Fault rheology near the downdip limit of the seismogenic zone: new insights from microstructural and geochemical studies in fault cores from the Kodiak Central Belt, Alaska

Geophysical evidence of high fluid pressures and the presence of fluidized microstructures provide two independent arguments supporting the existence of fluid-like materials within the core of slipping fault zones of the crust. The nature of these materials varies depending on the specific case, including H₂O-rich fluid, ultra-comminuted rock, and melt formed after frictional slip. The persistence of such fluid-like materials over several episodes of slip is questionable, because high fluid pressures may decrease after slip and associated host-rock damage, while frictional melts solidify almost instantaneously.

To shed light on this issue, we investigated several fault zones from the Kodiak Central Belt, Alaska, which were active under peak metamorphic conditions (3.0 ± 0.4 kbar, 320 ± 20 °C). At outcrop scale, these faults cut across metamorphosed turbidites and extend for tens of meters, with fault cores up to 5 cm of thickness. Their kinematics indicate a top-to-the-SE motion, consistent with the main deformation stage in the Kodiak Central Belt. Injections of the core material into dm-long cracks in the host rock, perpendicular to the main slip plane, are locally present.

At thin section-scale, the fault cores show a multilayered structure, indicative of multiple slip events. The microstructures of these layers are variable, including cataclasites with clasts of various size surrounded by a quartz-rich cement, as well as quartz or calcite veins. The fault slip surfaces, within layers dominated by quartz, are underlined by aligned micrometric chlorite and titanium-rich inclusions.

The cement is to a large extent composed of idiomorphic quartz crystals that exhibit successive growth increments, highlighted by rims of micrometric chlorite inclusions. These chlorite inclusions share the same composition as the larger grains forming the metamorphic foliation of the host rock. The growth history of idiomorphic quartz crystals is further revealed by sharp variations in the concentration in Al, accompanied by corresponding changes in cathodoluminescence intensity. Most crystals display isotropic growth microstructures, indicating that the crystal growth occurred without steric constraints or application of a significant deviatoric stresses. Additionally, crack-seal microstructures formed in a dilatation jog along a microfault slip plane show similarly cyclical variations in Al content of the quartz cement.

These microstructures indicate that quartz crystal growth spanned multiple slip events and occurred under variable physico-chemical conditions, which influenced the differential incorporation of Al and solid inclusions into the quartz. The geometry of the growth microstructures suggests that the density and viscosity of the fluid were sufficiently high to prevent the crystals from settling down by gravity during their growth. Based on these observations, we propose that the fault core remained predominantly in a fluid state over multiple slip cycles, with viscosity variations resulting primarily from the progressive growth of crystals within the fluid. This mechanical behavior, characterized by persistently low viscosity, may correspond to the sequence of repeated slow-slip events observed in subduction zones.