Article pubs.acs.org/est
Isolation and Microbial Reduction of Fe(III) Phyllosilicates from Subsurface Sediments Tao Wu,† Evgenya Shelobolina,† Huifang Xu,† Hiromi Konishi,† Ravi Kukkadapu,‡ and Eric E. Roden*,† †
Department of Geoscience, University of Wisconsin−Madison, Madison, Wisconsin 53706, United States Pacific Northwest National Laboratory, Richland, Washington 99354, United States
‡
S Supporting Information *
ABSTRACT: Fe(III)-bearing phyllosilicates can be important sources of Fe(III) for dissimilatory microbial iron reduction in clay-rich anoxic soils and sediments. The goal of this research was to isolate Fe(III) phyllosilicate phases, and if possible, Fe(III) oxide phases, from a weathered shale saprolite sediment in order to permit experimentation with each phase in isolation. Physical partitioning by density gradient centrifugation did not separate phyllosilicate and Fe(III) oxide phases (primarily nanoparticulate goethite). Hence we examined the ability of chemical extraction methods to remove Fe(III) oxides without significantly altering the properties of the phyllosilicates. XRD analysis showed that extraction with acid ammonium oxalate (AAO) or AAO in the presence of added Fe(II) altered the structure of Fe-bearing phyllosilicates in the saprolite. In contrast, citrate-dithionite-bicarbonate (CDB) extraction at room temperature or 80 °C led to minimal alteration of phyllosilicate structures. Reoxidation of CDB-extracted sediment with H2O2 restored phyllosilicate mineral d-spacing and Fe redox speciation to conditions similar to that in the pristine sediment. The extent of microbial (Geobacter sulfurreducens) reduction of Fe(III) phyllosilicates isolated by CDB extraction and H2O2 reoxidation (16 ± 3% reduction) was comparable to what took place in pristine sediments as determined by Mossbauer spectroscopy (20 ± 11% reduction). These results suggest that materials isolated by CDB extraction and H2O2 reoxidation are appropriate targets for detailed studies of natural soil/sediment Fe(III) phyllosilicate reduction.
■
INTRODUCTION The kinetics of microbial Fe(III) oxide reduction has been studied extensively, and models are available to predict the reduction of Fe(III) oxides in sedimentary environments.1−5 Fe(III)-bearing phyllosilicates (referred to hereafter as Fe(III) phyllosilicates) are also subject to microbial reduction,6,7 and the importance of Fe(III) phyllosilicates as an electron acceptor for dissimilatory Fe(III)-reducing bacteria in sediments has been increasingly recognized.8,9 Microbial reduction of structural Fe(III) in model phyllosilicates, including smectite, nontronite (Fe-rich smectite), illite, and chlorite has been studied in detail.10−14 However, the kinetics of natural Fe(III) phyllosilicate reduction in sediments is poorly understood, and more information about the rate and extent of natural Fe(III) phyllosilicate reduction is required to develop reaction models that include both Fe(III) oxide and Fe(III) phyllosilicate components. Reduction of structural Fe(III) changes both the physical and chemical properties of phyllosilicates including particle flocculation, swelling pressure, surface area, and cation exchange capacity,6,15−17 all of which may have a strong impact on sediment geochemistry. Most naturally occurring phyllosilicates have compositions varying from the idealized formulas.18,19 Consequently, model minerals without significant amounts of impurities and with formulas close to ideal forms are usually used in experimental studies. There are, however, important differences between minerals found in clay fractions © 2012 American Chemical Society
of soils and sediments compared to model minerals in relatively pure deposits. Clay minerals from soils and sediments often have interlayered structures or occur in a mixed order of stacking in which an individual crystal may consist of two or more layer silicates.18 Many investigations indicate that soil smectite is altered to a 14-Å phase that contains positively charged hydroxyl-Al groups that reduce cation exchange capacity and surface area.19 Such features may significantly alter the ability of microorganisms to reduce structural Fe(III), and further experimentation with naturally occurring as opposed to model Fe(III) phyllosilicates is needed. The goal of this research was to isolate and conduct microbial reduction experiments with Fe(III) phyllosilicate phases from a clay-rich subsurface sediment. A key challenge in this pursuit was the separation of phyllosilicates and Fe(III) oxides in order to permit experimentation with them in isolation. The ideal separation technique would allow complete isolation of Fe(III) phyllosilicates without altering their physical and chemical properties. Both physical separation and wet chemical extractions were evaluated as methods for phyllosilicates isolation in subsurface sediments from the Area 2 site at the U.S. Department of Energy Field Research Center (FRC) Received: Revised: Accepted: Published: 11618
