Artificial Photonic Hetero-synapses Based on ZnO/IGZO Heterojunction for Neuromorphic Computing

Artificial intelligence technology has made tremendous breakthroughs in recent years. Correspondingly, this technology brings high requirements to the task-processing ability of computers. Currently, the emergence of neuromorphic devices has attracted great attention and is seen as a potential candidate to break the traditional computer architecture. In this study, we develop an artificial optoelectronic synapse based on ZnO/IGZO heterojunction using a magnetron sputtering method. The proposed device demonstrates the capability of analog resistive switching behaviors as well as optoelectronic synaptic plasticity properties. Further, the demonstrated optoelectronic synaptic plasticity can effectively emulate hetero-synaptic behavior. This work provides an efficient approach to optoelectronic neuromorphic computing and biomimetic electronics.