1. Thesis Defense of Luminda Kulasiri Dept. of Physics, University of Cincinnati 05.09.2005 Search for exclusive two body decays of B → D h at Belle *S*S

2. Motivation- u V ub u B-B- u s c W-W- b V cs * DSDS *-*- 00 Decay via b → u spectator Diagram Clean measurement for V ub No penguin terms Model independent No yet seen Important input for measuring Sin(2β+ γ) V ub d B0B0 u s c W-W- b V cs * DSDS * -* - ++ d

3. Motivation- B0B0 Evidence for W-exchange First seen in B 0 → D s -  + decay -PRL 89, 231804(2002) Not seen Role of the final state interactions Br can be as large as 10 -4 D +  -, D 0  0 can turn out to be D s * - K + (B. Block et al. PRL 78, 3999, 1997) c V cb * V ud d b DSDS *+*+ K-K- W s s u

4. Past Measurements & Theoretical Predictions @ 90% C.L. -PDG 2004 Theoretical predictions A. Deandrea et al. Phy. Lett. B 318, 549(1993)

5. → → a 1 ~1.0, V ub ~.003, V cs ~0.97 D. Choudhry et al. Phy. Rev. D, 45, 217(1992) Theoretical predictions

6. KEKB Accelerator Asymmetric Collider with, 8.0 GeV e - x 3.5 GeV e + 22 mrad crossing angle L peak = 13.92 nb -1 s -1 ∫Ldt ~400 fb -1 E cm = 10.58 GeV operates at  (4S) resonance e + e - →  (4S) →BB  (4S)  center of mass frame

7. Belle Detector Silicon Vertex Detector(SVD) Central Drift Chamber(CDC) Aerogel Cherenkov cnt. K L /µ Detector CsI Calorimeter TOF counter SC solenoid 8GeV e - 3.5 GeV e +

8. Belle Detector Silicon Vertex Detector (SVD) Tracks low momentum particles with CDC Vertex reconstruction,   18 µm Central Drift Chamber (CDC) Mom. of charged particles is measured from the curvature of the track traversing in the magnetic field PID using dE/dx - energy loss by ionizationin the matter Aerogel Cerenkov Counter (ACC) Index of refraction ranges from 1.01 to 1.03 K/  ID between 1.2 – 4.0 GeV/c TOF counter K/  seperation using timing of plastic scintillation counters CsI Calorimeter Measure energy of e’ns and  via detection of scintillation light from e.m. showers. K L /µ Detector Detect high mom.(>600 MeV) K/µ SC Solenoid Generates 1.5 T mag. field

9. Particle ID at Belle Uses information from CDC, TOF, and ACC Combine the information using Likelihood method P i – likelihood for signal species P j – likelihood for background species Where i, j  {e, , K, , p}

10. Used 250 fb -1 data at  (4S) center of Mass resonance  274.8 million BB events Decay Chain Conjugate modes are also assumed B 0 → D s *+  -, B 0 → D s *- K +, B + → D s *+  0 D s *+ → D s +  D s + → {  +, K S K +, K * K + }  →K + K -, K S →  +  -, K * →K +  -

11. Reconstruction of , K 0 s, and K *0  →K + K - Kaon ID>0.6 1.0116<M(KK)<1.0272 GeV (  ±3  of the nominal mass) K *0 →K +  - K/  ID>0.6 |M(KK) – 0.8961| < 0.060 GeV (  ±3  of the nominal mass) Data

12. Helicity Angle Helicity angle(θ h ) – Angle between momentum of D S and momentum of K in  (K*) frame. Bg. evts. Signal evts. Flat dist. for background evts. Cos 2 (θ) dist. for signal evts. Selection requirement, |cos(θ h )| > 0.3

13. Reconstruction of , K 0 s, and K* 0 - Cont. K 0 s →  +  - Pion ID>0.6 0.4902<M(  +  - )<0.5051 GeV (  ±3  of the nominal mass)  2 <30 (vertex reconstruction fit) Other cuts: dr>0.009 cm; d  <0.2 rad dr – smaller of dr1 and dr2, where dr1 and dr2 are the smallest approach from the IP to the two tracks in x-y plane d  - angle betn. the momentum vector and decay vertex displacement vector in r-  plane Background Events Signal Events dr(cm) d  (rad) dr(cm) d  (rad)

14. Reconstruction of D S + and D S * + ±3  of the nominal mass (1968.5±0.6 MeV) 1.9539<M(  )<1.9833 GeV 1.9471<M(K S K)<1.9901GeV 1.9495<M(K*K)<1.9877GeV momentum selection 1.7<P(cms)<2.5 GeV  M=M(D S *)-M(D S )  M has better resolution than mass. 0.124<  M<0.164 GeV A large portion of the background is accounted by photons that are not really coming from D S *. E  (cms)>110 MeV  0 veto Photon Energy (E  ) Signal Bg. GeV

