Search for additional scalar bosons within the Inert Doublet Model in a final state with two leptons at the FCC-ee

TL;DR

This work assesses the discovery and exclusion potential of the Inert Doublet Model (IDM) at an $e^+e^-$ collider (FCC-ee) in final states with two same-flavour leptons, using $\\\sqrt{s}=240$ and 365 GeV. Signal samples are generated with MG5_aMC@NLO + PYTHIA and simulated with DELPHES (IDEA detector), while a parametric neural network (pNN) exploits kinematic differences across IDM mass points and interpolates to unseen masses. With integrated luminosities of 10.8 ab$^{-1}$ at 240 GeV and 2.7 ab$^{-1}$ at 365 GeV, almost the entire $(M_A-M_H$, $M_H)$ parameter space is excluded at 95% CL (up to $M_H=110$ GeV and $M_H=165$ GeV, respectively), and the discovery reach extends to $M_H=108$ GeV (240 GeV) and $M_H=157$ GeV (365 GeV) for a mass splitting $M_A-M_H=15$ GeV. The analysis showcases the FCC-ee’s sensitivity to IDM dark scalars and demonstrates the effectiveness of a pNN in concentrating signal while suppressing SM backgrounds, providing complementary constraints to DM relic density and direct-detection results.

Abstract

In this work, we investigate the discovery reach of a new physics model, the Inert Doublet Model, at an $e^+e^-$ machine with centre-of-mass energies $\sqrt{s}$ of 240 and 365 GeV. Within this model, four additional scalar bosons ($H$, $A$, $H^+$ and $H^-$) are predicted. Due to an additional symmetry, the lightest new scalar, here chosen to be $H$, is stable and provides an adequate dark matter candidate. The search for pair production of the new scalars is investigated in final states with two electrons or two muons, in the context of the future circular collider proposal, FCC-ee. Building on previous studies in the context of the CLIC proposal, this analysis extends the search to detector-level objects, using a parametric neural network to enhance the signal contributions over the Standard Model backgrounds, and sets limits in the $m_A-m_H$ vs $m_H$ plane. With a total integrated luminosity of 10.8 (2.7) ab$^{-1}$ for $\sqrt{s}=240$ (365) GeV, almost the entire phase-space available in the $m_A-m_H$ vs $m_H$ plane is expected to be excluded at 95% CL, reaching up to $m_H=110$ (165) GeV. The discovery reach is also explored, reaching $m_H= 108$ (157) GeV for $m_A-m_H=15$ GeV at $\sqrt{s}=240$ (365) GeV.

Figures (12)

• Figure 1: Leading-order Feynman diagrams of: (left) $HH\ell^+\ell^-$ production, (middle) $HH\ell^+\ell^-\nu\bar{\nu}$ production, and (right) $\HepParticle{Z}{}{}\xspace h\rightarrow HH\ell^+\ell^-$ production, with $\ell=e,\mu,\tau$.
• Figure 2: Allowed points in a general scan with $\lambda_2\,=\,0.1$(purple, $+$), as well as points used in the simulation (green, x). Some of the latter are forbidden by electroweak precision observable constraints, which were implemented via the oblique parameters $S,\,T,\,U$(red, $*$).
• Figure 3: Cross section from the MG5_aMC@NLO simulation at $\sqrt{s}\xspace=$240 $\text{GeV}$, for the points simulated in the $M_A-M_H$ vs $M_H$ plane in scenario S1. All cross sections are in fb, computed for $p_{\mathrm{T}}^{\ell}>0.5$ GeV and with a numerical integration error smaller than 0.1%. Left: $HH\ell\ell$ final state, right: $HH\ell\ell\nu\nu$ final state. The dashed line shows the kinematic limit for on-shell production.
• Figure 4: Cross section from the MG5_aMC@NLO simulation at $\sqrt{s}\xspace=$365 $\text{GeV}$, for the points simulated in the $M_A-M_H$ vs $M_H$ plane in scenario S1. All cross sections are in fb, computed for $p_{\mathrm{T}}^{\ell}>0.5$ GeV and with a numerical integration accuracy smaller than 0.1%. Left: $HH\ell\ell$ final state, right: $HH\ell\ell\nu\nu$ final state. The dashed line shows the kinematic limit for on-shell production.
• Figure 5: Cross section ratios S3/S2 in the $M_A-M_H$ vs $M_H$ plane, from the MG5_aMC@NLO simulation at $\sqrt{s}\xspace=$240 $\text{GeV}$ (left) and $\sqrt{s}\xspace=$365 $\text{GeV}$ (right), for the $HH\ell\ell\nu\nu$ final state. The dashed line shows the kinematic limit.