Pith. sign in

REVIEW 3 major objections 5 minor 47 references

Only phonons whose atomic motions match the lattice reconstruction couple strongly to electrons in twisted TMDs, peaking EPC near the angles where superconductivity appears.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

An interference selection rule selects reconstruction-matched breathing modes for strong EPC in moiré TMDs, peaking at large twist angles near observed superconductivity.

T0 review reviewed 2026-07-14 challenge →

load-bearing objection Solid geometric selection rule for moiré EPC, backed by Raman and large-scale ML; the “phonons can dominate large-angle pairing” claim is still provisional because only Γ-point λ is computed. the 3 major comments →

arxiv 2607.11425 v1 pith:EEPSKJON submitted 2026-07-13 cond-mat.supr-con

Interference-Enhanced Large Electron-Phonon Coupling from Raman-active Breathing Modes in Moir\'e Semiconductors

Ning Mao , Shaozheng Wang , Cheng Xu , Xumin Chang , Kenji Watanabe , Takashi Taniguchi , Claudia Felser , Shengwei Jiang
show 1 more author
Yang Zhang
This is my paper
classification cond-mat.supr-con
keywords moiré semiconductorselectron-phonon couplingtwisted WSe2twisted MoTe2lattice reconstructionRaman spectroscopysuperconductivitybreathing modes
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

The reading

Superconductivity has been seen in twisted bilayer WSe2 and MoTe2 near large twist angles, but it has been unclear whether phonons can contribute meaningfully to the pairing. Direct first-principles electron-phonon calculations are impossible for the thousands-of-atom moiré cells that matter. This paper combines gate-tuned Raman spectroscopy with machine-learning force fields and Hamiltonians to compute mode-resolved coupling in cells up to tens of thousands of atoms. Raman finds that only a handful of modes (low-frequency layer breathing and a high-frequency optical breathing mode) shift strongly with carrier filling. The calculations trace the selectivity to an interference rule: a phonon couples strongly only when its displacement pattern matches the static lattice-reconstruction field; mismatched modes cancel by destructive interference. The rule makes the total coupling constant peak near 7°, close to the experimental superconducting regime, implying a substantial phonon contribution to large-angle pairing.

Core claim

In twisted TMD semiconductors a phonon mode couples strongly to the electronic structure only when its displacement texture matches the static lattice-reconstruction pattern (quantified by a pattern-matching parameter χ near 1); otherwise the coupling is suppressed by destructive interference. This interference selection rule picks out the Raman-active low- and high-frequency breathing modes and causes the total electron-phonon coupling λ to peak at large twist angles near those where superconductivity is observed.

What carries the argument

The interference selection rule quantified by the pattern-matching parameter χ (the projection of a phonon eigenvector onto the static reconstruction displacement field). Modes with χ ≈ 1 produce large EPC matrix elements; modes with χ ≈ 0 are cancelled by interference.

Load-bearing premise

That the Γ-point coupling already computed, plus a rough guess that finite-momentum phonons contribute a comparable amount, is enough to give a lower-bound λ that can produce a critical temperature matching experiment.

What would settle it

A full finite-q Eliashberg calculation (or equivalent measurement of the complete phonon spectral function) for ~4–5° t-WSe2 that finds the total λ substantially below the Γ-point lower bound of ~0.17, or Raman/ARPES data showing that the breathing modes do not dominate the filling-dependent renormalization.

Watch this falsifier. Get emailed when new claim-graph text bears on it.

If this is right

  • Any complete theory of large-angle pairing in twisted TMDs must include the electron-phonon channel; pure correlation-only scenarios are incomplete.
  • The same interference rule supplies a design principle for screening other moiré semiconductors for phonon-driven superconductivity.
  • The low- and high-frequency breathing modes identified by Raman are the dominant contributors that should be retained in effective models.
  • Extending the machine-learning frozen-phonon framework across the full moiré Brillouin zone will yield the Eliashberg function needed for quantitative Tc.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • If the interference rule is general, hetero-bilayers or trilayers with deliberately engineered reconstruction textures could be used to switch phonon-mediated pairing on or off.
  • The same selection rule may explain why certain interlayer modes dominate transport or optical response in other reconstructed van der Waals stacks even when the bare density of states is low.
  • A hybrid calculation that feeds the mode-resolved λ into a Kohn-Luttinger or spin-fluctuation channel would test whether phonons and correlations cooperate or compete near integer filling.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 5 minor

