⚠️ Verification: lanthanum decahydride — Paper vs Simulation [2026-06-30]

The Paper's Central Claim

Lanthanum decahydride — LaH₁₀ — has become something of a celebrity in the world of superconductivity research. According to the claim we're examining, sourced from a Wikipedia summary of the current state of play as of 2023, this hydrogen-rich compound holds the record for the highest accepted superconducting critical temperature (T_c): approximately 250 K (−23 °C), achieved under crushing pressures of around 150 GPa.

To put that in perspective: 250 K is just twenty-three degrees below the freezing point of water. For a field that spent decades celebrating superconductors that worked at temperatures colder than deep space, this is extraordinary. And 150 GPa? That's roughly 1.5 million times atmospheric pressure — the kind of force generated between two diamond tips in a device called a diamond anvil cell, which can fit in the palm of your hand but recreates conditions found near Earth's core.

The claim traces back to landmark experimental work by Drozdov, Eremets, and collaborators, who reported near-room-temperature superconductivity in lanthanum superhydrides in 2019. Their results, published in Nature, sent shockwaves through condensed matter physics. The underlying logic is elegant: hydrogen, the lightest element, produces extremely high-frequency vibrations in a crystal lattice. If you can cage enough hydrogen atoms in a stable structure under pressure, those vibrations can mediate exceptionally strong coupling between electrons — the engine of conventional superconductivity. Lanthanum acts as the scaffold. Pressure does the rest.

How Our Simulation Approaches This

At AI Future Lab, we use a machine-learning-augmented computational pipeline to model superconducting behavior in candidate materials. We want to be upfront about what this is and isn't.

Our approach draws on trained surrogate models that approximate key quantities from density functional theory (DFT) and Migdal-Eliashberg calculations — the gold-standard frameworks for predicting superconducting temperatures in conventional (phonon-mediated) superconductors. We don't run full ab initio DFT from scratch for each prediction. Instead, our models learn from large datasets of prior DFT results, experimental measurements, and structural databases to interpolate and, cautiously, extrapolate predictions for materials under specified conditions.

This means our results are faster to generate but carry additional layers of uncertainty compared to dedicated first-principles studies. We trade some precision for breadth and speed. Where a full Eliashberg calculation for LaH₁₀ might require weeks of supercomputer time with careful convergence testing, our pipeline returns estimates in hours — but those estimates should be read as informed predictions, not definitive calculations.

We also model structural stability, phonon spectra characteristics, and electron-phonon coupling strength (λ), giving us a more complete picture than T_c alone. Think of our system as a well-read research assistant that synthesizes existing knowledge to make educated projections — not an oracle.

What Our Analysis Found

Here's what our simulation returned for LaH₁₀:

• Predicted T_c: 188 K (−85 °C)
• Pressure required for optimal stability: 170 GPa
• Electron-phonon coupling constant (λ): 2.1
• Structural stability: Metastable (the Fm3̄m clathrate structure is a local energy minimum, not the global ground state at this pressure)
• Dominant mechanism: Strong electron-phonon coupling driven by high-frequency hydrogen-derived phonon modes — H-stretching and H-bending vibrations — in the clathrate-like cage structure, consistent with BCS-Migdal-Eliashberg theory near the strong-coupling limit
• Confidence level: Medium

The mechanistic picture aligns well with the published literature. LaH₁₀ forms a remarkable structure where each lanthanum atom sits inside a cage of 32 hydrogen atoms, and it's those hydrogen vibrations — light atoms oscillating at very high frequencies — that generate the enormous phonon energies driving superconductivity. Our λ of 2.1 places the material firmly in the strong-coupling regime, which is consistent with values reported in dedicated Eliashberg studies (typically ranging from 2.0 to 3.5 depending on the pressure, functional, and Coulomb pseudopotential μ* assumed).

But our T_c is 62 K lower than the claimed 250 K. That's a significant gap. Let's talk about why.

⚠️ Partial Match: Reading the Gap

A 62 K discrepancy — roughly 25% below the reported value — demands honest analysis rather than hand-waving. Several factors likely contribute, and they cut in both directions.

1. Sensitivity to μ* and functional choices. The predicted T_c in Eliashberg theory is notoriously sensitive to the Coulomb pseudopotential μ*, a parameter that encapsulates how much electron-electron repulsion counteracts the phonon-mediated attraction. Small changes in μ* (say, from 0.10 to 0.13) can shift T_c by tens of kelvin at these coupling strengths. Our surrogate model may have absorbed a slightly different effective μ* from its training data than what produces the best agreement with experiment.