November 25, 2011 October 3, 2012 October 12, 2012 October 12, 2012 dx.doi.org/10.1021/es302639n | Environ. Sci. Technol. 2012, 46, 11618−11626
Environmental Science & Technology
Article
hour three consecutive times; and the CDB-80C extraction was carried out three times for 15 min. A separate AAO+Fe(II) time course extraction experiment was conducted, with sample collection at 4, 8, 16, and 24 h. The amount of Fe extracted was determined by Ferrozine analysis.33 The AAO-extracted solids were washed twice with deionized water. The AAO+Fe(II) and CDB-extracted solids were washed twice with the chelator (oxalate and citrate, respectively), followed by two washes with deionized water. The washed solids were freeze-dried and analyzed by XRD and TEM. Portions of the AAO+Fe(II) and CDB-extracted materials were treated with 3% H2O2 in order to oxidize any reduced structural Fe back to the +3 oxidation state.34 The total Fe and Fe(II) contents of the pristine and chemically extracted materials described above were determined by the HF-1,10-phenanthroline (HF-phenanthroline) method, which recovers Fe from both Fe(III) oxides and Fe(III)-bearing phyllosilicates.35 The HF-phenanthroline method used was a modification of the method described by Komadel and Stucki,35 in which hydroxylamine sulfate (rather than light) was used to reduce all Fe in the extract for determination of total Fe.36 The extent of Fe(III)-phyllosilicate reduction during AAO +Fe(II) and CDB extraction was determined by measuring the HF-phenanthroline extractable Fe(II) content of solids before and after extraction and anoxic washing (no reoxidation). The observed change in HF-phenanthroline extractable Fe(II) content was assumed to be the result of Fe(III) phyllosilicate reduction, as any Fe(II) produced during reductive dissolution of Fe(III) oxides was removed during the washing procedure. This quantity of Fe(II) was divided by the Fe(III) phyllosilicate content of the extracted sediment (see Table S2) to calculate the extent of reduction. This approach assumes that no phyllosilicate-associated Fe was lost during extraction (i.e., by dissolution), which was evaluated in other experiments where sediments that had been previously stripped of Fe(III) oxides (by CDB-80C) and reoxidized (by H2O2), were re-extracted with either CDB-80C or AAO+Fe(II). Microbial Reduction Experiments. Chemically extracted and pristine FRC Area 2 sediments were employed in microbial reduction experiments with the model Fe(III)-reducing bacterium Geobacter sulfurreducens.37,38 All experiments were conducted with washed, acetate/fumarate-grown cells in a PIPES-buffered (pH 6.8) growth medium containing 20 mM acetate as the electron donor and Fe(III) in sediments (10 g L−1, equivalent to ca. 1−4 mmol Fe(III) L−1) as the only electron acceptor. Fe(III) reduction was monitored by measuring the accumulation of HF-1,10-phenanthroline (HFphenanthroline) extractable Fe(II). Aqueous Fe in 0.2 μmfiltered samples was measured by ICP-OES. XRD, TEM, and Mö ssbauer Analyses. Oriented-aggregate mounts were prepared to identify phyllosilicates and other minerals by XRD. Specimens were mixed with a small amount of distilled water and sonicated for one hour. The suspension was gently transferred to a petrographic slide and air-dried. The (101) peak of quartz in all the XRD patterns was used as an internal calibrator. The samples were analyzed with a Scintag PADV X-ray Diffractometer (Cu Kα radiation). The Scintag diffractometer was operated at 45 kV, 40 mA and used a 2 mm divergence slit, 4 mm incident scatter slit, 1 mm diffracted beam scatter slit, and 0.5 mm receiving slit. A step size of 0.02° and a dwelling time of 2 s per step were used for collecting all the step-scan diffraction patterns.
located at Oak Ridge National Laboratory (ORNL) in Oak Ridge, TN. Following isolation, reduction of Fe(III) in the pristine sediment and isolated Fe(III) phyllosilicate phases was studied using the dissimilatory Fe(III)-reducing bacterium Geobacter sulf urreducens.
■
MATERIALS AND METHODS Subsurface Sediment. The FRC Area 2 site is a pathway for migration of uranium-contaminated groundwater to seeps in the upper reach of Bear Creek at ORNL.20 The sediments are comprised of weathered shale saprolite with abundant Fe(III) oxides (primarily nanocrystalline goethite with smaller amounts of hematite) and Fe(III) phyllosilicates (smectite, illite) in varying proportions.20 Materials from this and nearby locations at the FRC have been examined previously for their susceptibility toward microbial reduction by pure and mixed bacterial cultures,8,21−23 in a few instances specifically in the context of Fe(III) oxide vs Fe(III) phyllosilicate reduction.8,23 Core material from several intervals (25−50 cm length) between 3 and 8 and m depth were air-dried and combined. The dried sediment was ground with a mortar and pestle and passed through a 0.5 mm sieve prior to use in experiments. Some experiments employed clay size material (
