Pith. sign in

REVIEW 3 major objections 5 minor 48 references

Non-Bloch Dirac Points and Phase Diagram in the Stacked Non-Hermitian SSH Model

T0 review · 3 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A stacked non-Hermitian dimerized-chain model hosts linear band crossings, called non-Bloch Dirac points, that have real energy and integer topological charge, exist only under open boundaries, and move when the boundary shape changes.

desk verdict Solid analytic study with a new OBC-only Dirac point result; fix the Eq. (1)/(2) mismatch and justify the cylinder reduction before publication. read the letter →

arxiv 2412.02782 v1 pith:6Q2VHEGS submitted 2024-12-03 cond-mat.mes-hall cond-mat.quant-gas

Megan Schoenzeit , Chang Shu , Kai Zhang , Kai Sun This is my paper · ORCID
classification cond-mat.mes-hallcond-mat.quant-gas
keywords non-Hermitianskineffectnon-BlochDiracpointsgeneralizedBrillouinzoneopen-boundaryspectrumbulk-boundarycorrespondencetopologicalsemimetalexceptionalstackedSSHmodel
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper claims that a two-dimensional non-Hermitian model made of stacked non-reciprocal dimerized chains hosts linear band crossings, called non-Bloch Dirac points, that exist only under open boundary conditions. These crossings have real energy, vanish from the periodic-boundary Bloch spectrum where they split into pairs of exceptional points, and shift their positions when the shape of the boundary changes. Because the spectrum is real in the vicinity of each crossing, the authors map the local problem to a Hermitian semimetal and show that each non-Bloch Dirac point carries an integer topological charge. If the claim is right, the familiar bulk-boundary correspondence of Hermitian semimetals has to be replaced, in non-Hermitian systems, by a geometry-dependent relation between the open-boundary spectrum and edge physics.

What carries the argument

The argument runs on the generalized Brillouin zone (GBZ): replacing the $x$-direction Bloch phase $e^{ik_x}$ by a complex number $\beta_x$ whose magnitude is fixed by the open-boundary standing-wave condition $|\beta_{x,+}|=|\beta_{x,-}|$. With the $y$ direction kept periodic (the cylinder geometry), the characteristic equation gives an analytic spectrum $E=\pm\sqrt{t^2+w^2-\gamma^2-\eta^2+2w\sin\phi\sqrt{t^2-\gamma^2}}$, and the GBZ radius $R(k_y)=\sqrt{|(t+\gamma)/(t-\gamma)|}$. A similarity transformation $P$ maps the Hamiltonian to a Hermitian matrix whose low-energy expansion is two anisotropic Dirac cones; the integer charge is $\nu_\pm=\frac{i}{2\pi}\oint (Q_\pm)^{-1}dQ_\pm$.

What would settle it

Diagonalize the same Hamiltonian numerically on a finite cluster with open boundaries in both directions (or on a parallelogram-shaped cluster) and compare the real gap-closing points with the cylinder prediction $(k_{x,c},\pm k_{y,c})$; if the crossings move, acquire complex energy, or disappear, the cylinder-geometry reduction and the geometry-dependence claim are falsified.

Watch

Extended reading notes

Core claim

Using an exactly solvable stacked dimerized-chain model with asymmetric intra-cell hopping and on-site gain/loss, the authors derive the full open-boundary phase diagram from the complex energy gaps. The phase diagram contains a transition on which the open-boundary spectrum closes its real gap at two points, with linear dispersion in the two-dimensional parameter space formed by the y-momentum and the phase of the non-Bloch wave vector. These non-Bloch Dirac points are protected by chiral or mirror symmetry, carry an integer ($\mathbb{Z}$) winding number, and remain stable against off-diagonal long-range hopping while a diagonal long-range hopping opens a gap. Under periodic boundaries the same crossings split into pairs of exceptional points, and the projected positions of the crossings for different open-boundary geometries do not coincide.

