Pith. sign in

REVIEW 3 major objections 4 minor 17 references

Placing a two-dimensional metal next to a hyperbolic-polariton material suppresses its quasiparticle weight, cancels Fermi-velocity renormalization to leading order, and removes the leading second-order Cooper attraction.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · deepseek-v4-flash

2026-08-02 11:09 UTC pith:VTQNCMMG

load-bearing objection A careful analytic theory of Fermi-liquid coupling to hyperbolic polaritons; the self-energy results are new and credible, but the no-pairing conclusion rests on an unproven diagram-selection rule. the 3 major comments →

arxiv 2606.16674 v3 pith:VTQNCMMG submitted 2026-06-15 cond-mat.supr-con cond-mat.str-el

Effects on a metal that is proximately coupled to hyperbolic photon modes

Zhiyu Dong , Patrick A. Lee This is my paper
classification cond-mat.supr-con cond-mat.str-el PACS 71.10.Ay74.20.-z73.20.Mf78.20.Ci
keywords hyperbolic polaritonslongitudinal A·j couplingquasiparticle weightFermi velocity renormalizationspectral function sidebandCooper pairinghexagonal boron nitrideself-energy
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

This paper asks whether hyperbolic phonon-polariton modes in a nearby polar crystal, such as hexagonal boron nitride, can substantially alter the electronic properties of an adjacent two-dimensional metal. Because the coupling is through the longitudinal part of the vector potential, the interaction vertex is proportional to the energy difference between initial and final electron states, making it qualitatively different from electron-phonon coupling. The paper shows that this energy-dependent coupling produces a quasiparticle weight Z = 1/(1+λ0) but leaves the Fermi velocity unchanged to leading order, and it computes a spectral sideband that should be observable by photoemission. It then shows that the repulsive interaction mediated by these modes, while zero at first order on the Fermi surface, cancels exactly at the leading second order through three competing diagrams, so that hyperbolic polaritons do not induce pairing in a parabolic band.

Core claim

The central claim is that the longitudinal A·j coupling between a 2D Fermi liquid and hyperbolic polaritons is controlled by a single dimensionless parameter λ0, which suppresses the quasiparticle weight but does not renormalize the Fermi velocity at leading order. The cancellation follows from a precise relation between the frequency and momentum derivatives of the self-energy: Fω = -λ0 and Fk = +λ0, so the product Z(1+Fk) = 1. The paper further demonstrates that the HM-mediated interaction in the Cooper channel is repulsive and vanishes on the Fermi surface at first order; at second order, the crossed diagram, the Cooper ladder, and the two-photon A² vertex contribute with equal magnitude

What carries the argument

The key object is the interaction vertex g_{k,qκ} = e√(2ω_{qκ}ε∞)^{-1} (ξ_{k+q} - ξ_k)/q · f(q,κ,z), where the factor (ξ' - ξ) arises from converting the longitudinal current to the time derivative of density. The mode function f(q,κ,z) = (1/cosh h(qz)) times a normalization f_{qκ} sets the momentum cutoff 1/z. This vertex, combined with the hyperbolic-mode dispersion, yields the self-energy coefficients and the three second-order pairing diagrams whose cancellation is the paper's main nontrivial result.

Load-bearing premise

The quantitative value of the coupling λ0, and hence the magnitude of every predicted effect (quasiparticle weight, sideband intensity, pairing amplitude), rests on the hyperbolic-mode wave-function normalization and the κ-integration constant that the paper takes from the authors' earlier companion paper without re-deriving; if that normalization is inaccurate, all numbers scale by an order-one factor, though the qualitative cancellations are independent of it.

What would settle it

Measure the quasiparticle weight and Fermi velocity of a two-dimensional metal on hBN by ARPES as a function of distance z to the hBN layer: the paper predicts Z = 1/(1+λ0) with λ0 ∝ 1/z and an unrenormalized Fermi velocity, plus a sideband starting at the hyperbolic-mode frequency. A clear z-dependent velocity renormalization, or the absence of the sideband, would falsify the central claim. Additionally, a direct numerical evaluation of the three second-order Cooper-channel diagrams using the paper's own vertex but with an independently computed mode function would test whether the exact canc

Watch this falsifier. Get emailed when new claim-graph text bears on it.

Share X Bluesky LinkedIn Reddit HN

If this is right

  • A 2D metal on hBN should display an ARPES sideband starting at the hyperbolic-mode frequency, with the quasiparticle peak losing weight according to Z = 1/(1+λ0).
  • The Fermi velocity is not renormalized to leading order in λ0, so transport measurements (e.g., cyclotron mass or quantum oscillations) should show no change despite significant spectral-weight transfer.
  • The HM-mediated interaction cannot induce superconductivity at the leading second order for a parabolic band; the static part of the interaction is exactly canceled, so pairing would have to come from subleading dynamic terms.
  • The coupling strength λ0 depends only on the distance z to the polaritonic layer and the mode parameters, not on the metal's Fermi velocity or density, so the effect is universal across metals placed at the same distance.
  • Type-I and type-II hyperbolic modes contribute identically to the self-energy because the integrated mode-function normalization is the same, so materials with either type produce the same phenomenology.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The exact cancellation of the static second-order pairing relies on the parabolic-band identity ξ_q = q²/(2m); in a non-parabolic band (e.g., a tight-binding dispersion), the three diagrams will not cancel exactly, so a residual attractive channel may open — a testable prediction for materials with strong band curvature.
  • The same gauge argument implies that any purely static longitudinal coupling — not just hyperbolic polaritons — cannot renormalize a Fermi liquid's velocity or produce pairing at leading order; this suggests that retarded (finite-frequency) effects are essential for cavity-induced electronic phenomena.
  • For Bi2Se3 the paper estimates λ0 ≈ 0.9, which is not very small; subleading corrections beyond the leading-order cancellation may become numerically important, potentially restoring a small velocity renormalization or a weak pairing channel that would be observable in that material.
  • The spectral sideband predicted here should also be visible in tunneling (STS) as a dip-and-hump structure above the mode frequency, providing a complementary measurement to ARPES that could be performed on encapsulated samples.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 4 minor

Summary. The paper analyzes a 2D metallic Fermi liquid separated by a gap 2z from hBN hyperbolic phonon polaritons, coupled via the longitudinal part of the A·j interaction. Using the continuity equation, the vertex is Eq. (8), proportional to (ξ_{k+q}−ξ_k)/q. For a parabolic band the authors compute the self-energy derivatives F_ω = −λ0 and F_k = λ0 (Eqs. (14)–(17)), giving quasiparticle weight Z = 1/(1+λ0) and, via Eq. (19), no leading Fermi-velocity renormalization. λ0 is expressed in Eq. (15) in terms of z and material parameters, with numerical estimates λ0 ≈ 0.34 for hBN and ≈ 0.9 for Bi2Se3. The spectral function is calculated in Section III and shows a sideband above the average HM frequency ¯ω (Eqs. (20)–(25), Fig. 1). Section IV analyzes Cooper pairing: the first-order HM-mediated interaction is repulsive and vanishes on the Fermi surface; the authors claim that the remaining second-order static contributions from the crossed, Cooper-ladder, and A² diagrams cancel, so no leading-order pairing occurs. The paper attributes this cancellation to the static limit of the longitudinal vector potential being a pure gauge.

Significance. If the results hold, the paper identifies a qualitatively distinct electron-boson coupling regime: a vertex that grows with excitation energy suppresses quasiparticle weight and creates an observable spectral sideband while leaving v_F unrenormalized, and the same vertex suppresses leading Cooper pairing. This is clearly different from standard electron-phonon coupling and from previous Amperean-pairing proposals, and the material-specific estimates make the sideband a concrete ARPES target. The central self-energy derivation is internally consistent: the vertex follows from continuity, the F_ω and F_k calculations are explicit, and the velocity cancellation is shown algebraically for a parabolic band in Appendix A. The paper is also commendable for deriving λ0 from input parameters rather than fitting it, and for making falsifiable predictions. The main risks are two load-bearing points that need more support: the diagram-selection rule in Section IV and the normalization integral Eq. (13) imported from the companion paper Ref. [5].

major comments (3)
  1. [§IV and Fig. 2 caption] The no-pairing conclusion rests on the assertion that the bubble and vertex-correction diagrams 'vanish on the Fermi surface.' This is not derived, and the stated reason—that they 'contain at least one vertex proportional to ξ−ξ′'—is not obviously valid for second-order diagrams. In Eq. (8), every vertex connects an external line (ξ = 0) to an internal line, and therefore carries ξ_internal, not ξ−ξ′; the crossed diagram Eq. (27) itself retains factors ξ²_{p+q} and ξ²_{−p′+q}. Bubble and vertex-correction diagrams contain the same type of factors. Please enumerate the diagrams and show explicitly, by evaluation or power counting of the Cooper-channel kernels, that they vanish for external legs pinned to the Fermi surface. If they do not vanish, the cancellation in Eqs. (30)–(31) does not establish the absence of pairing.
  2. [§II, Eq. (13)] The magnitude of λ0—and hence the sideband spectral weight, lifetime estimates, and all pairing amplitudes—is set by the κ-integral in Eq. (13), which uses the mode-function normalization/interpolation f²_qκ taken from the companion paper Ref. [5] (listed as 'to appear'). Since Ref. [5] is not yet available, the O(1) prefactor in Eq. (13) cannot be checked from the present manuscript. Please provide a self-contained derivation of f²_qκ and of Eq. (13), or state precisely which approximation is being imported. This is a quantitative load-bearing point: if the prefactor is inaccurate, all numerical estimates scale by an O(1) factor, although the velocity cancellation and the pairing cancellation, which follow from the (ξ′−ξ)² vertex structure and the parabolic-band relation ξ_q = q²/2m, are more robust.
  3. [§IV, Eq. (31)] In the q_z < k_F regime, the reduction of V_cross + V_ladder to (2/(2m)²) relies on the relation ξ_{p+q} + ξ_{−p′+q} ≃ 2q²/2m and on the statement that 'after angular averaging over p, the first term vanishes.' This step is not shown, and it is not obvious for a Cooper pair with arbitrary incoming and outgoing directions. Please supply the angular-average derivation (along the lines of Appendix B for the first-order kernel) and state the phase-space conditions, including the treatment of p′ ≠ p, under which this reduction holds. This is part of the central cancellation claim.
minor comments (4)
  1. [§II, Eq. (4)] The notation f(q,κ,z) is used in Eq. (4), while f_qκ is used below; please define the relation between them and the mode-function normalization once, with consistent symbols.
  2. [§IV, after Fig. 2] The phrase 'three of the four diagrams involving bubble and vertex correction all vanish' is grammatically ambiguous. Please state explicitly which of the four diagrams in Fig. 2(b) vanish and which remain.
  3. [Ref. [16]] Reference [16] is an acknowledgment, not a published work; please label it as a private communication or move it to the acknowledgments.
  4. [§II, Eq. (14)] In the first line of Eq. (14), the angular integration is implicit; it would help to display the θ integral and state explicitly the conditions under which ξ²/(ξ+¯ω)² can be replaced by unity for both positive and negative ξ.

Circularity Check

0 steps flagged

No significant circularity: central results are derived from microscopic inputs; the self-cited normalization is not a fitted prediction.

full rationale

The central derivation is not circular. λ0 is not fitted: it is obtained from Eqs. (12)-(15) by integrating the vertex and mode function over q and κ with material inputs (ω_T, ω_L, η, ϵ∞, z, ω̄), and no parameter is adjusted to match the self-energy or pairing output. The velocity cancellation is a derived result: Fω = −λ0 and F_k = λ0 are computed from the same self-energy integral, and v'_F/v_F = Z(1+F_k) = 1 follows from the parabolic-band identity dξ_{k+q}/dξ_k = 1 + (q/k_F)cosφ and angular averaging (Appendix A, Eqs. A3–A7). The spectral-function sideband is computed from the same Σ′′ via Kramers–Kronig, with no fitted amplitude. The pairing cancellation is obtained by explicit summation of the crossed, ladder, and A² diagrams; the static cancellation reduces to the parabolic-band identity ξ_q²/q⁴ = 1/(2m)², not to a restatement of the input interaction. The only input taken from the authors' own companion paper is the interpolation formula for f²_{qκ} and the κ-integral Eq. (13); this is a parameter-free microscopic normalization that sets the overall magnitude of λ0 but is not tuned to the predicted quantities, and the qualitative cancellations do not depend on its precise value. The Section IV assertion that bubble and vertex-correction diagrams vanish is an unproven selection rule; that is a correctness risk rather than a circular step, because it is not equivalent by construction to the paper's conclusions.

Axiom & Free-Parameter Ledger

2 free parameters · 5 axioms · 0 invented entities

The paper introduces no new entities: the hyperbolic phonon polariton is an established mode. It does import a normalization function from the authors' prior work, assumes a parabolic single-band Fermi liquid, and relies on a stated regime ω̄ ≪ v_F/z. These are the main hidden costs; no parameters are fitted to the target predictions.

free parameters (2)
  • average HM frequency ω̄ = type II hBN: 170 meV; type I hBN: 96 meV; Bi2Se3 type-II ~ 10 meV
    The full HM dispersion ω_{qκ} is replaced by an average ω̄ to evaluate the self-energy, spectral function, and pairing integrals. The leading-order results assume ω̄ ≪ Λ_z = v_F/z, and the quantitative λ0 scales as 1/ω̄.
  • metal–hBN separation z = 0.6 nm for hBN estimates; 3.0 nm for Bi2Se3 estimates
    z sets the momentum cutoff via cosh⁻²(qz) and enters λ0 ∝ 1/z. It is a realistic interface-distance choice made by hand, not fitted to the target results.
axioms (5)
  • ad hoc to paper Mode-function normalization/interpolation f²_{qκ} = (2η²/ω_L²)/(1+(ω_T/ω_L)² x⁴) and the κ-integral Eq. (13) are taken from Ref. [5].
    This input is not derived in the present paper; it fixes λ0 and all spectral intensities. It is self-cited from a companion paper by one of the authors.
  • domain assumption Single parabolic band with trivial Bloch form factor F(k,q)=1; multi-band extension only noted.
    Used for the vertex Eq. (8), the self-energy, and the exact pairing cancellation via identities such as ξ_q = q²/(2m). The Bi2Se3 estimate involves a more complex band structure, so the no-pairing result is not directly transferable.
  • domain assumption The average HM frequency ω̄ is small compared with electronic scales Λ_z = v_F/z, so ω̄ can be dropped in energy denominators.
    This yields F_ω=−λ0 and F_k=λ0. The paper concedes the spectral-function calculation breaks down for α<1, i.e., when this assumption fails.
  • standard math Coulomb gauge, minimal coupling A·j with longitudinal current from continuity, and Peierls substitution for the A² vertex.
    Standard QED-in-media framework; both A·j and A² vertices are needed for gauge invariance and for the cancellation in the pairing section.
  • domain assumption Second-order diagram counting: bubble and vertex-correction diagrams vanish on the Fermi surface, leaving only the crossed, Cooper-ladder, and A² diagrams.
    Asserted in the Fig. 2 caption with a one-line reason ('contain at least one vertex proportional to ξ−ξ′'). This selection drives the pairing-cancellation conclusion and is not proven in detail.

pith-pipeline@v1.3.0-alltime-deepseek · 16026 in / 19383 out tokens · 208543 ms · 2026-08-02T11:09:44.220263+00:00 · methodology

0 comments
Cite this review

Pith. "Pith review of Effects on a metal that is proximately coupled to hyperbolic photon modes." pith.science (2026). https://pith.science/paper/VTQNCMMG

@misc{pith2026260616674,
  author       = {Pith},
  title        = {Pith review of: Effects on a metal that is proximately coupled to hyperbolic photon modes},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VTQNCMMG}},
  note         = {Machine review of arXiv:2606.16674}
}
Share X Bluesky LinkedIn Reddit HN
read the original abstract

The hyperbolic mode (HM) refers to a polariton mode in a polar insulator where the dielectric function is negative in some direction of propagation. Within a frequency window the light occupies a greatly expanded region in momentum space. The HM in hexagonal Boron Nitride (hBN) has been under intense study and we consider placing a metal directly on top of hBN and ask whether its physical properties can be strongly affected. While the problem resembles superficially the electron phonon coupling problem, there are important differences. Due to the longitudinal nature of the HM mode the coupling is driven by time dependent charge fluctuations which results in a coupling that strongly increases with the energy difference of the initial and final states. We find a significant frequency and momentum dependence of the self energy which allows us to identify the dimensionless coupling constant $\lambda_0$ that controls this effect. There is a suppression of the quasi-particle weight but it turns out that the leading correction to the velocity renormalization is canceled. We compute the single particle spectral function which shows a side band that can be measured experimentally. The virtual exchange of HM leads to a repulsive interaction which is ineffective to leading order because of the energy dependence. We are motivated to seek pairing by going to second order. Unfortunately we find that the leading contributions exactly cancel. This cancellation is not an accident and we give an explanation of why this cancellation would take place.

Figures

Figures reproduced from arXiv: 2606.16674 by Patrick A. Lee, Zhiyu Dong.

Figure 1
Figure 1. Figure 1: FIG. 1. Positive-frequency HM continuum contribution to the [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2. Diagrammatic illustration of the effective low-energy [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗

Review history (2 revisions) →

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

17 extracted references · 2 linked inside Pith

  1. [1]

    Basov, A

    D. Basov, A. Asenjo-Garcia, P. J. Schuck, X. Zhu, A. Ru- bio, A. Cavalleri, M. Delor, M. M. Fogler, and M. Liu, Polaritonic quantum matter, Nanophotonics14, 3723 (2025)

    2025

  2. [2]

    H. M. Bretscher, L. Graziotto, M. H. Michael, A. Mon- tanaro, I.-T. Lu, A. Grankin, J. W. McIver, J. Faist, D. Fausti, M. Eckstein, et al., Fluctuation engi- neering in cavity quantum materials, arXiv preprint arXiv:2604.08666 (2026)

    Pith/arXiv arXiv 2026

  3. [3]

    Appugliese, J

    F. Appugliese, J. Enkner, G. L. Paravicini-Bagliani, M. Beck, C. Reichl, W. Wegscheider, G. Scalari, C. Ciuti, and J. Faist, Breakdown of topological protection by cav- ity vacuum fields in the integer quantum hall effect, Sci- ence375, 1030 (2022)

    2022

  4. [4]

    Keren, T

    I. Keren, T. A. Webb, S. Zhang, J. Xu, D. Sun, B. S. Kim, D. Shin, S. S. Zhang, J. Zhang, G. Pereira, et al., Cavity- altered superconductivity, Nature650, 864 (2026)

    2026

  5. [5]

    P. A. Lee, Effect of the hyperbolic photon mode on the metal insulator transition in a proximate material, ArXiv , to appear (2026)

    2026

  6. [6]

    Ashida, A

    Y. Ashida, A. ˙Imamo˘ glu, and E. Demler, Cavity quan- tum electrodynamics at arbitrary light-matter coupling strengths, Physical Review Letters126, 153603 (2021)

    2021

  7. [7]

    Kohn and J

    W. Kohn and J. M. Luttinger, New mechanism for su- perconductivity, Phys. Rev. Lett.15, 524 (1965)

    1965

  8. [8]

    C. J. Eckhardt, A. Grankin, D. M. Kennes, M. Ruggen- thaler, A. Rubio, M. A. Sentef, M. Hafezi, and M. H. Michael, Surface-mediated ultrastrong cavity coupling of two-dimensional itinerant electrons, Physical Review Letters135, 156902 (2025)

    2025

  9. [9]

    Gatti, C

    G. Gatti, C. Berthod, J. Issing, M. Straub, S. Mand- loi, Y. Alexanian, J. Avila, P. Dudin, T. Kim, M. Wat- son, et al., Electron-phonon coupling across the tmd/hbn van der waals interface, arXiv preprint arXiv:2604.24139 (2026)

    Pith/arXiv arXiv 2026

  10. [10]

    Riolo, A

    R. Riolo, A. Tomadin, G. Mazza, R. Asgari, A. H. Mac- Donald, and M. Polini, Tuning fermi liquids with polari- tonic cavities, Proceedings of the National Academy of Sciences122, e2407995122 (2025)

    2025

  11. [11]

    Riolo, F

    R. Riolo, F. H. Koppens, P. Jarillo-Herrero, G. Mazza, A. H. MacDonald, and M. Polini, Polariton-induced superconductivity in two-dimensional metals, arXiv preprint arXiv:2511.00608 (2025)

    arXiv 2025

  12. [12]

    G. M. Andolina, A. De Pasquale, F. M. D. Pellegrino, I. Torre, F. H. Koppens, and M. Polini, Amperean super- conductivity cannot be induced by deep subwavelength cavities in a two-dimensional material, Physical Review B109, 104513 (2024)

    2024

  13. [13]

    S.-S. Lee, P. A. Lee, and T. Senthil, Amperean pairing instability in the u (1) spin liquid state with fermi surface and application¡? format?¿ toκ-(bedt-ttf) 2 cu 2 (cn) 3, Physical review letters98, 067006 (2007)

    2007

  14. [14]

    P. A. Lee, Amperean pairing and the pseudogap phase of cuprate superconductors, Physical Review X4, 031017 (2014)

    2014

  15. [15]

    Schlawin, A

    F. Schlawin, A. Cavalleri, and D. Jaksch, Cavity- mediated electron-photon superconductivity, Physical re- view letters122, 133602 (2019)

    2019

  16. [16]

    We thank Taige Wang for pointing out this cancellation, which was overlooked in our first draft

  17. [17]

    J.-S. Wu, D. N. Basov, and M. M. Fogler, Topological insulators are tunable waveguides for hyperbolic polari- tons, Physical Review B92, 205430 (2015)

    2015