Experimental studies of uranium, palladium, silver, and lead: Partitioning and phase stability.

다국어 초록 (Multilingual Abstract)

I seek an understanding of current and past planetary processes and events through constraining the chemical composition of the Earth's core. I use experimental techniques to simulate nature in a laboratory setting rather than to analyze natural samples. The range of techniques and equipment varies from one atmosphere furnaces to diamond anvil cells, but this thesis contains results mostly from piston cylinder and multianvil experiments. In chronological order of research conducted, the projects of my graduate work that are detailed in this thesis are (1) Phase stability of PbS with pressure; (2) Partitioning of U between molten sulfide and molten silicate; (3) Partitioning of Ag and Pd between molten sulfide and molten silicate.
1. PbS phase stability. This project was conceived to help resolve a classroom discussion about the Pb paradox; but it ended up a very different project more of interest for phase equilibrium and material science questions than for mantle/core geochemistry. Experimentally, it started as a difficult technical problem because molten PbS reacts with or wets most standard capsule materials, making it very difficult to contain. This technical issue was resolved with grade HP boron nitride capsules which proved to be relatively chemically inert and resistant to wetting. The study employed differential thermal analysis and conductivity measurements to define the phase boundaries of PbS. Liquidus slopes of PbS and other similarly structured and sized compounds correlate well with volume change on melting. However slopes do not correlate with published values of entropy of melting as is required from the Clapeyron equation, calling into question the accuracy of published entropy data. Galena was found to transform from cubic to orthorhombic structure at 26 kbar for all sub-liquidus temperatures.
2. U partitioning. The motivation for this study originated from the need to constrain the quantity of radioactive elements in the core in order to determine their contribution to planetary heat flux and geodynamo generation. K has been shown to partition into iron sulfide under magma ocean relevant conditions. Performing a similar study on U is a natural extension. This study employed piston cylinder and multianvil experiments to investigate the partitioning behavior of U in the liquid metallic sulfide-liquid silicate system. Previously existing analytical problems with measuring low U metal contents were overcome by use of LA-ICP-MS. This study concluded that, although there is variation in DUmetallic sulfide/silicate with S content, the absolute concentration of U in the core is too small to account for a significant amount of heat production or geodynamo power.
3. Pd-Ag partitioning. This study attempts to constrain the amount of 107Ag, the radiogenic daughter product of 107Pd (t1/2 = 6.5 million years), expected in Hawaiian basalts if they contain core a core component. In order to evaluate the feasibility of this finding, it is necessary to constrain how much Pd and Ag would have partitioned into the core during planetary differentiation. This study used experimental techniques from the U study to address partitioning in this system. The experimental Ds for Pd and Ag from this study are in concert with observed mantle observations, eliminating the necessity of processes in excess of a magma ocean to explain them. Applicability this study's experiments to an early Earth magma ocean is limited because of the high trace element concentrations used and the sensitivity of DPd to Pd concentration. However, insight from this study has allowed alternative interpretation of previously published data putting them into concert with mantle observations as well. Furthermore, this study highlights problems in experimentally determining partition coefficients for Pd as well as other elements. Variations in experimental chemical and physical conditions may have large impact on the shape and extent of relevant phase volumes. This can alter the composition of coexisting phases thereby impacting the calculation of D. D in this context is no longer a simple ratio of two static phases. Instead, it is a complexly changing relationship sensitive to changes in chemical and physical conditions.