Full text loading...
Cross-sectional views of the 3D reconstruction of the pore structures: (a) one slice of the reconstruction showing the cross section of the pore structures along the original projection direction; (b) one cross section of the reconstruction showing the pore structure within the material thickness only available by 3D reconstruction. This overview reconstruction was computed with images that were rebinned by two due to physical computer memory limitations. The arrow in (b) points to an intact pore only 6 nm from the top surface. The intensity and streaking effects outside the sample in the -direction in (b) is mostly due to the point spreading of the Pt/Cu redeposited particles and the residuals from artifacts of the initial back projection. It is not indicative of surface damage. The intensity of the Pt cap layer and the redeposited particles are rendered with saturation to give enough contrast to visualize the pores in gray scale. The saturated rendering produces a visually exaggerated elongation artifact of the particles. The scale bars in (a) and (b) are 20 nm.
Multiple views of a typical redeposited Pt/Cu particle on the 3D reconstructed surface of the low- dielectric: (a) the 3D volume rendering of the particle with max-intensity projections shown on the walls; (b) the max-intensity projection of the particle in direction; [(c)–(e)] the cross sectional slice through the center of mass: (c) cross section, (d) cross section, and (e) cross section of the reconstructed particle. This particle and others like it are used to estimate the PSF of the tomographic reconstruction process for these experiments.
Visualization of the 3D reconstruction of the porous low- dielectric with autosegmented isolines overlaid: (a) perspective view of orthogonal cross sections through the 3D reconstruction of the pore structure; (b) one slice of the reconstruction showing a large pore network on the top-left corner. This tomogram was reconstructed from the full-pixel data.
The pore-diameter distribution extracted from automatic pore segmentation: (a) histogram of pore diameters in log-scale with two-Gaussian fit. The pore-size statistics follow a log-normal distribution (Gauss1), where the second Gaussian function (Gauss2) accounts for the noise in the measurement; (b) comparison of the percentage volume fraction density between tomographic and EP measurements. The areas under the EP and tomography curves are 18.4% and 6.9%, respectively. Missing data points on this plot are zeros in linear scale.
Article metrics loading...