High-Pressure Studies of Oxides and Fluorides: Analogs for Ultra-High Pressure Behavior of Planetary Silicates.

Abstract

Analog materials are compounds that can be used as models of polymorphism in planetary silicates at lower, more experimentally accessible pressures. Super-Earths, defined here as extrasolar planets of 1-10 Earth masses, are expected to have pressure, temperature conditions much higher than the Earth. Terrestrial-type exoplanets are estimated to be Earth-like but may encompass a broader range of bulk compositions. Phases in the deep interior may therefore be substantially different, which in turn may affect the evolution and dynamics of these planets. In this work I studied analogs for two major compositions relevant to earth mantle: (Mg,Fe)SiO3 and SiO2. I combined laser-heated diamond anvil cell experiments combined with in situ synchrotron x-ray diffraction to reveal the structure of high-pressure phases at pressure and temperature conditions exceeding 2 Mbars and 5000 K.

Germanates have long been recognized as one of the most useful analogs for silicates. Here I focused on FeGeO3 as an analog for the high-pressure behavior of iron-rich silicates. In this work, I have reported both the room temperature and high-temperature behavior of FeGeO3 to 123 GPa and ~2100 K. FeGeO3 clinopyroxene was found to undergo a pressure-induced phase transition at ~18 GPa at room temperature. The diffraction pattern was consistent with the previously identified FeGeO3 (II) structure. On increasing pressure, the diffraction peaks broadened and there was evidence for another phase transition between 54 and 86 GPa. This unknown phase remained stable at room temperature up to 123 GPa, the peak pressure considered here. In contrast to previous studies, I did not find any evidence for formation of FeGeO3 in the perovskite structure at ~ 40 GPa and 300 K. On laser heating at 33, 54 and 123 GPa, FeGeO3 dissociated into a mixture of GeO2 (rutile/ CaCl2, α-PbO2 and pyrite-type phases respectively) and FeO (B1/ B1 + rB1). This result is consistent with previous findings for FeSiO3, which has been shown to dissociate into its respective oxides (SiO2 + FeO) up to 149 GPa.

Neighborite, NaMgF3, can serve as an analog for the-high pressure behavior in the MgSiO3 system at conditions relevant to large extra-solar planets. It is recognized to be a useful analog on the basis of structural parameters such as lattice parameter ratios, bond length ratios, and atomic positions. I examined the high-pressure phase transitions in NaMgF3 to 1.6 Mbars. This work elucidated for the first time the complex sequence of high-pressure phases beyond post-perovskite in an ABX3 system. I showed that the phase transition sequence in NaMgF3 is: NaMgF3 (perovskite)  NaMgF3 (post-perovskite)  NaMgF3 (Sb2S3-type)  NaF (B2-type) + NaMg2F5 (P21/c)  NaF (B2) + MgF2 (cotunnite-type). My results demonstrated the existence of a post-post-perovskite phase, followed a two-stage dissociation into binary fluorides. This work is the first experimental report of a dissociation of a post-perovskite in any ABX3 system.

Crystalline GeO2 follows a similar sequence of phase transitions as crystalline SiO2 but at substantially lower pressures. I carried out the first systematic study of structures and equations of state across a wide pressure range in this system. The α-PbO2-type (Pbcn) phase was synthesized by laser heating at 51 GPa from the α-quartz starting material and examined between 51 and 90 GPa, yielding the following EOS parameters: V0 = 53.8 (2) Å3, K0T = 293 (7) GPa with fixed K_0T^'=4. The pyrite-type phase was then synthesized and examined up to the peak pressure of 119.5 GPa. The equation of state fit to the data yielded: V0 = 50.3 (3) Å3, K0T = 342 (12) GPa with fixed K_0T^'=4. Theoretical calculations using different exchange correlation functions were also performed and found to agree with the experimental results. The experimental results were also compared with shock wave data for GeO2. I showed that the high-pressure phase observed during shock compression of rutile-type GeO2 was most consistent with transformation to either the CaCl2-type or the α-PbO2-type structure on shock-wave timescales.

Group 4 and 14 dioxides including HfO2, SnO2, and PbO2 provide a means to explore compact, highly coordinated structures in the AO2 system. Such phases may have novel properties including ultra-low compressibility and high hardness. I constrained the high-pressure behavior of these compounds using synchrotron x-ray diffraction and theoretical density functional theory-based calculations up to ~2 Mbars and 6 Mbars, respectively. I showed that the ninefold-coordinated cotunnite-type (Pnam) structure was stable in SnO2 and PbO2 up to the peak pressure considered (205 and 203 GPa respectively). Experimentally, HfO2 was found to transform into an Fe2P-type phase (P6 ̅2m, also nine coordinated) at 125 GPa with a Clapeyron slope of +6.9 MPa/K. Theoretical calculations were in excellent agreement with experimental findings and provided evidence for two additional phase transitions at 305 – 314 (Fe2P-type to cotunnite-type) and 390 – 469 GPa (cotunnite-type to Ni2In-type). The transition sequences predicted in these oxides were consistent among three different exchange-correlation functionals and could be explained by the energetic competition of stationary electronic flat bands and a pressure-induced shift of electronic states to lower energies.

Using these different analog systems, I identified novel phase transitions in silicate and oxide systems. Notably, a similar sequence of transitions is predicted to occur in MgSiO3 and SiO2 at ultrahigh pressures where it has implications for the mineralogy and dynamics in the deep interior of large, rocky extra-solar planets.