Accelerating Driven Liouville-von-Neumann time-dependent density function calculation

POSTER

Abstract

The driven Liouville von Neumann approach12 is a method to computationally explore electron dynamics and transport in nanoscale systems. It does so by imposing open boundary conditions on finite atomistic model systems, which drive them out of equilibrium. The approach is compatible with any underlying electronic structure treatment that can be phrased in terms of a single-particle framework, ranging from simple tight-binding descriptions to state-of-the-art density functional theory treatments of the interacting system. In the latter, the requirement to recalculate the Kohn-Sham Hamiltonian at every time-step results in substantial computational burden, which prohibits the simulation of large model systems over long time-scales. Here, I demonstrate how selectively skipping the re-evaluation of the time-dependent Kohn-Sham Hamiltonian, can significantly speed up the calculation. I further present a dynamic “freezing” scheme based on pre-determined current variation criteria, with minor impact on the accuracy of the electronic and spin current dynamics.

*This work is supported by NSF award number DMR-2318872, the United States–Isarel Binational Science Foundation NSF-BSF grant 2023602, the Israel Science Foundation grants 3645/24 and 3646/24, and the Heinemann Chair in Physical Chemistry.

Publication: 1. O. Hod and L. Kronik, "The Driven Liouville von Neumann Approach to Electron Dynamics in Open Quantum Systems", Isr. J. Chem. 63, e202300058 (2023).
2. A. Oz, A. Nitzan, O. Hod, and J. E. Peralta, "Electron Dynamics in Open Quantum Systems: The Driven Liouville-von Neuman Methodology within Time Dependent Density Functional Theory",J. Chem. Theory Comput. 19, 7496-7504 (2023).

Presenters

  • Rotem Widman

    • Tel Aviv University

Authors

  • Rotem Widman

    • Tel Aviv University
  • Oded Hod

    • Tel Aviv University
  • Juan E Peralta

    • Central Michigan University
    • Department of Physics, Central Michigan University, Mount Pleasant, Michigan 48859, USA
  • Kashinath T Chavan

    • Central Michigan University