Related Experiment Video
Updated: Sep 17, 2025
10:58
A Procedure for Implanting Organized Arrays of Microwires for Single-unit Recordings in Awake, Behaving Animals
Published on: February 14, 2014
13.4K
In-material physical computing based on reconfigurable microwire arrays via halide-ion segregation
Dengji Li1, Pengshan Xie1, Yuekun Yang2,3
1Department of Materials Science and Engineering, City University of Hong Kong, Hong Kong SAR, 999077, P. R. China.
Nature Communications
|July 2, 2025
Summary
Researchers developed novel neuromorphic hardware using perovskite microwires. This self-powered system demonstrates non-volatile light storage and performs complex computing tasks, paving the way for energy-efficient physical computing.
Area of Science:
- Materials Science
- Condensed Matter Physics
- Neuroscience
Background:
- Conventional computers use silicon transistors, limiting processing capabilities.
- Emerging materials offer new avenues for brain-inspired neuromorphic hardware.
- Perovskites exhibit unique optoelectronic properties suitable for advanced computing.
Purpose of the Study:
- To investigate photoelectricity-induced halide-ion segregation in perovskite microwires.
- To demonstrate the potential of in-material dynamics for physical computing.
- To develop self-powered, non-volatile memory and processing elements.
Main Methods:
- Epitaxial growth of mixed-halide perovskite CsPbBr1.5I1.5 microwire networks.
- In-situ measurements to confirm photoelectricity-induced halide-ion segregation.
- Testing the microwire array on graphical neural network and image restoration tasks.
Main Results:
- Demonstrated stable and controllable halide-ion segregation and recovery.
- Achieved reconfigurable, self-powered photoresponse for non-volatile light storage.
- Successfully performed graphical neural network and image restoration tasks without external circuits.
Conclusions:
- In-material dynamics in perovskites enable highly parallel and energy-efficient physical computing.
- This approach offers a promising direction for post-Moore era computing architectures.
- The developed neuromorphic hardware showcases significant potential for advanced information processing.