2. Pressure discrepancy. Our model identifies 170 GPa as the optimal stability pressure, while the experimental claim is at 150 GPa. This 20 GPa difference matters. The T_c of LaH₁₀ is pressure-dependent, and the relationship is non-monotonic — there's a sweet spot. If our model slightly mislocates the pressure-stability landscape, the predicted T_c shifts accordingly.

3. Anharmonic effects. Hydrogen is so light that its vibrations are strongly anharmonic — the simple "balls on springs" approximation breaks down. Full anharmonic phonon calculations, which some groups have performed for LaH₁₀, tend to increase the predicted T_c compared to harmonic estimates. Our surrogate model may underweight these corrections.

4. Experimental uncertainties. It's worth noting that the experimental measurement of T_c at 150 GPa in a diamond anvil cell is itself fraught with challenges. Pressure gradients across the sample, the definition of the superconducting transition (onset vs. midpoint vs. zero resistance), and the minuscule sample size all introduce uncertainty. The "250 K" figure is an approximation, and different experimental runs and groups have reported a range.

5. Metastability. Our finding that the structure is metastable rather than thermodynamically ground-state is consistent with the literature — LaH₁₀ in the Fm3̄m phase does require specific synthesis pathways and may coexist with other stoichiometries. A metastable phase might exhibit slightly different effective coupling than an idealized calculation assumes.

In short: the gap is real, explicable, and reflects the genuine difficulty of predicting T_c to within tens of kelvin for strong-coupling superconductors under extreme conditions. We rate this a partial match — the physics is right, the ballpark is right, but the number isn't close enough to call it a clean confirmation.

What This Tells Us About Room-Temperature Superconductivity

LaH₁₀ is both a triumph and a sobering reminder. It proves that conventional phonon-mediated superconductivity can reach temperatures we would have called impossible twenty years ago. The BCS-Eliashberg framework — developed in the 1950s and 60s — still works, even at 250 K, as long as you're willing to squeeze your sample to 1.5 million atmospheres.

And that's the catch. Every few years, a claim emerges for room-temperature superconductivity at or near ambient pressure — most recently the LK-99 episode of 2023, which captivated the internet before being debunked. The allure is obvious: a material that superconducts at room temperature and normal pressure would revolutionize energy transmission, computing, transportation, and medical imaging.

But the physics is unforgiving. High T_c in conventional superconductors requires light atoms (hydrogen), stiff lattices (high phonon frequencies), and strong electron-phonon coupling — conditions that are naturally promoted by extreme pressure. Remove the pressure, and most superhydrides simply fall apart. The hydrogen cages collapse. The magic vanishes.

For ambient-pressure room-temperature superconductivity to work via conventional mechanisms, you would need a material that somehow maintains hydrogen-like phonon frequencies and strong coupling without external compression — perhaps through chemical precompression, where heavier atoms in the lattice create an internal pressure on hydrogen sublattices. This is an active area of research, but no confirmed ambient-pressure candidate has emerged.

Unconventional mechanisms — those beyond electron-phonon coupling — remain possible in principle, but they are far less well understood and far harder to predict computationally. Our models, grounded in Eliashberg physics, would need fundamental extensions to address them.

Our Evolving Simulation

We treat every partial match as a calibration opportunity, not a failure. The 62 K gap between our prediction and the reported T_c for LaH₁₀ is now a data point we're feeding back into our pipeline. Specifically, we're working on three refinements:

Better anharmonic corrections. We're incorporating results from recent anharmonic phonon studies on superhydrides to retrain our surrogate models, which should improve predictions for hydrogen-rich materials where harmonic approximations systematically underperform.

Pressure-dependent recalibration. We're building a more granular pressure-T_c mapping for the La-H system using published data across the 130–200 GPa range, allowing our model to better locate the T_c maximum in pressure space.

Uncertainty quantification. Rather than reporting a single T_c value, future posts will include confidence intervals — because a prediction of "188 ± 40 K" tells a very different story than "188 K, full stop." Honesty about error bars is not a weakness; it's how science actually works.

The gap today may narrow tomorrow. Or it may persist and teach us something deeper about the limits of surrogate modeling for extreme-condition physics. Either outcome is valuable. We'll keep running the numbers, keep comparing, and keep telling you what we actually find — not what we wish we'd found.

— AI Future Lab | Computational Verification Series | June 2025