Higgs studies. P. Nieżurawski, A. F. Żarnecki, M. Krawczyk. Faculty of Physics Warsaw University. Higgs at the Photon Collider. p.

Transcription

1 Higgs studies at the TESLA Photon Collider P. Nieżurawski, A. F. Żarnecki, M. Krawczyk Faculty of Physics Warsaw University p.1/15

2 m h 12 GeV Process: γ + γ h b + b J z = γ h b γ b,w,t,... b p.2/15

3 m h 12 GeV Process: γ + γ h b + b J z = γ h b γ b,w,t,... b Hard background: γ + γ b + b γ b γ b p.2/15

4 m h 12 GeV Process: γ + γ h b + b J z = γ h b γ b,w,t,... b Hard background: γ + γ b + b γ + γ c + c γ b γ c σ Q 4 q σ LO ( J z = 2) σ LO (J z = ) γ b γ c p.2/15

5 m h 12 GeV Process: γ + γ h b + b J z = γ h b γ b,w,t,... b Hard background: γ + γ b + b γ + γ c + c γ b γ c σ Q 4 q σ LO ( J z = 2) σ LO (J z = ) γ b γ c Other background: Resolved photon(s) interactions γ + γ X + Q + Q p.2/15

6 m h 12 GeV Process: γ + γ h b + b J z = γ h b γ b,w,t,... b Hard background: γ + γ b + b γ + γ c + c γ b γ c σ Q 4 q σ LO ( J z = 2) σ LO (J z = ) γ b γ c Other background: Resolved photon(s) interactions γ + γ X + Q + Q Overlaying events (high intensity of photon-beams in the low-energy part of the spectrum) p.2/15

7 γ + γ F + F LO cross section for massless fermions σ(j z = 2) α2 s σ(j z = ) = p.3/15

8 γ + γ F + F LO cross section for massless fermions σ(j z = 2) α2 s σ(j z = ) = LO cross section for massive fermions S µ F = P µ F + O( m F E F ) = σ(j z = 2) α2 s σ(j z = ) m2 F s α 2 s p.3/15

9 γ + γ F + F NLO cross section for massless fermions = σ α2 α s s dσ de g (J z = 2) 1 E g σ(j z = ) E 3 g NLO cross section for massive fermions p.3/15

10 γ + γ F + F NLO cross section for massless fermions = σ α2 α s s dσ de g (J z = 2) σ(j z = ) E 3 g 1 E g NLO cross section for massive fermions NLO / LO bb (g) J z = N jets = bb (g) J z = N jets = bb (g) J z =2 N jets = bb (g) J z =2 N jets =3 W rec (GeV) p.3/15

11 Generation & Simulation. Selection. Photon-photon spectrum: CompAZ e e beams with s ee = 21 GeV Luminosity γγ (fb -1 /1GeV) J z =, 2 J z = J z = W γγ min = 8. GeV W max1 =131.2 GeV W max2 =161.5 GeV W γγ (GeV) p.4/15

12 Generation & Simulation. Selection. Photon-photon spectrum: CompAZ Signal: HDECAY, PYTHIA Background: program by G. Jikia Fragmentation: Lund in PYTHIA p.4/15

13 Generation & Simulation. Selection. Photon-photon spectrum: CompAZ Signal: HDECAY, PYTHIA Background: program by G. Jikia Fragmentation: Lund in PYTHIA Detector performance: SIMDET (parametric simulation) Jets: Durham algorithm with y cut =.2 p.4/15

14 Generation & Simulation. Selection. Photon-photon spectrum: CompAZ Signal: HDECAY, PYTHIA Background: program by G. Jikia Fragmentation: Lund in PYTHIA Detector performance: SIMDET (parametric simulation) Jets: Durham algorithm with y cut =.2 Selection of b b events: p.4/15

15 Generation & Simulation. Selection. Photon-photon spectrum: CompAZ Signal: HDECAY, PYTHIA Background: program by G. Jikia Fragmentation: Lund in PYTHIA Detector performance: SIMDET (parametric simulation) Jets: Durham algorithm with y cut =.2 Selection of b b events: 1) Assumed bb-tagging and mistagging efficiencies: ε bb = 7%, ε cc = 3.5% 2) Using ZVTOP-B-Hadron-Tagger p.4/15

16 Generation & Simulation. Selection. Photon-photon spectrum: CompAZ Signal: HDECAY, PYTHIA Background: program by G. Jikia Fragmentation: Lund in PYTHIA Detector performance: SIMDET (parametric simulation) Jets: Durham algorithm with y cut =.2 Selection of b b events: 1) Assumed bb-tagging and mistagging efficiencies: ε bb = 7%, ε cc = 3.5% 2) Using ZVTOP-B-Hadron-Tagger E vis > 9 GeV p.4/15

