Phase Relations for the Former Basaltic Crust of the Slab in the Perovskitite Facies of the Lower Mantle

Abstract

Irifune & Ringwood (1993) showed that subducted slabs become less dense than surrounding perovskititic lower mantle at depths between 660-800 km because of the survival of garnet in former basaltic crust. At these depths, this latter lithology consists of stishovite, calcium silicate perovskite and an aluminous phase with the Caferrite structure, plus garnet which is no longer majoritic but approaches compositions on the pyrope-almandine join. MORB had not yet transformed to perovskitite at the highest pressures that Ringwood & Irifune attained (27 GPa). Indeed, O'Neill and Jeanloz (1994) have recently proposed that garnet in the subducted slab may persist well into the lower mantle, to depths of ~ 1200 km (~ 50 GPa) There is as yet no experimental demonstration of the widely accepted belief that the basaltic crust of the slab would eventually transform to a perovskititic lithology stable within the lower mantle. We have accordingly investigated phase relations in some analog systems that are relevant to the transformation of MORB to perovskitite, and have also carried out experiments with a synthetic MORB glass. Experiments were carried out in a diamond anvil high pressure cell, and heated with the infrared beam of a YAG laser. The laser is unstabilised and so no attempt was made to measure or control temperature. Pressure was monitored by means of standard ruby fluorescence techniques.. Samples were recovered from the diamond anvil cell and a thin foil was prepared by ion-beam thinning. Experimental phases were characterised by transmission electron microscopy and energydispersive microanalysis. We have carried out reconnaisance experiments