Loading…

Optimizing short-term memory in transparent neuromorphic devices via microwave treatment: Enhancing transmittance and relaxation time

It is crucial to guarantee the visibility and attain a suitable relaxation time in order to implement reliable short-term memory (STM) for transparent neuromorphic devices. In this work, we demonstrated fully transparent and reliable STM devices via microwave treatment (MWT), which offers the benefi...

Full description

Saved in:
Bibliographic Details
Published in:Journal of alloys and compounds 2025-01, Vol.1010, Article 178036
Main Authors: Jang, Yuseong, Lee, Doowon, Chae, Myoungsu, Kim, Taegi, Kim, Hee-Dong
Format: Article
Language:English
Subjects:
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:It is crucial to guarantee the visibility and attain a suitable relaxation time in order to implement reliable short-term memory (STM) for transparent neuromorphic devices. In this work, we demonstrated fully transparent and reliable STM devices via microwave treatment (MWT), which offers the benefits of reduced stress and high efficiency. It was found that MWT improves transmittance in the visible light region and increases the relaxation time of the device. As a result of optimization of relaxation time in MWT process with 1000 W/120 s, the most optimal relaxation time margin of 0–6.95 s could be secured. Furthermore, we experimentally implemented a reservoir computing (RC) system that consisted of 16 states that can be precisely classified with 4 bits and 16 states, which thereby proved the RC system’s viability. Based on these results, we confirmed that MWT is an effective annealing method that can contribute to the optimization of the relaxation time in the STM device research. •Memristor-based short-term memory for reservoir computing (RC) system to conserve energy.•Microwave treatment (MWT) process: more energy efficient than conventional annealing method.•Enhancing transmittance and relaxation time by MWT.•Experimentally implemented RC system with 16 states, classifiable with 4 bits and 16 states.
ISSN:0925-8388
DOI:10.1016/j.jallcom.2024.178036