Assessment report: magnetophononic polaritons in atomically thin MnBi2Te4

Assessment report: magnetophononic polaritons in atomically thin MnBi2Te4

Paper: Enamul Haque et al., Highly Confined Magnetophononic Polaritons and Electron-Phonon Interactions in Atomically Thin MnBi2Te4, ACS Nano, DOI: 10.1021/acsnano.6c06828.
Assessment date: 29 October 2026.
Scope: outstanding issues only; concerns answered by the SI or corrected Figure S2 are omitted.
Material reviewed: the 17-page article; the supplied 34-page Supporting Information, nn6c06828_si_001-2.pdf; and the one-page replacement Figure S2, Fig.S2_of_nn6c06828_si_001.pdf. Main-article pages and SI PDF pages are identified separately throughout.

Expanded review: includes a detailed register of typos, minor technical and presentation mistakes, and reference corrections, with suggested replacements and confidence labels.

Scope and overall assessment

This report lists only concerns that remain after assessing the article, the complete supplied SI and the corrected Figure S2. Unless prefixed by SI, page references refer to the main article’s PDF pages 1-17 (printed A-Q). SI page references refer to its PDF pages 1-34. The supplied replacement Figure S2 is used as the current version of that figure.

The user also supplied a Monash repository record. Attempts to retrieve that record through the available web service failed, so this assessment does not certify its file inventory, version history or additional raw-data holdings. The locally attached SI and correction were successfully inspected. Missing-evidence statements below concern what remains undocumented in those reviewed materials, rather than an assumption that the SI is unavailable.

The paper presents a plausible research direction: infrared-active lattice modes in a magnetic layered material can support confined electromagnetic excitations, and lattice distortions can modify magnetic exchange. Its calculated phonon spectra, mode charges, exchange-displacement relations and electronic spectra provide useful starting points. However, the strongest conclusions require substantial clarification and validation. In particular, exchange modulation is treated as demonstrated magnetic switching; frozen electronic spectra are treated as evidence of a driven topological transition; and bulk electron linewidths are attributed to topological protection. There are also reproducible inconsistencies between printed equations, numerical parameters and plotted quantities.

Overall recommendation: major scientific revision or a detailed author response with reproducible calculations is warranted before relying on the quantitative confinement, pump-threshold, switching and topological-protection claims. This assessment does not establish that all underlying calculations are wrong, and it makes no finding of research misconduct.

The 53 numbered assessment issues use the following labels. The correction registers provide specific editorial actions; overlapping entries are not additional independent scientific defects.

• Direct inconsistency: identifiable from the article’s own text, equations or figures. This can be a reporting error rather than an implementation error.
• Unsupported inference: the displayed evidence does not establish the stated conclusion.
• Verification needed: implementation details, convergence evidence or additional numerical data are needed to decide the issue.

Highest-priority concerns

A. Magnetic switching, nonlinear driving and topology

1. An exchange sign change is not a demonstration of AFM-to-FM switching

Severity: major. Status: unsupported inference.
Location: abstract, pp. 1-3; Figure 4g, p. 8; switching discussion, pp. 10-12; conclusion, p. 13.

The calculation evolves lattice coordinates and evaluates exchange parameters. A positive effective interlayer exchange can favor FM alignment, but it does not establish that initially AFM Mn spins actually reorient on the available timescale. The response also depends on magnetic anisotropy, damping, temperature, initial perturbations and competing interactions. For perfectly collinear spins, changing an isotropic exchange coefficient alone need not generate an initial torque. The paper appropriately acknowledges on p. 12 that domain evolution and persistent switching are not modeled, but stronger claims remain elsewhere.

Required resolution: simulate the spin response to the calculated lattice trajectory, or consistently describe a transient distortion that favors FM order. Report the duration of that condition and whether a magnetic transition occurs within it.

Notes S4/S7 discuss local rectification and persistence, but neither they nor Note S8 evolves the Mn spins. Lattice dynamics and a static magnetic-force-theorem exchange evaluation remain distinct from spin switching.

2. Signed mean displacement and RMS amplitude are conflated

Severity: major. Status: direct inconsistency in definitions and interpretation.
Location: Figure 4f-g, p. 8; pp. 10-11; Methods, p. 14, Equation 11.

Equation 11 defines a nonnegative RMS amplitude, Q_RMS = sqrt(<Q(t)^2>). The Methods then add a sign, while the results use the quantity as a time-averaged structural displacement. An oscillation centered at zero has finite RMS but no rectified mean. Moreover, evaluating a nonlinear exchange function at a signed RMS is generally different from averaging exchange along the trajectory. For J(Q) = J0 + a Q + b Q^2, the average is J0 + a<Q> + b<Q^2>.

Required resolution: show signed means, RMS amplitudes, variances and averaging windows separately; calculate <J[Q(t)]> directly. Reassess the reported 2.94 Å sqrt(amu) and 2.09 MV/cm thresholds using those definitions.

Note S4 supplies a rectification argument, but the actual signed Raman trajectories and averaging windows are not given. Table S4 does not resolve the signed-RMS versus mean distinction.

3. Zero mean IR motion does not eliminate its quadratic exchange contribution

Severity: major. Status: direct inconsistency in reasoning.
Location: Figure 4c, p. 8; quadratic fits, p. 9; discussion of centered IR motion, p. 10.

The article argues that the IR coordinate oscillates around zero and therefore the transition can be accommodated only by the induced Raman breathing displacement. However, it also reports J(Q_IR) = J0 + alpha Q_IR^2. Consequently, <Q_IR> = 0 still produces Delta<J> = alpha<Q_IR^2>, which is generally nonzero. Its sign and size must be evaluated rather than excluded by the vanishing mean.

Required resolution: evaluate exchange along the combined IR and Raman trajectory, including relevant mixed terms, and quantify each contribution to the threshold.

Note S4 itself uses a nonzero average of squared IR motion to generate a Raman force. This strengthens the need to retain the corresponding quadratic direct-exchange contribution.

4. The simulated local field is not connected quantitatively to the confined polariton being advertised

Severity: major. Status: verification needed.
Location: Figure 4 caption, p. 8; pp. 11-12; light-driving Methods, p. 14.

The driving model uses near-zone-center ionic coordinates and an imposed electric field. This is a valid nonlinear-phononics calculation, but it does not by itself establish excitation of the highly confined finite-momentum polariton discussed in Figure 3. A spatially uniform far-field pulse does not supply the large in-plane momentum of that mode without a coupling mechanism. The Figure 4 caption explicitly says the conversion from normalized polariton amplitude to ionic displacement is not supplied.

