Constraining wet quartz rheology from constant-load experiments

Quartz rheology in the presence of H₂O is crucial for modelling (numerical and geophysical) the deformation behavior of the continental crust and gives important insights into crustal strength. Experimental studies in the past have determined stress exponent (n) values for flow law between ≤ 2 to 4, while the values for activation energy (Q) vary from ~120 to 242 kJ/mol. Here, we investigated the quartz rheology under high-pressure and high-temperature conditions, using a new generation hydraulically-driven Griggs-type apparatus. In order to develop a robust flow law for quartzite, we performed constant-load coaxial deformation experiments of natural coarse-grained (~ 200 μm) high purity (> 99 % SiO₂) quartzite from the Tana quarry (Norway). Our creep tests were carried out at 750 to 900 °C on the as-is (no added H₂O) and 0.1 wt.% of H₂O added samples under 1 GPa of confining pressure. In contrast to earlier strain rate stepping experiments, the constant-load procedure needs lower strain at each step (≤1−2%) to achieve steady-state conditions. As a consequence, there is a very low amount of recrystallization. Importantly, we can determine the Q-value independently of the stress exponent (n). Microstructures from the deformed samples were characterized using polarized light microscopy (LM), SEM-cathodoluminescence (CL), and Electron backscatter diffraction (EBSD).

Our creep results for both the as-is and 0.1 wt.% H₂O-added samples yield Q = 110 kJ/mol and n = 2. Our microstructural analysis suggests that the bulk sample strain is accommodated by grain-scale crystal-plasticity, i.e., dislocation glide (dominantly in prism <a>) with minor recovery by sub-grain rotation, accompanied by grain boundary migration and micro-cracking. It is inferred that strain incompatibilities induced by dislocation glide are accommodated by grain boundary processes, including dissolution precipitation and grain boundary sliding. These intra-grain and grain-boundary processes together resulted in a lower n-value of 2 for the quartzite.

Our new flow law predicts strain rates that are in much better agreement with the inferred natural values than the earlier flow laws. It further suggests that the strength of the continental crust considering quartz rheology is significantly lower than previously predicted.