Prediction of ambient pressure conventional superconductivity above 80 K in hydride compounds

The study addresses whether hydrogen-rich hydrides can exhibit conventional superconductivity at high temperatures under ambient pressure. Superconductivity, discovered in 1911, typically required low temperatures and often high pressures. Conventional superconductors pair via electron-phonon interactions, with Tc correlated to phonon frequencies and coupling strength, which can be enhanced using light elements like hydrogen. While atomic metallic hydrogen requires unattainably high pressures, chemical precompression in hydrides has enabled record high Tc values, but only under extreme pressures (e.g., H₃S above 100 GPa; LaH₁₀ ~250 K at ~170 GPa; CaH₉ ~215 K at ~172 GPa). The primary challenge is to realize superconductivity at low or ambient pressure, as many high-pressure phases become dynamically unstable upon decompression. Prior ambient-pressure candidates such as AlH₃ and (Be₂H₆)ₚ showed promising Tc but were thermodynamically unstable. The authors aim to discover thermodynamically and dynamically stable hydrides that superconduct at ambient pressure and liquid-nitrogen-range temperatures.

Hydrides have delivered record Tc under high pressure, notably H₃S (>100 GPa), LaH₁₀ (~250 K at ~170 GPa), and CaH₉ (~215 K at ~172 GPa), enabled by chemical precompression. However, ambient-pressure superconductivity remains rare. Prior proposals include AlH₃ (sp³ AuP-type) with predicted Tc up to 54 K and layered (Be₂H₆)ₚ with Tc ~72–84 K, but both lie significantly above the convex hull and decompose into other phases, hindering synthesis. The work relates to experimentally known Mg–Ru hydrides (e.g., Mg₂RuH₈) as structural/chemical templates, motivating exploration of chemically similar Mg–X–H systems at ambient pressure using high-throughput and machine-learning accelerated searches.

Screening and selection: A machine-learning-accelerated high-throughput search was conducted over >1 million compounds in the Alexandria database, including experimental and hypothetical structures derived from prototype substitutions. Candidates close to the convex hull of thermodynamic stability were selected for further study. First-principles calculations: Geometry optimizations and total energies were obtained with VASP using PAW potentials, a 520 eV plane-wave cutoff, and PBE for solids. Calculations started from ferromagnetic configurations but relaxed to spin-unpolarized ground states. Brillouin zones were sampled with Γ-centered k-point grids of density 2000–8000 k-points per reciprocal area depending on composition. Lattice dynamics and electron-phonon coupling: Harmonic phonon frequencies and electron-phonon matrix elements were computed within DFPT using Quantum ESPRESSO on 16×16×16 electronic k-grids and 10×10×10 phonon q-grids. Methfessel-Paxton smearing of 0.02 Ry was used. Norm-conserving pseudopotentials from the PSEUDO-DOJO project were employed with a 120 Ry cutoff. Anharmonicity: Quantum ionic and anharmonic effects were evaluated via the stochastic self-consistent harmonic approximation (SSCHA) using 2×2×2 supercells (~73 atoms), a 96 Ry cutoff, and 4×4×4 k-grids. SSCHA dynamical matrices on a 2×2×2 q-mesh were used to interpolate phonons onto the 10×10×10 grid for superconductivity calculations. Frequencies and polarization vectors were obtained from the Hessian of the SSCHA free energy. Additional checks at 300 K indicated negligible temperature effects on phonon spectra and dynamical stability. Superconductivity: Three approaches were used—(i) isotropic superconducting density-functional theory (SCDFT) including the energy dependence of the electronic DOS with an integration window δω ≈ 200 meV; (ii) fully anisotropic SCDFT; and (iii) isotropic Eliashberg theory. The spread among methods provides theoretical uncertainty, with isotropic Eliashberg typically yielding Tc values 10–20% larger than isotropic SCDFT and anisotropic SCDFT sometimes increasing Tc up to ~20% over isotropic SCDFT. Doping dependence of Tc was evaluated within a rigid-band approximation by rigidly shifting the Fermi level in isotropic SCDFT.