Load-bearing premise

The analytic phase diagram assumes that taking periodic boundary conditions in $y$ and open boundaries in $x$ gives the same spectrum as a fully open two-dimensional sample, justified only by the statement that the $y$-hopping is Hermitian; if that equivalence fails, the locations and even the existence of the non-Bloch Dirac points in a finite sample are not established.

Editorial extensions

If this is right

  • For these models, topological classification must start from the open-boundary (non-Bloch) spectrum rather than the periodic-boundary Bloch bands.
  • Non-Bloch Dirac points are symmetry-protected: perturbations that preserve chiral or mirror symmetry keep them gapless, and off-diagonal long-range hopping preserves them while diagonal long-range hopping gaps them.
  • Switching from open to periodic boundary conditions converts each non-Bloch Dirac point into a pair of exceptional points, so any bulk-boundary correspondence built on PBC bands will miss the edge physics.
  • Because the projected crossing locations depend on the orientation of the open edges (square vs parallelogram), the same bulk can appear to have different band-crossing momenta depending on the cut, a direct violation of the Hermitian projection doctrine.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The paper's cylinder-geometry reduction could be tested by adding non-Hermitian y-direction couplings; if PBC and OBC then disagree in y, the analytic phase boundaries and the non-Bloch Dirac point positions would need revision.
  • The same real-spectrum-plus-similarity-transformation construction suggests a route to defining non-Bloch Weyl points in three dimensions, with surface Fermi arcs whose endpoints are geometry-dependent.
  • The integer winding number computed around each non-Bloch Dirac point may be measurable in metamaterial or circuit realizations through the spatial profile of zero-energy edge modes, not just through the spectrum.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. The manuscript studies a two-dimensional stacked non-Hermitian SSH model with non-reciprocal hopping along x and Hermitian coupling along y. The authors derive an analytical spectrum for a cylinder geometry (OBC in x, PBC in y) using the generalized Brillouin zone, obtain a phase diagram in terms of complex-energy line gaps, and identify real-spectrum gap closings that they call non-Bloch Dirac points. They map the non-Hermitian Hamiltonian locally to a Hermitian Hamiltonian by a similarity transformation, assign an integer topological charge using chiral/mirror symmetry, and argue that the non-Bloch Dirac points disappear under PBC and have geometry-dependent locations. The analytical derivation from Eq. (7) is self-consistent, but the paper's central 'open-boundary' claim rests on an unproved cylinder reduction, and the parallelogram-geometry result in Sec. IV is stated without derivation.

Significance. If the boundary-condition issues are resolved, this would be a valuable contribution: it is one of the few analytically solvable 2D non-Hermitian models with explicit non-Bloch Dirac points, an exact phase diagram, and a concrete demonstration of geometry-dependent bulk-boundary correspondence. The derivation of Eq. (7) is analytic and parameter-free, the Hermitian mapping is explicit, and the topological charge is computed from a winding formula rather than asserted. The main value is the exact cylinder solution and the symmetry classification of the non-Bloch Dirac points. However, the abstract's claim that the Dirac points 'only appear under open boundary conditions' is stronger than what is currently established, because the calculation is performed on a cylinder and the fully open case is not proven.

