Engineering topics

Jung, Gang Seob

Publications and source records attributed to Jung, Gang Seob.

11 records

DOE OSTI2024

Atomic-Scale Dynamic Mechanisms of Embedded MoS 2 Wires

Nanowires composed of a 1:1 stoichiometry of transition metals and chalcogen ions can be fabricated from two-dimensional transition metal dichalcogenides (TMDs) by using electron beam irradiation. Wires fabricated through in situ experiments can be geometrically connected to TMD sheets in various ways, and their physical properties can vary accordingly. Understanding the structural transformation caused by electron beams is critical for designing wire-sheet structures for nanoelectronics. In this study, we report the behavior of nanowires formed inside a monolayer MoS 2 sheet by combining phase-contrast images and large-scale atomistic modeling. Here we investigate the effect of vacancies on the dynamic evolution of wires, such as rotations with different edge structures and breaking, by considering the interactions between MoS wires and MoS 2 nanosheets. The obtained insights can be applied to other monolayer TMDs to guide the behavior of TMD wires and fabricate favorable geometries for various applications.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

DOE OSTI2024

Enhancing high-fidelity neural network potentials through low-fidelity sampling

The efficacy of neural network potentials (NNPs) critically depends on the quality of the configurational datasets used for training. Prior research using empirical potentials has shown that well-selected liquid–solid transitional configurations of a metallic system can be translated to other metallic systems. This study demonstrates that such validated configurations can be relabeled using density functional theory (DFT) calculations, thereby enhancing the development of high-fidelity NNPs. Training strategies and sampling approaches are efficiently assessed using empirical potentials and subsequently relabeled via DFT in a highly parallelized fashion for high-fidelity NNP training. Our results reveal that relying solely on energy and force for NNP training is inadequate to prevent overfitting, highlighting the necessity of incorporating stress terms into the loss functions. To optimize training involving force and stress terms, we propose employing transfer learning to fine-tune the weights, ensuring that the potential surface is smooth for these quantities composed of energy derivatives. This approach markedly improves the accuracy of elastic constants derived from simulations in both empirical potential-based NNPs and relabeled DFT-based NNPs. Overall, this study offers significant insights into leveraging empirical potentials to expedite the development of reliable and robust NNPs at the DFT level.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2024

Molecular origin of viscoelasticity and influence of methylation in mesophase pitch

The viscoelastic and thermomechanical properties of pitches are responsible for their melt-spinning behavior, which is a critical step for manufacturing high-performance pitch-based carbon fibers. Here, we systematically explore the impact of methyl group modifications on the viscoelastic and thermal properties of mesophase pitches. We employ a range of atomistic modeling approaches, including Density Functional Theory (DFT), Density Functional Tight Binding (DFTB), and Classical Molecular Mechanics (MM), to provide detailed insights into the molecular interactions and structural changes. Further, our results revealed the molecular mechanisms that promote layered structures leading to the anisotropic nature of the viscoelastic behavior of mesophase pitch. Furthermore, we propose a modified molecular representation of naphthalene-based mesophase pitch based on the analysis of x-ray diffraction measurements. This study provides fundamental insights into the molecular structures of mesophase pitch and the role of methyl groups controlling its viscosity, which offer valuable insights into mesophase-based carbon fiber production.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

DOE OSTI2024

Active learning of neural network potentials for rare events

Developing an automated active learning framework for Neural Network Potentials, focusing on accurately simulating bond-breaking in hexane chains through steered molecular dynamics sampling and assessing model transferability.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2024

Large-scale atomistic model construction of subbituminous and bituminous coals for solvent extraction simulations with reactive molecular dynamics

Large-scale atomistic models for complex polycyclic aromatic hydrocarbon systems help understand the chemical properties and behaviors of complex feedstocks such as coal or petroleum. However, the development and utilization of large-scale models remain limited due to the difficulty in achieving the varied structural characteristics necessary to capture stochastic nature of these feedstocks. Here we demonstrate a systematic workflow to construct stochastic molecular systems from a broad analytical suite: high-resolution transmission electron microscopy (HRTEM), carbon-13 nuclear magnetic resonance spectroscopy ( 13 C NMR), laser desorption ionization mass spectroscopy (LDI-MS), and elemental analysis. We present a model construction and analysis utility of a new Python-based module. We selected one subbituminous and three high-volatile bituminous coals to construct large-scale models (~40,000 atoms). The constructed models were utilized to examine the affinity for solvent extraction (naphthalene or tetralin) and the effect of structural properties (e.g., aromatic cluster size, functional groups, and cross-linking) in reactive molecular dynamics simulations. Complex chemical reactions were monitored with bond order transitions, intermediates formation, and mass distributions. Reactive molecular dynamics simulations suggest a plausible chemical extraction process and products for the complex fossil feedstocks. The results indicated that radical formations with bond breaking of bridging oxygens and carbons were required at high temperatures to facilitate hydrogeneration and extraction of gas molecules from radical-free molecules. We observed that aliphatic chains of tetralin were easily decomposed and combined with radicals to form small size of molecules with aryl bonding, mainly increasing molecules in the 500–1000 Da, while naphthalene had little impact on chemical extraction process.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

