Search NASA⌕ Search

SEARCH · Search NASA

Results for “property predictions”

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 487 records · Page 27
All sourcesNASA NTRSDOE OSTI
DOE OSTI2021

Audacity of huge: overcoming challenges of data scarcity and data quality for machine learning in computational materials discovery

Machine learning (ML)-accelerated discovery requires large amounts of high-fidelity data to reveal predictive structure–property relationships. For many properties of interest in materials discovery, the challenging nature and high cost of data generation has resulted in a data landscape that is both scarcely populated and of dubious quality. Data-driven techniques starting to overcome these limitations include the use of consensus across functionals in density functional theory, the development of new functionals or accelerated electronic structure theories, and the detection of where computationally demanding methods are most necessary. When properties cannot be reliably simulated, large experimental data sets can be used to train ML models. In the absence of manual curation, increasingly sophisticated natural language processing and automated image analysis are making it possible to learn structure–property relationships from the literature. Finally, models trained on these data sets will improve as they incorporate community feedback.

36 MATERIALS SCIENCE↗
DOE OSTI2022

Thermodynamic modeling of aqueous lithium salt solutions with association electrolyte nonrandom two-liquid activity coefficient model

The high charge density of lithium ion and the resulting strong association phenomena make thermodynamic modeling of aqueous lithium electrolyte solutions extremely challenging. In this study, the association electrolyte nonrandom two-liquid activity coefficient model of Lin et al. (AIChE J. 2022, 68(2), e17422) is utilized to correlate and predict thermodynamic properties and solubility behavior of aqueous single electrolyte solutions of LiCl, LiBr, LiI, and LiNO 3 , and their mixed electrolyte solutions. Capturing self-association of water, cross-association of ion and water for hydration, and cross-association of cation and anion for ion-pairing, the association model accurately represents the literature experimental data up to saturation concentrations and at the temperature ranging from 263 K to 523 K. Here, this study further investigated the effect of anions of the lithium salts, and re-confirmed that the order of solution non-ideality as LiI > LiBr > LiCl > LiNO 3 because the anions with stronger association strengths are more likely to form ion pairs and thus lower the mean ionic activity coefficients.

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

A deep learning-accelerated data assimilation and forecasting workflow for commercial-scale geologic carbon storage

Fast assimilation of monitoring data to update forecasts of pressure buildup and carbon dioxide (CO 2 ) plume migration under geologic uncertainties is a challenging problem in geologic carbon storage. The high computational cost of data assimilation with a high-dimensional parameter space impedes fast decision-making for commercial-scale reservoir management. We propose to leverage physical understandings of porous medium flow behavior with deep learning techniques to develop a fast data assimilation-reservoir response forecasting workflow. Applying an Ensemble Smoother Multiple Data Assimilation (ES-MDA) framework, the workflow updates geologic properties and predicts reservoir performance with quantified uncertainty from pressure history and CO 2 plumes interpreted through seismic inversion. As the most computationally expensive component in such a workflow is reservoir simulation, we developed surrogate models to predict dynamic pressure and CO 2 plume extents under multi-well injection. The surrogate models employ deep convolutional neural networks, specifically, a wide residual network and a residual U-Net. The workflow is validated against a flat threedimensional reservoir model representative of a clastic shelf depositional environment. Intelligent treatments are applied to bridge between quantities in a true-3D reservoir model and those in a single-layer reservoir model. The workflow can complete history matching and reservoir forecasting with uncertainty quantification in less than one hour on a mainstream personal workstation.

25 ENERGY STORAGE↗
DOE OSTI2026

Creep in multi-principal element materials –– A review

The ongoing push towards enhanced energy efficiency and reduced emissions has necessitated the creation of materials with superior performance, especially under extreme conditions. Modern industries, such as aerospace, energy production, and nuclear power, rely heavily on materials that can withstand elevated temperatures without compromising structural integrity. At these heightened temperatures, materials, even when subjected to mechanical stresses well below their yield strength, may experience slow deformation leading to eventual rupture — a phenomenon known as creep. With the expansive design space that comes with the high entropy concept and their reported excellent high temperature strength, multi-principal element materials (MPEMs) have attracted interest in the scientific community for high-temperature applications. Here, this review offers a comprehensive examination of existing studies on creep in MPEMs, which includes multi-principal element−alloys, −bulk metallic glasses, −ceramics, and −superalloys, comparing published findings on MPEMs with pure elements, traditional alloys, bulk metallic glasses, and superalloys. The sub-topics covered include a comparison among different creep-testing methods, creep mechanisms, creep exponents, creep strain rates, activation volume, and creep-activation energy. Modeling efforts for predicting creep behavior of MPEMs are also reviewed. Methods for improving creep resistance by performing heat treatments and/or modifying microstructures are discussed. Overall, the current state of MPEMs has not yet surpassed the creep performance of commercial alloys. Finally, directions for future efforts are suggested, such as experimenting in various controlled environments, expanding the number of compositions tested, exploring advanced manufacturing techniques, and using machine-learning to predict creep properties based on compositions and microstructures.

