Isolated photon cross section and resummation of the logarithms of the cone radius

J.Ph.Guillet; M. Fontannaz

REVIEW 5 major objections 4 minor 13 references

Isolated photon cross section and resummation of the logarithms of the cone radius

T0 review · 5 major / 4 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read Resumming the logarithms of the cone radius restores unitarity of isolated photon cross sections.

desk verdict A useful extension of the cone-resummation program, but the abstract oversells 'unitarity restored' beyond what a single unbanded kinematic point can support. read the letter →

arxiv 2506.16092 v1 pith:TUVNHFJG submitted 2025-06-19 hep-ph

M. Fontannaz , J.Ph.Guillet This is my paper · ORCID

classification hep-ph

keywords isolatedphotoncrosssectionconeisolationhollow-conelogarithmresummationunitarityfragmentationcontributionQCDcorrectionsLHCpromptphotons

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

Large-transverse-momentum photon experiments isolate the photon by requiring little hadronic energy inside a cone, but the cone radius introduces logarithms $\ln R$ into perturbative cross sections. At small $R$, the NLO isolated cross section can exceed the fully inclusive one, violating the physical requirement that isolation cannot increase the rate. The paper argues that leading-logarithmic resummation of the $\ln R$ terms restores this inequality, and that the same is true, though less completely, for the hollow-cone criterion used in practice, where no isolation is imposed inside a small inner cone of radius $r$. If correct, standard cone-isolation calculations remain the right tool for LHC isolated-photon measurements, with the inner cone changing results only at the few-percent level.

What carries the argument

The machinery is the leading-logarithmic factorized resummation formula (2.3): a Born-level photon contribution, a direct contribution built from a quark/gluon hard cross section convolved with an evolution operator $E^{(0)}_{ab}(z/x; M, R p_T)$, and a fragmentation contribution convolving the same operator with photon fragmentation functions $D_b^\gamma(x, R p_T)$ evaluated at the low scale $M_F=R p_T$. The formula sums the $(\alpha_s \ln R)^k$ terms; for the hollow cone, the evolution operator is evaluated between $R p_T$ and $r p_T$, replacing the log by $\ln(R/r)$, and the fragmentation contribution must be resummed rather than truncated at $\mathcal{O}(\alpha_s)$, because the truncated form produces the spurious negative fragmentation term $\sim -\frac{\alpha_s}{2\pi} \ln(R^2/r^2) \ln(1/\epsilon)$.

What would settle it

Evaluate the resummed formula with a different photon fragmentation-function parametrization, or with the scale lowered to $R p_T \sim 5$ GeV, and scan $R$ from 0.5 down to 0.01 at the paper's kinematics; if $\sigma_{\rm cone} \le \sigma_{\rm inclusive}$ fails at any $R$, the claimed restoration is not genuine.

Watch

Extended reading notes

Core claim

The paper's central claim is that unitarity—the requirement $\sigma_{\rm inclusive} \ge \sigma_{\rm cone}$—is restored for isolated photon cross sections once the leading logarithms of the isolation-cone radius are resummed. At NLO with $\sqrt{s}=7$ TeV, $p_T^\gamma=100$ GeV and $\epsilon=0.04$, the unresummed cone-isolated cross section rises from 3.59 pb/GeV at $R=0.5$ to 4.56 at $R=0.06$, above the inclusive 4.29; after the LL resummation it becomes 4.24 at $R=0.06$, below the inclusive value. For the hollow-cone criterion ($R=0.4$, $r=0.1$), where the fragmentation contribution is much larger because the inner cone is not isolated, the same resummation with the scale ratio set by $\ln(R/r)$ improves the violation but does not remove it entirely at $r=0.1$; the paper's final conclusion is that the inner cone only slightly changes standard cone results at NLO, so cone-isolation calculations suffice for the experimental procedure.

Load-bearing premise

The calculation assumes the photon fragmentation functions are trustworthy when evaluated at the very low factorization scale $R p_T^\gamma$ (a few GeV for the smallest radii studied), a regime in which these functions are extrapolated below the data they were fitted to.

Editorial extensions

If this is right

Standard cone calculations with the LL supplement can be used for LHC isolation; the hollow-cone inner radius changes results by only a few percent.
The resummed correction is small for realistic cone radii: below 1% for $R\ge 0.3$ and about 7% at the extreme $R=0.06$, so existing NLO predictions need only modest shifts.
For hollow-cone isolation, resummation of the fragmentation contribution is mandatory: truncating the evolution at $\mathcal{O}(\alpha_s)$ gives a strongly negative contribution that breaks the ordering between hollow and standard cones.
At $r=0.1$, the NLO-resummed hollow-cone cross section still slightly violates unitarity, pointing to genuine higher-order contributions beyond the LL terms.

Reading between the lines

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

Beyond the paper: the unitarity check could serve as a model-independent validation test for any isolation prescription, for example smooth-cone Frixione isolation at very small radii.
Beyond the paper: the result's dependence on the photon fragmentation functions at $M_F=R p_T$ means a decisive test is to repeat the calculation with a second fragmentation-function parametrization or with low-scale data; if unitarity fails, the restoration is tied to the specific parametrization used.
Beyond the paper: a direct experimental probe would be an LHC measurement of isolated photon production with cone radii down to about 0.1-0.2, where the resummed prediction differs from plain NLO; agreement would confirm the low-scale fragmentation input.
Beyond the paper: the LL resummation structure is generic and may apply to other cone-jet observables—isolated $Z$ bosons, jet fragmentation functions, or photon-plus-jet rates—where a small cone parameter generates large logarithms.

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.

Desk Editor's Note

read the letter

The paper does something real: it extends the earlier Catani-Fontannaz-Guillet-Pilon resummation of small-cone logarithms to the fragmentation contribution in the hollow-cone isolation case. The convolution formulas in Section 4 (eqs. 4.3–4.5) are the new ingredient, and the numerical demonstration that the resummation reduces the unitarity violation (from 2.85 to 2.95 pb/GeV in the hollow case, closer to the cone value) is a legitimate, if modest, step forward. The authors are honest in the conclusion: they state plainly that NLO cross sections are not absolute for condition (3.1) and that the residual violation at r ≈ 0.1 points toward higher orders. That honesty is worth credit.

The soft spot is not in the math but in the packaging. The abstract says 'unitarity is restored when these logarithms are resummed,' and the paper's headline claim leans on Table 1, where at R = 0.06 the resummed NLO value sits 1.2% below the inclusive value. That margin is well inside the typical scale variation of an NLO cross section, and the resummed row shows no error band. The authors' own caveat in Section 4 (that the condition has to be checked by all-order cross sections) contradicts the flat abstract wording. This is a fixable rhetorical problem, not a fatal flaw, but it should be addressed in revision.

The lower-scale fragmentation input is a secondary worry: using BFG sets at MF = R pT with R = 0.06 means scales around 6 GeV, well below the data range. For the cone case the contribution is small, so the effect is likely minor; for the hollow-cone case (r = 0.1, MF = 10 GeV) it could matter, and the paper supplies no cross-check.

I agree with the reader's conditional verdict. The paper deserves a serious referee: the derivation is coherent, the numerical work is reproducible in principle, and the hollow-cone result is new. But the referee should ask for scale-uncertainty bands on the resummed values, a discussion of the low-scale fragmentation function reliability, and an abstract that matches the actual strength of the claim. I would not cite it in my own work, but I would bring it to a reading group to discuss the gap between an honest conclusion and an overreaching abstract.

Referee Report

5 major / 4 minor

Summary. The manuscript studies the cone-radius dependence of isolated prompt-photon cross sections in hadronic collisions, extending the framework of ref. [3] by resumming logarithms of the cone radius R in the fragmentation contribution and by applying the same treatment to a hollow-cone (ring) isolation criterion. The central numerical results are: (i) for the standard cone, the NLO+LL resummed cross section satisfies the unitarity bound sigma_isolated <= sigma_inclusive down to R = 0.06 at the chosen scales, whereas the fixed-order NLO result violates it for R <= 0.1; and (ii) for the hollow cone, the resummation increases the fragmentation component but does not fully restore unitarity for r ~ 0.1. The authors conclude that standard cone calculations adequately describe the more realistic hollow-cone isolation at NLO with the parameters used.

