基于电化学金属化机制的电子突触电导可控性研究进展

具有高密度和自适应特性的随机银纳米线网络能够模拟神经元的复杂网络拓扑，在信息存储、选择器和神经形态计算领域中应用广泛。文章采用旋涂工艺在图像化电极衬底上制备了银纳米线器件，研究了不同银纳米线浓度和直径对其电学性能的影响，并最终实现了具有超低工作电压（0.01 V）的神经形态器件。实验结果表明，采用直径为30 nm的银纳米线和浓度为5 wt%的复合溶液所制备的随机网络能够实现可控的大开关比阻变行为。进一步，将该大开关比阻变行为作为器件的初始化过程，实现了银纳米线阻变器件在外加超低脉冲电压刺激作用下的可控突触可塑性行为。最后，通过对单个纳米结点中银纳米细丝的形成与断裂过程进行研究，进一步提出随机银纳米线网络在不同电刺激条件下的工作机理。本工作为新型神经形态器件中随机网络的物理实现和应用提供了新思路。

Mimicking classical conditioning based on a single flexible memristor

Random Ag nanowire network with high density and adaptive properties can mimic the complex network topology of neurons and has a wide range of applications in the field of information storage, selectors, and neuromorphic computing. In this work, Ag nanowire devices are prepared on a patterned electrode substrate by a spin-coating process to investigate the impact of the concentrations and diameters of Ag nanowire on their electrical properties. The results show that the random network prepared by Ag nanowires with diameter of 30 nm and composite solutions with concentrations of 5 wt% can achieve controllable high switching resistance behavior. Furthermore, the resistive behavior with large switch is used as the initialization process of the device to achieve controllable synaptic plasticity behavior, which can respond to ultra-low voltage. Finally, the working mechanism of random Ag nanowire networks under different electrical stimulation is proposed by an investigation into the formation and rupture process of Ag nanofilaments in individual nanojunctions. This work provides new ideas for the physical realization and application of random network in the field of novel neuromorphic device.