REVIEW 3 major objections 4 minor 5 cited by
Ensuring that toponium is glued, not nailed
T0 review · 3 major / 4 minor · reviewed 2026-08-12 · deepseek-v4-flash
Pith's one-line read The LHC's near-threshold top-antitop excess is compatible with QCD toponium, not a new short-range force.
desk verdict A physically transparent qualitative diagnostic separating QCD toponium from a contact-bound state, but the quantitative 'compatible with CMS' claim rests on a normalization without an uncertainty budget. 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 central object is the modified phase-space factor in Eq. (A.8), the ttbar relative-velocity factor that replaces the free beta_t in the hard gg to ttbar cross section. It encodes the near-threshold dynamics through the nonrelativistic Green's function: the Sommerfeld Coulomb enhancement plus a sum over n=1..5 Coulomb-like eta_t bound states, where eta_t is the ground-state toponium. For generic potentials the paper solves the same Green's-function equation numerically (Eq. A.11) on a grid, and the contact-interaction case is treated with a regulated delta potential that supports exactly one bound state. The argument turns on comparing the Green's function at r=0, which controls the line shape, between QCD and short-range binding.
What would settle it
Look at the ttbar invariant-mass line shape near threshold with resolution good enough to see the dip between the eta_t peak and the continuum. If the binding energy is near 2.5 to 3 GeV, the QCD tower predicts the excited states add about 1 pbarn and fill the dip; observing a single isolated resonance with a clean dip, or a cross-section several times larger than the NNLO QCD prediction at the same binding energy, would falsify the QCD-toponium interpretation and point to a short-range force.
Extended reading notes
Core claim
The central claim is that the observed CMS excess is standard QCD toponium, and the tell-tale that distinguishes QCD 'glue' from an exotic short-range 'nail' is the excited-state tower. On the QCD side, the paper builds the near-threshold color-singlet gg to ttbar cross section from the nonrelativistic Green's function (Eq. A.8), which includes both the Sommerfeld enhancement and a sum over the first five Coulomb-like s-wave bound states, with the ground-state eta_t plus its n=2..5 excitations. With the binding energy in the 2-3 GeV range favored by pNRQCD, these excited states fill the threshold dip that a lone eta_t resonance would show, adding about 1 pbarn to the integrated cross section. The contrast with a contact interaction is sharp: a regulated delta potential that binds at a comparable energy produces a wavefunction concentrated at the origin, and because the Green's function in Eq. (1) is evaluated at r=0, the cross section is much larger than observed; tuning the contact coupling down to match the observed cross section leaves a state with BE around 1 GeV that is washed out by the 2 Gamma_t of about 3 GeV width. The paper therefore concludes that the CMS excess is compatible with QCD-glued toponium, and that a short-range interaction invoked to explain the excess, or a fraction of it, would not display a bound state. It also estimates the precision needed to constrain contact-interaction Wilson coefficients: about 1 percent on the eta_t mass, or better than about 10 percent on the excess cross section, would begin to be sensitive.
Load-bearing premise
The prediction assumes that near-threshold color-singlet ttbar production factorizes into a hard parton-level cross section times a modified phase-space factor built from the nonrelativistic Green's function with only the first five Coulomb-like bound states included; the paper itself notes this factorization is still under active investigation.
Editorial extensions
If this is right
- If the toponium binding energy reaches about 3 GeV, the ground-state peak separates from threshold by more than the width, and observing the dip between peak and continuum filled by excited states would confirm the QCD tower over a single short-range bound state.
- For smaller binding energies, the excited states add only a tenuous increase, at most about 1 pbarn, to the proton-proton cross section near threshold.
- A contact interaction that binds at a QCD-like energy would produce a much larger cross section than observed, so the CMS excess does not support an isolated short-range bound state.
- Constraining new-physics Wilson coefficients from the eta_t mass would require about 1 percent precision on the peak position, while about 10 percent precision on the excess cross section may already begin to be sensitive.
Reading between the lines
- As an editorial extension, a future e+e- scan across the top threshold could directly test the dip-filling prediction: a filled dip with Coulombic level spacing would confirm the glued tower, while an isolated peak would point to a short-range force.
