Structural and frictional properties of phyllosilicate-rich faults and implications for their seismogenic potential

Phyllosilicates minerals are a family of silicates characterized by a platy habitus derived from a layered structures where strong and weaker layers are alternated producing a marked strength anisotropy. Phyllosilicates such as micas, serpentines, talc, and clays, are abundant along fault zones belonging to all tectonic settings and can be found as primary product of the cataclastic deformation of phyllosilicate-rich rocks or as a secondary product of long-term fluid-rock interaction. The widespread occurrence of these minerals makes them a key controlling factor of the mechanics of faulting in the brittle crust. In the last 30 years an increasing attention has been given to the experimental characterization of the frictional properties of phyllosilicates. Experimental data show that phyllosilicate-rich fault rocks generally exhibit low friction, low healing, and rate-strengthening behavior over a large variety of boundary conditions. According to existing theory, these rocks due to their frictional properties cannot host earthquakes nucleation and thus deform aseismically. However, an increasing amount of geological, geophysical, and seismological data suggest that phyllosilicate-rich faults can be seismogenic or promote earthquakes nucleation elsewhere (e.g., on adjacent faults or asperities). Therefore, phyllosilicate-rich fault can represent an unexpected seismic hazard whose underlying mechanisms are still not well understood. To shed light on the seismogenic potential of phyllosilicate-rich fault rocks, I applied an interdisciplinary approach that integrates rock deformation experiments, structural and microstructural analysis with seismology. In the first chapter, I characterized the interplay between mineralogy, fabric, and frictional properties of phyllosilicates-granular mixtures, to understand how fault fabric influences frictional properties. The results show a strong influence of fault fabric on frictional properties highlighting how secondary fabric features, such as the shape of grains, the regularity of foliation, grain mantling, and the degree of localization, can be responsible for small but sensible changes in the frictional properties of rocks. In the second chapter, I investigated the role of the phyllitic basement of central Apennines in the Central Italy 2016-2017 seismic sequence. The results of the two works contained in this chapter show that the basement is rheologically heterogeneous and made of frictionally strong quartz- rich lenses enveloped by a frictionally weak phyllosilicate-rich matrix. The matrix controls the overall rheology of the basement favoring aseismic deformation that is compatible with the seismicity cutoff observed in correspondence of the basement. Furthermore, I found that the observed seismicity cutoff is not controlled by the canonical brittle-ductile transition, but it is rather controlled by the frictional properties of the basement rocks. Following the mainshock of the sequence, the basement shows a diffused microseismicity organized in clusters of co-located earthquakes. This behavior 4 can be explained as the seismic reactivation of small faults within the lenses loaded by accelerated but aseismic creep of the matrix caused by the post-mainshock loading-rate increase. The overall results of this chapters unveil the mechanisms behind the fault slip behavior of a rheologically heterogeneous phyllosilicate-rich shear zone. In the third chapter I shed light on the seismogenic potential of the clay-rich shallow subduction zones focusing on the nucleation of slow slip and tsunami earthquakes. The results of this chapters show that clay-rich faults can contemporaneously creep and nucleate slow slips in the form of slow ruptures driven by structural, frictional, and stress heterogeneities together with alternative frictional mechanisms which are imaged with an innovative technique that allows a direct observation of the fault surface during deformation. In the four and last chapter I assessed the potential of a phyllosilicate-rich fault selected for a limited fluid-induced reactivation that will take place within the activities of the Fault Activation and Earthquake Rupture (FEAR) project in the fully monitored environment provided by BedrettoLab (Swiss Alps). The results of this study highlight the strong influence of frictional, microstructural, and permeability properties on the fault slip behavior during fluid injection activity and therefore on its seismogenic potential. The overall findings presented in this PhD offer a new perspective on the seismogenic potential of phyllosilicate-rich faults with significant implication for earthquakes mechanisms and seismic hazard.