Plastic Creep and Brittle-Ductile Transition in Hydrated Rocks of the Plate Interface

Abstract:

Geophysical observations suggest that the formation of hydrous phyllosilicate-bearing rocks such as serpentinites favor aseismic slip on the plate interface. I review our current understanding of deformation of serpentines and similar phyllosilicates in the first 100 km of subduction and discuss some pending questions on measurements and modeling of the behavior and properties of the complex serpentinite rocks. Experimental studies suggest that serpentines have low enough mechanical strength to act as a "stabilizer" of stable creep, but the actual strength of serpentinites will depend on the exact nature of the crystallographic structure and fabric of the stable serpentine variety. Low-temperature, flat-layered lizardite has strong anisotropy in strength. Lizardite-serpentinite strength will depend crystal-preferred orientation (CPO), with isotropic texture having high strength (>300 MPa) and foliated serpentinites having small strength (<100 MPa), independent of temperature, pressure, and strain rate. Thus, the transition between brittle and plastic (or stable creep) behavior may result from progressive deformation. High-temperature serpentine antigorite has a complex corrugated-layered structure, and complex deformation modes were evidenced from experimental studies. Mechanical strength shows a strong stress dependence, suggesting dislocation-creep, and low temperature dependence, suggesting plastic behavior. Extrapolation of experimental results to natural strain rates suggests that antigorite-serpentinites have low strength (<100 MPa or lower), and will favor stable-creep. However, the extrapolation relies on mechanical flow laws that may not apply to serpentine. Electron microscopy observations reveals dislocation-like deformation mechanisms that are not sufficient to explain global deformation of antigorite aggregates, and that are likely accompanied by dissolution-precipitation at low natural strain-rates. Establishing reliable flow laws relevant to the subduction interface in the 30-100 km depth range will require further experimental investigations of such mechanisms.