Hunt down Higgs - PowerPoint PPT Presentation

search for higgs n.
Skip this Video
Loading SlideShow in 5 Seconds..
Hunt down Higgs PowerPoint Presentation
Hunt down Higgs

play fullscreen
1 / 61
Download Presentation
eliana-mckee
Views
Download Presentation

Hunt down Higgs

Presentation Transcript

  1. Search for Higgs Amitabh Lath Rutgers The State University of NJ Rutgers University September 7 2005

  2. What is the Higgs? • Short answer: Field/Particle responsible for mass. • Longer answer: Completes the Standard Model of Particles and Fields. • So what is this Standard Model of P & F? Rutgers University September 7,2005

  3. Our Periodic Table Rutgers University September 7,2005

  4. Our Periodic Table Quark stuff: hadrons. 3q  baryon q-anti(q)  meson Q=2/3 -1/3 0 1 + Higgs! (not yet found) Rutgers University September 7,2005

  5. Experimentalist View: Prologue Einstein, photoelectric effect J.J. Thomson, 1897 Street and Stevenson, 1937 cloud chamber Rutgers University September 7,2005

  6. Experimentalist View: Weak Effects Reines and Cowan Reactor (beta decay) Lederman, Schwartz and Steinberger group using AGS pion beam Rutgers University September 7,2005

  7. Experimentalist View: Quarks Kendall, Friedman, Taylor group at SLAC. Gell-Mann’s quarks Feynman’s “partons” Rutgers University September 7,2005

  8. Experimentalist View: Quarks Gell-Mann predicts Omega- particle. Found by bubble chamber group at BNL (Samios et al) Rutgers University September 7,2005

  9. Experimentalist View: Quarks The strange particle decay, KLmm not as prolific as predicted • Glashow, Iliopoulos, and • Maiani posit a “new” • particle. (GIM Mechanism) • Diagrams with “charm” destructively interfere. Found by Ting, Richter groups (almost simultaneously) at BNL, SLAC. Rutgers University September 7,2005

  10. Experimentalist View: 3rd Generation Note: Sq for each generation = 0 Fermilab collider program (cast of hundreds) Fermilab fixed-target experiment (Lederman) Beautiful analysis by Martin Perl at SLAC’s SPEAR Rutgers University September 7,2005

  11. Experimentalist View: Weak Bosons • Glashow, Weinberg and • Salaam introduce two • primordial fields: • Weak Isospin • Weak Hypercharge • that mix (weak mixing angle) • to give W,Z and g + Higgs! They also introduce a scalar field (Higgs). W, Z discovered by UA2 Rutgers University September 7,2005

  12. Electro-Weak Unification • Weak Isospin (3-component) • W1 + iW2 = W+ • W1 – iW2 = W- • Weak Hypercharge (1-component) g cosqw -sinqw W3 Z sinqw cosqw B0 Left Handed Int. only! = Weak Mixing Angle Rutgers University September 7,2005

  13. Higgs and ElectroWeak Unification • Adding scalar field to Lagrangian gives mass-like terms, leading to  • Masses of bosons determined by couplings, and parameters of the scalar field potential (v,l). • Mhiggs is a free parameter. All three things in this equation can be measured by experiments. MW/MZ = cosqW Rutgers University September 7,2005

  14. What is Higgs? • Scalar Field. Introduced in: A MODEL OF LEPTONS Steven Weinberg, Phys.Rev.Lett.19:12641266,1967 . • Gives mass to everything, but Higgs boson not seen. • Fermion masses? Need (arbitrary) couplings. Called Gi in Halzen & Martin. MH < 246 GeV/c2 (fit of all EW data and top mass) Rutgers University September 7,2005

  15. Different Types of Higgs • SM Higgs • One (complex) doublet. • Cross section for direct production at Tevatron depressingly low. • Theoretical problems. • Beyond SM Higgs • Different models fix (some) SM problems. • Tevatron can compete in some parts of parameter space. Rutgers University September 7,2005

  16. Problems with Higgs • Fundamental Scalar Field has problems. • What are the corrections to its mass? • Two basic types of solutions: • Higgs is a composite (Technicolor) • Other diagrams cancel problems. (SUSY) Rutgers University September 7,2005

