Scanning transmission electron microscopy (STEM) enables us to directly visualize atomic structures inside materials. However, atomic-scale observation of electron beam-sensitive materials has been extremely challenging because of their low resistance to electron irradiation. We have developed an ultra-efficient STEM imaging method, optimum bright-field (OBF) STEM, utilizing segmented STEM detectors. The OBF technique has much higher imaging efficiency than conventional STEM methods and can be applied to low-dose imaging of beam-sensitive materials with minimal irradiation damages. This paper introduces the imaging theory of OBF STEM and its application to various samples including lithium-ion battery materials.

In recent years, with the spread of aberration correction devices, observation and analysis with atomic resolution have been popular. However, it leads drastic increase in electron beam current density in parallel. As a result, previously inconspicuous electron beam damage is more likely to become more noticeable. On the other hand, the demand for observation and analysis of samples that are sensitive to an electron beam is increasing, and it is essential to reduce the dose for their observation and analysis. However, lowering the dose lowers the resulting signal-to-noise ratio. Therefore, there is a request for a method of suppressing damage to the electron beam without reducing the dose. In this paper, we proposed a new observation and/or analysis method that suppresses sample damage caused by electron beams, which is an intermittent irradiation method, using a newly developed high-speed electrostatic shutter. As a result, it was found that the damage caused by electrons was significantly reduced.

Single-molecule atomic-resolution time-resolved electron microcopy (SMART-EM) with a high-speed, high-resolution transmission electron microscope has ushered in the era of “cinematic molecular science,” in which the dynamic behavior of organic and inorganic molecules can be studied through real-time video images. High speed imaging without increase of electron dose results in very low electron dose per frame, which often causes a very low signal to noise ratio. We have established a method to improve the image quality optimized for SMART-EM by applying the Chambolle Total Variation denoising, which enables tracking of molecular motion with sub-millisecond temporal resolution and sub-angstrom localization precision. Herein we demonstrate in-situ observation of a nanomechanical shuttling motion of a single molecule and an atomic-level crystallization process as applications of the high-speed SMART-EM. We also introduce a new molecular model, atomic number correlated (ZC) molecular model, designed for spreading the microscopic molecular imaging to chemists who are still unfamiliar to electron microscopy.

With the technological development on automatic and high-precision microscopic measurements, demand for informatics technologies to handle a variety of large-scale data have been increasing. In order to further reduce the measurement cost and to prevent damage to the measurement specimen, informatics techniques are needed to accurately analyze the measured data with a short exposure time of the quantum beam. In this review article, we introduce machine learning methods for low-count spectrum image datasets. We first introduce a non-negative matrix factorization method, which is effective for spectrum image data analysis, and then introduce an extension of NMF for the Poisson noise model. We also introduce a preprocessing method of Poisson scaling dealing with Poisson noise. Using numerical experiments with synthetically generated data and real SEM-EDS data, we will demonstrate and discuss advantages and disadvantages of these methods.

Various methods for magnetic domain characterization have been proposed to understand and improve magnetic properties of magnets. SEM based techniques utilizing either secondary and backscattered electrons have been proposed, but they are not commonly used because they often lack spatial resolution and require modification of the instrument. Latest SEMs have enhanced low-accelerating-voltage performance and multiple imaging detectors including in-lens type detectors. Thus, they are capable of giving rich material information provided the signal acceptance is optimized by choosing the proper voltage and working distance. Here we introduce how magnetic domains of the Nd-Fe-B magnet can be visualized using latest SEM without modification of the detection system. Selective collection of low-energy secondary electrons by controlling acceleration voltage and working distance will allow observation of magnetic domains using both in-lens and Everhart-Thornley secondary electron detectors. This technique will give additional information on magnetic domains on top of compositional, elemental and crystal orientation information obtained by backscattered electron, energy dispersive X-ray and electron backscatter diffraction detectors in the SEM.

Optical microscopes are powerful tools to observe samples of micrometer order. Because of the definition of the resolution limit of the optical microscope, many researchers believed that optical microscopes never visualize nanometer structures. However, Super resolution microscopes are invented and commercialized in recent years. In this review, we focus on and the super resolution microscopes. Mechanism and resolution limit of commercially available super resolution microscopes are summarized.

The science of metallic subnanoparticles has significantly progressed over the past decade, reaching a level where we can strictly define the number of atoms and elemental ratios contained in the particles. This progress is primarily due to technological advances in synthetic chemistry, but improvements in observation techniques using electron microscopes have also contributed significantly. In this account article, the authors introduce the frontiers of metal subnanoparticle synthesis and the contributions of electron microscopy.

R3 Institute for Newly-Emerging Science Design at Osaka University has developed an online educational system for scanning/transmission electron microscopy (S/TEM). This system integrates the video monitor of an operation system of S/TEM, video movies of the hands of an instructor who is controlling both the buttons and encoders on the consoles and the pointing devices, and video movie of the fluorescence screen into an application for a web conference. This simultaneously and interactively remote-delivered system makes students be able to share online the actual operations for obtaining TEM images, electron diffraction patterns, characteristic x-ray spectra etc. with the instructor as if they were in a laboratory room for TEM. This educational system for TEM may be extended to various applications in research and to other kinds of scientific instruments.

The so-called iron-oxidizing bacteria have been recognized for their potential to form iron oxide (hydroxide) structures in natural aquatic environments. This paper describes the crystallinity, constituent elements, ultrastructure of these iron hydroxide structures, especially uniquely twisted extracellular stalks produced from Gallionella ferruginea. X-ray diffractometry, transmission electron microscopy-selected-area electron diffraction, and high-resolution transmission electron microscopy showed that the stalk fibers had an amorphous structure. Scanning electron microscopy and scanning transmission electron microscopy with energy dispersive X-ray spectroscopy revealed that the stalk fibers contain oxygen, iron, silicon and phosphorous, and elemental maps showed that major detectable elements, O, Fe, Si and P are uniformly located in the stalk fibers. Electron energy-loss spectroscopy revealed that the stalk fibers had a central carbon core of bacterial exopolymers and that aquatic iron interacted with oxygen at the surface of the carbon core, resulting in deposition of iron oxides at the surface. In addition we attempted to reconstruct three-dimensional images obtained from serial section scanning electron microscopy (array tomography) of the cell producing the fibers.