Short-term and long-term synaptic plasticity in Ag/HfO2/SiO2/Si stack by controlling conducting filament strength

Abstract

This article presents diverse resistive switching properties in Ag/HfO2/SiO2/Si memristor device for neuromorphic computing. We demonstrate that two characteristics of long term and short-term memory can be implemented in a single device depending on the strength of the applied signal. The elemental analysis, chemical characterization, and accurate film thickness of the device stack is confirmed by transmission electron microscopy (TEM), energy-dispersive X-ray spectroscopy (EDS), and X-ray photoelectron spectroscopy (XPS). The volatile resistive switching and non-volatile switching are dependent on the strength of the conducting filament. The short-term memory with a current decay is conducted by the low compliance current (CC) and a pulse with small voltage amplitude. On the other hand, the long-term memory with non-volatile properties is obtained by high CC and a pulse with large voltage amplitude. Potentiation and depression characteristics with multi-states are achieved for long-term plasticity, which is verified by the retention.