major comments (3)
  1. [II, Eqs. (1)-(2)] Equation (1) and Eq. (2) are not equivalent under the stated standard Pauli conventions. With the usual σx, σy matrices, Eq. (1) gives off-diagonal matrix elements i(γ−t)−iw e^{-ikx} and i(γ+t)+iw e^{ikx}, whereas Eq. (2) has t+γ+wβ and t−γ+w/β. Eq. (2) corresponds instead to (t+w cos kx)σx + (iγ−w sin kx)σy + iησz, i.e., the σx and σy terms are interchanged relative to Eq. (1). Since the entire subsequent derivation, including the symmetries and the Hermitian mapping, uses Eq. (2), the model definition must be corrected or the Pauli convention explicitly fixed. This is a load-bearing inconsistency in the definition of the model.
  2. [II, after Eq. (1)] The reduction from a fully open system to the cylinder is asserted without proof. The sentence 'Since the hopping along the y direction is Hermitian, taking PBC and OBC in y will yield the same results' is not sufficient for the claims made here. For a finite open strip, the allowed ky are discrete standing-wave values, so an exact gap closing at a specific ky,c occurs only if ky,c coincides with a discrete level; in the thermodynamic limit the spectrum becomes dense, but this limit is not stated. The sentence 'To analytically obtain the fully open boundary spectrum, we take PBC in y and OBC in x' is also self-contradictory. Please either prove the equivalence for the quantities computed, provide finite-size OBC-in-both-directions checks for the non-Bloch Dirac points, or explicitly reframe the results as cylinder results rather than fully open boundary results.
  3. [IV, Fig. 5(b)] The parallelogram-geometry calculation is stated without any derivation. After saying 'we calculated the positions of the non-Bloch DPs', the text only gives the notation kt_c,1 and kt_c,2 with no formula, no generalized Brillouin zone construction for the tilted boundary, and no numerical method. This is the central evidence for the paper's claim of geometry-dependent bulk-boundary correspondence. The calculation must be shown, or at least a reproducible numerical procedure must be provided, for both the rectangular and parallelogram geometries.
minor comments (5)
  1. [Abstract and Introduction] The name 'Altland-Zirnabuer' should be spelled 'Altland-Zirnbauer'.
  2. [II, phase boundary list] The listed boundary 'η = √ξ + 1 where −1 < ξ < 0' is inconsistent because √ξ is imaginary for negative ξ; from the preceding paragraph it should be η = √(ξ+1).
  3. [III A, Eq. (11) and the definition of P] As written, the similarity transformation P = e^{iπ/4 σz} diag{1, sqrt((t−γ)/(t+γ))} does not reproduce Eq. (11); the sign of the π/4 rotation appears to be opposite to what is needed. Please check the ordering and sign convention so that the mapping is reproducible.
  4. [IV, Fig. 5 caption] The text refers to 'link = 1/100' and 'link = 1/2', but the variable t_b mentioned in the main text is not defined in the caption; please clarify the notation.
  5. [III B, Eq. (14)] The long-range perturbation terms would benefit from a sentence describing the lattice interpretation of δ and a, and the expression 'awβ_x^2' should be typeset more clearly to avoid confusion with a parameter named aw.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the analytical phase diagram, non-Bloch Dirac points, and topological charge are derived directly from the model without fitted inputs or load-bearing self-citations.

full rationale

The paper's central derivation chain is self-contained. Starting from H(kx, ky), the authors impose the open-boundary condition along x through the non-Bloch ansatz beta_x and the standing-wave requirement |beta_x,+| = |beta_x,-|, which leads to the exact spectral formula in Eq. (7); the phase boundaries follow from analyzing the reality and imaginary character of that expression, with no parameter fitted to any target. The non-Bloch Dirac point condition is obtained by an independent discriminant calculation, Disc_beta_x f(0, beta_x, ky,c) = 0, and the topological index is computed from the explicit Hermitianized Hamiltonian h(k) in Eq. (11) via the standard winding integral in Eq. (13). The only notable epistemic weakness is the unproved statement that Hermitian y hopping makes PBC and OBC in y equivalent, which is a geometric modeling assumption rather than a circular reduction: no predicted quantity is fed back into the derivation. The use of prior work such as the auxiliary GBZ method [30] is as a computational tool, not as an imported uniqueness theorem or fitted input. Thus no load-bearing step reduces to its own input by construction, and the appropriate circularity score is 0.

