Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

arXiv:1805.02537 · Chinese Physics Letters

Topological Invariants in Terms of Green’s Function for the Interacting Kitaev Chain

Zhidan Li, Qiang Han

Published Jul 1, 2018• 1 citation
arXivHTMLPDF

Abstract

A one-dimensional closed interacting Kitaev chain and the dimerized version are studied. The topological invariants in terms of Green’s function are calculated by the density matrix renormalization group method and the exact diagonalization method. For the interacting Kitaev chain, we point out that the calculation of the topological invariant in the charge density wave phase must consider the dimerized configuration of the ground states. The variation of the topological invariant is attributed to the poles of eigenvalues of the zero-frequency Green functions. For the interacting dimerized Kitaev chain, we show that the topological invariant defined by Green’s functions can distinguish more topological nonequivalent phases than the fermion parity .

DOI
10.1088/0256-307x/35/7/077101

Related papers

Topological invariants and interacting one-dimensional fermionic systems

Salvatore R. Manmana, Andrew M. Essin, R. M. Noack +1 more

We study one-dimensional, interacting, gapped fermionic systems described by variants of the Peierls-Hubbard model, and we characterize their phases via a topological invariant constructed out of their Green's functions. We demonstrate that the existence of topologically protected, zero-energy states at the boundaries of these systems can be tied to the value of the topological invariant, just like when working with the conventional, noninteracting topological insulators. We use a combination of analytical methods and the numerical density matrix renormalization group method to calculate the values of the topological invariant throughout the phase diagrams of these systems, thus deducing when topologically protected boundary states are present. We are also able to study topological states in spin systems because, deep in the Mott insulating regime, these fermionic systems reduce to spin chains. In this way, we associate the zero-energy states at the end of an antiferromagnetic spin-1 Heisenberg chain with a topological invariant equal to 2.

arXiv:1205.5095•Physical Review B•Nov 16, 2012•138 citations

Topological invariants for the Haldane phase of interacting Su-Schrieffer-Heeger chains: Functional renormalization-group approach

Björn Sbierski, Christoph Karrasch

We present a functional renormalization-group approach to interacting topological Green's function invariants with a focus on the nature of transitions. The method is applied to chiral symmetric fermion chains in the Mott limit that can be driven into a Haldane phase. We explicitly show that the transition to this phase is accompanied by a zero of the fermion Green's function. Our results for the phase boundary are quantitatively benchmarked against density matrix renormalization-group data.

arXiv:1805.00839•Physical review. B./Physical review. B•Oct 1, 2018•24 citations

Green's-function windings are not phase invariants in interacting fermion chains

Yan Ru Pei

Windings of the zero-frequency single-particle Green's function are widely used to diagnose interacting topological phases of one-dimensional fermion chains. Such windings are known to change through Green's-function zeros, or under a change of the Hamiltonian, without a phase transition. We show that the winding of the elementary fermion field is not constant on a gapped symmetric phase even modulo the interacting classification, and even when both states compared have zero-free Green's functions. An exact identity moves a finite-depth circuit of symmetric gates from the field onto the state and the Hamiltonian, so dressing the field is the same as moving the state within its phase at constant many-body spectrum. With such dressings the winding changes by one in the trivial phase of a Majorana chain of class BDI, takes every value modulo eight in the trivial phase of the Fidkowski–Kitaev eight-chain model, and goes from minus one to minus two inside the Kitaev phase, so that even its parity, the Green's-function form of the class-D index, is not a phase label. In the phase of the fully dimerized spinful Peierls–Hubbard chain of Manmana and co-workers it changes by two, also with bond inversion kept. Every path between such endpoints crosses a Green's-function zero, here at constant spectrum, so invariance along zero-free gapped paths is untouched. We also prove sufficient conditions for invariance: Gaussian dressings of any state, finite-component dressings of cluster product states, and Clifford monomial dressings of stabilizer states, the last two modulo eight. A computed winding therefore cannot by itself certify an interacting phase.

