Reduction of iron(III) minerals by natural organic matter in groundwater

Construction of the entrance tunnel to the Äspö Hard Rock Laboratory, a prototype repository in Sweden for research into the geological disposal of spent nuclear fuel, has resulted in increased transport of organic carbon from the surface into the groundwater. This increased input of organic matter has induced accelerated oxidation of organic carbon associated with reduction of iron(III) minerals as the terminal electron acceptor in microbial respiration. Hydrochemical modeling of major solute ions at the site indicates an apparent first-order decay constant for organic carbon of 3.7 ± 2.6/yr. This rapid turnover is not accompanied by an equivalent mobilization of ferrous iron. Thermodynamic calculation of iron mineral solubility suggests that ferrous clay minerals may form in hydraulically transmissive fractures. The conditional potentials for the oxidation–reduction of such phases coincide with measured redox potentials at the site. The calculated potential is sufficiently low so that such phases would provide reducing capacity against future intrusion of O₂ into the groundwater, thus buffering a repository against oxic corrosion of the engineered barriers.     

The ferrozine method revisited: Fe(II)/Fe(III) determination in natural waters

The original ferrozine method has been modified to sequentially determine the Fe(II)/Fe(III) speciation in small volumes of fresh and marine water samples, at the submicromolar level. Spectrophotometric analyses of the Fe(II)–ferrozine complex are performed on a single aliquot before and after a reduction step with hydroxylamine. The procedure is calibrated using Fe(III) standards stable under normal conditions of analysis. It is shown also that the presence of high concentrations of dissolved NOM (natural organic matter) do not create any significant artifacts. The method was used to measure Fe(II) and Fe(III) depth distribution in salt marsh pore waters and in a stratified marine basin.     

Influence of biogenic Fe(II) on the extent of microbial reduction of Fe(III) in clay minerals nontronite, illite, and chlorite

Microbial reduction of Fe(III) in clay minerals is an important process that affects properties of clay-rich materials and iron biogeochemical cycling in natural environments. Microbial reduction often ceases before all Fe(III) in clay minerals is exhausted. The factors causing the cessation are, however, not well understood. The objective of this study was to assess the role of biogenic Fe(II) in microbial reduction of Fe(III) in clay minerals nontronite, illite, and chlorite. Bioreduction experiments were performed in batch systems, where lactate was used as the sole electron donor, Fe(III) in clay minerals as the sole electron acceptor, and Shewanella putrefaciens CN32 as the mediator with and without an electron shuttle (AQDS). Our results showed that bioreduction activity ceased within two weeks with variable extents of bioreduction of structural Fe(III) in clay minerals. When fresh CN32 cells were added to old cultures (6 months), bioreduction resumed, and extents increased. Thus, cessation of Fe(III) bioreduction was not necessarily due to exhaustion of bioavailable Fe(III) in the mineral structure, but changes in cell physiology or solution chemistry, such as Fe(II) production during microbial reduction, may have inhibited the extent of bioreduction. To investigate the effect of Fe(II) inhibition on CN 32 reduction activity, a typical bioreduction process (consisting of lactate, clay, cells, and AQDS in a single tube) was separated into two steps: (1) AQDS was reduced by cells in the absence of clay; (2) Fe(III) in clays was reduced by biogenic AH₂DS in the absence of cells. With this method, the extent of Fe(III) reduction increased by 45-233%, depending on the clay mineral involved. Transmission electron microscopy observation revealed a thick halo surrounding cell surfaces that most likely resulted from Fe(II) sorption/precipitation. Similarly, the inhibitory effect of Fe(II) sorbed onto clay surfaces was assessed by presorbing a certain amount of Fe(II) onto clay surfaces followed by AH₂DS reduction of Fe(III). The reduction extent consistently decreased with an increasing amount of presorbed Fe(II). The relative reduction extent [i.e., the reduction extent normalized to that when the amount of presorbed Fe(II) was zero] was similar for all clay minerals studied and showed a systematic decrease with an increasing clay-presorbed Fe(II) concentration. These results suggest a similar inhibitory effect of clay-sorbed Fe(II) for different clay minerals. An equilibrium thermodynamic model was constructed with independently estimated parameters to evaluate whether the observed cessation of Fe(III) reduction by AH₂DS was due to exhaustion of reaction free energy. Model-calculated reduction extents were, however, over 50% higher than experimentally measured, indicating that other factors, such as blockage of the electron transfer chain and mineralogy, restricted the reduction extent. Another important result of this study was the relative reducibility of Fe(III) in different clays: nontronite > chlorite > illite. This order was qualitatively consistent with the differences in the crystal structure and layer charge of these minerals.     

