Pith. sign in

REVIEW 2 major objections 4 minor 36 references

Boundary conditions for and ferromagnetic resonance spectra of magnetic bilayers coupled by interlayer Dzyaloshinskii-Moriya interactions

T0 review · 2 major / 4 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read The paper predicts that interlayer Dzyaloshinskii-Moriya interaction splits the ferromagnetic resonance of a magnetic bilayer into a doublet whose separation is proportional to $D_z/A$, and identifies an antiphase angular signature for…

desk verdict A solid new derivation of IL-DMI boundary conditions and an FMR splitting formula, but the angular signature ignores the static D⊥ torque and the claimed atomistic validation is missing. read the letter →

arxiv 2411.15010 v1 pith:D43VGULP submitted 2024-11-22 cond-mat.mes-hall cond-mat.mtrl-sci

Elena Y. Vedmedenko , Mikhail Kostylev This is my paper · ORCID
classification cond-mat.mes-hallcond-mat.mtrl-sci
keywords interlayerDzyaloshinskii-Moriyainteractionferromagneticresonancemagneticbilayerexchangeboundaryconditionsacousticandopticalmodesmultilayerschiralcoupling
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper sets out to show that interlayer Dzyaloshinskii-Moriya interaction (IL-DMI) does alter the ferromagnetic resonance (FMR) frequencies of a magnetic bilayer, in contrast to interfacial DMI which does not. To establish this, the authors derive new linearized exchange boundary conditions that include the IL-DMI torque at the internal interfaces, solve the associated boundary-value problem for two ferromagnetic layers separated by a nonmagnetic spacer, and confirm the result with a microscopic atomistic model. The central quantitative prediction is that the parent uniform precession mode splits into an acoustic and an optical mode with resonance-field separation $H_2-H_1$ proportional to $D_z/A$. They also predict that the two modes' resonance fields vary with the in-plane field angle as sine waves in antiphase, and argue that this angular signature makes IL-DMI separable from interlayer Heisenberg exchange in an FMR experiment.

What carries the argument

The load-bearing object is a set of interlayer boundary conditions (IL-BC) for the linearized Landau-Lifshitz equation, Eqs. (4) and (A7). They balance the intralayer exchange torque and surface-anisotropy torques against interlayer Heisenberg exchange and an IL-DMI torque that couples the x-component of one layer's dynamic magnetization to the y-component of the other. Substituting the standing-wave solutions into these conditions produces an $8\times 8$ determinant $\det[\boldsymbol{\Lambda}(\omega)]=0$ whose roots are the FMR fields; in the limit $D_z/A\ll 1$ the determinant reduces to the closed-form splitting Eq. (5). The physical mechanism is visible in the mode profiles: the dynamic magnetization changes phase by $\pm\pi/2$ at the internal interface, with the sign of the jump selecting the acoustic versus the optical mode and directly reflecting the chirality of the $\mathbf{D}\cdot(\mathbf{M}_1\times\mathbf{M}_2)$ coupling.

What would settle it

Angle-resolved FMR on a symmetric Co/Pd/Co or Co/Ag/Co bilayer with weak in-plane anisotropy should show two fundamental modes whose resonance-field angle sweeps are sine waves in antiphase, and the field separation should track Eq. (5). Observing in-phase sine waves, or a single unsplit mode, would refute the claim; so would wavevector-resolved spectroscopy that fails to see the predicted $\pm\pi/2$ phase jump between the layers at the internal interface.

