Image-analytical evaluation of the spatial distribution of particulate fillers in ceramic composites prepared via the polymer-derived ceramics route

Abstract

A robust and versatile image analysis routine is presented for the quantitative evaluation of the spatial distribution of a recently developed ceramic composite system prepared by using preceramic polymers as both low-loss binder and reagent in combination with particulate fillers, yielding particulate zirconia-strengthened mullite. Micrographs of polished cross-sections were used as a base for binarization, noise-reduction and Voronoi-based tessellation operations, employing open-source software. The resulting cell area distributions were used to evaluate the microstructural homogeneity by calculating a defined factor of homogeneity. Furthermore, the skewness of the cell area distributions was introduced as a novel characteristic number for the characterization of the filler distribution. The resulting values of the experimentally determined microstructural descriptors were compared to values calculated from theoretical distributions following a Poisson Voronoi model, including a hard-core model approach, to determine the quality of spatial dispersion. Significant differences in the spatial distribution of zirconia particulates in the mullite matrix were found depending on the method of introduction of fillers into the precursor material, spray-coating being identified as the superior compounding method. The comparison to a conventional nearest-neighbor metric demonstrates the potential of the developed tessellation-based image-analysis routine, yielding additional information about the nature of the microstructural setup of the material.