Y Yin

Assessment report: MnBi2Te4 magnetophononic polaritons

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: 28 October 2026.
Material reviewed: the supplied 17-page article, including its figures, tables, equations, methods and references.

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 issues identifiable from the supplied article and selected checks against primary literature and official computational documentation. Page references below are PDF page numbers; the article labels these pages A-Q. The separate Supporting Information (SI), raw data and calculation inputs were not available for inspection. Attempts to access the SI at its embedded publisher link were unsuccessful. Therefore, a statement that evidence is missing means missing from the supplied main article, unless a direct internal inconsistency is identified. Some requests below may already be addressed in the SI and should be checked there before being presented as unresolved deficiencies of the complete study.

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 42 numbered assessment issues use the following labels. A separate correction register after Issue 42 expands the editorial and reference findings into individual actions; entries that overlap earlier issues are not additional independent scientific defects.

Highest-priority concerns

Concern Why it matters Issues
Magnetic switching is not simulated A favorable exchange sign does not show that spins reverse during the pulse 1-4
Electronic and polaritonic topology need stronger evidence Band shape, edge intensity and phonon topology do not independently establish the claimed protected driven states 5-8
Polariton electrodynamics need a consistent multimode, finite-momentum treatment The reported confinement and propagation depend on the response model and its range of validity 11-13
Equations, parameters and plots disagree Factors near two affect dispersion and propagation quality; the wavelength expression is incorrect 14-16
Electron scattering is overinterpreted Bulk linewidths do not demonstrate protected edge transport or phonon renormalization 22-24
Reproducibility is incomplete in the supplied material Small magnetic/scattering differences and precise thresholds need convergence and implementation evidence 25-28, 35-36

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.

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.

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.

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.

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

Severity: major. Status: unsupported inference / SI verification needed.
Location: electronic-band discussion, p. 4; Figure 5, p. 10; topology discussion, p. 11.

Dirac-like dispersion is not itself a topological diagnostic. In-gap edge spectral weight is useful evidence, but edge states can be accidental or termination dependent. The supplied article does not report an electronic Chern number or an equivalent bulk invariant for the relevant phases. The phonon invariant discussed separately cannot fill this role.

Required resolution: provide the electronic invariant for each magnetic and displaced structure, demonstrate its convergence and identify the bulk gap and protecting symmetry. Resolve the chirality and edge localization of the displayed 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.

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.

8. Topological phonons do not automatically establish protected electromagnetic polaritons

Severity: major. Status: unsupported inference / SI verification needed.
Location: p. 9; phonon-topology Methods, p. 14; SI Notes S5-S6 cited there.

The article moves from a phonon Z2ph invariant and phononic edge states to a claim about topological polaritonic modes and suppressed backscattering. A phonon-band invariant does not automatically characterize the electromagnetic response of the coupled, lossy system. Its physical meaning also depends on the relevant band subspace, gap or crossing, symmetry and boundary. A local degeneracy or Berry-phase charge is not by itself a general guarantee against scattering in all propagation directions.

Required resolution: inspect the SI definition of Z2ph, identify the protected states and frequency/momentum window, and show that coupling to the electromagnetic field preserves the required structure. Demonstrate the claimed scattering protection or narrow the statement to the phonon calculation.

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.

10. The novelty of the switching mechanism needs a more specific comparison with prior work

Severity: moderate. Status: evidence-based positioning issue.
Location: reference-51 discussion, p. 13; abstract and conclusion.

Reference 51 already investigated resonantly driven IR phonons, anharmonic Raman excitation, interlayer exchange reversal and topological changes in 2 SL MnBi2Te4. The present paper acknowledges it but calls its perspective completely different and refers criticism to SI Note S10. New polariton and electron-scattering results can be valuable; they do not by themselves make the underlying nonlinear-phononics switching mechanism new. Earlier primary study.

Required resolution: compare the mechanisms, IR symmetry, electrodynamic treatment, damping, topology diagnostics and quantitative predictions directly. Identify the additional result attributable to a polariton calculation. The unavailable SI comparison should be reviewed before judging the fairness of its criticism.

B. Polariton equations, numerical consistency and propagation

11. 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.

12. 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.

13. 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.

14. 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.

15. 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.

16. 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.

17. 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.

18. 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.

19. 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.

20. 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.

21. 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.

C. Electron-phonon scattering and transport interpretation

22. 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.

