Realization of Artificial Nerve Synapses Based on Biological Threshold Resistive Random Access Memory

Area of Science:

• Materials Science
• Electronics Engineering
• Biotechnology

Background:

• One-selector one-resistor (1S1R) arrays are crucial for high-density storage and neuromorphic computing.
• Existing selectors suffer from low durability and poor consistency, hindering practical applications.

Purpose of the Study:

• To fabricate and characterize a novel selector device utilizing egg albumen (EA) for 1S1R array applications.
• To evaluate the performance of EA-based selectors in conjunction with EA-based resistive random-access memory (RRAM) for improved storage density and neuromorphic computing.

Main Methods:

• Fabrication of a selector device using egg albumen (EA).
• Characterization of the EA-based selector's electrical properties, including threshold switching, leakage current, ON/OFF ratio, and endurance.
• Integration of the EA selector with an EA-based RRAM to form a 1S1R unit for crossbar array applications.

Main Results:

• The EA-based selector demonstrated excellent bidirectional threshold switching characteristics.
• Achieved a low leakage current (10^-7 A), a high ON/OFF current ratio (10⁶), and remarkable endurance (>700 days).
• The 1S1R unit effectively mitigated leakage current issues in crossbar arrays and exhibited synapse-like behavior.

Conclusions:

• Egg albumen is a viable material for fabricating high-performance selectors for 1S1R arrays.
• The protein-based 1S1R array offers a feasible solution for high-density storage and neuromorphic computing.
• The developed technology holds significant potential for simulating brain functions for advanced computational tasks.