记忆电阻器
材料科学
钽
电阻随机存取存储器
神经形态工程学
铌
电阻式触摸屏
电极
堆栈(抽象数据类型)
光电子学
电解质
绝缘体(电)
电压
纳米技术
计算机科学
电子工程
电气工程
冶金
化学
工程类
程序设计语言
机器学习
物理化学
计算机视觉
人工神经网络
作者
Ivana Zrinski,Alexey Minenkov,Claudia Cancellieri,Cezarina Cela Mardare,Heiko Groiß,Achim Walter Hassel,Andrei Ionut Mardare
标识
DOI:10.1016/j.apsusc.2022.155917
摘要
Memristive devices exhibiting improved performance and memory characteristics are revealed by using a combinatorial screening approach. An Nb-Ta thin film library, with a total compositional spread ranging between 13 and 80 at.% Ta, is co-sputtered as the bottom electrode, while their anodic oxides served as active layers. Such metal–insulator-metal (MIM) specimens under investigation are finalized with patterned Pt top electrodes. Memristors based on mixed Nb2O5-Ta2O5 have been grouped according to their common switching characteristics in four different compositional zones. Reversible switching between threshold and non-volatile mode was observed for devices grown on Nb-39 at.% Ta alloy. The switching mode was dictated by applying different current compliances. Moreover, devices based on Nb-36 at.% Ta switch at minimum seven distinct resistive levels. Apart from the electrical measurements, the structure of samples and components distribution in cross-section geometry is studied via TEM techniques before and after electrical treatment. Electrolyte incorporation and the exact composition of the identified high-performance memristors (resulting from the combinatorial screening) are analyzed by HAXPES. The memristive devices based on the two singled-out Nb-Ta alloys show enhanced memory and electrical characteristics, thus being excellent candidates for neuromorphic applications.
科研通智能强力驱动
Strongly Powered by AbleSci AI