[S-MP47_1PO1] Physics and Chemistry of Minerals

[SMP47-P06] Experimental determination of post-spinel transition boundary in Fe₂SiO₄

*Taisuke MATSUZAWA¹, Hiroshi KOJITANI¹, Masaki AKAOGI¹ (1.Department of Chemistry, Gakushuin University)

Keywords:Fe₂SiO₄, post-spinel, spinel, high-pressure

It is widely accepted that (Mg,Fe)₂SiO₄ ringwoodite is the most abundant mineral in the mantle transition zone. Because spinel-type Fe₂SiO₄ is the endmember of (Mg,Fe)₂SiO₄ ringwoodite, many investigators have been studied on phase transitions of Fe₂SiO₄ spinel (Kawada 1977, Ohtani 1979, Morooka 1992, Katsura et al. 1998). Spinel-type Fe₂SiO₄ decomposes into 2Fe_xO+SiO₂(stishovite)+2(1-x)Fe above about 18GPa. However, the dissociation boundary has not yet been established well due in part to difficulty in oxygen fugacity control. In this study, we determined the post-spinel phase boundary in Fe₂SiO₄ by high-pressure experiments controlling oxygen fugacity with the Fe-FeO buffer.A starting material of high-pressure experiments was a mixture of Fe₂SiO₄ (fayalite), Fe_xO and Fe with molar ratios of 10:2:1, and it was packed in a Fe capsule. Oxygen fugacity of the sample at high pressure and high temperature was controlled by the Fe-FeO buffer. The high-pressure experiments were performed using a Kawai-type 6-8-type multi-anvil apparatus at 16-20GPa and 1000-1400℃. The starting samples were heated at the desired conditions for 3-6 hours, and then quenched and decompressed to ambient conditions. Recovered samples were identified by using powder XRD method and SEM-EDS, and then lattice parameters of FexO were determined by using powder XRD. The x values in Fe_xO were estimated from the composition-lattice parameter relationship of Fe_xO by McCammon (1993).The post-spinel transition boundary in Fe₂SiO₄ was determined to be P(GPa) = -0.0021T(℃) + 20.0 in the temperature range of 1000-1400 ℃. The phase boundary has a negative slope. Our boundary is almost consistent with those of Ohtani (1979) and Katsura et al. (1998). Katsura et al. (1998) interpreted that the negative slope of the boundary in the previous studies was apparent which was caused by slow kinetics of spinel decomposition. Because our study indicated that the decomposition of Fe₂SiO₄ spinel completed in the runs for 3 hours at 1000℃ and that x values of Fe_xO in the run products for 3 and 6 hours at the 1000℃ were approximately equal, we conclude that heating at 1000℃ for at least 3 hours was enough to reach the equilibrium. Because our transition boundary was determined by the runs for 6 hours at 1000℃, 3 hours at 1200℃, and 3 hours at 1400℃, we suggest that the negative slope of the post-spinel transition boundary in Fe₂SiO₄ is not apparent but the essential feature.