betaFeedback

WorksPhysics and Astronomy

Low-energy states of a Su-Schrieffer-Heeger chain connected perpendicularly to a metallic lead

Abstract

Recently, Su-Schrieffer-Heeger chains (SSHCs) coupled to metallic leads have attracted attention. In particular Mart'ınez et al. [Phys. Rev. Res. 8, 043118 (2026)] studied a topological SSHC connected at one of its ends perpendicularly to a metallic lead modeled by a tight-binding chain. Here, we extend their results by incorporating the finite-size splitting of the end modes of the isolated SSHC, which are determined analytically with high accuracy. We also derive simple analytical expressions, valid in the low-energy regime, for the zero-temperature conductance through the lead and the relevant spectral densities. We find that, when the chemical potential coincides with that of the half-filled SSHC, the conductance reaches its maximum value, equal to that of the lead in the absence of the SSHC. By contrast, the conductance vanishes when the chemical potential is shifted by an amount equal to half the end-mode splitting.

Introduction

In the last two decades, topological materials have aroused considerable interest [1, 2, 3, 4]. Several experimental studies have realized extensions of the Su-Schrieffer-Heeger chain (SSHC) [5, 6, 7, 8]. In its simplest form, the SSHC introduced to describe polyacetylene [9], consists of a dimerized chain of spinless fermions with alternating nearest-neighbor couplings. It supports two distinct topological phases, characterized by a Z2\mathbb{Z}_2 topological invariant. In particular, for an appropriate choice of unit cell, the Zak phase takes the values 00 and π\pi in the trivial and topological phases, respectively [4], corresponding to winding numbers of 00 and 11. In the topological phase of a finite chain, the absolute values of the first and last hoppings are smaller than those of the other hoppings. In this case, if the chain is infinite, a zero-energy mode appears at each end of the chain.

In recent years, numerous extended SSHCs have been investigated [5, 6, 7, 8, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21], some of which support integer winding numbers beyond 00 and 11 [5, 10, 12, 13, 15, 16, 19]. In addition, SSHCs (some including interactions) coupled to other systems, such as scanning tunnelling microscope (STM) tips [17, 22] or metallic leads [23, 24, 25], have also been investigated recently.

In particular, Martínez et al. [25] studied a topological SSHC coupled at one of its ends to the midpoint of a metallic lead modeled by a tight-binding chain. For a sufficiently long SSHC, the end state located at the opposite end of the chain, far from the metallic lead, becomes a topologically protected bound state in the continuum at zero energy, which is of particular interest. Their work extended a previous approach in which the states of an infinitely long SSHC were considered [24]. However, in a finite chain, the two end states hybridize, resulting in a finite energy splitting away from zero. This finite-size effect modifies the low-energy spectral densities and the conductance through the metallic lead.

The goal of the present work is to describe these modifications using a simple yet highly accurate analytical approach valid in the low-energy regime. The paper is organized as follows. In Section II, we introduce the model and the approximation used. Section III contains the results and Section IV summarizes and discusses our results.

Model and approximation

The starting model is the same as that considered by Martínez et al. [25]. The Hamiltonian is

H=HSSHC+HML+HC,(1)H = H_{\mathrm{SSHC}} + H_{\mathrm{ML}} + H_{\mathrm{C}}, \tag*{(1)}

where the first term corresponds to an isolated SSHC of NN unit cells (2N2N sites)

HSSHC=−v∑i=1N(ai†bi+H.c.)−w∑i=1N−1(bi†ai+1+H.c.),(2)H_{\mathrm{SSHC}} = -v \sum_{i=1}^{N} \left(a_i^\dagger b_i + \mathrm{H.c.}\right) - w \sum_{i=1}^{N-1} \left(b_i^\dagger a_{i+1} + \mathrm{H.c.}\right), \tag*{(2)}

the second term corresponds to the metallic lead

HML=−J∑j(cj†cj+1+H.c.),(3)H_{\mathrm{ML}} = -J \sum_j \left(c_j^\dagger c_{j+1} + \mathrm{H.c.}\right), \tag*{(3)}

and the last term is the coupling between them

HC=−tCc0†a1+H.c.(4)H_{\mathrm{C}} = -t_{\mathrm{C}} c_0^\dagger a_1 + \mathrm{H.c.} \tag*{(4)}

With an appropriate choice of phases of all the operators, one can assume v,w,J,tC>0v,w,J,t_{\mathrm{C}}>0.

