Atomic-resolution electromagnetic field observation is critical to the development of advanced materials and to the unveiling of their fundamental physics. For this purpose, a spherical-aberration corrected 1.2-MV cold field-emission transmission electron microscope has been developed. The microscope has the following superior properties: stabilized accelerating voltage, minimized electrical and mechanical fluctuation, and coherent electron emission. These properties have enabled to obtain 43-pm information transfer. On the bases of these performances, a 43-pm resolution has been obtained by correcting lens aberrations up to the third order. Observations of GaN [411] thin crystal showed a projected atomic locations with a separation of 44 pm.

Triple-biprism electron interferometry has been developed to flexibly control all interference parameters, i.e., interference region (size and shape), fringe spacing, and fringe angle. Each filament electrode of the upper and middle electron biprisms is installed in the individual image planes of the objective and first magnifying lenses, respectively. One more filament electrode of the lower biprism is set between the crossover plane and image plane behind the second magnifying lens. The azimuth angle between the upper and middle filaments is set at 90°, and lower filament is set at about 45° for both of the above filaments. Among many possible optical setups, we chose the setup with the easiest and most flexible control of all the interference parameters through theoretical investigation and experimental confirmation. The interferometer was constructed in a $1\text{×}{10}^6\mathrm{V}$ field emission electron microscope, and the overall performance was demonstrated and investigated experimentally.

To enable better statistical analysis of catalyst nanoparticles by high-resolution electron holography, we improved the particle detection accuracy of our previously developed automated hologram acquisition system by using an image classifier trained with machine learning. The detection accuracy of 83% was achieved with the small training data of just 232 images showing nanoparticles by utilizing transfer learning based on VGG16 to train the image classifier. Although the construction of training data generally requires much effort, the time needed to select the training data candidates was significantly shortened by utilizing a pattern matching technique. Experimental results showed that the high-resolution hologram acquisition efficiency was improved by factors of about 100 and 6 compared to a scan method and a pattern-matching-only method, respectively.

Mapping electrostatic potential around a Pt nanoparticle supported on TiO2 (110) - Volume 27 Issue S1