A Simplified Tantalum Oxide Memristor Model, Parameters Estimation and Application in Memory Crossbars

Valeri Mladenov,Stoyan Kirilov

doi:10.3390/technologies10010006

Abstract

In this paper, an improved and simplified modification of a tantalum oxide memristor model is presented. The proposed model is applied and analyzed in hybrid and passive memory crossbars in LTSPICE environment and is based on the standard Ta2O5 memristor model proposed by Hewlett–Packard. The discussed modified model has several main enhancements—inclusion of a simplified window function, improvement of its effectiveness by the use of a simple expression for the i–v relationship, and replacement of the classical Heaviside step function with a differentiable and flat step-like function. The optimal values of coefficients of the tantalum oxide memristor model are derived by comparison of experimental current–voltage relationships and by using a procedure for parameter estimation. A simplified LTSPICE library model, correspondent to the analyzed tantalum oxide memristor, is created in accordance with the considered mathematical model. The improved and altered Ta2O5 memristor model is tested and simulated in hybrid and passive memory crossbars for a state near to a hard-switching operation. After a comparison of several of the best existing memristor models, the main pros of the proposed memristor model are highlighted—its improved implementation, better operating rate, and good switching properties.

Highlights

Received: 26 November 2021The memristor elements, based mainly on amorphous transition metal oxides, as TiO2, HfO2, Ta2 O5, ZnO and other materials, have many potential applications, as in nonvolatile memory devices, artificial neural networks, and logical and reconfigurable electronic circuits [1,2,3,4]
Ac- the is initially adjusted according to experimental current–voltage to derive cording to thevalues alteration the stateelement variablemodel’s in a broad range
It is tuned according to experimental current–voltage characteristics of Ta2 O5 memristor nanostructures

Summary

Introduction

The memristor elements, based mainly on amorphous transition metal oxides, as TiO2 , HfO2 , Ta2 O5 , ZnO and other materials, have many potential applications, as in nonvolatile memory devices, artificial neural networks, and logical and reconfigurable electronic circuits [1,2,3,4]. The memristor was predicted in 1971 by Chua as the fourth fundamental, passive, and nonlinear one-port electronic element [5]. It has a memory effect and can retain its state after switching off the electric sources [5,6]. Along with the oxide materials with resistance switching abilities already analyzed, the amorphous Ta2 O5 , doped by oxygen vacancies, has excellent switching properties, a sound dependability, great operating speed, small energy consumption, comparatively extensive memorizing time, and a good compatibility with the commonly used Complementary Metal Oxide

Objectives

Findings

Discussion

Conclusion