Search NASA⌕ Search

SEARCH · Search NASA

Results for “information structures”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 181 records · Page 10
All sourcesNASA NTRSDOE OSTI
DOE OSTI2022

A Hybrid Structural Method for Investigating Low Molecular Weight Oligomeric Structures of Amyloid Beta

Spurred in part by the failure of recent therapeutics targeting amyloid β plaques in Alzheimer's Disease (AD), attention is increasingly turning to the oligomeric forms of this peptide that form early in the aggregation process. However, while numerous amyloid β fibril structures have been characterized, primarily by NMR spectroscopy and cryo-EM, obtaining structural information on the low molecular weight forms of amyloid β that presumably precede and/or seed fibril formation has proved challenging. Furthermore, these transient forms are heterogeneous, and depend heavily on experimental conditions such as buffer, temperature, concentration, and degree of quiescence during measurement. Here, we present the concept for a new approach to delineating structural features of early-stage low molecular weight amyloid β oligomers, using a solvent accessibility assay in conjunction with simultaneous fluorescence measurements.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2023

Hollow Ptychography: Toward Simultaneous 4D Scanning Transmission Electron Microscopy and Electron Energy Loss Spectroscopy

With the recent development of high-acquisition-speed pixelated detectors, 4D scanning transmission electron microscopy (4D-STEM) is becoming routinely available in high-resolution electron microscopy. 4D-STEM acts as a “universal” method that provides local information on materials that is challenging to extract from bulk techniques. It extends conventional STEM imaging to include super-resolution techniques and to provide quantitative phase-based information, such as differential phase contrast, ptychography, or Bloch wave phase retrieval. However, an important missing factor is the chemical and bonding information provided by electron energy loss spectroscopy (EELS). 4D-STEM and EELS cannot currently be acquired simultaneously due to the overlapping geometry of the detectors. Here, the feasibility of modifying the detector geometry to overcome this challenge for bulk specimens is demonstrated, and the use of a partial or defective detector for ptycholgaphic structural imaging is explored. Here, results show that structural information beyond the diffraction-limit and chemical information from the material can be extracted together, resulting in simultaneous multi-modal measurements, adding the additional dimensions of spectral information to 4D datasets.

4D-scanning transmitting electron microscopy (STEM↗
DOE OSTI2023

Two-Dimensional Energy Histograms as Features for Machine Learning to Predict Adsorption in Diverse Nanoporous Materials

A major obstacle for machine learning (ML) in chemical science is the lack of physically informed feature representations that provide both accurate prediction and easy interpretability of the ML model. In this work, we describe adsorption systems using novel two-dimensional energy histogram (2D-EH) features, which are obtained from the probe-adsorbent energies and energy gradients at grid points located throughout the adsorbent. The 2D-EH features encode both energetic and structural information of the material and lead to highly accurate ML models (coefficient of determination R2 ~ 0.94–0.99) for predicting single-component adsorption capacity in metal–organic frameworks (MOFs). Here, we consider the adsorption of spherical molecules (Kr and Xe), linear alkanes with a wide range of aspect ratios (ethane, propane, n-butane, and n-hexane), and a branched alkane (2,2-dimethylbutane) over a wide range of temperatures and pressures. The interpretable 2D-EH features enable the ML model to learn the basic physics of adsorption in pores from the training data. We show that these MOF-data-trained ML models are transferrable to different families of amorphous nanoporous materials. We also identify several adsorption systems where capillary condensation occurs, and ML predictions are more challenging. Nevertheless, our 2D-EH features still outperform structural features including those derived from persistent homology. The novel 2D-EH features may help accelerate the discovery and design of advanced nanoporous materials using ML for gas storage and separation in the future.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2021

Removal of spurious data in Bragg coherent diffraction imaging: an algorithm for automated data preprocessing

