower than the concentration of As in solution with any of the carbon donors. This is, however, evidence that bacteria were utilizing native organic material under anaerobic conditions, resulting in the release of As to solution. The maximum amount of As released without carbon addition was 35% of the maximum As released by treatment with HRC or lactic acid. The maximum Fe(II) in solution with the no carbon treatment was only 2.6% of the amount with the addition of whey or lactic acid. These differences in the extent of release of As and Fe(II) to solution in no carbon controls are well illustrated in Fig. 3.

Extraction of the Aquifer Solids
Sequential extractions are used to explore the association of a metal with the aquifer solid phase. Although there are limitations to the use of these procedures, including the non-specificity of the extractions and potential readsorption of the dissolved metal to another surface during the sequential extraction steps, the methods do provide an indication of the strength of association of the metal with various surfaces and mineral solid phases. Iron in the original OU5 aquifer solids was mostly associated with the extract for crystalline Fe oxides and with the residual fraction (Fig. 4).


Figure 4
View larger version (10K):
[in this window]
[in a new window]
  		Fig. 4. Distribution of Fe in the original aquifer solids. The fractions removed with these extractants are operationally defined as iron associated with (1) surface exchangeable sites, (2) carbonates, (3) organic matter, (4) Mn oxides, (5) non-crystalline Fe oxides, (6) crystalline Fe oxides, and (7) the residual fraction. Error bars are 95% CI of triplicate extractions.

 
Microbially reducible Fe(III), as defined by Lovley and Phillips (1987) [HA-HCl minus HCl extractable Fe(II)], accounted for 11.4% of the total iron in the original aquifer solids, or 696 mg Fe kg–1. If this entire mass went into solution in the test microcosms, the solution concentration would be 232 mg Fe(II) L–1. The measured concentration of Fe(II) in solution never exceeded this calculated concentration (Table 1). The biogenic Fe(II) produced in the microcosms was, however, the sum of Fe(II) in solution and Fe(II) extracted with 0.5 M HCl (Table 3). With the exception of the high melting point oil, more Fe(II) was produced than predicted from the chemical extraction procedure defining amorphous Fe(III) oxides, the assumed bioavailable fraction of sediment.


View this table:
[in this window]
[in a new window]
  		Table 3. Biogenic Fe(II) released in microcosms (average ± 95% CI). The microbially reducible Fe(III) in the aquifer solids was determined to be 696 ± 52 mg Fe(II) kg–1.

 
After 42 d of incubation with the 1000 mg C L–1 soluble C donors, an average of 16 ± 4% of the original Fe was released as Fe(II) to solution or was sorbed to or precipitated with HCl soluble minerals (Table 3 and Fig. 5). These minerals may include Fe(III) carbonate and oxides as sorbing surfaces for Fe(II) or newly formed Fe(II) containing minerals such as Fe(II) sulfides and Fe(II) carbonate (Heron et al., 1994; Lovley and Phillips, 1986). Incubation of the aquifer solids with these soluble carbon donors promoted the reduction of Fe(III) to Fe(II) with a shift in the distribution of the iron from the amorphous Fe oxides (HA-HCl minus HCl extractable), and for the whey treatment the crystalline Fe oxides (oxalate extractable), to HCl extractable iron solids. With the low solubility oils and lower dosing of the soluble C donors, there were significant decreases in Fe associated with amorphous Fe(III) oxides compared with the original sediment for treatments with lactic acid (94% decrease), HRC (54% decrease), and coconut oil (100% decrease).


Figure 5
View larger version (21K):
[in this window]
[in a new window]
  		Fig. 5. Distribution of Fe in aquifer solids incubated with soluble donors at 1000 mg C L–1 for 42 d compared with aquifer solids before incubation. Columns with the same letter within an extraction type are not significantly different by the Bonferroni–Dunn procedure. There was no measurable Fe(II) in the solution phase or extracted with HCl for the original aquifer solids before treatment.

 
Maximum Iron(II) and Arsenic Release from Aquifer Solids Incubated with Glucose
The OU5 aquifer solids were also incubated over a 21-d period with 200, 2000, and 20000 mg C L–1 as glucose. With the highest carbon addition, 1200 mg Fe(II) L–1 and 380 µg L–1 As were released to solution (Fig. 6).


Figure 6
View larger version (28K):
[in this window]
[in a new window]
  		Fig. 6. Iron(II) release from aquifer solids incubated with no carbon addition and glucose at 200, 2000, and 20 000 mg C L–1. Least significant difference (lsd) by the Bonferroni–Dunn procedure.