Summary. The manuscript combines filling-dependent Raman spectroscopy on dual-gated t-WSe2 (4.4° and 3.15°) with machine-learning force fields and Hamiltonians to compute mode-resolved electron–phonon coupling (EPC) in moiré supercells of thousands of atoms for t-WSe2 and t-MoTe2 (1.89°–13°). Raman finds only three filling-tunable moiré phonons (LBM, LSM, OBM). Theory attributes the mode selectivity to an interference selection rule: a phonon couples strongly only when its displacement texture matches the static lattice-reconstruction field, quantified by a pattern-matching parameter χ. Modes with χ≈1 (LBM and high-frequency OBM) dominate λ; mismatched modes (e.g. LSM) are suppressed. Consequently total λ peaks near 7.34°, near the large-angle regime where superconductivity is observed, while the DOS grows toward smaller angles. The authors conclude that lattice-reconstruction interference organizes moiré EPC and that phonons make a substantial, potentially dominant contribution to large-angle pairing.

Significance. The work addresses a timely open question—the relative roles of phonons versus electronic correlations in recently observed twisted-TMD superconductivity—by delivering mode-resolved EPC in realistic supercells that are inaccessible to direct DFT. Strengths include: (i) independent experimental identification of the same LBM/OBM that theory flags via high χ and high λ; (ii) a transparent, non-circular geometric metric χ that correlates with independently computed band shifts and λ; (iii) near-DFT MLFF force constants (MAE < 0.005 eV/Ų down to 5.08°); and (iv) a falsifiable organizing principle (reconstruction-matching interference) that predicts both mode selectivity and the non-monotonic angle dependence of λ. If the finite-q extension and pairing estimates hold, the result would reframe large-angle pairing discussions and supply a practical screening tool for phonon-favorable moiré systems.

major comments (3)
  1. Discussion (and abstract claim of a “substantial, potentially dominant” phonon contribution): only the Γ-point contribution is computed (λ = 0.17 for 4.4° t-WSe2). The Tc ≈ 0.3 K estimate that “matches experiment” explicitly assumes a finite-q piece of comparable size plus µ* ≈ 0.1. No finite-q matrix elements, no Eliashberg spectral function, and no q-resolved λ(q) are provided. Because this assumption is load-bearing for the pairing claim that frames the paper, either (a) compute representative finite-q EPC (or a controlled lower/upper bound) or (b) clearly demote the pairing language to a qualitative suggestion and state that a quantitative phonon-driven Tc remains unproven until finite-q data exist.
  2. §IV.A–C and Methods: EPC matrix elements |g| and λj are obtained from frozen-phonon band shifts evaluated with machine-learning Hamiltonians on supercells of thousands of atoms. The MLFF force constants are benchmarked against DFT (Fig. 2h), but no analogous validation of the ML Hamiltonians’ band energies or deformation potentials is reported for the large moiré cells (only a preprint citation). Systematic errors in the reconstruction-matching direction would directly affect χ–λ correlation and the absolute scale of λ. A DFT benchmark on a smaller-angle cell (or on selected high-χ modes) is needed to underwrite the quantitative EPC numbers.
  3. §IV.A and Fig. 4b: g(M,Γ) and λ are evaluated at the M-point van Hove singularity of the first valence band (ν = −1) with a 1 meV DOS broadening. The paper does not show how sensitive the reported λ peak near 7.34° or the mode ranking is to Fermi-level placement, broadening, or inclusion of the second band. A short sensitivity check would confirm that the angle dependence and the LBM/OBM dominance are robust rather than artifacts of these free parameters.
minor comments (5)
  1. Fig. 1f–i vs. experimental angles: calculated dispersions/EPC are shown for 4.4° and 5.08°, while Raman is at 4.4° and 3.15°. A brief statement that 5.08° is representative of the experimental 4.4° device would avoid confusion.
  2. Eq. (2) and surrounding text: the normal-mode amplitude Q and the precise conversion to physical displacements should be stated once for reproducibility; units of Q (Å√amu) appear only in the text.
  3. Fig. 4d: the second high-χ peak in the low-frequency window is said to dominate coupling at Γ (Fig. S10) but not at M; a one-sentence clarification in the main text would help readers who do not consult the SI.
  4. Methods (Raman): ALS baseline parameters and Lorentzian fitting are given; it would be useful to report the uncertainty on the extracted peak positions that underlie the ~0.1 meV filling shifts.
  5. Typographical: “t-WSez” appears once in Methods (Raman measurement); “moir´ e” accent rendering is inconsistent in a few places.

