Show Abstract ▼

Abstract: Fluid release from dehydration reactions is considered to have significant effects on the strength and dynamics of tectonic faults at convergent plate boundaries. It is classically assumed that the production of fluid leads to increased pore fluid pressures that perturb a fault's stress state and thereby facilitates and enhances deformation. This important assumption has never been supported by direct microstructural observations. Here, we investigate the role of gypsum dehydration in the deformation of evaporitic rocks using synchrotron-based time-resolved X-ray computed microtomography (4D) imaging. This approach enables the documentation of coupled chemical, hydraulic and mechanical processes on the grain scale. In our experiments with deforming halite-gypsum-halite sandwiches we observe that the fluid released by dehydrating gypsum accumulates at the gypsum-halite interface before a distributed hydraulic failure of the halite layer drains the fluid. From our observations we conclude that perceivedly impermeable halite layers in evaporites are unlikely to trap overpressured fluid, e.g., in thin-skinned tectonic detachment horizons. Moreover, as the hydraulic failure is diffuse and not localized, our experiments suggest that dehydration reactions alone may not explain intermediate depth seismicity in subduction zones. Our data demonstrate the significant potential that in-situ 4D imaging has for the grain-scale investigation of fundamental tectonic processes.

Show Abstract ▼

Abstract: Mantle oxygen fugacity (fO2) governs the physico-chemical evolution of the Earth, however current estimates from commonly used basalt redox proxies are often in disagreement. In this study we compare three different potential basalt fO2 proxies: Fe3+/Fetot, V/Sc and V isotopes, determined on the same submarine lavas from a 700 km section of the Reykjanes Ridge, near Iceland. These samples provide a valuable test of the sensitivities of fO2 proxies to basalt petrogenesis, as they formed at different melting conditions and from a mantle that towards Iceland exhibits increasing long-term enrichment of incompatible elements. New trace element data were determined for 63 basalts with known Fe3+/Fetot. A subset of 19 lavas, covering the geographical spread of the ridge transect, was selected for vanadium isotope analyses. Vanadium is a multi-valence element whose isotopic fractionation is theoretically susceptible to redox conditions. Yet, the δ51 VAA composition of basaltic glasses along the Reykjanes Ridge covers only a narrow range ( δ51 VAA = −1.09 to −0.86‰; 1SD = 0.02–0.09) and does not co-vary with fractionation-corrected Fe3+/Fetot (0.134–0.151; 1SD = 0.005) or V/Sc (6.6–8.5; 1SD = 0.1-1.3) ratios. However, on a global scale, basaltic δ51 VAA may be controlled by the extent of melting. The V/Sc compositions of primitive (MgO > 7.5 wt%) basalts show no systematic change along the entire length of the Reykjanes Ridge. Typical peridotite melting models in which source Fe3+/Fetot is constant at 5% and that account for the increased mantle potential temperature nearer the plume center and the fO2 dependent partitioning of V, can reproduce the V/Sc data. However, while these melting models predict that basalt Fe3+/Fetot ratios should decrease with increasing mantle potential temperature towards Iceland, fractionation-corrected Fe3+/Fetot of Reykjanes Ridge lavas remain nearly constant over the ridge length. This discrepancy is explained by source heterogeneity, where an oxidized mantle pyroxenite component contributes to melting with increasing proximity to Iceland. Comparison of observed and modeled Fe3+/Fetot indicate that source variation in fO2 is present under the Reykjanes Ridge, with higher Fe3+/Fetot closer to Iceland. This source variability in fO2 cannot be resolved by V isotopes and redox-sensitive trace element ratios, which instead appear to record magmatic processes.

Open Access

Show Abstract ▼

Abstract: We have determined the chemical speciation of dissolved sulfur and the sulfur concentration at fixed oxygen and sulfur fugacities for a wide range of silicate melt compositions (from Fe-rich basalt to dacite). Each melt was equilibrated at 1300 °C and 1-atmosphere pressure at oxygen fugacities (fO2) between −1.67 and +1.6 log units relative to the Fayalite–Magnetite–Quartz (FMQ) buffer and absolute sulfur fugacities between −5.1 and −1.2 log units. The fO2 and fS2 of the experiments were controlled by using gas mixtures of CO–CO2–SO2. The speciation of sulfur in the quenched glasses was determined using both X-ray Absorption Near-Edge Spectroscopy (XANES), and from the dependence of equilibrium sulfur concentration on the fS2/fO2 ratio measured by secondary-ion mass spectrometry (SIMS) and electron microprobe. The speciation of dissolved sulfur in each melt undergoes an abrupt transformation from S2− to S6+ with increasing fO2, and this transition is shifted ∼0.5 log units higher in fO2 as melt FeO concentration increases from ∼5 wt% to ∼18 wt%. Since sulfide concentrations at constant fO2 and fS2 are consistently greater for more FeO-rich melts, the compositional effect on speciation may be explained by the well-known sensitivity of the sulfide capacity ( CS2− ) of the melt to FeO concentration. S6+/S2− ratios for the glasses exhibit a linear relationship with Fe3+/Fe2+, indicating that the redox couples for iron and sulfur can be directly related to one another. We used thermodynamic data to model the interrelationship between Fe and S oxidation states in terms of the equilibrium FeS+8FeO1.5=8FeO+FeSO4 Fitting the data to our experiments at 1300 °C we obtained the following expression for the temperature-dependence of speciation: log⁡(S6+S2−)=8log⁡(Fe3+Fe2+)+8.7436×106T2−27703T+20.273 This equation fits the data for all our compositions and is also consistent with earlier results at 1050 °C and 950 °C. We used the interdependence of S and Fe oxidation states to infer electron transfer between Fe2+ and S6+ during quenching of glasses from Mauna Kea, Hawaii. The effect is sufficient to cause significant overestimation of equilibrium Fe3+/ΣFe in natural glasses and corresponding overestimate of fO2 by about 0.8 log units. Glasses equilibrated under the most oxidizing conditions (containing S6+ only) have equilibrium S concentrations that are negatively correlated with their mole fractions of tetrahedral (Si + Ti) cations.

