Magnetic anisotropy related to hybridization between Fe 3$d$ and As 4$p$ orbitals in a bcc Fe-As thin film

Akira Yasui; Atsushi Fujimori; Karumuri Sriharsha; Le Duc Anh; Masaaki Tanaka; Masaki Kobayashi; Miho Kitamura; Ryo Okano; Seiji Aota; Takahito Takeda

REVIEW 3 major objections 3 minor 2 references

Magnetic anisotropy related to hybridization between Fe 3$d$ and As 4$p$ orbitals in a bcc Fe-As thin film

T0 review · 3 major / 3 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read A bcc Fe-As thin film's in-plane magnetic anisotropy comes from an anisotropic magnetic dipole term driven by strain-induced p–d hybridization, not from orbital-moment anisotropy.

desk verdict New XMCD data on Fe-As are worth a look, but the paper's own equations say its central in-plane anisotropy conclusion has the wrong sign. read the letter →

arxiv 2505.05924 v1 pith:4IRKEMLX submitted 2025-05-09 cond-mat.mtrl-sci

Takahito Takeda , Karumuri Sriharsha , Seiji Aota , Ryo Okano , Le Duc Anh , Yukiharu Takeda , Akira Yasui , Miho Kitamura

show 4 more authors

Yuki K. Wakabayashi Atsushi Fujimori Masaaki Tanaka Masaki Kobayashi

This is my paper · ORCID

classification cond-mat.mtrl-sci

keywords bccFe-Asthinfilmin-planemagneticanisotropyXMCDsumrulesdipoletermp-dhybridizationepitaxialstrainMCD-HAXPESmagnetocrystalline

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 claims that the in-plane magnetic anisotropy of a bcc Fe-As thin film on GaAs(111)B is caused by an anisotropic magnetic dipole term in the Fe 3d spin density, a mechanism distinct from the orbital-moment anisotropy usually invoked for Fe alloys. Element-specific XMCD at the Fe $L_{2,3}$ edge shows that the orbital moment is isotropic within experimental accuracy, while the magnetic dipole term $m_T$ differs between out-of-plane and in-plane magnetization. The authors attribute that difference to compressive epitaxial strain, which lowers the symmetry at the As sites, splits the As 4p orbitals, and makes the Fe 3d–As 4p hybridization stronger in the film plane than along [111]; XMCD at the As edge confirms the As moment is induced antiparallel to Fe through this hybridization. Converting the measured $m_T$ anisotropy with the extended Bruno model gives $E_\mathrm{UMA} = -5.5 \pm 0.5$ MJ/m$^3$, which they take as the in-plane easy axis. If the mechanism holds, strain and non-magnetic dopant choice become knobs for controlling magnetic anisotropy in 3d transition metal films.

What carries the argument

The load-bearing object is the magnetic dipole term $m_T$, extracted from angle-dependent XMCD sum rules: $m_T$ measures the anisotropy of the spin-density distribution, and in this film it is the only moment component that changes with magnetization direction. The angle dependence is analyzed through $m_T(\theta) = m_{T,\perp}\cos^2\theta + m_{T,\parallel}\sin^2\theta$ with $m_{T,\perp} + 2m_{T,\parallel} = 0$, so measuring at $\theta = 0^\circ$ and $\theta = 60^\circ$ determines the full anisotropy. The companion identity is the extended Bruno model, which splits the uniaxial anisotropy energy into an orbital-moment term and a spin-density term; because $\Delta m_\mathrm{orb} \approx 0$, the second term carries the entire anisotropy.

What would settle it

Grow bcc Fe-As films with the same As fraction on substrates that impose different in-plane strains (or on a strain-relaxed buffer) and measure $\Delta m_T$ by angle-dependent XMCD: the strain-splitting mechanism predicts $\Delta m_T$ and the in-plane easy axis should weaken or vanish as the compressive strain goes to zero. An independent check would be a magnetometry measurement of the easy-axis direction on the same sample: if the easy axis is perpendicular while the paper's formula gives negative $E_\mathrm{UMA}$, the sign convention in the model needs revision.

