Artificial Synapses with Short- and Long-Term Memory for Spiking Neural Networks Based on Renewable Materials

Summary

This paper introduces flexible biomemristor devices based on lignin, a renewable material, that emulate essential synaptic behaviors for use in artificial synapses and flexible electronic devices.

Key findings

• Lignin-based memristors emulate analog memory switching, short-term plasticity, long-term plasticity, spike-rate-dependent plasticity, and short-term to long-term transition.
• The devices are flexible and operate without degradation under mechanical bending.
• Lignin is an abundant, biocompatible, biodegradable, and environmentally benign material.

Methods

• Fabrication of flexible biomemristor devices using lignin via a simple solution process.
• Characterization of synaptic behaviors including analog memory switching and plasticity.
• Mechanical bending tests to assess device flexibility and durability.

Limitations

• The study focuses on emulation of biological synapses, not a full neuromorphic system.
• The long-term stability and performance in complex network environments are not detailed.

Applications

• Artificial synapses for spiking neural networks.
• Flexible electronic devices.
• Bioinspired neuromorphic systems.
• Energy-efficient AI hardware.