Edge states and topological invariants of non-Hermitian systems

Edge states and topological invariants of non-Hermitian systems

Shunyu Yao Institute for Advanced Study, Tsinghua University, Beijing, 100084, China    Zhong Wang wangzhongemail@gmail.com Institute for Advanced Study, Tsinghua University, Beijing, 100084, China Collaborative Innovation Center of Quantum Matter, Beijing, 100871, China

For Hermitian systems, the creation or annihilation of topological edge modes is accompanied by the gap closing of Bloch Hamiltonian; for non-Hermitian systems, however, the edge-state transition points can differ from the gap closing points of Bloch Hamiltonian, which indicates breakdown of the usual bulk-boundary correspondence. We study this intriguing phenomenon via exactly solving a prototype model, namely the one-dimensional non-Hermitian Su-Schrieffer-Heeger model. The solution shows that the usual Bloch waves give way to eigenstates localized at the ends of an open chain, and the Bloch Hamiltonian is not the appropriate bulk side of bulk-boundary correspondence. It is shown that the standard Brillouin zone (a unit circle for one-dimensional systems) is replaced by a deformed one (a non-unit circle for the solved model), in which topological invariants can be precisely defined, embodying an unconventional bulk-boundary correspondence. This topological invariant correctly predicts the edge-state transition points and the number of topological edge modes. The theory is of general interest to topological aspects of non-Hermitian systems.

Introduction.–Topological materials are characterized by robust boundary states immune to perturbationsHasan and Kane (2010); Qi and Zhang (2011); Chiu et al. (2016); Bernevig and Hughes (2013); Bansil et al. (2016). According to the principle of bulk-boundary correspondence, the existence of boundary states is dictated by the bulk topological invariants, which, in the band-theory framework, are defined in terms of the Bloch Hamiltonian. The Hamiltonian is often assumed to be Hermitian. In many physical systems, however, non-Hermitian HamiltoniansBender and Boettcher (1998); Bender (2007) are more appropriate. For example, they are widely used in describing open systemsRotter (2009); Malzard et al. (2015); Carmichael (1993); Zhen et al. (2015); Diehl et al. (2011); Cao and Wiersig (2015); Choi et al. (2010); San-Jose et al. (2016); Lee and Chan (2014); Lee et al. (2014), wave systems with gain and lossMakris et al. (2008); Longhi (2009); Klaiman et al. (2008); Regensburger et al. (2012); Bittner et al. (2012); Rüter et al. (2010); Lin et al. (2011); Feng et al. (2013); Guo et al. (2009); Liertzer et al. (2012); Peng et al. (2014); Fleury et al. (2015); Chang et al. (2014); Hodaei et al. (2017, 2014); Feng et al. (2014); Gao et al. (2015); Xu et al. (2016); Ashida et al. (2017); Kawabata et al. (2017); Chen et al. (2017); Ding et al. (2016) (e.g. photonic and acoustic Ozawa et al. (2018); Lu et al. (2014); El-Ganainy et al. (2018); Longhi (2018)), and solid-state systems where electron-electron interactions or disorders introduce a non-Hermitian self energy into the effective Hamiltonian of quasiparticleKozii and Fu (2017); Papaj et al. (2018); Shen and Fu (2018). With these physical motivations, there have recently been growing efforts, both theoreticallyShen et al. (2017); Rudner and Levitov (2009); Esaki et al. (2011); Leykam et al. (2017); Lee (2016); Hu and Hughes (2011); Gong et al. (2017); Gong and Wang (2010); Rudner et al. (2016); Liang and Huang (2013); Kawabata et al. (2018); Ni et al. (2018); Zyuzin and Zyuzin (2018); Cerjan et al. (2018); Zhou et al. (2017a); González and Molina (2017); Klett et al. (2017, 2018); Menke and Hirschmann (2017); Yuce (2016, 2015); Xu et al. (2017); Hu et al. (2017); Li et al. (2017); Wang et al. (2015); Ke et al. (2017); Martinez Alvarez et al. (2017); Rivolta et al. (2017); Gong et al. (2018) and experimentallyZeuner et al. (2015); Zhan et al. (2017); Xiao et al. (2017); Weimann et al. (2017); Parto et al. (2017); Zhou et al. (2017b), to investigate topological phenomena of non-Hermitian Hamiltonians.

Among the key issues is the fate of bulk-boundary correspondence in non-Hermitian systems. Recently, numerical results indicate that the usual bulk-boundary correspondence may break downLee (2016); Xiong (2017). In particular, we observe a puzzling phenomenon in the data of Ref.Lee (2016) that the transition point, which divides phases with and without topological edge modes, does not seem to coincide with jump of any bulk topological invariant. Instead, the spectrum gap of Bloch Hamiltonian does not close at the transition point. The questions we address below are: How to determine the transition point? Is there a generalized bulk-boundary correspondence? What is the topological invariant responsible for the topological edge states?

