1 / 56

Why charmless sl. Decays? Analysis Techniques Overview of BaBar results (3 new results in 2005)

Jochen Dingfelder , SLAC SLAC Experimental Seminar, June 9 th 2005. Why charmless sl. Decays? Analysis Techniques Overview of BaBar results (3 new results in 2005) Form Factors Vub Prospects for 500 fb-1. |V ub |/|V cb | band ( ±2 σ ) σ =12%. Why Charmless Semileptonic B Decays ?.

truong
Télécharger la présentation

Why charmless sl. Decays? Analysis Techniques Overview of BaBar results (3 new results in 2005)

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Jochen Dingfelder , SLAC SLAC Experimental Seminar, June 9th 2005 Why charmless sl. Decays? Analysis Techniques Overview of BaBar results (3 new results in 2005) Form Factors Vub Prospects for 500 fb-1

  2. |Vub|/|Vcb| band (±2σ) σ=12% Why Charmless Semileptonic B Decays ? • Study b  u transitions Goal: Precise determination of |Vub| • Test CKM description of quark coupling and CP violation • Complementary constraint of UT to angle measurements; • try to reach similar precision as sin2b. s(|Vcb|) ≈ 2% s(|Vub|) ≈ 12%

  3. Why Charmless Semileptonic B Decays ? • Study b  u transitions Goal: Precise determination of |Vub| • Test CKM description of quark coupling and CP violation • Complementary constraint of UT to angle measurements; • try to reach similar precision as sin2b. Constraints w/o angles s(|Vcb|) ≈ 2% s(|Vub|) ≈ 12% |Vub|/|Vcb| band (±2σ) σ=12%

  4. u/d u/d Xu= p, r, ... B0,± b u l Vub n Why Charmless Semileptonic B Decays ? • Vub is best measured in charmless semileptonic decays B  Xuln • Factorization of leptonic and hadronic currents • Learn about electroweak interaction (coupling) • Learn about strong interaction study structure of B meson •  allows tests of e.g. Lattice QCD • So far B pln and rln have been measured (CLEO, Belle, BaBar (rln) ). • Also wln (Belle) and hln (CLEO) have been seen. • Will present preliminary BABAR results from 2004-05 (82 fb-1 or 211 fb-1)

  5. Semileptonic B decays : The “Big Picture” |Vub|2 / |Vcb|2 ≈ 1 / 50 • Exclusive Decays • Low signal rate, better bkg reduction and kinematic constraints • Need Form FactorsF(q2) to describe the hadronization process u  p, r, … • Measurement as function of q2 • Inclusive Decays • Large signal rate, high b cln bkg • “Easy” to calculate (OPE/HQE) • Need Shape Function (b-quark motion inside B meson) . Constrain SF param. mb, mp2 with b  sg or b  cln.

  6. CsI (Tl) BaBar DIRC e+ (3.1 GeV) • 5-layer SVT tracker • 40-layer Drift CHamber dE/dx • Novel RICH based on total internal reflection (DIRC) • CsI(Tl) crystal calorimeter (e±, γ) • RPC and LST chambers in flux return for muon ID e- (9 GeV) IFR SVT DCH How Do We Detect These Decays ? B  Xu l+n (l = e,m) • Good e,m ID(p*l>1GeV)and hadron ID • (e.g. p/K separation  reject kaon tracks) • Reliable track and g reco. (DCH,EMC) • Angular coverage ≈ 92% of 4πin CMS • (challenge forνreco)  p’s , g’s

  7. CsI (Tl) BaBar DIRC e+ (3.1 GeV) • 5-layer SVT tracker • 40-layer Drift CHamber dE/dx • Novel RICH based on total internal reflection (DIRC) • CsI(Tl) crystal calorimeter (e±, γ) • RPC and LST chambers in flux return for muon ID e- (9 GeV) IFR SVT DCH How Do We Detect These Decays ? B  Xu l+n (l = e,m) • Good e,m ID(p*l>1GeV)and hadron ID • (e.g. p/K separation  reject kaon tracks) • Reliable track and g reco. (DCH,EMC) • Angular coverage ≈ 92% of 4πin CMS • (challenge forνreco)  p’s , g’s

