[Paper Review] Searches for Higgs boson pair production with ATLAS

A. Ferrari|arXiv (Cornell University)|Sep 24, 2018

Particle physics theoretical and experimental studies10 references3 citations

TL;DR

This paper presents three ATLAS searches for Higgs boson pair production (HH) in proton-proton collisions at 13 TeV, using 36.1 fb⁻¹ of data from LHC Run-2. The analyses target non-resonant and resonant HH production in final states HH→bbbb, HH→bbγγ, and HH→WWγγ, setting 95% CL exclusion limits and constraining the Higgs self-coupling to −5.0 < κλ < 12.1.

ABSTRACT

The search for Higgs boson pair ($HH$) production is at the core of the ATLAS experimental program, as it probes the Brout-Englert-Higgs mechanism, as well as new physics beyond the Standard Model. Based on the proton-proton collision data recorded by the ATLAS experiment at CERN's Large Hadron Collider in 2015 and 2016, three public results on the search for $HH o bbbb$, $HH o bbγγ$ and $HH o WWγγ$ were recently released and are summarised in these proceedings.

Motivation & Objective

• To probe the Higgs self-coupling by searching for Higgs boson pair production (HH), a key test of the Higgs potential in the Standard Model.
• To search for new physics beyond the Standard Model through non-resonant and resonant HH production in multiple final states.
• To measure deviations from the Standard Model prediction of 33.41 fb for HH production, which is enhanced in many BSM theories.
• To develop and apply novel data-driven techniques to suppress dominant multi-jet backgrounds, especially in the HH→bbbb channel.
• To set exclusion limits on resonant HH production via spin-0 or spin-2 resonances decaying to Higgs pairs.

Proposed method

• Utilizes 36.1 fb⁻¹ of proton-proton collision data from ATLAS at √s = 13 TeV during LHC Run-2.
• Employs event topologies with resolved and boosted jet reconstruction: anti-kT jets with R=0.4 (resolved) and R=1.0 (boosted), with b-tagging at R=0.2 for boosted jets.
• Applies kinematic selection based on m4j, m2j, ΔRjj, and pT requirements to identify Higgs boson candidates and suppress top-quark and QCD backgrounds.
• Uses data-driven techniques to estimate multi-jet backgrounds in the HH→bbbb channel, particularly addressing trigger inefficiencies in 2016 data.
• Models signal and background distributions using simulations (for HH and single-H) and fits to data (for continuum multi-jet/photon backgrounds), with Crystal-Ball and exponential functions.
• Applies event selection cuts tailored to each final state: e.g., pT(γγ) > 100 GeV for HH→WWγγ, and mγγ within 3.4 GeV of mH for consistency with H→γγ decay.

Experimental results

Research questions

• RQ1 What are the limits on non-resonant Higgs boson pair production cross-section in the HH→bbbb, HH→bbγγ, and HH→WWγγ final states?
• RQ2 Can resonant Higgs boson pair production be observed via spin-0 or spin-2 resonances decaying to HH in the 260–3000 GeV mass range?
• RQ3 How do the observed event rates compare to Standard Model predictions in each final state, and what constraints do they place on the Higgs self-coupling?
• RQ4 To what extent can data-driven background estimation techniques reduce uncertainties in the HH→bbbb channel with large multi-jet backgrounds?
• RQ5 What is the combined sensitivity of multiple HH final states to new physics in the Higgs sector?

Key findings

• The combined observed 95% CL limit on the non-resonant Higgs boson pair production cross-section is 0.223 pb, equivalent to 6.7 times the SM prediction.
• The Higgs self-coupling κλ is constrained at 95% CL to the range −5.0 < κλ < 12.1, based on the statistical combination of the three most sensitive searches.
• For resonant HH production, 95% CL exclusion limits are set on the cross-section as a function of the hypothetical resonance mass, with exclusion extending up to 3 TeV in the HH→bbbb channel.
• In the HH→bbγγ channel, exclusion limits are set across a mass range of 800–3000 GeV, with a change in event selection observed at 1500 GeV.
• In the HH→WWγγ channel, the observed limit on non-resonant HH production is 230 times the SM cross-section, with expected limits ranging from 160 to 340 times SM at 95% CL.
• The fourth search, HH→bbττ, published post-conference, sets a 95% CL limit of 12.7 times the SM expectation for non-resonant HH production.