Search NASA⌕ Search

SEARCH · Search NASA

Results for “TEM image analysis”

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 19 records
All sourcesNASA NTRSDOE OSTI
DOE OSTI2022

Understanding the Influence of Receptive Field and Network Complexity in Neural Network-Guided TEM Image Analysis

Abstract Trained neural networks are promising tools to analyze the ever-increasing amount of scientific image data, but it is unclear how to best customize these networks for the unique features in transmission electron micrographs. Here, we systematically examine how neural network architecture choices affect how neural networks segment, or pixel-wise separate, crystalline nanoparticles from amorphous background in transmission electron microscopy (TEM) images. We focus on decoupling the influence of receptive field, or the area of the input image that contributes to the output decision, from network complexity, which dictates the number of trainable parameters. For low-resolution TEM images which rely on amplitude contrast to distinguish nanoparticles from background, we find that the receptive field does not significantly influence segmentation performance. On the other hand, for high-resolution TEM images which rely on both amplitude and phase-contrast changes to identify nanoparticles, receptive field is an important parameter for increased performance, especially in images with minimal amplitude contrast. Rather than depending on atom or nanoparticle size, the ideal receptive field seems to be inversely correlated to the degree of nanoparticle contrast in the image. Our results provide insight and guidance as to how to adapt neural networks for applications with TEM datasets.

42 ENGINEERING↗
DOE OSTI2026

Automated Bacterial Identification and Morphological Feature Analysis in Low‐Dose Cryo‐EM Using YOLOv11

Bacteria rapidly adapt to environmental cues through morphological and ultrastructural changes that correlate with physiology and behavior. Cryogenic transmission electron microscopy (cryo‐TEM) can capture these phenotypic changes in near‐native, vitrified states, but manual analysis of low‐dose micrographs is labor intensive and limits throughput. Here, we present an end‐to‐end workflow that combines low‐dose cryo‐TEM imaging with a YOLOv11‐based instance‐segmentation model to automatically identify bacteria and quantify key structural features directly from the micrographs. This workflow enables (i) robust bacterial localization and counting from low‐magnification atlas/montage images, (ii) automated measurements of cell‐envelope (outer–inner membrane) thickness and anisotropy from higher‐magnification views, and (iii) detection and quantification of bacteria–flagella interactions, including overlap length and curvature metrics for interacting versus noninteracting flagella. Using Pantoea sp. YR343 grown under distinct media conditions, we show that the automated measurements agree with manual annotations while substantially reducing analysis time. Together, these tools provide a practical framework for scalable bacterial identification and quantitative phenotyping in low‐dose cryo‐TEM datasets and establish a foundation for extending cryo‐TEM image analysis toward higher‐throughput studies of microbial heterogeneity and biointerfaces.

YOLOv11↗
DOE OSTI2025

MultiTaskDeltaNet: change detection-based image segmentation for operando ETEM with application to carbon gasification kinetics

Transforming in situ transmission electron microscopy (TEM) imaging into a tool for spatially-resolved operando characterization of solid-state reactions requires automated, high-precision semantic segmentation of dynamically evolving features. However, traditional deep learning methods for semantic segmentation often face limitations due to the scarcity of labeled data, visually ambiguous features of interest, and scenarios involving small objects. To tackle these challenges, we introduce MultiTaskDeltaNet (MTDN), a novel deep learning architecture that creatively reconceptualizes the segmentation task as a change detection problem. By implementing a unique Siamese network with a U-Net backbone and using paired images to capture feature changes, MTDN effectively leverages minimal data to produce high-quality segmentations. Furthermore, MTDN utilizes a multi-task learning strategy to exploit correlations between physical features of interest. In an evaluation using data from in situ environmental TEM (ETEM) videos of filamentous carbon gasification, MTDN demonstrated a significant advantage over conventional segmentation models, particularly in accurately delineating fine structural features. Notably, MTDN achieved a 10.22% performance improvement over conventional segmentation models in predicting small and visually ambiguous physical features. This work bridges key gaps between deep learning and practical TEM image analysis, advancing automated characterization of nanomaterials in complex experimental settings.

