Back to Search Start Over

Connector theory for reusing model results to determine materials properties

Authors :
Marco Vanzini
Ayoub Aouina
Martin Panholzer
Matteo Gatti
Lucia Reining
LSI - Spectroscopie théorique (ST)
Laboratoire des Solides Irradiés (LSI)
Commissariat à l'énergie atomique et aux énergies alternatives (CEA)-École polytechnique (X)-Centre National de la Recherche Scientifique (CNRS)-Commissariat à l'énergie atomique et aux énergies alternatives (CEA)-École polytechnique (X)-Centre National de la Recherche Scientifique (CNRS)
Ecole Polytechnique Fédérale de Lausanne (EPFL)
European Theoretical Spectroscopy Facility (ETSF)
Johannes Kepler Universität Linz - Johannes Kepler University Linz [Autriche] (JKU)
Uni Software Plus GmbH
Synchrotron SOLEIL (SSOLEIL)
Centre National de la Recherche Scientifique (CNRS)
the Austrian science Fund FWF under Project No. J 3855-N27
European Project: 320971,EC:FP7:ERC,ERC-2012-ADG_20120216,SEED(2013)
Source :
npj Computational Materials, npj Computational Materials, 2022, 8, pp.98. ⟨10.1038/s41524-022-00762-2⟩
Publication Year :
2022
Publisher :
HAL CCSD, 2022.

Abstract

The success of Density Functional Theory (DFT) is partly due to that of simple approximations, such as the Local Density Approximation (LDA), which uses results of a model, the homogeneous electron gas, to simulate exchange-correlation effects in real materials. We turn this intuitive approximation into a general and in principle exact theory by introducing the concept of a connector: a prescription how to use results of a model system in order to simulate a given quantity in a real system. In this framework, the LDA can be understood as one particular approximation for a connector that is designed to link the exchange-correlation potentials in the real material to that of the model. Formulating the in principle exact connector equations allows us to go beyond the LDA in a systematic way. Moreover, connector theory is not bound to DFT, and it suggests approximations also for other functionals and other observables. We explain why this very general approach is indeed a convenient starting point for approximations. We illustrate our purposes with simple but pertinent examples.

Details

Language :
English
ISSN :
20573960
Database :
OpenAIRE
Journal :
npj Computational Materials, npj Computational Materials, 2022, 8, pp.98. ⟨10.1038/s41524-022-00762-2⟩
Accession number :
edsair.doi.dedup.....03f0fbefdefb4ea49950e512503fb52b