Tunneling That ClicksBonus ⑥ / Thirty years of room-temperature superconductivity

Two things we want to happen at room temperature; one genuinely came close ── the episode paired against bonus ①

Thirty years of room-temperature superconductivity BCS was supposed to have a ceiling. Cuprates broke it,
and high-pressure hydrides reached 250 K (−23 °C) ── with ordinary BCS, no less.
And in 2023 the field went through a major retraction.

Tools you'll need: \(T_c\sim\theta_De^{-1/\lambda}\) from Episode 5, the non-perturbative from Episode 6, bonus ① The heart of this episode: read what is on the exponent, then go and build it

The last episode of the series. Bonus ① judged one of the two things we want to happen at room temperature ── cold fusion ── and concluded that there is no convincing evidence that it happens. This episode looks at the other one. Room-temperature superconductivity. Its history has an entirely different shape ── discovered at 4 K in 1911, given a theoretical ceiling estimate of about 40 K in 1968, broken by cuprates in 1986, and carried by high-pressure hydrides to 203 K in 2015 and 250 K (−23 °C) in 2019. And the hydrides are explained by the ordinary BCS we learned in Episode 5. The theory was never wrong ── it just needed a stiffer lattice. Then between 2020 and 2023 the field experienced a major series of retractions. Set the two "things we want at room temperature" side by side and you have a specimen of how science self-corrects.

01There was supposed to be a ceiling

Episode 5's formula, once more:

$$T_c\approx1.13\,\theta_D\,e^{-1/\lambda}$$

There are only two ways to raise \(T_c\) ── raise \(\theta_D\) (lattice stiffness) or raise \(\lambda\) (coupling strength). Both seemed to be walled in.

The McMillan limit (1968)

Making \(\lambda\) large requires softening the lattice, and softening it too far destroys the crystal (lattice instability). From that tug-of-war McMillan estimated that conventional superconductors could not much exceed 30–40 K.
In fact, from niobium–germanium in 1973 (23 K) until 1986, the record barely moved for thirteen years. The limit looked real.

021986 ── cuprates break the wall

YearMaterialT_cSignificance
1911mercury (Onnes)4.2 Kdiscovery
1973Nb₃Ge23 Kthe conventional record. Stuck for thirteen years
1986La-Ba-Cu-O (Bednorz and Müller)30 Kbreaks the limit. Nobel Prize the following year (the fastest ever)
1987YBCO93 Kabove liquid nitrogen (77 K) ── the threshold of practicality
1993Hg-Ba-Ca-Cu-O133 K (138 K under pressure)long held as the ambient-pressure record

Cuprates are not phonon-mediated (their isotope effect is small or anomalous). That the pairing has d-wave symmetry is established, but after nearly forty years the mechanism is still unresolved ── spin-fluctuation theories are prominent, but no consensus has been reached. One of the great open problems in condensed matter physics.

032015 ── hydrides, and BCS was right

Here is the interesting turn. Cuprates went outside BCS, but the other road was still open ── raise \(\theta_D\) as far as it will go.

Ashcroft's proposal

The lattice frequency goes as \(\omega_D\propto1/\sqrt{M}\). So using the lightest atom there is ── hydrogen raises \(\theta_D\) by orders of magnitude.

Ashcroft predicted in 1968 that metallic hydrogen should be a high-temperature superconductor, and in 2004 proposed that "compressing pure hydrogen is hard, but combining it with other elements provides chemical pre-compression" ── i.e. use hydrides. He read what was riding on the exponent and went after it directly.

YearMaterialT_cPressure
2015H₃S (Drozdov, Eremets and colleagues)203 K155 GPa
2019LaH₁₀about 250 K (−23 °C)about 170 GPa

The decisive point was that the isotope effect was confirmed in H₃S ── replace hydrogen with deuterium and \(T_c\) falls. As Episode 5 showed, that is evidence for a phonon mechanism. Which means ──

Conclusion: BCS was never wrong

Superconductivity in high-pressure hydrides is explained directly by the ordinary BCS of Episode 5 (strictly, its strong-coupling version, Eliashberg theory). No new mechanism is needed.
The McMillan limit was broken not because the theory was wrong but because nobody had found a way to "strengthen the coupling while keeping the lattice stiff." Hydrogen is light, so \(\omega\) is high, and in hydrides the electron coupling is strong too ── you can raise both at once.

Put another way: read what is on the exponent, go and build the quantity that rides there, and succeed. As a closing example for this series there could not be a better one.

04Try it ── the record's history, and what the exponent demands

The upper panel of the figure is the history of \(T_c\), with lines for liquid nitrogen (77 K) and room temperature (293 K). Retracted claims can be shown separately (toggle with the button).

The lower panel works backwards. Using the strong-coupling version of Episode 5's formula (Allen–Dynes), it shows what is required for a \(T_c\) of 250 K. Presets take you from aluminium to LaH₁₀ ── and you see that what you need is an \(\omega_{\log}\) of order 1000 K, i.e. hydrogen.

