Long-Range Kitaev Chain Overview

• Long-range Kitaev chain is a one-dimensional p-wave superconducting model featuring algebraically decaying pairing amplitudes that extend beyond nearest neighbors.
• It exhibits hybrid exponential–algebraic correlation decay, modified bulk-boundary correspondence, and novel edge excitations such as massive Dirac modes and half-integer winding numbers.
• The framework spans various extensions including finite-range, self-consistent, and effective platforms like planar Josephson junctions, offering diverse experimental probes of nonlocal superconductivity.

The long-range Kitaev chain is a one-dimensional spinless $p$-wave superconducting lattice model in which the pairing, and in broader variants also the hopping or the effective interaction-induced gap matrix, extends beyond nearest neighbors with algebraically decaying amplitudes. It generalizes the standard Kitaev chain recovered in the limit of infinitely short-ranged couplings, but its phase structure is substantially richer: depending on the decay exponent, symmetry class, and the precise implementation of nonlocality, it can exhibit hybrid exponential–algebraic correlations in gapped phases, massive subgap edge modes, half-integer bulk indices, area-law violation, anomalous entanglement dynamics, Floquet $\pi/T$ modes, and disorder-induced reentrance (Vodola et al., 2014).

1. Canonical formulation and major variants

In its canonical pairing-only form, the model is a chain of spinless fermions with nearest-neighbor hopping and chemical potential, but with long-range superconducting pairing decaying as a power law. A standard representative Hamiltonian is

$\begin{split} H_L &= - t \sum_{j=1}^{L} \left(a^\dagger_j a_{j+1} + \mathrm{H.c.}\right) - \mu \sum_{j=1}^L \left(n_j - \frac{1}{2}\right) \ &\quad +\frac{\Delta}{2} \sum_{j=1}^L \sum_{\ell=1}^{L-1} d_\ell^{-\alpha} \left( a_j a_{j+\ell} + a^\dagger_{j+\ell} a^\dagger_j\right), \end{split}$

where the pairing amplitude scales as $d_\ell^{-\alpha}$, and $\alpha\to\infty$ yields the ordinary nearest-neighbor Kitaev chain (Vodola et al., 2014). In open chains one takes $d_\ell=\ell$, while on a ring one uses the shortest lattice distance.

The literature uses several non-equivalent extensions under the same umbrella term. Finite-range “extended Kitaev chains” allow hopping and pairing only up to a fixed neighbor distance $r$, often with algebraic amplitudes; in BDI symmetry this permits winding numbers as large as $r$ and, correspondingly, up to $r$ Majorana zero modes per edge (Alecce et al., 2017). Other works allow distinct decay exponents for hopping and pairing, $j_r^\alpha$ and $\pi/T$0, and show that edge-mode asymptotics is then controlled by the smaller of $\pi/T$1 and $\pi/T$2 (Jäger et al., 2020). Quasiperiodic variants add Aubry–André–Harper modulation to the onsite potential and reveal massive Dirac edge modes, half-integer winding numbers, and weakened bulk–boundary correspondence (Fraxanet et al., 2020).

A further distinction concerns whether long-range pairing is imposed directly or generated self-consistently from an underlying long-range interaction. In the self-consistent long-range Kitaev chain, the pairing matrix is not simply $\pi/T$3, but develops a short-range bulk band plus edge-to-edge long-range structure, with important consequences for edge-mode hybridization (Haink et al., 30 Sep 2025). Finally, effective long-range Kitaev chains also arise as low-energy reductions of higher-dimensional devices, notably planar Josephson junctions on Rashba 2DEGs, where both hopping and pairing become extended and tunable in real space (Liu et al., 2018).

With real hopping and pairing amplitudes the relevant symmetry class is typically BDI, supporting an integer topological invariant. Breaking effective time-reversal symmetry reduces the problem to class D with a $\pi/T$4 classification (Alecce et al., 2017).

