Search

Instrumentation developments in electron energy-loss spectroscopy (EELS) in the scanning transmission electron microscope (STEM) one decade ago paved the way for combining milli-electronvolt energy resolution in spectroscopy with \r{A}ngstr\"om-sized electron probes, unlocking unexplored realms for solid state physics at the nanometer-scale. The fundamental understanding of the scattering processes involved has made it possible to separate atomically localized signals, providing insights into vibrations at the atomic scale. In this Letter, we outline the experimental, conceptual and theoretical advances in this field and also make comparisons with tip-based optical techniques before discussing future perspectives of this exciting technique. Optimization of dark-field signal collection will play a fundamental role in making this technique more widely applicable to a range of materials problems. The benefits of adding momentum-resolution will also be discussed and a powerful acquisition scheme proposed., Comment: 16 pages, 3 figures, invited perspective article

The vast majority of materials science knowledge exists in unstructured natural language, yet structured data is crucial for innovative and systematic materials design. Traditionally, the field has relied on manual curation and partial automation for data extraction for specific use cases. The advent of large language models (LLMs) represents a significant shift, potentially enabling efficient extraction of structured, actionable data from unstructured text by non-experts. While applying LLMs to materials science data extraction presents unique challenges, domain knowledge offers opportunities to guide and validate LLM outputs. This review provides a comprehensive overview of LLM-based structured data extraction in materials science, synthesizing current knowledge and outlining future directions. We address the lack of standardized guidelines and present frameworks for leveraging the synergy between LLMs and materials science expertise. This work serves as a foundational resource for researchers aiming to harness LLMs for data-driven materials research. The insights presented here could significantly enhance how researchers across disciplines access and utilize scientific information, potentially accelerating the development of novel materials for critical societal needs.

A detailed analysis of ptychography for 3D phase reconstructions of thick specimens is performed. We introduce multi-focus ptychography, which incorporates a 4D-STEM defocus series to enhance the quality of 3D reconstructions along the beam direction through a higher overdetermination ratio. This method is compared with established multi-slice ptychography techniques, such as conventional ptychography, regularized ptychography, and multi-mode ptychography. Additionally, we contrast multi-focus ptychography with an alternative method that uses virtual optical sectioning through a reconstructed scattering matrix ($\mathcal{S}$-matrix), which offers more precise 3D structure information compared to conventional ptychography. Our findings from multiple 3D reconstructions based on simulated and experimental data demonstrate that multi-focus ptychography surpasses other techniques, particularly in accurately reconstructing the surfaces and interface regions of thick specimens.

While capable of imaging the atoms constituting thin slabs of material, the achievable resolution of conventional electron imaging techniques in a transmission electron microscope (TEM) is very sensitive to the partial spatial coherence of the electron source, lens aberrations and mechanical instabilities of the microscope. The desire to break free from the limitations of the apparatus spurred the popularity of ptychography, a computational phase retrieval technique that, to some extent, can compensate for the imperfections of the equipment. Recently it was shown that ptychography is capable of resolving specimen features as fine as the blurring due to the vibrations of atoms, a limit defined not by the microscope, but by the investigated sample itself. Here we report on the successful application of a mixed-object formalism in the ptychographic reconstruction that enables the resolution of fluctuations in atomic positions within real space. We show a reconstruction of a symmetric {\Sigma}9 grain boundary in silicon from realistically (molecular dynamics) simulated data. By reconstructing the object as an ensemble of 10 different states we were able to observe movements of atoms in the range of 0.1-0.2 \AA in agreement with the expectation. This is a significant step forward in the field of electron ptychography, as it enables the study of dynamic systems with unprecedented precision and overcomes the resolution limit so far considered to be imposed by the thermal motion of the atoms.

Phonon scattering at grain boundaries (GBs) is significant in controlling nanoscale device thermal conductivity. However, GBs could also act as waveguides for selected modes. To measure localized GB phonon modes, meV energy resolution is needed with sub-nm spatial resolution. Using monochromated electron energy loss spectroscopy (EELS) in the scanning transmission electron microscope (STEM) we have mapped the 60 meV optic mode across GBs in silicon at atomic resolution and compared it to calculated phonon densities of states (DOS). The intensity is strongly reduced at GBs characterised by the presence of five- and seven-fold rings where bond angles differ from the bulk. The excellent agreement between theory and experiment strongly supports the existence of localized phonon modes and thus of GBs acting as waveguides., Comment: Merged article (8 pages, 3 figures, 1 table) and Supplementary Information (8 pages, 8 figures)

Ptychography is a computational imaging technique that has risen in popularity in the x-ray and electron microscopy communities in the past half decade. One of the reasons for this success is the development of new high performance electron detectors with increased dynamic range and readout speed, both of which are necessary for a successful application of this technique. Despite the advances made in computing power, processing the recorded data remains a challenging task, and the growth in data rate has made the size of the resulting datasets a bottleneck for the whole process. Here we present an investigation into lossy compression methods for electron diffraction patterns that retain the necessary information for ptychographic reconstructions, yet lead to a decrease in data set size by three or four orders of magnitude. We apply several compression methods to both simulated and experimental data - all with promising results.

