REVIEW 4 major objections 5 minor 8 references
Probing Phonon Modes in Reconstructed twisted Homo and Hetero Bilayer System
T0 review · 4 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash
Pith's one-line read In reconstructed twisted bilayer graphene and graphene/hBN stacks, each Raman G-band satellite can be traced to a specific stacking region, making the twist-dependent peak pattern a direct readout of lattice reconstruction.
desk verdict Systematic Raman sweep of moiré phonons in TBLG and Gr/hBN with plausible stacking-region attribution, but the inferred twist-angle axis and qualitative theory match will need scrutiny before the regime boundaries can be trusted. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The load-bearing object is the spectral function $A^{E_g}_0(\omega)=\frac{1}{\pi}\sum_\nu \frac{|\langle\Psi^M_{0\nu}|\Psi^{E_g}_0\rangle|^2\Gamma_{0\nu}}{(\omega-\omega_{0\nu})^2+\Gamma_{0\nu}^2}$, with $\Gamma_{0\nu}=1$ cm$^{-1}$, which converts calculated moiré phonon eigenvalues $\omega_{0\nu}$ and eigenvectors $\Psi^M_{0\nu}$ into a predicted Raman line shape by projecting onto the bilayer-like $E_g$ mode $\Psi^{E_g}_0$ in the moiré unit cell. Alongside it, the paper classifies the relaxed structure by interlayer-spacing landscape into stacking regions: AB, AA, SP1, and SP2 for TBLG, and AB, AB', AA, and SP for Gr/hBN, then computes region-wise projection weights. This machinery turns a large phonon eigenvector problem into a small set of region-labeled peaks that can be compared with Lorentzian fits of the measured spectra.
What would settle it
A direct test: perform tip-enhanced or nano-Raman mapping on a 0.65-degree TBLG sample and check whether the four fitted G components localize to the predicted regions, with the 1673 and 1709 cm$^{-1}$ features in SP1-like soliton domains and the 1679/1702 cm$^{-1}$ pair in AB+SP2 domains. If the four components are spatially uniform or carry different region labels, the central attribution is refuted; alternatively, recomputing the spectral function with ab initio electron-phonon matrix elements and mode-dependent linewidths could shift or eliminate the four-peak structure.
Extended reading notes
Core claim
On the paper's own terms, the central discovery is a stacking-region-resolved account of moiré phonons. For a 0.65-degree TBLG, the relaxed interlayer-spacing landscape contains AB, AA, SP1, and SP2 regions; projecting moiré phonon eigenvectors onto the bilayer-like $E_g$ mode produces four spectral features near 1673, 1679, 1702, and 1709 cm$^{-1}$, assigned respectively to SP1, AB+SP2, AB+SP2, and SP1. This quadruplet matches the measured four-peak G spectrum at intermediate angles, and the angle evolution is explained by the shrinking of SP2 and then SP1: at 0.3 degrees the spectrum becomes three peaks, and at zero twist only the AB peak survives. For Gr/hBN, the same machinery yields three-peak spectra at small and large angles and five peaks at intermediate angles, with the central G peak immune to twist: the low-frequency satellite comes from AB', the high-frequency satellite from AA+AB, the middle peak from AB+SP, and an intermediate-angle-only fourth peak from SP. The claim is that these peak-to-region assignments convert Raman spectroscopy into a room-temperature probe of lattice reconstruction in both homo- and hetero-bilayer stacks.
Load-bearing premise
The argument's load-bearing premise is that the Raman intensity of a moiré phonon mode is proportional to how strongly that mode projects onto the bilayer-like $E_g$ vibration, with a single universal linewidth of 1 cm$^{-1}$ for every stacking region; if resonance or mode-dependent electron-phonon coupling actually controls the Raman spectrum, the peak-to-stacking-region mapping fails, even though the observation of multiple peaks would remain.
