神经形态工程学
整改
超级电容器
材料科学
电容
二极管
计算机科学
CMOS芯片
电容感应
电阻式触摸屏
光电子学
电压
纳米技术
电气工程
电极
化学
工程类
人工神经网络
人工智能
物理化学
计算机视觉
操作系统
作者
Dong Wang,Bofan Yang,Ziye Zhou,Zhihan Zhang,Zhiwei Wu,Xiaodong Huang
标识
DOI:10.1002/ange.202421913
摘要
Biological system utilizes unidirectional ion flow to produce and transmit signals. To realize bioinspired artificial intelligence and thus seamless human‐machine interaction, ion rectification devices should be developed. Here, a reconfigurable CMOS‐compatible supercapacitor‐diode (CAPode) is developed by resettling the pseudo‐capacitive and electrochemical‐double‐layer‐capacitive components of a lithium‐ion pseudocapacitor into the positive and negative voltage regions respectively through engineering the redox peaks. This CAPode exhibits good ion rectification and charge‐storage bifunction with high rectification ratio (RR) (RRI ~ 20, RRII ~ 0.83), large areal capacitance (17 mF cm‐2) and long cycling stability (5000 cycles). More importantly, two main computing paradigms in the biological system are efficiently realized based on this CAPode by empowering the supercapacitor function into the diode: (I) multivalued ionic logic gates are constructed based on the tunable ion rectification characteristics induced by the bifunction of this CAPode for mimicking the dendritic computing; (II) all‐CAPode based reservoir computing is implemented based on the reconfigurable volatile and nonvolatile charge‐storage characteristics of this CAPode for mimicking the neuromorphic computing. This work paves a new way towards seamless and high‐efficiency human‐machine interaction.
科研通智能强力驱动
Strongly Powered by AbleSci AI