17 Generation & Simulation. Selection. Photon-photon spectrum: CompAZ Signal: HDECAY, PYTHIA Background: program by G. Jikia Fragmentation: Lund in PYTHIA Detector performance: SIMDET (parametric simulation) Jets: Durham algorithm with y cut =.2 Selection of b b events: 1) Assumed bb-tagging and mistagging efficiencies: ε bb = 7%, ε cc = 3.5% 2) Using ZVTOP-B-Hadron-Tagger E vis > 9 GeV N jets = 2, 3 p.4/15

18 Generation & Simulation. Selection. Photon-photon spectrum: CompAZ Signal: HDECAY, PYTHIA Background: program by G. Jikia Fragmentation: Lund in PYTHIA Detector performance: SIMDET (parametric simulation) Jets: Durham algorithm with y cut =.2 e e beams with s ee = 21 GeV Selection of b b events: 1) Assumed bb-tagging and mistagging 2) Using ZVTOP-B-Hadron-Tagger Number of events 1 8 m h =12 GeV L γγ (W γγ >8GeV)= 84 fb -1 Higgs signal NLO Background: bb (g) J z = bb (g) J z =2 cc (g) J z = cc (g) J z =2 E vis > 9 GeV N jets = 2, 3 6 For comparison: LO Background P z /E vis < P z /E vis p.4/15

19 Generation & Simulation. Selection. Photon-photon spectrum: CompAZ Signal: HDECAY, PYTHIA Background: program by G. Jikia Fragmentation: Lund in PYTHIA Detector performance: SIMDET (parametric simulation) Jets: Durham algorithm with y cut =.2 Selection of b b events: 1) Assumed bb-tagging and mistagging efficiencies: ε bb = 7%, ε cc = 3.5% 2) Using ZVTOP-B-Hadron-Tagger E vis > 9 GeV N jets = 2, 3 P z /E vis <.1 cos θ i <.75 for each jet p.4/15

20 B-tagging ZVTOP-B-Hadron-Tagger b-tag(jet 2 ) jet events γγ bb(g) b-tag(jet 1 ) Number of γ + γ b + b events per 1 year of collider running p.5/15

21 B-tagging ZVTOP-B-Hadron-Tagger b-tag(jet 2 ) jet events γγ cc(g) b-tag(jet 1 ) Number of γ + γ c + c events per 1 year of collider running p.5/15

22 B-tagging ZVTOP-B-Hadron-Tagger b-tag(jet 2 ) 1.8 Ratio bb/cc (2-jet events) b-tag(jet 1 ) 1-3 S B = #(γγ b b) #(γγ c c) p.5/15

23 B-tagging ZVTOP-B-Hadron-Tagger b-tag(jet 2 ) 1.8 Ratio bb/cc (2-jet events) b-tag(jet 1 ) jet events: ε bb = 81% ε cc = 1.8% 3-jet events: ε bb = 77% ε cc = 1.3% Earlier assumed: ε bb = 7% ε cc = 3.5% p.5/15

24 Results Γ(h γγ)br(h bb) = Γ(h γγ)br(h bb) Nobs N obs N bkgd Consecutive approaches: LO cross section for γ + γ Q Q. (1.7%) Number of events/2gev m h =12 GeV e e beams with s ee = 21 GeV L γγ (W γγ >W γγ min )= 84 fb -1 W γγ min = 8 GeV Background: bb Higgs signal cc Reconstructed invariant mass (GeV) p.6/15

25 Results Γ(h γγ)br(h bb) = Γ(h γγ)br(h bb) Nobs N obs N bkgd Consecutive approaches: LO cross section for γ + γ Q Q. (1.7%) NLO cross section for γ + γ Q Q(g). (1.9%) Number of events/2gev e e beams with s ee = 21 GeV m h =12 GeV Higgs signal L NLO Background: γγ (W γγ >8GeV)= 84 fb -1 bb (g) J z = bb (g) J z =2 cc (g) J z = cc (g) J z = For comparison: LO Background W rec (GeV) p.6/15

26 Results Γ(h γγ)br(h bb) = Γ(h γγ)br(h bb) Nobs N obs N bkgd Consecutive approaches: LO cross section for γ + γ Q Q. (1.7%) NLO cross section for γ + γ Q Q(g). (1.9%) NLO cross section for γ + γ Q Q(g). B-tagging algorithm. (1.8%) Number of events/2gev m h =12 GeV L γγ (W γγ >8GeV)= 84 fb -1 e e beams with s ee = 21 GeV Higgs signal NLO Background: bb (g) J z = bb (g) J z =2 cc (g) J z = cc (g) J z =2 NZK W rec (GeV) p.6/15

