C. Ciobanu, page 1 Highlights from High P T Physics at CDF Catalin Ciobanu, Univ. of Illinois Urbana-Champaign, for the CDF Collaboration Fermilab Users Meeting, June 8, 2005 Topics from electroweak, top, and exotic physics C. Ciobanu, page 2 CDF at the Tevatron l Tevatron and CDF performance: Last month was very exciting: Init. luminosity > 1.3 cm -2 s -1 Single store: > 5 pb -1 Delivered integrated luminosity: 1fb -1 CDF recorded ~0.8 fb -1 High-luminosity upgrades on schedule l Current analyses use fb -1 Records falling like dominos! C. Ciobanu, page 3 Electroweak Physics l W and Z leptonic decays Yardstick for all high-P T lepton analyses Backgrounds for many new physics processes Used for calibration of energy/momentum scales l Large datasets better and broader physics reach: W/Z cross-sections: Phys.Rev.Lett.94:091803,2005 Z boson asymmetry: Phys.Rev.D71:051104,2005 W charge asymmetry: Phys.Rev.D71:051104,2005 W , Z production: Phys.Rev.Lett.94:041803,2005 ZZ, ZW production: Phys.Rev.D71:091105,2005 WW production: hep-ex/ l Highlights: W boson mass measurement Diboson measurements C. Ciobanu, page 4 Introduction to m W l W mass fundamental parameter of SM Radiative corrections: depend on m t and m H,? M W =34GeV l Measure m W, m t precisely constrain m H ! l Currently, m W known to a precision of 34 MeV: LEP: 80,447 42 MeV Tevatron: 80,454 59 MeV (Run I) Uncertainty: 79 MeV (CDF), 84 MeV (D0), with L 120 pb -1 l Run II Goal: uncertainty < 40 MeV per experiment C. Ciobanu, page 5 W Mass l How do we measure W mass: Clean signals: W e, W Most information comes from P T ( l ), l = e, Neutrino P T ( ) inferred via hadronic recoil Calculate transverse mass: W e W l Backgrounds (contribute m W ~20 MeV) C. Ciobanu, page 6 W Mass Status Sourcee: Run2 (IB) : Run2 (IB) Lepton Energy Scale and Resolution 70 (80)30 (87) Recoil Scale and Resolution50 (37)50 (35) Backgrounds20 (5)20 (25) Production and Decay Models30 (30) Statistics45 (65)50 (100) Total105 (110)85 (140) CDF RUN II PRELIMINARY CDF RUN II PRELIMINARY m T ()(GeV) m T (e)(GeV) l First pass at W mass with 200 pb -1 : Uncertainty m W =76 MeV lower than CDF Run I Work in progress to reduce m W : Adjust tracker aligment Recoil resolution and P T (W) Passive material and radiation First m W results to be finalized soon W transverse mass m T fit: C. Ciobanu, page 7 Dibosons: W , Z , WZ, ZZ SM expect: (W ) = 19.3 1.4 pb SM expect: (Z ) = 4.5 0.3 (pb) l Test of WW couplings l Constrain ZZ and Z vertices (absent in SM) l Background of gauge-mediated SUSY Breaking models Phys.Rev.Lett.94:041803,2005 l WZ, ZZ important step towards Higgs searches l W Z unavailable at e + e - colliders => unique meas. of WWZ (ppZZ/ZW+X) WW) THEORY NLO = 12.4 0.8 pb hep-ex/ l Can also search for lepton+jets signature (l jj): Higher BR than dilepton final state Larger uncertainties Smaller S/B ratio Constrain anomalous triple gauge couplings C. Ciobanu, page 9 Top Physics at CDF l Top is the most intriguing particle we know: Why so heavy? m t, m W, m H are related via loop diagrams Top decays before hadronizing (4 s) SM Yukawa coupling ~1 (coincidence?) l CDF top group very dynamic: Single-top-quark search: Phys.Rev.D71:012005,2005 W polarization in top decays: Phys.Rev.D.71:031101,2005 Top pair cross section b-tag: Phys.Rev.D71:052003,2005 Top pair cross section kinematic+b-tag: Phys.Rev.D71:072005,2005 Top pair cross section kinematic (NN): hep-ex/ ,2005 Top pair cross section with Soft Muon Tagging: hep-ex/ ,2005 Anomalous kinematics in tt dilepton events: FERMILAB-PUB E Br(t Wb)/B(t Wq): FERMILAB-PUB E l Highlights: Top mass measurement Top cross-section measurement GeV/c 