DPG Phi
Verhandlungen
Verhandlungen
DPG

Dresden 2020 – wissenschaftliches Programm

Die DPG-Frühjahrstagung in Dresden musste abgesagt werden! Lesen Sie mehr ...

Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe

MM: Fachverband Metall- und Materialphysik

MM 58: Topical Session: Data Driven Materials Science - Machine Learning for Materials Characterization (joint session MM/CPP)

MM 58.5: Vortrag

Donnerstag, 19. März 2020, 17:00–17:15, BAR 205

Automatic Identification of Crystallographic Interfaces from Scanning Transmission Electron Microscopy Data by Artificial Intelligence — •Byung Chul Yeo1, Christian H. Liebscher2, Matthias Scheffler1, and Luca Ghiringhelli11Fritz-Haber-Institut der Max-Planck-Gesellschaft, Berlin, Germany — 2Max-Planck-Institut für Eisenforschung, Düsseldorf, Germany

Characterizing crystallographic interfaces in synthetic nanomaterials is an important step for the design of novel materials, e.g., catalysts, gas sensors, etc. In principle, trained materials scientists can assign interface structures of materials by looking at high-resolution imaging and diffraction data obtained by aberration-corrected scanning transmission electron microscopy (STEM). However, the high-acquisition rates in STEM pose a challenge to a purely human-based identification of interfaces or defects. As of today, STEM datasets are being massively accumulated, but they cannot be fully exploited due to the lack of automatic analysis tools. Here, we present a newly developed artificial-intelligence tool for accurately extracting the key features of (poly)crystalline materials, i.e., crystal-structure prototype, lattice constant, and (relative) orientation from atomic-resolution STEM images. The tool is based on a convolutional neural network and operates on both high-angle annular dark-field (HAADF) and convergent beam electron diffraction (CBED) images. The network is trained on 13 200 simulated STEM images, including structures distorted by thermal noise, and our model achieves excellent predictive performance for automatically identifying crystal structure and lattice misorientations.

100% | Mobil-Ansicht | English Version | Kontakt/Impressum/Datenschutz
DPG-Physik > DPG-Verhandlungen > 2020 > Dresden