  8. CsI (Tl) BaBar DIRC e+ (3.1 GeV) • 5-layer SVT tracker • 40-layer Drift CHamber dE/dx • Novel RICH based on total internal reflection (DIRC) • CsI(Tl) crystal calorimeter (e±, γ) • RPC and LST chambers in flux return for muon ID e- (9 GeV) IFR SVT DCH How Do We Detect These Decays ? B  Xu l+n (l = e,m) • Good e,m ID(p*l>1GeV)and hadron ID • (e.g. p/K separation  reject kaon tracks) • Reliable track and g reco. (DCH,EMC) • Angular coverage ≈ 92% of 4πin CMS • (challenge forνreco)  p’s , g’s

  9. CsI (Tl) BaBar DIRC e+ (3.1 GeV) • 5-layer SVT tracker • 40-layer Drift CHamber dE/dx • Novel RICH based on total internal reflection (DIRC) • CsI(Tl) crystal calorimeter (e±, γ) • RPC and LST chambers in flux return for muon ID e- (9 GeV) IFR SVT DCH How Do We Detect These Decays ? B  Xu l+n (l = e,m) • Good e,m ID(p*l>1GeV)and hadron ID • (e.g. p/K separation  reject kaon tracks) • Reliable track and g reco. (DCH,EMC) • Angular coverage ≈ 92% of 4πin CMS • (challenge forνreco)  p’s , g’s DIRC p/K separation

  10. CsI (Tl) BaBar DIRC e+ (3.1 GeV) • 5-layer SVT tracker • 40-layer Drift CHamber dE/dx • Novel RICH based on total internal reflection (DIRC) • CsI(Tl) crystal calorimeter (e±, γ) • RPC and LST chambers in flux return for muon ID e- (9 GeV) IFR SVT DCH How Do We Detect These Decays ? B  Xu l+n (l = e,m) • Good e,m ID(p*l>1GeV)and hadron ID • (e.g. p/K separation  reject kaon tracks) • Reliable track and g reco. (DCH,EMC) • Angular coverage ≈ 92% of 4πin CMS • (challenge forνreco)  p’s , g’s EMC

  11. D* D0 p+ p+ p+ n p+ Y(4S) Y(4S) Y(4S) n n n l- l- l- p- l- Tagging Methods Hadronic Tag: Fully reconstruct hadronic decay of one B: B D(*) + (p+,p0,K+,K0) ≈ 1000 modes  know kinematics of other B Semileptonic Tag: Reconstruct B  D(*) l n and study recoil - Full reconstruction of D(*) - Partial reconstruction of D* (only l, psoft) Two n tag-B kinematics incomplete No Tag: High statistics High backgrounds and cross-feed  Fully reconstruct signal side (n reco.)

  12. Tagging Methods: Event Yields • Advantages of Tagging: • Determine tag-Bcharge, flavor, momentumconstraints for signal B • Separating the two B decays  reducecombinatorics / cross-feed • Reduce non-BB(continuum) background looser cutsfull phase space • BUT: Significantly lower signal rates ! × 4 ×10 for pln ×1/3 ×1/3

  13. Hadronic Tags

  14. M2miss p+ln b  cln Hadronic Tags: B p (r, w, h,...) l n • Tag side: fully reco. B  PB sig = -PB tag • Signal side: Plep > 1 GeV • n fully constrained : M2miss < 0.5 GeV2 • Pro: Clean sample, almost bkg free • Loose cuts  keep full phase space • Con:Very low signal rate ! Variety of signal modes studied ! 82 fb-1

  15. 82 fb-1 Plus competitive limits for h, a0,a + Systematics dominated by MC statistics Hadronic Tags: B p (r, w, h,...) l n

  16. Semileptonic Tag

  17. Semileptonic Tag Method • Reconstruct B D(*)ln and study semileptonic recoil • Tagging efficiency measured with “Double Tags” (two D(*)ln) D(*)ln decays Correct sl. decay  |cosqBY| < 1, bkg broader ( tag side: cosqBY, signal side: cosqB,pl) cos2fB < 1 for signal, bkg flat

  18. B+p0ln with SL Tag : Signal Extraction • Cut-and-count analysis in cosqB,pl and mD • Signal region: -1.1 < cosqB,pl < 1.0 • Subtract mD sidebands remove cominatoric background • Subtract other background using MC normalized in -20 < cosqBpl < -1.5 82 fb-1 45 p0ln

