Off-axis electron holography is a quantitative phase contrast technique that allows the measurement of electromagnetic fields both within and around the sample. The sensitivity of the electric and magnetic fields (and sources) is thus largely dependent on the quality of the measured phase. Phase retrieval from holograms is usually achieved by a standard Fourier transform method. The aperture used in the reciprocal-space reconstruction process alters the nature of the noise from independence to correlated among pixels. This correlation is further enhanced if additional operator is used, for example, in the calculation of the charge density using the Laplacian operator, and of the magnetic induction and electric field using the gradient operator on the phase. Therefore, it is of importance to trace the noise and correlation of the phase in the processing of experimental holograms, in order to understand the limit of the noise (phase resolution), and further detect infinitesimal tiny electromagnetic signals. In this work, we systematically analyze the noise and correlation of the phase as a function of a variety of parameters, including the detector, electron dose, reconstruction aperture size, Gaussian smoothing and binning. Furthermore, we also assess the effect of the Laplacian and gradient operators on the phase. In particular, we provide detailed analyses on the possibility of detection of a single electric charge and Bohr magnetron. This finding would provide a solid basis for experimental measurements of very weak electromagnetic signals using off-axis electron holography and other phase contrast techniques in transmission electron microscopy.
Existing grain detection algorithms for EBSD data primarily identify grains based on predefined boundary criteria, leading to poor fidelity of grain detection. Significant deviations may exist between the mean orientation of a grain and the orientations of individual pixels. Here, we propose a variational method based post-processing framework that refines an existing grain detection and addresses this issue by framing the problem as a piecewise constant Mumford-Shah problem, and solving it using a variational method grounded in energy minimization principles. The algorithm employs an iterative thresholding method to enhance pixel orientation similarity within grains while balancing grain boundary length. Compared to existing grain detection algorithms, this approach demonstrates superior fidelity of grain detection, as evidenced by validation on a high-strength low-alloy bainitic steel EBSD map. Setting the threshold misorientation angle as 5° for grain detection, the fraction of pixels with orientation deviation angle from grain mean orientation larger than 5° were reduced from 0.186 for the conventional flood fill algorithm to 0.009 for the proposed one. The detected grain number even decreases from 1093 to 1033. In addition, the influences of different grain detection results on texture analysis and parent phase reconstruction were discussed. Based on the grains detected by the proposed algorithm, the orientation distribution function and pole figure textures characteristic aligns better with those of the EBSD data, and more physically reasonable parent phase reconstruction results were obtained for the case examined here, compared with those based on the conventional flood fill algorithm. This proposed framework offers a powerful tool for grain-detection refinement from EBSD data, and enables more reliable subsequent crystallographic analysis.
Monochromated electron energy loss spectroscopy in advanced scanning transmission electron microscopes has become a powerful tool for probing local electronic and vibrational spectra in many materials. However, acquiring high-quality multi-dimensional hyperspectral image datasets remains challenging due to sample drift, low signal-to-noise ratio (SNR), and detector artifacts. These limitations often require shorter acquisition times to reduce their influence, which can in turn compromise the spatial and energy resolution of the final results. To overcome these challenges, we develop a new approach that aligns and integrates multi-frame, fast-scan datasets using optimized image-registration algorithms and selection criteria to achieve high-SNR, low-drift results. By mitigating time-dependent experimental instabilities, this approach effectively preserves the intrinsic energy and spatial resolution of the instrument in the final integrated dataset. Ultimately, this provides an efficient and versatile strategy for capturing high-fidelity multi-dimensional hyperspectral images without compromising the attainable resolution.
