Automated Symbolic Derivation Enables Scalable Ensemble DFT for Strongly Correlated Molecules and Materials
- 1Department of Chemistry, Ulsan National Institute of Science and Technology (UNIST, UNIST-gil 50, Ulsan, Republic of Korea, 44919
- 2Center for Multidimensional Carbon Materials, Institute for Basic Science (IBS, UNIST-gil 50, Ulsan, Republic of Korea, 44919
Abstract
Strong electron correlation governs the chemistry of bond breaking, polyradicals, and excited states, yet accurately describing it requires computational methods whose cost increases steeply with system size. We present a scalable ensemble density functional theory (eDFT) methodology that extends accurate treatment of strong correlation to molecular systems of realistic size at standard Kohn-Sham DFT cost. Automated derivation of the method’s working equations and a data-driven acceleration scheme enable linear scaling with active-space size while retaining multiconfigurational accuracy. Applied to systems inaccessible to standard Kohn-Sham DFT methods, such as multiple bond dissociation, open-shell polyradicals, and a recently synthesized magnetic quantum material, the framework reproduces experimental magnetic gaps and near-exact wavefunction benchmarks using active spaces up to ten electrons in ten orbitals, thus removing a long-standing barrier to routine simulation of strongly correlated matter.
Related
- Improving coupled cluster theory for strongly correlated molecules with Kohn-Sham density encoding.
- Where Is the p-Wave Superfluid? Correlated Phase Diagram of Microwave-Shielded Polar Molecules
- Initial-State Preparation for Strongly Correlated Molecules from Matrix Product States: Benchmarking and Environment-Aware Divide-and-Conquer Compilation
- Automated Many-Body Simulations of Strongly Correlated Systems Using a Correlation-Aware Agentic Framework
- COMPLAX: A Python Tool Package for Automated Microsolvation of Organic Molecules
BibTeX
@misc{komarov2026automated,
title = {{Automated Symbolic Derivation Enables Scalable Ensemble DFT for Strongly Correlated Molecules and Materials}},
author = {Konstantin Komarov and Michael Filatov(Gulak) and Seung Kyu Min},
howpublished = {ChemRxiv},
year = {2026},
month = sep,
doi = {10.26434/chemrxiv.15008475/v1},
url = {https://chemrxiv.org/doi/full/10.26434/chemrxiv.15008475/v1},
note = {CC BY},
}
RIS
TY - UNPB
TI - Automated Symbolic Derivation Enables Scalable Ensemble DFT for Strongly Correlated Molecules and Materials
AU - Konstantin Komarov
AU - Michael Filatov(Gulak)
AU - Seung Kyu Min
PY - 2026
DA - 2026/09/08
PB - ChemRxiv
DO - 10.26434/chemrxiv.15008475/v1
UR - https://chemrxiv.org/doi/full/10.26434/chemrxiv.15008475/v1
AB - Strong electron correlation governs the chemistry of bond breaking, polyradicals, and excited states, yet accurately describing it requires computational methods whose cost increases steeply with system size. We present a scalable ensemble density functional theory (eDFT) methodology that extends accurate treatment of strong correlation to molecular systems of realistic size at standard Kohn-Sham DFT cost. Automated derivation of the method’s working equations and a data-driven acceleration scheme enable linear scaling with active-space size while retaining multiconfigurational accuracy. Applied to systems inaccessible to standard Kohn-Sham DFT methods, such as multiple bond dissociation, open-shell polyradicals, and a recently synthesized magnetic quantum material, the framework reproduces experimental magnetic gaps and near-exact wavefunction benchmarks using active spaces up to ten electrons in ten orbitals, thus removing a long-standing barrier to routine simulation of strongly correlated matter.
ER -