Synapse Neurotransmitter Channel-Inspired AlO_x Memristor with "V" Type Oxygen Vacancy Distribution

Area of Science:

• Materials Science
• Neuroscience
• Computer Engineering

Background:

• Memristors offer potential for neuromorphic computing but face challenges in consistency and power consumption.
• Biological synapses provide a low-cost, accurate model for brain-like computing.
• Developing artificial synaptic devices is crucial for advancing neuromorphic systems.

Purpose of the Study:

• To develop a memristor device that simulates biological synapses for efficient neuromorphic computing.
• To address memristor consistency and power consumption issues using a novel device structure.
• To investigate the potential of memristors in multi-value storage and high-accuracy recognition tasks.

Main Methods:

• Fabrication of a five-layer AlO_x memristor device with a V-shaped oxygen distribution.
• Simulation of synaptic functions including Long-term Potentiation/Depression (LTP/LTD) and Spike Timing Dependent Plasticity (STDP).
• Electrical characterization of the device for power consumption, switching speed, and nonlinearity.

Main Results:

• The AlO_x device successfully simulated biological synapse structure and function, including ion flow and neurotransmitter release.
• Achieved low power consumption (56.7 pJ/392 fJ) and high switching speed (25 ns/60 ns).
• Demonstrated high recognition accuracy (98.18%) in neuromorphic simulations with improved nonlinearity.

Conclusions:

• The developed AlO_x memristor device shows significant promise for neuromorphic computing applications.
• The device effectively mimics synaptic behavior, offering a pathway to overcome current memristor limitations.
• This work highlights the potential of memristors for multi-value storage and efficient, accurate brain-inspired computing.