Graphene oxide-based bioinspired neuromorphic transistors with artificial synaptic plasticity

Abstract

The explosive growth of information and the emergence of artificial intelligence have sparked significant interest in brain-like computing devices due to their unique properties in today's society. Transistor-based three-terminal artificial synaptic devices are anticipated to overcome the limitations of current von Neumann architectures in the realm of neural computing. In this study, we propose a graphene oxide synaptic transistor that demonstrates pronounced bipolar characteristics and exhibits a low threshold voltage of −0.84 V in p-type devices. The device demonstrates neurosynaptic behavior, providing evidence for the presence of excitatory postsynaptic current and its dependence on pulse amplitude and width. Furthermore, successful simulations have also been conducted to model paired-pulse facilitation and depression. The spike voltage-dependent plasticity was observed, and the device exhibited favorable repeatability in terms of short-term plasticity, encompassing both short-term potentiation and short-term depression. The present study showcases the viability of synaptic transistors based on graphene oxide, thereby opening up new avenues in the realm of neural computing.