Automatic 3D image registration for nano-resolution chemical mapping using synchrotron spectro-tomography

Area of Science:

• Materials Science
• Chemistry
• Physics

Background:

• Nano-resolution synchrotron X-ray spectro-tomography is a powerful technique for analyzing 3D structural and chemical heterogeneity.
• X-ray energy-dependent intensity variations in voxels provide local chemical fingerprints.
• Experimental challenges like hardware instability and sample deformation can compromise data quality.

Purpose of the Study:

• To address data quality degradation caused by unanticipated sample deformation in X-ray spectro-tomography.
• To develop and implement an automatic 3D image registration method for correcting sample deformation.
• To reveal the redox heterogeneity in partially delithiated Li_xTa_0.3Mn_0.4O₂ battery cathode particles with improved accuracy.

Main Methods:

• Utilized nano-resolution synchrotron X-ray spectro-tomography.
• Reconstructed tomographic datasets at various X-ray energy levels.
• Implemented an automatic 3D image registration algorithm to correct for sample deformation.

Main Results:

• Identified sample deformation as a significant factor degrading X-ray spectro-tomography data quality.
• Successfully applied the 3D image registration method to correct for deformation-induced artifacts.
• Achieved significantly improved fidelity in revealing the redox heterogeneity of Li_xTa_0.3Mn_0.4O₂ battery cathode particles.

Conclusions:

• Automatic 3D image registration is effective in mitigating data quality issues caused by sample deformation in X-ray spectro-tomography.
• The corrected data enables more accurate characterization of chemical heterogeneity in energy storage materials.
• This approach enhances the reliability of spectro-tomography for advanced materials research.