Engineering topics

Jeong, Wonseok

Publications and source records attributed to Jeong, Wonseok.

7 records

DOE OSTI2024

Spectroscopy-guided discovery of three-dimensional structures of disordered materials with diffusion models

Spectroscopy techniques such as x-ray absorption near edge structure (XANES) provide valuable insights into the atomic structures of materials, yet the inverse prediction of precise structures from spectroscopic data remains a formidable challenge. In this study, we introduce a framework that combines generative artificial intelligence models with XANES spectroscopy to predict three-dimensional atomic structures of disordered systems, using amorphous carbon (a-C) as a model system. In this work, we introduce a new framework based on the diffusion model, a recent generative machine learning method, to predict 3D structures of disordered materials from a target property. For demonstration, we apply the model to identify the atomic structures of a-C as a representative material system from the target XANES spectra. We show that conditional generation guided by XANES spectra reproduces key features of the target structures. Furthermore, we show that our model can steer the generative process to tailor atomic arrangements for a specific XANES spectrum. Finally, our generative model exhibits a remarkable scale-agnostic property, thereby enabling generation of realistic, large-scale structures through learning from a small-scale dataset (i.e. with small unit cells). Our work represents a significant stride in bridging the gap between materials characterization and atomic structure determination; in addition, it can be leveraged for materials discovery in exploring various material properties as targeted.

36 MATERIALS SCIENCE↗

DOE OSTI2024

Integrating Machine Learning Potential and X-ray Absorption Spectroscopy for Predicting the Chemical Speciation of Disordered Carbon Nitrides

Precise determination of atomic structural information in functional materials holds transformative potential and broad implications for emerging technologies. Spectroscopic techniques, such as X-ray absorption near-edge structure (XANES), have been widely used for material characterization; however, extracting chemical information from experimental probes remains a significant challenge, particularly for disordered materials. We present an integrated approach that combines atomic simulations, data-driven techniques, and experimental measurements to investigate chemical speciation of amorphous carbon nitride systems as a case study. Here, we discuss the development of machine learning potentials that can efficiently explore the vast configuration space of amorphous carbon nitrides. By employing statistical methods, this structural database enables the elucidation of the most representative local structures and how they evolve with chemical compositions and density. Density functional theory simulations are used to establish a correlation between the local structure and spectroscopic signatures, which then serve as the basis for interpreting and extracting chemical content from experimental data. Although our framework is specifically demonstrated for XANES and carbon nitrides, the approach described herein is readily adaptable as applied to other experimental characterization probes and materials classes.

36 MATERIALS SCIENCE↗

DOE OSTI2023

Electrochemical Degradation of Pt 3 Co Nanoparticles Investigated by Off-Lattice Kinetic Monte Carlo Simulations with Machine-Learned Potentials

In fuel cell applications, the durability of catalysts is critical for large-scale industrial implementation. However, limited synthesis controllability and spectroscopic resolution impede a comprehensive understanding of degradation mechanisms at the atomic level. In this study, we develop a machine-learned potential (MLP) to simulate the degradation processes for Pt3Co nanoparticles. The precision of MLP is determined to be comparable to that of density functional theory calculations. Using off-lattice kinetic Monte Carlo simulations with MLP, we successfully replicate established experimental trends and offer a logical resolution to ongoing debates regarding atomic orderings. Based on the simulation results, we suggest design principles for Pt3Co nanoparticles that combine high activity and durability. Finally, we validate the wide applicability of our method by successfully applying it to Pt3Ni and Pt3Co0.5Ni0.5 nanoparticles. The research serves as a guideline for developing MLPs for alloy electrochemical catalysts and lays the foundation for designing more durable and active fuel-cell catalysts.

36 MATERIALS SCIENCE↗

DOE OSTI2023

Harnessing Neural Networks for Elucidating X-ray Absorption Structure–Spectrum Relationships in Amorphous Carbon

Improved understanding of structural and chemical properties through local experimental probes, such as X-ray absorption near-edge structure (XANES) spectroscopy, is crucial for the understanding and design of functional materials. In recent years, significant advancements have been made in the development of data science approaches for the automated interpretation of XANES structure–spectrum relationships. However, existing studies have primarily focused on crystalline solids and small molecules, while fewer efforts have been devoted to disordered systems. Thus, in this work, we demonstrate the development of neural network models for predicting and interpreting XANES spectra of amorphous carbon (a-C) from local structural descriptors. Comparison between different structural descriptors expectedly shows that the inclusion of both bond length and bond angle information is necessary for an accurate prediction of the spectra. Among the descriptors considered in this work, we find that the local many-body tensor representation yields the highest accuracy and greatest interpretability so that it can be leveraged to understand the importance of structural motifs in determining XANES spectra. Furthermore, we also discuss performance of neural network models for predicting both local structure features, such as bond lengths and bond angles, and global chemical composition, such as the sp:sp 2 :sp 3 ratio.

