神经形态工程学
极化子
材料科学
纳秒
皮秒
电子
光电子学
石墨烯
凝聚态物理
记忆电阻器
化学物理
格子(音乐)
纳米技术
铁电性
微秒
振幅
极化率
量子隧道
电压
等离子体子
作者
Jijian Liu,Guoquan Gao,Dan Guo,Lan Jiang,Weikang Dong,Keming Li,Shuang Du,Yanzhong Wang,Ping Wang,Tianyu Zang,Zhihang Zhang,Minghui Li,Qingmei Hu,Yadi Guan,Chunyu Zhao,Shoujun Zheng,Tong Zhu,Yao Zhou,Jiadong Zhou
出处
期刊:ACS Nano
[American Chemical Society]
日期:2026-01-27
卷期号:20 (5): 4313-4322
标识
DOI:10.1021/acsnano.5c17513
摘要
Artificial synapse (AS) offers a promising approach to emulate biomimetic nervous systems and potentially overcomes the von Neumann bottleneck. Despite their potential, current neuromorphic devices still suffer from challenges, including electrolysis risks, material degradation, ferroelectric fatigue, and irreversible conductance changes from ion migration and trapping. Here, we propose a self-trapping mechanism due to the intrinsic structural distortion of FePSe3 induced by strong electron–phonon (e-ph) coupling, which preferentially captures electrons, forming polarons within 1 picosecond (ps) and extending carrier lifetimes to tens of nanoseconds (ns). A memristor based on polarons formed in FePSe3 via the nondestructive capture and release of electrons transferred from graphene (Gr) was realized, exhibiting a large memory window exceeding 124 V and stable electrical performance over more than 103 switching cycles. Furthermore, FePSe3-Gr devices show good synaptic plasticity stimulated by different amplitudes and numbers of electrical pulses, indicating the capacity to be applied in AS devices. Meanwhile, the synaptic reset function is observed due to the saturation formation of polarons under optical injection. Our findings present a microscopic approach for stable, high-performance AS devices, advancing their application potential in neuromorphic systems.
科研通智能强力驱动
Strongly Powered by AbleSci AI