2. Bulk spectrum and static phase structure

Because the canonical Hamiltonian is quadratic, it is exactly solvable by Fourier and Bogoliubov transformations. For a closed chain with antiperiodic boundary conditions, the momentum-dependent pairing function is

$\pi/T$5

and the quasiparticle dispersion is

$\pi/T$6

The non-analyticity of $\pi/T$7 near $\pi/T$8 is the source of the distinctive long-range behavior (Vodola et al., 2014).

In the pairing-only model with the convention $\pi/T$9, the phase diagram is asymmetric in $\begin{split} H_L &= - t \sum_{j=1}^{L} \left(a^\dagger_j a_{j+1} + \mathrm{H.c.}\right) - \mu \sum_{j=1}^L \left(n_j - \frac{1}{2}\right) \ &\quad +\frac{\Delta}{2} \sum_{j=1}^L \sum_{\ell=1}^{L-1} d_\ell^{-\alpha} \left( a_j a_{j+\ell} + a^\dagger_{j+\ell} a^\dagger_j\right), \end{split}$0 once $\begin{split} H_L &= - t \sum_{j=1}^{L} \left(a^\dagger_j a_{j+1} + \mathrm{H.c.}\right) - \mu \sum_{j=1}^L \left(n_j - \frac{1}{2}\right) \ &\quad +\frac{\Delta}{2} \sum_{j=1}^L \sum_{\ell=1}^{L-1} d_\ell^{-\alpha} \left( a_j a_{j+\ell} + a^\dagger_{j+\ell} a^\dagger_j\right), \end{split}$1 is finite. For $\begin{split} H_L &= - t \sum_{j=1}^{L} \left(a^\dagger_j a_{j+1} + \mathrm{H.c.}\right) - \mu \sum_{j=1}^L \left(n_j - \frac{1}{2}\right) \ &\quad +\frac{\Delta}{2} \sum_{j=1}^L \sum_{\ell=1}^{L-1} d_\ell^{-\alpha} \left( a_j a_{j+\ell} + a^\dagger_{j+\ell} a^\dagger_j\right), \end{split}$2, the system is gapped for $\begin{split} H_L &= - t \sum_{j=1}^{L} \left(a^\dagger_j a_{j+1} + \mathrm{H.c.}\right) - \mu \sum_{j=1}^L \left(n_j - \frac{1}{2}\right) \ &\quad +\frac{\Delta}{2} \sum_{j=1}^L \sum_{\ell=1}^{L-1} d_\ell^{-\alpha} \left( a_j a_{j+\ell} + a^\dagger_{j+\ell} a^\dagger_j\right), \end{split}$3, and the lines $\begin{split} H_L &= - t \sum_{j=1}^{L} \left(a^\dagger_j a_{j+1} + \mathrm{H.c.}\right) - \mu \sum_{j=1}^L \left(n_j - \frac{1}{2}\right) \ &\quad +\frac{\Delta}{2} \sum_{j=1}^L \sum_{\ell=1}^{L-1} d_\ell^{-\alpha} \left( a_j a_{j+\ell} + a^\dagger_{j+\ell} a^\dagger_j\right), \end{split}$4 are the natural critical loci. For $\begin{split} H_L &= - t \sum_{j=1}^{L} \left(a^\dagger_j a_{j+1} + \mathrm{H.c.}\right) - \mu \sum_{j=1}^L \left(n_j - \frac{1}{2}\right) \ &\quad +\frac{\Delta}{2} \sum_{j=1}^L \sum_{\ell=1}^{L-1} d_\ell^{-\alpha} \left( a_j a_{j+\ell} + a^\dagger_{j+\ell} a^\dagger_j\right), \end{split}$5, however, the line $\begin{split} H_L &= - t \sum_{j=1}^{L} \left(a^\dagger_j a_{j+1} + \mathrm{H.c.}\right) - \mu \sum_{j=1}^L \left(n_j - \frac{1}{2}\right) \ &\quad +\frac{\Delta}{2} \sum_{j=1}^L \sum_{\ell=1}^{L-1} d_\ell^{-\alpha} \left( a_j a_{j+\ell} + a^\dagger_{j+\ell} a^\dagger_j\right), \end{split}$6 becomes gapped, so that phases adiabatically disconnected in the short-range chain can be connected without bulk gap closing (Vodola et al., 2014). In other parameter conventions used elsewhere, the same structure appears at $\begin{split} H_L &= - t \sum_{j=1}^{L} \left(a^\dagger_j a_{j+1} + \mathrm{H.c.}\right) - \mu \sum_{j=1}^L \left(n_j - \frac{1}{2}\right) \ &\quad +\frac{\Delta}{2} \sum_{j=1}^L \sum_{\ell=1}^{L-1} d_\ell^{-\alpha} \left( a_j a_{j+\ell} + a^\dagger_{j+\ell} a^\dagger_j\right), \end{split}$7 rather than $\begin{split} H_L &= - t \sum_{j=1}^{L} \left(a^\dagger_j a_{j+1} + \mathrm{H.c.}\right) - \mu \sum_{j=1}^L \left(n_j - \frac{1}{2}\right) \ &\quad +\frac{\Delta}{2} \sum_{j=1}^L \sum_{\ell=1}^{L-1} d_\ell^{-\alpha} \left( a_j a_{j+\ell} + a^\dagger_{j+\ell} a^\dagger_j\right), \end{split}$8, reflecting only normalization differences in the Hamiltonian (Bhattacharya et al., 2018).