Assumptions & free parameters 6 free parameters · 4 assumptions · 0 invented entities

The model parameters (r, c, gamma, w, eta, delta, a) are chosen by hand to define the Hamiltonian; they are not fitted to external data. The derivation relies on standard non-Bloch band theory and on the unproven but plausible reduction to a cylinder geometry. No new physical entities are introduced.

free parameters (6)
  • r = 7/8 in figures
    Intra-cell hopping amplitude, model parameter scanned in the phase diagram.
  • c = 1/4 in figures
    Inter-chain coupling along y, model parameter.
  • gamma = 1/4 in figures
    Non-reciprocity strength, model parameter.
  • w = 1 (set by energy scale)
    Inter-cell hopping; set to 1 without loss of generality by scaling the total energy.
  • eta = 0 at the non-Bloch DP
    On-site gain/loss strength; the phase diagram spans eta.
  • delta, a = 1/10 in Fig. 4
    Long-range hopping strengths in the generalized model; used to test robustness of the DPs.
assumptions (4)
  • domain assumption The open-boundary spectrum is described by the non-Bloch band theory, in particular the generalized Brillouin zone condition |beta_x,+| = |beta_x,-|.
    Invoked in Section II to derive Eq. (7); standard result from refs [12,14].
  • domain assumption PBC and OBC in y give the same spectrum because the y-direction hopping is Hermitian.
    Stated in Section II after Eq. (1); load-bearing for reducing the 2D OBC problem to a quasi-1D cylinder.
  • standard math A similarity transformation on the GBZ Hamiltonian preserves the spectrum and the topological classification.
    Used in Section III A to map to a Hermitian Hamiltonian; standard linear algebra.
  • domain assumption The Hermitian classification of Dirac points in the Altland-Zirnbauer classes applies to the Hermitianized Hamiltonian.
    Invoked in Section III A with refs [6,45] to assign a Z topological charge.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Non-Bloch Dirac Points and Phase Diagram in the Stacked Non-Hermitian SSH Model." pith.science (2026). https://pith.science/paper/6Q2VHEGS

@misc{pith2026241202782,
  author       = {Pith},
  title        = {Pith review of: Non-Bloch Dirac Points and Phase Diagram in the Stacked Non-Hermitian SSH Model},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/6Q2VHEGS}},
  note         = {Machine review of arXiv:2412.02782}
}
read the original abstract

Topological semimetals exhibit protected band crossings in momentum space, accompanied by corresponding surface states. Non-Hermitian Hamiltonians introduce geometry-sensitive features that dissolve this bulk-boundary correspondence principle. In this paper, we exemplify this phenomenon by investigating a non-Hermitian 2D stacked SSH chain model with non-reciprocal hopping and on-site gain/loss. We derive an analytical phase diagram in terms of the complex energy gaps in the open-boundary spectrum. The phase diagram reveals the existence of non-Bloch Dirac points, which feature a real spectrum and only appear under open boundary conditions but disappear in Bloch bands under periodic boundary conditions. Due to the reality of the spectrum in the vicinity of non-Bloch Dirac points, we can locally map it to Hermitian semimetals within the Altland-Zirnabuer symmetry classes. Based on this mapping, we demonstrate that non-Bloch Dirac points are characterized by an integer topological charge. Unlike the band crossings in Hermitian semimetals, the locations of the non-Bloch Dirac points under different boundary geometries do not match each other, indicating a geometry-dependent bulk-boundary correspondence in non-Hermitian semimetals. Our findings provide new pathways into establishing unconventional bulk-boundary correspondence for non-Bloch Dirac metals in non-Hermitian systems.

Figures

Figures reproduced from arXiv: 2412.02782 by the authors.

Figure 1
Figure 1. FIG. 1. Graphical representation of 2D stacked SSH model with [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2. Phase diagram (left image) and open boundary energy spectra for each phase (right panels). Only the [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. FIG. 3. The non-Bloch Dirac points in the band structure. Here we [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: FIG. 4. Here we take [PITH_FULL_IMAGE:figures/full_fig_p005_4.png]
Figure 5
Figure 5. Figure 5: FIG. 5. Crossing R3-R4 (a) real energy band structure (a), bulk [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

48 extracted references · 22 canonical work pages

  1. [1]