In the limit N→∞N \to\infty, in the topological phase that takes place when v≪wv \ll w, HSSHCH_{\mathrm{SSHC}} has two zero-energy modes, one at the left end of the SSHC ∣l⟩|l\rangle and another at right one ∣r⟩|r\rangle, where [17, 24]

l†=p∑i=1Nzj−1aj†(5)l^\dagger= p\sum_{i=1}^{N} z^{j-1}a_j^\dagger \tag*{(5)}
r†=p∑j=1NzN−jbj†(6)r^\dagger= p\sum_{j=1}^{N} z^{N-j}b_j^\dagger \tag*{(6)}
z=−v/w,p=1−z21−z2N.(7)z=-v/w,\quad p=\sqrt{\frac{1-z^2}{1-z^{2N}}}. \tag*{(7)}

For finite chains and absolute value of the energies below half of the gap Eg=w−vE_g=w-v one has [17]

HSSHC≃HSSHCLE=−tlrl†r+H.c.,H_{\mathrm{SSHC}}\simeq H_{\mathrm{SSHC}}^{\mathrm{LE}}=-t_{lr}l^\dagger r+\mathrm{H.c.},
tlr=1−z21−z2NzN−1v,(8)t_{lr}=\frac{1-z^2}{1-z^{2N}}z^{N-1}v, \tag*{(8)}

where the superscript LE stands for low energy. This is the only approximation made in this work. The relative error introduced by the states at energies outside the gap (E<−EgE<-E_g or E>EgE>E_g) is of order zN+2/(1+z)z^{N+2}/(1+z) [17] and decays exponentially with increasing NN. Comparison with numerical data for the eigenergies near zero is excellent (Fig. 6 of Ref. [17]). An extension of this approach to the SSHC including other hoppings also shows an excellent agreement with numerical data for the low-energy eigenstates [19].

Projecting Eq. (4) in the low-energy sector one obtains

HC≃HCLE=−tclcl†c0+H.c.,H_C\simeq H_C^{\mathrm{LE}}=-t_{cl}c_l^\dagger c_0+\mathrm{H.c.},
tcl=ptc.(9)t_{cl}=pt_c. \tag*{(9)}

Results

For convenience, we write the low-energy part of the Hamiltonian as HLE=H0+VH^{\mathrm{LE}}=H^0+V, where H0=HSSHCLE+HMLH^0=H_{\mathrm{SSHC}}^{\mathrm{LE}}+H_{\mathrm{ML}} and V=HCLEV=H_C^{\mathrm{LE}}. We define the Green function operators G^=(ω−HLE)−1\hat{G}=(\omega-H^{\mathrm{LE}})^{-1} and G^0=(ω−H0)−1\hat{G}^0=(\omega-H^0)^{-1}. We are particularly interested in the Green function at site 0 of the metallic lead, G^00(ω)=⟨⟨c0;c0†⟩⟩ω\hat{G}_{00}(\omega)=\langle\langle c_0;c_0^\dagger\rangle\rangle_\omega, which as shown below, determines the conductance through the metallic lead. The Green functions for V=0V=0 are the matrix elements of G^0\hat{G}^0 and are known. The Green functions of the complete system can be found from the matrix elements of the equation

G^=G^0+G^0VG^,(10)\hat{G}=\hat{G}^0+\hat{G}^0V\hat{G}, \tag*{(10)}

derived from the definitions of G^\hat{G} and G^0\hat{G}^0.

