Yuanyuan Zhu

Based on historical developments and the current state of the art in gas-phase transmission electron microscopy (GP-TEM), we provide a perspective covering exciting new technologies and methodologies of relevance for chemical and surface sciences. Considering thermal and photochemical reaction environments, we emphasize the benefit of implementing gas cells, quantitative TEM approaches using sensitive detection for structured electron illumination (in space and time) and data denoising, optical excitation, and data mining using autonomous machine learning techniques. These emerging advances open new ways to accelerate discoveries in chemical and surface sciences.

Graphical abstract

Crystalline materials exhibit long-range ordered lattice unit, within which resides nonperiodic structural features called defects. These crystallographic defects play a vital role in determining the physical and mechanical properties of a wide range of material systems. While computer vision has demonstrated success in recognizing feature patterns in images with well-defined contrast, automated identification of nanometer scale crystallographic defects in electron micrographs governed by complex contrast mechanisms is still a challenging task. Here, building upon an advanced defect imaging mode that offers high feature clarity, we introduce DefectSegNet - a new convolutional neural network (CNN) architecture that performs semantic segmentation of three common crystallographic defects in structural alloys: dislocation lines, precipitates and voids. Results from supervised training on a small set of high-quality defect images of steels show high pixel-wise accuracy across all three types of defects: 91.60 ± 1.77% on dislocations, 93.39 ± 1.00% on precipitates, and 98.85 ± 0.56% on voids. We discuss the sources of uncertainties in CNN prediction and the training data in terms of feature density, representation and homogeneity and their effects on deep learning performance. Further defect quantification using DefectSegNet prediction outperforms human expert average, presenting a promising new workflow for fast and statistically meaningful quantification of materials defects.

As gas–solid heterogeneous catalytic reactions are molecular in nature, a full mechanistic understanding of the process requires atomic‐scale characterization under realistic operating conditions. Although atomic resolution imaging has become routine in modern high‐vacuum (scanning) transmission electron microscopy ((S)TEM), both image quality and resolution nominally degrade on introduction of the reaction gases. In this work, we systematically assess the effects of different gases at various pressures on the quality and resolution of images obtained at room temperature in the annular dark field STEM imaging mode by using a differentially pumped (DP) gas cell. This imaging mode is largely free from inelastic scattering effects induced by the presence of gases and retains good imaging properties over a wide range of gas mass/pressures. We demonstrate the application of ESTEM with atomic resolution images for a complex oxide alkane oxidation catalyst MoVNbTeO_x (M1) immersed in light and heavy gas environments.