Can Superconductivity Be Achieved at Room Temperature and Ambient Pressure?

The Resistance That Won't Die

In a basement laboratory in Leiden in April 1911, Heike Kamerlingh Onnes watched the electrical resistance of a thread of frozen mercury vanish—not fade, not diminish, but drop to a value indistinguishable from zero—the moment his thermometer crossed 4.2 kelvin. He wrote the observation in his lab notebook with the flat understatement of a man who did not yet know he'd found a hole in the fabric of solid-state physics: kwik nagenoeg nul — "mercury practically zero." Practically zero was, it turned out, actually zero. A current started circulating in a superconducting ring can, in principle, run for longer than the age of the universe without measurable decay.

More than a century later, we still cannot make that happen in a wire on your desk at room temperature, in open air, for free. We can get superconductivity at temperatures that would have stunned Onnes—above the boiling point of liquid nitrogen, even flirting with the temperature of a chilly day outside—but only in materials squeezed to pressures found near the center of the Earth, or in compounds too delicate and expensive to wind into a magnet coil. The dream that has haunted condensed matter physics for a hundred years is deceptively simple to state: a material that carries electricity with zero resistance, at 20°C, at one atmosphere of pressure, cheaply and stably enough to rewire civilization. It remains, stubbornly, a dream.

This is one of the great open problems of matter and energy, not because we lack superconductors — we have thousands — but because we do not fully understand, even now, what rules govern how high the transition temperature can go, and whether nature permits it to go all the way to room temperature without extreme pressure as the toll for admission.

svg[The generic cuprate phase diagram: doping away from the parent antiferromagnetic insulator opens a superconducting dome, but the region above it — the pseudogap and the "strange metal" — remains only partly understood.]

That diagram — antiferromagnetic insulator on the left, an underdoped pseudogap region whose electronic structure is only partly gapped for reasons still argued over, a dome of superconductivity peaking near "optimal doping," and a mysterious strange metal phase overhead where resistivity rises linearly with temperature all the way from near-zero to room temperature and beyond, defying ordinary metal physics — is the central object that any theory of cuprates has to explain. It connects directly to the modern physics of correlated electron systems: materials where the electron-electron interaction is not a small correction to free-electron behavior but the dominant physics, and where phases of matter — magnetism, superconductivity, charge order, exotic "strange" metallicity — compete and intertwine on nearly equal energetic footing.

The trouble is this: BCS theory, in its original form, has no room for a Mott insulator as the parent compound of a superconductor, and no obvious mechanism to produce transition temperatures near 100 K from phonons alone in these materials. The electron-phonon coupling in cuprates, measured directly by various spectroscopic probes, appears too weak by most estimates to produce BCS pairing at the observed $T_c$ through phonons doing all the work. And more tellingly, the symmetry of the pairing is wrong for a straightforward BCS story: conventional superconductors pair electrons into an isotropic, "s-wave" state, the gap is roughly the same magnitude in every direction on the Fermi surface, but cuprates pair into a d-wave state — the superconducting gap has four lobes, changes sign around the Fermi surface, and actually vanishes entirely along certain crystal directions. This was established with remarkable clarity through phase-sensitive Josephson junction experiments in the 1990s (notably by Dale Van Harlingen, Donald Ginsberg, and collaborators, and independently confirmed by tricrystal SQUID experiments led by Chang Tsuen and John Kirtley's collaborators at IBM). D-wave pairing is exactly what you'd expect if the pairing glue were not phonons but spin fluctuations — electrons scattering off the fluctuating antiferromagnetic spin correlations left over from the parent Mott insulator. That is suggestive, but suggestive is not proof: three and a half decades of increasingly sophisticated numerical and analytical work have failed to close the case.

