Focused ion beam-scanning electron microscopy on solid-oxide fuel-cell electrode: Image analysis and computing effective transport properties

In the present work, microstructural and transport properties of a three-dimensional (3D) microstructure of lanthanum strontium manganite (LSM) are deduced using dual-beam focused ion beam-scanning electron microscopy (FIB-SEM) facility. A series of two-dimensional (2D) cross-sectional images are collected from the LSM sample using FIB-SEM and then reconstructed to 3D structures from the 20 images in a systematic approach. For the first time, the effect of different image processing steps including threshold value, median filter radius, morphological operators, surface triangulation, smoothing filter, etc., on porosity, internal surface area, electronic conductivity and diffusivity are studied. Variation of 33% and 25% on porosity E and internal surface area S. respectively is observed because of improper selection of threshold value, median filter radius, and morphological operator. The number of triangular surfaces used in 3D reconstructions also varied the porosity sand internal surface area S by 14.5% and 4.4%, respectively. Computational domains for calculating effective transport properties are generated using body-fitted cut-cell based finite volume meshes on reconstructed 3D volumes. The normalized effective transport properties are computed on computational domains reconstructed by the FIB-SEM as well as by a numerical model. For the FIB-SEM reconstruction case, the normalized effective properties in z-direction are 25-44% smaller than those properties in x and y directions. This difference is significant and reveals the anisotropy in FIB-SEM reconstructed volume compared to numerically reconstructed volume. The presence of large crater, milling direction and smaller 3D FIB-SEM reconstructed volume could be the main reasons for this local anisotropy. (C) 2010 Elsevier B.V. All rights reserved.