Andreev Crystals: Periodic Superconducting States

• Andreev crystals are periodic arrangements formed by coherent electron–hole conversion, manifesting in structures like vortex lattices, magnetic chains, and Josephson arrays.
• They exhibit distinct realizations such as triangular Abrikosov lattices with CdGM states and directional Andreev spectroscopy in single crystals to resolve multiband superconductivity.
• These systems offer practical insights into superconducting anisotropy, subgap band formation, and nonreciprocal transport, with potential applications in advanced quantum devices.

Andreev crystals are periodic or symmetry-structured manifestations of Andreev physics—coherent electron–hole conversion at superconducting interfaces—in which bound states, conductance spectra, or collective modes are organized by a lattice, a vortex array, a magnetic superstructure, or a phase-biased junction network. In the cited literature, the expression is used in several related senses rather than as a single universally fixed term: a triangular Abrikosov vortex lattice whose cores host Caroli–de Gennes–Matricon (CdGM) states in hole-doped BaFe$_2$As$_2$ (Shan et al., 2010); single-crystal Andreev spectroscopies that use crystallographic orientation to resolve anisotropic or multiband superconductivity (Aslam et al., 2016, Tortello et al., 2010); quasi-one-dimensional superconductors with periodic magnetic regions whose local Andreev states hybridize into Bloch subgap bands (Rouco et al., 2020, Rouco et al., 2021); and hybrid Josephson arrays in which short-junction Andreev states form directional bands under uniform phase bias (Dahl et al., 15 Aug 2025).

1. Conceptual scope and common mechanisms

The common element in all usages is Andreev reflection: at a normal-metal/superconductor interface, a subgap electron is converted into a hole while a Cooper pair is transferred into the condensate. When this process is confined by geometry, magnetic texture, vortex structure, or crystal orientation, it produces Andreev bound states (ABS), surface resonances, or subgap bands. In vortex cores, the relevant levels are the CdGM states with energies $E_\mu \approx \mu \omega_0$, $\mu=\pm 1/2,\pm 3/2,\ldots$, and $\omega_0\sim \Delta^2/E_F$ (Shan et al., 2010). In multiple-Andreev-reflection (MAR) junctions, the diagnostic structure is the subharmonic series $V_n=2\Delta/(en)$ (Kuzmichev et al., 2014, Abdel-Hafiez et al., 2014). In magnetic or hybrid arrays, periodic hybridization converts localized ABS into subgap bands whose topology and transport depend on phase, symmetry, or magnetization texture (Rouco et al., 2020, Rouco et al., 2021, Dahl et al., 15 Aug 2025).

Usage of the term	Defining structure	Representative papers
Vortex-lattice sense	Periodic Abrikosov lattice with CdGM states in each core	(Shan et al., 2010)
Single-crystal spectroscopic sense	Directional PCAR or MAR on anisotropic or multigap superconductors	(Aslam et al., 2016, Kuzmichev et al., 2014)
Magnetic-chain sense	Periodic magnetic regions in a 1D superconducting wire	(Rouco et al., 2020, Rouco et al., 2021)
Hybrid-array sense	Periodic Josephson fingers forming phase-biased ABS bands	(Dahl et al., 15 Aug 2025)
Related topological/interfacial sense	Crystal-symmetry-organized ABS or magnetic crystallization waves	(Yoshida et al., 2023, Todoshchenko et al., 21 Feb 2025)

This plurality of meanings is significant. In some papers, “Andreev crystal” denotes a literal periodic array of bound-state-hosting objects. In others, it denotes a methodological program: using single crystals, directional point contacts, field rotation, or break-junction stacks to make the superconducting order parameter visible through Andreev spectroscopy.

2. Vortex-lattice realization in Ba$_{0.6}$K$_{0.4}$Fe$_2$As$_2$

The most literal experimental realization of an Andreev crystal in the vortex-lattice sense was reported in hole-doped Ba$_2$0K$_2$1Fe$_2$2As$_2$3 by low-temperature STM/STS at $_2$4 K and magnetic fields $_2$5 T and $_2$6 T (Shan et al., 2010). There the defining hallmark was the simultaneous observation of a triangular Abrikosov lattice and Andreev bound states in every measured vortex core. The magnetic flux per vortex was $_2$7 Wb at $_2$8 T and $_2$9 Wb at $E_\mu \approx \mu \omega_0$0 T, consistent with $E_\mu \approx \mu \omega_0$1 Wb. The corresponding lattice spacings, from $E_\mu \approx \mu \omega_0$2, were $E_\mu \approx \mu \omega_0$3 nm at $E_\mu \approx \mu \omega_0$4 T and $E_\mu \approx \mu \omega_0$5 nm at $E_\mu \approx \mu \omega_0$6 T, in agreement with the STM maps (Shan et al., 2010).