We start from solving a one-dimensional(1D) lattice model with non-Hermiticity. Interestingly, the analytic solution shows that all the bulk eigenstates of an open chain become localized near the boundary (“non-Hermtian skin effect”), in sharp contrast to the extended Bloch waves in Hermitian cases. We will show that such a “non-Hermitian skin effect” is closely related to the breakdown of the usual bulk-boundary correspondence.

In essence, previous topological invariantsRudner and Levitov (2009); Esaki et al. (2011); Leykam et al. (2017); Shen et al. (2017); Lieu (2018) are formulated in terms of the Bloch Hamiltonian defined in the standard Brillouin zone, which, for 1D systems, is the unit circle parameterized by () on the complex plane. As we will explain, it is replaced by a “deformed Brillouin zone”, which is found to be a non-unit circle in the present model [Fig.3(b)]. In view of this non-Bloch-wave nature of bulk states, we introduce an unconventional topological invariant defined in the deformed Brillouin zone, which faithfully determines the topological edge modes. It embodies the unusual bulk-boundary correspondence of non-Hermitian systems.

Model.–The lattice model is pictorially shown in Fig.1. It is one of the non-Hermitian versions of Su-Schrieffer-Heeger(SSH) modelSu et al. (1980)111Related models has been studied for various purposes in Ref.Zhu et al. (2014); Yin et al. (2018); Lieu (2018)., which are relevant to quite a few experimentsWeimann et al. (2017); Zeuner et al. (2015); Poli et al. (2015). The Bloch Hamiltonian is222Compared to Ref. Lee (2016), a basis change is taken, bringing the physical interpretation closer to SSH.


where , , and are the Pauli matrices. The model has a chiral symmetryChiu et al. (2016) , which ensures that the eigenvalues appear in pairs: . The energy gap closes at the exceptional points , which requires () or ().

Figure 1: A non-Hermitian SSH model. Each unit cell contains two sites with asymmetric intra-cell hopping .

As found in Ref.Lee (2016), the open-boundary spectrum is fundamentally different from that of periodic boundary. We solve the real-space Hamiltonian for an open chain, taking as a parameter [Fig.2]. Zero modes exist for an interval of , and they are robust to perturbation [Fig.2(d)], which indicates their topological origin. A zero-mode transition point is located at (, as shown below), which is a unremarkable point viewed from whose spectrum is gapped there (). Although not explicitly mentioned333It was presumably due to insufficient numerical precision of Ref.Lee (2016). According to our analytic and improved numerical calculations, the zero-mode line in their Fig. 3(a) should span the entire interval, instead of two disconnected ones presented there. , this phenomenon are already appreciable in the spectrum in Ref. Lee (2016).

Figure 2: Energy spectrum of of an open chain with length (unit cell). , . (a) as functions of . The zero-mode line is colored red (twofold degenerate, ignoring an indiscernible split). The transition point () and the gap-closing points of () are indicated by arrows. (b,c) The real and imaginary parts of . (d) The same as (a) except that the value of at the leftmost bond is replaced by , which generates additional nonzero modes, but the zero-mode line is unaffected.

To gain insights, let us analytically solve an open chain. The wavefunction is written as . The real-space eigen-equation leads to and in the bulk of chain. We take the ansatz that , where each takes the exponential form (omitting the index temporarily)


which satisfies


Therefore, we have


which has two solutions, namely , where corresponds to . In the limit, we have


They can also be seen from Eq.(3). These two solutions correspond to and , respectively.

Restoring the index in , we have


which are equivalent because of Eq.(4). The general solution is written as a linear combination:


which should satisfy the boundary condition


Together with Eq.(6), they lead to


in which we have defined . We are concerned about the spectrum for a long chain, which necessitates for the bulk states. If not, suppose that , we would be able to discard the tiny term in Eq.(9), and the equation becomes essentially independent of . Eq.(4) tells us that , which leads to (via )


for bulk states. As an illustration, we plot the - curve for in Fig.3(a). As seen from Fig.2, the spectrum is real for this set of parameter, therefore, no imaginary part of is needed (the reality of spectrum in certain parameter intervals is due to the PT symmetryBender and Boettcher (1998); Bender (2007)). There is indeed an energy interval with (line in Fig.3(a))). As is increased from , moves towards left, and finally hits the axis, which apparently satisfies . Inserting Eq.(5) into this equation, we have


where the first two solutions are relevant for the parameters used in this paper (with ). At these locations of , the bulk spectrum touches zero energy, and transitions occur. This explains the numerical puzzle in Fig.2.

Figure 3: (a) versus from Eq.(4). , (dark color) and (light color). (b) curve on complex plane, denoted as , which is a circle with radius .

