BSM at Belle: B K ℓ ℓ and search for leptonic B decays

Transcription

1 BSM at Belle: B K ℓ ℓ and search for leptonic B decays * + - Stefano Villa, April 24, 2006 based on (i.e. same as) the talk given at Les Rencontres de Physique de la Vallée d Aoste, La Thuile, Italy, March 10, 2006 NEW: updated with Belle result on B

2 Summary B physics at Belle The B K*ℓ+ℓ- channel forward-backward asymmetry measurement of Wilson coefficients B ℓ ℓ searches B ℓ+ℓ- searches Future prospects Conclusions Notation: 0 0 B B d Charge-conjugate modes always included SM: Standard Model BSM: Beyond the Standard Model

3 B physics at Belle Super-Cond. Solenoid CsI calorimeter (ECL) Time Of Flight counter Aerogel Cherenkov Counter Central Drift Chamber B production BB pairs produced at KEKB in e+e (3.5 GeV on 8 GeV) collisions at the (4S) resonance. Collected so far more than 500 fb-1 Charged tracks reconstruction/id: electron ID: loss in CDC, shower shape Silicon Vertex Detector /KL Detector (KLM) in ECL and response of ACC; eff 90%, -misid rate 0.1% B signal selection: muon ID: based on ECL and KLM; typically based on event shape variables eff 90%, -misid rate 1% with signal window defined using K selected using ACC, TOF and CDC; 2 2 M bc = E beam p B m B eff 90% and -misid rate 6%. Other charged tracks identified as and E = E E ( 0 ) B beam

4 B K ℓ ℓ * + -

5 B K ℓ ℓ : a window on BSM physics * + - SM: : FCNC process, forbidden at tree level at lowest order via electromagnetic penguin or box diagrams b sℓℓ Lepton pair yields useful observables for testing the theory: forward-backward asymmetry (AFB ) invariant mass ( q2 ) BSM: Sensitive to new physics via insertion of heavy particles in the internal lines.

6 B K ℓ ℓ : Wilson coefficients * + - New Physics at the one loop level can be described in terms of an effective Hamiltonian: Local operators (see next page) Ci ( ) Wilson coefficients: effective strength of short distance interactions To leading order, only O7, O9 and O10 contribute to b sℓℓ Ci computed perturbatively up to NNLO: Ci = Ai + higher order terms The B K*ℓ+ℓ- amplitude depends on A7, A9 and A10 under the assumption that higher order terms behave like in the SM. SM VALUES: A7 = 0.330, A9 = 4.069, A10 = H.H. Asatryan et al. Phys. Lett. B 507, 162 (2001); A. Ali et al. Phys. Rev. D 66, (2002)

7 Operators in Heff electromagnetic operator semileptonic vector semileptonic axial-vector

8 Constraints on Wilson coefficients The absolute value of C7 is constrained by B Xs ; constraints on C9 and C10 (donut-shape) are derived from the B Xsℓ+ℓ- branching fractions. Allowed region at 90% CL, based on NNLO and experimental bounds on B Xs and B Xsℓ+ℓ- Br s; A7<0 C10NP=C10 C10SM A. Ali et al. Phys.Rev. D 66, (2002) SUSY Extended-MFV SM: To determine sign of C7 and to C9NP=C9 C9SM measure C9 and C10 need to look at the differential distributions in B K*ℓ+ℓ-

9 Forward-backward asymmetry in K ℓ ℓ * + - ℓ Bℓ- - ( ): angle between B and ℓ in the dilepton B K* rest frame AFB is a function of q2 of the dilepton system AFB non-zero due to interference of vector (C7, C9) and axial vector (C10) couplings More generally, one can extract the coefficients by fitting the double-differential decay width: d2 / dq2 d cos Bℓ- ℓ+

10 B K ℓ ℓ selection * + - Dataset: 357 fb-1 = 386M BB pairs *+ Modes: K K+ 0, Ks +; K*0 K+ lepton = e, Charmonium (J/, (2S)) veto Dominant background: BB with both B's decaying semileptonically: suppressed using Emiss and cos B* B Kℓ+ℓ- used as null test : AFB 0 in SM, small BSM D.A. Demir et al. Phys.Rev. D66 (2002) Signal yield: Nsig = Consistent with Belle measurement (140fb-1): Br(B K*ℓ+ℓ-)=( )x10 7 A. Ishikawa et al. Phys.Rev. Lett. 91, (2003)

11 Extraction of AFB and Wilson coeffs. Extract the ratio of Wilson coefficients A9 /A7, A10 /A7 ( A7 = A7SM= 0.330) from an unbinned maximum likelihood fit on events in the signal window with a pdf including g(q2, ) = d2 / dq2 d cos. Several event categories: signal + cross feeds from misreconstructed B K(*)ℓ+ℓ- or other b sℓℓ 4 background sources dominated by dilepton (80%) AFB simply obtained by integration: background subtracted AFB in bins of q2 Null test: extract AFB for B Kℓ+ℓ- consistent with 0!

12 Fit results hep-ex/ submitted to PRL J/ fix A7 = 0.330; AFB >0 at 3.4 * Abkg-sub ( B K ℓℓ) = 0.56 ± 0.13 (stat.) FB A7A10 sign flipped slightly worse fit, but OK Both A7A10 and A9A10 signs flipped A9A10 sign flipped 3 4 (2S) Wilson coefficients: +3.4 A9/A7 = ± 1.1 A10/A7 = ± < A9 A10 / A72 < 26.4 SM: excluded! 2 1 (A7SM) (95% CL, any A7 ) SM

13 Positive A7 solution Best fit for positive A7 (non-sm like): SM

14 LEPTONIC B DECAYS

15 _ B ℓ ℓ - - SM: B lifetime Direct Measurement of decay constant fb! Br(B ) 1 x Other leptons suppressed ~(m ) : ℓ by 1/225 for B, 10-7 for B e e BSM: Possible enhancements of BF in MSSM (charged Higgs): can explore the (M, tan ) plane. H Pati-Salam models: can set limit on the mass of LQ

16 _ -_ B e e and B - - One highly energetic lepton Charmonium veto Large missing E and p Signal window defined on E and Mbc of the companion B Cut on lepton momentum in B rest frame B at 90% CL BELLE results Br(B ) < 2.0x Br(B e e) < 5.4x10 hep-ex/ , 140 fb-1-1 Belle-conf-0247, 60 fb update coming soon!

17 _ B - - NEW: submitted to PRL, hep-ex/ Reconstruct the companion B in exclusive D( )0h+ and D( )0D( )+s channels to get a pure (55%) B+B- sample (6.8x105 evts) Reconstruct signal from remaining particles in the event Final selection based on remaining energy in ECL: E ECL 0 for signal Dataset: 414 fb-1

18 _ B - - To validate the EECL cut, use control sample of double tagged events: Bsig substituted by B D*0ℓ FIT RESULT: signal background total

19 _ B - : results - First evidence of a purely leptonic B decay BELLE result Br(B ) = ( (stat) (syst)) x fb = (stat) (syst) GeV obtained using Vub =(4.39±0.33)x10-3 (HFAG) -4 SM: Br(B )=(1.59±0.40)x10-4 fb = 0.216±0.022 GeV (from lattice QCD) First direct determination of fb

20 _ B - : constraints on BSM - (two Higgs doublet model, type II) W.S. Hou, Phys. Rev. D 48, 2342 (1993)

21 0 B ℓ ℓ + - SM: Box or annihilation diagram Br(B0 + -) = ( ) x Br(B0 e+e-) = ( ) x Br(B0 e ) 0 (neutrino osc.) Helicity suppressed ~(mℓ)2 BELLE results 90% CL limits based on 78 fb-1 BSM: Enhancement of BF(B0 + -, e+e-) in high tan MSSM (2 orders of magnitude) and SUSY B0 e allowed in Pati-Salam (leptoquark) and SUSY models Phys. Rev D 68, (R) (2003) Br(B + -) < 1.6x10-7 Br(B e+e-) < 1.9x10-7 Br(B e )< 1.7x10-7 limit on the Pati-Salam LQ mass: MLQ > 46 TeV/c2 at 90% CL

22 Super Belle: expected performance Goal: L=5x1035cm-2 s-1; in 1 year L=5 ab-1 expected performance on B K ℓ ℓ with 1 year of data taking no syst. errors included * + - from AFB(K*ℓ+ℓ-) C10NP C9NP zero of AFB(q2) is very sensitive to BSM effects. Will be able to measure it. A9/A9~11% A10/A10~13% A. Ishikawa at Lake Louise 2006

23 Conclusions Belle performed the first measurement of Wilson Coefficients in B K*ℓ+ℓ- : Integrated forward-backward asymmetry significantly >0 First determination of sign of A9A10 Results compatible with SM prediction and ruling out many BSM scenarios B leptonic decays set constraints on BSM parameter space MH tan in MSSM M LQ in Pati-Salam models First evidence of B! Still a lot to come from Belle and hopefully Super Belle! but something else happened since la Thuile...

24 NEW! Cecilia was born! BELLA bimba mc = 3650 g LC = 53 cm t0 = Apr. 7, 22:26

25 BACKUP SLIDES

26 Details of the fit The Probability Density Function: : efficiency functions, estimated from data and MC f : event by event signal and background probability, from Mbc fit

27 Systematic uncertainties source negative A7 solution positive A7 solution A9/A7 A10/A7 A9/A7 A10/A7 A mb ± 0.63 ± 0.42 Form factor model ± 0.66 ± 1.72 ± q2 resolution ± 0.28 ± 0.39 ± 0.28 ± 0.39 efficiency ± 0.08 ± 0.03 ± 0.10 ± 0.06 signal fraction total