Pith. sign in

REVIEW 4 major objections 6 minor 5 references

Fabrication and Characterization of the Moir\'e surface state on a topological insulator

T0 review · 4 major / 6 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read A two-step MBE growth produces a Moiré superlattice on the Sb2Te3 topological surface state, evidenced by a shifted Dirac point and Landau levels that require a second Dirac cone.

desk verdict A clever two-step MBE growth method likely creates a twisted Sb2Te3 homojunction with a real Moiré pattern, but the electronic-state evidence is overinterpreted. read the letter →

arxiv 2509.03322 v1 pith:P5TX7HBL submitted 2025-09-03 cond-mat.mtrl-sci

Yi Zhang , Dang Liu , Qiaoyan Yu , Ruijun Xi , Xingsen Chen , Shasha Xue , Jice Sun , Xian Du
show 14 more authors
Xuhui Ning Tingwen Miao Pengyu Hu Hao Yang Dandan Guan Xiaoxue Liu Liang Liu Yaoyi Li Shiyong Wang Canhua Liu Haijiao Ji Noah F. Q. Yuan Hao Zheng Jinfeng Jia
This is my paper · ORCID
classification cond-mat.mtrl-sci
keywords MoirésuperlatticetopologicalinsulatorSb2Te3molecularbeamepitaxytwo-stepgrowthLandaulevelsDiracconetwistronics
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 is trying to establish that a Moiré superlattice — a periodic interference pattern created by stacking two atomic lattices with a small twist — can be grown directly on the surface of a topological insulator, something theory predicts but experiments had not realized. The route is a two-step molecular beam epitaxy growth: a low-temperature step forms Sb2Te3 films with rotational domains, and a high-temperature step lets the top quintuple layer overgrow the domain boundary with an ~11° twist, producing a 2.24 nm Moiré period. The evidence is spectroscopic as well as structural: the surface Dirac point shifts from 85 to 121 meV in the Moiré region, and under magnetic fields new Landau-level peaks appear above 6.3 T that a single Dirac cone cannot explain. A two-Dirac-cone model, with the second cone carrying an enhanced Zeeman term, fits the fans for four Moiré periods. If the interpretation is right, topological-insulator twistronics becomes accessible by growth rather than by manual stacking, and the method should transfer to other van der Waals materials.

What carries the argument

The structural machinery is the Moiré interference itself, quantified by the twist-angle relation $\lambda_m = a/\sqrt{2(1-\cos\theta)}$, which converts the measured 11.0° rotation of the top quintuple layer into a predicted 2.21 nm period that matches the observed 2.24 nm. The electronic machinery is a two-Dirac-cone model. The first cone is the ordinary topological surface state with Hamiltonian $H_1 = v_1(\sigma_x\pi_y - \sigma_y\pi_x)$ and Landau levels $E_{n,1}=E_{D1}+\mathrm{sgn}(n)v_{F1}\sqrt{2e\hbar|n|B}$. The second cone adds a Zeeman term, $H_2 = v_2(\sigma_x\pi_y - \sigma_y\pi_x) + gB\sigma_z$, giving $E_{n,2}=E_{D2}+\mathrm{sgn}(n)\sqrt{2v_{F2}^2e\hbar|n|B + g^2B^2}$ and a zeroth level $E_{0,2}=E_{D2}-gB$; it is this Zeeman term that produces the extra Landau-level features and the nonzero slope of the zeroth level versus field. The model does the work of explaining why one Dirac cone cannot account for the measured fans, and it yields the fitted Fermi velocities, Dirac-point offsets, and the enhanced $g$ factor.

What would settle it

Measure the twist angle of the buried interface directly — for example by electron diffraction or cross-sectional transmission electron microscopy on identically grown films — and confirm both the 11.0° rotation and the 2.24 nm period. If the buried interface is not a coherent twist, or if the extra Landau-level peaks do not follow the two-Dirac-cone fan when the period is known independently, the central claim fails.