  17. Beyond SM Higgs (Technicolor) • “pion” like composite Higgs. • Techniquarks make technipions • QCD-like strong force (ergo “technicolor”). • Ruled out! • The electron-positron asymmetries at the Z pole were not kind to this theory. • However, a lot of physicists have a soft spot for compositeness. T = isospin breaking terms Rutgers University September 7,2005 S = Isospin cons. terms

  18. Beyond SM Higgs (Supersymmetry) • The “GIM Mechanism” solution to the Higgs problem. • Got a problem diagram? Add another to cancel it out. • Fermion sFermion • Boson  Bosino • A very expensive theory. • Doubles the particle spectrum. • Consolation prizes: LSPDark Matter? Unification of couplings? Rutgers University September 7,2005

  19. The MSSM Higgs • Minimal SuperSymmetric Standard Model. • Simplest realistic SUSY theory. • 3 neutral(h,A,H), 2 charged Higgs Bosons (masses related). • tanb ratio of up/down type couplings. • The LEP experiment (e+e- collider) covered a large swath of parameter space by ruling out higgs up to 90 GeV. Rutgers University September 7,2005

  20. MSSM Higgs Production • At large tanb, both gg and bb production contribute. • s rises like tanb2. • A, and h/H are produced simultaneously. • ~100 pb! • For inclusive mode, ttdecay is very interesting. Double this! ~SM Higgs Rutgers University September 7,2005

  21. Higgs Decays • Higgs decays to bb 90% and tt 10% (Higgs couples to mass) • higgs bb? No hope for inclusive. • higgs tt ? Yes! • Decays to WW, ZZ take over if Higgs is heavy enough. Rutgers University September 7,2005

  22. Higgs Decays • Most Higgs searches look for b-quarks. • SM, SUSY higgses (ask SVS). • Problem: “real” b-quarks from hadronic interaction swamp those from Higgs. • About 6 orders of magnitude! • Usual solution: Look at “associated” stuff:(Higgs+W/Z, Higgs+extra b-quarks) • Our solution: Forget b. Look for tt. • Give up (some) sensitivity to SM Higgs. • Gain a clean channel for MSSM Higgs. Rutgers University September 7,2005

  23. CDF Tevatron Main Injector The Tevatron Accelerator p-pbar collisions at 1.96 TeV C.M. Energy Rutgers University September 7,2005

  24. CDF Collaboration Missing Conway… Rutgers University September 7,2005

  25. The CDF Detector beampipe Collision point Superconducting Magnet Rutgers University September 7,2005

  26. Particle Detection • Electron track, contained cluster, E/P~1 g, no track • Quark matter (p) - track, extended (hadron) cluster • Muon penetrating track • Weak, no charge (n) Missing momenta Rutgers University September 7,2005

  27. Particle Detection s • Electron track, contained cluster, E/P~1 g, no track • Quark matter (p) - track, extended (hadron) cluster • Muon penetrating track • Weak, no charge (n) Missing momenta ~nb ~mb ~nb HIGGS? ~pb Rutgers University September 7,2005

  28. The tau lepton. • 3rd generation lepton. • Heavy. Mass = 1.77 GeV/c2 • Lots of decays • Leptonic: t enn, t  mnn(17% each) • Hadronic: t  pn, pp0n, pppn, pp0p0n(64%) • Shorthand: te, tm, th Rutgers University September 7,2005

  29. Looking for Higgs t t n H n n p, quark matter e or m Signal: • One t goes to electron or muon. • We know how to detect e, m • Second t goes to pions Backgrounds? • Ztt (irreducible) • We, m + “jets” (quark stuff including p) Rutgers University September 7,2005

  30. Detecting taus? • Taus are leptons that can decay to “jets” of pions. • jets from q,g are fatter. • Signal cone. • Isolation cone (annulus). • Up to 30o. • Veto tracks, p0… • Should get: • Characteristic 1,3 track enhancement. • |Q| = 1 • m < 1.8 te ,tm Rutgers University September 7,2005 Standard cuts

  31. Problem with taus. • Fakes from jets(as with almost everything at hadron machines). • th IS a jet; albeit a narrow one (pencil jet). • Jet production 100’s mb. Compare to 100’s of pb for t production. • Don’t know enough about JETS  need fake rate from data. This is where the problems start. • Different sources of jets give very different fake estimates. Rutgers University September 7,2005

  32. Fake Rates. Prologue. • CDF Run 1(U.C. Berkeley) • Jet triggered samples give lower fake rates than lepton triggered ones. • Fake rates rising with jet ET. • What is going on? Huge difference! Rutgers University September 7,2005

