Related Experiment Video
Updated: Nov 20, 2025
07:40
Using Affordable LED Arrays for Photo-Stimulation of Neurons
Published on: November 15, 2011
18.9K
High photosensitivity light-controlled planar ZnO artificial synapse for neuromorphic computing.
Wei Xiao1, Linbo Shan1, Haitao Zhang1
1Key Laboratory of Special Function Materials & Structure Design of the Ministry of Education, School of Physical Science & Technology, Lanzhou University, Lanzhou 730000, China. wangqi77@lzu.edu.cn.
Nanoscale
|January 20, 2021
Summary
This study presents a highly photosensitive artificial synapse using ZnO thin film, achieving high accuracy in neuromorphic computing tasks. The device successfully emulates brain functions, paving the way for advanced computing systems.
Area of Science:
- Materials Science
- Neuroscience
- Computer Engineering
Background:
- Neuromorphic computing systems require efficient artificial synapses for high-bandwidth, low-power computation.
- Existing synapse devices often suffer from low photosensitivity, limiting accuracy in recognition and classification tasks.
- Artificial synapses mimic human brain functions, crucial for developing advanced computing architectures.
Purpose of the Study:
- To develop a planar light-controlled artificial synapse with high photosensitivity.
- To emulate various synaptic functions of the human brain.
- To demonstrate the device's effectiveness in a neural network for classification tasks.
Main Methods:
- Fabrication of a ZnO thin film using radiofrequency sputtering.
- Characterization of synaptic functions including memory, plasticity, and learning behaviors.
- Implementation in a three-layer neural network for classification using backpropagation.
Main Results:
- Achieved high photosensitivity (Ion/Ioff > 1000) with high photocurrent and low dark current.
- Successfully emulated key synaptic functions like memory, plasticity, and learning.
- Attained high classification accuracies of 90%, 92%, and 86% for different datasets.
- Identified oxygen vacancies and chemisorbed oxygen as critical performance factors.
Conclusions:
- The developed ZnO-based artificial synapse offers high photosensitivity and emulates complex brain functions.
- The device enables high-accuracy classification in neuromorphic computing applications.
- Understanding defect mechanisms in ZnO films is key to optimizing artificial synapse performance.
