Skip to Content

Non-Adiabatic Electron-Ion Dynamics and Electronic Stopping

Andre Schleife, University of Illinois at Urbana-Champaign

Usage Details

Elif Ertekin, Andre Schleife, Taishan Zhu, Joshua Schiller, Xiao Zhang, Prashun Gorai, Cheng-Wei Lee, Joshua Leveillee, Ethan Shapera, Alina Kononov, Edoardo Di Napoli, Jan Winkelmann, Kisung Kang, Tatiane Pereira dos Santos, Erick Hernandez

Understanding the impact of highly energetic particle radiation on matter from first principles is both challenging and most promising for (computational) materials science: Not only are experiments with high-energy particles dangerous and/or expensive, it is also not unusual that materials in real-life applications have to withstand remarkable radiation doses. Metal walls of nuclear reactors in future fusion or current fission systems, for instance, are subject to ion bombardment. Solar cells and semiconductor components in satellites are exposed to fast ions impacting from outer space. The underlying physics asks for an accurate description of electronic excitations caused by fast projectiles on the ultra-fast time scale associated with electron dynamics.

Very recently, we achieved an efficient numerical implementation of first-principles Ehrenfest molecular dynamics and time-dependent density functional theory into the Qbox code. In order to answer critical open questions related to the stopping of fast ions in metals and nonmetals, I will use these parameter-free approaches that are capable to describe such highly non-adiabatic electron-ion processes.

This will be the basis for computer-based predictions of how the radiation-hardness of materials can be improved. The influence of strain and point defects in the target material on the electronic stopping is a key issue in this field and, in particular, alloys such as MgZnO will be assessed as to whether they provide an efficient way to mitigate radiation damage. This will allow for an understanding of radiation damage from a very fundamental level, providing unprecedented insight that is crucial for further progress. However, the computational approaches outlined above are highly demanding, especially with respect to the required CPU time.