GSA Annual Meeting in Seattle, Washington, USA - 2017

Paper No. 208-11
Presentation Time: 10:45 AM

STORAGE OF C IN OLIVINE AND CARBONATED MELTING IN THE MORB SOURCE REGION (Invited Presentation)


HIRSCHMANN, Marc M., Earth Sciences, U. Minnesota, Minneapolis,, MN 55455, AMSTRONG, Lora, Earth Sciences, U. Minnesota, Minneapolis, MN, MN 55455 and HAURI, Erik H., Department of Terrestrial Magnetism, Carnegie Institution of Washington, 5241 Broad Branch Rd NW, Washington, DC 20015-1305, mmh@umn.edu

Unlike H2O, which is hosted in the mantle chiefly in the principal silicate minerals, carbon is thought to reside chiefly in accessory minerals such as carbonate, graphite or diamond. Early studies in the 1980s found conflicting evidence for C storage in olivine, but were hampered by high detection limits. Using high sensitivity SIMS analyses, Shcheka et al. (2006) found appreciable C only in forsterite precipitated from sodium carbonate melt near the limit of its stability (13 ppmw, 11 GPa), and considerably less at lower pressure (<4 ppmw, 7 GPa, < 0.4 ppmw, 1.5 GPa). This seemed to confirm that the ~15-40 ppm C (50-150 ppmw CO2) stored in depleted (MORB-source) mantle is present chiefly in accessory minerals, except perhaps in the deepest upper mantle (350-400 km). However, compared to Shcheka et al. (2006) at similar pressures, Rosenthal et al. (2015) found greater C in minerals grown from carbonated basaltic melts up to 3 GPa. To investigate further, we performed multi anvil experiments at 3, 4.8, and 6.6 GPa and 1250-1550 °C, growing olivine from 13C-doped carbonate rich (~15-30 wt.%) silicate melts. Experiments were conducted in Re capsules, except for 1 in a graphite capsule. Large (>100 micron) olivine, and in one case, cpx, crystals were grown from melt by programmed cooling at 0.5 or 1°C/min and then held at 24 hours at the final temperature prior to quench. Nanosims analyses indicate up to 5.5 ppmw in olivine at 3 GPa and up to 10 ppmw at 6.6 GPa. Considerable variation in C contents of olivine from experiments under similar conditions leaves open the possibility that olivine-melt equilibrium has not been approached. However, if the experiments approximate equilibrium, then for a depleted source with 60% olivine and 15-40 ppm C; 50-150 ppmw CO2), much (15-40%) of the C in the oceanic mantle may be hosted in silicate at 200 km depth and proportionally less (8-25%) at 100 km. Progressive release of C from olivine yields a source of small fractions of deep carbonated melt over a wide depth interval during deep upwelling beneath ridges. Though significant C may be liberated to melts from accessory phases, marginal solubility of C in olivine indicates that peridotite/melt fractionation of C is controlled by silicate mineral/melt partitioning and that the effective partition coefficient of C during melting is small but finite.