New paper published in Cell Press Physical Science on the use of sub-nanometer pores for capacitive deionization to enable membrane-free seawater desalination. Big pores are mighty powerful when it comes to capacitive deionization (CDI). CDI is highly appreciated as a potentially energy-efficient desalination technology, rendering saline water into desalinated (potable/processable) water. However, once we move from saline media with low salt concentrations (like in brackish water regimes) towards higher salt concentrations (as you find in ordinary seawater), CDI become less attractive: the desalination capacity and charge efficiency (think of it as salt removal per invested charge) drop drastically. This issue is linked to the limited permselectivity of carbon pores commonly found in CDI electrodes. Put simply: the invested charge is not only used for adding “extra “ions into the pore (thereby: lowering the feedwater ion concentration) but also to eject ions that are already inside the pore (which basically increases the ion concentration in the effluent stream). We can address this issue by implementing an ion exchange membrane (adding costs and a more complex design) or using charge-transfer materials (giving rise to desalination batteries). But is there a way to keep low-cost, nanoporous carbon and still enable direct, membrane-free seawater desalination? The answer is a resounding YES. In our 2020 paper in Sustainable Energy & Fuels, we showed already the proof of concept of using quasi-ionophobic, and thereby permselective, carbon pores. Now, our work extends the scope and demonstrates this effect’s intricate pore size dependency. The key is a subtle play of pore size and hydrated ion diameter, which allows the pore to only uptake (extra) ions once an electric potential is applied. This work was a great collaboration with the team of Guang Feng at HUST, China, and Christian Prehal at ETH Zürich that puts together simulation and experimental work.