Imaging material yielding and phase transitions in shock-compressed matter.

Abstract

Decades of research have investigated the mechanical properties of matter under extreme conditions, exploring how materials deform, yield, and fail when subjected to high strain rates. While plasticity codes and computational methods can provide atomic-scale precision, reliable experimental benchmarks are needed to validate these models. However, for decades only bulk and surface measurements (e.g., velocimetry and reflectivity) were available to the community, and data interpretation was mainly done relying on conventional plasticity models and drawing parallels with results obtained under quasi-hydrostatic conditions. In recent years the advent of X-ray free electron lasers has provided insight into the microscopic structure of shock-compressed matter, revealing substantial differences with static compression experiments. Information at the mesoscale was, however, still missing, and the onset and propagation of the deformation at this length-scale was only accessible using computational methods; in this paper we fill that void, providing experimental data at the relevant time- and length-scales. Here, we combine imaging and structural characterization in situ using a nanosecond pump laser coupled with a femtosecond X-ray probe in a novel experimental configuration at the LCLS. Our data provides a comprehensive characterization of the deformation of silicon, a simple, yet highly debated, model system for high-strength materials. Information spanning the macroscopic and mesoscopic scale down to the lattice level allows us to resolve and directly visualize the nucleation and growth of the high-pressure phase for the first time, providing a temporal constraint on its kinetics. These novel insights are crucial to unambiguously determine how silicon yields under shock-compression, as they are able to connect the structural evolution at the atomic level with the complex multi-wave dynamic observed at the macroscopic scale, reliably benchmarking decade-old theoretical predictions.