Pith. sign in

REVIEW 4 major objections 4 minor 34 references

Probing the Axion-Photon-Dark Photon Interaction at Future $e^+e^-$ Colliders

T0 review · 4 major / 4 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 Future e+e- colliders could probe the axion-photon-dark photon coupling down to 10⁻⁴ GeV⁻¹ for dark photon masses around 10-230 GeV, using single-photon events with missing energy.

desk verdict Solid single-photon reach study for axion–dark-photon couplings at future e+e- colliders, but the advertised sensitivity ignores background systematics and the abstract promises a polarization analysis the text does not deliver. read the letter →

arxiv 2509.08733 v2 pith:HYFXRQUA submitted 2025-09-10 hep-ph

Chuan-Ren Chen , Yuan-Feng Hsieh , Van Que Tran This is my paper · ORCID
classification hep-ph
keywords axiondarkphotonsingle-photonplusmissingenergyfuturee+e-collidersILCCEPCFCC-eerecoilmass
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

Future electron-positron colliders can test a dark sector in which a light axion and a dark photon interact with the Standard Model photon through the dimension-five operator $\frac{1}{2} g_{a\gamma'\gamma} a F_{\mu\nu}\tilde F'^{\mu\nu}$. The paper's central claim is that the resulting single-photon-plus-missing-energy signal would let the ILC, CEPC, and FCC-ee probe the coupling $g_{a\gamma'\gamma}$ down to about $10^{-4}\,\mathrm{GeV}^{-1}$ for dark photon masses between roughly 10 and 230 GeV, about an order of magnitude below existing LEP II limits. The signal stands out from the Standard Model $\nu\bar\nu\gamma$ background through a sharp drop-off in the recoil-mass distribution, which also provides a direct measurement of the dark photon mass. These machines would therefore complement hadron-collider searches for dark photons decaying to muon pairs.

What carries the argument

The central object is the dimension-five interaction term $\frac{1}{2}g_{a\gamma'\gamma}aF_{\mu\nu}\tilde F'^{\mu\nu}$, which couples the axion to one Standard Model photon and one dark photon. The paper combines this with the kinetic-mixing parameter $\varepsilon$; in the regime $m_{\gamma'}\gtrsim 10$ GeV, $m_a=1$ MeV, and $\varepsilon$ comparable to $g_{a\gamma'\gamma}m_{\gamma'}$, the decay $\gamma'\to a\gamma$ dominates, making single-photon-plus-missing-energy the leading signature. The discriminating observable is the recoil mass $M_{\mathrm{recoil}}^2 = s - 2\sqrt{s}E_\gamma$, whose maximum value sets a sharp edge used both to separate signal from the $\nu\bar\nu\gamma$ background and to measure $m_{\gamma'}$; the analysis also applies a $Z$-mass recoil veto and a photon polar-angle cut $|\cos\theta_\gamma|<0.75$.

What would settle it

Run the single-photon search at $\sqrt{s}=250$ GeV with $2\,\mathrm{ab}^{-1}$ of data, require $|\cos\theta_\gamma|<0.75$, and reject recoil masses within $[m_Z-2\Gamma_Z,\,m_Z+2\Gamma_Z]$; if the observed event count matches the SM $\nu\bar\nu\gamma$ background within uncertainties, all $g_{a\gamma'\gamma}$ values on the paper's $5\sigma$ contour are excluded. The claim fails if the background systematic uncertainty is larger than the roughly 1 percent assumed, so a measurement of the $\nu\bar\nu\gamma$ background rate at the 1 percent level is the decisive check.

