Early differentiation of the Earth and the problem of mantle siderophile elements: a new approach

Citation metadata

Author: V. Rama Murthy
Date: July 19, 1991
From: Science(Vol. 253, Issue 5017)
Publisher: American Association for the Advancement of Science
Document Type: Article
Length: 3,120 words

Document controls

Main content

Article Preview :

RECENT PROGRESS IN THE DEVELopment of quantitative models of the accretion of planets has indicated that the earth was largely or totally molten during its accretion [1]. For this initial state, the effect of pressure on melting in a planet of terrestrial size requires that a substantial amount of the planet would be at temperatures much higher than the low-pressure melting points of silicates and metallic iron. Differentiation to form an iron core and a silicate mantle would essentially be a high-temperature process, the temperature corresponding to the liquids temperatures of iron and silicates at the prevailing pressures as the planet accretes. On the basis of a detailed investigation of the physical process of core separation in a largely molten earth, Stevenson [2] concluded that core segregation occurred rapidly under complete chemical equilibrium between the iron metal and mantle silicates. In such a case, the abundances of siderophile elements in the mantle would be controlled by distribution coefficients ([K.sub.d]'s) applicable to the temperatures and pressures at which the core-mantle equilibrium was established and not by the [K.sub.d]'s at the relatively low temperatures measured in the laboratory. This point has not been considered so far in the use of abundances of siderophile elements in the mantle as constraints to theories of the early chemical differentiation of the earth [for example [3, 4]].

In this report, I examine a model of core formation in which the siderophile element partitioning between the core and the mantle occurred at temperatures close to the liquidus temperatures in the accreting earth. Other than the consideration of the effect of temperature, all assumptions remain the same as in earlier discussions, namely, that the partitioning of trace components between two phases occurs at equilibrium and that the phases relevant to core-mantle differentiation in the earth are dominantly Fe-metal and liquid silicates.

The temperatures at which core-mantle equilibrium in the earth was established cannot be specified exactly. Most modern theories of accretion suggest that the earth would be molten by the time it had grown to about one-tenth of its present mass [1, 5, 6]. Core formation will commence at this stage and continue through the stochastic accretion process by the infall of [10.sup.25]- to [10.sup.26]-gram planetary embryos [1]. Thus we can expect that successive fractions of the core separate from the mantle silicates at their liquids temperatures in this growing proto-earth. Because the interior pressure in the planet in the initial stages of accretion would be less than that in the present earth, the silicates and metal would have melted at temperatures lower than those at depth in the present earth. On the basis of data on melting of mantle silicates as a function of pressure, it can be inferred that core formation commenced at about 2500 K when the earth is about one-tenth of its present mass. As accretion continued, the sinking Fe-metal droplets would have equilibrated with mantle silicates at progressively higher temperatures because of the increase in the melting points of mantle silicates in response...

Source Citation

Source Citation   

Gale Document Number: GALE|A11145350