Current work

I am a postdoctoral researcher at SIMaP laboratory, Grenoble INP, working on algorithm development for 3D reconstructions of multi-phase nanoscale materials from 4D Electron Diffraction data, and more specifically Scanning Precession Electron Diffraction (SPED) data. Diffraction contrast allows indvidual grains to be distinguished within a polycrystalline sample, which is not possible using common imaging signals in the electron microscope (BF/ADF). Tilted projections of individual grains are calculated for tomographic reconstruction by combining SPED with tilt-series tomography, enabling 3D reconstrion of the sample volume and analysis of the material microstructure.

This work involves the development of grain reconstruction algorithms from the diffraction data using Python-based machine learning and image processing. The backbone of the processing workflow is based around grain orientations, which can be calculated using the Automated Crystal Orientation Mapping (ACOM) technique. The dimensionality of the individual crystal orientations maps is reduced through orientation-based clustering into grain components, which are tracked throughout the different tilt-series reference frames, allowing for both grain coupling and orientation refinement. Development of Virtual Reconstruction (VR) algorithms enable orientation-specific contrast and unambiguous determination of the same grain throughout the tilt-series.

This work is explained further in the following publications:

• Reconstructing Dual-Phase Nanometer Scale Grains within a Pearlitic Steel Tip in 3D through 4D-Scanning Precession Electron Diffraction Tomography and Automated Crystal Orientation Mapping
• New Features in Crystal Orientation and Phase Mapping for Transmission Electron Microscopy
• Reconstructing grains in 3D through 4D Scanning Precession Electron Diffraction

For those interested in undertaking a similar analysis, a GUI is expected to be released in the future- more to follow! In the mean time check out orix for crystallographic analysis using Python.