36 MATERIALS SCIENCE↗
DOE OSTI2024

Machine learning-based discovery of molecular descriptors that control polymer gas permeation

While machine learning has found increasing use in predicting the properties of polymeric materials with only a knowledge of chain architecture, determining the molecular factors underpinning properties (“interpretable AI”) has remained less well explored. We show that encoding chain chemistry in commonly employed formats, e.g., binary-valued fingerprints, leads to uniqueness issues during the hashing process to save storage space. This is because the hashing algorithm can map several chemical moieties into the same bit. These issues carry over into the ML algorithms, especially for “inverse” design and interpretable AI, and cannot be avoided by changing the length of the fingerprint. Using MACCS key featurizations of monomer repeats resolves some of these issues, and we show that a few substructures consistently appear in top features for maximizing permeability across several gases and ML models. These are carbon–carbon double bonds (as in polyacetylenes) especially when they are associated with methyl groups (found in branching architectures). Here these results, derived from the limited data set of ~ 500 polymers with experimental gas permeation data, are in agreement with physical insight and thus provide a robust foundation which could further enable study of these material classes through detailed experiments and simulations.

36 MATERIALS SCIENCE↗
DOE OSTI2021

Technical and economic feasibility of molten chloride salt thermal energy storage systems

A techno-economic study is performed to assess the feasibility of molten chloride salt thermal energy storage (TES) systems for next generation concentrating solar power. Refractory liners internally insulate tanks to allow tank shells to be constructed from carbon steel. The liner must not be wetted by salt to maintain predictable thermal properties and manageable heat loss out of the tank. The commercial scale tank liner is an anchored brick and mortar design with expansion joints to accommodate thermal expansion. Furthermore, finite element analysis is performed to optimize the thermal and mechanical profile of the tank. Equalizing the shell temperature between the water-cooled foundation and the shell wall is necessary to minimize differential thermomechanical stress and lower overall stress values below industrial allowable limits. The cost of the TES system is estimated to be $60/kWhth, which is four times greater than Department of Energy targets. Solutions to reduce system cost and overall risk are proposed.

14 SOLAR ENERGY↗
DOE OSTI2025

Boosting Noise2Inverse via enhanced model selection for denoising computed tomography data

Synchrotron-based x-ray tomographic imaging enables the examination of the internal structure of materials at high spatial and temporal resolution. Experimental constraints can impose dose and time limits on the measurements, introducing a higher level of noise and artifacts in the reconstructed images. Deep learning has emerged as a powerful tool to remove noise from reconstructed images. Recently, the Noise2Inverse method was designed specifically for denoising reconstructed images without requiring paired noisy and clean images. This method creates multiple statistically independent reconstructions used to pair the data in which training involves transforming one reconstruction into the other, and vice versa. Originally designed to be used after a fixed number of epochs, we see in practice that this approach may not produce the optimal model and may unnecessarily waste computational resources. Therefore, we propose an alternative method of identifying the best model during training that aligns with the Noise2Inverse method. During validation, we compare the model output of the multiple reconstructions among each other. We hypothesize that the best model is the one that produces images with the highest similarity, implying a convergence in the predicted material properties and absorption values. To compare model outputs, we consider the absolute error, square error, structural similarity index (SSIM), peak signal-to-noise ratio (PSNR), and cosine similarity. We evaluate our method on two simulated tomography datasets and two, real-world, low-contrast, high-energy x-ray tomography datasets. We show our approach is more effective at determining the best model, up to an increase of 12.50% and 12.53% in SSIM and PSNR, respectively, while only requiring a fifth of the training time compared to the original approach.

CT↗
DOE OSTI2025

Fabrication and characterization of boron-terminated tetravacancies in monolayer hBN using STEM, EELS and electron ptychography

Tetravacancies in monolayer hexagonal boron nitride (hBN) with consistent edge termination (boron or nitrogen) form triangular nanopores with electrostatic potentials that can be leveraged for applications such as selective ion transport and neuromorphic computing. In order to quantitatively predict the properties of these structures, an atomic-level understanding of their local electronic and chemical environments is required. Moreover, robust methods for their precision manufacture are needed. Here we use electron irradiation in a scanning transmission electron microscope (STEM) at a high dose rate to drive the formation of boron-terminated tetravacancies in monolayer hBN. Characterization of the defects is achieved using aberration-corrected STEM, monochromated electron energy-loss spectroscopy (EELS), and electron ptychography. Z-contrast in STEM and chemical fingerprinting by core-loss EELS enable identification of the edge terminations, while electron ptychography gives insight into structural relaxation of the tetravacancies and provides evidence of enhanced electron density around the defect perimeters indicative of bonding effects.