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

Severity: major. Status: unsupported inference.
Location: p. 13, paragraph discussing the approximately 15% change in electron self-energy.

The article infers strong changes in phonon resonance, phonon damping, oscillator strength, band edges and topology from a modest change in the imaginary electron self-energy. Electron linewidth is a different observable. Band-energy shifts require the real self-energy and appropriate Debye-Waller contribution; phonon shifts and damping require the phonon self-energy. Comparing an electron linewidth with a phonon energy does not supply those missing calculations. Official EPW theory.

Required resolution: calculate the quantities used in each inference and propagate them into the relevant spectra or invariant. Otherwise limit the conclusion to the calculated electron linewidth differences.

24. 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.

25. 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.

26. 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.

27. 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.

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

Severity: major. Status: verification needed.
Location: Figure 7 caption, p. 12; Methods, p. 14.

The caption says Hubbard corrections were used in eigenvalues and wavefunctions but not explicitly in interpolation. Interpolation need not add a second U correction when the input matrix elements are complete. However, omitting the displacement derivative of the Hubbard potential while using U-corrected electronic states is a different approximation. The main paper does not distinguish these possibilities. Primary DFPT+U methodology.

Required resolution: specify inclusion of the Hubbard-potential and projector derivatives, the relevant QE/EPW implementation, and consistency between the PBE/EV93 electronic calculations and the perturbing potential. Benchmark any approximation affecting the reported differences.

29. 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.

30. 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.

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

Severity: moderate. Status: unsupported inference / missing transport specification.
Location: conductivity and screening discussion, p. 13; Figure S12 cited there.

The article uses reduced electrical conductivity to infer weaker screening of incident light. Pump screening depends on the frequency-dependent complex response at the pump frequency, carrier density, chemical potential and geometry. At 20 K, intrinsic carrier populations in a gapped material are strongly suppressed, so a band-edge lifetime alone cannot determine conductivity. The article acknowledges uncertainty in the 20 K transport result, which further limits its interpretive weight.

Required resolution: specify carrier population, transport approximation and 2D sheet normalization, and calculate the THz response if screening is part of the proposed mechanism. Distinguish bulk conduction from edge conduction.

32. 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.

33. 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.

34. 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

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

Severity: major. Status: verification needed.
Location: Equation 9, p. 9; Figure 4e-g; Methods, p. 14.

The potential contains harmonic and cubic terms, while the plotted trajectories reach substantial mass-weighted amplitudes. A cubic truncation is not globally bounded; that does not invalidate a local expansion, but its validity must be demonstrated over every trajectory excursion. The main article does not show force residuals, higher-order sensitivity, stability against omitted symmetry-allowed terms or convergence with respect to additional coupled modes. A mass-weighted coordinate is also not a direct atomic displacement.

Required resolution: report actual maximum atomic displacements and bond changes, fitted energy/force ranges, time-step convergence, initial conditions, damping and pulse-width convention. Validate the trajectory and threshold against a stabilized higher-order potential where necessary.

36. 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.

37. 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.

38. 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.

39. The supplied material is insufficient for independent reproduction

Severity: moderate to major. Status: reporting/access gap.
Location: Methods and Data Availability, p. 14; SI description, pp. 14-15.

Raw data are available on request, but the main article does not provide a permanent calculation archive with exact relaxed structures, pseudopotential identifiers, software versions, force constants, mode normalization, Wannier projections/windows, exchange fits, nonlinear coefficients, time traces, numerical spectra and convergence results. Much may be described in SI, but descriptions alone do not replace reusable data and input files.

Required resolution: provide the SI and deposit the inputs, relevant outputs and analysis scripts with stable identifiers. At minimum, make Figures 3-7 and the threshold calculation reproducible from the deposited material.

E. Confirmed notation, symmetry and presentation issues

40. 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.

41. 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.

42. 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 16-20, 33-34 and 40-42 rather than adding new scientific concerns.