Required resolution: specify whether 2.09 MV/cm is the incident or local field; give the geometry, momentum-coupling mechanism, field enhancement and relevant polariton mode profile. Establish which benefit arises specifically from confinement, beyond direct coherent IR-phonon excitation.

Note S4 explicitly restricts polaritonic character to a very narrow near-Gamma region, creating an additional tension with the finite-q confinement examples; see Issue 49.

5. Electronic topology is inferred from band shape and edge spectra without a reported electronic bulk invariant

Severity: major. Status: bulk topological diagnostic missing.
Location: main pp. 4 and 10-11; SI Figure S11, p. 25.

The phase classification is not supported by a reported electronic Chern number or another appropriate bulk invariant. Figure S11 identifies edge-localized branches, but edge localization does not replace that bulk diagnostic. Dirac-like curvature and edge localization alone do not demonstrate the claimed phase classification or robustness.

Required resolution: calculate the electronic invariant and bulk gap for each magnetic/displaced configuration, demonstrate convergence and identify any protecting symmetry. Identify the symmetry and invariant that protect the displayed edge branches.

6. A frozen distorted FM spectrum does not establish time-dependent topological switching

Severity: major. Status: unsupported inference.
Location: Figure 5e, p. 10; discussion, pp. 11-12.

The edge spectrum at a selected displacement near 3 Å sqrt(amu) is a conditional result for a frozen structure and imposed magnetic configuration. It does not show that the actual driven lattice and spins simultaneously reach that configuration, remain there sufficiently long, or produce the asserted conducting response. It also does not establish an adiabatic approximation throughout the pulse.

Required resolution: follow the bulk gap and invariant along the physically realized lattice/spin trajectory. Justify an instantaneous-band description, or calculate the nonequilibrium response if it is needed. Label a frozen-spectrum result as conditional until this connection is established.

Figure S11 remains a frozen-configuration calculation rather than a dynamically realized spin/lattice trajectory.

7. The phase labels and the meaning of edge protection need reconciliation

Severity: major. Status: verification needed.
Location: pp. 4 and 11; Figure 5.

The 2 SL FM phase is called a zero-plateau QAH phase, while 3 SL AFM is called QAH; the associated edge modes are described as symmetry protected. These descriptions require explicit Hall/Chern and symmetry information. Established even-layer AFM zero-plateau terminology cannot simply be transferred to an imposed FM bilayer. Conversely, the label alone is insufficient to prove an error because different zero-Hall mechanisms exist. The earlier work cited as reference 51 distinguishes AFM and FM thin-film cases using Chern classifications. Rodriguez-Vega et al. (2022).

Required resolution: tabulate magnetic order, gap, Chern number, any additional invariant, Hall response and edge-protecting symmetry for every phase discussed.

Figure S11 strengthens edge localization evidence but gives no electronic Chern number that resolves the phase labels.

8. Topological phonons do not automatically establish protected electromagnetic polaritons

Severity: major. Status: polariton protection remains unestablished.
Location: main p. 9; SI Figure S7b, p. 15; Figure S9/Notes S5-S6, pp. 18-20.

The underlying Wannier-center flow is not displayed alongside the binary Z2ph assignments. In addition, Notes S5-S6 do not fully specify the band subspace, integration manifold, symmetry constraints or invariant algorithm needed to reproduce and interpret the plotted quantity. SI reference 22 is a study of Z2 monopole charges of nodal-line semimetals; its applicability to this 2D phonon problem requires explanation. A localized phonon edge mode does not automatically establish a protected electromagnetic polariton or suppression of backscattering.

Required resolution: define and reproduce the invariant, identify the relevant gap/crossing and symmetry, and connect it to the coupled electromagnetic mode. Correct the figure references: the invariant is in actual Figure S7, whereas actual Figure S8 concerns hBN.

9. Strong or ultrastrong coupling is not quantified

Severity: moderate. Status: unsupported inference.
Location: abstract; pp. 9-13, especially the quantum-electrodynamics outlook on p. 13.

The terms strong light-matter coupling and potential strong-to-ultrastrong multimode coupling are used without a coupling rate, resolved hybridization, mode splitting, loss comparison or normalized coupling ratio for a specified system. A strong classical drive, large ionic response or exchange modulation is not itself a quantitative strong-coupling criterion. The p. 13 statement is appropriately prospective, but surrounding language can blur that distinction.

Required resolution: distinguish large driven nonlinear response from a demonstrated coupling regime. Quantify the latter for a defined photonic/emitter system, or retain it only as a possible future direction.

B. Polariton equations, numerical consistency and propagation

10. Independent oscillator curves need validation against the full multimode response

Severity: major. Status: verification needed.
Location: Equations 4-5 and Figure 3, pp. 6-7; Methods, p. 14.

Equation 4 displays an individual oscillator contribution, and Figure 3 treats the three branches separately. The Methods assert consistency with a mode-sum treatment but do not demonstrate it. Multiple oscillators can produce several collective branches, but their electromagnetic dispersion is determined by the total response; the roots need not equal curves obtained by inserting each oscillator separately. The cited reference 26 explicitly writes a summed conductivity. Juraschek and Narang, Equation 4.

Required resolution: show the total conductivity and solve its collective-mode condition. Compare with the independent-mode approximation and delimit the frequency intervals in which each approximation is accurate, including contributions from weaker IR modes.

The SI supplies no full multimode conductivity/pole calculation resolving this concern.

11. The long-wavelength sheet model needs a demonstrated finite-momentum validity range

Severity: major. Status: verification needed.
Location: Figures 2-3, pp. 3 and 6; Equations 1 and 4-6; pp. 7-9.

The article applies a local long-wavelength description to momenta reaching roughly 1-1.5 nm^-1 and to multilayer slabs. Small momentum relative to a Brillouin zone is not sufficient to establish that intrinsic screening and finite thickness are negligible. The cited theory identifies a sheet-limit condition q t << 1 and requires modified electrostatics as thickness or momentum increases. Rivera et al., few-layer discussion. The 19 meV example is also well above the second mode’s approximately 11.33 meV zone-center frequency.

Required resolution: report slab thickness and screening length, evaluate the relevant dimensionless parameters, and compare the model with direct finite-q calculations. Include nonlocal screening and finite-thickness effects wherever required.

Figure S6 provides finite-q phonons, while Note S4 limits polariton character near Gamma; neither supplies a quantified finite-thickness/nonlocal error bound for the Figure 3 extrapolation.

12. The conductivity denominator must distinguish the bare zone-center frequency from a Coulomb-shifted LO frequency

Severity: major if implemented incorrectly. Status: ambiguous formulation.
Location: Equation 4 and its explanation, p. 6; Equation 5, p. 7.

