Reconstructing porous structures from FIB-SEM image data: Optimizing sampling scheme and image processing

Nano-porous materials can be imaged spatially by focused ion beam scanning electron microscopy (FIB-SEM). This method generates a stack of SEM images that has to be segmented (or reconstructed) to serve as basis for structural characterization. To this end, we apply two state-of-the-art algorithms. We study the influence of the original image's voxel size on estimates of morphological characteristics and effective permeabilities. Special attention is paid to analyzing anisotropies due to the FIB-SEM typical anisotropic sampling. Quantitative comparison of morphological descriptors and flow properties of reconstructed data is enabled by the use of synthetic FIB-SEM sets for which a ground truth is available. Moreover, in that case, reconstruction parameters can be chosen optimally, too.

Automated summary

The spatial imaging of nano-porous materials using focused ion beam scanning electron microscopy (FIB-SEM) involves generating a stack of SEM images that are segmented and reconstructed for structural characterization using advanced algorithms. The influence of the original image's voxel size on estimates of morphological characteristics and effective permeabilities is studied, with a focus on anisotropies due to the FIB-SEM typical anisotropic sampling. Quantitative comparison of morphological descriptors and flow properties of reconstructed data is facilitated by using synthetic FIB-SEM sets with a known ground truth.