Hanaya Okuda , Manami Kitamura, Miki Takahashi, Asuka Yamaguchi

The transition from smectite to illite at around the depth with temperature of ~150 °C has long been hypothesized as one of controlling factors of the updip limit depth of the seismogenic zone of subduction zones. Although illite has a higher friction coefficient than smectite, previous friction experiments have little reported that the smectite-illite transition accompanies the transition from velocity strengthening (not seismogenic) to velocity weakening (potentially seismogenic). In this study, we simulated in-situ temperature-pressure-mineral conditions along the shallow plate boundary in the Nankai Trough, and conducted friction experiments to further constrain the role of smectite-illite transition in seismogenesis in subduction zones. We found that friction coefficient of the simulated sediment increased with the progress of illitization and followed the Reuss average of friction coefficients of the mineral phases. The obtained friction coefficients and the Coulomb wedge model inferred that the pore pressure conditions around the plate boundary may be over-pressured but not as high as lithostatic pressure. Higher friction coefficients of sediment due to diagenetic processes will be required to sustain the nearly lithostatic pore pressure along the plate boundary. The velocity dependence of friction coefficient (a−b) was always positive except under the velocity conditions lower than ~10 μm/s at 171 °C. Therefore, the shallow slow earthquakes beneath the outer prism cannot be explained only by the frictional instability of unconsolidated sediment. Since the smectite-illite transition mostly completes at ~30 km landward from the trough axis, the transition from velocity strengthening to velocity weakening observed at 171 °C is caused by temperature-dependent frictional behavior of illite-rich material. Hence the updip limit depth of the seismogenic zone of subduction zones would be mainly controlled by frictional properties of illite-rich material.