Abstract
Ions are fundamental to solid-liquid-phase processes, whether as essential components or contaminants, making precise and real-time monitoring necessary. Electrochemical sensors have been identified as promising tools, particularly for field-deployable applications. However, conventional electrochemical sensing is inherently restricted to species that participate in redox reactions and are often single use, limiting its scope. In this study, electrochemical impedance spectroscopy (EIS) is presented as a promising alternative for ion detection, utilizing physico-chemical interactions at the electrode/electrolyte interface. A first-principles model was developed to describe the impedance behavior of electrochemical interfaces, demonstrating how ion-specific interfacial processes influence electrochemical response. Based on this framework, an extensive EIS dataset was compiled, and an AI-assisted model was trained to predict electrolyte composition with high accuracy, achieving detection limits in the parts-per-billion (ppb) range. The findings indicate that EIS has significant potential as a complementary method for ion sensing, providing a novel perspective on selectivity and sensitivity beyond traditional electrochemical approaches. It is anticipated that this work will serve as a foundation for more advanced models of impedance behavior and EIS interpretation, as well as for the development of next-generation impedance-based sensors with broader applicability in complex environments, including biological fluids and industrial liquids.
Supplementary materials
Title
Supplementary information
Description
This Supplementary Information provides detailed descriptions of the numerical models, experimental methods, and machine learning approaches used in the study. It includes the mean-field model for ion distribution and electric field calculations, the impedance model for interfacial ion dynamics, and the methodology for experimental electrochemical impedance spectroscopy (EIS) measurements. Additionally, it covers the preprocessing and training of a machine learning model for ion concentration prediction, along with a summary of key modeling parameters to ensure reproducibility.
Actions