We emphasize that indicates that all the bulk eigenstates are localized at the left end of the chain. This phenomenon is dubbed the “non-Hermitian skin effect” (see Fig.4 for illustration). To see the bulk spectrum, let us parameterize the circle by (), then Eq.(4) leads to


which apparently recovers the spectrum of SSH model when .

Before proceeding, we comment on a standard method of finding zero modes, which has often been used in literature of non-Hermitian systems. Let us focus on our model for concreteness. Consider a semi-infinite system with a left end. One can see that appears as a zero-energy eigenstate. The normalizable condition is then imposed in the standard approach. Consequently, the transition points satisfy . However, it would predict , instead of , as a transition point. In contrast, our solution of SSH model shows that the zero mode merges into the bulk spectrum at


which correctly produces . This represents an unconventional bulk-boundary correspondence.

Figure 4: The wavefunction profiles of a zero mode (main figure) and eight randomly chosen bulk states (inset), illustrating the “non-Hermitian skin effect” found in the analytic solution, namely, all the bulk eigenstates are localized near the boundary. .

Shortcut.–Now we introduce a shortcut solution from hindsight (The longer solution is presented first because it is easier to generalize, e.g., to include longer-range hoppings). Let us take a similarity transformation of the real-space Hamiltonian: and


where is a diagonal matrix whose elements are , with given by Eq.(10). Now becomes a Hermitian SSH model for . In space:


with . The transition points via are identical to Eq.(11). Apparently, the case can also be solved this way, though is no longer Hermitian.

Non-Bloch topological invariant.–The curve reflects the non-Bloch nature of the bulk states of open chains. Notably, it remains a closed loop and can be intuitively viewed as a deformation of the standard Brillouin zone (unit circle). Departing from the usual framework, we define a topological invariant on as follows. First, a “non-Bloch” Hamiltonian is obtained from by the replacement :


where . One then find the right and left eigenvectors by


Chiral symmetry ensures that and are also right(left) eigenvectors, with eigenvalues and . To obtain these eigenvectors, one can diagonalize the matrix as , then each column of and is a right and left eigenvector, respectively. As a generalization of the usual “ matrix”Chiu et al. (2016), we define


which, due to the chiral symmetry , is off-diagonal: . Now we introduce a winding number


which differs essentially from the Hermitian cases in that is a general curve [e.g. Fig.3(b)] instead of the unit circle. It is useful to mention that the conventional formulations using may sometimes produce correct topological numbers, when happens to be a unit circle, even if the model is non-Hermitian444For example, it is the case for the model numerically studied in Ref. Lieu (2018)..

Figure 5: Numerical result of topological invariant. is the number of grid point on . .

The numerical results for our model in Eq.(1) is shown in Fig.5, which is consistent with both the numerical and analytical spectrum obtained above. Quantitatively, counts the total number of robust zero modes at the left and right ends. For example, corresponding to Fig.2, there are two zero modes for , and none elsewhere. The analytic solution shows that, for , both modes live at the left end; for , one for each end; and for , both at the right end. Thus, the -gap closing points (called “half vortices” in Ref.Leykam et al. (2017)) are where zero modes migrate from one end to the other, conserving the total mode number. In fact, one can see at , indicating penetration into the bulk.

To provide a more generic example, we take , where introduces hopping between a pair separated by another pair. Now the curve is no longer a circle, yet correctly predicts the total zero-mode number [Fig.6].

Finally, we remarked that Eq.(19) can be generalized to multi-band systems. Each pair of bands (labeled by ) correspond to a curve, and the matrix [Eq.(18)] becomes , each one defining a winding number . The topological invariant is .

Figure 6: (a) Upper: Spectrum for the modified model; . Lower: topological invariant calculated using 200 grid points on . . (b) for .

Conclusion.–Through the analytic solution of non-Hermitian SSH model (which itself will be quite useful as a benchmark for theories), we explained why the usual bulk-boundary correspondence breaks down, and how to generalize it to non-Hermitian systems in a precise manner. The bulk states lose the Bloch-wave nature and become localized near the boundary, which necessitates an unconventional correspondence not manifested in . We formulate this generalized correspondence by introducing a precise topological invariant. The physics presented is quite general and closely relevant to a vast variety of non-Hermitian systems, which will be left for future investigations.

Acknowledgements.–This work is supported by NSFC under Grant No. 11674189.