Circularity Check

0 steps flagged

No definitional circularity: χ is an independent geometric projector; λ and Raman shifts are computed/measured separately and only then correlated.

full rationale

The load-bearing chain is: (i) MLFF-relaxed structures give a static reconstruction field u_relax; (ii) the same MLFF yields phonon eigenvectors u_frozen; (iii) χ is the cosine similarity of those two fields (Eq. 3); (iv) frozen-phonon band shifts with ML Hamiltonians give independent matrix elements g and mode-resolved λ_j (Eqs. 2, 4); (v) filling-dependent Raman measures frequency shifts of a few modes. χ is not defined from λ or from the Raman data, and λ is not fitted to χ—the paper reports a post-hoc correlation between independently obtained quantities. The peak of total λ near 7.34° is a sum over computed mode contributions, not a fit to the experimental SC angles. Self-citations [33, 39] supply the MLFF and ML-Hamiltonian tools and are benchmarked against DFT IFCs (MAE < 0.005 eV/Ų), but they do not force the selection rule or the λ(θ) shape. The Discussion’s Tc estimate assumes finite-q EPC comparable to the Γ-point λ=0.17; that is an untested extrapolation (correctness risk), not a reduction of the result to its inputs by construction. No self-definitional loop, fitted-input-as-prediction, uniqueness import, or ansatz smuggling is present. Score 1 reflects only the minor, non-load-bearing methodological self-citations.

Axiom & Free-Parameter Ledger

4 free parameters · 5 axioms · 1 invented entities

The central claim rests on the validity of ML force fields and ML Hamiltonians for reconstructing forces, phonons and band shifts in multi-thousand-atom cells; on the harmonic frozen-phonon approximation for EPC matrix elements; on the geometric definition of χ as a faithful predictor of coupling; and on a small set of numerical choices (Fermi-level placement at M, 1 meV DOS broadening, µ* ≈ 0.1 for the Tc estimate). No new particles or forces are postulated; χ is a derived diagnostic, not an independent entity.

free parameters (4)
  • DOS broadening = 1 meV
    1 meV Gaussian/Lorentzian broadening used when integrating λj over the mBZ (Sec. IV.C); affects absolute λ scale.
  • Coulomb pseudopotential µ* = ≈0.1
    Assumed ≈0.1 when converting λ to Tc ≈ 0.3 K (Discussion); standard but free choice that sets the numerical match to experiment.
  • Normal-mode amplitude Q
    Finite-displacement amplitude used to extract g from band differences (Eq. 2); must be small enough for linearity yet large enough for numerical stability.
  • Fermi-level placement = M-point of first valence band
    Fixed at the M-point van-Hove singularity of the first valence band (ν = −1) for all λ calculations; motivated by experiment but still a modeling choice.
axioms (5)
  • domain assumption Machine-learning force fields trained on small-cell DFT data accurately reproduce interatomic force constants and phonon eigenvectors of multi-thousand-atom moiré supercells.
    Invoked throughout Sec. III and Methods; benchmarked only down to 5.08° (MAE < 0.005 eV/Ų).
  • domain assumption The frozen-phonon finite-displacement method with ML Hamiltonians yields reliable first-order EPC matrix elements g(k,Γ).
    Eqs. (1)–(2) and Sec. IV.A; assumes harmonic response and that ML band structures capture deformation potentials.
  • domain assumption Lattice reconstruction is fully captured by the in-plane displacement field u(r) plus out-of-plane corrugation h(r) obtained from energy minimization.
    Sec. III.A; standard continuum picture but required for the definition of χ.
  • domain assumption Off-diagonal interband EPC matrix elements are negligible near the isolated first moiré band at M.
    Stated in Sec. IV.A to justify using only the diagonal g(M,Γ).
  • standard math Harmonic approximation for phonons remains valid after strong lattice reconstruction.
    Used to obtain dynamical matrices from second derivatives of the ML potential (Sec. III.B).
invented entities (1)
  • pattern-matching parameter χ no independent evidence
    purpose: Scalar projection of a phonon eigenvector onto the static reconstruction displacement field; used to predict which modes have large EPC without computing every g.
    Defined in Eq. (3); introduced by the paper as the quantitative embodiment of the interference selection rule. Independent evidence is partial: it correlates with computed λ and with Raman-active modes, but χ itself is not measured directly.

reviewed 2026-07-14 · how reviews work

0 comments
Cite this review

Pith. "Pith review of Interference-Enhanced Large Electron-Phonon Coupling from Raman-active Breathing Modes in Moir\'e Semiconductors." pith.science (2026). https://pith.science/paper/EEPSKJON