Editorial extensions
If this is right
- In TBLG, the number and position of G-band Raman components encode the twist angle: one peak below 0.2 degrees and above 2 degrees, up to four peaks between 0.3 and 1 degree, so Raman alone can bracket the reconstruction regime at room temperature.
- In Gr/hBN, the twist-invariant central G peak combined with splitting and merging of the M and M' satellites provides a hetero-specific Raman fingerprint of twist angle and lattice reconstruction.
- The disappearance of SP2 and then SP1 with decreasing twist angle explains why the multi-peak structure collapses back to a single peak, identifying stacking-region population as the controlling variable for phonon renormalization.
- Because the calculations identify which stacking regions contribute to each moiré phonon, the framework can guide design of stacks in which particular domain phonons dominate, with consequences for heat flow and phonon-assisted processes.
- Lattice reconstruction, rather than the bare moiré periodicity alone, drives the observed phonon renormalization in both homo- and hetero-bilayer systems.
Reading between the lines
- If the projection proxy survives more demanding checks, the same interlayer-spacing-region projection method could be applied to other reconstructed moiré systems, such as twisted transition-metal dichalcogenides, turning Raman peak counts into a general domain-meter.
- The strong region specificity of the modes suggests that nanoscale Raman mapping should reveal spatial localization of individual G components to specific stacking domains, a testable prediction the paper does not itself perform.
- The single phenomenological linewidth is likely the least secure point; allowing mode-dependent linewidths, for example from ab initio electron-phonon coupling, could shift fitted peak positions and should be checked against measured relative intensities.
- Angle-resolved thermal transport measurements could test whether the reconstructed stacking regions, with their distinct phonon frequencies, act as phonon scattering or confinement centers.
Signed reviews
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. This manuscript reports a combined Raman spectroscopy and classical force-field phonon study of reconstructed twisted bilayer graphene (TBLG) and graphene/hBN moiré superlattices. The authors classify TBLG into small, intermediate, and large twist-angle regimes based on the Raman G peak line shape, observing a single G peak at small angles, a three- and then four-peak structure in the intermediate range (~0.3–1°), two peaks at 1°, and a single peak at large angles. For Gr/hBN they observe a twist-angle-invariant central G peak with satellite M and M′ peaks that split from two to four (five total including G) in an intermediate range and recombine at larger angles. The theory uses LAMMPS relaxation and the PARPHOM phonon package to compute moiré phonon modes at the moiré Γ point, projects them onto bilayer-like E_g modes, and attributes observed features to stacking regions (AB, AA, SP1, SP2 for TBLG; AA, AB, AB′, SP for Gr/hBN) on the basis of region-wise projection weights. The paper concludes that the data establish a direct link between twist angle, lattice reconstruction, moiré phonons, and interlayer coupling.
Significance. The systematic coverage of multiple twist angles in both homo- and hetero-bilayer geometries is valuable and goes beyond earlier single-angle studies. The theoretical analysis is not fitted to the experimental peak positions: the eigenvalues and projections are computed independently from a classical force field, and the region-wise projections provide a concrete, falsifiable assignment scheme for the observed peaks. The explicit treatment of reconstructed interlayer separation landscapes and the AB/SP classification is a useful contribution. If the angle calibration and the peak-fitting statistics are strengthened, the qualitative picture—regime-dependent G splitting in TBLG and invariant G with angle-dependent satellites in Gr/hBN—would be a useful benchmark for future phonon engineering. However, the current evidence for the quantitative 'direct link' claim is limited by the absence of uncertainties on twist angles and fit parameters, the large offset between calculated and measured phonon frequencies, and the extrapolation of the theory to angles where it was not actually computed.
major comments (4)
- [Fig. 1a and SI S3 (twist-angle determination)] The TBLG twist angle is inferred as θ = |θ_L − θ_R| from the 2D FWHM at the two Gr/hBN interfaces, but no uncertainty or independent validation of this calibration is given. The entire assignment into small (≤0.2°), intermediate (0.3°–1°), and large (≥2°) regimes rests on these inferred angles, which are separated by only a few tenths of a degree. Please report the calibration scatter or uncertainty, propagate it into θ, and cross-check at least one device with a direct moiré-period measurement such as AFM/STM, transport, or nano-Raman mapping; otherwise the placement of a device into a given regime is not anchored.
