Mantle–melt partitioning of the highly siderophile elements: New results and application to Mars

Abstract

Trace elements and extant and extinct isotopic attributes in Martian meteorites have been used to argue that Mars accreted quickly, differentiated into core and mantle, and established several mantle reservoirs, possibly within 10 Ma of T0. The partitioning of trace elements in the deep mantle has been relatively unstudied, despite the need for such knowledge in understanding magma ocean crystallization and the origin of depleted and enriched mantle reservoirs. The siderophile element composition of the Martian mantle and lithophile isotopic systems such as Sr, Hf, and Nd are thought to record evidence for early metal–silicate equilibrium and deep magma ocean at an intermediate depth and pressure of 800 km or 14 GPa. We have carried out experiments across this pressure range to better understand the mineral/melt partitioning of a wide range of elements. These new data are used to evaluate differentiation models for Mars and to help interpret the available isotopic data. The relatively incompatible nature of Re compared to mildly compatible Os means that the crystallization of a deep magma ocean will lead to residual liquids with superchondritic Re/Os, and solids with subchondritic Re/Os. Such material available in the mantle could be the source of enriched isotopic reservoir that produced shergottites with +γOs values. Slightly subchondritic Re/Os ratios in the crystallizing solids would provide a reservoir that could produce −γOs values. Melting of mixtures of these two enriched and depleted endmembers could explain the Nd‐Os isotopic correlations and systematics of shergottites.

A multi-siderophile element connection between volcanic hotspots and Earth’s core

“…Such signatures would contrast with the lower Ru/Ir ratios observed among Hawaiʻi and Iceland OIB. Although there is a paucity of data indicating how Ru and Ir would behave during subsequent basal magma ocean crystallization, an experimental study of Ru and Os has revealed no significant difference in partitioning among deep-mantle phases (Righter et al, 2020). Since IPGE generally behave similarly during mantle melting, this is a priori evidence that Ru/Ir ratios would not be strongly altered by this process.…”

Earth Without Life 2: A Global Network Model of Abiotic Phosphorus Cycling on Earth Through Time

“…In some contexts P likely behaves as a volatile element during planetary accretion. Under some melt conditions, it is an incompatible element, and sometimes it also behaves as a siderophile element [10]. This latter behavior may best explain its depletion in Earth's upper geological reservoirs (e.g.…”

“…The oxidation state of P in the environment is governed by the kinetics and stoichiometry of electrons from other easily exchanged donor and acceptor elements, most importantly Fe, S and C [18]. Consideration of the oxidation state of the BSE over time suggests that P remaining in the BSE after core formation would have continuously experienced global redox conditions maintaining it in the +5 oxidation state [8,10], though localized reservoirs hosting reduced P are also possible. In modern terrestrial surface environments, P is generally encountered in the +5 state, though there is some evidence for widely distributed, though minor, reduced P [19], and the widespread distribution of genes for the oxidation of reduced phosphorus suggests there is a significant biologically exploitable reservoir of reduced P [20].…”