The conductance GG through the metallic lead can be expressed in terms of the spectral density ρ0(ω)=(−1/π)Im⁡G^00(ω)\rho_0(\omega)=(-1/\pi)\operatorname{Im}\hat{G}_{00}(\omega), following the Meir-Wingreen formalism [26]. The resulting expression is analogous to that for transport through a quantum dot embedded between metallic leads [27], and is given by

G=πG0∫dω(−df(ω)dω)Γ(ω)ρ0(ω),G=\pi G_0\int d\omega\left(-\frac{df(\omega)}{d\omega}\right)\Gamma(\omega)\rho_0(\omega),
G0=e2h,Γ(ω)=4J2−ω2,(11)G_0=\frac{e^2}{h},\quad\Gamma(\omega)=\sqrt{4J^2-\omega^2}, \tag*{(11)}

where f(ω)f(\omega) is the Fermi function.

At zero temperature Eq. (11) reduces to

G=πG04J2−μ2ρ0(μ),(12)G=\pi G_0\sqrt{4J^2-\mu^2}\rho_0(\mu), \tag*{(12)}

where μ\mu is the chemical potential.

Taking the matrix element 00 of Eq. (10) one obtains

G^00=G^000−G^000tclG^l0,(13)\hat{G}_{00}=\hat{G}_{00}^0-\hat{G}_{00}^0t_{cl}\hat{G}_{l0}, \tag*{(13)}

and similarly from the element l0l0

G^l0=−G^ll0tclG^00.(14)\hat{G}_{l0}=-\hat{G}_{ll}^0t_{cl}\hat{G}_{00}. \tag*{(14)}

Replacing this result in the previous equation one obtains

G^00=G^000+tcl2G^000G^ll0G^00,(15)\hat{G}_{00}=\hat{G}_{00}^0+t_{cl}^2\hat{G}_{00}^0\hat{G}_{ll}^0\hat{G}_{00}, \tag*{(15)}

which allows to express G^00\hat{G}_{00} in terms of the Green functions of H0H^0. Using

G^000=−i4J2−ω2,G^ll0=ωω2−tlr2,(16)\hat{G}_{00}^0=\frac{-i}{\sqrt{4J^2-\omega^2}},\quad\hat{G}_{ll}^0=\frac{\omega}{\omega^2-t_{lr}^2}, \tag*{(16)}

where the first equation is valid for ∣ω∣<J|\omega|<J, one obtains

G^00(ω)=−i4J2−ω2+iC(ω),\hat{G}_{00}(\omega)=\frac{-i}{\sqrt{4J^2-\omega^2}+iC(\omega)},
C(ω)=tcl2ωω2−tlr2.(17)C(\omega)=\frac{t_{cl}^2\omega}{\omega^2-t_{lr}^2}. \tag*{(17)}

There are in addition two poles of G^00(ω)\hat{G}_{00}(\omega) at ω=±[2J2+4J4+tcl4]1/2\omega=\pm\left[2J^2+\sqrt{4J^4+t_{cl}^4}\right]^{1/2}. Replacing in Eq. (12) one obtains for the conductance at zero temperature

G(μ)=4J2−μ24J2−μ2+C2(μ).(18)G(\mu)=\frac{4J^2-\mu^2}{4J^2-\mu^2+C^2(\mu)}. \tag*{(18)}

In Figure 1, we show this function and ρ0(ω)\rho_0(\omega) for selected parameter values. Note that their dependence on the original SSHC parameters, vv and ww, is only indirect, through tlrt_{lr} [see Eq. (8)] and tclt_{cl} [see Eqs. (7) and (9)]. We recall that our results are valid for ∣μ∣,∣ω∣<Eg|\mu|, |\omega| < E_g. Nevertheless, the metallic-lead hopping JJ can be much smaller than the SSHC gap 2Eg2E_g. The conductance reaches the quantum of conductance G0G_0 for μ=0\mu= 0 and vanishes for μ=±tlr\mu= \pm t_{lr}. For other values of the chemical potential the conductance is smaller but near to G0G_0, particularly for small values of tclt_{cl}. These properties might be used to detect experimentally the splitting 2tlr2t_{lr} of the low-energy modes.