The text describes a conductivity based on LO phonon frequencies, alongside nonanalytic Coulomb-corrected dispersions. In a local susceptibility the denominator must use the appropriate frequency before the collective Coulomb shift; in a true 2D zone-center limit LO and TO coincide. Inserting an already Coulomb-shifted finite-q LO frequency and then generating the collective mode can count that interaction twice. The cited theory explicitly cautions against this. Rivera et al., susceptibility derivation.

Required resolution: state exactly which frequency is used at every stage. If the implementation uses only the common Gamma-point frequency, explain this clearly; the notation alone does not prove double counting occurred.

Note S4 discusses the zone-center limit but does not unambiguously identify every frequency actually inserted into the susceptibility.

13. Table 2 and Equations 5-6 do not reproduce the reported dispersion example

Severity: major. Status: direct internal numerical inconsistency under the printed model.
Location: Table 2, p. 4; Equations 5-6 and the 19 meV example, p. 7; Figure 3a.

Combining the printed equations in the lossless single-mode limit gives

q = (E^2 - E0^2) / (2 E0 hbar v_g),     E = hbar omega.

For the 1 SL second mode, Table 2 gives E0 = 11.33 meV and v_g/c = 8.7e-5. Using hbar c = 197326.98 meV nm, the result at E = 19 meV is q = 0.598 nm^-1, while the 1 SL red curve in Figure 3a is visually near 0.3 nm^-1. For the 1 SL first mode, the same check at 20 meV gives about 3.16 nm^-1, while Figure 3a is near 1.5 nm^-1. The text’s separate 0.32 nm^-1 example does not clearly identify thickness; it is close to the value obtained from the printed parameters for 2 SL AFM, so it cannot independently be treated as a contradiction without clarifying that assignment.

Required resolution: reconcile the equations, velocity normalization, oscillator strength and plotted data. The text should also identify which thickness its 19 meV numerical example refers to. This is an inconsistency of the published presentation, not a determination of which underlying quantity is wrong.

14. The plotted propagation quality factors appear inconsistent with the stated damping convention

Severity: major. Status: direct inconsistency between the printed model and visually read plots.
Location: Equation 4, p. 6; Table 2, p. 4; Figure 3c.

For the individual oscillator exactly as printed, Q = Im(sigma)/Re(sigma) above resonance is

Q = (E^2 - E0^2) / (2 gamma_E E),     gamma_E = hbar gamma.

At 20 meV, using the 1 SL damping value 0.09 meV, this gives approximately 100.5, 75.5 and 36.7 for the first, second and third modes. Figure 3c visually approaches about 200 and 150 for the first two modes, suggesting a factor near two. A linewidth convention could explain this, but the definitions and plotted calculation must then be made consistent.

Required resolution: provide the numerical curve data and define whether each gamma is an amplitude-decay rate, HWHM, FWHM or population-decay rate. Recalculate the figures or correct the equations/table as appropriate.

15. The polariton wavelength formula and quoted numerical value are inconsistent

Severity: moderate. Status: confirmed mathematical error.
Location: 19 meV example, p. 7.

The article prints lambda_effective = 2/(pi k), followed by approximately 18 nm for momentum 0.32 nm^-1. For the conventional angular wavevector used in q/q0, the relation is lambda = 2 pi/q, giving 19.63 nm at 0.32 nm^-1. The printed formula instead gives approximately 1.99 nm. An 18 nm wavelength corresponds to approximately 0.349 nm^-1. The free-space wavelength at 19 meV is approximately 65.25 micrometers and is consistent with the text.

Required resolution: correct the expression, identify the precise wavevector and use a consistent convention in all confinement and wavelength comparisons.

16. The stated dependence of wavelength on mode effective charge is reversed

Severity: moderate. Status: direct algebraic inconsistency.
Location: p. 7, discussion following Equation 6.

The article states that a larger mode charge gives a shorter wavelength. At fixed frequency and fixed remaining parameters, its Equation 5 instead gives q proportional to 1/|Z_n|^2, hence lambda proportional to |Z_n|^2. Comparisons across different modes also change resonance frequencies, so they cannot establish the opposite charge dependence without controlling those variables.

Required resolution: correct the fixed-frequency explanation and distinguish it from comparisons at fixed momentum or among different resonances.

17. The discussion of resonance, conductivity and low dissipation conflicts with Equation 4

Severity: moderate. Status: direct inconsistency in the model’s interpretation.
Location: p. 7, discussion of conductivity and s-SNOM detection.

The text describes a sharp conductivity fall near resonance and low dissipation near the LO frequency. In the displayed local oscillator, at the zone-center resonance the real conductivity is maximal, the imaginary part crosses zero, and the propagation ratio Im(sigma)/Re(sigma) vanishes. A fall on one side of a peak must be distinguished from the peak itself. The equation also makes momentum grow with detuning rather than diverge at the zone-center resonance.

Required resolution: specify which resonance, frequency side and loss measure are intended; reconcile the explanation with the actual complex response and total multimode calculation.

18. The damping values and resulting long-propagation claim lack a documented derivation

Severity: major. Status: verification needed.
Location: Table 2, p. 4; Equation 4, p. 6; p. 7; outlook, pp. 13-14.

A single damping value is supplied for each structure, although the claims involve several branches and momenta. The supplied Methods do not explain how these linewidths were calculated, at which temperature, or which decay mechanisms they include. The outlook lists three-phonon lifetime calculations as future work. Harmonic phonon calculations alone do not determine anharmonic lifetimes. An electron-state linewidth is also not the phonon linewidth required here.

Required resolution: identify whether gamma is calculated, fitted or assumed; provide its mode, momentum and temperature dependence or justify the approximation. Show sensitivity of propagation quality and pump response to plausible damping uncertainty.

The SI adds a damping choice for the hBN benchmark, but no branch-resolved derivation of the MnBi2Te4 damping. Its hBN and lifetime conventions introduce further conflicts (Issues 44-45).

19. Propagation quality, distance and lifetime need consistent units and conventions

Severity: moderate. Status: reporting/interpretation gap.
Location: Equation 3 and Figure 3 caption, p. 6; Table 2; pp. 7 and 10.

The dimensionless ratio Q = Im(sigma)/Re(sigma) is described as a distance or as decay in wavenumber. For a field proportional to exp(i q x), it equals Re(q)/Im(q) within the sheet model. The amplitude propagation length is 1/Im(q) and the intensity length is 1/[2 Im(q)]; in wavelengths these are Q/(2 pi) and Q/(4 pi). Thus Q = 100 is not automatically 100 propagated wavelengths. Likewise, gamma in meV is not literally an inverse time without hbar and a decay convention.