Reductive biotransformation of Fe in shale-limestone saprolite containing Fe(III) oxides and Fe(II)/Fe(III) phyllosilicates

A <2.0-mm fraction of a mineralogically complex subsurface sediment containing goethite and Fe(II)/Fe(III) phyllosilicates was incubated with Shewanella putrefaciens (strain CN32) and lactate at circumneutral pH under anoxic conditions to investigate electron acceptor preference and the nature of the resulting biogenic Fe(II) fraction. Anthraquinone-2,6-disulfonate (AQDS), an electron shuttle, was included in select treatments to enhance bioreduction and subsequent biomineralization. The sediment was highly aggregated and contained two distinct clast populations: (i) a highly weathered one with “sponge-like” internal porosity, large mineral crystallites, and Fe-containing micas, and (ii) a dense, compact one with fine-textured Fe-containing illite and nano-sized goethite, as revealed by various forms of electron microscopic analyses. Approximately 10-15% of the Fe(III)_TOT was bioreduced by CN32 over 60 d in media without AQDS, whereas 24% and 35% of the Fe(III)_TOT was bioreduced by CN32 after 40 and 95 d in media with AQDS. Little or no Fe²⁺, Mn, Si, Al, and Mg were evident in aqueous filtrates after reductive incubation. Mössbauer measurements on the bioreduced sediments indicated that both goethite and phyllosilicate Fe(III) were partly reduced without bacterial preference. Goethite was more extensively reduced in the presence of AQDS whereas phyllosilicate Fe(III) reduction was not influenced by AQDS. Biogenic Fe(II) resulting from phyllosilicate Fe(III) reduction remained in a layer-silicate environment that displayed enhanced solubility in weak acid. The mineralogic nature of the goethite biotransformation product was not determined. Chemical and cryogenic Mössbauer measurements, however, indicated that the transformation product was not siderite, green rust, magnetite, Fe(OH)₂, or Fe(II) adsorbed on phyllosilicate or bacterial surfaces. Several lines of evidence suggested that biogenic Fe(II) existed as surface associated phase on the residual goethite, and/or as a Fe(II)-Al coprecipitate. Sediment aggregation and mineral physical and/or chemical factors were demonstrated to play a major role on the nature and location of the biotransformation reaction and its products.     

Effect of divalent cations on the kinetics of Fe(III) complexation by organic ligands in natural waters

We have investigated the kinetics of Fe(III) complexation by several organic ligands including fulvic acid, citrate and ethylenediaminetetraacetic acid (EDTA). Particular attention was given to examination of the effect of competitive divalent cations (Me: Ca²⁺ and Mg²⁺) at concentrations typical of seawater on the complexation rate. All experiments were conducted in 0.5 M NaCl solution buffered with 2 mM bicarbonate at pH 8.0 in the absence and presence of Me (25 μM-250 mM). The rate constants of complex formation determined by using the competitive ligand (5-sulfosalicylic acid) method combined with visible spectrophotometry ranged from 3.3 × 10⁴ to 3.2 × 10⁶ M⁻¹ s⁻¹. The mechanism of complexation was then examined based on a kinetic model. When EDTA was used as a ligand, Me at concentrations comparable to the ligand markedly retarded the rate of iron complex formation due to the predominance of an adjunctive pathway (where iron-ligand complex is formed via direct association of iron to Me-ligand complex). In contrast, the competing effect of Me on iron complexation by citrate and fulvic acid was observed only when the Me concentration was in excess of the ligand by more than a factor of 10-1000. The kinetic model suggests that iron complexation by fulvic acid occurs predominantly via a disjunctive pathway (where iron complexation by ligand occurs after dissociation of Me from Me-ligand complex) at concentrations of divalent cations and natural organic matter typical of natural waters including seawater and freshwater.     