Watch

Extended reading notes

Core claim

The discovery, stated on the paper's own terms, is that IL-DMI has a different symmetry from interfacial DMI: its energy $E_{\rm IL-DMI} = -\mathbf{D}\cdot(\mathbf{M}_1\times\mathbf{M}_2)$ couples the layers chirally and therefore enters the exchange boundary conditions, whereas interfacial DMI does not enter the FMR frequencies. Solving the boundary-value problem shows that the uniform mode of a single layer splits into a doublet even when the Heisenberg interlayer exchange is absent. The acoustic mode has the lower frequency and the optical mode the higher one, and the dynamic magnetizations of the two layers precess neither in phase nor exactly out of phase: the mode profiles show phase jumps of $\pm\pi/2$ at the internal interface. In the limit of small $D_z/A$, the splitting is $H_2-H_1 = \frac{1}{\gamma\mu_0}\frac{4\alpha_{\rm inh}\omega_M}{d_1}\frac{\omega^2}{4\omega^2+\omega_M^2}\frac{D_z}{A}$, and angle-resolved FMR traces of the two modes oscillate in antiphase, which the paper proposes as the practical fingerprint of IL-DMI.

Load-bearing premise

The derivation assumes the IL-DMI energy has the bilinear form $-\mathbf{D}\cdot(\mathbf{M}_1\times\mathbf{M}_2)$ with a fixed D vector and that the static magnetizations are perfectly aligned with the applied field; if the real interaction has a different angular dependence, if D is not fixed by the film geometry, or if the field is too small to enforce alignment, the predicted doublet splitting and antiphase angular signature could be modified or absent.

Editorial extensions

If this is right

  • In a bilayer such as Co/Pd/Co, an FMR trace should show the single-layer parent absorption replaced by two lines, with the field separation increasing linearly with $D_z$ at a fixed microwave frequency.
  • Rotating the in-plane applied field should make the acoustic and optical resonance fields trace sine waves with a nonzero phase shift between them; for vanishing in-plane anisotropy the traces are in antiphase, a direct IL-DMI signature.
  • Because the interlayer Heisenberg contribution is angle-independent while the IL-DMI contribution is angle-dependent, the same measurement can separate $A_{12}$ from $D$.
  • With strong ferromagnetic interlayer exchange present, the optical mode loses its perfect antisymmetry, so its absorption peak should become visible above the noise floor in broadband FMR.
  • The ordering of the two branches as a function of field carries the sign of $D_z$, so the acoustic/optical assignment from the phase profiles also gives the chirality of the coupling.

Reading between the lines

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

  • If the linearized IL-DMI energy is correct, the same boundary-condition method should extend to finite-wavevector spin waves: the $\pm\pi/2$ interface phase jump would become nonreciprocal dispersion, a testable prediction the paper leaves implicit.
  • The predicted antiphase angular dependence offers a low-cost screening method: standard angle-resolved broadband FMR could rank candidate nonmagnetic spacers by their IL-DMI strength before more elaborate spin-polarized or transport measurements.
  • The model's assumption of perfect static alignment means the doublet splitting should shrink or distort at applied fields comparable to the anisotropy fields; measuring the splitting down to low fields would test the theory beyond the regime it currently covers.
  • If the true IL-DMI has higher-order angular dependence than the bilinear $\mathbf{D}\cdot(\mathbf{M}_1\times\mathbf{M}_2)$ form, the sine-wave angular traces would acquire harmonics; harmonic analysis of the FMR angle dependence could therefore expose deviations from the assumed symmetry.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 4 minor

Summary. The paper derives interlayer exchange boundary conditions that include the interlayer Dzyaloshinskii-Moriya interaction (IL-DMI) for linear magnetization dynamics in a magnetic bilayer separated by a nonmagnetic spacer. Solving the resulting boundary-value problem, the authors obtain FMR frequencies and resonance fields for the acoustic and optical modes, including the approximate closed-form expression Eq. (5) for the mode splitting in the limit of weak D_z and vanishing interlayer Heisenberg exchange and anisotropy. They further predict that the resonance fields of the two modes vary in antiphase as the in-plane direction of the applied field is rotated relative to the D vector, and they propose this angular dependence as an experimental fingerprint of IL-DMI. The paper also states that the numerical determinant solution and the approximate formula are validated by a microscopic atomistic model, although no details of that model are provided.