The strange metal phase compounds the puzzle. In an ordinary metal, resistivity from electron-phonon scattering rises linearly with temperature at high $T$, then saturates or drops toward a residual value at low $T$, following the physics of Fermi-liquid quasiparticles — well-defined, long-lived electron-like excitations that are the bedrock assumption of almost all of standard solid-state physics. In the strange metal phase of cuprates, resistivity is linear in $T$ from a few kelvin all the way past room temperature, with no sign of saturation, and — most disturbingly — the scattering rate implied by that resistivity approaches a universal bound, roughly $\hbar/\tau \sim k_B T$, called the Planckian limit, as if the electrons are scattering as fast as quantum mechanics will allow. There may be no well-defined quasiparticles at all in this regime — the entire conceptual scaffolding of "electrons as particles that occasionally bump into things" may simply not apply. If the normal state you're pairing from isn't even a normal Fermi liquid, then extending BCS's Cooper-pairing logic — built on the premise of a well-defined Fermi surface with sharp electron and hole excitations — becomes conceptually shaky at its foundation, not just quantitatively.

Add to this the discovery, from 2015 onward, of hydrogen-rich compounds — hydrogen sulfide, then lanthanum hydride (LaH₁₀), then a series of others — that superconduct at temperatures rivaling or exceeding the cuprates, in some cases above 250 K, tantalizingly close to room temperature. These hydride superconductors appear, on current evidence, to be conventional BCS superconductors after all: hydrogen's extreme lightness gives an enormous phonon frequency (recall $T_c$ scales with $\omega_D$ in the BCS formula), and under the crushing pressures used to synthesize them — over a million atmospheres, achieved by squeezing tiny samples between the tips of two diamonds in a diamond anvil cell — hydrogen atoms are forced into dense metallic lattices with exceptionally strong electron-phonon coupling. It's BCS theory, vindicated and pushed to its extreme, rather than refuted. But the achievement comes at a brutal price: those pressures are far beyond anything usable in a wire, a magnet, or a power line. The samples are microscopic, often unstable outside the anvil cell, and difficult even to characterize with full confidence. So the hydride story simultaneously reinforces BCS theory's power and underscores how far we remain from a practical room-temperature superconductor: we may know how to reach the temperature, but not yet how to do it without a diamond anvil doing the equivalent of a small planet's gravity on your sample.

And the field has also had to reckon, painfully, with the credibility of some of its own claimed breakthroughs. A 2020 paper reporting room-temperature superconductivity near 15°C in a carbonaceous sulfur hydride under extreme pressure was retracted by Nature in 2022 after concerns about data processing methods, and follow-up work by other groups failed to reproduce the result. The 2023 sensation around "LK-99," a copper-doped lead apatite claimed by a South Korean team to superconduct at ambient pressure and room temperature, ignited a frenzy of replication attempts within days — and collapsed within weeks, as multiple independent laboratories showed the observed effects (partial levitation, resistance drops) were consistent with ferromagnetic impurities and sample inhomogeneity rather than superconductivity. These episodes are worth naming plainly, not as gossip but as data about the field itself: the prize is so large, and the phenomenon so exacting to verify correctly (you need zero resistance, the Meissner effect, and a sharp specific-heat anomaly all coinciding at the same temperature, ideally in more than one lab, before you believe it), that both honest error and motivated overreach are recurring hazards on this frontier.

No single accepted theory yet explains cuprate high-$T_c$ superconductivity, but there are several serious, actively pursued research programs, and it's worth laying out their appeal and their trouble honestly.

Spin-fluctuation mediated pairing. The leading phenomenological picture treats the cuprates as doped Mott insulators where residual antiferromagnetic spin correlations, rather than phonons, provide the pairing glue — an electron moving through the lattice disturbs the local spin arrangement, and a second electron is attracted to that disturbance. This naturally produces d-wave pairing, matches a great deal of neutron scattering data showing a magnetic resonance mode that tracks superconductivity, and connects cuprates conceptually to other unconventional superconductors — heavy-fermion compounds, some iron-based superconductors, even certain organic superconductors — that also sit near antiferromagnetic instabilities and also pair with sign-changing gaps. Its trouble is that "spin-fluctuation theory" is really a family of related approaches (the fluctuation-exchange approximation, various renormalization-group treatments, slave-boson mean-field theories of the resonating-valence-bond idea originally proposed by Philip Anderson in 1987) rather than one sharp, universally agreed calculation, and getting from the qualitative picture to a quantitatively correct $T_c$ from first principles has proven extraordinarily hard, because the physics is at strong coupling — precisely the regime where the mathematical tools that work beautifully for BCS (perturbation theory in a small coupling constant) break down.