Finite-range and truly long-range models with both hopping and pairing display a broader hierarchy of phases. In the finite-range case, winding numbers up to $\begin{split} H_L &= - t \sum_{j=1}^{L} \left(a^\dagger_j a_{j+1} + \mathrm{H.c.}\right) - \mu \sum_{j=1}^L \left(n_j - \frac{1}{2}\right) \ &\quad +\frac{\Delta}{2} \sum_{j=1}^L \sum_{\ell=1}^{L-1} d_\ell^{-\alpha} \left( a_j a_{j+\ell} + a^\dagger_{j+\ell} a^\dagger_j\right), \end{split}$9 are possible, but they are not generically stable: as the decay exponent increases, higher-order topological quantum critical lines superpose and vanish, and the accessible sequence of phases collapses toward the nearest-neighbor pattern $d_\ell^{-\alpha}$0 (Kartik et al., 2020). In the truly long-range limit, the distinction $d_\ell^{-\alpha}$1 versus $d_\ell^{-\alpha}$2 becomes structural. For $d_\ell^{-\alpha}$3, the bulk pseudo-spin map remains regular and integer winding numbers are available; for $d_\ell^{-\alpha}$4, the long-range contribution is singular at $d_\ell^{-\alpha}$5, and the standard momentum-space invariant becomes ill-defined (Kartik et al., 2020).

This separation between weak and strong long-range regimes is therefore not merely quantitative. It determines whether the low-energy theory remains short-range-like, whether conventional topological indices survive, and whether critical lines continue to be associated with ordinary bulk gap closings.

3. Topology, winding numbers, and edge excitations

In the short-range reference problem, the topological superconducting phase supports Majorana zero modes localized at the ends. Long-range pairing preserves this scenario only in part. For $d_\ell^{-\alpha}$6, open chains still support boundary modes in the topological region, but their localization is hybrid: near the boundary the probability density decays exponentially, while asymptotically it develops a power-law tail,

$d_\ell^{-\alpha}$7

and the finite-size splitting crosses over from exponential to algebraic behavior (Vodola et al., 2014).

For $d_\ell^{-\alpha}$8, several analyses find that the two edge Majoranas no longer decouple in the thermodynamic limit. Instead, they hybridize into a finite-energy subgap state, often described as a massive Dirac mode or massive edge mode (Giuliano et al., 2017). In quasiperiodic long-range chains with Aubry–André–Harper modulation, these massive modes remain sharply visible in the central gap, can undergo true zero-energy crossings, and are associated with half-integer winding numbers such as $d_\ell^{-\alpha}$9 (Fraxanet et al., 2020). Related static analyses of long-range pairing chains likewise identify symmetry-protected massive Dirac end modes for $\alpha\to\infty$0, while $\alpha\to\infty$1 retains massless Majorana zero modes (Bhattacharya et al., 2018).