Core spectroscopy revealed a pronounced negative-bias conductance peak centered at $E_\mu \approx \mu \omega_0$7 mV at the vortex center. Along a $E_\mu \approx \mu \omega_0$8 Å line through a vortex, this single peak evolved into two sub-peaks with dominant spectral weight at negative bias and then faded continuously into the zero-field superconducting spectrum (Shan et al., 2010). In the same study, bright domains exhibited two gaps, $E_\mu \approx \mu \omega_0$9 meV and $\mu=\pm 1/2,\pm 3/2,\ldots$0 meV, whereas dark domains showed predominantly the smaller gap and higher zero-bias conductance. The observed $\mu=\pm 1/2,\pm 3/2,\ldots$1 meV core feature was interpreted as compatible with the $\mu=\pm 1/2,\pm 3/2,\ldots$2 CdGM level, implying $\mu=\pm 1/2,\pm 3/2,\ldots$3 meV and, through $\mu=\pm 1/2,\pm 3/2,\ldots$4, effective $\mu=\pm 1/2,\pm 3/2,\ldots$5 estimates of $\mu=\pm 1/2,\pm 3/2,\ldots$6 meV for $\mu=\pm 1/2,\pm 3/2,\ldots$7 meV and $\mu=\pm 1/2,\pm 3/2,\ldots$8 meV for $\mu=\pm 1/2,\pm 3/2,\ldots$9 meV (Shan et al., 2010).

The same work emphasized the contrast with electron-doped Ba(Fe,Co)$\omega_0\sim \Delta^2/E_F$0As$\omega_0\sim \Delta^2/E_F$1, where vortices were highly disordered and Andreev bound states were not observed in the cores (Shan et al., 2010). The ordered lattice in the hole-doped material was attributed to weaker disorder-induced pinning from off-plane K dopants, in contrast to in-plane Co dopants in the FeAs layers. Within this usage, an Andreev crystal is therefore a periodic Abrikosov vortex array whose individual cores remain clean enough for CdGM quantization to be spectroscopically resolved.

3. Directional Andreev spectroscopy on single crystals

A second major usage treats “Andreev crystals” as single-crystalline platforms in which PCAR, MAR, or related Andreev spectroscopies resolve anisotropy, multigap structure, and pairing symmetry by exploiting crystal orientation. In this sense the crystal is not itself an array of bound states; rather, crystallographic direction enters the Andreev process as a selection rule on $\omega_0\sim \Delta^2/E_F$2, $\omega_0\sim \Delta^2/E_F$3, interface transparency, and orbital weight.

Field-angle-dependent PCAR on La(O,F)BiSeS provided a clear example. With current injected predominantly along the $\omega_0\sim \Delta^2/E_F$4 axis, the superconducting gap was $\omega_0\sim \Delta^2/E_F$5 meV at $\omega_0\sim \Delta^2/E_F$6 K, the gap decreased faster for $\omega_0\sim \Delta^2/E_F$7 than for $\omega_0\sim \Delta^2/E_F$8, and at $\omega_0\sim \Delta^2/E_F$9 kG the extracted values were $V_n=2\Delta/(en)$0 meV and $V_n=2\Delta/(en)$1 meV, giving an anisotropy factor $V_n=2\Delta/(en)$2. The spectra and angular magnetoresistance exhibited two-fold symmetry, interpreted as evidence for anisotropic and unconventional pairing (Aslam et al., 2016).

