Photo-Assisted Ferroelectric Domain Control for α-In₂Se₃ Artificial Synapses Inspired by Spontaneous Internal...

Area of Science:

• Materials Science
• Condensed Matter Physics
• Nanotechnology

Background:

• Artificial synapses are crucial for developing neuromorphic computing systems.
• Indium selenide (In2Se3) exhibits ferroelectric properties suitable for synaptic applications.
• Tuning ferroelectricity with electrical and optical stimuli offers diverse pathways for synaptic plasticity.

Purpose of the Study:

• To demonstrate and characterize synaptic memory behavior in α-In2Se3 using electrical gate voltages under white light.
• To investigate the role of optically induced internal electric fields in enhancing synaptic plasticity.
• To evaluate the performance of α-In2Se3-based artificial synapses in artificial learning simulations.

Main Methods:

• Fabrication of α-In2Se3/hexagonal boron nitride (hBN) heterostructure ferroelectric field-effect transistors (FeFETs).
• Application of electrical gate voltages and white light illumination to induce and modulate ferroelectricity.
• Characterization of synaptic behavior, including long-term depression (LTD) curves and recognition accuracy in simulations.

Main Results:

• An internal electric field was identified at the polarization-flipped conductance channel in α-In2Se3/hBN FeFETs under white light.
• The enhanced internal built-in electric field under illumination improved the linearity of LTD curves.
• Artificial learning simulations showed significantly elevated recognition accuracy for handwritten digits using the α-In2Se3/hBN FeFETs.

Conclusions:

• Optically assisted ferroelectricity in α-In2Se3 enhances synaptic plasticity and memory functions.
• The identified internal electric field plays a critical role in the improved synaptic characteristics.
• α-In2Se3-based FeFETs show promise for high-performance artificial learning applications.