Watch

Extended reading notes

Core claim

The central claim is that a twisted top layer of Sb2Te3 produces a genuine Moiré topological surface state with measurable electronic consequences. In the grown films, two rotational domains differ by 11.0° ± 0.3°, and the eighth quintuple layer extends over the seventh to create a Moiré pattern whose 2.24 nm period matches the twist-angle formula. Scanning tunneling spectroscopy shows the Dirac point moves from 85 to 121 meV on the Moiré region. Under magnetic fields up to 10.5 T, the Moiré region develops Landau-level peaks absent on pristine Sb2Te3, beginning at 6.3 T where the cyclotron orbit first fits inside the ~19 nm Moiré patch; position-dependent spectra rule out charged-defect splitting. The fans from four different Moiré periods (2.08, 2.17, 2.24, 2.41 nm) are fit by a two-Dirac-cone model — one cone for the original surface state and one satellite cone with a magnetic-field-dependent Zeeman energy — while the one-cone model fails, and the fitted g factor of the satellite cone is roughly an order of magnitude larger than that of the original cone.

Load-bearing premise

The load-bearing premise is that the topmost layer is a coherent, uniformly twisted Sb2Te3 layer rather than a strained or polycrystalline overlayer; the 11.0° twist angle is inferred from atomically resolved top-surface images and the matching Moiré period, with no independent cross-sectional or diffraction confirmation of the buried interface.

Editorial extensions

If this is right

  • The two-step growth method, being an in-vacuum growth route rather than manual stacking, can be applied to other van der Waals materials to build Moiré superlattices without contamination or interface defects.
  • On the topological surface state, the Moiré pattern acts as a band-engineering tool: it shifts the Dirac point and creates a satellite Dirac cone whose Landau quantization is visible in tunneling spectra.
  • The enhanced g factor of the satellite cone, about an order of magnitude larger than the intrinsic cone's, makes the Moiré region a candidate platform for spin manipulation of topological surface states.
  • Because the onset of the new Landau-level peaks is tied to the cyclotron orbit fitting in the ~19 nm Moiré patch, larger Moiré domains should show the extra features at lower magnetic fields.
  • With future doping, the Moiré topological surface state should be able to access the correlated and superconducting phases predicted for this system.

Reading between the lines

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

  • A testable extension follows from the size argument: if the 6.3 T onset is really a cyclotron-orbit effect, growing Moiré domains of different radii should move the onset as $B \propto 1/r^2$; if it does not, the onset has a different origin.
  • An independent structural probe of the buried twist interface would separate the Moiré interpretation from a strain or polycrystalline-overlayer explanation, since all current angle evidence comes from top-surface atomic images.
  • The two-Dirac-cone fit treats the second cone as an independent object, but a Moiré band with mini-Brillouin-zone effects could produce fan deviations at low Landau index; comparing the four fitted $g$ factors against twist angle would test whether the Zeeman term is really Moiré-induced.
  • If the $g$ factor is Moiré-induced, its value should depend on period or twist angle; the four fitted values (1.06, 2.23, 4.25, 8.50 meV/T) vary widely, so a systematic twist-angle dependence would be a direct way to check the mechanism.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 6 minor

Summary. The manuscript reports the fabrication of a Moiré superlattice on the surface of the topological insulator Sb2Te3 using a two-step molecular beam epitaxy growth method. The authors argue that the topmost quintuple layer of one rotational domain overgrows a neighboring domain, creating a twisted homojunction whose surface exhibits a 2.24 nm Moiré period. Scanning tunneling microscopy and spectroscopy reveal a Dirac-point shift between the Moiré and normal regions, and Landau-level spectra measured under magnetic fields up to 10.5 T show additional peaks in the Moiré region. These data are modeled with a two-Dirac-cone Hamiltonian, yielding a satellite Dirac cone with an enhanced Landé g factor. The paper claims this is the first experimental realization of a Moiré topological surface state and suggests the growth method is broadly applicable to other van der Waals materials.

Significance. If the structural and electronic interpretations are correct, this work would be an important experimental milestone: it would demonstrate a route to Moiré engineering on topological insulator surfaces using epitaxial growth, avoiding the contamination and interface defects associated with manual stacking. The two-step growth method itself is plausible and potentially transferable, and the quantitative match between the measured Moiré period and the twist-angle formula is a genuine strength. The Landau-level data are extensive and the comparison between normal and Moiré regions is a useful approach. However, the central electronic-state claim relies on a five-parameter fit without reported uncertainties, and the structural identification of a buried twist is inferred rather than directly imaged. These limitations currently temper the impact of the result.

