The reactivity and isotopic fractionation of Fe-bearing minerals during sulfidation : an experimental approach

摘要

The presence of Fe bearing minerals at the sediment-water interface (within marine settings) promotes a variety of biological and abiological redox reactions during early diagenesis. The highly reactive nature of a portion of these Fe bearing minerals, with respect to organic and inorganic species, influences both porewater chemistry and the biogeochemical cycling of trace metals. Of particular importance is the reaction between 'reactive' Fe minerals and dissolved sulfide (which ultimately produces pyrite). This is a major process in the modern environment, but has also been prevalent throughout Earth's history and forms the basis for identifying different paleodepositional redox conditions in the ancient rock record. Initial experimental studies of the sulfidic reductive dissolution of pure synthetic Fe(III) oxides have provided detailed insight into the mechanism and rates at which different minerals release Fe(II)aq into anoxic waters; whilst also describing the formation of reduced sulfide products including FeS and elemental sulfur. However, it remains unclear how realistic laboratory studies of the sulfidation of pure Fe minerals are in relation to natural sediment assemblages containing different minerals. Comparison of natural sediments with the reactivity of pure minerals studied under laboratory conditions thus forms the basis for the first part of this study. Sediment cores were obtained from Aarhus Bay (Denmark) and the Umpqua River Shelf (North Pacific Basin, N. America), representing contrasting conditions in terms of the reactive Fe species present. Aarhus Bay sediment samples contain a high proportion of the most reactive Fe oxide minerals (e.g. ferrihydrite, lepidocrocite) at the surface, which decrease with depth throughout the core, leaving a near-constant concentration of slower reacting Fe oxide minerals (e.g. goethite, magnetite, hematite). These trends are reflected in decreased rates of reductive dissolution with depth in the core, as determined via sulfidation experiments of sediment sampled from different depth intervals. In contrast, the Fe oxide content of Umpqua River Shelf sediments is more homogeneously distributed, with the slower reacting Fe oxide species dominating the sediment assemblage. As such, rates of reaction with respect to dissolved sulfide do not differ vastly with depth. Based on the determination of rate constants during these experiments, this examination of the reactivity of Fe oxides suggests that the natural Fe oxide assemblages react on a similar timescale (over the same order of magnitude) to that of synthetic minerals, suggesting that existing schemes for the reactivity of Fe oxide minerals towards dissolved sulfide provide a realistic framework for evaluating rates of reactivity in natural environments. In a range of similar experiments, the rate and mechanism of the sulfide-mediated dissolution of synthetic Fe carbonate (siderite) has also been assessed, in addition to a sample of natural crystalline siderite from the 1.88 billion year-old Biwabik Iron Formation, North America. These experiments have been performed as a consequence of the prevalence of siderite in ancient sediments. Siderite is commonly assumed to be highly reactive towards dissolved sulfide. However, despite the common occurrence of siderite in ancient sediments, its reactivity has not previously been determined, a fact which impacts upon the use of Fe speciation in ancient sediments as a paleodepositional redox indicator. Although carbonates follow a different mechanism of dissolution than Fe oxides, probably via the direct formation of FeS at the mineral surface, the experiments performed here show that synthetic Fe carbonate dissolution in the presence of dissolved sulfide is faster than for most Fe oxide minerals (including ferrihydrite and lepidocrocite). Furthermore, although the reactivity of the ancient crystalline siderite sample was slower than for the synthetic siderite, this mineral was also relatively reactive, suggesting that all Fe carbonate minerals can be considered 'highly reactive' towards dissolved sulfide. The final part of this thesis concerns an examination of Fe isotope fractionations during the sulfide-promoted reductive dissolution of a variety of synthetic Fe oxide minerals. The isotopic composition of Fe in natural rocks and sediments is commonly used to infer the processes responsible for Fe cycling during deposition and diagenesis. In particular, experiments with Fe reducing bacteria have demonstrated that isotopic fractionations of up to -1.3‰ may occur between the original oxide mineral and Fe(II) released to solution, and thus light Fe isotope values in ancient sediments have been used to reconstruct the antiquity or occurrence of bacterial Fe reduction. However, in all of these cases, the potential for an isotopic fractionation during the sulfide-promoted reductive dissolution of Fe oxides has been ignored. Thus it is important to quantify potential fractionations associated with this process in order to better evaluate Fe isotope compositions observed in the rock record. During both the reductive and dissolution steps of this abiotic reaction, a significant isotopic fractionation is observed, the magnitude of which is dependent on the mineral phase under reaction, and the specific experimental parameters. This detailed study represents the first time that an isotopic fractionation has been demonstrated with regard to the reductive step, in addition to the subsequent dissolution step of the overall reaction. Isotopic fractionations in the dissolved phase are not as large as those sometimes found in association with bacterial Fe reduction, but are in the same range (up to ~-0.8‰), suggesting that the influence of this reaction needs to be taken into careful consideration when evaluating Fe isotope compositions in modern and ancient sediments. Overall, this study builds upon existing experiments which have assessed the reactivity of individual Fe(III) oxide minerals towards dissolved sulfide, to provide new insight with regard to (bio)geochemical Fe mineral cycling. In particular, this study provides kinetic and isotopic constraints that have the potential to greatly enhance reconstructions of syngenetic and diagenetic reactions occurring in modern and ancient environments.