  33. Solution: Multi-Parameter. • Instead of parameterizing fake rate in one variable, we now use several. • Big discrepancy in rates disappears. • Is this the right thing to do? • Why bother? “Relative” fake rates Rutgers University September 7,2005

  34. Check of Fake Rate Predictor • Newfangled multi-parameter fake rate function is scary. • Jet variables (energy, direction, etc)  BLACKBOX  probability of a fake t. • Let’s test it on different samples of jets. • Jet triggered sample (assume one jet is “lepton”) • Photon+jet (use photon as lepton) • W+jet. • MC Jets. • Let’s look at different distributions, not just absolute numbers. Rutgers University September 7,2005

  35. Pure sample of jets Predictor Fake rate prediction Checking Jett Fake Rates Pure sample of jets t reconst Fake rate measured Rutgers University September 7,2005

  36. Checking Fake Rates (QCD) This is a sanity check at best, since we built our “black box” -- fake predictor-- using Jet Triggered data. Rutgers University September 7,2005

  37. Checking Fake Rates (g +jet) This is the most heartening set of plots. Rutgers University September 7,2005

  38. Checking Fake Rate (W+jet) Rutgers University September 7,2005

  39. tt signature Classic Hadronic tau signature • Bothte th andtmth are included in these plots • Ntracks = 1,3 and |Qt| = 1 not applied to the • left plot to demonstrate tau signature Rutgers University September 7,2005

  40. Expected vs. Observed Events source thtethtµall Z → τ τ 215.8±3.9 186.5±3.4 402.4±5.3 Z → ll 6.5±0.5 8.3±0.8 14.8±1.0 tt, VV 1.0±0.1 0.9±0.1 1.9±0.1 jet → τ 44.6±0.2 30.0±0.4 73.2±0.5 total bg 268.9±3.9 225.7±3.7 492.3±5.4 Observed 260 229 489 Errors are stat. only No big whopping discrepancy… Rutgers University September 7,2005

  41. Higgs-Z Separation. Define p(ET) as (EX,EY, 0 ,ET) Define mvis = m(p(t1) + p(t2) + p(ET)) Use these shapes to fit the data,  Is there a “Higgs bump”? Total acceptance ~1.5% (from simulations). Rutgers University September 7,2005

  42. Mvis Plot • We use this “mass-like” variable to get Higgs/Z separation. • Nice fit! • Note: Z does not peak at MZ! Rutgers University September 7,2005

  43. More on Higgs Fit • Fix Z, other bg • by known s and • luminosity. • Fit Att shapes • from simulation • We don’t see a • higgs signal. ? Rutgers University September 7,2005

  44. Events in high mass tail Tau (3-prong) These look like good Z/g* tt events (Or higgs signal?) More data coming soon! Electron Rutgers University September 7,2005

  45. Expected and Observed Limits Rutgers University September 7,2005

  46. Expected and Observed Limits fluctuated high around m=130 GeV fluctuated low near MZ Rutgers University September 7,2005

  47. Compare to Run 1 Limits m=120 (914 exp, 700 obs)* (BR to tt) = 82.3 exp, 63 obs (836 exp, 643 obs)* (BR to tt ) = 75.2 exp, 57.9 obs m=140 We’re a factor of 8 better, with approx three times the luminosity. • Get the low pT leptons (trig) • Reconstruct p0. • Understand bg (esp W+jet) • Fit, don’t count. Rutgers University September 7,2005

  48. Extracting MSSM Parameters • We need production cross-section for inclusive A/H/h production. • We only require two taus, no extra jets, b-quarks, leptons, neutrinos…nothing. • Most of the recent work has concentrated on Higgs production associated with “spectator” b-quarks. • Why? Nobody thought tt could compete. • We need bb  A and ggA. Rutgers University September 7,2005

  49. Getting the Cross Section • Not trivial! Consensus emerging slowly among theorists. • These results have spurred some activity. • The method (as of June 2005): • Get bbhiggs and gghiggs for SM. • Multiply by a MSSM/SM factor. • This ensures corrections apply properly. This is for tanb=30 Rutgers University September 7,2005

  50. The Exclusion Plot • We did not see a higgs signal, but we can put limits on mass/coupling (tanb). • The exclusion depends in “scenarios” chosen by theorists  points in the SUSY parameter space that are considered good benchmarks. Rutgers University September 7,2005