27 Missing P T Neutrinos from semileptonic decays of D- and B-mesons. Events /.5GeV Events=6719 µ=119.3 GeV σ= 2.3 GeV e e beams with s ee = 21 GeV E νs (GeV)= > Events=4572 µ=119.4 GeV σ= 2.2 GeV P T /E T < Events=2683 µ=119.5 GeV σ= 1.6 GeV P T /E T < W rec (GeV) p.7/15

28 Missing P T Neutrinos from semileptonic decays of D- and B-mesons. ν ν p.7/15

29 Missing P T Neutrinos from semileptonic decays of D- and B-mesons. ν P L ν p.7/15

30 Missing P T Neutrinos from semileptonic decays of D- and B-mesons. ν ν eff P T miss P L P L miss ν W corr W 2 rec + 2P T (E vis + P T ) p.7/15

31 Final results (W rec W corr ) Γ(h γγ)br(h bb) = Γ(h γγ)br(h bb) Nobs N obs N bkgd NLO cross section for γ + γ Q Q(g). (1.9%) - with W rec Number of events/2gev e e beams with s ee = 21 GeV m h =12 GeV Higgs signal L NLO Background: γγ (W γγ >8GeV)= 84 fb -1 bb (g) J z = bb (g) J z =2 cc (g) J z = cc (g) J z = For comparison: LO Background W rec (GeV) p.8/15

32 Final results (W rec W corr ) Γ(h γγ)br(h bb) = Γ(h γγ)br(h bb) Nobs N obs N bkgd NLO cross section for γ + γ Q Q(g). (1.9%) - with W rec (1.7%) - with W corr Number of events/2gev 25 2 e e beams with s ee = 21 GeV m h =12 GeV L γγ (W γγ >8GeV)= 84 fb -1 Higgs signal NLO Background: bb (g) J z = bb (g) J z =2 cc (g) J z = cc (g) J z = For comparison: LO Background W corr (GeV) p.8/15

33 Final results (W rec W corr ) Γ(h γγ)br(h bb) = Γ(h γγ)br(h bb) Nobs N obs N bkgd NLO cross section for γ + γ Q Q(g). (1.9%) - with W rec (1.7%) - with W corr NLO cross section for γ + γ Q Q(g). B-tagging algorithm. (1.8%) - with W rec Number of events/2gev m h =12 GeV L γγ (W γγ >8GeV)= 84 fb -1 e e beams with s ee = 21 GeV Higgs signal NLO Background: bb (g) J z = bb (g) J z =2 cc (g) J z = cc (g) J z =2 NZK W rec (GeV) p.8/15

34 Final results (W rec W corr ) Γ(h γγ)br(h bb) = Γ(h γγ)br(h bb) Nobs N obs N bkgd NLO cross section for γ + γ Q Q(g). (1.9%) - with W rec (1.7%) - with W corr NLO cross section for γ + γ Q Q(g). B-tagging algorithm. (1.8%) - with W rec (1.6%) - with W corr Number of events/2gev e e beams with s ee = 21 GeV m h =12 GeV L γγ (W γγ >8GeV)= 84 fb -1 NZK. Higgs signal NLO Background: bb (g) J z = bb (g) J z =2 cc (g) J z = cc (g) J z = W corr (GeV) p.8/15

35 Final results (W rec W corr ) Γ(h γγ)br(h bb) = Γ(h γγ)br(h bb) Nobs N obs N bkgd NLO cross section for γ + γ Q Q(g). (1.9%) - with W rec (1.7%) - with W corr NLO cross section for γ + γ Q Q(g). B-tagging algorithm. (1.8%) - with W rec (1.6%) - with W corr Number of events/2gev e e beams with s ee = 21 GeV m h =12 GeV L γγ (W γγ >8GeV)= 84 fb -1 NZK. Higgs signal NLO Background: bb (g) J z = bb (g) J z =2 cc (g) J z = cc (g) J z =2 Plans: m h up to 16 GeV, W corr (GeV) p.8/15

36 γγ W + W W + W 4 jets event selection: balanced transverse momentum: P T /E T <.1 #events 1 4 All WW events WW qqqq WW qqlν WW llνν p T /E T p.9/15

37 γγ W + W W + W 4 jets event selection: balanced transverse momentum: P T /E T <.1 4 hadronic jets reconstructed (Durham algorithm) #events All WW events WW qqqq WW qqlν WW llνν p T /E T p.9/15