major comments (4)
  1. [Figure 2 and accompanying text (structural determination)] The claim that the Moiré pattern originates from a coherent 11.0° twist between the topmost quintuple layer and the underlying 7-QL film is inferred from atomic-resolution images of two normal domains and from the agreement of the measured 2.24 nm period with the twist-angle formula. The buried interface itself is never directly imaged, and the same surface period could in principle arise from strain modulation, an interfacial reconstruction, or a local stacking variation at the domain boundary. Because every electronic signature in the paper is interpreted through this structural premise, the authors should provide direct evidence for the twisted interface (e.g., cross-sectional STM or TEM, LEED, or atomically resolved imaging across the boundary that shows both lattices coexisting at the same location) or explicitly rule out alternative origins for the periodic modulation.
  2. [Section 4, Eqs. (2)-(4) and Table 1] The two-Dirac-cone model introduces five free parameters per Moiré region (ED1, vF1, ED2, vF2, and g), and the manuscript does not report the number of Landau-level peaks used, the fit uncertainties, or a quantitative comparison with the one-cone model. With five adjustable parameters, a good fit to a limited set of peaks is not a strong validation of the model. The statement that the one-cone model 'clearly fails' should be supported by a figure and goodness-of-fit metrics. Furthermore, the fitted g values vary from 1.06 to 8.50 meV/T with no monotonic trend versus the Moiré period, suggesting that this parameter may be absorbing unrelated physics rather than representing a robust Zeeman term.
  3. [Section 3, paragraph on the 6.3 T onset] The authors explain the appearance of the new Landau-level features at B ≥ 6.3 T by the cyclotron orbit fitting within the 19 nm effective radius of the Moiré region. This finite-size threshold indicates that the observed peaks may be quantum-confinement states of a finite Moiré patch rather than intrinsic Landau levels of a periodic superlattice. The text should clarify whether the two-Dirac-cone model is intended to describe an infinite periodic Moiré system or a finite confined region, and if the latter, how the model accounts for the boundary conditions.
  4. [Section 3, Fig. 2e-f and Table 1] The reported Dirac-point shift from 85 to 121 meV is presented as evidence that the Moiré pattern modifies the electronic structure, but the ED1 values for the other Moiré regions in Table 1 are 153-160 meV, and no statistics or error bars are provided for any of the extracted energies. A shift of this magnitude could also result from local doping variations or tip-induced band bending. The authors should report measurements on multiple normal regions and compare the extracted ED1 and vF1 values with theoretical predictions for a twisted topological insulator surface (e.g., Refs. 14 and 15) to establish that the observed changes are attributable to the Moiré potential.
minor comments (6)
  1. [Section 2, Moiré period formula] The formula is written as λ_m = a/√2(1−cos θ), which is ambiguous and would give an incorrect value if read literally; it should be λ_m = a/√[2(1−cos θ)].
  2. [Figure 3 caption] The caption begins with 'Figure 3. Figure 3.'; the duplication should be removed.
  3. [References] Reference 20 contains a typo in the journal name: 'Phys. Rev. ett.' should be 'Phys. Rev. Lett.'.
  4. [Table 1] The g-factor values are given in meV/T; the authors should state the conversion to a dimensionless g factor or define an effective g value to make the magnitude of the enhancement more transparent.
  5. [Section 5 (summary)] The claim that the two-step growth method 'can be widely applied to other van der Waals materials' is not supported by any demonstration in this manuscript; it should be framed as a proposal rather than an established feature.
  6. [Section 3, height measurement] The statement that 'the height of 1 QL of Sb2Te3 in the Moiré region is slightly higher than that of the normal region' is not quantified; a line-profile height comparison should be included in the Supporting Information.

Circularity Check

1 steps flagged · score 4.0 of 10

The structural Moiré fabrication is self-contained, but the enhanced g-factor claim is a fitted parameter restated as a discovery; no parameter-free prediction is tested.

  1. fitted input called prediction [Main text after Figure 3, equations (3)-(4), Table 1]
    "The second Dirac cone’ Landau quantization follows25,28: H2 = v2(σxπy − σyπx) + gBσz ... E0,2 = ED2 − gB ... We find that the two-Dirac-cone model agrees well with the data but the oneDirac-cone model clearly fails to fit all Landau-level peaks. The deduced parameters are listed in Table 1, from which one may note that the nonzero g factor is essential to fitting the second Dirac cone ..."

