Machine Learning Method Reveals Hidden Strong Metal‐Support Interaction in Microscopy Datasets

Abstract

Forming an ultra‐thin, permeable encapsulation oxide‐support layer on a metal catalyst surface is considered an effective strategy for achieving a balance between high stability and high activity in heterogenous catalysts. The success of such a design relies not only on the thickness, ideally one to two atomic layers thick, but also on the morphology and chemistry of the encapsulation layer. Reliably identifying the presence and chemical nature of such a trace layer has been challenging. Electron energy‐loss spectroscopy (EELS) performed in a scanning transmission electron microscope (STEM), the primary technique utilized for such studies, is limited by a weak signal on overlayers when using conventional analysis methods, often leading to misinterpreted or missed information. Here, a robust, unsupervised machine learning data analysis method is developed to reveal trace encapsulation layers that are otherwise overlooked in STEM‐EELS datasets. This method provides a reliable tool for analyzing encapsulation of catalysts and is generally applicable to any spectroscopic analysis of materials and devices where revealing a trace signal and its spatial distribution is challenging.

Engineering Heterogeneous Catalysis with Strong Metal–Support Interactions: Characterization, Theory and Manipulation

“…The insights from the following computational tools are of tremendous interest: i) accurate modeling of interfacial sites; [42,112] ii) modeling the incorporation of adsorbate interactions and environmental conditions; [19,113] iii) ab initio molecular dynamics; [114] iv) machine learning algorithms. [115] d) Extension to related fields. Beyond the field of thermocatalysis, supported catalysts are ubiquitous in photocatalysis and electrocatalysis.…”

“…In addition, constructing the proper SMSI state could be a complicated combination of metal+support+adsorbate+inducing conditions. The insights from the following computational tools are of tremendous interest: i) accurate modeling of interfacial sites; [42, 112] ii) modeling the incorporation of adsorbate interactions and environmental conditions; [19, 113] iii) ab initio molecular dynamics; [114] iv) machine learning algorithms [115] Extension to related fields .…”

Machine learning in electron microscopy for advanced nanocharacterization: current developments, available tools and future outlook

“…Specially from articles comparing, for a single application or for a similar purpose, some of these methods and confronting them towards finding the best performer. [233][234][235][236][237][238][239][240][241][242][243][244] In a general basis, NMF was the preferred way to proceed when the major goal lied beyond denoising or data inspection. That is because it can admit not only the direct physical significance of the extracted components, but to add additional constraints with ease.…”

Pixel-wise classification in graphene-detection with tree-based machine learning algorithms

“…Especially, image classification using deep learning (DL) has emerged as a game changer technique that has allowed drastic reduction of analysis time from hours to seconds. For example, the convolutional neural network (CNN) has been used in biomedical fields, such as abdominal CT scan, cell, hippocampus, and pancreas segmentations [29][30][31][32], and in analyzing big image data obtained from satellites [33,34] providing significant aid to error-prone human eyes [35][36][37][38].…”