How to harvest drinking water from the air — no energy needed

Posted: June 26, 2026

Across the world, access to water is severely limited. Now, thanks to worsening droughts and the growth of energy-hungry industries like AI, resources from rivers to reservoirs are increasingly stretched, too. 

One option to secure more water is desalination. That industry is finally making headway after decades of low growth, with new innovations like modular deep-sea plants installed on the ocean floor. But desalination is energy-intensive and requires lots of infrastructure—not to mention access to the ocean. 

So, enterprising scientists are also looking for water in the opposite direction: high in the sky. That’s because millions of billions of gallons of water swirl through the atmosphere in the form of vapor. It's a tantalizing resource, abundant but inaccessible. That is, until recently. Harvesting it is now becoming more realistic than ever, as researchers refine techniques to simply squeeze water out of thin air.

Squeezing water from the air with hydrogels 

In May 2024, a team of researchers from Stanford University placed a small metal-and-glass device in Chile’s Atacama Desert, one of the driest places on Earth. The prototype was about the size of a cookie sheet and looked like a dark window with an aluminum frame. As the sun moved across the sky, and the researchers tilted their device to face it, water began trickling into a plastic bottle taped to a funnel at the bottom.^[1]

The contraption had been capturing moisture from the ambient air during the previous night, using one key material: hydrogel composites. These are soft, porous composites made of absorbent polymers and salt that work like a microscopic sponge to soak up atmospheric vapor. During the following day, inside the now sealed-off prototype, a sheet of aluminum painted black absorbed heat to warm up the hydrogel and release its captured water, which then condensed back into liquid—and drinkable—form. 

Hydrogels can hold several times their weight in water, which is why they’ve also found use in everything from contact lenses to disposable diapers to medical dressings. But, as the researchers found, they can also degrade quickly. The material in the Atacama prototype started breaking down after only about 30 cycles of soaking up and releasing water. This can even make the water unsafe to drink, as either the salt or the polymer mix with the condensed liquid.   

After several more years of lab tests, the Stanford scientists have now identified their problem: the gel’s contact with a metal surface, specifically their black-painted foil, releases highly reactive ions that attack the polymer chains and turn the gel into goo. A new anti-corrosion coating now dramatically extends its lifespan, keeping the hydrogel stable for more than eight months at an extreme 167 degrees Fahrenheit, able to perform more than 190 water-harvesting cycles.^[2]

“The radicals are very efficient at eating the polymer away,” said Carlos Diaz-Marin, an assistant professor of energy science and engineering in the Stanford Doerr School of Sustainability and a co-lead author of the research. “To our knowledge, nobody had thought of durability and degradation of these materials, despite it being a critical parameter for water production.” 

The off-grid potential of hydrogels 

The water harvester could bring drinking water to rural communities in arid inland areas, where desalination is not an option. Even in coastal regions, it would have some big advantages, given the energy intensity of conventional desalination plants (which, to be fair, have much higher capacity, too). 

It would also have a minimal environmental footprint compared to water that needs to be pumped or trucked in. And, since it runs entirely passively on solar irradiation, the Stanford prototype doesn’t even need a grid connection—or any external power source at all. 

The Stanford team aren’t the only researchers working on perfecting water harvesters. Last year, another team at MIT tested its own atmospheric water harvester for over a week in California. It similarly consisted of a hydrogel panel in a glass chamber coated with a cooling layer. But, drawing inspiration from origami, the hydrogel itself was folded into small domes that look like black bubble wrap to increase its surface area, expanding and shrinking as it draws and releases moisture. Even in the dry desert air of Death Valley, the model squeezed roughly two-thirds of a cup of water from the air every day.^[3]

Xuanhe Zhao, a professor of mechanical engineering and civil and environmental engineering at MIT, said the harvester’s passive operations would make the design feasible in regions where even a solar cell is hard to come by. An array of panels, Zhao hopes, could eventually even supply a whole household once the design becomes more efficient. 

Can atmospheric water harvesting scale commercially? 

These systems have a clear use case in areas without modern water infrastructure, or with growing water stress. By 2050, an additional 1 billion people are expected to live with extremely high water stress, many of them in developing countries. 

But the researchers also hope their models could be used to supplement traditional water supplies from utilities to help meet surging demand from water-intensive industries. A single large data center needs around 300,000 gallons of water per day and, and a majority of data centers built since 2022 are in areas of high water stress. 

A handful of companies already offer commercially available water harvesters, from utility-scale plants to single-household units, although the industry seems to be in the very early stages. Most companies are also tight-lipped about the specific materials they use in their devices. One, Source Global, says it has installed panels with its own proprietary desiccant at more than 450 sites. Another, Atoco, uses metal organic frameworks instead of hydrogels and is specifically exploring whether its model can produce water for data center cooling.

The Stanford team is now working to further improve the efficiency and cost of its own device, which can already produce just over half a gallon of water per day from a thin layer of hydrogel roughly the size of a bath towel—enough for a person’s basic health during emergencies, but not much more.  

Diaz-Marin, the co-lead on the project, says the higher durability of his team’s latest design moves it closer to a point where it could produce water at a competitive cost, potentially as low as one cent per liter. That’s 10 times higher than the U.S. household rate, but a fraction of the price of bottled water. If successful, he could imagine licensing the design or spinning up a startup. Either way, the ambition is sky-high. 

“We see a path [for] this technology to perhaps even being competitive with tap water,” he said.

^{[1] https://www.sciencedirect.com/science/article/pii/S2666998625001115#bib1
[2] https://www.nature.com/articles/s41467-026-71987-8
[3] https://www.nature.com/articles/s44221-025-00447-2}