Earth & Planetary Science Research Frontiers 2023 100 Fe 3+ -rich oxidizing magma formed by Fe 2+ redox disproportionation under high pressure The valence state of iron oxides (Fe 2+ and Fe 3+ ) in the mantle of terrestrial planets significantly affects the oxidation state of the mantle. This is because iron is the most abundant multi-valence state element in the mantle of terrestrial planets. Oxygen fugacity is controlled by the exchange reaction between Fe 2+ and Fe 3+ (i.e., 2FeO + O 2 = Fe 2 O 3 ); therefore, the valence state of the iron oxides determines the oxidation state of the mantle. Because volcanic gas compositions change substantially depending on the oxidation state of the mantle, constraining the valence state of iron oxides in the mantle is a quite important issue for understanding the atmospheric composition of terrestrial planets and perhaps planetary habitability. Previous geological constraints on the oxidation state of the Earth’s mantle inferred from the Ce anomaly of zircons suggest that the Earth’s mantle was already oxidized close to the present level 4 billion years ago [1]. This has been profoundly challenging to the canonical model of the oxidation state of the early Earth’s mantle because of the equilibration between the metallic core and the silicate mantle during its formation. The redox disproportionation of Fe 2+ in a molten mantle (hereafter magma ocean) under high pressures has been proposed to explain the oxidation of the early Earth’s mantle [2]. This mechanism produces Fe 3+ and metallic iron (Fe 0 ) from Fe 2+ (i.e., 3Fe 2+ → 2Fe 3+ + Fe 0 ). After removing Fe 0 from the metallic core, the Fe 3+ /(Fe 2+ + Fe 3+ ) ratio of the magma ocean increases and becomes more oxidizing than in the initial state. The redox disproportionation of Fe 2+ in silicate melts was recently confirmed at pressures above 10 GPa [3]. However, determining the Fe 3+ /(Fe 2+ + Fe 3+ ) ratio of silicate melt equilibrating with metallic iron under the Earth’s magma ocean conditions (peridotitic melt under high- pressure conditions >27 GPa) remains unconstrained and technically challenging due to its high-liquidus (completely molten) temperatures (>2400℃). We optimized the cell assembly for high-pressure melting experiments and successfully melted peridotite samples under metal-saturated conditions up to 28 GPa [4] (Fig. 1). The Fe 3+ /(Fe 2+ + Fe 3+ ) ratio of the silicate portion of the recovered samples was determined using X-ray absorption near-edge structure (XANES) measurements at SPring-8 BL27SU . The XANES spectra of the quenched samples showed a remarkable change at pressures between 15 and 28 GPa (Fig. 2), suggesting that the Fe 3+ /(Fe 2+ + Fe 3+ ) ratio changes significantly under different pressure conditions. We found that the calculated Fe 3+ /(Fe 2+ + Fe 3+ ) ratio of the quenched silicate melts from the XANES spectra increased from 0.13 to 0.52 with pressure (Fig. 3). These values are approximately one order of magnitude higher than the Fe 3+ /(Fe 2+ + Fe 3+ ) ratio of the Earth’s upper mantle. These results suggest that the surface of the Earth’s magma ocean after removing metallic iron into the core should be highly oxidizing, perhaps more than in the present state. The oxidation state of the surface of the Earth’s magma ocean having an Fe 3+ /(Fe 2+ + Fe 3+ ) ratio of 0.2 – 0.5 would have created highly oxidizing conditions, more so than at present. However, the estimated oxygen fugacity of such a magma ocean around 4–7 log-units above the iron-wüstite buffer is partially consistent with those estimated for the Earth’s upper mantle 4.0 – 4.4 billion years ago based on zircon records [1] within their analytical uncertainties. To explain the discrepancy in the oxidation state between the magma ocean and the present Earth’s upper mantle, we propose that metallic iron in late accreting impactors after the formation of the Earth may have reduced the Earth’s mantle [4]. The positive relationship between the pressure and redox disproportionation of Fe 2+ in the magma provides another unique insight into the origin of the difference in the oxidation states of terrestrial planets and their atmospheric compositions, particularly during their formation. More specifically, Moon- or Mars- sized rocky bodies may have a relatively reduced Fig. 1. The back-scattered electron image of the quenched sample at 28 GPa and 2600℃. The spheric metal surrounded by silicate indicates both phases were completely molten. Metal Capsule Silicate × × 300 15.0 kV COMPO SEM WD 6.5 mm 10 μ μ m Research Frontiers 2023 101 mantle and an atmosphere enriched in CH 4 and CO if the mantle is equilibrated with a metallic core. When a planet grows larger than Mars, the mantle may be significantly oxidized, and the atmospheric composition becomes rich in CO 2 and SO 2 [5]. The relationship between planetary size and atmospheric composition indicated in this study can be tested by future observations of exoplanetary atmospheres, particularly around young planets that likely have atmospheres equilibrated with the magma ocean. Fig. 2. Comparison of obtained XANES spectra of the silicate part of the recovered sample between (a) 15 GPa and (b) 28 GPa. The Fe 3+ /(Fe 2+ + Fe 3+ ) ratio was calculated from the area ratio of the first green-colored peak around 707 eV to all of the peaks. The inset of each figure shows an oxygen XRF map and analytical point indicated by + . Fig. 3. Calculated Fe 3+ /(Fe 2+ + Fe 3+ ) ratio of quenched silicate melts equilibrating with metallic iron from XANES spectra. Data of andesitic melt were taken from [3]. The dashed curve indicates thermodynamic prediction based on experimental data [4]. At pressures above 20 GPa, Fe 3+ in the melt is more stabilized than Fe 2+ , and the Fe 3+ /(Fe 2+ + Fe 3+ ) ratio rapidly increases with pressure. Hideharu Kuwahara a, * and Ryoichi Nakada b a Geodynamics Research Center, Ehime University b Kochi Institute for Core Sample Research, Institute for Extra-cutting-edge Science and Technology Avant-garde Research (X-star), Japan Agency for Marine-Earth Science and Technology (JAMSTEC) *Email: kuwahara.hideharu.vd@sci.ehime-u.ac.jp References [1] D. Trail et al .: Nature 480 (2011) 79. [2] M. M. Hirschmann: Earth Planet. Sci. Lett. 341-344 (2012) 48. [3] K. Armstrong et al .: Science 365 (2019) 903. [4] H. Kuwahara, R. Nakada, S. Kadoya, T. Yoshino and T. Irifune: Nat. Geosci. 16 (2023) 461. [5] F. Gaillard and B. Scaillet: Earth Planet. Sci. Lett. 403 (2014) 307. (a) Normalized Absorption 0 0.2 0.4 0.6 0.8 1.0 0 0.2 0.4 0.6 0.8 1.0 0 0.2 0.4 0.6 0.8 1.0 0 0.2 0.4 0.6 0.8 1.0 0 0.2 0.4 0.6 0.8 1.0 0 0.2 0.4 0.6 0.8 1.0 705 710 715 705 710 715 Energy (eV) Energy (eV) Obtained spectrum Obtained spectrum Fitting result Fitting result (b) Normalized Absorption Sample Sample Analytical Analytical point point O O O O O O 0 0.1 0.2 0.3 0.4 0.5 0.6 0 Fe 3+ /(Fe 2+ + Fe 3+ ) 5 10 15 20 25 30 Pressure (GPa) 2300 2900 2800 2700 2600 2500 2400 Temperature (K) Peridotitic melt [this study] Basaltic melt [this study] Andesitic melt [3] Thermodynamic prediction curve
home
- JP
- EN