Numerically exact and quasi-exact simulations of the Hubbard model. Much of the theoretical cuprate literature reduces the problem to the Hubbard model — a deliberately minimal toy model of electrons hopping between lattice sites with a single number, $U$, encoding the Coulomb repulsion cost of two electrons sharing a site. It's a caricature of real cuprate chemistry, but a caricature many physicists believe captures the essential physics. Over the last decade, large international collaborations pooling several numerical methods — dynamical mean-field theory, density-matrix renormalization group, determinant quantum Monte Carlo, tensor-network methods — have cross-checked each other's results on this same model with unprecedented rigor. The verdict, as of the mid-2020s, is genuinely double-edged: these calculations do find a superconducting phase with d-wave pairing symmetry in the doped Hubbard model, at least in some parameter regimes and geometries, which is an important consistency check on the spin-fluctuation picture. But they also find that the model's superconductivity competes closely — sometimes losing — to stripe phases (coexisting charge and spin density modulations) and other competing orders, and the results depend sensitively on details (next-nearest-neighbor hopping, the precise value of $U$, ladder geometry versus full two-dimensional lattices) in ways that make some researchers doubt whether the single-band Hubbard model, however faithfully simulated, truly contains all essential cuprate physics, or whether real cuprates need additional ingredients — longer-range interactions, the specific role of apical oxygen atoms, multi-band effects — left out of the minimal model.

Quantum criticality and the strange metal. A different, complementary research program focuses less on the mechanism of pairing itself and more on the strange metal phase from which superconductivity emerges, arguing that the superconductivity might be almost a secondary phenomenon — a low-temperature instability of an even stranger, more fundamental "parent" quantum state. This program draws on the idea of a quantum critical point hidden under the superconducting dome, a zero-temperature phase transition (perhaps involving the pseudogap's mysterious partial gapping, or a hidden order parameter never directly observed) whose quantum fluctuations, fanning out to finite temperature, would naturally produce both the linear-in-$T$ Planckian resistivity and an enhanced, non-phonon pairing glue. This connects, intriguingly and at a deep conceptual level, to ideas from a completely different corner of physics: string-theory-inspired holographic duality techniques (the AdS/CFT correspondence) have been used by theorists including Subir Sachdev, Sean Hartnoll and others to construct toy "holographic superconductors" and strange-metal models that reproduce Planckian scattering, offering a strange, unexpected bridge between gravitational physics and correlated electron matter. It's a beautiful and mathematically rich program, but its critics note, fairly, that it has so far produced qualitative, universal-scaling insight more than material-specific, falsifiable predictions — it can explain that Planckian behavior is natural near quantum criticality, but has not yet nailed down which cuprate compound should have which $T_c$.

Phonon revival and multi-channel pairing. It would be a mistake to write phonons out of the cuprate story entirely. Some researchers argue the mainstream narrative has been too quick to dismiss electron-phonon coupling, pointing to strong isotope effects observed in underdoped cuprates and to angle-resolved photoemission spectroscopy (ARPES) data showing "kinks" in the electron dispersion consistent with coupling to specific optical phonon modes, particularly involving the apical oxygen atoms above and below the copper-oxide planes. The likeliest resolution many researchers now favor is not "phonons versus spin fluctuations" as a binary choice but some cooperative combination — spin fluctuations providing the dominant pairing glue while phonons contribute a supporting, non-negligible boost, the two channels reinforcing each other in a way genuinely new theoretical tools are only beginning to disentangle.

The materials-design program for hydrides and beyond. Running in parallel to the theoretical arguments over cuprates is a distinct, more empirical research program: use machine-learning-accelerated crystal structure search and density functional theory to predict which hydrogen-rich (or boron-, or other light-element-rich) compounds should have the strongest possible electron-phonon coupling, then attempt synthesis under high pressure. This program is squarely built on trusting BCS/Migdal-Eliashberg theory (the more sophisticated, strong-coupling extension of BCS developed by Gerasim Eliashberg) as essentially correct for these hydrides, and using it predictively — an approach that has scored real, independently verified successes, notably the 2015 prediction and subsequent 2019 confirmation of near-room-temperature superconductivity in LaH₁₀ near 250 K under roughly 170 gigapascals. Its central open challenge is not "does the theory work" but "can chemistry be found that keeps this behavior at lower, eventually ambient, pressure" — whether some clever combination of chemical pre-compression (using a cage of other atoms to squeeze hydrogen internally, without needing external mechanical pressure) can substitute for the diamond anvil.