15. Selection of the B Candidate Two quantities M(B) and  E are defined as, where E beam =5.29 GeV,  P i – mom. of B,  E i – energy of B Signal Region, –0.05<  E<0.05 GeV for h  { , K} –0.10<  E<0.05 GeV for h  {  0 } 5.27<M(B)<5.29 GeV/c 2 GeV/c 2 B→D s *+  - (D s + →  + ) GeV Signal MC  E vs. M(B)

16. Background Suppression – Fisher Discriminant Largest background source is e + e - → qq events Fisher Discriminant is a powerful tool to discriminate signal and background A linear combination of 9 variables Optimized to discriminate signal from background

17. Background Suppression – Fisher Discriminant (cont.) Cos(θ th ) – Angle between thrust axis of the B cand., and the thrust axis of the remaining particles. Cos(  B ) – Angle between the B momentum & beam axis qr x (Q DS ) – qr contains flavor information of the other B; q = ±1; 0<r<1; Where is a unit vector s. t. it maximizes T is the mom. of the i th particle in CM frame Combine all 9 variables into F

18. Background Suppression – Fisher Discriminant (cont.) All the parameters are optimized to get the maximum discrimination between signal and background Used Figure of Merit (FOM) plots to decide the best selection FD Arbitrary units continuum Signal S - Signal B - background

19. First used Sig. MC – fitting M(B) Observed inconsistency among the yields of D S sub modes Minimized MC dependence by using inclusive  M Efficiency for  mode obtained using sig. MC Total eff. obtained by multiplying by eff. of the other cuts N – inclusive  M yield ;  - efficiency ; Reconstruction Efficiencies Following relationships can be obtained

20. Sideband Study Sidebands of D s and  M used 3  from lower and upper side of the signal region Can be used to compare data and MC Background shapes and rates can be obtained Bg. shapes of data and MC agree each other Observed a disagreement in bg. levels ~12% – 22% Bg. of D s not random – real D s but not from D s * →D s  Observations:

21. Simultaneous Fitting Simultaneous fitting of 3 D S sub decay modes Common branching fraction for all 3 DS sub-decay modes  M-1D Signal-Gaussian shape with mean &  fixed to sig. MC shape Bg. – linear shape, by fitting data excluding the signal region  E-1D Signal-Gaussian shape with mean &  fixed to sig. MC shape Bg. – sideband shapes of  M Significance

22. Simultaneous fitting - cont.  E fit Solid line (red) – total fit dotted line (blue) - background Ds*+-Ds*+- Ds*-K+Ds*-K+ Ds*+0Ds*+0

23. DS*+-DS*+- DS*+K-DS*+K- DS*+0DS*+0 Simultaneous fitting - cont.  M fit Solid line (red) – total fit dotted line (blue) - background

24. Systematic Uncertainties Since 3 D s modes, Total Syst. error = common syst. errors + indept. syst. errors Common Errors (%)  rcon.  0 recon. Br(D s *→D s  ) N BB Total D s *+  - 5 - 3 1 5.9 D s *- K + 5 - 3 1 5.9 D s *+  0 5 3 1 7.7

25. Systematic Uncertainties-cont. Independent Errors Br(D s ) /Br(D s  →   ) Br(sub D s ) K s Fitting MC stat. Tracking PID Common Total Br(D s  →   ) D s *+  -  + K s K + K * K + 0.00 22.2 9.1 1.4 0.4 0.2 - 3.0 - 8.20 4.5 3.4 2.9 5.9 12.04 25.0 D s *- K +  + K s K + K * K + 0.00 22.2 9.1 1.4 0.4 0.2 - 3.0 - 9.40 6.1 4.5 3.9 5.9 14.25 25.0 D s *+  0  + K s K + K * K + 0.00 22.2 9.1 1.4 0.4 0.2 - 3.0 - 14.9 8.1 4.4 4.0 7.7 17.98 25.0

26. R D*  for Sin(2  1 +  3 )  D*  - strong phase,  c – Cabibbo angle D. Becirevic, Nucl. Phys. Proc. Suppl. 94, 337 (2001 - good agreement with the expected result which is ~0.02

27. Estimation of V ub -Good agreement with the world average for |V ub | which is (3.67±0.47)x10 -3

28. Summary  Used 278.4 million events   M fits give more consistent results  Both  M and  E results agree within errors  Obtained estimates for V ub and R D* 

29. D S * +  - 19.0±5.7 evts D S * - K + 11.0±4.8 evts D S * +  0 9.4±5.5 evts Combined yields from  M fit (274.8 m evts.)

30. Combined yields from M(B) fit (274.8 m evts.) D S * +  - 21.4±5.9 evts. D S * +  0 8.1±5.7 evts. D S * - K + 10.6±4.9 evts.

31. Combined yields from  E fit (274.8 m evts.) 4.4  8.4 evts. 15.1  5.6 evts. 8.2  4.7 evts.