36 MATERIALS SCIENCE↗

DOE OSTI2023

E-beam-enhanced solid-state mechanical amorphization of α-quartz: Reduced deformation barrier via localized excess electrons as network modifiers

Under hydrostatic pressure, α-quartz (α-SiO 2 ) undergoes solid-state mechanical amorphization wherein the interpenetration of [SiO 4 ] 2- tetrahedra occurs and the material loses crystallinity. This phase transformation requires a high hydrostatic pressure of 14GPa because the repulsive forces resulting from the ionic nature of the Si–O bonds prevent the severe distortion of the atomic configuration. Herein, we experimentally and computationally demonstrate that e-beam irradiation changes the nature of the interatomic bonds in α-quartz and enhances the solid-state mechanical amorphization at nanoscale. Specifically, during in situ uniaxial compression, a larger permanent deformation occurs in α-quartz submicron pillars compressed during e-beam irradiation than in those without e-beam irradiation. Microstructural analysis reveals that the large permanent deformation under e-beam irradiation originates from the enhanced mechanical amorphization of α-quartz and the subsequent viscoplastic deformation of the amorphized region. Further, atomic-scale simulations suggest that the delocalized excess electrons introduced by e-beam irradiation move to highly distorted atomic configurations and alleviate the repulsive force, thus reducing the barrier to the solid-state mechanical amorphization. In conclusion, these findings deepen our understanding of electron–matter interactions and can be extended to new glass forming and processing technologies at nano- and microscale.

36 MATERIALS SCIENCE↗

DOE OSTI2022

Athermal glass work at the nanoscale: Engineered electron-beam-induced viscoplasticity for mechanical shaping of brittle amorphous silica

Amorphous silica deforms viscoplastically at elevated temperatures, which is common for brittle glasses. The key mechanism of viscoplastic deformation involves interatomic bond switching, which is thermally activated. Here, for this study, we precisely control the mechanical shaping of brittle amorphous silica at the nanoscale via engineered electron–matter interactions without heating. We observe a ductile plastic deformation of amorphous silica under a focused scanning electron beam with low acceleration voltages (few to tens of kilovolts) during in-situ compression studies, with unique dependence on the acceleration voltage and beam current. By simulating the electron–matter interaction, we show that the deformation of amorphous silica depends strongly on the volume where inelastic scattering occurs. The electron–matter interaction via e-beam irradiation alters the Si–O interatomic bonds, enabling the high-temperature deformation behavior of amorphous silica to occur athermally. Finally, by systematically controlling the electron–matter interaction volume, it is possible to mechanically shape the brittle amorphous silica on a small scale at room temperature to a level comparable to glass shaping at high temperatures. The findings can be extended to develop new fabrication processes for nano- and microscale brittle glasses.

36 MATERIALS SCIENCE↗

DOE OSTI2022

Accelerated computation of lattice thermal conductivity using neural network interatomic potentials

We report with the development of the density functional theory (DFT) and ever-increasing computational capacity, an accurate prediction of lattice thermal conductivity based on the Boltzmann transport theory becomes computationally feasible, contributing to a fundamental understanding of thermal conductivity as well as a choice of the optimal materials for specific applications. However, steep costs in evaluating third-order force constants limit the theoretical investigation to crystals with high symmetry and few atoms in the unit cell. Currently, machine learning potentials are garnering attention as a computationally efficient high-fidelity model of DFT, and several studies have demonstrated that the lattice thermal conductivity could be computed accurately via the machine learning potentials. However, test materials were mostly crystals with high symmetries, and the applicability of machine learning potentials to a wide range of materials has yet to be demonstrated. Furthermore, establishing a standard training set that provides consistent accuracy and computational efficiencies across a wide range of materials would be useful. To address these issues, herein we compute lattice thermal conductivities at 300 K using neural network interatomic potentials. As test materials, we select 25 materials with diverse symmetries and a wide range of lattice thermal conductivities between 10 -1 and 10 3 Wm -1 K -1 . Among various choices of training sets, we find that molecular dynamics trajectories at 50–700 K consistently provide results at par with DFT for the test materials. In contrast to pure DFT approaches, the computational cost in the present approach is uniform over the test materials, yielding a speed gain of 2–10 folds. When a smaller reduced training set is used, the relative efficiency increases by up to ~50 folds without sacrificing accuracy significantly. The current work will broaden the application scope of machine learning potentials by establishing a robust framework for accurately computing lattice thermal conductivity with machine learning potentials.

36 MATERIALS SCIENCE↗