Bragg coherent diffraction imaging (BCDI) provides a powerful tool for obtaining high-resolution structural information from nanocrystalline materials. Here a BCDI sample consisting of a large number of randomly oriented nanoscale crystals is considered. Ideally, only one crystal is oriented to produce a Bragg peak on the detector. However, diffraction from other crystals often produces additional signals on the detector. Before the measured diffraction patterns can be processed into structural images, scientists routinely need to manually identify and remove the `alien' intensities from sources other than the intended crystal. With the development of modern high-coherence storage rings, such as the upgraded Advanced Photon Source (APS), the already slow process of manual preprocessing will be untenable for the large volumes of data that will be produced. An automated method of identifying and deleting alien intensities is proposed. This method exploits the fact that BCDI of a perfect crystal produces diffraction data with inversion symmetry around the Bragg peak. This approach uses the machine learning clustering method DBSCAN to distinguish between diffraction from multiple sources, and then calculates cluster size and inversion symmetry to assess whether clusters of intensity belong to desired data or alien signals. This approach can dramatically reduce the amount of time spent manually processing data, allowing BCDI data processing capabilities to keep pace with the technological advances of fourth-generation synchrotron light sources.

36 MATERIALS SCIENCE↗
DOE OSTI2022

The weathered bedrock vadose zone: A hidden control on water availability in the western United States (Final Technical Report)

This report summarizes the research completed during the project entitled "The weathered bedrock vadose zone: A hidden control on water availability in the western United States" (Award: DE-SC0018039). The report covers work completed between 9/15/17 and 9/14/20. The objective of the exploratory project work was to evaluate the overarching hypothesis that the subsurface underlying actively eroding landscapes is systematically structured as a result of a balance between weathering and erosion and that this structure leads to dynamic moisture storage within weathered rock. To accomplish project objectives, we characterized hillslope weathering profile structure and conducted subsurface moisture monitoring across four study sites with comparable bedrock types but differing erosion rate and hydroclimate: The Angelo Coast Range Reserve (ACRR), Sagehorn-Russell Ranch (SRR), Antelope Valley Ranch (AVR), and the LBNL Watershed Function Scientific Focus Area (SFA). New characterization and monitoring infrastructure at the AVR and SFA sites was established through this project and enables long-term monitoring of hydrologic and carbon fluxes at the sites for comparison with long-term monitoring in other sites. Deep drilling, downhole logging, and characterization of chemistry and porosity of rock recovered from drilling reveals two weathering fronts across the sites. A shallower front is marked by the extent of pervasive fracturing and oxidation of pyrite and organic carbon. A deeper weathering front is marked by the appearance of open fractures and discoloration. At the AVR site, multiple ridge tops of different hillslope length were investigated to establish the relationship between the weathering front and topography. Drilling reveals that the fraction of the hillslope relief that is weathered scales with hillslope length, consistent with the theoretical framework that guided the project relating groundwater drainage with landscape evolution. With the establishment of new boreholes, we have documented significant water storage in weathered bedrock using low-field borehole nuclear magnetic resonance (NMR) across all four sites. In the upper 4 m of bedrock, the average water contents at ACRR, AVR, SRR, and SFA are approximately 25%, 16%, 25%, 23% respectively. At the ACRR site, where long-term monitoring via neutron probe surveys and time-domain transmission sensors are available, we have demonstrated that NMR reliably records moisture content and importantly, moisture content dynamics in weathered bedrock. Across all sites, our NMR monitoring has also revealed that a considerable fraction of the water storage occurs in fractures or pores that are significantly larger than the fine-grained bedrock matrix of the parent rock. To identify the proportion of water storage occurring in fractures and large pores, we analyzed the NMR signal by evaluating both the sum of echoes and an inversion of the NMR signal for a distribution of T2 relaxation times. Both methods reveal that at least 20% of the water storage occurs in fractures and larger pores across all sites. To constrain the interpretation of the in-situ field-scale NMR measurements, we compare the NMR response in the field to independent pore-structure information obtained via helium pycnometry. Results support the inference that dynamic seasonal water storage is dominantly restricted to the fracture network, which can reach 7% of the total volume. This dynamic range in water content is comparable to that of some soils, underscoring the need to understand how this region of the Earth System is structured, and functions as a hydrologic reservoir.

