Search for decays of the Higgs boson into scalar particles decaying into four or six $b$-quarks using $pp$ collisions at $\sqrt{s}= 13\,\mathrm{TeV}$ with the ATLAS detector

TL;DR

This ATLAS study searches for exotic Higgs decays into light scalars that subsequently decay to $b$-quarks in $ZH$ events at $\sqrt{s}=13$ TeV using $140\,\mathrm{fb}^{-1}$. It employs specialized heavy-flavor techniques (merged $B$-jets, soft secondary vertices) and multivariate discriminants to target $H\rightarrow 2a\rightarrow 4b$ and $H\rightarrow a_1a_2\rightarrow 4b/6b$ across $12\le m_a \le 60$ GeV, with separate $2\ell$ and $0\ell$ channels. No significant excess is observed, and upper limits at 95% CL are set on $\sigma(ZH)/\sigma_{\text{SM}}(ZH) \times \mathcal{B}$ for each decay topology, reaching as low as a few percent in favorable mass regions. The analysis introduces novel reconstruction and tagging strategies, including low-mass merged $B$-jets and soft-vertex collections, and combines multiple channels to maximize sensitivity to multi-$b$ final states, thereby strengthening constraints on extended Higgs sectors and hidden-sector scenarios. The results complement prior searches and contribute important bounds on light scalar cascades in the Higgs sector.

Abstract

A search for exotic decays of the Higgs boson $H$ into new scalar or pseudoscalar particles that subsequently decay into $b$-quarks is presented. The search considers $ZH$ production with several decay scenarios for the Higgs boson: first to a pair of identical scalars, $H\rightarrow 2a \rightarrow 4b$, second to a pair of scalars with different masses ($m_{a1}<m_{a2}$), either directly, $H\rightarrow a_1a_2 \rightarrow 4b$, or via a longer decay chain, $H \rightarrow a_1a_2 \rightarrow 3a_1 \rightarrow 6b$. The analysis uses proton-proton collision data at $\sqrt{s} = 13$ TeV collected with the ATLAS detector at the Large Hadron Collider, corresponding to an integrated luminosity of $140~\mathrm{fb}^{-1}$. No significant excess above the Standard Model prediction is observed. The search sets upper limits at 95% confidence level on the ratio of the Higgs boson production cross-section to the SM prediction times the branching ratio of Higgs bosons decaying into $4b$ or $6b$, between 4% and 25% for $σ(ZH)/σ_{\text{SM}}(ZH) \times \mathcal{B}$($H \rightarrow 2a \rightarrow 4b$), between 24% and 38% for $σ(ZH)/σ_{\text{SM}}(ZH) \times \mathcal{B}$($H \rightarrow a_1a_2 \rightarrow 4b$), and between 10% and 20% for $σ(ZH)/σ_{\text{SM}}(ZH) \times \mathcal{B}$($H \rightarrow a_1a_2 \rightarrow 3a_1 \rightarrow 6b$), depending on the masses of the scalar particles.

Figures (15)

• Figure 1: Representative Feynman diagrams for $ZH$ production with the \ref{['fig:2a-4b']}$H \rightarrow 2a \rightarrow 4b$, \ref{['fig:a1a2-4b']}$H \rightarrow a_1a_2 \rightarrow 4b$, and \ref{['fig:a1a2-6b']}$H \rightarrow a_1a_2 \rightarrow 3 a_1 \rightarrow 6b$ processes considered in this search.
• Figure 2: Schematic overview summarizing the heavy-flavor algorithms used and the criteria used to define the $b$-objects.
• Figure 3: Diagram of the NN used for an example of a quadruplet selection showing the case of a hypothesis with $B$-jets, $b$-jets, and soft-v. All filled blocks are dense multi-layer perceptrons (MLP) that share common weights between MLPs of the same color. The $a$-boson MLPs are parameterized as a function of the mass hypothesis $m_a$. Unfilled rectangles are inputs from the four $b$-objects and the boson. The value of the quadruplet selection score, $\hat{y}_{\text{pred}}^{m_{a}}\xspace$, is used to classify the $b$-object quadruplet.
• Figure 4: Diagram depicting an example of how the NN selects a quadruplet. Each $b$-object is represented as an MLP ($b$-object MLP). The $a$-boson is represented as a pairing between $b$-objects, which also corresponds to an MLP ($a$-boson MLP). All physically motivated pairings are considered, including loop pairings representing merged reconstructed objects. The $a$-bosons themselves form new pairings representing the fully reconstructed Higgs boson, which also corresponds to an MLP (Higgs MLP). The value of this last pairing is used to select the quadruplet. The middle diagram shows the two reconstructed $a$-boson candidates and the rightmost diagram shows the final Higgs boson candidate.
• Figure 5: The quadruplet selection score $\hat{y}_{\text{pred}}^{m_{a}}\xspace$ distribution assuming the signal hypothesis $m_{a} = 25~\text{Ge V}\xspace$ for the $b$-object categories \ref{['fig:2B_NN']}$2B$ and \ref{['fig:1B2b_NN']}$1B2b$. Signal events for the $m_{a} = 25~\text{Ge V}\xspace$ mass hypothesis are also shown, assuming $\mathcal{B}=1$ and scaling this signal by a factor of ten. The category labeled "Other" includes processes with small contributions to the total yield (diboson, $\Pqt{}\Paqt Z\xspace$, and $\Pqt{}\Paqt W\xspace$). The lower panels show the ratio of the observed to the estimated SM background. The hashed area represents the total uncertainty in the background.