Ultraviolet Optoelectronic Synapse Based on AlScN/p-i-n GaN Heterojunction for Advanced Artificial Vision Systems

Area of Science:

• Semiconductor research
• Optoelectronics
• Ferroelectric materials

Background:

• Ferroelectric materials offer novel optoelectronic potential.
• Aluminum Scandium Nitride (AlScN) is a promising ferroelectric semiconductor.
• Limited light sensitivity hinders multi-state optical responders.

Purpose of the Study:

• To overcome light sensitivity limitations in ferroelectric semiconductors.
• To fabricate a two-terminal AlScN/p-i-n GaN heterojunction ultraviolet optoelectronic synapse.
• To leverage hole capture for multi-state modulation.

Main Methods:

• Fabrication of a two-terminal AlScN/p-i-n GaN heterojunction.
• Utilizing hole capture at the AlScN/p-GaN hetero-interface.
• Characterization of memristor and synaptic characteristics.

Main Results:

• Achieved an on/off ratio of 9.36 × 10^5.
• Demonstrated synaptic behavior for artificial vision.
• Attained 93.7% image recognition accuracy with 0.26 weight nonlinearity.

Conclusions:

• The AlScN/GaN heterojunction overcomes light sensitivity limitations.
• The device exhibits excellent memristor and synaptic properties.
• Paves the way for advanced artificial vision systems and on-chip computing.