Composite Octet Searches with Jet Substructure

TL;DR

This work targets the collider signatures of a new confining gauge sector that produces a color-octet vector meson $\rho_G^a$ and a color-octet pseudoscalar $\pi_G^a$. It presents a simplified model with kinetic mixing between $\rho_G$ and gluons and a $\rho_G$-$\pi_G$-$\pi_G$ coupling, exploring decays $\pi_G \to gg$ or $\pi_G \to b\bar{b}$ and the resulting multi-jet final states. The authors compare a simple jet-mass dijet strategy to a sophisticated jet-substructure approach (mass-drop and $N$-subjettiness) for resonant production $q\bar{q}\to\rho_G\to\pi_G\pi_G$, using 7 TeV LHC simulations and a benchmark point to demonstrate substantially enhanced discovery potential with substructure tagging for intermediate mass ratios $m_{\pi_G}/m_{\rho_G}$. They conclude that jet-substructure techniques extend the reach across a broad parameter space, with simple jet mass winning in highly hierarchical spectra ($m_{\pi_G}/m_{\rho_G}\lesssim 0.2$) and traditional multi-jet searches regaining sensitivity at larger ratios, while nonresonant production channels provide complementary avenues for discovery.

Abstract

Many new physics models with strongly interacting sectors predict a mass hierarchy between the lightest vector meson and the lightest pseudoscalar mesons. We examine the power of jet substructure tools to extend the 7 TeV LHC sensitivity to these new states for the case of QCD octet mesons, considering both two gluon and two b-jet decay modes for the pseudoscalar mesons. We develop both a simple dijet search using only the jet mass and a more sophisticated jet substructure analysis, both of which can discover the composite octets in a dijet-like signature. The reach depends on the mass hierarchy between the vector and pseudoscalar mesons. We find that for the pseudoscalar-to-vector meson mass ratio below approximately 0.2 the simple jet mass analysis provides the best discovery limit; for a ratio between 0.2 and the QCD-like value of 0.3, the sophisticated jet substructure analysis has the best discovery potential; for a ratio above approximately 0.3, the standard four-jet analysis is more suitable.

Figures (8)

• Figure 1: Left panel: the branching ratios of $\rho_G^\mu$ into different modes as a function of the mixing angle. Right panel: the $\rho_G^\mu$ width over its mass as a function of $\tan{\theta}$. The daughter particle masses are neglected.
• Figure 2: Left panel: the production cross section of $\rho_G$ at the LHC for $\tan{\theta}=0.15$. The range of cross section is for two different renormalization scales: $\frac{m_{\rho_G}}{2}$ (upper) and $2 m_{\rho_G}$ (lower). Right panel: the production cross section of different decaying modes, where the renormalization scale is fixed to be $m_{\rho_G}$.
• Figure 3: Dijet constraints on the model parameters at 95% C.L from Atlas dijet narrow resonance searches with 163 pb$^{-1}$ luminosity dijetupdate. The shaded regions are excluded. The two boundary lines are for (lower) 100% efficiency to detect $\pi_G^a$ as a single jet and (upper) 0% efficiency to detect $\pi_G^a$ as a single jet. Branching ratios are given by Eq. (\ref{['eq:decay']}). The left panel shows limits for $m_{\pi_G}/m_{\rho_G} =0.1$ while the right panel shows limits for $m_{\pi_G}/m_{\rho_G} = 0.3$.
• Figure 4: Left panel: the numbers of events of signal and backgrounds at the 7 TeV LHC after the jet substructure analysis. Right panel: the same as the left panel but for the analysis without using jet substructure.
• Figure 5: Left panel: the discovery significance for different masses of $\rho_G$ and $\pi_G$ for $\pi_G \rightarrow gg$. We scanned four variables to find the optimized significance: the mass drop variable $\mu$, the symmetric splitting cut $r_{xy}$, the $p_T$ cut of the fat jets, and the $N$-subjettiness variable $\tau_2/\tau_1$. The numbers besides each contour line are the significance in $\sigma$. Right panel: the same as the left panel but instead of using sophisticated jet substructure analysis, only the jet masses are used in this plot.