Optimizing Structure Factors From Four-dimensional Scanning Transmission Electron Microscopy via the Scattering...

Area of Science:

• Materials Science
• Condensed Matter Physics
• Electron Microscopy

Background:

• Quantitative potential retrieval in scanning transmission electron microscopy (STEM) is crucial for material characterization.
• Thicker samples present challenges due to significant multiple scattering effects.

Purpose of the Study:

• To develop a novel gradient-descent-based optimization algorithm for quantitative potential retrieval from thick samples in 4D STEM.
• To overcome limitations imposed by multiple scattering in electron microscopy.

Main Methods:

• Utilized a scattering matrix (S-matrix) to model multiple scattering effects.
• Employed a gradient-descent optimization algorithm without common regularization terms.
• Used optimum bright-field STEM data as an initial potential guess and a loss function based on measured/estimated 4D STEM data.

Main Results:

• Successfully retrieved quantitative potential from thick samples across various simulated conditions (thickness, dose, incoherence).
• Demonstrated the algorithm's efficacy on experimental 4D STEM data.
• Investigated the extension of the algorithm for segmented detectors.

Conclusions:

• The proposed gradient-descent optimization algorithm effectively enables quantitative potential retrieval in STEM for thicker samples.
• This method advances material characterization capabilities by mitigating multiple scattering artifacts.