Formation and dissolution of conductive channels in an Ag₂S-islands network

Abstract

A network that shows dynamical change in response to each input pattern works as a reservoir. In this study, we propose to use a network of Ag2S islands for reservoir computing. Input of a small number of electrons to an Ag2S island from a neighbored island causes Ag nanowire growth, which eventually connects the two islands. Shrinkage of an Ag nanowire also occurs depending on the distribution of a local electric potential. Thus, growth and shrinkage of Ag nanowires among Ag2S islands shows dynamical change in conductive channels responding to input bias, which is demonstrated by an Ag2S-island network surrounded by input/output electrodes. The Ag2S islands that are electrically connected to a surrounding electrode are identified using a conductive-atomic force microscope, confirming the operating mechanism.

Neuromorphic nanowire networks: principles, progress and future prospects for neuro-inspired information processing

“…One of the first resistive switching memory devices to be realised, the atomic switch [52], typically uses Ag 2 S as the electrolyte material and Ag as the active metal electrode. Recognising the potential for a unique neuromorphic system by combining the memristive synthetic synapse functionality with the recurrent network structure of an NWN, Gimzewski, Stieg and colleagues at UCLA-CNSI and NIMS-MANA extended the atomic switch concept further by demonstrating Ag-Ag 2 S NWNs [45,59] (see also [93]). In an innovative approach that integrates top-down fabrication with bottom-up self-assembly (Figure 8), Ag nanowires were grown on a substrate that was lithographically pre-patterned with seed Cu posts (100 nm high, 1 μm pitch).…”

Ag2S island network reservoir that works with direct optical signal inputs

“…Bias application to V in causes diffusion of Ag + cations in each Ag 2 S island, which results in growth/ shrinkage of an Ag filament between Ag 2 S islands. These phenomena induce the formation of random conductive channels, 22) which is observed as nonlinear resistance change between V in and V out electrodes. Using the characteristics, basic operations of reservoir computing have been demonstrated, 23) where voltage waves were used as input signals.…”

Measurement of changes in resistance of a Ag2+δ S nano-island on removal of dopant δ-Ag atoms

“…Adding an extra I/V measuring system to the C-AFM can potentially remove this limit. Dynamically controlling resistance by growing Ag nanowires can be used in such things as unconventional computing, [35][36][37][38] where conductive channels are modified not only by an Ag filament in a gap between Ag 2+δ S nano-islands but also by the Ag 2+δ S nano-islands themselves. 39) The technique demonstrated in this study will contribute to the development of such new computing systems.…”