SEARCHES BSM HIGGS IN CDF -------------------------------------------- Silde 1 Title page Slide 2, 3 Contents Slide 4 The SM has demonstrated excellent agreement with experimental results. However, EW symmetry breaking mechanism still remains untested. The Higgs mechanism seems very popular, but there is no experimental evidence of the hypothesized new scalar boson. Slide 5 Moreover the SM has some other problems, like the fact that EW fits prefer a lighter Higgs mass that the one currently excluded by LEP. Also, hierarchy problem poses tensions due to loop corrections to the mass. These issues, however, are solved with SUSY Slide 6 SUSY arises from the proposal of a new symmetry: for each known particle, a hypothetical superpartner has opposite statistics. SM theoretical problems vanish, but a large number of free parameters are introduced Slide7 MSSM is a SUSY model with minimal Higgs content: 2 Higgs doublets that generate 5 physical states Two parameters describe at tree level the Higgs sector: tanBeta- the ration of the VEV of the doublets mA- the mass of the pseudoscalar Higgs Slide 8 The tevatron is sensitive to large values of tabBeta, because the coupling of Higss to b-quarks is enhanced by a factor ~tanBeta^2. The main production mechanisms are gloun fusion (b-quark loop) and b-quark anihilation (with extra b-quark content in the event) Slide 9 Cross sections of the prod mechsnism when extra b-quarks are requested Slide 10: 90% of the produced MSSM Higgs decay to b-quark pairs: bigger event yield, but backgrounds are more difficult and selections demands extra event content 10% decay to tau pairs. This is a small yield, but backgrounds are better control and both productions mechanisms can be probed Slide 12 The Tevatron collides protons and antiprotons at a CM energy of 1.96 TeV. Performance keeps improving, more than 6.5 fb-1 already delivered Slide 13 CDF has recorded more than 5.5 fb-1, but the results shown in the talk only cover up to the first 2.5 fb-1 Slide 14 CDF has the usual collider experiment structure Slide 16 Ditau final state is difficult, because tau have a rich decay spectrum. This analysis focus on three final states: hadronic tau and electron, hadronic tau and muon, and muon and electron. Slide 17 This analysis will select events compatible with the signatures searched for in an inclusive way (ie, no extra requirements) We run a profile likelihood fit on Visible mass templates. Visible mass is the mass of the sum of the decay products and missing et four-momentums Slide 18 Hadronic taus a challenging in hadron colliders due to the similarity with jets. For this reason, tau candidates with tracks or pi0s between the signal (10 degress) and the isolation cones (30 degrees) are vetoed. Moreover, the signal cone size decreases with the energy of the tau Silde 19 identification cuts applied to taus Slide 20 Electron reconstruction details. Z->ee mass spectrum is used as a control sample Slide 21 Muon reconstruction details. Z->mumu mass spectrum is used as a control sample Slide 22 Background estimation methods Slide 23 Event selection cuts Slide 24 Signal acceptance per channel, including both prod mechanisms. Overall acceptance ranges from 1% to 3% Slide 25 Sample of plots before fit. Four different distributions show that the agreement of data and our background estimation is really good Slide 26 fitting procedure Slide 27 Systematic uncertainties taken into account for the fit. These are introduced as gaussian constraints in to the likelihood Slide 28 Final 95% CL in cross section x branching ration as a function of Higgs mass. Very good agreement between expected and observed Slide 29 Results need to be interpreted in MSSM. Description of the procedure based on the combination of different theoretical calculations Slide 30 tanBeta vs mA 95% CL in the benchmark scenarios Slide 32 Summary of Tevatron results of this channel Slide 33 Tevatron Combination uses two methods, bayesian and frequentist. Results thru both methods are similar (within 10%) and the most conservative approach is picked Slide 34 Combination plots: 95% CL exclusion limits in cross section times branching ratio as a function of Higgs mass Slide 35 Description of the parameters as defined in the usual benchmark scenarios Slide 36 95% CL exclusion limits in the 4 benchmarks scenarios Slide 38 B-quark search strategy: selection with only 2 b-jets doesn't provide enough sensitivity. Instead, at least 3 b-jets are requested. Background estimation is performed in the 2 b-jet sample. Signal discrimination is performed following the CLs method, with a 2D fit Slide 39 One of the most difficult issues of this analysis is triggering on the events. This analysis uses the datasample collected with two different triggers that require displaced verteces Slide 40 Jet b-tagging relies on finding displaced vertexes. The operational point balances an efficiency of around a 50% and a mistag rate below 5% Slide 41 Event selection requires at least 3 jets identified as being originated from b-quarks. The main backgrounds are Z+jets, ttbar productions and QCD multijet Slide 42 The discriminant variables have been defined to incorporated the maximum relevant information as possible at the same time that a minimal number of histograms has to be fitted. Two varibles were selected: m12: the invariant mass of the two leading jets system (higgs?) xtag: unstacking of a 3x3 2D histogram representing in one axis the sum of the leading jets mass (m1+m2) and m3 on the other axis Slide 43 Fits for the null hypothesis: showing and excellent agreement of data with background predictions Slide 44 The 95% CL exclusion limits were calculated with the MCLIMIT package. This packages uses the CLs method to extract limits, and incorporates systematic uncertainties thru pseudoexperiments, when assessing sensitivity (expected limits) Slide 45 95% CL exclusion limits obtained without considering the width of the Higgs Slide 46 Same limits, but now in the MSSM Higgs parameter plane (width not considered yet) Slide 47 Interpretation of the limits in MSSM is tricky in this case. Radiative corrections depend on the scenario. Now the width in taken into account by convoluting the the cross sections with a BW. Sensitivity worsens a lot. Slide 49 Charged Higgs searches in the hadrionic decay channel are performed in the ttbar sample. One of the top quarks would decay to a Higgs, instead of a W and a b-quark Slide 50 A list of systematic uncertainties is provided. This systematics being considered, a binned likelihood fit to the di-jet mass spectrum is used to extract limits Slide 51 Result of this analysis: exclusion limits Slide 53 SM predicts a very small branching ratio to photons. However, some scenarios suppress the couplings of the Higgs to ferminons. In this case, the di photon final state is dominant at low mass Slide 54 Monte Carlo events were generateed with Pythia. Selected events contained at least one central photon and no more than one plug photon (two photons overall). Large invariant mass of the diphoton system was requested Slide 55 The resolution of the diphoton mass wa sstudied in detail, reaching down to 3 GeV. List of systematic uncertainties included Slide 56 Results of this analysis. Background was estimated on a different sample and fitted in the signal region Slide 58 No signal has been found yet... but we are still searching!!! Thanks for reading all this text :) \ \ \ \ \ \ \ \ \ \ }