In iron pnictides, directional PCAR and MAR established a broad multigap program. Ba(Fe$V_n=2\Delta/(en)$3Co$V_n=2\Delta/(en)$4)$V_n=2\Delta/(en)$5As$V_n=2\Delta/(en)$6 showed two nodeless gaps, with low-temperature averages $V_n=2\Delta/(en)$7–$V_n=2\Delta/(en)$8 meV and $V_n=2\Delta/(en)$9–$_{0.6}$0 meV, and additional conductance structures from which a characteristic boson energy $_{0.6}$1 meV was extracted; these spectra were modeled within a three-band $_{0.6}$2 Eliashberg framework (Tortello et al., 2010). In BaFe$_{0.6}$3Ni$_{0.6}$4As$_{0.6}$5, c-axis PCAR tracked a doping evolution from underdoped plateau-like spectra to overdoped sharp in-gap peaks, with optimal-doping fits yielding $_{0.6}$6 or $_{0.6}$7 meV, $_{0.6}$8 or $_{0.6}$9 meV, and an electron-gap anisotropy parameter $_{0.4}$0 (Ren et al., 2011). A three-dimensional BTK analysis further distinguished Ca(Fe$_{0.4}$1Co$_{0.4}$2)$_{0.4}$3As$_{0.4}$4, where the small gap required nodes or zeros, from ab-plane Ba(Fe$_{0.4}$5Co$_{0.4}$6)$_{0.4}$7As$_{0.4}$8, where two nodeless gaps sufficed; by contrast, c-axis Ba-122 films could not be fitted by a two-nodeless-gap 3D model and were discussed in terms of hot spots or 3D minima (Gonnelli et al., 2013).

Break-junction MAR on LiFeAs resolved three gaps directly from subharmonic gap structures: $_{0.4}$9–$_2$0 meV, $_2$1–$_2$2 meV, and $_2$3–$_2$4 meV, with estimated anisotropies $_2$5, $_2$6, and $_2$7, respectively (Kuzmichev et al., 2014). Combined lower-critical-field and IMARE work on 122 arsenides supported a nodeless multigap state with $_2$8–$_2$9 meV, $_2$0-space anisotropy of approximately $_2$1–$_2$2, and $_2$3 meV (Abdel-Hafiez et al., 2014). Sub-kelvin PCAR on FeSe yielded an anisotropic two-gap fit with $_2$4 meV, $_2$5 meV, weight $_2$6, barrier $_2$7, and anisotropy parameter $_2$8; the study argued that $_2$9 K operation was necessary to resolve the multiband structure unambiguously (Bashlakov et al., 2019).

Non-centrosymmetric and topological-candidate systems sharpened the distinction between conventional Andreev spectroscopy and genuinely unconventional ABS phenomenology. Directional PCAR on BiPd found $_2$00 meV together with orientation-selective secondary gaps $_2$01 meV and $_2$02 meV, plus a pronounced zero-bias conductance peak attributed to ABS from a mixed-parity order parameter (Mondal et al., 2012). By contrast, PCAR on SnAs gave $_2$03 meV and $_2$04, consistent with weak-coupling BCS superconductivity (Howlader et al., 2021). In Nb$_2$05Bi$_2$06Se$_2$07, low- and high-resistance Andreev spectra displayed zero-bias conductance enhancement and conductance dips at the superconducting gap that were described as inconsistent with conventional BTK and discussed as consistent with $_2$08-wave pairing in a topological-superconductor candidate (Kurter et al., 2017).

4. One-dimensional magnetic Andreev crystals

A third usage is strictly theoretical and highly specific: a quasi-one-dimensional $_2$09-wave superconducting wire with a periodic arrangement of mesoscopic magnetic regions. In this construction, each magnetic region is a semiclassical impurity that does not generate normal backscattering but instead imparts a spin- and particle–hole-dependent phase $_2$10. The local ABS associated with individual regions hybridize with an overlap controlled by $_2$11, where $_2$12 is the period and $_2$13 the coherence length, thereby forming Bloch subgap bands (Rouco et al., 2020, Rouco et al., 2021).

For collinear antiferromagnetic arrangements, the tight-binding description yields

$_2$14

with $_2$15 and $_2$16 (Rouco et al., 2020). The excitation gap therefore closes at half-integer $_2$17, where $_2$18, producing Dirac phase boundaries. The same work related the spectral asymmetry index of the spin-resolved Hamiltonian to the spin polarization and used it as the observable that quantifies gap closing and reopening. Heterojunctions between neighboring antiferromagnetic phases support spin-polarized interface bound states and exhibit fractionalization of the interface spin, in close analogy to Jackiw–Rebbi mass inversion (Rouco et al., 2020).

