Three-Dimensional Observation of Dopant Atoms in Quantitative Scanning Transmission Electron Microscopy

One of the ultimate goals in electron microscopy is 3-dimensional (3D) atomic scale information. The main challenge in obtaining 3D information in microscopy originates from the limited depth resolution. The depth resolution of scanning transmission electron microscopy (STEM) remains above ~10 nm, even with state-of the-art STEM with aberration correction [1]. Through-focal series and confocal approaches can improve the resolution to about 5 nm, which is still not sufficient for atomic scale 3D studies. An alternative approach is to extract dopant depth information based on the STEM image intensity. The atomic number difference between the dopant and the host crystal will alter the intensity of the atomic column that contains the dopant. This approach, however, faces a few important challenges. Probe channeling and dynamical scattering along the depth direction can significantly complicate the dopant information [2], and sets a limit on the usable TEM sample thickness [3]. Furthermore, criteria that define the dopant visibility in the image must be established. To tackle these issues, and test the practical feasibility, a fully quantitative experiment in STEM is required.