Phase Changes and Transport Properties of Geophysical Materials Under Shock Loading

Citation

Holland, Kathleen Gabrielle (1999) Phase Changes and Transport Properties of Geophysical Materials Under Shock Loading. Dissertation (Ph.D.), California Institute of Technology. doi:10.7907/cbd3-mk54. https://resolver.caltech.edu/CaltechTHESIS:10062023-181634008

Abstract

The lower mantle of the Earth is believed to be largely composed of (Mg, Fe)O (magnesiowiistite) and (Mg,Fe)SiO₃ (perovskite); thus the high pressure phase of (Mg,Fe)₂SiO₄ (olivine), which is believed to be perovskite plus magnesiowiistite is of geophysical interest. Radiative temperatures of single-crystal olivine starting material [(Mg_(0.9), Fe_(0.1))₂SiO₄] decreased abruptly from 7040 ± 315 to 4300 ± 270 K upon shock compression above 80 GPa. The data indicate that an upper bound to the solidus of the magnesiowiistite and perovskite assemblage at 4300 ± 270 K is 130 ± 3 GPa. These conditions correspond to those for partial melting at the base of the mantle, as has been suggested to occur within the recently discovered ultra-low-velocity zone (ULVZ) beneath the Central Pacific. We construct speculative high pressure phase diagrams for the MgO - SiO₂ system using experimental data from our work, and other mineral physics experiments.

In separate experiments, time dependent shock temperatures were measured for stainless steel (SS) films sandwiched between two transparent Al₂O₃ anvils. The anvil material was the same as the driver material so that there would be symmetric heat flow from the sample. Inferred Hugoniot temperatures, T_h, of 5000 - 8500±500 Kat 222- 321 GPa are consistent with previous measurements in SS. Temperatures at the film­ anvil interface (T_i), which are directly measured (rather than T_h) indicate that T_i did not decrease measurably during the approximately 250 ns that the shock wave took to traverse the Al₂O₃ anvil. Thus an upper bound is obtained for the thermal diffusivity of Al₂O₃ at the metal/anvil interface of K ≤ 14 ± 5 cm²/s at 208 GPa and 2110 K. This is a factor of 1.6 lower than previously calculated values, resulting in a decrease of the inferred T_h by at least 400 K. The observed shock temperatures are combined with temperatures calculated from measured Hugoniots and are used to calculate the thermal conductivity of Al₂O₃. There was no measurable radiant-intensity decrease during the time when the shock wave propagated through the anvil; we infer from this that Al₂O₃ remained transparent while in the shocked state. Thus an Al₂O₃ anvil is sufficiently transparent for shock temperature measurements for metals, to at least 240 GPa.

Finally, shock temperature experiments employing a six-channel pyrometer were conducted on 200, 500, and 1000 Å thick films of Fe sandwiched between 3 mm thick anvils of Al₂O₃ and LiF, to measure the thermal diffusivity ratios of Al₂O₃/Fe and LiF/Fe, at high temperatures and pressures. Temperature decays of 3000 ± 800 K in 250 ns were observed at Fe pressures of 194 - 303 GPa, which reflect the conduction of heat from the thin metal films into the anvil material. These results were achieved in experiments employing LiF anvils at 164 - 166 GPa and 4190 - 4220 K, and Al2O3 anvils at 196 - 303 GPa and 1410 - 2750 K. Thermal modeling of interface temperature versus time yields best fit thermal diffusivity ratios ranging from 15 ± 30 to 80 ± 20 (Fe/anvil) over the pressure and temperature range of the experiments. Calculated thermal conductivities for Fe, using electron gas theory, of 110 - 212 W /mK are used to calculate thermal conductivities for the anvil materials ranging from 6 to 12 W/mK. Debye theory predicts higher values of 8 to 34 W/mK. Data from previous experiments on thick (≥ 100µm) films of Fe and stainless steel are combined with our present results from experiments on thin (≤ 1000 Å) films to infer a 5860 ± 390 K Hugoniot temperature for the onset of melting of iron at 243 GPa. Our results address the question of whether radiation observed in shock temperature experiments on metals originates from the metal at the metal/ anvil interface or from the shocked anvil. We conclude that the photon flux from the shocked iron/anvil sandwich recorded in all experiments originates from the metal. Within the uncertainties of the shock temperature data, the uncertainties in shock temperatures resulting from the radiation from the anvils is negligible. This is in direct disagreement with previous conclusions of Kondo.

Thesis Files