In this research topic, we have realized a resistance switching device operating based on a phenomenon that occurs at the interface between a chalcogenide alloy based on (Bi,Sb)-(Cu,Ag,Ni)-(S,Se,Te) and a metal such as Al and Ag. Our goal is to demonstrate machine learning using an artificial neural network, constructed by interfacial-resistance-switching devices.
Recently, there has been a drastic increase in the interest in artificial intelligence and machine learning such as deep learning. Deep learning carries out nonlinear transformations in a stack of artificial neural networks to obtain a target output from input data. A network constructed using so-called memristors can improve the efficiency for such neural computations. A memristor is a two-terminal element, the resistance of which depends not only on its actual state but also on the history of the current passing through it. Memristors can be realized by resistance switching devices, where the resistance switches between high and low values due to, for instance, the creation and annihilation of a conductive filament in a solid electrolyte.
We have realized a resistance switching device operating based on a phenomenon that occurs at the interface between a chalcogenide alloy based on (Bi,Sb)-(Cu,Ag,Ni)-(S,Se,Te) and a metal such as Al and Ag. The resistance changes due to a spontaneous generation of high-resistivity compounds at the metal-chalcogenide interface and their electrochemical annihilation under a voltage bias. A unique property of the interfacial-resistance switching is that the switching voltage can be tuned by utilizing the dependence of the involved redox reactions on the history of applied voltage. Such a characteristic may be utilized for new functionalities in artificial neural networks. Our goal is to demonstrate machine learning using an artificial neural network, which is constructed by the interfacial-resistance-switching devices.
Author: Y. Takagaki , B. Jenichen , M. Ramsteiner , A. Trampert
Title: Interfacial resistance switching characteristics in metal-chalcogenide junctions using Bi-Cu-Se, Bi-Ag-Se, and Sb-Cu-Te alloys
Source: J. Alloy. Comp. , 824 , 153880 ( 2020 )
Author: Y. Takagaki , M. Ramsteiner , U. Jahn , B. Jenichen , A. Trampert
Title: Memristive resistive switch based on spontaneous barrier creation in metal-chalcogenide junctions
Source: J. Phys. D: Appl. Phys. , 52 , 385101 ( 2019 )