Zero-temperature conductance and spectral density of the metallic lead

Figure 1. (Color online) Top: zero-temperature conductance as a function of chemical potential. Bottom: spectral density of the site of the metallic lead connected to the SSHC as a function of energy. parameters are J=1J = 1, tlr=0.1t_{lr} = 0.1 and two values of tclt_{cl}.

Similarly, the spectral density at site 0 is reduced but remains qualitatively similar to that of a disconnected metallic lead (tcl=0t_{cl}=0), except at ω=±tlr\omega= \pm t_{lr}, where it vanishes, and in their vicinity. The reduction of the spectral weight within the interval −2J<ω<2J-2J < \omega< 2J, relative to the disconnected-lead case, is compensated by the appearance of two delta-function peaks at

ω=±[2J2+4J4+tcl4]1/2\omega= \pm\left[2J^2+\sqrt{4J^4+t_{cl}^4}\right]^{1/2}

Proceeding in a similar way as before, one obtains

G^ll(ω)=G^ll0(ω)(1−itcl2G^ll0(ω)itcl2G^ll0(ω)+4J2−ω2)(19)\widehat{G}_{ll}(\omega)=\widehat{G}^{0}_{ll}(\omega)\left(1-\frac{it_{cl}^{2}\widehat{G}^{0}_{ll}(\omega)}{it_{cl}^{2}\widehat{G}^{0}_{ll}(\omega)+\sqrt{4J^{2}-\omega^{2}}}\right) \tag*{(19)}

where G^ll0(ω)\widehat{G}^{0}_{ll}(\omega) is given in Eqs. (16). From ρl(ω)=(−1/π)Im⁡G^ll(ω)\rho_l(\omega)=(-1/\pi)\operatorname{Im}\widehat{G}_{ll}(\omega) one obtains the density of states of the left mode in presence of the metallic lead

left mode in presence of the metallic lead

ρl(ω)=1πtcl2ω24J2−ω2tcl4ω4+(4J2−ω2)(tcl2−ω2)2(20)\rho_l(\omega)=\frac{1}{\pi}\frac{t_{cl}^{2}\omega^{2}\sqrt{4J^{2}-\omega^{2}}}{t_{cl}^{4}\omega^{4}+(4J^{2}-\omega^{2})(t_{cl}^{2}-\omega^{2})^{2}} \tag*{(20)}

Finally G^rr(ω)=⟨⟨r;r⟩⟩ω\widehat{G}_{rr}(\omega)=\langle\langle r;r\rangle\rangle_\omega can be related to G^ll(ω)=⟨⟨l;l⟩⟩ω\widehat{G}_{ll}(\omega)=\langle\langle l;l\rangle\rangle_\omega using appropriate equations of motion

ω⟨⟨r;r⟩⟩ω=1−tlr⟨⟨l;r⟩⟩ω,ω⟨⟨l;r⟩⟩ω=−tlr⟨⟨l;l⟩⟩ω.(21)\begin{aligned} \omega\langle\langle r;r\rangle\rangle_\omega&=1-t_{lr}\langle\langle l;r\rangle\rangle_\omega,\\ \omega\langle\langle l;r\rangle\rangle_\omega&=-t_{lr}\langle\langle l;l\rangle\rangle_\omega. \tag*{(21)} \end{aligned}

From here

G^rr(ω)=ω+tlr2G^ll(ω)ω2,\widehat{G}_{rr}(\omega)=\frac{\omega+t_{lr}^{2}\widehat{G}_{ll}(\omega)}{\omega^{2}},
ρr(ω)=tlr2ω2ρl(ω).(22)\rho_r(\omega)=\frac{t_{lr}^{2}}{\omega^{2}}\rho_l(\omega). \tag*{(22)}

The spectral densities for both end states are shown in Figure 2. In contrast to the density at site 0 of the metallic lead, the density of the left end state vanishes at ω=0\omega= 0 and is peaked at ω=±tlr\omega= \pm t_{lr}. In comparison the density of the right end mode is narrower and with a rather flat maximum in the interval −tlr/2<ω<tlr/2-t_{lr}/2 < \omega< t_{lr}/2.

