Analytic two-loop amplitudes for q ​ q ¯ → γ ​ γ q\bar{q}\to\gamma\gamma and g ​ g → γ ​ γ gg\to\gamma\gamma mediated by a heavy-quark loop

Matteo Becchetti, Federico Coro, Christoph Nega, Lorenzo Tancredi, Fabian J. Wagner · 2025

Read Analytic two-loop amplitudes for q ​ q ¯ → γ ​ γ q\bar{q}\to\gamma\gamma and g ​ g → γ ​ γ gg\to\gamma\gamma mediated by a heavy-quark loop by Matteo Becchetti, Federico Coro, Christoph Nega, Lorenzo Tancredi, Fabian J. Wagner (2025) 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-2502-00118.txt.

Introduction

Matteo Becchetti, Federico Coro, Christoph Nega, Lorenzo Tancredi, and Fabian Wagner work out two-photon production in full, on paper, two loops deep. Their 2025 paper covers two routes. A quark and an antiquark can meet and make two photons. Two gluons can do the same. Both routes run through loops of heavy quarks.

Elliptic integrals blocked this work in the past. So past work used semi-numerical methods for part of the job. New theory for the differential equations of such integrals clears the path here. The team builds a canonical basis for each integral in play. The answers come as iterated integrals over elliptic and polylog forms. Once they are written this way, the physical terms show neat cancellations. Series expansions then give fast, sound numbers across much of the phase space.

Catalog record

CREATORMatteo Becchetti, Federico Coro, Christoph Nega, Lorenzo Tancredi, Fabian J. Wagner
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE132 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2502-00118

