1. Malcolm John 1  and  at Babar Malcolm John LPNHE – Universités Paris 6&7 On behalf of the B A B AR collaboration

2. Malcolm John 2 PEP-II and B A B AR have performed beyond expectation CP violation in the B system is well established –sin(2  ) fast becoming a precision measurement As for the other two angles (the subject of this presentation) : –Many analysis strategies in progress –The CKM angle  is measured but greater precision will come –First experimental results on  are available –First experimental results on 2  are available Results presented here are based on datasets up-to 227 M BB –B A B AR and PEP-II aim to achieve 550 M BB (500 fb  1 ) by summer 2006 Conclusions (for those who can't wait…)

3. Malcolm John 3 (0,0)(0,1) (,)(,) V ub V ud * V td V tb * V cd V cb *   

4. Malcolm John 4 B A B AR : where? what? who? B A B AR collaboration consists 11 countries and ~590 physicists ! DIRC (PID) 144 quartz bars 11000 PMs 1.5T solenoid ElectroMagnetic Calorimeter 6580 CsI(Tl) crystals Drift Chamber 40 layers Instrumented Flux Return iron / RPCs [ → LSTs ] Silicon Vertex Tracker 5 layers, double sided strips At the PEP-II B-factory at SLAC e + (3.1 GeV) (9.0 GeV) e -

5. Malcolm John 5 PEP-II : performance PEP-II / B A B AR : August 2004 L peak = 9.2×10 33 cm  2 s  1 (3 times design luminosity !) PEP-II delivered 253 fb  1 B A B AR recorded 244 fb  1 At  (4S) resonance221 fb  1 Analysis dataset 211 fb  1 PEP-II / B A B AR : August 2004 L peak = 9.2×10 33 cm  2 s  1 (3 times design luminosity !) PEP-II delivered 253 fb  1 B A B AR recorded 244 fb  1 At  (4S) resonance221 fb  1 Analysis dataset 211 fb  1 Beams circulating in PEP-II again this month for the beginning of the 2005-2006 run. –Aim for 500 fb  1 by summer 2006 Belle and KEKB are running well –Their dataset could be 600 fb  1 in the same time period

6. Malcolm John 6 Analysis techniques : Continuum suppression For every pair-production, expect three –Many techniques available it fight this background. –They can be amalgamated in linear discriminators or neural networks. Variables that distinguish spherical B events from jet-like continuum. Variables that distinguish  (4S)→bb from e + e  →qq Other variables, like  H (D 0 ) bb signal bb background qq continuum

7. Malcolm John 7 Analysis techniques : m ES and  E Precise kinematics, unique to machines operating at a threshold –The initial energy of the system is well known from the precise tuning of the beam

8. Malcolm John 8 Large data sample feeds a plethora of analyses New particle searches : Pentaquarks, exotic baryons, D sJ Tau physics : lifetime measurements, rare decay searches Charm physics : D 0 mixing, precision D 0 measurements New physics searches in s-penguin decays Precision measurement of SM CP violation : sin(2  SM CP violation : measurement of CKM angle  SM CP violation : measurement of CKM angle  Search for rare leptonic B decays Semileptonic B decays and the determination of |V ub | and |V cb | Radiative penguin B decays Precision measurement of SM CP violation : sin(2  SM CP violation : measurement of CKM angle  SM CP violation : measurement of CKM angle  SM CP violation : measurement of CKM angles 2 

9. Malcolm John 9 t =0 Time-dependent analysis requires B 0 flavour tagging We need to know the flavour of the B at a reference t=0. B 0  (4S) The two mesons oscillate coherently : at any given time, if one is a B 0 the other is necessarily a B 0 In this example, the tag- side meson decays first. It decays semi-leptonically and the charge of the lepton gives the flavour of the tag-side meson : l  = B 0 l  = B 0. Kaon tags also used. tag B 0 l  (e-,  -)  =0.56  z =  t  c rec  t picoseconds later, the B 0 (or perhaps its now a B 0 ) decays. B 0 At t=0 we know this meson is B 0

10. Malcolm John 10 C  0 : Direct CP violation Formalism of CP violation with B mesons Time evolution of a S  0 : Indirect CP violation Time-dependent asymmetry Final State Amplitudes from mixing

