Photon-induced Background for Dilepton Searches and Measurements in pp Collisions at 13 TeV

TL;DR

The paper addresses the need to model photon-induced backgrounds to high-mass dilepton searches at the LHC, alongside the Drell-Yan process. It performs FEWZ-based calculations of DY and PI cross sections at 13 TeV within typical detector acceptances, comparing photon-PDF sets such as MRST2004qed Mem0/Mem1 and NNPDF23qed (and CT14QED), and introduces two kinematic acceptances (\"Search\" and \"DY\"). Key findings show that PI contributions are highly PDF-dependent: Mem1 generally minimizes PI effects, while NNPDF23qed yields large, mass-dependent uncertainties; a simple KISS estimate scaling as O(\alpha/\alpha_s) agrees with FEWZ results within PDF bands. The work guides photon-PDF choices for high-mass analyses and motivates future dilepton measurements to constrain photon PDFs and probe gamma-gamma production mechanisms in light of the 750 GeV diphoton hint.

Abstract

The production of high invariant mass opposite sign lepton pairs in proton-proton collisions at the LHC is dominated by the Drell-Yan process. In addition to this photon or Z exchange mediated mechanism, gamma-gamma collisions, where photons radiated by the incoming protons collide, can produce lepton pairs. This is an important additional source of background for high mass resonant (like Z') or non-resonant (like contact interactions) searches. In this paper detailed calculations of the Drell-Yan and photon-induced cross sections in the typical acceptance of a multi-purpose LHC detector at center of mass energy 13 TeV are presented. The hint for a diphoton excess at a mass around 750 GeV, reported by the ATLAS and CMS experiments from the analysis of the 2015 data at 13 TeV, raises the possibility that such an excess could be produced through gamma-gamma collisions. A good theoretical understanding and measurements in the dilepton channels at these energies can help to elucidate this production hypothesis.

Figures (8)

• Figure 1: Differential cross sections for DY and PI at 13 TeV in the search acceptance.
• Figure 2: Top: ratios of cross sections (DY+PI)/DY for different PDFs in the Search acceptance. Bottom: same, but including the PDF uncertainties for NNPDF23qed.
• Figure 3: Top: ratios of cross sections (DY+PI)/DY for the two members of MRST2004qed in the DY acceptance. Bottom: ratio of (DY+PI) for the two members (blue), and ratio of the DY cross sections for the two members (red), showing that the DY predictions are almost identical while the (DY+PI) predictions differ substantially.
• Figure 4: Top: ratios of cross sections (DY+PI)/DY for the two members of MRST2004qed in the Search acceptance. Bottom: ratio of (DY+PI) for the two members (blue), and ratio of the DY cross sections for the two members (red), showing that the DY predictions are almost identical while the (DY+PI) predictions differ substantially.
• Figure 5: Ratios of cross sections (DY+PI) for MRST2004qed in the Search and DY acceptances. At high mass the ratio approaches unity.