Welcome to the new version of CaltechAUTHORS. Login is currently restricted to library staff. If you notice any issues, please email coda@library.caltech.edu
Published February 2002 | public
Journal Article

The Case for Irreversible Chemical Stratification of the Mantle


Chemical differentiation of the Earth into a buoyant, olivine-rich upper mantle, along with protocrustal materials, a perovskite-rich deeper layer, and an iron-rich core occurred continuously during accretion. Dense komatiitic liquids and eclogitic solids sank to mid-mantle depths. The large-ion lithophile elements and primordial gases accumulated in the proto-upper mantle. During subsequent evolution, most of the crustal elements were sweated out of the upper mantle; the layer at the base of the mantle collected light dross from the core and dense dregs from the mantle and reacted with the core. This fractionation and gravitational sorting of primordial materials according to density, solubility, silicate compatibility, and melting point became irreversible as the planet grew because of the effect of pressure on thermal expansion. Chemical boundaries are hard to detect by seismic techniques, but evidence favors one such boundary near 1000 km. Below this, the mantle is probably depleted in volatiles and the heat-producing elements, and represents the accreted material minus the buoyant and fusable compounds and the accompanying trace elements. Observations also favor a thick, chemically distinct layer at the base of the mantle that may extend, in places, more than 1000 km from the core-mantle boundary. This layer exhibits large-scale sluggish behavior as appropriate for high Prandtl number, low Rayleigh number convection. This kind of chemical and gravitational stratification resolves various geodynamic and geochemical paradoxes, and is more consistent with petrology and mineral physics than one- and two-layer models, and reversible stratification.

Additional Information

© 2002 by V. H. Winston & Son, Inc. Received: 02 Oct 2007. Version of record first published: 14 Jul 2010. I thank Gillian Foulger for an instructive review. This research was partially supported by the National Science Foundation. Contribution No. 8875, Division of Geological and Planetary Sciences, California Institute of Technology.

Additional details

August 19, 2023
October 19, 2023