Bio-Inspired Photosensory Artificial Synapse Based on Functionalized Tellurium Multiropes for Neuromorphic Computing

  • Adila Rani
  • , M. Junaid Sultan
  • , Wanqi Ren
  • , Atanu Bag
  • , Ho Jin Lee
  • , Nae Eung Lee
  • , Tae Geun Kim

Research output: Contribution to journalArticlepeer-review

Abstract

Nanomaterials like graphene and transition metal dichalcogenides are being explored for developing artificial photosensory synapses with low-power optical plasticity and high retention time for practical nervous system implementation. However, few studies are conducted on Tellurium (Te)-based nanomaterials due to their direct and small bandgaps. This paper reports the superior photo-synaptic properties of covalently bonded Tellurium sulfur oxide (TeSOx) and Tellurium selenium oxide (TeSeOx)nanomaterials, which are fabricated by incorporating S and Se atoms on the surface of Te multiropes using vapor deposition. Unlike pure Te multiropes, the TeSOx and TeSeOx multiropes exhibit controllable temporal dynamics under optical stimulation. For example, the TeSOx multirope-based transistor displays a photosensory synaptic response to UV light (λ = 365 nm). Furthermore, the TeSeOx multirope-based transistor exhibits photosensory synaptic responses to UV–vis light (λ = 365, 565, and 660 nm), reliable electrical performance, and a combination of both photodetector and optical artificial synaptic properties with a maximum responsivity of 1500 AW−1 to 365 nm UV light. This result is among the highest reported for Te-heterostructure-based devices, enabling optical artificial synaptic applications with low voltage spikes (1 V) and low light intensity (21 µW cm−2), potentially useful for optical neuromorphic computing.

Original languageEnglish
Article number2310013
JournalSmall
Volume20
Issue number31
DOIs
StatePublished - 1 Aug 2024

Keywords

  • defected TeSO and TeSeO structure
  • photo-synaptic devices
  • Te multiropes

Fingerprint

Dive into the research topics of 'Bio-Inspired Photosensory Artificial Synapse Based on Functionalized Tellurium Multiropes for Neuromorphic Computing'. Together they form a unique fingerprint.

Cite this