Kinetic and equilibrium Fe isotope fractionation between aqueous Fe(II) and Fe(III)

Equilibrium and kinetic Fe isotope fractionation between aqueous ferrous and ferric species measured over a range of chloride concentrations (0, 11, 110 mM Cl⁻) and at two temperatures (0 and 22°C) indicate that Fe isotope fractionation is a function of temperature, but independent of chloride contents over the range studied. Using ⁵⁷Fe-enriched tracer experiments the kinetics of isotopic exchange can be fit by a second-order rate equation, or a first-order equation with respect to both ferrous and ferric iron. The exchange is rapid at 22°C, ∼60-80% complete within 5 seconds, whereas at 0°C, exchange rates are about an order of magnitude slower. Isotopic exchange rates vary with chloride contents, where ferrous-ferric isotope exchange rates were ∼25 to 40% slower in the 11 mM HCl solution compared to the 0 mM Cl⁻ (∼10 mM HNO₃) solutions; isotope exchange rates are comparable in the 0 and 110 mM Cl⁻ solutions.The average measured equilibrium isotope fractionations, Δ_{Fe(III)-Fe(II)}, in 0, 11, and 111 mM Cl⁻ solutions at 22°C are identical within experimental error at +2.76±0.09, +2.87±0.22, and +2.76±0.06 ‰, respectively. This is very similar to the value measured by Johnson et al. (2002a) in dilute HCl solutions. At 0°C, the average measured Δ_{Fe(III)-Fe(II)} fractionations are +3.25±0.38, +3.51±0.14 and +3.56±0.16 ‰ for 0, 11, and 111 mM Cl⁻ solutions. Assessment of the effects of partial re-equilibration on isotope fractionation during species separation suggests that the measured isotope fractionations are on average too low by ∼0.20 ‰ and ∼0.13 ‰ for the 22°C and 0°C experiments, respectively. Using corrected fractionation factors, we can define the temperature dependence of the isotope fractionation from 0°C to 22°C as: where the isotopic fractionation is independent of Cl⁻ contents over the range used in these experiments. These results confirm that the Fe(III)-Fe(II) fractionation is approximately half that predicted from spectroscopic data, and suggests that, at least in moderate Cl⁻ contents, the isotopic fractionation is relatively insensitive to Fe-Cl speciation.     

Precipitation of secondary Fe(III) minerals from acid mine drainage

Oxidation of FeS₂ in mine waste releases ${\mathrm{SO}}_4^{2-}$, Fe(II) and H⁺, resulting in acid mine drainage (AMD). Subsequent oxidation and precipitation of Fe produces different Fe(III) phases where the mineralogical composition depends on pH and the ambient concentrations of metal ions and complexing ligands. The oxidation and precipitation of Fe in AMD has been studied under various conditions with the intent of understanding the role these processes play in the natural attenuation of metal contaminants in the AMD. The combined process of Fe oxidation and precipitation in AMD from the Kristineberg mine, northern Sweden, has been investigated with pH-stat experiments at pH 5.5 and 7 at 10 and 25 °C. The precipitates formed have been characterised in terms of mineralogy and surface area. Similar phases formed at both temperatures, while the oxidation and precipitation occurred more readily at the higher temperature and higher pH. At pH 7, mainly lepidocrocite (γ-FeOOH) was precipitated while at a lower pH of 5.5, a mixture of schwertmannite, goethite, ferrihydrite and lepidocrocite formed. The ambient Zn(II) concentration was immediately reduced to acceptable levels (according to Swedish EPA) at pH 7 whereas a 2–3 weeks ageing period was necessary to achieve the same effect at pH 5.5. The presence of natural organic matter (NOM) reduced the attenuating effect at pH 5.5 after ageing but increased it slightly at pH 7. Addition of Zn(II) at pH 8 resulted in a mixed Fe(III)–Zn(II) precipitate of unknown composition with some Zn(II) adsorbed at the surface. The Fe(III) precipitates formed are potentially useful for the natural attenuation of metal contaminants in AMD although based on these investigations, the degree of success depends upon pH and NOM concentration.     

