Related Experiment Video
Updated: Feb 28, 2026
08:15
Capturing Dynamic Finger Gesturing with High-resolution Surface Electromyography and Computer Vision
Published on: March 28, 2025
1.3K
Organic neuromorphic vision devices with multilevel memory for palmprint identification
Chenxi Liu1,2, Yongfeng Gu3, Yongjie Ren3
1Tianjin Key Laboratory of Molecular Optoelectronic Sciences, Department of Chemistry, Institute of Molecular Aggregation Science, Tianjin University Tianjin 300072 China jideyang@tju.edu.cn dingmy@tju.edu.cn.
Chemical Science
|February 26, 2026
Summary
Molecular engineering of neuromorphic vision devices using polar groups in dielectrics enhances performance. Polarity-engineered hafnium oxide phototransistors achieve high sensitivity and multilevel memory for intelligent vision systems.
Area of Science:
- Materials Science
- Neuroscience
- Electrical Engineering
Background:
- Neuromorphic visual devices offer a solution to the von Neumann bottleneck.
- The impact of interfacial molecular engineering, particularly polar groups in polymer dielectrics, on neuromorphic vision systems is not well understood.
Purpose of the Study:
- To investigate the role of interfacial molecular engineering in neuromorphic vision systems.
- To develop polarity-engineered hafnium oxide (HfO2)-based phototransistors with enhanced photodetection and memory capabilities.
Main Methods:
- Fabrication of HfO2-based phototransistors using polymer gate dielectrics with tuned polar functional groups (PPO and PVP).
- Characterization of device performance, including photoresponsivity, ON/OFF ratio, cycling endurance, retention time, and multilevel memory states.
- Integration of optimized devices with machine learning algorithms for biometric authentication.
Main Results:
- Engineered polar groups in dielectrics significantly enhanced photoresponsivity compared to low-polarity dielectrics.
- Achieved ultrahigh photodetection sensitivity and multilevel nonvolatile memory with 256 distinct conductance states (8-bit resolution).
- Demonstrated exceptional device performance: ON/OFF ratio > 10^5, endurance > 700 P/E cycles, retention time > 3x10^4 s.
- Enabled reliable biometric authentication with >98% accuracy using CASIA-palmprint database.
Conclusions:
- Molecular design principles for neuromorphic electronics were established.
- Presented an energy-efficient paradigm for vision systems integrating sensing, memory, and in situ processing.
- Paved the way for next-generation intelligent devices with unified optoelectronic functionalities.