38 γγ W + W W + W 4 jets event selection: balanced transverse momentum: P T /E T <.1 4 hadronic jets reconstructed (Durham algorithm) σ M /M [%] cos Θ =.95 cut on jet angle cos θ jet <.95 to preserve good mass resolution min Θ jet p.9/15

39 γγ W + W W + W 4 jets event selection: balanced transverse momentum: P T /E T <.1 4 hadronic jets reconstructed (Durham algorithm) # events 6 WW ZZ P W =.1 cut on jet angle cos θ jet <.95 to preserve good mass resolution 4 two W ± reconstructed with probability P W >.1 2 P W = 2 MW Γ 2 W P W W 1,W 2 m 2 jj M 2 W 2 + M 2 W Γ2 W p.9/15

40 γγ W + W W + W 4 jets event selection: balanced transverse momentum: P T /E T <.1 4 hadronic jets reconstructed (Durham algorithm) # events 6 WW ZZ P W =.1 cut on jet angle cos θ jet <.95 to preserve good mass resolution 4 two W ± reconstructed with probability P W >.1 2 P W = 2 MW Γ 2 W P W W 1,W 2 m 2 jj M 2 W 2 + M 2 W Γ2 W selection efficiency between 2% for and 16% for (W γγ = 2 4 GeV ) probability for both W ± to decay into hadrons is 47% p.9/15

41 γγ W + W W + W 4 jets event selection: balanced transverse momentum: P T /E T <.1 4 hadronic jets reconstructed (Durham algorithm) cut on jet angle cos θ jet <.95 to preserve good mass resolution two W ± reconstructed with probability P W >.1 #events P W = W 1,W 2 m 2 jj M 2 W 2 MW Γ 2 W 2 + M 2 W Γ2 W M 4j [GeV] selection efficiency between 2% for and 16% for (W γγ = 2 4 GeV ) probability for both W ± to decay into hadrons is 47% invariant mass resolution: Γ GeV (Breit-Wigner like) p.9/15

42 γγ ZZ ZZ lljj selection (l = e, µ): balanced transverse momentum: P T /E T <.1 2 leptons (e ± or µ ± ) + 2 hadronic jets reconstructed too large background in 4-jet channel cut on lepton and jet angle cos θ jet <.95 leptons and jets reconstruct into two Z with probability P Z >.1 #events M lljj [GeV] selection efficiency about 5% (BR(ZZ q q l + l ) 9.4%) invariant mass resolution: Γ GeV (Breit-Wigner like) p.1/15

43 Parametrization Invariant mass resolution for selected W + W and ZZ events is parametrized as a function of W γγ. p.11/15

44 Parametrization Invariant mass resolution for selected W + W and ZZ events is parametrized as a function of W γγ. Measured invariant mass distribution can be then described by convolution of: p.11/15

45 Parametrization Invariant mass resolution for selected W + W and ZZ events is parametrized as a function of W γγ. Measured invariant mass distribution can be then described by convolution of: Analytical luminosity Spectra CompAZ p.11/15

46 Parametrization Invariant mass resolution for selected W + W and ZZ events is parametrized as a function of W γγ. Measured invariant mass distribution can be then described by convolution of: Analytical luminosity Spectra CompAZ Cross section formula for signal + background + interf. p.11/15

47 Parametrization Invariant mass resolution for selected W + W and ZZ events is parametrized as a function of W γγ. Measured invariant mass distribution can be then described by convolution of: Analytical luminosity Spectra CompAZ Cross section formula for signal + background + interf. Invariant mass resolution p.11/15

48 Parametrization Invariant mass resolution for selected W + W and ZZ events is parametrized as a function of W γγ. Measured invariant mass distribution can be then described by convolution of: Analytical luminosity Spectra CompAZ Cross section formula for signal + background + interf. Invariant mass resolution mass spectra can be calculated for any s ee and M h without MC simulation p.11/15

49 Parametrization Invariant mass resolution for selected W + W and ZZ events is parametrized as a function of W γγ. Measured invariant mass distribution can be then described by convolution of: Analytical luminosity Spectra CompAZ Cross section formula for signal + background + interf. Invariant mass resolution mass spectra can be calculated for any s ee and M h without MC simulation # events 1 NZK. simulation m h =18 GeV Parameterization: # events 3 simulation m h =3 GeV Parameterization: NZK. 75 m h =18 GeV no Higgs 2 m h =3 GeV no Higgs M 4j [GeV] M llqq [GeV] p.11/15

