[Paper Review] Two photon background for Higgs boson searches at the LHC

T. Binoth|ArXiv.org|May 18, 2000

Particle physics theoretical and experimental studies3 citations

TL;DR

This paper presents a full next-to-leading order (NLO) calculation of the two-photon background in hadronic collisions at the LHC, including fragmentation contributions from QCD partons, which are critical for accurate background estimation in Higgs boson searches via the H→γγ decay channel. The study provides a precise prediction for the invariant mass distribution of photon pairs and demonstrates a residual scale dependence of 10–20% under variations of renormalization and factorization scales, with isolation cuts significantly reducing fragmentation contributions.

ABSTRACT

The search for an intermediate mass Higgs boson at the LHC needs the quantitative understanding of the two-photon background. A calculation of two-photon production in hadronic collisions at full next-to-leading order is described. It includes photons originating from the hadronization of QCD partons which play an important role at the LHC. A prediction for the invariant mass distribution for photon pairs at the LHC is presented and finally the residual scale dependencies are discussed.

Motivation & Objective

• To quantitatively understand the dominant two-photon background in Higgs boson searches at the LHC, particularly for intermediate-mass Higgs bosons.
• To include next-to-leading order (NLO) corrections for photon pair production, especially fragmentation contributions from QCD partons.
• To provide a reliable prediction for the invariant mass distribution of photon pairs (Mγγ) at the LHC.
• To assess the stability of the prediction under variations of renormalization, factorization, and fragmentation scales.
• To evaluate the effectiveness of photon isolation criteria in suppressing fragmentation-driven background.

Proposed method

• A full NLO calculation of two-photon production in hadronic collisions, including virtual and real emission corrections at order α²αs.
• Incorporation of photon fragmentation functions Dγ/j (j = q, q̄, g) to account for photons from hadronization of QCD partons.
• Use of the MRST2 parton distribution functions and BFG photon fragmentation functions in the computation.
• Implementation of isolation criteria by restricting transverse hadronic energy within a cone (R = 0.4, ETmax = 5 GeV) around each photon.
• Scale variation analysis using μ, μFACT, and μFRAG to estimate theoretical uncertainties.
• Comparison of results with Tevatron data to validate the calculation for infrared-safe observables.

Experimental results

Research questions

• RQ1 What is the contribution of QCD parton fragmentation to the two-photon background at the LHC?
• RQ2 How does the inclusion of NLO corrections, especially for fragmentation processes, affect the Mγγ distribution?
• RQ3 To what extent do isolation criteria suppress the fragmentation-driven background in γγ final states?
• RQ4 What is the residual scale dependence of the NLO prediction for the Mγγ distribution?
• RQ5 How stable is the prediction under variations of renormalization, factorization, and fragmentation scales?

Key findings

• The inclusion of NLO fragmentation contributions significantly improves the accuracy of the two-photon background prediction at the LHC.
• Without isolation, the one-fragmentation contribution dominates the Mγγ distribution; with isolation (ETmax = 5 GeV, R = 0.4), the direct production component becomes dominant.
• The Mγγ distribution at the LHC is well described by the NLO calculation, with good agreement to Tevatron data for infrared-safe observables.
• The residual scale dependence of the isolated Mγγ distribution is 10–20%, with anti-diagonal variations of μ and μFACT providing a conservative uncertainty estimate.
• Accidental stability is observed when μ and μFACT are varied together, due to compensation between NLO corrections and parton flux changes.
• The calculation is implemented in the DIPHOX event generator and provides a foundation for higher-order corrections and resummation in infrared-sensitive distributions.