REVIEW 3 major objections 4 minor 42 references
Theory of surface-induced multiferroicity in magnetic materials, thin films and multilayers
T0 review · 3 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash
Pith's one-line read A Ginzburg-Landau analysis shows that surfaces make collinear magnets ferroelectric even when their bulk is not.
desk verdict Worth a referee's time, but the detectability claim is not backed by the paper's own coupling constant ranges. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The key machinery is the Ginzburg-Landau free energy with boundary conditions, augmented by a linear magnetoelectric term and a quadratic electric term. The object that carries the first mechanism is the dimensionless magnetization profile $f(z)=\tanh((z+a_0)/(\sqrt{2}\xi))$, whose gradient $\nabla(m^2)$ plays the role of the inversion-symmetry-breaking field; substituting it into $p=-\chi_E\nabla(\gamma m^2)$ yields the polarization. For the second mechanism, the central object is the Lifshitz invariant $m_x\,dm_y/dz-m_y\,dm_x/dz$ (and its three-dimensional generalizations), multiplied by a surface profile $f(z)$ representing the spatially inhomogeneous DMI strength; minimizing the coupled equations gives $p_z=-D\gamma\chi_E f(z)\,d\varphi/dz$ and the skyrmion-like expression $p=\chi_E D f\gamma\cos^2\theta|\partial\theta/\partial z|$. The DMI profile and the spin-angle gradients carry the boundary effect into the polarization.
What would settle it
A first-principles calculation of $\gamma\chi_E$ for a clean (001) surface of a simple collinear magnet such as bcc iron would settle the quantitative claim: if the computed value is below about $10^{-19}\,\mathrm{sm/A}$, the polarization predicted by the paper's formula $|p_{\max}|\sim\chi_E|\gamma|m_\infty^2/\xi$ is too small to detect, and the central measurable prediction fails.
Extended reading notes
Core claim
At the level of a Ginzburg-Landau free energy, the paper shows that a collinear ferromagnet or antiferromagnet with a (001) surface has an equilibrium magnetization $m(z)=m_\infty\tanh((z+a_0)/(\sqrt{2}\xi))$, whose square has a nonzero gradient in a boundary layer. Adding the magnetoelectric coupling $F_{ME}=p\cdot(\gamma\nabla(m^2)+\gamma'[m(\nabla\cdot m)-(m\cdot\nabla)m]+\dots)$ plus the electric term $p^2/(2\chi_E)$ and minimizing gives $p(z)=-\chi_E\gamma\nabla(m^2)$, namely $p(z)=-\frac{m_\infty^2}{\sqrt{2}\xi}\chi_E\gamma\,\tanh\!\big(\frac{z+a_0}{\sqrt{2}\xi}\big)\big[1-\tanh^2\!\big(\frac{z+a_0}{\sqrt{2}\xi}\big)\big]$, peaked at $z=\sqrt{2}\xi\tanh^{-1}(1/\sqrt{3})$. The second route holds when a surface-induced DMI exists: the Lifshitz-invariant coupling yields $p_z=-D\gamma\chi_E f(z)\,d\varphi/dz$ for in-plane spin rotation, and $p=\chi_E D f\gamma\cos^2\theta|\partial\theta/\partial z|$ for localized three-dimensional textures, giving the same qualitative surface-bound polarization. The paper claims both mechanisms produce polarization detectable with current techniques for realistic parameter ranges.
Load-bearing premise
The load-bearing premise is that the magnetic materials in question possess a nonzero coupling between magnetization gradients and electric polarization, with strength $\gamma\chi_E$ at least around $10^{-19}\,\mathrm{sm/A}$; if that coupling is absent or smaller, the surface polarization may be theoretically present but experimentally invisible.
Editorial extensions
If this is right
- A clean surface of any collinear ferromagnet or antiferromagnet with nonzero $\gamma\chi_E$ should show a polarization layer of thickness set by $\xi$, even when the bulk is centrosymmetric and nonpolar.
- The predicted maximum polarization scales as $|p_{\max}|\sim \chi_E|\gamma| m_\infty^2/\xi$; for the reported coupling range it spans $10^{-10}$ to $10^2\,\mu\mathrm{C/cm}^2$, and becomes measurable once $\gamma\chi_E\gtrsim10^{-19}\,\mathrm{sm/A}$.
- Temperature controls the profile: $\xi$ grows near $T_c$, so the polarization peak moves away from the surface and weakens as the transition is approached from below.
- In thin films and multilayers, an engineered interfacial DMI creates a second route: polarization is proportional to the surface spin rotation rate $d\varphi/dz$, so chiral textures (including skyrmion-like states) carry a local electric polarization.
