Coupled Exsolution of Fluid and Spinel from Olivine: Evidence for O in the Mantle?

  1. Robert N. Schock
  1. H. W. Green II

Published Online: 18 MAR 2013

DOI: 10.1029/GM031p0226

Point Defects in Minerals

Point Defects in Minerals

How to Cite

Green, H. W. (1985) Coupled Exsolution of Fluid and Spinel from Olivine: Evidence for O in the Mantle?, in Point Defects in Minerals (ed R. N. Schock), American Geophysical Union, Washington, D. C.. doi: 10.1029/GM031p0226

Author Information

  1. Department of Geology, University of California, Davis, Davis, California 95616

Publication History

  1. Published Online: 18 MAR 2013
  2. Published Print: 1 JAN 1985

ISBN Information

Print ISBN: 9780875900568

Online ISBN: 9781118664070



  • Mineralogical chemistry—Congresses;
  • Crystals—Defects—Congresses


At least some of the carbon in the earth's upper mantle is dissolved in the silicates. Recent experimental studies on olivine by Freund and coworkers have indicated carbon solubility of several hundred ppm at atmospheric pressure and very rapid diffusivities. These workers have proposed that carbon and hydrogen dissolve in olivine via a process involving a charge transfer (CT) reaction in which O ions are farmed. O is predicted to be much smaller than O2−, hence pressure should enhance such reactions. To explore this hypothesis, I here consider the chemistry of composite fluid and solid precipitates in olivine from peridotite xenoliths in kimberlite pipes. The precipitates consist of hemispheres of CO2−rich fluid and platelets of spinel. The two phases exsolved simultaneously, suggesting a link between the dissolution of carbon and the trivalent cations of the spinel. Qualitative analysis of the spinel shows that Al is subordinate to both Cr and Fe, suggesting that the exsolution process is essentially represented by $$\eqalign{ & ({\rm Fe}_{{\rm Mg}}ˆ \bullet ,\,{\rm Cr}_{{\rm Mg}}ˆ \bullet )_2 {\rm C}_{{\rm Si}}ˆ{\rm o} {\rm O}_{\rm 4} \, + \,{\rm Fe}_{{\rm Mg}}ˆ{\rm o} {\rm V}_{{\rm Mg}}ˆ{{\rm ' '}} {\rm C}_{{\rm Si}}ˆ{\rm o} {\rm O}_{\rm 4} \cr & \,\,\,\,\,\,\,\,\,\,\,\, \to \,{\rm Fe(Fe,}\,{\rm Cr)}_{\rm 2} {\rm O}_{\rm 4} \, + \,2{\rm CO}_{\rm 2} \cr} $$ if oxygen is exclusively O2−, or $$\eqalign{ & 3({\rm Fe}_{{\rm Mg}}ˆ{\rm o} ,\,{\rm Cr}_{{\rm Mg}}ˆ{\rm o} {\rm )}_{\rm 2} {\rm C}_{{\rm Si}}ˆ{{\rm ' '}\,\,{\rm ''}} {\rm O}_{\rm 4}ˆ \bullet \cr & \to \,2{\rm Fe(Fe,}\,{\rm Cr)}_{\rm 2} {\rm O}_{\rm 4} \, + \,2{\rm CO}_{\rm 2} \, + \,{\rm c} \cr} $$if O is involved. These two reactions differ in the fluid-solid ratio of the products and in the development of elemental carbon (or CO) in the second equation. The volumetric fluid-solid ratio of the precipitates is about 3:1, and the bubbles appear to have an amorphous film lining their surfaces, suggesting that the second equation is the better choice at this time. Quantitative analysis of the two phases should offer a clear choice. Previous studies involving pressure-induced reduction of aliovalent cations have invoked other CT mechanisms which are macro-scopically equivalent to the O hypothesis. Other decompression-induced reactions and the controversy about the fO2 of the mantle also may reflect operation of such CT mechanisms.