Tracking the picoscale spatial motion of atomic columns during dynamic structural change.

Area of Science:

• Materials Science
• Nanotechnology
• Microscopy

Background:

• Characterizing dynamic atomic structural changes in materials is crucial for understanding functionality, especially in catalytic nanoparticles.
• Advancements in direct electron detection enable high-speed imaging (1000 frames/sec) in bright-field transmission electron microscopy (BF TEM), offering millisecond temporal resolution for atomic structure analysis.

Purpose of the Study:

• To develop and validate computational methods for analyzing large, noisy datasets from high-speed TEM image series.
• To present a robust two-dimensional Gaussian fitting algorithm for tracking atomic columns in temporally resolved BF TEM data.

Main Methods:

• Developed a two-dimensional Gaussian fitting algorithm to track atomic column positions and intensities.
• Applied the algorithm to experimental BF TEM image series of ceria (CeO2) nanoparticle surfaces.
• Compared the algorithm's accuracy for position and intensity measurements against traditional centroid fitting and evaluated performance as a function of electron beam dose.

Main Results:

• The Gaussian fitting algorithm accurately tracks atomic column positions and intensities in high-speed TEM image series.
• Accuracy of position localization was compared to centroid fitting.
• Intensity measurement accuracy was evaluated across a range of electron beam doses per frame (125-5000 e-/Ų).

Conclusions:

• The developed Gaussian fitting algorithm provides a reliable method for analyzing dynamic atomic structural changes in materials using high-speed TEM.
• The methodology and code are broadly applicable to other temporally resolved TEM datasets for tracking atomic column dynamics.
• This work facilitates a deeper understanding of material functionality through precise characterization of atomic-level dynamics.