THE PARADOX OF SEISMIC RUPTURE IN SERPENTINIZED PERIDOTITES

Seismogenic rupture occurs within the mantle section of the oceanic lithosphere. This is well documented by both seismological analysis of intraslab earthquakes and by ultramafic pseudotachylytes in oceanic peridotites discovered in the last 15 years. Previous studies reported that some of these pseudotachylytes formed in serpentinized peridotites although the timing of fluid-rock interactions remains unclear. This is a first order paradox because, under 500 MPa confining pressure, the yield strength of peridotites drops sharply from ~2,000 MPa to ~500 MPa (i.e., 75% decrease in strength) when the degree of serpentinization increases from 0 to 10% (Escartin et al., 2001). A similar drop in yield strength occurs under 200 MPa confining pressure (66% decrease). Since the differential stresses, calculated from olivine clasts dislocation densities range from 300 to 600 MPa, if the host peridotite was indeed partially serpentinized, then the confining pressure during rupture should not have exceeded 600 MPa (~20 km depth). Regardless of the actual depth of rupture, seismogenic rupture typically takes place in strong rocks and therefore should not be expected in serpentinized peridotites, even under very high strain rates. Further, the melting temperature of serpentinized peridotite is ~200ºC less than dry peridotite. Therefore, the volume of frictional melt produced along the fault (generation) plane during a seismic event of given energy varies depending on the degree of hydration (serpentinization) of the peridotite. The degree of thermal efficiency, which is the fraction of work converted to heat (as opposed to radiated energy), is also a factor that affects frictional melt production. Here we show that determining whether the peridotitic host-rock­ was dry or serpentinized during a seismic event is an elusive yet fundamental piece of information to understand seismic rupture in mantle rocks.