Enhanced Artificial Synaptic Properties Enabled by Arrays of Electrolyte-Gated Electrospun InZnO Nanowires Enhanced Artificial Synaptic Properties Enabled by Arrays of Electrolyte-Gated Electrospun InZnO Nanowires AITranslate

Basic Information:

DOI：https://doi.org/10.1021/acsaelm.2c00326

Chinese Library Classification Number:

Citation Information:

Multiterminal artificial synapses constitute a vibrant research field in the hardware of artificial intelligence. Arrays consisting of one-dimensional nanowire can render artificial synapses with enhanced electrical properties, exhibiting potential for highly compact circuit integration. Herein, we fabricated three-terminal electrolyte-gated synaptic transistors based on indium zinc oxide (InZnO) nanowire arrays (NWAs) via a low-cost electrospinning technique combined with a facile nanowire transfer process. The nanowire diameter can be as narrow as 40 nm, presenting an opportunity for low-cost compact circuit integration. Moreover, based on the synaptic transistors composed of InZnO NWAs and a LiClO4 electrolyte, we obtained figures of merit of artificial synapses, such as the excitatory postsynaptic current, paired-pulse facilitation, high-pass filtering capability, and long-term potentiation. Furthermore, we systematically compared the electrical performance of NWAs with the randomly aligned InZnO nanowire networks (NWNs), and the electrical characteristics of NWA synaptic transistors are superior compared with those of the NWN counterpart. In addition, an artificial neural network for handwritten digit recognition based on the experimentally measured potentiation/depression characteristics has been simulated with a recognition accuracy of 93.1%. Our study demonstrates the opportunity to realize low-power artificial synapses with aligned nanowire arrays obtained from cost-effective fabrication techniques.