fO2 of Earth’s mantle includes data for continental mantle xenoliths, abyssal peridotites, MORBs, arc lavas, and lamprophyres (Frost and McCammon, 2008).
The fO2 of Planetesimals and the Role of Volatiles.
Planetesimals of km‐size diameter formed by condensation and melting from the Solar Nebula. Thus, their chemical composition and oxidation state likely resemble those of the portion of Solar Nebula where they formed, taking into consideration the proposed correlation between heliocentric distance and H2/H2O and C/O ratios reflected in the oxidized nature of the planetesimals upon melting (Grossman et al., 2008). On the basis of N‐body simulations of particle interaction and planetary accretion, the bulk composition of planetesimals and their redox state change with respect to the heliocentric distances in the Solar Nebula. The bodies formed at a distance less than 1.1–1.7 AU exposed to higher temperatures were highly reduced at IW‐4 log units (high C/O ratio) at which all Fe would be stable as metal (plus some Si) whereas the bodies formed beyond 1.1–1.7 AU were more oxidized (low C/O and H2/H2O ratios) with a proportion of Fe as FeO dissolved in silicates (Rubie et al., 2015). Melting of the interior of planetesimals enhanced by C‐O‐H species might also contribute to variations of their redox state with consequent volatile depletion and oxidation (or reduction) of the surrounding mineral matrix before magma can reach the surface and degas (Fu et al., 2017). The variations in oxidation state of the planetesimals are inferred from the intrinsic oxygen fugacity estimated for various meteorite classes (Brett and Sato, 1984), as discussed in Section 2.2.1.3.
The Oxidation State of Earth’s Interior, Other Planets, and Meteorites.
Fig. 2.1 summarizes the estimated fO2 of the Solar Nebula, achondrites, and chondrites along with estimates of the fO2 for the interiors of planets including Mars, Mercury, Venus, Earth, and the Moon (Righter et al., 2016). Aubrites and achondrites are the most reduced materials, followed by ordinary chondrites and the oxidized carbonaceous chondrites. These partially overlap with the fO2 determined for Mercury, Martian basalts, and Earth’s mantle rocks; the latter, however, displays a uniquely broad range of values. Although there is geochemical evidence supporting bulk Earth’s origin by accretion from enstatite chondrite material (Javoy, 1995), the fO2 displayed by both appears quite different. Indeed, considerations of the Si isotopic variation and Mg/Si ratio of enstatite chondrites with respect to fertile mantle rocks (Fitoussi & Bourdon, 2012), as well as initial volatile content, are in favor of a multi‐component mixing model involving more than one type of material, with a dominating enstatite chondrite (EH or EL) component, ~30% of carbonaceous chondrites, ~24% achondrites (angrites or eucrites), and a minor amount of ordinary L‐type chondrites (Liebske & Khan, 2019). Importantly, it has been pointed out that Si isotopes are fractionated during core formation (e.g., Georg et al., 2007), nebular processes (Savage & Moynier, 2013), and evaporation during accretion (Pringle et al., 2014). Computational models suggest that the resulting fO2 after accretion must have been low enough (IW‐2) to allow core–mantle separation (Rubie et al., 2015; Schaefer & Fegley, 2017). Early Earth differentiation set a stratified redox interior, with a metallic core below and an FeO‐bearing silicate mantle above the IW buffer. Core‐mantle separation followed by magma ocean crystallization, meteorite impact processes (e.g., Late Heavy Bombardment) involving more oxidized impactors (“heterogeneous accretion”; e.g., O’Neill, 1991) and loss of primordial atmospheres are all likely to have played a major role in the gradula and stepwise oxidation of Earth’s interior (Armstrong et al., 2019), although tight quantitative constraints remain elusive. Recent models suggest that the escape of H2 from the early atmosphere must have been limited being the H2/H2O ratio < 0.3 in the gas released from the magma ocean, and this implies more oxidizing conditions (> IW+1) than those established during equilibrium with the core (Pahlevan et al. 2019).
