Machine learning the electronic structure of matter across temperatures

Aug 4, 2026

Upload by Himbo

Machine learning models are presented to predict the electronic structure of materials over a wide temperature range, using neural networks trained on density functional theory (DFT) data. Instead of relying on DFT-based representations used by other ML approaches, the models directly predict the local density of states (LDOS). This enables simultaneous access to observables such as electronic density and electronic total free energy, while treating electronic and ionic temperatures independently. Performance is validated on a metallic test system, capturing energetic effects across broad ionic and electronic variations even with partial temperature training.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages18

Reading Time45 min

AccessFree to download

Introduction

Computational challenges of temperature-dependent electronic-structure calculations

Prior ML approaches and their limitations

LDOS-based neural-network models and electronic-temperature treatment

arXiv :2306 .06032v1 [ cond-mat .mtrl-sci ] 9 Jun 2023 Machine learning the electronic structure of matter across temperatures Lenz Fiedler, 1, 2, ∗ Normand A. Modine,3 Kyle D. Miller,3 and Attila Cangi 1, 2,† 1 Center for Advanced Systems Understanding (CASUS), D-02826 Görlitz, Germany 2 Helmholtz-Zentrum Dresden-Rossendorf, D-01328 Dresden, Germany 3 Sandia National Laboratories, Albuquerque, NM 87185, USA (Dated: June 12, 2023) We introduce machine learning (ML) models that predict the electronic structure of materials across a wide temperature range. Our models employ neural networks and are trained on density functional theory (DFT) data. Unlike other ML models that use DFT data, our models directly predict the local density of states (LDOS) of the electronic structure. This provides several advantages, including access to multiple observables such as the electronic density and electronic total free energy. Moreover, our models account for both the electronic and ionic temperatures independently, making them ideal for applications like laser-heating of matter. We validate the efficacy of our LDOS-based models on a metallic test system. They accurately capture energetic effects induced by variations in ionic and electronic temperatures over a broad temperature range, even when trained on a subset of these temperatures. These findings open up exciting opportunities for investigating the electronic structure of materials under both ambient and extreme conditions. I. INTRODUCTION Predicting the electronic structure of matter is essential for advancing scientific progress across various applications. Electronic structure calculations, which typically employ density functional theory (DFT) [1, 2], have become a routine tool in materials science and chemistry due to their accuracy and computational efficiency [3, 4] . However, as the demand for high-fidelity simulation data in emerging research areas increases, conventional DFT simulations face significant limitations. DFT calculations exhibit unfavorable scaling with both system size and temperature [5], limiting their applicability for current scientific challenges, particularly in studying materials under extreme conditions and within the warm dense matter regime [6–8] . Progress in this area not only contributes to the fundamental sciences by advancing our understanding of astrophysical objects [9–12], but also propels technological developments by enabling the modeling of inertial confinement fusion capsule heating processes [13], radiation damage processes in reactor walls [14–17], and advanced manufacturing [18, 19] . Additionally, it supports diagnostics of scattering experiments conducted at free-electron laser facilities [20, 21] and promotes the emerging field of hot-electron chemistry for accelerating chemical reactions [22, 23] . A particularly relevant phenomenon in these applications involves rapidly driven electrons leading to transient non-equilibrium conditions resulting in hot electrons and cool nuclei which have also been observed in semiconducting and dielectric materials [24, 25] . To address these computational limitations, the electronic structure community has increasingly turned to machine learning (ML) techniques [26] . ML algorithms can accurately predict complicated relationships using tractable data samples. The application of ML to DFT has led to numerous approaches, with most focusing on predicting specific observables of interest [27] or replacing DFT entirely with ML interatomic potentials (ML- IAPs), which capture the electronic total energy or total free energy landscape of a system and enable extended simulations of ionic dynamics [28, 29] . While existing ML-based approaches show promise in accurately predicting observables or capturing the energy landscape of quantum systems, most of them do not provide direct access to the electronic structure of matter. Knowledge of the electronic structure, however, offers several

FAQs

What do the proposed machine learning models predict?

They predict the electronic structure of materials across a wide temperature range by directly outputting the local density of states (LDOS).

How are electronic and ionic temperatures handled in the models?

The models account for electronic and ionic temperatures independently, allowing predictions that reflect both temperature contributions.

What evidence supports the effectiveness of the LDOS-based models?

Validation on a metallic test system shows the models accurately capture energetic effects induced by variations in ionic and electronic temperatures over a broad range, even when trained on only a subset of those temperatures.

Aug 4, 2026

Upload by Himbo