Watch

Extended reading notes

Core claim

The central discovery is that in bcc Fe-As the magnetic anisotropy is carried by the magnetic dipole term, not by the orbital moment. Using XMCD sum rules at two incidence angles, the authors find $m_\mathrm{orb} = 0.22 \pm 0.01$ $\mu_B$/Fe at both $\theta = 0^\circ$ and $\theta = 90^\circ$, while $m_T$ changes from $0.02 \pm 0.01$ to $-0.01 \pm 0.01$ $\mu_B$/Fe. This finite $\Delta m_T$, together with a near-zero $\Delta m_\mathrm{orb}$, selects the spin-density term in the extended Bruno model as the source of the uniaxial anisotropy. They estimate $E_\mathrm{UMA} = -5.5 \pm 0.5$ MJ/m$^3$, which is several times larger than the shape anisotropy, and conclude that the film has in-plane magnetization because of this intrinsic term. Microscopically, the compressive strain (out-of-plane lattice constant 0.501 nm, in-plane 0.399 nm) splits the As 4p orbitals into $p_z$ and $p_x/p_y$ states and increases the Fe 3d–As 4p overlap in the plane, producing an anisotropic spin-density distribution and hence an anisotropic $m_T$.

Load-bearing premise

The quantitative anisotropy energy rests on the extended Bruno model with assumed spin-orbit coupling $\xi = 0.05$ eV and exchange splitting $\Delta_\mathrm{ex} = 1.5$ eV, plus the sign convention that negative $E_\mathrm{UMA}$ means an in-plane easy axis; if those parameters or the convention are wrong, the magnitude—and possibly the direction—of the claimed anisotropy would change.

Editorial extensions

If this is right

Orbital-moment-only models of magnetic anisotropy will not describe Fe-As; the spin-density (magnetic dipole) term must be included.
The intrinsic uniaxial anisotropy ($-5.5$ MJ/m$^3$) dominates both shape anisotropy ($-1.3$ MJ/m$^3$) and cubic anisotropy, so the in-plane easy axis is set by the $m_T$ mechanism.
Epitaxial strain magnitude and sign should provide a direct control knob for the magnetic dipole term and therefore for the anisotropy.
Other magnetic 3d films doped with non-magnetic p-block elements should exhibit the same strain-dependent in-plane anisotropy mechanism if their p states hybridize with the host d states.
Hole doping by As couples carrier concentration to magnetism in Fe-As, which could be relevant for electrically tunable magnetocrystalline anisotropy.

Reading between the lines

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

If the strain-splitting picture is right, then a film with the same As concentration grown with no compressive strain should show a much smaller $\Delta m_T$ and a weaker or rotated easy axis; measuring that would separate strain effects from chemical doping effects.
The mechanism suggests a search criterion for other materials: dopants with p states near the host d bands and with large strain-induced $p_z$–$p_{x,y}$ splitting should produce magnetic-dipole-term anisotropy.
A density-functional calculation of the Fe 3d–As 4p hybridization under the measured strain could test whether the proposed orbital occupancies ($4p_x^1 4p_y^1 4p_z^{1.6}$) and the antiparallel As moment are quantitatively consistent with the observed $\Delta m_T$.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Desk Editor's Note

read the letter

Dear [name],

I've read the Fe-As anisotropy paper. The experimental side is the strongest part: new element-resolved XMCD on bcc Fe-As showing a small As moment, a clean angle-dependent separation of m_orb and m_T, and a plausible story that anisotropic p-d hybridization under strain gives an anisotropic magnetic dipole term. That is a genuinely new measurement, and the MCD-HAXPES complement helps.

The problem is the quantitative conclusion. The paper defines EUMA as E_U,perpendicular minus E_U,parallel, with perpendicular along [111] and parallel in-plane, then reports EUMA = -5.5 MJ/m^3 and says the negative sign means IMA. That is backwards: if EUMA is the difference, a negative value means the out-of-plane direction is lower in energy (PMA). The confusion comes from Eq. (10) having -K_u(M.u)^2 while Eq. (12) writes E_U,theta = K_u(M.u)^2 with no minus sign. They have conflated K_u with E_perpendicular minus E_parallel. If the -5.5 value is actually K_u in the E=-K_u cos^2 theta convention, then negative K_u does give IMA, so the physical conclusion might be right, but that is not what the text says. As written, a reader following their definitions would draw the opposite anisotropy direction.

There is also a minor circularity: n_h and Delta_n_h are set by forcing the XMCD sum-rule magnetization to match SQUID, so those numbers are not independent. That limits the quantitative meaning of mFe and Delta_n_h, though the m_T anisotropy that drives the MA estimate is measured directly.

Overall, the data are valuable and the mechanism is plausible, but the sign error is load-bearing. I would not accept the paper in this form. The authors need to fix the definition of EUMA/K_u and check what the extended Bruno model actually outputs. If they do that, the core claim may hold. I would send it to a referee who knows the anisotropy conventions, but it is not publishable as written.

All the best.

Referee Report

3 major / 3 minor

Summary. The manuscript reports an XMCD and MCD-HAXPES study of a bcc Fe-As thin film grown on GaAs(111)B, which exhibits in-plane magnetic anisotropy (IMA). The Fe L2,3 XMCD spectra resemble those of Fe metal, while the As L2,3 XMCD shows a small magnetic moment on As that is antiferromagnetically coupled to Fe, attributed to Fe 3d–As 4p hybridization. Sum-rule analysis yields an isotropic orbital magnetic moment and an anisotropic magnetic dipole term m_T. The authors use the extended Bruno model with the measured Δm_T to estimate a uniaxial anisotropy energy EUMA = -5.5 ± 0.5 MJ/m3 and state that this negative value indicates IMA, proposing that anisotropic p-d hybridization under epitaxial strain produces the m_T anisotropy and thereby the observed IMA.

Significance. If the conclusion holds, the paper would present a new mechanism for magnetic anisotropy in Fe-based alloys: anisotropy of the magnetic dipole term driven by p-d hybridization with non-magnetic dopants, rather than the usual orbital-moment or spin-orbit mechanisms. The element-specific As L-edge XMCD data and the observation of anisotropic m_T are valuable experimental contributions. However, the central quantitative claim is undermined by a sign inconsistency between the definition of EUMA and its interpretation, which affects the main physics conclusion.

major comments (3)

[Section 3, Eqs. (10) and (12) and the definition of EUMA] The paper defines EUMA = E_U,perp - E_U,par, with ⊥ along [111] and ∥ along [112-], and then states that a negative value of EUMA indicates IMA. This is internally inconsistent: under the paper's own definition, a negative E_perp - E_par means the perpendicular direction is lower in energy, i.e., perpendicular magnetic anisotropy (PMA). The inconsistency likely stems from the sign difference between Eq. (10), which contains -K_u(M·u/M)^2, and Eq. (12), which writes E_U,θ = K_u(M·u/M)^2 without the minus sign. Thus the reported value EUMA = -5.5 MJ/m3, if taken literally, contradicts the observed IMA and the paper's own SQUID and XMCD-H data. The authors must correct the sign convention, reconcile Eqs. (10) and (12), and re-derive the sign of the anisotropy energy before the quantitative claim can be accepted.
[Section 3, Eqs. (8)-(9) and Table I] The value of Δn_h is obtained by enforcing msat,XMCD = msat,SQUID, i.e., by fitting the sum-rule magnetization to the SQUID value. The resulting mFe, mspin, and m_T in Table I are therefore calibration outcomes, not independent measurements. In particular, the absolute values of m_T (and hence the magnitude of EUMA) scale with the assumed n_h and the correction factor r. The angular dependence of m_T is less affected by this normalization, but the paper should clearly state that this is a calibration procedure and discuss how the uncertainty in Δn_h and r propagates into the EUMA estimate.
[Section 3, extended Bruno model parameters] The numerical estimate EUMA = -5.5 MJ/m3 rests on the extended Bruno model with ξ = 0.05 eV taken from Fe3+ ions and Δ_ex = 1.5 eV from metallic Fe. The film is a metallic Fe-As alloy with a Fe 3d occupation near 6.7, so the transferability of these parameters is not obvious. The paper should justify these choices for the specific Fe-As system or provide a range of values to show that the sign and order of magnitude of EUMA are robust.

minor comments (3)

[Throughout] There are duplicated equation numbers (two equations labeled (10)) and a typo in Eq. (3)' where the label appears as (3)′ and the text references Eq. (3)′ inconsistently. Please renumber and clean up the equation labels.
[Section 3, sum-rule analysis] Table I lists m_T at θ = 90°, but the angular-dependent XMCD data were measured at θ = 0° and 60°. The extrapolation to θ = 90° via Eqs. (3) and (4) should be stated explicitly in the table caption or text.
[Fig. 6(b) and text] The text refers to XMCD-H curves of As obtained by subtracting L3 from L2 XMCD. The sign of the L2 XMCD is opposite to L3, and the subtraction might produce artifacts; please specify the normalization procedure and show the raw L2 and L3 curves separately or as supplementary material.

Circularity Check

1 steps flagged · score 4.0 of 10

Partial circularity: mFe/Δnh are rescaled to the SQUID saturation moment, but the central ΔmT-based IMA claim is independent; a sign inconsistency in EUMA is a correctness risk, not circularity.

fitted input called prediction [Section 3, Eqs. (8)-(9), Table I]
"Here, we use the approximation that nh is represented by Δnh + 6.61 (nh of Fe metal is 6.61 [31]) and r is 1 (r of Fe metal is 1 [31]). Δnh is estimated at 0.12±0.03 from Eqs. (1), (2), (7) and (9)."

The XMCD sum-rule moments in Eqs. (1)-(2) all scale linearly with the hole number nh. Eq. (9) is not an independent measurement but the constraint msat,SQUID = (m_ferro/(m_ferro+m_para)) × 0.75 mFe, obtained by setting the XMCD-derived saturation magnetization equal to the SQUID value. Solving for Δnh therefore rescales the XMCD integrals so that the quoted mFe (2.37 μB/Fe) and the 3d occupancy change reproduce SQUID by construction; Table I's mFe and Δnh are fitted inputs presented as estimates. The same nh rescales mT⊥ and mT∥ and thus the extended-Bruno EUMA magnitude. However, the sign and relative anisotropy of ΔmT come from angle-dependent XMCD integrals and the fixed relation mT⊥ + 2mT∥ = 0, so the central IMA attribution is not forced by the SQUID fit.

full rationale

The paper is mostly self-contained against external benchmarks: XMCD, SQUID, MCD-HAXPES, STEM, and XRD data are original, and the extended Bruno model is cited to standard literature. One secondary quantity is obtained by construction: the hole number Δnh (and with it mFe, mspin, and the magnitudes of mT) is solved by forcing the XMCD sum-rule magnetization to match SQUID msat through Eq. (9). This is a fitted input called an estimate, but it does not invalidate the anisotropy analysis because the direction of ΔmT is measured from angle-dependent XMCD and is unaffected by the nh scale; the paper's central conclusion that an anisotropic magnetic dipole term contributes to IMA is therefore not a reduction to its inputs. The self-citation to Aota et al. [28] supplies sample growth and structural identity, but the MA mechanism is derived from the present data and external models, so that citation is not load-bearing circularity. A separate, non-circular correctness risk: Eq. (10) writes the uniaxial term as −Ku(M·u/Ms)^2 while Eq. (12) writes EU,θ = Ku(M·u/Ms)^2; the statement 'The negative value of the EUMA indicates that Fe-As has IMA' is consistent only with the plus-sign convention of Eq. (12), and a reader following Eq. (10)'s sign would reach the opposite conclusion. This is an internal sign inconsistency rather than a circular derivation, so it does not raise the circularity score beyond the partial-fit level.

Assumptions & free parameters 1 free parameters · 5 assumptions · 0 invented entities

The central quantitative result rests on sum-rule calibration against SQUID, a symmetry assumption relating theta=0 and theta=90 dipole terms, and model parameters taken from Fe metal/Fe3+ ions. The most fragile input is the extended Bruno model and its sign convention, where the paper is internally inconsistent. No new entities are postulated.

free parameters (1)

nh (number of Fe 3d holes) = 6.73 +/- 0.03 (Delta_nh = 0.12 +/- 0.03)
Determined by equating XMCD sum-rule magnetization to SQUID msat (Eqs. 8-9); not independently measured.

assumptions (5)

domain assumption XMCD sum rules (Eqs. 1, 2) with r=1 and nh near 6.6 from Fe metal are valid for Fe-As.
Sum rules require calibration of the hole number; the paper sets r=1 from Ref. [31] and later adjusts nh to match SQUID.
domain assumption The magnetic dipole term obeys mT_perp + 2 mT_par = 0 and axial symmetry (Eqs. 3, 4), so measurements at theta=0 and theta=60 suffice.
Standard relation for the dipolar term, but no theta=90 measurement is reported; the extrapolation assumes uniaxial symmetry.
ad hoc to paper Extended Bruno model (Eq. 10) with xi=0.05 eV and Delta_ex=1.5 eV quantitatively applies to Fe-As.
Parameters from Fe3+ ions and Fe metal; no verification for Fe-As; sign of the resulting EUMA is inconsistent with the IMA conclusion in the text.
ad hoc to paper The cubic anisotropy energy of the Fe-As film is approximately that of bulk bcc Fe (K1 near 5x10^-4 MJ/m3).
Paper states the order of magnitude is assumed as 10^-4 MJ/m3, without measuring K1/K2 for the strained alloy.
domain assumption As magnetic moment is negligible in Eq. (8), mAs approx 0.
As XMCD intensity is small but finite; the paper neglects mAs when converting SQUID magnetization to mFe.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Magnetic anisotropy related to hybridization between Fe 3$d$ and As 4$p$ orbitals in a bcc Fe-As thin film." pith.science (2026). https://pith.science/paper/4IRKEMLX

