We have shown within the first years of the ERC-funded project, that phase transitions can be tracked from time-resolved x-ray diffraction data collected during continuous compression of mantle minerals. We have also demonstrated that the quality of our data allows for measuring subtle effects related to phase transitions that were undetectable previously.
One example include results on (Mg,Fe)O, the second most abundant mineral in Earth’s lower mantle, which undergoes a change of electronic structure at pressures corresponding to the mid-lower mantle. This spin crossover is predicted to cause changes in a range of physical properties of the lower mantle, with major impacts on geophysics and geodynamic. Within DEEP-MAPS we used novel time-resolved x-ray diagnostics combined with a dynamically-driven DAC to constrain the softening of P-wave velocities in (Mg,Fe)O throughout the iron spin crossover, for the first time at high temperature. High-temperature results are pivotal to distinguish between different proposed theoretical descriptions of the spin crossover. We use our result to benchmark new theoretical calculations and combined them with an expected temperature field inside the Earth based on a large-scale geodynamic model to predict the seismic signature of the spin crossover. After applying a filter to simulate a realistic resolution of geophysical techniques in the lower mantle, we compare our prediction to results from real Earth observations (seismic tomography model SP12RTS) and find strong evidence for the presence of the spin crossover. We also show that the seismic changes associated with the spin crossover might allow us to constrain mantle temperatures once seismic resolution improves.
Another example is work on a phase transition in SiO2 stishovite, a major phase in subducted oceanic crust, that occurs in Earth’s lower mantle. Detecting the transition in seismic data is of relevance for tracking SiO2 cycling from the surface through the deep mantle. In this study, we used the novel experimental setup to continuously compress a sintered sample of SiO2 stishovite to high pressures across the tetragonal-orthorhombic phase transition in the lower mantle. Different experiments have been performed under different sample stress states. The dense pressure-coverage achieved allowed us to (1) pinpoint the phase transition pressure and quantify the effect of stress on the phase transition pressure, and (2) directly calculate the bulk modulus (incompressibility) from our experimental data. We found a very strong effect of stress on the phase transition pressure, suggesting that the different mantle depths where scattering of Earthquake waves is observed could be partially explained by different stress states in the mantle. In addition, and surprisingly, we found that the polycrystalline stishovite is less compressible than single-crystal stishovite, an observation we explain by grain-grain interactions across stiff grain boundaries. This finding has implications for our understanding of how to scale from laboratory measurements on single crystal samples to rocks in the Earth.