The expansive production of data in materials science, their widespread sharing and repurposing requires educated support and stewardship. In order to ensure that this need helps rather than hinders scientific work, the implementation of the FAIR-data principles (Findable, Accessible, Interoperable, and Reusable) must not be too narrow. Besides, the wider materials-science community ought to agree on the strategies to tackle the challenges that are specific to its data, both from computations and experiments. In this paper, we present the result of the discussions held at the workshop on "Shared Metadata and Data Formats for Big-Data Driven Materials Science". We start from an operative definition of metadata, and what features a FAIR-compliant metadata schema should have. We will mainly focus on computational materials-science data and propose a constructive approach for the FAIRification of the (meta)data related to ground-state and excited-states calculations, potential-energy sampling, and generalized workflows. Finally, challenges with the FAIRification of experimental (meta)data and materials-science ontologies are presented together with an outlook of how to meet them.

The prosperity and lifestyle of our society are very much governed by achievements in condensed matter physics, chemistry and materials science, because new products for sectors such as energy, the environment, health, mobility and information technology (IT) rely largely on improved or even new materials. Examples include solid-state lighting, touchscreens, batteries, implants, drug delivery and many more. The enormous amount of research data produced every day in these fields represents a gold mine of the twenty-first century. This gold mine is, however, of little value if these data are not comprehensively characterized and made available. How can we refine this feedstock; that is, turn data into knowledge and value? For this, a FAIR (findable, accessible, interoperable and reusable) data infrastructure is a must. Only then can data be readily shared and explored using data analytics and artificial intelligence (AI) methods. Making data 'findable and AI ready' (a forward-looking interpretation of the acronym) will change the way in which science is carried out today. In this Perspective, we discuss how we can prepare to make this happen for the field of materials science.

We present a method that lowers the dose required for a ptychographic reconstruction by adaptively scanning the specimen, thereby providing the required spatial information redundancy in the regions of highest importance. The proposed method is built upon a deep learning model that is trained by reinforcement learning (RL), using prior knowledge of the specimen structure from training data sets. We show that equivalent low-dose experiments using adaptive scanning outperform conventional ptychography experiments in terms of reconstruction resolution., Comment: 12 pages, 8 figures

4D-STEM, in which the 2D diffraction plane is captured for each 2D scan position in the scanning transmission electron microscope (STEM) using a pixelated detector, is complementing and increasingly replacing existing imaging approaches. However, at present the speed of those detectors, although having drastically improved in the recent years, is still 100 to 1,000 times slower than the current PMT technology operators are used to. Regrettably, this means environmental scanning-distortion often limits the overall performance of the recorded 4D data. Here we present an extension of existing STEM distortion correction techniques for the treatment of 4D-data series. Although applicable to 4D-data in general, we use electron ptychography and electric-field mapping as model cases and demonstrate an improvement in spatial-fidelity, signal-to-noise ratio (SNR), phase-precision and spatial-resolution.

Increasing interest in three-dimensional nanostructures adds impetus to electron microscopy techniques capable of imaging at or below the nanoscale in three dimensions. We present a reconstruction algorithm that takes as input a focal series of four-dimensional scanning transmission electron microscopy (4D-STEM) data. We apply the approach to a lead iridate, Pb$_2$Ir$_2$O$_7$, and yttrium-stabilized zirconia,Y$_{0.095}$Zr$_{0.905}$O$_2$ , heterostructure from data acquired with the specimen in a single plan-view orientation, with the epitaxial layers stacked along the beam direction. We demonstrate that Pb-Ir atomic columns are visible in the uppermost layers of the reconstructed volume. We compare this approach to the alternative techniques of depth sectioning using differential phase contrast scanning transmission electron microscopy (DPC-STEM) and multislice ptychographic reconstruction., Comment: 27 pages, 6 figures in main manuscript. 6 pages, 5 figures in supplementary. All computer code for electron microscopy simulation and reconstruction included in directory anc. Submitted to Microscopy & Microanalysis

The overdetermination of the mathematical problem underlying ptychography is reduced by a host of experimentally more desirable settings. Furthermore, reconstruction of the sample-induced phase shift is typically limited by uncertainty in the experimental parameters and finite sample thicknesses. Presented is a conjugate gradient descent algorithm, regularized optimization for ptychography (ROP), that recovers the partially known experimental parameters along with the phase shift, improves resolution by incorporating the multislice formalism to treat finite sample thicknesses, and includes regularization in the optimization process, thus achieving reliable results from noisy data with severely reduced and underdetermined information., Comment: 18 pages, 7 figures, 3 tables

Optical and electrical properties of two-dimensional transition metal dichalcogenides (TMDCs) grown by chemical vapor deposition (CVD) are strongly determined by their microstructure. Consequently, the visualization of spatial structural variations is of paramount importance for future applications. Here we demonstrate how grain boundaries, crystal orientation, and strain fields can unambiguously be identified with combined lateral force microscopy (LFM) and transverse shear microscopy (TSM) for CVD-grown tungsten disulfide (WS2) monolayers, on length scales that are relevant for optoelectronic applications. Further, angle-dependent TSM measurements enable us to acquire the fourth-order elastic constants of monolayer WS2 experimentally. Our results facilitate high-throughput and nondestructive microstructure visualization of monolayer TMDCs, insights into their elastic properties, thus providing an accessible tool to support the development of advanced optoelectronic devices based on such two-dimensional semiconductors.