Non-destructive and precise thickness measurement of nanometre metal thin films, especially at the sub-10 nm scale where accurate measurement poses significant difficulties, is crucial for micro-nano device manufacturing. However, in thickness measurement for nanometre metal thin films based on the sample current method, the description of the relationship between secondary electron yield (SEY, σ) and film thickness relies on physical approximations of the secondary electron emission process. This leads to the construction of nonlinear piecewise functions, which consequently impairs the model's accuracy. To address this limitation, a physics-informed neural networks (PINN)-based high-precision thickness measurement model for nanometre metal thin films is proposed. The model takes the physical mechanism of secondary electron emission as a constraint and integrates a small amount of experimental SEY data for training. By embedding physical equations into neural networks, the model successfully learns the complex non-linear relationship between SEY and film thickness. In tests on Au, Ag, and Ti films, the PINN model demonstrates superior stability in thickness prediction error compared to the physical approximation model (PAM), maintaining the error within 10%. This method enables real-time thickness measurement during scanning electron microscope (SEM) surface observation, representing a precise and non-destructive nanometre thin film thickness measurement technique.
As oscillating AFM probe approaches the sample, the amplitude and phase shift of the cantilever begin to change at distances of tens of microns away from the surface. Consequently, before the start of semicontact tip-surface interactions, the dynamic characteristics of the cantilever are shifted from their undisturbed values. This change is attributed mainly to the air damping in the gap between the cantilever and the sample. The damping is due to the peculiarities of geometry of sample surface as well as the properties of the probe. The observed behavior is studied by harmonic oscillator model. The results show that damping in the near-surface gap is described by introducing a single additional parameter into the model. The values of this parameter are calculated for various rectangular cantilevers. The numerical simulations and experimental data treatment illustrates a significance of influence of air damping in the gap on the determination of the physical and mechanical constants of the material.
An approach to modeling cathode lenses with arbitrary electrode configurations is developed, in which the electric field on the symmetry axis is calculated using the boundary element method. Criteria for the telescopic mode of idealized and real cathode lenses are determined. A telescopic mode is observed for the first time in a lens with a small-diameter cathode. Algorithms for numerical modeling of cathode lenses are optimized, and a lens with a small-diameter cathode is numerically studied. A design for an electron gun with a thermionic cathode that provides high brightness is proposed.
Femtosecond electron beams serve as effective tools for investigating ultrafast structural and dynamic processes in matter, providing complementary capabilities to femtosecond laser beams. In this paper we propose and demonstrate the feasibility of a scheme combining an undulator with THz modulation to generate isolated few-femtosecond electron bunches. We have developed a theoretical method that incorporates the transport dynamics of low-energy relativistic electrons interacting with the THz modulation field in the undulator and the space charge effects within the bunch itself. The results indicate that, under currently achievable conditions in infrared laser technology and undulator engineering, the proposed scheme can generate single isolated ultrafast electron bunches with kinetic energy 3 MeV, bunch length about 9 fs (rms) with core charge about 8 fC. The normalized slice emittance of the core is about 0.4μm and the rms beam size remains within 1 mm. We have also evaluated the potential influence of several relevant physical quantities on the final bunch length and arrival time. We believe that the proposed scheme and the developed theoretical model presented in this paper can provide useful insights for generating few-femtosecond electron bunches or even shorter attosecond electron bunches in accelerator-based ultrafast electron facilities.
High-resolution imaging and accurate measurement of atomic column positions are essential for achieving an atomistic understanding of crystalline structures. Annular bright-field scanning transmission electron microscopy (ABF-STEM) and integrated differential phase contrast scanning transmission electron microscopy (iDPC-STEM) are powerful techniques for achieving atomic resolution characterization, offering the advantage of direct image interpretation for the positions of atomic columns and enabling simultaneous detection of both heavy and light atomic columns. Nevertheless, the image contrast and determination of atomic column positions in ABF-STEM and iDPC-STEM are susceptible to variations in specimen conditions and experimental conditions, posing significant challenges for accurately determining atomic configuration. Compared to the image contrast deviations observed in ABF-STEM and iDPC-STEM imaging, atomic column displacements are less perceptible, as they typically occur at the picometer scale, presenting a greater challenge to the reliability of ABF-STEM and iDPC-STEM imaging techniques. However, the underlying mechanisms governing atomic column displacements have not yet been thoroughly investigated. In this study, SrTiO3 and LaSrAlO4 are selected as representative examples of highly symmetric and lower symmetric crystalline structures, respectively, to systematically investigate the variations in atomic column displacement observed in ABF-STEM and iDPC-STEM images through a combination of experimental observations and simulation analyses. The oxygen atomic columns in SrTiO3 remain undisplaced. In contrast, the displacement of apical oxygen atomic columns in LaSrAlO4 from their ideal positions is observed to vary with specimen thickness under both ABF-STEM and iDPC-STEM imaging; however, the two techniques yield markedly different observations. The propagation dynamics of the electron probe within SrTiO3 and LaSrAlO4 indicate that crystallographic symmetry plays a dominant role in the formation of deviations in oxygen atomic column positions. This study highlights the necessity of exercising caution when quantifying atomic column displacements from ABF-STEM and iDPC-STEM images, particularly in cases involving lower crystal symmetry and variations in sample thickness.
