Heavy-fermion superconductors — intermetallic compounds of rare-earth or actinide elements whose charge carriers acquire effective masses of order hundreds of the free-electron mass — have been suspected, since the 1979 discovery of the first of them, of pairing through electronic magnetism rather than lattice vibrations. The tunneling experiments of Huth, Jourdan, and Adrian on UPd2Al3 (1999) formulated the evidential bar for settling the question: spectroscopic signatures that appear at the energy scale of the magnetic excitations, are tied to the superconducting state (disappearing when the gap closes), and cannot be ascribed to the known phonon spectrum. Later work supplied the strongest Eliashberg-type case for UPd2Al3 (tunneling combined with neutron-scattering detection of a magnetic exciton coupling to the superconducting heavy electrons), a neutron spin resonance in CeCoIn5, quasiparticle-interference-based evidence for magnetic f-electron-mediated pairing, and theoretical alternatives — octupolar magnetic fluctuations, valence fluctuations, and correlation-driven Kondo-Mott pairing without a well-defined exchanged boson — while nodeless superconductivity in the prototype compound CeCu2Si2 complicated the magnetic-glue picture. The class-wide question of what binds heavy quasiparticles into Cooper pairs, and whether phonon-based explanations can be excluded, remains open.

Background

Heavy-fermion compounds are intermetallics containing rare-earth or actinide elements in which nearly localized f-shell electrons (the 4f electrons of Ce or Yb, the 5f electrons of U or Pu) hybridize with ordinary conduction electrons through the Kondo effect — the screening of a localized magnetic moment by the conduction sea. In a lattice of such moments (a Kondo lattice), this hybridization produces quasiparticles with effective masses of order 100–1000 free-electron masses, hence 'heavy fermions'. The first heavy-fermion superconductor, CeCu2Si2, was discovered in 1979 (F. Steglich et al., Phys. Rev. Lett. 43, 1892), and dozens of Ce-, Yb-, U-, and Pu-based compounds have since been found to superconduct at temperatures from below 0.1 K to about 18 K, often at the border of, or coexisting with, antiferromagnetic order.

In a conventional superconductor the attractive interaction that binds Cooper pairs is mediated by phonons, quantized lattice vibrations, and this glue can be observed directly: in Eliashberg theory the electron-boson spectral functionα2F(ω)\alpha^2F(\omega)appears in the tunneling conductance of a superconducting junction (the McMillan-Rowell analysis), which is how the phonon spectrum of lead was spectroscopically confirmed as its pairing glue. Heavy-fermion superconductors, by contrast, are 'unconventional': their order parameters generally have non-s-wave symmetry (for example the d-wave gap of CeCoIn5, which changes sign under 90-degree rotation), and their pairing appears in close proximity to magnetism. Mathur et al. (Nature 394, 39 (1998)) found that in CePd2Si2 and CeIn3 superconductivity forms a dome around the pressure-tuned antiferromagnetic quantum critical point — the zero-temperature endpoint where antiferromagnetic order just vanishes — and argued that the pairs are bound by magnetic spin-spin interactions rather than by phonons. Such phase-diagram evidence, however, is circumstantial: it shows where superconductivity appears, not what boson (if any) provides the attraction.

The compound UPd2Al3 became the decisive test case. It orders antiferromagnetically near 14 K and superconducts below about 2 K, and Jourdan, Huth, and Adrian (Nature 398, 47 (1999)) succeeded in fabricating controlled tunneling junctions, UPd2Al3/AlOx/Pb, on epitaxial films — a prerequisite absent for essentially all other heavy-fermion superconductors, whose pristine surfaces and sub-kelvin gaps had precluded controlled directional tunneling spectroscopy. They observed a strong-coupling feature in the tunneling conductivity at roughly a millielectronvolt, an energy too low for any known phonon of this compound and matching the energies of its antiferromagnetic excitations, and disappearing with the closing of the superconducting gap. Sato et al. (Nature 410, 340 (2001)) then detected by inelastic neutron scattering, which measures the dynamic spin structure factorS(Q,ω)S(\mathbf{Q},\omega)as a function of momentumQ\mathbf{Q}and energyω\omega, a dispersive 'magnetic exciton' — a collective excitation of the ordered f-electron moments that propagates through the lattice via exchange forces — that couples strongly to the superconducting heavy electrons; combining this with the tunneling data, they argued that the exciton plays the role of the pairing glue, in analogy to phonons in conventional superconductors.