Complexing of silica by iron(III) in natural waters

Determination of amorphous silica solubility in acidified ferric nitrate solutions confirms the presence of ferric silicate complexing. A dissociation constant for the reaction:

$\text{FeH}{}_3\text{SiO}{}_4{}^{\mathrm{2+}}\text{→}\text{Fe}{}^{\mathrm{3+}}\text{+}\text{H}{}_3\text{SiO}{}_4{}^{\mathrm{?}}$

of 10^?9.8 ± 0.3 pK units at room temperature (22 ± 3°C) is obtained, in close agreement with reported values at 25°C corrected to zero ionic strength of 10^?9.9 by Weber and Stumm and 10^?9.5 by Olson and O'Melia. Iron-silicate complexing may be of significance to the mobilization of silica in acid waters associated with oxidizing sulphide deposits and coal strip mining and the precipitation of secondary silicate mineral phases.     

Impact of natural organic matter on H2O2-mediated oxidation of Fe(II) in a simulated freshwater system

The oxidation of Fe(II) by H₂O₂ has been studied in the presence of Suwannee River fulvic acid, a standard form of natural organic matter, by adding inorganic Fe(II) to solutions containing both H₂O₂ and fulvic acid and monitoring the total Fe(II) concentration using a luminol chemiluminescence method. At pH 8.4 and in the absence of competing metals, Suwannee River fulvic acid significantly retards the rate of Fe(II) oxidation due to gradual formation of a species that is oxidized more slowly than inorganic Fe(II) by both O₂ and H₂O₂. It is suggested that rapid formation of a weak Fe(II)-fulvic acid complex that is not readily oxidized by H₂O₂ is the cause of the reduction in the initial oxidation rate, and that the subsequent further reduction in oxidation rate is a result of the formation of a second type of Fe(II)-fulvic acid complex that is resistant to both O₂ and H₂O₂ oxidation. A kinetic model has been developed that supports this conceptual model. The results demonstrate that, under certain conditions, natural organic matter may stabilize Fe(II) in the presence of elevated H₂O₂ concentrations, significantly increasing the lifetime of ferrous iron and reducing the flux of hydroxyl radicals produced through this oxidation pathway.     

Nanoparticles in natural systems II: The natural oxide fraction at interaction with natural organic matter and phosphate

Information on the particle size and reactive surface area of natural samples and its interaction with natural organic matter (NOM) is essential for the understanding bioavailability, toxicity, and transport of elements in the natural environment. In part I of this series (Hiemstra et al., 2010), a method is presented that allows the determination of the effective reactive surface area (A, m²/g soil) of the oxide particles of natural samples which uses a native probe ion (phosphate) and a model oxide (goethite) as proxy. In soils, the natural oxide particles are generally embedded in a matrix of natural organic matter (NOM) and this will affect the ion binding properties of the oxide fraction. A remarkably high variation in the natural phosphate loading of the oxide surfaces (Γ, μmol/m²) is observed in our soils and the present paper shows that it is due to surface complexation of NOM, acting as a competitor via site competition and electrostatic interaction. The competitive interaction of NOM can be described with the charge distribution (CD) model by defining a ≡NOM surface species. The interfacial charge distribution of this ≡NOM surface species can be rationalized based on calculations done with an evolved surface complexation model, known as the ligand and charge distribution (LCD) model. An adequate choice is the presence of a charge of −1 v.u. at the 1-plane and −0.5 v.u. at the 2-plane of the electrical double layer used (Extended Stern layer model).The effective interfacial NOM adsorption can be quantified by comparing the experimental phosphate concentration, measured under standardized field conditions (e.g. 0.01 M CaCl₂), with a prediction that uses the experimentally derived surface area (A) and the reversibly bound phosphate loading (Γ, μmol/m²) of the sample (part I) as input in the CD model. Ignoring the competitive action of adsorbed NOM leads to a severe under-prediction of the phosphate concentration by a factor ∼10 to 1000. The calculated effective loading of NOM is low at a high phosphate loading (Γ) and vice versa, showing the mutual competition of both constituents. Both constituents in combination usually dominate the surface loading of natural oxide fraction of samples and form the backbone in modeling the fate of other (minor) ions in the natural environment.Empirically, the effective NOM adsorption is found to correlate well to the organic carbon content (OC) of the samples. The effective NOM adsorption can also be linked to DOC. For this, a Non-Ideal Competitive adsorption (NICA) model is used. DOC is found to be a major explaining factor for the interfacial loading of NOM as well as phosphate. The empirical NOM-OC relation or the parameterized NICA model can be used as an alternative for estimating the effective NOM adsorption to be implemented in the CD model for calculation of the surface complexation of field samples. The biogeochemical impact of the NOM-PO₄ interaction is discussed.     