- The same wavefunction-at-origin comparison should carry over to any system where a contact interaction competes with a Coulomb-like force, so the glued-versus-nailed diagnostic is a template for exotic heavy-flavour or dark-matter bound states.
- The dependence on only the first five Coulombic states is testable: computing the line shape with a full Richardson or Cornell potential, which the paper notes can push excited states above threshold, would show whether the dip remains filled, and if it does not, the comparison with the CMS excess would shift.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper argues that the near-threshold ttbar cross-section excess reported by CMS is compatible with standard QCD-bound toponium ('glued') rather than with a state bound by a new short-range interaction ('nailed'). The argument combines three ingredients: (i) the binding energy of the eta_t is estimated from pNRQCD potentials to lie in the range giving a peak separated from threshold at the level of the top width; (ii) the Coulomb-like QCD spectrum contains an excited-state tower that fills the threshold dip and contributes roughly one additional pbarn to the integrated cross-section; and (iii) a short-range contact interaction, having a wavefunction much more concentrated at the origin, either produces a much larger cross-section for the same binding energy or, if matched to the observed cross-section, yields a state too broad to be visible. The line shapes are computed by replacing the free phase space with a modified phase-space factor built from the nonrelativistic Green's function (Eq. A.8), folded with a simple gluon PDF (Eq. A.2).
Significance. If the quantitative compatibility claim holds, the paper makes a useful contribution: it sharpens a falsifiable line-shape diagnostic (an isolated resonance with a filled-in dip signals the QCD excited-state tower, whereas a single short-range bound state would not fill the dip), and it explains why the wavefunction-at-the-origin difference between Coulomb-like and contact potentials controls the cross-section ordering. The author is transparent about using leading-order parton cross-sections, a 2004-era PDF parametrization, and a factorized phase-space treatment, and explicitly flags the factorization assumption as an active research subject. These choices make the numerical results estimates rather than precision predictions, but the qualitative contrast is well motivated and worth publishing if the central quantitative claim is suitably qualified.
major comments (3)
- [Section 3, Table 1 and Appendix A (Eqs. A.1, A.2, A.8)] The central quantitative claim in conclusion (iii) of Section 6 rests on the absolute normalization of the cross-section, but the manuscript provides no uncertainty budget for any of the ingredients that set that normalization. The gluon luminosity is the 2004 CTEQ5-style parametrization of Eq. (A.2) with no PDF error bands; Eq. (2) is a leading-order alpha_s^2 hard cross-section; only the gg initial state is included; and the soft and hard scales are fixed at specific values with no variation. In addition, the alpha_s <= 0.5 saturation described in Appendix A introduces an unquantified artifact in the potential calculation. Since the QCD entries in Table 1 (8.3 and 9.1 pb) lie only about 1.2-2.0 pb above the CMS central value of 7.1(0.8) pb, a moderate downward shift from any of these sources would remove the claimed compatibility. An error estimate, or a reformulation of conclusion (iii) as an order-of-magnitude statement, is needed before the quantitative comparison can be considered established.
- [Appendix A, after Eq. (A.8)] The factorization ansatz underlying all numerical line shapes is explicitly acknowledged in the manuscript to 'remain the subject of active investigation' [37], yet no estimate is given for the size of the corrections to this factorization. The additional cross-section attributed to toponium is literally the difference between the modified phase space and the bare phase space (Table 1: 4.9 pb for Sommerfeld-only vs 8.3-9.1 pb with bound states), so factorization corrections are first-order, not a small perturbation. The n_max = 5 truncation is also not varied, and color-octet contributions are only mentioned qualitatively in Figure 4. Since conclusion (iii) is quantitative, the manuscript should either propagate an estimate of these uncertainties or explicitly downgrade the compatibility claim to a plausibility argument.
- [Section 4, Eqs. (5)-(7) and Table 1] The short-range cross-sections used to argue that a new interaction would not display a bound state for the observed excess are computed for two illustrative regulator scales (M = 1 TeV and M = 13 GeV), with the Wilson coefficient tuned to produce a chosen binding energy, but no regulator dependence or matching uncertainty is provided. The statement in conclusion (iii) that a short-range interaction invoked to explain the excess 'will not display a bound state' is therefore tied to these particular examples and to the same absolute normalization criticized above. A scan over mediator masses and couplings, or an analytic argument removing the normalization dependence, is required to make the exclusion robust.
minor comments (4)
- [Section 2, Eq. (3)] The sign convention for the binding energy is inconsistent: Eq. (3) defines '-BE' with a negative sign, while Table 2 and the text discuss positive BE values; please make the sign convention uniform.