08 HYDROGEN↗
DOE OSTI2026

The ever-evolving active site: transformation of single atoms to extended structures during the Rh-catalyzed reverse water–gas shift reaction

At low temperatures (< 400°C), single atoms of Rh supported on rutile TiO 2 (rTiO 2 ) are responsible for the formation of CO during the reverse water gas shift (RWGS), while methane production is associated with the Rh-TiO 2 interface due to the correlation between methane formation rates and the volume-averaged Rh nanoparticle diameter. As the temperature is increased to >540°C, there is a notable increase in CO selectivity and methane production rates tend towards zero. The time to reach zero depends on the temperature but independent of the initial Rh structure (single atoms and/or nanoparticles), which is controlled by the catalyst preparation method (wetness impregnation versus colloidal nanoparticles). At 600°C and > 4 h time on stream, the catalytic behaviour becomes completely agnostic to the initial Rh structure as well as weight loading, and the catalysts are highly selective for the reverse water gas-shift reaction. Post-reaction HR-TEM image analysis confirms Rh nanoparticles crystallize/order during the reaction; at 400°C, most of the Rh particles are disordered, while at 600°C, they are more ordered (i.e., development of defined faceting). Infrared spectroscopy of CO adsorption on Rh nanoparticles confirms the appearance of defined facets after annealing in nitrogen at high temperatures. Annealing the Rh/rTiO 2 catalysts prior to the RWGS reaction demonstrates the structural transformation of Rh depends only on time and temperature and not on reactant or product fugacity. Sites responsible for stabilizing Rh single atoms are no longer competent at higher temperatures, enabling their integration into existent nanoparticles. As the reaction temperature is increased to temperatures >540°C, the dominant active site for CO production evolves from single atoms to extended Rh structures.

36 MATERIALS SCIENCE↗
DOE OSTI2022

Characterization of Engineered Pinholes in Dielectric Stacks of High-Performance Poly-Silicon Passivating Contacts

Passivating contact structures are expected to be implemented in > 50 % of industrially manufactured silicon photovoltaics produced over the next decade. One type of passivating contact utilizes a c-Si/dielectric/doped poly-Si layer stack. In cases where the dielectric layer/s are thick enough to impede tunneling of charge carriers, it was shown that disruptions (pinholes) in the dielectric are required to facilitate transport.[1] Pinholes in SiOx layers have typically been formed via high temperature annealing that allows localized breakup of this layer. This approach has drawbacks; 1) high temperature annealing increases cost and 2) pinhole formation is sensitive to both the SiOx thickness and wafer surface morphology making control of density and dimensions difficult. Recently, an alternative method of pinhole formation relying on metal assisted chemical etching (MACE) was shown to allow control over pinhole areal densities.[2] Such control is critical as both the density and dimensions of the pinholes determine saturation current density and contact resistivity. The contacts formed with the MACE process are known as polysilicon on locally etched oxide or PLEO and also result in low saturation current densities and low contact resistivity due to SiOx passivation and engineered pinhole enabled charge carrier transport, respectively. Photovoltaic devices with efficiencies greater than 20 % have recently been produced with these contacts, demonstrating the viability of contacts with engineered pinholes. Experimentally, preferential local collection of excess charge carriers has been observed and definitively connected to thermally induced pinholes in thick SiOx layers within c-Si/SiOx/poly-Si contact structures.[3] Recent work has shown that enhanced local collection of excess carriers can result from enhanced diffusion of dopants through pinholes in SiOx layers.[4] This study was restricted to simulation and characterization of poly-Si with the same dopant type as the c-Si substrates. In this work we focus our analysis on boron doped poly-Si with different stacks of dielectric layers (PLEO and PLENO where the N refers to a SiNy layer on top of the SiOx) on n-type crystalline silicon. This provides insight into how engineered pinholes affect charge collection across the p-n junction. Like previous observations of locally enhanced carrier transport through thermally induced pinholes, the EBIC images in Figure 1 shows that engineered pinholes in PLENO (1a) and PLEO (1b) also exhibit locally enhanced carrier transport at pinhole locations (bright spots in the images). We will also present detailed correlative nanoscale structural and chemical analysis using TEM imaging and EELS analysis with sample preparation guided by EBIC imaging. This allows us to connect the density and dimensions of engineered pinholes with charge carrier transport and measured photovoltaic device parameters. Our findings will accelerate the adoption of PLEO and PLENO contacts in industrially manufactured photovoltaic devices.