Required resolution: report distance in nanometers or micrometers, number of wavelengths, and amplitude/population lifetimes separately. The 13.3 ps Raman lifetime cited on p. 10 should not be transferred to a different IR mode without justification.

SI pp. 16-17 contain additional speed and lifetime inconsistencies; see Issues 43-44.

20. Record-confinement, detectability and dielectric-normalization claims need bounded comparisons

Severity: moderate. Status: verification needed.
Location: Table 2; pp. 7-9; EELS outlook, p. 13.

The assertion that the wavelength is shorter than any previously measured phonon polariton is not established by the few benchmarks cited. s-SNOM/EELS feasibility also depends on wavevector access, spectral resolution, signal strength and loss, rather than confinement alone. For multilayers, Table 2’s dielectric constants and polarizability in Å^2 require an explicit definition and conversion to the sheet response; supercell dielectric constants depend on vacuum normalization. No demonstrated device environment is attached to these predictions.

Required resolution: use like-for-like benchmarks at specified frequency, thickness and loss; avoid an unverified record claim. Provide measurable spectra or a realistic detection estimate, and state the dielectric/polarizability normalization and substrate assumptions.

The SI identifies a 2024 hBN study with confinement around 4000 in its reference list. Comparisons should include the appropriate conditions, and should not treat the lower selected hBN benchmark as a universal limit.

C. Electron-phonon scattering and transport interpretation

21. Bulk band-edge linewidths are incorrectly attributed to topological protection

Severity: major. Status: unsupported causal inference.
Location: Figures 6-7, pp. 11-12; p. 13.

The paper explains reduced VBM/CBM scattering in 3 SL by saying these states are topologically protected. The quantities shown are periodic bulk electronic states, not the chiral edge channels associated with a QAH phase. A topological bulk classification does not generally protect every bulk band-edge state against electron-phonon scattering. Thickness and magnetic order also change the dispersion, phonon branches, matrix elements and available final states.

Required resolution: describe the result as a correlation unless causality is isolated. Calculate edge-state scattering or compare a controlled topological transition while accounting for ordinary changes in matrix elements and phase space.

The additional band/transport plots do not isolate topology as the cause of bulk linewidth changes.

22. Imaginary electron self-energy does not establish the claimed phonon or electronic renormalization

Severity: major. Status: unsupported interpretation and incomplete implementation detail.
Location: main p. 13; SI Figure S13/Note S9, p. 28.

For the gap changes reported in Figure S13/Note S9, the reviewed methods do not document the real Fan/Debye-Waller calculation sufficiently to reproduce these shifts, and Note S9 repeatedly treats gap shifts as a measure of scattering rate. Real energy shifts and imaginary linewidths are different observables. The assertion that no 0 K renormalization is expected also neglects the possibility of zero-point renormalization; see Issue 50.

The claimed changes in phonon resonance/damping and oscillator strength still do not follow from the reported imaginary electron self-energy. They require the corresponding phonon-response calculation. EPW theory.

Required resolution: document the real and imaginary self-energy treatments separately, including zero-point/Debye-Waller contributions, and calculate each claimed phonon consequence rather than inferring it from electron linewidth.

23. Carrier scattering by a polariton-forming phonon is not automatically scattering by a normalized polariton

Severity: moderate. Status: interpretation gap.
Location: Figures 6-7; pp. 12-13; Figure 4 caption.

DFPT/EPW electron-phonon matrix elements quantify coupling to lattice modes. Calling those modes polariton forming is reasonable, but their phonon-only scattering rates are not automatically the scattering rates of a coupled electromagnetic normal mode with a specified occupation and phononic weight. The Figure 4 caption already recognizes an analogous normalization issue for polariton amplitude.

Required resolution: use precise phonon-scattering terminology, or project the interaction onto properly normalized polariton modes and state the mode composition and excitation conditions.

24. The written long-range interpolation procedure would double count if followed literally

Severity: major if implemented literally. Status: definite ambiguity in the reported algorithm.
Location: Methods, p. 14, Equations 13-14 and the following sentence.

Equation 13 uses the derivative of the full self-consistent DFPT potential. The text then says to add the analytical long-range term in Equation 14. The usual polar-interpolation procedure instead subtracts the analytical long-range part from the coarse-grid DFPT matrix, interpolates the short-range remainder, and restores the long-range term on the fine grid. Official EPW description.

Required resolution: state the subtraction/restoration procedure and actual program settings; benchmark interpolated matrix elements against direct DFPT. This wording problem does not prove that the actual EPW run double counted the interaction.

Notes S7-S8 describe nonlinear forces and exchange, not the long-range subtraction/restoration algorithm. This implementation question remains.

25. The printed long-range kernel is not reconciled with isolated-2D electrostatics

Severity: major. Status: verification needed.
Location: p. 14, 18 Å vacuum statement and Equation 14; Figure 7 caption.

Equation 14 has a bulk-style dielectric denominator, while the results are described as isolated-2D interpolation. A vacuum gap by itself does not document removal of slab-image electrostatic effects. The 2D Fröhlich interaction has different momentum-dependent screening, and a consistent implementation needs an appropriate kernel and boundary conditions. Sohier, Calandra and Mauri (2016).

Required resolution: give the actual 2D long-range expression, Coulomb truncation/boundary settings, code version and vacuum convergence. If a 2D implementation was used, correct or qualify the printed formula accordingly.

The supplied SI does not provide the missing 2D kernel, truncation settings or vacuum convergence.

26. Convergence of low-temperature scattering differences is not demonstrated

Severity: major. Status: verification needed.
Location: Methods, p. 14; approximately 15% AFM/FM difference, p. 13.

The Methods print the coarse q grids as “661” and “441”, apparently meaning 6 x 6 x 1 and 4 x 4 x 1, and cite a twice-as-dense interpolation grid. The grid dimensions should be confirmed and typeset explicitly. Their convergence statement concerns phonon accuracy, which is not sufficient to establish convergence of low-temperature, on-shell electron-scattering integrals. Final integration meshes, delta-function broadening and numerical errors at the exact VBM/CBM are not specified in the main paper. The 0.001 Ry SCF smearing is a separate parameter and cannot be assumed to describe the self-energy integration.

Required resolution: show joint mesh/broadening convergence and uncertainties substantially smaller than the claimed 15% effect. A modest coarse grid is not automatically invalid if interpolation and integration have been validated.

Convergence of low-temperature scattering integrals and validation of matrix elements across all structures remain undocumented; agreement of interpolated band dispersions alone does not establish them.

