EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Mimicking Silent Synapse Recruitment: A SiO x /Cu‐Pancake Memristive Device For Analog Neuromorphic Computing

Rouven Lamprecht, Ole Gronenberg, Sahitya Yarragolla, Tobias Gergs, Richard Marquardt, Victor Witte, Jürgen Carstensen, George Popkirov, Jan Trieschmann, Lorenz Kienle, Hermann Kohlstedt

Advanced Functional Materials · 2026-08 · 36 (63)

DOI: 10.1002/adfm.77024 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Advanced Functional Materials, online erschienen am 11.07.2026. Verfasst von Rouven Lamprecht und 10 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 150 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Deutsche Forschungsgemeinschaft.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Analog memristive devices are essential for continuous synaptic weight updates in energy‐efficient neuromorphic computing. Achieving robust and predictable multi‐level conductance modulation requires precise physical control over the active switching volume to avoid stochastic instabilities. While various non‐filamentary switching concepts exist, structurally engineering and stabilizing a distributed nanoscale morphology to ensure reliable, area‐wide analog switching remains a fundamental materials challenge. To address this gap, we introduce an electroforming‐free TiN//Cu//TiN device that leverages the controlled thermal evolution of embedded copper into laterally extended, pancake‐like nanoparticles (Cu‐PCs). By combining in situ transmission electron microscopy with potential‐ and time‐resolved impedance spectroscopy, as well as area‐scaling analyses, we demonstrate that this unique morphology creates a highly heterogeneous Schottky barrier landscape. Applying an electrical bias drives local oxygen vacancy redistribution, dynamically lowering these interfacial barriers and progressively recruiting previously inactive, high‐resistance Cu‐PC pathways into a parallel conduction ensemble. This progressive “recruitment” provides an intrinsic physical basis for continuous analog tunability, functionally mimicking the activation of biological silent synapses and establishing targeted metal nanoparticle morphology engineering as a powerful paradigm for robust neuromorphic hardware.

Förderung

  • Deutsche Forschungsgemeinschaft

Zitieren

APA

Lamprecht, R., Gronenberg, O., Yarragolla, S., Gergs, T., Marquardt, R., Witte, V., Carstensen, J., Popkirov, G., Trieschmann, J., Kienle, L., & Kohlstedt, H. (2026). Mimicking Silent Synapse Recruitment: A SiO x /Cu‐Pancake Memristive Device For Analog Neuromorphic Computing. Advanced Functional Materials, 36(63). https://doi.org/10.1002/adfm.77024

Vancouver

Lamprecht R, Gronenberg O, Yarragolla S, Gergs T, Marquardt R, Witte V, et al. Mimicking Silent Synapse Recruitment: A SiO x /Cu‐Pancake Memristive Device For Analog Neuromorphic Computing. Advanced Functional Materials. 2026;36(63). doi:10.1002/adfm.77024

BibTeX
@article{Lamprecht2026,
  author = {Lamprecht, Rouven and Gronenberg, Ole and Yarragolla, Sahitya and Gergs, Tobias and Marquardt, Richard and Witte, Victor and Carstensen, Jürgen and Popkirov, George and Trieschmann, Jan and Kienle, Lorenz and Kohlstedt, Hermann},
  title = {Mimicking Silent Synapse Recruitment: A SiO x /Cu‐Pancake Memristive Device For Analog Neuromorphic Computing},
  journal = {Advanced Functional Materials},
  year = {2026},
  volume = {36},
  number = {63},
  doi = {10.1002/adfm.77024}
}
Kostenlos starten →