EBIC↗
NASA NTRS2004

Microstructural Evaluations of Baseline HSR/EPM Disk Alloys

Six alloys representing two classes of powder metallurgy nickel-based superalloys were examined by transmission electron microscopy (TEM) and phase extraction. Alloys KM4, CH98, IN-100 and 456 are based on a Ni-18Co-12Cr composition while alloys Rene' 88 DT and SR 3 have lower Al and Co and higher Cr contents. The lambda size distributions were determined from quantitative image analysis of the TEM images. The volume fraction of lambda and carbides and the composition of the phases were determined by a combination of phase extraction and TEM. The results showed many similarities in lambda size distributions, grain boundary serrations, and grain boundary carbide frequencies between alloys KM4, CH98, 456, Rene' 88 DT and SR 3 when heat treated to give an approximate grain size of ASTM 6. The density of grain boundary carbides in KM4 was shown to substantially increase as the grain size increased. IN-100 and 456 subjected to a serration cooling heat treatment had much more complex lambda size distributions with very large intergranular and intragranular secondary lambda as well as finer than average cooling and aging lambda. The grain boundary carbides in IN-100 were similar to the other alloys, but 456 given the serration cooling heat treatment had a more variable density of grain boundary carbides. Examination of the phases extracted from the matrix showed that there were significant differences in the phase chemistries and elemental partitioning ratios between the various alloys.

Gabb, Timothy P.↗
DOE OSTI2020

Nucleation, growth, and superlattice formation of nanocrystals observed in liquid cell transmission electron microscopy

This work reviews the advancements and prospects of liquid cell transmission electron microscopy (TEM) imaging and analysis methods in understanding the nucleation, growth, etching, and assembly dynamics of nanocrystals. The bonding of atoms into nanoscale crystallites produces materials with nonadditive properties unique to their size and geometry. The recent application of in situ liquid cell TEM to nanocrystal development has initiated a paradigm shift, (1) from trial-and-error synthesis to a mechanistic understanding of the “synthetic reactions” responsible for the emergence of crystallites from a disordered soup of reactive species (e.g., ions, atoms, molecules) and shape-defined growth or etching; and (2) from post-processing characterization of the nanocrystals’ superlattice assemblies to in situ imaging and mapping of the fundamental interactions and energy landscape governing their collective phase behaviors. Imaging nanocrystal formation and assembly processes on the single-particle level in solution immediately impacts many existing fields, including materials science, nanochemistry, colloidal science, biology, environmental science, electrochemistry, mineralization, soft condensed-matter physics, and device fabrication.

36 MATERIALS SCIENCE↗
DOE OSTI2023

A deep learning model for automatic analysis of cavities in irradiated materials

Transmission electron microscopy (TEM) is a commonly used technique in materials science for defect investigation. Quantitative analysis of defects is important for understanding the properties of a material, but manual analysis of TEM micrographs can be time-consuming and prone to error, especially when the defects have irregular shapes rather than spherical shapes. Many existing methods or deep learning models do not handle a wide range of sizes for the same object type within a single image. In this work, we present a framework that enables users to train an instance segmentation model called Mask R- CNN on any microstructure dataset, perform multi-detection on the same image at different scales, and obtain properties (e.g., size, area) of the objects based on the desired shape (e.g., circle, ellipse, rectangle). Additionally, we have developed a parallel detection module that uses multiple GPUs to increase the efficiency of the object detection process. We demonstrate the capabilities of our framework using a set of TEM images of cavities with different shapes, size distributions, and background contrasts. Finally, we show that the performance of our model in terms of density, size, and swelling of the cavities is comparable to the human average and that our model achieves the highest recall value compared to existing methods due to the use of image multi-rescaling.

