Experimental determination of carbon solubility in Fe-Ni-S melts

Zhou Zhang, Patrick Hastings, Anette von der Handt, Marc M Hirschmann

Research output: Contribution to journalArticlepeer-review

7 Scopus citations

Abstract

To investigate the effect of metal/sulfide and Ni/Fe ratio on the C storage capacity of sulfide melts, we determine carbon solubility in Fe-Ni-S melts with various (Fe + Ni)/S and Ni/Fe via graphite-saturated high-pressure experiments from 2–7 GPa and 1200–1600 °C. Consistent with previous results, C solubility is high (4–6 wt.%) in metal-rich sulfide melts and diminishes with increasing S content. Melts with near M/S = 1 (XS > 0.4) have <0.5 wt.% C in equilibrium with graphite. C solubility is diminished modestly with increased Ni/Fe ratio, but the effect is most pronounced for S-poor melts, and becomes negligible in near-monosulfide compositions. Immiscibility between S-rich and C-rich melts is observed in Ni-poor compositions, but above ∼18 wt.% Ni there is complete miscibility. Because mantle sulfide compositions are expected to have high Ni concentrations, sulfide-carbide immiscibility is unlikely in natural mantle melts. An empirical parameterization of C solubility in Ni-Fe-S melts as a function of S and Ni contents allows estimation of the C storage capacity of sulfide in the mantle. Importantly, as the metal/sulfide (M/S) ratio of the melt increases, C storage increases both because C solubility increases and because the mass fraction of melt is enhanced by addition of metal from surrounding silicates. Under comparatively oxidized conditions where melts are near M/S = 1, as prevails at <250 km depth, bulk C storage is <3 ppm. In the deeper, more reduced mantle where M/S increases, up to 200 ppm C in typical mantle with 200 ± 100 ppm S can be stored in Fe-Ni-S melts. Thus, metal-rich sulfide melts are the principal host of carbon in the deep upper mantle and below. Residual carbon is present either as diamond or, if conditions are highly reduced and total C concentrations are low, solid alloy.

Original languageEnglish (US)
Pages (from-to)66-79
Number of pages14
JournalGeochimica et Cosmochimica Acta
Volume225
DOIs
StatePublished - Mar 15 2018

Bibliographical note

Funding Information:
We thank Tony Withers and Jed Mosenfelder for assistance with high pressure experiments and Fred Davis for discussions. We are grateful for the constructive reviews from Arno Rohrbach and two anonymous reviewers. The first author benefited from a University of Minnesota Doctoral Dissertation Fellowship and a Scripps Institution of Oceanography Postdoctoral Fellowship during completion of this work. We gratefully acknowledge the support of grants NSF EAR1119295 and EAR1426772.

Funding Information:
We thank Tony Withers and Jed Mosenfelder for assistance with high pressure experiments and Fred Davis for discussions. We are grateful for the constructive reviews from Arno Rohrbach and two anonymous reviewers. The first author benefited from a University of Minnesota Doctoral Dissertation Fellowship and a Scripps Institution of Oceanography Postdoctoral Fellowship during completion of this work. We gratefully acknowledge the support of grants NSF EAR1119295 and EAR1426772 .

Publisher Copyright:
© 2018 Elsevier Ltd

Copyright:
Copyright 2018 Elsevier B.V., All rights reserved.

Keywords

  • Deep carbon
  • Mantle redox
  • Metal-rich sulfide melt

Fingerprint Dive into the research topics of 'Experimental determination of carbon solubility in Fe-Ni-S melts'. Together they form a unique fingerprint.

Cite this