@misc{pith2026260711425,
  author       = {Pith},
  title        = {Pith review of: Interference-Enhanced Large Electron-Phonon Coupling from Raman-active Breathing Modes in Moir\'e Semiconductors},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/EEPSKJON}},
  note         = {Machine review of arXiv:2607.11425}
}
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Superconductivity was recently observed in twisted WSe2 and MoTe2, raising a central question: is the pairing driven by electronic correlations, by phonons, or by both? Answering it requires determining the electron-phonon coupling (EPC) in these moir\'e semiconductors, whose calculation in realistic supercells of thousands of atoms lies beyond the reach of direct first-principles methods. Here we combine filling-dependent Raman spectroscopy with machine-learning first-principles calculations to obtain the EPC mode by mode in supercells of up to tens of thousands of atoms. Raman reveals only a few moir\'e phonons whose frequencies shift strongly with filling; we trace this to an interference selection rule: a phonon couples strongly only when its displacement texture matches the static lattice-reconstruction pattern, and is otherwise suppressed by destructive interference. The rule selects the low- and high-frequency breathing modes seen in Raman and makes the coupling peak at large twist angles, near those at which superconductivity appears. Lattice-reconstruction interference thus emerges as an organizing principle for moir\'e EPC, pointing to a substantial, potentially dominant, phonon contribution to large-angle pairing.

Figures

Figures reproduced from arXiv: 2607.11425 by Cheng Xu, Claudia Felser, Kenji Watanabe, Ning Mao, Shaozheng Wang, Shengwei Jiang, Takashi Taniguchi, Xumin Chang, Yang Zhang.

Figure 1
Figure 1. Figure 1: FIG. 1 [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: g for θ = 1.89◦ . C. Breathing and shear modes While the eigenvalues are largely inherited from the bilayer phonons, the eigenvectors carry the distinctive signature of the moir´e reconstruction. A prominent ex￾ample is the LBM, which characterizes the relative vibra￾tion of the two layers along the out-of-plane z-direction. In pristine bilayers, the LBM manifests as a rigid, out-of￾phase oscillation where… view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3 [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4 [PITH_FULL_IMAGE:figures/full_fig_p007_4.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

47 extracted references · 3 linked inside Pith

  1. [1]

    Y. Xia, Z. Han, K. Watanabe, T. Taniguchi, J. Shan, and K. F. Mak, Nature637, 833 (2025)

    2025

  2. [2]

    Y. Guo, J. Pack, J. Swann, L. Holtzman, M. Cothrine, K. Watanabe, T. Taniguchi, D. G. Mandrus, K. Barmak, J. Hone,et al., Nature637, 839 (2025)

    2025

  3. [3]

    Y. Guo, J. Cenker, A. Fischer, D. Mu˜ noz-Segovia, J. Pack, L. Holtzman, L. Klebl, K. Watanabe, T. Taniguchi, K. Barmak,et al., Nature652, 622 (2026)

    2026

  4. [4]

    Y. Xia, Z. Han, J. Zhu, Y. Zhang, P. Kn¨ uppel, K. Watan- abe, T. Taniguchi, K. F. Mak, and J. Shan, Preprint at https://doi.org/10.48550/arXiv.2508.02662 (2025)

    Pith/arXiv arXiv doi:10.48550/arxiv.2508.02662 2025

  5. [5]

    F. Xu, Z. Sun, J. Li, C. Zheng, C. Xu, J. Gao, T. Jia, K. Watanabe, T. Taniguchi, B. Tong,et al., Preprint at https://doi.org/10.48550/arXiv.2504.06972 (2025)

    Pith/arXiv arXiv doi:10.48550/arxiv.2504.06972 2025

  6. [6]

    Z. Sun, F. Xu, J. Li, Y. Jiang, J. Gao, C. Xu, T. Jia, K. Cheng, J. Zhang, W. Tian,et al., Preprint at https://doi.org/10.48550/arXiv.2603.16412 (2026)

    doi:10.48550/arxiv.2603.16412 2026

  7. [7]

    Schrade and L

    C. Schrade and L. Fu, Phys. Rev. B110, 035143 (2024)

    2024

  8. [8]

    Tuo, M.-R

    C. Tuo, M.-R. Li, Z. Wu, W. Sun, and H. Yao, Nat. Commun.16, 9525 (2025)

    2025

  9. [9]

