SOURCE-LINKED INTELLIGENCE

Modeling disorder in crystalline materials using systematically improvable correlated methods

CORDIS · observation · Publication date unknown

novel ab-initio quantum chemistry methods that challenge the prevailing dominance of density functional theory (DFT) in this area. By harnessing recent breakthroughs in Green's function techniques, machine learning, and advanced mathematical models, we propose a comprehensive suite of tools for predicting spectral and thermodynamic properties of disordered materials. The methods developed in this proposal, based on finite temperature Green’s function approaches, will enable accurate and cost-effective simulations of disordered systems, overcoming the limitations of current approaches which rely predominantly on DFT. The proposed techniques will facilitate the study of a wide range of materials where disorder is a key factor, such as high entropy alloys, superconductors, and catalytic surfaces. Additionally, we will develop machine learning models trained on ab-initio Green's function data and integrate them with ab-initio molecular dynamics to study dynamic processes in disordered systems. Finally, we will also deliver tools for learning material specific DFT functionals from the

Read original source ↗ Open in workspace

recordType

award

status

SIGNED

region

EU

value

3500000

unit

EUR