Search NASA⌕ Search
Home/Archive/DOE OSTI

DOE OSTI · 3367750

Machine Learning an Ab-Initio Based Bond-Order Potential for Bismuthene

Dutta, Partha Sarathi [Univ. of Illinois, Chicago, IL (United States); Argonne National Laboratory (ANL), Argonne, IL (United States). Center for Nanoscale Materials (CNM)] (ORCID:0000000165337454)·Koneru, Aditya [Univ. of Illinois, Chicago, IL (United States); Argonne National Laboratory (ANL), Argonne, IL (United States). Center for Nanoscale Materials (CNM)] (ORCID:0000000263211900)·Muhammed, Adil [Univ. of Illinois, Chicago, IL (United States); Argonne National Laboratory (ANL), Argonne, IL (United States). Center for Nanoscale Materials (CNM)]·Chan, Henry [Argonne National Laboratory (ANL), Argonne, IL (United States). Center for Nanoscale Materials (CNM)] (ORCID:0000000281987737)·Balasubramanian, Karthik [Univ. of Illinois, Chicago, IL (United States); Argonne National Laboratory (ANL), Argonne, IL (United States). Center for Nanoscale Materials (CNM)]·Manna, Sukriti [Univ. of Illinois, Chicago, IL (United States); Argonne National Laboratory (ANL), Argonne, IL (United States). Center for Nanoscale Materials (CNM)] (ORCID:0000000231937803)·Loeffler, Troy [Univ. of Illinois, Chicago, IL (United States); Argonne National Laboratory (ANL), Argonne, IL (United States). Center for Nanoscale Materials (CNM)] (ORCID:0000000302441793)·Sasikumar, Kiran [Avant-Garde Materials Simulation Deutschland GmbH (Germany)]·Darancet, Pierre [Argonne National Laboratory (ANL), Argonne, IL (United States). Center for Nanoscale Materials (CNM)] (ORCID:0000000258461673)·Sankaranarayanan, Subramanian K. R. S. [Univ. of Illinois, Chicago, IL (United States); Argonne National Laboratory (ANL), Argonne, IL (United States). Center for Nanoscale Materials (CNM)] (ORCID:000000029708396X)
View original source ↗Read source document ↗Export metadata ↓

Abstract

Bismuthene is a heavy 2D material whose strong spin–orbit coupling and recently observed single-element ferroelectricity have intensified interest in its structural, vibrational, and transport properties. Accurate modeling of these behaviors requires a short-range interatomic potential that can reproduce the underlying bonding physics at a fraction of the computational cost of first-principles methods. However, such a potential is currently unavailable. Here, in this work, we construct a Tersoff bond-order potential for β-bismuthene using a reinforcement-learning framework that integrates a continuous Monte Carlo Tree Search with a simplex-based local optimizer. The optimized parameter sets reproduce first-principles lattice constants, cohesive energy, the equation of state, elastic constants, and phonon dispersion. We validate the models by performing thermal-conductivity calculations and uniaxial fracture simulations our findings confirm the reliability of the resulting models across multiple thermomechanical regimes. Comparison of the three best solutions reveals how differences in pairwise interactions, angular terms, and bond-order behavior govern phonon features and mechanical responses. We demonstrate an interpretable and computationally efficient potential for bismuthene and demonstrate a general reinforcement-learning strategy for developing bond-order models in emerging 2D materials.