 
Total biogenic Fe(II) for the high glucose dose reached 3900 mg kg–1, 64% of the total solids Fe (Fig. 7). With the higher dosing of glucose, iron was removed from not only the amorphous Fe oxides but also from the crystalline Fe oxides, with 35 and 76% of the crystalline Fe in the original sediment being removed with 2000 and 20 000 mg C L–1, respectively. The addition of 2000 mg C L–1 as acetate led to production of 300 mg kg–1 biogenic Fe(II), whereas an equivalent loading of glucose produced 1500 mg kg–1 Fe(II). With acetate, deletion of Fe from the amorphous Fe oxide phase was observed.


Figure 7
View larger version (24K):
[in this window]
[in a new window]
  		Fig. 7. Distribution of Fe in aquifer solids after 21 d of incubation with glucose at 200, 2000, and 20 000 mg C L–1 and acetate at 2000 mg C L–1 compared with aquifer solids before incubation. Columns with the same letter within extraction type are not significantly different by the Bonferroni–Dunn procedure. There was no measurable Fe(II) in the solution phase or extracted with HCl for the original aquifer solids before treatment.

 
Distribution of Biogenic Iron(II) and Carbon Type
The effect of carbon donor concentration on the total biogenic Fe produced, and the distribution of Fe(II) between the solid and solution phases is illustrated with the glucose and acetate microcosms (Fig. 8). Not only was there an affect on total biogenic Fe(II) with increasing carbon dose for glucose, but the distribution of the biogenic Fe(II) also shifted from all solid phase associated with the lowest dosage of glucose, to 75% being in the solution phase at the highest dose of 20000 mg C L–1. The use of acetate as a carbon donor at 2000 mg C L–1 illustrates that the total reactivity and distribution of Fe(II) between the solid and solution phases is not exclusively concentration dependent. Acetate yielded the lowest production of Fe(II) of all the carbon donors tested, with 94% of the biogenic Fe(II) generated from this donor associated with the solid phase. The concentration of carbon donor added also affected system pH. The initial sediment–ground water mixture had a pH of 7.8. After incubation, the pH of the microcosms were 5.6, 6.5, 7.8, and 7.4 for the highest to lowest dosing of glucose, and acetate, respectively. The decrease in pH may have prevented the precipitation of Fe(II) minerals or sorption of Fe(II) to Fe(III) minerals, accounting for the higher proportion of Fe(II) in solution with higher dosing.


Figure 8
View larger version (17K):
[in this window]
[in a new window]
  		Fig. 8. Biogenic Fe(II) produced by different dosing of glucose and acetate. Error bars are 95% confidence intervals on total biogenic Fe(II).

 
For microcosms with the addition of low solubility oils (150 mg C L–1) or the low dose lactic acid, HRC, or whey (100 mg C L–1), 1 to 3% of the Fe(II) produced was associated with the solution phase with the rest of the biogenic Fe(II) associated with 0.5 M HCl extractable solid phase (Fig. 9). With 1000 mg C L–1 of the soluble donors, the percentage of biogenic Fe associated with the solution phase increased.


Figure 9
View larger version (21K):
[in this window]
[in a new window]
  		Fig. 9. Biogenic Fe(II) produced by a variety of carbon donors (100–1000 mg C L–1). Error bars are 95% confidence intervals on total biogenic Fe(II).

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS
 REFERENCES
 
Iron Release
Unlike other TEAs, Fe(III) occurs within the structure of minerals of varying water solubilities. Some bacteria utilize Fe(III) by direct attachment. If this was the only mode for bacteria to access Fe(III), bioavailable iron would be limited to water soluble Fe minerals and Fe associated with mineral surfaces. Some investigators have assumed that it is the amorphous Fe minerals that are bioavailable due to their large surface area and high water solubility compared with crystalline minerals. The amorphous Fe oxides present in the OU5 sediment, however, accounted for only 11.4% of the total iron pool in the aquifer solids, yet up to 64% of the total Fe was reduced to Fe(II) during these experiments with the highest dose of glucose. Even with moderate additions of electron donor (1000–2000 mg C L–1) using a number of carbon donors (Table 3), the amount of Fe(II) in solution and associated with acid soluble Fe(II) minerals exceeded the Fe associated with the amorphous Fe(III) oxides.

