Related Experiment Video
Updated: Aug 6, 2026
08:07
Assembly and Characterization of Biomolecular Memristors Consisting of Ion Channel-doped Lipid Membranes
Published on: March 9, 2019
Defect-Engineered Floating-Gate Synapses With Programmable Relaxation Dynamics for Multi-Timescale Neuromorphic
Chunyang Li1,2,3, Lu Li4, Zhongyi Li1,2,3
1Key Laboratory of Biomimetic Robots and Systems, Ministry of Education, Beijing Institute of Technology, Beijing, China.
Small (Weinheim an Der Bergstrasse, Germany)
|July 25, 2026
Summary
Researchers developed tunable artificial synapses that mimic the human brain
Area of Science:
- Neuromorphic Engineering
- Materials Science
- Artificial Intelligence
Background:
- The human brain's ability to process temporal information across vast timescales is crucial for cognition.
- Current artificial neuromorphic hardware struggles to replicate this multi-timescale processing capability.
- Hierarchical memory systems in the brain enable complex temporal information processing.
Purpose of the Study:
- To develop artificial synapses with programmably tunable relaxation dynamics for emulating multi-timescale cognition.
- To engineer two-dimensional (2D) material-based synaptic devices for neuromorphic applications.
- To create a physical reservoir computing framework capable of processing complex time-series data.
Main Methods:
- Precisely engineering defect density in 2D materials to modulate charge trapping/de-trapping kinetics.
- Developing synaptic transistors with tunable nonvolatile to volatile memory characteristics.
- Integrating devices with complementary timescales into a parallel dynamic memory superposition architecture.
- Implementing a reservoir computing framework for time-series data analysis.
Main Results:
- Achieved systematic modulation of synaptic device relaxation times across orders of magnitude.
- Demonstrated neuromorphic functionalities including paired-pulse facilitation and multilevel conductance states.
- The heterogeneous reservoir successfully extracted long-term and short-term patterns simultaneously without signal entanglement.
- Superior performance in multifrequency oscillator prediction tasks compared to single-timescale systems.
Conclusions:
- Defect engineering in 2D materials provides a viable method for programming temporal dynamics in neuromorphic systems.
- The developed synaptic devices and architecture offer a pathway toward energy-efficient edge intelligence.
- This approach enables adaptive real-world time-series processing for advanced AI applications.