F1. Technical notation and small factual mistakes

ID Location Printed problem Suggested correction or action Status
T01 p. 2, Figure 1 caption The total angular momentum is given only as j = abs(l + s). This omits the lower-j branch, although the figure includes both p and d spin-orbit components. For a single electron, state j = l +/- 1/2, retaining only allowed nonnegative values; more generally j runs from abs(l-s) to l+s. Confirmed technical error
T02 p. 3, crystal-structure paragraph P-3m1 (#166) combines a symbol with the wrong number. Use P-3m1 (No. 164) for that space group. If R-3m (No. 166) was intended, correct the symbol and explain the slab/bulk setting. See Issue 40. Confirmed
T03 p. 5, representation-construction sentence Gamma_vib = Gamma_equiva + Gamma_vect is printed as a direct sum. Use the tensor product of the site/permutation and vector representations before decomposition, with the definitions stated. Confirmed notation error
T04 p. 5, 1 SL decomposition and Equation 2 The E-mode coefficients include the twofold dimension as though it were also an irrep multiplicity. Write multiplicities as 3 A1g + 4 A2u + 3 Eg + 4 Eu for 1 SL and count each E irrep as dimension two. Apply the same convention to multilayers. Confirmed notation inconsistency
T05 p. 9, immediately before Equation 9 The cubic-invariant selection rule contains three IR factors. Replace the last IR factor with the relevant Raman representation: Gamma_R x Sym^2(Gamma_IR) must contain A1g. Equation 9 itself has the intended Raman-times-IR-squared terms. Confirmed
T06 p. 5, explanation of Equation 1 q_hat is called a phonon unit vector. Call it the unit wavevector or propagation-direction vector, distinguishing it from an atomic phonon polarization/eigendisplacement vector. Clarify
T07 p. 9, introduction to Equation 7; Equations 7-10 Lowercase q is used for displacement amplitude in a sentence, while q elsewhere denotes wavevector and the exchange equations use Q. Use Q consistently for the mass-weighted mode coordinate and q for momentum/wavevector. Define every omitted mode subscript. Confirmed notation inconsistency
T08 pp. 10-11 and 14, Equation 11 An RMS coordinate is described and plotted as a signed time average. Reserve Q_RMS for a nonnegative magnitude, use <Q> for the signed mean, and define any signed-amplitude convention separately. See Issue 2. Confirmed definition conflict
T09 p. 7, wavelength example The printed wavelength expression is 2/(pi k), inconsistent with the quoted wavevector convention and result. Use lambda = 2 pi/Re(q). For 0.32 nm^-1 this gives 19.63 nm, not 18 nm. Define k if retaining it instead of q. Confirmed
T10 p. 12, scattering-rate relation tau^-1 = 2 Im(Sigma) omits the energy-to-rate conversion. Include 1/hbar and the linewidth/sign convention; include quasiparticle weight if used. See Issue 34. Confirmed dimensional omission
T11 p. 4, Table 2 footnote Gamma, tabulated in meV, is described literally as inverse lifetime. Define gamma_E = hbar gamma_rate and specify amplitude versus population lifetime, HWHM versus FWHM. Confirmed unit/convention gap
T12 p. 4, Table 2 The heading D(c) can suggest a velocity unit, although Equation 3 defines D = v_g/c. Label the row D = v_g/c (dimensionless). If listing v_g instead, state the velocity unit explicitly. Clarify
T13 Figures 2-3 and Table 2 Frequency symbols such as omega are labeled directly in meV. This is common shorthand but the energy conversion is not consistently stated. Label energy axes hbar omega (meV), or explicitly state that phonon frequencies are expressed as energies throughout. Distinguish THz frequency from angular frequency. Clarify; notation convention
T14 p. 7, speed-of-light comparison The numerical speed is written as 10^8 without units. Give c approximately 3 x 10^8 m/s; describe D = 10^-5 to 10^-4 as a speed ratio or a reduction by roughly 10^4-10^5. Confirmed missing unit; wording clarification
T15 p. 14, Methods The phonon grids are printed as 661 and 441; the latter appears again later in the same sentence. These are not an extraction artifact. After checking the inputs, write the intended dimensions explicitly, apparently 6 x 6 x 1 and 4 x 4 x 1. Confirmed typesetting ambiguity; intended grids to verify
T16 p. 14, Mn pseudopotential configuration 4d^5 is listed where the relevant Mn valence shell is normally 3d^5. Check the actual pseudopotential metadata and correct the shell if this is a transcription error. Do not infer a faulty calculation from the text alone. Likely typo; input check needed
T17 p. 14, Equations 13-14 m0 occurs in the normalization but is not defined in the supplied explanatory text. Define m0 and explain how atomic masses and mass-weighted eigenvectors enter the perturbation and mode normalization. Clarify
T18 pp. 4-5, dielectric notation Table 2 calls the high-frequency quantity epsilon_D, while the discussion uses epsilon_infinity. Use one notation, or explicitly state their equivalence. Clarify
T19 p. 4, Table 2 epsilon_D and epsilon have identical values for every structure, but their definitions do not make clear whether they are distinct quantities. Explain the distinction or remove a redundant row. Also define the relation of epsilon_z to the in-plane electronic response. Clarify; not proof of incorrect values
T20 p. 1, hBN example citing reference 8 The cited monolayer-hBN polariton benchmark is described as far-infrared. For the cited in-plane hBN mode, use mid-infrared; the primary study gives an example near 1376.5 cm^-1, about 7.26 micrometers. Primary hBN paper and SI. Confirmed spectral-description correction
T21 p. 12, photonic-versus-elastic energy discussion Small photonic energy is attributed to a large deceleration factor, whereas the paper defines D as an ultralow v_g/c. Specify whether the intended quantity is small D or a large slowdown ratio c/v_g. Do not use these reciprocal quantities interchangeably. Confirmed terminology inconsistency

