Superconducting nodes: dimensionality, disorder and experimental tests

Scope

A superconducting node is a zero of the quasiparticle excitation energy. In a simple weak-pairing single-band description it lies where the normal-state Fermi surface intersects a zero of the gap. A point, a line, a small but finite minimum and an extended Bogoliubov Fermi surface are different possibilities. In multiband systems, inspect the full Bogoliubov spectrum rather than only one scalar gap. Nodal dimension, symmetry protection and the pairing representation are separate questions. R1 R8

A low-temperature power law constrains low-energy excitations within a measured window; it is not a unique symmetry label. The dispersion near a node, probe weighting, disorder, nonlocal response, multiple gaps and backgrounds can change the observed dependence. A fit should state its lowest temperature and energy resolution, and compare with alternatives before assigning a representation. R2 R3 R6

Mathematical assumptions

In a clean local spectrum with $m$ independent linearly dispersing directions transverse to a smooth nodal manifold, $E^2=\sum_{i=1}^{m}v_i^2q_i^2$ gives momentum volume below $E$ proportional to $E^m$, hence $N(E)\propto E^{m-1}$. All transverse velocities must be finite and nonzero; the nodal measure and degeneracy must not introduce another low-energy singularity. A three-dimensional line node has two transverse directions; an isolated linear point has three. These powers count states, not probe matrix elements. A point with $E^2=q_z^2+(q_x^2+q_y^2)^2$ instead gives $N(E)\propto E$. To illustrate a finite-window ambiguity, define $r^2=q_x^2+q_y^2$ and the dimensionless imposed spectrum $E^2=q_z^2+\delta^2r^2+r^4$. Its only zero remains a point for every $\delta\ge0$. For $\delta>0$, direct phase-volume differentiation gives $N(E)=2\pi E\arctan(2E/\delta^2)$, before momentum-normalization and degeneracy factors. Thus $N\propto E^2$ for $E\ll\delta^2$, while $N\propto E$ for $E\gg\delta^2$ within this model. At $\delta=0$, $N(E)=\pi^2E$. The limits of vanishing $\delta$ and low energy give different apparent exponents. Nine anisotropic linear cases and fifteen crossover cases check these relations. The added term is assumed allowed, not inferred from a material symmetry. Durst and Lee provide the two-dimensional linear nodal context; the three-dimensional crossover is a local illustration, not their result or a material fit. R3

With a temperature-independent low-energy density of states $N(E)=AE^p$ and $k_B=1$, the positive-energy quasiparticle contribution satisfies $C(T)=\int_0^\infty N(E)E^2 e^{E/T}/[T^2(1+e^{E/T})^2]dE\propto T^{p+1}$. Direct energy integrals and numerical derivatives of internal energy agree for $p=0,1,2$. Thus ideal linear line nodes give $C\propto T^2$, whereas ideal linear point nodes in three dimensions give $C\propto T^3$. These are electronic contributions in the stated regime, not the total measured heat capacity. Two hard-gap toy densities also show increasing suppression as temperature falls below the gap; they do not model a BCS coherence peak or an actual sample. A nonzero residual density of states changes this asymptotic inference: Durst and Lee obtain finite $N(0)$ in their scalar impurity-self-energy approximation. It does not justify assigning every observed linear heat-capacity term to that mechanism. Apply clean nodal powers only above relevant broadening scales and within the nodal energy cutoff. R3 R7

On an imposed circular Fermi surface, $\Delta_s(\theta)=s_0+s_4\cos4\theta$ can have zeros while retaining fully symmetric character. For real amplitudes with $s_4\ne0$, $|s_0/s_4|<1$ gives eight zeros, equality gives four merged quadratic zeros, and a larger ratio gives none. In contrast, $\Delta_d=\cos2\theta+\eta\cos6\theta$ retains diagonal zeros for the tested small real $\eta$ values in the B1g sector. Seven numerical root checks verify these examples. Changing a model harmonic ratio is not a calculation of how irradiation changes a real gap; a disorder self-energy and gap equation are needed for that connection. R1 R9

Agterberg, Brydon and Timm's even-parity multiband model supplies another caution. With its inversion and particle-hole symmetries, the Hamiltonian can be transformed to an antisymmetric form with real Pfaffian P and $\det H=P^2$. Opposite signs of P enclose protected zero-energy surfaces. For the specified pure-quintet example, $P=(\epsilon_+\epsilon_-)^2+4\Delta_0^2(\epsilon_+\epsilon_-+c_{xz}^2+c_{yz}^2)$. Five independent eight-dimensional Hamiltonian determinants match this expression, including a change of sign between two nearby momenta. This tests the named model identity. It does not show that every multiband or time-reversal-breaking superconductor has such a surface, nor establish a material's energetic stability. R8

