Geoscience Reference
In-Depth Information
of our knowledge, the chondritic model works for the relative distribution of the refrac-
tory lithophile elements in the Earth. In particular, the chondritic concentrations of the
147 Sm- 143 Nd and 176 Lu- 176 Hf pairs of parent-daughter nuclides in chondrites are reliable
indicators of planetary values and can be used as a robust reference for the Nd and Hf iso-
topic evolution of the Bulk Silicate Earth. This is in contrast to the 87 Rb- 87 Sr, 238 U- 204 Pb,
and 187 Re- 187 Os pairs in which at least one of the nuclides is either volatile (Rb, Pb) or
siderophile (Os).
One aspect of the composition of the Earth and its mantle relates to the formation of
its core. As indicated by the extinct radioactivity of 182 Hf (see below), the metallic core
almost certainly formed within a few tens of My of the Earth's accretion. Core segregation
is now seen as a runaway process releasing enormous amounts of gravitational energy,
enough to melt the iron-nickel alloy and a substantial fraction of the mantle. It might
be expected that during the segregation of the metallic core, highly siderophile elements
strongly partitioned into liquid iron and were nearly completely removed from the mantle.
Geochemical evidence tells us otherwise. Some moderately (Ni, Cr) to strongly (Pt, Re,
Os) siderophile elements are far more abundant in the mantle than they would be if the
mantle had been in contact with iron at the time it migrated into the core. Isotopic analyses
for the parent-daughter 187 Re- 187 Os pair indicate that the Re/Os abundance ratios in the
mantle are chondritic, while the metal/silicate partition coefficients are very high (
10 000)
and different for both elements. A widely accepted explanation of this is that, shortly after
the core segregated, the Earth was subjected to intense meteoritic bombardment, and the
highly siderophile elements observed in the mantle reflect this chondritic input, termed
the late veneer. This process replenished mantle Re, Os, Ni, etc., in relative abundances
identical to those found in meteorites.
A good model of the elemental composition of the Earth can provide some idea of its
core composition. The properties of the Earth's magnetic field indicate that the core is
composed of iron. Seismic shear waves do not propagate in the upper part of the core,
which must therefore be liquid. The planetary inventory of nickel, the presence of nickel
in iron meteorites, and metallurgical data, all indicate that the core is essentially an iron-
nickel alloy (8% Ni). The densities obtained from wave propagation speeds also point to
the presence of light elements such as oxygen, sulfur, and less probably magnesium and
silicon. To evaluate core composition, McDonough (1999) suggested comparing the rela-
tive abundances of lithophile elements (elements that are a priori absent from the core) and
siderophile elements of equivalent volatility (measured by the condensation temperatures
of these elements). For elements such as Fe, Ge, As, and S, the difference in abundance of
lithophile and siderophile elements of equivalent volatility standardized to that of the Earth
reveals a deficit represented by the elemental composition of the core. For all its elegance,
this exercise depends on an assumption that, although reasonable, it is difficult to test and
leads to considerable uncertainties.
The issue of core composition is surprisingly germane to the question of where and when
terrestrial seawater originated. It is reasonably well established by astrophysical models of
star evolution that either the Earth's environment was too hot to allow any atmosphere to
accrete or that any original atmosphere was rapidly blown off from the Earth's surface
by the intense radiation emitted by the nascent Sun. It seems reasonable to assume that
>
 
 
Search WWH ::




Custom Search