Mineralogical and morphological constraints on the reduction of Fe(III) minerals by Geobacter sulfurreducens

The biologically-mediated reduction of synthetic samples of the Fe(III)-bearing minerals hematite, goethite, lepidocrocite, feroxhyte, ford ferrihydrite, akaganeite and schwertmannite by Geobacter sulfurreducens has been investigated using microbiological techniques in conjunction with X-ray Diffraction (XRD), Transmission Electron Microscopy (TEM) and X-ray Photoelectron Spectroscopy (XPS). This combination of approaches offers unique insights into the influence of subtle variations in the crystallinity of a given mineral on biogeochemical processes, and has highlighted the importance of (oxyhydr)oxide crystallite morphology in determining the changes occurring in a given mineral phase. Problems arising from normalising the biological Fe(III) reduction rates relative to the specific surface areas of the starting materials are also highlighted. These problems are caused primarily by particle aggregation, and compounded when using spectrophotometric assays to monitor reduction. For example, the initial rates of Fe(III) reduction observed for two synthetic feroxyhytes with different crystallinities (as shown by XRD and TEM studies) but almost identical surface areas, differ substantially. Both microbiological and high-resolution TEM studies show that hematite and goethite are susceptible to limited amounts of Fe(III) reduction, as evidenced by the accumulation of Fe(II) during incubation with G. sulfurreducens and the growth of nodular structures on crystalline goethite laths during incubation. Lepidocrocite and akaganeite readily transform into mixtures of magnetite and goethite, and XRD data indicate that the proportion of magnetite increases within the transformation products as the crystallinity of the starting material decreases. The presence of anthraquinone-2,6-disulfonate (AQDS) as an electron shuttle increases both the initial rate and longer term extent of biological Fe(III) reduction for all of the synthetic minerals examined. High-resolution XPS indicates subtle but measurable differences in the Fe(III):Fe(II) ratios at the mineral surfaces following extended incubation. For example, for a poorly crystalline schwertmannite, deconvolution of the Fe2p_3/2 peak suggests that the Fe(III):Fe(II) ratio of the near-surface regions varies from 1.0 in the starting material to 0.9 following 144 h of incubation with G.sulfurreducens, and to 0.75 following the same incubation period in the presence of 10 μM AQDS. These results have important implications for the biogeochemical cycling of iron.     

Influence of water-extractable organic matter from Opalinus Clay on the sorption and speciation of Ni(II), Eu(III) and Th(IV)