Watch

Extended reading notes

Core claim

The paper argues that whenever the dimension-five axion-photon-dark photon vertex exists, the processes $e^+e^-\to\gamma'\to\gamma a$ and $e^+e^-\to\gamma^*\to a\gamma'$ followed by $\gamma'\to\gamma a$ produce a single photon plus missing energy. For dark photon masses $m_{\gamma'}\gtrsim 10$ GeV and an axion mass of 1 MeV, the dark photon decay is dominated by the channel $\gamma'\to a\gamma$, so nearly every signal event has one hard photon and large missing energy. Using LEP II single-photon data, the paper derives an upper limit $g_{a\gamma'\gamma}\lesssim 10^{-3}$ GeV$^{-1}$ for $m_{\gamma'}\lesssim 100$ GeV, and it projects that ILC, CEPC, and FCC-ee will be sensitive down to about $10^{-4}$ GeV$^{-1}$ for $m_{\gamma'}$ between 10 and 230 GeV at the $5\sigma$ level, with longitudinal beam polarization at the ILC improving the significance by a factor of four. The recoil-mass distribution exhibits a sharp edge at $M_{\mathrm{recoil}}^{\max}$, from which the dark photon mass is reconstructed as $m_{\gamma'} = \sqrt{s - (M_{\mathrm{recoil}}^{\max})^2}$.

Load-bearing premise

The projections assume that for dark photon masses above about 10 GeV and an axion mass of 1 MeV, the dark photon decays almost exclusively to a photon plus an invisible axion, so every signal event gives one hard photon and missing energy; if that branching fraction is not near 100 percent, or the axion does not escape the detector, the projected $10^{-4}$ GeV$^{-1}$ sensitivity collapses.

Editorial extensions

If this is right

  • A null result from the ILC single-photon search with $2\,\mathrm{ab}^{-1}$ at $\sqrt{s}=250$ GeV would exclude $g_{a\gamma'\gamma}\gtrsim 10^{-4}$ GeV$^{-1}$ for dark photon masses of order 20-200 GeV, roughly an order of magnitude below the LEP II bound.
  • The sharp edge in the recoil-mass spectrum gives a direct, model-independent way to measure the dark photon mass in single-photon events without reconstructing the invisible axion.
  • Longitudinal beam polarization at the ILC would increase the signal significance by a factor of four, making the ILC's projected reach the strongest of the three colliders considered.
  • For dark photon masses above about 230 GeV, these $e^+e^-$ machines lose sensitivity in single-photon events because the dark photon cannot be produced on shell at $\sqrt{s}=240$-$250$ GeV.
  • LEP II data already place the strongest existing constraint in this model, excluding $g_{a\gamma'\gamma}$ above roughly $10^{-3}$ GeV$^{-1}$ for $m_{\gamma'}\lesssim 100$ GeV.

Reading between the lines

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

  • If the branching ratio $\mathrm{Br}(\gamma'\to a\gamma)$ is not close to one, the projected $10^{-4}$ GeV$^{-1}$ reach would degrade; mapping the reach as a function of $m_a$ and $\varepsilon$ would show how robust the claim is.
  • The recoil-mass edge technique is more general than this model: it could be used to measure the mass of any invisibly decaying resonance produced with a single photon at a lepton collider, such as a $Z'$ or a heavy neutral lepton.
  • Because the two production channels scale as $\varepsilon^2$ and $g_{a\gamma'\gamma}^2$ respectively, combining a single-photon measurement with a $\gamma'\to\ell^+\ell^-$ search at the same machine could separate the kinetic-mixing and axionic couplings.
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.

Referee Report

4 major / 4 minor

Summary. The paper studies the effective interaction vertex a-γ-γ' described by a dimension-five operator plus kinetic mixing, focusing on single-photon plus missing-energy signals at LEP II and at future e+e- colliders (ILC, CEPC, FCC-ee). The authors derive cross sections and decay widths, extract a LEP II bound from published L3 single-photon data, simulate signal and background with MadGraph, propose cuts on recoil mass and photon polar angle, and project 5σ sensitivity contours in the (m_γ', g_{aγγ'}) plane. They also propose determining m_γ' from the sharp edge of the recoil-mass distribution. The stated central result is that future colliders can probe g_{aγγ'} down to about 10^-4 GeV^-1 for m_γ' in the O(10-100) GeV range, under the assumption that the dark photon decays dominantly to γ plus an invisible, very light axion.

Significance. If the projected reach is robust, the paper identifies a clean and potentially powerful search channel for a well-motivated but less-studied coupling, and it provides a useful LEP II constraint derived from published data with explicit assumptions. The recoil-mass edge method is a nice observable for mass reconstruction. However, the central quantitative claim depends on a statistics-only significance estimate and on a branching-ratio assumption that is not mapped onto the plotted parameter space; the polarization statement in the abstract is not supported by the body of the paper. With these points addressed, the work would be a solid phenomenological contribution.

major comments (4)
  1. [Section III.B, significance formula and Fig. 5] The projected sensitivity is computed as Z = N_s / sqrt(N_b) with no systematic uncertainty term. After the cuts in Eqs. (7)-(8), the dominant e+e- -> nu nu gamma background leaves on the order of 10^6 events at the quoted luminosities, so sqrt(N_b) is about 10^3 while a 1% normalization uncertainty contributes about 10^4 events to the error budget. Because the signal yield scales as g^2, a 1-5% background systematic shifts the 5-sigma contour upward by roughly a factor of 3-7 in g. The paper neither provides a systematic budget nor explains why systematics are negligible. Since Fig. 5 and the abstract's 10^-4 GeV^-1 claim are the central results, this is a load-bearing issue that must be addressed, for example by including a systematic term in Z or by quoting contours for several assumed background uncertainties.
  2. [Abstract (submitted version) and Section III] The abstract accompanying the manuscript states that longitudinal beam polarization at the ILC can enhance the signal significance by a factor of four and provides the strongest projected reach. No polarization analysis appears anywhere in Sections II-IV, and the abstract in the full text omits this statement. This is a discrepancy between the claimed result and the actual content. Either the polarization study must be added, or the abstract claim must be removed.
  3. [Section II and Fig. 5] The projected contours in Fig. 5 are computed under the assumption Br(γ' -> a γ) ≈ 1, realized for ε ≪ g_{aγγ'} m_γ'. However, Fig. 5 is plotted in the (m_γ', g_{aγγ'}) plane without specifying the value of ε used or the region of that plane in which the branching-ratio assumption actually holds. Since Br depends on the ratio of the widths in Eqs. (2) and (5), the same point in Fig. 5 can give significantly different signal yields depending on ε. The paper should state the benchmark choice of ε for the contours or superimpose the region where Br(γ' -> a γ) ≈ 1 is valid.
  4. [Section III.B, Eq. (6) and Fig. 3] The recoil-mass edge relation m_γ' = sqrt(s - (M_recoil^max)^2) follows for the three-body process e+e- -> a γ' (on shell) -> a a γ. The two-body off-shell s-channel contribution e+e- -> γ'* -> γ a, which the authors also list as a signal channel, produces a photon with essentially fixed energy and recoil mass near m_a, not an edge at sqrt(s - m_γ'^2). The text and Fig. 3 do not state whether the simulated recoil distribution includes both channels and, if so, which contribution dominates after the cuts. The mass-reconstruction claim needs this separation to be made explicit.
minor comments (4)
  1. [Fig. 3 caption] The caption lists m_γ' = 200 GeV while the text in Section III.B refers to m_γ' = 220 GeV; these should be made consistent.
  2. [Section III.A] The phrase 'progress (2) dominates' should read 'process (2) dominates'.
  3. [Section III.B] The signal and background efficiencies after cuts are quoted only in relative terms ('background reduced by ~65%', 'signal efficiency 65% to 80%'); reporting the absolute cut flow and the resulting N_s and N_b would make the significance estimate reproducible.
  4. [Section II] The condition for γ' -> a γ dominance is stated verbally ('if the numerical values for ε and g are of the same size'); a quantitative comparison of Γ(γ' -> a γ) with Γ(γ' -> l+l-) would help the reader identify the valid parameter region.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the reach is computed from the explicit Lagrangian and external LEP data, not from fitted inputs or self-citation.