27. DFT+U consistency in the electron-phonon perturbation is unresolved

Severity: major. Status: explicit cross-document methodological inconsistency.
Location: main Figure 7 caption and Methods, pp. 12 and 14; SI Figure S7 caption, p. 15; corrected Figure S2.

The main article says Hubbard corrections enter electronic eigenvalues/wavefunctions. The corrected Figure S2 explicitly says its electronic-band comparison includes Hubbard corrections. In contrast, the actual SI Figure S7 caption explicitly states that Hubbard corrections were not included in those phonon/interpolation calculations, attributing this to implementation limitations with magnetism and SOC. The documents do not identify consistently which phonon/e-ph datasets share that omission.

Required resolution: provide a dataset-by-dataset map of functional, U, SOC, DFPT perturbation and interpolation settings. Specify inclusion/omission of the Hubbard-potential/projector derivatives and benchmark its effect. A corrected U-including electronic interpolation plot cannot validate a U-omitting phonon/e-ph perturbation. DFPT+U methodology.

28. Low temperature does not by itself make the long-range interaction weak

Severity: moderate. Status: incorrect general explanation.
Location: p. 12, explanation for the small PI-STDI difference.

Temperature changes phonon occupation and scattering phase space; it does not generally suppress the electrostatic Fröhlich vertex itself. Spontaneous emission can still occur for sufficiently energetic carriers, including states plotted away from the band edges in Figure 7. A small integrated correction at low temperature may be real, but it needs a specific explanation.

Required resolution: resolve the difference by phonon branch and momentum, and distinguish coupling strength from occupation and energy-conservation effects.

29. Single-particle linewidths do not establish interband dissipation or recombination

Severity: moderate. Status: unsupported inference.
Location: p. 13, claim about reduced dissipation from VBM to CBM; related application claims.

The calculated single-particle linewidth is not an interband transition or recombination rate. VBM-to-CBM motion is an excitation, and the reported bandgaps of 109-612 meV are much larger than individual phonon energies in the displayed spectrum. Any multiphonon, phonon-assisted optical or recombination pathway requires its own energy-conserving calculation and carrier population.

Required resolution: separate band-edge relaxation, interband excitation and recombination; calculate the claimed process or remove the causal statement.

30. DC conductivity and carrier lifetime do not establish screening of the THz pump

Severity: moderate. Status: unsupported optical-screening inference.
Location: main p. 13; SI Figure S12, p. 27.

The transport interpretation still requires the carrier density/chemical potential chosen for the claimed physical situation, the transport approximation and 2D sheet normalization, and numerical reliability near intrinsic low-temperature band edges.

DC conductivity does not directly establish screening of a 2.74 THz pump, which requires the complex frequency-dependent response, carrier population and geometry. Nor does a bulk band-edge lifetime establish protected edge conduction.

Required resolution: identify the operating chemical potential/density and transport units/convergence, and calculate the relevant THz response if screening is used to explain switching.

31. Artificially ordered 300 K results cannot establish actual high-temperature protected transport

Severity: moderate. Status: limitation acknowledged but not consistently carried into conclusions.
Location: p. 13; SI-band/transport discussion; high-temperature outlook.

The paper explicitly keeps the low-temperature magnetic order at 300 K, well above its stated ordering temperature near 20-25 K. This is a legitimate hypothetical comparison, but it does not establish the magnetic or topological state of the real material at that temperature. Magnetic disorder and possible loss of the relevant gap/protection are outside that calculation.

Required resolution: retain fixed-order results as conditional trends and constrain practical high-temperature claims. A prediction of real high-temperature performance requires a model of magnetic disorder or a mechanism stabilizing the assumed order.

32. Figure 6 does not fully define the scattering dataset or its mode comparisons

Severity: moderate. Status: reporting gap.
Location: Figure 6 and caption, p. 11.

The caption explains pink, red and yellow markers but leaves the abundant blue series undefined. It does not clearly explain the multiple points per phonon index, their electronic momenta, or whether the same electronic-state selection is used in every panel. Highest-occupied/lowest-unoccupied-band resolution is described explicitly only for panels c-d. Mode activity labels refer to zone-center symmetry, whereas a scattering integral samples finite-q modes whose character must be tracked.

Required resolution: define all colors, bands, momenta and aggregation rules. Support the statement that polariton-forming modes scatter more weakly than almost all other modes with integrated, consistently normalized near-edge contributions and mode tracking.

Tables S1-S2 now give activity labels, but contain internal contradictions requiring correction before mode-resolved comparisons are trusted; see Issue 47.

33. The linewidth-to-rate conversion omits hbar

Severity: minor to moderate. Status: confirmed dimensional omission.
Location: p. 12, relation between scattering rate and imaginary electron self-energy.

The paper writes tau^-1 = 2 Im(Sigma) while plotting self-energy in meV. In the on-shell approximation with unit quasiparticle weight, the corresponding rate is 2 |Im(Sigma)|/hbar. If quasiparticle weight is retained, it must be included consistently. EPW conventions.

Required resolution: state sign, energy-width and lifetime conventions, including any use of natural units, and verify every derived lifetime.

D. Model robustness, experimental benchmarks and reproducibility

34. The nonlinear potential needs validation over the actual driven trajectory

Severity: major. Status: dynamic validation incomplete.
Location: main Equation 9 and Figure 4; SI Note S7/Table S4, pp. 20-24.

The cubic potential needs force-fit validation over the complete driven trajectory, convergence against higher-order terms and omitted modes, and evidence of dynamical stability. The static bond changes and displacement comparisons in Note S7 do not establish those properties or an optical damage threshold. A cubic truncation is not globally bounded, so its local validity must be demonstrated over the sampled coordinates. The factor-of-two inconsistency between the SI equations of motion and the main potential also needs resolution (Issue 42).

Required resolution: reconcile the equations first, disclose the trajectory/integration settings and damping, and validate the fitted potential over all sampled coordinates with uncertainty in the threshold.

35. Structural and exchange uncertainties are large enough to require threshold sensitivity tests

Severity: major. Status: verification needed, with a visible baseline discrepancy.
Location: Figure 4c/g; Table 1; Methods, p. 14.

Figure 4c gives an equilibrium interlayer exchange near -0.0208 meV, while Figure 4g appears near -0.027 meV at zero displacement, a difference of roughly 30%. Total-energy mapping and magnetic-force-theorem extraction may explain it, but this must be reconciled before quoting a precise threshold. The Methods also do not specify a van der Waals treatment or show sensitivity to interlayer distance, U or exchange-correlation functional, although the mechanism depends on small exchange energies and interlayer geometry.

