Approaching experimental time and length scales in molecular dynamics simulations of carbon at extreme conditions

Understanding carbon's behavior under extreme conditions is essential for various applications, including inertial confinement fusion (ICF) and planetary science. By

utilizing state-of-the-art quantum-accurate molecular dynamics simulations with machine learning interatomic potentials and exascale supercomputers, we made a breakthrough in uncovering the fundamental mechanisms of phase transitions in various carbon phases, both amorphous and crystalline, under dynamic compression. Our simulations, which approach experimental time and length scales, serve as a critical tool in designing high-impact experiments currently underway at world-class dynamic compression facilities such as NIF, Omega EP, and EuXFEL.