58 GEOSCIENCES↗
DOE OSTI2023

Machine Learning Full NMR Chemical Shift Tensors of Silicon Oxides with Equivariant Graph Neural Networks

The nuclear magnetic resonance (NMR) chemical shift tensor is a highly sensitive probe of the electronic structure of an atom and furthermore its local structure. Recently, machine learning has been applied to NMR in the prediction of isotropic chemical shifts from a structure. Current machine learning models, however, often ignore the full chemical shift tensor for the easier-to-predict isotropic chemical shift, effectively ignoring a multitude of structural information available in the NMR chemical shift tensor. Here we use an equivariant graph neural network (GNN) to predict full 29 Si chemical shift tensors in silicate materials. The equivariant GNN model predicts full tensors to a mean absolute error of 1.05 ppm and is able to accurately determine the magnitude, anisotropy, and tensor orientation in a diverse set of silicon oxide local structures. When compared with other models, the equivariant GNN model outperforms the state-of-the-art machine learning models by 53%. The equivariant GNN model also outperforms historic analytical models by 57% for isotropic chemical shift and 91% for anisotropy. The software is available as a simple-to-use open-source repository, allowing similar models to be created and trained with ease.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2021

Probing extreme states of matter using ultra-intense x-ray radiation

Extreme states of matter, that is, matter at extremes of density (pressure) and temperature, can be created in the laboratory either statically or dynamically. In the former, the pressure–temperature state can be maintained for relatively long periods of time, but the sample volume is necessarily extremely small. When the extreme states are generated dynamically, the sample volumes can be larger, but the pressure–temperature conditions are maintained for only short periods of time (ps to μs). In either case, structural information can be obtained from the extreme states by the use of x-ray scattering techniques, but the x-ray beam must be extremely intense in order to obtain sufficient signal from the extremely-small or short-lived sample. In this article I describe the use of x-ray diffraction at synchrotrons and XFELs to investigate how crystal structures evolve as a function of density and temperature. After a brief historical introduction, I describe the developments made at the Synchrotron Radiation Source in the 1990s which enabled the almost routine determination of crystal structure at high pressures, while also revealing that the structural behaviour of materials was much more complex than previously believed. I will then describe how these techniques are used at the current generation of synchrotron and XFEL sources, and then discuss how they might develop further in the future at the next generation of x-ray lightsources.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
DOE OSTI2022

Current limitations of solid-state NMR in carbohydrate and cell wall research

High-resolution investigation of cell wall materials has emerged as an important application of biomolecular solid-state NMR (ssNMR). Multidimensional correlation experiments have become a standard method for obtaining sufficient spectral resolution to determine the polymorphic structure of carbohydrates and address biochemical questions regarding the supramolecular organization of cell walls. Here, using plant cellulose and matrix polysaccharides as examples, we will review how the multifaceted complexity of polysaccharide structure is impeding the resonance assignment process and assess the available biochemical and spectroscopic approaches that could circumvent this barrier. We will emphasize the ineffectiveness of the current methods in reconciling the ever-growing dataset and deriving structural information. We will evaluate the protocols for achieving efficient and homogeneous hyperpolarization across the cell wall material using magic-angle spinning dynamic nuclear polarization (MAS-DNP). Critical questions regarding the line-broadening effects of cell wall molecules at cryogenic temperature and by paramagnetic biradicals will be considered. Finally, the MAS-DNP method will be placed into a broader context with other structural characterization techniques, such as cryo-electron microscopy, to advance ssNMR research in carbohydrate and cell wall biomaterials.

09 BIOMASS FUELS↗
DOE OSTI2026

Imaging Three-Dimensional Molecular Structure and Dynamics with Multiparticle Covariance and Cumulant Coulomb Explosion Analysis

Coulomb explosion imaging (CEI) provides a direct means of imaging molecular geometry by correlating fragment ion momenta following the fragmentation of a molecular polycation. Here, we demonstrate the use of three-body covariance and four-body cumulant analysis to extract three-dimensional (3D) structural information from the X-ray-induced Coulomb explosion of tert-butyl iodide (C4H9I). Site-selective ionization at the iodine 4d edge with intense femtosecond soft X-ray pulses from an X-ray free-electron laser (XFEL) enables rapid charge buildup and molecular breakup. By correlating ionic fragments in the molecular frame, we isolate complete dissociation channels and reveal subtle structural changes, such as umbrella-type motion of the branched alkyl chain, during the ionization process. Comparison with point-charge simulations of the Coulomb explosion shows close agreement, validating the approach. Furthermore, these results establish covariance/cumulant mapping as a powerful strategy for imaging complex three-dimensional molecular structures and point the way toward time-resolved CEI using both XFEL and tabletop sources for capturing ultrafast structural dynamics.