- Discovery of a family of ambient-pressure hydrides Mg₂XH₈ (X = Rh, Ir, Pd, Pt), related to experimentally realized Mg₂RuH₈, predicted to be conventional superconductors. - Thermodynamic screening indicates these phases are on or near the convex hull, suggesting experimental accessibility; stability appears to improve with heavier X (noting uncertainties from exchange-correlation and incomplete hull knowledge). - Electronic structure: All compounds are metallic with pronounced van Hove singularities near the Fermi level; H contributes roughly 20% of the Fermi-level DOS. Rh and Ir have EF near a DOS peak; Pd and Pt near a shoulder above a DOS minimum. This leads to strong sensitivity of superconducting properties to doping. - Phonons: All four compounds are dynamically stable in the harmonic approximation. Acoustic modes span ~0–20 meV; Mg-dominated modes ~20–30 meV; H-dominated modes extend to high energies (Rh/Ir show separated H-mode manifolds around ~30–100 meV and ~160–230 meV; Pd/Pt show a more compact H spectrum ~40–190 meV). - Anharmonicity: SSCHA (tested on Ir) shows weak anharmonic effects, slightly softening select high-energy H modes, with negligible impact on λ and Tc; a similar mild effect is expected for the other compounds. - Electron-phonon coupling: At least ~2/3 of the integrated coupling constant λ arises from H modes. Characteristic phonon frequencies are ~50 meV (Ir) to ~60 meV (Rh), high for ambient-pressure hydrides, enabling elevated Tc. - Superconducting properties: Across isotropic/anisotropic SCDFT and isotropic Eliashberg methods, predicted Tc ranges are approximately 45–50 K (Rh), 51–67 K (Pd), 66–77 K (Ir), and 64–80 K (Pt). The Pt compound is predicted to exceed 100 K under suitable electron doping (rigid-band approximation). - Practical implications: The presence of sharp DOS features near EF and strong electron-phonon coupling implies sensitivity to doping and sample quality. Thin-film, high-throughput synthesis and screening are proposed as effective experimental routes.

The work targets the central challenge of realizing high-temperature superconductivity in hydrides at ambient pressure. By combining machine-learning-guided screening with first-principles calculations of lattice dynamics and superconductivity, the authors identify Mg₂XH₈ (X = Rh, Ir, Pd, Pt) as promising, thermodynamically accessible and dynamically stable candidates. The mechanism is conventional electron-phonon pairing with significant H-character at the Fermi level coupling to high-energy H vibrations, yielding elevated characteristic frequencies and strong λ. The electronic DOS landscape with van Hove features near EF explains the predicted high Tc values and the strong sensitivity to doping, which can enhance Tc further (notably for Pt). The minimal impact of anharmonicity on phonons and λ supports the robustness of the Tc predictions. For experimental realization, the authors emphasize high-throughput thin-film synthesis and screening, and note magnesium’s essential role; isoelectronic substitutions (e.g., B, C) either destabilize the lattice or significantly reduce Tc. Overall, the results suggest that, although rare, ambient-pressure hydride superconductors meeting thermodynamic and dynamical stability criteria are achievable, opening pathways for discovery and optimization via doping and compositional tuning.

This study identifies a family of ambient-pressure, conventional superconductors Mg₂XH₈ (X = Rh, Ir, Pd, Pt) with predicted Tc values up to ~80 K, and potentially exceeding 100 K under electron doping (Pt). The compounds are near the convex hull and dynamically stable, with superconductivity driven by strong coupling between electronic states with notable H character and high-energy H vibrations. Anharmonic effects are weak, lending confidence to Tc estimates. The work provides concrete targets and practical guidance for experimental synthesis, particularly through thin-film high-throughput methods, and highlights doping as a lever to optimize Tc. Future research directions include experimental synthesis and characterization, systematic doping studies, exploration of strain and alloying effects, refinement of Coulomb screening treatments, and expansion of ML-accelerated searches to related chemistries.

- Thermodynamic stability assessments rely on an incomplete convex hull and approximate exchange-correlation functionals; true stability may differ. - Superconducting Tc predictions vary across methods (isotropic/anisotropic SCDFT and Eliashberg), reflecting theoretical uncertainty; systems with sharp DOS features near EF are particularly sensitive to computational details. - The rigid-band approximation used to assess doping effects neglects potential structural/electronic changes upon doping. - Coulomb interaction treatments significantly affect Tc in materials with large DOS at EF; estimating an effective μ is nontrivial and introduces uncertainty. - Some previously proposed ambient-pressure hydrides were found thermodynamically unstable; experimental synthesis of the proposed compounds may still face kinetic and processing challenges. - Reported phase diagrams and stability trends may shift with improved databases or more exhaustive structure searches.