11. Malcolm John 11 (0,0)(0,1) (,)(,) V ub V ud V cd V cb * * V td V tb V cd V cb * *    B → J/  K S

12. Malcolm John 12 sin(2  ) measurement with charmonium (214 M BB ) Limit on direct CPV from B 0 /B 0 mixing = ± 1 = 0 = sin(2  ) Time-dependent asymmetry with an amplitude  sin(2  ) (1  2 w) sin(2  ) w = mis-tag fraction 7730 events

13. Malcolm John 13 (0,0)(0,1) (,)(,) V ub V ud V cd V cb * * V td V tb V cd V cb * *    B →  B →  B → 

14. Malcolm John 14 The route to sin  Access to  from the interference of a b→u decay (  ) with B 0 B 0 mixing (  ) B 0 B 0 mixing Tree decayPenguin decay  Inc. penguin contribution How can we obtain α from α eff ? Time-dep. asymmetry :  NB : T = "tree" amplitude P = "penguin" amplitude

15. Malcolm John 15 How to estimate |  eff | : Isospin analysis Use SU(2) to relate decay rates of different hh final states (h  {  }) Need to measure several related B.F.s Gronau, London : PRL65, 3381 (1990) Difficult to reconstruct. Limiting factor in analysis 2|  eff |

16. Malcolm John 16 Time-dependent A CP of B  →     Good  /K separation up to 4.5 GeV/c Blue : Fit projection Red : qq background + B 0 →K  cross-feed BR result in fact obtained from 97M BB

17. Malcolm John 17 Now we need B  →     Analysis method reconstructs and fits B  →     and B  → K    together B→B→ B→B→ B→KB→K B→KB→K Inserts show background components B→hB→h B→hB→h

18. Malcolm John 18 …and B  →     61 ± 17 events in signal peak (227M BB ) –Signal significance = 5.0  –Detection efficiency 25% Time-integrated result gives : B±→±0B±→±0 3 B.F.s –B 0 →  –B  →   –B 0 →     2 asymmetries –C  –C  Using isospin relations and Large penguin pollution ( P/T ) –Isospin analysis not currently viable in the B→  system |  eff |< 35°

19. Malcolm John 19 Isospin analysis using B→  Extraction of  follows the same logic as for the B→  system –Except  is a vector-vector state –  is not generally a CP eigenstate –Angular analysis needed Longitudinal Helicity state h=0 CP+1 eigenstate Transverse Helicity state h= ±1 non-CP eigenstate However, is measured to be  1 in B→  –Transverse component taken as zero in analysis

20. Malcolm John 20 very clean tags Time dependent analysis of B→     Maximum likelihood fit in 8-D variable space 32133 events in fit sample

21. Malcolm John 21 Similar analysis used to search for     –Dominant systematic stems from the potential interference from B→a 1 ±  ± (~22%) Searching for B→     c.f. B→     B.F.= 4.7 x 10  6 and B→     B.F.= 1.2 x 10  6 B (B→      = 33 x 10  6

22. Malcolm John 22 Taking the world average and thanks to we apply the isospin analysis to B→  The small rate of means –|  eff | is small[er] –P/T is small in the B→  system (…Relative to B→  system) Isospin analysis using B→  |  eff |< 11°

23. Malcolm John 23 Another approach : B→(  ) 0 Unlike      and    ,     is not a CP eigenstate –Must consider 4 (flavour/charge) configurations –Equivalent "isospin analysis" not viable (triangles→pentagons, 6→12 unknowns…) However, a full time-dependent Dalitz plot analysis of can work! –Enough information to constrain  –+–+ +–+– 0000 Snyder, Quinn : PRD 48, 2139 (1993) Interference at equal masses- squared gives information on strong phases between resonances         

24. Malcolm John 24 Time-dependent Dalitz fit Extract  and strong phases using interferences between amplitudes Time evolution of can be written as : Assuming amplitude is dominated by  ,   and  , we write –The "f"s are functions of the Dalitz-plot and describe the kinematics of B→  (S→VS) –The "A"s are the complex amplitudes containing weak and strong phases. They are independent of the Dalitz variables script {  } refers to {       } Complicated stuff! –At least 17 parameters to fit-for in a 6-D variable space –Large backgrounds. Over 80% of selected events are continuum