The influence of water-extractable organic matter from 6 Opalinus Clay (OPA) samples from Mont Terri and Benken (Switzerland) on the sorption of Ni(II), Eu(III) and Th(IV) has been measured using an ion exchange technique. OPA is considered to be one of the potential host rocks for the deep geological disposal of high-level and long-lived intermediate-level radioactive waste in Switzerland. Within the range of estimated uncertainties, no significant differences in sorption were observed in most cases as compared with suitable synthetic waters devoid of organic C. Only in certain individual cases were slight reductions in sorption (less than a factor of 5) for Eu(III) and Ni(II) found. The results of accompanying laser fluorescence spectroscopy experiments did not show any influence of the extracts on Cm(III) speciation. This would suggest that the reduction of sorption occasionally observed in the ion exchange experiments is probably not caused by the formation of complexes between the radionuclides and the organic matter in the extracts, but is rather due to an underestimation of systematic uncertainties. From these findings, and from UV–VIS spectroscopic characterisation of the organic matter in the extracts, it can be concluded that only a negligible fraction of the organic matter present may be in the form of humic or fulvic acids. It is consequently justified to put aside overly conservative assumptions with respect to the complexing behaviour of the organic matter used towards the metal ions investigated and their chemical analogues. In view of the site-specific character of the present study, these conclusions may not be arbitrarily applied to other geological formations considered as possible host rocks for the disposal of radioactive waste.     

Dissolution of cinnabar (HgS) in the presence of natural organic matter

Cinnabar (HgS) dissolution rates were measured in the presence of 12 different natural dissolved organic matter (DOM) isolates including humic, fulvic, and hydrophobic acid fractions. Initial dissolution rates varied by 1.3 orders of magnitude, from 2.31 × 10⁻¹³ to 7.16 × 10⁻¹² mol Hg (mg C)⁻¹ m⁻²s⁻¹. Rates correlate positively with three DOM characteristics: specific ultraviolet absorbance (R² = 0.88), aromaticity (R² = 0.80), and molecular weight (R² = 0.76). Three experimental observations demonstrate that dissolution was controlled by the interaction of DOM with the cinnabar surface: (1) linear rates of Hg release with time, (2) significantly reduced rates when DOM was physically separated from the surface by dialysis membranes, and (3) rates that approached constant values at a specific ratio of DOM concentration to cinnabar surface area, suggesting a maximum surface coverage by dissolution-reactive DOM. Dissolution rates for the hydrophobic acid fractions correlate negatively with sorbed DOM concentrations, indicating the presence of a DOM component that reduced the surface area of cinnabar that can be dissolved. When two hydrophobic acid isolates that enhanced dissolution to different extents were mixed equally, a 20% reduction in rate occurred compared to the rate with the more dissolution-enhancing isolate alone. Rates in the presence of the more dissolution-enhancing isolate were reduced by as much as 60% when cinnabar was prereacted with the isolate that enhanced dissolution to a lesser extent. The data, taken together, imply that the property of DOM that enhances cinnabar dissolution is distinct from the property that causes it to sorb irreversibly to the cinnabar surface.     

Iron isotope fractionation during microbially stimulated Fe(II) oxidation and Fe(III) precipitation

Interpretation of the origins of iron-bearing minerals preserved in modern and ancient rocks based on measured iron isotope ratios depends on our ability to distinguish between biological and non-biological iron isotope fractionation processes. In this study, we compared ⁵⁶Fe/⁵⁴Fe ratios of coexisting aqueous iron (Fe(II)_aq, Fe(III)_aq) and iron oxyhydroxide precipitates (Fe(III)_ppt) resulting from the oxidation of ferrous iron under experimental conditions at low pH (<3). Experiments were carried out using both pure cultures of Acidothiobacillus ferrooxidans and sterile controls to assess possible biological overprinting of non-biological fractionation, and both SO₄²⁻ and Cl⁻ salts as Fe(II) sources to determine possible ionic/speciation effects that may be associated with oxidation/precipitation reactions. In addition, a series of ferric iron precipitation experiments were performed at pH ranging from 1.9 to 3.5 to determine if different precipitation rates cause differences in the isotopic composition of the iron oxyhydroxides. During microbially stimulated Fe(II) oxidation in both the sulfate and chloride systems, ⁵⁶Fe/⁵⁴Fe ratios of residual Fe(II)_aq sampled in a time series evolved along an apparent Rayleigh trend characterized by a fractionation factor α_{Fe(III)aq-Fe(II)aq} ∼ 1.0022. This fractionation factor was significantly less than that measured in our sterile control experiments (∼1.0034) and that predicted for isotopic equilibrium between Fe(II)_aq and Fe(III)_aq (∼1.0029), and thus might be interpreted to reflect a biological isotope effect. However, in our biological experiments the measured difference in ⁵⁶Fe/⁵⁴Fe ratios between Fe(III)_aq, isolated as a solid by the addition of NaOH to the final solution at each time point under N₂-atmosphere, and Fe(II)_aq was in most cases and on average close to 2.9‰ (α_{Fe(III)aq-Fe(II)aq} ∼ 1.0029), consistent with isotopic equilibrium between Fe(II)_aq and Fe(III)_aq. The ferric iron precipitation experiments revealed that ⁵⁶Fe/⁵⁴Fe ratios of Fe(III)_aq were generally equal to or greater than those of Fe(III)_ppt, and isotopic fractionation between these phases decreased with increasing precipitation rate and decreasing grain size. Considered together, the data confirm that the iron isotope variations observed in our microbial experiments are primarily controlled by non-biological equilibrium and kinetic factors, a result that aids our ability to interpret present-day iron cycling processes but further complicates our ability to use iron isotopes alone to identify biological processing in the rock record.     

