神经形态工程学
记忆电阻器
材料科学
非易失性存储器
异质结
纳米技术
电子线路
光电子学
备用电源
集成电路
微尺度化学
电压
电气工程
计算机科学
数学教育
数学
机器学习
人工神经网络
工程类
作者
Renu Yadav,Saroj Poudyal,Ramesh Rajarapu,Bubunu Biswal,Prahalad Kanti Barman,S. Kasiviswanathan,Kostya S. Novoselov,Abhishek Misra
出处
期刊:Small
[Wiley]
日期:2023-12-27
标识
DOI:10.1002/smll.202309163
摘要
Memristors-based integrated circuits for emerging bio-inspired computing paradigms require an integrated approach utilizing both volatile and nonvolatile memristive devices. Here, an innovative architecture comprising of 1D CVD-grown core-shell heterostructures (CSHSs) of MoO2 -MoS2 is employed as memristors manifesting both volatile switching (with high selectivity of 107 and steep slope of 0.6 mV decade-1 ) and nonvolatile switching phenomena (with Ion /Ioff ≈103 and switching speed of 60 ns). In these CSHSs, the metallic core MoO2 with high current carrying capacity provides a conformal and immaculate interface with semiconducting MoS2 shells and therefore it acts as a bottom electrode for the memristors. The power consumption in volatile devices is as low as 50 pW per set transition and 0.1 fW in standby mode. Voltage-driven current spikes are observed for volatile devices while with nonvolatile memristors, key features of a biological synapse such as short/long-term plasticity and paired pulse facilitation are emulated suggesting their potential for the development of neuromorphic circuits. These CSHSs offer an unprecedented solution for the interfacial issues between metallic electrodes and the layered materials-based switching element with the prospects of developing smaller footprint memristive devices for future integrated circuits.
科研通智能强力驱动
Strongly Powered by AbleSci AI