9.02 The composition and major reservoirs of the earth around the time of the moon-forming giant impact

The Earth formed over tens of millions of years from an unknown number of stochastic collisions with other planetary objects. Its chemical and isotopic compositions provide an integrated view of this accretion and early differentiation history. Geochemical data for the Earth and other terrestrial planets can be compared with the results of experimental simulations, theory, and astronomical observations to elucidate the conditions of planet and core formation. This chapter summarizes the latest thinking in this area and particularly focuses on the portion that we think is best understood, the state of the Earth at around the time of the last major stage of accretion, the Moon-forming Giant Impact. The age of the Moon is known to lie in the range 30–55 My after the start of the solar system because of evidence of decay of formerly live ¹⁸²Hf (half-life = 8.9 My) to ¹⁸²W within the lunar interior. As such, Earth’s accretion and differentiation was more protracted than that of Mars, Vesta, and magmatic iron meteorite parent bodies for which W and other isotope data provide evidence of internal differentiation certainly within 5 My and probably <1 My. There is no evidence that the Earth grew at a slower rate when it was of an equivalent size, merely that significant accretion persisted for longer. The Earth formed from material that was on average depleted in moderately volatile elements such as potassium relative to the concentrations found in the Sun. Isotopic data provide evidence that some of this loss took place during the growth of the Earth and/or its constituent accreting protoplanets. Noble gas isotopic compositions have been used to argue for a component of solar nebular volatiles in the Earth’s interior, a major fraction of which was probably lost during accretion. However, it is also possible that at certain times the early Earth was associated with a blanketing atmosphere such that the energy from accretion led to magma oceans. The abundances of siderophile elements in the Earth’s mantle provide evidence that some of the core formed from metal that segregated at high pressures and temperatures possibly in such an environment. Tungsten isotopes also indicate that the majority of the siderophile-element budget that was added to the Earth during its accretion isotopically equilibrated with the silicate reservoir before being segregated into the core, which is consistent with magma oceans. The small difference between the W isotopic compositon of the silicate Earth and chondrites could reflect efficient mixing and decay over the period of accretion. However, it may also partly reflect disequilibrium such as when large fragments of planetary core material mixed directly with the Earth’s core. The age of the Moon provides an independent evaluation of which explanation is correct but the uncertainty is still too large to offer very strong constraints. Experimental petrology provides evidence that the core equilibrated in stages at various pressures and temperatures during accretion of the Earth. As it did so the phases present in the mantle would have changed. The oxidation state of the mantle is thought to have evolved as a result although it is also possible that there was an evolution in the composition of accreted material. Perovskite crystallization in particular may have changed the oxidation state of the silicate Earth. Mars and Vesta differentiated at lower pressures and therefore did not undergo this self-oxidation. Even though their cores are smaller and the Fe content of their mantles is higher, the magmas are less oxidized than those of the Earth. Recently, new stable isotope tracers of accretion and core formation have been developed. It has been found that there are no significant differences between the lithium and magnesium isotopic compositions of ‘normal’ mantle-derived materials from the Earth, Moon, Mars, and Vesta. As such, there is no evidence for significant evaporative losses during the accretion processes that built the Earth and Moon. However, the iron isotopic compositions of terrestrial and lunar basalts are slightly heavy for reasons that are unclear. One possiblility is that the high-pressure perovskite-driven self-oxidation process led to enrichment in isotopically heavy iron in the silicate Earth and light iron in the core. The similar composition of the Moon would then be hard to explain unless there was some level of isotopic mixing and equilibration between the Earth and the proto-lunar disk as recently proposed to explain the identical oxygen isotopic compositions. Lead and thallium isotopes provide evidence for losses from the silicate Earth postdating the Giant Impact by tens of millions of years. These may be caused by a late stage loss of sulfide from the mantle following the addition of large amounts of sulfur from the impactor’s core. Being relatively volatile however, these elements might also have been atmophile and lost from protoatmospheres as recorded in xenon isotopes. This history of accretion and differerntiation links up with evidence from ¹⁴⁶Sm–¹⁴²Nd studies that the Hadean mantle was differentiated at an early stage. However, recent proposals that the deep mantle contains hidden primordial reservoirs left over from this period need to be treated with some caution because the degree to which the Earth is exactly chondritic in its Sm–Nd systematics is unclear.