Login / Signup

Ab Initio Static Exchange-Correlation Kernel across Jacob's Ladder without Functional Derivatives.

Zhandos A MoldabekovMaximilian BöhmeJan VorbergerDavid BlaschkeTobias Dornheim
Published in: Journal of chemical theory and computation (2023)
The electronic exchange─correlation (XC) kernel constitutes a fundamental input for the estimation of a gamut of properties such as the dielectric characteristics, the thermal and electrical conductivity, or the response to an external perturbation. In this work, we present a formally exact methodology for the computation of the system specific static XC kernel exclusively within the framework of density functional theory (DFT) and without employing functional derivatives─no external input apart from the usual XC-functional is required. We compare our new results with exact quantum Monte Carlo (QMC) data for the archetypical uniform electron gas model under both ambient and warm dense matter conditions. This gives us unprecedented insights into the performance of different XC functionals, and it has important implications for the development of new functionals that are designed for the application at extreme temperatures. In addition, we obtain new DFT results for the XC kernel of warm dense hydrogen as it occurs in fusion applications and astrophysical objects. The observed excellent agreement to the QMC reference data demonstrates that presented framework is capable to capture nontrivial effects such as XC-induced isotropy breaking in the density response of hydrogen at large wave numbers.
Keyphrases
  • density functional theory
  • molecular dynamics
  • monte carlo
  • electronic health record
  • air pollution
  • particulate matter
  • climate change
  • deep learning
  • machine learning
  • oxidative stress
  • endothelial cells