Van der Waals heterostructure-based flexible two-terminal ferroelectric field-effect transistors for in-sensor...

Area of Science:

• Materials Science
• Neuroscience
• Computer Engineering

Background:

• Artificial intelligence drives demand for efficient, bio-inspired computing systems.
• In-sensor optoelectronic computing leverages light for data processing within sensors.
• Flexible electronics are crucial for next-generation wearable and integrated devices.

Purpose of the Study:

• To propose and demonstrate a novel bio-inspired flexible optoelectronic synapse.
• To utilize a van der Waals (vdW) ferroelectric field-effect transistor (FeFET) for synaptic emulation.
• To explore the potential for in-sensor optoelectronic computing applications.

Main Methods:

• Fabrication of a flexible two-terminal optoelectronic synapse using a MoS2/h-BN/graphene/CuInP2S6 (CIPS) vdW FeFET with a metal-ferroelectric-metal-insulator-semiconductor (MFMIS) architecture.
• Modulation of the ferroelectric polarization state and synaptic plasticity using drain voltage (Vds) and light pulses.
• Simulation of an optoelectronic neural network utilizing an array of these vdW FeFET synapses.

Main Results:

• The vdW FeFET successfully emulated various neural synaptic functions.
• The device exhibited tunable synaptic plasticity modulated by electrical and optical stimuli.
• Simulated optoelectronic neural networks achieved image recognition accuracy exceeding 90%.

Conclusions:

• The developed flexible optoelectronic synapse shows significant promise for advanced in-sensor computing.
• The bio-inspired vdW FeFET architecture is a viable platform for neuromorphic engineering.
• This work paves the way for future flexible and efficient optoelectronic computing systems.