The main goal of this proposal was to test a specific hypothesis for deep focus earthquakes, which occur between 300 and 700 km depths in the deep Earth. The hypothesis of interested states that as the major mantle mineral olivine [(Mg,Fe)2SiO4] is brought into depth by subduction processes, it undergoes phase transformations to high pressure forms, known as wadsleyite and ringwoodite (spinel-related structures) in the mantle transition zone. During these phase transformations, the high-pressure phases nucleate within olivine grains, forming weakened zones to introduce rapid unstable faulting. As a team of mineral physicists and seismologists, we conducted laboratory brittle deformation experiments on olivine-structured materials to simulate faulting induced by the transformations and detected failure events that generated acoustic emission (AE). With in-situ x-ray diffraction and imaging, we followed the stress-strain behavior of deforming samples as they underwent structural changes. Then we applied seismological tools to recorded AE event waveforms to determine where and how faulting developed. We also used scanning and transmission miicroscopy to examine microstructure of recovered samples. We have studied three different materials: Mg2GeO4 (an analog olivine, which transforms directly to spinel), Mn2GeO4 (another analog olivine, which transforms to the wadsleyite-type structure), and (Mg0.5Fe0.5)2SiO4, and found that transformational faulting occurs in all of them, each generating hundreds to thousands of AE events in a single experiment. We have developed broad-band AE transducers to increase detection sensitivity and to examine characteristic frequencies of the AE events. We have also developed a larger deformation apparatus that increased sample volume by a factor of 20. We have developed a machine-learning-based algorithm to detect and automatically pick the first arrivals of AE events. By using the arrival times and the hypoDD event relocation method we were able to locate AE events, to a relative location accuracy of 0.01-0.02 mm. We have also developed a new waveform inversion method to determine events? magnitudes and source parameters using their recorded waveforms. Work is in progress to measure the corner frequencies of AE events, in hope to extend the scaling law of seismic events down to nano-seismological magnitudes. These seismological advances allowed us to map distribution of AE events in space and time, essentially following the process of transformational faulting. For this three-year project (with one year of no-cost extension), we have published 7 peer-reviewed papers, with additional 2 currently under review and several in preparation. One female student has earned her PhD through this project, two post-docs were trained. The new developments of experimental techniques at GSECARS beamlines are available to all users. The seismological tools will find wider applications not only in laboratory experiments, but also in observational seismology. We have also conducted two workshops in 2018 and 2020 to disseminate our findings between experimentalists and seismologists.

Last Modified: 10/12/2021
Modified by: Lupei Zhu