F2. Figure, caption and cross-reference corrections

ID Location Problem Suggested correction or action Status
T22 p. 2, Figure 1 caption The PDOS is assigned to the left panels, but it is on the right. Replace left with right. Confirmed
T23 p. 3, Figure 2e caption; repeated p. 5 The caption/text say five Raman modes, but six labeled groups appear: three Eg and three A1g groups. Replace five with six, or explain and identify an intentionally excluded group. Confirmed count mismatch
T24 p. 7, propagation-quality discussion A sentence refers to Figure 3c for the imaginary optical conductivity; Figure 3c shows Q and Figure 3b shows conductivity. Cite Figure 3b for conductivity and Figure 3c for Q. Confirmed wrong panel reference
T25 p. 8, Figure 4a The positive-J vertical axes run in the opposite numerical direction to the conventional upward-increasing axes in the other panels. Consider using upward-increasing axes consistently, or explicitly flag the direction. Recheck prose describing parabola orientation against numerical J, not visual curvature. Presentation recommendation; axis reversal is not itself an error
T26 p. 8, Figure 4f-g captions and labels The curves are labeled as time-averaged displacement/exchange but are discussed using signed RMS amplitudes. Align the figure labels with the actual statistic and show the signed mean separately. See Issue 2. Confirmed definition/label conflict
T27 p. 10, Figure 5e The panel header says 2 SL AFM at finite Q, while the caption and discussion describe induced FM order. State explicitly whether the header names the initial phase or the spin configuration used for the displayed spectrum; use a consistent label for the computed state. Clarify
T28 p. 10, Figure 5 caption Parentheses around the light-induced-displacement explanation and around the final finite-Q condition are not properly balanced. Rewrite the caption as complete sentences and close each parenthetical expression. Confirmed punctuation error
T29 p. 11, Figure 6 caption Blue markers are used extensively but not defined in the caption. Define the blue series and the state/momentum represented by each point. See Issue 33. Confirmed incomplete legend
T30 pp. 4 and 12, Figure 1/7 labels and prose Imposed FM configurations are alternately called high-field, artificial and FM, without consistently identifying whether an external field is part of the calculated Hamiltonian. Use imposed FM for an imposed spin alignment, and reserve field values for calculations that actually include a specified field. Clarify

F3. Spelling, grammar and punctuation

The short replacements below preserve the intended meaning where possible. They do not endorse any scientific claim that is questioned elsewhere in this report.

