Reaching Bio-Voltages and Controlling Synaptic Dynamics in Liquid-Based Neuromorphic Devices.

Area of Science:

• Neuromorphic Engineering
• Materials Science
• Neuroscience

Background:

• Neuromorphic resistive switching devices aim to emulate biological synapses and neurons.
• Liquid switching media offer soft, flexible, and brain-like characteristics for artificial systems.
• Existing artificial synapses often require an electroforming step and lack versatile behavior.

Purpose of the Study:

• To develop copper solution-based artificial synapses with nonvolatile and volatile, excitatory and inhibitory behaviors.
• To investigate the impact of different copper sulfate solutions, concentrations, electrode materials, and spacings.
• To explore the potential for bio-voltage neuromorphic memristor-based interfaces.

Main Methods:

• Fabrication of artificial synapses using copper sulfate solutions with varying concentrations.
• Testing device performance with different electrode materials and spacings in aqueous and glyceline-based media.
• Characterization of synaptic behaviors including potentiation, depression, nonvolatile, and volatile dynamics.

Main Results:

• Copper solution-based artificial synapses demonstrated nonvolatile and volatile, excitatory and inhibitory behavior without electroforming.
• Aqueous solutions achieved low operation voltage with high endurance and data retention.
• Transitioning to glyceline solvation induced a switch between nonvolatile and volatile dynamics, enhancing stability and showing potentiation/depression in volatile devices.

Conclusions:

• Copper solution-based artificial synapses offer a versatile platform for neuromorphic computing.
• The ability to exhibit both volatile and nonvolatile synaptic plasticity in a single device is a significant advancement.
• These findings hold promise for the development of bio-voltage neuromorphic memristor-based interfaces.