Cardiff University
Browse
- No file added yet -

Data for Trainable Segmentation for Transmission Electron Microscope Images of Inorganic Nanoparticles

Download (473.5 MB)
figure
posted on 2024-09-18, 11:28 authored by Thomas SlaterThomas Slater, Kevin Treder

The data are 3 sets of images that were used to investigate trainable segmentation using the open-source python package ParticleSpy (10.5281/zenodo.3763072). The images were obtained using either high-angle annular dark field (HAADF) scanning transmission electron microscopy (STEM) or high-resolution transmission electron microscopy (HRTEM).

Specififcally, the 3 datasets are:

1: HAADF-STEM images of a mix of metallic nanoparticles, namely PtNi nanoparticles, Pd nanocubes and Au nanospheres. The images were obtained using a JEOL ARM 200 scanning transmission electron microscope (STEM). All images were acquired with a 200 kV accelerating voltage, approximately 13 pA beam current, 23 mrad convergence angle, and 80 mrad ADF detector inner angle.

Sample synthesis: Gold nanoparticles of nominally 20 nm diameter were purchased from Sigma-Aldrich (product ref. 741965). PtNi nanoparticles were synthesized using a procedure previously published (Wang et al., Reference Wang, Slater, Leteba, Roseman, Race, Young, Kirkland, Lang and Haigh2019), using the OLEA-aged particles from this previous publication. For the synthesis of the Pd nanocubes, 60 mg of ascorbic acid (≥99.0%, Sigma-Aldrich), 105 mg of polyvinylpyrrolidone (PVP, Sigma-Aldrich, M.W. 55,000 g/mol), 5 mg of KBr (VMR chemicals), and 185 mg of KCl (VMR chemicals) were placed in a 20 mL vial and dissolved in 8 mL of deionized water. The resulting solution was heated in air under magnetic stirring at 80 °C. After 10 min, a solution containing 57 mg of Na2PdCl4 in 3 mL of deionized water was quickly added, and the reaction was allowed to proceed at 80 °C for 3 h. The product was collected by centrifugation and washed two times with acetone and ten times with water to remove excess PVP. All nanoparticle solutions were sequentially deposited dropwise onto a single holey carbon covered copper TEM grid (200 mesh) to create a grid containing all three nanoparticles.

2. HAADF-STEM images of Pt atomic clusters on a carbon support. The images were obtained using a JEOL ARM 200 scanning transmission electron microscope (STEM). All images were acquired with a 200 kV accelerating voltage, approximately 13 pA beam current, 23 mrad convergence angle, and 80 mrad ADF detector inner angle.

Sample synthesis: The deposition of Pt atoms on g-C3N4via magnetron sputtering was carried out using a bespoke AJA International system. g-C3N4 (1 g) was placed into a tailor-made stirring sample-holder, and then loaded into the magnetron sputtering chamber (reaching 3 × 10−8 torr background pressure in 40 min). After waiting 10 min for background pressure stabilisation, Ar gas (99.9999%) was introduced into the chamber, reaching 3 × 10−3 torr. Upon using an applied power (370 V and 16 mA), highly energetic ions were accelerated against the Pt target leading to a cascade of collisions at the surface of the Pt target, thus ejecting primarily individual neutral Pt atoms from the target, which were then deposited onto the g-C3N4 framework. The g-C3N4 powder was stirred constantly during the Pt atoms deposition to allow for atomic dispersion throughout the whole powder (Fig. 1a). In this fashion, Pt atoms deposition was carried out for 12 min, yielding Pt/g-C3N4 (0.5 wt% of Pt onto 1 g of g-C3N4).

3. HRTEM images of Au nanoparticles on a Ge film. The images were obtained using a JEOL ARM 200 transmission electron microscope (TEM). All images were acquired with a 200 kV accelerating voltage.

Sample Synthesis: This sample is a standard provided by the microscope manufacturer (JEOL).

Research results based upon these data are pubished at https://doi.org/10.1111/jmi.13110


Funding

Automating electron microscopy: machine learning for cluster identification (2022-01-31 - 2023-03-31); Slater, Thomas. Funder: Engineering and Physical Sciences Research Council

History

Specialist software required to view data files

The images can be opened natively using Gatan's Digital Micrograph software, which is available free of charge from Gatan's website. Opening the images with non-proprietary software is possible using the open-source python package Hyperspy.

Usage metrics

    School of Chemistry

    Categories

    No categories selected

    Licence

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC