Pith. sign in

REVIEW 3 major objections 6 minor 3 references

Correlated topological-polarization surface states in the narrow-gap insulator FeSb2

T0 review · 3 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read The metallic surface states of the narrow-gap insulator FeSb2 are polar surface states of topological-polarization origin, controlled by bulk electron correlations, and electrostatic gating drives them into a ferromagnetic or possibly…

desk verdict Not a slam dunk, but a serious multi-technique case for correlated topological surface states in FeSb2; the transport evidence is correlative rather than unique, yet the paper deserves a real referee. read the letter →

arxiv 2608.05887 v1 pith:FFUSAG3N submitted 2026-08-06 cond-mat.str-el cond-mat.mtrl-sci

Takahiro Iwagaki , Hideki Matsuoka , Ginta Hoshino , Kanata Watanabe , Shungo Aoyagi , Shunsuke Kitou , Yuiga Nakamura , Motoaki Hirayama
show 2 more authors
Takashi Koretsune Naoya Kanazawa
This is my paper · ORCID
classification cond-mat.str-elcond-mat.mtrl-sci
keywords topologicalpolarizationFeSb2correlatedsurfacenonreciprocaltransportorbitalreconstructionobstructedatomicinsulatorelectrostaticgatingaltermagnetism
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

This paper argues that the metallic surface states of the narrow-gap insulator FeSb2 are polar surface states of topological-polarization origin, and that strong electron correlations in the bulk control when they exist and what they do. If true, FeSb2 becomes a working example of a correlated topological surface, a place where two usually separate worlds, band topology and strong interactions, meet on one surface. The authors support the claim by showing that the bulk's valence electron density carries part of its charge in interatomic regions, by imaging a temperature-driven reconstruction of the iron 3d orbital occupation that strengthens the polarization at low temperature, and by detecting nonreciprocal surface transport that turns on only in that low-temperature regime. Electrostatic gating then drives the same surface into a ferromagnetic or possibly altermagnetic state, while the polarization survives.

What carries the argument

The central object is topological polarization, the quantized displacement of valence Wannier charge centers into the interatomic region, equivalent to a nonzero Zak phase in the bulk bands. This displaced charge is what appears at a surface termination as floating polar surface states. The paper's measurement machinery is nonreciprocal transport: in a polar system the resistance acquires a term $R = R_0[1 + \gamma_0 (\mathbf{P}\times\mathbf{H})\cdot\mathbf{I}]$, detected as a second-harmonic resistance that is forbidden in the centrosymmetric bulk. The correlation control comes from the temperature dependence of the iron 3d orbital occupation, imaged in real space and reproduced by dynamical mean-field calculations, which modulates the amount of bonding charge available for the polarization.

What would settle it

An angle-resolved photoemission measurement on a clean FeSb2(101) surface that failed to find floating surface bands inside the gap, or a control experiment on a non-topological polar film that showed the same nonreciprocal transport, would falsify the assignment. A more targeted test: measure the second-harmonic resistance as a function of temperature through the orbital reconstruction onset and check whether its onset tracks the reconstructed orbital occupation shift in a single sample.

Watch

Extended reading notes

Core claim

In FeSb2, the metallic surface states seen in earlier experiments are assigned to polar surface states of topological-polarization origin. The bulk valence Wannier charge centers lie in interatomic positions rather than on the Fe and Sb atoms, so cutting the crystal exposes displaced charge as floating surface states within the gap. The paper further claims a bulk-edge correspondence in a correlated system: as temperature rises, a correlation-driven reconstruction of the Fe 3d orbital occupation mixes antibonding weight into the occupied states and depletes the bonding charge that generates the polarization, and the nonreciprocal transport signal of the polar surface disappears in step with that reconstruction. Finally, gate voltages above a few volts drive the correlated surface across a quantum phase transition into a ferromagnetic or possibly altermagnetic state, with the nonreciprocal signal persisting into the ordered phase.

Load-bearing premise

The claim rests on interpreting the measured second-harmonic nonreciprocal resistance as the signature of the topological-polarization surface state; if that signal actually came from the electric-double-layer interface or from a trivial inversion-breaking surface state, the bulk-edge correspondence conclusion would not follow from the data.

Editorial extensions

If this is right

  • The previously unexplained metallic surface states of FeSb2 are identified as topological-polarization states, resolving their origin and connecting them to the bulk electronic structure.
  • Bulk-edge correspondence holds in a correlated insulator: the surface's polar transport tracks the bulk orbital reconstruction, so the surface state is not a purely surface-localized phenomenon.
  • Electrostatic gating is a viable control knob for a correlated topological surface, driving it into a ferromagnetic or possibly altermagnetic state with transition temperatures up to 50 K.
  • Because topological polarization needs no heavy elements or spin-orbit coupling, the same physics should appear in other 3d transition-metal compounds with obstructed atomic insulator character, including FeSi.
  • The correlated surface, with its narrow bands and tunable carrier density, sits between correlated oxides and moiré systems as a platform for two-dimensional strongly correlated phases such as superconductivity or density waves.

Reading between the lines

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

  • A direct test of the topological-polarization assignment would be spin-resolved photoemission on a clean (101) surface to look for the predicted floating surface bands and their polarization; the paper infers the surface state from transport and calculations rather than imaging its dispersion.
  • If the surface is indeed altermagnetic, the anomalous Hall effect should show a characteristic dependence on the in-plane field direction relative to the crystal axes, which the current data do not yet isolate.
  • The persistence of the nonreciprocal signal across the magnetic transition suggests that the polarization and the order parameter are coupled; measuring the nonreciprocal coefficient as a function of gate voltage through the transition could reveal whether the magnetic order enhances or competes with the polarization.
  • The same nonreciprocal transport protocol could be applied to other correlated narrow-gap insulators predicted to be obstructed atomic insulators, offering a bulk-transport substitute for surface-sensitive spectroscopy in materials where clean surfaces are hard to prepare.
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

3 major / 6 minor

Summary. The paper reports a combined computational and experimental study of FeSb2, a narrow-gap 3d insulator, and argues that its (101) surface hosts metallic polar surface states of topological-polarization origin, generated by the bulk Zak phase / obstructed-atomic-insulator character rather than by spin-orbit coupling. PBE slab calculations show surface bands crossing the bulk gap with charge floating off the outermost atoms; a WCC Wilson-loop analysis places valence Wannier centers in interatomic positions. Temperature-resolved synchrotron CDFS maps show a redistribution of Fe 3d valence electron density between 100 and 200 K, quantified by CHOD as a decrease of the d_xy/d_yz occupation and an increase of the d_z2 occupation, which the authors attribute to correlations and reproduce qualitatively with DFT+DMFT (U=5 eV, J=0.8 eV). In ionic-liquid-gated 23.6-nm FeSb2(101) films, the authors observe a gate-tunable surface conduction channel, a second-harmonic nonreciprocal resistance appearing below about 50 K and peaking near 25 K, and, above 4 V gate bias, a hysteretic anomalous Hall effect indicating an emergent surface magnetic order with Tc up to 50 K. They interpret these observations as demonstrations that the topological-polarization surface states inherit the bulk correlation-driven orbital reconstruction (bulk-edge correspondence) and that gating drives the surface into a ferromagnetic or possibly altermagnetic state.

Significance. If the interpretation is correct, this is a significant advance: it would be one of the first explicit demonstrations that a topological polarization, rather than spin-orbit band inversion, can stabilize metallic surface states in a correlated 3d compound, and that the surface states can be tuned electrostatically through a magnetic quantum phase transition. The paper's computational backbone is partly parameter-free: the PBE WCC and slab calculations use no adjustable parameters, the CDFS maps are raw diffraction-derived density maps, and the DFT+DMFT calculations use standard values of U and J. The experiments are internally consistent in that the gate-induced resistance drop, carrier-density increase, nonreciprocal transport, and anomalous Hall effect are qualitatively reproduced in a second film. However, the central claim of 'direct evidence' currently relies on a correlation between temperature scales and on a transport fingerprint that is not unique to topological-polarization surface states; the strongest claim is therefore conditional. If the identified gaps are addressed in revision, the work would merit publication in a high-impact journal.

major comments (3)
  1. [First-principles calculations on bulk and surface electronic states in FeSb2; Fig. 1g–i] The topological-polarization assignment is based entirely on PBE-DFT WCC and slab calculations, while the correlation-driven reconstruction that the paper claims governs this polarization is established by DFT+DMFT with U = 5.0 eV, J = 0.8 eV (Methods). The WCC/Zak-phase invariant and the floating surface-state wavefunctions are never recomputed with these interactions. Since the paper's central concept is a correlated topological polarization whose magnitude is asserted to be modulated by the bulk orbital reconstruction, the survival of the invariant under strong correlations is precisely what must be demonstrated; without such a calculation or a clear adiabatic-continuity argument, the step from 'noninteracting WCC' to 'correlated bulk-edge correspondence' is not established.
  2. [Nonreciprocal transport from the polar surface state; Methods, Reproducibility] The measured second-harmonic resistance is interpreted through R = R0[1 + γ0(P×H)·I], but this form applies to any inversion-broken conducting channel. The EDLT interface, involving a strongly polar [DEME][TFSI] ionic liquid on a chemically reactive FeSb2 surface, could itself produce such a signal. The Methods report that in film B the nonreciprocal signal reverses sign upon gating, with the origin left unexplained; a fixed bulk polarization would not be expected to reverse sign under gating, whereas a trivial interface channel could. The transport data therefore do not uniquely identify the topological-polarization surface state as the current-carrying channel, and the phrase 'direct evidence' overstates what the experiment demonstrates without additional corroboration, such as a direct probe of the surface-state dispersion or a systematic control on the interface condition.
  3. [Nonreciprocal transport from the polar surface state; Fig. 3d,e compared with Fig. 2a–d] The claimed bulk-edge correspondence is supported mainly by a temperature coincidence—nonreciprocal transport emerges below about 50 K, whereas the CDFS/CHOD orbital reconstruction occurs between 100 and 200 K. The paper does not provide a quantitative relation between γ(T) and the reconstructed occupancies, and the two temperature scales differ by roughly a factor of two. A direct link, such as a calculation of the WCC or surface-state weight as a function of the correlated orbital occupations, or a control measurement on a surface whose bulk reconstruction is suppressed, is needed to elevate the observed concurrence from correlation to evidence.
minor comments (6)
  1. [Methods, CHOD analysis] The minimization expression contains garbled notation ('minimize,,.' and misplaced sub-/superscripts); please restore the intended mathematical form.
  2. [Fig. 1 and main text, bulk band structure discussion] 'e.g., G-Z' and 'X=-S?' should be typeset as Γ–Z and X–S (or an explicit k-path label) to be readable.
  3. [Author contributions] 'G.H. and K.K.' names a K.K. who is not in the author list; this is presumably a typo for T.K. or K.W. and should be corrected.
  4. [Methods, Device fabrication] 'The ionic liquids used in this study was' should read 'The ionic liquid used in this study was' or 'The ionic liquids used in this study were'.
  5. [Supplementary figure references] Supplementary Fig. 3 is referenced both for thermal expansion and for θ–2θ XRD; please renumber to avoid ambiguity.
  6. [Data availability] The repository URL appears as a placeholder 'https://...'; the full link should be provided.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: DFT/WCC, CDFS/CHOD, and transport are independent legs; the film-B sign reversal is a validation gap, not a circular reduction.

full rationale

The central derivation chain is not circular. The topological-polarization assignment rests on an independent PBE slab/WCC calculation (Fig. 1g–i) plus standard bulk-edge correspondence; the transport experiment is a separate measurement, and its nonreciprocal signal is a generic consequence of inversion breaking, not an input to the WCC calculation. The orbital reconstruction is first seen in raw CDFS valence-electron-density maps (Fig. 2a–c) and only quantified by the CHOD fit (ref. 49), so the fit does not manufacture the trend. DFT+DMFT is a separate parameterized calculation (U = 5 eV, J = 0.8 eV) using standard external solvers; it is not fitted to the transport data. The main experimental evidence (gamma onset, anomalous Hall effect) is compared with theory after the fact, not derived from it. The admitted sign reversal of the nonreciprocal signal in film B (Methods) is an unexplained limitation that weakens the uniqueness of the transport interpretation, but it is a falsifiability concern, not a circular reduction. The self-citations (CHOD method, FeSi analogy) are not load-bearing: the CHOD method is a data-analysis tool whose qualitative conclusion is visible in the unmodeled maps, and the FeSi comparison is contextual. No equation in the paper defines a target quantity in terms of itself, and no fitted parameter is renamed as a prediction.

Assumptions & free parameters 4 free parameters · 7 assumptions · 0 invented entities

The central claim rests on standard band-theory results (bulk-boundary correspondence), DFT approximations, DMFT parameters, and interpretive assumptions about transport and Hall signals. No fundamentally new entity is introduced; the main burden is the assignment of measured signals to specific surface-state properties, not the postulation of new particles or forces.

free parameters (4)
  • DFT+DMFT on-site Coulomb interaction U = 5.0 eV
    Set by hand/prior convention; controls the correlation-driven spectral redistribution used to support the orbital reconstruction interpretation.
  • DFT+DMFT Hund coupling J = 0.8 eV
    Set by hand/prior convention; used in the DMFT impurity solver.
  • CHOD orbital populations and radial magnification kappa = not stated numerically
    CHOD minimizes a weighted residual between a model density and the CDFS density; the resulting d-orbital occupancies in Fig. 2d are fit outputs, with a magnification factor kappa and a cutoff radius of 0.6 A selected.
  • Surface doping active thickness = 1 nm
    Used to convert measured sheet carrier density to electrons per formula unit; chosen as an EDLT screening length. The magnetic-transition threshold argument depends on this number.
assumptions (7)
  • standard math Bulk-boundary correspondence: displaced Wannier charge centers in the bulk necessarily produce surface states at a termination.
    Used to interpret the slab bands and WCCs in Fig. 1g-i and to assign the surface states to topological polarization.
  • domain assumption PBE DFT is an adequate approximation for the ground-state electronic structure and topological polarization of FeSb2.
    All WCC and slab calculations use PBE; no experimental measurement of the surface-state dispersion is provided to validate this.
  • domain assumption The second-harmonic resistance formula R = R0[1 + gamma0 (P x H) dot I] applies, and the measured R2omega originates from surface polarization rather than from other inversion-breaking effects.
    Central to the nonreciprocal transport interpretation in Fig. 3d-e; no direct microscopic check is provided.
  • domain assumption A hysteretic anomalous Hall effect is sufficient evidence of surface ferromagnetic or altermagnetic order.
    No magnetization measurement is reported, and extrinsic AHE mechanisms are not fully excluded.
  • domain assumption CDFS/CHOD analysis with [Ar] and [Kr]4d10 core configurations yields reliable orbital occupancies.
    The quantitative orbital reconstruction in Fig. 2d relies on this model; the Methods section acknowledges residual artifacts such as double scattering, absorption, and extinction.
  • domain assumption Gate-induced carriers are confined to about 1 nm below the surface, so the per-formula-unit doping estimate is valid.
    Used to argue that the doping level is below the bulk magnetic threshold and therefore points to the surface states.
  • domain assumption DFT+DMFT with the chosen U and J values captures the temperature-dependent spectral function and correlation-induced incoherence of FeSb2.
    Used to connect the orbital reconstruction to strong correlations and to argue the gap is filled by incoherent spectral weight at high temperatures.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Correlated topological-polarization surface states in the narrow-gap insulator FeSb2." pith.science (2026). https://pith.science/paper/FFUSAG3N

@misc{pith2026260805887,
  author       = {Pith},
  title        = {Pith review of: Correlated topological-polarization surface states in the narrow-gap insulator FeSb2},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/FFUSAG3N}},
  note         = {Machine review of arXiv:2608.05887}
}
read the original abstract

Strong electron correlations and band topology each generate rich quantum phases, but conflicting elemental requirements have largely kept them apart. Topological polarization offers a route to unite them, producing polar surface states from bonding charge without spin-orbit coupling and thereby extending band topology to correlated 3d transition-metal compounds. Here we demonstrate that epitaxial thin films of the narrow-gap insulator FeSb2 host metallic polar surface states of topological-polarization origin, governed by the strong correlations of the bulk. Nonreciprocal surface transport emerges only below the onset temperature of a correlation-driven reconstruction of the bulk Fe 3d orbital occupation, providing direct evidence of bulk-edge correspondence in a correlated topological system. Moreover, electrostatic gating drives this correlated surface across a quantum phase transition into a ferromagnetic or possibly altermagnetic state. Our results establish topological polarization as a design principle for correlated topological phases in a broad range of materials.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

3 extracted references · 2 canonical work pages

  1. [23]

    Xu, Y. et al. Three-dimensional real space invariants, obstructed atomic insulators and A new principle for active catalytic sites. Preprint at https://arxiv.org/abs/2111.02433 (2021). 24. Gao, J. et al. Unconventional materials: the mismatch between electronic charge centers and atomic positions. Sci. Bull. (Beijing) 67, 598–608 (2022). 25. Xu, Y. et al....

    arXiv 2021

  2. [43]

    Chikina, A. et al. Correlated electronic structure of colossal thermopower FeSb2 : An ARPES and ab initio study. Phys. Rev. Res. 2, 023190 (2020). 44. Tomczak, J. M., Haule, K. & Kotliar, G. Signatures of electronic correlations in iron silicide. Proc. Natl. Acad. Sci. U. S. A. 109, 3243–3246 (2012). 45. Tomczak, J. M. Thermoelectricity in correlated narr...

    work page Pith review arXiv 2020

  3. [64]

    #, 𝑑!", 𝑑#!, 𝑑!!+

    Krause, L. et al. Accurate high-resolution single-crystal diffraction data from a Pilatus3 X CdTe detector. J. Appl. Crystallogr. 53, 635–649 (2020). 65. Kresse, G. & Furthmüller, J. Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set. Comput. Mater. Sci. 6, 15–50 (1996). 66. Blöchl, P. E. Projector...

    work page 2020

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.