Image Segmentation for FIB-SEM Serial Sectioning of a Si/C-Graphite Composite Anode Microstructure Based on...

Area of Science:

• Materials Science
• Electrochemistry
• Image Analysis

Background:

• Electrode microstructure significantly impacts lithium-ion battery performance, energy density, and power density.
• Controlling electrode microstructure during manufacturing is crucial for high-performance batteries.
• Understanding microstructure aids in designing cost-effective, high-performance batteries.

Purpose of the Study:

• To develop a systematic algorithm workflow for analyzing anode microstructures from focused ion beam scanning electron microscope (FIB-SEM) images.
• To address common issues in raw FIB-SEM images and propose effective preprocessing and segmentation methods.
• To reconstruct three-dimensional (3D) microstructures for improved battery design.

Main Methods:

• Preprocessing of raw FIB-SEM images to resolve typical issues.
• Application of a Fourier transform-based filter to reduce curtain artifacts.
• Implementation of a global-thresholding method for identifying active materials and pores.
• 3D reconstruction by concatenating segmented 2D images.

Main Results:

• A systematic algorithm workflow for FIB-SEM image analysis of battery microstructures was established.
• A Fourier transform filter effectively mitigated curtain artifacts in the images.
• A stable global-thresholding method achieved high segmentation performance for active materials and pores.
• 3D microstructures were successfully reconstructed, although with some manual user interaction required.

Conclusions:

• The proposed algorithm workflow provides a robust method for analyzing battery electrode microstructures.
• The developed image processing techniques, particularly the global-thresholding method, offer stable and effective segmentation.
• While not fully automated, the workflow aids in understanding and designing high-performance lithium-ion batteries.