Significance. If the numerical results withstand a proper uncertainty treatment, the paper carries a useful practical message for LHC prompt-photon analyses: small-cone logarithms can be resummed and standard cone results remain close to hollow-cone results. The use of the established framework of ref. [3] and the explicit evolution-kernel resummation formulas are strengths, and the unitarity bound provides a physical, falsifiable benchmark. The paper does not supply reproducible code or a public implementation, and the main quantitative claims rest on a narrow set of parameter choices without uncertainty bands.

major comments (5)

[Section 4, Eq. (4.5)] As printed, the definitions z_c = 1/(1+epsilon) and v_c = 1/(1+epsilon) make the second theta factor theta(z_c - z) theta(z - v_c) vanish identically, yet the text reports a contribution 0.561 from "the second term". This is an internal inconsistency in the central formula of the hollow-cone resummation. The authors presumably intend different isolation parameters for the outer ring and the inner cone; these must be defined explicitly and consistently, and the derivation of the three terms should be provided.
[Table 1 and abstract] The claim that "unitarity is restored" is supported by a single numerical point at R = 0.06, where the resummed value 4.24 pb/GeV is only 1.2% below the inclusive value 4.29 pb/GeV. No scale variation is shown for the resummed prediction, while the earlier columns of the same table exhibit sensitivity to the factorization scale M_F. With standard factor-of-two scale variations typically giving several-percent changes at NLO, this margin is not sufficient to establish restoration; please provide a scale-uncertainty band or at least a scan over mu, M, and M_F.
[Eqs. (2.3), (4.2), (4.4)] The resummed fragmentation term uses the fragmentation functions D_b^gamma(x, R p_T) (or D_b^gamma(x, r p_T)) evaluated at the very low factorization scale R p_T. For p_T = 100 GeV and R = 0.06 this scale is 6 GeV; for r = 0.1 it is 10 GeV. The BFG parameterization [7] is fitted to e+e- data and is not validated at such scales. Since the fragmentation resummation is the mechanism that moves the cone result from 4.55 to 4.24 pb/GeV in Table 1, the low-scale fragmentation-function input is load-bearing. Please quantify the sensitivity, for example by varying M_F around R p_T, comparing BFG sets, or using an alternative photon fragmentation function.
[Section 4, Table 4] The statement that the inner cone "only slightly" changes the standard cone results is not directly supported by the table without uncertainty estimates: at R = 1, r = 0.1 the hollow-cone total is 2.95 pb/GeV versus 3.10 pb/GeV for the cone, a 5% deficit; at R = 0.5, r = 0.1 it is 3.49 versus 3.60. These differences are of the same order as the resummation effects and plausible scale variations. In addition, the table does not state explicitly which value of epsilon is used for the cone column and which for the hollow-cone column; the text introduces epsilon = 0.5 for the inner cone while earlier sections use epsilon = 0.04. Please clarify, and when testing inequality (3.1) ensure the same E_T^cut is used for both criteria.
[Sections 2 and 4, Eqs. (2.3)-(2.5), (4.2)-(4.5)] The paper relies almost entirely on ref. [3] for the definition of E^(0)_ab, the LL resummation, and the subtraction procedure, presenting only final formulas. Since the new results of this paper (the fragmentation resummation and the hollow-cone extension) are built on these formulas, and since Eq. (4.5) is not derived, the manuscript should include at least the essential definitions and the path from the Mellin-space evolution to the z-space convolution, or clearly point to specific equations in ref. [3]. As it stands, an independent reader cannot verify the central calculation.

minor comments (4)

