Stability of electron ptychography at low electron dose

Dearg M, Michaelides N, Gilbert J, Ding Z, Aslam Z, Hopkinson D, Allen C, Clark L

Abstract Electron ptychography provides a promising avenue towards dose‐efficient, high‐resolution materials characterisation. Prior work demonstrates the feasibility of this approach, but an overarching view on the reliability of ptychographic images in low‐dose scenarios is required. Here, we address this limitation with a systematic study of image clarity across dose, thickness and convergence semi‐angle, on a range of materials science specimens. With the now widespread adoption of 4D‐STEM and ptychographic imaging, the establishment of the practical parameter space in which one can anticipate a reliably interpretable phase image is urgently needed. In some cases, our parameter space exploration confirms high‐resolution imaging at doses of 200 Å.
LAY DESCRIPTION Electron ptychography is an increasingly popular imaging technique, valued for its ability to image at much lower doses than conventional STEM imaging methods. In this manuscript we quantify the gains possible via ptychography in a comparison of methods (SSB, WDD, ePIE, iCoM), across different materials (GaN, STO, ZSM‐5) over thickness and dose series with both simulated and experimental data analysed. The phase imaging methods used include direct (non‐iterative) ptychography methods: SSB, WDD; iterative ptychography method (ePIE) and a related non‐ptychographic phase retrieval approach (iCoM). The samples studied include Gallium Nitride (previously studied in related papers), Strontium Titanate (a classic test material used in the electron microscopy community for its stable crystal structure) and Zeolite Socony Mobil–5 (ZSM‐5) a particularly beam‐sensitive material on which there is interesting prior experimental results in the literature. Through a series of simulations and experimental studies, we are able to characterise the behaviour of these imaging techniques across an important range of experimentally relevant conditions ‐ which we hope will be useful to colleagues in their future experimental design processes.

Keywords:

3101 Biochemistry and Cell Biology

,

31 Biological Sciences

,

34 Chemical Sciences

,

3406 Physical Chemistry

,

40 Engineering

,

4016 Materials Engineering

,

Biomedical Imaging