•Nov 7, 2026•0 citations

Characterization of topological phases of dimerized Kitaev chain via edge correlation functions

Yucheng Wang, Jian-Jian Miao, Hui-Ke Jin +1 more

We study analytically topological properties of a noninteracting modified dimerized Kitaev chain and an exactly solvable interacting dimerized Kitaev chain under open boundary conditions by analyzing two introduced edge correlation functions. The interacting dimerized Kitaev chain at the symmetry point $\mathrm{\ensuremath{\Delta}}=t$ and the chemical potential $\ensuremath{\mu}=0$ can be exactly solved by applying two Jordan-Wigner transformations and a spin rotation, which permits us to calculate the edge correlation functions analytically. We demonstrate that the two edge correlation functions can be used to characterize the trivial, Su-Schrieffer-Heeger-like topological and topological superconductor phases of both the noninteracting and interacting systems and give their phase diagrams.

arXiv:1707.08430•Physical review. B./Physical review. B•Nov 20, 2017•41 citations

Majorana zero modes and long range edge correlation in interacting Kitaev chains: analytic solutions and density-matrix-renormalization-group study

Jian-Jian Miao, Hui-Ke Jin, Fu‐Chun Zhang +1 more

We study Kitaev model in one-dimension with open boundary condition by using exact analytic methods for non-interacting system at zero chemical potential as well as in the symmetric case of Δ = t, and by using density-matrix-renormalization-group method for interacting system with nearest neighbor repulsion interaction. We suggest and examine an edge correlation function of Majorana fermions to characterize the long range order in the topological superconducting states and study the phase diagram of the interating Kitaev chain.

arXiv:1608.08382•Scientific Reports•Jan 5, 2018•30 citations

Exact ground states and topological order in interacting Kitaev/Majorana chains

Hosho Katsura, Dirk Schuricht, Masahiro O. Takahashi

We study a system of interacting spinless fermions in one dimension that, in the absence of interactions, reduces to the Kitaev chain [Kitaev, Phys. Usp. 44, 131 (2001)]. In the noninteracting case, a signal of topological order appears as zero-energy modes localized near the edges. We show that the exact ground states can be obtained analytically even in the presence of nearest-neighbor repulsive interactions when the on-site (chemical) potential is tuned to a particular function of the other parameters. As with the noninteracting case, the obtained ground states are twofold degenerate and differ in fermionic parity. We prove the uniqueness of the obtained ground states and show that they can be continuously deformed to the ground states of the noninteracting Kitaev chain without gap closing. We also demonstrate explicitly that there exists a set of operators each of which maps one of the ground states to the other with opposite fermionic parity. These operators can be thought of as an interacting generalization of Majorana edge zero modes.