  • Hasan and Kane (2010) M. Z. Hasan and C. L. Kane, “Colloquium : Topological insulators,” Rev. Mod. Phys. 82, 3045–3067 (2010).
  • Qi and Zhang (2011) Xiao-Liang Qi and Shou-Cheng Zhang, “Topological insulators and superconductors,” Rev. Mod. Phys. 83, 1057–1110 (2011).
  • Chiu et al. (2016) Ching-Kai Chiu, Jeffrey C. Y. Teo, Andreas P. Schnyder,  and Shinsei Ryu, “Classification of topological quantum matter with symmetries,” Rev. Mod. Phys. 88, 035005 (2016).
  • Bernevig and Hughes (2013) B Andrei Bernevig and Taylor L Hughes, Topological insulators and topological superconductors (Princeton University Press, Princeton, NJ, 2013).
  • Bansil et al. (2016) A. Bansil, Hsin Lin,  and Tanmoy Das, ‘‘Colloquium : Topological band theory,” Rev. Mod. Phys. 88, 021004 (2016).
  • Bender and Boettcher (1998) Carl M Bender and Stefan Boettcher, “Real spectra in non-hermitian hamiltonians having p t symmetry,” Physical Review Letters 80, 5243 (1998).
  • Bender (2007) Carl M Bender, “Making sense of non-hermitian hamiltonians,” Reports on Progress in Physics 70, 947 (2007).
  • Rotter (2009) Ingrid Rotter, “A non-hermitian hamilton operator and the physics of open quantum systems,” Journal of Physics A: Mathematical and Theoretical 42, 153001 (2009).
  • Malzard et al. (2015) Simon Malzard, Charles Poli,  and Henning Schomerus, “Topologically protected defect states in open photonic systems with non-hermitian charge-conjugation and parity-time symmetry,” Phys. Rev. Lett. 115, 200402 (2015).
  • Carmichael (1993) H. J. Carmichael, “Quantum trajectory theory for cascaded open systems,” Phys. Rev. Lett. 70, 2273–2276 (1993).
  • Zhen et al. (2015) Bo Zhen, Chia Wei Hsu, Yuichi Igarashi, Ling Lu, Ido Kaminer, Adi Pick, Song-Liang Chua, John D Joannopoulos,  and Marin Soljačić, “Spawning rings of exceptional points out of dirac cones,” Nature 525, 354 (2015).
  • Diehl et al. (2011) Sebastian Diehl, Enrique Rico, Mikhail A Baranov,  and Peter Zoller, “Topology by dissipation in atomic quantum wires,” Nature Physics 7, 971–977 (2011).
  • Cao and Wiersig (2015) Hui Cao and Jan Wiersig, “Dielectric microcavities: Model systems for wave chaos and non-hermitian physics,” Rev. Mod. Phys. 87, 61–111 (2015).
  • Choi et al. (2010) Youngwoon Choi, Sungsam Kang, Sooin Lim, Wookrae Kim, Jung-Ryul Kim, Jai-Hyung Lee,  and Kyungwon An, “Quasieigenstate coalescence in an atom-cavity quantum composite,” Phys. Rev. Lett. 104, 153601 (2010).
  • San-Jose et al. (2016) Pablo San-Jose, Jorge Cayao, Elsa Prada,  and Ramón Aguado, “Majorana bound states from exceptional points in non-topological superconductors,” Scientific reports 6, 21427 (2016).
  • Lee and Chan (2014) Tony E. Lee and Ching-Kit Chan, “Heralded magnetism in non-hermitian atomic systems,” Phys. Rev. X 4, 041001 (2014).
  • Lee et al. (2014) Tony E. Lee, Florentin Reiter,  and Nimrod Moiseyev, “Entanglement and spin squeezing in non-hermitian phase transitions,” Phys. Rev. Lett. 113, 250401 (2014).
  • Makris et al. (2008) K. G. Makris, R. El-Ganainy, D. N. Christodoulides,  and Z. H. Musslimani, “Beam dynamics in symmetric optical lattices,” Phys. Rev. Lett. 100, 103904 (2008).
  • Longhi (2009) S. Longhi, “Bloch oscillations in complex crystals with symmetry,” Phys. Rev. Lett. 103, 123601 (2009).
  • Klaiman et al. (2008) Shachar Klaiman, Uwe Günther,  and Nimrod Moiseyev, “Visualization of branch points in -symmetric waveguides,” Phys. Rev. Lett. 101, 080402 (2008).
  • Regensburger et al. (2012) Alois Regensburger, Christoph Bersch, Mohammad-Ali Miri, Georgy Onishchukov, Demetrios N Christodoulides,  and Ulf Peschel, “Parity–time synthetic photonic lattices,” Nature 488, 167 (2012).
  • Bittner et al. (2012) S. Bittner, B. Dietz, U. Günther, H. L. Harney, M. Miski-Oglu, A. Richter,  and F. Schäfer, “Pt-symmetry and spontaneous symmetry breaking in a microwave billiard,” Phys. Rev. Lett. 108, 024101 (2012).
  • Rüter et al. (2010) Christian E Rüter, Konstantinos G Makris, Ramy El-Ganainy, Demetrios N Christodoulides, Mordechai Segev,  and Detlef Kip, “Observation of parity–time symmetry in optics,” Nature physics 6, 192 (2010).
  • Lin et al. (2011) Zin Lin, Hamidreza Ramezani, Toni Eichelkraut, Tsampikos Kottos, Hui Cao,  and Demetrios N. Christodoulides, “Unidirectional invisibility induced by -symmetric periodic structures,” Phys. Rev. Lett. 106, 213901 (2011).