Explore related subjects

deformationphononsstiffnessthermal conductivitytwo-dimensional materials

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Dutta, Partha Sarathi [Univ. of Illinois, Chicago, IL (United States); Argonne National Laboratory (ANL), Argonne, IL (United States). Center for Nanoscale Materials (CNM)] (ORCID:0000000165337454), Koneru, Aditya [Univ. of Illinois, Chicago, IL (United States); Argonne National Laboratory (ANL), Argonne, IL (United States). Center for Nanoscale Materials (CNM)] (ORCID:0000000263211900), Muhammed, Adil [Univ. of Illinois, Chicago, IL (United States); Argonne National Laboratory (ANL), Argonne, IL (United States). Center for Nanoscale Materials (CNM)], Chan, Henry [Argonne National Laboratory (ANL), Argonne, IL (United States). Center for Nanoscale Materials (CNM)] (ORCID:0000000281987737), Balasubramanian, Karthik [Univ. of Illinois, Chicago, IL (United States); Argonne National Laboratory (ANL), Argonne, IL (United States). Center for Nanoscale Materials (CNM)], Manna, Sukriti [Univ. of Illinois, Chicago, IL (United States); Argonne National Laboratory (ANL), Argonne, IL (United States). Center for Nanoscale Materials (CNM)] (ORCID:0000000231937803), Loeffler, Troy [Univ. of Illinois, Chicago, IL (United States); Argonne National Laboratory (ANL), Argonne, IL (United States). Center for Nanoscale Materials (CNM)] (ORCID:0000000302441793), Sasikumar, Kiran [Avant-Garde Materials Simulation Deutschland GmbH (Germany)], Darancet, Pierre [Argonne National Laboratory (ANL), Argonne, IL (United States). Center for Nanoscale Materials (CNM)] (ORCID:0000000258461673), Sankaranarayanan, Subramanian K. R. S. [Univ. of Illinois, Chicago, IL (United States); Argonne National Laboratory (ANL), Argonne, IL (United States). Center for Nanoscale Materials (CNM)] (ORCID:000000029708396X). 2026-03-16. Machine Learning an Ab-Initio Based Bond-Order Potential for Bismuthene. https://doi.org/10.1021/acs.jpcc.5c08318

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

KEEP EXPLORING

Related reports

DOE OSTI2025

SAN-Based Block Polymers as a Platform for Manufacturing Strong Isoporous Membranes

Ultrafiltration (UF) membranes are ubiquitous in water purification and bioprocessing. However, co-designing their mechanical and transport properties remains challenging because of the broad pore size distributions at the surface and within the bulk that result from nonsolvent-induced phase separation (NIPS) – their typical manufacturing process. These distributions influence the hydrodynamic resistance to water flow and the stress concentrations around the pores. Developing advanced UF membranes requires innovative molecular designs that offer control over the surface and bulk pores, as well as the mechanical properties of the load-bearing, polymer. Here, we introduce a platform for designing UF membranes by leveraging solution self-assembly of block polymers and chain architectures with pendant polar groups. The block polymers consist of a poly(styrene-co-acrylonitrile) hydrophobic block, which is known for its strength, and a poly(4-vinyl pyridine) hydrophilic block, which drives solution self-assembly. We focus on a series of block polymers with constant molecular weight, M n ≈ 115 kDa, SAN fraction, 75 wt.%, and varying acrylonitrile content, 0 to 40 mol%, to demonstrate that: (i) RAFT dispersion copolymerization of acrylonitrile and styrene provides a facile route to synthesize strong block polymers, (ii) incorporation of acrylonitrile into the hydrophobic block enhances membrane strength by facilitating chain entanglements and dipole-dipole interactions, and (iii) acrylonitrile alters the balance between membrane permeance and rejection, even when the membranes feature similar surface and bulk pores. Overall, our results provide insights into the molecular design of UF membranes with enhanced mechanical and separation properties, contributing to the development of materials for water and energy technologies.

deformation↗
DOE OSTI2023

Learning Canonical Embeddings for Unsupervised Shape Correspondence With Locally Linear Transformations

We present a new approach to unsupervised shape correspondence learning between pairs of point clouds. We make the first attempt to adapt the classical locally linear embedding algorithm (LLE)-originally designed for nonlinear dimensionality reduction-for shape correspondence. The key idea is to find dense correspondences between shapes by first obtaining high-dimensional neighborhood-preserving embeddings of low-dimensional point clouds and subsequently aligning the source and target embeddings using locally linear transformations. We demonstrate that learning the embedding using a new LLE-inspired point cloud reconstruction objective results in accurate shape correspondences. More specifically, the approach comprises an end-to-end learnable framework of extracting high-dimensional neighborhood-preserving embeddings, estimating locally linear transformations in the embedding space, and reconstructing shapes via divergence measure-based alignment of probability density functions built over reconstructed and target shapes. Our approach enforces embeddings of shapes in correspondence to lie in the same universal/canonical embedding space, which eventually helps regularize the learning process and leads to a simple nearest neighbors approach between shape embeddings for finding reliable correspondences. Comprehensive experiments show that the new method makes noticeable improvements over state-of-the-art approaches on standard shape correspondence benchmark datasets covering both human and nonhuman shapes.

deformation↗