An alternative explanation for the distribution of highly siderophile elements in the Earth.

Abstract

The presence of highly siderophile elements (Ru, Rh, Pd, Re, Os, Ir, Pt, and Au) in the mantle has been a long standing enigma in the Earth sciences. Highly siderophile elements exhibit extremely low silicate/ metal partition coefficients and so should have partitioned into the core, leaving the mantle depleted and fractionated compared with precursor material. Late accretion of undifferentiated material after completion of core formation may have overprinted the residue inherited from metal-silicate equilibrium partitioning. Here, we present a model based on the osmium isotopic composition of the mantle that sheds new light on the distribution of highly siderophile elements in Earth. As the Earth grew from the accretion of chondritic material of unspecified nature, gravitational and radioactive heating permitted early segregation of metal from silicate. This resulted in fractionation of highly siderophile element abundances in the residual mantle relative to chondritic abundances. After completion of core formation, the model supposes that a late carbonaceous veneer delivered biogenic and highly siderophile elements to the Earth. This late veneer was mixed inhomogeneously with the fractionated residue left over after core formation. Part of the deep mantle was isolated early from shallow convection and thus preserved primordial noble gas and highly siderophile element signatures. Allègre and Luck, 1980). This contradiction disappears (Morgan, 1986) if highly siderophile elements were delivered to the mantle by a late accreting veneer after completion of core formation (Morgan, 1986;Chyba, 1990;Dauphas et al., 2000). Nevertheless, recent theoretical (Murthy, 1991) and experimental (Li and Agee, 1996;Righter and Drake, 1997;Cottrell and Walker, 2002) studies on the high pressure/high temperature partition coefficients of moderately and highly siderophile elements have revived discussion of the idea that the distribution of highly siderophile INTRODUCTIONMetal/silicate differentiation was one of the most important events in early Earth history. However, we have a very poor understanding of the underlying mechanisms that governed this segregation (Stevenson, 1990). Low pressure experiments indicate that highly siderophile elements in the residual mantle left over after core formation should have been much more depleted and fractionated (Jones and Drake, 1986) than is observed (Kimura et al., 1974;Jagoutz et al., 1979; Os ratio among the various meteorite classes. This implies that it may be possible to deduce the ultimate source of mantle osmium from its isotopic composition (Morgan, 1985;Meisel et al., 1996;Walker et al., 2001). Analysis of fertile lherzolites suggests that the 187 Os/ 188 Os ratio of the primitive upper mantle that has not experienced melt extraction is 0.1296 ± 0.0008 ). This value, first proposed by Meisel et al. (1996), has been confirmed by numerous recent studies of lithospheric peridotite suites (McBride et al., 1996;Handler et al., 1997;Peslier et al., 2000;Reisberg et al., 2000;Snow ...

Inference on the nature and the mass of Earth's late veneer from noble metals and gases

“…Core formation resulted in a fractionated noble metal pattern in the residual mantle. If a late carbonaceous veneer was mixed with this fractionated residue, then the osmium signature of the mantle is reconciliable with a carbonaceous late veneer [ Dauphas et al , 2002a]. Note however that our conclusions do not rely heavily on this assumption.…”

“…We focus our discussion on osmium because its geochemistry is better known than that of other noble metals (Table 1). It has recently been suggested on the basis of the osmium isotopic composition of the mantle that there might be a concentration contrast for noble metals between the deep and the shallow mantle, the upper mantle being a factor of 3 ± 2 richer than the lower mantle [ Dauphas et al , 2002]. The osmium concentration of the upper mantle is estimated to be 0.018 × 10 −9 mol.g −1 [ McDonough and Sun , 1995], hence the osmium concentration of the deep mantle should be approximately 0.006 × 10 −9 mol.g −1 and there should be 3 − 7 × 10 16 mol of osmium in the whole mantle.…”

Hydrogen isotopic composition of water from fossil micrometeorites in howardites

“…First the PUM may not represent the bulk silicate Earth (BSE) (Dauphas et al, 2002). Second, given the materials from which the PUM estimate has been derived (i.e., Proterozoic to Phanerozoic mantle xenoliths and massifs), it is at least possible that this estimate includes recycled components such as radiogenic subducted slabs mixed back to the mantle via convection (Walker et al, 2002a) and/or pyroxenitic components (Allègre and Turcotte, 1986;Meibom et al, , 2004.…”

Inference on terrestrial genesis from molybdenum isotope systematics

“…It is thus estimated that approximately 87% of the molybdenum inventory of the upper mantle was left by core extraction and that 13% was delivered by the late veneer. Based on the osmium isotopic composition of the mantle, it was recently suggested that the late veneer was not completely homogenized in the whole mantle, its contribution being a factor of 3 or −1 higher in the upper mantle than in the lower mantle [ Dauphas et al , 2002c]. In such a case, the contribution of the residual mantle and the late veneer in the deep mantle would be 95% and 5%, respectively.…”