    Qin, W.-X

    W. Qin, W.-X. Qiu, and F. Wu, Phys. Rev. Lett.135, 246002 (2025)

    2025

  10. [10]

    Chubukov and C

    A. Chubukov and C. Varma, Phys. Rev. B111, 014507 (2025)

    2025

  11. [11]

    Zhu, Y.-Z

    J. Zhu, Y.-Z. Chou, Y. Huang, and S. Das Sarma, Phys. Rev. B112, L020507 (2025)

    2025

  12. [12]

    Fischer, L

    A. Fischer, L. Klebl, V. Cr´ epel, S. Ryee, A. Rubio, L. Xian, T. O. Wehling, A. Georges, D. M. Kennes, and A. J. Millis, Phys. Rev. X15, 041055 (2025)

    2025

  13. [13]

    F. Wu, A. H. MacDonald, and I. Martin, Phys. Rev. Lett. 121, 257001 (2018)

    2018

  14. [14]

    J. Cai, E. Anderson, C. Wang, X. Zhang, X. Liu, W. Holtzmann, Y. Zhang, F. Fan, T. Taniguchi, K. Watanabe,et al., Nature622, 63 (2023)

    2023

  15. [15]

    Y. Zeng, Z. Xia, K. Kang, J. Zhu, P. Kn¨ uppel, C. Vaswani, K. Watanabe, T. Taniguchi, K. F. Mak, and J. Shan, Nature622, 69 (2023)

    2023

  16. [16]

    H. Park, J. Cai, E. Anderson, Y. Zhang, J. Zhu, X. Liu, C. Wang, W. Holtzmann, C. Hu, Z. Liu,et al., Nature 622, 74 (2023)

    2023

  17. [17]

    F. Xu, Z. Sun, T. Jia, C. Liu, C. Xu, C. Li, Y. Gu, K. Watanabe, T. Taniguchi, B. Tong,et al., Phys. Rev. X13, 031037 (2023)

    2023

  18. [18]

    Christos, P

    M. Christos, P. M. Bonetti, and M. S. Scheurer, Phys. Rev. Lett.135, 046503 (2025)

    2025

  19. [19]

    Zhu, Y.-Z

    J. Zhu, Y.-Z. Chou, M. Xie, and S. Das Sarma, Phys. Rev. B111, L060501 (2025)

    2025

  20. [20]

    F. Xie, L. Chen, S. Sur, Y. Fang, J. Cano, and Q. Si, Phys. Rev. Lett.134, 136503 (2025)

    2025

  21. [21]

    J. Z. Lu, Z. Zhu, M. Angeli, D. T. Larson, and E. Kaxiras, Phys. Rev. B106, 144305 (2022)

    2022

  22. [22]

    X. Liu, R. Peng, Z. Sun, and J. Liu, Nano Lett.22, 7791 (2022)

    2022

  23. [23]

    Ramos-Alonso, B

    A. Ramos-Alonso, B. Remez, D. Bennett, R. M. Fernan- des, and H. Ochoa, Phys. Rev. Lett.134, 026501 (2025)

    2025

  24. [24]

    Maity, M

    I. Maity, M. H. Naik, P. K. Maiti, H. Krishnamurthy, and M. Jain, Phys. Rev. Res.2, 013335 (2020)

    2020

  25. [25]

    L. P. Krisna and M. Koshino, Phys. Rev. B107, 115301 (2023)

    2023

  26. [26]

    Pizzi, S

    G. Pizzi, S. Milana, A. C. Ferrari, N. Marzari, and M. Gibertini, ACS Nano15, 12509 (2021)

    2021

  27. [27]

    Lin, Q.-H

    M.-L. Lin, Q.-H. Tan, J.-B. Wu, X.-S. Chen, J.-H. Wang, Y.-H. Pan, X. Zhang, X. Cong, J. Zhang, W. Ji,et al., ACS Nano12, 8770 (2018)

    2018

  28. [28]

    Bathen, T

    N.-L. Bathen, T. Deilmann, A. Senki´ c, H. Lambers, R. Dana, K. Watanabe, T. Taniguchi, F. M. Ross, J. Klein, and U. Wurstbauer, Nano Lett.26, 3745 (2026)

    2026

  29. [29]

    Koshino and Y.-W

    M. Koshino and Y.-W. Son, Phys. Rev. B100, 075416 (2019)

    2019

  30. [30]

    Xie and J

    B. Xie and J. Liu, Phys. Rev. B108, 094115 (2023)

    2023

  31. [31]

