Enhancing memristor performance with 2D SnOx/SnS2 heterostructure for neuromorphic computing

IF 6.8 2区材料科学 Q1 MATERIALS SCIENCE, MULTIDISCIPLINARY

Science China Materials Pub Date : 2025-01-02 DOI:10.1007/s40843-024-3208-3

Yangwu Wu (, ), Sifan Li (, ), Yun Ji (, ), Zhengjin Weng (, ), Houying Xing (, ), Lester Arauz, Travis Hu, Jinhua Hong (, ), Kah-Wee Ang, Song Liu (, )

{"title":"Enhancing memristor performance with 2D SnOx/SnS2 heterostructure for neuromorphic computing","authors":"Yangwu Wu \n (, ), Sifan Li \n (, ), Yun Ji \n (, ), Zhengjin Weng \n (, ), Houying Xing \n (, ), Lester Arauz, Travis Hu, Jinhua Hong \n (, ), Kah-Wee Ang, Song Liu \n (, )","doi":"10.1007/s40843-024-3208-3","DOIUrl":null,"url":null,"abstract":"<div><p>Layered metal dichalcogenides (LMDs) neuromorphic memristor devices offer a promising alternative to conventional von Neumann architectures, addressing speed and energy efficiency constraints. However, challenges remain in controlling resistive switching and operating voltage in crystalline LMD memristors due to environmental stabilization issues, which hinder neural network hardware development. Herein, we introduce an optimization method for memristor operation by controlling oxidation through ozone treatment, creating a SnO<sub><i>x</i></sub>/SnS<sub>2</sub> resistive layer. These optimized memristors demonstrate low switching voltages (∼1 V), rapid switching speeds (∼20 ns), high switching ratios (10<sup>2</sup>), and the ability to emulate synaptic weight plasticity. Cross-sectional transmission electron microscopy and energy-dispersive X-ray spectroscopy identified defects and Ti conductive filaments in the resistive switching layer, contributing to uniform switching and minimized operating variation. The device achieved 90% accuracy in MNIST handwritten recognition, and hardware-based image convolution was successfully implemented, showcasing the potential of SnO<sub><i>x</i></sub>/SnS<sub>2</sub> memristors for neuromorphic applications.</p><div><figure><div><div><picture><source><img></source></picture></div></div></figure></div></div>","PeriodicalId":773,"journal":{"name":"Science China Materials","volume":"68 2","pages":"581 - 589"},"PeriodicalIF":6.8000,"publicationDate":"2025-01-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Science China Materials","FirstCategoryId":"88","ListUrlMain":"https://link.springer.com/article/10.1007/s40843-024-3208-3","RegionNum":2,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"MATERIALS SCIENCE, MULTIDISCIPLINARY","Score":null,"Total":0}

引用次数: 0

Abstract

Layered metal dichalcogenides (LMDs) neuromorphic memristor devices offer a promising alternative to conventional von Neumann architectures, addressing speed and energy efficiency constraints. However, challenges remain in controlling resistive switching and operating voltage in crystalline LMD memristors due to environmental stabilization issues, which hinder neural network hardware development. Herein, we introduce an optimization method for memristor operation by controlling oxidation through ozone treatment, creating a SnO_x/SnS₂ resistive layer. These optimized memristors demonstrate low switching voltages (∼1 V), rapid switching speeds (∼20 ns), high switching ratios (10²), and the ability to emulate synaptic weight plasticity. Cross-sectional transmission electron microscopy and energy-dispersive X-ray spectroscopy identified defects and Ti conductive filaments in the resistive switching layer, contributing to uniform switching and minimized operating variation. The device achieved 90% accuracy in MNIST handwritten recognition, and hardware-based image convolution was successfully implemented, showcasing the potential of SnO_x/SnS₂ memristors for neuromorphic applications.

查看原文本刊更多论文

利用二维SnOx/SnS2异质结构增强忆阻器性能，用于神经形态计算

层状金属二硫族化合物（LMDs）神经形态忆阻器器件为传统的冯·诺伊曼结构提供了一个有前途的替代方案，解决了速度和能源效率的限制。然而，由于环境稳定性问题，在控制晶体LMD忆阻器的电阻开关和工作电压方面仍然存在挑战，这阻碍了神经网络硬件的发展。本文介绍了一种通过臭氧处理控制氧化的忆阻器操作优化方法，形成SnOx/SnS2电阻层。这些优化的忆阻器具有低开关电压（~ 1 V）、快速开关速度（~ 20 ns）、高开关比（102）和模拟突触重量可塑性的能力。横截面透射电子显微镜和能量色散x射线能谱识别了电阻开关层中的缺陷和Ti导电丝，有助于均匀开关和最小化操作变化。该器件在MNIST手写识别中达到90%的准确率，并成功实现了基于硬件的图像卷积，展示了SnOx/SnS2记忆电阻器在神经形态应用中的潜力。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Science China Materials Materials Science-General Materials Science

CiteScore

11.40

自引率

7.40%

发文量

949

期刊介绍： Science China Materials (SCM) is a globally peer-reviewed journal that covers all facets of materials science. It is supervised by the Chinese Academy of Sciences and co-sponsored by the Chinese Academy of Sciences and the National Natural Science Foundation of China. The journal is jointly published monthly in both printed and electronic forms by Science China Press and Springer. The aim of SCM is to encourage communication of high-quality, innovative research results at the cutting-edge interface of materials science with chemistry, physics, biology, and engineering. It focuses on breakthroughs from around the world and aims to become a world-leading academic journal for materials science.