The use of large-area and multi-detector energy-dispersive X-ray spectrometer (EDXS) systems introduces a number of complex challenges with regard to the accuracy of EDXS quantification. This is due to the increased absorption effects and variable shadowing across the detector surface that these detectors entail. The application of conventional techniques, such as specimen tilting to enhance the take-off angle, proves inadequate for ensuring accurate quantification in these circumstances. In order to address these issues, we developed a Python-based ray tracing simulation tool, edXTrace. This tool models energy-dependent X-ray absorption and shadowing effects with arbitrary detector geometries, sample holder and specimen shapes. Furthermore, a supplementary script for DigitalMicrograph is provided, which integrates the calculated absorption correction into the TEM workflow, thus enhancing the quantitative accuracy of large-area EDXS systems. This approach offers TEM operators a practical and adaptable solution for quantitative EDXS analysis, which is particularly beneficial in the context of complex specimen-detector geometries that are prevalent in modern TEM configurations. Using a NiO test specimen we demonstrate the use of edXTrace with respect to quantitative analysis results.
In this study, the determination of the structure of the tertiary phase which forms following heat treatment in an alloy based on Ti-Zr-Mo-Nb-Ta-Al is reported. Positive identification of the structure was accomplished using convergent beam electron diffraction (CBED) techniques to deduce the point and the space groups. Analyses of the zone axis symmetries showed that the point group of the phase was 6/mmm. Further analyses of dynamic absences resulted in the identification of the space group P6/mcc (no. 192). Structure models were created based on the experimental findings which were utilized to simulate the zone axis CBED patterns using the JEMS software package. The simulated patterns showed good agreement with the experimental findings.
Modern scanning transmission electron microscopes (STEMs) can generate large image datasets of semiconductor samples. The recorded position-averaged convergent beam electron diffraction (PACBED) images are useful for measuring parameters such as sample thickness, rotation, or mistilt. Typically, these measurements are performed manually, which makes them time-consuming and often inaccurate. This is why machine learning methods are an attractive option for analyzing large datasets of PACBED images quickly and consistently. Previous approaches focus on machine learning models that analyze each sample parameter individually. This paper investigates multitask deep neural networks that simultaneously predict multiple sample parameters, outperforming models trained on a single parameter. Our deep learning models are trained using a combination of synthetic images of III-V semiconductor materials and silicon, simulated by a multi-slice algorithm and small amounts of experimental data. In addition, we explore how the amount of experimental training data impacts the practical performance of various neural network architectures. On our experimental test datasets, the best deep neural network achieves a mean absolute error of 4.19° for predicting sample rotation α, 0.43milliradians for predicting sample mistilt β, and 2.93nanometers for predicting sample thickness t. These values are within the range of experimental measurement uncertainty. For material classification m, the best models perform without errors on the trained materials.
