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

MM: Fachverband Metall- und Materialphysik

MM 70: Topical Session (Symposium MM): Big Data in Materials Science - Managing and exploiting the raw material of the 21st century

MM 70.2: Vortrag

Freitag, 16. März 2018, 10:00–10:15, H 0107

High-throughput classification and categorization of structures from atomistic simulations — •Lauri Himanen, Patrick Rinke, and Adam Foster — Department of Applied Physics, Aalto University, Espoo, Finland

Our capability of producing, storing and analysing computational materials science data has grown tremendously. As the high-throughput screening of materials is becoming ever more popular, materials databases are being filled with atomic and electronic structure data.

To enable structure-related database queries, specific structural classes need to be defined. Unfortunately the required information is not always provided, and when it is, it is often based on an unspecified definition. To cope with large heterogeneous datasets of atomistic calculations, automated and verifiable methods for analyzing and categorizing atomistic structures are becoming necessary.

We discuss different methods that can be used in extracting structural information from various structural classes. These techniques involve finding a standardized unit cell, finding a translational basis for periodic materials within complex atomic environments and cluster analysis for separating different structural components. We also propose a material map that can be used to categorize the structural space and apply the introduced methods in the automatic classification of pristine crystals, surfaces and 2D materials.