  • Feng et al. (2013) Liang Feng, Ye-Long Xu, William S Fegadolli, Ming-Hui Lu, José EB Oliveira, Vilson R Almeida, Yan-Feng Chen,  and Axel Scherer, “Experimental demonstration of a unidirectional reflectionless parity-time metamaterial at optical frequencies,” Nature materials 12, 108 (2013).
  • Guo et al. (2009) A. Guo, G. J. Salamo, D. Duchesne, R. Morandotti, M. Volatier-Ravat, V. Aimez, G. A. Siviloglou,  and D. N. Christodoulides, “Observation of -symmetry breaking in complex optical potentials,” Phys. Rev. Lett. 103, 093902 (2009).
  • Liertzer et al. (2012) M. Liertzer, Li Ge, A. Cerjan, A. D. Stone, H. E. Türeci,  and S. Rotter, “Pump-induced exceptional points in lasers,” Phys. Rev. Lett. 108, 173901 (2012).
  • Peng et al. (2014) B Peng, ŞK Özdemir, S Rotter, H Yilmaz, M Liertzer, F Monifi, CM Bender, F Nori,  and L Yang, “Loss-induced suppression and revival of lasing,” Science 346, 328–332 (2014).
  • Fleury et al. (2015) Romain Fleury, Dimitrios Sounas,  and Andrea Alù, “An invisible acoustic sensor based on parity-time symmetry,” Nature communications 6, 5905 (2015).
  • Chang et al. (2014) Long Chang, Xiaoshun Jiang, Shiyue Hua, Chao Yang, Jianming Wen, Liang Jiang, Guanyu Li, Guanzhong Wang,  and Min Xiao, “Parity–time symmetry and variable optical isolation in active–passive-coupled microresonators,” Nature photonics 8, 524 (2014).
  • Hodaei et al. (2017) Hossein Hodaei, Absar U Hassan, Steffen Wittek, Hipolito Garcia-Gracia, Ramy El-Ganainy, Demetrios N Christodoulides,  and Mercedeh Khajavikhan, “Enhanced sensitivity at higher-order exceptional points,” Nature 548, 187 (2017).
  • Hodaei et al. (2014) Hossein Hodaei, Mohammad-Ali Miri, Matthias Heinrich, Demetrios N Christodoulides,  and Mercedeh Khajavikhan, “Parity-time–symmetric microring lasers,” Science 346, 975–978 (2014).
  • Feng et al. (2014) Liang Feng, Zi Jing Wong, Ren-Min Ma, Yuan Wang,  and Xiang Zhang, “Single-mode laser by parity-time symmetry breaking,” Science 346, 972–975 (2014).
  • Gao et al. (2015) Tiejun Gao, E Estrecho, KY Bliokh, TCH Liew, MD Fraser, Sebastian Brodbeck, Martin Kamp, Christian Schneider, Sven Höfling, Y Yamamoto, et al., ‘‘Observation of non-hermitian degeneracies in a chaotic exciton-polariton billiard,” Nature 526, 554 (2015).
  • Xu et al. (2016) Haitan Xu, David Mason, Luyao Jiang,  and JGE Harris, “Topological energy transfer in an optomechanical system with exceptional points,” Nature 537, 80 (2016).
  • Ashida et al. (2017) Yuto Ashida, Shunsuke Furukawa,  and Masahito Ueda, “Parity-time-symmetric quantum critical phenomena,” Nature communications 8, 15791 (2017).
  • Kawabata et al. (2017) Kohei Kawabata, Yuto Ashida,  and Masahito Ueda, “Information retrieval and criticality in parity-time-symmetric systems,” Phys. Rev. Lett. 119, 190401 (2017).
  • Chen et al. (2017) Weijian Chen, Şahin Kaya Özdemir, Guangming Zhao, Jan Wiersig,  and Lan Yang, “Exceptional points enhance sensing in an optical microcavity,” Nature 548, 192 (2017).
  • Ding et al. (2016) Kun Ding, Guancong Ma, Meng Xiao, Z. Q. Zhang,  and C. T. Chan, “Emergence, coalescence, and topological properties of multiple exceptional points and their experimental realization,” Phys. Rev. X 6, 021007 (2016).
  • Ozawa et al. (2018) T. Ozawa, H. M. Price, A. Amo, N. Goldman, M. Hafezi, L. Lu, M. Rechtsman, D. Schuster, J. Simon, O. Zilberberg,  and I. Carusotto, “Topological Photonics,” ArXiv e-prints  (2018), arXiv:1802.04173 [physics.optics] .
  • Lu et al. (2014) Ling Lu, John D. Joannopoulos,  and Marin Soljacic, “Topological photonics,” Nat Photon 8, 821–829 (2014).
  • El-Ganainy et al. (2018) Ramy El-Ganainy, Konstantinos G Makris, Mercedeh Khajavikhan, Ziad H Musslimani, Stefan Rotter,  and Demetrios N Christodoulides, “Non-hermitian physics and pt symmetry,” Nature Physics 14, 11 (2018).
  • Longhi (2018) S. Longhi, “Parity-Time Symmetry meets Photonics: A New Twist in non-Hermitian Optics,” ArXiv e-prints  (2018), arXiv:1802.05025 [physics.optics] .
  • Kozii and Fu (2017) V. Kozii and L. Fu, “Non-Hermitian Topological Theory of Finite-Lifetime Quasiparticles: Prediction of Bulk Fermi Arc Due to Exceptional Point,” ArXiv e-prints  (2017), arXiv:1708.05841 [cond-mat.mes-hall] .
  • Papaj et al. (2018) M. Papaj, H. Isobe,  and L. Fu, “Bulk Fermi arc of disordered Dirac fermions in two dimensions,” ArXiv e-prints  (2018), arXiv:1802.00443 [cond-mat.dis-nn] .
