How high-temperature superconductivity left the helium room

September 3, 2026

Donald Pooke, CEO of HTS-110, discusses how advances in high-temperature superconducting (HTS) technology are changing the magnet game.

Earlier this year, two of our engineers carried a superconducting magnet across a laboratory and set it down on a bench. The whole system weighs 56 kilograms. It runs from a standard single-phase wall socket; it is air-cooled, and it holds a stable field of up to three tesla, matching the field strengths of clinical MRI scanners. What matters as much as the numbers is that it is entirely self-contained: not a magnet tethered to a separate compressor and a rack of services, but a single standalone instrument you can pick up and put down.

One superconducting magnet we shipped back in 2009 to Brookhaven National Laboratory in New York was a retrofitted synchrotron storage ring magnet. It was, at the time, the first known HTS retrofit of an accelerator magnet, and getting it there was an international achievement. But it was a different kind of magnet doing a different job, built as a lower-field magnet to make a point about energy.

In a synchrotron, copper bending and steering magnets define the storage ring the electrons travel around. They run at modest fields, but because copper is resistive, a ring of them can easily draw on the order of a megawatt of power simply to stay energised. Our argument was straightforward: replace the copper with HTS coils and you save energy. Even putting a cryocooler on every magnet leaves you ahead, and centralised cooling puts you further ahead still. We built demonstrators on that premise, one at Brookhaven and one in Taiwan. Brookhaven mapped the magnet’s field and confirmed it reproduced the field of the original copper coils, at 3.9 kilowatts against the copper magnet’s 13.1.

Fig. 2: The 3 tesla benchtop magnet system on the bench at the Experimental Nuclear Magnetic Resonance Conference (ENC 2026), Pacific Grove, California, where it was unveiled.

The common thread connecting that massive synchrotron magnet to the one on the bench is the ongoing quest for operating efficiency. It traces back to a single physical fact, and almost everything else follows from it.

The one idea: A few degrees warmer

Superconductors carry electrical current with zero resistance, but only when they are cold enough. How cold is the whole story.

Conventional low-temperature superconductors, such as niobium-titanium and niobium-tin, are usually operated at around four kelvin, and certainly below ten, to carry a useful current. For a long time that meant a bath of liquid helium. as although today, four-kelvin cryocoolers exist and many LTS laboratory magnets today run ‘dry’ on a cryocooler rather than in a helium bath. But what made dry magnets practical in the first place was HTS technology itself. LTS magnet systems use HTS current leads to carry the current into the cold space without dragging heat in with it.

Here is the benefit that higher operating temperature changes. A four-kelvin cryocooler typically delivers only a watt or two of cooling, which leaves very little thermal margin. Move up to twenty, thirty, fifty kelvin and the same machine gives you something like ten times or more the cooling power. That headroom is what lets you build a magnet that is robust, that recovers quickly, and that can be ramped hard.

Fig. 3: Nominal published capacity of a Cryomech PT410 pulse-tube cryocooler: about one watt at 4.2 K on the second stage with about 40 watts at 45 K on the first, running simultaneously — roughly 40 times the heat lift in the same machine.

The high-temperature superconductors our magnets are wound from are ceramics, and they are superconductingwhen far warmer than the low-temperature alloys. REBCO, the second-generation conductor we often use today, has a transition temperature above ninety kelvin. BSCCO a first-generation conductor transitions at around 110 kelvin, which is where the “110” in our name, HTS-110, comes from. In practice we run our magnets well above four kelvin. For example our benchtop relaxometry magnet operates around 50-60 kelvin, reachable with a closed-cycle cryocooler that runs off single-phase electricity and never needs refilling with cryogens.

Fig. 4: Specific heat capacity of copper relative to its value at 4.2 K. The heat capacity of copper in the range 20–50 K is approximately 70-900 times greater than its value at 4.2 K. Because HTS coils include significant amounts of copper for conduction cooling, an HTS coil pack operated in that temperature range would typically require tens to hundreds of times more heat energy to increase its temperature by a given amount than would an LTS coil pack operated at 4.2 K.

It is worth understanding why the second-generation conductor is so much better. The reason is structural. These materials are built from layers of copper-oxide planes separated by spacer layers, which form the gap between the conducting planes. The wider that gap, the sooner the resistance-free superconductivity gives out under field; the best conductors have the smallest gap. REBCO has the smallest of all, because part of its structure is itself a copper chain, which is why it sustains the highest fields at the highest temperatures. The work that identified this behaviour was done by Jeff Tallon and others at New Zealand Institute for Industrial Research, and it is the reason REBCO magnets have become the new frontier for high-field performance.

Fig. 5: The unit cell of YBCO, drawn schematically. Superconductivity forms in the copper-oxide planes. The block layer separating one pair of planes from the next — barium oxide, copper-oxide chain, barium oxide — sets the spacing d₈.

Fig. 6: Selected points after Tallon, Williams and Loram (1999); the grey line is their published exponential fit. Y-123 appears at both its true block-layer spacing (0.824 nm) and its effective spacing (0.412nm), halved because its copper-oxide chains are metallic. Ca-123 — the same compound with that chain order destroyed — and Bi-2223 sit near the trend at their true spacings, as measured data does.

What the gap buys you

Operating warmer is not the point in itself; it is the enabler for a set of features we design our HTS magnets for. Each one tends to matter most to a particular kind of user.

Compact, with access: HTS conductors carry high current densities even at high field, so you can pack more turns into a smaller envelope. This leaves more optical and physical access to the central field for the user, while passive shielding keeps stray fields contained so magnets can be located near sensitive instruments.
Fast ramping: The large thermal margin that comes with operating warmer lets us ramp a magnet quickly (so far, at up to 0.5 tesla per second). For a fast-ramp dipole, this means more measurements per shift. HTS gives us the headroom to do that comfortably.
Higher field: This is one physical advantage of HTS over LTS. Low-temperature conductors run out of capability somewhere around 20 tesla. HTS keeps going well beyond: coils above 40 tesla have already been demonstrated, with the material’s ceiling far higher again.
Simpler running, lower cost: A cryocooler-driven HTS magnet has no helium bath to maintain and no top-up schedule. Operating warmer simplifies the cryostat itself, and reduces the energy requirement of the cryocooler. Taken together with no cryogen to buy, reduced maintenance, and lower energy requirements, this keeps the total cost of ownership down. These magnets run in driven mode rather than the persistent mode of an LTS system, so they stay powered and controlled — an arrangement now well validated for high-resolution spectroscopy and process work.
Field quality: In high-resolution NMR, as in our 300 and 400 MHz systems, HTS magnets deliver stable fields that shim to sub-ppm homogeneity – typically better than 5 ppm bare, then brought under 1 ppm with standard shim sets and a field lock.

Fig. 7: Superconducting field records by conductor. Outline bars mark the NbTi conductor’s upper critical field (11.5 T at 4.2 K; 14.5 T at 1.9 K) — practical NbTi magnets run lower. Solid bars are built magnet records: 23.5 T for the best all-LTS magnet; 32.4 T for an all superconducting magnet with a REBCO insert; and 48.7 T for a REBCO coil operating inside a 31 T resistive background magnet.

From the frontier inward

HTS-110 was formed in 2004 to commercialise two decades of high-temperature superconductivity research carried out in New Zealand. For most of the company’s life, the work concentrated on developing the most demanding instruments at the largest facilities, like the Brookhaven dipole. But there has always been a second direction I find more interesting. Once you no longer need helium, we can ask: where else can superconductivity go if it were allowed to leave the specialist facility altogether?

The benchtop moment

The three-tesla magnet I opened with is our clearest answer to that question. It is the heart of a new benchtop NMR relaxometer we have launched with our long-standing Italian partners at Stelar, who build the instrument around it.

The application is NMR relaxometry, a technique for investigating molecular dynamics, used in everything from developing MRI contrast agents to analysing battery electrolytes and checking food quality. Researchers in this field told us something simple: they did not want to walk across campus to a shared NMR facility; they wanted an instrument next to their bench in their own laboratory.

Fig. 8: The new benchtop relaxometry magnet in use, and the first-generation system it replaces (inset). The earlier magnet needed more than four kilowatts of three-phase power, facility water and a substantial installation. The new one runs from single-phase power with air cooling and no cryogenic liquids.

Our first-generation magnet for this work could do the physics, but it needed more than four kilowatts of three-phase power and a substantial installation. The new one needs none of that, relying instead on single-phase power, air cooling, no facility water, and no cryogenic liquids. We did it by moving to the second-generation REBCO conductor and shrinking the cryogenic envelope, along with the cooling and power it demands. It delivers the same field with a fraction of the facility burden.

Where warmer magnets go next

I do not want to overstate where this leads, as high-temperature superconductivity is a field with many players and a long way still to run. The largest prizes, fusion energy among them, are being chased by many great teams. What I would argue is narrower and, I think, just as durable: the value of operating a few degrees warmer compounds over time.

Every gain in conductor performance and every reduction in cryogenic overhead pushes the technology further out of the specialist facility and closer to the people who need the measurement. A magnet that used to define a building can now sit on a bench. That direction of travel – warmer, smaller, simpler – is the part of this story I would bet on continuing.

References

J. Muratore et al., ‘Magnetic Field Measurements of an HTS Retrofit Synchrotron Dipole’, IEEE Transactions on Applied Superconductivity (2011); Brookhaven National Laboratory report BNL-94492-2010-JA.
J. L. Tallon, G. V. M. Williams and J. W. Loram, arXiv:cond-mat/9911423 (1999).
J. L. Tallon, G. V. M. Williams, C. Bernhard, D. M. Pooke and co-workers, Physical Review B 53, R11972(R) (1996).
P. E. Bradley and R. Radebaugh, ‘Properties of Selected Materials at Cryogenic Temperatures’, NIST Publication 913059.
PT410 pulse-tube cryocooler, manufacturer’s published capacity specification. Cryomech, now part of Bluefors; bluefors.com.
S. Hahn et al., ‘45.5-tesla direct-current magnetic field generated with a high-temperature superconducting magnet’, Nature 570, 496–499 (2019).
National High Magnetic Field Laboratory, 48.7-tesla field with a REBCO insert coil in a resistive background magnet (2025); nationalmaglab.org.
J. Liu et al., ‘World record 32.35 tesla direct-current magnetic field generated with an all-superconducting magnet’, Superconductor Science and Technology 33, 03LT01 (2020).
M. N. Wilson, Superconducting Magnets, Oxford University Press; with manufacturer-published specifications of
10. GHz-class LTS NMR systems.