Neuromorphic computing targets realizing biomimetic or intelligence systems capable of processing abundant tasks in parallel analogously to our brain, and organic electrochemical transistors (OECTs) that rely on the mixed ionic–electronic synergistic couple possess significant similarity to biological systems for implementing synaptic functions. However, the lack of reliable stretchability for synaptic OECTs, where mechanical deformation occurs, leads to consequent degradation of electrical performance. Herein, we demonstrate stretchable synaptic OECTs by adopting a three-dimensional poly(3-hexylthiophene) (P3HT)/styrene-ethylene-butylene-styrene (SEBS) blend porous elastic film for neuromorphic computing. Such architecture shows the full capability to emulate biological synaptic behaviors. Adjusting the accumulated layer numbers of porous film enables tunable OECT output and hysteresis, resulting in transition in plasticity. Especially, with a trilayer porous film, large-scale conductance and hysteresis are endorsed for efficient mimicking of memory-dependent synapse behavior. Benefitted from the interconnected three-dimensional porous structures, corresponding stretchable synaptic OECTs exhibit excellent mechanical robustness when stretched at a 30% strain, and maintain reliable electrical characteristics after 500 stretching cycles. Furthermore, near-ideal weight updates with near-zero nonlinearities, symmetricity in long-term potentiation (LTP) and depression, and applications for image simulation are validated. This work paves a universal design strategy toward high-performance stretchable neuromorphic computing architecture and could be extended to other flexible/stretchable electronics.
Publications
- Article type
- Year
- Co-author
Article type
Year
Research Article
Issue
Nano Research 2023, 16(7): 10206-10214
Published: 24 April 2023
Downloads:184
Total 1