- [Table 2] Table 2 is difficult to read: the raised/lowered scale-variation values are not separated clearly from the central values. Please reformat the table so that each eigenstate and its scale variation are presented in a reader-friendly way.
- [Figure 2 caption] The sentence 'Above 3.5 (100 GeV)^2 where beta_t ~ 0.2' is unclear; the axis unit and the physical threshold location should be stated explicitly.
- [Section 6 and reference [19]] The claim that the LO line shape is 'surprisingly close' to the high-order calculation of Ref. [19] is not quantified; a direct comparison or a table of peak positions and widths would make this valuable point more convincing.
Circularity Check
Minor self-citations appear in the BSM discussion, but the central glued-vs-nailed comparison is computed from external QCD inputs and benchmarked to CMS data, so no load-bearing circularity.
-
other
[Section 5, Section 4, and Appendix A (Refs. [27], [33])]
"Recently [27] we have examined the e ffect of such contact interactions ... for numerical work, and also for short- but not zero-range V, a regulated form of the contact potential is a Yukawa exchange, VBSM = αcontact/|r'−r| e^{−Λ|r'−r|} can be used [27] ... The NNLO one can be found in [33] and refs. therein."
These are the only self-referential elements. Ref. [27] supplies the Yukawa regulator used in the numerical contact-potential comparison and the HEFT-bound context invoked against the nailed option; Ref. [33] supplies the NNLO potential expression used in the spectrum. They do not drive the central QCD-compatibility claim: the binding energies are cross-checked against external high-order calculations ([17,18,19]), and the CMS excess is used as a benchmark rather than a fitted input. The step is therefore a minor, peripheral self-citation rather than a load-bearing circularity, and the score reflects that mild self-reference without treating the main derivation as circular.
full rationale
The derivation chain runs from the parton-level gg->tt cross-section (Eq. 2), the gluon luminosity parametrization (Eqs. A.1-A.2), and the Fadin-Khoze modified phase space (Eq. A.8) into which the t-quark width and Coulomb/pNRQCD bound-state energies are inserted. The binding energies in Table 2 are computed from standard LO/NLO/NNLO potentials with explicit scale choices, and the CMS near-threshold excess is used only as a comparison value, not as an input fitted to fix the model. The contact-interaction ('nailed') comparison solves the Schrödinger Green's function (Eq. A.11) with a delta or Yukawa potential, with C tuned to a chosen BE; the resulting cross-section is a distinct predicted observable. The claim that a ~7 pb excess would not display a short-range bound state follows from the computed BE-dependence (BE ~1 GeV, below the 3 GeV width), not from fitting CMS. The factorization ansatz behind Eq. A.8 is explicitly flagged in Appendix A as 'remains the subject of active investigation'; this is an accuracy limitation, not a circular reduction. The paper is also anchored to external benchmarks: the QCD line shape is compared with high-order calculations [18,19], and the CMS data provide an external falsification target. The only self-referential inputs are Refs. [27] and [33], which are peripheral to the central glued-vs-nailed comparison. Accordingly, the circularity score is 2, reflecting minor self-citation, not a structurally circular derivation.
Assumptions & free parameters
free parameters (4)
- nmax (number of included toponium excited states) =
5
- Renormalization scales (mu_soft and mu_hard) =
mu_soft ~ 25 GeV (alpha_s ~ 0.147); mu_hard ~ 172 GeV (alpha_s ~ 0.108)
- alpha_s saturation ceiling =
0.5
- Contact-interaction Wilson coefficient C(Lambda) in illustrative BSM cases =
Tuned to produce BE=1 or 2 GeV at M=13 GeV (with C=3.7 in one case); C(1 TeV) values tuned to BE around 2.5 GeV but…
assumptions (7)
- domain assumption Nonrelativistic Schrodinger equation with pNRQCD static potentials (Eq. A.3-A.4) describes the ttbar bound-state spectrum near threshold.
