IUC09 – Infrastructure interfaces with condensed matter physics (collaboration with FAIRmat)

← Back to Participant Projects and Use Cases

Main Task Area: TA-WSD
Other related Task Areas: TA-OMS
Possible connections within NFDI: FAIRmat
Material/Data: Atom Probe Tomography (APT) for metallic alloys
Main Success Scenario: Through the physical modeling of the measurement process, users get more precise information and a deeper understanding of their experimental results, in particular to distinguish defect states in the microstructure.
Added value for the MatWerk community: Workflows are designed in coordination with FAIRmat such that the exchange of data, metadata, ontologies, and concepts for data curation between two NFDI consortia is ensured without a loss of information. In this way, both consortia can mutually benefit from ongoing improvements in the infrastructure for condensed-matter physics (FAIRmat) and the interpretation of microstructure data (NFDI-MatWerk).

Main requirements

  • Workflows how to process measured data (e.g. detector hit positions, time-of-flight per ion) into atomic position and atom types (chemical species) including all metadata with contextualization (e.g. sample preparation, electric field) (FAIRmat)
  • Close interaction with device software via open-source parsers and converters
  • Machine learning/statistical analysis of APT atomic data to detect and interpret defect structures
  • Interface between formats and queries of (meta)data in FAIRmat and NFDI-MatWerk
  • Ontology matching between both communities
  • Workflow combining experimental data, image processing and theoretical models / properties of defects in microstructured materials (NFDI-MatWerk)

Related Participant Projects

Description

The interface to FAIRmat is most relevant for NFDI-MatWerk, since condensed-matter physics is the basic science perspective on solid-state phenomena. The resulting consequence for the infrastructure is explored in this IUC. FAIRmat focuses their effort on fundamental chemo-physical effects. In the case of atom probe tomography (APT), for example, the correlation of details of the electric field and the electronic density in the sample needs to be analyzed and exploited in order to predict the emission of atoms from steps at the surface. Such predictions are the key to arrive at more accurate and precise predictions of the ion launch and trajectories; and thereby are a route to improve current methods for backing out atomic positions. In contrast, NFDI-MatWerk addresses the relevance of the microstructure for engineering materials and their applications. Therefore, the focus of NFDI-MatWerk is on the correlation of atomic positions and defect states as a function of the chemical potential and the sample history (processing, heat treatment). For both perspectives, several gigabyte of data needs to be handled and needs to be evaluated with machine learning approaches.