Open Access

Show Abstract ▼

Abstract: Glaciers and ice sheets are a significant source of nanoparticulate Fe, which is potentially important in sustaining the high productivity observed in the near-coastal regions proximal to terrestrial ice cover. However, the bioavailability of particulate iron is poorly understood, despite its importance in the ocean Fe inventory. We combined high-resolution imaging and spectroscopy to investigate the abundance, morphology and valence state of particulate iron in glacial sediments. Our results document the widespread occurrence of amorphous and Fe(II)-rich and Fe(II)-bearing nanoparticles in Arctic glacial meltwaters and iceberg debris, compared to Fe(III)-rich dominated particulates in an aeolian dust sample. Fe(II) is thought to be highly biolabile in marine environments. Our work shows that glacially derived Fe is more labile than previously assumed, and consequently that glaciers and ice sheets are therefore able to export potentially bioavailable Fe(II)-containing nanoparticulate material to downstream ecosystems, including those in a marine setting. Our findings provide further evidence that Greenland Ice Sheet meltwaters may provide biolabile particulate Fe that may fuel the large summer phytoplankton bloom in the Labrador Sea, and that Fe(II)-rich particulates from a region of very high productivity downstream of a polar ice sheet may be glacial in origin.

Open Access

Show Abstract ▼

Abstract: The redox state of volcanic products determines their leverage on the oxidation of Earth's oceans and atmosphere, providing a long-term feedback on oxygen accumulation at the planet's surface. An archive of redox conditions in volcanic plumbing systems from a magma's mantle source, through crustal storage, to eruption, is carried in pockets of melt trapped within crystals. While melt inclusions have long been exploited for their capacity to retain information on a magma's history, their permeability to fast-diffusing elements such as hydrogen is now well documented and their retention of initial oxygen fugacities (fO2fO2) could be similarly diffusion-limited. To test this, we have measured Fe3+/ΣFe by micro-XANES spectroscopy in a suite of 65 olivine-hosted melt inclusions and 9 matrix glasses from the AD 1783 Laki eruption, Iceland. This eruption experienced pre-eruptive mixing of chemically diverse magmas, syn-eruptive degassing at the vent, and post-eruptive degassing during lava flow up to 60 km over land, providing an ideal test of whether changes in the fO2fO2 of a magma may be communicated through to its cargo of crystal-hosted melt inclusions. Melt inclusions from rapidly quenched tephra samples have Fe3+/ΣFe of 0.206±0.0080.206±0.008 (ΔQFM of +0.7 ± 0.1), with no correlation between their fO2fO2 and degree of trace element enrichment or differentiation. These inclusions preserve the redox conditions of the mixed pre-eruptive Laki magma. When corrected for fractional crystallisation to 10 wt.% MgO, these inclusions record a parental magma [Fe3+/ΣFe](10) of 0.18 (ΔQFM of +0.4), significantly more oxidised than the Fe3+/ΣFe of 0.10 that is often assumed for Icelandic basalt magmas. Melt inclusions from quenched lava selvages are more reduced than those from the tephra, having Fe3+/ΣFe between 0.133 and 0.177 (ΔQFM from −0.4 to +0.4). These inclusions have approached equilibrium with their carrier lava, which has been reduced by sulfur degassing. The progressive re-equilibration of fO2fO2 between inclusions and carrier melts occurs on timescales of hours to days, causing a drop in the sulfur content at sulfide saturation (SCSS) and driving the exsolution of immiscible sulfide globules in the inclusions. Our data demonstrate the roles of magma mixing, progressive re-equilibration, and degassing in redox evolution within magmatic systems, and the open-system nature of melt inclusions to fO2fO2 during these processes. Redox heterogeneity present at the time of inclusion trapping may be overprinted by rapid re-equilibration of melt inclusion fO2fO2 with the external environment, both in the magma chamber and during slow cooling in lava at the surface. This can decouple the melt inclusion archives of fO2fO2, major and trace element chemistry, and mask associations between fO2fO2, magmatic differentiation and mantle source heterogeneity unless the assembly of diverse magmas is rapidly followed by eruption. Our tools for understanding the redox conditions of magmas are thus limited; however, careful reconstruction of pre- and post-eruptive magmatic history has enabled us to confirm the relatively oxidised nature of ocean island-type mantle compared to that of mid-ocean ridge mantle.