27 FEB. 16:00 - 17:00 Conference Room 1
Prof. Tomoo Katsura 桂智男氏 (Bayerisches Geoinstitut, University of Bayreuth)
Measurement of Si self-diffusion coefficient as a function of pressure, temperature and water content, and its application to upper-mantle rheology
Creep of mantle minerals is expected to be controlled by diffusion of their constituent species under very low strain-rate conditions, which is the case in deep mantle. Since silicon has the lowest diffusion coefficient in silicate minerals, diffusion of silicon should be the rate-limiting process of creep. For this reason, we measured lattice and grain-boundary self-diffusion coefficients of silicon in synthetic single-crystal and fine-grained aggregates of forsterite, respectively, as a function of pressure, temperature and water content. The activation energy, activation volume, and water-content exponent of the lattice and grain-boundary diffusions are found to be E_Lat=410±30 and E_(G.B.)=250±10 kJ/mole, V_Lat=1.7±0.4 and V_(G.B.)=1.8±0.2 cm3/mole, and r_Lat=0.32±0.03 and r_(G.B.)=0.22±0.05. These numbers suggest: (1) The activation energy of the lattice diffusion is smaller than that of creep obtained by deformation experiment (500 kJ/mole). (2) The activation energy of lattice diffusion is much larger than that of the grain-boundary diffusion. (3) The pressure effects on both diffusions are very small. (4) The water content dependence is significantly smaller than that given by deformation experiment (1.2). From these results, we conclude: (1) Creep of olivine in deformation apparatus is not controlled by diffusion. (2) The effects of water on upper-mantle dynamics should be limited. (3) Because of the small pressure and water-content dependence, viscosity in the upper mantle should monotonically increase with increasing depth. (4) Not the transition from dislocation to Coble creep but that from Coble to dislocation creep should occur with depth in the upper mantle.