佩多:嘘
果胶
油藏计算
纳米技术
计算机科学
材料科学
人工智能
环境科学
生化工程
人工神经网络
工程类
化学
图层(电子)
生物化学
循环神经网络
作者
L. Zhou,Li Song,Junqing Wei,Yaodong Liu,Saijun Fan,Kuibo Lan,Yulin Feng,Fang Wang,Kailiang Zhang
标识
DOI:10.1021/acs.jpcb.5c05699
摘要
Organic polymer-based memristors attract attention in emerging artificial synapses thanks to the simple fabrication process, low cost, and biocompatibility, which can be used for reliable reservoir computing (RC). This article introduces an effective strategy of construction of an organic artificial neural synapse by using conductive polymer PEDOT:PSS and polysaccharide-pectin as a functional layer. After receiving an electrical stimulation, metal ions (Cu2+, from the top electrode) can gain electrons and form uniform and controllable conductive filaments through the PEDOT:PSS-pectin layer. Especially, the as-fabricated Cu/PEDOT:PSS-pectin/ITO (CPPI) device simulates abundant functions like neural synapses, such as long-term potentiation (LTP), long-term depression (LTD), paired-pulse facilitation (PPF), paired-pulse depression (PPD), excitatory postsynaptic current (EPSC), spike-timing-dependent plasticity (STDP), and the transition from short-term to long-term plasticity. Furthermore, a physical RC system is constructed, achieving a recognition rate of 99.0% for machine-written numbers (from "0" to "9"). Then, a physical model of a CPPI device is built and integrated with the RC system, achieving a recognition rate of 91.1% for MNIST handwritten digits. This study advances the development of organic polymers in artificial synapses and provides a novel material platform for designing memristors tailored to next-generation brain-inspired computing.
科研通智能强力驱动
Strongly Powered by AbleSci AI