We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:


Current browse context:


Change to browse by:

References & Citations


(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Condensed Matter > Mesoscale and Nanoscale Physics

Title: Deep learning of interface structures from the 4D STEM data: cation intermixing vs. roughening

Abstract: Interface structures in complex oxides remain one of the active areas of condensed matter physics research, largely enabled by recent advances in scanning transmission electron microscopy (STEM). Yet the nature of the STEM contrast in which the structure is projected along the given direction precludes separation of possible structural models. Here, we utilize deep convolutional neural networks (DCNN) trained on simulated 4D scanning transmission electron microscopy (STEM) datasets to predict structural descriptors of interfaces. We focus on the widely studied interface between LaAlO3 and SrTiO3, using dynamical diffraction theory and leveraging high performance computing to simulate thousands of possible 4D STEM datasets to train the DCNN to learn properties of the underlying structures on which the simulations are based. We validate the DCNN on simulated data and show that it is possible (with >95% accuracy) to identify a physically rough from a chemically diffuse interface and achieve 85% accuracy in determination of buried step positions within the interface. The method shown here is general and can be applied for any inverse imaging problem where forward models are present.
Comments: 18 pages, 4 figures
Subjects: Mesoscale and Nanoscale Physics (cond-mat.mes-hall); Computational Physics (physics.comp-ph); Data Analysis, Statistics and Probability (physics.data-an)
Cite as: arXiv:2002.09039 [cond-mat.mes-hall]
  (or arXiv:2002.09039v1 [cond-mat.mes-hall] for this version)

Submission history

From: Rama K Vasudevan [view email]
[v1] Thu, 20 Feb 2020 22:11:39 GMT (618kb)

Link back to: arXiv, form interface, contact.