THE EFFECT OF DEHYDRATION on ANTIGORITE RHEOLOGY
To test the ‘dehydration embrittlement’ hypothesis we conducted temperature-ramping experiments on antigorite serpentinite. Cold-pressed powdered samples of antigorite were deformed to a high differential stress at 400 °C and 1.0 GPa, within the antigorite stability field, where we have shown that deformation localizes (Chernak and Hirth, 2010). Temperature was then increased at different rates, 1800 °C/hr and 180 °C/hr, to cross the reaction boundary while the samples continued to deform; samples were deformed at strain rates of 1.5 x 10-4 s-1, 10-5 s-1 and 10-6 s-1. Two additional experiments were conducted in a similar manner at 300 °C, 1.5 GPa and 1.5 x 10-5 s-1 but samples remained “statically” at high stress during the temperature increase.
Our results show that although the decrease in stress during temperature ramping is large, stress relaxes stably, even after dehydration. In addition, we find that stress relaxes over several minutes, which is not characteristic of an earthquake. We find that the slopes of the unloading curves are approximately the same for constant values of the ratio (strain rate/ramp rate) and that the unloading slope is greater for higher values of this ratio. In addition, we find that the unloading curves with the greatest slopes are similar to the apparatus compliance, suggesting that we are generating “slow earthquakes” in our experiments over the course 5 to 10s of minutes. Strain rate stepping experiments indicate that antigorite has velocity strengthening behavior at 700 °C and 1.0 providing an explanation for why unstable slip does not occur. Our results thus suggest that antigorite dehydration does not result in “dehydration embrittlement” but that it may promote slow earthquakes and/or slow slip events.