    C. L. Kane and E. J. Mele, Phys. Rev. Lett. 95, 146802 (2005)

    2005

  2. [2]

    η = √ξ + 1where −1 < ξ <0

    work page

  3. [3]

    η = √ξ − 1 where ξ >1

    work page

  4. [4]

    KdpqKj9erxD5bRIILPzt2UwRxKA=

    ξ = (η − 1)2 where η >0. These phase boundaries partition the (ξ, η) plane into six dis- tinct regions R1-R6, each representing a phase with unique spectral characteristics. Next, we outline the key characteristics of each phase in Fig. 2 and the transitions between them. According to the previous analysis, phase R1 is imaginary-line-gapped and R2 is real...

    work page

  5. [5]

    B. A. Bernevig, T. L. Hughes, and S.- C. Zhang, Science 314, 1757 (2006), https://www.science.org/doi/pdf/10.1126/science.1133734

    doi:10.1126/science.1133734 2006

  6. [6]

    A. P. Schnyder, S. Ryu, A. Furusaki, and A. W. W. Ludwig, Phys. Rev. B 78, 195125 (2008)

    2008

  7. [7]

    Qi and S.-C

    X.-L. Qi and S.-C. Zhang, Rev. Mod. Phys. 83, 1057 (2011)

    2011

  8. [8]

    Fu and C

    L. Fu and C. L. Kane, Phys. Rev. B 76, 045302 (2007)

    2007

Show all 48 references
  1. [9]

    C.-K. Chiu, J. C. Y . Teo, A. P. Schnyder, and S. Ryu, Rev. Mod. Phys. 88, 035005 (2016)

    2016

  2. [10]

    Ryu and Y

    S. Ryu and Y . Hatsugai, Phys. Rev. Lett.89, 077002 (2002)

    2002

  3. [11]

    N. P. Armitage, E. J. Mele, and A. Vishwanath, Rev. Mod. Phys. 90, 015001 (2018)

    2018

  4. [12]

    H. Gao, J. W. Venderbos, Y . Kim, and A. M. Rappe, Annual Review of Materials Research 49, 153 (2019)

    2019

  5. [13]

    B. Lv, T. Qian, and H. Ding, Reviews of Modern Physics 93, 025002 (2021)

    2021

  6. [14]

    T. E. Lee, Phys. Rev. Lett. 116, 133903 (2016)

    2016

  7. [15]

    Yao and Z

    S. Yao and Z. Wang, Phys. Rev. Lett. 121, 086803 (2018)

    2018

  8. [16]

    F. K. Kunst, E. Edvardsson, J. C. Budich, and E. J. Bergholtz, Phys. Rev. Lett. 121, 026808 (2018)

    2018

  9. [17]

    Yokomizo and S

    K. Yokomizo and S. Murakami, Phys. Rev. Lett. 123, 066404 (2019)

    2019

  10. [18]

    C. H. Lee and R. Thomale, Phys. Rev. B 99, 201103(R) (2019)

    2019

  11. [19]

    D. S. Borgnia, A. J. Kruchkov, and R.-J. Slager, Phys. Rev. Lett. 124, 056802 (2020)

    2020

  12. [20]

    Zhang, Z

    K. Zhang, Z. Yang, and C. Fang, Phys. Rev. Lett. 125, 126402 (2020)

    2020

  13. [21]

    Okuma, K

    N. Okuma, K. Kawabata, K. Shiozaki, and M. Sato, Phys. Rev. Lett. 124, 086801 (2020)

    2020

  14. [22]

    Kawabata, K

    K. Kawabata, K. Shiozaki, M. Ueda, and M. Sato, Phys. Rev. X 9, 041015 (2019)

    2019

  15. [23]

    Zhang, Z

    K. Zhang, Z. Yang, and C. Fang, Nature communications 13, 2496 (2022)

    2022

  16. [24]

    Wang, J.-S

    Z.-Y . Wang, J.-S. Hong, and X.-J. Liu, Phys. Rev. B 108, L060204 (2023)

