In Situ Observations of Phase Changes in Shock Compressed Forsterite

Abstract

Shockwave data on mineral‐forming compounds such as Mg2SiO4 are essential for understanding the interiors of Earth and other planets, but correct interpretation of these data depends on knowing the phase assemblage being probed at high pressure. Hence, direct observations of the phase or phases making up the measured states along the forsterite Hugoniot are essential to assess whether kinetic factors inhibit the achievement of the expected equilibrium, phase‐separated assemblage. Previous shock recovery experiments on forsterite, which has orthorhombic space group Pbnm, show discrepant results as to whether forsterite undergoes segregation into its equilibrium phase assemblage of compositionally distinct structures upon shock compression. Here we present the results of plate impact experiments on polycrystalline forsterite conducted at the Dynamic Compression Sector of the Advanced Photon Source. In situ X‐ray diffraction measurements were used to probe the crystal structure(s) in the shock state and to investigate potential decomposition into periclase and bridgmanite. In contrast to previous interpretations of the forsterite shock Hugoniot, we find that forsterite does not decompose but instead reaches the forsterite III structure, which is a metastable structure of Mg2SiO4 with orthorhombic space group Cmc21.

A Metastable Fo-III Wedge in Cold Slabs Subducted to the Lower Part of the Mantle Transition Zone: A Hypothesis Based on First-Principles Simulations

“…Provided that the Fo-II phase does not break down to the phase assemblages such as the Brg + Pe and Aki + Pe, it eventually transforms to the Fo-III phase. These theoretically calculated phase transition sequences, Fo → Fo-III in the P interval of~22-24.9 GPa and Fo → Fo-II → Fo-III at P >~32.2 GPa, are broadly compatible with the experimental observations of Finkelstein et al [37] and Newman et al [39]. In the shockwave experiments of Newman et al [39], the transition pressure for the Fo → Fo-III phase transition was constrained as~44 GPa, much higher than our predicted values (~22.1-24.9 GPa).…”

“…Comparing our LDA results with the experimental data, it is evident that the experimentally determined volumes have been surprisingly overestimated, by~1.5% on average, which might be attributable to the high defect density in the Fo-III crystals, as observed by Finkelstein et al [37]. The Fo-III phase was experimentally discovered lately [37,39]. More importantly, it formed from the Fo starting materials at very high P (P > ~58 GPa in the hydrostatic compression experiments or P > ~44(3) GPa in the dynamic shockwave experiments), and became amorphous as the experimental P decreased to ambient P. Consequently, nearly all of its thermodynamic properties are presently undetermined.…”

“…However, some new high-P polymorphs, compositionally identical to but structurally different from Ol, have been discovered in the last decade or so, which thus introduces new variables to the metastable Ol wedge hypothesis. These new metastable high-P polymorphs were discovered in meteorites [35,36], in in-situ X-ray diffraction single-crystal compression experiments [37,38], in polycrystalline shock-wave compression experiments [39] and in first-principles simulation experiments [40]. Using micro-Raman spectroscopy and high-resolution transmission electron microscopy combined with molecular dynamics simulations, Van de Moortèle et al [35] reported a new polymorph (ζ-(Mg, Fe) 2 SiO 4 ) in some shocked Ol grains of the Martian meteorites NWA 2737 and NWA 1950, but did not document the structural details.…”

“…Summarily, our extensive phonon dispersion data suggest that the Fo-III phase is the best candidate as a metastable phase forming from the Fo starting material, and the Fo-II phase is likely an intermediate transitional structure with no dynamic stability. Interestingly, recent shock-wave compression experiments with polycrystalline Fo staring materials performed at 44(3) and 73 (5) GPa have showed a direct transformation from the Fo phase to the Fo-III phase, without involving the Fo-II phase (Newman et al [39]). In addition, the identity of the Fo-IV phase has not been established, either theoretically or experimentally.…”

“…This claim is strongly supported by the excellent agreement between our values and those obtained by Hernandez et al [44] using similar theoretic techniques (the GGA method) but with slightly different simulating parameters (e.g., the planewave cutoff energy of 500 eV): for the Fo, Wds, Rwd, Aki, Brg and Pe phases, the absolute relative difference is~1.9-19.9% for the α,~0.7-2.1% for the S,~0.1-2.2% for the C V , and~0.1-1.4% for the C P . The Fo-III phase was experimentally discovered lately [37,39]. More importantly, it formed from the Fo starting materials at very high P (P >~58 GPa in the hydrostatic compression experiments or P >~44(3) GPa in the dynamic shockwave experiments), and became amorphous as the experimental P decreased to ambient P. Consequently, nearly all of its thermodynamic properties are presently undetermined.…”