High-Performance Memristive Synapse Based on Space-Charge-Limited Conduction in LiNbO₃

Area of Science:

• Materials Science
• Neuroscience
• Computer Engineering

Background:

• Neuromorphic computing aims to mimic the human brain's structure and function.
• Developing efficient and versatile synaptic devices is crucial for advancing neuromorphic technology.
• Existing synaptic devices face challenges in performance and mimicking biological plasticity.

Purpose of the Study:

• To fabricate a high-performance memristive synapse for neuromorphic computing.
• To investigate the oxygen vacancy-mediated valence charge migration (V_O-VCM) mechanism in synaptic devices.
• To emulate key synaptic functions and assess the device's potential for pattern recognition.

Main Methods:

• Fabrication of an Al/LiNbO3/Pt memristive synapse.
• Utilizing voltage-controlled V_O-VCM to induce space-charge-limited conduction and self-rectifying hysteresis.
• Modulating applied voltage pulse parameters (polarity, amplitude, width, interval) to tune device characteristics.
• Emulating synaptic functions like short-term memory, long-term memory, and spike-time-dependent plasticity.
• Conducting simulation studies for hand-written image pattern recognition.

Main Results:

• The Al/LiNbO3/Pt memristive synapse exhibited voltage-controlled V_O-VCM, leading to space-charge-limited conduction and asymmetric hysteresis.
• The device demonstrated voltage pulse-tunable multi-state memory characteristics.
• Successful emulation of short-term memory, dynamic range-tunable long-term memory, and spike-time-dependent synaptic plasticity.
• High accuracy (up to 95.2%) achieved in simulation studies for hand-written image pattern recognition.

Conclusions:

• The V_O-VCM mechanism is effective for controlling memristive synapse behavior.
• The fabricated Al/LiNbO3/Pt memristive synapse shows promise for brain-inspired neuromorphic computing.
• The device's ability to emulate synaptic functions and perform pattern recognition highlights its potential.