CP Violation and Rare B Decays at Belle

Transcription

1 CP Violation and Rare B Decays at Belle DESY Seminar 11 Sep, 2001 M. Hazumi (Osaka University, Osaka, Japan) Representing the Belle Collaboration

2 Outline Introduction KEKB Accelerator and Belle Detector Observation of Large CP Violation Rare B Decays Today s Main Topic Conclusion

3 Introduction KEKB Accelerator and Belle Detector Observation of Large CP Violation Rare B Decays Conclusions

4 The Belle Collaboration A World-Wide Activity Involving 50 Institutions KEK aphpehuv room for additional 300 physicists available J

5 Closer Look on 9. Sep, 2001 Osaka KEK Typhoons!

6 Much closer look at KEK KEK KEKB Collider ~1km

7 The Belle Detector

8 d s b = Kobayashi-Maskawa and CPV &39GXHWRFRPSOH[SKDVHVLQ& 0PDWUL[ V V V ud cd td V V V us cs ts V V V cb tb ub d s b *RDOVRI%IDFWRU\ V ud V * ub φ 3 φ 2 V td V * tb V cd V * cb 1) Discover / Establish CPV 2) Precise/Redundant Measurement of CKM elements : angles and lengths 3) Beyond SM φ 1 CKM matrix Unitarity triangle

9 The Rules of the CPV Game Find a decay mode which has two decay paths with different weak phases (i.e. one of them has Vub or Vtd). Two amplitudes should be similar for sizable interference. Two paths should have static phase difference. Example : B + K + π 0 and B K π 0 Γ(B + K + π 0 ) - Γ(B K π 0 ) = 4 T P sin(φ 3 )sin(s -s) Large Uncertainty!

10 Time-dependent CP Violation 6DQGD%LJL &DUWHU B 0 V* td V tb V cb B 0 B 0 V tb + V* td J/ψ K S J/ψ K S V* 2 td m B is the origin of static phase difference. No hadronic uncertainty!

11 Large BB Mixing! ARGUS (1987) (103pb -1 on resonance) r = (N(BB)+N(BB)) /N(BB) = 0.21 ± 0.08 m B0 = (0.472 ± 0.017) ps -1 (PDG2000)

12 Appropriate BB Mixing! sin mt = 1 Maximal CP at 3.3 ps B0 Lifetime = 1.55ps Large enough and small enough (It is a miracle to me.)

13 Principle of the Measurement t from z e e + t=0 Flavor-tag decay (B 0 or B 0?) z J/ψ K S f CP B - B B + B PRUH% V PRUH% V t z/c βγ = (1-2w)sin2φ 1 = (1-2w)sin2φ 1

14 Experimental Challenges 1) Copious B pair production, efficient B reconstruction 2) Efficient and correct flavor tagging 3) Observation of time-dependent CP asymmetry in B decays to a CP eigenstate with good vertex resolution e- e+ Υ(4s) prod.pt. Z z 1 Z 2 B(cp) decays B(tag) decays z c t (~200µm at Belle) +Z decay rate proper time difference t (ps)

15 Introduction KEKB Accelerator and Belle Detector Observation of Large CP Violation Rare B Decays Conclusions

16 KEKB asymmetric e + e collider Low Energy Ring (LER) for Positron WIGGLER RF RF NIKKO Area HER TSUKUBA Area (Belle) HER LER Interaction Region (TRISTAN Accumulation Ring) Electron RF Positron RF FUJI Area e + /e - High Energy Ring (HER) for Electron LER Linac WIGGLER RF RF OHO Area Two separate storage rings e + (LER) : 3.5 GeV e (HER) : 8.0 GeV E CM : GeV at Υ(4S) Luminosity World Record target: cm -2 s -1 achieved:4.5x10 33 cm -2 s -1 ±11 mrad crossing angle Small beam sizes: σ y 3 µm; σ x 100 µm

17 KEKB performance The world record 230pb -1 /day Integrated luminosity/day 32.8fb -1 Total accumulated luminosity ~80fb -1 expected by Summer 2002

18 KEKB s Special Features Small beam sizes low beam currents 4.5x10 33 with less than 1 Amp in each ring ± 11 mrad beam crossing angle e + No strong bending magnets near the IR Fewer spent particles into Belle Synchrotron X-rays easily handled 22mrad e -