Superoxide-mediated Fe(II) formation from organically complexed Fe(III) in coastal waters

Fe(III) complexed by organic ligands (Fe(III)L) is the primary form of dissolved Fe in marine and coastal environments. Superoxide, typically produced in biological and photochemical processes, is one of the reducing agents that contributes to transformation of Fe(III)L to bioavailable, free dissolved Fe(II) (Fe(II)′). In this work, the kinetics of superoxide-mediated Fe(II)′ formation from Fe(III)L in a simulated coastal water system were investigated and a comprehensive kinetic model was developed using citrate and fulvic acid as exemplar Fe-binding ligands. To simulate a coastal environment in laboratory experiments, Fe(III)L samples with various ligand/Fe ratios were incubated for 5 min to 1 week in seawater medium. At each ratio and incubation time, the rate of superoxide-mediated Fe(II)′ formation was determined in the presence of the strong Fe(II) binding ligand ferrozine by spectrophotometrically measuring the ferrous-ferrozine complex generated at a constant concentration of superoxide. The Fe(II)′ formation rate generally decreased with incubation time, as Fe(III)L gradually dissociated to form less reactive Fe(III) oxyhydroxide. However, when the ligand/Fe ratio was sufficiently high, the dissociation of Fe(III)L (and subsequent Fe precipitation) was suppressed and Fe(II)′ was formed at a higher rate. The rate of Fe(II)′ produced during the experiment was explained by the kinetic model. The model confirmed that both the ligand/Fe ratio and incubation time have a significant effect on the pathway via which Fe(II)′ is formed from Fe(III)-fulvic acid complexes.     

Formation, reactivity, and aging of ferric oxide particles formed from Fe(II) and Fe(III) sources: Implications for iron bioavailability in the marine environment

Freshly formed amorphous ferric oxides (AFO) in the water column are potentially highly reactive, but with reactivity declining rapidly with age, and have the capacity to partake in reactions with dissolved species and to be a significant source of bioavailable iron. However, the controls on reactivity in aggregated oxides are not well understood. Additionally, the mechanism by which early rapid aging occurs is not clear. Aging is typically considered in terms of changes in crystallinity as the structure of an iron oxide becomes more stable and ordered with time thus leading to declining reactivity. However, there has been recognition of the role that aggregation can play in determining reactivity, although it has received limited attention. Here, we have formed AFO in seawater in the laboratory from either an Fe(II) or Fe(III) source to produce either AFO(II) or AFO(III). The changes in reactivity of these two oxides following formation was measured using both ligand-promoted dissolution (LPD) and reductive dissolution (RD). The structure of the two oxides was examined using light scattering and X-ray adsorption techniques. The dissolution rate of AFO(III) was greater than that of AFO(II), as measured by both dissolution techniques, and could be attributed to both the less ordered molecular structure and smaller primary particle size of AFO(III). From EXAFS analysis shortly (90 min) following formation, AFO(II) and AFO(III) were shown to have the same structure as aged lepidocrocite and ferrihydrite respectively. Both oxides displayed a rapid decrease in dissolution rate over the first hours following formation in a pattern that was very similar when normalised. The early establishment and little subsequent change of crystal structure for both oxides undermined the hypothesis that increasing crystallinity was responsible for early rapid aging. Also, an aging model describing this proposed process could only be fitted to the data with kinetic parameters that were inconsistent with such a mechanism. The similar aging patterns and existence of diffusion limited cluster aggregation (DLCA) suggested that loss of Fe centre accessibility due to aggregation is the likely cause of early rapid aging of AFO. A simple model describing the loss of surface area during the aggregate growth, measured using dynamic light scattering (DLS), produced aging patterns that matched the reactivity loss of AFO(III) measured using RD but not LPD. The difference between the two measures of dissolution rate could not be explained, but indicated that different measures of reactivity respond differentially to various parameters controlling reactivity. Analysis of aggregate structure using aggregation kinetics and static light scattering (SLS) suggested that restructuring during aggregation was occurring at an aggregate level for AFO(III), but only minimally so for AFO(II). While our investigations support the contention that aggregation is responsible for early rapid aging, the role of aggregate structure is remains unclear.     