Figure: above = the history of superconducting transition temperatures (blue = conventional, red = cuprates, purple = high-pressure hydrides, grey = retracted claims). Below = T_c from the Allen–Dynes formula. Move ω_log and λ to see what is required to reach 250 K
conventional (BCS) cuprates high-pressure hydrides retracted

052020–2023 ── a chain of retractions

YearEventOutcome
2020288 K at 267 GPa in carbonaceous sulfur hydride (CSH) ── room temperature at last, published in NatureConcerns raised about the background treatment of the magnetic susceptibility data. Retracted in September 2022 over the authors' objections
March 2023294 K in nitrogen-doped lutetium hydride, and at only 1 GPa. Enormous attentionMultiple groups failed to replicate it. Retracted in November 2023. A subsequent university investigation found research misconduct
July 2023A Korean group claimed ambient-pressure, ambient-temperature superconductivity in "LK-99." A worldwide social-media frenzySettled in about three weeks. Replication attempts worldwide showed the resistance drop came from a structural transition in a copper sulfide (Cu₂S) impurity, and the levitation was diamagnetism rather than the Meissner effect
Why did LK-99 take three weeks and CSH two years? This contrast matters. What sets the speed of self-correction is not correctness but how easy the replication is.
LK-99 was ambient pressure, cheap, with a published recipe, so laboratories worldwide could make it in days. Hence three weeks.
CSH and LuNH, by contrast, required diamond anvils at 170 GPa, samples a few micrometres across, and are within reach of perhaps a dozen laboratories worldwide. Replication took years, and the claims survived in the meantime.
"The speed of replication is inversely proportional to the difficulty of the experiment" ── and that is also, in fact, the structure behind cold fusion in bonus ① going thirty years unresolved (calorimetry is cheap, but measuring a few percent of excess heat correctly is hard, so attempts pile up without converging).

06Side by side with bonus ①

Two things we want to happen at room temperature
cold fusion (bonus ①)room-temperature superconductivity (this episode)
theoretical obstacleyes. 50 orders from the Gamow factor, and the branching ratio is decisivenone. Explicable by BCS/Eliashberg
progress over thirty yearsessentially none4 K → 250 K (under pressure)
independent replicationnot establishedH₃S and LaH₁₀ confirmed by multiple groups
retraction episodesthe E-Cat litigationthe CSH and LuNH retractions, LK-99
current statusno evidenceit is real. But it needs extreme pressure

The decisive difference is the first row. There is no theoretical reason forbidding room-temperature superconductivity. That is why it is worth looking for. Cold fusion faces the branching-ratio wall, and no amount of cleverness about barriers gets past it.
"Not yet found" and "forbidden by theory" are entirely different things. A retraction episode does not turn the former into the latter.

07What remains is pressure, not temperature

What is needed now is 170 GPa ── 1.7 million atmospheres. It can only be made inside a diamond anvil cell, with samples a few micrometres across. You cannot build a power line out of it.

So the current goal is not "a higher \(T_c\)" but "the same \(T_c\) at a much lower pressure." The lead is chemical pre-compression ── an all-out materials-design effort, screening ternary hydrides exhaustively with first-principles computation.

Whether an ambient-pressure, room-temperature superconductor exists is unknown to anyone. No theorem forbids it, and nothing guarantees it either. That is precisely what "still worth looking for" means.

◇ ◇ ◇
The honest line

Established: the history of \(T_c\) (mercury 4.2 K 1911, Nb₃Ge 23 K 1973, La-Ba-Cu-O 30 K 1986, YBCO 93 K 1987, Hg-based 133 K 1993); McMillan's estimate for conventional \(T_c\) (1968); that cuprates are d-wave with an unresolved mechanism; Ashcroft's proposals on metallic hydrogen and hydrides (1968, 2004); H₃S at 203 K (155 GPa, 2015) and the confirmation of a phonon mechanism via the isotope effect; LaH₁₀ at about 250 K (about 170 GPa, 2019) with confirmation by multiple groups; the 2022 retraction of the CSH paper and the 2023 retraction of the nitrogen-doped lutetium hydride paper together with the subsequent finding of research misconduct; and that the LK-99 claim was shown by replication to arise from an impurity. All published facts.

Points to note: (1) The \(T_c\) in the figure is computed from the Allen–Dynes formula (with \(\mu^*=0.1\) fixed) and differs from real first-principles calculations. The material presets also vary between sources. (2) Measuring \(T_c\) under high pressure is hard ── confirming the Meissner effect as well as zero resistance is difficult, and the measurements themselves are sometimes debated (LaH₁₀ has been confirmed by several groups, but discussion of systematic errors in high-pressure experiments generally continues). (3) The mechanism of cuprate superconductivity is unresolved, and the body asserts nothing beyond "not phonons." (4) On the retractions, the body records only published facts and includes no judgement about anyone's motives. (5) "Whether an ambient-pressure room-temperature superconductor is possible is unknown" is meant literally ── at present there is neither a theorem forbidding it nor an argument guaranteeing it. (6) The comparison table with bonus ① sets out the state of the evidence and the presence or absence of a theoretical obstacle; it is not a comparison of researchers' integrity.