arXiv:1507.04444•Physical Review B•Sep 18, 2015•161 citations
A one-dimensional closed interacting Kitaev chain and the dimerized version are studied. The topological invariants in terms of Green’s function are calculated by the density matrix renormalization group method and the exact diagonalization method. For the interacting Kitaev chain, we point out that the calculation of the topological invariant in the charge density wave phase must consider the dimerized configuration of the ground states. The variation of the topological invariant is attributed to the poles of eigenvalues of the zero-frequency Green functions. For the interacting dimerized Kitaev chain, we show that the topological invariant defined by Green’s functions can distinguish more topological nonequivalent phases than the fermion parity .
We study one-dimensional, interacting, gapped fermionic systems described by variants of the Peierls-Hubbard model, and we characterize their phases via a topological invariant constructed out of their Green's functions. We demonstrate that the existence of topologically protected, zero-energy states at the boundaries of these systems can be tied to the value of the topological invariant, just like when working with the conventional, noninteracting topological insulators. We use a combination of analytical methods and the numerical density matrix renormalization group method to calculate the values of the topological invariant throughout the phase diagrams of these systems, thus deducing when topologically protected boundary states are present. We are also able to study topological states in spin systems because, deep in the Mott insulating regime, these fermionic systems reduce to spin chains. In this way, we associate the zero-energy states at the end of an antiferromagnetic spin-1 Heisenberg chain with a topological invariant equal to 2.
We present a functional renormalization-group approach to interacting topological Green's function invariants with a focus on the nature of transitions. The method is applied to chiral symmetric fermion chains in the Mott limit that can be driven into a Haldane phase. We explicitly show that the transition to this phase is accompanied by a zero of the fermion Green's function. Our results for the phase boundary are quantitatively benchmarked against density matrix renormalization-group data.
Windings of the zero-frequency single-particle Green's function are widely used to diagnose interacting topological phases of one-dimensional fermion chains. Such windings are known to change through Green's-function zeros, or under a change of the Hamiltonian, without a phase transition. We show that the winding of the elementary fermion field is not constant on a gapped symmetric phase even modulo the interacting classification, and even when both states compared have zero-free Green's functions. An exact identity moves a finite-depth circuit of symmetric gates from the field onto the state and the Hamiltonian, so dressing the field is the same as moving the state within its phase at constant many-body spectrum. With such dressings the winding changes by one in the trivial phase of a Majorana chain of class BDI, takes every value modulo eight in the trivial phase of the Fidkowski–Kitaev eight-chain model, and goes from minus one to minus two inside the Kitaev phase, so that even its parity, the Green's-function form of the class-D index, is not a phase label. In the phase of the fully dimerized spinful Peierls–Hubbard chain of Manmana and co-workers it changes by two, also with bond inversion kept. Every path between such endpoints crosses a Green's-function zero, here at constant spectrum, so invariance along zero-free gapped paths is untouched. We also prove sufficient conditions for invariance: Gaussian dressings of any state, finite-component dressings of cluster product states, and Clifford monomial dressings of stabilizer states, the last two modulo eight. A computed winding therefore cannot by itself certify an interacting phase.
We study analytically topological properties of a noninteracting modified dimerized Kitaev chain and an exactly solvable interacting dimerized Kitaev chain under open boundary conditions by analyzing two introduced edge correlation functions. The interacting dimerized Kitaev chain at the symmetry point $\mathrm{\ensuremath{\Delta}}=t$ and the chemical potential $\ensuremath{\mu}=0$ can be exactly solved by applying two Jordan-Wigner transformations and a spin rotation, which permits us to calculate the edge correlation functions analytically. We demonstrate that the two edge correlation functions can be used to characterize the trivial, Su-Schrieffer-Heeger-like topological and topological superconductor phases of both the noninteracting and interacting systems and give their phase diagrams.
We study Kitaev model in one-dimension with open boundary condition by using exact analytic methods for non-interacting system at zero chemical potential as well as in the symmetric case of Δ = t, and by using density-matrix-renormalization-group method for interacting system with nearest neighbor repulsion interaction. We suggest and examine an edge correlation function of Majorana fermions to characterize the long range order in the topological superconducting states and study the phase diagram of the interating Kitaev chain.
We study a system of interacting spinless fermions in one dimension that, in the absence of interactions, reduces to the Kitaev chain [Kitaev, Phys. Usp. 44, 131 (2001)]. In the noninteracting case, a signal of topological order appears as zero-energy modes localized near the edges. We show that the exact ground states can be obtained analytically even in the presence of nearest-neighbor repulsive interactions when the on-site (chemical) potential is tuned to a particular function of the other parameters. As with the noninteracting case, the obtained ground states are twofold degenerate and differ in fermionic parity. We prove the uniqueness of the obtained ground states and show that they can be continuously deformed to the ground states of the noninteracting Kitaev chain without gap closing. We also demonstrate explicitly that there exists a set of operators each of which maps one of the ground states to the other with opposite fermionic parity. These operators can be thought of as an interacting generalization of Majorana edge zero modes.