Engineering topics

Reuter, Karsten

Publications and source records attributed to Reuter, Karsten.

6 records

DOE OSTI2025

A foundation model for atomistic materials chemistry

Atomistic simulations of matter, especially those that leverage first-principles (ab initio) electronic structure theory, provide a microscopic view of the world, underpinning much of our understanding of chemistry and materials science. Over the last decade or so, machine-learned force fields have transformed atomistic modeling by enabling simulations of ab initio quality over unprecedented time and length scales. However, early machine-learning (ML) force fields have largely been limited by (i) the substantial computational and human effort required to develop and validate potentials for each particular system of interest and (ii) a general lack of transferability from one chemical system to the next. Here, we show that it is possible to create a general-purpose atomistic ML model, trained on a public dataset of moderate size, that is capable of running stable molecular dynamics for a wide range of molecules and materials. We demonstrate the power of the MACE-MP-0 model-and its qualitative and at times quantitative accuracy-on a diverse set of problems in the physical sciences, including properties of solids, liquids, gases, chemical reactions, interfaces, and even the dynamics of a small protein. The model can be applied out of the box as a starting or "foundation" model for any atomistic system of interest and, when desired, can be fine-tuned on just a handful of application-specific data points to reach ab initio accuracy. Establishing that a stable force-field model can cover almost all materials changes atomistic modeling in a fundamental way: experienced users obtain reliable results much faster, and beginners face a lower barrier to entry. Foundation models thus represent a step toward democratizing the revolution in atomic-scale modeling that has been brought about by ML force fields.

Batatia, Ilyes↗

DOE OSTI2025

Spin-informed universal graph neural networks for simulating magnetic ordering

The screening and discovery of magnetic materials are hindered by the computational cost of first-principles density-functional theory (DFT) calculations required to find the ground state magnetic ordering. Although universal machine-learning interatomic potentials (uMLIPs), also known as atomistic foundation models, offer high-fidelity models of many atomistic systems with significant speedup, they currently lack the inputs required for predicting magnetic ordering. In this work, we present a data-efficient, spin-informed graph neural network framework that incorporates spin degrees of freedom as inputs and preserves physical symmetries, extending the functionality of uMLIPs to simulate magnetic orderings. This framework speeds up DFT calculations through better initial guesses for magnetic moments, determines the ground-state ordering of bulk materials and even generalizes to magnetic ordering in surfaces. Furthermore, we implement a closed-loop anomaly detection approach that effectively addresses the classic "chicken-and-egg" problem of creating a high-quality dataset while developing a uMLIP, unearthing anomalies in large benchmark datasets and boosting model accuracy.

Xu, Wenbin↗

DOE OSTI2024

Factors Affecting the Electron Conductivity in Single Crystal Li 7 La 3 Zr 2 O 12 and Li 7 P 3 S 11

One of the serious challenges in all solid-state Li ion batteries is neutral Li intrusion into the solid-state electrolyte that can ultimately cause catastrophic failure. One possibility for this is due to n-type electron conductivity that induces the reaction Li + + e – → Li 0 at sites where the potential is less than the Li + /Li potential. This paper reports hybrid density functional theory calculations of the electronic conductivity in two prototype single crystalline solid-state electrolytes, cubic Li 7 La 3 Zr 2 O 12 (c-LLZO) and Li 7 P 3 S 11 (LPS). The formation energies of important point defects that can affect electron conductivity are determined, and we find that the mechanism of n-type electron conductivity for both solid-state electrolytes is via “small” electron polaron hopping, where the quotes signify that substantial Li ion rearrangement is associated with the polaron formation and its migration. In both electrolytes, the formation energies for the small polarons at the Fermi energy are too high to generate measurable electron conductivity at room temperature. For c-LLZO, the concentration of electron polarons necessary to ensure charge neutrality from positively charged oxygen vacancies formed in synthesis can be significantly higher. Hence, the electron conductivity could be significant when measured with ion-blocking metal electrodes, and we discuss how the synthesis conditions could affect this magnitude. However, in the solid-state battery, these polarons are replaced by negatively charged Li vacancies so that the electron conductivity should remain minimal. For LPS single crystals, the inherent minimal electron conductivity is independent of synthesis conditions. Here we also show that the cost of forming Li 0 in bulk c-LLZO is enormous due to strain effects so that it could only potentially form at voids, grain boundaries, or around vacancy defects which relax the lattice strain.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

DOE OSTI2024

Discovering High Entropy Alloy Electrocatalysts in Vast Composition Spaces with Multiobjective Optimization

High entropy alloys (HEAs) are a highly promising class of materials for electrocatalysis as their unique active site distributions break the scaling relations that limit the activity of conventional transition metal catalysts. Existing Bayesian optimization (BO)-based virtual screening approaches focus on catalytic activity as the sole objective and correspondingly tend to identify promising materials that are unlikely to be entropically stabilized. Here, we overcome this limitation with a multiobjective BO framework for HEAs that simultaneously targets activity, cost-effectiveness, and entropic stabilization. With diversity-guided batch selection further boosting its data efficiency, the framework readily identifies numerous promising candidates for the oxygen reduction reaction that strike the balance between all three objectives in hitherto unchartered HEA design spaces comprising up to 10 elements.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

DOE OSTI2022

Di- and Tetrameric Molybdenum Sulfide Clusters Activate and Stabilize Dihydrogen as Hydrides

NaY zeolite-encapsulated dimeric (Mo 2 S 4 ) and tetrameric (Mo 4 S 4 ) molybdenum sulfide clusters stabilize hydrogen as hydride binding to Mo atoms. Density functional theory (DFT) calculations and adsorption measurements suggest that stabilization of hydrogen as sulfhydryl (SH) groups, as typical for layered MoS 2 , is thermodynamically disfavored. Competitive adsorption of H 2 and ethene on Mo was probed by quantifying adsorbed CO on partly hydrogen and/or ethene covered samples with IR spectroscopy. During hydrogenation, experiment and theory suggest that Mo is covered predominately with ethene and sparsely with hydride. DFT calculations further predict that under reaction conditions, each Mo x S y cluster can activate only one H 2 , suggesting that the entire cluster (irrespective of its nuclearity) acts as one active site for hydrogenation. The nearly identical turnover frequencies (24.7 ± 3.3 mol ethane ·h -1 ·mol cluster -1 ), apparent activation energies (31-32 kJ·mol -1 ), and reaction orders (~0.5 in ethene and ~1.0 in H 2 ) show that the active sites in both clusters are catalytically indistinguishable.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

DOE OSTI2021

Zeolite-Stabilized Di- and Tetranuclear Molybdenum Sulfide Clusters Form Stable Catalytic Hydrogenation Sites

Supercages of faujasite (FAU)-type zeolites serve as a robust scaffold for stabilizing dinuclear (Mo 2 S 4 ) and tetranuclear (Mo 4 S 4 ) molybdenum sulfide clusters. The FAU-encaged Mo 4 S 4 clusters have a distorted cubane structure similar to the FeMo-cofactor in nitrogenase. Both clusters possess one unpaired electron per Mo atom. Additionally, they show identical catalytic activity per sulfide cluster. Their catalytic activity is stable (> 150 h) for ethene hydrogenation, while layered MoS 2 structures deactivate significantly under the same reaction conditions.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