Imaging↗
DOE OSTI2020

Data-driven surrogates for high dimensional models using Gaussian process regression on the Grassmann manifold

This paper introduces a surrogate modeling scheme based on Grassmannian manifold learning to be used for cost-efficient predictions of high-dimensional stochastic systems. The method exploits subspace-structured features of each solution by projecting it onto a Grassmann manifold. This point-wise linear dimensionality reduction harnesses the structural information to assess the similarity between solutions at different points in the input parameter space. The method utilizes a solution clustering approach in order to identify regions of the parameter space over which solutions are sufficiently similarly such that they can be interpolated on the Grassmannian. In this clustering, the reduced-order solutions are partitioned into disjoint clusters on the Grassmann manifold using the eigen-structure of properly defined Grassmannian kernels and, the Karcher mean of each cluster is estimated. Then, the points in each cluster are projected onto the tangent space with origin at the corresponding Karcher mean using the exponential mapping. For each cluster, a Gaussian process regression model is trained that maps the input parameters of the system to the reduced solution points of the corresponding cluster projected onto the tangent space. Using this Gaussian process model, the full-field solution can be efficiently predicted at any new point in the parameter space. In certain cases, the solution clusters will span disjoint regions of the parameter space. In such cases, for each of the solution clusters we utilize a second, density-based spatial clustering to group their corresponding input parameter points in the Euclidean space. The proposed method is applied to two numerical examples. Here, the first is a nonlinear stochastic ordinary differential equation with uncertain initial conditions where the surrogate is used to predict the time history solution. The second involves modeling of plastic deformation in a model amorphous solid using the Shear Transformation Zone theory of plasticity, where the proposed surrogate is used to predict the full strain field of a material specimen under large shear strains.

42 ENGINEERING↗
DOE OSTI2026

PARP1-HPF1 structure and dynamics on nicked DNA suggest a mechanism for acute and localized ADP-ribosylation

PARP1 detection of DNA strand breaks allosterically leads to PARP1 synthesis of poly(ADP-ribose) modifications that signal DNA damage. HPF1 engages activated PARP1 to control modification site selection. Understanding of the mechanism of DNA break detection and catalytic activation is incomplete, due largely to limited structural information for full-length PARP1. Here, single-particle cryo-EM provides views of the full complement of PARP1 domains engaging a DNA single-strand break in the presence of HPF1 and a fragment of binding partner Timeless. Cryo-EM, single-molecule DNA dynamics, and small-angle X-ray scattering analysis indicate that PARP1 remains dynamic even when the multi-domain structure is organized on a DNA break, with the minimal catalytic region displaying high mobility relative to domains engaging damage. We propose that the organization of PARP1 domains on a DNA break releases a tethered, constitutively active catalytic region to modify molecules in a radius surrounding the DNA break site.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2024

Data Generation for Machine Learning Interatomic Potentials and Beyond

The field of data-driven chemistry is undergoing an evolution, driven by innovations in machine learning models for predicting molecular properties and behavior. Recent strides in ML-based interatomic potentials have paved the way for accurate modeling of diverse chemical and structural properties at the atomic level. The key determinant defining MLIP reliability remains the quality of the training data. A paramount challenge lies in constructing training sets that capture specific domains in the vast chemical and structural space. This Review navigates the intricate landscape of essential components and integrity of training data that ensure the extensibility and transferability of the resulting models. We delve into the details of active learning, discussing its various facets and implementations. We outline different types of uncertainty quantification applied to atomistic data acquisition and the correlations between estimated uncertainty and true error. The role of atomistic data samplers in generating diverse and informative structures is highlighted. Furthermore, we discuss data acquisition via modified and surrogate potential energy surfaces as an innovative approach to diversify training data. The Review also provides a list of publicly available data sets that cover essential domains of chemical space.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2020

Isolating the Contributions of Specific Network Sites to the Diffuse Vibrational Spectrum of Interfacial Water with Isotopomer-Selective Spectroscopy of Cold Clusters