Significance. If the results are valid, the paper provides a practical and conceptually useful way to detect IL-DMI by angle-resolved FMR and to separate it from interlayer Heisenberg exchange. The derivation from an explicit IL-DMI energy is transparent, the approximate formula Eq. (5) is explicit and physically interpretable, and the numerical determinant solution is a well-defined consistency check for the boundary-value problem. The claimed distinction between IL-DMI and IF-DMI in FMR is an important message for the multilayer magnetism community. However, the central angular fingerprint is derived by linearizing about a collinear static state that is not an equilibrium when the D vector has a component transverse to the static magnetization; this issue must be resolved before the proposed experimental protocol can be considered reliable.

major comments (2)
  1. [§II, Eqs. (3)–(4); Appendix I, Eq. (A4)] The linearized boundary conditions and effective fields in Eqs. (3)–(4) retain only the D_z component of the IL-DMI vector, but the exact fields in Eq. (A4) contain static terms proportional to D_y M_s and D_x M_s. For an in-plane D vector and φ_HD not equal to 0 or π, each layer experiences a static torque of order D sin(φ_HD)/(d_1 μ_0) that cannot be balanced by H0, because a field collinear with the assumed equilibrium magnetization exerts no torque. For the authors' Co parameters (D = 0.1 mJ/m², d_1 ≈ 3 nm), this corresponds to a transverse effective field of roughly 2×10⁴ A/m (≈ 240 Oe), which is not negligible compared with the resonance fields shown in Fig. 3. Thus the state M1 = M2 ∥ H0 is not an equilibrium once φ_HD is varied, and the linearization leading to Eq. (4) and the angular traces in Figs. 3(c)–3(d) is not self-consistent. The caveat in Sec. III that co-alignment may fail at small fields does not address this: the problem is not the field magnitude but the absence of any restoring torque for the transverse D_y component. The authors should recompute the static equilibrium including the full D vector and re-examine whether the predicted φ_HD dependence, especially the positions of the extrema and the antiphase relation, survives the resulting canting.
  2. [§II and §III, microscopic atomistic model] The manuscript states that a microscopic atomistic model was constructed to independently evaluate the numerical determinant solution and Eq. (5), and later that the numerical results 'align well with the microscopic model and those from Ref. 9.' However, no Hamiltonian, no atomic-layer parameters, no integration scheme, and no comparison data or error estimates are provided anywhere in the paper. As written, this validation claim cannot be checked by a reader. Either present the atomistic model and the comparison in sufficient detail in an appendix, or remove the validation claim and present the boundary-value problem solution as the primary and self-contained check of the theory.
minor comments (4)
  1. [Throughout] The text contains several typographical errors, including 'into into', 'tt remains', 'repectively', 'intentionnaly', 'mulitplied', and 'freqeuncy'. A careful proofreading pass is needed.
  2. [§III, discussion of Fig. 2] The text refers to 'D = 0.3 J/m2' when discussing Fig. 2(a); the correct unit should be mJ/m², consistent with the rest of the paper.
  3. [§I and §III] The paragraph describing the phase difference between the dynamic magnetizations of the two layers is repeated almost verbatim in the Introduction and in Sec. III. Please consolidate the repetition.
  4. [Eq. (3)] The notation 'effective field' for H_D would benefit from an explicit statement of normalization: as written, the prefactor D_z/(μ_0 d_1 M_s^2) is dimensionless, so it should be clarified whether H_D is normalized by M_s or whether H_eff in Eq. (1) is defined in dimensionless units.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the FMR splitting and angular signatures are derived from an explicit IL-DMI energy functional, with no fitted target and no load-bearing self-citation.