25. Malcolm John 25 B→       : data/MC, 213M BB Fit finds 1184 ± 58 m ES  E’  ’ (DP variable) m’ (DP variable) NN tt Signal Self-crossfeed B-background Continuum-background

26. Malcolm John 26 Hint of direct CP-violation Fit result → Physics results : B→(  ) 0 213M BB Mirror solution not shown Weak constraint at C.L.<5% 2.9  Likelihood scan of  using :  {  } T =tree amp. P =penguin  

27. Malcolm John 27 Combining results on  Combining results in a global CKM fit Mirror solutions are clearly disfavoured  is measured. –Although improve- ments will come

28. Malcolm John 28 (0,0)(0,1) (,)(,) V ub V ud V cd V cb * * V td V tb V cd V cb * *    B ± → D (*) K (*) GLW, ADS and D0-Dalitz methods

29. Malcolm John 29 How to access  ? Decays where interferes with Colour favoured b  c amplitudeColour suppressed b  u amplitude –charged Bs only (time-independent, direct CPV) –no penguins pollution Need same final state Strong phase between diagrams → f CP GLW method Crucial parameter : (not well measured)

30. Malcolm John 30 GLW method : B→D 0 K (214 M BB ) Main background from kinematically similar B→D 0  which has B.F. 12x larger –So the signal and this main background are fitted together –2D fit to DE and the Čerenkov angle of the prompt track A 2  cut on  C in these plots R CP+ = 0.87±0.14(stat.) ±0.06(syst.) R CP  = 0.80±0.14(stat.) ±0.08(syst.) A CP+ = 0.40±0.15(stat.) ±0.08(syst.)A CP  = 0.21±0.17(stat.) ±0.07(syst.) R CP+ = 1.73±0.36(stat.) ±0.11(syst.) R CP  = 0.64±0.25(stat.) ±0.07(syst.) A CP+ =  0.08±0.20(stat.) ±0.06(syst.) A CP  =  0.35±0.38(stat.) ±0.10(syst.) B ± →D 0 K* ± Only a loose bound on r B with current statistics: (r B ) 2 = 0.19 ± 0.23

31. Malcolm John 31 Accessing  without using CP states Using CP final states of the D 0 yields an expected A CP of only ~10% –We can potentially do better using "wrong-sign" final states Colour favoured b  c amplitudeColour suppressed b  u amplitude → K    ADS method Cabibbo suppressed c  d amplitudeCabibbo favoured c  s amplitude  And with enough events (i.e. large r B ), expect large asymmetry D 0 →K  suppression factor: r D = 0.060 ±0.003 Phys.Rev.Lett. 91:17 1801

32. Malcolm John 32 Any  Any  ADS method : B→D ( * )0 K (227 M BB ) The number of events depends foremost on the value of The smallness of r B makes the extraction of  with the GLW/ADS methods difficult DK : r B < 0.23 D*K : (r B ) 2 < (0.16) 2 (90% C.L.)

33. Malcolm John 33 b→u sensitivity with an unsuppressed D 0 decay Consider once again B→D ( * )0 K decays, this time the with D 0 →K S . –Obtain  D information from a fit to the D 0 Dalitz plot Colour favoured b  c amplitudeColour suppressed b  u amplitude → K S  → K S  D 0 Dalitz model Sensitivity to   =75°,  =180°, r B =0.125 K   DCS  

34. Malcolm John 34 The D 0 →K S      Dalitz model Determine on clean, high statistics sample of 81500 D*  →D 0   events –ASSUME no D-mixing or CP violation in D decays –Build model from 15 known resonances (+2 unidentified scalar  resonances)    d.o.f. = 3824/(3054-32) = 1.27 K   K   DCS  

35. Malcolm John 35 D 0 Dalitz method : B→D ( * )0 K (227 M BB ) 282 ± 20 44 ± 8 B+ BB BB BB K  DCS 89 ± 11 Maximum likelihood fit extracts r B ( * ), ,  ( * ) from a fit to m ES,  E, Fisher and the D 0 →K S  Dalitz model.