The review chapter (M. Huth and M. Jourdan, Advances in Solid State Physics 39, 351 (1999)) poses the pairing question at the class level: its introduction asks what the residual interactions of the quasiparticles are that induce the transition into the superconducting state, noting that conventional phonon exchange is 'extrinsic' to the electron system whereas an intrinsic pairing mechanism — one 'based on the polarization properties of the background (Fermi sea of Landau quasiparticles) against charge or spin perturbations' — seems plausible for heavy-fermion superconductors, and its abstract records that 'evidence for spin-fluctuation coupling has not yet been experimentally observed' for the class. The evidential bar applied in the source's own affirmative claim for UPd2Al3 — in which the analysis of the tunneling differential conductivity giving 'direct evidence for Cooper-pair coupling via the exchange of antiferromagnetic spin fluctuations' was combined with inelastic neutron scattering data on the magnetic excitations — is: to establish antiferromagnetic spin fluctuations rather than phonons as the pairing interaction, one needs spectroscopic signatures that (i) appear at the characteristic energy of the magnetic excitations, (ii) are tied to the superconducting state, disappearing when the gap closes, and (iii) cannot be ascribed to the known phonon spectrum. Later work broadened the candidate list: a sharp neutron spin resonance — a peak in the magnetic fluctuation spectrum that appears only below the superconducting transition — in CeCoIn5 (Stock et al., Phys. Rev. Lett. 100, 087001 (2008)) and a quasiparticle-interference analysis, which infers the gap's momentum-space structure from the interference pattern that superconducting electrons form around surface scatterers in scanning-tunneling-microscope images, combined with solved superconducting gap equations (Van Dyke et al., Proc. Natl. Acad. Sci. 111, 11663 (2014)); octupolar (higher-rank, spin-quadrupole/octupole rather than dipolar) magnetic fluctuations proposed as the glue of CeCu2Si2 (Ikeda et al., Phys. Rev. Lett. 114, 147003 (2015)); valence-crossover fluctuations — fluctuations of the f-electron shell's charge state — proposed for high-pressure CeCu2Si2 (Seyfarth et al., Phys. Rev. B 85, 205105 (2012)); and Kondo-lattice/Mott-physics descriptions in which pairing emerges from strong correlations without a well-defined propagating boson at all (Steglich and Wirth, Rep. Prog. Phys. 79, 084502 (2016)). Nodeless superconductivity reported in the prototype CeCu2Si2 further challenged the expectation that a magnetically mediated gap must have sign-changing nodes. The unresolved question remains the source's, posed at the class level: what interaction binds heavy quasiparticles into Cooper pairs, and can phonon explanations be excluded at this evidential bar?

Problem Statement

Identify the interaction that mediates Cooper pairing in heavy-fermion superconductors — in particular, establish whether it is an intrinsic electronic collective mode such as antiferromagnetic spin fluctuations rather than phonons — at the natural scope of the class, without restriction to a specific compound, observable, or method.

An answer must, for each heavy-fermion superconductor or family it claims to cover: (i) exhibit pairing-related spectroscopic signatures whose energies (and, where measurable, momenta) coincide with those of the proposed electronic collective mode; (ii) show that these signatures are tied to the superconducting state, disappearing when the superconducting gap closes at the transition temperature or critical field; and (iii) demonstrate quantitatively that the signatures cannot be ascribed to the compound's known phonon spectrum. Where the interaction identified is an intrinsic electronic mechanism other than dipolar antiferromagnetic spin fluctuations — for example a charge/valence-fluctuation channel, of the kind admitted by the source's own framing of the intrinsic alternative as a polarization of the quasiparticle background 'against charge or spin perturbations' — the same evidential bar applies at equivalent stringency; where the identified interaction has no well-defined exchanged boson, requirement (i) is met by pairing-related spectroscopic signatures tied to the superconducting state at the characteristic energy scale of the correlation-driven mechanism, together with an explicit argument that no collective-mode signature is expected, while requirements (ii) and (iii) apply unchanged. The single overarching objective is a defensible identification of the pairing interaction across the heavy-fermion class; the per-compound spectroscopic identifications and phonon-exclusion analyses jointly serve this one objective and are not independent deliverables, and a result meeting the evidential bar for a single compound or a single family alone — a family being a group of heavy-fermion superconductors sharing the same f-electron element and crystal-structure type — is partial progress toward the class-level identification rather than its closure; closure requires meeting the evidential bar across the families that make up the class.