- [Figs. 1b–c and 3a–c (peak fitting)] The spectral decomposition into one, three, four, or five Lorentzians is presented without per-peak parameter uncertainties, residual plots, number of devices or spot spectra, or an objective model-selection criterion. Since the central claims are the number of peaks and their evolution with twist angle, the fits need to be supported by error bars on positions and widths and by a statistical justification for adding peaks (for example an F-test or AIC/BIC comparison); otherwise the regime boundaries and the reported 'splitting' are difficult to evaluate quantitatively.
- [Spectral function paragraph and Figs. 2c, 4c (theory-experiment comparison)] The comparison between theory and experiment is only qualitative after an implicit shift: the calculated TBLG peaks at 1673–1709 cm⁻¹ lie roughly 90–130 cm⁻¹ above the experimental G peak, and the Gr/hBN calculated features near 1720–1750 cm⁻¹ are similarly far from the measured G/M/M′ complex, a discrepancy the text does not quantify or correct for. The text calls the TBLG offset 'slightly overestimated,' which materially understates the issue. In addition, the displayed spectral function assumes Raman intensity is proportional to the projection of each moiré mode onto the bilayer-like E_g mode, with a single phenomenological linewidth of 1 cm⁻¹, an assumption that is not tested against resonance or electron-phonon effects. Because the paper's strongest claim is the attribution of each observed peak to a specific stacking region, the authors should either validate the projection-intensity proxy with an independent calculation or explicitly present the attributions as a qualitative interpretation.
- [Theoretical evolution argument after Fig. 2c and SI S7] The evolution from the four-peak spectrum at 0.65° to the three-peak spectrum at 0.3° and the single peak at 0° is not computed; the paper states that such calculations are computationally expensive and instead infers the evolution from the twist-angle dependence of interlayer separation landscape region fractions. This is an extrapolation rather than a phonon calculation: the disappearance of SP2 and SP1 regions may change the number of contributing projections, but the peak positions and spectral weights at those angles are not verified. Please present calculations for at least one smaller angle, or explicitly present this part of the explanation as a conjecture rather than as a direct theoretical result.
minor comments (5)
- [Section 'Raman Spectra of Twisted Bilayer Graphene' (angle-regime definitions)] The regime definitions are inconsistent: the text gives 'small angle (θ ≤ 0.2°), intermediate angle (0.2° < θ < 1°), and large angle (θ ≥ 2°)', while the abstract and introduction use 0.3°–1° for the intermediate range, and no data or discussion covers 1° < θ < 2°; please harmonize the definitions and explain the gap.
- [Figs. 1d and 3d] These peak-position evolution plots have no error bars even though the positions are extracted from fits; error bars are needed for the reader to judge whether the apparent peak splitting and merging are statistically significant.
- [Data availability] The data availability statement says data are available 'upon request'; for reproducibility, consider depositing raw spectra, fitting code, and PARPHOM/LAMMPS input files in a public repository.
- [References] References 46 and 49 are the same paper (Kim et al., Phys. Rev. Lett. 108, 246103) and should be merged to avoid duplication.
- [Spectral function equation] The spectral function is presented as an unnumbered displayed equation; please number it, define all symbols in the text (Ψ_{0ν}^M, Γ_{0ν}, and the meaning of the subscript 0), and clarify how it is evaluated numerically.