Decoding the structural information contained in the interfacial vibrational spectrum of water requires understanding how the spectral signatures of individual water molecules respond to their local hydrogen bonding environments. In this study, we isolated the contributions for the five classes of sites that differ according to the number of donor (D) and acceptor (A) hydrogen bonds that characterize each site. These patterns were measured by exploiting the unique properties of the water cluster cage structures formed in the gas phase upon hydration of a series of cations M+·(H2O)n (M = Li, Na, Cs, NH4, CH3NH3, H3O, and n = 5, 20-22). This selection of ions was chosen to systematically express the A, AD, AAD, ADD, and AADD hydrogen bonding motifs. The spectral signatures of each site were measured using two-color, IR-IR isotopomer-selective photofragmentation vibrational spectroscopy of the cryogenically cooled, mass selected cluster ions in which a single intact H2O is introduced without isotopic scrambling, an important advantage afforded by the cluster regime. The resulting patterns provide an unprecedented picture of the intrinsic line shapes and spectral complexities associated with excitation of the individual OH groups, as well as the correlation between the frequencies of the two OH groups on the same water molecule, as a function of network site. The properties of the surrounding water network that govern this frequency map are evaluated by dissecting electronic structure calculations that explore how changes in the nearby network structures, both within and beyond the first hydration shell, affect the local frequency of an OH oscillator. The qualitative trends are recovered with a simple model that correlates the OH frequency with the network-modulated local electron density in the center of the OH bond.

Yang, Nan↗
DOE OSTI2021

H 2 O-assisted O 2 reduction by H 2 on Pt and PtAu bimetallic nanoparticles: Influences of composition and reactant coverages on kinetic regimes, rates, and selectivities

Hydrogen peroxide (H 2 O 2 ) can replace hazardous oxidants in industrial processes but is currently too expensive for many such applications. While direct synthesis of H 2 O 2 (H 2 + O 2 → H 2 O 2 ) may reduce costs in comparison to incumbent technology, current catalysts lack the requisite stability and selectivity. Here, we examine the direct synthesis of H 2 O 2 on bimetallic Pt 1 Au x (0 ≤ x ≤ 230) and Pt catalysts at steady-state in pure water and relate kinetic parameters for H 2 O 2 and H 2 O formation to possible active site structures informed by complementary characterization methods. X-ray photoelectron spectra show significant Pt surface enrichment compared to the bulk composition. Analysis of infrared spectra of mixed monolayers of 12 CO* and 13 CO* indicate that Pt and Au form substitutional surface alloys. The Pt 1 Au x nanoparticles with the greatest mole fractions of Au predominantly expose Pt monomers (i.e., isolated Pt atoms), yet Pt atoms exposed upon all these nanoparticles possess electronic structures distinct from bulk Pt. Despite these differences, rate measurements are consistent with product formation through proton-electron transfer pathways for all Pt 1 Au x catalysts. In situ XAS indicate that Pt remains metallic during H 2 O 2 synthesis. Under the most oxidizing conditions, selectivities toward H 2 O 2 increase strongly with the Au to Pt ratio from 2% for monometallic Pt to 85% for Pt 1 Au 170 . However, selectivities are similar among all catalysts within reducing conditions. Comparisons of apparent activation enthalpies for the formation of H 2 O 2 and H 2 O across these catalysts and the range of conditions suggest that Pt monomers within Au provide the greatest selectivities for H 2 O 2 formation, because these active sites present high barriers for O-O bond rupture. Further, selectivities decrease with increasing ratios of H 2 to O 2 pressures, because Pt atoms aggregate and form oligomers that readily dissociate dioxygen intermediates. The combined use of spectroscopy, kinetics, and concepts employed in reaching these conclusions take inspiration from the legacy of Prof. Michel Boudart, and specifically his elegant methods for interrogating bimetallic catalysts.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2026

Unveiling the electronic structure and chemical bonding of the deprotonated cisplatin anion [(NH 3 )(NH 2 )PtCl 2 ] − via low-temperature photoelectron spectroscopy and theoretical calculations