19 Belle Detector Performance u u u u u Silicon Vertex Detector (SVD) u Impact parameter resolution : 55µm for p=1gev/c at normal incidence Central Drift Chamber (CDC) u (σ Pt /Pt) 2 = (0.0019Pt) 2 + (0.0034) 2 (Pt in GeV/c) 2 K/π separation with u de/dx in CDC (σ de/dx =6.9%) u TOF (σ TOF = 95ps) u Aerogel Cerenkov (ACC) u Efficiency = ~85%, Fake rate = ~10% up to 3.5GeV/c u u γ, e ± with CsI crystals (ECL) 6m σ E /E ~ E=1GeV e ± : effic. > 90% w/ ~0.3% fake for p > 1GeV/c K L and µ ± with KLM (RPC chambers) u µ ± : effic. > 90% with ~2% fake at p > 1GeV/c

20 Belle SVD (3 layers of DSSDs) S/N = 20~40, Occupancy < 4%

21 Impact parameter resolution Resolution (µm) Z Impact Parameter Resolution (Real Data) 36 42/pβ*sin(θ) 5/2 (µm) - γγ 4π - cosmic ray Belle SVD Resolution (µm) R-Phi Impact Parameter Resolution (Real Data) /pβ*sin(θ) 3/2 (µm) - γγ e + e - - cosmic ray - e + e - µ + µ Z rφ pβ*sin(θ) 5/2 (GeV/c) pβ*sin(θ) 3/2 (GeV/c) Sufficient to measure CP violation

22 Introduction KEKB Accelerator and Belle Detector Observation of Large CP Violation Rare B Decays Conclusions

23 Summary of data samples 3 data sets were used: Set 1: 10.5 fb -1 (June 1999-July 2000) Set 2: 10.6 fb -1 (Oct 2000-May 2001) Set 3: 8.0 fb -1 (May 2001-July 2001) Total: 29.1 fb -1 (corresponds to 31.3 million BBbar) All data samples were analyzed and reconstructed with the same consistent procedure.

24 8VHaDOOORZEDFNJURXQGFF PRGHV B CP J/ψ K S ( π + π & π 0 π 0 ) ψ(2s)( l + l & J/ψπ + π ) K S χ c1 ( J/ψγ) K S η c ( K S K + π, K + K π 0 ) K S J/ψ K L J/ψ K* 0 ( K S π 0 ) (mostly) CP odd B Å CP Eigenstate Decays b CP even d d K 0 c s ξ f = ξ f =

25 *ROGHQPRGH: B 0 J/ψ K s ( π + π ) σ~10mev/c 2 σ~11mev/c 2 J/ψ µ + µ J/ψ e + e σ~4mev/c 2 K s π + π Belle %LJJHVWFRQWULEXWRUWRWKHI FS HYHQWVDPSOH

26 B 0 J/ψ K S ( π + π ) (cont d) 457 events ~3% bkgd Energy difference: E E + E S Beam-constrained mass: 2 r r m E 2) ( p + p bc E J / ψ K CM = ( CM J / ψ KS 2 ) 2

27 B 0 J/ψ Ks( π + π ) Event Display

28 All except J/ψ K L B 0 J/ψ Ks( π + π ) 457 candidates ~12 background (Purity = 97%) Belle B 0 other modes 290 candidates 46 background (Purity = 84%)

29 B 0 J/ψ K L J/ψ l + l + K L 2) Assume BÅJ/ψ K L : compute P KL 3) Remove reconstructed B J/ψ K, J/ψ K*, 4) Cut on a likelihood based on kinematical and shape quantities 5) Plot P* = P J/ψ + P KL B

30 B 0 J/ψ K L signal yield P* B = P J/ψ * + P KL * Belle N sig = 346 events B 0 J/ψ K L 569 total events 346 signal (Purity = 61%) N bkg = 223 evts P B (cms)

31 Flavor-tag the other B meson Use inclusive flavor-specific properties: ŒInclusive Leptons: Œhigh-p l bå c l ν Œintermed-p l + s l + ν ŒInclusive Hadrons: Œhigh-p π + B 0 ÅD (*) π +, D (*) ρ +, etc. Œintermed-p K + K + X, π + π 0 Œlow-p π D 0 π $OVRLQFOXGHFRUUHODWLRQV

32 2-level Multi-dimensional Flavor Tagging π s Λ K Lepton Lookup tables q = +1 B tag q = 1 B tag r=(1-2w) MC 6 regions

33 Determination of wrong tag fraction w l obtain w from B B mixing amplitude: (1-2w l )cos( m d t)

34 Dilution : Measured vs. Expected Uses all events Efficiency > 99% ε effective = 27.0 ± 1.2% Includes correlations Use MC-r as a classifier Use data-w for CP fits 1-2w l r 0&GHWHUPLQHG r 1-2w PHDVXUHGIURPGDWD

