Visible-light modulated ferroelectric optoelectronic field-effect transistor for superior neuromorphic computing.

Area of Science:

• Materials Science
• Neuroscience
• Electrical Engineering

Background:

• High-performance neuromorphic computing and bionic vision demand simpler, low-power devices.
• Traditional oxide semiconductors are limited to ultraviolet light, hindering visible-light applications.

Purpose of the Study:

• To develop a visible-light-modulated ferroelectric optoelectronic field-effect transistor (Fe-OFET) for neuromorphic synapse simulation and multilevel storage.
• To improve visible-light response and absorption efficiency in optoelectronic devices.

Main Methods:

• An all-solution method was used to prepare a TiO2/IGZO/Al2O3/PZT based Fe-OFET.
• The device's optoelectronic performance and ferroelectric non-volatility were characterized.
• Synaptic plasticity functions like excitatory postsynaptic current (EPSC) and paired-pulse facilitation (PPF) were simulated.

Main Results:

• The TiO2 layer extended the device's light current response to the red light band, improving overall visible-light absorption.
• Stable ferroelectric non-volatility and optoelectronic performance were achieved.
• High accuracies were obtained in neuromorphic tasks: 95.5% for target motion recognition (CNN) and 90.0% for MNIST recognition (hybrid ANN with MHA).

Conclusions:

• The developed Fe-OFET offers a process-simple, visible-light-compatible solution for neuromorphic hardware.
• This work provides significant support for integrating visual neuromorphic hardware and computing.