  • Shen and Fu (2018) H. Shen and L. Fu, “Quantum Oscillation from In-gap States and non-Hermitian Landau Level Problem,” ArXiv e-prints  (2018), arXiv:1802.03023 [cond-mat.str-el] .
  • Shen et al. (2017) H. Shen, B. Zhen,  and L. Fu, “Topological Band Theory for Non-Hermitian Hamiltonians,” ArXiv e-prints  (2017), arXiv:1706.07435 [cond-mat.mes-hall] .
  • Rudner and Levitov (2009) M. S. Rudner and L. S. Levitov, “Topological transition in a non-hermitian quantum walk,” Phys. Rev. Lett. 102, 065703 (2009).
  • Esaki et al. (2011) Kenta Esaki, Masatoshi Sato, Kazuki Hasebe,  and Mahito Kohmoto, “Edge states and topological phases in non-hermitian systems,” Phys. Rev. B 84, 205128 (2011).
  • Leykam et al. (2017) Daniel Leykam, Konstantin Y. Bliokh, Chunli Huang, Y. D. Chong,  and Franco Nori, ‘‘Edge modes, degeneracies, and topological numbers in non-hermitian systems,” Phys. Rev. Lett. 118, 040401 (2017).
  • Lee (2016) Tony E. Lee, “Anomalous edge state in a non-hermitian lattice,” Phys. Rev. Lett. 116, 133903 (2016).
  • Hu and Hughes (2011) Yi Chen Hu and Taylor L. Hughes, “Absence of topological insulator phases in non-hermitian -symmetric hamiltonians,” Phys. Rev. B 84, 153101 (2011).
  • Gong et al. (2017) Zongping Gong, Sho Higashikawa,  and Masahito Ueda, “Zeno hall effect,” Phys. Rev. Lett. 118, 200401 (2017).
  • Gong and Wang (2010) Jiangbin Gong and Qing-hai Wang, “Geometric phase in -symmetric quantum mechanics,” Phys. Rev. A 82, 012103 (2010).
  • Rudner et al. (2016) M. S. Rudner, M. Levin,  and L. S. Levitov, “Survival, decay, and topological protection in non-Hermitian quantum transport,” ArXiv e-prints  (2016), arXiv:1605.07652 [cond-mat.mes-hall] .
  • Liang and Huang (2013) Shi-Dong Liang and Guang-Yao Huang, “Topological invariance and global berry phase in non-hermitian systems,” Phys. Rev. A 87, 012118 (2013).
  • Kawabata et al. (2018) Kohei Kawabata, Yuto Ashida, Hosho Katsura,  and Masahito Ueda, “Parity-time-symmetric topological superconductor,” arXiv preprint arXiv:1801.00499  (2018).
  • Ni et al. (2018) X. Ni, D. Smirnova, A. Poddubny, D. Leykam, Y. Chong,  and A. B. Khanikaev, “Exceptional points in topological edge spectrum of PT symmetric domain walls,” ArXiv e-prints  (2018), arXiv:1801.04689 [cond-mat.mes-hall] .
  • Zyuzin and Zyuzin (2018) A. A. Zyuzin and A. Yu. Zyuzin, “Flat band in disorder-driven non-hermitian weyl semimetals,” Phys. Rev. B 97, 041203 (2018).
  • Cerjan et al. (2018) Alexander Cerjan, Meng Xiao, Luqi Yuan,  and Shanhui Fan, “Effects of non-hermitian perturbations on weyl hamiltonians with arbitrary topological charges,” Phys. Rev. B 97, 075128 (2018).
  • Zhou et al. (2017a) L. Zhou, Q.-h. Wang, H. Wang,  and J. Gong, ‘‘Dynamical quantum phase transitions in non-Hermitian lattices,” ArXiv e-prints  (2017a), arXiv:1711.10741 [cond-mat.stat-mech] .
  • González and Molina (2017) J. González and R. A. Molina, “Topological protection from exceptional points in weyl and nodal-line semimetals,” Phys. Rev. B 96, 045437 (2017).
  • Klett et al. (2017) Marcel Klett, Holger Cartarius, Dennis Dast, Jörg Main,  and Günter Wunner, ‘‘Relation between -symmetry breaking and topologically nontrivial phases in the su-schrieffer-heeger and kitaev models,” Phys. Rev. A 95, 053626 (2017).
  • Klett et al. (2018) M. Klett, H. Cartarius, D. Dast, J. Main,  and G. Wunner, “Topological edge states in the Su-Schrieffer-Heeger model subject to balanced particle gain and loss,” ArXiv e-prints  (2018), arXiv:1802.06128 [quant-ph] .
  • Menke and Hirschmann (2017) Henri Menke and Moritz M. Hirschmann, “Topological quantum wires with balanced gain and loss,” Phys. Rev. B 95, 174506 (2017).
  • Yuce (2016) C. Yuce, “Majorana edge modes with gain and loss,” Phys. Rev. A 93, 062130 (2016).
  • Yuce (2015) Cem Yuce, “Topological phase in a non-hermitian pt symmetric system,” Physics Letters A 379, 1213–1218 (2015).
  • Xu et al. (2017) Yong Xu, Sheng-Tao Wang,  and L.-M. Duan, “Weyl exceptional rings in a three-dimensional dissipative cold atomic gas,” Phys. Rev. Lett. 118, 045701 (2017).
  • Hu et al. (2017) Wenchao Hu, Hailong Wang, Perry Ping Shum,  and Y. D. Chong, “Exceptional points in a non-hermitian topological pump,” Phys. Rev. B 95, 184306 (2017).
  • Li et al. (2017) C. Li, X. Z. Zhang, G. Zhang,  and Z. Song, “Topological phases in Kitaev chain with imbalanced pairing,” ArXiv e-prints  (2017), arXiv:1707.04718 [quant-ph] .