ID Location Printed wording/problem Suggested replacement Status
T31 p. 4, first paragraph An extra closing parenthesis follows the N_SL = 1 expression. Retain one matching pair of parentheses. Confirmed punctuation error
T32 p. 9, introduction to the nonlinear potential The expression introducing V(Qx,Qy,Qa,QX,QY;t) has an unmatched opening parenthesis before V. Remove the extra opening parenthesis or add the intended matching close. Confirmed punctuation error
T33 p. 14, paragraph before Equation 11 stimulation trajectories simulation trajectories Likely word-substitution typo
T34 p. 9, explanation of quadratic exchange variates the coupling changes the coupling or causes the coupling to vary Copy edit
T35 p. 4, orbital-order discussion, repeated alternations is used for changes in orbital ordering/hybridization. Use alterations or changes unless a repeated alternation is specifically intended. Likely wrong-word choice
T36 p. 9, comparison of parabolas is inversed is inverted; check the numerical curvature sign, especially given the reversed axis in Figure 4a. Copy edit
T37 p. 7 and p. 13 It is worth to mention It is worth mentioning or We note that Confirmed grammar error
T38 p. 12, comparison with experiment This observation is well consistent with This observation is consistent with or agrees well with Copy edit
T39 p. 12, final paragraph the calculations of Im Sigma is not possible calculations of Im Sigma are not possible within the present method, or use the singular the calculation ... is. Confirmed subject-verb disagreement
T40 p. 14, introduction to Equation 13 The matrix element is described with two successive predicates: it is the ... matrix is defined by. Use is the electron-phonon matrix element, defined by .... Matrix element is also more precise for the indexed scalar g. Confirmed sentence-structure error
T41 p. 10, Figure 5 caption and diagram; p. 11 equivalence superexchange, inequivalence superexchange, and inconsistent splitting of superexchange into two words. Use equivalent superexchange pathways and inequivalent superexchange pathways, with superexchange written consistently as one word. Confirmed grammar/terminology issue
T42 p. 9, left column Besides, the high confinement mentioned above, ... Besides the high confinement mentioned above, ... or In addition to the high confinement discussed above, .... Confirmed punctuation/grammar error
T43 p. 7, layer-number discussion Increasing number of layers leads to ... Increasing the number of layers leads to .... Copy edit
T44 p. 9, explanation of suppression of linear terms remaining the modulation effect at second order leaving a second-order modulation or so that the leading modulation is second order. Copy edit
T45 p. 14, vacuum description we added an 18 A vacuum we included an 18 A vacuum gap, using the angstrom symbol in the manuscript. Specify whether 18 A is the gap or a supercell dimension. Copy edit plus definition clarification

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.

