Unconventional Pairing Mechanisms and Competing Evidence

A pairing mechanism is a causal claim about the irreducible interaction kernel that overcomes repulsion in a symmetry channel and sets the observed scales. Gap symmetry narrows that kernel but does not identify it. A credible mechanism comparison must jointly reproduce momentum and frequency structure, scale dependence, controlled perturbations, and competing orders while surviving plausible phonon, spin, charge, excitonic, orbital, and mixed-interaction alternatives.

Required background. Pairing symmetry diagnostics fixes what the order parameter does. Eliashberg theory supplies a retarded-kernel example.

Helpful background. Evidence triangulation supplies multi-probe inference and covariance discipline.

From interaction to eigenchannel

Near $T_c$, a general linearized equation is

$$\lambda\Delta_a(k)=-T\sum_{k',b} \Gamma^{\mathrm{irr}}_{ab}(k,k') G_b(k')G_b(-k')\Delta_b(k').$$

The leading eigenfunction predicts the pairing structure within the adopted normal-state propagator and vertex. To claim a mechanism, the calculation must also show why the proposed contribution to $\Gamma^{\mathrm{irr}}$ dominates in the relevant window and why approximations to self-energy and vertex are controlled.

An attractive phonon kernel is retarded and carries isotope and lattice-mode information, but Coulomb, anharmonic, nonadiabatic, and multiband effects can alter simple isotope expectations. Spin-fluctuation exchange can favor a sign-changing gap because repulsion at a characteristic wavevector becomes attractive after the gap changes sign; it must reproduce the measured susceptibility and survive feedback and vertex tests. Charge, orbital, excitonic, and composite channels have analogous obligations. “The calculated symmetry is right” is only one row of the comparison.

Scalapino 2012, §§2–5 explains how momentum-dependent repulsion can generate sign-changing pairing; Monthoux, Pines, and Lonzarich 2007, pp. 667–670 reviews magnetic pairing across material families while also displaying the model-dependence of the comparison.

Evidence that changes likelihoods

Useful tests perturb or observe different parts of the forward model:

Test	What it constrains	Important alternatives or nuisance terms
Isotope substitution	Ionic-frequency and structural dependence of $T_c$, gap, or mode spectrum	Volume change, disorder, anharmonicity, competing order, multiband redistribution
Momentum-resolved gap sign	Eigenfunction of the pairing kernel	Surface selection, matrix elements, domains, reconstruction
Spin or charge spectrum	Candidate retarded glue and its wavevector structure	Feedback from superconductivity, common cause, broad continua
Controlled impurities	Pair breaking and sign structure	Local moments, carrier-density change, inhomogeneity
Pressure or strain	Correlated change of electronic, magnetic, and lattice scales	Multiple parameters move simultaneously; first-order transitions
Strong-coupling structures	Frequency dependence in tunnelling or optics	Nonunique inversion, backgrounds, self-energy channels unrelated to pairing

No row is a universal yes/no test. The analysis must propagate calibration and covariance and compare full predicted responses, including null regions, rather than selected peak locations.

Claim levels and stopping rule

As of 10 August 2026, mechanism assignments remain material-specific and mutable. This page therefore gives no timeless list of “proved” unconventional glues. Use the following claim ladder:

1. Compatibility: a kernel can reproduce a subset of data with stated parameters.
2. Preference: under a frozen likelihood and alternatives, the kernel predicts held-out observables better.
3. Dominant contribution: multiple perturbations and probes quantitatively isolate its spectral and momentum weight.
4. Causal identification: interventions and independent methods exclude competitive or cooperative explanations over the claimed regime.

Stop at the highest rung supported by the data. A mixture of phonon and electronic channels is not a failure of classification; forcing a single named cause may be the failure.

For new material-specific measurements, corrections, and explicit supersession notes after that cutoff, use the dated Quantum Matter and Emergence Research synthesis. This method page fixes the durable comparison standard; it does not freeze a changing experimental record.

The chapter map routes the claim through independent symmetry, retardation, and competing-hypothesis gates.

Mechanism inference is stronger than finding a gap or a compatible kernel. Momentum, frequency, controlled perturbations, and alternatives must be tested with a common observable convention and uncertainty model. Original schematic, not to scale.

The canonical paired-matter claim test matrix makes the stopping rule explicit.

Exercise

Design a discriminating test. Both a phonon kernel and a spin-fluctuation kernel fit the same $T_c$ and gap magnitude. Name two additional measurements and state what must be modelled.

Solution

One useful pair is isotope substitution plus a phase-sensitive gap-sign measurement. The isotope analysis must model structural volume shifts, disorder, anharmonic phonons, and correlated carrier changes; a nonzero exponent is not automatically causal. The phase-sensitive experiment must model junction orientation, tunnelling matrix elements, domains, and surface order. A sign-changing gap disfavors a simple momentum-independent attraction but does not exclude a structured electron–phonon interaction. Momentum-resolved spin and phonon spectra under pressure would add an independent frequency-and-wavevector test.

References

• Monthoux, P., Pines, D., and Lonzarich, G. G. (2007). “Superconductivity without phonons.” Nature 450, 1177–1183. doi:10.1038/nature06480.
• Scalapino, D. J. (2012). “A common thread: The pairing interaction for unconventional superconductors.” Reviews of Modern Physics 84, 1383–1417. doi:10.1103/RevModPhys.84.1383.