50 Γ γγ measurement Average statistical precision expected after 1 year of PC running One parameter fit to invariant mass distribution for W + W and ZZ events Γ γγ only Γ γγ /Γ γγ.1 35 GeV NZK M h [GeV] assuming SM branching ratios p.12/15

51 Γ γγ measurement Average statistical precision expected after 1 year of PC running One parameter fit to invariant mass distribution for W + W and ZZ events Γ γγ only Γ γγ /Γ γγ.1 35 GeV 362 GeV NZK M h [GeV] assuming SM branching ratios p.12/15

52 Γ γγ measurement Average statistical precision expected after 1 year of PC running One parameter fit to invariant mass distribution for W + W and ZZ events Γ γγ only Γ γγ /Γ γγ.1 35 GeV 362 GeV 418 GeV NZK M h [GeV] assuming SM branching ratios p.12/15

53 Γ γγ measurement Average statistical precision expected after 1 year of PC running One parameter fit to invariant mass distribution for W + W and ZZ events Γ γγ only Γ γγ /Γ γγ.1 35 GeV 362 GeV 418 GeV 5 GeV NZK M h [GeV] assuming SM branching ratios p.12/15

54 Γ γγ measurement Average statistical precision expected after 1 year of PC running One parameter fit to invariant mass distribution for W + W and ZZ events Γ γγ only Γ γγ /Γ γγ.1 35 GeV 362 GeV 418 GeV 5 GeV NZK M h [GeV] assuming SM branching ratios p.12/15

55 Γ γγ measurement Average statistical precision expected after 1 year of PC running One parameter fit to invariant mass distribution for W + W and ZZ events Γ γγ only Γ γγ /Γ γγ.1 35 GeV 362 GeV 418 GeV 5 GeV 2HDM NZK. Sensitive to possible new physics only up to M h 28 GeV M h [GeV] assuming SM branching ratios new physics modeled by SM-like 2HDM (II) with M H + = 8 GeV p.12/15

56 Γ γγ measurement Average statistical precision expected after 1 year of PC running One parameter fit to invariant mass distribution for W + W and ZZ events Γ γγ only Γ γγ /Γ γγ.1 35 GeV 362 GeV 418 GeV 5 GeV 2HDM NZK. Sensitive to possible new physics only up to M h 28 GeV.5 For higher Higgs masses Γ γγ is little sensitive to contribution of new heavy charged particles! M h [GeV] assuming SM branching ratios new physics modeled by SM-like 2HDM (II) with M H + = 8 GeV p.12/15

57 Γ γγ measurement Average statistical precision expected after 1 year of PC running Two parameter fit to invariant mass distribution for W + W and ZZ events Γ γγ and φ γγ Γ γγ /Γ γγ.1 35 GeV 362 GeV 418 GeV 5 GeV 2HDM NZK. Sensitive to possible new physics only up to M h 28 GeV.5 For higher Higgs masses Γ γγ is little sensitive to contribution of new heavy charged particles! M h [GeV] φ γγ fit increases the error only slightly assuming SM branching ratios new physics modeled by SM-like 2HDM (II) with M H + = 8 GeV p.12/15

58 φ γγ measurement Average statistical precision expected after 1 year of PC running Two parameter fit to invariant mass distribution for W + W and ZZ events Γ γγ and φ γγ φ γγ [rad].1 35 GeV 362 GeV 418 GeV 5 GeV 2HDM NZK. Phase measurement significantly improves our sensitivity to new heavy charged particles at large Higgs boson masses M h [GeV] assuming SM branching ratios Example: heavy charged Higgs boson of the SM-like 2HDM(II) with M H + = 8 GeV p.13/15

59 Γ γγ and φ γγ measurement Two parameter fit to W + W and ZZ invariant mass distribution; 1 PC year. Expected statistical error contours (1σ) in φ γγ - Γ γγ, for M h = 3 GeV: 4 th generation lepton M L = 8 GeV φ γγ -φ SM.4.2 H + (2HDM) D (Q= - 1 / 3 ) U (Q= + 2 / 3 ) L (Q= -1) NZK. 1σ stat M h = 3 GeV SM-like 2HDM (II) M H + = 8 GeV SM Γ γγ /Γ SM separation not possible without phase measurement! p.14/15

60 Conclusions Comparison of Γ γγ results from different analyses Our plans: h b b up to 16 GeV H b b in MSSM Γ γγ /Γ γγ.1 h bb [JS-R] [NZK] 21 GeV h WW, ZZ [NZK] 35 GeV 362 GeV 418 GeV 5 GeV NZK. CP of h in h ZZ M h [GeV] p.15/15