Biorealistic Neuronal Temperature-Sensitive Dynamics within Threshold Switching Memristors: Toward Neuromorphic...

Area of Science:

• Materials Science
• Neuroscience
• Electronics

Background:

• Neuromorphic computing aims to emulate biological neural systems.
• Temperature sensitivity in biological neurons is crucial for their function.
• Existing neuromorphic devices often lack biomimetic thermosensitivity.

Purpose of the Study:

• To investigate the biomimetic thermosensitive properties of V3O5 memristive devices.
• To emulate the firing characteristics of thermosensitive biological neurons using V3O5 devices.
• To develop a neuromorphic thermosensation system for bioinspired thermal perception.

Main Methods:

• Fabrication and characterization of two-terminal V3O5 memristive devices.
• Analysis of temperature-dependent electrical characteristics.
• Numerical simulations using the Morris-Lecar neuron model to compare dynamics.
• Development of a neuromorphic thermosensation system.

Main Results:

• V3O5 memristive devices exhibit biomimetic thermosensitive properties similar to biological neurons.
• The spiking response of V3O5-based relaxation oscillators mimics biological neuron firing.
• Simulations confirm the temperature-dependent dynamics align with biological neuron models.
• A robust neuromorphic thermosensation system was successfully demonstrated.

Conclusions:

• V3O5 memristors show significant potential for biorealistic emulation of neural dynamics.
• V3O5 is a promising functional material for solid-state neurons in neuromorphic applications.
• This work advances bioinspired robotics and in silico neuroscience through novel sensing capabilities.