A Step Toward Batteries Made From Plentiful Sodium to Store Renewable Power

The sun sets, the wind drops, and the power still has to come from somewhere. As renewable power and electrified infrastructure grow, so does the need for energy storage that is low cost, made from abundant materials and able to charge fast. A team at MIT reports a step toward that kind of battery, built around a very common element: sodium.

In a study published in Joule in August 2026, a 15-member MIT team led by Ju Li and including Jeremiah Johnson reports an electrolyte solvent, N,N-dimethylsulfamoyl fluoride (DMFSA), that helps sodium-metal batteries with both stability and speed. According to MIT News, it was found with help from an AI design tool that designed 100,000 candidates on a computer.

Why sodium?

MIT News reports sodium is about 1,000 times more abundant than lithium and, pound for pound, about one-hundredth the cost. That is the appeal: storage built from something plentiful. The catch is that sodium metal is so reactive that batteries built with it struggle to be both long-lived and fast.

What was standing in the way?

The electrolyte, the liquid that carries ions between a battery’s electrodes, has to be stable against those electrodes and also move ions quickly. Solvents that are very stable tend to move ions slowly, and fast ion movers tend to eat away at the electrodes. As Li put it to MIT News, the electrolyte “is supposed to be an ion conductor.” The hard part is finding one that conducts well without attacking anything.

Where did the AI come in?

The search began from a known good molecule. According to MIT News, the Li group and collaborators found in 2021 that a solvent called DMTMSA was unusually stable in lithium batteries, so the team went looking for smaller relatives. PhD student Chia-Wei Hsu built an AI-guided algorithm that designed 100,000 candidate molecules on a computer within 24 hours, then narrowed the pool to 200 using criteria including shape similarity to DMTMSA and comparable electronic properties. Twenty-seven representative candidates were made and tested head-to-head in the lab. “We tested them all under the same conditions to make it a fair, head-to-head competition,” postdoc Weiyin Chen told MIT News. The winner, DMFSA, was also the smallest.

The paper itself describes the pipeline as molecular design to generate candidates and a three-dimensional Zernike model, a mathematical way of describing molecular shape, to screen them. The AI did not run the battery tests. It shrank an enormous search into a list a lab could handle.

Did the winner work?

In a cell pairing a high-voltage sodium nickel manganese iron oxide cathode with a sodium-metal anode, DMFSA showed what the paper’s abstract calls “extraordinarily stable cycling.” The abstract also reports that small solvents in this family showed up to about 20 times higher ionic conductivity than larger counterparts, plus a higher electrode potential, which together supported better high-rate charging and discharging. The team argues the lasting lesson is a design rule: choose small solvents from a family with a proven stable backbone, a principle they suggest could apply to other alkali-metal batteries. Jinhyuk Lee of McGill University, who was not involved, told MIT News the work “addresses one of the most persistent challenges in battery research.”

How far is this from a battery people can use?

A long way, honestly. This is lab-scale cell testing, not a commercial battery, and long-term safety, cost and manufacturing at scale are untested. Moving toward real use would require testing in larger, practical cells. MIT News reports that a new search is already underway, this time starting from DMFSA. If that line of work keeps paying off, the people who benefit will be the ones who never think about electrolytes at all: a town whose solar power keeps flowing after dark, stored in something as ordinary as sodium.