MSA ROEBLING MEDAL LECTURE: MINERALOGY IN ACTION -- CRITICAL CONTROLS ON GEOLOGICAL PHENOMENA AND ROCK MEMORY

The mineralogy of a rock, in addition to providing chemical information, carries information about the conditions under which geological phenomena occurred. However, during metamorphism, when the external conditions vary in such a way that the existing assemblage becomes metastable, the kinetics of reaction to the stable assemblage can become very important for phenomena such as deformation and can be not only rate-controlling but also may control the physical mechanism(s) activated. This process can be important when the kinetics of reaction are either increasing or decreasing. The critical issue is the relative rate of reaction initiation and change of geological conditions (P, T or strain rate). In this talk I will focus on subduction zone phenomena and will give one (experimental) example of control of physical mechanism of failure and one of microstructural memory of very great depth in a natural rock.

(1) In the former case, deep subduction of cold oceanic lithosphere transports olivine out of its stability field. As subduction continues, the olivine inevitably encounters conditions where nucleation of one of its denser high-pressure polymorphs is kinetically enabled (both are exothermic reactions). Sinking slabs are under stress, hence nucleation of the stable phase occurs at stress concentrations and stimulates a local thermal runaway of the reaction driven by latent heat release. The nanocrystalline reaction product of this transformation results in self-organized failure and generation of deep earthquakes.

(2) In the latter case, a deeply subducted, inherently buoyant, fragment of pelitic gneiss that had been transformed to a stishovite-bearing mineral assemblage, warmed sufficiently that it broke free from the heavier material that had dragged it down and began the journey of exhumation. The changing conditions (primarily down P) stimulated exsolution of kyanite and spinel rods in stishovite that were preserved all the way to the surface despite the fact that the host underwent sequential transformation to coesite and quartz.

Lastly, I will discuss a new discovery in which shear heating after stress-induced failure produces a profound change in flow mechanism that may explain the large drop in friction in high-speed sliding experiments and, in nature, the mechanism of sliding in large earthquakes.