Design of 3-D Interface in Synaptic Transistors With Attojoule Power by Electrospinning Aligned Nanofibers for Neuromorphic Computing

Neuromorphic devices, which are modeled after the human brain, have garnered increasing attention for their remarkable computing speed and high parallel computing capabilities. While electrolyte gate-controlled transistors have been previously reported to simulate synaptic function, their layered structure posed a significant challenge to the modulation of the channel layer. To address this limitation, this study proposes a novel electrolyte-gated electric double layer (EDL) synaptic transistor with 3-D interface contacts. This device utilizes aligned InMgO nanofibers as the channel layer, and doping lithium into the organic solid electrolyte to enhance the EDL effect. The aligned structure further optimizes synaptic properties. Notably, the device's power consumption is remarkably low, at 101 aJ, due to the 3-D interface and double layer effect. Through successful simulation of several crucial synaptic behaviors, including signal filtering characteristics, Ebbinghaus forgetting curve, and its application to image memory, these findings provide significant insights into the future development of complex neuromorphic systems.