2D hBN↗
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 OSTI2022

Coarse-Grained Density Functional Theory Predictions via Deep Kernel Learning

Scalable electronic predictions are critical for soft materials design. Recently, the Electronic Coarse-Graining (ECG) method was introduced to renormalize all-atom quantum chemical (QC) predictions to coarse-grained (CG) resolutions using deep neural networks (DNNs). While DNNs can learn complex representations that prove challenging for kernel-based methods, they are susceptible to overfitting and the overconfidence of uncertainty estimations. Here, we develop ECG within a GPU-accelerated Deep Kernel Learning (DKL) framework to enable CG QC predictions using range-separated hybrid density functional theory (DFT), obtaining a 107 speedup relative to naive all-atom QC. By treating the predicted electronic properties as random Gaussian Processes, DKL incorporates CG mapping degeneracy by learning the distribution of electronic energies as a function of CG configuration. DKL-ECG accurately reproduces molecular orbital energies from range-separated DFT while facilitating efficient training via active learning using the uncertainties provided by DKL. Further, we show that while active learning algorithms enable efficient sampling of a more diverse configurational space relative to random sampling, all explored query methods exhibit comparable performance for the examined system. We attribute this result to the significant overlap of the feature space and output property distributions across multiple temperatures.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Neural Network Water Model Based on the MB-Pol Many-Body Potential

The MB-pol many-body potential accurately predicts many properties of water, including cluster, liquid phase, and vapor–liquid equilibrium properties, but its high computational cost can make applying it in large-scale simulations quite challenging. Here, in order to address this limitation, we developed a “deep potential” neural network (DPMD) model based on the MB-pol potential for water. We find that a DPMD model trained on mostly liquid configurations yields a good description of the bulk liquid phase but severely underpredicts vapor–liquid coexistence densities. By contrast, adding cluster configurations to the neural network training set leads to a good agreement for the vapor coexistence densities. Liquid phase densities under supercooled conditions are also represented well, even though they were not included in the training set. These results confirm that neural network models can combine accuracy and transferability if sufficient attention is given to the construction of a representative training set for the target system.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2024

Electron Radiation-Induced Nickel Nanoparticle Formation in KCl-ZnCl 2 Eutectic Molten Salt Using X-ray Absorption Spectroscopy

Understanding the speciation and local structure of metal solutes in molten salts is essential for predicting thermal properties and the redox and corrosion potentials of molten salts for next-generation nuclear reactors and concentrated solar power plant applications. Here, we employ X-ray absorption spectroscopy combined with electron paramagnetic resonance (EPR) measurements to investigate the effects of electron irradiation on 0.1 wt % NiCl 2 dissolved in molten eutectic KCl-ZnCl 2 salt. Radiation-driven reduction of Ni 2+ leading to nucleation and growth of Ni nanoparticles is studied as a function of electron dose (3–15 MGy) and temperature. Quantitative X-ray absorption near edge structure and linear combination fitting are utilized to investigate the extent of reduction as a function of the electron dose and temperature. A multiple-scattering (MS) approach is used for extended X-ray absorption fine structure analysis to deduce the size of the Ni nanoparticles formed and to understand the effect of the dose on the local structure of the nanoparticles. EPR studies on solidified NiCl 2 dissolved in molten eutectic KCl-ZnCl 2 salt show the formation of magnetic Ni nanoparticles.

36 MATERIALS SCIENCE↗
DOE OSTI2025

Resolving the Valence of Iron Oxides by Resonant Photoemission Spectroscopy

Precisely determining the oxidation states of metal cations within variable-valence transition metal oxides remains a significant challenge, yet it is crucial for understanding and predicting the properties of these technologically important materials. Iron oxides, in particular, exhibit a remarkable diversity of electronic structures due to the variable valence states of iron (Fe 2+ and Fe 3+ ). A quantitative analysis using conventional X-ray photoelectron spectroscopy (XPS) is challenging because of the strong overlap of the Fe 2p XPS peaks from different oxidation states. Here, in this study, we show how this problem can be resolved using Resonant Photoemission Spectroscopy (ResPES), which unambiguously distinguishes Fe oxidation states and spectroscopically estimates the composition ratio of Fe cation valence states in the complex Fe oxides. We demonstrate this in the model case of a FeO 2 monolayer film on Pt(111), showing that the FeO 2 film consists of an equal mixture of Fe 2+ and Fe 3+ cations, yielding an average valence of +2.5, contrary to the +3 valence proposed based on density functional theory (DFT).

