Multiphase polarization in ion-intercalation nanofilms: general theory including various surface effects and memory applications

TL;DR

This paper develops a comprehensive 2D phase-field model to study multiphase polarization in ion-intercalation nanofilms, revealing how surface effects influence switching behavior and memory applications.

Contribution

It introduces a detailed 2D phase-field model incorporating surface effects to analyze multiphase polarization and its potential for nonvolatile memory devices.

Findings

Surface conditions significantly affect switching performance.

Manipulating mean concentration improves switching efficiency.

Model reproduces experimental memristor results.

Abstract

Ion concentration polarization (CP, current-induced concentration gradient adjacent to a charge-selective interface) has been well studied for single-phase mixed conductors (e.g., liquid electrolyte), but multiphase CP has been rarely addressed in literature. In our recent publication, we proposed that CP above certain threshold currents can flip the phase distribution in multiphase ion-intercalation nanofilms sandwiched by ion-blocking electrodes. We call this phenomenon as multiphase polarization (MP). We then proposed that MP can further lead to nonvolatile interfacial resistive switching (RS) for asymmetric electrodes with ion-modulated electron transfer, which theory can reproduce the experimental results of LTO memristors. In this work, we derive a comprehensive 2D phase-field model for coupled ion-electron transport in ion-intercalation materials, with surface effects including electron transfer kinetics, non-neutral wetting, energy relaxation, and surface charge. Then we use the model to study MP. We present time evolution of phase boundaries, and analyze the switching time, current, energy, and cyclic voltammetry, for various boundary conditions. We find that the switching performance can be improved significantly by manipulating surface conditions and mean concentration. Finally, we discuss the prospects of MP-based memories and possible extensions of the current model.