A full Next to Leading Order study of direct photon pair production in hadronic collisions

TL;DR

The paper delivers a full next-to-leading order QCD calculation for direct photon-pair production in hadronic collisions, incorporating direct, single-fragmentation, and double-fragmentation mechanisms as well as the $gg\to\gamma\gamma$ box, implemented in the DIPHOX program. It validates the approach against fixed-target WA70 and Tevatron D0 data and provides LHC predictions under realistic isolation criteria, while exploring scale dependencies and the impact of isolation on different production channels. The authors highlight infrared-sensitive observables that require all-order soft-gluon resummation and discuss potential instabilities with stringent isolation, outlining directions for future improvements such as NNLO corrections and alternative isolation schemes. Overall, the work advances precise background predictions for $h\to\gamma\gamma$ searches at the LHC and informs the treatment of fragmentation and isolation in NLO calculations.

Abstract

We discuss the production of photon pairs in hadronic collisions, from fixed target to LHC energies. The study which follows is based on a QCD calculation at full next-to-leading order accuracy, including single and double fragmentation contributions, and implemented in the form of a general purpose computer program of "partonic event generator" type. To illustrate the possibilities of this code, we present the comparison with observables measured by the WA70 and D0 collaborations, and some predictions for the irreducible background to the search of Higgs bosons at LHC in the channel $h \to γγ$. We also discuss theoretical scale uncertainties for these predictions, and examine several infrared sensitive situations which deserve further study.

Figures (16)

• Figure 1: Diphoton differential cross section $d\sigma/dp_T$ vs. $p_T$, the transverse energy of each photon, in $\pi^{-}$-proton collisions at $\sqrt{S}= 22.9$ GeV. Data points from the WA70 collaboration wa70cross. The solid line is the full contribution with scales $M=\mu=M_f=0.275 \, (p_T (\gamma_1) + p_T (\gamma_2))$.
• Figure 2: Diphoton differential cross section $d\sigma/dp_T$ vs $p_T$, the transverse energy of each photon, at Tevatron, $\sqrt{S}=1.8$ TeV. Preliminary data points (statistical errors and systematics in quadrature) from the D0 collaboration d0 are compared to the theoretical predictions: the full NLO prediction is shown as the solid line. The ratio data/(full NLO theory) is shown below.
• Figure 3: Diphoton differential cross section $d\sigma/dm_{\gamma \gamma}$ vs. $m_{\gamma\gamma}$, the mass of the photon pair, at Tevatron, $\sqrt{S}=1.8$ TeV. Preliminary data points (statistical errors and systematics in quadrature) from the D0 collaboration d0 are compared to the theoretical predictions: the full NLO prediction is shown as the solid line.
• Figure 4: Diphoton differential cross section $d\sigma/dq_T$ vs. $q_T$, the transverse momentum of the photon pair, at Tevatron, $\sqrt{S}=1.8$ TeV. Preliminary data points (statistical errors and systematics in quadrature) from the D0 collaboration d0 are compared to the theoretical predictions: the full NLO prediction is shown as the solid line
• Figure 5: Diphoton differential cross section $d\sigma/d \phi_{\gamma\gamma}$ vs. $\phi_{\gamma\gamma}$, the azimuthal angle between the two photons, at Tevatron, $\sqrt{S}=1.8$ TeV. Preliminary data points (statistical errors and systematics in quadrature) from the D0 collaboration d0 are compared to the theoretical predictions: the full NLO prediction is shown as the solid line while open squares (open circles) represent the single (double) fragmentation contribution.