Circularity Check
No significant circularity: theoretical moiré phonon peaks are computed independently and compared qualitatively to experiment.
full rationale
The paper's central derivation is self-contained against the experimental Raman data. Moiré phonon eigenvalues and eigenvectors are obtained by diagonalizing a dynamical matrix built from classical force fields in LAMMPS, via the PARPHOM package; the spectral function then weights each mode by its squared projection onto a bilayer-like Eg mode with a phenomenological 1 cm^-1 linewidth. The manuscript does not report any adjustment of the force-field parameters, projection weights, or linewidth to reproduce the measured G-peak positions, and the comparison is explicitly qualitative ('the model accurately captures the qualitative splitting of the G peak in intermediate angle regime'). The stacking-region attributions are read off from region-wise projection weights, not imposed by the experiment. The cited prior work by the same group (Refs. 54 and 56) supplies the computational tool and the higher-angle behavior already studied there; those are reproducible calculations, not restatements of the present spectra, so citing them does not make the argument circular. The TBLG twist angle is inferred as θ = |θ_L − θ_R| from 2D-peak FWHM calibration of the two Gr/hBN interfaces; that is an experimental uncertainty or accuracy concern, not a circularity, because the inferred angle is not defined through the G-peak splitting that the theory is meant to explain. No equation in the paper reduces a predicted quantity to an input by construction, and no fitted parameter is renamed as a prediction.
Assumptions & free parameters
free parameters (1)
- phenomenological linewidth Gamma0nu =
1 cm^-1
assumptions (4)
- domain assumption The classical force field (LAMMPS with the chosen interlayer potential) accurately describes lattice reconstruction and phonon eigenvalues/eigenvectors in twisted bilayers.
- domain assumption Raman intensity of moiré phonon modes is proportional to the squared projection onto the bilayer-like Eg mode with mode-independent matrix elements.
- domain assumption The evolution of the phonon spectrum from 0.65 degrees to 0.3 degrees and lower can be inferred from the interlayer separation landscape without direct phonon calculations.
- domain assumption The twist angle of the TBLG is determined as |left - right| from the FWHM of the Gr/hBN 2D peaks, which assumes the FWHM-angle calibration of Ref. 51 is valid at these small angles.
Cite this review
Pith. "Pith review of Probing Phonon Modes in Reconstructed twisted Homo and Hetero Bilayer System." pith.science (2026). https://pith.science/paper/RVKVHXHX
@misc{pith2026250619669,
author = {Pith},
title = {Pith review of: Probing Phonon Modes in Reconstructed twisted Homo and Hetero Bilayer System},
year = {2026},
howpublished = {\url{https://pith.science/paper/RVKVHXHX}},
note = {Machine review of arXiv:2506.19669}
}
read the original abstract
Twist angle engineering in van der Waals homo and hetero-bilayers introduces profound modifications in their electronic, optical and mechanical properties due to lattice reconstruction. In these systems, the interlayer coupling and atomic rearrangement strongly depend on the twist angle, leading to the formation of periodic Moire superlattices. At small twist angles, significant lattice relaxation results in the emergence of domain structures separated by one dimensional soliton networks, influencing electronic band structures and phonon modes. Here we systematically investigate the impact of lattice reconstruction on phonon renormalization in twisted bilayer graphene (TBLG,homo) and graphene-hBN Moire superlattices(hetero). Using Raman spectroscopy, we identify distinct phonon behaviours across different twist angle regimes. In TBLG, we observe the evolution of the G peak, including broadening, splitting, and the emergence of additional peaks in the small angle range 0.3 to 1 degree, attributed to Moire modified phonon interactions. At large twist angles, the peaks gradually merge back into a single feature, reflecting the reduced impact of lattice reconstruction. Similarly, in hBN graphene Moire superlattices, we detect Moire induced Raman peaks above and below the G peak, while the central G peak remains largely invariant to twist angle variation. The theoretical calculations uncover Moire phonon modes originating from different stacking regions providing insights into phonon renormalization. Our results establish a direct link between twist angle, lattice reconstruction, Moire phonons, and interlayer coupling, offering a fundamental framework for understanding phonon engineering in twisted bilayer systems. These findings pave the way for controlling phononic, optoelectronic and heat flow properties in next generation van der Waals heterostructures.