36 MATERIALS SCIENCE↗
NASA NTRS1984

Structural characterizaiton and gas reactions of small metal particles by high-resolution, in-situ TEM and TED

The existing in-situ transmission electron microscopy (TEM) facility was improved by adding a separately pumped mini-specimen chamber. The chamber contains wire-evaporation sources for three metals and a specimen heater for moderate substrate temperatures. A sample introduction device was constructed, installed, and tested, facilitating rapid introduction of a specimen into the mini-chamber while maintaining the background pressure in that chamber in the 10(-9) millibar range. Small particles and clusters of Pd, grown by deposition from the vapor phase in an in-situ TEM facility on amorphous and crystalline support films of alumina and on ultra-thin carbon films, were analyzed by conventional high-resolution TEM and image analysis in terms of detectability, number density, and size distribution. The smallest particles that could be detected and counted contained no more than 6 atoms; size determinations could be made for particles 1 nm in diameter. The influence of various oxygen plasma treatments, annealing treatments, and of increasing the substrate temperature during deposition was investigated. The TEM technique was employed to demonstrate that under otherwise identica l conditions the lattice parameter of Pd particles in the 1 to 2 nm size range and supported in random orientation on ex-situ prepared mica films is expanded by some 3% when compared to 5 nm size particles. It is believed that this expansion is neither a small-particle diffraction effect nor due to pseudomorphism, but that it is due to a annealing-induced transformation of the small as-deposited particles with predominantly composite crystal structures into larger particles with true f.c.c. structure and thus inherently smaller lattice parameter.

Source record↗
DOE OSTI2021

Deep Learning Segmentation of Complex Features in Atomic-Resolution Phase-Contrast Transmission Electron Microscopy Images

Phase-contrast transmission electron microscopy (TEM) is a powerful tool for imaging the local atomic structure of materials. TEM has been used heavily in studies of defect structures of two-dimensional materials such as monolayer graphene due to its high dose efficiency. However, phase-contrast imaging can produce complex nonlinear contrast, even for weakly scattering samples. It is, therefore, difficult to develop fully automated analysis routines for phase-contrast TEM studies using conventional image processing tools. For automated analysis of large sample regions of graphene, one of the key problems is segmentation between the structure of interest and unwanted structures such as surface contaminant layers. In this study, we compare the performance of a conventional Bragg filtering method with a deep learning routine based on the U-Net architecture. We show that the deep learning method is more general, simpler to apply in practice, and produces more accurate and robust results than the conventional algorithm. We provide easily adaptable source code for all results in this paper and discuss potential applications for deep learning in fully automated TEM image analysis.

47 OTHER INSTRUMENTATION↗
DOE OSTI2025

Probing the role of local tunnel variations in early-stage lithiation of α-MnO₂ nanowires via in situ TEM

