Search for resonant pair production of Higgs bosons decaying to two bottom quark-antiquark pairs in proton-proton collisions at 8 TeV

TL;DR

This CMS study searches for a narrow resonance decaying to a pair of Higgs bosons, each decaying to $b\bar{b}$, in $pp$ collisions at $\sqrt{s}=8$ TeV with $17.9\,\text{fb}^{-1}$. Employing region-specific event selections and a GaussExp-based background model, the analysis derives 95% CL upper limits on $\sigma(pp\to X)\times\mathcal{B}(X\to HH\to bbbb)$ for $m_X$ between $270$ and $1100$ GeV, finding no evidence for a signal. The results exclude a radion with $\Lambda_R=1$ TeV for $m_X$ in $300$–$1100$ GeV and KK gravitons for $m_X$ in $380$–$830$ GeV, thereby constraining warped extra-dimension scenarios. The approach combines data-driven control regions, a mass-constrained Higgs fit that improves $m_X$ resolution by $20$–$40\%$, and CLs statistics to set robust limits on resonant Higgs-pair production in multi-$b$-jet final states.

Abstract

A model-independent search for a narrow resonance produced in proton-proton collisions at sqrt(s) = 8 TeV and decaying to a pair of 125 GeV Higgs bosons that in turn each decays into bottom quark-antiquark pairs is performed by the CMS experiment at the LHC. The analyzed data correspond to an integrated luminosity of 17.9 inverse femtobarns. No evidence for a signal is observed. Upper limits at a 95% confidence level on the production cross section for such a resonance, in the mass range from 270 to 1100 GeV, are reported. Using these results, a radion with decay constant of 1 TeV and mass from 300 to 1100 GeV, and a Kaluza-Klein graviton with mass from 380 to 830 GeV are excluded at a 95% confidence level.

Figures (5)

• Figure 1: Illustration of the SR, SB, VR and VS kinematic regions in the ($m_{{{H}\xspace}_1}$, $m_{{{H}\xspace}_2}$) plane used to motivate and validate the parametric model for the QCD multijet background. The quantities $m_{{{H}\xspace}_1}$ and $m_{{{H}\xspace}_2}$ are the two reconstructed Higgs boson masses. The distribution in data events after b-tagging and kinematic selections is shown with the SR blinded.
• Figure 2: A maximum likelihood fit to the $m_X\xspace$ distribution of simulated signal events for the 700$\,\text{Ge\spaceV}$ mass hypothesis. The distribution is fitted to a Gaussian core smoothly extended on both sides to exponential tails. Here $n$ is the number of degrees of freedom in the fit.
• Figure 3: The $m_X\xspace$ distributions in data (after the ${t}\overline{{t}}$ background has been subtracted) in the SB of the MMR (top left), the CR of the MMR (top right), the VS of the LMR (bottom left), and the VR of the LMR (bottom right). The distributions are fitted to the GaussExp parametric model. The shaded regions correspond to $\pm$1$\sigma$ variations of this fit. Here $n$ is the number of degrees of freedom in each fit. The pull, for a given bin, is defined as the number of data events minus the value of the fit model, divided by the uncertainty in the number of data events.
• Figure 4: The $m_{\mathrm{X}}$ distribution in data in the SR between 260 and 650$\,\text{Ge\spaceV}$ of the LMR (top left), between 400 and 900$\,\text{Ge\spaceV}$ of the MMR (top right), and between 600 and 1200$\,\text{Ge\spaceV}$ in the HMR (bottom). All distributions are fitted to the background-only hypothesis for illustration, showing the relative contributions of the QCD multijet (dashed-dotted red) and ${t}\overline{{t}}$ (dashed green) processes. The pull, for a given bin, is defined as the number of data events minus the value of the background-only fit, divided by the uncertainty in the number of data events. Also for illustration, we overlay the signal models of the spin-0 resonance (dotted blue) corresponding to mass hypotheses and production cross sections of 350$\,\text{Ge\spaceV}$ and 653$\text{\,fb}$ for the LMR, 700$\,\text{Ge\spaceV}$ and 17.6$\text{\,fb}$ for the MMR, and 900$\,\text{Ge\spaceV}$ and 8.1$\text{\,fb}$ for the HMR. These cross sections correspond to the observed upper limits, which are computed for signal mass hypotheses from 270 to 450$\,\text{Ge\spaceV}$ in the LMR, from 450 to 730$\,\text{Ge\spaceV}$ in the MMR, and from 730 to 1100$\,\text{Ge\spaceV}$ in the HMR.
• Figure 5: The observed and expected upper limits on the cross section for $\mathrm{p}\mathrm{p} \to X \to{H}\xspace{H}\xspace \to {b}\xspace\overline{{b}}\xspace\xspace{b}\xspace\overline{{b}}\xspace\xspace$ at a 95% confidence level, where the resonance X has spin-0 (top) and spin-2 (bottom). The theoretical cross section for the RS1 radion, with $\Lambda_{R} =$1$\,\text{Te\spaceV}$, $kL = 35$, and no radion-Higgs boson mixing, decaying to four b jets via Higgs bosons is overlaid on the left plot. The theoretical cross section for the first excitation of the KK-graviton for the same parameters is overlaid on the bottom plot.