The largest reservoir of carbon on Earth is within its interior. A long-standing question is: Where and how is this carbon stored in the mantle? The search for carbon hosts has focused on homogeneous mineral phases as well as some carbon-rich fluids or melts. However, in a recent paper released in Nature Geoscience (4 August 2013), Jun Wu and Peter Buseck from Arizona State University provide a new perspective for answering this question .
Carbonates can account for the carbon degassing from the shallow upper mantle through partial melting, but the lower mantle has been thought by some scientists to be too reduced to contain carbonates. In the lower mantle, graphite, diamond, and iron carbide (the most obvious hosts of reduced carbon), are inefficient in terms of carbon liberation and would require combination with, and subsequent extraction from, readily fusible forms of carbon (like carbonates) to explain observed global fluxes .
Wu and Buseck, instead, envisage crystallographic defects in mantle minerals as an alternative carbon sink, broadening the inference that grain boundaries of mantle minerals could be localized sites for carbon enrichment .
To test this hypothesis, the authors adopted a novel in situ, high-pressure TEM method, which takes advantage of the remarkable property of electron-induced self-contraction of graphitic nanocages . Nanocrystals of two different analog materials to assumed mantle minerals were squeezed inside these “nanopresses” at pressures up to ~7 and ~8 GPa and temperatures of 605 and 770 °C, respectively, in an electron microscope. In this way, previously unobtainable in situ TEM observations at high pressures were made.
With adequate diffusion, the carbon atoms displaced from the wall into the core of the nanocages were found to concentrate along stacking faults in both materials with concentrations of 3-5%. This measurement corroborates the role of crystal defects such as stacking faults as effective sites for storing carbon in mantle conditions. Considering the similar mechanism of impurity-defect interaction, the authors believe such a role can be reasonably extended to other defect types, including twin boundaries and dislocations. Estimates based on some observed defect densities indicate that, with an adequate carbon supply, the overall carbon content could possibly reach as much as ~2,500 parts per million by weight (ppmw) - values that are comparable to ~1000 ppmw estimated from volcanic activities , for a mantle with an equivalent population of defects.
This new mechanism for carbon storage implies it is not essential that mantle carbon on the order of 1,000 ppmw must reside in separate carbon-rich phases. It also offers an easy approach to justify the efficient carbon degassing on a large scale of the whole mantle, because carbon stored in this way has higher diffusivities than when in crystalline forms and thus can be readily released during depressurization. Finally, these results demonstrate the utility of this new method of high-pressure, high-temperature measurements within a TEM, adding to the stable of existing high-pressure tools for the earth sciences.
Featured image taken from Figure 1, Wu and Buseck 2013.