    2023

  17. [25]

    R. Lin, T. Tai, L. Li, and C. H. Lee, Frontiers of Physics 18, 53605 (2023)

    2023

  18. [26]

    S. Yao, F. Song, and Z. Wang, Phys. Rev. Lett. 121, 136802 (2018)

    2018

  19. [28]

    Kawabata, N

    K. Kawabata, N. Okuma, and M. Sato, Phys. Rev. B 101, 195147 (2020)

    2020

  20. [29]

    Yokomizo and S

    K. Yokomizo and S. Murakami, Phys. Rev. B 107, 195112 (2023)

    2023

  21. [30]

    H.-Y . Wang, F. Song, and Z. Wang, Phys. Rev. X 14, 021011 (2024)

    2024

  22. [31]

    Yokomizo and S

    K. Yokomizo and S. Murakami, Phys. Rev. Res. 2, 043045 (2020)

    2020

  23. [32]

    L. Xiao, T. Deng, K. Wang, G. Zhu, Z. Wang, W. Yi, and P. Xue, Nature Physics 16, 761 (2020)

    2020

  24. [33]

    Z. Yang, K. Zhang, C. Fang, and J. Hu, Phys. Rev. Lett. 125, 226402 (2020)

    2020

  25. [34]

    Deng and W

    T.-S. Deng and W. Yi, Phys. Rev. B 100, 035102 (2019)

    2019

  26. [35]

    Okuma and M

    N. Okuma and M. Sato, Annual Review of Condensed Matter Physics 14, 83 (2023)

    2023

  27. [36]

    H. Xue, Q. Wang, B. Zhang, and Y . D. Chong, Phys. Rev. Lett. 124, 236403 (2020)

    2020

  28. [37]

    Zhang, Z

    K. Zhang, Z. Yang, and C. Fang, Nature Communications 13, 2496 (2022)

    2022

  29. [38]

    Wang, J.-S

    Y .-C. Wang, J.-S. You, and H. H. Jen, Nature Communications 13, 4598 (2022)

    2022

  30. [39]

    W. Wang, M. Hu, X. Wang, G. Ma, and K. Ding, Phys. Rev. Lett. 131, 207201 (2023)

    2023

  31. [40]

    Q. Zhou, J. Wu, Z. Pu, J. Lu, X. Huang, W. Deng, M. Ke, and Z. Liu, Nature Communications 14, 4569 (2023)

    2023

  32. [41]

    T. Wan, K. Zhang, J. Li, Z. Yang, and Z. Yang, Science Bulletin 68, 2330 (2023)

    2023

  33. [42]

    Y . Qin, K. Zhang, and L. Li, Phys. Rev. A109, 023317 (2024)

    2024

  34. [43]

    Two-dimensional non-hermitian skin effect in an ultracold fermi gas,

    E. Zhao, Z. Wang, C. He, T. F. J. Poon, K. K. Pak, Y .- J. Liu, P. Ren, X.-J. Liu, and G.-B. Jo, “Two-dimensional non-hermitian skin effect in an ultracold fermi gas,” (2023), arXiv:2311.07931

    2023

  35. [44]

    W. P. Su, J. R. Schrieffer, and A. J. Heeger, Phys. Rev. Lett.42, 1698 (1979)

    1979

  36. [45]

    Lieu, Phys

    S. Lieu, Phys. Rev. B 97, 045106 (2018)

    2018

  37. [46]

    Yuce and H

    C. Yuce and H. Ramezani, Phys. Rev. A 100, 032102 (2019)

    2019

  38. [47]

    H. C. Wu, L. Jin, and Z. Song, Phys. Rev. B 103, 235110 (2021)

    2021

  39. [48]

    Chiu and A

    C.-K. Chiu and A. P. Schnyder, Phys. Rev. B90, 205136 (2014)

    2014

  40. [49]

    J. C. Budich and E. J. Bergholtz, Phys. Rev. Lett. 125, 180403 (2020)

    2020

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.