Bridging Database Analysis with Microrheology to Reveal Super-Hydrodynamic Conductivity Scaling Regimes in Ionic Liquids

Abstract

Ion transport through electrolytes critically impacts the performance of batteries and other electrochemical devices. Many frameworks used to predict and tune ion transport, such as the Nernst-Einstein model, assume hydrodynamic transport mechanisms, and hence focus on maximizing electrolyte conductivity by minimizing bulk viscosity. However, the emergence of solid-state electrolytes illustrates that selective, non-hydrodynamic ion transport provides promising avenues for enhancing ionic transport in electrolytes. Increasingly, selective ion transport mechanisms, such as hopping, are proposed for concentrated electrolytes, including ionic liquid-derived materials. Yet viscosity-conductivity scaling relationships in ionic liquids are still often analyzed with hydrodynamic models. Here, we report a data-centric analysis of how well hydrodynamic transport models describe the scaling between viscosity and conductivity in neat ionic liquids by merging three databases to bridge physical properties and chemical descriptors. With this expansive data set, we constrained our scaling analysis using ion sizes defined using simulated molecular volumes, as opposed to prior approaches that estimate sizes from activity coefficients or rely on ad-hoc estimates. Remarkably, we find that many commonly studied ionic liquids exhibit positive deviations from the Nernst-Einstein model, implying that ions move faster than hydrodynamic limitations should allow. We experimentally verify these positive deviations in a common class of ionic liquids using microrheology and conductivity measurements. Our results highlight overlooked super-hydrodynamic regimes in ionic liquid viscosity-conductivity scaling and point to opportunities to understand mechanisms of correlated ion motion in ionic liquids. We further show data science and machine learning tools can improve predictions of conductivity from molecular properties, including demonstrating predictions can be made using only computational features. Our findings reveal that many ionic liquids exhibit super-hydrodynamic viscosity-conductivity scaling, which could be harnessed to influence the behavior of electrochemical devices.