35 Vertex Reconstruction For CP-side, use J/ψÅl + l Reject poorly fit events. δz CP 75 µm (rms) For Tag-side use well fit tracks iterate: discard worst track δz tag 140 µm (rms) Require z CP - z tag <2mm ( 10τ B ) σ t 1.5 ps Tails 3%; Effic. 85% HYWVXVHGLQWKH&3ILW

36 Validation: B lifetimes BÅD*lν τ B0 = 1.55 ± 0.02ps τ B+ = 1.64 ± 0.03ps PDG: 1.55 ± 0.03ps PDG: 1.65 ± 0.03ps

37 Summary so far CP eigenstates with high purity. l Purity ~ 90% except for J/ψ K L (~60% for J/ψ K L ) Efficient flavor tagging. l Effective efficiency = 27.0% Efficient vertexing with good resolution 1137 candidate events B lifetime measured precisely (high stat. control sample) Everything is ready. Now let s open the box!

38 Combine q, ξ f & t raw data

39 CP is violated in B decays Large effect Apparent even in the raw data

40 Event-by-event Likelihood )) ( ) ( ) ( ) ( ) ((1 t d t t R t P f t t R t P f L bk bk bk sig sig bk i + = ) ( ) (1 2 / ) ( / t f e f t P bkg t bk bkg + = δ τ τ τ τ )) sin( ) 2 sin( ) 2 (1 (1 2 1 ) ( 1 / t m w q e t P d f t B sig B = φ ξ τ τ Taken from data

41 sin2φ 1 value that maximizes Π i L i sin2φ 1 = 0.99 ± 0.14 (stat) ±0.06(sys) curves from unbinned fit

42 Confidence intervals 1 sin2φ (95% C.L.) Belle central values 1σ 95% Ciuchini et al (indirect constraints) 1σ 95%

43 asymmetry plot: all data Result from global fit (sin2φ 1 =0.99)

44 Control sample: B 0 non-cp states use: B 0 ÅD (*) π +, D * ρ +, D * l + ν, J/ψK*(K + π ) ~17K events sin2φ ±0.04 (statistical error only)

45 Asymmetry Compare CP 1 and CP+1 CP-1 sin2φ ±0.17 CP+1 opposite! 1.31±0.23 (statistical errors only)

46 sin2φ 1 from various subsamples q= evts q= evts (statistical errors only)

47 Test of Direct CP Term If? is allowed to float, (i.e. a cos(?m t) term) λ =1.03 ± 0.09 sin 2φ = 0.99 ± The CPV asymmetry is unchanged.

48 CP fit : Systematic Error Vertex algorithm Flavor tagging Resolution function K L background fraction Background shapes m d and τ B0 errors Total ±0.04 ±0.03 ±0.02 ±0.02 ±0.01 ±0.01 ±0.06

49 Conclusions on CPV lcp is violated in B decay >6σ!! lsin 2φ 1 is large: 0.99 ± 0.14 ± 0.06

50 Introduction KEKB Accelerator and Belle Detector Observation of Large CP Violation Rare B Decays Conclusions

51 Goals of Rare B Decay Program 1) Discover / Establish Direct CPV 2) Precise Measurements of CKM elements : angles and lengths 3) Beyond SM B ππ, ρπ B πlν V ud V * ub φ V td V * tb B D 2 s π f B : B lν (B ργ) B - D CP K - B ππ/kπ B 3 body φ 3 φ 1 V cd V * cb B 0 D (*)- l + ν, dileptons B 0 η Ks, φks B 0 D *+ D (*)- (K) EM/EW Penguin : B K x γ, B Kl + l -

52 Belle 2001 Summer Results (2001 Lepton-Photon Conferences) In addition to the observation of large CPV, 35 contributed papers (5 charm, 3 τ, 2 two-γ physics) [4 final results - sent to journal ] 14 First Observations! [see Web: for the datails of analyses and results ]

53 Importance of Vub Measurement Belle s 1σ band (both sol ns)

54 Towards V ub : B 0 π - l + ν B 0 π - l + ν Important to check large sin2φ 1 update 21.3 fb -1 Missing mass recon. M miss2 = E 2 miss - p 2 miss < 2 GeV p l > 1.2 GeV p l + p π > 3.1 GeV 107 ± 16 signals GeV/c 2 Br = (1.28 ± 0.2 ± 0.26) x10-4 ε(trk/g rec.) GeV/c 2

55 Towards V ub : B 0 Ds π - B 0 V ub Tree dominant (No Penguin) 6.3 ± 3 evens Br < 1.1 x10-4 Ds π (CLEO:< 2.7 x10-4 ) About to see signal! Number of events Ds φπ, K *0 K +, KsK + D ρ D s π 21.3 fb -1 D π E (GeV)