Figure 2. Spectral density of the left (top) and right (bottom) end state of the SSHC connected at the left to a metallic lead. Parameters are J=1J = 1, tlr=0.1t_{lr} = 0.1 and tcl=0.5t_{cl} = 0.5.

Summary and Discussion

We have analyzed the effect of the splitting of the end modes of a finite SSHC connected perpendicularly to a metallic lead on the spectral densities and conductance through the lead at low energies. We have followed an analytical approach valid for energies below the gap of the isolated SSHC including in particular accurately the localized modes at the ends of the SSHC. At higher energies, provided the bandwidth of the metallic lead is sufficiently large, hybridization between the extended SSHC states and the metallic-lead states at comparable energies modifies the results.

The zero-temperature conductance through the metallic lead exhibits sharp zeros when the chemical potential coincides with the energies of the split end modes. This feature could be exploited to determine these energies experimentally.

The spectral density of the end state farthest from the metallic lead exhibits a narrow peak centered at zero energy, with a relatively flat maximum extending over an energy range of the order of half the splitting. In contrast, the spectral density of the end state coupled to the lead is broader and vanishes at zero energy.

References

  1. [1]S. Ryu, A. P. Schnyder, A. Furusaki, and A. W. W. Ludwig, Topological insulators and superconductors: tenfold way and dimensional hierarchy, New Journal of Physics 12, 065010 (2010).DOI
  2. [2]X.-L. Qi and S.-C. Zhang, Topological insulators and superconductors, Rev. Mod. Phys. 83, 1057 (2011).DOI
  3. [3]Y. Ando, Topological insulator materials, Journal of the Physical Society of Japan 82, 102001 (2013).DOI
  4. [4]J. K. Asbóth, L. Oroszlány, and A. Pályi, A Short Course on Topological Insulators, Lecture Notes in Physics, Vol. 919 (Springer International Publishing, 2016).
  5. [5]C. Li, R. Wang, Q. Gong, and Y. Li, Manipulating winding numbers and multiple topological bound states via long-range coupling in chiral photonic lattices, Phys. Rev. Res. 6, 043087 (2024).DOI
  6. [6]P. St-Jean, V. Goblot, E. Galopin, A. Lemaître, T. Ozawa, L. Le Gratiet, I. Sagnes, J. Bloch, and A. Amo, Lasing in topological edge states of a one-dimensional lattice, Nature Photonics 11, 651 (2017).
  7. [7]P. Gagel, O. A. Egorov, F. Dzimira, J. Beierlein, M. Emmerling, A. Wolf, F. Jabeen, S. Betzold, U. Peschel, S. Höfling, C. Schneider, and S. Klembt, An electrically pumped topological polariton laser, Nano Letters 24, 6538 (2024).arxiv.org/abs/2402.00639
  8. [8]Z. Yu, F. Jin, J. Ren, S. Mandal, B. Zhang, and R. Su, Topological exciton-polaritons with negative coupling, Nature Communications 17, 1269 (2025).DOI
  9. [9]W. P. Su, J. R. Schrieffer, and A. J. Heeger, Solitons in polyacetylene, Phys. Rev. Lett. 42, 1698 (1979).DOI
  10. [10]R. G. Dias and A. M. Marques, Long-range hopping and indexing assumption in one-dimensional topological insulators, Phys. Rev. B 105, 035102 (2022).DOI
  11. [11]A. A. Aligia, Topological invariants based on generalized position operators and application to the interacting Rice-Mele model, Phys. Rev. B 107, 075153 (2023).
