Dual-Functional Optoelectronic Synaptic Device Based on MoTe₂/h-BN Transistor Through UV Light Induced Doping

Area of Science:

• Materials Science
• Computer Engineering
• Neuroscience

Background:

• Neuromorphic computing aims to enhance computational efficiency by mimicking brain architecture.
• Von Neumann architectures face limitations in computational efficiency.
• Synaptic devices are key components for brain-inspired computing.

Purpose of the Study:

• To propose a dual-functional optoelectronic synaptic device integrating synaptic and logic operations.
• To investigate the device's performance in neural network applications and logic gate functions.
• To explore UV light-induced doping for versatile plasticity modulation.

Main Methods:

• Fabrication of a three-terminal MoTe2/h-BN transistor.
• Characterization of synaptic behavior, including stability and repeatability.
• Implementation in a virtual three-layered neural network for image classification.
• Testing of optoelectronic logic gate functionalities (AND, OR, XOR).

Main Results:

• The device exhibits stable and repeatable synaptic behavior.
• Achieved 95.4% accuracy (n-type) and 94.2% accuracy (p-type) in handwritten digit classification.
• Demonstrated functionality as AND, OR, and XOR optoelectronic logic gates.
• UV light doping enabled switching between n- and p-type modes for tailored plasticity.

Conclusions:

• The proposed MoTe2/h-BN device offers a promising platform for hybrid neuromorphic systems.
• Integration of synaptic and logic functions enhances data processing efficiency.
• This advancement contributes to the development of more efficient brain-inspired computing architectures.