Iron-based superconductors: families and evidence

Scope

Iron-based superconductors contain Fe–pnictogen or Fe–chalcogen layers in several structural families. Multiple Fe orbitals contribute to their low-energy electronic structure, and parent magnetic or structural transitions must be distinguished from superconductivity in a substituted sample. Kamihara and colleagues' fluorine-doped LaOFeAs discovery established a high-temperature 1111 platform. Its original abstract reports a transition near 26 K; the exact operational criterion remains unassigned in this review. R1

A 122 arsenide, a binary 11 chalcogenide and a single-layer interface are different experimental systems. Rotter and colleagues studied K-substituted BaFe2As2, while Hsu and colleagues reported superconductivity in PbO-type FeSe. Neither result establishes that every related composition has the same gap, transition temperature or magnetic ground state. Starting mixture, refined phase composition and nominal formula should remain separately identifiable. R2 R3

Mathematical assumptions

The coefficient $y$ in $\mathrm{BaFe}_{2-y}\mathrm{Co}_y\mathrm{As}_2$ and the site fraction $x$ in $\mathrm{Ba}(\mathrm{Fe}_{1-x}\mathrm{Co}_x)_2\mathrm{As}_2$ obey $y=2x$. Eleven exact atom-count checks verify this normalization. Likewise, nominal $\mathrm{FeSe}_{1-x}$ becomes $\mathrm{Fe}_{1+\delta}\mathrm{Se}$ with $\delta=x/(1-x)$ when normalized to one Se. Five checks preserve the elemental ratio. These conversions do not measure crystal occupancy, mobile-carrier density or the location of a defect. R7

Bulk fractions also have a precise but limited meaning. For superconducting and magnetic fractions $f_S$ and $f_M$ of the same volume, their overlap lies between $\max(0,f_S+f_M-1)$ and $\min(f_S,f_M)$. Two explicit arrangements with 95% and 90% fractions give 85% or 90% overlap. This set-theory bound neither establishes atomic-scale coexistence nor removes probe calibration and sample-matching requirements. A microscopic interpretation must use additional spatial, temporal and model evidence. R8

A multiband gap $\Delta_i(\mathbf{k})$ is defined on specified bands. Mazin and colleagues' LaFeAsO1−xFx model connects hole and electron pockets by repulsive singlet pair transfer; its leading eigenvector has opposite signs. This is a material-specific proposal, not a consequence of merely containing iron. Orbital weights and accessible momentum transfers enter the response, as the separate Maier–Scalapino two-orbital calculation makes explicit.

Opposite signs do not require equal gap magnitudes. In a local near-transition illustration with positive constant one-spin densities of states $N_h,N_e$, a shared cutoff, reciprocal interband interaction $V>0$ and zero intraband terms, use $\lambda_{ij}=-V_{ij}N_j$. The leading mode has $\lambda_+=V\sqrt{N_hN_e}$ and $\Delta_h/\Delta_e=-\sqrt{N_e/N_h}$. In synthetic normalized units, $N_h=1$, $N_e=4$ and $V=0.1$ give $\lambda_+=0.2$ and a gap vector proportional to $(2,-1)$. Six exact rational checks and six attractive-interband controls verify the eigenvectors and density-of-states transformation. This is a linearized model example, not a fitted iron-based gap ratio or a zero-temperature prediction. Band edges inside the pairing window, energy-dependent densities of states and unequal cutoffs require a more general kernel. The multiband and s± pages give the broader assumptions and probe limits. R6 R11

Experimental criteria

The revised Rotter author manuscript gives a resistive midpoint of 38.1 K for Ba0.6K0.4Fe2As2, a first deviation near 39 K and a reported zero-resistance point of 37.2 K. Figure 4 also labels 37.2 K as its 10% threshold, so the term zero retains an unresolved instrumental-threshold caveat. Its 90–10% width is 1.5 K. Quoting only the rounded 38 K headline loses distinctions that matter in a searchable record. R2

The same revision measures magnetic cycles at 0.5 and 1 mT, replacing the earlier manuscript's 10 mT example. Diamagnetism begins near 38.3 K; shielding and field-cooled responses differ and depend on density and demagnetization assumptions. The powder contains a minor FeAs impurity. Selected revised-manuscript figures have been checked, but the final published methods and raw data have not been completely compared here. R2

