PaperlayerView exact revision on arXiv

Source-linked AI summary

Search for massive resonances in dijet systems containing jets tagged as W or Z boson decays in pp collisions at sqrt(s) = 8 TeV

CMS Collaboration

arXiv:1405.1994v2hep-ex

TL;DR

The paper searches for massive resonances in all-jet final states containing boosted W/Z decays, where hadronic vector-boson products merge into massive jets. Using CMS data and improved jet-substructure tagging, it finds no peak and sets 95% confidence-level mass limits on several benchmark models, with reinterpretation constrained by model-dependent efficiency assumptions.

  • Problem

    The search addresses whether high-mass resonances decaying to qW, qZ, WW, WZ, or ZZ appear in boosted all-jet events.

  • Method

    CMS analyzes 19.7 fb−1 of 8 TeV pp multijet data using W/Z jet-substructure tagging, including an improved N-subjettiness-based tagger, and data-modeled backgrounds.

  • Results

    No peak is observed; 95% confidence-level limits exclude q*→qW and qZ below 3.2 and 2.9 TeV, GRS→WW below 1.2 TeV, and W′→WZ below 1.7 TeV.

  • Takeaways & Limitations

    Mass limits are set for the first time on W′→WZ and GRS→WW in the all-jets final state, and the reported limits are the most stringent to date for the listed channels.

  • Takeaways & Limitations

    Reinterpreting nominal acceptances and W/Z-tagging efficiencies for arbitrary models requires an additional 10% uncertainty.

Abstract

from arXiv · show

A search is reported for massive resonances decaying into a quark and a vector boson (W or Z), or two vector bosons (WW, WZ, or ZZ). The analysis is performed on an inclusive sample of multijet events corresponding to an integrated luminosity of 19.7 inverse femtobarns, collected in proton-proton collisions at a centre-of-mass energy of 8 TeV with the CMS detector at the LHC. The search uses novel jet-substructure identification techniques that provide sensitivity to the presence of highly boosted vector bosons decaying into a pair of quarks. Exclusion limits are set at a confidence level of 95% on the production of: (i) excited quark resonances q* decaying to qW and qZ for masses less than 3.2 TeV and 2.9 TeV, respectively, (ii) a Randall-Sundrum graviton G[RS] decaying into WW for masses below 1.2 TeV, and (iii) a heavy partner of the W boson W' decaying into WZ for masses less than 1.7 TeV. For the first time mass limits are set on W' to WZ and G[RS] to WW in the all-jets final state. The mass limits on q* to qW, q* to qZ, W' to WZ, G[RS] to WW are the most stringent to date. A model with a "bulk" graviton G[Bulk] that decays into WW or ZZ bosons is also studied.

1 Introduction

The search targets high-mass resonances producing a quark and a W/Z boson or two vector bosons, using boosted-jet substructure in 8 TeV CMS data. It examines benchmark q*, graviton, and W′ models and sets updated or new mass limits.

  • Motivation: 1 TeV-scale resonances can produce quark–vector-boson or diboson final states whose quark decays merge into single massive jets.The search uses pp collisions at √s = 8 TeV and 19.7 fb−1 collected with CMS.
  • Analysis strategy: Jet-substructure techniques enhance sensitivity to W/Z jets by distinguishing them from quark and gluon remnants.The analysis improves on a previous 7 TeV CMS study with an N-subjettiness-based W/Z tagger.
  • Signal models: The benchmark models include q*→qW/qZ, RS and bulk gravitons→WW/ZZ, and W′→WZ.These processes yield events with one or two W/Z-tagged all-jet systems.
  • Prior constraints: Previous searches excluded q* masses below 2.4 TeV in qW and 2.2 TeV in qZ at the LHC.Earlier Tevatron limits were 0.54 TeV, while dijet searches constrained q* below approximately 3.5 TeV.
  • Model differences: The bulk-graviton model predicts predominantly longitudinal W/Z bosons, whereas the RS model favors transverse polarization.An improved bulk-graviton cross-section calculation predicts a fourfold smaller yield than previous studies assumed.
  • Prior constraints: Earlier W′→WZ searches set a lower mass limit of 1.1 TeV, while leptonic searches constrained mW′ above 2.9 TeV.The leptonic limit can vary by approximately 0.1 TeV with the chirality of W′ couplings.

2 The CMS detector, data, and simulated event samples

CMS reconstructs jets with tracking and calorimetry, while simulated signal and multijet samples support the analysis. Background distributions are ultimately modeled from data.

