Researchers in Germany and Japan have demonstrated a cooling system in which heat supplies the force normally provided by an electric motor. Built around two ultrathin metal films, the prototype opens a possible route to using waste heat or solar heat to keep other components cool.
The team at the Karlsruhe Institute of Technology (KIT) and Japan’s University of Tsukuba reported its findings in Nature Energy on August 28, 2026. Their work establishes a working principle for heat-driven solid-state cooling, although the laboratory setup still uses electrical equipment and remains far from a finished appliance.
How can heat make something colder?
Think about the warmth coming from the back of a refrigerator. A conventional refrigerator uses an electrically powered compressor and a circulating refrigerant to move heat from its interior into the surrounding room.
The new design gives that job a different starting point, pairing two films made from nickel-titanium-based shape-memory alloys. These materials can recover a preset shape when heated, allowing the first film to contract and pull on the second. In practical terms, heat creates the mechanical force needed to operate the cooling material.
Stretching the second film changes its crystal structure and releases heat, which the device transfers away. When the load is released, the film cools and can absorb heat from the component being cooled. Repeating this sequence produces what scientists call “elastocaloric cooling.”
What the prototype actually achieved
In tests using electrical resistance to heat the driving film to 86°C (about 187°F), the researchers measured a 4°C temperature difference across the cooling device. The refrigerant film showed a separate temperature span of 12.9°C, which does not mean the whole device, or a room around it, became that much colder.
With an external heat source at 130°C (266°F), the device achieved a smaller temperature difference of 2.2°C. That test demonstrated the central idea, but the source was a laboratory heater, so operation using actual industrial waste heat or sunlight remains a future application.
“This showed us that the principle doesn’t just work in theory,” lead author Yi-Ting Hsiau said in KIT’s announcement. The doctoral researcher described measuring cooling from the heat-driven system as the decisive moment for the team.
What “heat-driven” means for electricity use
So, could this eventually cool electronics without a conventional electric drive? That is the opportunity the researchers are exploring, since the heated film can supply the force that stretches the refrigerant film.
But the complete experimental setup still included electrically operated equipment, including actuators that moved heat exchangers into contact with the films. A practical system would need to account for those supporting functions when calculating its electricity demand.
The distinction matters for anyone picturing an air conditioner that never affects the electric bill. What has been demonstrated is a new way to power the mechanical cooling cycle, while household energy savings and performance at larger scales still need to be established.
Why cooling needs another approach
There is a substantial environmental reason to pursue the idea. According to a July 2026 analysis from the International Energy Agency, global electricity demand for space cooling has grown by 50% since 2015, reaching roughly 2,900 terawatt-hours.
That is more electricity than the entire European Union consumes, the agency reported. Cooling also contributed around 1,400 gigawatts of peak electricity demand across markets in 2025, showing why that familiar blast of air conditioning on a sticky afternoon matters beyond the home.
Solid-state cooling offers another potential advantage because its working refrigerant is a solid material, avoiding the volatile refrigerants used in conventional systems. Many common refrigerants contribute to global warming if released, so developing alternatives could address both electricity demand and refrigerant emissions.
Where the technology could go next
KIT says the researchers are already working on connecting multiple films in parallel to increase cooling capacity. The present prototype was designed to establish feasibility, which means demonstrating that the principle works before optimizing how much useful cooling it can deliver.
Possible applications include cooling computer processors with available waste heat and protecting sensitive automotive electronics using heat from the drivetrain. Solar heat is another potential energy source, although matching the available heat to a device’s operating requirements will be essential.
The interesting possibility is giving unwanted heat a second job, helping drive cooling where it is needed. Whether that can become a compact, practical product will depend on how successfully the team scales the design.
The study was published in Nature Energy.
Photo: KIT / Ella Maru Studio