full rationale

No circular derivation step is present. The central claim (future sensitivity to g_aγ'γ around 10^-4 GeV^-1) is computed directly from the explicit Lagrangian in Eq. (1) using MadGraph matrix elements, the stated cuts in Eqs. (7)-(8), and the projected luminosities in Table I; the significance Z = N_s/√N_b is applied to raw event counts, and no fitted parameter is subsequently relabeled as a prediction. The LEP II constraint is obtained from external L3 single-photon data [30,31] via the CLs method, not from the paper's own model output. The dark-photon mass reconstruction in Eq. (6) is a kinematic identity following from M_recoil^2 = s - 2√s E_γ and involves no fitted input. The self-citations ([4], [22], [33]) motivate the operator or provide ancillary Z-pole and muon g-2 comparisons, but the numerical sensitivity chain does not depend on any unverified self-cited uniqueness theorem or fitted value; the relevant Lagrangian and decay widths are stated in the paper. The background-systematics caveat raised by a skeptical reader would affect the robustness of the projected reach but is not an instance of circular reasoning.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The axion and dark photon are pre-existing beyond-SM candidates; the paper introduces no new particles, forces, or conserved quantities. The model parameters (g, epsilon, m_a) are inputs from the effective Lagrangian and are scanned or set by hand rather than fitted to the central result.

free parameters (3)
  • m_a (axion mass) = 1 MeV (chosen by hand)
    Set to 1 MeV to ensure the axion is extremely light and escapes detection; the reach is insensitive to values up to about a GeV as long as the axion is light and stable on detector scales.
  • g_{a gamma' gamma} (axion-photon-dark photon coupling)
    The coupling is the target of the search; the paper scans its value and projects sensitivity as a function of it, rather than fitting it to data.
  • epsilon (kinetic mixing parameter)
    The kinetic mixing parameter is scanned to show constraints from LEP II and Z-pole data; in the future projections it is set small enough that the gamma' -> a gamma branching ratio is near unity.
assumptions (4)
  • domain assumption The effective Lagrangian in Eq. (1), with kinetic mixing epsilon and dimension-5 coupling g_{a gamma' gamma}, fully describes the relevant dark-sector interactions.
    The paper assumes no other operators or particles affect the signal processes; this is the standard effective field theory assumption for light dark sectors (Section II).
  • ad hoc to paper For m_gamma' greater than about 10 GeV and comparable epsilon and g, the decay gamma' -> a gamma dominates (branching ratio close to 1).
    This is the parameter-space focus stated in Section II; the projected reach assumes this branching ratio when computing signal yields.
  • domain assumption The axion is extremely light and escapes the detector, giving missing energy.
    Assumed in Section III A; if the axion decays inside the detector, the single-photon plus missing-energy signature is altered.
  • domain assumption The SM background is dominated by e+e- -> nu nu gamma, and the significance Z = N_s/sqrt(N_b) with no systematic uncertainties is applicable.
    Used in Section III B; at the planned luminosities of 2-13 ab^-1 systematic uncertainties are likely to dominate, which the paper does not address.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Probing the Axion-Photon-Dark Photon Interaction at Future $e^+e^-$ Colliders." pith.science (2026). https://pith.science/paper/HYFXRQUA

@misc{pith2026250908733,
  author       = {Pith},
  title        = {Pith review of: Probing the Axion-Photon-Dark Photon Interaction at Future $e^+e^-$ Colliders},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/HYFXRQUA}},
  note         = {Machine review of arXiv:2509.08733}
}
abstract

We study the interaction between photons, dark photons, and axions at future lepton colliders, focusing on single-photon events with missing energy as the experimental signature. We find that future facilities such as the ILC, CEPC, and FCC-ee will be sensitive to the axion--photon--dark photon coupling down to the order of $10^{-4}\, \mathrm{GeV}^{-1}$ for dark photon masses around $O(10~\mathrm{GeV})$, assuming that the axion is extremely light and escapes detection. We further show that longitudinal beam polarization at the ILC can enhance the signal significance by a factor of four, providing the strongest projected reach in the model parameter space. Existing constraints from LEP II are analyzed for comparison. Furthermore, the mass of dark photon can be determined by measuring the sharp drop-off in the distribution of the recoil mass.

