[Paper Review] Search for invisible Higgs boson production with the CMS detector at the LHC

Matthew Chasco|arXiv (Cornell University)|Oct 3, 2013

Particle physics theoretical and experimental studies1 references4 citations

TL;DR

This paper presents a direct search for invisible Higgs boson decays using 5.1 fb⁻¹ and 19.6 fb⁻¹ of proton-proton collision data at √s = 7 TeV and 8 TeV collected by the CMS detector. The analysis focuses on ZH → ℓ⁺ℓ⁻ + invisible Higgs final states, using missing transverse energy and two leptons as signatures, and sets a 95% confidence level upper limit of 75% on the invisible branching fraction for a 125 GeV Higgs boson, with an expected limit of 91%.

ABSTRACT

Results are presented for the search for invisible Higgs boson production using the full LHC dataset corresponding to integrated luminosity of 5.1 fb^-1 and 19.6 fb^-1 of proton-proton collision data at sqrt(s)= 7 TeV and 8 TeV (respectively) collected by the CMS detector. The invisible Higgs is searched for in final states of missing transverse energy, with two leptons from a recoiling Z boson. No significant excess is found beyond standard model predictions, and limits are obtained on the branching fraction of the Higgs boson to invisible particles.

Motivation & Objective

• To directly probe invisible decays of the Higgs boson at the LHC using the full 7 and 8 TeV datasets.
• To test the hypothesis that the Higgs boson could decay into undetected particles, such as dark matter or graviscalar modes, indicating new physics beyond the Standard Model.
• To set model-independent limits on the branching fraction of the Higgs boson to invisible particles using the ZH → ℓ⁺ℓ⁻ + H(invisible) signature.
• To improve sensitivity by employing a shape analysis on the transverse mass variable m_T to distinguish signal from dominant ZZ background.
• To constrain the invisible branching fraction across a Higgs mass range of 105–145 GeV.

Proposed method

• The analysis targets the ZH → ℓ⁺ℓ⁻ + invisible Higgs final state, where the Z boson decays into two charged leptons and the Higgs decays invisibly, producing missing transverse energy.
• Events are selected based on two isolated leptons (e⁺e⁻, μ⁺μ⁻, or eμ) and significant missing transverse energy, with additional requirements to suppress QCD and top quark backgrounds.
• The transverse mass m_T is used as a discriminating variable, defined as m_T² = (E_T,ℓℓ + E_T,miss)² - |p_T,ℓℓ + p_T,miss|², to exploit kinematic differences between ZH and ZZ final states.
• Backgrounds are modeled using Monte Carlo simulations with POWHEG, MadGraph, and PYTHIA, and validated with data-driven methods using control samples in orthogonal regions.
• Systematic uncertainties are propagated using log-normal priors, with key sources including PDF, QCD scale variations, luminosity, and lepton reconstruction efficiencies.
• A modified frequentist CL_S method with profile-likelihood test statistics is used to set upper limits on the invisible branching fraction at 95% confidence level.

Experimental results

Research questions

• RQ1 What is the upper limit on the branching fraction of the Higgs boson to invisible particles, assuming a Standard Model production rate?
• RQ2 Can a direct signal of invisible Higgs decays be observed in the ZH → ℓ⁺ℓ⁻ + invisible final state using 7 and 8 TeV LHC data?
• RQ3 How effective is the transverse mass m_T variable in distinguishing the ZH signal from the dominant ZZ background?
• RQ4 What are the constraints on invisible Higgs decays across a range of Higgs masses from 105 to 145 GeV?
• RQ5 How do systematic uncertainties, particularly from PDF and QCD scale variations, affect the final limit on the invisible branching fraction?

Key findings

• No significant excess is observed in the data beyond Standard Model expectations, with observed yields consistent with background-only predictions.
• For a Higgs boson of mass 125 GeV, the observed 95% confidence level upper limit on the invisible branching fraction is 75%, with an expected limit of 91%.
• The limit on the invisible branching fraction increases with Higgs mass, reaching 85% (observed) and 105% (expected) at 145 GeV.
• The transverse mass m_T provides the best sensitivity among kinematic variables, yielding the lowest expected limit for each Higgs mass point.
• Systematic uncertainties are dominated by theoretical sources, particularly QCD scale variations in W±Z and ZZ processes, and PDF uncertainties.
• The control sample method using orthogonal ee, μμ, and eμ samples yields scale factors of α_μμ ≈ 0.6 and α_ee ≈ 0.4, with a 25% uncertainty assigned to account for b-tagging effects.