Unipolar modulated volatile memristor: Achieving multi-scale plasticity, associative learning, and dynamic reservoir computing

Read full text Summarise with VIRAT
Authors: Jie Lu, Kexiang Wang, Haoyu Li, Zeyang Xiang, Junxiu Zhou, Yun Wen, Junyu Wang, Xinyu Cao, Weitian Xun, Yegang Lu, Ran Jiang
Cited by: 0
References: 58

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Abstract The von Neumann architecture is nearing its physical limits regarding energy efficiency and parallelism. Consequently, brain-inspired computing hardware is viewed as a crucial solution to address these limitations. This research presents a volatile memristor featuring an Ag/Al2O3/Ti/Al2O3/ITO configuration for applications in brain-inspired computing. The device demonstrates conductivity modulation under unipolar pulses, attributed to the dynamic competition between electric field-driven drift and heat-induced diffusion, governed by the oxygen vacancy reservoir formed by the in situ oxidized Ti interlayer. The device exhibits multi-scale plasticity, encompassing short-term facilitation (STF), paired-pulse facilitation (PPF), and dependence on frequency and duty cycle. Switching between potentiation and depression can be achieved under the same polarity by tuning the strength of forward stimulation (amplitude, width, or interval). The device emulates the classical conditioned reflex of the spotted butterfly and establishes a dynamic reservoir, attaining accuracies of 97.05% and 83.22% in the MNIST and Fashion-MNIST image recognition tasks, with normalized mean square errors (NMSE) for the two second-order nonlinear system tasks of approximately 0.123 and 0.108, respectively. This study outlines a methodology for fabricating unipolar volatile memristors, thus enhancing the understanding of brain-inspired computing hardware.

Fields of Study & Subjects

Advanced Memory and Neural Computing Neural Networks and Reservoir Computing Ferroelectric and Negative Capacitance Devices

Publication details

Journal / source
Nano Research
Volume
19
Issue
7
Pages
94908641–94908641
Publication date
23 March 2026
Language
EN