Abstract:The design of better exchange-correlation functionals for Density Functional Theory (DFT) is a central challenge of modern electronic structure theory. However, current developments are limited by the mathematical form of the functional, with efficient semilocal functionals being inaccurate for many technologically important systems and the more accurate hybrid functionals being too expensive for large solid-state systems due to the use of the exact exchange operator. In this work, we use machine learning combined with exact physical constraints to design an exchange functional that is both orbital-dependent and nonlocal, but which can be evaluated at roughly the cost of semilocal functionals and is significantly faster than hybrid DFT in plane-wave codes. By training functionals with several different feature sets, we elucidate the roles of orbital-dependent and nonlocal features in learning the exchange energy and determine that both types of features provide vital and independently important information to the model. Having trained our new exchange functional with an expressive, nonlocal feature set, we substitute it into existing hybrid functionals to achieve hybrid-DFT accuracy on thermochemical benchmark sets and improve the accuracy of band gap predictions over semilocal DFT. To demonstrate the scalability of our approach as well as the practical benefits of improved band gap prediction, we compute charged defect transition levels in silicon using large supercells. Due to its transferability and computational efficiency for both molecular and extended systems, our model overcomes the cost-accuracy trade-off between semilocal and hybrid DFT, and our general approach provides a feasible path toward a universal exchange-correlation functional with post-hybrid DFT accuracy and semilocal DFT cost.

Graph Atomic Cluster Expansion for semilocal interactions beyond equivariant message passing

Graph Atomic Cluster Expansion for Semilocal Interactions beyond Equivariant Message Passing

Cartesian atomic cluster expansion for machine learning interatomic potentials

Charge-constrained Atomic Cluster Expansion

Atomic cluster expansion for quantum-accurate large-scale simulations of carbon

Atomic Cluster Expansion without Self-Interaction

Permutation-adapted complete and independent basis for atomic cluster expansion descriptors

Performant implementation of the atomic cluster expansion (PACE) and application to copper and silicon

Single Scale Cluster Expansions with Applications to Many Boson and Unbounded Spin Systems

Cluster expansion constructed over Jacobi-Legendre polynomials for accurate force fields

Atomic cluster expansion of scalar, vectorial and tensorial properties and including magnetism and charge transfer

A Multilevel Method for Many-Electron Schrödinger Equations Based on the Atomic Cluster Expansion

Cluster expansion methods from physical concepts

Learning local equivariant representations for large-scale atomistic dynamics

Learning Equivariant Non-Local Electron Density Functionals

Unified theory of atom-centered representations and message-passing machine-learning schemes

On the Atomic Cluster Expansion: interatomic potentials and beyond

Nonlocal Machine-Learned Exchange Functional for Molecules and Solids

Gaussian expansion of Yukawa non-local kinetic energy functionals: application to metal clusters