full rationale

The paper's central claim, that IL-DMI alters the frequencies of fundamental FMR modes and produces an anti-phase angular signature, is derived rather than assumed. Equation (5), H2 - H1 = (1/(gamma mu0)) (4 alpha_inh omega_M/d1) (omega^2/(4 omega^2 + omega_M^2)) (D_z/A), follows from solving the linearized Landau-Lifshitz equation with the boundary conditions (4)/(A7), which are themselves obtained variationally from the stated IL-DMI energy Eq. (2)/A2. This is a mathematical consequence of a concrete input Hamiltonian, not a definitional restatement: the same input does not automatically determine the FMR frequency shift, and the paper contrasts IL-DMI with IF-DMI, which is known not to contribute to FMR frequencies. The angular anti-phase behavior of H1(phi) and H2(phi) is likewise computed from the boundary-value problem, not fitted to data; it is a prediction of the model. The numerical 'microscopic' atomistic checks use the same interaction energy and therefore are consistency checks rather than independent confirmation, but a consistency check is not circular reasoning. Self-citations such as Refs. [4, 6, 9, 10, 18, 20, 21, 30] are used for parameter ranges, material constants, and prior independent results; none is the load-bearing justification for the IL-DMI frequency shift itself. The acknowledged breakdown of perfect M parallel H0 co-alignment at small fields is a stated limitation and a potential correctness risk, not a circular step. No step in the derivation chain can be exhibited as reducing Eq. (5) or the angular signature to the input by construction.

Assumptions & free parameters 0 free parameters · 4 assumptions · 0 invented entities

The paper introduces no new free parameters fitted to data; material parameters are taken from prior literature (e.g., Ms, A, Ks, Kin) and scanned values of D and A12 are inputs, not fits. The main assumptions are the IL-DMI energy form, perfect static alignment, and the neglect of IF-DMI, all of which are domain assumptions from prior work.

assumptions (4)
  • domain assumption The IL-DMI energy has the form E = -D dot (M1 x M2) with a constant vector D.
    This form is taken from prior literature (Refs. 4, 7-12); the derivation of FMR splittings depends on this specific energy functional.
  • domain assumption The static magnetization vectors are perfectly co-aligned with the applied field H0, making the dynamics linearizable around a uniform state.
    Used throughout Section II and Appendix I; the authors note this may fail at small fields.
  • domain assumption Interfacial DMI (IF-DMI) does not contribute to FMR, so it is neglected.
    Based on Refs. 19-21; the paper does not rederive this.
  • standard math The Rado-Weertman boundary conditions apply at external surfaces without modification.
    Standard boundary conditions from Refs. 23-25 used as input.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Boundary conditions for and ferromagnetic resonance spectra of magnetic bilayers coupled by interlayer Dzyaloshinskii-Moriya interactions." pith.science (2026). https://pith.science/paper/D43VGULP

@misc{pith2026241115010,
  author       = {Pith},
  title        = {Pith review of: Boundary conditions for and ferromagnetic resonance spectra of magnetic bilayers coupled by interlayer Dzyaloshinskii-Moriya interactions},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/D43VGULP}},
  note         = {Machine review of arXiv:2411.15010}
}
read the original abstract

Interfacial Dzyaloshinskii-Moriya interaction (IF-DMI) leads to non-collinear spin configurations within the magnetic layers of multilayer heterostructures, while its interlayer counterpart (IL-DMI) minimizes chiral states between the layers. Here, we demonstrate that the symmetries of these interactions are very different, even though both arise from pairwise exchange interactions between magnetic sites mediated by nonmagnetic atoms. By deriving new boundary conditions for the exchange operator and solving the associated boundary value problem, we show that, unlike IF-DMI, which does not contribute to the FMR frequencies, IL-DMI alters the frequencies of the fundamental FMR modes and can be separated from other contributions in an FMR experiment.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

36 extracted references · 36 canonical work pages

  1. [1]

    E. Y. Vedmedenko et al. J. Phys. D 53, 1 (2020)

    work page 2020

  2. [2]

    Dzyaloshinskiy, J

    I. Dzyaloshinskiy, J. Phys. Chem. Solids 4, 24 (1958)

    work page 1958

  3. [3]

    P. M. Lévy and A. Fert, Phys. Rev. B 23, 4667 (1981)

    work page 1981

  4. [4]