Experimental criteria

Hardy and colleagues measured the penetration-depth change in YBa2Cu3O6.95 using a resonator with a 1.3 K lower operating temperature. The reported strong linear term extended approximately from 3 to 25 K; those are different intervals. The publisher abstract supports a classic nodal constraint, with defects relevant to comparisons, but the full experimental analysis has not been reviewed here. R2

Hashimoto and colleagues' P-substituted BaFe2(As0.67P0.33)2 study combined penetration depth and thermal conductivity. The author version describes measurements down to about 0.15 K with a tunnel-diode oscillator and about 1.6 K with a microwave resonator. Its zero-field thermal fit resolves a residual contribution of approximately $25\,\mathrm{mW\,K^{-2}\,m^{-1}}$. The low-temperature penetration-depth upturn was discussed as a small paramagnetic contribution, not silently discarded. These combined observations support line nodes; the measurements do not alone locate every node or uniquely select d-wave over nodal s-wave. R5

The inherited Reid citation concerns KFe2As2, not P-substituted BaFe2As2. Reid and colleagues compared heat flow along two crystal directions using different samples and measurements down to 50 mK. A 0.05 T field made solder contacts normal; the residual zero-field term was obtained by extrapolation, with thermal fits below 0.3 K. Cross-sample scattering comparisons also used resistivity normalization and assumptions relating resistivity to scattering. The authors favored d-wave from several constraints. That interpretation must retain these conditions and its material identity rather than become an automatic classification of another compound. R4

Mizukami and colleagues irradiated P-substituted BaFe2As2 with 2.5 MeV electrons and followed penetration depth across multiple samples and doses. Selected data show a nodal-like response becoming nearly temperature independent, then developing low-energy excitations again. One irradiated sample was measured to 80 mK; the reported flat response had roughly 1 Å precision. The proposed two-band interpretation removes accidental nodes before an impurity band reaches zero energy: the final gapless response is not simply restored symmetry-enforced nodes. Retain the separate sample batches, partial defect annealing on warming, and assumed interband/intraband scattering ratio. R6

Rotating-field heat measurements require a response calculation. Vorontsov and Vekhter's quasi-two-dimensional model includes vortex scattering and finds that the same nodal direction can give a maximum or a minimum depending on temperature and field. The selected calculation neglects Zeeman splitting and uses an approximate vortex-state treatment; its illustrated reversal scale is not universal. Reading every fourfold minimum as a node would discard precisely the physics that changes the sign. R7

Limits and disputes

“Universal” residual heat transport refers to a specified low-temperature, low-frequency scattering regime. Durst and Lee distinguish thermal transport from electrical and spin responses, whose corrections differ; an impurity-independent formula cannot be transferred indiscriminately across all probes or multiband materials. Likewise, a symmetry-enforced zero is robust only while the protecting symmetry, relevant Fermi-surface crossing and phase remain applicable. Strong perturbations may change the phase or electronic structure. R3 R9

Topology requires an appropriate invariant and protecting symmetries, with any surface prediction tied to surface orientation and the projected bulk nodes. Schnyder and Ryu's time-reversal-symmetric noncentrosymmetric model links nodal-line topology to surface flat bands, distinct from a fully gapped bulk classification. A tunneling feature, a normal-state Weyl band or the word “topological” in a gap-evolution title is not sufficient to establish superconducting Weyl nodes. The cited abstract is checked here; its full surface calculation remains open. R10

Discovery implications

Store the lowest measured temperature, fit interval, residual background, field orientation, contact protocol, dose history and sample identity with each observation. Separate the measured response, inferred low-energy spectrum and proposed representation. Our numerical examples test fixed spectra, roots and a named Pfaffian identity, not the complete experiments. Full scientific acceptance, complete source-version comparisons, raw-data reconstruction and disorder/vortex self-consistency remain open. R4 R5 R6 R7

Related topics

Continue with s-wave pairing, d-wave pairing, singlet and triplet classification and topological superconductivity.

Material cases

Values describe individual reports, not a ranking. Unknown means the checked source did not supply the field at this review scope.