  • Detector: CMS combines a silicon tracker with electromagnetic and hadronic calorimeters to reconstruct particle jets.The tracker operates in a 3.8 T magnetic field, while ECAL and HCAL provide complementary measurements.
  • Detector coverage: The tracker covers |η| < 2.5, while ECAL and HCAL extend to |η| < 3 and |η| < 5, respectively.The calorimeter cells are grouped into towers whose dimensions vary with pseudorapidity.
  • Simulation: Signal samples use JHUGEN, PYTHIA, and HERWIG++ with GEANT4-based CMS detector simulation.The q*→W+jet and Z+jet processes are generated with PYTHIA.
  • Signal modeling: RS-graviton signals use k/MPl = 0.1, giving widths near 1% of the resonance mass and below the dijet-mass experimental resolution.HERWIG++ models the RS signal because it provides a more precise production-angle description.
  • Signal modeling: Bulk-graviton samples span k/MPl from 0.1 to 0.5, with reference samples generated at k/MPl = 0.2.The resonance width has no impact on the signal distribution over the considered range because of detector resolution.
  • Background modeling: Multijet simulations from HERWIG++ and MADGRAPH provide guidance and cross-checks, while the background distribution is modeled from data.The samples are interfaced to PYTHIA for parton showering and hadronization.

3 Event reconstruction and selections

The analysis selects high-mass dijet events and identifies boosted W/Z jets using pruned mass and N-subjettiness, while accounting for acceptance, efficiency, and modeling effects.

  • Event selection: At least two jets with pT > 30 GeV and |η| < 2.5 are required, with |∆η| < 1.3 and mjj > 890 GeV for the two leading jets.The dijet-mass requirement yields 99% trigger efficiency with negligible systematic uncertainty.
  • Efficiency and modeling: Signal efficiency is determined from full simulation, incorporating correlations between acceptance and W/Z-tagging efficiency.Reinterpretation using nominal acceptance times tagging efficiency requires an additional 10% uncertainty; the approximation agrees with full simulation to better than 10%.
  • W/Z-jet identification: The two highest-pT jets are pruned, and a pruned mass between 70 and 100 GeV tags a W/Z candidate.Fully merged W/Z decays peak near mj ≈80–90 GeV, unlike the approximately 20 GeV peak from multijet events and non-merged bosons.
  • W/Z-jet identification: The ratio τ21 = τ2/τ1 discriminates two-pronged W/Z decays from single multijet jets, whose distributions peak near 0.8 while signal favors smaller values.The analysis uses N-subjettiness without pruning after a one-pass optimization of subjet-axis directions.
  • Efficiency and modeling: Signal predictions differ across HERWIG++, PYTHIA, and JHUGEN/PYTHIA because of vector-boson polarization and showering or hadronization modeling.These modeling differences are included in the systematic-uncertainty estimation.

4 The search for a peak in the mass spectrum

The search models the dijet-mass spectrum as a possible resonance peak over a smoothly falling multijet background, using tagged event categories and likelihood fits. Data show no statistically compelling excess, and the resulting spectra constrain benchmark resonance models.

  • 4 The search for a peak in the mass spectrum: The multijet background is described by a smoothly falling empirical distribution separately for each event category.A Fisher F-test checks whether additional parameters are needed in the background model.
  • 4 The search for a peak in the mass spectrum: The analysis examines singly and doubly W/Z-tagged dijet-mass distributions in data, with multijet simulations shown normalized to data for comparison.The singly and doubly tagged distributions form the basis of the search.
  • 4 The search for a peak in the mass spectrum: Signal templates cover resonance masses of 1, 2, and 3 TeV for all models, plus 4 TeV for q* models, with linear interpolation between reference distributions.The W′ templates also contain a low-mass component near 0.8 TeV at large resonance masses, where multijet background dominates and the search is insensitive.
  • 4 The search for a peak in the mass spectrum: A binned maximum-likelihood fit searches for a peak while floating the signal scale and background parameters as functions of resonance mass.The expected bin yield is λ_i = µN_i(S) + N_i(B), combining signal and multijet-background contributions.
  • 4 The search for a peak in the mass spectrum: The largest local excesses are 1.8 standard deviations for q*→qW at 1.5 TeV and 1.3 standard deviations for G_RS→WW at 1.9 TeV.The G_bulk→WW/ZZ model yields no excess larger than one standard deviation.
  • 4 The search for a peak in the mass spectrum: Pseudo-experiments associate a local fluctuation of at least two standard deviations anywhere in the mass range with a global significance of one standard deviation.The dijet-mass distributions are used to set upper limits on production cross section times decay branching fraction.

5 Systematic uncertainties

Systematic uncertainties are dominated by signal modelling, especially W/Z-tagging efficiency, JES, JER, and luminosity. Tagging-efficiency uncertainties are constrained with a top-quark control sample, while other uncertainties affect the dijet-mass analysis.

  • Signal modelling, W/Z-tagging efficiency, JES, JER, and integrated luminosity provide the largest systematic uncertainties.
  • 7.5% and 54% uncertainties apply to HP and LP tagging scale factors, respectively, including control-sample statistics, JES, and JER contributions.
  • 1–2% JES uncertainty propagates to a 1% uncertainty in mjj across the examined resonance-mass range.
  • Acceptance uncertainties from PDF variations are assigned an envelope of 5%–15% over the relevant resonance-mass range.

6 Results

The analysis sets cross-section limits with the asymptotic LHC CLs method, combining HP and LP categories in a correlated likelihood. Observed limits are compared with benchmark predictions to derive resonance-mass exclusions.

