A Modified SiO₂-Based Memristor with Reliable Switching and Multifunctional Synaptic Behaviors

Area of Science:

• Materials Science
• Nanotechnology
• Electrical Engineering

Background:

• Dielectric silicon dioxide (SiO2) is promising for emerging memory devices due to its high on/off ratio and low operating voltage.
• However, SiO2-based memristors relying on conducting filaments suffer from limited endurance and stability.

Purpose of the Study:

• To improve the endurance and stability of SiO2-based memristors.
• To explore the potential of these modified memristors for neuromorphic computing applications.

Main Methods:

• Constructed a passivated layer of SiO2 using Ag-doped SrTiO3 as a silver ion reservoir.
• Investigated filament formation control and memristor switching characteristics.
• Implemented and evaluated synaptic functions and a 7x7 pixel array for learning behavior.

Main Results:

• The modified memristor demonstrated excellent endurance and stable operation in ambient conditions for 20 days without degradation.
• Successfully implemented synaptic functions including excitatory postsynaptic current and memory transition.
• A 7x7 memristor array mimicked simple learning and forgetting behaviors.

Conclusions:

• Ag-doped SrTiO3 passivation offers a viable approach to enhance SiO2-based memristors.
• The improved memristors show potential for robust neuromorphic computing applications.