MRS Meetings and Events

 

MD02.08.10 2023 MRS Spring Meeting

S/TEMfit—A Tool for Real-Space Analysis of Large-Scale S/TEM Images

When and Where

Apr 14, 2023
11:00am - 11:15am

Marriott Marquis, Second Level, Foothill G1/G2

Presenter

Co-Author(s)

Ewout van der Veer1,Ruben Hamming-Green1,Beatriz Noheda1,Majid Ahmadi1,Bart Kooi1

University of Groningen1

Abstract

Ewout van der Veer1,Ruben Hamming-Green1,Beatriz Noheda1,Majid Ahmadi1,Bart Kooi1

University of Groningen1
The analysis of atomic resolution (scanning) transmission electron microscopy images has been a topic of much research interest in recent years due to its ability to locally determine material properties such as elastic strain, electric polarization, phase distributions, etc. Real-space techniques, in particular, have taken flight to circumvent the limitations of traditional methods like geometric phase analysis. In this context, several software packages have been developed, including Atomap, STATstem and mpfit. Although these tools have shown great utility in the analysis of highly idealized images containing relatively few atoms, they tend to fall short on larger images and those with defects like out-of-zone or amorphous regions. This precludes the accurate analysis of, for example, highly strained epitaxial thin films.<br/>We demonstrate the analysis of such strained films consisting of BaTiO<sub>3</sub> grown on a SrTiO<sub>3</sub> substrate using a Sn-doped SrTiO<sub>3</sub> buffer layer with tunable lattice parameter using an in-house developed software package—called S/TEMfit. The package is designed for the <i>a posteriori</i> analysis of as wide a range of S/TEM images as possible in a performant and automated manner. It is fundamentally based on principal component analysis for image filtering, a thresholding procedure for atom detection and 2D Gaussian fitting.<br/>We show that S/TEMfit allows us to visualize continuous strain relaxation in our film and resolve inter-layer lattice parameter differences down to at least 2 pm, despite the presence of significantly defective regions in the images. It further allows us to image electric polarization and identify different structural phases in the image.

Keywords

nanostructure | transmission electron microscopy (TEM)

Symposium Organizers

Soumendu Bagchi, Los Alamos National Laboratory
Huck Beng Chew, The University of Illinois at Urbana-Champaign
Haoran Wang, Utah State University
Jiaxin Zhang, Oak Ridge National Laboratory

Symposium Support

Bronze
Patterns and Matter, Cell Press

Publishing Alliance

MRS publishes with Springer Nature