Transition from synaptic simulation to nonvolatile resistive switching behavior based on an Ag/Ag:ZnO/Pt memristor

Area of Science:

• Materials Science
• Neuroscience
• Computer Engineering

Background:

• The development of brain-inspired computing, or neuromorphic computing, aims to create intelligent systems that mimic the human brain's structure and function.
• Memristors, electronic components with memory properties, are crucial for building artificial neural networks due to their ability to emulate synaptic plasticity.

Purpose of the Study:

• To fabricate and characterize ZnO-based memristors capable of performing synaptic functions.
• To investigate the resistive switching behavior and potential applications in neuromorphic computing and data storage.

Main Methods:

• Fabrication of ZnO-based memristors using radio frequency magnetron sputtering.
• Incorporation of silver nanocrystals (NCs) into the ZnO oxide layer.
• Testing of memristor devices for synaptic functions like potentiation, depression, and facilitation, as well as bipolar resistive switching.

Main Results:

• Successful imitation of several synaptic functions, including nonlinear transmission, short-term potentiation, long-term potentiation/depression, and pair-pulse facilitation.
• Observation of a transition from synaptic behaviors to bipolar resistive switching under repeated stimulation.
• Speculation of the switching mechanism involving the formation and rupture of conductive silver (Ag) filaments and electrochemical metallization.

Conclusions:

• The fabricated Ag/Ag:ZnO/Pt memristor exhibits both resistive switching and multiple synaptic behaviors.
• These memristors show significant potential for applications in advanced neuromorphic computing and efficient data storage systems.