High temporal resolution dynamic transmission electron microscopy (DTEM) requires substantially higher peak currents than conventional TEM sources can sustain without severe brightness loss, motivating a redesign of the electron gun. This work targets improving the brightness in a 200 keV electrostatic electron gun for DTEM by mitigating aberrations through optimizing conductor geometry. The cathode-anode assembly is reconfigured to act simultaneously as an accelerator and condenser pre-lens, confining the beam size in regions of strong field curvature. A third-order off-axis transfer map with Green's-function evaluation of image errors, validated against Superfish and GPT, quantifies how space-charge-driven divergence growth couples into spherical aberration and guides voltage/geometry choices. Along a minimum-spot operating contour, the spherical aberration coefficient is reduced from >150mm to ∼5-10 mm at 200 keV while preserving a ≤10μm exit-aperture spot for MTE = 0.1-1 eV and ∼1 mA peak current. The results establish geometry-driven gun redesign as a practical route to higher-brightness, low-aberration electron sources for next-generation time-resolved microscopy.
Transmission electron microscopy (TEM) is a fundamental tool for many research fields like catalysis or material research. Many experiments require measurements under environmental TEM (ETEM) conditions, in which an atmosphere of a selected gas is in the vicinity of the sample, resulting in a higher pressure in the octagon of the microscope. However, the comparably high pressure in the octagon also influences how long volatile organic components (VOCs), which are often part of the sample itself, remain in the sample environment and can result in carbon contamination and undesired reactions during (E)TEM experiments. We have analyzed contamination in an ETEM octagon by investigating the impact of different organic solvents used in the sample synthesis and preparation, but also how tedious the removal of VOCs is once these contaminants are present inside a microscope. With proton-transfer-reaction mass spectrometry, we have quantified the outgassing contamination of samples consisting of THF, toluene or mesitylene depending on their drying time. We found that high amounts of contaminants adsorb on the inner surfaces of the octagon and we measured how well these VOCs desorb when, after removing the sample, pumping the octagon to typical TEM pressures. For persistent solvents or samples that were not sufficiently dried before insertion, contaminant removal is an extensive process that can be accelerated by plasma cleaning. Our investigation of residual solvents suggests that changes in the sample preparation (i.e. solvent choice and drying time ) can improve the contamination effects in both TEM and ETEM experiments.
Advances in transmission electron microscopy are fundamentally driven by improvements in electron detection technology. Timepix4 is a novel hybrid pixelated detector that can operate in both Frame-based and Data-driven modes. The increased number of pixels, higher clock frequency, and faster frame rates with respect to its predecessors make this detector an ideal candidate for multi-purpose use in TEM, particularly for 4D STEM applications. This work provides a comprehensive experimental characterization of the Timepix4 detector for 60 and 200 keV electrons. Key characteristics investigated include cluster size, detector quantum efficiency, linearity, readout limits, and temporal response. The detector demonstrates reasonable spatial fidelity while offering flexible trade-offs between high per-pixel flux tolerance and per-event timing/zero-suppressed streaming. This detailed study aids in understanding the detector's performance in both operation modes and guides researchers in selecting optimal settings to maximize the detector's performance. Finally, two application results are presented: one utilizing Data-driven mode to acquire atomic resolution images, and the other using Frame-based mode to perform orientation maps on a polycrystalline specimen.
Off-axis electron holography is used to measure the temperature dependence of the mean inner potential and electron-beam-induced charging effects in polystyrene spheres with diameters of between 230 and 600 nm from room temperature to 5.3 K. Integrating the model independent and line fitting simulation approaches, enabled by the high quality of the recorded holograms, it has been possible to disentangle the contribution the beam induced charge from that of the mean inner potential (MIP), finding that both quantities (at least for the smaller spheres) increase with decreasing temperature. In particular the MIP reaches a value of (6.48 ± 0.24) V at 5.3 K, corresponding to an increase of 16.8% ±4.2%, with respect to the measurements carried at room temperature, giving the value (5.55 ± 0.07) V.