[Throughout] There are typographical errors: "criterium" should be "criterion", "preceeding" should be "preceding", "holow-cone" should be "hollow-cone", and the CMS reference [2] contains an extra digit in the volume/page string.
[Table 1] The column headers "NLO p_T/2" and "NLO R p_T" are ambiguous; they should state explicitly which scale is M_F, e.g., "NLO with M_F = p_T/2" and "NLO with M_F = R p_T".
[Eq. (4.5)] The notation ln(R/r)^2 is ambiguous; it should be written as [ln(R/r)]^2 or ln^2(R/r) to avoid confusion about the argument of the logarithm.
[Table 4] The table lacks units in the column headers; please add "(pb/GeV)" for all cross-section columns.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the resummation is an independent QCD evolution calculation, and unitarity is used as an external benchmark rather than a fitted target.

full rationale

The paper's central claim is that resumming logarithms of the cone radius restores the inequality sigma_inclusive > sigma_cone. The resummed formulas, eqs. (2.3), (4.2) and (4.4), are derived from DGLAP evolution kernels and the BFG fragmentation functions; they are not constructed to enforce the inequality. The unitarity condition (3.1) is an external physical benchmark: the isolated/hollow-cone cross sections are compared against an independently computed inclusive NLO value (4.29 pb/GeV in Table 1), and the resummed cone value 4.24 pb/GeV is a numerical output, not a fitted parameter. The same check for the hollow cone still shows violation for r=0.1 (Table 4), which would not happen if the result were circularly engineered. The self-citations to refs. [3]-[5] and [7] provide prior derivations, the JetPhox implementation, and the BFG parametrization; none of them is a uniqueness theorem or an ansatz that already contains the target inequality. The authors themselves note at the end of Sec. 4 that condition (3.1) must ultimately be checked by all-order cross sections and that the NLO evaluation is 'no more absolute'; this is a caveat on numerical support, not a circularity. The low factorization-scale behavior of D_b^gamma(x, R pT) and the absence of a scale-uncertainty band are correctness/robustness concerns, not reductions of the predictions to their inputs.

Assumptions & free parameters 1 free parameters · 4 assumptions · 0 invented entities

The calculation inherits standard QCD factorization and the leading-log resummation formalism from the authors' previous paper [3]. The only hand-chosen inputs are the renormalization and factorization scales. No new physical entities are introduced.

free parameters (1)

Renormalization and factorization scales = mu = pT/2, M = pT, MF = R pT or r pT
Chosen by hand following standard practice in prompt photon calculations. The paper checks the MF dependence at NLO in Table 1, but no systematic scale-variation uncertainty is quoted.

assumptions (4)

domain assumption QCD factorization for prompt photon production, splitting the cross section into PDFs, hard scattering, and fragmentation functions.
Eq. (2.1) assumes this factorization, which is standard in the field but not derived here.
domain assumption The leading-log resummation formula eq. (2.3) taken from ref. [3] is correct.
The paper relies on the prior result for the evolution operator E^(0) and the resummed expression; it does not rederive it.
domain assumption The BFG fragmentation functions [7] and CTEQ6M PDFs [6] are accurate enough at the scales used.
These are standard inputs fitted to other data, but their uncertainties are not propagated or tested, especially at the low scale MF = r pT.
domain assumption The physical constraint sigma_inclusive >= sigma_iso >= sigma_cone, called unitarity, is a meaningful check for NLO-plus-resummed cross sections.
Stated in eq. (3.1); the authors themselves note in Section 5 that this condition is only exact at all orders, so using it at NLO is an assumption about what resummation should restore.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Isolated photon cross section and resummation of the logarithms of the cone radius." pith.science (2026). https://pith.science/paper/TUVNHFJG

@misc{pith2026250616092,
  author       = {Pith},
  title        = {Pith review of: Isolated photon cross section and resummation of the logarithms of the cone radius},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TUVNHFJG}},
  note         = {Machine review of arXiv:2506.16092}
}

read the original abstract

The cone isolation criterion used in large-pt photon experiments generates logarithms of the cone radius in the theoretical cross sections. When a small radius is used, unitarity is violated as the inclusive cross section is smaller than the isolated one. We show that unitarity is restored when these logarithms are resummed. We also study a criterion which offers a more precise description of the experimental procedure : no isolation is imposed in a very small cone inside the standard one. In this case as well unitarity is violated and the resummation of logarithms is required.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