DOE OSTI2023

On the role of methyl groups in the molecular architectures of mesophase pitches

The role of methyl groups on the liquid–crystal structure of mesophase pitches was investigated by combining experimental characterizations and atomic-scale computational modeling, using three pitches synthesized from different precursors. Of the three pitches, C-9 alkyl benzene and naphthalene-based pitches have 13 and 7 methyl groups per average polyaromatic hydrocarbon, respectively. By contrast, mesophase produced from a coal-tar pitch has about one methyl group. The coal tar–based mesophase pitch is hydrogen deficient or more aromatic compared with C-9 alkyl benzene- and naphthalene-based pitches. Additionally, X-ray diffraction data showed that average coherent domain sizes of C-9 alkyl benzene (3.7 nm) and naphthalene-based (3.6 nm) pitches with more methyl groups are larger than that of coal tar–based mesophase (2.4 nm). Based on the identified features, the influence of the methyl group on the layering structures was investigated via molecular dynamics simulations. The results revealed that methyl groups are critical in mesophase layering in C-9 alkyl benzene- and naphthalene-based pitches, by reducing CH-π interaction. However, similar alignment could be achieved without the same degree of methyl substitutions for the coal tar-based pitch because of stronger π-π interaction than the other precursors. The insights from this study contribute to our understanding of the formation of conventional mesophase pitch and have implications for the processing of coal-derived materials. In conclusion, this knowledge is vital to produce valuable products like carbon fiber and graphite from pitches.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

DOE OSTI2023

Tailoring Chemical Absorption-Precipitation to Lower the Regeneration Energy of a CO 2 Capture Solvent

Solvent-based CO 2 capture consumes significant amounts of energy for solvent regeneration. To improve energy efficiency, this study investigates CO 2 fixation in a solid form through solvation, followed by ionic self-assembly-aided precipitation. Based on the hypothesis that CO 3 2- ions may bind with monovalent metal ions, we introduced Na + into an aqueous hexane-1,6-diamine solution where CO 2 forms carbamate and bicarbonate. Then, Na + ions in the solvent act as a seed for ionic self-assembly with diamine carbamate to form an intermediate ionic complex. The recurring chemical reactions lead to the formation of an ionic solid from a mixture of organic carbamate/carbonate and inorganic sodium bicarbonate (NaHCO 3 ), which can be easily removed from the aqueous solvent through sedimentation or centrifugation and heated to release the captured CO 2 . Mild-temperature heating of the solids at 80–150 °C causes decomposition of the solid CO 2 -diamine-Na molecular aggregates and discharge of CO 2 . This sorbent regeneration process requires 6.5–8.6 GJ/t CO 2 . It was also found that the organic carbamate/carbonate solid, without NaHCO 3 , contains a significant amount of CO 2 , up to 6.2 mmol CO 2 /g-sorbent, requiring as low as 2.9–5.8 GJ/t CO 2 . In conclusion, molecular dynamic simulations support the hypothesis of using Na + to form relatively less stable, yet sufficiently solid, complexes for the least energy-intensive recovery of diamine solvents compared to bivalent carbonate–forming ions.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

DOE OSTI2023

AL-ASMR: Active Learning of Atomistic Surrogate Models for Rare Events

Atomistic simulation with artificial intelligence (AI) is an emerging tool for understanding materials' properties and behaviors and predicting novel materials with optimized/targeted properties. Neural network potentials (NNPs) are outstanding in this field as they have shown a comparable accuracy to ab initio electronic structure calculations for reproducing potential energy surfaces while being several orders of magnitude faster. However, such NNPs can perform poorly outside of their training domain and typically fail catastrophically in the prediction of rare events in molecular dynamics (MD) simulations. For effective AL loops to distinguish the informative data from enhanced sampled configurations, we developed a decision engine by configurational similarity and uncertainty quantification (UQ) with data augmentation.

