Interpreting the 650 GeV and 95 GeV Higgs Anomalies in the N2HDM

TL;DR

This work analyzes whether the Next-to-2-Higgs-Doublet Model (N2HDM) can jointly explain the 95 GeV Higgs hints seen at LEP and LHC and the 650 GeV resonance reported by CMS. Focusing on CP-even cascades with a heavy $h_3$ around $650$ GeV decaying as $h_3\to h_2 h_1$ (where $h_2\equiv H_{ m SM}$ and $h_1$ is the 95 GeV state), the authors explore Type-II and Type-Y Yukawa structures and impose comprehensive theoretical and experimental constraints. They find that a CP-even interpretation remains viable at about the $2\sigma$ level in both types, while a CP-odd explanation is not supported when fitting all anomalies. The paper provides benchmark points and makes testable predictions for Run-3 and HL-LHC in the $\gamma\gamma b\bar{b}$, $b\bar{b}$, and $\tau^+\tau^-$ channels, offering a concrete path to experimentally validate or refute the N2HDM scenario.

Abstract

Recent experimental hints from the Large Hadron Collider (LHC) in di-photon and partially in the $τ^+τ^-$ final states suggest the possible existence of an additional Higgs boson with a mass around 95 GeV. Interestingly, these observations are consistent with earlier results from the Large Electron-Positron (LEP) collider, which pointed to an excess in $b\bar b$ final states within a similar mass range. Additionally, CMS has observed an excess in the $γγb\bar{b}$ final state, indicating a possible resonance near 650 GeV decaying into a pair of SM-like Higgs bosons or into a SM-like Higgs boson accompanied by a lighter scalar with mass near 95 GeV. In this work, we investigate whether these anomalies can be simultaneously explained within the Next-to-2-Higgs-Doublet Model (N2HDM), an extension of the Standard Model (SM) scalar sector featuring two complex Higgs doublets and an additional real singlet. Assuming the existence of a CP-even Higgs state compatible with the 95 GeV excesses (restricted to the $γγ$ and $b\bar b$ channels), we analyse the Type-II and Type-Y Yukawa structures, taking the observed 650 GeV resonance to be a CP-even Higgs state. An extensive parameter scan is performed, incorporating the latest constraints. Our results show that a heavy CP-even Higgs resonance around 650 GeV, produced predominantly via gluon-gluon fusion and subsequently decaying into a 125 GeV Higgs boson together with another scalar at approximately 95 GeV, can be simultaneously accommodated within both the N2HDM Type-II and Type-Y frameworks at a significance level of 2$σ$. This interpretation leads to distinctive and testable predictions from the N2HDM for the ongoing LHC Run-3 and the forthcoming High-Luminosity LHC (HL-LHC) phase.

Figures (3)

• Figure 1: The cross section $\sigma(pp \to h_3)\times {\rm BR}(h_3 \to h_2 h_1 \to \gamma\gamma b\bar{b})$, denoted as $\sigma_{\gamma\gamma b\bar{b}}$, shown as a function of the signal strengths ($\mu^{h_1}_{\gamma\gamma}$, $\mu^{h_1}_{bb}$ and $\mu^{h_1}_{\tau^+\tau^-}$) of the lighter Higgs boson $h_1$ at approximately 95 GeV, within the N2HDM frameworks. The upper(lower) panels correspond to Type-II(Type-Y) Yukawa interactions. The shaded horizontal blue bands represent the experimentally allowed 2$\sigma$ regions for the cross section of the 650 GeV resonance observed by CMS in the $\gamma\gamma b\bar{b}$ final state. The vertical shaded brown regions represent the experimentally allowed 1$\sigma$ intervals for the signal strengths of the 95 GeV resonance in each decay channel. Colours of the points quantify the ${\rm BR}(h_3 \to h_1 h_2)$. The squares with magenta color indicate points excluded by the CMS Higgs resonance searches in the $\tau^+\tau^- b\bar{b}$ final state CMS:2021yci.
• Figure 2: The cross section $\sigma_{\gamma\gamma b\bar{b}}$ as a function of $\tan\beta$ values versus the BRs of $h_3 \to h_1 h_2$ (left panel), $h_1 \to b\bar{b}$ (middle) and $h_2 \to \gamma\gamma$ (right) in the N2HDM Type-II (top) and Type-Y (bottom). Magenta squares are the same as in Fig \ref{['fig1']}.
• Figure 3: Surviving parameter points for $\sigma(pp \to h_3)\times {\rm BR}(h_3 \to h_2(\to b\bar{b}) h_1(\to \gamma\gamma))$, overlaid onto the observed 95$\%$ CL exclusion limits CMS:2025qit. All points satisfy the full set of theoretical and experimental constraints discussed above within the N2HDM Type-II (left) and Type-Y (right).