Understanding lithium-ion transport in tunnel-structured manganese oxides is essential for designing high-performance lithium-ion battery electrode materials. Here, we elucidate the early-stage lithiation mechanism of potassium-stabilized α-MnO 2 nanowires using in situ transmission electron microscopy (TEM) coupled with electron energy-loss spectroscopy (EELS), high-resolution TEM (HRTEM), and geometric phase analysis (GPA). Real-time TEM imaging reveals clear volume expansion at the reaction front, while EELS analysis uncovers lithium-ion diffusion far beyond this region, where no visible expansion is observed, indicating fast, defect-assisted transport. GPA and HRTEM analyses show that localized tensile and compressive strain fields, originating from pre-existing local tunnel structural variations, persist after lithiation. The tensile-strained regions enable lithium-ion insertion with minimal lattice distortion, offering additional free volume that facilitates rapid lithium-ion accommodation ahead of the structural transformation. Our results demonstrate a local tunnel variation-mediated fast diffusion pathway that precedes bulk reaction, underscoring the critical role of local strain in enabling early-stage lithium transport. Given the structural versatility of MnO 2 and its ability to accommodate diverse atomic arrangements beyond the well-known tunnel phases (β-, γ-, δ-, λ-, R-phases), our findings highlight the importance of understanding and engineering local structural environments. This work provides fundamental insights into the interplay between defects, strain, and ion dynamics, and presents defect engineering as a promising approach to enhance both rate performance and structural stability in manganese-based cathodes.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
DOE OSTI2025

Enhanced Monte Carlo Simulations for Electron Energy Loss Mitigation in Real-Space Nanoimaging of Thick Biological Samples and Microchips

High-resolution imaging using Transmission Electron Microscopy (TEM) is essential for applications such as grain boundary analysis, microchip defect characterization, and biological imaging. However, TEM images are often compromised by electron energy spread and other factors. In TEM mode, where the objective and projector lenses are positioned downstream of the sample, electron–sample interactions cause energy loss, which adversely impacts image quality and resolution. This study introduces a simulation tool to estimate the electron energy loss spectrum (EELS) as a function of sample thickness, covering electron beam energies from 300 keV to 3 MeV. Leveraging recent advances in MeV-TEM/STEM technology, which includes a state-of-the-art electron source with 2-picometer emittance, an energy spread of 3 × 10 -5 , and optimized beam characteristics, we aim to minimize energy spread. By integrating EELS capabilities into the BNL Monte Carlo (MC) simulation code for thicker samples, we evaluate electron beam parameters to mitigate energy spread resulting from electron–sample interactions. Based on our simulations, we propose an experimental procedure for quantitively distinguishing between elastic and inelastic scattering. The findings will guide the selection of optimal beam settings, thereby enhancing resolution for nanoimaging of thick biological samples and microchips.

36 MATERIALS SCIENCE↗
DOE OSTI2024

Application of Weak-Beam Dark-Field STEM for Dislocation Loop Analysis

Nanoscale dislocation loops formed by irradiation can significantly contribute to both irradiation hardening and embrittlement of materials when subjected to extreme nuclear reactor environments. Here, this study explores the application of weak-beam dark-field (WBDF) scanning transmission electron microscopy (STEM) methods for quantitative irradiation-induced defect analysis in crystalline materials, with a specific focus on dislocation loop imaging and analysis. A high-purity Fe-5 wt% Cr model alloy was irradiated with 8 MeV Fe 2+ ions at 450°C to a fluence of 8.8 × 10 19 m -2 , inducing dislocation loops for analysis. While transmission electron microscopy (TEM) has traditionally been the primary tool for dislocation imaging, recent advancements in STEM technology have reignited interest in using STEM for defect imaging. This study introduces and compares three WBDF STEM methods, demonstrating their effectiveness in suppressing background contrasts, isolating defect information for dislocation loop type classification, providing finer dislocation line images for small loop analysis, and presenting inside–outside contrast for identifying loop nature. Experimental findings indicate that WBDF STEM methods surpass traditional TEM approaches, yielding clearer and more detailed images of dislocation loops. The study concludes by discussing the potential applications of WBDF STEM techniques in defect analysis, emphasizing their adaptability across various material systems beyond nuclear materials.

