4.5 Article

Nonstoichiometry and growth of some Fe carbides

Journal

CONTRIBUTIONS TO MINERALOGY AND PETROLOGY
Volume 166, Issue 3, Pages 935-957

Publisher

SPRINGER
DOI: 10.1007/s00410-013-0900-7

Keywords

Cohenite; Eckstrom-Adcock carbide; Nonstoichiometry; Growth rate; Phase diagram; Pressure

Funding

  1. National Science Foundation
  2. NSF [EAR 1219891]
  3. Packard Fellowship
  4. Division Of Earth Sciences [1219891] Funding Source: National Science Foundation

Ask authors/readers for more resources

Iron carbides containing from 31 to 17 atomic % carbon, with cohenite XRD structure and optical properties, were grown in experiments in Fe-Ni-S-C, Fe-Ni-C, and in Fe-C at 1, 6, and 7 GPa. X-ray cell volumes increase with C content. Compositions listed above vary considerably outside the nominal (Fe,Ni)(3)C stoichiometry of cohenite/cementite. Cohenites coexisting with Fe-C liquid are carbon poor. The Eckstrom-Adcock carbide, nominally Fe7C3, was found to show compositions from 29 to 36 atomic % C at 7 GPa in Fe-C. Both these materials are better regarded as solutions than as stoichiometric compounds, and their properties such as volume have compositional dependencies, as do the iron oxides, sulfides, silicides, and hydrides. The fraction of C dissolved in cohenite-saturated alloy is found to become smaller between 1 and 7 GPa. If this trend continues at higher pressures, the deep mantle should be easier to saturate with carbide than the shallow mantle, whether or not carbide is metastable as at ambient pressure. At temperatures below the cohenite-graphite peritectic, cohenite may grow as a compositionally zoned layer between Fe and graphite. The Eckstrom-Adcock carbide joins the assemblage at 7 GPa. Phases appear between Fe and C in an order consistent with metasomatic interface growth between chemically incompatible feed stocks. Diffusion across the carbide layer is not the growth rate limiting step. Carbon transport along the grain boundaries of solid Fe source stock at 1 GPa, to form C-saturated Fe alloy, is observed to be orders of magnitude faster than the cohenite layer growth. Growth stagnates too rapidly to be consistent with diffusion control. Furthermore, lateral variations in carbide layer thickness, convoluted inert marker horizons, and variable compositional profiles within the layers suggest that there are local transport complexities not covered by one-dimensional diffusive metasomatic growth. In contrast to many transport phenomena which slow with pressure, at 7 GPa and 1,162 A degrees C, carbide growth without open grain boundaries is faster than at 1 GPa with fast grain boundary channels, again suggesting C transport is less of a constraint on growth than C supply. C supply at 7 GPa is enhanced by graphite metastability and the absence of fast grain boundary channels to divert C into the Fe instead of growing carbide. At both 1 and 7 GPa, the growth rate of carbide is found to systematically vary depending on which of two stock pieces of graphite are used to form the growth couple, suggesting that some property of each specific graphite, like C release rate, possibly from amorphous binder material, may influence the cohenite growth process. At temperatures near and above the cohenite-graphite peritectic at 1-1.5 GPa, complex intergrowths involving Fe-C liquids and extensive thermal migration transport were encountered, eroding the organized spatial resolution, and the range of cohenite compositions found grown below this peritectic from growth couples of crystalline Fe and graphite. The migration of graphite to a position in the metasomatic sequence between liquid and cohenite demonstrates that the solubility of graphite in liquid increases with temperature above the peritectic, whereas the solubility of graphite in cohenite below the peritectic decreases with temperature. The variable solubility of graphite in cohenite, shown by thermal migration, emphasizes that cohenite does have compositional variations.

Authors

I am an author on this paper
Click your name to claim this paper and add it to your profile.

Reviews

Primary Rating

4.5
Not enough ratings

Secondary Ratings

Novelty
-
Significance
-
Scientific rigor
-
Rate this paper

Recommended

No Data Available
No Data Available