  • Wang et al. (2015) Xiaohui Wang, Tingting Liu, Ye Xiong,  and Peiqing Tong, “Spontaneous -symmetry breaking in non-hermitian kitaev and extended kitaev models,” Phys. Rev. A 92, 012116 (2015).
  • Ke et al. (2017) Shaolin Ke, Bing Wang, Hua Long, Kai Wang,  and Peixiang Lu, “Topological edge modes in non-hermitian plasmonic waveguide arrays,” Optics Express 25, 11132–11143 (2017).
  • Martinez Alvarez et al. (2017) V. M. Martinez Alvarez, J. E. Barrios Vargas,  and L. E. F. Foa Torres, “Non-Hermitian robust edge states in one-dimension: Anomalous localization and eigenspace condensation at exceptional points,” ArXiv e-prints  (2017), arXiv:1711.05235 [cond-mat.mes-hall] .
  • Rivolta et al. (2017) Nicolas X. A. Rivolta, Henri Benisty,  and Bjorn Maes, “Topological edge modes with symmetry in a quasiperiodic structure,” Phys. Rev. A 96, 023864 (2017).
  • Gong et al. (2018) Z. Gong, Y. Ashida, K. Kawabata, K. Takasan, S. Higashikawa,  and M. Ueda, “Topological phases of non-Hermitian systems,” ArXiv e-prints  (2018), arXiv:1802.07964 [cond-mat.mes-hall] .
  • Zeuner et al. (2015) Julia M. Zeuner, Mikael C. Rechtsman, Yonatan Plotnik, Yaakov Lumer, Stefan Nolte, Mark S. Rudner, Mordechai Segev,  and Alexander Szameit, “Observation of a topological transition in the bulk of a non-hermitian system,” Phys. Rev. Lett. 115, 040402 (2015).
  • Zhan et al. (2017) Xiang Zhan, Lei Xiao, Zhihao Bian, Kunkun Wang, Xingze Qiu, Barry C. Sanders, Wei Yi,  and Peng Xue, “Detecting topological invariants in nonunitary discrete-time quantum walks,” Phys. Rev. Lett. 119, 130501 (2017).
  • Xiao et al. (2017) L. Xiao, X. Zhan, Z. H. Bian, K. K. Wang, X. Zhang, X. P. Wang, J. Li, K. Mochizuki, D. Kim, N. Kawakami, W. Yi, H. Obuse, B. C. Sanders,  and P. Xue, “Observation of topological edge states in parity–time-symmetric quantum walks,” Nature Physics 13, 1117 (2017).
  • Weimann et al. (2017) S Weimann, M Kremer, Y Plotnik, Y Lumer, S Nolte, KG Makris, M Segev, MC Rechtsman,  and A Szameit, ‘‘Topologically protected bound states in photonic parity–time-symmetric crystals,” Nature materials 16, 433 (2017).
  • Parto et al. (2017) M. Parto, S. Wittek, H. Hodaei, G. Harari, M. A. Bandres, J. Ren, M. C. Rechtsman, M. Segev, D. N. Christodoulides,  and M. Khajavikhan, “Complex Edge-State Phase Transitions in 1D Topological Laser Arrays,” ArXiv e-prints  (2017), arXiv:1709.00523 [physics.optics] .
  • Zhou et al. (2017b) H. Zhou, C. Peng, Y. Yoon, C. W. Hsu, K. A. Nelson, L. Fu, J. D. Joannopoulos, M. Soljacic,  and B. Zhen, “Observation of Bulk Fermi Arc and Polarization Half Charge from Paired Exceptional Points,” ArXiv e-prints  (2017b), arXiv:1709.03044 [physics.optics] .
  • Xiong (2017) Y. Xiong, “Why does bulk boundary correspondence fail in some non-hermitian topological models,” ArXiv e-prints  (2017), arXiv:1705.06039v1 [cond-mat.mes-hall] .
  • Lieu (2018) Simon Lieu, “Topological phases in the non-hermitian su-schrieffer-heeger model,” Phys. Rev. B 97, 045106 (2018).
  • Su et al. (1980) W_P Su, JR Schrieffer,  and AJ Heeger, “Soliton excitations in polyacetylene,” Physical Review B 22, 2099 (1980).
  • (85) Related models has been studied for various purposes in Ref.Zhu et al. (2014); Yin et al. (2018); Lieu (2018).
  • Poli et al. (2015) Charles Poli, Matthieu Bellec, Ulrich Kuhl, Fabrice Mortessagne,  and Henning Schomerus, “Selective enhancement of topologically induced interface states in a dielectric resonator chain,” Nature communications 6, 6710 (2015).
  • (87) Compared to Ref. Lee (2016), a basis change is taken, bringing the physical interpretation closer to SSH.
  • (88) It was presumably due to insufficient numerical precision of Ref.Lee (2016). According to our analytic and improved numerical calculations, the zero-mode line in their Fig. 3(a) should span the entire interval, instead of two disconnected ones presented there.
  • (89) For example, it is the case for the model numerically studied in Ref. Lieu (2018).
  • Zhu et al. (2014) Baogang Zhu, Rong Lü,  and Shu Chen, “ symmetry in the non-hermitian su-schrieffer-heeger model with complex boundary potentials,” Phys. Rev. A 89, 062102 (2014).
  • Yin et al. (2018) C. Yin, H. Jiang, L. Li, R. Lü,  and S. Chen, ‘‘Geometrical meaning of winding number and its characterization of topological phases in one-dimensional chiral non-Hermitian systems,” ArXiv e-prints  (2018), arXiv:1802.04169v1 [cond-mat.mes-hall] .

