High-throughput continuous rotation electron diffraction data acquisition via software automation.

Area of Science:

• Materials Science
• Crystallography
• Electron Microscopy

Background:

• Single-crystal electron diffraction (SCED) is a powerful technique for analyzing nano-sized crystals.
• Current methods can be limited by crystal movement and require specialized user expertise.

Purpose of the Study:

• To implement and validate the continuous rotation electron diffraction (cRED) method for high-throughput nano-crystal structure determination.
• To enhance the accessibility and reproducibility of SCED data collection.

Main Methods:

• Development of dedicated software for automated transmission electron microscope and camera control.
• Implementation of crystal tracking via intermittent defocusing during continuous rotation.
• Integration of serial electron diffraction with deep convolutional neural network for automated crystal screening.
• Demonstration using zeolite mordenite samples.

Main Results:

• cRED data collection was transformed into a high-throughput method with increased success rates and larger rotation ranges.
• Automated logging of experimental parameters and file generation reduced human error.
• Deep convolutional neural network effectively screened for suitable crystals.
• Refined crystal structure confirmed the quality of cRED data.

Conclusions:

• The implemented cRED method significantly improves the efficiency and accessibility of SCED for nano-crystal structure analysis.
• Automated screening and data collection protocols enhance reproducibility and broaden user accessibility.
• The combination of cRED and AI-driven screening represents a significant advancement in electron diffraction techniques.