Influence of Al, C, N and H on the iron redox state in the Earth’s lower mantle: A geochemical quantum model

Iron is the most abundant element by weight in our planet, the dominant component of the core and the only major transition metal in the mantle. Its speciation in the present-day lower mantle remains one of the most controversial aspects to deal with in modelling the deep interiors of the Earth. Here, we present an unconventional approach that relies upon quantum mechanics and the bonding Bader theory to predict the iron oxidation state at lower mantle conditions (24/1900–90/2700 GPa/K) in bridgmanite, the major mineral phase. This approach provides insights into the lower mantle geochemistry on a global scale and a unified viewpoint. The chemical species that on account of their electronic/steric features and mobility expectably induce redox effects on iron in bridgmanite are Al, N, C and H. Hydrogen causes reduction, whereas the other species promote oxidation. The combination of the probability of occurrence of the Al-N-C-H driven reactions with the availability of the involved species points to iron never achieving full oxidation; instead, it reaches a maximum average oxidation number of ∼ 2.4. This is equivalent to a Fe3+/Fetot ratio that varies with depth from 15.9 to 12.1 % (if Al-N-C-H are accounted for), and from 19.3 to 29.0 % (if only Al is considered). Iron in the lower mantle is therefore more reduced than previously expected, in terms of ferric fraction, because of the important reducing action of H. If we assume that Fe3+ is always associated with iron disproportionation (3Fe2+→2Fe3++Fe0), then the Al-N-C–H atom exchange reactions yield an estimate of metallic iron fraction in the lower mantle as large as ∼ 0.4 wt%. This figure increases up to ∼ 0.8 wt% when neglecting N-C–H effects on ferric iron formation and is fully comparable to the latest experimental result (0.7 wt%) using aluminium only.