Supplemental Material

This Supplemental Material contains two supplemental figures: Fig.7, Fig.8.

Figure 7: Left panels: Theoretical values of solved via [see the discussion below Eq.(9)]; Right panels: Numerical results of energy spectrum for open chains. Common parameters are .
Figure 8: (a) The modulus of energy for an open chain with length .(b) Numerical results of the topological invariant. . According to the analytical solution, in the regime , there are four transition points . The theory is consistent with the numerical results. The topological invariant correctly predicts the number of zero modes.
Comments 0
Request Comment
You are adding the first comment!
How to quickly get a good reply:
  • Give credit where it’s due by listing out the positive aspects of a paper before getting into which changes should be made.
  • Be specific in your critique, and provide supporting evidence with appropriate references to substantiate general statements.
  • Your comment should inspire ideas to flow and help the author improves the paper.

The better we are at sharing our knowledge with each other, the faster we move forward.
The feedback must be of minimum 40 characters and the title a minimum of 5 characters
Add comment
Loading ...
This is a comment super asjknd jkasnjk adsnkj
The feedback must be of minumum 40 characters
The feedback must be of minumum 40 characters

You are asking your first question!
How to quickly get a good answer:
  • Keep your question short and to the point
  • Check for grammar or spelling errors.
  • Phrase it like a question
Test description