Publication date: 10th April 2024
There is an urge to correctly understand the motion of ions under large electric fields due to implications in several technologies such as resistive switching memories, capacitors, and flash sintering. In spite, of over more than 70 years of analysis to this problem, a fundamental understanding is still missing. Herein, we conducted molecular dynamics simulations on two model oxide systems; a simple dielectric (CeO2) and a ferroelectric phase (f-HfO2) under a range of temperatures and electric fields. From these simulations, we extracted activation barriers for the “effective” diffusivities of the oxide ion and found out that in both oxides, the barriers can be correctly predicted using recent analytical models. These models are the ones that take into account local atomic polarization effects. Beyond this analysis, we tried to answer the question of the possibility of separating the drift from the diffusion contributions to the field-enhanced ionic motion. This was conducted by analyzing the time and space density correlation function of the oxide ion. We found out that complete decoupling of diffusion from drift is not possible and eventually one has to live with a “drift-mixed diffusion” or a “diffusion-mixed drift”. As an alternative approach to describe the motion of ions under electric fields, we analyzed the cumulants of motion up to the fourth cumulant. Our analysis revealed that these cumulants do not vanish indicated a highly non-Gaussian behavior for the motion of ions under electric fields. Even more problematic, higher cumulants (third and fourth) do not lend themselves to simple Arrhenius behavior. Our work does not completely address the needed understanding of field-enhanced ionic transport. But it furnishes the ground for future studies that can build on our analyses.
M.Y. acknowledges an AUC Faculty Support Grant (SSE-MENG-M.Y.-FY21-FY22-RG (1-21).M.Y. acknowledges an AUC Faculty Support Grant (SSE-MENG-M.Y.-FY21-FY22-RG (1-21) and a confernce travel grant.