Imitation of a Dual-Modal Synapse Based on a Hf0.5Zr0.5O2 Ferroelectric Tunnel Junction for Neuromorphic Computing

Abstract

Bioinspired artificial synapses have attracted considerable attention in the field of neuromorphic computing. Hf_0.5Zr_0.5O₂ thin films exhibit robust ferroelectricity even at a thickness of less than 10 nm, which contributes to energy efficiency and fast write/read speed. In this work, we report a low-cost and low-energy consumption dual-modal synapse with both excitation and inhibition in a Hf_0.5Zr_0.5O₂ ferroelectric tunnel junction. This transition is controlled by ferroelectric polarization switching and electron trapping by adjusting pulse duration. Furthermore, its excitation effect can simulate logic operations and pain perception nociceptors. The combined excitation and inhibition effects can simulate the dual-pulse Bienenstock–Cooper–Munro learning rule and image edge recognition. This contributes to the development of in-memory computing, image processing, and analog perception systems.