Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
O: Fachverband Oberflächenphysik
O 57: Plasmonics and nanooptics: Light-matter interaction, spectroscopy II
O 57.4: Vortrag
Mittwoch, 11. März 2026, 11:30–11:45, HSZ/0403
Toward Plasmonic Neuronal Architectures at the Nanometer Scale — •Christopher Weiß1, Tobias Eul2, Emily Kruel1, Mario Pfeiffer1, Bert Lägel1, Benjamin Stadtmüller2, and Martin Aeschlimann1 — 1Department of Physics and Research Center OPTIMAS, RPTU University Kaiserslautern-Landau, Germany — 2Experimentalphysik II, Institute of Physics, University of Augsburg, Germany
Classical von-Neumann computers face severe energy and speed limitations when operating large artificial neural networks, motivating alternative computing concepts beyond the traditional architecture. Plasmonic nanostructures offer a promising route toward ultrafast and highly integrated neuromorphic systems by enabling strong optical confinement at the nanoscale. We present a concept for a plasmonic neuronal cell that combines multiplexed signal reception, static weighting, and nonlinear activation within a single device. Optical inputs encoded in the orbital angular momentum of light are directed into separate dielectric-loaded surface-plasmon-polariton waveguides. Synaptic weighting is implemented through nanoscale gaps, whose attenuation characteristics are analyzed using finite-difference time-domain simulations and experimentally validated by photoemission electron microscopy. Nonlinear activation is provided by plasmon-enhanced two-photon photoemission. These results establish the essential functional components required for ultrafast plasmonic neuromorphic architectures.
Keywords: Neuromorphic computing; Surface plasmon polaritons (SPPs); Orbital angular momentum (OAM) multiplexing; Plasmonic Dielectric-loaded SPP waveguides; Photoemission electron microscopy (PEEM)