Figures

Figures reproduced from arXiv: 2509.08733 by the authors.

Figure 1
Figure 1. Feynman diagrams for the single-photon plus missing energy signal. [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. Constraints from LEP single photon data (red line) projected on the ( [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Recoil mass distribution (left) and photon polar angle distribution (right). [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Signal cross section after selection cuts for [PITH_FULL_IMAGE:figures/full_fig_p009_4.png]
Figure 5
Figure 5. Figure 5: Sensitivity projections at ILC (red dashed), FCC-ee (blue dashed), and CEPC [PITH_FULL_IMAGE:figures/full_fig_p010_5.png]

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

34 extracted references · 5 canonical work pages

  1. [23]

    K. Ding, Y. Li, X. Liu, Y. Liu, C.-T. Lu, and B. Zhu, (2025), arXiv:2504.00383 [hep-ph]

    arXiv 2025

  2. [4]

    R. Fang, Z. Liu, and V. Q. Tran, (2025), In preparation

    work page 2025

  3. [1]

    K. Choi, S. H. Im, and C. Sub Shin, Ann. Rev. Nucl. Part. Sci.71, 225 (2021), arXiv:2012.05029 [hep-ph]

    arXiv 2021

  4. [2]

    Fabbrichesi, E

    M. Fabbrichesi, E. Gabrielli, and G. Lanfranchi, (2020), 10.1007/978-3-030-62519- 1, arXiv:2005.01515 [hep-ph]

    arXiv 2020

  5. [3]

    Kaneta, H.-S

    K. Kaneta, H.-S. Lee, and S. Yun, Phys. Rev. Lett.118, 101802 (2017), arXiv:1611.01466 [hep-ph]

    arXiv 2017

  6. [5]

    K.Kaneta, H.-S.Lee, andS.Yun,Phys.Rev.D95,115032(2017),arXiv:1704.07542 [hep-ph]

    arXiv 2017

  7. [6]

    K. Choi, S. Lee, H. Seong, and S. Yun, Phys. Rev. D101, 043007 (2020), arXiv:1806.09508 [hep-ph]. 11

    work page Pith review arXiv 2020

  8. [7]

    O. E. Kalashev, A. Kusenko, and E. Vitagliano, Phys. Rev. D99, 023002 (2019), arXiv:1808.05613 [hep-ph]

    arXiv 2019

Show all 34 references
  1. [8]

    K. Choi, H. Seong, and S. Yun, Phys. Rev. D102, 075024 (2020), arXiv:1911.00532 [hep-ph]

    2020 arXiv

  2. [9]

    A. Hook, G. Marques-Tavares, and Y. Tsai, Phys. Rev. Lett.124, 211801 (2020), arXiv:1912.08817 [hep-ph]

    2020 arXiv

  3. [10]

    Arias, A

    P. Arias, A. Arza, J. Jaeckel, and D. Vargas-Arancibia, JCAP05, 070 (2021), arXiv:2007.12585 [hep-ph]

    2021 arXiv

  4. [11]

    A. Hook, G. Marques-Tavares, and C. Ristow, JHEP06, 167 (2021), arXiv:2105.06476 [hep-ph]

    2021 arXiv

  5. [12]

    Domcke, K

    V. Domcke, K. Schmitz, and T. You, JHEP07, 126 (2022), arXiv:2108.11295 [hep- ph]

    2022 arXiv

  6. [13]

    J. C. Gutiérrez, B. J. Kavanagh, N. Castelló-Mor, F. J. Casas, J. M. Diego, E. Martínez-González, and R. V. Cortabitarte, (2021), arXiv:2112.11387 [hep-ph]

    2021 arXiv

  7. [14]

