Partially (in)visible Higgs decays at the LHC

TL;DR

The paper investigates constraints on non-SM Higgs decays, specifically $H\rightarrow AA$ with a light scalar $A$, leading to partially invisible final states. It adopts a simplified-model framework and uses Higgs-strahlung, with $Z\to\ell^+\ell^-$, to trigger and reconstruct the event while applying jet-substructure techniques to separate signal from sizable backgrounds. Through detailed event generation, detector modeling, and CLs-based limit setting across several topology scenarios, it finds sensitivity to branching ratios down to about 10% in favorable cases, while ISR-dominated or highly missing-energy–driven scenarios yield weaker constraints. The study demonstrates a viable path to directly constrain partially invisible Higgs decays at the LHC and underscores the role of advanced jet-substructure observables for mass reconstruction and signal discrimination.

Abstract

Both Atlas and CMS have reported a discovery of a Standard Model-like Higgs boson $H$ of mass around 125 GeV. Consistency with the Standard Model implies the non-observation of non-SM like decay modes of the newly discovered particle. Sensitivity to such decay modes, especially when they involve partially invisible final states is currently beyond scrutiny of the LHC. We systematically study such decay channels in the form of $H\rightarrow AA\rightarrow jets+missing energy$, with $A$ a light scalar or scalar, and analyze to what extent these exotic branching fractions can be constrained by direct measurements at the LHC. While the analysis is challenging, constraints as good as ${BR}\lesssim 10%$ can be obtained.

Figures (6)

• Figure 1: Higgs decay topologies in the simplified models that we study for the purpose of this paper. "vis" is a placeholder for $u\bar{u},d \bar{d}$ flavor quark pairs that give rise to visible hadronic energy.
• Figure 2: $\Delta{\slashed{E}}_T/{\slashed{E}}_T$ as a function of the true $\slashed{E}_T$ that follows from particle flow and early LHC data pflow. We also display the fitted function that is employed for our analysis.
• Figure 3: Missing transverse energy (a) and fat jet transverse momentum (b) of the decay scenarios Fig. \ref{['fig:topref']}(a)-(d) and the contributing backgrounds after all analysis steps have been applied.
• Figure 4: Transverse mass or transverse cluster mass distributions after filtering and filtering+trimming, depending on the scenario of Fig. \ref{['fig:topref']}. We assume $\text{BR}(\text{scenario~(i)})=1,~i=(a)-(d)$ for illustration purposes.
• Figure 5: 95% confidence level exclusion of the various scenario's branching ratio (or signal strength $\xi=\sigma\times \text{BR}/\sigma_{\text{SM}}$ to be more specific) from a direct measurement along the lines of Sec. \ref{['sec:analysis']}. We assume a background template uncertainty of initially 10%, which is we assume to saturated at 5% at end-of-lifetime LHC luminosities of ${\cal{L}}=1000/{\text{fb}}$.