    H. Li, Z. Tang, J. Fu, W.-H. Dong, N. Zou, X. Gong, W. Duan, and Y. Xu, Phys. Rev. Lett.132, 096401 (2024)

    2024

  32. [32]

    Zhong, S

    Y. Zhong, S. Liu, B. Zhang, Z. Tao, Y. Sun, W. Chu, X.- G. Gong, J.-H. Yang, and H. Xiang, Nat. Comput. Sci. 4, 615 (2024)

    2024

  33. [33]

    N. Mao, C. Xu, J. Li, T. Bao, P. Liu, Y. Xu, C. Felser, L. Fu, and Y. Zhang, Commun. Phys.7, 262 (2024)

    2024

  34. [34]

    Y. Jia, J. Yu, J. Liu, J. Herzog-Arbeitman, Z. Qi, H. Pi, N. Regnault, H. Weng, B. A. Bernevig, and Q. Wu, Phys. Rev. B109, 205121 (2024)

    2024

  35. [35]

    Zhang, C

    X.-W. Zhang, C. Wang, X. Liu, Y. Fan, T. Cao, and D. Xiao, Nat. Commun.15, 4223 (2024)

    2024

  36. [36]

    J. Liu, Z. Fang, H. Weng, and Q. Wu, npj Comput. Mater.11, 248 (2025)

    2025

  37. [37]

    Z. Qi, H. Pi, Y. Zhang, J. Liu, N. Regnault, H. Weng, B. A. Bernevig, J. Yu, and Q. Wu, Phys. Rev. B113, 125116 (2026)

    2026

  38. [38]

    S. Yang, J. Chen, C.-F. Liu, and M. Chen, Phys. Rev. B 110, 235410 (2024)

    2024

  39. [39]

    T. Bao, N. Mao, W. Duan, Y. Xu, A. Del Maestro, and Y. Zhang, Preprint at https://doi.org/10.48550/arXiv.2501.12452 (2025)

    Pith/arXiv arXiv doi:10.48550/arxiv.2501.12452 2025

  40. [40]

    B. Lian, Z. Wang, and B. A. Bernevig, Phys. Rev. Lett. 122, 257002 (2019)

    2019

  41. [41]

    Ochoa, Phys

    H. Ochoa, Phys. Rev. B100, 155426 (2019)

    2019

  42. [42]

    Birkbeck, J

    J. Birkbeck, J. Xiao, A. Inbar, T. Taniguchi, K. Watan- abe, E. Berg, L. Glazman, F. Guinea, F. von Oppen, and 9 S. Ilani, Nature641, 345 (2025)

    2025

  43. [43]

    Kresse and J

    G. Kresse and J. Hafner, Phys. Rev. B47, 558 (1993)

    1993

  44. [44]

    Zhang, J

    L. Zhang, J. Han, H. Wang, R. Car, and W. E, Phys. Rev. Lett.120, 143001 (2018). 10 Appendix A: Methods

    2018

  45. [45]

    Atomically thin flakes of WSe 2, h-BN, graphite and TaSe 2 were exfoliated from bulk crystals onto Si/SiO 2 substrates with a 285 nm oxide layer

    Sample fabrication Twisted WSe2 devices were fabricated using a standard tear-and-stack dry-transfer method. Atomically thin flakes of WSe 2, h-BN, graphite and TaSe 2 were exfoliated from bulk crystals onto Si/SiO 2 substrates with a 285 nm oxide layer. Graphite flakes with thicknesses of 3–5 nm were used as top and bottom gates, while few-layer TaSe 2 f...

  46. [46]

    Raman measurement Raman spectra were acquired in backscattering geometry using a home-built confocal microscope integrated with a closed-cycle helium cryostat (attoDRY 2100, base temperature 1.7 K). Excitation was derived from a 532 nm solid- state laser, with the beam focused along the crystallographicc-axis onto the dual-gated t-WSez devices through a l...

  47. [47]

    We employ Projector Augmented Wave (PAW) potentials, treating six valence electrons for the W, Se, Te, and Mo atoms

    Machine-learning force field for phonons A second-stage structural relaxation is performed using the Vienna Ab initio Simulation Package (VASP) [43]. We employ Projector Augmented Wave (PAW) potentials, treating six valence electrons for the W, Se, Te, and Mo atoms. The calculations use a plane-wave energy cutoff of 300 eV, and Brillouin-zone sampling is ...

This paper was first reviewed by grok-4.5 on July 14, 2026.