  12. [12]A. Ghosh, A. M. Martin, and S. Majumder, Quench dynamics of edge states in a finite extended Su-Schrieffer-Heeger system, Phys. Rev. E 108, 034108 (2023).DOI
  13. [13]R. K. Malakar and A. K. Ghosh, Engineering topological phases of any winding and Chern numbers in extended Su-Schrieffer-Heeger models, Journal of Physics: Condensed Matter 35, 335401 (2023).DOI
  14. [14]M. Ghuneim and R. Weda Boman-tara, Topological phases of tight-binding trimer lattice in the BDI symmetry class, Journal of Physics: Condensed Matter 36, 495402 (2024).DOI
  15. [15]D. Joshi and T. Nag, Adiabatic charge transport in extended Su-Schrieffer-Heeger models, Phys. Rev. B 112, 075411 (2025).DOI
  16. [16]P.-J. Chang, J. Pi, M. Zheng, Y.-T. Lei, X. Pan, D. Ruan, and G.-L. Long, Topological phases of extended Su-Schrieffer-Heeger-Hubbard model, Scientific Reports 15, 18023 (2025).DOI
  17. [17]A. A. Aligia, A. M. Lobos, L. Peralta Gavensky, and C. J. Gazza, Probing boundary spins in the Su-Schrieffer-Heeger-Hubbard model, Phys. Rev. B 112, 235156 (2025).
  18. [18]X. He, A. Shi, J. Liu, and J. Gong, Quantum metrology via adiabatic control of topological edge states, Phys. Rev. A 114, 032205 (2026).DOI
  19. [19]P. A. Grizzi, A. A. Aligia, and P. Roura-Bas, Exact analytical edge states in the extended Su-Schrieffer-Heeger model, Phys. Rev. B 114, 235404 (2026).DOI
  20. [20]A. Ritz-Zwilling and O. Dmytruk, Fermionic many-body topology in cavity-coupled Su-Schrieffer-Heeger chains, Phys. Rev. B 114, 165423 (2026).
  21. [21]P. A. Grizzi, P. Roura-Bas, and A. A. Aligia, Topological phase diagram of an extended Su-Schrieffer-Heeger model (2026), arXiv:2611.18456 [cond-mat.mes-hall].arxiv.org/abs/2611.18456
  22. [22]C. Zhao, G. Catarina, J.-J. Zhang, J. C. G. Henriques, L. Yang, J. Ma, X. Feng, O. Gröning, P. Ruffieux, J. Fernández-Rossier, and R. Fasel, Tunable topological phases in nanopgraphene-based spin-1/2 alternating-exchange Heisenberg chains, Nature Nanotechnology 19, 1789 (2024).
  23. [23]N. Wagner, L. Crippa, A. Amaricci, P. Hansmann, M. Klett, E. J. König, T. Schäfer, D. D. Sante, J. Cano, A. J. Millis, A. Georges, and G. Sangiovanni, Mott insulators with boundary zeros, Nature Communications 14, 7531 (2023).DOI
  24. [24]A. Bissonnette, N. Delnour, A. Mckenna, H. Eleuch, M. Hilke, and R. MacKenzie, Boundary-induced topological transition in an open Su-Schrieffer-Heeger model, Phys. Rev. B 109, 075106 (2024).DOI
  25. [25]D. Martínez, R. P. A. Lima, A. López, F. Domínguez-Adame, and P. A. Orellana, Topologically protected bound states in the continuum in the Su-Schrieffer-Heeger chain, Phys. Rev. Res. 8, 043118 (2026).DOI
  26. [26]Y. Meir and N. S. Wingreen, Landauer formula for the current through an interacting electron region, Phys. Rev. Lett. 68, 2512 (1992).DOI
  27. [27]A. A. Aligia and C. R. Proetto, Kondo and anti-Kondo resonances in transport through nanoscale devices, Phys. Rev. B 65, 165305 (2002).DOI

Paper details

vlnc:W_yn4azdzj
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC0-1.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents