A high-temperature superconductor is a material that conducts electricity with zero resistance at temperatures above 77 K (−196 °C), the boiling point of liquid nitrogen, which makes cooling far cheaper than with the liquid helium needed by conventional superconductors (Wikipedia: High-temperature superconductivity). The best-known examples are copper-oxide ceramics called cuprates, such as YBCO, which superconducts at about 93 K (Wu et al., Phys. Rev. Lett. 1987). Iron-based compounds, magnesium diboride and compressed hydrides extend the family, but no material has yet been confirmed to superconduct at room temperature and ambient pressure.

• "High-Tc" usually means a critical temperature (Tc) above 77 K, so the material can be cooled with liquid nitrogen instead of liquid helium.
• Cuprates hold the ambient-pressure record: mercury-based cuprates superconduct above 130 K (Schilling et al., Nature 1993), and a 2026 study reported 151 K in pressure-quenched Hg-1223 (Deng et al., PNAS 2026).
• Iron-based superconductors, whose higher-Tc members were found in 2008, reach about 55 K in bulk and are a second family of unconventional superconductors (Ren et al., Chin. Phys. Lett. 2008).
• Hydrides such as H3S and LaH10 superconduct at 203 K and about 250 K, but only at pressures well above a million atmospheres (Drozdov et al., Nature 2015; Drozdov et al., Nature 2019).
• LK-99 was not a superconductor: independent work traced its behavior to a copper sulfide impurity and found pure crystals to be insulating.
• The main uses are magnets and cables, including compact fusion magnets, while most MRI scanners still use low-temperature niobium-titanium wire.

What does "high-temperature" mean in superconductivity?

Superconductivity is a state in which a material carries current with no electrical resistance and expels magnetic fields (the Meissner effect). It appears only below a material-specific critical temperature, Tc. For most of the twentieth century, known superconductors had Tc values of a few kelvin to a few tens of kelvin. Niobium-titanium, the workhorse magnet alloy, has a Tc of about 10 K (Wikipedia: Niobium-titanium), so it must be cooled with liquid helium.

"High temperature" is therefore relative. The conventional threshold is 77 K, the boiling point of liquid nitrogen (Wikipedia). Above that line, cooling becomes a matter of a cheap, abundant cryogen rather than scarce helium. Some writers use "high-Tc" loosely to mean any of the cuprate or iron-based families, even when a particular compound's Tc is below 77 K. In this article we use the strict meaning where it matters and name the family otherwise.

What are cuprate superconductors?

Cuprates are layered copper-oxide ceramics. Their defining structural unit is the CuO2 plane, a square lattice of copper and oxygen in which superconducting charge carriers move. Layers of other cations (barium, yttrium, bismuth, strontium, mercury and so on) sit between the planes and act as charge reservoirs. The parent compounds are antiferromagnetic insulators; superconductivity appears only when carriers are added by doping, for example by changing oxygen content.

The family began in 1986, when J. Georg Bednorz and K. Alex Müller reported possible high-Tc superconductivity in the Ba-La-Cu-O system (Z. Phys. B 1986), at around 35 K (Wikipedia). They received the 1987 Nobel Prize in Physics for the discovery (Wikipedia). The parent of that system, La2CuO4, remains a textbook reference compound.

YBCO

In 1987, Wu, Chu and colleagues reported superconductivity at 93 K in the Y-Ba-Cu-O system at ambient pressure (Phys. Rev. Lett. 1987), the first material found to superconduct above 77 K (Wikipedia: YBCO). In YBa2Cu3O7 (YBCO), Tc depends strongly on oxygen content: YBa2Cu3O7−x superconducts only for roughly 0 ≤ x ≤ 0.65 and peaks near x ≈ 0.07 (Wikipedia: YBCO). YBCO and related rare-earth compounds (often written REBCO) are the basis of "coated conductor" tapes, in which a thin superconducting film is deposited on a flexible metal tape with buffer layers (Wikipedia: YBCO).

BSCCO

Bismuth strontium calcium copper oxide was reported around 1988 by Hiroshi Maeda and colleagues and was the first high-Tc superconductor without a rare-earth element. Its members have Tc values of about 33 K (Bi-2201), 96 K (Bi-2212) and 108 K (Bi-2223) (Wikipedia: BSCCO). Because its plate-like grains align during mechanical deformation, BSCCO became the material for first-generation (1G) superconducting wire.

Mercury cuprates and the ambient-pressure record

Schilling and co-workers reported superconductivity above 130 K in the Hg-Ba-Ca-Cu-O system in 1993 (Nature 1993). Under quasi-hydrostatic pressure the same family reached up to 164 K (Gao et al., Phys. Rev. B 1994). For about three decades the three-layer compound HgBa2Ca2Cu3O8+δ (Hg-1223), at around 133 K, held the ambient-pressure record (Wikipedia). In 2026, Chu and colleagues at the University of Houston reported a record ambient-pressure Tc of 151 K in Hg-1223 using a "pressure-quench" protocol that retains a pressure-enhanced superconducting state after the pressure is released (Deng et al., PNAS 2026). Because it relies on specially processed, metastable samples, independent replication is the next test.

Four decades after their discovery, there is still no consensus theory of how electrons pair in cuprates. Conventional Bardeen-Cooper-Schrieffer (BCS) theory, in which lattice vibrations (phonons) glue electrons into pairs, does not account for their behavior, and magnetic interactions are widely thought to play a central role.

What are iron-based superconductors?

In 2008, Hideo Hosono's group reported superconductivity at 26 K in fluorine-doped LaFeAsO (Kamihara et al., J. Am. Chem. Soc. 2008). Within months, substituting samarium for lanthanum raised Tc to 55 K (Ren et al., Chin. Phys. Lett. 2008). The discovery surprised many researchers because iron, being strongly magnetic, was expected to suppress superconductivity.

Iron-based superconductors are grouped by structure: the 1111 type (such as LaFeAsO), the 122 type (such as BaFe2As2), the 111 type (such as LiFeAs) and the 11 type (FeSe) (Wikipedia: Iron-based superconductor). All share layers of iron tetrahedrally coordinated by pnictogen (such as arsenic) or chalcogen (such as selenium) atoms. Unlike the cuprates, their parent compounds are usually metals rather than insulators, several electronic bands cross the Fermi level, and the leading pairing proposal is an "s±" state in which the sign of the superconducting gap differs between bands (Mazin et al., Phys. Rev. Lett. 2008). Bulk FeSe superconducts at only about 8 K, but single layers grown on SrTiO3 show much higher onset temperatures, a result that has driven intense study of interface effects (Wikipedia).

Strictly, iron-based compounds fall short of the 77 K liquid-nitrogen line in bulk form, but they are usually discussed alongside the cuprates as the second major family of unconventional high-Tc superconductors.

How does MgB2 fit in?

In 2001, Nagamatsu and colleagues reported superconductivity at 39 K in MgB2, a simple, inexpensive compound that had been sitting on chemical shelves for decades (Nature 2001). Unlike the cuprates, MgB2 is understood as a phonon-mediated (BCS-type) superconductor with two distinct superconducting gaps. Its Tc is well below 77 K, but it can be operated with cryocoolers or small amounts of helium, and it has been studied for conduction-cooled MRI magnets and cables.

Do hydrides superconduct near room temperature?

Hydrogen-rich compounds under extreme pressure currently hold the overall Tc records. Drozdov, Eremets and colleagues reported conventional superconductivity at 203 K in the sulfur hydride system (Nature 2015), supported in the same study by zero resistance, an isotope effect and magnetic susceptibility measurements (arXiv:1506.08190). The pressure was about 155 GPa (Wikipedia), and the superconducting phase is identified as H3S. In 2019 the same group reported Tc of about 250 K in LaH10 at roughly 170 GPa (Nature 2019; arXiv:1812.01561), roughly 1.7 million atmospheres. In 2025, researchers posted a preprint reporting superconductivity with onset temperatures of 271–298 K in LaSc2H24 at 195–266 GPa (Song et al., arXiv:2510.01273); if it is independently reproduced, it would be the first room-temperature superconductor, though still only at extreme pressure.

Hydrides are notable because their high Tc values were anticipated by theory: light hydrogen atoms produce high-frequency phonons, and density functional theory (DFT) calculations of electron-phonon coupling helped guide the experiments. They are not practical conductors, however, since samples are micrometer-scale and exist only inside diamond anvil cells.

The field has also seen high-profile retractions. A 2020 Nature paper claiming room-temperature superconductivity in a carbonaceous sulfur hydride (Nature 2020, retracted) and a 2023 paper claiming near-ambient superconductivity in nitrogen-doped lutetium hydride (Nature 2023, retracted) were both withdrawn. The H3S and LaH10 results are separate work and have not been retracted.

What happened with LK-99?

On July 22, 2023, Sukbae Lee, Ji-Hoon Kim and Young-Wan Kwon posted a preprint titled "The First Room-Temperature Ambient-Pressure Superconductor," claiming a Tc of at least 400 K (127 °C) in a copper-doped lead apatite they called LK-99 (arXiv:2307.12008). The claim drew worldwide attention and rapid replication attempts.

The resolution came within weeks. Prashant Jain pointed out that copper(I) sulfide, a by-product of the reported synthesis, undergoes a phase transition at 104 °C that produces sharp changes in resistivity and heat capacity resembling the LK-99 data (arXiv:2308.05222). A team at the Max Planck Institute for Solid State Research grew phase-pure single crystals of Pb10−xCux(PO4)6O and found them highly insulating and optically transparent, not superconducting (Puphal et al., arXiv:2308.06256). In December 2023, a verification committee of the Korean Society of Superconductivity and Cryogenics concluded that there was no evidence that LK-99 is a room-temperature, ambient-pressure superconductor (Korea Herald, 2023).

The episode is a useful reminder of the standard of evidence for superconductivity: zero resistance and a bulk Meissner effect, measured on well-characterized phase-pure samples and reproduced independently. Partial levitation or a resistance drop alone is not enough.

How do the main superconductor families compare?

What are high-temperature superconductors used for?

The largest superconducting markets still rely on low-temperature materials. Niobium-titanium is used in most MRI systems, which accounted for about 80% of the superconductivity market's value in 2014, and the Large Hadron Collider magnets contain 1,200 tonnes of Nb-Ti cable (about 470 tonnes of it Nb-Ti alloy, the rest copper) cooled to 1.9 K to reach fields up to 8.3 T (Wikipedia: Niobium-titanium). Nb3Sn, an A15-structure compound, is used where higher fields are needed but is harder to fabricate.

High-temperature superconductors matter most where very high fields or higher operating temperatures change the engineering. The clearest recent example is fusion: in September 2021, MIT's Plasma Science and Fusion Center and Commonwealth Fusion Systems ramped a large magnet built from high-temperature superconducting tape to 20 tesla (MIT News, 2021), a field strength that enables more compact tokamak designs. Other uses include power cables and fault-current limiters in electrical grids, high-field research and NMR magnets, and current leads that feed low-temperature magnets.

The practical obstacles are mostly materials problems. Cuprates are brittle ceramics; their current-carrying capacity drops sharply across misaligned grain boundaries, which is why coated conductors require carefully textured films; and conductor cost per unit of current remains higher than for Nb-Ti.

What are the open questions?

• Mechanism: there is still no agreed microscopic theory for cuprate superconductivity, and the role of competing orders (charge density waves, the pseudogap) is debated.
• Ambient-pressure hydrides: whether hydride-like high Tc can be stabilized at low or ambient pressure is a central target of computational screening.
• Prediction: DFT plus electron-phonon calculations can estimate Tc for conventional superconductors, but no reliable first-principles method exists for unconventional families. Computed structures and stabilities are a starting point, not a Tc prediction.
• Reproducibility: the LK-99 episode, the hydride retractions and the still-unreplicated record claims show why independent replication and full data release matter.

Frequently asked questions

What is the highest-temperature superconductor?

At ambient pressure, mercury-based cuprates hold the record: about 133 K in conventionally prepared Hg-1223 (Wikipedia) and 151 K reported in 2026 for pressure-quenched Hg-1223 (Deng et al., PNAS 2026). Under pressures near 170 GPa, LaH10 superconducts at about 250 K (Nature 2019), and a 2025 preprint reports onsets up to 298 K in LaSc2H24 at 195–266 GPa (arXiv:2510.01273).

Is there a room-temperature superconductor?

None has been confirmed at ambient pressure. LK-99 and two hydride papers that were later retracted did not hold up, and a 2025 report of 298 K in LaSc2H24 requires pressures near 200 GPa or more and is still a preprint (arXiv:2510.01273).

Why are cuprates called unconventional superconductors?

Their electron pairing is not explained by standard phonon-based BCS theory, and their parent compounds are antiferromagnetic insulators. The pairing has d-wave symmetry rather than the simple s-wave form of conventional superconductors.

Is MgB2 a high-temperature superconductor?

Not by the 77 K definition. Its Tc of 39 K (Nature 2001) is high for a conventional superconductor, and it is attractive for low-cost wire.

Why don't MRI scanners use high-temperature superconductors?

Nb-Ti wire is mature, ductile and cheaper per unit of current, so it remains the standard. High-temperature superconductors and MgB2 are studied for designs that reduce liquid helium use.

Exploring superconductor data on LatticeGraph

LatticeGraph aggregates computed (DFT) structures and stability data for compounds in the cuprate, iron-based and A15 classes, with literature synthesis recipes and related patents. Computed values are not measured Tc data; for how databases can disagree, see why DFT databases disagree and 13 ways computational materials science goes wrong.