A team of MIT engineers has unveiled a groundbreaking new method to produce hydrogen fuel that could drastically slash the carbon emissions normally associated with the process by using recycled soda cans, seawater, and a splash of caffeine.

In a study published in Cell Reports Sustainability, the researchers report that their aluminium-based method could emit as little as 1.45 kilograms of CO₂ per kilogram of hydrogen, more than seven times less than conventional fossil-fuel methods, which typically release around 11 kilograms of carbon dioxide.

The discovery builds on a quirky but potent chemistry experiment first announced last year, in which researchers triggered a reaction between aluminium, seawater, and gallium-indium alloy to release hydrogen gas. But the big question was: could this be scaled up?

Now, after conducting a comprehensive “cradle-to-grave” life cycle analysis, the MIT team says yes with encouraging results.

“We’re in the ballpark of green hydrogen,” said lead author Aly Kombargi to the MIT website, who recently earned a PhD in mechanical engineering from MIT. “This offers a scalable, low-emission route to hydrogen for vehicles and remote power systems.”

Ordinary aluminium doesn’t usually react with water, thanks to a protective oxide layer. But treat that metal with a small dose of gallium-indium alloy, and the shield comes off—freeing the aluminium to interact with water, releasing hydrogen gas and leaving behind aluminium oxide.

Using seawater helps too, as the salt naturally helps recover the gallium-indium for reuse—turning the cycle into an environmentally and economically viable loop. Even better, the process works best with recycled aluminium, meaning yesterday’s soda can could help fuel tomorrow’s clean energy system.

In the team’s proposed vision, the hydrogen supply chain would look very different from today’s high-pressure, volatile gas systems. Instead, pretreated aluminium pellets would be trucked to hydrogen fueling stations located near the coast. There, seawater would be mixed in on demand, releasing hydrogen on-site and eliminating the need to transport the gas itself.

The researchers estimate a production cost of around $9 per kilogram of hydrogen, competitive with solar and wind-powered alternatives.

And the process has an added bonus: the byproduct, boehmite, is a valuable industrial material used in semiconductors and ceramics, potentially creating a secondary revenue stream.

The team has already built a prototype hydrogen generator the size of a water bottle, capable of powering an electric bike. They’ve previously used the technology to fuel a small car and are now exploring maritime applications, including self-sustaining underwater vehicles that feed on surrounding seawater.

The research was funded in part by the MIT Portugal Program and co-authored by Brooke Bao, Enoch Ellis, and MIT mechanical engineering professor Douglas Hart.

“We still have work to do,” Kombargi said, “but the chemistry works, the economics make sense, and the environmental benefits are clear. That’s a rare trifecta in clean energy.”