Some DIRB can access Fe(III) in crystalline oxides or clays through the use of electron shuttling compounds that are produced by some species of bacteria. Alternatively, DIRB may utilize chemicals in the sediments, such as humic acids, for transferring electrons from the organism to Fe in crystalline structures. The observed difference in Fe release with the various carbon sources may be due to the presence of electron shuttles that are differentially generated depending on the patterns of growth and composition of microbial communities that develop in response to the type of carbon used. The addition of glucose (2000 or 20000 mg C L–1) provided an environment for the rapid growth of fermenting microbes that produced a mixture of metabolites, some of which appear to act as electron shuttling compounds. The reductive dissolution of 64% of the total Fe in the sediment, including dissolution of 76% of the crystalline Fe oxides with the highest glucose dose, cannot be explained by any other mechanisms than the production of electron shuttling compounds as by-products of metabolism. The pH of this system was depressed due to microbial activity (to 5.6). A two-fold decrease in pH would cause a million-fold increase in the activity of Fe(III) in equilibrium with Fe oxides (soil-Fe) (Lindsay, 1979). This dissolved Fe(III) would be available for reduction to Fe(II). The activity of Fe(III) dissolved from the oxides due to pH changes alone is, however, sub part per billion and does not explain the high concentration of Fe(II) in solution.

Not only did the total amount of biogenic Fe(II) increase at higher glucose doses, but the proportion of Fe(II) in solution also increased. The proportionally higher solution concentration of Fe(II) at higher doses may be explained by decreasing pH values as a function of glucose concentration that would prevent Fe(II) from precipitating or sorbing to surfaces, and to the presence of chelating agents that also increase with carbon dose.

Although there was an initial decrease in pH with the addition of 1000 mg C L–1 HRC and whey, at the time of sampling (42 d) the pH had increased to the original value of 7.8. There were, therefore, no differences in pH between high and low dosing of HRC, whey, and lactic acid, nor was the total biogenic Fe(II) consistently, statistically different between the two doses for these carbon sources (Fig. 9). However, the proportion of Fe(II) in solution increased from a few percent up to 32% for these donors, indicating that the increase in the proportion of Fe(II) in solution with higher donor addition was due to the production of chelating agents that complexed with the biogenic Fe(II), preventing sorption or precipitation of Fe(II) to the HCl soluble solid phases. For all treatments there was no measurable Fe(III) in solution, indicating no solubilization of Fe(III) minerals due to pH effects nor complexing of Fe(III) by chelating agents.

With 2000 mg C L–1 as glucose, 24% of the total Fe underwent reductive dissolution. At the same concentration of carbon using calcium acetate, less than 5% Fe reduction occurred over the 36-d study. Acetate is a major product of fermentation and is metabolized to carbon dioxide, coupled with the reduction of Fe(III), by DIRB of the family Geobacteraceae (Lovley, 2000). Members of this family have been the most commonly recovered DIRB from sediments and subsurface environments. Lovley (2000) proposed the cooperative activity of fermentative microbes (producing acetate) and Geobacter [utilizing acetate and Fe(III)] for the degradation of organic matter in anoxic environments. The addition of acetate directly to the microcosms, however, resulted in the lowest production of Fe(II) of the carbon donors tested. Without the complex mixture of metabolic products present with the fermentation of glucose that acted as electron shuttles and chelators, the DIRB in the microcosms with acetate addition appear only capable of accessing Fe(III) by direct attachment.

Hydrogen release compound (HRC), which is a slow release lactate carbon donor, and lactic acid produced the same amount of biogenic Fe(II). Lactate is utilized as an electron donor by Geobacter species (Lovley, 2000; Coates et al., 1996), by other DIRB such as Shewanella species (Lovley, 2000), and by Rhodoferax ferrireducens (Finneran et al., 2003). Lactate is a simple fatty acid and, like acetate, its catabolism is less likely to result in high concentrations of diverse types of metabolic products than a substrate such as glucose. Yet unlike acetate, microcosms treated with lactate or HRC (1000 mg C L–1) produced statistically the same amount of biogenic Fe(II) as with whey (1000 mg C L–1) and glucose (2000 mg C L–1). All treatments produced more Fe(II) than was associated with non-crystalline Fe oxides.

