Search for heavy resonances in the W/Z-tagged dijet mass spectrum in pp collisions at 7 TeV

TL;DR

The paper searches for heavy resonances decaying to quarks and vector bosons (qW/qZ) or vector-boson pairs (WW/WZ/ZZ) in fully hadronic final states using CMS data from pp collisions at sqrt(s)=7 TeV with an integrated luminosity of $5.0 ext{ fb}^{-1}$. It exploits jet substructure and W/Z tagging to identify hadronically decaying W/Z within jets, greatly reducing QCD backgrounds, and performs a Bayesian limit setting across benchmark models including excited quarks (q*), RS gravitons (G_RS), and W' bosons. The study produces 95% CL limits on cross sections and sets the most stringent mass bounds to date for qW/qZ final states in the fully hadronic channel (M(q*→qW) > 2.38 TeV and M(q*→qZ) > 2.15 TeV, observed), with complementary cross-section limits for WW/WZ/ZZ and insights into the relative sensitivity of ZZ versus WW channels. The results demonstrate the effectiveness of W/Z tagging in high-mass dijet searches and provide important constraints on new physics scenarios in hadronic final states.

Abstract

A search has been made for massive resonances decaying into a quark and a vector boson, qW or qZ, or a pair of vector bosons, WW, WZ, or ZZ, where each vector boson decays to hadronic final states. This search is based on a data sample corresponding to an integrated luminosity of 5.0 inverse femtobarns of proton-proton collisions collected in the CMS experiment at the LHC in 2011 at a center-of-mass energy of 7 TeV. For sufficiently heavy resonances the decay products of each vector boson are merged into a single jet, and the event effectively has a dijet topology. The background from QCD dijet events is reduced using recently developed techniques that resolve jet substructure. A 95% CL lower limit is set on the mass of excited quark resonances decaying into qW (qZ) at 2.38 TeV (2.15 TeV) and upper limits on the cross section for resonances decaying to qW, qZ, WW, WZ, or ZZ final states.

Figures (7)

• Figure 1: Comparisons of the dijet invariant mass distributions between data and Monte Carlo (pythia6 and herwig++) simulations. The three sets of lines correspond to the inclusive dijet category (no $\mathrm{W}/{Z}$-tag required), single $\mathrm{W}/{Z}$-tagged, and double $\mathrm{W}/{Z}$-tagged events. The simulations are normalized to the number of data events in each category.
• Figure 2: Pruned jet mass (top) and mass drop (bottom) in data, signal and background simulations. The signal simulation distributions are plotted as smooth curves connecting the histogram entries (using the same binning as the data distribution). All simulation distributions have been scaled to match the number of data events.
• Figure 3: The branching fraction into dijet final states B($\mathrm{W}/{Z} \to \text{jets}$) times angular acceptance ($|\eta| < 2.5$, $|\Delta\eta|<1.3$). The $\mathrm{W}/{Z}$-tagging efficiencies are excluded from the acceptance.
• Figure 4: Efficiency of requiring 1 $\mathrm{W}/{Z}$-tag (top) and 2 $\mathrm{W}/{Z}$-tags (bottom) in signal and background simulations, and in data for events passing the angular acceptance requirement ($|\eta| < 2.5$, $|\Delta\eta|<1.3$).
• Figure 5: The signal dijet invariant mass distributions for 1.5$\,\text{Te\spaceV}$$G_\mathrm{RS}\xspace\to \mathrm{W}\mathrm{W}$, $G_\mathrm{RS}\xspace\to {Z}{Z}$, $\mathrm{W^{'}} \to \mathrm{W}{Z}$, ${q}^*\to {q}\mathrm{W}$, and ${q}^*\to {q}{Z}$ resonances, computed using anti-$k_{\mathrm{T}}$ jets with $R=0.5$. Crystal-Ball functions fit to the simulated distributions are shown. The distributions in the plot are scaled to the same integral.