[S-IT14] Deep Earth Sciences

[SIT14-20] Viscosity of the lowermost lower mantle inferred from deformation of NaNiF₃ as an analogue of (Mg,Fe)SiO₃

*Daisuke Yamazaki¹ (1.Institute for Planetary Materials, Okayama University)

Seismological observation revealed that the inhomogeneous velocity structure in the lowermost lower mantle of the Earth. For example, high velocity region observed in the circum-Pacific region may correspond to the cold area of the subducting slab whereas low velocity region in the central Pacific region may correspond to the warm upwelling plume region. In addition, variation of viscosity in the lowermost lower mantle is estimated such as low viscous in the circum-Pacific region high viscosity in the central Pacific region based on geophysical observations such as geoid. To explain these observations from the material science of the view, the knowledge of physical properties of constituting materials are essentially required. Bridgmanite, (Mg,Fe)SiO₃ perovskite, is the dominant phase of the lower mantle and it transforms into post-perovskite at the pressure corresponding to the D” layer at relatively lower temperature, suggesting that the cold region of the lower most lower mantle is mainly composed of post-perovskite and the other may be dominated by bridgmanite. In this study, therefore, we experimentally investigated the viscosity of perovskite and post-perovskite of NaNiF₃ as an analogue of (Mg,Fe)SiO3 because of experimental difficulty of post-perovskite phase. We conducted deformation experiments to estimate viscosity contrast between perovskite and post-perovskite at high pressure and temperature. By using in-situ X-ray observation technique, we determined stress, strain rate and lattice preferred orientation simultaneously. Then, we found two main results that (i) viscosity of post-perovskite is comparable to that of perovskite and (ii) effect of lattice preferred orientation of post-perovskite on viscosity is negligibly small. The present result of lattice preferred orientation effect is inconsistent with previous estimation of anisotropic atomic diffusivity. The present results indicates that it is difficult to explain viscosity variation in the lowermost lower mantle by phase transition between perovskite and post-perovskite and hence alternative is needed.