Arsenic Release
Arsenic, as both arsenate and arsenite, is primarily associated with iron oxides in the environment. Studies reported in the literature indicate that As(V) release to solution is a result of the reductive dissolution of Fe(III) oxides with sorbed or structural As (Cummings et al., 1999). Simultaneous release of Fe(II) and As(V) with the subsequent reduction of arsenic to As(III) has been used as evidence of the dependence of As release on dissolution of iron minerals. In the present study, however, the behavior of As did not follow that of Fe(II). The following three observations would indicate that the reduction and solubilization of As is independent of the reductive dissolution of Fe(III) oxides: (i) the initial release of As was independent of Fe(III) dissolution (Fig. 3); (ii) As was released to solution under anaerobic conditions with the native organic matter, whereas only small amounts of Fe(II) were released to solution with the no carbon control (Fig. 3); and (iii) As(III) accounted for 100% of the total As in solution at all sampling times, except when whey was used as the carbon source. Langner and Inskeep (2000) suggested that the depletion of As(V) in solution due to its reduction to As(III) would result in the continued desorption of As(V) from solid surfaces, a process independent of Fe(III) mineral dissolution. These authors conducted a study where As sorbed to ferrihydrite was incubated with an As reducing bacteria (Clostridium CN8) that did not reduce iron. The bacteria rapidly reduced the As(V) in solution but the depletion of this solution phase As(V) did not result in the solubilization of As from the non-reacting ferrihydrite, leading to the conclusion that this proposed mechanism was of minor importance compared with the reductive dissolution of Fe(III) minerals. However, Ahmann et al. (1997) provided evidence of direct mobilization of As from sediments by dissimilatory arsenic reducing bacteria (DARB), a mechanism supported by the results of the present study. Studies supporting the dependence of As release on the reductive dissolution of Fe(III) minerals have used As sorbed onto laboratory prepared Fe(III) minerals (ferrihydrite, goethite, FeAsO4) in the presence of Shewanella alga, a DIRB (Cummings et al., 1999), an As reducing bacteria enriched from soil with naturally elevated levels of As (Jones et al., 2000; Langner and Inskeep, 2000), and an unenriched extraction of microbes from the same soil (Jones et al., 2000). In the present study, native bacteria including fermenters, DIRB, and perhaps DARB were stimulated with the addition of various carbon donors to provide a complex interplay in the biogeochemistry of iron and arsenic not observable in highly controlled laboratory studies.


   CONCLUSIONS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
REFERENCES
 
The release of Fe(II) and arsenic to solution was mediated by microbial activity. The extent of dissolution of Fe(III) minerals was related to the amount and type of carbon donor added. Apparently, the type of electron donor used influences the structure of the community of bacteria that developed within the microcosms, which in turn determined the types and concentrations of metabolic products. After incubation of the HAFB, Utah, OU5 sediments with acetate as the carbon source, 300 (±28) mg Fe(II) kg–1 were produced through microbial activity. With the same level of carbon addition using glucose, the biogenic Fe(II) concentration was 1500 ± 300 mg Fe(II) kg–1. With carbon sources that lead to a complex mixture of metabolic products, some of the metabolites appear to be used as electron shuttling chemicals and as chelating agents. The electron shuttling compounds facilitate the reductive dissolution of even crystalline Fe(III) oxides. Without the presence of electron shuttling chemicals to transport electrons from the bacteria onto and into Fe(III) minerals, only surficial Fe that can be directly contacted by the bacteria is bioavailable, as illustrated when acetate was used as the C donor. The presence of chelating compounds with the higher doses of complex carbon sources resulted in a higher proportion of the biogenic Fe(II) being held in solution as compared to a simple substrate such as acetate.

Regardless of carbon donor type and concentration, As concentrations in the water exceeded drinking water standards. The release of As to solution was independent of the reductive dissolution of Fe(III) minerals contrary to the majority of other studies, which have shown As is solubilized as Fe(III) oxides dissolve under reducing conditions. The As dissolution observed during this study appears to have been the result of the direct use of As as an electron acceptor by dissimilatory arsenic reducing bacteria.

Care should be taken in the selection of an electron donor for biostimulation to promote reductive dechlorination of chlorinated solvents, as many have the potential to be used in competitive terminal electron accepting processes as was shown in this study. For the HAFB OU5 sediment, the selection of electron donor greatly influenced the extent of reductive dissolution of Fe(III) minerals, including those in crystalline form. In addition to highly favoring iron reduction in the OU5 sediment, possibly making them ineffective for the stimulation of TCE reductive dechlorination, most of the electron donors evaluated in this study resulted in the stimulation of direct use of arsenic as a TEA, potentially causing additional contamination of the aquifer produced as a by-product of the biostimulation process. The magnitude of the competitive bioavailable iron pool, and the extent of arsenic release that would result from biostimulation in the OU5 sediment were not predicted accurately using conventional sediment chemical extraction techniques. It is recommended that limited bioassay procedures like those described in this paper be conducted on site aquifer solids to screen an array of electron donors before full-scale biostimulation, to ensure the optimal use of the added organic compound for dechlorination purposes, and to prevent unintended ground water contamination from competing electron accepting processes.


   REFERENCES
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 REFERENCES
 





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 Similar articles in this journal
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 Google Scholar
Google Scholar
Right arrow Articles by McLean, J. E.
Right arrow Articles by Sorensen, D. L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by McLean, J. E.
Right arrow Articles by Sorensen, D. L.
GeoRef
Right arrow GeoRef Citation
Agricola
Right arrow Articles by McLean, J. E.
Right arrow Articles by Sorensen, D. L.
Related Collections
Right arrow Biogeochemical Processes
Right arrow Bioremediation and Biodegradation
Right arrow Water Pollution

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