- The existence of a surface magnetoelectric response means thin-film devices based on nominally non-multiferroic magnets may display parasitic or exploitable ferroelectric responses at interfaces.
Reading between the lines
- The symmetry argument implies surface-bound polarization should be generic to any magnetoelectric material, so published surface-sensitive data on nominally non-multiferroic magnets may already contain an overlooked electric signal.
- Temperature sweeps provide a direct test: as $T\to T_c$ from below, $\xi$ diverges, so the predicted peak moves farther from the surface while $|p_{\max}|\sim 1/\xi$ diminishes; a measured profile with this scaling would confirm the mechanism.
- Engineering the interfacial DMI, for instance through heavy-metal/ferromagnet bilayers, should turn the chirality of surface spin textures into a voltage-addressable polarization, linking skyrmion chirality to ferroelectric switching.
- Because the coupling $\gamma$ is material-specific and not derived here, a first-principles calculation of $\gamma$ for a clean Fe or Ni surface would settle whether the predicted polarization is observable or purely formal.
Signed reviews
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The manuscript develops a Ginzburg-Landau theory for electric polarization induced near surfaces of collinear magnetic materials. In Sec. II the authors solve the one-dimensional GL equation for the magnetization profile near a (001) surface and couple it to polarization through a term p·(γ∇(m^2)+...). They obtain a closed-form p(z) peaked a few correlation lengths from the surface. In Sec. III they consider two DMI-based routes in thin films: an in-plane anisotropy case with a one-dimensional twist and an easy-axis anisotropy case with an axisymmetric soliton; they solve the resulting equations numerically and plot polarization profiles. The paper estimates possible polarization magnitudes and asserts that the predicted polarization is detectable with current experimental techniques.
Significance. If the qualitative claim holds, the paper identifies a general mechanism: any magnetic material with a nonzero magnetoelectric coupling γ and a surface-induced magnetization gradient develops a boundary-localized polarization even in the absence of DMI. This is a useful complement to material-specific first-principles studies. The analytic part of Sec. II is clean, the numerical scheme in Appendix B is described in sufficient detail to be reproducible, and no result is fitted to a target polarization. The main limitation is that the coupling γ is not constrained for any material, so the quantitative predictions span many orders of magnitude and the detectability claim is conditional rather than unconditional.
major comments (3)
- [Sec. II, Eq. (4)] The prefactor in Eq. (4) is off by a factor of two. Using m(z)=m∞ tanh[(z+a0)/(√2 ξ)], one finds d(m^2)/dz = √2 (m∞^2/ξ) tanh u (1−tanh^2 u), so p(z) = −χE γ √2 (m∞^2/ξ) tanh u (1−tanh^2 u), not −χE γ m∞^2/(√2 ξ) tanh u (1−tanh^2 u). Consequently the maximum polarization should be pmax = −2√2/(3√3) χE γ m∞^2/ξ, which is twice the value implied by the displayed Eq. (4). The approximate estimate pmax ≈ −χE γ m∞^2/ξ remains numerically valid, but the printed formula needs correction.
- [Secs. II and IV; footnote 34] The claim in the Discussion that 'the polarization we predict is detectable within the current accuracy of the experimental techniques' is not supported by the analysis. The polarization amplitude is proportional to γχE, and the paper itself gives the literature range 10^-23 to 10^-13 sm/A for γχE and the threshold γχE > 10^-19 sm/A for measurability. Because no material-specific value or independent constraint on γ is provided, a material at the lower end of this range yields p ~ 10^-10 µC/cm^2, far below current experimental sensitivity. Footnote 34 states that the authors 'focus on the necessary value of coupling constant that would provide measurable polarization,' which is precisely a conditional statement; the unconditional detectability assertion in the Discussion overreaches. The authors should either identify a material or material class where γχE exceeds the threshold, or explicitly reframe the prediction as conditional on the coupling.
- [Sec. III, Figs. 2 and 3] The numerical DMI calculations are made dimensionless and illustrated with A=1, D=1, k=1, γ=1, χE=1.5, but no mapping is given from these parameter choices to physical materials, and no systematic variation of D, k, the profile function f(z), or the anisotropy is presented. The estimates p ~ 10^-10 to 10^4 µC/cm^2 in Secs. III A and III B rest on the same unconstrained product DγχE as in Sec. II, and on the assumed gradients Δφ/Δz ≈ 10^5 m^-1 and |∂θ/∂z| ≈ 10^4 m^-1, whose compatibility with the chosen dimensionless parameters is not demonstrated. The statement that both profiles f(z) lead to the same physics qualitatively is asserted but not shown. This limits the predictive content of the DMI route.
minor comments (4)
- [Fig. 1 and Sec. II] Figure 1 uses 'τ = 1, 2, ...' but τ is defined as T/Tc, and the curves are not explained in the caption; clarify which reduced temperatures are shown and specify the units of the axes.
