The signal

On May 5th the US patent office handed Volkswagen a patent for a heat pump that has no refrigerant. There is no R-134a inside, no R-1234yf, no hydrofluorocarbon of any flavor. There is a strip of nickel-titanium alloy that gets cyclically squeezed between two coolant channels by a pair of carriages riding on guide rails. When the alloy is stretched, it cools. When it relaxes, it warms. That is the whole machine. The patent is US 12,617,250, the carmaker filed it in Germany in 2023, and it is the first elastocaloric heat pump claim ever granted to an auto OEM.

Carrier, the world’s largest air-conditioning company, has been on a parallel path. As of this month, sixteen of Carrier’s US patents are for variants of solid-state cooling that swap refrigerant for a stressed alloy, an electrically polarized polymer, or a pressurized plastic crystal. The most recent one, an electrocaloric fiber that can be woven into fabric, was granted in January. In May 2024, Carrier Ventures put €15 million into Exergyn, the Irish shape-memory-alloy startup that has been working on the same problem since 2012, in a round that valued the company north of €200 million. Exergyn has now raised $64.3 million across ten investors. None of this has been front-page news.

And in Cambridge, Massachusetts, on January 6th, a Harvard chemist named Jarad Mason was granted US patent 12,516,892 for a phenomenon called the “inverted barocaloric effect”: a way to drive a cooling cycle in a plastic crystal at less than 100 bar of pressure, where the prior art needed thousands. That is the engineering breakthrough that turns a lab curiosity into something a compressor can plausibly push.

Four different physics, three different US patents in the last five months, one strategic capital deployment by the biggest HVAC company in the world. They are all pointed at the same target: a refrigerator with no refrigerant inside.

What’s actually inside these patents

The shared engineering DNA matters because it’s the only reason this counts as a story. All four of these inventions exploit the caloric effect: a solid material that absorbs or releases heat when you change its entropy by applying an external field. The field can be magnetic (a magnetocaloric alloy demagnetizes and cools), electric (an electrocaloric polymer depolarizes and cools), mechanical (a shape-memory alloy unloads and cools), or hydrostatic (a plastic crystal decompresses and cools). The thermodynamics is the same. The hardware is the same: a solid working medium plumbed between a hot side and a cold side, cycled by some kind of actuator.

Read the actual claims and you can see the convergence in real time. Volkswagen’s patent describes a NiTi strip riding between two cooling channels on a four-bar mechanism. Exergyn’s US 12,111,084, granted last October, is a tubular SMA core with a fluid pump and “buckling supports” to keep the alloy from collapsing under repeated compression, a years-old failure mode in elastocaloric prototypes that the patent finally addresses. Iowa State’s US 12,158,291 takes a different shot at the same problem: laminate the NiTi strips into a polymer composite that bends, so the metal only ever sees pure compression or pure tension, never the unstable axial loading. The paper underlying that patent reports a 50% reduction in the force needed to drive the cycle.

MIT’s US 12,270,610, granted in April 2025 to Asegun Henry’s group, plumbs the pressure-driven version with a high-pressure and low-pressure fluid source feeding a bed of “barocaloric material … in particulate form.” The examples named in the claims include natural rubber, acetoxy silicone, and neopentyl glycol, a chemical you can buy by the kilogram. Mason’s Harvard patent then sharpens the cycle further: the inverted effect lets it run at pressures a normal hydraulic system can produce, which is the difference between a science demonstration and a unit you could mount in a building.

None of these patents are about software. They are about a piece of metal or polymer or crystal that gets hot when you push on it, and the boring mechanical engineering of cycling that push at hundreds of hertz without breaking the material.

Why now

The deadline is real. The EU’s revised F-gas Regulation, 2024/573, took effect on March 11, 2024. It mandates an 80% cut in hydrofluorocarbon supply by 2030 against the 2011–13 baseline, and from 2030 the global warming potential of any refrigerant in a new stationary system other than a chiller is capped at 150. That cap rules out the entire current generation of HFCs. The Kigali Amendment, ratified by 159 countries, runs the same phase-down globally with a longer fuse.

