Decaying shock studies of phase transitions in MgO‐SiO₂ systems: Implications for the super‐Earths' interiors

Abstract

We report an experimental study of the phase diagrams of MgO, MgSiO3, and Mg2SiO4 at high pressures. We measured the shock compression response, including pressure‐temperature Hugoniot curves of MgO, MgSiO3, and Mg2SiO4 between 0.2–1.2 TPa, 0.12–0.5 TPa, and 0.2–0.85 TPa, respectively, using laser‐driven decaying shocks. A melting signature has been observed in MgO at 0.47 ± 0.04 TPa and 9860 ± 810 K, while no such phase changes were observed either in MgSiO3 or in Mg2SiO4. Increases of reflectivity of MgO, MgSiO3, and Mg2SiO4 liquids have been detected above 0.55 TPa (12760 K), 0.15 TPa (7540 K), 0.2 TPa (5800 K), respectively. In contrast to SiO2, melting and metallization of these compounds do not coincide, implying the presence of poorly electrically conducting liquids close to the melting lines. This has important implications for the generation of dynamos in super‐Earth's mantles.

Thermodynamically complete equation of state of MgO from true radiative shock temperature measurements on samples preheated to 1850 K

“…Their shock reflectivity measurements were done at two wavelengths, 532 and 1064 nm. Despite all these improvements, Bolis et al [23] observed largely the same main features of MgO already reported by McWilliams et al [22]. Yet, they attributed the large light intensity increase encountered during shock decay, attributed by McWilliams et al to the B2 to B1 solid-solid transition, instead to a transition from melt to B2.…”

“…Upon compression along its principal Hugoniot, MgO goes from the B1 (NaCl structure) solid phase into the B2 (CsCl structure) solid phase, then into the B2-liquid coexistence region, and finally reaches the pure liquid phase above approximately 620-700 GPa [18,21]. Interpolation between the melting points measured for B1 phase at the highest attainable static pressure of ≈40 GPa [3,4] and those measured for B2 phase at approximately 470 to 650 GPa [22,23] or ≈700 GPa [21] is subject to large errors comparable to the scatter between various theoretical predictions.…”

“…None of the recently reported shock experiments to approximately 1 TPa pressure [17,18,[21][22][23][24] was designed to probe even indirectly the actual melting of B1 phase MgO. These studies probed decaying [17,22,23], steady [18], or almost steady [21] states along the principal Hugoniot or quasi-isentrope [24] and so they only characterize, at best, the B2 melting curve and the B1-B2 solid-solid phase transition.…”

“…These studies probed decaying [17,22,23], steady [18], or almost steady [21] states along the principal Hugoniot or quasi-isentrope [24] and so they only characterize, at best, the B2 melting curve and the B1-B2 solid-solid phase transition. Moreover, all the laser shock experiments so far published have very large uncertainties on MgO pressure and temperature due to the severe challenges associated with such measurements.…”

“…Bolis et al [23] repeated the measurements of McWilliams et al [22] with some improvements in experimental design and data analysis. First, they eliminated target preheating by employing pushers with high-Z x-ray absorbing layers optimized to minimize spurious shock wave reverberations in the studied samples.…”