Jung, Gang Seob↗

DOE OSTI2023

Artificial neural network potentials for mechanics and fracture dynamics of two-dimensional crystals **

Understanding the mechanics and failure of materials at the nanoscale is critical for their engineering and applications. The accurate atomistic modeling of brittle failure with crack propagation in covalent crystals requires a quantum mechanics-based description of individual bond-breaking events. Artificial neural network potentials (NNPs) have emerged to overcome the traditional, physics-based modeling tradeoff between accuracy and accessible time and length scales. Previous studies have shown successful applications of NNPs for describing the structure and dynamics of molecular systems and amorphous or liquid phases of materials. However, their application to deformation and failure processes in materials is still uncommon. In this study, we discuss the apparent limitations of NNPs for the description of deformation and fracture under loadings and propose a way to generate and select training data for their employment in simulations of deformation and fracture simulations of crystals. We applied the proposed approach to 2D crystalline graphene, utilizing the density-functional tight-binding method for more efficient and extensive data generation in place of density functional theory. Then, we explored how the data selection affects the accuracy of the developed artificial NNPs. It revealed that NNP’s reliability should not only be measured based on the total energy and atomic force comparisons for reference structures but also utilize comparisons for physical properties, e.g. stress–strain curves and geometric deformation. In sharp contrast to popular reactive bond order potentials, our optimized NNP predicts straight crack propagation in graphene along both armchair and zigzag (ZZ) lattice directions, as well as higher fracture toughness of ZZ edge direction. Our study provides significant insight into crack propagation mechanisms on atomic scales and highlights strategies for NNP developments of broader materials.

2D materials↗

DOE OSTI2023

Low-thermal-budget synthesis of monolayer molybdenum disulfide for silicon back-end-of-line integration on a 200 mm platform

Two-dimensional (2D) materials are promising candidates for future electronics due to their excellent electrical and photonic properties. Although promising results on the wafer-scale synthesis (≤150 mm diameter) of monolayer molybdenum disulfide (MoS 2 ) have already been reported, the high-quality synthesis of 2D materials on wafers of 200 mm or larger, which are typically used in commercial silicon foundries, remains difficult. The back-end-of-line (BEOL) integration of directly grown 2D materials on silicon complementary metal–oxide–semiconductor (CMOS) circuits is also unavailable due to the high thermal budget required, which far exceeds the limits of silicon BEOL integration (<400 °C). This high temperature forces the use of challenging transfer processes, which tend to introduce defects and contamination to both the 2D materials and the BEOL circuits. Here we report a low-thermal-budget synthesis method (growth temperature < 300 °C, growth time ≤ 60 min) for monolayer MoS 2 films, which enables the 2D material to be synthesized at a temperature below the precursor decomposition temperature and grown directly on silicon CMOS circuits without requiring any transfer process. We designed a metal–organic chemical vapour deposition reactor to separate the low-temperature growth region from the high-temperature chalcogenide-precursor-decomposition region. We obtain monolayer MoS 2 with electrical uniformity on 200 mm wafers, as well as a high material quality with an electron mobility of ~35.9 cm 2 V -1 s -1 . Finally, we demonstrate a silicon-CMOS-compatible BEOL fabrication process flow for MoS 2 transistors; the performance of these silicon devices shows negligible degradation (current variation < 0.5%, threshold voltage shift < 20 mV). We believe that this is an important step towards monolithic 3D integration for future electronics.

77 NANOSCIENCE AND NANOTECHNOLOGY↗

DOE OSTI2023

Graph neural networks predict energetic and mechanical properties for models of solid solution metal alloy phases

Here, we developed a PyTorch-based architecture called HydraGNN that implements graph convolutional neural networks (GCNNs) to predict the formation energy and the bulk modulus for models of solid solution alloys for various atomic crystal structures and relaxed volumes. We trained the GCNN surrogate model on a dataset for nickel–niobium (NiNb) generated by the embedded atom model (EAM) empirical interatomic potential for demonstration purposes. The dataset was generated by calculating the formation energy and the bulk modulus as a prototypical elastic property for optimized geometries starting from initial body-centered cubic (BCC), face-centered cubic (FCC), and hexagonal compact packed (HCP) crystal structures, with configurations spanning the possible compositional range for each of the three types of initial crystal structures. Numerical results show that the GCNN model effectively predicts both the formation energy and the bulk modulus as function of the optimized crystal structure, relaxed volume, and configurational entropy of the model structures for solid solution alloys.

36 MATERIALS SCIENCE↗