Exercises
  1. Was the McMillan limit broken because BCS theory was wrong?
    See the answer
    For cuprates, they are outside BCS; for high-pressure hydrides, BCS still applies. The latter show the isotope effect and are explained by Eliashberg theory. What broke was not the theory but the assumption that "there is no way to strengthen the coupling while keeping the lattice stiff." Hydrogen satisfies both at once.
  2. Why hydrogen? Where in Episode 5's formula should you look?
    See the answer
    At \(\theta_D\) in \(T_c\approx1.13\theta_De^{-1/\lambda}\). Since \(\omega_D\propto1/\sqrt M\), the lightest atom raises the lattice frequency by orders of magnitude. And in hydrides \(\lambda\) is large too. You gain on both the prefactor and the exponent at once.
  3. Why was LK-99 settled in three weeks and CSH in two years?
    See the answer
    Because the ease of replication differed by orders of magnitude. LK-99 was ambient pressure, cheap and with a published recipe, so the world could make it at once. CSH and LuNH need diamond anvils at 170 GPa, within reach of perhaps a dozen laboratories worldwide. What sets the speed of self-correction is not correctness but the difficulty of the experiment.
  4. What is the most important difference between cold fusion and room-temperature superconductivity?
    See the answer
    The presence or absence of a theoretical obstacle. Cold fusion faces 50 orders from the Gamow factor and, more decisively, the branching-ratio wall (tunneling changes the entrance but not the exit). There is no theoretical reason forbidding room-temperature superconductivity. "Not yet found" and "forbidden by theory" are different things.

Bonus ⑥ summary / Tunneling That Clicks, completeRead what is on the exponent, then go and build it

There are two roads to raising \(T_c\approx1.13\theta_De^{-1/\lambda}\) ── stiffen the lattice or strengthen the coupling. They seemed incompatible, and McMillan put the limit at 40 K. Cuprates broke it in 1986, and their mechanism is still unresolved after forty years.

But another road remained ── use hydrogen, the lightest atom, exactly as Ashcroft proposed. H₃S reached 203 K in 2015 and LaH₁₀ about 250 K in 2019. The isotope effect was confirmed and ordinary BCS (in its strong-coupling form) explained it. The theory was never wrong ── they read what was riding on the exponent, went after it, and got there.

Then the chain of retractions from 2020 to 2023. CSH and LuNH were retracted, and LK-99 was shown within three weeks to come from an impurity. What set the speed of self-correction was not correctness but the ease of replication. Set beside bonus ①, the decisive difference is the first row ── nothing in theory forbids room-temperature superconductivity; cold fusion faces the branching-ratio wall. "Not yet found" and "forbidden by theory" are entirely different things.

This document is Bonus ⑥ of the "Tunneling That Clicks" series, a reading piece for physics-loving high-schoolers and undergraduates. The history of superconducting transition temperatures, McMillan's estimate for conventional \(T_c\) (1968), the discovery of cuprates (Bednorz–Müller 1986, Nobel Prize 1987) and the YBCO and Hg-based records, that cuprates are d-wave with an unresolved mechanism, Ashcroft's proposals on metallic hydrogen and hydrides (1968, 2004), H₃S at 203 K (155 GPa, 2015) with the isotope effect confirming a phonon mechanism, LaH₁₀ at about 250 K (2019), the CSH paper (retracted 2022) and the nitrogen-doped lutetium hydride paper (retracted 2023, with a subsequent finding of research misconduct), and that the LK-99 claim was shown by replication to arise from an impurity are all published facts. That the figure's \(T_c\) comes from the Allen–Dynes formula (\(\mu^*=0.1\) fixed) with material presets varying between sources, that discussion of systematic errors in high-pressure measurements continues, that the cuprate mechanism is unresolved, that the body records only published facts about the retractions without judging motives, that whether an ambient-pressure room-temperature superconductor is possible is currently unknown, and that the comparison with bonus ① concerns the state of evidence and theoretical obstacles rather than researchers' integrity ── all spelled out in the body's "honest line." ── To print, use your browser's "Print" and "Save as PDF" (in the print version the sliders and answers are frozen and hidden). Adjacent: Bonus ⑤, Do living things use tunneling? / Bonus ⑦, When the thing that tunnels extends through spacetime / Contents / sister series Temperature That Clicks, Renormalization That Clicks, Black Holes That Click, Mass That Clicks, The Physics Cube.

Print / make a PDF: ⌘+P (Ctrl+P on Windows). On screen, move ω_log and λ to see what is required to reach 250 K. The presets travel from Al to LaH₁₀, and the button toggles the display of retracted claims. "See the answer" opens each solution.