Correlative tomography

Abstract

Increasingly researchers are looking to bring together perspectives across multiple scales, or to combine insights from different techniques, for the same region of interest. To this end, correlative microscopy has already yielded substantial new insights in two dimensions (2D). Here we develop correlative tomography where the correlative task is somewhat more challenging because the volume of interest is typically hidden beneath the sample surface. We have threaded together x-ray computed tomography, serial section FIB-SEM tomography, electron backscatter diffraction and finally TEM elemental analysis all for the same 3D region. This has allowed observation of the competition between pitting corrosion and intergranular corrosion at multiple scales revealing the structural hierarchy, crystallography and chemistry of veiled corrosion pits in stainless steel. With automated correlative workflows and co-visualization of the multi-scale or multi-modal datasets the technique promises to provide insights across biological, geological and materials science that are impossible using either individual or multiple uncorrelated techniques.

Figure 1

Correlative tomography workflow applied to study the nucleation and growth of pits in sensitized stainless steel linking together (A) medium and (B) high resolution X-ray CT with (C) serial sectioning FIB-SEM followed by EBSD imaging and (D) STEM-EDX imaging. B2 and C1 show the co-registry of surface texture as seen by X-ray CT and SEM imaging respectively with red arrows marking key features used in the correlation; B2 and C1 show the surface as viewed by X-ray CT and SEM respectively with the sub-surface extent of the intergranular corrosion displayed as a semi-transparent overlay on to the B2. C3 shows an EBSD map of the crystal orientation in Euler colours. D1 is a high angle annular dark image of the extracted slice and D2 presents STEM-EDX combining Fe, Ni, Cr, Mo, Si elemental maps. The magnifications and resolutions of the images collected by each instrument are indicated.

Figure 2

a) Virtual slice through the 3D image collected at high resolution (0.8 μm voxel size) for the selected pit (approximate extent marked by red dashed line), b) side and c) underside 3D segmented visualization of the pit (red), local intergranular corrosion front network (light blue), surface perforations (dark blue) and sample surface (purple).

Figure 3

a) 3D visualization of a region of the crack front obtained by serial section FIB-SEM with the grain boundary (A) and CSL boundaries (B) indicated. b) EBSD map from final slice adjacent to the serial section FIB-SEM volume, showing crystal orientation Euler colours. (A) marks the grain boundary, (B) marks the CSL boundaries and (C) marks two examples of slip bands (referred to in main text). c) a high angle annular dark field (HAADF) image of the extracted slice, d) STEM-EDX image of the same slice showing a chemical map with a combination of Fe, Ni, Cr, Mo and Si. e) and f) high resolution elemental maps showing Cr distribution along and ahead of the corrosion front along e) the Σ11boundary and f) the slip band are indicated in d) by D1 and D2 respectively.