Chen, Hao [Lawrence Berkeley National Laboratory (↗
DOE OSTI2023

Novel Structural Motif To Promote Mg-Ion Mobility: Investigating ABO 4 Zircons as Magnesium Intercalation Cathodes

There is an increasing need for sustainable energy storage solutions as fossil fuels are replaced by renewable energy sources. Multivalent batteries, specifically Mg batteries, are one energy storage technology that researchers continue to develop with hopes to surpass the performance of Li-ion batteries. However, the limited energy density and transport properties of Mg cathodes remain critical challenges preventing the realization of high-performance multivalent batteries. In this work, ABO 4 zircon materials (A = Y, Eu and B = V, Cr) are computationally and experimentally evaluated as Mg intercalation cathodes. Remarkably good Mg-ion transport properties were predicted and Mg-ion intercalation was experimentally verified in sol-gel synthesized zircon YVO 4 , EuVO 4 , and EuCrO 4 . Among them, EuVO 4 exhibited the best electrochemical performance and demonstrated repeated reversible cycling. While we believe that the one-dimensional diffusion channels and redox-active species tetragonal coordination limit the value of many zircons as high-performance cathodes, their unique structural motif of overlapping polyhedra along the diffusion pathway appears instrumental for promoting good Mg-ion mobility. The motif results in a favorable "6-5-4" change in coordination that avoids unfavorable sites with lower coordination along the diffusion pathway and a structural design metric for future Mg cathode development.

25 ENERGY STORAGE↗
DOE OSTI2025

Raman Digital Twin of Monolayer Janus Transition Metal Dichalcogenides

Monolayer transition metal dichalcogenides (TMDs) are a key class of two-dimensional (2D) materials with broad technological potential. Their Janus counterparts exhibit unique properties due to broken out-of-plane symmetry and further enrich the functionalities of TMDs. However, experimental synthesis and identification of Janus TMDs remain challenging. It is thus highly desirable to have a rapid, simple, and in situ characterization technique to monitor, in real time, the conversion process from the parent to Janus structure. Raman spectroscopy stands out for such a task as it is a powerful, nondestructive, and very commonly used tool to characterize 2D materials both in situ and ex situ. To realize the full potential of Raman spectroscopy on rapid characterization of Janus TMDs, we present a computational “Raman digital twin” library for various monolayer Janus TMDs in both 2H and Td phases. We focus on group-6 TMDs: MoS 2 , WS 2 , MoSe 2 , WSe 2 , MoTe 2 , WTe 2 and their Janus variants: MoSSe, MoSTe, MoSeTe, WSSe, WSTe, and WSeTe. Using first-principles density functional theory (DFT), we calculate their vibrational properties and predict distinct Raman fingerprints. These phonon and Raman signatures reflect each material’s structural symmetry and atomic composition, enabling clear identification via Raman spectroscopy. Our theoretical work supports experimental efforts by providing benchmarks for material identification, structural analysis, and quality control. In conclusion, the computational library expedites the discovery and development of Janus 2D materials, facilitating tighter integration between theoretical predictions and experimental validation.

Chalcogenides↗
DOE OSTI2021

Large-Area, Highly Crystalline DNA-Assembled Metasurfaces Exhibiting Widely Tunable Epsilon-Near-Zero Behavior

Metasurfaces prepared via bottom-up nanoparticle assembly enable the deliberate manipulation of light in the optical regime, resulting in media with unprecedented optical responses. Here, we report a scalable method to grow highly crystalline 2D metasurfaces comprised of colloidal gold nanocubes, over macroscopic areas, using DNA-mediated assembly under equilibrium conditions. Using an effective medium description, we predict that these plasmonic metasurfaces behave as dielectric media with high refractive indices that can be dynamically tuned by tuning DNA length. Furthermore, we predict that when coupled with an underlying thin gold film, the real permittivity of these metasurfaces exhibits a crossover region between positive and negative values, known as the epsilon-near-zero (ENZ) condition, which can be tuned between 1.5 - 2.6 μm by changing DNA length. Optical characterization performed on the DNA-assembled metasurfaces reveals that the predicted optical properties agree well with measured response. Overall, we propose an efficient method for realizing large-area plasmonic metasurfaces that enable dynamic control over unique optical characteristics. High-index and ENZ metasurfaces operating in the telecommunications regime could have significant implications in high-speed optical computing, optical communications, optical imaging, and other areas.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