Probing Electron Beam Induced Transformations on a Single-Defect Level via Automated Scanning Transmission Electron...

Area of Science:

• Materials Science
• Condensed Matter Physics
• Electron Microscopy

Background:

• Real-time analysis of scanning transmission electron microscopy (STEM) data is crucial for understanding dynamic atomic processes.
• Current methods often lack the speed and precision required for complex atomic manipulation and defect engineering.

Purpose of the Study:

• To develop and implement a robust, real-time analysis approach for STEM data streams.
• To enable automated experiments for exploring atomic dynamics under electron beam irradiation.
• To facilitate precise atomic manipulation and defect creation in advanced materials.

Main Methods:

• Ensemble learning and iterative training (ELIT) of deep convolutional neural networks for real-time data analysis.
• Integration of ELIT with an operational STEM microscope and beam control.
• Automated experimental procedures for targeted atomic manipulation and defect identification.

Main Results:

• Demonstrated atomically precise engineering of single vacancy lines in transition metal dichalcogenides.
• Successfully created and identified topological defects in graphene.
• Achieved on-the-fly analysis of STEM data with real-time feedback capabilities.

Conclusions:

• The ELIT-based approach provides a powerful tool for real-time STEM data analysis and automated experimentation.
• This method opens new avenues for probing electron beam chemistry, atomic manipulation, and atom-by-atom assembly.
• Enables advanced materials engineering and the study of fundamental material properties at the atomic scale.