An experimental protocol development of three-dimensional transmission electron microscopy methods for ferrous alloys: Towards quantitative microstructural characterization in three dimensions

The majority of engineering steels are ferromagnetic and structually inhomogeneous on special scales ranging from nanometers to micrometers, and physical properties of engineering steels arise from three-dimensional (3D) features of the microstructure. Thus, obtaining 3D representation with a large field of view is desired for transmission electron microscopy (TEM) based microstructure characterization to establish microstructure - physical properties relationships with reasonable statistical relevancy. Here, we venture to use a conventional sample preparation process, i.e., mechanical polishing followed by electro-polishing, and experimental protocols optimization for electron tomography (ET) for ferromagnetic materials, especially engineering steels' microstructural characterization are carried out. We found that the sample thickness after the mechanical polishing step is a critical experimental parameter affecting the success rate of tilt-series image acquisition. For example, for a ferritic heat-resistant 9Cr steel with lath martensite structure, mechanically thinning down to 30 μm or thinner was necessary to acquire an adequate tilt-series image of carbide precipitates in the high-angle annular dark-field scanning TEM (HAADF-STEM) mode. On the other hand, tilt-series image acquisition from dislocation structures remains challenging because the electron beam deflection during specimen-tilt was unavoidable and significant in the HAADF-STEM mode. To overcome the electron beam deflection problem, we evaluate several relatively accessible approaches including the "Low-Mag and Lorentz" TEM/STEM modes; although they are rarely used for ET, both the modes reduce or even zero the objective lens current and likely weaken the magnetic interference between the ferromagnetic specimen and the objective lens magnetic field. The advantages and disadvantages of those experimental components are discussed.