Electricity-free solid-state cooling turns heat directly into cold

Heat Becomes Cold: Scientists Unveil the World’s First Cooling System That Runs on Waste Heat — No Electricity Required

For more than a century, producing cold has rested on a single recipe: an electric motor spins a compressor, the compressor squeezes a chemical refrigerant, and the refrigerant ferries heat from where it is not wanted to somewhere it matters less. That architecture, essentially unchanged for over a hundred years, hums inside refrigerators, air conditioners and the data centers that underpin modern digital life — and its appetite grows every year. Now researchers at the Karlsruhe Institute of Technology (KIT) in Germany and the University of Tsukuba in Japan have broken with that tradition. In a study published in the journal Nature Energy on 28 August 2026, they describe what they call a world first: a heat-driven, elastocaloric cooling system that generates cold directly from heat, with no electricity, no compressor and no refrigerant. Its energy source is the resource our civilization squanders in staggering quantities — waste heat — together with energy harvested straight from the sun.

The stakes are enormous. According to the researchers, cooling and heating now account for almost half of global energy consumption, a share that keeps climbing as heatwaves intensify, cities swell and server farms multiply. Meanwhile, many common refrigerants are themselves potent greenhouse gases that contribute to global warming when they escape into the atmosphere. The technology carrying this burden is the vapor-compression cycle, in which an electricity-driven compressor transfers heat carried by a refrigerant from one location to another — the same basic principle that has served fridges, air conditioners and industrial chillers since the early twentieth century. Elastocaloric solid-state cooling has long been considered one of the most credible alternatives. It relies on a remarkable property of shape-memory alloys: they tend to cool down once a previously applied mechanical load is released. No fluids, no compressor, no fluorinated chemistry — just a metal that changes its crystal structure on command. But until now, even these futuristic systems shared one trait with their ancestors: an electric motor did the heavy lifting.

The physics behind the cooling effect is as striking as it is elegant. When a superelastic shape-memory alloy is mechanically stressed, its crystal lattice abruptly reorganizes from the austenite phase into the martensite phase — a diffusionless reshuffling of atoms that releases latent heat and warms the material. The decisive moment comes when the load is released. The lattice snaps back, the reverse transformation absorbs latent heat from the surroundings, and the material’s temperature plunges by several degrees Celsius within a fraction of a second. Cycled rapidly between loading and unloading, a thin strip of nickel-titanium becomes a heat pump built from solid metal: no refrigerant to leak, no compressor oil, no evaporation and condensation. Among the family of so-called caloric effects — magnetocaloric, electrocaloric and barocaloric — the elastocaloric effect is prized for the magnitude of its temperature change and for relying on nickel-titanium, an alloy that medicine and industry have manufactured at scale for decades.

Yet one stubborn caveat has shadowed elastocaloric cooling from the beginning: something still has to do the squeezing. Every elastocaloric system demonstrated to date has depended on an electrically driven actuator to generate the force that triggers the phase transformation — meaning that even a “solid-state” cooler ultimately needed a motor, and therefore grid electricity, to run. That electrical umbilical cord has left the technology blind to its most tantalizing fuel: abundant low-temperature heat. Industrial waste streams, vehicle exhaust, engine coolant and rooftop solar-thermal collectors all deliver heat in enormous quantities, and cooling demand peaks precisely when the sun blazes — but heat cannot turn an electric motor. The question that has hovered over the field is disarmingly simple: could the actuator itself be driven by heat, so that a cooling system could feed on the very warmth that would otherwise be thrown away? The KIT-led team has now answered it in experiment for the first time.

Their solution is a piece of materials choreography. The prototype couples two ultra-thin nickel-titanium films with complementary functions. The first film works as a heat-responsive actuator built on the classic shape-memory effect: once it is heated up, it starts to shrink, converting thermal energy directly into mechanical work — a genuine pulling force generated without any electric motor in the loop. That motion immediately transfers to the second film, which serves as the elastocaloric refrigerant. Under cyclic loading and unloading, the second film undergoes reversible alterations in its crystal structure that generate cold. In one stroke, heat replaces the electric actuator that previously stood at the heart of every elastocaloric cooling system. “The crucial innovation is that we combine two complementary functions of shape memory alloys, with one film converting heat into mechanical work and the other film converting this work into cold,” said Dr. Jingyuan Xu, who leads the Young Investigator Group of the ZEco Thermal Lab at KIT’s Institute of Microstructure Technology (IMT). “This way, we’re establishing a new approach to drive solid-state cooling, thereby opening up exciting possibilities for the use of waste heat and solar energy.”

