Skip to content
DraconiticResearch
Search Sources API
203,234 studies31 sourcesUpdated 18 min ago
PreprintCC BY-NC-ND 4.0arXiv

Wavelength-selective bidirectional plasticity in self-powered Se/ZnMgO visual synapses

Abstract

Artificial visual synapses combine sensing, memory and processing in a single device, but most of them need an external voltage to read the synaptic weight and light is generally able to strengthen the weight rather than to weaken it. We report here Se/ZnMgO thin film photodiodes in which the photocurrent at zero bias writes and reads the synaptic weight, with a direction of change set by the photon energy alone. At equal photon flux, 365 nm pulses increase the photocurrent by 38 % while 455 nm pulses decrease it, and the depression reaching 30 % at higher flux, with up to 19 distinguishable levels per branch and no electrical energy spent. Spectrally resolved light soaking shows that short wavelengths potentiate while low-energy light erases. The interesting property is therefore the photon energy-dependent competition between potentiation and depression, the crossover being set by the defect population of the interface. Optical simulations show that both wavelengths are absorbed within the first tens of nanometres of Se, and the stored state is erased by a reverse bias in a voltage-dependent way and by sub-gap light, which indicates that potentiation corresponds to interface defects filled with electrons, in agreement with drift-diffusion simulations. The devices also show relearning, their response is modulated by bending on flexible substrates, and a network built from the measured weight updates reaches 89 % accuracy on handwritten digit recognition without noise and 85 % with the measured noise, against 97.5 % for floating-point weights.

Related

  1. RADICAL: Reconfiguration-Aware Direct-Connect AI Cluster using Spatial-and-Wavelength Selective Switching
    Brett George, Weiyang Wang, Yuyang Wang +5
    arXivPreprintarXiv licenceFull textOptics & photonicsarXiv
  2. Wavelength-selective second-harmonic splitting and focusing in engineered χ ⁽²⁾ photonic crystals
    Zongheng Cui, Tianxiang Xu, Shan Liu +2
    Optics ExpressPeer-reviewedCC BY 4.0Optics & photonicsCrossref
  3. Stochastic Resonance in a Small-World Network of Rulkov Neurons with Electrical and Chemical Synapses Under an Alternating Electric Field and Synaptic Transmission Delay.
    Chunhua Yuan, Yunrong Jiang, Junfeng Wang +1
    Biomimetics (Basel, Switzerland)Peer-reviewedCC BYFull textStatistical & nonlinear physicsEurope PMC
  4. Two-State Dynamical Synapses as a Physical Substrate for Equilibrium Propagation
    Dongshu Liu, Julie Grollier
    arXivPreprintarXiv licenceFull textApplied physics & instrumentationarXiv
  5. Generation of tunable wavelength coherent states and heralded single photons for quantum optics applications
    N. Bruno, Anthony Martin, R.T. Thew
    Optics CommunicationsPeer-reviewedLicence not statedFull text18 citedQuantum physics & quantum informationHAL
BibTeX
@misc{likao2026wavelengthselective,
  title = {{Wavelength-selective bidirectional plasticity in self-powered Se/ZnMgO visual synapses}},
  author = {Zacharie Jehl Li-Kao and Miho Akasaka and Myeongok Kim and Kunal J. Tiwari and Antoine Moore and Mykhailo A. Koltsov and Sergio Giraldo and Saiki Kitagawa and Marcel Placidi and Taizo Kobayashi},
  howpublished = {arXiv},
  year = {2026},
  month = oct,
  doi = {10.48550/arxiv.2610.21618},
  eprint = {2610.21618},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/2610.21618},
  note = {CC BY-NC-ND 4.0},
}
RIS
TY  - UNPB
TI  - Wavelength-selective bidirectional plasticity in self-powered Se/ZnMgO visual synapses
AU  - Zacharie Jehl Li-Kao
AU  - Miho Akasaka
AU  - Myeongok Kim
AU  - Kunal J. Tiwari
AU  - Antoine Moore
AU  - Mykhailo A. Koltsov
AU  - Sergio Giraldo
AU  - Saiki Kitagawa
AU  - Marcel Placidi
AU  - Taizo Kobayashi
PY  - 2026
DA  - 2026/10/15
PB  - arXiv
DO  - 10.48550/arxiv.2610.21618
UR  - https://arxiv.org/abs/2610.21618
AB  - Artificial visual synapses combine sensing, memory and processing in a single device, but most of them need an external voltage to read the synaptic weight and light is generally able to strengthen the weight rather than to weaken it. We report here Se/ZnMgO thin film photodiodes in which the photocurrent at zero bias writes and reads the synaptic weight, with a direction of change set by the photon energy alone. At equal photon flux, 365 nm pulses increase the photocurrent by 38 % while 455 nm pulses decrease it, and the depression reaching 30 % at higher flux, with up to 19 distinguishable levels per branch and no electrical energy spent. Spectrally resolved light soaking shows that short wavelengths potentiate while low-energy light erases. The interesting property is therefore the photon energy-dependent competition between potentiation and depression, the crossover being set by the defect population of the interface. Optical simulations show that both wavelengths are absorbed within the first tens of nanometres of Se, and the stored state is erased by a reverse bias in a voltage-dependent way and by sub-gap light, which indicates that potentiation corresponds to interface defects filled with electrons, in agreement with drift-diffusion simulations. The devices also show relearning, their response is modulated by bending on flexible substrates, and a network built from the measured weight updates reaches 89 % accuracy on handwritten digit recognition without noise and 85 % with the measured noise, against 97.5 % for floating-point weights.
ER  -