[en] Electron tomography is a unique technique for imaging the microstructure of materials with a nanometer resolution. The signal to noise ratio of electron tomograms is, however, often too low for a reliable segmentation-based image analysis. We derive a general relation between the grey-tone correlation function of the tomograms and the 2-point correlation function of the morphology, which enables to analyze quantitatively the grey-tone correlation function with a morphological model of the material. The methodology is applied to SBA-15 ordered mesoporous silica. The three-dimensional grey-tone correlation function obtained from electron tomography is analyzed in terms of a hexagonal array of Gaussian independent pores. The model enables to relate the morphology obtained from the 2-point correlation function to macroscopic characterization data of the material, notably small-angle X-ray scattering and nitrogen adsorption.
Departement de Chimie Appliquée, ULg
Fonds de la Recherche Scientifique (Communauté française de Belgique) - F.R.S.-FNRS