The measurements that matter have now been delivered. At an actuator temperature of 86 degrees Celsius — a level comfortably within reach of solar-thermal collectors, engine coolant loops and countless industrial processes — the prototype achieved a temperature difference of 4 degrees Celsius on the component level, while the temperature change inside the elastocaloric refrigerant itself amounted to nearly 13 degrees Celsius. Those figures, the researchers emphasize, represent the first experimental proof that a heat-driven elastocaloric system can genuinely produce cold rather than merely exist on paper. The setup also operated reliably when fed by an external heat source providing 130 degrees Celsius, demonstrating that the concept can digest real-world heat sources rather than laboratory idealizations. “The decisive moment for us was when we were able to measure the cold that had indeed been generated by a heat-driven system,” said Yi-Ting Hsiau, lead author of the study and a doctoral researcher at the IMT. “This showed us that the principle doesn’t just work in theory.”

Conceptually, the device is a heat engine fused with a solid-state heat pump: the actuator film is the engine, converting a temperature difference into mechanical work, while the elastocaloric film is the pump, spending that work to absorb heat from the cold side. Because the working substance is a solid rather than a pressurized gas, there is nothing to leak, no fluorinated chemistry to regulate and no compressor to lubricate, maintain or eventually fail. And because the trigger is thermal, the system can in principle be attached directly to a waste-heat pipe or placed beneath a solar absorber, converting energy that would otherwise simply warm the atmosphere into useful refrigeration. The researchers are careful to frame the current device as a feasibility study rather than an optimized machine — but as a demonstration that heat can replace the electric actuator driving solid-state cooling, it is a first.

The application landscape is correspondingly broad. Processors in computers could, in a sense, cool themselves: the chips that need cooling also radiate the waste heat that would drive the elastocaloric cycle, a self-regulating loop for servers and data centers whose thermal appetite grows with every exabyte. In automobiles, sensitive electronics could be cooled using heat drawn from the drivetrain rather than by taxing the vehicle’s electrical system. Solar-driven air conditioning is an equally natural fit, since the same sunshine that heats a building supplies the energy to cool it, aligning cooling supply with cooling demand hour by hour. In every case the principle is the same: warmth that already exists does the work, and no fresh electricity needs to be spent.

Nobody involved presents the prototype as market-ready. The published device was designed explicitly as a feasibility study and is deliberately not optimized for maximum cooling capacity; the temperature spans it achieved, while scientifically decisive, remain far below what a domestic air conditioner must deliver. The path forward — already under way in the lab — is parallelization: connecting multiple shape-memory film pairs side by side, multiplying cooling power much as individual photovoltaic cells are tiled into solar panels. Scaling up will also mean confronting the fatigue that shape-memory alloys suffer under relentless mechanical cycling and engineering heat exchangers that can feed ultrathin films efficiently. The study was conducted in collaboration with the University of Tsukuba in Japan, which the researchers say paves the way for heat-driven solid-state cooling fit for practical use. Funding came from the Carl Zeiss Foundation through the CZS Nexus project, the Baden-Württemberg Foundation’s elite postdocs program, and the Hector Fellow Academy.

For Xu, the milestone is a starting line rather than a finish line. “We believe that this is only the beginning,” Xu said. “By scaling up this technology we want to develop compact cooling systems that leverage abundantly available heat sources for sustainable cooling.” If that ambition holds, the implications stretch far beyond a laboratory bench in Karlsruhe. Cooling is quietly becoming one of the defining energy problems of a warming century, and a machine that turns discarded heat directly into cold attacks the crisis from both ends — trimming electricity demand while putting neglected thermal energy to work. For a hundred years, cooling has been the story of electricity pushing heat around through chemical refrigerants. Two ribbons of nickel-titanium, driven by nothing more exotic than waste heat or sunlight, have now demonstrated a different ending: heat itself, set against heat, becomes cold.