Physicochemical Controls on the Compositions of the Earth and PlanetsOPEN ACCESS 

Paolo A. Sossi, Remco C. Hin, Thorsten Kleine, Alessandro Morbidelli & Francis Nimmo

Space Science Reviews, Volume 221, article number 118, Published: 25 November 2025

LINK (OPEN ACCESS)
PDF (OPEN ACCESS)

“Despite the fact that the terrestrial planets all formed from the protoplanetary disk, their bulk compositions show marked departures from that of material condensing from a canonical H2-rich solar nebula. Metallic cores fix the oxygen fugacities (fO2s) of the planets to between ∼5 (Mercury) and ∼1 log units below the iron-wüstite (IW) buffer, orders of magnitude higher than that of the nebular gas. Their oxidised character is coupled with a lack of volatile elements with respect to the solar nebula. Here we show that condensates from a canonical solar gas at different temperatures (T0 ) produce bulk compositions with Fe/O (by mass) ranging from ∼0.93 (T0 = 1250 K) to ∼0.81 (T0 = 400 K), far lower than that of Earth at 1.06. Because the reaction Fe(s) + H2O(g) = FeO(s) + H2(g) proceeds only below ∼600 K, temperatures at which most moderately volatile elements (MVEs) have already condensed, oxidised planets are expected to be rich in volatiles, and vice-versa. That this is not observed suggests that planets i) did not accrete from equilibrium nebular condensates and/or ii) underwent additional volatile depletion/f O 2 changes at conditions distinct from those of the solar nebula. Indeed, MVE abundances in small telluric bodies (Moon, Vesta) are consistent with evaporation/condensation at ΔIW-1 and ∼1400–1800 K, while the extent of mass-dependent isotopic fractionation observed implies this occurred near- or at equilibrium. On the other hand, the volatile-depleted elemental- yet near-chondritic isotopic compositions of larger telluric bodies (Earth, Mars) reflect mixing of bodies that had themselves experienced different extents of volatile depletion, overprinted by accretion of volatile-undepleted material. On the basis of isotopic anomalies in Cr- and Ti in the BSE, such undepleted matter has been proposed to be CI chondrites, which would comprise 40% by mass if the proto-Earth were ureilite-like. However, this would result in an overabundance of volatile elements in the BSE, requiring significant loss thereafter, which has yet to be demonstrated. On the other hand, 6% CI material added late to an enstatite chondrite-like proto-Earth would broadly match the BSE composition. However, because the Earth is an end-member in isotopic anomalies of heavier elements, no combination of existing meteorites alone can account for its chemical- and isotopic composition. Instead, the Earth is most likely made partially or essentially entirely from an NC-like missing component. If so, the oxidised-, yet volatile-poor nature of differentiated bodies in the inner solar system, including Earth and Mars, is a property intrinsic to the NC reservoir.”