These results are tied to a broader modification of bulk–boundary correspondence. In the long-range AAH model, the bulk winding number can remain fixed while edge behavior changes dramatically, and Chern numbers in nonzero-energy gaps no longer count edge-mode crossings in the usual way (Fraxanet et al., 2020). This does not eliminate topology, but it weakens the short-range one-to-one correspondence between integer invariants and edge spectra.

A separate line of work shows that finite-range BDI chains can host several independent Majorana zero modes per edge. When hopping and pairing extend to the $\alpha\to\infty$2-th neighbor, the maximal number of zero modes per edge is $\alpha\to\infty$3, and this counting can be derived by a generalized transfer-matrix construction (Alecce et al., 2017).

An important recent refinement is that self-consistent long-range interacting versions need not reproduce the same edge physics as imposed long-range pairing. In the self-consistent model, edge-mode wavefunctions remain exponentially localized, yet their splitting decays algebraically with chain length as $\alpha\to\infty$4 for all $\alpha\to\infty$5; the mechanism is non-local edge-mode hybridization through the self-consistent gap matrix, not power-law bulk tails of the wavefunctions (Haink et al., 30 Sep 2025). This suggests that “long-range Kitaev chain” is not a single universality class of edge behavior, but a family of models whose topology depends sensitively on how nonlocality is implemented.

4. Correlations, entanglement, and criticality

One of the defining signatures of long-range pairing is that gapped phases need not look short-range. In the canonical model, for $\alpha\to\infty$6 the gapped phases show exponential decay at short and intermediate distances but algebraic tails at long distances,

$\alpha\to\infty$7

whereas for $\alpha\to\infty$8 correlations become purely algebraic at all scales despite a nonzero spectral gap; in particular, $\alpha\to\infty$9 for all $d_\ell=\ell$0 (Vodola et al., 2014).

Critical behavior is likewise modified. Along the line $d_\ell=\ell$1 of the pairing-only model, the system is Ising-like for $d_\ell=\ell$2, with $d_\ell=\ell$3. As $d_\ell=\ell$4 decreases below $d_\ell=\ell$5, the entanglement entropy still admits a logarithmic fit,

$d_\ell=\ell$6

but $d_\ell=\ell$7 drifts continuously and reaches $d_\ell=\ell$8 at $d_\ell=\ell$9. Simultaneously, finite-size ground-state energy scaling ceases to match conformal-field-theory expectations because high-energy modes near $r$0 contribute nonuniversally or with divergent derivatives (Vodola et al., 2014). In this sense, long-range pairing can preserve gaplessness while breaking conformal symmetry.

Entanglement is particularly sensitive to nonlocality. The same work shows logarithmic area-law violation throughout wide gapped regions, especially for $r$1, and also near $r$2 for $r$3 (Vodola et al., 2014). In more general chains with long-range hopping and pairing subject to Kac normalization, the asymptotic ground-state entanglement can be logarithmic, fractal, or volume-law. For $r$4 at $r$5, the entropy scales as

$r$6

while at the mean-field point $r$7 a volume law

$r$8

emerges from an extensive set of zero modes (Solfanelli et al., 2023). This extends the long-range Kitaev chain from an anomalous topological superconductor to a controlled setting for nonstandard entanglement scaling.

A complementary result concerns gapless topological criticality. In an extended critical fermionic chain with long-range interactions and next-nearest-neighbor terms, the transition between $r$9 and $r$0 topological superconductors remains second order with $r$1 and $r$2, and the critical edge modes stay massless even for strong long-range coupling (Zhong et al., 2024). This sharply contrasts with the massive edge modes of gapped long-range phases and shows that long-range interactions affect critical and gapped topology differently.

5. Nonequilibrium dynamics, transport, Floquet structure, and disorder

Long-range pairing strongly reshapes nonequilibrium propagation. After a global quench, mutual information still displays a dominant ballistic light-cone structure, but a significant amount of information appears at time-like separations, and local observables can equilibrate very slowly because of quasi-particle velocity distributions with both ultrafast and ultraslow components (Regemortel et al., 2015). In the canonical half-chain quench to $r$3, the entanglement entropy grows linearly in time for $r$4, as in Calabrese–Cardy theory, but only logarithmically for $r$5, providing a direct dynamical signature of broken conformal behavior (Vodola et al., 2014).