References

  1. M. Huth, M. Jourdan, 'Heavy-fermion superconductivity induced by antiferromagnetic spin fluctuations', Advances in Solid State Physics 39, 351–360 (1999). DOI: 10.1007/bfb0107493, https://doi.org/10.1007/bfb0107493
  2. M. Jourdan, M. Huth, H. Adrian, 'Superconductivity mediated by spin fluctuations in the heavy-fermion compound UPd2Al3', Nature 398, 47–49 (1999). DOI: 10.1038/17977, https://doi.org/10.1038/17977
  3. N. D. Mathur, F. M. Grosche, S. R. Julian, I. R. Walker, D. M. Freye, R. K. W. Haselwimmer, G. G. Lonzarich, 'Magnetically mediated superconductivity in heavy fermion compounds', Nature 394, 39–43 (1998). DOI: 10.1038/27838, https://doi.org/10.1038/27838
  4. N. K. Sato, N. Aso, K. Miyake, R. Shiina, P. Thalmeier, G. Varelogiannis, C. Geibel, F. Steglich, P. Fulde, T. Komatsubara, 'Strong coupling between local moments and superconducting "heavy" electrons in UPd2Al3', Nature 410, 340–343 (2001). DOI: 10.1038/35066519, https://doi.org/10.1038/35066519
  5. F. Steglich, J. Aarts, C. D. Bredl, W. Lieke, D. Meschede, W. Franz, H. Schäfer, 'Superconductivity in the presence of strong Pauli paramagnetism: CeCu2Si2', Physical Review Letters 43, 1892–1896 (1979). DOI: 10.1103/PhysRevLett.43.1892, https://doi.org/10.1103/PhysRevLett.43.1892
  6. C. Stock, C. Broholm, J. Hudis, H. J. Kang, C. Petrovic, 'Spin resonance in the d-wave superconductor CeCoIn5', Physical Review Letters 100, 087001 (2008). DOI: 10.1103/PhysRevLett.100.087001, https://doi.org/10.1103/PhysRevLett.100.087001
  7. D. Parker, P. Thalmeier, 'Theory of tunneling spectroscopy in UPd2Al3', Physical Review B 75, 184502 (2007). DOI: 10.1103/PhysRevB.75.184502, https://doi.org/10.1103/PhysRevB.75.184502 (preprint arXiv:cond-mat/0612402, https://arxiv.org/abs/cond-mat/0612402)
  8. J. S. Van Dyke, F. Massee, M. P. Allan, J. C. Séamus Davis, C. Petrovic, D. K. Morr, 'Direct evidence for a magnetic f-electron–mediated pairing mechanism of heavy-fermion superconductivity in CeCoIn5', Proceedings of the National Academy of Sciences 111, 11663–11667 (2014). DOI: 10.1073/pnas.1409444111, https://doi.org/10.1073/pnas.1409444111 (preprint arXiv:1405.5883, https://arxiv.org/abs/1405.5883)
  9. H. Ikeda, M.-T. Suzuki, R. Arita, 'Emergent loop-nodal s±-wave superconductivity in CeCu2Si2: similarities to the iron-based superconductors', Physical Review Letters 114, 147003 (2015). DOI: 10.1103/PhysRevLett.114.147003, https://doi.org/10.1103/PhysRevLett.114.147003 (preprint arXiv:1504.07312, https://arxiv.org/abs/1504.07312)
  10. G. Seyfarth, A.-S. Rüetschi, K. Sengupta, A. Georges, D. Jaccard, S. Watanabe, K. Miyake, 'Heavy fermion superconductor CeCu2Si2 under high pressure: multiprobing the valence crossover', Physical Review B 85, 205105 (2012). DOI: 10.1103/PhysRevB.85.205105, https://doi.org/10.1103/PhysRevB.85.205105 (preprint arXiv:1111.4873, https://arxiv.org/abs/1111.4873)
  11. F. Steglich, S. Wirth, 'Foundations of heavy-fermion superconductivity: lattice Kondo effect and Mott physics', Reports on Progress in Physics 79, 084502 (2016). DOI: 10.1088/0034-4885/79/8/084502, https://doi.org/10.1088/0034-4885/79/8/084502
  12. Y. Song, W. Wang, C. Cao, Z. Yamani, Y. Xu, Y. Sheng, W. Löser, Y. Qiu, Y.-f. Yang, R. J. Birgeneau, P. Dai, 'High-energy magnetic excitations from heavy quasiparticles in CeCu2Si2', npj Quantum Materials 6, 60 (2021). DOI: 10.1038/s41535-021-00358-x, https://doi.org/10.1038/s41535-021-00358-x
  13. V. Zambra, A. Nathwani, M. Nauman, S. K. Lewin, C. E. Frank, N. P. Butch, A. Shekhter, B. J. Ramshaw, K. A. Modic, 'Giant transverse magnetic fluctuations at the edge of re-entrant superconductivity in UTe2', Nature Communications 17, 3742 (2026). DOI: 10.1038/s41467-026-71899-7, https://doi.org/10.1038/s41467-026-71899-7
  14. C. Pfleiderer, 'Superconducting phases of f-electron compounds', Reviews of Modern Physics 81, 1551 (2009). DOI: 10.1103/RevModPhys.81.1551, https://doi.org/10.1103/RevModPhys.81.1551
  15. B. D. White, J. D. Thompson, M. B. Maple, 'Unconventional superconductivity in heavy-fermion compounds', Physica C 514, 246–278 (2015). DOI: 10.1016/j.physc.2015.02.044, https://doi.org/10.1016/j.physc.2015.02.044
  16. O. Stockert, F. Steglich, 'Spin resonances in heavy-fermion superconductors', Physica C 615, 1354375 (2023). DOI: 10.1016/j.physc.2023.1354375, https://doi.org/10.1016/j.physc.2023.1354375