Scanning tunneling microscopy (STM) has become a cornerstone of atomic-scale research, as modern experiments demand not only high spatial resolution but also flexible instrumentation capable of supporting electrical and optical measurements. Custom-built STMs provide distinct advantages in specialized environments, including cryogen-free superconducting magnets, water-cooled magnet systems, millikelvin cryostats, and platforms integrating optical or terahertz excitation. These systems enable experiment-specific optimization of mechanical design, vibration isolation, thermal coupling, and low-noise signal pathways, routinely achieving sub-picometer stability and supporting advanced measurement modalities. Recent advances, including terahertz-driven STM, electron spin resonance combined STM, and spectroscopic measurements under extreme conditions, highlight the unique scientific opportunities enabled by custom-designed STM platforms. This review outlines the engineering principles underpinning high-performance custom-built STMs and highlights key discoveries enabled by their versatility. We further address ongoing challenges, including drift suppression, electromagnetic interference, and reliable tip preparation. Finally, emerging directions toward next-generation STM platforms promise enhanced stability, ultrafast temporal resolution, and expanded access to atomic-scale quantum phenomena.
Contact-resonance atomic force microscopy (CR-AFM) has emerged as a leading technique for nanoscale subsurface detection owing to its noninvasive capability and high sensitivity. However, most applications have been limited to specimens with nearly flat surfaces, since large local inclinations introduce artifacts in the CR-frequency maps. Consequently, CR-based subsurface imaging remains severely constrained by topography-induced artifacts, which hinders its application to structured surfaces such as metasurfaces and microlens arrays. Here, we present a quantitative subsurface reconstruction strategy based on an analytical model that accounts for optical-lever sensitivity, applied force, and normal and lateral contact stiffness under varying local inclinations. This approach reduces topography-induced artifacts and enables high-resolution modulus and subsurface mapping on textured or inclined surfaces. The reconstructed modulus maps cover materials from approximately 5 to 163 GPa, with a propagated Type B standard uncertainty below 4% for the tested specimens. Moreover, it reconstructs subsurface features in specimens with topographically complex, textured surfaces-such as graphene-oxide (GO) membrane covered and single-point diamond turning (SPDT) machined silicon-revealing previously obscured buried structures. These results support the use of the proposed framework for artifact-reduced modulus and subsurface reconstruction on the tested topographically complex specimens, and indicate its potential for subsurface defect detection in structured micro/nano-scale surfaces.
This study introduces a two-stage (Stage 1-Stage 2), knowledge-guided, and drift-aware computer vision (CV) pipeline that redefines the traditionally manual and observer-dependent analysis workflow of AFM measurements. The proposed approach establishes, for the first time in the literature, a nanoscale point-tracking method capable of automatically, robustly, and reliably following the same surface locations across consecutive loading conditions. Stage 1 ensures the physical consistency of raw AFM data, extracts the multilayer morphology of the surface, and generates reliable point correspondences between consecutive load steps using SIFT-ORB feature extraction combined with RANSAC-based geometric verification. Stage 2 converts these correspondences into a drift-aware chained tracking structure that preserves the identity of each point and enables pixel-level computation of tribological responses, including local friction forces and load-friction curves. The method is validated on a representative two-dimensional material system, demonstrating its capability to both characterize lateral drift emerging in consecutive scans and reliably disentangle the layer-dependent tribological responses of the surface. By processing AFM data through a holistic sensing-to-inference information flow, the pipeline introduces the first drift-aware and knowledge-guided CV-based nanoscale characterization approach that produces machine-readable, reproducible, and statistically consistent tribological outputs.
Scanning electron microscopy faces a fundamental resolution-aberration trade-off, which is exacerbated in scanning ultrafast electron microscopy by the long working distances required for pump-probe integration and low electron flux. We report an optically modulated free-electron computational ghost imaging framework to overcome these constraints. By employing laser-induced ponderomotive modulation to generate structured electron illumination, we record the total secondary-electron yield via a bucket detector. Critically, spherical aberration is integrated into the forward model for wavefront engineering, rendering the system intrinsically aberration-resilient. Image reconstruction is performed using a modified stochastic gradient descent algorithm. Numerical simulations demonstrate high-fidelity surface characterization under long-working-distance and low-flux conditions, where conventional SEM performance degrades. This paradigm provides a robust pathway for high-resolution imaging in aberration-dominated and signal-starved electron microscopy platforms.