This is the regulatory cliff every HVAC executive has on a whiteboard somewhere. Cooling is, per the International Energy Agency, about 10% of global electricity consumption today and the single largest source of buildings-sector demand growth through 2035. The industry has to find something to put in the box that isn’t an HFC, isn’t propane (flammable, regulatory headache in dense buildings), and isn’t CO2 (works, but requires supercritical operating pressures that don’t fit existing form factors). Caloric materials are the only option on the table where the working medium cannot leak by definition: it’s a solid.

The science has been racing toward this for two decades. Papers indexed in OpenAlex that pair an elasto/magneto/electro/barocaloric term with cooling or refrigeration went from 12 in 2000 to 499 in 2025. The slope steepened sharply after 2017, when groups in Spain (Lluís Mañosa at Barcelona) and the UK (Xavier Moya at Cambridge) reported “colossal” barocaloric effects in cheap plastic crystals like neopentyl glycol, with entropy changes of 380 to 510 joules per kilogram per kelvin, comparable to the best magnetocaloric alloys at a fraction of the cost. That 2019 Nature paper is the moment the field stopped being exotic.

Who’s positioned

The patent positioning tells you who is taking it seriously. Carrier’s sixteen-patent portfolio spans all four physics: magnetocaloric regenerators, electrocaloric polymer films, electrocaloric fibers, and elastocaloric heat exchangers. That is the strategy of someone who doesn’t yet know which horse wins and refuses to bet on only one. Haier US Appliance Solutions (the former GE Appliances) has nine. Mitsubishi Electric has four. United Technologies, before its merger into RTX, filed at least three. Volkswagen now has the only auto-OEM caloric grant in the file.

On the lab side, Iowa State has six US patents and an active Department of Energy ARPA-E line. The University of Maryland’s Ichiro Takeuchi, who served as principal investigator on a long-running NSF/DOE elastocaloric program with Reinhard Radermacher and Yunho Hwang, has spun out Maryland Energy & Sensor Technologies LLC and is its CTO. France’s CEA holds a paramagnetic garnet ceramic patent. The Technical University of Denmark has three. Cambridge Enterprise, the licensing arm of Cambridge University where Moya works, holds at least one.

The most interesting recent commercial signal is the EU’s SMACool consortium, funded at €3.9 million through the European Innovation Council in 2024. SMACool’s stated goal is a residential air conditioner built around Exergyn’s NiTi cores, with the Universities of Naples and Ljubljana on the science side. It is the first government-backed program in the world to target a household caloric AC by a date certain: September 2027.

What an R&D director should take from this

The headline isn’t that solid-state cooling is coming. The headline is that the major HVAC, appliance, and auto incumbents have already finished the “if” debate and are doing the “which” debate, in public, on the patent record. Carrier’s portfolio is the tell: when the largest player buys options across four mutually exclusive physics, they are positioning to license rather than invent.

The window to be a supplier, not a target, is narrow. The materials side (magnetocaloric alloys from BASF and Vacuumschmelze, NiTi wire and ribbon from specialty SMA mills, high-purity neopentyl glycol, electrocaloric ceramics) is largely held by a handful of European and Japanese metallurgical specialty firms. Anyone in that supply chain in the US should be reading EU F-gas rule annexes the way semiconductor people read CHIPS Act schedules. Anyone selling vapor-compression compressors into stationary refrigeration should be reading them faster.

Method

This brief is built from US utility patent grants in the USPTO bulk grant feed and scholarly works indexed in OpenAlex, queried in mid-May 2026. Patent counts cover the search terms elastocaloric, magnetocaloric, electrocaloric, and barocaloric in title or abstract; the 2015–2025 total is 182 grants. Assignee tallies use the USPTO-supplied assignee field with simple name normalization (Carrier-affiliated entities collapsed; Haier US Appliance Solutions and pre-acquisition GE Appliances variants counted together). The Harvard, MIT, Volkswagen, and Iowa State patents were verified directly via the published documents. Literature counts come from OpenAlex full-text search across all venues for any caloric term paired with cooling, refrigeration, or heat pump. Regulatory references are to EU Regulation 2024/573 and IEA Future of Cooling. The Carrier Ventures / Exergyn round figures are from Cooling Post and PitchBook; Exergyn cumulative fundraising figure ($64.3M) is from PitchBook as of May 2026.