13 extracted references · 13 canonical work pages

[3]

Catani, M

S. Catani, M. Fontannaz, J. Ph. Guillet and E. Pilon, JHEP 1309 (2013) 007

work page 2013
[7]

Bourhis, M

L. Bourhis, M. Fontannaz and J. Ph. Guillet, Eur. Phys. J. C2 (1998) 529

work page 1998
[1]

D83 (2011)052005

Atlas coll., Phys.Rev. D83 (2011)052005

work page 2011
[2]

D84 (2011)0520011 ; JHEP 01 (2012) 133

CMS coll., Phys.Rev. D84 (2011)0520011 ; JHEP 01 (2012) 133

work page 2011
[4]

Belghobsi, M

Z. Belghobsi, M. Fontannaz, J.Ph. Guillet, G. Heinrich, E. Pilon, M. Werlen, Phys.Rev. D79 (2009) 114024

work page 2009
[5]

Catani, M

S. Catani, M. Fontannaz, J. Ph. Guillet and E. Pilon, JHEP 05 (2002) 028

work page 2002
[6]

Pumplin, D

J. Pumplin, D. R. Stump, J. Huston, H. L. Lai, P. M. Nadolsky and W. K. Tung, JHEP 0207 (2002) 012

work page 2002
[8]

Frixione, Phys

S. Frixione, Phys. Lett. B429 (1998) 369

work page 1998

Show all 13 references

[9]

Hall and J

E. Hall and J. Thaler, JHEP 09 (2018), 164

2018
[10]

Belanger, F

G. Belanger, F. Boudjema, J. P. Guillet and E. Pilon,\\ ``Proceedings, 7th Les Houches Workshop on Physics at TeV Colliders: Les Houches, France, May 30-June 17, 2011''

2011
[11]

Amoroso, P

S. Amoroso, P. Azzurri, J. Bendavid, E. Bothmann, D. Britzger, H. Brooks, A. Buckley, M. Calvetti, X. Chen and M. Chiesa, et al
[12]

X. Chen, T. Gehrmann, N. Glover, M. H\"ofer and A. Huss, JHEP 04 (2020), 166

2020
[13]

Aurenche, M

P. Aurenche, M. Fontannaz, J. Ph. Guillet, Eur. Phys. J. C44 , 395 (2005) ; hep-ph/0503259

2005 arXiv

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.

[1] [3]

Catani, M

S. Catani, M. Fontannaz, J. Ph. Guillet and E. Pilon, JHEP 1309 (2013) 007

work page 2013

[2] [7]

Bourhis, M

L. Bourhis, M. Fontannaz and J. Ph. Guillet, Eur. Phys. J. C2 (1998) 529

work page 1998

[3] [1]

D83 (2011)052005

Atlas coll., Phys.Rev. D83 (2011)052005

work page 2011

[4] [2]

D84 (2011)0520011 ; JHEP 01 (2012) 133

CMS coll., Phys.Rev. D84 (2011)0520011 ; JHEP 01 (2012) 133

work page 2011

[5] [4]

Belghobsi, M

Z. Belghobsi, M. Fontannaz, J.Ph. Guillet, G. Heinrich, E. Pilon, M. Werlen, Phys.Rev. D79 (2009) 114024

work page 2009

[6] [5]

Catani, M

S. Catani, M. Fontannaz, J. Ph. Guillet and E. Pilon, JHEP 05 (2002) 028

work page 2002

[7] [6]

Pumplin, D

J. Pumplin, D. R. Stump, J. Huston, H. L. Lai, P. M. Nadolsky and W. K. Tung, JHEP 0207 (2002) 012

work page 2002

[8] [8]

Frixione, Phys

S. Frixione, Phys. Lett. B429 (1998) 369

work page 1998