    The Zeeman term gB in Eq. (4) is a free parameter of the fitting model. The Moiré-region Landau fan is fit by adjusting ED2, vF2, and g, and Table 1 lists the resulting fitted g values (e.g., 8.50 meV/T for the 2.24 nm period). The paper then presents the nonzero slope E0,2 = ED2 − gB and the factor-of-ten enhancement of g as properties of the Moiré-induced Dirac cone. Those claims are numerically identical to the fitted parameter, restated as a physical discovery. No independent measurement or parameter-free theoretical value of g is supplied; the cited theory (ref 28) provides only the Hamiltonian form, not the magnitude. Thus the stated agreement of the two-Dirac-cone model is a goodness-of-fit summary of the very data used to determine g, not an external confirmation.

full rationale

The structural core of the paper is not circular: the 11.0° twist angle and the 2.24 nm Moiré period are measured on different images and matched through the standard formula λ_m = a/√2(1−cosθ), so the period consistency is an independent cross-check rather than a definitional identity. The choice of a two-Dirac-cone model is also not circular, because a satellite Dirac cone from a Moiré potential on a topological insulator is an external prediction of refs 14 and 15. However, the paper's central electronic claim about the Moiré-induced satellite cone—specifically the large g factor and the nonzero slope of its zeroth Landau level—is obtained by fitting the same Landau-level data that are then said to confirm the model. The g parameter is free in Eq. (4), the fit determines it, and the paper reports the fitted value as a discovery. That is a fitted input presented as a result, and the 'agreement' of the model is therefore not a parameter-free prediction. The structural fabrication and the qualitative existence of new Landau-level features retain independent experimental content, so the circularity is partial rather than total, supporting a moderate score of 4 rather than 6 or higher.

Assumptions & free parameters 5 free parameters · 6 assumptions · 0 invented entities

The fabrication claim rests on standard MBE and STM assumptions and the standard Moiré period formula. The electronic-state interpretation adds five fitted parameters per Moiré region and assumes the theoretical satellite-Dirac-cone picture. No new particles, forces, or conserved quantities are introduced; the 'satellite Dirac cone' is a predicted band-structure feature, not an independently evidenced new entity.

free parameters (5)
  • ED1 (first Dirac cone energy) = 120-160 meV across four regions
    Fitted from Landau-level fans in Figure 4; not independently measured in zero-field spectra.
  • vF1 (first cone Fermi velocity) = 4.66-5.51 x 10^5 m/s
    Fitted from the Landau fan; the normal-region fit gives 3.43 x 10^5 m/s, so the Moiré-region value is not independently constrained.
  • ED2 (satellite Dirac cone energy) = 155-211 meV across four regions
    Fitted from Moiré-region Landau levels; no direct zero-field spectral feature is shown for this cone.
  • vF2 (satellite cone Fermi velocity) = 4.50-5.09 x 10^5 m/s
    Fitted from the same Landau-level data; varies by more than ten percent across the four regions.
  • g (effective Landé factor of satellite cone) = 1.06-8.50 meV/T
    Fitted and essential to the two-Dirac-cone model, but varies eightfold across the four regions and has no stated uncertainty.
assumptions (6)
  • domain assumption Sb2Te3 films thicker than 4 QLs host a topological surface state with a single Dirac cone at the Gamma point.
    Invoked to justify that the 8-QL film has a topological surface state; cited to Refs. 20 and 21.
  • standard math The Moiré period is given by lambda = a / sqrt(2(1 - cos theta)).
    Used to relate the measured twist angle of 11.0 degrees to the expected Moiré period of 2.21 nm; from Ref. 1.
  • domain assumption The Landau-level spectrum of the surface state follows the Dirac form in Eqs. 1-4, including a Zeeman term gB sigma_z for the satellite cone.
    This is the fitting model; the massive-Dirac-with-Zeeman form is taken from prior literature (Refs. 25 and 28), not derived in this paper.
  • domain assumption The Moiré potential generates one additional satellite Dirac cone, as predicted by theory.
    The two-Dirac-cone fit presupposes the theoretical prediction in Refs. 14, 15, and 28; the paper does not independently derive that this potential yields exactly one satellite cone.
  • domain assumption The new Landau-level peaks in the Moiré region are intrinsic to the Moiré superlattice and not caused by charged defects or other disorder.
    Supported only by position-dependent spectra cited as SI Figures S3, which are not shown in the main text; this assumption underlies the central electronic interpretation.
  • standard math The cyclotron radius formula r_k = sqrt((2k+1) hbar / eB) is valid for estimating the onset field of the first Landau level.
    Used to connect the 19 nm Moiré-region radius to the 6.3 T onset field; taken from Ref. 27.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Fabrication and Characterization of the Moir\'e surface state on a topological insulator." pith.science (2026). https://pith.science/paper/P5TX7HBL