@misc{pith2026250505924,
  author       = {Pith},
  title        = {Pith review of: Magnetic anisotropy related to hybridization between Fe 3$d$ and As 4$p$ orbitals in a bcc Fe-As thin film},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/4IRKEMLX}},
  note         = {Machine review of arXiv:2505.05924}
}

abstract

The magnetic anisotropy (MA) of Fe-based ferromagnetic thin films has been extensively studied for device applications. The examined material is a new Fe-based ferromagnetic thin film, bcc Fe$_{1-x}$As$_x$ (Fe-As) with the in-plane MA (IMA) grown on a GaAs (111)B substrate. The magnetic properties of the Fe-As thin film have been investigated by Xray magnetic circular dichroism (XMCD) and magnetic circular dichroism in hard X-ray photoemission spectroscopy (MCD-HAXPES) to elucidate the role of As ions in the IMA. The XMCD spectra at the Fe $L_{2,3}$ edge and MCD-HAXPES spectra of the Fe 2$p$ core level exhibit ferromagnetic and metallic features like Fe metal. The XMCD at the As $L_{2,3}$ edge demonstrates that the As ions contribute to the ferromagnetism of bcc Fe-As through the hybridization between the Fe 3$d$ and As 4$p$ orbitals. The estimations of the magnetic moments of Fe using the XMCD sum rules have revealed that the orbital magnetic moment is isotropic and the magnetic dipole term is anisotropic. The anisotropy of the magnetic dipole term can be attributed to the anisotropic $p-d$ hybridization due to epitaxial strain, contributing to the IMA of bcc Fe-As. Our findings enlighten the mechanism of the MA of the non-magnetic ion-doped bcc Fe thin film, which can be applied to other magnetic 3$d$ transition metal thin films doped with non-magnetic elements.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

2 extracted references · 2 canonical work pages

[2]

Tudu and A

B. Tudu and A. Tiwari, Recent Developments in Perpendicular Magnetic Anisotropy Thin Films for Data Storage Applications, Vacuum 146, 329 (2017). [3] I. Žutić, J. Fabian, and S. Das Sarma, Spintronics: Fundamentals and Applications, Rev. Mod. Phys. 76, 323 (2004). [4] S. V Halilov, A. Y. Perlov, P. M. Oppeneer, A. N. Yaresko, and V. N. Antonov, Magnetocry...

work page 2017
[14]

Microscopic Origin of Magnetic Anisotropy in Au/Co/Au Probed with X-Ray Magnetic Circular Dichroism,

J. Okabayashi, J. W. Koo, H. Sukegawa, S. Mitani, Y. Takagi, and T. Yokoyama, Perpendicular Magnetic Anisotropy at the Interface between Ultrathin Fe Film and MgO Studied by Angular-Dependent x-Ray Magnetic Circular Dichroism, Appl. Phys. Lett. 105, 122408 (2014). [15] C. ‐J. Lin and G. L. Gorman, Evaporated CoPt Alloy Films with Strong Perpendicular Magn...

work page arXiv 2014

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.

[1] [2]

Tudu and A

B. Tudu and A. Tiwari, Recent Developments in Perpendicular Magnetic Anisotropy Thin Films for Data Storage Applications, Vacuum 146, 329 (2017). [3] I. Žutić, J. Fabian, and S. Das Sarma, Spintronics: Fundamentals and Applications, Rev. Mod. Phys. 76, 323 (2004). [4] S. V Halilov, A. Y. Perlov, P. M. Oppeneer, A. N. Yaresko, and V. N. Antonov, Magnetocry...

work page 2017

[2] [14]

Microscopic Origin of Magnetic Anisotropy in Au/Co/Au Probed with X-Ray Magnetic Circular Dichroism,

J. Okabayashi, J. W. Koo, H. Sukegawa, S. Mitani, Y. Takagi, and T. Yokoyama, Perpendicular Magnetic Anisotropy at the Interface between Ultrathin Fe Film and MgO Studied by Angular-Dependent x-Ray Magnetic Circular Dichroism, Appl. Phys. Lett. 105, 122408 (2014). [15] C. ‐J. Lin and G. L. Gorman, Evaporated CoPt Alloy Films with Strong Perpendicular Magn...

work page arXiv 2014