Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Heterogeneous materials”

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 1,495 records · Page 83

Development of Machine Learning Algorithms to Segment and Study Images of Astromaterial Samples

Introduction: Micrometer-scale chemical analyses of chondritic meteorites and mission-returned asteroid samples can reveal details of the physical and chemical processes operating in the early solar system, including processes that gave rise to planets, moons, and minor bodies. These primitive astromaterials are comprised of chondrules, calcium- and aluminum-rich inclusions (CAI), and many other silicates, oxides, metals, sulfides, and fine-grained materials. The chemical and mineralogical complexity of these samples, vast populations of different components, and heterogeneity across mm to km scales, all limit our understanding of the origin and evolution of these materials. Here, we describe recent efforts to use machine learning techniques to automate the segmentation of chemical maps of chondritic meteorites, designed to aid studies of asteroid samples returned by spacecraft. By automating the task of segmentation it will become possible to rapidly analyze and interpret the sizes, shapes, mineralogy, chemistry, and other properties of every chondrule, calcium- and aluminum-rich inclusion (CAI) and other clast within and between asteroid samples. Sample return missions significantly accelerate and heighten the need to develop such new data analysis techniques, and associated data repositories. Techniques: Neural networks require abundant training data, i.e. images which have been segmented by a human user. We have manually segmented data available from previous petrologic and chemical work at NASA Johnson Space Center and the American Museum of Natural History [1-4]. These data were derived from energy- and wavelength-dispersive X-ray spectroscopy (EDS, WDS) mapping of samples from many chondrite groups. The Deeplabv3+ [5] neural network architecture was trained on human-labeled masks and used to create machine-labeled masks. Several different algorithms were investigated, with inputs ranging from common RGB image formats through to hyperspectral datasets, with raw data comprising greyscale maps of Mg, Ca, and Al, with or without Si, Fe, Ti for both EDS and WDS data, and extending to other elements in EDS only. Each greyscale image was paired with a binary mask for each labelled particle type. Results: The trained algorithms can segment (Fig 1), classify, and measure the dimensions of thousands of particles in chemical maps of a standard 1-inch round petrographic section in seconds to minutes, rather than many hours needed by a human. Accuracy of the algorithms varied from chondrite to chondrite and across particle types. Further results and details of the algorithms will be presented at the workshop. Future directions: Machine learning has the potential to revolutionize our understanding of complex particle populations contained within primitive astromaterial, with segmentation being a critical first step. Example applications include better understanding of particle transport, nebular reservoirs, parent body accretion, and a deeper understanding of the relationships between particle populations and bulk rock elemental and isotopic compositions. In addition to benefits that machine learning can bring to individual researchers, building a community data repository of thousands to millions of particles across hundreds of samples will open up many other possibilities. For example, with a large enough dataset it will be possible to search for exceptionally closely matching particles across disparate samples. Such a capability would enable a single CAI from OSIRISREx or Hayabusa/II samples to be matched to chondritic CAIs that exhibit near-identical size, texture, and mineralogy, down to the level of similar core phenocrysts, zonation, and rim sequences. Such comparative analyses will help to disentangle precursor chemistry, chronology, gas/dust reservoirs during heating, and accretion. Such an endeavor would be impossible without machine learning and a large community data repository of astromaterial chemical/mineralogic maps.

Machine Learning↗

Anisotropic Thermal Conductivity of Exfoliated Black Phosphorus

We ascertain the anisotropic thermal conductivity of passivated black phosphorus (BP), a reactive 2D nanomaterial with strong in-plane anisotropy. We measure the room temperature thermal conductivity by time-domain thermoreflectance for three crystalline axes of exfoliated BP. The thermal conductivity along the zigzag direction (86 ± 8 W m -1 K -1 ) is ~2.5 times higher than that of the armchair direction (34 ± 4 W m -1 K -1 ).

36 MATERIALS SCIENCE↗

Combining ToF‐SIMS and Multivariate Analysis to Resolve Active Sites on Ni‐Based HER Catalysts

Unambiguous identification of active sites in heterogeneous catalysis remains a major challenge, particularly for materials with ultrathin, chemically mixed surface layers. Here, we demonstrate a generalizable approach that combines time-of-flight secondary ion mass spectrometry (ToF-SIMS) with multivariate statistical analysis (principal component analysis [PCA] and multivariate curve resolution [MCR]) to resolve catalytically relevant motifs at the nanoscale. Using Ni electrodes as a model system, PCA distinguished hydroxide-enriched domains from oxide- and metal-rich regions, while MCR decomposed depth profiles and 3D images into hydroxide, oxide, and metallic layers with nanometer resolution. A unique secondary-ion fragment, NiO 3 H 3 − (m/z 108.94), emerged as a marker of hydroxide-rich environments and correlated with hydrogen evolution reaction (HER) activity across a series of Ni electrodes. Complementary density functional theory (DFT) calculations revealed that Ni(OH) 2 clusters adjacent to metallic Ni offer the most favorable water dissociation energetics, establishing the structural origin of the marker. While illustrated here for Ni-based HER, this workflow provides a broadly applicable framework to isolate and rank near-surface patterns that govern catalytic activity, thereby extending ToF-SIMS from a qualitative probe to a predictive tool for active site identification.

