Optimizing Atomic Number Contrast in Multislice Electron Ptychography

Abstract

Here we explore the atomic number ($Z$) dependence of multislice electron ptychography and approaches to optimize Z sensitivity. Specifically, we show that ptychography's $Z$-dependence is highly dependent on the integrated area of an atom column considered. A monotonic $Z$-dependence is found when the reconstructed projected atomic potentials are integrated over a small region. When increasing the integration area, $Z$-contrast changes significantly, becoming highly non-monotonic and following trends in the orbital shell-structure. Moreover, the reconstructed projected potential aligns with the transmission function with an overall deviation of only 2.4\%. The non-monotonic $Z$-dependence is further shown to be useful to accentuate contrast between certain elements, allowing for distinguishability of elements that are only a single atomic number apart, and even in $>$ 20 nm thick samples. This is demonstrated for $\beta$-CuZn ($Z$ = 29 and 30), with the differentiability between the elements explored for different signal quantification methods. The impact of electron dose and finite effective source size are also considered. These results demonstrate that the atom column integration area can optimize ptychographic $Z$-contrast for specific applications and experimental conditions.