Predicting Atomistic Transitions with Transformers

Tischler, Henry [Los Alamos National Laboratory (LANL), Los Alamos, NM (United States); Univ. of Denver, CO (United States)] (ORCID:0009000480915311); Li, Wenting [Univ. of Texas, Austin, TX (United States)] (ORCID:0000000313822030); Tang, Qi [Georgia Institute of Technology, Atlanta, GA (United States); Los Alamos National Laboratory (LANL), Los Alamos, NM (United States)] (ORCID:0000000196141075); Perez, Danny [Los Alamos National Laboratory (LANL), Los Alamos, NM (United States)] (ORCID:0000000330285249); Vogel, Thomas [Los Alamos National Laboratory (LANL), Los Alamos, NM (United States)] (ORCID:0000000302053205)

doi:10.1080/26941899.2026.2685344

DOE OSTI · 3382482

Predicting Atomistic Transitions with Transformers

Tischler, Henry [Los Alamos National Laboratory (LANL), Los Alamos, NM (United States); Univ. of Denver, CO (United States)] (ORCID:0009000480915311)·Li, Wenting [Univ. of Texas, Austin, TX (United States)] (ORCID:0000000313822030)·Tang, Qi [Georgia Institute of Technology, Atlanta, GA (United States); Los Alamos National Laboratory (LANL), Los Alamos, NM (United States)] (ORCID:0000000196141075)·Perez, Danny [Los Alamos National Laboratory (LANL), Los Alamos, NM (United States)] (ORCID:0000000330285249)·Vogel, Thomas [Los Alamos National Laboratory (LANL), Los Alamos, NM (United States)] (ORCID:0000000302053205)

View original source ↗Read source document ↗Export metadata ↓

Abstract

Accurate knowledge of the atomistic transition pathways in materials and material surfaces is crucial for many material science problems. However, conventional simulation techniques used to find these transitions are extremely computationally intensive. Even with large-scale, accelerated material simulations, the computational cost constrains the applicable domain in practice. Machine learning models, with the potential to learn the complex emergent behaviors governing atomistic transitions as a fast surrogate model, have great promise to predict transitions with a vastly reduced computational cost. Here, we demonstrate how transformers can be trained to predict atomistic transitions in nano-clusters. We show how we evaluate physical validity of the predictions and how a multitude of additional, different microstates can be generated by slightly varying the data provided to the model.

Explore related subjects

36 MATERIALS SCIENCE 74 ATOMIC AND MOLECULAR PHYSICS 97 MATHEMATICS AND COMPUTING 99 GENERAL AND MISCELLANEOUS atomistic transitions material defects surrogate models transformers

Keep this discovery

BibTeX RIS

Tischler, Henry [Los Alamos National Laboratory (LANL), Los Alamos, NM (United States); Univ. of Denver, CO (United States)] (ORCID:0009000480915311), Li, Wenting [Univ. of Texas, Austin, TX (United States)] (ORCID:0000000313822030), Tang, Qi [Georgia Institute of Technology, Atlanta, GA (United States); Los Alamos National Laboratory (LANL), Los Alamos, NM (United States)] (ORCID:0000000196141075), Perez, Danny [Los Alamos National Laboratory (LANL), Los Alamos, NM (United States)] (ORCID:0000000330285249), Vogel, Thomas [Los Alamos National Laboratory (LANL), Los Alamos, NM (United States)] (ORCID:0000000302053205). 2026-06-26. Predicting Atomistic Transitions with Transformers. https://doi.org/10.1080/26941899.2026.2685344

Cite the original work for its findings. Save a collection to share your selection of sources.

Discover connections

36 MATERIALS SCIENCE subject 74 ATOMIC AND MOLECULAR PHYSICS subject 97 MATHEMATICS AND COMPUTING subject 99 GENERAL AND MISCELLANEOUS subject atomistic transitions subject material defects subject surrogate models subject transformers subject

Connections use source metadata and explicit phrase matches, not verified experimental comparisons.

