JpGU-AGU Joint Meeting 2017

Presentation information

[JJ] Oral

S (Solid Earth Sciences) » S-MP Mineralogy & Petrology

[S-MP44] [JJ] Physics and Chemistry of Minerals

Sun. May 21, 2017 9:00 AM - 10:30 AM A03 (Tokyo Bay Makuhari Hall)

convener:Hiroaki Ohfuji(Geodynamics Research Center, Ehime University), Seiji Kamada(Frontier Research Institute for Interdisciplinary Sciences, Tohoku University), Chairperson:Seiji Kamada(Frontier Research Institute for Interdisciplinary Sciences, Tohoku University)

9:15 AM - 9:30 AM

[SMP44-02] High-pressure phase transitions of MgCO3 under the lower mantle conditions

*Fumiya Maeda1, Seiji Kamada2, Tatsuya Sakamaki1, Naohisa Hirao3, Yasuo Ohishi3, Akio Suzuki1 (1.Department of Earth Sciences, Graduate School of Science, Tohoku University, 2.Frontier Research Institute for Interdisciplinary Sciences, Tohoku University, 3.Japan Synchrotron Radiation Research Institute)

Keywords:magnesite, high-pressure polymorph, lower mantle, LHDAC

MgCO3 is one of the important carbonate minerals in the deep Earth because it can be a carbon carrier from the surface to the mantle in subduction processes. Such deep carbonates may be involved in melting of subducted rocks and formation of deep diamonds in the mantle transition zone or lower mantle. MgCO3 has especially been suggested to be the most stable carbonate under the high-pressure and temperature conditions.
Recent experimental and theoretical studies reported high-pressure phase transitions of MgCO3 under the lower mantle conditions. A report of the high-pressure polymorph was started from ‘magnesite II’ at 115 GPa in Isshiki et al. (2004). Several recent studies supported a monoclinic MgCO3 (‘phase II’) as a post-magnesite phase above ~80 GPa.
Phase II was revealed to be composed of (C3O9)-ring units which were constituted by three CO4 tetrahedra sharing three oxygen atoms. However, the structure of phase II has a little difference between the previous studies: the space groups of phase II were reported as C2/m and P21/c in Oganov et al. (2008) and Boulard et al. (2011), respectively. Moreover the latest study by Pickard and Needs (2015) observed a new high-pressure polymorph having a triclinic lattice as a post-magnesite phase above 85 GPa, which changed to phase II at 101 GPa.
In addition to the difference of the post-magnesite phase and phase-II structures, none of the above studies decided the phase boundary of the high-pressure phase transitions at high temperature. Therefore, we have been studied the phase relation of MgCO3 up to the lowermost mantle conditions based on high-pressure and temperature experiments. We especially focused on the phase boundary of the high-pressure polymorphs at high temperature and a true post-magnesite phase.
The starting material was a natural magnesite from Bahia in Brazil. The experimental conditions were up to 138 GPa and 2900 K generated using a double-sided laser-heated diamond anvil cell (LHDAC). Culet diameters of diamond anvils used were between 130 and 250 μm. The sample was loaded into a sample chamber in a tungsten gasket which was pre-indented to 40–60 μm in thickness and drilled a 60–80-μm hole in diameter. Laser heating was conducted using a fiber laser. Pt or Au was used as a laser absorber. Run products were detected using synchrotron X-ray diffraction (XRD) measurements at beamline BL10XU of SPring-8 in Hyogo, Japan. Experimental pressures were measured using a thermal equation of state of Pt or Au (Fei et al., 2007) and thermal pressures were calculated using Mie-Grüneisen-Debye model (e.g., Fei et al., 1992). XRD patterns were analyzed using IPAnalyzer and PDIndexer software (Seto et al., 2010).
We observed the two high-pressure polymorphs of MgCO3, which might be monoclinic phase II and triclinic phase reported in Oganov et al. (2008) and Pickard and Needs (2015), respectively. Phase II was observed mainly above 90 GPa and the lattice constants were estimated to be a = 8.209 Å, b = 6.575 Å, c = 6.978 Å, β = 104.06°, and V = 365.3 Å3 at 100 ± 2 GPa and 2080 ± 230 K when fitted using the Oganov’s space group, C2/m. The triclinic phase might be appeared as a post-magnesite phase around 90 GPa: The XRD patterns were not explained only by magnesite and phase II. Strong peaks near 104 diffraction of magnesite are considered to be derived from the triclinic phase although we could not fit their patterns and estimate the lattice constants due to lack of the peak number to fit the triclinic unit cell.
We could estimate the phase boundaries of high-pressure polymorphs based on the above observations. The triclinic post-magnesite phase may have a very narrow stability field in the P-T phase diagram. The triclinic phase might buffer the significant structural change from magnesite (CO4 triangles) to phase II (C3O9 rings composed of three CO4 tetrahedra).