  • Upper limits on resonance-production cross sections are set at 95% CL using the asymptotic approximation of the LHC CLs method.
  • HP and LP categories are combined in a common likelihood with anticorrelated tagging-efficiency uncertainties and correlated remaining signal uncertainties.
  • The q*→qW and q*→qZ mass limits increase by 0.8 and 0.7 TeV, respectively, relative to the previous 7 TeV search.
  • Mass limits are set for the first time in this channel on W′→WZ and GRS→WW, while no limits are set for GRS→ZZ or either bulk-graviton channel.
  • Figure 7 compares expected and observed 95% CL cross-section limits with predicted qW, qZ, and WZ benchmark cross sections as functions of resonance mass.

7 Summary

Using 19.7 fb^-1 of 8 TeV CMS multijet data, the search targets boosted vector bosons through W/Z-tagged jets and smooth-background fits. No resonance peak is observed, yielding the stated mass exclusions and sensitivity boundaries.

  • Pruned-jet mass and the N-subjettiness ratio τ21 identify merged vector-boson decays and suppress the multijet background.
  • The remaining background is estimated by fitting smooth analytic functions to the selected data.
  • 3.2 and 2.9 TeV lower mass limits are obtained for q*→qW and q*→qZ, respectively, at 95% CL.
  • GRS→WW is excluded below 1.2 TeV and W′→WZ below 1.7 TeV, while no bulk-graviton mass limits are set because predicted cross sections are small.
  • The analysis is sensitive to resonances heavier than 1 TeV, as summarized in Table 2.

Yerevan Physics Institute, Yerevan, Armenia

This section lists contributors and institutional affiliations associated with the paper, including researchers from Austria, Belgium, Brazil, Bulgaria, China, and other locations.

  • The listed contributors include V. Khachatryan, A.M. Sirunyan, and A. Tumasyan.
  • Researchers affiliated with the Institut für Hochenergiephysik der OeAW in Vienna, Austria, are listed.
  • The contributor list includes researchers associated with institutions in Belgium and Brazil.
  • Additional listed affiliations include institutions in Bulgaria, China, Egypt, Estonia, Finland, France, Georgia, Germany, Greece, Hungary, and Croatia.

P.N. Lebedev Physical Institute, Moscow, Russia

This section lists contributors affiliated with the P.N. Lebedev Physical Institute in Moscow, Russia and the Skobeltsyn Institute of Nuclear Physics at Lomonosov Moscow State University.

  • V. Andreev, M. Azarkin, I. Dremin, M. Kirakosyan, A. Leonidov, G. Mesyats, S.V. Rusakov, and A. Vinogradov are listed.
  • A. Belyaev, E. Boos, M. Dubinin7, L. Dudko, A. Ershov, A. Gribushin, V. Klyukhin, O. Kodolova, I. Lokhtin, S. Obraztsov, S. Petrushanko, V. Savrin, A. Snigirev, and other contributors are listed with the Skobeltsyn Institute.

State Research Center of Russian Federation, Institute for High Energy Physics, Protvino, Russia

This section lists contributors associated with Russian and Serbian institutions, as well as the CIEMAT research center in Madrid, Spain.

  • I. Azhgirey, I. Bayshev, S. Bitioukov, V. Kachanov, A. Kalinin, D. Konstantinov, V. Krychkine, V. Petrov, R. Ryutin, A. Sobol, L. Tourtchanovitch, S. Troshin, N. Tyurin, A. Uzunian, and A. Volkov are listed.
  • P. Adzic34, M. Dordevic, M. Ekmedzic, and J. Milosevic are listed under the Faculty of Physics and Vinca Institute of Nuclear Sciences in Belgrade, Serbia.
  • The CIEMAT section in Madrid lists J. Alcaraz Maestre, C. Battilana, E. Calvo, M. Cerrada, M. Chamizo Llatas2, and numerous additional contributors.

Universidad Aut´onoma de Madrid, Madrid, Spain

This section lists C. Albajar, J.F. de Trocóniz, and M. Missiroli.

  • C. Albajar, J.F. de Trocóniz, and M. Missiroli are listed.

Universidad de Oviedo, Oviedo, Spain

This section lists contributors associated with the Universidad de Oviedo in Oviedo, Spain.

  • H. Brun, J. Cuevas, J. Fernandez Menendez, S. Folgueras, I. Gonzalez Caballero, and L. Lloret Iglesias are listed.

Instituto de F´ısica de Cantabria (IFCA), CSIC-Universidad de Cantabria, Santander, Spain

This section lists contributors associated with the CMS collaboration and affiliated institutions, including Instituto de Física de Cantabria and numerous international universities and laboratories.

  • The listed contributors include members associated with Instituto de Física de Cantabria and the University of Cantabria in Santander, Spain.
  • The section includes contributors from universities and research institutions in Europe, North America, and Asia.
  • The entries identify individual CMS collaboration members and, in several cases, their institutional affiliations.
Loading 1405.1994v2…