@misc{pith2026250903322,
  author       = {Pith},
  title        = {Pith review of: Fabrication and Characterization of the Moir\'e surface state on a topological insulator},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/P5TX7HBL}},
  note         = {Machine review of arXiv:2509.03322}
}
read the original abstract

A Moire superlattice on the topological insulator surface is predicted to exhibit many novel properties but has not been experimentally realized. Here, we developed a two-step growth method to successfully fabricate a topological insulator Sb2Te3 thin film with a Moire superlattice, which is generated by a twist of the topmost layer via molecular beam epitaxy. The established Moire topological surface state is characterized by scanning tunneling microscopy and spectroscopy. By application of a magnetic field, new features in Landau levels arise on the Moire region compared to the pristine surface of Sb2Te3, which makes the system a promising platform for pursuing next-generation electronics. Notably, the growth method, which circumvents contamination and the induced interface defects in the manual fabrication method, can be widely applied to other van der Waals materials for fabricating Moire superlattices.

Figures

Figures reproduced from arXiv: 2509.03322 by the authors.

Figure 3
Figure 3. [PITH_FULL_IMAGE:figures/full_fig_p011_3.png] view at source ↗

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

5 extracted references · 5 canonical work pages

  1. [2]

    C.; Wang, D.; Jin, C.; Bakti Utama, M

    (13) Regan, E. C.; Wang, D.; Jin, C.; Bakti Utama, M. I.; Gao, B.; Wei, X.; Wang, F. et al. Mott and Generalized 15 Wigner Crystal States in WSe₂/WS₂ Moiré Superlattices. Nature 2020, 579 (7799), 359–363. (14) Cano, J.; Fang, S.; Pixley, J. H.; Wilson, J. H. Moiré Superlattice on the Surface of a Topological Insulator. Phys. Rev. B 2021, 103 (15), 15515...

    work page 2020

  2. [4]

    Parts a-d show the Landau fans measured on four different Moiré regions with periods of 2.08, 2.17, 2.24, and 2.41 nm, respectively

    Landau fan fitting results under different Moiré periods. Parts a-d show the Landau fans measured on four different Moiré regions with periods of 2.08, 2.17, 2.24, and 2.41 nm, respectively. A two-Dirac-cone model fits well with the data, and the red and black curves represent original and satellite Dirac cones, respectively. 13 2.08nm 2.17nm 2.24nm 2.4...

    work page 2011

  3. [8]

    A.; Das, I.; Efetov, D

    (7) Lu, X.; Stepanov, P .; Yang, W.; Xie, M.; Aamir, M. A.; Das, I.; Efetov, D. K. et al. Superconductors, Orbital Magnets and Correlated States in Magic-Angle Bilayer Graphene. Nature 2019, 574 (7780), 653–657. (8) Serlin, M.; Tschirhart, C. L.; Polshyn, H.; Zhang, Y .; Zhu, J.; Watanabe, K.; Young, A. F. et al. Intrinsic Quantized Anomalous Hall Effect ...

    work page 2019

  4. [13]

    (23) Yin, Y .; Wang, G.; Liu, C.; Huang, H.; Chen, J.; Liu, J.; Jia, J. et al. Moiré-Pattern-Modulated Electronic Structures in Sb₂Te₃/Graphene Heterostructure. Nano Res. 2022, 15, 1115–1119. (24) Hsieh, D.; Qian, D.; Wray, L. A.; Xia, Y .; Hor, Y . S.; Cava, R. J.; Hasan, M. Z. Landau Levels in a Topological Insulator. Phys. Rev. B 2011, 84 (16), 161306...

    work page Pith review arXiv 2022

  5. [17]

    Correlated and Topological Physics in ABC -Trilayer Graphene Moiré Superlattices

    (6) Chen, G. Correlated and Topological Physics in ABC -Trilayer Graphene Moiré Superlattices. Quantum Front. 2022, 1,

    work page 2022

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.