We became aware of uncertainties about lattice parameters for osmium in Table 1 and B.K. Godwal (with the Department of Earth and Planetary Science at UC Berkeley) has reanalyzed the radial diamond anvil cell diffraction data, measured in situ at high pressure and compressive stress, with the Moment Pole Stress model in an advanced MAUD Rietveld technique (Wenk et al., 2014), taking into account lattice distortion under stress. [Table presented] The new results compare much better with data from hydrostatic experiments (Godwal et al., 2012). Since the emphasis of the paper was on mechanical twinning, we did not pay much attention to lattice distortion. Note that standard deviations are based on the Rietveld fit.

The lower mantle is estimated to be composed of mostly bridgmanite and a smaller percentage of ferropericlase, yet very little information exists for two-phase deformation of these minerals. To better understand the rheology and active deformation mechanisms of these lower mantle minerals, especially dislocation slip and the development of crystallographic preferred orientation (CPO), we deformed mineral analogs neighborite (NaMgF3, iso-structural with bridgmanite) and halite (NaCl, iso-structural with ferropericlase) together in the deformation-DIA at the Advanced Photon Source up to 51% axial shortening. Development of CPO was recorded in situ with X-ray diffraction, and information on microstructural evolution was collected using X-ray microtomography. Results show that when present in as little as 15% volume, the weak phase (NaCl) controls the deformation. Compared to single phase NaMgF3 samples, samples with just 15% volume NaCl show a reduction of CPO in NaMgF3 and weakening of the aggregate. Microtomography shows both NaMgF3 and NaCl form highly interconnected networks of grains. Polycrystal plasticity simulations were carried out to gain insight into slip activity, CPO evolution, and strain and stress partitioning between phases for different synthetic two-phase microstructures. The results suggest that ferropericlase may control deformation in the lower mantle and reduce CPO in bridgmanite, which implies a less viscous lower mantle and helps to explain why the lower mantle is fairly isotropic.