Required resolution: explain the two exchange baselines, reference structures and spin normalization. Propagate structural, functional and exchange-extraction uncertainties into the field threshold and topological classification; provide the relevant SI tests if they exist.

The two exchange baselines, numerical integration parameters and sensitivity of the threshold remain unreconciled in Note S8.

36. Circular driving does not generically produce a static shear rectification

Severity: moderate. Status: model-derived clarification needed.
Location: Equation 9 and symmetry discussion, p. 9; field profile, p. 10; Figure 4f.

For ideal circular IR motion, Qx = A sin(omega t) and Qy = A cos(omega t), the breathing force proportional to Qx^2 + Qy^2 has a nonzero cycle average. The shear forces proportional to Qx^2 - Qy^2 and 2 Qx Qy oscillate at twice the drive frequency and have zero cycle average. Finite pulses, asymmetry and nonlinear feedback can modify this result, but symmetry allowance alone does not establish rectified shear.

Required resolution: distinguish shear oscillation RMS from a signed DC displacement, and explain any nonzero shear mean from the actual trajectory.

37. The provenance of the claimed layer-resolved Raman benchmark needs checking

Severity: moderate. Status: source/benchmark verification needed.
Location: Figure 2e and caption, p. 3; p. 5; reference 27.

Figure 2e labels experimental comparisons as 2 SL and 3 SL MnBi2Te4. Reference 27 studies the MnBi2nTe3n+1 homologous family, with bulk crystals and a reported 10 nm MnBi2Te4 flake; its index n denotes composition/intercalation, not simply the number of MnBi2Te4 septuple layers. Its accessible primary text therefore does not clearly substantiate the specifically labeled 2 SL/3 SL benchmarks. Cho et al., primary paper.

Required resolution: identify the exact source figure/table, sample thickness and magnetic state for each experimental bar. If bulk or thicker-film frequencies were reused as an approximate benchmark, label them accordingly. Do not infer a sample-label error from the title alone; trace the actual values, including the source SI.

The new SI does not identify the source sample/figure for each experimental bar; the benchmark-provenance request remains.

38. The supplied material is insufficient for independent reproduction

Severity: moderate to major. Status: numerical reproducibility incomplete in reviewed files.
Location: main Data Availability, p. 14; SI Notes S7-S8 and Table S4; supplied repository link.

The reviewed PDFs do not provide an executable input/output package with structures, exact pseudopotential identifiers, versions, meshes/broadenings, Wannier windows, fitted force data, trajectories and numerical curve data. The Monash record could not be retrieved, so this report makes no claim that such data are absent from that repository.

Required resolution: identify the repository’s exact relevant files and connect each figure to reusable inputs and numerical outputs; report the outstanding calculation settings and convergence evidence.

E. Confirmed notation, symmetry and presentation issues

39. The space-group number and nonlinear selection rule contain errors

Severity: minor to moderate. Status: confirmed textual errors; computational consequences unverified.
Location: structure description, p. 3; symmetry discussion, p. 9.

The article identifies P-3m1 as number 166. P-3m1 is number 164; R-3m is number 166. The relevant slab structure and magnetic symmetry should be distinguished from the bulk setting. Primary monolayer structural calculation.

The nonlinear-coupling discussion also prints (Gamma_IR x Gamma_IR) x Gamma_IR contains A1g. Three inversion-odd IR factors have odd parity and cannot contain an even A1g invariant. The final factor should be the relevant Raman representation, as in Gamma_R x Sym^2(Gamma_IR). Equation 9 itself has the appropriate Raman-times-IR-squared form, so the displayed selection-rule error may be purely textual.

Required resolution: correct both statements and report the structural and magnetic symmetries actually used in the calculations.

Notes S1/S3 add magnetic-group claims that need a separate consistency check; see Issue 48.

40. Vibrational-representation notation, mode counts and figure labels need correction

Severity: minor to moderate. Status: confirmed notation/presentation defects, with some generalization needing verification.
Location: pp. 2-5; Figures 1-2; Equation 2.

The general vibrational-representation construction uses a direct sum where the site/vector construction ordinarily requires a tensor product. The displayed E-mode coefficients also mix multiplicity with the twofold dimension: for 1 SL the conventional notation is 3 A1g + 4 A2u + 3 Eg + 4 Eu, whose dimensions total 21; writing 6 Eg + 8 Eu as ordinary multiplicities would overcount. Equation 2 should use one convention consistently and state its applicable layer/stacking range.

The Figure 2 caption and text say five Raman modes, but Figure 2e displays six labeled groups. Figure 1’s caption locates PDOS on the left, whereas it is on the right. These mistakes reduce confidence in mode tracking and should be corrected against the actual eigenvectors/data.

Required resolution: provide an unambiguous symmetry/mode table, distinguish acoustic from optical modes, verify each branch index and correct the captions.

Tables S1-S2 add explicit mode assignments but introduce additional representation, acoustic-mode and degeneracy inconsistencies; see Issue 47.

41. Several citations and the Mn pseudopotential configuration need an explicit audit

Severity: minor to moderate. Status: confirmed citation mismatch / likely configuration typo.
Location: p. 5, p. 9, Methods p. 14; references 28, 38 and 57.

The phonon-topology statement cites reference 28, a crystal-growth/magnetic-structure study, rather than an evident methodology for Z2ph. A statement about experiments includes reference 38, which is a group-theory textbook. Reference 57 concerns spin anisotropy rather than directly establishing the stated comparison of magnetoelastic effects. These references should be matched to the precise claims they support. The detailed audit below also identifies publication-year inconsistencies, an incorrectly parsed author name in reference 20, and a mismatch between the named EV93 functional and reference 34.

The listed Mn valence configuration includes 4d^5, unusual for the Mn 3d^5 shell relevant here. This is likely a transcription error, but the actual pseudopotential file is needed to decide; it is not evidence that an incorrect shell was used computationally.

Required resolution: verify citations, supply primary references for the specific claims, and correct the Mn configuration against the actual pseudopotential metadata.

F. Detailed correction register: typos, minor mistakes and references

This register gives specific corrections suitable for an author or copy editor. Confirmed means that the printed problem is identifiable in the supplied PDF; it does not establish a corresponding error in the calculations. Clarify means the notation or description needs an explicit definition. Check means the intended correction requires the SI, original inputs or fuller source inspection. Copy edit denotes language improvements, not an independent scientific defect. Some entries expand Issues 15-19, 32-33 and 39-41 rather than adding new scientific concerns.

F1. Technical notation and small factual mistakes

F2. Figure, caption and cross-reference corrections

F3. Spelling, grammar and punctuation

The short replacements below preserve the intended meaning where possible.