KEEP EXPLORING

Related reports

DOE OSTI2026

Revealing the Hidden Third Dimension of Point Defects in Two-Dimensional MXenes

Point defects govern many important functional properties of two-dimensional (2D) materials. However, resolving the three-dimensional (3D) arrangement of these defects in multi-layer 2D materials remains a fundamental challenge, hindering rational defect engineering. Here, we overcome this limitation using an artificial intelligence-guided electron microscopy workflow to map the 3D topology and clustering of atomic vacancies in Ti3C2TX MXene. Our approach reconstructs the 3D coordinates of vacancies across hundreds of thousands of lattice sites, generating robust statistical insight into their distribution that can be correlated with specific synthesis pathways. This large-scale data enables us to classify a hierarchy of defect structures-from isolated vacancies to nanopores-revealing their preferred formation and interaction mechanisms, as corroborated by molecular dynamics simulations. This work provides a generalizable framework for understanding and ultimately controlling point defects across large volumes, paving the way for the rational design of defect-engineered functional 2D materials.

2D materials↗

DOE OSTI2026

Scattering theory of frequency-entangled biphoton states facilitated by cavity polaritons

The use of quantum light to probe exciton properties in semiconductor and molecular nanostructures typically occurs in the low-intensity regime. A substantial enhancement of exciton-photon coupling can be achieved with photonic cavities, where excitons hybridize with cavity modes to form polariton states. Here, to provide a theoretical framework for interpreting emerging experimental efforts in this direction, we develop a scattering theory describing the interaction of frequency-entangled photon pairs with cavity polariton and bipolariton states under various coupling regimes. Employing the Tavis-Cummings model in combination with our scattering approach, we present a quantitative analysis of how the interaction of the entangled photon pair with the polariton or bipolariton modifies its joint spectral amplitude (JSA). Specifically, we examine the effects of the cavity-mode steady-state population, exciton-cavity coupling strength, and different forms of the input photon JSA. Our results show that the entanglement entropy of the scattered photons is highly sensitive to the interplay between the input JSA and the spectral line shapes of the polariton resonances, emphasizing the cavity filtering effects. We suggest that biphoton-scattering quantum light spectroscopy best serves as a sensitive probe of polariton and bipolariton states in the photon-vacuum cavity state. Our approach is not only robust to various regimes of cavity-exciton coupling, but also amenable to extensions beyond the Tavis-Cummings model, enabling the representation of a broad class of molecular systems and solid state quantum materials.

36 MATERIALS SCIENCE↗

DOE OSTI2026

Extreme confinement of hydrogen gas within fullerenelike nanoporous carbon

Nanoporous carbons and carbon nanostructures can store hydrogen at cryogenic temperatures but lack the volumetric and gravimetric capacity to be industrially significant. Recent inelastic neutron scattering experiments suggest a highly dense phase of hydrogen at temperatures well above the melting point of solid hydrogen. However, it remains unclear how pore geometry and intermolecular interactions enable these dense phases to exist, with dispersion (van der Waals) or electrostatic/induction suggested to be the key effects in slit and curved pores but their relative contributions have yet to be quantified. In this paper, we perform benchmark electronic structure calculations allowing the interactions between planar and curved aromatic molecules with hydrogen to be accurately determined. Dispersion was found to dominate over electrostatic and inductive effects with some many-body charge transfer (Dobson type-A) effects needed to capture the most highly curved structures. Density functional methods that include type-A many-body effects were found to accurately describe the intermolecular interactions at a fraction of the cost of coupled-cluster simulations and these approaches were used to calculate the energies inside large carbon bowl and slit pores. The interaction energies inside the bowl pores were found to depend on the orientation of the hydrogen molecule. This rotational barrier, modeled as a quantum hindered rotor, could reproduce the peak splitting observed in inelastic neutron scattering experiments, with weak splitting arising from bowl-like fullerene pores and strong splitting from highly confining nanotubelike pores. Increasing the fraction of such curved pores in nanoporous carbons may therefore offer a pathway to enhance their hydrogen-storage capacity. Moreover, the preferential adsorption of ortho hydrogen on nanotubelike pores could enable the storage of high-density hydrogen without the need to remove heat produced during the ortho-para hydrogen conversion.

36 MATERIALS SCIENCE↗