TFS-STEM: Combined tilt- and focal series for STEM tomography with a computational correction for beam blurring

• Applicants: Dr.-Ing. Tim Dahmen; Professor Dr. Niels De Jonge
• Subject Area: Synthesis and Properties of Functional Materials ; Image and Language Processing, Computer Graphics and Visualisation, Human Computer Interaction, Ubiquitous and Wearable Computing
• Term: from 2018 to 2022
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 402741306

The primary method currently used for obtaining insight into the threedimensional (3D) structure at the nanometer scale of unique samples from biology and materials science is tilt-series transmission electron microscopy (TEM). In this project, we propose to establish combined tilt- and focal series (CTFS) as a new recording scheme for threedimensional (3D) scanning transmission electron microscopy (STEM) imaging. In this scheme, the specimen is rotated in relatively large tilt increments, and for every tilt direction, a through-focal series is recorded. A 3D volume is then reconstructed using adapted algorithms of computed tomography. This leads to an increased resolution in the axial direction and thus to a more faithful representation of 3D shapes. Particularly for the imaging of thick specimen of 1 um and thicker, TFS exhibits the advantage of a reduced overall tilt range and reduced beam blurring over TEM tomography. The project aims to: (i) understand the CTFS parameters influencing resolution and electron dose, (ii) apply the method and experimentally measure the achievable resolution for different samples from both materials science and biology of a range of thicknesses, and (iii) improve the reachable resolution for very thick specimen by developing a computational method to compensate beam blurring. By realizing those goals, we aim to establish a new standard 3D electron microscopy method that can contribute fundamentally to the fields of material science and biology. The project comprises 5 work packages: -WP1: Analysis of the parameter space of TFS. -WP2: Investigation and stabilization of the reconstruction algorithm. -WP3: Experimental analysis of the reachable resolution for thin and medium-thick samples (50 - 300 nm). -WP4: Experimental analysis of the maximum reachable resolution for micrometers-thick specimens. -WP5: Development of algorithms for reducing the effect of beam blurring in 3D datasets of thick samples . The project is submitted to the category Künstliche Intelligenz, Bild- und Sprachverarbeitung, but it proposes interdisciplinary research and combines methods from computer science, physics, electron microscopy, material sciences and applied mathematics.

• DFG Programme: Research Grants