The dehydrogenated cisplatin anion, [(NH 3 )(NH 2 )PtCl 2 ] − , was investigated via low-temperature photoelectron spectroscopy and theoretical calculations. Seven and four spectral peaks are respectively resolved at 193 and 266 nm, yielding rich electronic structure information for both the anion and neutral. From the threshold and maximum of the lowest electron binding energy band, the experimental adiabatic (ADE) and vertical detachment energies (VDE) are determined to be 3.3 ± 0.1 and 3.525 ± 0.025 eV, respectively. Theoretical calculations indicate the dominant isomer adopting a cis-geometry, in which the platinum center is coplanar with two chlorine and two nitrogen ligands. The calculated VDE of 3.54 eV based on this structure agrees well with the experimental value. Charge analyses reveal that the excess electron in the anion is primarily localized on the Pt and Cl atoms. A suite of theoretical analysis tools was employed to elucidate the bonding characteristics and interaction strength between Pt and its ligands.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2025

Elucidating molecular scale interactions underlying the freezing behavior of salt solutions in silica nanopores

Elucidating the influence of nanoscale confinement on the freezing behavior of salt solutions is of fundamental interest to environmental security and materials science and engineering. Specific structural information such as the coordination environment of ions in confined aqueous medium, effects on the extended hexagonal network of water molecules and their mutual non-bonding interactions in confinement are sparse in the literature along with the change in the dynamical characteristics of water in the presence of ions. To address these knowledge gaps, the current study is focused on investigating the influence of reduced dimensionality arising from nanoscale confinement on the structural evolution of salt solutions on freezing and the associated fluid–surface interactions. In this regard, operando wide-angle X-ray scattering (WAXS) measurements and classical molecular dynamics (MD) simulations are conducted with water and 0.5 M CaCl 2 , MgCl 2 and KCl solutions confined in 4 nm sized SBA-15 silica pores upon cooling from 300 K to 200 K. The freezing point of the salt solutions is depressed to 235 K which is about 10 K lower than that of confined water. The translational dynamics of confined salt solutions indicates shifts in the fragile-to-strong dynamical crossover to a lower temperature compared to pure water. The strong electrostatic attraction between the cations and the surrounding water molecules contributes to the freezing point depression in confined salt solutions. These insights unlock the molecular-scale basis and mechanisms underlying the freezing behavior of confined salt solutions.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2022

Atomically Resolved Electrically Active Intragrain Interfaces in Perovskite Semiconductors

Deciphering the atomic and electronic structures of interfaces is key to developing state-of-the-art perovskite semiconductors. However, conventional characterization techniques have limited previous studies mainly to grain-boundary interfaces, whereas the intragrain-interface microstructures and their electronic properties have been much less revealed. Herein using scanning transmission electron microscopy, we resolved the atomic-scale structural information on three prototypical intragrain interfaces, unraveling intriguing features clearly different from those from previous observations based on standalone films or nanomaterial samples. These intragrain interfaces include composition boundaries formed by heterogeneous ion distribution, stacking faults resulted from wrongly stacked crystal planes, and symmetrical twinning boundaries. The atomic-scale imaging of these intragrain interfaces enables us to build unequivocal models for the ab initio calculation of electronic properties. Our results suggest that these structure interfaces are generally electronically benign, whereas their dynamic interaction with point defects can still evoke detrimental effects. This work paves the way toward a more complete fundamental understanding of the microscopic structure-property-performance relationship in metal halide perovskites.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
DOE OSTI2024

Intricate short-range order in GeSn alloys revealed by atomistic simulations with highly accurate and efficient machine-learning potentials

GeSn alloys hold promise for silicon-compatible integrated applications in electronics, photonics, and topological quantum devices. However, understanding their intricate structures using density functional theory (DFT) calculations is hindered by spatiotemporal constraints. To overcome this limitation, we develop highly accurate and efficient machine-learning interatomic potentials based on a neuroevolution potential approach with farthest point sampling on a comprehensive DFT data set. The application of the developed machine-learning potential in large-scale atomistic simulations bridges the spatiotemporal gap between modeling and advanced characterizations, and facilitates the discovery of structural intricacies in GeSn alloys. Through extensive statistical sampling, we identify a type of short-range order (SRO) that is distinguished by both its structural signature and electronic band gap from the SRO structure previously predicted. Modeling based on a large simulation cell reveals the coexistence of nano SRO domains with various degrees of ordering, demonstrating a complex spatial heterogeneity of SRO structure. Finally, our study not only reinforces the significance of fine-level structural information in alloys, but it also constitutes an effective framework for exploring SRO in a broad range of complex alloys based on highly accurate and effective machine-learning potentials.

36 MATERIALS SCIENCE↗