We apply inline electron holography to investigate the electrostatic potential across an individual BaZr0.9Y0.1O3 grain boundary. With holography, we measure a grain boundary potential of -1.3 V. Electron energy loss spectroscopy analyses indicate that barium vacancies at the grain boundary are the main contributors to the potential well in this sample. Furthermore, geometric phase analysis and density functional theory calculations suggest that reduced atomic density at the grain boundary also contributes to the experimentally measured potential well.

Low-loss electron energy loss spectroscopy (EELS) in the scanning transmission electron microscope (STEM) probes the valence electron density and relevant optoelectronic properties such as band gap energies and other band structure transitions. The measured spectra can be formulated in a dielectric theory framework, comparable to optical spectroscopies and ab-initio simulations. Moreover, Kramers-Kronig analysis (KKA), an inverse algorithm based on the homonym relations, can be employed for the retrieval of the complex dielectric function (CDF). However, spurious contributions traditionally not considered in this framework typically impact low-loss EELS modifying the spectral shapes and precluding the correct measurement and retrieval of the dielectric information. A relativistic KKA algorithm is able to account for the bulk and surface radiative-loss contributions to low-loss EELS, revealing the correct dielectric properties. Using a synthetic low-loss EELS model, we propose some modifications on the naive implementation of this algorithm that broadens its range of application. The robustness of the algorithm is improved by regularization, appliying previous knowledge about the shape and smoothness of the correction term. Additionally, our efficient numerical integration methodology allows processing hyperspectral datasets in a reasonable amount of time. Harnessing these abilities, we show how simultaneous relativistic KKA processing of several spectra can share information to produce an improved result., Comment: 11 pages, 7 figures

Under the constraint of constant illumination, an information criterion is formulated for the Fisher information that compressed sensing measurements in optical and transmission electron microscopy contain about the underlying parameters. Since this approach requires prior knowledge of the signal's support in the sparse basis, we develop a heuristic quantity, the detective quantum efficiency (DQE), that tracks this information criterion well without this knowledge. It is shown that for the investigated choice of sensing matrices, and in the absence of read-out noise, i.e. with only Poisson noise present, compressed sensing does not raise the amount of Fisher information in the recordings above that of Shannon sampling. Furthermore, enabled by the DQE's analytical tractability, the experimental designs are optimized by finding out the optimal fraction of on-pixels as a function of dose and read-out noise. Finally, we introduce a regularization and demonstrate, through simulations and experiment, that it yields reconstructions attaining minimum mean squared error at experimental settings predicted by the DQE as optimal., Comment: 18 pages, 13 figures. New Monte Carlo simulations in Figure 13 showing the behavior of the single-pixel camera under various magnitudes of read-out noise

In this paper, we present a systematic investigation of symmetry-breaking in the plasmonic modes of triangular gold nanoprisms. Their geometrical C3 symmetry is one of the simplest possible that allows degeneracy in the particle's mode spectrum. It is reduced to the non-degenerate symmetries Cv or E by positioning additional, smaller gold nanoprisms in close proximity, either in a lateral or a vertical configuration. Corresponding to the lower symmetry of the system, its eigenmodes also feature lower symmetries (Cv), or preserve only the identity (E) as symmetry. We discuss how breaking the symmetry of the plasmonic system not only breaks the degeneracy of some lower order modes, but also how it alters the damping and eigenenergies of the observed Fano-type resonances.

Three-dimensional charge density maps computed by first-principles methods provide information about atom positions and the bonds between them, data which is particularly valuable when trying to understand the properties of point defects, dislocations, and interfaces. This letter presents a method by which 3-dimensional maps of the electrostatic potential, related to the charge density map by the Poisson equation, can be obtained experimentally at 0.1 nm resolution or better. This method requires data acquired by holographic transmission electron microscopy (TEM) methods such as off-axis electron holography or focal series reconstruction for different directions of the incident electron beam. The reconstruction of the 3-dimensional electrostatic (and absorptive) potential is achieved by making use of changes in the dynamical scattering within the sample as the direction of the incident beam varies., Comment: 4 pages, 3 figues

Solving crystal structures from kinematical X-ray or electron diffraction patterns of single crystals requires many more diffracted beams to be recorded than there are atoms in the structure, since the phases of the structure factors can only be retrieved from such data if the atoms can be resolved as sharply peaked objects. Here a method is presented by which the fact that multiple scattering encodes structure factor phases in the diffracted intensities is being used for solving the crystallographic phase problem. The retrieval of both amplitudes and phases of electron structure factors from diffraction patterns recorded with varying angle of incidence will be demonstrated. No assumption about the scattering potential itself is being made. In particular, the resolution in the diffraction data does not need to be sufficient to resolve atoms, making this method potentially interesting for electron crystallography of 2-dimensional protein crystals and other beam-sensitive complex structures., Comment: John Spence Memorial Issue