36 MATERIALS SCIENCE↗
DOE OSTI2023

Machine learning for automated experimentation in scanning transmission electron microscopy

Abstract Machine learning (ML) has become critical for post-acquisition data analysis in (scanning) transmission electron microscopy, (S)TEM, imaging and spectroscopy. An emerging trend is the transition to real-time analysis and closed-loop microscope operation. The effective use of ML in electron microscopy now requires the development of strategies for microscopy-centric experiment workflow design and optimization. Here, we discuss the associated challenges with the transition to active ML, including sequential data analysis and out-of-distribution drift effects, the requirements for edge operation, local and cloud data storage, and theory in the loop operations. Specifically, we discuss the relative contributions of human scientists and ML agents in the ideation, orchestration, and execution of experimental workflows, as well as the need to develop universal hyper languages that can apply across multiple platforms. These considerations will collectively inform the operationalization of ML in next-generation experimentation.

36 MATERIALS SCIENCE↗
DOE OSTI2025

A comparative analysis of YOLOv8 and U-Net image segmentation approaches for transmission electron micrographs of polycrystalline thin films

Metallic thin films offer a platform to experimentally study the dynamics of microstructural evolution, but the required transmission electron microscopy (TEM)-based imaging generates complex images that are challenging to segment and quantify. This work provides a comparative analysis of a new YOLOv8 model and an established U-Net model for bright-field TEM images of polycrystals, employing a framework leveraging physical observables to evaluate performance against two hand-traced benchmark datasets. This methodology obviates the comparison of large, diversely structured, and manually labeled datasets that are required to assess performance on a per-image/per-pixel basis. It is found that the YOLOv8 model, adapted for real-time instance segmentation, has up to 43× faster inferencing (NVIDIA GeForce RTX 4090) compared to U-Net and reconstructs hand-traced grain size distributions (GSDs) with excellent fidelity, finding mean diameter within 3% for grains near an optimal magnification; for grains that deviate from the optimal pixel-diameter, the size of small- (large)-diameter grains is systematically over- (under)-estimated. This is partially mitigated by including scale-aware augmentations during training. Moreover, when the bias is corrected post-inference by a rigid shift in distribution, the YOLOv8 model reproduces ground truth GSDs with exceptional fidelity, with statistical tests indicating <5% probability that the distributions are distinct. Based on ground truth data, calibration curves pertaining to this shift can be constructed for a given model. This issue is not present in the U-Net model’s results, indicating that for quantitative measurements where the true size of objects is of interest, special procedures must be implemented for YOLO-based models.

36 MATERIALS SCIENCE↗
DOE OSTI2025

Computer Vision Pipeline for Image Analysis for Freeze‐Fracture Electron Microscopy: Rosette Cellulose Synthase Complexes Case

In materials science, plant biology, agriculture, and environmental research, the automated analysis of high-magnification, complex microscopy images, such as those generated by freeze-fracture electron microscopy (FF-TEM), remains a critical challenge that limits the scalability of data interpretation. We present a deep learning computer vision pipeline for high-throughput detection and morphological characterization analysis of cellulose synthase complexes (CSCs, or rosettes) in FF-TEM images. The pipeline integrates preprocessing, detection, human-in-the-loop verification, and semantic segmentation to quantify features such as rosette diameter and inter-lobe spacing. The approach was trained and tested on a curated dataset of high-resolution FF-TEM micrographs of Physcomitrium patens, expanded via strategic tiling and augmentation to over 650 images. We compare YOLOv8 and YOLOv9 architectures and demonstrate that YOLOv9 achieves superior performance in both localization accuracy (mAP50-95 = 0.854) and inference speed. The resulting distributions revealed biological variability consistent with prior manual studies, validating the approach for high-throughput applications. Our results show that the pipeline achieves human-expert level accuracy while dramatically reducing analysis time, enabling scalable, reproducible structural characterization of intramembrane protein complexes. The pipeline is broadly applicable to other domains requiring precise interpretation of complex microscopy data and establishes a foundation for future artificial intelligence (AI)-assisted workflows in biological imaging.

59 BASIC BIOLOGICAL SCIENCES↗