36. Malcolm John 36 Measurement of gamma –Twofold ambiguity in  extraction D 0 Dalitz method : B→D ( * )0 K : result DK : r B < 0.19 D*K : r B = 0.155(90% C.L.) +0.070  0.077 ± 0.040 ± 0.020  = 70° ± 26° ± 10° ± 10° (+n  )  B = 114° ± 41° ± 8° ± 10° (+n  )  B = 303° ± 34° ± 14° ± 10° (+n  ) D*K DK  180° 0°  180° 180°  180°  0° 0.1 0.3 rBrB 68% 95% Bayesian C.L.s 3 rd error is due attributed to the Dalitz model

37. Malcolm John 37  (0,0)(0,1) (,)(,) V ub V ud V cd V cb * * V td V tb V cd V cb * *   B 0 → D (*) K (*) B 0 → D ± K S  B 0 → D (*) 

38. Malcolm John 38 Other ways to access b→u So, what happens if we start with a neutral B 0 ? b  c amplitudeb  u amplitude mixing Eventually this will be a time-dependent analysis –Early days yet though. Using non-CP modes of the D 0, we search for : → f CP AND self tagging b  c amplitude self tagging b  u amplitude AND

39. Malcolm John 39 D0 mass sideband B 0 →D ( * )0 K ( * )0 (124 M BB ) ( ) b  u amplitude b  c amplitude b  c  b  u Very challenging for 

40. Malcolm John 40 Avoiding colour suppression : B 0 →D ± K S  ± (88 M BB ) b  c amplitudeb  u amplitude mixing D ± K* ± → Method has two advantages : –Avoids colour suppression in b→u –Integrating over Dalitz plane removes ambiguities in eventual  extraction First experimental step complete : –Branching fraction measurement –Currently, Too few events for TD analysis –  1/3 of events are NOT in the K* region

41. Malcolm John 41 Another way to sin  B 0 →D*    favoured b  c amplitudesuppressed b  u amplitude ~  ~  Both and decay to, neither with a colour-suppressed diagram mixing but there's is not enough information to solve the system… The suppression factor, r B must be taken from elsewhere. is estimated from –SU(3) symmetry used Input

42. Malcolm John 42 B 0 →D*    : very small A CP offset by copious statistics Full reconstruction (110M BB )Partial reconstruction (178 M BB ) Total yields (all tags) yields D*  Combinat. BB Peaking BB Continuum Find events with two pions and examine the missing mass X

43. Malcolm John 43 sin  results Partial reconstruction (178 M BB ) Full reconstruction (110M BB ) L L L L L : result uses lepton tags only

44. Malcolm John 44 PEP-II and B A B AR have performed beyond expectation CP violation in the B system is well established –sin(2  ) fast becoming a precision measurement As for the other two angles (the subject of this presentation) : –Many analysis strategies in progress –The CKM angle  is measured but greater precision will come –First experimental results on  are available –First experimental results on 2  are available Results presented here are based on datasets up-to 227 M BB –B A B AR and PEP-II aim to achieve 550 M BB (500 fb  1 ) by summer 2006 Conclusions

45. Malcolm John 45 (0,0)(0,1) (,)(,) V ub V ud V cd V cb * * V td V tb V cd V cb * *    BACK-UP

46. Malcolm John 46 PEP-II : performance Trickle injection in both beams Luminosity LER current HER current Trickle injection in the 3.1GeV beam Before implementation of trickle injection 5Hz "trickle" injection used in 2004

47. Malcolm John 47 Comparison with Belle : CPV in B 0      3.2  5.2  Belle report observation of CPV in B 0      >3  discrepancy between B A B AR & Belle Belle 3.2  evidence for Direct CP violation not supported by B A B AR measurements

48. Malcolm John 48 GLW method : B→D 0 K* (227 M BB ) Recontruct K*→K S . –Clean, no kinematically similar background –Lower B.F. and lower efficiency : Fewer events than D 0 K analysis R CP+ = 1.73±0.36(stat.) ±0.11(syst.) R CP  = 0.64±0.25(stat.) ±0.07(syst.) A CP+ =  0.08±0.20(stat.) ±0.06(syst.) A CP  =  0.35±0.38(stat.) ±0.10(syst.) (r B ) 2 = 0.19 ± 0.23 R CP+ = 1.09±0.26(stat.) ±0.09(syst.) A CP+ =  0.02±0.24(stat.) ±0.05(syst.) Similar analysis, 121 M BB

49. Malcolm John 49  from B→DK Combining results from GLW, ADS and D0-Dalitz methods –UTFit collaboration – hep-ph/0501199  = 60° ± 28° (+n  )

50. Malcolm John 50  from Belle