Deformation and softening mechanism in naturally deformed rocks at the brittle-ductile transition zone in upper crust: pervasive micro-faulting accommodated by pressure solution of quartz

Although the conventional two-mechanism strength profile of upper crustal rheology has been applied for a long time to geodynamical problems, the high differential stresses as much as a few hundreds of MPa have never been reported from geophysical observations such as those inferred from heat flow along active faults and stress drop during earthquakes. These facts suggest that there must be some softening mechanisms around the brittle-ductile transition zone in upper crust, where inland earthquakes most frequently occur. In order to unravel softening mechanisms in these regions, we have been examining natural microstructures in exhumed metamorphic rocks, which experienced pervasive deformation at brittle-ductile conditions (T=c. 300 oC). The Sambagawa metamorphic rocks experienced localized deformation under brittle-ductile transition conditions at D2 phase during exhumation. At outcrop scales, low-angle normal faults were pervasively developed with a dominant top-to-the-NNW movement recorded in quartz slickenfibre. In quartz schist deformed at D2 phase, shear bands coated by phengite were pervasively developed. We have extensively studied quartz c-axis fabrics and microstructures in the micro-faulted quartz schist. In quartz lenses surrounded by D2 shear bands, quartz microstructures and c-axis fabrics formed at D1 phase indicative of dislocation creep were well preserved. However, in the matrix (i.e. domains outside the lenses), quartz c-axis fabrics became weakened, and in some cases, became completely random. We have analyzed the degree of undulation of recrystallized quartz grain boundaries using the index called normalized perimeter of grains to that of the equivalent ellipse obtained with the NIH image, and compared the degree among different quartz schist samples and domains in the same sample. As a result, it has been found out that there is a nice positive correlation between the degree of grain boundary undulation and c-axis fabric intensity in quartz aggregates, which are both negatively correlated with the area fraction of phengite. The facts suggest that the randomization of quartz c-axis fabrics were perhaps caused by dissolution-precipitation of quartz, and rather straight grain boundaries of quartz (i.e. lower normalized perimeter of grains) with random or weaker quartz c-axis fabrics resulted from the precipitation of quartz from the solution, which was pinned by newly grown phengite. All these composite deformation processes revealed by detailed microstructural analyses indicate that pervasive micro-faulting (i.e. formation of shear bands) was accommodated by pressure solution creep of quartz. This is consistent with a model frictional-viscous flow of phyllosilicate-bearing fault rocks proposed by Bos and Spiers (2002), by which the bulk strength of rocks is controlled by the coefficient of internal friction of phyllosilicate and becomes less than one-third of that predicted by the Byerlee's law.