Exploring microstructures in lower mantle mineral assemblages with synchrotron x-rays

Abstract

Understanding dynamics across phase transformations and the spatial distribution of minerals in the lower mantle is crucial for a comprehensive model of the evolution of the Earth's interior. Using the multigrain crystallography technique (MGC) with synchrotron x-rays at pressures of 30 GPa in a laser-heated diamond anvil cell to study the formation of bridgmanite [(Mg,Fe)SiO₃] and ferropericlase [(Mg,Fe)O], we report an interconnected network of a smaller grained ferropericlase, a configuration that has been implicated in slab stagnation and plume deflection in the upper part of the lower mantle. Furthermore, we isolated individual crystal orientations with grain-scale resolution, provide estimates on stress evolutions on the grain scale, and report {110} twinning in an iron-depleted bridgmanite, a mechanism that appears to aid stress relaxation during grain growth and likely contributes to the lack of any appreciable seismic anisotropy in the upper portion of the lower mantle.

Fig. 1. Schematic of the experimental setup showing the “spotty” Debye-Scherrer rings that typify multigrain crystallography measurements.

The spotty rings result when a relatively small number of distinctly oriented grains comprise the diffracting volume (black, red, and blue shapes). The detector image shows the projection of all 240 frames in the ω 0° to 60° rotation series. Each crystallite in the sample volume contributes a discrete subset of the observed reflection spots as do the diamond anvils (depicted by diamond symbols). Each distinct Bragg condition manifests as an intensity peak at the coordinates (2θ, η, and ω); 2θ is the Bragg angle converted from the radial distance to the beam center; η is the azimuth along the Debye-Scherrer ring.

Fig. 2. Identification and separation of br and fp diffraction spots at 30 GPa.

(A) Projection of 240 diffraction images following the phase transition in experiment 2 after background subtraction. Arrows indicate the locations of two reflections belonging to a single br and fp grain. Middle inset shows zoomed-in views of the selected diffraction peaks from (A), highlighting their compact Gaussian character after heating, indicating no observable plastic deformation in those grains. Note that the fp peaks are much smaller and numerous than the br, reflecting a much higher number yet small size of fp grains in the diffraction volume (B).

Fig. 3. Grain spatial and stress/strain distribution.

(A) 3D distribution of br and fp grains constrained within the 100-μm radius from sample center after each heating cycles 1 to 4 in experiment 2. Dot sizes scaled by the number of reflections assigned to each grain (legend shown in lower right). Grain colors are in % equivalent elastic strain. (B) Average equivalent stress in br (blue) and fp (red) plotted against heating cycle for both phases showing the decreasing trend in stress with each heating cycle for all indexed grains (triangles) compared to those constrained to the sample chamber (circles). Black arrows represent incident beam direction and the rotational axis of the sample in (A). Along the beam direction remains the least constrained spatial coordinate due to limited DAC rotation angles.

Fig. 4. SEM and EDS of decompressed recovered sample from experiment 2.

(A) SEM backscatter electron image of sample within the laser heating spot and corresponding maps of Si (yellow), Mg (blue), and Fe (purple). Iron appears bright in SEM-BE image. (B) Spectra collected at two locations labeled 1 and 2 in (A).

Fig. 5. Olivine and br and fp grain orientations from experiments 1 and 2 represented as PFs (columns 1 to 3 and 5) and IPFs (column 4).

(Row 1) Olivine textures extracted from 49 unique orientations in experiment 1. (Rows 2 to 3) br textures retrieved from experiments 1 (15 grains) and 2 (48 grains) after the first heating cycle at 1800 to 2000 K. (Row 4) br textures (16 grains) after third heating cycle in experiment 1 at 1800 K. (Row 5) Experiment 2 (21 grains) at 2200 to 2600 K. Increasing from blue to red indicates concentration of lattice plane normals. White and orange circles in br inverse pole figure (IPF) in row 5 represent individual orientations used for further analysis described in the following section. Column 5 displays a consistently nearly random distribution of fp with (100) PFs. Units for PFs and IPFs are in multiples of random distribution (m.r.d). PFs and IPFs were calculated and plotted using the MTEX (23, 24).

Fig. 6. Spatially resolved twinning relation observed in br.

Two pairs (one pair indicated by orange circles, the other black in IPFs, highlighted in Fig. 5, fifth row) of br grain centroids, also shown at bottom (large). Grain color shows variation in average equivalent stress (GPa) with one high and one low stress grain in each pair located within the 100-μm-diameter sample chamber (labeled). (001) PFs (right) show single c-axis maxima in each pair indicative of {110} twinning. Units for pole PFs and IPFs are in m.r.d.