Opto-Ferroelectric Coupling Enhanced α-In₂Se₃ Transistors With Light-Controlled Mode Switchover

Area of Science:

• Materials Science
• Electrical Engineering
• Computer Science

Background:

• Neuromorphic computing excels at pattern recognition but struggles with digital logic integration.
• Ferroelectric transistors offer unique properties but often lack compatibility with digital systems.

Purpose of the Study:

• To develop a transistor with light-controlled mode switching between digital logic and neuromorphic computing.
• To address the incompatibility between digital logic and neuromorphic computing units.

Main Methods:

• Utilized an opto-ferroelectric coupling effect in a ferroelectric transistor with α-In2Se3/h-BN dielectrics and MoS2 channel.
• Investigated transistor performance in both dark (digital logic) and light (optoelectronic synapse) conditions.
• Implemented inverter and NOR circuits and evaluated image classification accuracy and energy consumption.

Main Results:

• In dark conditions, the transistor demonstrated digital logic capabilities with a 10^9 on/off ratio, 33 mV/dec subthreshold swing, and 5.6 × 10^-13 A leakage current.
• Under light, the transistor switched to an optoelectronic synapse mode for neuromorphic computing, achieving 94.43% image classification accuracy.
• The device exhibited low energy consumption of 0.47 pJ/spike for neuromorphic tasks.

Conclusions:

• The developed opto-ferroelectric transistor successfully integrates digital logic and neuromorphic computing functionalities.
• This technology offers a promising solution for next-generation computing devices by bridging the gap between different processing modes.
• The light-controlled mode switchover enhances the versatility and applicability of the transistor in advanced computing systems.