Memristor switching and integration in ensembles of silicon nanocrystallites

Isaac Balberg

doi:10.1063/5.0001546

Abstract

We suggest a memristor device that is based on ensembles of Si nanocrystallites that are embedded in an SO2 matrix, for which we show that its operation is well accounted for by the theory of space charge limited currents. This operation consists of a memristive film that exhibits electronic charge integration properties as well as a potential to provide a quantum confinement controlled spiking behavior. As such, the present system is probably the closest available two-terminal electronic film analog that may imitate the neuron’s function. This system can be considered then, not only as “purely electronic” and “bio-realistic” but also as having the great advantage of being compatible with the silicon microelectronic technology. Corresponding devices have the potential to become practical by their downscaling, on the one hand, and by providing a controllable spiking mechanism on the same device, on the other hand.

Highlights

The research of resistive switching in two-terminal devices has taken two parallel routes in the 1960s and early 1970s of the previous century.[1,2] On the experimental end, many solid state phenomena associated with such devices were reported,[3] and on the theoretical end, in 1971, Chua[4] concluded from a symmetry consideration that there can be a passive electric circuit element that he called a memristor and that this element should exhibit a reversible memory behavior with hysteresis and switching
Two-terminal Si based neuromorphic designs that are driven by the space charge limited currents (SCLC) mechanism and have a neuron-like integration capability have not been mentioned far
This structural information, which is summarized in the figure, shows the three main regimes of x: the x < xc regime, the x → xc regime, and the x > xc regime. Each of these regimes is associated with different optical and transport properties and mechanisms.[35,38]. As these properties were already reported previously,[38–43] we summarize here [in SM(2) of the supplementary material] only those that can shed light on the mechanism that leads to the present results and their interpretation

Summary

Introduction

The research of resistive switching in two-terminal devices has taken two parallel routes in the 1960s and early 1970s of the previous century.[1,2] On the experimental end, many solid state phenomena associated with such devices were reported,[3] and on the theoretical end, in 1971, Chua[4] concluded from a symmetry consideration that there can be a passive electric circuit element that he called a memristor and that this element should exhibit a reversible memory behavior with hysteresis and switching.

Objectives

Results

Conclusion