Counting Atoms in Gold Nanoparticles Using Low Dose Four-Dimensional Scanning Electron Microscopy with Supervised Machine Learning

scite shows how a scientific paper has been cited by providing the context of the citation, a classification describing whether it supports, mentions, or contrasts the cited claim, and a label indicating in which section the citation was made.

Header Abstract Content Author and article information References Comments

Record: found
Abstract: found
Conference Proceedings: found

Is Open Access

Counting Atoms in Gold Nanoparticles Using Low Dose Four-Dimensional Scanning Electron Microscopy with Supervised Machine Learning

Published

conference-abstract

Author(s): Cheng Zhao ¹ , Alireza Sadri ² , Bryan Esser ³ , Zhi Tong Liew ⁴ , Alison Funston ⁴ , Scott Findlay ² , Joanne Etheridge ² ^, ³ ^,

Publication date (Electronic): 21 January 2025

Conference name: 13th Asia Pacific Microscopy Congress 2025 (APMC13)

Conference date: 2-7 Febuary 2025

Keywords: atom countin, low dose 4DSTE, machine learnin

Bookmark

Abstract

Due to their unique nanoscale properties, metal nanoparticles are of great interest for applications such as drug delivery, biological imaging, and chemical sensing [ 1]. Although wet chemical synthesis is a common technique for synthesising these nanoparticles, the precise mechanisms responsible for symmetry breaking and subsequent growth into a specific shape still remains under debate, despite extensive progress in synthetic methods over the years. It is well established that varying surfactants can have a significant impact on the morphology of these particles.

In this work, we aim to study the influence of surfactants on the surface energies of nanoparticle facets. Our approach uses quantitative scanning transmission electron microscopy (STEM) to obtain a measure of surface mobility, as a proxy to estimate surface stability and hence energy [ 2], by calculating the difference in number of atoms between two consecutive scans with known electron doses. This requires an accurate measurement of atom locations and numbers under the lowest possible dose, in order to minimize the impact of the electron beam (see Fig.1). Such low dose datasets are comprised of noisy and sparse diffraction patterns which makes thickness estimation of the sample very challenging.

To improve robustness against noise, we employed a modified Capsule Neural Network (CapsNet) architecture and trained the network with diffraction patterns within Au unit cells of various thicknesses and noise realizations. The training considers invariance to prior knowledge of location of unit cells by using histograms of binned data as the input. We then applied the trained network to estimate the number of atoms in different datasets to investigate the atom mobility on a gold nanoparticle, with the use of double aberration-corrected microscopes (FEI Titan3 80-300 FEG TEM and a recently installed Thermo Fisher Scientific Spectra φ FEG TEM).

Fig. 1

Schematics for the determination of facet mobility. a) and b) two consecutive 4D-STEM scans with known electron doses, and c) difference in number of atoms in each atom column. The mobility of the atoms on this facet can be measured as the energy of one scan under which one atom movement can be detected.

Content

Author and article information

Conference

Publication date (Electronic): 21 January 2025

Electronic Location Identifier: e142

Affiliations

[1 ]Department of Materials Science and Engineering, Monash University, Clayton, Victoria 3800, Australia

[2 ]School of Physics and Astronomy, Monash University, Clayton, Victoria 3800, Australia

[3 ]Monash Centre for Electron Microscopy, Monash University, Clayton, Victoria 3800, Australia

[4 ]School of Chemistry, Monash University, Clayton, Victoria 3800, Australia

Author notes

*Corresponding author: joanne.etheridge@ 123456monash.edu

Article

DOI: 10.14293/APMC13-2025-0142

SO-VID: ae1af138-28ab-4ecf-97ff-7ce8a4daa66e

License:

Published under Creative Commons Attribution 4.0 International ( CC BY 4.0). Users are allowed to share (copy and redistribute the material in any medium or format) and adapt (remix, transform, and build upon the material for any purpose, even commercially), as long as the authors and the publisher are explicitly identified and properly acknowledged as the original source.

Conference name: 13th Asia Pacific Microscopy Congress 2025

Conference acronym: APMC13

Conference number: 13

Conference location: Brisbane, Australia

Conference date: 2-7 Febuary 2025

History

Funding

Funded by: Australian Research Council (ARC)

Award ID: FT190100619

Funded by: Australian Research Council (ARC)

Award ID: FL220100202

This work was supported by the Australian Research Council (ARC) grants FT190100619 and FL220100202. The authors acknowledge the use of the instruments and scientific and technical assistance at the Monash Centre for Electron Microscopy, a Node of Microscopy Australia. This research used equipment funded by ARC grants LE0454166 and LE170100118.

References

L M Liz-Marzán, Materials Today 7 (2004), p. 26–31. https://doi.org/10.1016/S1369-7021(04)00080-X
H Katz-Boon et al., Nano Letters 15 (2015), p. 1635-1641. https://doi.org/10.1021/acs.nanolett.5b00124
S Sabour et al., Advances in Neural Information Processing Systems 30 (2017). https://doi.org/10.48550/arXiv.1710.09829

Comments

Comment on this article

[1] L M Liz-Marzán, Materials Today 7 (2004), p. 26–31. https://doi.org/10.1016/S1369-7021(04)00080-X

[2] H Katz-Boon et al., Nano Letters 15 (2015), p. 1635-1641. https://doi.org/10.1021/acs.nanolett.5b00124

[3] S Sabour et al., Advances in Neural Information Processing Systems 30 (2017). https://doi.org/10.48550/arXiv.1710.09829

Counting Atoms in Gold Nanoparticles Using Low Dose Four-Dimensional Scanning Electron Microscopy with Supervised Machine Learning

Conference

Affiliations

Author notes

Article

History

Funding

Categories

Comment on this article