- domain assumption The color-singlet ttbar production cross-section factorizes as the parton-level Born term times a modified phase-space factor built from the nonrelativistic Green's function (Eq. A.8).
- domain assumption The gluon luminosity is well described by the CTEQ5 parameterization in Eq. (A.2).
- standard math Watson's final-state interaction theorem ensures the eta_t pole position is the same in e+e- and gg production.
- domain assumption A regulated delta-function potential with renormalized coupling (Eqs. 4-5) and a Yukawa realization represent a TeV-scale BSM contact interaction.
- domain assumption The line shape near threshold is controlled almost entirely by m_t, Gamma_t, and m_eta_t.
- standard math The running coupling is given by Eq. (A.6) with Lambda_QCD=0.22 GeV, and is saturated at alpha_s=0.5 in the numerical diagonalization.
Cite this review
Pith. "Pith review of Ensuring that toponium is glued, not nailed." pith.science (2026). https://pith.science/paper/JLYV6B6C
@misc{pith2026241119180,
author = {Pith},
title = {Pith review of: Ensuring that toponium is glued, not nailed},
year = {2026},
howpublished = {\url{https://pith.science/paper/JLYV6B6C}},
note = {Machine review of arXiv:2411.19180}
}
read the original abstract
Hints of toponium might be incipient in LHC data, as given the vast numbers of t quarks produced, some survive on the exponential-decay tail long enough to fasten ttbar together. I here discuss a few differences between the standard Quantum Chromodynamics (QCD) binding (the ``glue'') and exotic short-range binding (the ``nail''). If the binding energy below threshold reaches the 3 GeV range the peak of the eta_t is distinct enough that a cross-section dip should be apparent in the line shape, should there only be one isolated resonance, but is filled by the excited QCD states adding about a pbarn to the cross section of ttbar production. Their effect for smaller binding energies is a tenuous increase in the cross section. A new-physics short-range interaction, on the other hand, yields a larger cross-section for equal binding energy (or hardly a visible bound state for similar cross section). This is due to its larger ttbar relative wavefunction at small distances. Finally, assuming that standard QCD plays out, I comment on what size of constraints on new-physics coefficients one can expect at given precision.
Figures
Figures from the paper (4 more)
Forward citations
Cited by 5 Pith papers
-
Extracting a Toponium Signal at the LHC with Spin and Quantum Information Tools
Spin and quantum-information observables add only marginal statistical power beyond kinematic variables for isolating toponium in near-threshold top-pair events, but improve interpretability.
-
New physics in toponium's shadow?
Consistently including toponium bound-state effects inside the new-physics amplitude, not just the Standard Model one, reshapes the allowed mass–coupling region for a top-philic pseudoscalar near the top-antitop threshold.
-
Topped baryons from QCD sum rules
QCD sum rules in HQET predict ground-state singly topped baryon masses near 174 GeV, some 1.1-1.5 GeV above the top quark pole mass.
-
Contrasting Pseudoscalar Higgs and Toponium States at the LHC and Beyond
The paper shows that quantum interference between a 365 GeV pseudoscalar Higgs and the QCD continuum significantly changes its measured ttbar rate, offering observables to distinguish such a Higgs from toponium.
-
Mass spectra and wave functions of toponia
Toponium masses from the Salpeter equation with Cornell and Coulomb potentials lie near 343.6-345.3 GeV with zero spin splittings, and predicted decay widths to photons, gluons, and leptons differ by up to a factor of...
