In search of lost planets – the paleocosmochemistry of the inner solar system

Abstract

The depletion of moderately volatile elements in planetesimals and planets is generally considered to be a result of removal of hot nebula gases. This theory can be tested with Sr isotopes. The calculated initial 87Sr/ 86Sr of the angrite parent body (APB), eucrite parent body (EPB), the Moon and the Earth are significantly higher than the initial Sr isotopic composition of the solar system despite the volatile-depleted nature of all of these objects. Calculated time-scales required to accomplish these increases in 87Sr/ 86Sr with a solar Rb/Sr in a nebula environment are >2 Myr for the APB, >3 Myr for the EPB and >10 Myr for the Moon. These times are more than an order of magnitude longer than that expected for cooling the nebula in the terrestrial planet-forming region and correspond to the period during which most of the mass already should have been accreted into sizeable planetesimals and even planets. Therefore, incomplete condensation of the nebula does not provide an adequate explanation for the depletion in moderately volatile elements. The data are better explained by a protracted history of depletion via more than one mechanism, including processes completely divorced from the earliest cooling of the circumstellar disk. The Sr model ages are maximum formation ages of the APB and EPB and indicate that these are most probably secondary objects. With independent estimates of their minimum age, a time-integrated Rb/Sr can be calculated for the precursor materials from which they formed. These are consistent with accretion of the APB and EPB from objects that at one stage may have resembled carbonaceous chondrite parent bodies in terms of volatile budgets. At some late stage there were large losses of volatiles, the most likely mechanism for which is very energetic collisions between planetesimals and proto-planets that, in the case of the Asteroid Belt, have since been lost. The same applies to the Moon, which presently has Rb/Sr=0.006 even though the material from which it formed had a time-integrated Rb/Sr ratio of ∼0.07, consistent with a precursor planet (Theia) that was even less volatile element-depleted than the present Earth (Rb/Sr=0.03). The time-integrated Rb/Sr of Theia is similar to the present Rb/Sr of Mars (0.07). There is suggestive evidence of a similar time-integrated value for the proto-Earth (∼0.09). Therefore, prior to the later stages of planet formation involving giant impacts between large objects, the inner solar system may have had relatively uniform concentrations of moderately volatile elements broadly similar to those found in volatile-depleted chondrites. Correlations of the present Rb/Sr ratios in planets and planetesimals with ratios of other volatile elements to Sr can be used to infer the time-integrated composition of precursor materials. The time-integrated inferred K/U ratios of the proto-Earth, as well as Theia, were ∼20 000, so that early radioactive heat production may have been ∼40% greater than that calculated by extrapolating back from the Earth’s present K/U. Higher C and S bulk concentrations may have led to concentrations in proto-cores of 0.6–1.5% C and 4–10% S. These are significantly higher than those anticipated from the degree of volatile depletion of the present silicate Earth (∼0.12% C, ∼1.3% S). If the late history of accretion did not involve large-scale re-equilibration of silicates and metal, the present core may have inherited such high C and S concentrations. In this case, S would be the dominant light element in the present core.