Compact Model for Bipolar and Multilevel Resistive Switching in Metal-Oxide Memristors.

Area of Science:

• Materials Science
• Electrical Engineering
• Computational Physics

Background:

• Memristive devices based on thin metal oxide films offer promising applications in neuromorphic computing and non-volatile memory.
• Accurate modeling is crucial for understanding and optimizing memristor performance, including switching dynamics and parameter variations.

Purpose of the Study:

• To develop and validate a combined circuitry (compact) model for metal oxide film memristive elements.
• To simulate bipolar switching and multilevel conductivity tuning, accounting for statistical parameter variability.

Main Methods:

• Development of an equivalent circuit and system of equations for the memristive element.
• Software implementation of the model using MATLAB.
• Simulation of static current-voltage characteristics and transient switching processes.

Main Results:

• The model successfully simulates bipolar switching and multilevel conductivity tuning of memristors.
• Statistical variability (device-to-device and cycle-to-cycle) is incorporated into the simulations.
• Simulated results show good agreement with experimental data for TiOₓ and TiO₂/Al₂O₃ based memristors.

Conclusions:

• The proposed compact model provides a robust framework for simulating metal oxide memristive devices.
• The model's ability to capture parameter variability is essential for realistic device design and fabrication.
• Validation against experimental data confirms the model's accuracy and utility for TiOₓ and TiO₂/Al₂O₃ memristors.