F4. Reference and citation audit

Bibliographic corrections and citation-scope concerns are separated below. A valid background citation is not automatically an incorrect reference merely because it studies another material. The question is whether it supports the particular claim attached to it. The audit is targeted; it is not a certification that every field in all 68 reference entries has been checked.

F5. Editorial safeguards for applying the corrections

• Do not count ordinary PDF extraction artifacts, line-end hyphenation or the download watermark as manuscript typos. The grid strings, angular-momentum expression and selected caption/notation errors above were checked against rendered pages.
• Do not count the ACS Nano XXXX, XXX, XXX-XXX online-first footer as a scientific error; it can reflect the article’s publication stage.
• Do not silently change an unusual physical parameter or mode label as a copy edit. Check it against the calculation inputs, eigenvectors or SI first.
• Treat language-only entries as improvements in readability. Their number should not be used to inflate the number or severity of independent scientific problems.

G. Outstanding issues in the SI

42. The SI equations of motion do not follow the main-text potential with the same cubic coefficients

Severity: major. Status: confirmed equation-level inconsistency; implementation to verify.
Location: main Equation 9, p. 9; SI Equation S2, p. 21; Equation S3, p. 22; Table S4.

For the printed main potential term gamma_A Qa (Qx^2 + Qy^2), differentiation gives an IR force -2 gamma_A Qa Qx and a Raman force -gamma_A (Qx^2 + Qy^2). Equation S3 instead prints -gamma_A Qa Qx and -(gamma_A/2)(Qx^2 + Qy^2). Its Eg cross-coupling forces likewise correspond to half the main-text coefficient. Equation S2 uses the force relation consistent with the full main-text coefficient, so the difference cannot be dismissed without identifying which convention was fitted and integrated. The Raman self-anharmonic terms do not show the same global rescaling.

Required resolution: provide one potential, coefficient convention, fitted-force relation and differentiated equation set used by the code. Recompute the field-displacement and exchange threshold if the implemented coupling differs. This inconsistency directly affects a quantitative central claim.

43. The SI speed conversion is wrong by a factor of one million

Severity: moderate. Status: confirmed unit-conversion error.
Location: SI p. 16, paragraph below Figure S8.

The SI equates approximately 0.003-0.030 m/ps with 3-30 nm/ps. These differ by a factor of 10^6. From the stated v_g/c = 10^-5 to 10^-4, the correct order is 3 x 10^3 to 3 x 10^4 m/s, or 3-30 nm/ps, or 3 x 10^-9 to 3 x 10^-8 m/ps. The nm/ps values are consistent with the stated speed ratio; the m/ps values are not.

Required resolution: correct the units and recompute all propagation distances using a single consistent rate convention. These parameters are in main Table 2, not Table 1 as cited in this SI discussion.

44. The SI lifetimes are inconsistent with the tabulated energy damping and borrow a different mode’s lifetime

Severity: major. Status: confirmed convention/provenance conflict.
Location: SI p. 17; main Table 2; main/SI damped equations.

The SI claims average lifetimes about 26-45 ps and uses the measured 13.3 ps Raman breathing lifetime as a magnetophononic-polariton lifetime. For the main table’s energy damping 0.09-0.15 meV and an oscillator with 2 gamma Qdot, amplitude decay times hbar/gamma_E are about 7.31-4.39 ps, and population decay times are about 3.66-2.19 ps. The 26-45 ps range is near h/gamma_E, suggesting a possible angular-versus-cyclic-frequency convention mismatch, but the calculation files are needed to determine the cause. The table reference to main Table 1 also does not provide the stated lifetimes.

Required resolution: specify conversion, measured mode and population/amplitude convention, and justify the IR lifetime independently. A Raman-mode measurement alone does not establish the IR driver or polariton lifetime, persistence or switching window.

45. The hBN propagation benchmark uses a different damping from the main text

Severity: major for the claimed quantitative advantage. Status: confirmed input discrepancy.
Location: main p. 7; SI Figure S8 caption, p. 16.

The main comparison cites hBN damping about 0.14 meV. The SI explicitly computes its hBN curves using 0.49 meV, about 3.5 times larger. In the same single-oscillator convention, Q at fixed detuning scales inversely with damping, so this choice changes the apparent propagation advantage substantially. The discrepancy is separate from the factor-of-two Q inconsistency in Issue 14.

Required resolution: identify the temperature, mode and source for each linewidth and compare under matched conventions and conditions. Do not use an arbitrarily larger hBN loss to support an intrinsic superiority claim.

46. Table S3’s 3 SL AFM frequency block appears copied from the 2 SL FM column

Severity: moderate to major. Status: confirmed cross-table mismatch.
Location: SI Table S3, p. 12; Table S2, p. 11; main Table 2, p. 4.

Under 3 SL AFM, Table S3 lists indices 14, 24 and 34 with energies 6.08, 11.31 and 16.50 meV. Those match the main 2 SL FM column. Main 3 SL AFM instead uses indices 20, 35 and 50 with about 5.96, 11.22 and 16.25 meV; Table S2 gives about 5.96, 11.21 and 16.24 meV. Small rounding differences do not explain the wrong block and indices.

Required resolution: correct the apparent column-copy error against source data, and verify which dataset entered all polariton and mode-charge comparisons.

47. The SI mode tables contain contradictory acoustic, activity and degeneracy assignments

Severity: major for symmetry/mode identification. Status: confirmed tabular conflicts; correct assignments require eigenvectors.
Location: SI Tables S1-S2, pp. 9 and 11; Note S1/Figure S5, pp. 7-8.

Examples requiring a full table audit include:

• Table S1’s 2 SL FM zero-frequency rows are labeled A1g, A2u and Eg, rather than consistently representing the three rigid translations of the assumed centrosymmetric slab, A2u + Eu.
• In 1 SL, rows 17-18 are both assigned Eu but have 16.36 and 16.78 meV. This is not the claimed exact Gamma doublet, and neither simply reproduces the main third-mode entry 16.37 meV as a pair. State whether the table mixes finite-q LO data, different calculations or transcription errors.
• Table S1, 2 SL FM row 35, has R in the group-representation column rather than an irrep.
• Table S2, 3 SL FM row 55, marks an A2u mode Raman active, conflicting with the stated centrosymmetric selection rule.
• Note S1 calls 2 SL AFM modes 40 and 41 Raman active; Table S1 labels mode 40 A2u/IR and puts an A1g Raman mode at 39. Note S1 also mentions a 43rd mode in a 42-mode system.

