Soft Contributions Stabilize NNLO QCD Corrections to Quarkonium Production and Decay

Luca Maxia, Hua-Sheng Shao, Lukas Simon, Guoxing Wang · 2026

Read Soft Contributions Stabilize NNLO QCD Corrections to Quarkonium Production and Decay by Luca Maxia, Hua-Sheng Shao, Lukas Simon, Guoxing Wang (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2606-28494.txt.

Introduction

Luca Maxia, Hua-Sheng Shao, Lukas Simon, and Guoxing Wang fix a long-standing flaw in quarkonium theory. Quarkonium, a heavy quark bound to its antiquark, should be the cleanest system in strong-force physics. Yet second-order corrections in non-relativistic QCD often turn huge and negative, predicting impossible negative decay rates and ballooning error bars instead of steady improvement.

Maxia and colleagues identify the source of the blowups and propose a simple remedy centered on soft effects, the low-energy gluon effects their analysis singles out. Applied to S-wave color-singlet quarkonium production and decay, the approach restores much better convergence, with successive approximations settling down as they should. Predictions also agree better with measured data, returning a flagship framework to health.

Catalog record

CREATORLuca Maxia, Hua-Sheng Shao, Lukas Simon, Guoxing Wang
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE43 KB
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2606.28494

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Soft Contributions Stabilize NNLO QCD Corrections to Quarkonium Production and Decay Luca Maxia, Hua-Sheng Shao, Lukas Simon, Guoxing Wang Abstract Next-to-next-to-leading order (NNLO) QCD corrections to quarkonium production and decay are known to exhibit perturbative instabilities within non-relativistic QCD. We identify the origin of this problem and propose a simple remedy. Applying our approach to S S -wave color-singlet quarkonium processes, we achieve substantially improved perturbative convergence and agreement with experimental data. As the simplest class of hadrons consisting of a heavy quark-antiquark pair, quarkonium plays a special role in deepening our understanding of strong interactions in nuclear and particle physics. In contrast to light or open heavy-flavor hadrons, quarkonium production and decay processes are widely regarded as tractable within non-relativistic quantum field theory frameworks. For instance, non-relativistic QCD (NRQCD) 22 is a rigorous low-energy effective field theory of QCD, obtained by integrating out degrees of freedom associated with energy scales higher than the heavy-quark mass m_{Q} . It therefore enables systematic improvements in the calculation of quarkonium production and decay rates and is, to date, the most widely adopted theoretical framework for quarkonium studies. While many open issues remain 24 ; 51 ; 28 , a prominent purely theoretical challenge has emerged over the past decade in NRQCD studies of quarkonium: the convergence of the perturbative expansion in the strong coupling \alpha_{s} is increasingly called into question for a wide range of quarkonium processes. Thanks to advances in multi-loop techniques 67 ; 21 ; 11 over the last ten years, a significant number of quarkonium production and decay processes have been computed at next-to-next-to-leading order (NNLO) accuracy in \alpha_{s} . However, several common issues have been observed in theoretical predictions once NNLO QCD corrections are included. In particular, NNLO corrections are often large and negative, which can lead to negative and thus unphysical partial decay widths or cross sections. Typical examples include \eta_{c}\to\gamma\gamma 36 ; 40 ; 10 , J/\psi\to e^{+}e^{-} 35 ; 16 ; 36 ; 20 ; 38 , J/\psi\to\gamma\gamma\gamma 69 , and e^{+}e^{-}\to J/\psi J/\psi 60 ; 45 . For other observables that remain positive, such as the \gamma\gamma^{*}\to\eta_{c} transition form factor 40 ; 13 , the inclusion of NNLO corrections significantly increases the renormalization-scale uncertainty compared to next-to-leading order (NLO), contrary to the usual expectation of improved perturbative stability. Moreover, due to the large size of NNLO corrections, noticeable differences can also arise in exclusive processes depending on whether the \alpha_{s} expansion is truncated at the amplitude level or at the squared-amplitude level (see, e.g., refs. 45 ; 69 ; 40 ). There may be several possible explanations for these issues, depending on the specific problem, such as a large value of \alpha_{s} , the presence of large logarithms, or the emergence of new enhanced channels 58 ; 63 ; 56 , each of which requires different mitigation strategies. In this letter, however, we argue that an important common perturbative ingredient has been overlooked in the literature on quarkonium processes calculated in NRQCD at NNLO accuracy in \alpha_{s} . To