Reference graph
Works this paper leans on
-
[1]
The colour bar shows the squared modulus of the projection weights in (4b, 4c)
classified based on ILS in 4a. The colour bar shows the squared modulus of the projection weights in (4b, 4c). d. Schematic of contributions of different stacking regions to the peaks observed in different angle regimes. Unlike Gr/Gr, where there are three main stacking regions (AA, AB, and SP), the presence of boron and nitrogen atoms in the hBN sublatti...
1979
-
[6]
(53) Pizzocchero, F.; Gammelgaard, L.; Jessen, B
https://doi.org/10.1038/ncomms9429. (53) Pizzocchero, F.; Gammelgaard, L.; Jessen, B. S.; Caridad, J. M.; Wang, L.; Hone, J.; Bøggild, P.; Booth, T. J. The Hot Pick -up Technique for Batch Assembly of van Der Waals Heterostructures. Nat Commun 2016,
-
[7]
(54) Mandal, S.; Maity, I.; Krishnamurthy, H
https://doi.org/10.1038/ncomms11894. (54) Mandal, S.; Maity, I.; Krishnamurthy, H. R.; Jain, M. PARPHOM: PARallel PHOnon Calculator for Moir\’e Systems
-
[8]
(55) Thompson, A. P.; Aktulga , H. M.; Berger, R.; Bolintineanu, D. S.; Brown, W. M.; Crozier, P. S.; in ’t Veld, P. J.; Kohlmeyer, A.; Moore, S. G.; Nguyen, T. D.; Shan, R.; Stevens, M. J.; Tranchida, J.; Trott, C.; Plimpton, S. J. LAMMPS - a Flexible Simulation Tool for Particle -Based Materials Modeling at the Atomic, Meso, and Continuum Scales. Comput...
work page 2022
-
[271]
(56) Mandal, S.; Maity, I.; Krishnamurthy, H
https://doi.org/10.1016/j.cpc.2021.108171. (56) Mandal, S.; Maity, I.; Krishnamurthy, H. R.; Jain, M. Phonon Linewidths in Twisted Bilayer Graphene near the Magic Angle. Phys Rev B 2024, 110 (12), 125421. https://doi.org/10.1103/PhysRevB.110.125421. 19 Acknowledgements: This research was funded by an IISc start-up grant, Lakshmi Narayanan Young In vestiga...
arXiv 2021
-
[2020]
https://doi.org/10.1126/sciadv.abd1919. 15 (19) Klein, D. R.; Zondiner, U.; Keren, A.; Birkbeck, J.; Inbar, A.; Xiao, J.; Sidorova, M.; Al Ezzi, M. M.; Peng, L.; Watanabe, K.; Taniguchi, T.; Adam, S.; Ilani, S.; Paulson, J. A. Imaging the Sub -Moiré Potential Landscape Using an Atomic Single Electron Transistor
-
[2024]
https://doi.org/10.48550/arXiv.2410.22277 (accessed 2024 -10- 29). (20) Utama, M. I. B.; Koch, R. J.; Lee, K.; Leconte, N.; Li, H.; Zhao, S.; Jiang, L.; Zhu, J.; Watanabe, K.; Taniguchi, T.; Ashby, P. D.; Weber-Bargioni, A.; Zettl, A.; Jozwiak, C.; Jung, J.; Rotenberg, E.; Bostwick, A.; Wang, F. Visualization of the Flat Electronic Band in Twisted Bilayer...
-
[4534]
https://doi.org/10.1021/nl201370m. 17 (41) He, R.; Chung, T. F.; Delaney, C.; Keiser, C.; Jauregui, L. A.; Shand, P. M.; Chancey, C. C.; Wang, Y.; Bao, J.; Chen, Y. P. Observation of Low Energy Raman Modes in Twisted Bilayer Graphene. Nano Lett 2013, 13 (8), 3594 –3601. https://doi.org/10.1021/nl4013387. (42) Gupta, A. K.; Tang, Y.; Crespi, V. H.; Eklund,...
Reviewed August 15, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.