Reference graph
Works this paper leans on
- [37]
-
[1]
[CMS], Preprint CMS-PAS-HIG-22-013 (unpublished)
-
[2]
Jeppe [CMS], delivered at the 17th Intnal
L. Jeppe [CMS], delivered at the 17th Intnal. Workshop on Top Quark Physics (Top2024), 22/9-27/9 2024, [arXiv:2411.18414 [hep-ex]]
arXiv 2024
- [3]
- [4]
- [5]
-
[6]
B. Fuks, K. Hagiwara, K. Ma and Y . J. Zheng, Phys. Rev. D 104, 034023 (2021)
work page 2021
-
[7]
J. A. Aguilar-Saavedra, Phys. Rev. D 110, 054032 (2024)
work page 2024
Show all 39 references
-
[8]
Y . Yang, B. Yang, J. Li, Z. Lu and J. Sun, [arXiv:2410.22841 [hep-ph]]
-
[9]
Petraki, M
K. Petraki, M. Postma and J. de Vries, JHEP 04, 077 (2017)
2017
-
[10]
Sommerfeld Enhancement in Bound State Dark Matter,
J. H. Streuer, “Sommerfeld Enhancement in Bound State Dark Matter,” thesis submitted to the Univ. of Amsterdam, 2017
2017
-
[11]
V . S. Fadin, V . A. Khoze and T. Sjostrand, Z. Phys. C48, 613-622 (1990)
1990
-
[12]
V . S. Fadin and V . A. Khoze, Sov. J. Nucl. Phys.53, 692-698 (1991)
1991
-
[13]
V . S. Fadin and V . A. Khoze, Sov. J. Nucl. Phys.48, 309-313 (1988)
1988
-
[14]
M. A. Braun, Zh. Eksp. Teor. Fiz. 54, 1220-1227 (1968)
1968
-
[15]
Brambilla, A
N. Brambilla, A. Pineda, J. Soto and A. Vairo, Rev. Mod. Phys. 77, 1423 (2005)
2005
-
[16]
Vairo, PoS CHARM2020, 015 (2021)
A. Vairo, PoS CHARM2020, 015 (2021)
2021
-
[17]
Beneke and Y
M. Beneke and Y . Kiyo, Phys. Lett. B 668, 143-147 (2008)
2008
-
[18]
Beneke et al
M. Beneke et al. Phys. Rev. Lett. 115, 192001 (2015)
2015
- [19]
-
[20]
G. L. Wang, T. F. Feng and Y . Q. Wang, [arXiv:2411.17955 [hep-ph]]
-
[21]
Maltoni, D
F. Maltoni, D. Pagani and S. Tentori, JHEP 09, 098 (2024)
2024
-
[22]
Galindo and P
A. Galindo and P. Pascual, Quantum Mechanics, english Ed. (1990) Springer-Verlag
1990
-
[23]
Beg Memorial V olume
R. Jackiw, MIT-CTP-1937, published in “Beg Memorial V olume” 35-53, World Scientific Publishing, 1991
1937
-
[24]
Gosdzinsky and R
P. Gosdzinsky and R. Tarrach, Am. J. Phys. 59, 70-74 (1991)
1991
-
[25]
R. M. Cavalcanti, Rev. Bras. Ens. Fis. 21, 336 (1999)
1999
-
[26]
Fabian Brau and Francesco Calogero, J. Phys. A: Math. Gen. 36, 12021 (2003)
2003
- [27]
-
[28]
S. J. Brodsky, G. P. Lepage and P. B. Mackenzie, Phys. Rev. D 28, 228 (1983)
1983
-
[29]
Di Giustino, S
L. Di Giustino, S. J. Brodsky, P. G. Ratcli ffe, X. G. Wu and S. Q. Wang, Prog. Part. Nucl. Phys. 135, 104092 (2024)
2024
-
[30]
Armour, submitted to the univ
K. Armour, submitted to the univ. of California, San Diego, 2004, based on the CTEQ5 sets
2004
-
[31]
H. L. Lai et al. [CTEQ], Eur. Phys. J. C 12, 375-392 (2000)
2000
-
[32]
M. V . Garzelli et al., JHEP 05 (2024), 321
2024
-
[33]
F. J. Llanes-Estrada, O. I. Pavlova and R. Williams, Eur. Phys. J. C 72 (2012), 2019 doi:10.1140/epjc/s10052-012-2019-9
2012 doi
-
[34]
Patrignani et al
C. Patrignani et al. [PDG], Chin. Phys. C 40 (2016), 100001
2016
-
[35]
K. G. Chetyrkin, B. A. Kniehl and M. Steinhauser, Nucl. Phys. B 510 (1998), 61-87
1998
-
[36]
G. M. Prosperi, M. Raciti and C. Simolo, Prog. Part. Nucl. Phys. 58, 387- 438 (2007)
2007
-
[38]
M. J. Strassler and M. E. Peskin, Phys. Rev. D 43, 1500-1514 (1991)
1991
-
[39]
Bezanson et al
J. Bezanson et al. SIAM Rev. 59, 65-98 (2017). 6
2017
Reviewed August 12, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.