Oxygen Vacancy-Induced Synaptic Plasticity in InAlZnO Nanofiber Transistors for Low-Power Neuromorphic Electronics

Abstract

As an alternative for the future computer, the brain-like neuromorphic computing systems have been widely concerned for the characteristics of power efficiency, self-learning, and parallel computing. Meanwhile, the simulation of the synaptic behavior based on electronic devices has attracted widespread attention in past decades. In this report, indium-aluminum–zinc-oxide (InAlZnO) nanofibers were prepared by electrospinning, and the synaptic transistors based on InAlZnO nanofibers were integrated and investigated. With the excitatory response modes, the basic biological functions, including excitatory postsynaptic current (EPSC), paired-pulse facilitation (PPF), high-pass filtering properties, short-term memory (STM), and long-term memory (LTM), were simulated by the synaptic transistor. It is found that low-power consumption (~75 fJ/spike) and nonvolatility of the channel conductance for the synaptic transistor based on the InAlZnO nanofibers have been achieved. The pattern recognition training based on the long-term potentiation (LTP)/depression characteristics was developed by CrossSim simulation platform, and a recognition accuracy of 95.6% was achieved by using the Modified National Institute of Standards and Technology database. This work demonstrates a new approach for the establishment of the large-scale, energy-efficient artificial synapses for neuromorphic computing.