- [Sec. III, after Eq. (5)] The sentence following Eq. (5) mislabels the free-energy terms: it refers to the second term as 'the energy of the electric polarization with susceptibility χE', but in Eq. (5) the second term is the DMI term DfL and the polarization energy appears only through F_E. This should be reworded.
- [Sec. III, units paragraph] The units listing for γ ('in units of inverse polarization') and the parenthetical '(in ,' before 'the second is the energy' appear incomplete; add the missing units for A and χE.
- [Appendix B] In the finite-difference description, the identification of nodes (i,0) with (i,N_z) and (i,N_z+1) with (i,2) is stated without a figure reference; label the computational grid nodes as referenced in Fig. 4 so that the periodic boundary conditions are unambiguous.
Circularity Check
No significant circularity: the central polarization profile is a direct variational minimum of the stated free energy, and the detectability threshold is a stated conditional, not a fitted target.
full rationale
The paper's central derivations are self-contained variational calculations. Eq. (4) for p(z) is obtained by minimizing the free energy (3) with respect to p, using the tanh solution of the GL equation (2) for the magnetization; no parameter is fitted to the predicted polarization. Likewise, Eqs. (7) and (10) follow directly from minimizing the DMI free energy with respect to p. The magnetoelectric coupling gamma is an input taken from a literature range, and the statement that a measurable polarization requires gamma_chiE above about 1e-19 sm/A is a threshold condition, not a postdiction; footnote 34 explicitly says 'We focus on the necessary value of coupling constant that would provide measurable polarization.' That is an acknowledged model limitation, not a circular step. The self-citation to Ref. 16 (phase dislocations) is used only as background contrast in the Introduction and is not load-bearing for the new GL/DMI mechanisms. The GL equation is attributed to Mills (Ref. 28), and the boundary conditions are stated. Thus the derivation does not reduce to its inputs by construction.
Assumptions & free parameters
free parameters (5)
- gamma_chiE_product =
range 1e-23 to 1e-13 sm/A from literature; no specific material
- m_infinity_bulk_magnetization =
10-100 kA/m for the FM path; 10-1000 kA/m for DMI paths
- xi_correlation_length =
about 10 nm
- DMI_strength_D_and_anisotropies_K_k =
D=1, K=1 or -1, k=1 in dimensionless Figures 2 and 3
- DMI_profile_scale_lambda =
lambda=4 in Fig. 3; otherwise arbitrary
assumptions (5)
- standard math The magnetization near a surface follows the Ginzburg-Landau solution m = m_infinity tanh((z+z0)/(sqrt(2) xi)) with boundary condition f(0)=a0 f'(0) from Mills.
- domain assumption The free energy contains a magnetoelectric coupling p dot (gamma grad(m^2) + gamma' Lifshitz terms) with a nonzero gamma in collinear magnets.
- domain assumption The electric susceptibility chiE is constant and the polarization is subdominant, so its gradient/stiffness energy can be neglected.
- domain assumption The DMI near surfaces is described by Lifshitz invariants with an inhomogeneous profile f(z) that decays into the bulk.
- ad hoc to paper For K>0, the axisymmetric localized solution with theta(0)=pi, theta(infinity)=0 and periodic boundary conditions in z is appropriate for the film geometry.
Cite this review
Pith. "Pith review of Theory of surface-induced multiferroicity in magnetic materials, thin films and multilayers." pith.science (2026). https://pith.science/paper/5L72H3PG
@misc{pith2026190800602,
author = {Pith},
title = {Pith review of: Theory of surface-induced multiferroicity in magnetic materials, thin films and multilayers},
year = {2026},
howpublished = {\url{https://pith.science/paper/5L72H3PG}},
note = {Machine review of arXiv:1908.00602}
}
read the original abstract
We present a theoretical study of the onset of electric polarization close to a surface in magnetic materials and in thin films and multilayers. We consider two different paths that lead to the onset of multiferroic behavior at the boundary in materials that are bulk collinear ferromagnets or antiferromagnets. These two paths are distinguished by the presence or absence of a surface induced Dzyaloshinskii-Moriya interaction which can be taken into account through Lifshitz invariants in the free energy of the system. Experimental consequences are discussed in the light of the developed theory.
