Search

Redox-reactive, biogeochemical phases generated by reductive microbial activity in hyporheic zone sediments from a dynamic groundwater–river interaction zone were evaluated for their ability to reduce soluble pertechnetate [ 99 Tc(VII)O 4 − ] to less soluble Tc(IV). The sediments were bioreduced by indigenous microorganisms that were stimulated by organic substrate addition in synthetic groundwater with or without sulfate. In most treatments, 20 μmol L −1 initial aqueous Tc(VII) was reduced to near or below detection (3.82 × 10 −9 mol L −1 ) over periods of days to months in suspensions of variable solids concentrations. Native sediments containing significant lithogenic Fe(II) in various phases were, in contrast, unreactive with Tc(VII). The reduction rates in the bioreduced sediments increased with increases in sediment mass, in proportion to weak acid-extractable Fe(II) and sediment-associated sulfide (AVS). The rate of Tc(VII) reduction was first order with respect to both aqueous Tc(VII) concentration and sediment mass, but correlations between specific reductant concentrations and reaction rate were not found. X-ray microprobe measurements revealed a strong correlation between Tc hot spots and Fe-containing mineral particles in the sediment. However, only a portion of Fe-containing particles were Tc-hosts. The Tc-hot spots displayed a chemical signature (by EDXRF) similar to pyroxene. The application of autoradiography and electron microprobe allowed further isolation of Tc-containing particles that were invariably found to be ca 100 μm aggregates of primary mineral material embedded within a fine-grained phyllosilicate matrix. EXAFS spectroscopy revealed that the Tc(IV) within these were a combination of a Tc(IV)O 2 -like phase and Tc(IV)–Fe surface clusters, with a significant fraction of a TcS x -like phase in sediments incubated with SO 4 2− . AVS was implicated as a more selective reductant at low solids concentration even though its concentration was below that required for stoichiometric reduction of Tc(VII). These results demonstrate that composite mineral aggregates may be redox reaction centers in coarse-textured hyporheic zone sediments regardless of the dominant anoxic biogeochemical processes.

Sediments with basaltic provenance, such as those at the Hanford nuclear reservation, Washington, USA, are rich in Fe-bearing minerals of mixed valence. These minerals are redox reactive with aqueous O2 or Fe(II), and have the potential to react with important environmental contaminants including Tc. Here we isolate, identify and characterize natural Fe(II)/Fe(III)-bearing microparticles from Hanford sediments, develop synthetic analogues and investigate their batch redox reactivity with aqueous Tc(VII). Natural Fe-rich mineral samples were isolated by magnetic separation from sediments collected at several locations on Hanford’s central plateau. This magnetic mineral fraction was found to represent up to 1 wt% of the total sediment, and be composed of 90% magnetite with minor ilmenite and hematite, as determined by X-ray diffraction. The magnetite contained variable amounts of transition metalsconsistent with alio- and isovalent metal substitutions for Fe. X-ray microprobe analysis showed that Ti was the most significant substituent, and that these grains could be described with the titanomagnetite formula Fe3_xTixO4, which falls between endmember magnetite (x = 0) and ulvo¨ spinel (x = 1). The dominant composition was determined to be x = 0.15 by chemical analysis and electron probe microanalysis in the bulk, and by L-edge X-ray absorption spectroscopy and X-raymore » photoelectron spectroscopy at the surface. Site-level characterization of the titanomagnetites by X-ray magnetic circular dichroism showed that despite native oxidation, octahedral Fe(II) was detectable within 5 nm of the mineral surface. By testing the effect of contact with oxic Hanford and Ringold groundwaters to reduced Ringold groundwater, it was found that the concentration of this near-surface structural Fe(II) was strongly dependent on aqueous redox condition. This highlights the potential for restoring reducing equivalents and thus reduction capacity to oxidized Fe-mineral surfaces through redox cycling in the natural environment. Reaction of these magnetically-separated natural phases from Hanford sediments with a solution containing 10 lmol L_1 Tc(VII) showed that they were able to reductively immobilize Tc(VII) with concurrent oxidation of Fe(II) to Fe(III) at the mineral surface, as were synthetic x = 0.15 microparticle and nanoparticle analogue phases. When differences in the particle surface area to solution volume ratio were taken into consideration, measured Tc(VII) reduction rates for Fe3_xTixO4(x = 0.15) natural material, synthetic bulk powder and nanoparticles scaled systematically, suggesting possible utility for comprehensive batch and flow reactivity studies.« less

Technetium is an important environmental contaminant introduced by the processing and disposal of irradiated nuclear fuel and atmospheric nuclear tests. Under oxic conditions technetium is soluble and exists as pertechnatate anion (TcO4−), while under anoxic conditions Tc is usually insoluble and exists as precipitated Tc(IV). Here we investigated abiotic Tc(VII) reduction in mineralogically heterogeneous, Fe(II)-containing sediments. The sediments were collected from a 55 m borehole that sampled a semi-confined aquifer at the Hanford Site, USA that contained a dramatic redox transition zone. One oxic facies (18.0–18.3 m) and five anoxic facies (18.3–18.6 m, 30.8–31.1 m, 39.0–39.3 m, 47.2–47.5 m and 51.5–51.8 m) were selected for this study. Chemical extractions, X-ray diffraction, electron microscopy, and Mossbauer spectroscopy were applied to characterize the Fe(II) mineral suite that included Fe(II)-phyllosilicates, pyrite, magnetite and siderite. The Fe(II) mineral phase distribution differed between the sediments. Sediment suspensions were adjusted to the same 0.5 M HCl extractable Fe(II) concentration (0.6 mM) for Tc(VII) reduction experiments. Total aqueous Fe was below the Feaq detection limit (

