Electrochemical Water Separation Study Calls for a New Way to Measure Ion Selectivity

Every glass of clean water hides a sorting problem. Beneath the surface of desalination, groundwater remediation and resource recovery lies the challenge of pulling one kind of ion out of a solution crowded with lookalikes: sodium from calcium, chloride from sulfate, nitrate from bicarbonate. A new study published in Nature Water argues that the field has been measuring its progress at this task with tools too crude for the job, and it proposes a unified framework that could change how engineers compare and design electrochemical separation technologies.

The research, led by K. Li and J. E. Dykstra of Wageningen University & Research together with A. N. Shocron of Ben-Gurion University of the Negev, P. Śledzik and S. Porada of Wroclaw University of Science and Technology, and H. H. M. Rijnaarts, tackles a deceptively simple question: when two competing technologies claim to be selective, how do you know which one actually performs better? The authors’ answer is that, until now, you often could not tell, because conventional selectivity metrics provide limited practical value when applied across different systems.

Selectivity matters far beyond the laboratory. In agriculture, irrigation water must be upgraded so that sodium and chloride are stripped away while nutrients remain. In industry, recovering valuable metals such as silver from streams contaminated with copper requires membranes and electrodes that discriminate between ions differing by only a single charge or a fraction of a nanometre in size. In environmental protection, the targeted removal of nitrate, fluoride or heavy metals from drinking water depends on the same underlying physics. The new paper situates itself squarely in this landscape of water treatment and resource circularity, where the difference between a good separator and a mediocre one is measured in energy, cost and sustainability.

To make their case, the researchers performed a rigorous comparative analysis of two leading electrochemical technologies: electrodialysis, or ED, and membrane capacitive deionization, known as MCDI. Electrodialysis drives ions through a stack of ion-exchange membranes using an electric field, with charged membranes acting as gates that admit some ions and block others. Membrane capacitive deionization works differently: ions are stored electrically in porous carbon electrodes, with ion-exchange membranes shielding each electrode from counter-charged species. Both technologies can be made selective, but they achieve selectivity through fundamentally different mechanisms, and the study reveals that their selective separation behaviour is not directly comparable under the traditional metrics.

The heart of the problem is representation. A typical selectivity number, such as a separation factor between two ions, compresses an entire separation process into a single value, discarding information about purity, water recovery and the degree of removal. A technology might achieve impressive selectivity while recovering only a trickle of usable water, or deliver high recovery while quietly failing to remove the target ion. Because each subfield reports its results in its own format, cross-technology comparison becomes an exercise in apples-to-oranges reasoning. The authors demonstrate this by mapping electrodialysis and MCDI results side by side, showing how the same underlying performance looks misleadingly different depending on the metric chosen.

Their proposed solution is an integrative framework in which selectivity, purity and water recovery can be visualized together on a single plot. Instead of collapsing performance into one number, the framework preserves the trade-offs that engineers actually face: how much of the target ion is removed, how much of the competing ion is left untouched, and how much water comes out the other end. By plotting separation results in this unified space, the researchers show that the advantages of different removal regimes become visible at a glance, allowing a flexible and concise analysis that single metrics cannot provide.

Perhaps the most striking findings concern two advantages that the analysis identifies as unique to membrane capacitive deionization. The first is electrode-enhanced selectivity: in MCDI, the carbon electrodes themselves contribute to discrimination between ions, beyond what the ion-exchange membranes provide. The second is more remarkable still. Because MCDI operates cyclically, adsorbing ions during charging and releasing them during discharge, the timing of the cycle can be tuned to exploit differences in how quickly different ions are absorbed and desorbed. The study shows that this cyclic operation enables what the authors call infinite selectivity, a regime in which target ions are completely removed while competing ions achieve zero removal, based purely on the distinct absorption and desorption dynamics of each ion species.

This time-dependent selectivity builds on a body of theory showing that in porous electrodes, ions do not all arrive at the same rate. Smaller, more mobile ions race ahead during the early stages of charging, while others catch up later. By cutting the adsorption phase at precisely the right moment, an operator can harvest a stream enriched almost exclusively in the target ion. The paper’s simulations, developed with COMSOL for the electrodialysis model and Maple for the MCDI model, trace these dynamics in detail, and the authors have made both codebases openly available through Code Ocean, a move that supports reproducibility and invites the community to build on their framework.

The implications reach across the water technology sector. For designers of electrodialysis stacks, the framework offers a way to quantify honestly how monovalent-selective membranes perform under realistic multi-ionic conditions. For the capacitive deionization community, it provides a common language for reporting the gains from modified electrodes, coating chemistries and cycle-timing strategies. And for decision-makers choosing between technologies for a specific application, whether softening hard groundwater, recovering nutrients from greenhouse drainage water or removing micropollutants, the single-plot visualization promises a more transparent basis for comparison than the fragmented metrics of the past.

The study also joins a growing conversation about measurement itself in separation science. Recent work in the literature has questioned how selectivity is reported in membrane-based separations generally, and this paper extends that critique into electrochemistry with a concrete alternative. By demonstrating that electrodialysis and MCDI behave in fundamentally different ways under equivalent conditions, the authors make a compelling argument that the field needs not just better materials but better ways of seeing what those materials do. As water scarcity intensifies and resource recovery becomes an industrial imperative, the ability to compare separation technologies on shared, meaningful terms may prove as important as the technologies themselves.

Subject of Research: Selective ion removal and selectivity metrics in electrochemical water separation technologies

Article Title: Comparative analysis of selective ion removal in electrochemical technologies reveals the need for a new data representation framework

Article References: Li, K., Shocron, A. N., Śledzik, P., Rijnaarts, H. H. M., Porada, S., & Dykstra, J. E. (2026). Comparative analysis of selective ion removal in electrochemical technologies reveals the need for a new data representation framework. Nature Water. https://doi.org/10.1038/s44221-026-00715-9

Image Credits: AI Generated

DOI: 10.1038/s44221-026-00715-9

Keywords: ion selectivity, electrodialysis, membrane capacitive deionization, water treatment, desalination, resource recovery, ion-exchange membranes, porous carbon electrodes, separation metrics, water recovery, electrochemistry, Nature Water