有机聚合物掺杂离子液体忆阻器用于低电压人工光电突触

Ionic Liquid-doped Organic Polymer Memristors for Low-voltage Artificial Optoelectronic Synapses

基于忆阻器的神经形态计算有望克服冯诺依曼瓶颈,有助于实现节能数据处理。有机聚合物因其优异的半导体和光电特性以及溶液加工优势,是低成本忆阻器和突触应用的理想候选者。然而,聚合物忆阻器的高操作电压和阻态随机切换问题限制了其在神经形态芯片中的应用,增加了外围电路系统的复杂性。本研究利用掺杂离子液体[EMIM][PF_(6)]的有机聚合物MEH-PPV为活性层构建了垂直结构忆阻器,通过电输入调制各种短期和长期突触可塑性行为,显著降低了工作电压并增加了可调谐导电态范围。该忆阻器在mV级电压下实现光调控生物突触的关键特征和类脑学习与记忆特性,集成光信号传感、存储和处理功能。在此基础上,将该有机突触阵列应用于人类视觉神经系统的图案识别与记忆任务。研究结果为下一代低功耗高性能有机神经形态设备的设计提供了重要指导。

Memristor-based neuromorphic computing showcases great potential for overcoming the von Neumann bottleneck and enabling energy-efficient data processing.Organic polymers,due to their excellent semiconductor and optoelectronic properties as well as solution processability,are ideal candidates for low-cost memristor and synap-tic applications.However,the high operating voltage and random switching of resistive states in polymer memristors limit their performance in neuromorphic chips and increase the complexity of peripheral circuits.In this study,the polymer MEH-PPV doped with ionic liquid[EMIM][PF_(6)]was used as the active layer to construct a vertically struc-tured memristor.This design provides a variety of short-term and long-term synaptic plasticity behaviors,which can be modulated by electrical and optical inputs,achieving a significant reduction in operating voltage and an expanded range of tunable conductance states.The memristor exhibited key features of optically modulated biological synapse and brain-inspired learning and memory characteristics at mV-level voltage while integrating photodetection,memo-ry,and processing functions into a single compact device.Additionally,this organic synaptic array was applied to image recognition and memory tasks in the human visual neural system.The promising results provide important in-sights for the design of next-generation low-power and high-performance organic neuromorphic devices.