Hsu's final FeSe report identifies zero resistance near 8 K in a sample labelled by nominal FeSe0.88 composition. Magnetic susceptibility at 30 G and a broad heat-capacity anomaly provide complementary evidence. Its field-dependent transition uses 50% of normal resistance, with applied fields through 9 T; the roughly 16.3 T upper-critical-field value is a fit extrapolation. Nominal selenium deficiency and a fitted critical field must not be promoted to directly measured phase occupancy or field limits. R3

McQueen and colleagues subsequently found nearly stoichiometric superconducting FeSe and strong preparation sensitivity. Their selected neutron-refinement analysis gives Fe1.01±0.02Se, with limited ability to distinguish tiny deviations. Nominal composition, oxygen contamination and annealing temperature vary across the comparison. A sample labelled Fe1.03Se showed no superconductivity down to 0.6 K; that lower measurement bound is not a superconducting transition. The proposed defect-site explanation remains a hypothesis. R7

Pressure and magnetism

Medvedev's author manuscript associates a resistive onset of 36.7 K with 8.9 GPa in Fe1.01Se. Broadened transitions and curve offsets require careful interpretation; at 15 GPa the authors report residual resistance associated with a second phase. Their resistance method uses a diamond-anvil cell, platinum contacts, a van der Pauw arrangement and ruby pressure measurements at low and room temperature. The silicone-oil medium specified for the Mössbauer experiment must not automatically be assigned to the separate resistance experiment. R4

The earlier Mössbauer nondetection of static order is not a universal FeSe phase diagram. Bendele and colleagues later reported pressure-induced static magnetism using zero-field muon-spin rotation and AC susceptibility in nominal FeSe0.94 and FeSe0.98 samples. Their selected revised manuscript specifies Daphne 7373, pressure-cell background treatment and low-temperature pressure calibration. Its Figure 3 supports a pressure-dependent coexistence interpretation near and above 0.8 GPa, with model-dependent spatial conclusions. Different probes, samples and pressure environments must remain explicit. R4 R8

Evidence for gap structure

Christianson's neutron experiment concerns K-doped Ba0.6K0.4Fe2As2. Its superconductivity-related resonance and the relevant coherence-factor model constrain candidate signs, without directly determining every iron-based gap. Background, momentum, orbital matrix elements and independently measured gap scales remain necessary parts of that interpretation. A resonance measurement cannot be substituted for a family-wide phase verdict. R5 R6

At a FeSe/SrTiO3 interface, Lee's replica bands and mean-field gap-closing scale retain their spectroscopic labels. Faeth's later photon-energy comparison supports an intrinsic electron–phonon interpretation within its stated analysis, but does not determine all contributions to pairing enhancement. A gap-closing temperature is not interchangeable with zero resistance or a magnetic transition. These interface results remain linked to the detailed multiband discussion rather than silently entering a bulk Tc comparison. R9 R10

Limits and disputes

Several orbitals, correlations, nematicity and magnetic proximity help organize the evidence without selecting one universal pairing mechanism. A mass renormalization is not automatically a larger pairing eigenvalue. Complete scientific acceptance remains open, including final-version comparisons, full supplementary methods and raw-data analysis. Differences between a source's claim and this article's review scope are preserved in the references. The historical mechanism narrative also needs its published qualification: Mazin and colleagues' note 8 says that the magnetic instability they initially described as ferromagnetic was misidentified. Keep that qualification with the original abstract; it is not a universal statement that doping suppresses ferromagnetism in every iron-based superconductor. R5 R6 R7 R8

Discovery implications

Test a specific model with controlled composition, structure, pressure or disorder changes, and identify the confounding changes. Preserve actual measurement criteria and sample provenance when comparing results. For a proposed coexistence regime, report what each probe resolves and how sample fractions were calibrated. For a candidate pairing mediator, explain which alternative survives the same observation. R2 R4 R8

Related topics

Continue with s± pairing, multiband pairing, magnetic substitution, nematic order, and two-dimensional interfaces.

Material cases

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