Scalar Resonances near 650 and 95 GeV in the GNMSSM with Correct Dark Matter Relic Abundance

TL;DR

The paper tackles the question of whether the 95 GeV scalar hints and the 650 GeV diphoton+$b\bar b$ excess can be described within a single beyond-the-Standard-Model framework that also yields the correct dark matter relic density. It performs a comprehensive 12-parameter scan of the General NMSSM (GNMSSM) using SARAH/SPheno/MicrOMEGAs and MultiNest, enforcing Higgs-sector masses ($m_{h_s}=95.4$ GeV, $m_H=650$ GeV, $m_h=125$ GeV) and a suite of experimental constraints including LZ dark matter limits and Higgs-boson data. The analysis identifies two viable DM-relic scenarios with a bino-dominated LSP: Scenario I, where relic density is set by $A_s$ funnel annihilation or coannihilation with $\tilde\chi^0_2$ into $h_s A_H$, and Scenario II, where coannihilation with Higgsino-like states drives the abundance into $h_s A_s$ final states. Both scenarios can accommodate the observed excesses at approximately the 2$\sigma$ level and make concrete predictions for heavy Higgs phenomenology (e.g., $A_H$ in the 450–650 GeV range) and future collider tests, while remaining consistent with DM direct-detection bounds and current LHC searches. The study thus presents a coherent, testable framework tying together low-mass scalar hints, a heavy resonance decay topology, and dark matter within the GNMSSM, with explicit benchmark points for HL-LHC exploration.

Abstract

Recent CMS analyses report an excess in the diphoton-plus-$b \bar{b}$ channel, indicative of a heavy resonance around 650 GeV decaying into a Standard Model (SM)-like Higgs boson and a lighter scalar near 95 GeV. The case for a 95 GeV state is further supported by diphoton excesses observed by both CMS and ATLAS, as well as a $b\bar{b}$ excess previously observed at the Large Electron-Position collider. This study present a unified interpretation of these anomalies within the framework of the General Next-to-Minimal Supersymmetric Standard Model that naturally accommodates a light singlet-dominated $CP$-even scalar boson near 95 GeV and an heavier doublet-like scalar boson near 650 GeV. Through a comprehensive scan of the parameter space, we demonstrate that the model can explain these excesses at $2σ$ level while satisfying constraints from the dark matter relic density, direct detection experiments, the properties of the 125 GeV Higgs boson, $B$-physics observables, and searches for electroweakinos at the Large Hadron Collider (LHC). The interpretation features a Bino-dominated lightest neutralino as the dark matter candidate, whose relic abundance is achieved primarily via $A_s$ funnel annihilation or coannihilation with $\tilde{S}$-like $\tildeχ^0_2$s into $h_sA_H$ final states. Our findings provide clear predictions for testing this scenario at the high-luminosity LHC and future colliders.

Figures (5)

• Figure 1: Scattering plots of samples satisfying all applied constraints, projected onto the planes of the cross section $\sigma(gg\to H\to hh_s\to \gamma\gamma b\bar{b})$ versus the signal strengths $\mu_{\gamma\gamma}$ (left) and $\mu_{b\bar{b}}$ (right). The orange dashed lines and shaded bands indicate the central value of $\sigma(gg\to H\to hh_s\to \gamma\gamma b\bar{b})$ in Eq. (\ref{['XSbbrrExp']}) with the corresponding $2\sigma$ uncertainty, while the blue lines show the central values of $\mu_{\gamma\gamma}$ in Eq. (\ref{['diphoton-rate']}) (left) and $\mu_{b\bar{b}}$ in Eq. (\ref{['LEPrate']}) (right). The purple dash-dotted contours mark the combined $2\sigma$ regions on each plane. Grey dots represent samples excluded by HiggsSignals, HiggsBounds, or by DM relic density and direct detection constraints. Amber squares and blue dots denote Scenario I and Scenario II, respectively. Colored diamonds correspond to the four benchmark points, with details listed in Table \ref{['BP1BP2']} and Table \ref{['BP3BP4']}. \ref{['BP3BP4']}.
• Figure 2: Two-dimensional posterior PDF distributions of the heavy Higgs boson mass $m_H$ versus (a) $\mu_{\gamma\gamma}$, (b) $\mu_{b\bar{b}}$, (c) $\sigma(gg\to H\to hh_s\to \gamma\gamma b\bar{b})$, (d) $\sigma(gg\to H\to hh_s\to \tau\bar{\tau}b\bar{b})$, (e) $\sigma(gg\to A_H\to Zh \to Zb\bar{b})$, (f) $\sigma(gg\to A_H\to Zh_s \to \ell\ell b\bar{b})$, (g) $\sigma(gg\to H\to h_sh_s\to b\bar{b}b\bar{b})$, (h) $C_{Ht\bar{t}}$, and (i) $C_{A_Ht\bar{t}}$. Solid and dashed contours indicate the $1\sigma$ and $2\sigma$ credible regions, respectively. Colored diamonds denote the four benchmark points corresponding to those in Fig. \ref{['Fig1']}, with details provided in Table \ref{['BP1BP2']} and Table \ref{['BP3BP4']}.
• Figure 3: Two-dimensional profile likelihoods projected onto the $(\lambda, \kappa)$, $(m_N, \delta)$, $(m_N, A_\lambda)$, and $(\mu, \mu^\prime)$ planes. Dashed-dotted and dashed contours indicate the $1\sigma$ and $2\sigma$ confidence intervals, respectively. The best-fit point has a total $\chi^2$ value of $\chi^2_{650+95} + \chi^2_{125} \simeq 162$. Colored diamonds denote the four benchmark points shown in Figs. \ref{['Fig1']} and \ref{['Fig2']}.
• Figure 4: Same as Fig.\ref{['Fig3']}, but projected onto $(M_1, m_N)$, $(M_1, \mu_{tot})$, $(M_1, M_2)$ planes.
• Figure 5: Spin-independent (SI) and spin-dependent (SD) DM–nucleon scattering cross sections, $\sigma^{\rm SI}_p$ (left) and $\sigma^{\rm SD}_n$ (right), as functions of the LSP mass $m_{\tilde{\chi}_1^0}$. Dashed black lines show the 90% C.L. upper limits from the 2025 LZ results LZ:2024zvoLZ2024slides, dashed-dotted grey lines indicate the projected LZ sensitivities for future runs LZ:2018qzl, and dotted lines denote the neutrino floor Billard:2013qya. Amber squares and blue dots correspond to Scenario I and Scenario II, respectively. Colored diamonds represent the four benchmark points shown in Figs. \ref{['Fig1']}–\ref{['Fig4']}.