Answering the mechanism question and answering the room-temperature-at-ambient-pressure question require different kinds of evidence, and it's worth separating them.

For the mechanism of cuprate superconductivity, the decisive evidence would most likely come from convergence, not a single knockout experiment: numerical simulations of increasingly realistic multi-band models (going beyond the single-band Hubbard model to include the specific orbital physics of real copper-oxide planes) that reproduce not just d-wave pairing but the correct $T_c$, the correct doping dependence, and the correct connection to the pseudogap and strange-metal phases, cross-validated against ever more precise spectroscopic measurements — resonant inelastic x-ray scattering (RIXS), which can now map spin excitations with a precision unavailable a decade ago, and ARPES with ever finer energy and momentum resolution. Quantum simulators — ultracold atoms trapped in optical lattices engineered to mimic the Fermi-Hubbard model exactly, an approach pursued by groups including Markus Greiner's and Waseem Bakr's laboratories — offer another route: build a defect-free, fully controllable analog of the theorist's toy model in the lab, using actual atoms instead of solving equations on a computer, and simply measure whether it superconducts and how. This sidesteps the approximations of any numerical method, though reaching temperatures cold enough (in units of the relevant hopping energy) to see pairing directly has been a persistent technical struggle, one that has seen real recent progress. A firm, quantitative match between such a "quantum simulation" and real cuprate behavior — or a clear demonstration that the minimal Hubbard model cannot reproduce cuprate $T_c$'s, forcing theorists to admit missing ingredients — would each, in its own way, be genuinely decisive.

For the practical goal of ambient-pressure room-temperature superconductivity, the criterion for "settled" is unambiguous and has been for decades: independent replication, in at least two or three separate laboratories, of a material showing all three simultaneous fingerprints — zero DC resistance below some $T_c \geq 293$ K, the Meissner effect (magnetic field expulsion, ideally shown via both magnetization measurements and direct imaging), and a sharp thermodynamic anomaly (a jump in specific heat) at that same temperature — measured on a sample stable at one atmosphere of pressure. Any claim lacking Meissner data, or resting on unusual single-lab measurements of partial resistance drops or anomalous magnetic susceptibility alone, should be treated as provisional at best, exactly the lesson of the retracted 2020 hydride claim and the LK-99 episode. Given how thoroughly and how fast the community now mobilizes to check extraordinary claims — LK-99 was checked, debated, and substantially debunked within roughly a month, an extraordinary pace for materials science — the infrastructure for verification is arguably in a healthier state now than at any earlier point in the field's history, even though the underlying physics remains as hard as ever.

As for timing: honest researchers in the field decline to give a confident date, and that reticence is itself informative — this is a problem that has resisted focused effort by thousands of very capable physicists for decades. Machine-learning-guided materials discovery, increasingly sophisticated correlated-electron simulation on both classical and early quantum computers, and continued high-pressure hydride chemistry are all active, well-funded, rapidly moving fronts; each could plausibly produce a surprise within the next decade. But "plausibly within a decade" has been said, honestly, for several decades running, and there is no guarantee nature has left room-temperature ambient-pressure superconductivity available at all outside of exotic pressure conditions.

The stakes are not abstract. Roughly five to ten percent of all electricity generated worldwide is lost to resistive heating in transmission and distribution before it ever reaches a light bulb or a factory floor — a bill of hundreds of billions of dollars and a proportional slice of humanity's carbon output, paid every single year simply because copper and aluminum wires are not perfect conductors. A cheap, stable, ambient-pressure room-temperature superconductor would let that loss largely vanish. It would make today's cumbersome liquid-helium-cooled superconducting magnets — the ones inside every MRI scanner, every large particle accelerator, every design for a compact fusion reactor pursuing the goals discussed under controlled nuclear fusion — into simple, cheap, uncooled components, radically shrinking the size and cost of high-field magnets and potentially accelerating the timelines of fusion devices that depend on ever-stronger magnetic confinement. It would transform electric motors, generators, maglev transport, and quantum computing hardware, much of which already depends on superconducting circuits (the superconducting qubits in current quantum processors are themselves a direct descendant of Cooper pairing physics) that currently require dilution refrigerators reaching millikelvin temperatures.

