Broadband Neuromorphic Phototransistors Based on Oxygen Vacancy Modulation in Indium-Gallium-Zinc Oxide Films

Area of Science:

• Materials Science
• Neuroscience
• Optoelectronics

Background:

• Optoelectronic neuromorphic devices show promise for artificial vision but struggle with broadband photodetection, especially in the near-infrared (NIR) spectrum.
• Achieving broadband detection often requires complex heterostructures, hindering simplicity, scalability, and stability.

Purpose of the Study:

• To develop a simple and effective method for broadband photosensing and synaptic functionalities in neuromorphic devices.
• To engineer indium-gallium-zinc oxide (IGZO) thin films by controlling oxygen vacancies to enhance performance.

Main Methods:

• Fabrication of IGZO thin films with engineered oxygen vacancy concentrations.
• Characterization of phototransistor performance across visible and NIR spectra.
• Evaluation of synaptic emulation capabilities (short-term memory, long-term memory, paired-pulse facilitation) using NIR light.
• Implementation of an artificial neural network using device data for handwritten digit classification.

Main Results:

• Engineered IGZO phototransistors demonstrated broadband detection from visible to NIR (850 nm).
• Oxygen vacancy engineering significantly boosted NIR photosensitivity (from 9.17 to 244.8 A W⁻¹).
• Devices successfully mimicked synaptic behaviors under NIR stimulation, achieving 90.38% accuracy on MNIST dataset with an artificial neural network.

Conclusions:

• Oxygen-vacancy-engineered IGZO phototransistors offer a robust and scalable solution for broadband neuromorphic vision.
• The approach simplifies device fabrication while enhancing performance, particularly in the crucial NIR range.
• These devices are suitable for developing low-power, compact, and efficient neuromorphic vision systems.