8th SmoleQ Greeting Seminar (2025.05.26 at 17:00 (JST))

 

Speaker: 

Fumiaki Ichihashi

Title: 

Implementing Machine Learning to Extend Electron Holography Measurements.

Abstract:

Elucidation of the reaction mechanism of heterogeneous catalysts is essential to realize highly efficient artificial photosynthesis, hydrogen production, and carbon-neutral fuel cells. The potential distribution around catalytic nanoparticles may contribute to the reaction mechanism of the catalyst, and a nanoscale electric field analysis technique has been needed. Hitachi possesses an electron holography microscope with an atomic resolution of 1.2 MV that can measure nanoscale electromagnetic fields and has measured the electric potential distribution around catalytic nanoparticles. In this presentation, we will report the results of introducing machine learning into automated electron holography measurements to improve data collection efficiency by about 100 times compared to conventional methods, as well as the results of using machine learning to reduce noise superimposed by environmental cells used in catalyst measurements in a gas environment.

A part of this work was supported by JST CREST Grant Number JPMJCR 1664, Japan. A part of this work was supported by Innovative Science and Technology Initiative for Security Grant Number JPJ004596, ATLA, Japan.