Dissociation kinetics of Fe(III)- and Al(III)-natural organic matter complexes at pH 6.0 and 8.0 and 25 °C

The rate at which iron- and aluminium-natural organic matter (NOM) complexes dissociate plays a critical role in the transport of these elements given the readiness with which they hydrolyse and precipitate. Despite this, there have only been a few reliable studies on the dissociation kinetics of these complexes suggesting half-times of some hours for the dissociation of Fe(III) and Al(III) from a strongly binding component of NOM. First-order dissociation rate constants are re-evaluated here at pH 6.0 and 8.0 and 25 °C using both cation exchange resin and competing ligand methods for Fe(III) and a cation exchange resin method only for Al(III) complexes. Both methods provide similar results at a particular pH with a two-ligand model accounting satisfactorily for the dissociation kinetics results obtained. For Fe(III), half-times on the order of 6-7 h were obtained for dissociation of the strong component and 4-5 min for dissociation of the weak component. For aluminium, the half-times were on the order of 1.5 h and 1-2 min for the strong and weak components, respectively. Overall, Fe(III) complexes with NOM are more stable than analogous complexes with Al(III), implying Fe(III) may be transported further from its source upon dilution and dispersion.     

Cu(II) adsorption at the calcite-water interface in the presence of natural organic matter: Kinetic studies and molecular-scale characterization

We have characterized the adsorption of Suwannee River humic acid (SRHA) and Cu(II) on calcite from preequilibrated solutions at pH 8.25. Sorption isotherms of SRHA on calcite follow Langmuir-type behavior at SRHA concentrations less than 15 mg C L⁻¹, whereas non-Langmuirian uptake becomes evident at concentrations greater than 15 mg C L⁻¹. The adsorption of SRHA on calcite is rapid and mostly irreversible, with corresponding changes in electrostatic properties. At pH 8.25, Cu(II) uptake by calcite in the presence of dissolved SRHA decreases with increasing dissolved SRHA concentration, suggesting that formation of Cu-SRHA aqueous complexes is the primary factor controlling Cu(II) sorption at the calcite surface under the conditions of our experiments. We also observed that surface-bound SRHA has little influence on Cu(II) uptake by calcite, suggesting that Cu(II) coordinates to calcite surface sites rather than to surface-bound SRHA.Cu K-edge X-ray absorption near edge structure (XANES) and extended X-ray absorption fine-structure (EXAFS) spectroscopic results show that the local coordination of Cu adsorbed at the calcite surface is very similar in the presence and absence of SRHA. Ca backscatterers at ∼3.90 Å indicate that Cu(II) forms tetragonally distorted inner-sphere adsorption complexes in both binary and ternary systems. Subtle differences in the XANES and EXAFS between binary sorption samples and ternary sorption samples, however, prevent us from ruling out the formation of ternary Cu-SRHA surface complexes. Our findings demonstrate that SRHA plays an important role in controlling the fate and transport of Cu(II) in calcite-bearing systems.