But the deeper stakes are intellectual, and they connect to one of the largest live questions in all of physics: how collections of many strongly interacting quantum particles organize themselves into qualitatively new behavior that cannot be predicted just by knowing the rules governing any one particle. Cuprates are a flagship example of emergence — Mott insulators, antiferromagnets, strange metals and d-wave superconductors all arising from what is, at bottom, simply electrons and ions obeying the ordinary rules of quantum electrodynamics at low energy. Understanding how that emergence works, precisely and predictively, would be a triumph on the scale of BCS theory itself, and the tools built to attack it — the numerical methods for strongly correlated systems, the classification of unconventional pairing symmetries, the study of quantum criticality and Planckian transport — bleed directly into neighboring frontiers, including the study of exotic topological quantum matter, where some of the same correlated-electron machinery is used to hunt for Majorana modes, fractionalized excitations, and other states of matter with no classical analog at all. Solving the cuprate mechanism problem would very likely hand physics a general theory of strongly correlated quantum matter — a prize whose value would extend far past superconductivity itself.

physicist[John Bardeen|1908–1991|Co-developed BCS theory with Cooper and Schrieffer in 1957, becoming the only person ever to win the Nobel Prize in Physics twice — the first for the transistor, the second for superconductivity.]

physicist[Philip W. Anderson|1923–2020|Proposed the resonating-valence-bond theory of cuprates in 1987, arguing the physics of the doped Mott insulator, not phonons, was the key to high-$T_c$ superconductivity — a still-influential, still-contested framework.]

history 1911 — Onnes discovers superconductivity in mercury at 4.2 K. 1957 — Bardeen, Cooper, and Schrieffer publish BCS theory. 1986 — Bednorz and Müller discover cuprate superconductivity above 30 K. 2015 — Eremets and colleagues report superconductivity near 203 K in compressed hydrogen sulfide.

Stand back from the details and a strange, humbling picture emerges. We have superconductors that work near room temperature and we have superconductors that work at ambient pressure — we have simply never had both at once in the same material, verified beyond doubt. It is one of the few genuinely open frontiers in physics where the basic experimental fact — does a working device exist — is as uncertain as the theory behind it, and where the theory of the best-known high-temperature examples, the cuprates, has resisted forty years of extraordinarily talented effort without full resolution. That combination of empirical accessibility (anyone with access to a decent lab could in principle be the one to find the next surprise) and theoretical depth (the problem sits at the genuine edge of what many-body quantum mechanics lets us calculate) is part of why this question exerts such a pull on physicists and public alike. It is not a question awaiting one clever experiment so much as a question awaiting the maturation of an entire toolkit — better simulations, better spectroscopy, better materials chemistry, and perhaps a genuinely new theoretical idea nobody has had yet.

What remains unresolved, stated plainly:

• Is the dominant pairing glue in cuprates spin fluctuations, a cooperative mix of spin fluctuations and phonons, or something not yet identified — and can any theory predict cuprate $T_c$ quantitatively from first principles?
• What is the strange metal phase, physically — is it evidence of a hidden quantum critical point, a breakdown of the quasiparticle concept altogether, or something else entirely?
• Can chemical pre-compression or entirely new light-element compounds deliver hydride-like $T_c$ values without needing hundred-gigapascal external pressure?
• Is there a fundamental thermodynamic or structural reason no stable ambient-pressure material can superconduct above roughly 250–300 K, or is this simply a materials-search problem still awaiting its answer?
• Could a fundamentally new class of pairing mechanism — engineered through synthetic quantum matter, twisted two-dimensional materials, or ideas from topological physics — sidestep the whole cuprate/hydride dichotomy?

Ask the tutor: If spin fluctuations really are the glue holding Cooper pairs together in the cuprates, why has that idea, unlike BCS theory, never produced an equally precise formula for predicting a material's transition temperature — and what would it take to get one?