The subsequent extension to “spectral properties of Andreev crystals” generalized this framework to helical magnetic textures and to regimes beyond nearest-neighbor coupling using Eilenberger theory (Rouco et al., 2021). In the helical case, the in-gap spectrum consists of a pair of energy-symmetric bands. Ferromagnetic configurations allow band crossings without hybridization, whereas generic helical configurations produce a Dirac point at $_2$19 and $_2$20, followed by an inverted gap upon further increase of magnetic strength (Rouco et al., 2021). The Eilenberger treatment computed local density of states and local spin polarization for infinite chains and semi-infinite junctions, and showed that junctions may exhibit interface bound states together with fractionalization of the surface spin polarization. In this usage, an Andreev crystal is a bona fide one-dimensional crystal of hybridized ABS.

5. Hybrid Josephson-junction arrays

A more recent transport-theoretic usage concerns periodic arrays of short superconducting segments embedded in a proximitized semiconducting channel. When the superconducting finger length $_2$21 is comparable to the coherence length, Andreev bound states localized near opposite NS interfaces hybridize within a segment and then across the array, generating subgap bands below the parent gap (Dahl et al., 15 Aug 2025). The formalism uses concatenated scattering matrices for superconducting segments and normal spacers, together with the Andreev amplitudes

$_2$22

and shows that high transparency together with a constant phase bias between neighboring superconductors produces directional bands (Dahl et al., 15 Aug 2025).

The characteristic resonance condition of the minimal NSNSN structure is $_2$23, with $_2$24, reproducing the short-junction ABS alignment $_2$25 in the high-transparency limit (Dahl et al., 15 Aug 2025). In extended arrays, the nonlocal conductances $_2$26 and $_2$27 form distinct subgap bands in the $_2$28 plane. The central claim is that one band consists only of right-moving, and the other only of left-moving, electronic states, enabling a flux- and bias-voltage-tunable one-way filter (Dahl et al., 15 Aug 2025). Phase control can be implemented either by superconducting loops, where $_2$29, or by a uniform current-induced phase gradient.

This version of the concept departs from vortex physics and from spectroscopic nomenclature. It treats the Andreev crystal as an engineered band structure whose transport nonreciprocity derives from coherent phase-biased hybridization of ABS in a periodic Josephson environment.

6. Related topological and interfacial extensions

Several adjacent literatures extend the same logic—crystal structure organizing Andreev phenomena—without always using exactly the same definition. In helical crystals with point group $_2$30, interlayer pairings along the helical hopping produce surface ABS and zero-energy peaks in the surface local density of states. For the $_2$31 surface, three irreducible representations exhibit zero-energy peaks; for the zigzag surface, four do. A one-dimensional winding number,

$_2$32

diagnoses these states and establishes their topological origin (Yoshida et al., 2023).

In type-II Weyl-semimetal/superconductor junctions, crystalline anisotropy produces a distinctly different “Andreev-crystal” analogy: an incident electron can undergo double Andreev reflection into two hole beams, one retro and one specular, symmetric about the interface normal, with a critical orientation $_2$33 separating double-Andreev and normal-reflection regimes (Hou et al., 2017). Here the crystal analogy is optical: the tilted Weyl cone acts as an electronic birefringent medium for Andreev processes.

A farther extension appears at the solid/liquid $_2$34He interface, in direct lineage from A. F. Andreev’s crystallization-wave theory. Below the Néel temperature, magnetic crystallization waves acquire an additional spin-supercurrent contribution to their inertial mass,

$_2$35

so that the inertial mass differs from the gravitational mass of the interfacial mode (Todoshchenko et al., 21 Feb 2025). This usage is not about Andreev reflection at all, but it preserves the deeper theme of periodic interfacial dynamics shaped by Andreev’s ideas.

Taken together, these works show that “Andreev crystals” does not denote a single material class. It denotes a family of constructions in which coherent electron–hole superpositions are ordered by periodicity, vortex matter, magnetic texture, crystal symmetry, or phase bias. The most restrictive definition is the periodic array of bound-state-hosting objects—vortex cores, magnetic impurities, or Josephson elements. The broadest definition is methodological: a crystalline setting in which Andreev spectroscopy resolves the momentum structure, topology, or collective organization of superconducting quasiparticles.