Required resolution: regenerate frequencies, activity labels and indices from the same eigenvectors, separating translations from optically active vibrations. Until reconciled, do not use these tables as definitive evidence that the present work’s assignments are correct and an earlier paper’s assignments are wrong.

48. The SI magnetic-symmetry explanation does not resolve the mode-label contradictions

Severity: moderate to major. Status: conceptual inconsistency / verification needed.
Location: SI Notes S1-S3, pp. 7-13.

Note S1 says the three SLs are identical by symmetry, although inversion can exchange the top and bottom layers without making the middle layer equivalent to the outer layers. Note S3 says the 3 SL AFM/FM magnetic point group and symmetry operations remain the same, then attributes changed irreps to different unitary symmetry groups without providing a consistent full group/subgroup construction. Generic spin-phonon coupling does not by itself repair the acoustic and inversion-parity errors in Issue 47.

Note S2 also says nonanalytic corrections are essential to account for interlayer hybridization of nonpolar Eg modes, although interlayer short-range forces already enter the harmonic dynamical matrix. It does not separate genuine finite-q mode mixing from incorrect labeling or double counting.

Required resolution: supply actual structural/magnetic operations, their action on displacements, and mode projections before/after ordering changes. Distinguish frequency reordering and branch-index changes from a change of symmetry representation or optical activity.

49. The SI restricts polaritonic character in a way not reconciled with the main confinement claims

Severity: major. Status: cross-document physical-interpretation conflict.
Location: SI Note S4, p. 15; main Figure 3 and finite-q wavelength examples.

Note S4 states that for q much larger than a very small photon-phonon crossing momentum the mode becomes purely mechanical, and that polaritonic character persists only in an extremely narrow region near Gamma. The main article nevertheless attributes large finite-q confinement, propagation and switching benefits to polaritons. Phonon-dominated quasistatic modes can still be described as phonon polaritons, so the issue is not resolved by declaring the large-q mode either entirely photonic or entirely nonpolaritonic. The cited microscopic theory explicitly discusses confined 2D LO modes in this framework. Rivera et al..

Required resolution: define the polariton criterion, quantify the crossover and electromagnetic/phononic response, and apply the definition consistently to the actual q values advertised in the main text. Separate this question from the far-field momentum-coupling problem in Issue 4.

50. Zero-temperature band renormalization and scattering are misinterpreted in Note S9

Severity: moderate to major. Status: incorrect general expectation / missing implementation detail.
Location: SI Note S9/Figure S13, p. 28.

Note S9 calls the absence of a 0 K gap change expected. Zero-point lattice motion can produce a nonzero band-energy renormalization even when thermal phonon occupation vanishes; its absence is not a general consequence of zero temperature. The same note treats the magnitude of a gap shift as a direct measure of scattering strength and ambiguously uses 20 meV as a renormalized gap rather than clearly distinguishing the final gap from its shift.

Required resolution: report the bare gap, zero-point correction, finite-temperature correction and final gap separately, with the real Fan/Debye-Waller implementation. Explain any cancellation or approximation leading to negligible 0 K change. Keep energy shifts distinct from lifetimes. EPW band-renormalization conventions.

51. Static displacement and generic field ranges do not ensure dynamical stability or absence of optical damage

Severity: moderate. Status: overinterpretation of a supplied structural check.
Location: SI Note S7, pp. 22-24.

The static bond changes and displacement ratios do not ensure stability of the driven structure or an experimental damage threshold. The printed global RMS ratio also leaves an atom-indexed nearest-neighbor distance outside the averaged numerator, so its aggregation needs definition if denominators vary by atom. A consistent local-ratio RMS could be sqrt(mean_i[(u_i/d_NN,i)^2]), or a global RMS could use a defined common reference length.

The proposed universal field ranges for nonlinear phononics versus symmetry breaking are not material-independent thresholds. The claim that femtosecond pulses avoid ionization/damage also needs material-, frequency- and fluence-specific evidence. The SI itself acknowledges that absorption and heat flow lie outside its static calculation.

Required resolution: define the statistic and verify dynamic stability, fitted forces and pulse-energy/damage conditions; avoid presenting the static check as a guarantee.

52. Rectified force is incorrectly treated as proof of a latched, nonoscillatory Raman displacement

Severity: major for the persistence/switching argument. Status: unsupported dynamical inference.
Location: SI Note S4, p. 17.

The decomposition of squared IR motion into a DC component plus an oscillatory component is valid in its scalar example. It does not establish that the Raman mode stops oscillating and remains at a fixed displacement for 13.3 ps. A Raman oscillator responds with its own inertia, restoring force, damping and transient/free oscillations; a time-dependent rectified force does not remove them. The use of the Raman breathing lifetime as the IR/polariton lifetime further compounds the persistence estimate.

Required resolution: show signed Raman-coordinate time traces and directly averaged exchange after the pulse, using consistently derived equations and mode-specific damping. Establish any interval of nearly static distortion numerically rather than inferring it from the existence of a DC force.

53. SI bibliography and production mistakes need correction

Severity: minor to moderate. Status: confirmed production errors plus explicit source checks.
Location: SI contents/captions, Notes S1-S11 and references.

Quantitative checks that can be reproduced directly from the article

These checks use the printed equations and parameters; they are not reruns of the DFT calculations. Plot values are approximate visual readings and require the original numerical data for a definitive reconciliation.

Minimum evidence needed to resolve the main conclusions after SI review

1. Obtain reusable numerical inputs/outputs and the repository file inventory for the remaining implementation and convergence questions.
2. Reconcile main Equation 9 with SI Equations S2-S3 and their fitted coefficients; reconcile Equations 4-6, Table 2 and Figure 3 from a single documented set of numerical inputs and linewidth conventions; provide the full multimode response and its finite-q validity range.
3. Show signed lattice trajectories and directly averaged exchange, including quadratic IR contributions, with uncertainty in the threshold.
4. Establish the spin response during the interval of FM-favoring exchange, or limit the claim to that energetic tendency.
5. Provide electronic invariants along the relevant structural/magnetic path and distinguish conditional frozen spectra from an established driven conducting state.
6. Validate isolated-2D electrostatics, polar interpolation, DFPT+U consistency and low-temperature scattering convergence.
7. Separate bulk electron linewidths from protected edge transport, phonon damping, optical screening and recombination; calculate any claimed connection.
8. Identify/release the numerical inputs and outputs, and correct the mathematical, mode-table, unit, benchmark and citation issues listed above.

The evidence currently supports investigating a lattice-mediated route to modifying exchange and electronic spectra. Whether the quoted confined polaritons produce the claimed magnetic and topological switching at the stated field remains a quantitative question requiring the validations above.