    E. Y. Vedmedenko, J. A. Arregi, P. Riego, and A. Berger, Phys. Rev. Lett. 122, 257202 (2019)

    work page 2019

  5. [5]

    Goebel, I

    B. Goebel, I. Mertig, and O. A. Tretiakov, Phys. Rep. 895, 1 (2021)

    work page 2021

  6. [6]

    E. Y. Vedmedenko, L. Udvardi, P. Weinberger, and R. Wiesendanger, Phys. Rev. B 75, 104431 (2007)

    work page 2007

  7. [7]

    Fernandez-Pacheco, E

    A. Fernandez-Pacheco, E. Y. Vedmedenko, F. Ummelen, D. Petit, R. Cowburn, Nat. Mater. 18, 679 (2019)

    work page 2019

  8. [8]

    D.-S. Han, K. Lee, J.-P. Hanke, Y. Mokrousov, K.-W. Kim, W.Yoo, Y. L. W. Hees, T.-W. Kim, R. Lavrijsen, C.-Y. You, H. J. M. Swagten, M.-H. Jung, and M. Kläui, Nat. Mater. 18, 7032019 (2019)

    work page 2019

Show all 36 references
  1. [9]

    J. A. Arregi, P. Riego, A. Berger, and E. Y. Vedmedenko, Nat. Commun. 14, 6927 (2023)

    2023

  2. [10]

    Yun et al., Adv

    J. Yun et al., Adv. Funct. Mater. 33, 2302731 (2023)

    2023

  3. [11]

    C. O. Avci, C.-H. Lambert, G. Sala, and P. Gambardella, Phys. Rev. Lett. 127, 167202 (2021)

    2021

  4. [12]

    K. Wang, L. Qian, S.-C. Ying, and G. Xiao, Commun. Phys. 4, 10 (2021)

    2021

  5. [13]

    M. N. Baibich, J. M. Broto, A. Fert, F. Nguyen Van Dau, F. Petroff, P. Etienne, G. Creuzet, A. Friederich, and J. Chazelas, Phys. Rev. Lett. 61, 2472 (1988)

    1988

  6. [14]

    Bruno and C

    P. Bruno and C. Chappert, Phys. Rev. B 46, 261 (1992)

    1992

  7. [15]

    Ruerig, R

    M. Ruerig, R. Schäfer, A. Hubert, R. Mosler, J. A. Wolf, S. Demokritov, P. Grünberg, Phys. Stat. Sol. A 125, 635 (1991). 12

    1991

  8. [16]

    X. M. Liu, H. T. Nguen, J. Ding, M. G. Cottam, and A. O. Adeyeye, Phys. Rev. B 90, 064428 (2014)

    2014

  9. [17]

    R. A. Duine, K.-J. Lee, S. S. P. Parkin, and M. D. Stiles, Nat. Phys. 14, 217 (2018)

    2018

  10. [18]

    I. S. Maksymov and M. Kostylev, Physica E 69, 253 (2015)

    2015

  11. [19]

    J. H. Moon, S.-M. Seo, K.-J. Lee, K.-W. Kim, J. Ryu, H.-W. Lee, R. D. McMichael, and M. D. Stiles, Phys. Rev. B 88, 184404 (2013)

    2013

  12. [20]

    Kostylev, J

    M. Kostylev, J. Appl. Phys. 115, 233902 (2014)

    2014

  13. [21]

    A. A. Stashkevich, M. Belmeguenai, Y. Roussigné, S. M. Cherif, M. Kostylev, M. Gabor, D. Lacour, C. Tiusan, and M. Hehn, Phys. Rev. B 91, 214409 (2015)

    2015

  14. [22]

    Y. Wang, J. Xia, C. Wan, X. Han, and C. Yu, Phys. Rev. B 109, 054416 (2024)

    2024

  15. [23]

    Hoffmann, Phys

    F. Hoffmann, Phys. Stat. Sol 41, 807 (1970)

    1970

  16. [24]

    M. Vohl, P. Barna𝑠𝑠̂, P. Grünberg, Phys. Rev. B 39, 12003 (1989)

    1989

  17. [25]

