The combination of electron back-scatter diffraction (EBSD) with focused ion beam scanning electron microscopy (FIB-SEM) tomography extends high-resolution site-specific microstructural characterization to the third dimension. Phase relationships, local mis-orientations and grain properties such as size, morphology and boundary character can be studied in 3D. In an iterative process, the ion beam removes a slice of material before SEM image and EBSD map are acquired. This way a data stack is generated slice by slice that can be used afterwards to reconstruct the information about the volume of interest. However, established 3D EBSD workflows have been limited by resolution, accuracy and speed. These restrictions are now addressed by an advanced solution which will be presented here.
To assure homogeneous slice thicknesses and minimum sample drift, the software controlling the tomography acquisition features a unique system to track and adjust the milling progress. This is achieved by a pair of angled fiducial lines patterned in advance that gets regularly imaged throughout the run. The measurement of the distance between these lines allows to calculate the thickness of the milled slice and estimate the current sample drift. The milling progress is then adjusted accordingly to keep the slice thickness constant and compensate for the drift. While this assures optimal resolution in Z direction, best image resolution in XY can only be achieved by lowering the primary beam energy in combination with detectors capable of filtering the detected electrons. These conditions, however, are not ideal for high signal-to- noise EBSD map acquisition that typically needs a higher primary energy and beam current. Therefore, another unique feature is introduced that allows to store a second set of conditions for the EBSD mapping. The software automatically switches the conditions during the acquisition and advanced auto-functions assure perfect repositioning and fine-tuning of the beam. This way, SEM images and EBSD maps are acquired under ideal conditions respectively without sacrificing resolution.
Recent developments in the field of EBSD hardware are perfectly compatible with the described solution and facilitate the collection of 3D data additionally. Application examples will be presented to illustrate the above mentioned features.