Figures
Figures from the paper (1 more)
Reference graph
Works this paper leans on
- [1]
-
[2]
K. F. Wang, J.-M. Liu, and Z. F. Ren, Adv. in Phys. 58 , 321 (2009)
work page 2009
- [3]
- [4]
- [5]
- [6]
- [7]
- [8]
Show all 42 references
-
[9]
Hur et al
N. Hur et al. , Nature 429 , 392 (2004)
2004
-
[10]
Lawes et al
G. Lawes et al. , Phys. Rev. Lett. 95 , 087205 (2005)
2005
-
[11]
Katsura, N
H. Katsura, N. Nagaosa, and A. V. Balatsky, Phys. Rev. Lett. 95 , 057205 (2005)
2005
-
[12]
Mostovoy, Phys
M. Mostovoy, Phys. Rev. Lett. 96 , 067601 (2006)
2006
-
[13]
Wu et al
H. Wu et al. , Phys. Rev. Lett. 102 , 026404 (2009)
2009
-
[14]
I. A. Sergienko, C. Sen, and E. Dagotto, Phys. Rev. Lett. 97 , 227204 (2006)
2006
-
[15]
L. C. Chapon et al. , Phys. Rev. Lett. 96 , 097601 (2006)
2006
-
[16]
J. J. Betouras, G. Giovannetti, and J. van den Brink, Phys. Rev. Lett. 98 , 257602 (2007)
2007
-
[17]
Soumyanarayanan, N
A. Soumyanarayanan, N. Reyren, A. Fert, and C. Panagopoulos, Nature 539 , 510 (2016)
2016
-
[18]
Gerhard et al , Nat
L. Gerhard et al , Nat. Nanotechnol. 5 , 792-797 (2010)
2010
-
[19]
Maruyama et al
T. Maruyama et al. , Nat. Nanotech. 4 ,158-161 (2009)
2009
-
[20]
Z\^ u ti\' c et al
I. Z\^ u ti\' c et al. , Materials Today 22 , 85 (2019)
2019
-
[21]
Langridge et al
S. Langridge et al. , Phys. Rev. Lett. 112 , 167201 (2014)
2014
-
[22]
Sonntag et al
A. Sonntag et al. , Phys. Rev. Lett. 112 , 017204 (2014)
2014
-
[23]
C. G. Duan et al. , Phys. Rev. Lett. 101 , 137201 (2008)
2008
-
[24]
Nakamura, et al
K. Nakamura, et al. Phys. Rev. Lett. 102 , 187201 (2009)
2009
-
[25]
Mills, Phys
D.L. Mills, Phys. Rev. B 3 , 3887 (1971)
1971
-
[26]
K. Xia, W. Zhang, M. Lu, and H. Zhai, Phys. Rev. B 55 , 12561 (1997)
1997
-
[27]
Crepieux and C
A. Crepieux and C. Lacroix, J. Magn. Magn. Mat. 182 , 341 (1998)
1998
-
[28]
A. N. Bogdanov and U. K. Rossler, Phys. Rev. Lett. 87 , 037203 (2001)
2001
-
[29]
A. O. Leonov and M. Mostovoy, Nat. Commun. 6 , 8275 (2015)
2015
-
[30]
Due to the physical assumption that the dominant order parameter is the magnetic one, the stiffness energy part of the polarisation can be neglected
-
[31]
A. N. Bogdanov, U. K. Rossler, and K.-H. Mueller, J. Magn. Magn. Mat. 238 , 155 (2002)
2002
-
[32]
A. N. Bogdanov and D. A. Yablonsky, Sov. Phys. JETP 68 , 101 (1989); A. N. Bogdanov JETP Lett. 62 , 247 (1995)
1989
-
[33]
Hu, L.-Q
J.-M. Hu, L.-Q. Chen, and C.-W. Nan, Adv. Mater. 28 , 15 (2016)
2016
-
[34]
Details of the calculation in Supplemental Material
-
[35]
Valencia et al
S. Valencia et al. , Nat. Mater. 10 , 753 (2011)
2011
-
[36]
Chen et al
G. Chen et al. , Nat. Commun. 4 , 2671 (2013)
2013
-
[37]
Moreau-Luchaire et al
C. Moreau-Luchaire et al. , Nat. Nanotech. 11 , 444 (2016)
2016
-
[38]
Emori et al
S. Emori et al. , Nat. Mater. 12 , 611 (2013)
2013
-
[39]
Emori et al
S. Emori et al. , Phys. Rev. B 90 , 184427 (2014)
2014
-
[40]
Legrand et al
W. Legrand et al. , Sci. Adv. 4 : eaat0415 (2018)
2018
-
[41]
Maccariello et al
D. Maccariello et al. , Nat. Nanotech. 13 , 233 (2018)
2018
-
[42]
D. L. Maslov and A. V. Chubukov,
Reviewed August 14, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.