    G. D. Rado and J. T. Weertman, J. Phys. Chem. Solids 11, 315 (1959)

    1959

  18. [26]

    Pogoryelov, M

    Y. Pogoryelov, M. Pereiro, S. Jana, A. Kumar, S. Akansel, M. Ranjbar, D. Thonig, D. Primetzhofer, P. Svedlindh, J. Åkerman, O. Eriksson, O. Karis, and D. A. Arena, Phys. Rev. B 101, 054401 (2020)

    2020

  19. [27]

    This substitution into det (𝜦𝜦(𝜔𝜔)) = 0 will yield an imaginary part of the FMR frequency

    It is well-known that for linear dynamics, the effect of the Gilbert damping term is equivalent to a simple substitution 𝐻𝐻 0 → 𝐻𝐻0 + 𝑖𝑖Δ𝐻𝐻 into the final equations, where Δ𝐻𝐻 = 𝜔𝜔𝛼𝛼𝐺𝐺/𝛾𝛾, with 𝛼𝛼𝐺𝐺 as the Gilbert damping constant, and i the imaginary unit. This substitution i...

    1996

  20. [28]

    Y. Wei, Y. Pogoryelov, M. Pereiro, S. Jana, A. Kumar, S. Akansel, M. Ranjbar, D. Thonig, D. Primetzhofer, P. Svedlindh, J. Åkerman, O. Eriksson, O. Karis, and D. A. Arena, J. Appl. Phys. 115, 17D129 (2014)

    2014

  21. [29]

    Lueng, P

    C. Lueng, P. J. Metaxas, and M. Kostylev, IEEE Trans. Magn. 52, 4002503 (2016)

    2016

  22. [30]

    Cascales-Sandoval, A

    M. Cascales-Sandoval, A. Hierro-Rodríguez, S. Ruiz-Gómez , L. Skoric , C. Donnelly, M. A. Niño, E. Y. Vedmedenko, D. McGrouther, S. McVitie , S. Flewett, N. Jaouen , M. Foerster , A. Fernández-Pacheco, Appl. Phys. Lett. 123, 172407 (2023)

    2023

  23. [31]

    Upon closer inspection, this root corresponds to a mode with a perfectly vanishing amplitude, which we consider unphysical

    The numerical solution of the determinant also yields a third root, which we choose to ignore. Upon closer inspection, this root corresponds to a mode with a perfectly vanishing amplitude, which we consider unphysical. This conclusion aligns with the presence of only two appro...

  24. [32]

    Therefore, in Fig, 2(b), the acoustic branches are above the respective optical ones

    The larger resonance frequency for a given applied field corresponds to a smaller resonance field for a given frequency. Therefore, in Fig, 2(b), the acoustic branches are above the respective optical ones. [33] N.G. Bebenin, A.V. Kobelev, A.P. Tankeyev, V.V. Ustinov, J. Magn....

    1997

  25. [34]

    When addressing the AFM IL-HE case, we assume the same FM state of static magnetization as for the FM IL -HEI scenario. This assumption implies that H 0 is sufficiently strong to co- align the static magnetization vectors of the AFM-coupled layers, resulting in a perfectly uni...

  26. [35]

    Recall that we assume H 0 lies in the plane of the structure and is strong enough to perfectly align the static magnetization vectors of both magnetic layers along H0

  27. [36]

    In this case, the two sine waves are also in anti-phase

    The only exception to this rule occurs when the in- plane bulk anisotropy axes of the two magnetic layers are perfectly orthogonal to one another and D =0. In this case, the two sine waves are also in anti-phase. However, their amplitudes are very small (approximately 2 Oe for...

  28. [37]

    Rado- Weertmann torques

    The precise value of d 0 is not important within this formalism, as the theory treats A12, D and d0 as independent quantities. Thus, d0 serves as a purely geometric parameter that defines the co - ordinates of the interfaces without affecting the exchange conditions at those i...

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.