HER active sites↗

Ultra-Stable and High-Cobalt-Loaded Cobalt@Ordered Mesoporous Carbon Catalysts: All-in-One Deoxygenation of Ketone into Alkylbenzene

Catalytic deoxygenation represents a straightforward and core methodology for fine-chemical production and biomass upgrading. Generally, the application of homogeneous metal complexes or heterogeneous noble-metal catalysts prevails in academia and the chemical industry. We introduce cobalt@ordered mesoporous carbon (Co@OMC) catalysts, which are constructed conveniently by a mechanochemical coordination self-assembly based on a renewable tannin precursor. Importantly, the Co@OMC catalysts with a high loading of in situ confined Co species promote the selective deoxygenation of various ketones, aldehydes, and alcohols efficiently into the corresponding alkanes under mild conditions. Therefore, a simple, inexpensive, and heterogeneous catalyst for selective deoxygenation can be expected, meanwhile the solid-state synthesis affords a green, rapid, and scalable pathway to Co@OMC catalysts.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Reduction of NO by CO on unsupported Ir: Bridging the materials gap

Here, temperature programmed desorption (TPD) and density functional theory (DFT) are used to investigate adsorption sites and reaction of coadsorbed NO and CO on planar Ir(210) and faceted Ir(210) with tailored sizes of three-sided nanopyramids exposing (311), (311) and (110) faces. Both planar and faceted Ir(210) are highly active for reduction of NO by CO with high selectivity to N 2 , which is accompanied by simultaneous oxidation of CO. Evidence is found for structure sensitivity in adsorption sites and reaction of coadsorbed NO and CO on faceted Ir(210) versus planar Ir(210). Strong interaction between NO and CO at high NO exposure and one-monolayer CO pre-coverage results in "explosive" evolution of N 2 and CO 2 , on planar Ir(210) and size effects in reduction of NO by CO on faceted Ir(210) for average facet size ranging from 5 to 14 nm without change in facet structure.

36 MATERIALS SCIENCE↗

Nanofabrication and Demonstration of a Direct-Write Microevaporator

Direct-write vapor deposition is a new technique that would enable one-step 3D maskless nanofabrication on a variety of substrates. A novel silicon chip-based microevaporator is developed that allows evaporant to exit through 2000–300 nm nozzles while held at distances comparable to the nozzle diameter from the substrate by a three-axis nanopositioning stage in vacuum. This results in a localized deposition on the substrate, which may be scanned relative to the substrate to produce direct-write patterns. The performance of the microevaporator is tested by creating localized depositions of various materials and the line-writing potential is demonstrated. The relationship between linewidth and source-to-substrate distance is investigated by the application of Knudsen's cosine law and Monte-Carlo simulations, and then utilized to approximate the source-to-substrate distance from performed depositions.

36 MATERIALS SCIENCE↗

Enhanced accuracy through ensembling of randomly initialized auto-regressive models for dynamical systems

Computational mechanics simulations using traditional finite element methods (FEM) require prohibitively expensive computational resources for real-time engineering applications, design optimization, and digital twin implementations. While machine learning (ML) surrogate models offer significant computational speedups, autoregressive ML models for time-dependent mechanical systems suffer from error accumulation that compromises long-term prediction reliability - a critical concern for engineering applications where accuracy over extended time horizons is essential for safety and performance assessments. Here, we propose a deep ensemble framework specifically designed to address this challenge in computational mechanics applications, where multiple ML surrogate models with random weight initializations are trained in parallel and their predictions aggregated during inference. This approach leverages statistical diversity to maximize information gain from a fixed set of training data and to mitigate error propagation, while maintaining the computational efficiency that makes ML surrogates attractive for engineering practice. We validate the framework on three representative problems spanning critical areas of computational mechanics: stress field evolution in heterogeneous microstructures under complex loading (relevant to advanced materials design and composite analysis), planetary-scale shallow water dynamics (applicable to environmental and geotechnical engineering), and Gray-Scott reaction-diffusion systems (relevant to mass transport and chemical process engineering). Across all test cases, the ensemble approach demonstrates consistent error reduction of 15-33% compared to individual models. The codes for this work are available on GitHub (https://github.com/Graham-Brady-Research-Group/AutoregressiveEnsemble_SpatioTemporal_Evolution).

autoregressive prediction↗

Influence of concentration polarization and thermodynamic non-ideality on salt transport in reverse osmosis membranes

The classic Merten and Lonsdale transport model for reverse osmosis membranes was reformulated to explicitly demonstrate the effects of concentration polarization and solution phase thermodynamic non-idealities on salt transport. A framework presented here accounts for the concentration dependence of ion activity coefficients in salt solutions, which was not explicitly included in the classic model. This approach was applied to four salt solutions, NaCl, MgCl 2 , CaCl 2 , and Na 2 SO 4 , tested in cross-flow conditions for a commercial RO membrane, Dow Filmtec™ BW30XFR. Salt transport coefficients corrected for concentration polarization and non-ideal thermodynamic effects, B a , were calculated as a function of permeate flux and compared with apparent salt transport coefficients, B . Finally, these corrections were significant, resulting in B a values greater than B values by a factor of 1.3–2.1 for 2:1 and 1:2 salts (i.e., MgCl 2 , CaCl 2 , and Na 2 SO 4 ). B a values for NaCl (a 1:1 salt), however, were similar to or somewhat smaller than B values.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Altering Linear Scaling Relationships on Metal Catalysts via Ligand–Adsorbate Hydrogen Bonding

