Enhanced Optoelectronic Synaptic Performance in Sol–Gel Derived Al-Doped ZnO Thin Film Devices
We report the fabrication and characterization of Al-doped ZnO (AZO) optoelectronic synaptic devices based on sol-gel-derived thin films with varying Al concentrations (0~4.0 wt%). Structural and optical analyses reveal that moderate Al doping modulates the crystal orientation, optical bandgap, and defect levels of ZnO films. Notably, 2.0 wt% Al doping yields the widest bandgap (3.31 eV), stable PL emission, and uniform deep-level absorption without inducing significant lattice disorder. Synaptic performance, including learning-forgetting dynamics and persistent photoconductivity (PPC), is strongly dependent on Al concentration. The 2.0 wt% AZO device exhibits the lowest forgetting rate and longest memory retention due to optimized trap formation, particularly Al-oxygen vacancy complexes that enhance carrier lifetime. Visual memory simulations using a 3 × 3 pixel array under patterned UV illumination further confirm superior long-term memory (LTM) behavior at 2.0 wt%, with stronger excitatory postsynaptic current (EPSC) retention during repeated stimulation. These results demonstrate that precise doping control via the sol-gel method enables defect engineering in oxide-based neuromorphic devices. Our findings provide an effective strategy for designing low-cost, scalable optoelectronic synapses with tunable memory characteristics suitable for future in-sensor computing and neuromorphic vision systems.
Authors
- Dabin Jeon
- Sung‐Nam Lee (ORCID: https://orcid.org/0000-0002-4800-5199)
- Seung Hun Lee (ORCID: https://orcid.org/0000-0002-2337-7826)
Institutions
- Korea University of Technology and Education (KR)
Publication Details
- Journal
- Materials
- Published
- 2025-06-20
- DOI
- https://doi.org/10.3390/ma18132931
- Citations
- 4
- Primary Topic
- Advanced Memory and Neural Computing
- Type
- article
- Field-Weighted Citation Impact
- 2.92
Funders
- Ministry of SMEs and Startups
- Ministry of Science and ICT, South Korea