2 C. Ciobanu, page 10 Top Phenomenology l Only at the Tevatron! (so far) l Strong (pair) production main channel (ttX) ~ 6-7 pb l Top decay: SM: t Wb almost 100% of the time CDF: V tb >>V ts, V td Classified according to W decays Lepton+jets, dilepton, all hadronic l 5% Dilepton (ee, , e ) l 30% Lepton+jets (e+j, +j) l 45% All hadronic C. Ciobanu, page 11 Top Quark Mass Run I average: Using pb -1 per exp. Lowest uncertainty: D0 l+jets result Run 2 goal: 300 pb -1 We should do better than Run I Difficult measurement: Jet energies not precisely measured Final state difficult to reconstruct Many possible configurations ISR and FSR further complicate the picture Small statistics (esp. dileptons) C. Ciobanu, page 12 Jet Energy Scale Uncertainty l Jet energies imprecisely measured. Poor resolution statistical error. 50 GeV quark measured as 30 GeV jet Uncertain scale systematic error. Jets are hard to calibrate (3-5%) (no nice resonance to use) Run II 2004 Run II Run I C. Ciobanu, page 13 Top Mass in l+jets events l Lepton+jets channel: ttbar b-tag event efficiency 60% Category2-tag1-tag(T)1-tag(L)0-tag j1-j3E T >15 E T >21 j4E T >8E T >1515>E T >8E T >21 S:B18:14.2:11.2:10.9:1 Observe l Reconstruct mass: 2 mass fitter one number per event Best m t and combination assignment ttbar Monte Carlo C. Ciobanu, page 14 Fit to Data l Simultaneous fit in the 4 channels Luminosity: 318 pb -1 l Systematic uncert. dominated by JES: m t = /-2.8 (stat) +/- 3.4 (syst) GeV/c 2 = /-4.4 GeV/c 2 Jet Energy Systematic SourceUncert. (GeV/c 2 ) Relative to Central Region0.6 Corrections to Hadrons (Absolute Scale) 2.2 Corrections to Partons (Out-of- Cone) 2.1 Total3.1 .The answer is : C. Ciobanu, page 15 Using M W to constrain JES l Build templates using invariant mass m jj of all non-tagged jet pairs l 1 3.1 GeV uncertainty l Rather than assuming JES and measuring M W... l Assume M W and measure JES l Parameterize P(m jj ;JES) same as P(m t reco ;M top ) C. Ciobanu, page 16 2D Templates l Account for correlations: P(m reco |M t,JES) and P(m jj |M t, JES) templates for every point (M t, JES) More complex, but tractable l (2D) Fit data simultaneously for M t, JES l This reduces JES uncertainty: In turn, top mass uncertainty is reduced l Bonus: Using this method, JES scales down directly with luminosity! C. Ciobanu, page 17 Fit to Data (318 pb -1 ) Future is bright ahead! Today with 3% prior on JES C. Ciobanu, page 18 Implications l New result is world best measurement of M top ! Thanks to: S. Heinemeyer G. Weiglein M. Grunewald C. Ciobanu, page 19 Other Techniques and Results l Dynamical Likelihood Method Similar to Run I best measurement (D): (Nature 429, 638 (2004)) Calculates probability for each event to be signal with a given top mass Calculation based on full Standard Model matrix-element Very statistically powerful! m t =173.8 (stat) 3.3 (syst.)GeV/c = 4.1 GeV l Stay tuned for results from dilepton and all-hadronic analyses l Publications out by end of summer C. Ciobanu, page 20 Top Pair Cross-Section l Top pair cross-section: Depends on top mass m t We need to measure both m t and (tt) l Use e+jets, +jets events l Traditionally, require b-tag l Alternately, relax b-tag condition: Increase statistics But decrease S/B l Use 7-input Neural-Network: Energy and angular information M top =178 GeV/c 2 C. Ciobanu, page 21 tt Cross-Section: b-tagging (tt) = 7.9 0.9 (stat) 0.9 (syst) pb Displaced vertices (silicon vertex detector)Soft lepton (muon) tagging (tt) = 5.3 (stat) (syst) pb +3.3 1.0 M top =175 GeV/c 2 C. Ciobanu, page 22 Exotic Top Decays? l Standard Model decay: t Wb l Two Higgs doublet extension of SM: H , A(CP-odd), H,h (CP-even) t H + b with H + , H + cs, H + Wbb l Search in 4 channels: Dileptons, lepton+jets (1, or 2 b-tags), lepton+ (decaying hadronically) CDF Run II Preliminary Most conservative BR combination C. Ciobanu, page 23 New Physics at CDF l Where to look? Everywhere we can! l Events with leptons (even s) provide clean signals l Higgs searches: MSSM Higgs A + - (310 pb -1 ) l Search for massive resonances: Spin 0: Scalar neutrino production and decay (344 pb -1 ) Spin 1: Z-bosons search (448 pb -1 ) Spin 2: Randall-Sundrum gravitons C. Ciobanu, page 24 MSSM Higgs: A l Enhanced production of A at hadron colliders A decays: bb (~90%), (~8-9%) bb: larger BR, but need production in association of b-quarks to control bgrds allows searches in all production channels l One tau decays leptonically, the other one hadronically l Backgrounds: Z Events with jet fakes (pp H/A/h) tan 2 C. Ciobanu, page 25 A Results Limits set using a likelihood fit of m vis (mass-like variable) Signal region: M>120 Expect: 8.4 events Observe: 11 events C. Ciobanu, page 26 Scalar Neutrino Search l R-parity violation Lepton-number violating terms: ijk, ijk l Resonance-like production Start with e+ channel Very clean signature Sensitive to both and l High acceptance in the region of interest Dominant background is Z l Expect 71.3 events. Observe 74 Set limits C. Ciobanu, page 27 Scalar Neutrino Search l Interpret x-section limit: For 132 =0.05 311 =0. 16 M>450 GeV Can set a limit in 2D 311 vs M( ) Assume 2 Br( e)=0.01 CouplingChannelLimit (GeV) eeM>680 ee M>460 ee - M>665 - M>375 C. Ciobanu, page 28 Z Searches l Most extensions to SM predict new gauge interactions Most give neutral or singly charged bosons Tight constraints on Z 0 -Z mixing from LEP l Search for decays: Z e + e Use dielectron mass M ee and cos * Dataset: 448 pb -1 l No evidence of signal: Set 95% C.L. limits: Z ModelM Z Limit (GeV) Seq. Z 845 E6 Z 720 E6 Z 690 E6 Z 715 E6 Z I 625 = world best = world best excluding LEP C. Ciobanu, page 29 Z Searches l Adding angular information helps: At 448 pb -1 : use cos * +25% data (seq. Z) l 2004 paper by Carena et al defines 4 general model classes ( PRD70:093009,2004) : B-xL, q+xu, 10+x5, d-xu Within each class, Z defined by: mass M Z, strength g z, parameter x LEP II excluded below line l Comparisons to LEP 2 possible: q+xu, 10+x5, d-xu more exclusion than LEP II As well as LEP II for B-xL More sensitive than LEP to low g z couplings C. Ciobanu, page 30 Conclusions l The high-P T physics program at CDF is broad: 10 Run 2 high-P T papers published in 2005 Top program entering new territory Worlds best top mass measurement Unprecedented top-quark datasets Electroweak: W mass result being finalized New precision on Diboson results Exotics: New MSSM Higgs and heavy boson searches Many searches for New Phenomena ongoing CDF has 800 pb -1 on tape More precise measurements and increased sensitivity for New Physics searches C. Ciobanu, page 31 W Mass Uncertainty l Production and Decay Model Uncertainties: PDFs, QED corrections, P T (W) and (W) model: m W ~30 MeV l Tracker Calibration: Use J/ and (1S) : m W ~25 MeV l Electron Energy Calibration: Energy scale set using using E/p peak from W electrons m W ~35 MeV Non-linear calorimeter response: m W ~25 MeV E/p tail used for tuning passive material upstream (silicon) m W ~55 MeV l Hadronic Recoil Measurement (impacts neutrino measurement): Recoil u measured by summing over all calorimeter towers Lepton energy is removed: m W ~10 MeV Hadronic Recoil Model: Parametrize R = u meas /u true Just u true contributes m W ~20 MeV Underlying event: m W ~40 MeV Jet energy resolution: m W ~20 MeV