Co-percolation to tune conductive behaviour in dynamical metallic nanowire networks

Abstract

Nanowire networks act as self-healing smart materials, whose sheet resistance can be tuned via an externally applied voltage stimulus. This memristive response occurs due to modification of junction resistances to form a connectivity path across the lowest barrier junctions in the network. While most network studies have been performed on expensive noble metal nanowires like silver, networks of inexpensive nickel nanowires with a nickel oxide coating can also demonstrate resistive switching, a common feature of metal oxides with filamentary conduction. However, networks made from solely nickel nanowires have high operation voltages which prohibit large-scale material applications. Here we show, using both experiment and simulation, that a heterogeneous network of nickel and silver nanowires allows optimization of the activation voltage, as well as tuning of the conduction behavior to be either resistive switching, memristive, or a combination of both. Small percentages of silver nanowires, below the percolation threshold, induce these changes in electrical behaviour, even for low area coverage and hence very transparent films. Silver nanowires act as current concentrators, amplifying conductivity locally as shown in our computational dynamical activation framework for networks of junctions. These results demonstrate that a heterogeneous nanowire network can act as a cost-effective adaptive material with minimal use of noble metal nanowires, without losing memristive behaviour that is essential for smart sensing and neuromorphic applications.

Collective capacitive and memristive responses in random nanowire networks: Emergence of critical connectivity pathways

“…To illustrate the nonlinear mechanisms governing the formation and conduction processes in disordered NWNs, we developed two modelling schemes that will describe the network dynamics in the two distinctive regimes: a (i) fast-switching capacitive model (CPM) and a (ii) slow-switching MR model (MRM). Both models have already been used to capture the main dynamical features of numerous NWN samples and their outcomes are successfully supported by experimental data [22,25,26]. However, these models were never closely compared and, as we shall demonstrate, they will unveil distinct switching behaviours that play an important role in the synaptic-like response of electrically stressed NWNs.…”

“…CPM simulation [25] begins by placing the whole capacitor network in contact with electrodes that source and drain a certain amount of charge Q, representing the charge that builds up due to the applied bias voltage. The applied charge is incremented from an initial value Q i up to a pre-defined maximum value of Q max in steps of ∆Q.…”

Recent Developments and Perspectives for Memristive Devices Based on Metal Oxide Nanowires

“…[275][276][277][278][279][280][281][282][283] In the following, resistive switching in stacked and planar devices based on NW networks is described. Resistive switching networks based on random ordered nanowires, dendritic, and fractal structures have attracted great attention because of the peculiar conduction properties that make these structures promising for resistive switching and neuromorphic applications.…”

“…[275,277,[280][281][282][283] Thus, the global resistive switching behavior of the network arises from resistive switching events located at the numerous wire interconnections. Indeed, the NW network global properties arise from junctions between individual wires that determine the network connectivity.…”

Brain‐Inspired Structural Plasticity through Reweighting and Rewiring in Multi‐Terminal Self‐Organizing Memristive Nanowire Networks

“…Despite a bright future prospective for the development of next‐generation artificial intelligence (AI) systems, it is hard for such rigid top‐down architectures to emulate most typical features of biological neural networks such as high connectivity, adaptability through reconnection and rewiring, and long‐range spatio‐temporal correlation. Alternative ways using unconventional systems consisting of many interacting nano‐parts have been proposed for the realization of biologically plausible architectures where the emergent behavior arises from a complexity similar to that of biological neural circuits . However, these systems were unable to demonstrate bio‐realistic implementation of structural plasticity including reweighting and rewiring and spatio‐temporal processing of input signals similarly to our brain.…”