Ramped quenches reveal further long-range-specific effects. When a linear ramp of the chemical potential crosses a single quantum critical point, long-range pairing can produce a region with three distinct DQPT time scales, whereas the short-range chain shows only one. This three-scale region shrinks under noise. For ramps crossing two critical points, the critical sweep velocity above which DQPTs disappear is enhanced by long-range pairing and reduced by noise (Baghran et al., 2024).

Transport through normal leads offers a distinct diagnostic. In an NSN geometry, the long-range Kitaev chain exhibits a Fano factor that tends to $r$6 in the short-range topological phase, $r$7 in the long-range topological phase with massive edge modes, and exactly $r$8 on the critical line $r$9. The interpretation is that short-range Majorana zero modes yield noiseless local Andreev reflection, whereas finite-energy long-range edge modes generate rare Poissonian Cooper-pair processes (Giuliano et al., 2017).

Periodic driving of the chemical potential produces a Floquet counterpart of long-range topology. For $r$0, periodic $r$1-kicks generate topologically protected end modes not only at quasienergy $r$2 but also at $r$3, and a bulk integer-counting invariant correctly predicts their number. For $r$4, the driven system supports massive Floquet Dirac end modes rather than massless Floquet Majoranas (Bhattacharya et al., 2018).

Disorder enriches the picture further in open, bath-coupled systems. In a disordered long-range pairing chain attached to metallic leads and Lindblad baths, increasing onsite disorder can produce reentrant behavior of the massive topological phase and, according to spectral and transport diagnostics, a disorder-induced direct transition between the massive phase and a short-range-like topological phase (Cinnirella et al., 2024). This suggests that disorder does not merely destroy long-range topology; at intermediate strength it can also transmute its edge phenomenology.

6. Realizations, effective platforms, and broader landscape

Long-range Kitaev physics is not confined to a single microscopic platform. A prominent solid-state route is the planar Josephson junction in a Rashba 2DEG with in-plane Zeeman field, whose low-energy Andreev bands map to an effective one-dimensional Kitaev chain with long-range hopping and pairing. In that setting, the couplings are tunable through the chemical potential, phase difference $r$5, Zeeman field, and junction geometry, and topology under disorder can be tracked by the real-space Clifford pseudospectrum index of Hastings and Loring, as implemented in this context following work also associated with Fulga and collaborators (Liu et al., 2018).

Other proposed or discussed realizations include trapped-ion simulators of long-range Ising chains, which are Jordan–Wigner related to fermionic long-range pairing models, helical Shiba chains and magnetic adatom arrays on superconductors, weak-coupling semiconductor–superconductor nanowires with effective long-range pairing and hopping, and ultracold atoms or polar molecules with engineered quasi-periodic potentials and long-range interactions (Vodola et al., 2014). In these settings, observables such as entanglement growth, lead conductance, Fano factors, Floquet spectroscopy, and disorder-averaged subgap transport provide experimentally differentiated probes of long-range regimes.

Across this landscape, a recurring theme is that “long-range Kitaev chain” does not denote a single deformation of the original model. Pairing-only chains, hopping-plus-pairing chains, finite-range BDI extensions, quasiperiodic chains, Floquet chains, and self-consistent interaction-induced models agree on the importance of algebraic nonlocality, but they do not agree on every consequence. Massive Dirac edge modes, half-integer winding numbers, and weakened bulk–boundary correspondence arise naturally in imposed long-range pairing models (Fraxanet et al., 2020), whereas self-consistent constructions can retain exponentially localized edge wavefunctions while modifying only their finite-size splitting (Haink et al., 30 Sep 2025). This suggests that the long-range Kitaev chain is best understood not as a single model, but as a broad theoretical framework for studying how algebraically decaying couplings reorganize one-dimensional topological superconductivity.