SSRL XANES tomography

This spectro-tomography dataset contains 67 tomograms of an NMC battery cathode particle at different x-ray energies. For each tomogram, 180 projection images were acquired at different angles. The x-ray energy points were selected such that we will be able to reconstruct the elemental maps of Mn, Co, and Ni (6 energies, one above and one below the respective K-edges of Mn, Co, and Ni), and a XANES scan across the Ni K-edge (for valence state distribution). The idea is to correlate the compositional heterogeneity with the redox heterogeneity. To read the data use read_xrm_script.py python script.

There are two independent image alignment tasks: 1) we need to correct the image jitter in a tomography dataset and 2) we need to register the tomograms at different x-ray energies.

For task 1, the SSRL team uses an iterative alignment protocol described in [B44], which works fine except that this process is very slow.

For task 2, the SSRL team use TXM-Wizard to register the projection images in the same sample orientation angle by angle. Images at different x-ray energies are of different magnification and contrast. SSRL’s approach is essentially calling the function described in section 3 of [B29] multiple times. An improved approach with better quality control and assessment is highly desirable. Note that the pixel size is nearly proportional to the x-ray energy.



Image preview





SSRL nano-CT

Sample name


X-ray energy

6534, 6558, 7704, 7728, 8178, 8570 eV

Pixel size

33.4 nm at 8670 eV