better understand the problem, we first examine the partial decay width of \eta_{c}\to\gamma\gamma , which is a single-scale process and therefore free of logarithms arising from hierarchical scales. Its \alpha_{s} expansion up to NNLO 36 ; 40 ; 10 is \frac{\Gamma_{\mathrm{NNLO}}(\eta_{c}\to\gamma\gamma)}{\Gamma_{\mathrm{LO}}(\eta_{c}\to\gamma\gamma)}=1-1.07\alpha_{s}-11.16\alpha_{s}^{2}\,, (1) where \Gamma_{\mathrm{LO}} is the leading order (LO) width. We have used n_{f}=3 quark flavors and set the renormalization and NRQCD factorization scales to the charm-quark mass, \mu_{R}=\mu_{\Lambda}=m_{c} , in the above equation. The negative NNLO width can be attributed to the large and negative coefficient of \alpha_{s}^{2} , whose magnitude is more than one order larger than that of the \alpha_{s} term. This issue appears generically in many quarkonium processes known at NNLO in the literature that we have examined (see the Supplemental Material A ). Such an issue has been broadly acknowledged in the literature, yet no fully satisfactory solution has been proposed so far. Nevertheless, there are several suggestive hints. In NNLO QED corrections to para-positronium decays into two photons 34 ; 9 , the large and negative \alpha^{2} coefficient in the short-distance coefficient (SDC) cancels against a large and positive contribution arising from \mathcal{O}(\alpha^{2}) soft corrections to the wavefunction at the origin, yielding a final \mathcal{O}(1) coefficient at NNLO. In addition, ref. 33 presents a method to compute wavefunctions at the origin within potential NRQCD (pNRQCD) 57 ; 25 ; 18 ; 17 . In this framework, large cancellations are also observed between the SDCs of \eta_{c}\to\gamma\gamma and J/\psi\to e^{+}e^{-} and the non-Coulombic contributions to the wavefunctions at the origin, which include both perturbative and non-perturbative (NP) effects. These cancellations are interpreted as a consequence of the elimination of renormalon ambiguities 19 in the NP QCD potential. Finally, ref. 50 adopts a weak-coupling approach to compute quarkonium long-distance matrix elements (LDMEs) perturbatively, incorporating both Coulombic and non-Coulombic corrections. This framework has been used to predict the leptonic decay of the \Upsilon(1S) meson up to three loops in QCD 20 . However, the authors of ref. 50 express some reservations regarding the applicability of the weak-coupling approach to the charmonium system. Inspired by the para-positronium decay into two photons 34 ; 9 , we propose a simple solution to address the aforementioned perturbative instability problem for S -wave color-singlet quarkonium states, restricting ourselves to leading power (LP) in the heavy-quark relative velocity v expansion within NRQCD. We remind the reader that, within the NRQCD factorization framework, only the combination of SDCs and LDMEs is physical. At NNLO in QCD, the SDCs alone are infrared (IR) divergent; these divergences must be canceled by the renormalization of the LDMEs, in close analogy with the renormalization of parton distribution functions in perturbative QCD factorization. A common practice in the literature is to renormalize the LDMEs using the modified minimal subtraction ( \overline{\rm MS} ) scheme, which retains only the IR poles in dimensional regularization. This procedure introduces a dependence on the NRQCD factorization scale \mu_{\Lambda} in the SDCs, which should be (partially) compensated by the renormalization group (RG) evolution of the LDMEs, leaving a residual higher-order dependence on \mu_{\Lambda} . Physically, \mu_{\Lambda} represents the ultraviolet (UV) cutoff of NRQCD, whose natural value is of the order of the heavy-quark mass m_{Q} . However, we argue that the widely used \overline{\rm MS} scheme is not optimal for studying quarkonium processes perturbatively within NRQCD. Using the \overline{\rm MS} scheme, in which the finite term vanishes, is analogous to setting the finite soft contribution to the wavefunction at the origin to zero in the para-positronium case; in this case, the NNLO cancellation observed in refs. 34 ; 9 would no longer be manifest. Unlike positronium, quarkonium involves both perturbative and NP aspects that require distinct treatments. Even within the purely perturbative sector, Coulomb resummation is necessary to improve perturbative stability. However, as in any resummation procedure, ambiguities arise in deciding which contributions should be expanded to fixed order in \alpha_{s} and which should be resummed to all orders in \alpha_{s} 1 . A proper determination of the fixed-order perturbative part in the renormalized LDMEs is therefore crucial for stabilizing quarkonium NNLO QCD corrections. In dimensional regularization, it is understood that Coulomb singularities do not generate 1/\epsilon poles,

Read the complete text