Air conditioners and refrigerators quietly eat a huge share of the world’s electricity, and the gases inside many of them trap heat in the atmosphere. A team at the Hong Kong University of Science and Technology (HKUST) now reports a step toward a different kind of cooling: a solid metal that gets cold when it is squeezed, built in a way that is cheap, fast to make and tough enough to run for more than 100 million cycles.
The work, published in the journal Science and described in an HKUST press release dated November 6, 2026, comes from a group led by Prof. Sun Qingping of the Department of Mechanical and Aerospace Engineering. It targets one of the biggest obstacles for this technology: manufacturing a reliable cooling element at a price that makes mass production realistic.
How elastocaloric cooling works
Elastocaloric cooling relies on shape memory alloys, metals that change their internal crystal structure when stretched or compressed. According to HKUST, the reversible phase change absorbs and releases heat, so loading and unloading the alloy can pump heat from one place to another. Because the working material is a solid, the approach avoids the liquid refrigerants used in conventional vapor-compression systems, which the university notes are energy-intensive and can contribute to the greenhouse effect.
Researchers have demonstrated strong cooling capacity in the lab for years. The sticking point, as the HKUST team explains, is the refrigerant structure itself. Most designs use commercial tubes, plates or wires, or thin-walled parts cut by wire electrical discharge machining (WEDM). Those routes tend to be expensive and slow, and the parts are prone to fatigue cracks and buckling when they are loaded repeatedly.
The “Stamping and Stacking” method
The HKUST group’s answer is deliberately simple. Instead of machining a long, delicate tube, the team takes commercial shape memory alloy sheets, stamps them into thin porous chips, and stacks the chips to form a tubular refrigerant structure. The layout is inspired by the stacked, porous architecture of bamboo and honeycombs, which are known for combining strength with low weight.
The press release highlights three advantages:
- Fast, cheap production. Stamping can make 100 chips per minute, which HKUST says is more than 100 times faster and 95% less expensive than the WEDM approach.
- Durability. Because the structure is assembled from separate chips, the interfaces between neighbors help stop cracks from spreading along the whole length and prevent catastrophic splitting. In tests, it withstood more than 100 million cycles, which the team estimates equals a service life above 10 years under the tested operating conditions.
- Useful cooling. A 30-millimeter unit made of 200 stacked chips powered a prototype that reached a maximum temperature span of 11.7 K and a specific cooling power of 3.85 W per gram.
The paper is titled “Stacked porous NiTi structure with low cost and ultrahigh fatigue life for elastocaloric cooling,” and the alloy at the center of it is nickel-titanium (NiTi), a widely used shape memory material.
A third result in a year
HKUST describes this as the team’s third breakthrough in about twelve months. In January, the group reported what it called the world’s first sub-zero elastocaloric freezer, capable of reaching -12 degrees Celsius, published in Nature. In August, it reported a fatigue-resistant NiTiCuCo alloy and a cooling device with no measured degradation, published in Joule. Together, the three papers address cooling performance, material lifetime and, now, manufacturing cost.
The team adds that the approach is not tied to stamping alone. Similar structures could be made from thin alloy sheets that are laser cut or etched, which could give manufacturers several routes to the same design. The work was supported by Hong Kong’s Research Grants Council and the Innovation and Technology Commission’s RAISe+ Scheme, and the team has filed multiple international patents.
What it means
Cooling demand is rising as the world warms, and that creates a feedback problem: more air conditioning means more electricity and, with today’s refrigerants, more potent greenhouse gases. A solid-state alternative that needs no liquid refrigerant could eventually ease both issues, but only if it can be built cheaply and last for years in a real appliance.
That is why a manufacturing result matters as much as a record temperature. A cooling element that costs far less to produce and survives a decade of operation moves elastocaloric technology a little closer to products people can buy. It is still early. The prototype’s 11.7 K temperature span is a laboratory figure, and the team itself says it must still improve heat transfer, develop more energy-efficient actuation systems and engineer higher-performance alloys before the technology can compete on efficiency and capacity.
For businesses and consumers, the practical takeaway is to watch for pilot products rather than expect shelf-ready fridges soon. For researchers, the stacked-chip idea offers a template that others could adapt to different alloys and cutting methods.
Bottom line
HKUST’s stamped-and-stacked cooling structure does not make elastocaloric refrigerators a reality overnight, but it removes a key barrier: cost and durability at scale. If follow-up work can lift efficiency and cooling capacity, solid metal could one day take over some of the work now done by greenhouse-gas refrigerants.