Browse the full library · About the Tower of Babel

Full text

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

Analytic two-loop amplitudes for q ​ q ¯ → γ ​ γ q\bar{q}\to\gamma\gamma and g ​ g → γ ​ γ gg\to\gamma\gamma mediated by a heavy-quark loop Matteo Becchetti, Federico Coro, Christoph Nega, Lorenzo Tancredi, Fabian J. Wagner Abstract We address the analytic computation of the two-loop scattering amplitudes for the production of two photons in parton-parton scattering, mediated by loops of heavy quarks. Due to the presence of integrals of elliptic type, both partonic channels have been previously computed using semi-numerical methods. In this paper, leveraging new advances in the theory of differential equations for elliptic Feynman integrals, we derive a canonical basis for all integrals involved and compute them in terms of independent iterated integrals over elliptic and polylogarithmic differential forms. We use this representation to showcase interesting cancellations in the physical expressions for the scattering amplitudes. Furthermore, we address their numerical evaluation by producing series expansion representations for the whole amplitudes, which we demonstrate to be fast and numerically reliable across a large region of the phase space. Preprint: TUM-HEP 1554/25 1 Introduction Scattering amplitudes for the production of two final state particles in partonic collisions play a crucial role in the study of particle interactions. This is true both phenomenologically, where such amplitudes provide essential building blocks to model events observed at particle colliders, and formally, since starting at four points, their non-trivial dependence on the external kinematics provides important theoretical data for the study of general properties of perturbative Quantum Field Theory. The complexity of scattering amplitudes strongly scales with the increase in the perturbative order (i.e., the number of virtual loops) and the number of different mass scales involved. More mass scales are generated when considering the production of more particles (as in 2\to n scattering with n\geq 2 ), but also when one allows more massive virtual states to circulate in the loops (top quarks, electroweak vector bosons, etc.). From this perspective, amplitudes for the production of two photons or two jets in massless QCD are among the simplest. In fact, due to momentum conservation, they depend only on one non-trivial dimensionless ratio. This class of amplitudes was computed up to second order in perturbation theory long ago [ 1 , 2 ] , recently reaching the three-loop order both for the production of two photons [ 3 , 4 ] , two jets [ 5 , 6 , 7 ] as well as a photon and a jet [ 8 ] . Thanks to the simplicity of the underlying kinematics and the absence of any massive particles in the loops, they can be expressed in terms of a class of very well-understood special functions, namely multiple polylogarithms [ 9 , 10 , 11 , 12 ] (MPLs). Even more impressively, the subclass of so-called Harmonic Polylogarithms [ 13 ] (HPLs) only turns out to be sufficient for their analytical representation up to the three-loop order. Modeling the production of two photons and two jets in hadron collisions to a precision consistent with the three-loop order requires many more ingredients in addition to the virtual amplitudes. First of all, defining infrared finite observables for any QCD process to N 3 LO requires the inclusion of real and real-virtual amplitudes, i.e., amplitudes for the production of additional final state partons at a lower loop order. Recently, all these quantities have become available, including the last outstanding building blocks constituted by the two-loop amplitudes for the production of two photons and a jet [ 14 , 15 , 16 ] and three jets [ 17 , 18 , 19 ] . While all amplitudes are there, devising a scheme to properly organize and cancel the infrared divergences between real and virtual contributions to N 3 LO remains an outstanding task, particularly when the production of colored partons is considered. In addition to the ingredients necessary for a complete N 3 LO description of diphoton and dijet production, when working at this precision, more effects can potentially have a sizable impact on these observables. One such contribution is generated by heavy quarks (especially the top quarks) circulating in the loops. These mass effects have the potential to become sizable when one considers the production of photons and jets at high energies and transverse momenta, which are of the same order as the top quark mass. For diphoton production, in particular, these effects have been included up to the two-loop order in a series of recent calculations [ 20 , 21 , 22 ] . Interestingly, from a more formal point of view, allowing for the propagation of massive virtual states deeply changes the analytic properties of the underlying two-loop amplitudes. Not only do they become functions of two independent dimensionless ratios, but also a new type of special functions becomes relevant for their analytic calculation: iterated integrals over kernels related to an elliptic curve. Integrals of this type have been known to appear in perturbative quantum field theory (QFT) for a long time, and the first place where they become relevant is arguably the calculation of the electron self-energy to two loops [ 23 ] . The last two decades have led to the discovery of a large number of physically relevant scattering amplitudes involving not only these special functions, but also further higher-dimensional and higher-genus generalization thereof. This, in turn, has motivated intense work to understand their properties and devise methodologies for their calculations. Similar functions describe scattering amplitudes in string theory, and methods exported from this field and even from pure mathematics have provided the basis to generalize our knowledge of multiple polylogarithms to iterated integrals defined on elliptic geometries [ 24 , 25 , 26 , 27 ] and beyond. Despite the progress in understanding these functions analytically, the complexity of non-polylogarithmic amplitudes has also stimulated the development of powerful numerical and semi-numerical methods that aim to bypass the problem of having to deal with the specific properties of the geometries involved [ 28 , 29 , 30 , 31 ] . In fact, most of the results in the literature involving amplitudes of elliptic type have been obtained leveraging these numerical methods, including the recent studies for diphoton production quoted above [ 20 , 22 ] and similar ones for the production of a Higgs boson and a jet [ 32 ] . While numerical methods are powerful, there remain strong reasons to study the structure of these amplitudes analytically, especially in view of the calculation of more complex 2\to 3 massive amplitudes whose large kinematical phase-space can constitute a roadblock for automated numerical methods. Moreover, understanding the analytic structure of scattering amplitudes can lead to significant simplifications at the amplitude level, which might be crucial for phenomenology applications. In recent years, growing evidence has been found of large cancellations among complex combinations of iterated integrals of elliptic or even Calabi-Yau type, when physical quantities are considered. These have been demonstrated explicitly for two-point correlators [ 33 , 34 , 35 ] , while hints of similar simplifications can be observed also in elliptic higher-point amplitudes [ 36 , 37 ] , suggesting that getting analytic control over physical quantities might be simpler than naively expected. A crucial tool in these investigations

Read the complete text