Linear scaling relationships for adsorption energies of related molecules provide a simple tool for the prediction of catalytic properties and also reveal inherent constraints in heterogeneous catalyst design. In an effort to predict materials that preferentially stabilize intermediates capable of forming hydrogen bonds (H-bonds) and thus reduce these constraints, scaling relations for thiolate-coated fcc (111) surfaces were developed. Here, we demonstrate how ligand–adsorbate H-bonds lead to the stabilization of certain adsorbates and how this affects the linear scaling of adsorption energies for similarly bound intermediates. For H-bonds that occur remote from the surface, ligand–adsorbate H-bond strength is independent of metal composition and instead depends only on the acidities of the H-bond-forming functional groups. As the distance of H-bond-accepting groups from the surface decreases and interactions with the metal strengthen, the effect of surface composition on H-bond strength increases while additional factors, such as conformational changes and steric hinderance, can offset the stabilizing effect of the ligands. In this work, these competing factors may influence both the slope and y-intercept of adsorption strength scaling. These findings aid the rational design of enhanced catalytic materials by enabling the screening of ligand- or spectator-modified materials that do not conform to the linear scaling relations of bare metal surfaces.

36 MATERIALS SCIENCE↗

Understanding the Mechanism of Secondary Cation Release from the (001) Surface of Li(Ni 1/3 Mn 1/3 Co 1/3 )O 2 : Insights from First-Principles

The transformations of complex metal oxides in aqueous settings must be studied to form a chemical understanding of how technologically relevant nanomaterials impact the environment upon disposal. Owing to the inherent heterogeneity and structural complexity of the ternary intercalation material Li(Ni x Mn y Co 1-x-y )O 2 (NMC), the mechanisms of chemical processes at the solid–water interfac e are challenging to model. Here, density functional theory (DFT) + solvent ion methodology is used to study the energetics of stepwise release of two surface metals following unique pathways. Further, the study spans different combinations of metal removal and also considers unique patterns of defects formed by modeling the NMC surface in supercells. The approach here also considers the equilibration of the surface with the surroundings between successive metal removals. A key finding is that a second metal removal prefers to proceed at a metal lattice site adjacent to the initial defect, and this is attributed in part to how the resulting slab with two metal vacancies maintains the most antiferromagnetic couplings between the remaining Ni/Mn.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Understanding Charge Transfer in Carbon Nanotube–Fullerene Bulk Heterojunctions

Semiconducting single-walled carbon nanotube/fullerene bulk heterojunctions exhibit unique optoelectronic properties highly suitable for flexible, efficient, and robust photovoltaics and photodetectors. We investigate charge-transfer dynamics in inverted devices featuring a polyethylenimine-coated ZnO nanowire array infiltrated with these blends and find that trap-assisted recombination dominates transport within the blend and at the active layer/nanowire interface. We find that electrode modifiers suppress this recombination, leading to high performance.

36 MATERIALS SCIENCE↗

High Throughput Screening of 3D Printable Resins: Adjusting the Surface and Catalytic Properties of Multifunctional Architectures

Identification of 3D printable materials is crucial to expand the breadth of physical and chemical properties attainable by additive manufacturing. Stereolithography (SLA), a widespread 3D printing method based on resin photopolymerization, is ideally suited for exploring a large variety of monomers to produce functional three-dimensional solids of diverse properties. However, for most commercial SLA printers, screening monomers and resin compositions requires large volumes (~150 mL) in each printing cycle, making the process costly and inefficient. Herein, a high throughput block (HTB) adaptor was developed to screen arrays of monomers and resin compositions, consuming lower volumes (<2 mL) and less time per print (<1/16 based on a 4 × 4 matrix) than using the original hardware. By use of this approach, a library of materials with different surface hydrophobicities were 3D printed by including long chain acrylates in the resins. In addition, several metal salts were dissolved in an acrylic acid-based resin, 3D printed, and screened as benzyl alcohol oxidation catalysts. Cu(II)-based resins produced the most active structures. Combinations of Cu(II) and long chain acrylate monomers were then used to 3D print complex catalytic architectures with varying degrees of hydrophobicity. Linear relationships were observed between 3D printed surface area, surface hydrophobicity, and catalyst performance. For a high surface Schwarz P topology ca. 60% enhancement in the catalytic activity of Cu(II) was attained by replacing the parent resin with one containing hydrophobic isodecyl groups, indicating that the immediate environment of the catalytic site affected its performance. The HTB enables fast screening of resins for 3D printing multifunctional architectures with intrinsic catalytic activity, tunable surface properties, and minimal waste.

36 MATERIALS SCIENCE↗