Regensburg 2019 – wissenschaftliches Programm
Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
KFM: Fachverband Kristalline Festkörper und deren Mikrostruktur
KFM 10: Ferroics - Domains and Domain Walls
KFM 10.5: Vortrag
Dienstag, 2. April 2019, 10:30–10:45, H47
Deconvoluting conductance contributions at charged ferroelectric domain walls — •Theodor Secanell Holstad1, Donald Malcolm Evans1, Didrik René Småbråten1, Joshua Agar2, Stephan Krohns3, Zewu Yan4, Edith Bourret4, Sverre Magnus Selbach1, and Dennis Meier1 — 1Department of Materials Science and Engineering, Norwegian University of Science and Technology, Norway. — 2Department of Materials Science and Engineering, University of Lehigh, USA. — 3Center for Electronic Correlations and Magnetism, University of Augsburg, Germany. — 4Materials Sciences Division, Lawrence Berkeley National Laboratory, USA.
Ferroelectric domain walls are spatially mobile interfaces that naturally occur in materials that develop a spontaneous electric polarization. Because of their unique electronic properties, such walls hold great promise as functional 2D systems, but the characterization of their intrinsic transport properties remains a challenging task. Here, we combine scanning probe microscopy (SPM) with machine learning to gain new insight into the local electronic domain wall properties and to enhance the informational value of local conductance measurements. As a model case, we study the transport at ferroelectric domain walls in (Er0.99, Zr0.01)MnO3: At low voltages, head-to-head domain walls appear to be conducting. At higher voltages, however, the same domain walls are insulating. Using machine learning, we disentangle different conduction contributions and explain these seemingly contradictory results, highlighting how machine detection enhances the SPM output.