ID Location/reference Finding and correction Status / evidence
R01 Reference 1, p. 15 The entry pairs 2014 with Nanophotonics volume 4, pages 44-68. The published article PDF identifies the volume/page citation as 2015, 4, 44-68. Normalize the bibliography to that issue year; the 2014 embedded in the DOI is not the issue year. Confirmed volume/year mismatch. Primary article PDF; DOI 10.1515/nanoph-2014-0003.
R02 Reference 2, p. 15 The entry gives 2018, but the published article is Nanophotonics 2019, 8(12), 2129-2175. Its manuscript dates also fall in 2019. Replace the year and reimport the bibliographic record from the article. Confirmed wrong year against the published PDF. Primary published article; DOI 10.1515/nanoph-2019-0232.
R03 Reference 3, p. 15 The entry uses 2023 alongside volume 9(1), pages 9-28. The issue citation is Nature Reviews Materials 2024, 9, 9-28. The online publication date is December 2023, so 2023 is a real date, but it should not be mixed silently with the 2024 issue metadata. Metadata normalization, not a nonexistent reference. Publisher dates and recommended citation.
R04 Reference 20, p. 16 The author is rendered as Hugo Wang, H., treating part of the given name/nickname as the family name. The source lists Huaiyu (Hugo) Wang, with Wang as the surname. Confirmed name-parsing error. Use Wang, H. or an accurately formatted form preserving the nickname if required. Primary author list.
R05 Table 1 footnote, p. 3; Methods, p. 14; reference 34, p. 16 The named EV93 functional is linked to an Engel-Vosko 1994 fourth-order-gradient paper, PRB 50, 10498. The standard EV93 reference is Engel and Vosko, Phys. Rev. B 47, 13164-13174 (1993). Add/correct the foundational citation after confirming the actual functional implementation. Confirmed mismatch with the standard EV93 reference, not proof that the code used a wrong functional. Official Libxc entry, 1993 paper, currently cited 1994 paper.
R06 p. 9, citation following Z2ph; reference 28 Reference 28 is a study of crystal growth and magnetic structure. It is not an evident source for the phonon topological invariant invoked at that location. Strong citation mismatch to correct or explain. Cite the actual invariant methodology and specify its definition, then keep reference 28 where it supports magnetism/structure. See Issue 8.
R07 p. 5, sentence citing experiments; reference 38 Reference 38 is a group-theory textbook, not an experimental report establishing 2D polariton conditions. Confirmed mismatch in source type. Keep the textbook for symmetry analysis; use the relevant primary 2D polariton experiments for the experimental statement.
R08 p. 13, ordinary magnetoelastic comparison; reference 57 Reference 57 is titled Spin Anisotropy of Ferromagnetic Films. Its connection to the exact magnetoelastic comparison is not explained. Citation-scope check, not an assertion that the entire source is irrelevant. Explain the supporting result or replace it with a direct magnetoelastic source.
R09 p. 1, device/application sentence; references 9-10 Reference 9 concerns an organic polariton transistor and reference 10 a trapped-ion simulation. They provide broader polariton context, rather than direct demonstrations of the stated capabilities of material phonon polaritons. Scope qualification. Say that these are examples from polaritonics more broadly, or add primary phonon-polariton device/computing references for a narrower claim.
R10 p. 7, graphene wavelength comparison; reference 39 Reference 39 is a review of graphene-based plasmonic sensors, while the text later calls it the real experiment being compared. Citation-description mismatch. Identify the specific primary experiment and its frequency, doping and geometry, or describe reference 39 accurately as a review.
R11 Figure 2e, pp. 3 and 5; reference 27 The plotted 2 SL/3 SL experimental assignments are not clearly traced to a source measurement. The reference studies a homologous chemical family, whose composition index must not be confused with septuple-layer thickness. Source-provenance check. Trace each numerical bar to the original figure/table and sample thickness, including the source SI; see Issue 38.
R12 p. 11, high-field feasibility; references 43-45 The cited high-field sources cover different pulse spectra and amplitudes. A maximum field in one frequency range does not establish the same field, pulse width and heating conditions at the modeled 2.74 THz drive. Citation-scope/feasibility check. Attribute the field and frequency to the specific experiment; distinguish technical availability from demonstrated switching in this material.
R13 p. 13, EELS proposal; references 52-54 References 52 and 54 study electronic plasmons, and reference 53 is a broad electron-microscopy review. They support the technique generally but are less direct precedents for monolayer phonon-polariton detection than reference 40, already in the bibliography. Improve citation precision. Add reference 40 at this claim and explain the additional energy/momentum-resolution requirements for MnBi2Te4.
R14 p. 7 and p. 13, references to Figure S7/S8 Figure S7 is invoked for mode charges/IR spectra, an hBN quality-factor comparison and temperature-dependent electronic bands, while related hBN properties are elsewhere assigned to Figure S8. One multipanel figure could contain different quantities, but the supplied text makes the assignments difficult to follow. SI cross-reference audit needed. Check every figure/panel number against the actual SI; do not declare these references incorrect before that inspection.
R15 p. 12, SI-note reference The discussion cites Note 4, whereas the surrounding article uses Note S4 and other S-prefixed note identifiers. Minor cross-reference-format inconsistency. Confirm the intended note and use the exact SI identifier consistently.
R16 p. 9 and p. 11, literature comparisons Residual forms such as [ref 41] and [ref 33] appear within otherwise superscript-numbered citation prose. Confirmed formatting inconsistency. Replace editorial reference tags with the journal’s standard numbered citation format.

F5. Editorial safeguards for applying the corrections

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.

Check Result from printed inputs Published comparison Interpretation
1 SL second mode, E = 19 meV, E0 = 11.33 meV, D = 8.7e-5 q = 0.598 nm^-1 1 SL red curve in Figure 3a near 0.3 nm^-1 Dispersion/normalization requires reconciliation; the separate text example has an ambiguous thickness assignment
1 SL first mode, E = 20 meV, E0 = 6.17 meV, D = 4.7e-5 q = 3.16 nm^-1 Figure 3a near 1.5 nm^-1 Apparent factor near two
1 SL first-mode Q at 20 meV, gamma_E = 0.09 meV Q = 100.5 Figure 3c near 200 Check linewidth convention
1 SL second-mode Q under the same conditions Q = 75.5 Figure 3c near 150 Check linewidth convention
Wavelength for angular wavevector 0.32 nm^-1 2 pi/q = 19.63 nm Approximately 18 nm and incorrect printed expression Formula and numerical value require correction
Free-space wavelength at 19 meV 65.25 micrometers Approximately 65 micrometers This conversion is consistent
Pump frequency 2.74 THz 11.33 meV Second mode near 11.33 meV This frequency assignment is consistent

Minimum evidence needed to resolve the main conclusions

  1. Obtain and inspect the SI, especially the electrodynamics, topology, nonlinear dynamics and exchange notes, before finalizing any absence-based criticism.
  2. 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. Release the numerical data and calculation inputs, and correct the mathematical, symmetry, 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.