An Atomistic Modelling Framework for Valence Change Memory Cells

TL;DR

This paper introduces an atomistic simulation framework combining structural modeling, vacancy dynamics, and quantum transport to analyze resistive switching in Valence Change Memory cells, providing insights into device operation and optimization.

Contribution

It presents a novel integrated atomistic modeling approach for VCM cells, combining structural, vacancy, and quantum transport simulations to study resistive switching mechanisms.

Findings

Reproduces stochastic switching behavior with conductance ratios of about ten.

Shows conductance changes are due to vacancy dynamics near electrodes.

Demonstrates the framework's potential for device optimization.

Abstract

We present a framework dedicated to modelling the resistive switching operation of Valence Change Memory (VCM) cells. The method combines an atomistic description of the device structure, a Kinetic Monte Carlo (KMC) model for the creation and diffusion of oxygen vacancies in the central oxide under an external field, and an ab-initio quantum transport method to calculate electrical current and conductance. As such, it reproduces a realistically stochastic device operation and its impact on the resulting conductance. We demonstrate this framework by simulating a switching cycle for a TiN/HfO$_2$/TiN VCM cell, and see a clear current hysteresis between high/low resistance states, with a conductance ratio of one order of magnitude. Additionally, we observe that the changes in conductance originate from the creation and recombination of vacancies near the active electrode, effectively modulating a tunnelling gap for the current. This framework can be used to further investigate the mechanisms behind resistive switching at an atomistic scale and optimize VCM material stacks and geometries.