56 Towards φ 3 : B D CP K 1) B - D 0 K (ICHEP2000) [PRL] 2) B D CP K decay D CP K + K, π + π (Br ~ 1%) D CP = (D 0 + D 0 )/ 2 First Observation 21.3 fb -1 Candidates/ 10 MeV B - D o K - 20 (a) D o K - π (b) D o K - K + 2 σ 85 ± 10 ev 12 ± 4 ev b B μu W μu μu s c K D 0 B b μu W V ub Br(B D CP K ) Br(B + D CP K + ) u μu μc s μd 0 K 0 5 (c) D o π - π + 5 ± 5 ev E (GeV) Br(B D CP K ) + Br(B + D CP K + ) = (stat) ± 0.15(syst) 0.35

57 D (*)0 X 0 decays : First Observation Color suppressed modes: ~0.1x Non-suppr. mode W μd μu ß d b μb 0 c + D μd μb 0 b μd W c 0 D - Step towards D 0 K/DKs μu d 0 ß μd Br = 2.9 ±0.4± ±0.6± ±1.3±0.8 (x10-4 ) Events / 20 MeV (a) (e) D Belle (c) D 0 π 0 0 Belle Belle 20 D 0 ω D 0 η 128±18 ev 16 30±9 ev 10 18±7 ev σ Events / 20 MeV Events / 20 MeV σ σ E (GeV) E (GeV) Evidence for: D *0 π 0 (3.2σ), D *0 ω(3.6σ), D *0 η(3.8σ) E (GeV)

58 Bcp events at Belle as of July 2001 For sin2φ1 analysis candidates purity int. lumi. J/ψ Ks( π + π ) 457ev. 97% 29.1fb -1 Other (cc) Ks 290ev. 84% 29.1fb -1 J/ψ K L 569ev. 61% 29.1fb -1 sin2φ1 = 0.99 ±0.14 ±0.06 Rare decays φ ( K + K ) Ks 10ev. 80% 21.3fb -1 η Ks 26ev. 63% 10.4fb -1 D*D ( * ) 27ev. 82% 21.3fb -1 Ks π 0 12ev. 67% 10.4fb -1 Penguin Tree b sss (uus) b suu, sdd uus b ucd, ccd b dds, uus, dds

59 Clear Manifestation of New Physics! Acp(J/ψ Ks) Acp(Bcp with Penguin) b c c No new phase vs. s CP eigenstates (or flavor non-specific states) with any other decay diagrams, such as the following; b s or d q q b s or d γ

60 B0 η Ks (10.4fb -1 ) Origin of the large branching fraction is not understood yet. It may include contribution from new physics!

61 B 0 D *+ D - φ 1 φ 1 measurements with other diagrams Partial Reconstruction Technique : Use only slow π and D *+ π s α D - (full) D 0 Br=1.84 ±0.46±0.68 (x10-3 ) To be used for CP corresponds to ~100 ev ytitle D - (full) ytitle ytitle fb -1 (New) All(no tag) continuum Continuum subtracted a c Lepton tag ± 11 ev 224 ± 87 ev a First Observation cosα cosα -1

62 B 0 D *+ D (*)- φ 1 Full Reconstruction : better S/N B0 D *+ D - B0 D *+ D *- Confirm partial Recon. Br=1.04 ±0.38±0.22 (x10-3 ) Entries/(2 MeV/c 2 ) First Observation a) Belle 11.2 ± 4 ev σ Mbc (GeV/c 2 ) Mbc (GeV/c 2 ) Entries/(2 MeV/c 2 ) 21.3 fb -1 (New) Br=1.21 ±0.41±0.27 Belle (x10-3 ) 5 4 a) 11.0 ± 3.7 ev D *+ D 0 π + only add D *+ D + π 0 mode Need angular anal. for φ 1 σ D 0 K - π + K - π + π 0 K - π + π + π

63 EW Penguin : b s l + l - Awaited mode sensitive to SUSY after b sγ (consistent to SM) Exclusive mode: B K (*) l + l - B K * µ + µ : < 3.0 ( ) B K * e + e - B K µ + µ : 0.99 B K e + e x : < 5.1 (5.5 ) : < 1.2 (2.6 ) First Observation 29.1 fb -1 B K + µ + µ B Ks µ + µ combined Events / 2.5 MeV/c ev σ M bc (GeV/c 2 )

64 Conclusions Large CPV in B decays! sin2φ 1 = 0.99 ± 0.14 ± First Observations mainly on Rare B Decays ~80fb -1 by Summer 2002! Precision measurements of sin2φ1 Vub, φ2 to constrain the Triangle Mixing-induced CPV in penguin diagrams Search for direct CPV More rare decays Exciting new results in the near future!