2.3. MANTLE OXIDATION STATE OVER TIME AND ITS EFFECT ON THE C–O–H VOLATILE SPECIATION
The present‐day large oxygen fugacity range in Earth’s mantle (Fig. 2.1) reflects the evolution of the mantle through the pressure‐driven crystallization of the magma ocean to form Fe‐bearing minerals potentially able to set the fO2 locally (e.g., bridgmanite, ferropericlase, majorite, garnet; McCammon, 2005), trigger volatile outgassing and ingassing by recycling. Given that there is controversy regarding the dominant oxygen‐buffering species (e.g., Fe vs. C) even in the modern convecting mantle (e.g., Ballhaus & Frost, 1994; Cottrell & Kelley, 2013; Eguchi & Dasgupta, 2018), it is not surprising that to date it remains unclear how the redox state of the terrestrial mantle has evolved over time, and how this might have influenced volcanism, and ultimately, the atmospheric composition on Earth. Whether the chemical interaction between mantle rocks and volatile species like C played the most important role in determining conditions for the geological activity and habitability of our planet depends on the abundance of Fe (Fe0, Fe2+, and Fe3+) species compared to the abundance of C in its different forms (C0, CO, CO2, CH4) relevant in processes such as partial melting of rocks, melt migration, and diamond formation. However, three important variables influence the circulation of volatiles, among them carbon, in a planet: (i) the primordial volatile budget; (ii) the bulk oxygen fugacity of the early accreted terrestrial body; and (iii) the mineralogy of the planet’s interior. The deep volatile cycle in any planet is inextricably linked to changes of these variables through time.
2.3.1. The Redox State of the Upper Mantle and the Mobilization of Deep C
Carbon can occur in the Earth’s interior in different forms as elemental species (graphite and diamond), linked with iron as metal carbides (Fe3C, Fe7C3, etc.), in association with hydrogen (e.g., methane CH4 and a variety of aqueous organic and inorganic species) or in oxidized forms such as CO2 carbonates (either liquid or solid). Its fluxing properties are strongly related to the availability of oxygen (redox state) within the surrounding mineral assemblages: elemental carbon passively follows the convection rates of the solid mantle while oxidized carbon present in fluids or melts has its own dynamics, moving through grain boundaries at rates ruled by the fluid mechanics and chemical interactions with minerals (Hammouda & Laporte, 2002; Stagno et al., 2019). The redox state of the convecting oceanic mantle and the continental lithosphere has been recently determined employing oxy‐thermobarometry on natural rocks (mantle xenoliths) from depths of 30–200 km, which is based on the knowledge of ferric/ferrous ratios in redox‐sensitive silicates (Fig. 2.2; Frost & McCammon, 2008; Stagno et al., 2013; Stagno et al., 2015; Stagno, 2019). These studies point out that the upper mantle xenoliths, as sampled at the time of emplacement of the magmatic host rock (predominantly in the Phanerozoic), mostly fall in the diamond stability field (Fig. 2.2), and that fO2 of the lithospheric mantle decreases with the depth to conditions where Fe(Ni) alloy is stable. A large fO2 interval at a given depth is evident, reflecting small variations in the chemistry (or mineral mode) of the plotted rocks and/or the local effect of metasomatism. The experimental calibration of oxybarometers that relies on the Fe3+ of natural garnets can be used to estimate the fO2 of ultrahigh‐pressure metamorphic eclogites and cratonic mantle eclogite xenoliths (Stagno et al., 2015). These results are shown in Fig. 2.2 and demonstrate that peridotite and eclogite rocks record similar mantle redox states. A portion of the mantle likely coexisted with graphite/diamond along with C‐O‐H fluids (Luth & Stachel, 2014) and/or CO2‐bearing silicate melts such as kimberlite, whereas a small
