Dislocation recovery experiments were conducted on predeformed olivine single crystals at temperatures of 1,450 to 1,760 K, room pressure, and oxygen partial pressures near the Ni-NiO buffer to determine the annihilation rate constants for [001](010) edge dislocations. The obtained rate constants were found to be comparable to those of previously determined [001] screw dislocations. The activation energies for the motion of both dislocations are identical. This result suggests that the motion of screw dislocations in olivine is not controlled by cross-slip but by the same rate-limiting process of the motion of edge dislocations, i.e., climb, under low-stress, high-temperature conditions. The diffusivity derived from dislocation climb indicates that dislocation recovery is controlled by Si pipe diffusion, rather than Si lattice diffusion. Our results suggest that the conventional climb-controlled model for olivine can be applied to motions of not only edge but also screw dislocations. Therefore, the previous proposed cross-slip model cannot explain the softness of asthenosphere.

Segregation of liquid metal from solid silicate is a necessary pathway for core formation in a large rocky planetary body during the planet growth. The mechanism and extent of such process have an important effect on the geophysical and geochemical properties of the planetary body. Percolative flow of core forming melts through a silicate mantle has been ruled out as a possible mechanism of core formation by hydrostatic annealing experiments at pressures less than 50 GPa. However, an evolving mantle is not static, but continually deforming. Here, using element migration in the melts as an effective indicator for melt connectivity, we conclusively demonstrated that iron alloy melts could form an interconnected network in a solid bridgmanite matrix under deformation, even at a small total strain of ~0.1. Depending on the grain size of bridgmanite, percolation as a core formation mechanism could leave mantle disequilibrium/equilibrium with the core. The result showed that ~0.4 vol.% liquid metal was trapped in the silicate mantle and the stranded metal alloy could explain the highly siderophile elements (HSE) chondritic abundance in the Earth’s mantle without late veneer. Plain Language Summary We use element migration as an indicator of melt interconnection to demonstrate that core-forming melt can form an interconnected network in a bridgmanite matrix under deformation. The fast segregation velocity of melts in silicate makes the stress-induced percolation a feasible core formation mechanism. The melts left in the mantle after draining could explain the highly siderophile elements abundance in the Earth mantle.

Dislocation recovery experiments were conducted on predeformed olivine single crystals at temperatures of 1,450 to 1,760 K, room pressure, and oxygen partial pressures near the Ni-NiO buffer to determine the annihilation rate constants for [001](010) edge dislocations. The obtained rate constants were found to be comparable to those of previously determined [001] screw dislocations. The activation energies for the motion of both dislocations are identical. This result suggests that the motion of screw dislocations in olivine is not controlled by cross-slip but by the same rate-limiting process of the motion of edge dislocations, i.e., climb, under low-stress, high-temperature conditions. The diffusivity derived from dislocation climb indicates that dislocation recovery is controlled by pipe diffusion. The conventional climb-controlled model for olivine can be applied to the motions of not only edge but also screw dislocations. The softness of the asthenosphere cannot be explained by cross-slip controlled olivine dislocation creep.