    Carenza, G

    P. Carenza, G. Lucente, and E. Vitagliano, Phys. Rev. D107, 083032 (2023), arXiv:2301.06560 [hep-ph]

    2023 arXiv

  8. [15]

    A. Hook, G. Marques-Tavares, and C. Ristow, JHEP05, 086 (2024), arXiv:2306.13135 [hep-ph]

    2024 arXiv

  9. [16]

    H. Hong, U. Min, M. Son, and T. You, JHEP03, 155 (2024), arXiv:2310.19544 [hep-ph]

    2024 arXiv

  10. [17]

    Broadberry, S

    E. Broadberry, S. Das, A. Hook, and G. M. Tavares, (2024), arXiv:2408.03370 [hep-ph]

    2024 arXiv

  11. [18]

    Biswas, A

    S. Biswas, A. Chatterjee, E. Gabrielli, and B. Mele, Phys. Rev. D100, 115040 (2019), arXiv:1906.10608 [hep-ph]

    2019 arXiv

  12. [19]

    Deniverville, H.-S

    P. Deniverville, H.-S. Lee, and Y.-M. Lee, Phys. Rev. D103, 075006 (2021), arXiv:2011.03276 [hep-ph]

    2021 arXiv

  13. [20]

    S. D. Lane, H.-S. Lee, and I. M. Lewis, (2023), arXiv:2305.00013 [hep-ph]. 12

    2023 arXiv

  14. [21]

    Jodłowski, Phys

    K. Jodłowski, Phys. Rev. D108, 115017 (2023), arXiv:2305.05710 [hep-ph]

    2023 arXiv

  15. [22]

    Chen, Y.-F

    C.-R. Chen, Y.-F. Hsieh, and C. S. Nugroho, (2024), arXiv:2405.19087 [hep-ph]

    2024

  16. [24]

    Fujiiet al

    K. Fujiiet al. (LCC Physics Working Group), (2019), arXiv:1908.11299 [hep-ex]

    2019 arXiv

  17. [25]

    Abadaet al

    A. Abadaet al. (FCC), Eur. Phys. J. ST228, 261 (2019)

    2019

  18. [26]

    Abdallah et al

    W. Abdallah et al. (CEPC Study Group), Radiat. Detect. Technol. Meth- ods8, 1 (2024), [Erratum: Radiat.Detect.Technol.Methods 9, 184–192 (2025)], arXiv:2312.14363 [physics.acc-ph]

    2024

  19. [27]

    Navaset al

    S. Navaset al. (Particle Data Group), Phys. Rev. D110, 030001 (2024)

    2024

  20. [28]

    Aaijet al

    R. Aaijet al. (LHCb), Phys. Rev. Lett.124, 041801 (2020), arXiv:1910.06926 [hep- ex]

    2020

  21. [29]

    A. M. Sirunyanet al. (CMS), Phys. Rev. Lett.124, 131802 (2020), arXiv:1912.04776 [hep-ex]

    2020 arXiv

  22. [30]

    Acciarriet al

    M. Acciarriet al. (L3), Phys. Lett. B470, 268 (1999), arXiv:hep-ex/9910009

    1999 arXiv

  23. [31]

    Achardet al

    P. Achardet al. (L3), Phys. Lett. B587, 16 (2004), arXiv:hep-ex/0402002

    2004 arXiv

  24. [32]

    Curtin, R

    D. Curtin, R. Essig, S. Gori, and J. Shelton, JHEP02, 157 (2015), arXiv:1412.0018 [hep-ph]

    2015 arXiv

  25. [33]

    M. Du, Z. Liu, and V. Q. Tran, JHEP05, 055 (2020), arXiv:1912.00422 [hep-ph]

    2020 arXiv

  26. [34]

    Alwall, R

    J. Alwall, R. Frederix, S. Frixione, V. Hirschi, F. Maltoni, O. Mattelaer, H. S. Shao, T. Stelzer, P. Torrielli, and M. Zaro, JHEP07, 079 (2014), arXiv:1405.0301 [hep-ph]. 13

    2014 arXiv

Pith tools

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