HR: 0830h
AN: T11C-0397    [PDF]
TI: Kinetics of grain-growth in wadsleyite: implications for point defect chemistry
AU: * Nishihara, Y
EM: yu.nishihara@yale.edu
AF: Yale University, Department of Geology and Geophysics, New Haven, CT 06520 United States
AU: Shinmei, T
EM: shinmei@sci.ehime-u.ac.jp
AF: Yale University, Department of Geology and Geophysics, New Haven, CT 06520 United States
AU: Karato, S
EM: shun-ichiro.karato@yale.edu
AF: Yale University, Department of Geology and Geophysics, New Haven, CT 06520 United States
AB: We investigate the kinetics of grain-growth in wadsleyite for two reasons. First, grain-growth kinetics controls the grain-size of wadsleyite in the mantle transition zone which in turn controls the rheology in that region. Second, the detailed knowledge of grain-growth kinetics will provide us with important constraints on the defect-related properties of this mineral which may control other properties such as diffusion, electrical conductivity and creep. We carried out the grain-growth experiments by using KIWI 1000-ton Kawai-type multi-anvil apparatus installed at Yale University. Starting material was synthesized from powdered San Carlos olivine. The grain-growth experiments were conducted at 15 GPa and 1100-$1500\deg$C for 1-24 hours. We used Mo, Ni and Re foil capsules, in order to control the oxygen fugacity by metal-oxide buffer. For ''wet'' experiments (water-saturated), a mixture of talc and brucite was packed into a capsule together with a wadsleyite sample separated by metal foils. We used a Au-Pd outer capsule which is known to be a good barrier for hydrogen diffusion. Water content in each sample was determined after an experiment by FTIR analysis of a doubly polished thin section. Grain-size was measured on a polished section using an intercept method. One of the difficulties in these experiments is to reduce the amount of water in wadsleyite. Even in nominally ''dry'' experiments in which no water is added, a significant amount of water (upto $\sim$25,000 H/10$^{6}$ Si) was detected, which comes presumably from some components in the sample assembly such as the cement. This water-uptake by wadsleyite can be minimized by surrounding it with a Au-Pd capsule. In this truly ''dry'' sample assembly, the water content of wadsleyite (after an experiment) is reduced to less than $\sim$100 H/10$^{6}$ Si, a water content similar to typical ''dry'' experiments on olivine. Compared at similar water content, the kinetics of grain-growth in wadsleyite is significantly slower than that in olivine. The oxygen fugacity is shown to have a strong effect on grain-growth kinetics. At ''wet'' conditions, the grain-growth becomes significantly faster with increasing oxygen fugacity. In contrast, at ''dry'' conditions, preliminary results suggest that the effect of oxygen fugacity is opposite to the ''wet'' conditions. This suggests that the dominant defect that controls the rate of grain-growth may be different between ''dry'' and ''wet'' conditions. In contrast to the strong dependence of grain-growth kinetics on oxygen fugacity, the total amount of water dissolved in wadsleyite depends very weakly on oxygen fugacity. This implies that the defect that controls the rate of grain-growth in wadsleyite is different from the dominant defect. Our study demonstrates that chemical environment including water and oxygen fugacity has important influence on grain-growth kinetics in wadsleyite. Therefore careful control and/or characterization of chemical environment is essential in experimental studies on this type of defect-related properties, and the evaluation of effects of these parameters in Earth is needed when these results are applied to Earth.
DE: 3630 Experimental mineralogy and petrology
DE: 3900 MINERAL PHYSICS
DE: 3904 Defects
DE: 3924 High-pressure behavior
DE: 3934 Optical, infrared, and Raman spectroscopy
SC: Tectonophysics [T]
MN: 2003 Fall Meeting