Experiments were performed herein to investigate the rates and products of heterogeneous reduction of Tc(VII) by Fe(II) adsorbed to hematite and goethite, and by Fe(II) associated with a dithionite–citrate–bicarbonate (DCB) reduced natural phyllosilicate mixture [structural, ion-exchangeable, and edge-complexed Fe(II)] containing vermiculite, illite, and muscovite. The heterogeneous reduction of Tc(VII) by Fe(II) adsorbed to the Fe(III) oxides increased with increasing pH and was coincident with a second event of Fe 2 + (aq) adsorption. The reaction was almost instantaneous above pH 7. In contrast, the reduction rates of Tc(VII) by DCB-reduced phyllosilicates were not sensitive to pH or to added Fe 2 + (aq) that adsorbed to the clay. The reduction kinetics were orders of magnitude slower than observed for the Fe(III) oxides, and appeared to be controlled by structural Fe(II). The following affinity series for heterogeneous Tc(VII) reduction by Fe(II) was suggested by the experimental results: aqueous Fe(II) ∼ adsorbed Fe(II) in phyllosilicates [ion-exchangeable and some edge-complexed Fe(II)] ≪ structural Fe(II) in phyllosilicates ≪ Fe(II) adsorbed on Fe(III) oxides. Tc-EXAFS spectroscopy revealed that the reduction products were virtually identical on hematite and goethite that were comprised primarily of sorbed octahedral TcO2 monomers and dimers with significant Fe(III) in the second coordination shell. The nature of heterogeneous Fe(III) resulting from the redox reaction was ambiguous as probed by Tc-EXAFS spectroscopy, although Mossbauer spectroscopy applied to an experiment with 56Fe-goethite with adsorbed 57Fe(II) implied that redox product Fe(III) was goethite-like. The Tc(IV) reduction product formed on the DCB-reduced phyllosilicates was different from the Fe(III) oxides, and was more similar to Tc(IV) oxyhydroxide in its second coordination shell. The heterogeneous reduction of Tc(VII) to less soluble forms by Fe(III) oxide-adsorbed Fe(II) and structural Fe(II) in phyllosilicates may be an important geochemical process that will proceed at very different rates and that will yield different surface species depending on subsurface pH and mineralogy.

The sorption of Co2+ (10−6 mol/L) was measured on subsurface mineral materials in the absence and presence of a sorbed leonardite humic acid (LHA) to 1. (1) evaluate the sorptive role of mineral-bound humic substances, and 2. (2) establish approaches to model metal ion binding in composite materials. The subsurface materials were a 1. (1) maximum LHA sorption occurred, and 2. (2) Co sorption to the mineral phase was weak and dominated by ion exchange. The LHA appeared simply to augment, rather than to change the intrinsic adsorption behavior of the mineral sorbents. Accordingly, predictions of the Kd for Co on the LHA-coated subsurface materials (DCP, CP) based on a linear additivity model agreed well with the experimental data, suggesting that the complex humic-mineral association acted as a noninteractive sorbent mixture at low aqueous Co concentrations.

The sorption of seven divalent metals (Ba, Sr, Cd, Mn, Zn, Co, and Ni) was measured on calcite over a large initial metal (Me) concentration range (10−8 to 10−4 mol/L) in constant ionic strength (I = 0.1), equilibrium CaCO3(s)-CaCO3(aq) suspensions that varied in pH. At higher initial Me concentrations (10−5 to 1−4 mol/L) geochemical calculations indicated that the equilibrium solutions were saturated with discrete solid phases of the sorbates: CdCO3(s), MnCO3(s), Zn5(OH)6(CO3)2(s), Co(OH)2(s), and Ni(OH)2(s), implying that aqueous concentrations were governed by solubility. However, significant sorption of all the metals except for Ba and Sr was observed at aqueous concentrations below saturation with Me-solid phases. Divalent metal ion sorption was dependent on aqueous Ca concentration, and the following selectivity sequence was observed: Cd > Zn ≥ Mn > Co > Ni > Ba = Sr. The metals varied in their sorption reversibility, which was correlated with the single-ion hydration energies of the metal sorbates. The strongly hydrated metals (Zn, Co, and Ni) were most desorbable. A sorption model that included aqueous speciation and Me2+-Ca2+ exchange on cation-specific surface sites was developed that described most of the data well. The chemical nature of the surface complex used in this model was unspecified and could represent either a hydrated or dehydrated surface complex, or a surface precipitate. A single exchange constant for Cd, Mn, Co, and Ni could describe the sorption of that metal over a wide range in pH, Ca concentration, and surface concentration. Zinc, however, exhibited nonlinear sorption behavior and required exchange constants that varied with surface coverage. Our data suggested that (i) Cd and Mn dehydrate soon after their adsorption to calcite and form a phase that behaves like a surface precipitate, and (ii) Zn, Co, and Ni form surface complexes that remain hydrated until the ions are incorporated into the structure by recrystallization.