A Deep Learning Approach to Identify Local Structures in Atomic‐Resolution Transmission Electron Microscopy Images

Madsen, Jacob; Liu, Pei; Kling, Jens; Wagner, Jakob Birkedal; Hansen, Thomas Willum; Winther, Ole; Schiøtz, Jakob

doi:10.1002/adts.201800037

Cited by 168 publications

(152 citation statements)

References 43 publications

(105 reference statements)

Supporting

Mentioning

150

Contrasting

Order By: Relevance

“…28 The constructed network has a symmetric encoderdecoder structure and utilizes skip connections to provide low-level local features to the high-level global features during upsampling, as well as atrous convolutions in its bottleneck layers to probe features at multiple scales. 29 The models were trained using simulated STEM data 30,31 from three major types of graphene edges, namely, the zigzag, armchair, and bearded edge. Lattice defects such as vacancies and substitutional Si impurities were randomly introduced along the edges.…”

mentioning

confidence: 99%

Atomic Mechanisms for the Si Atom Dynamics in Graphene: Chemical Transformations at the Edge and in the Bulk

Ziatdinov

Dyck

Jesse

et al. 2019

Adv Funct Materials

View full text Add to dashboard Cite

The dynamic behavior of e‐beam irradiated Si atoms in the bulk and at the edges of single‐layer graphene is examined using scanning transmission electron microscopy (STEM). A deep learning network is used to convert experimental STEM movies into coordinates of individual Si and carbon atoms. A Gaussian mixture model is further used to establish the elementary atomic configurations of the Si atoms, defining the bonding geometries and chemical species and accounting for the discrete rotational symmetry of the host lattice. The frequencies and Markov transition probabilities between these states are determined. This analysis enables insight into the defect populations and chemical transformation networks from the atomically resolved STEM data. Here, a clear tendency is observed for the formation of a 1D Si crystal along zigzag direction of graphene edges and for the Si impurity coupling to topological defects in bulk graphene.

show abstract

mentioning

confidence: 99%

Atomic Mechanisms for the Si Atom Dynamics in Graphene: Chemical Transformations at the Edge and in the Bulk

Ziatdinov

Dyck

Jesse

et al. 2019

Adv Funct Materials

View full text Add to dashboard Cite

show abstract

“…The method is capable of extracting high-level information such as defect type, lattice orientation and strain, as well as characteristics of the electron probe. Our method is based on two algorithms, building on recent advances in deep-learning and on computational geometry and graph theory.The deep learning recognition model is similar to recently published results [2,3]. A neural network is trained to identify the smallest distinguishable repeated substructures within the image, i.e.…”

mentioning

confidence: 97%

“…The deep learning recognition model is similar to recently published results [2,3]. A neural network is trained to identify the smallest distinguishable repeated substructures within the image, i.e.…”

mentioning

confidence: 97%

Automated Real-time Analysis of Atomic-resolution STEM Images

2019

Self Cite

View full text Add to dashboard Cite

Understanding the microstructural characteristics of materials, such as lattice defects, increasingly relies on the analysis of large numbers of images from electron and scanning probe microscopy [1]. It is now becoming routine to record series of atomic-resolution images, resulting in the generation of massive datasets. The new challenge is then analysing this data. The workflow for such analyses typically comprises the identification of atomic positions and subsequent derivation of physical quantities, such as defect concentration and strain. Conventionally, this analysis is done manually, which is slow and laborious, and the results are prone to human errors and bias.We demonstrate an automatic method for extracting information from atomically resolved images. We take advantage of GPU acceleration and fast graph-based algorithms to enable real-time structural analysis. The method is capable of extracting high-level information such as defect type, lattice orientation and strain, as well as characteristics of the electron probe. Our method is based on two algorithms, building on recent advances in deep-learning and on computational geometry and graph theory.The deep learning recognition model is similar to recently published results [2,3]. A neural network is trained to identify the smallest distinguishable repeated substructures within the image, i.e. atoms or atomic columns of a particular species. We take advantage of the recent finding that deep neural networks trained using simulated data can generalise to experimental data [2]. Furthermore, by using randomisation in the generation of the synthetic images, the neural network is capable of making predictions with minimal prior assumptions of the types of defects present. Fig. 1 shows the results of the neural network applied to a noisy image of graphene with a silicon substitutional defect. A simple routine converts the predictions of the neural network to a set of 2D points representing the centres of the detected substructures, each point associated with a substructure class. We further explore the precision of the detected atom locations and their sensitivity to the imaging parameters including noise.The geometric relationship between these points encodes further information, for example, whether the substitutional silicon atom in graphene has three or four carbon neighbours. To facilitate fast geometric analysis, we create a geometric graph from the detected points. It is crucial that the graph is stable to small perturbations of the atomic positions. We identified a type of geometric graph, called a stable Delaunay graph [4], fulfilling this criterion while being fast to construct for large numbers of points. Using simple rules localised segments of the graph can be extracted representing an atom and its surrounding neighbourhood. Each segment of the graph is compared to a library of known graph templates, representing, for example, different defect types. The similarity between a segment and a template is calculated using the symmetry invariant root-m...

show abstract

“…Au/CeO 2 model systems have been studied extensively in the past decades, mainly focusing on the static structure at the atomic level. However, the surface dynamics on individual nanoparticles at the atomic level is rarely reported [23]. Only a few works show descriptive results as a function of time under the influence of the surroundings [2,[24][25][26].…”

mentioning

confidence: 99%

Reversible and concerted atom diffusion on supported gold nanoparticles

et al. 2020

Self Cite

View full text Add to dashboard Cite

Traditionally, direct imaging of atom diffusion is only available by scanning tunneling microscopy and field ion microscopy on geometry-constrained samples: flat surfaces for STM and needle tips for FIM. Here we show time-resolved atomic-scale HRTEM investigations of CeO 2 -supported Au nanoparticle surfaces to characterize the surface dynamics of atom columns on gold nanoparticles. The observed surface dynamics have been categorized into four types: layer jumping, layer gliding, re-orientation and surface reconstruction. We successfully captured atoms moving in a concerted manner with a time resolution of 0.1 s. A quantitative approach for measuring the dynamics in various gaseous surroundings at elevated temperatures is presented. An approach for measuring quantitative electron beam effects on the surface dynamics is presented by counting atom column occupation as a function of time under a range of dose rates in high vacuum.

show abstract

A Deep Learning Approach to Identify Local Structures in Atomic‐Resolution Transmission Electron Microscopy Images

Cited by 168 publications

References 43 publications

Atomic Mechanisms for the Si Atom Dynamics in Graphene: Chemical Transformations at the Edge and in the Bulk

Atomic Mechanisms for the Si Atom Dynamics in Graphene: Chemical Transformations at the Edge and in the Bulk

Automated Real-time Analysis of Atomic-resolution STEM Images

Reversible and concerted atom diffusion on supported gold nanoparticles

Contact Info

Product

Resources

About