  19. 14 p+ln 26 p+ln 21 p+ln 211 fb-1 Mainly B0B0 bkg ≈ 30% rln X-feed B0 p+ln with SL Tag : Signal Extraction Extract signal yields by binned c2 fit to cos2fBin3 bins of q2 : q2 = (pB – pp)2 , s(q2) ≈ 0.7 GeV2 Fit parameters = signal and background normalizations

  20. Partial SL (D*ln) Tag: B p+l n • Tag only l+ and p-soft from D*-D0p-soft •  D* momentum pD* = f (psoft) • Tag side n mass:Mn2 = (pB - pD* - pl)2 • Signal side: Look for l- p+hard • Check consistency of remaining particles (X) with D0 decay (multivariate discriminator) • Signal yieldfromfit to signal side mn2 MX pln BB bkg B(B0->p-l+n) = (1.46 ± 0.27 ±0.34 ±0.03FF) × 10-4 q2<8 8<q2<16 q2>16 82 fb-1

  21. No Tag

  22. u/d u/d Xu= p/r B0,± Y = p/r + l b u l n • Harsh suppression of b  cln background • e+e- qq background Untagged B pln, B  rln : Analysis Strategy • Neutrino Reconstruction:Reconstruct n from full event & ensure good reco. quality • Select signal decay candidates: • Y = hadron (p/r) + lepton (e/m) • Max-LH fitof signal and background in DE, mES, and q2 • Fit all 4 signal modes simultaneously (p+, p0, r+, r0)

  23. u/d u/d Xu= p/r B0,± Y = p/r + l b u l n • Harsh suppression of b  cln background • e+e- qq background Untagged B pln, B  rln : Analysis Strategy • Neutrino Reconstruction:Reconstruct n from full event & ensure good reco. quality • Select signal decay candidates: • Y = hadron (p/r) + lepton (e/m) • Max-LH fitof signal and background in DE, mES, and q2 • Fit all 4 signal modes simultaneously (p+, p0, r+, r0)

  24. u/d u/d Xu= p/r B0,± Y = p/r + l b u l n • Harsh suppression of b  cln background • e+e- qq background Untagged B pln, B  rln : Analysis Strategy • Neutrino Reconstruction:Reconstruct n from full event & ensure good reco. quality • Select signal decay candidates: • Y = hadron (p/r) + lepton (e/m) • Max-LH fitof signal and background in DE, mES, and q2 • Fit all 4 signal modes simultaneously (p+, p0, r+, r0)

  25. pmiss resolution s = 70 MeV Landau: 280 MeV Tail due to losses |pmiss| - |pn|(GeV) Neutrino Reconstruction Neutrino “Quality Cuts”: • Net charge: |SQ| ≤ 1, qmiss > 0.6 rad  limit losses due to acceptance • Missing mass: |M2miss/2Emiss| < 0.4 GeV2 B0 pln MC no add. KL, n add. KL add. n M2miss/2Emiss (GeV)

  26. pmiss resolution s = 70 MeV Landau: 280 MeV Tail due to losses |pmiss| - |pn|(GeV) Neutrino Reconstruction Neutrino “Quality Cuts”: • Net charge: |SQ| ≤ 1, qmiss > 0.6 rad  limit losses due to acceptance • Missing mass: |M2miss/2Emiss| < 0.4 GeV2 B0 pln MC no add. KL, n add. KL add. n Scale neutrino momentum, so that DE=0  Improves q2 resolution M2miss/2Emiss (GeV) w/o DE=0

  27. u/d u/d Xu= p/r B0,± Y = p/r + l b u l n • Harsh suppression of b  cln background • e+e- qq background Untagged B pln, B  rln : Analysis Strategy • Neutrino Reconstruction:Reconstruct n from full event & ensure good reco. quality • Select signal decay candidates: • Y = hadron (p/r) + lepton (e/m) • Max-LH fitof signal and background in DE, mES, and q2 • Fit all 4 signal modes simultaneously (p+, p0, r+, r0)

  28. Background Suppression Continuum Suppression: Off-res. data statistics low Need to rely on MC Minimize impact of continuum bkg L2 =Spi*cos2qi *< 1.5 GeV  Suppress 80% b  c l n b  c l n Suppression: Need to suppress large charm bkg in rln channel kinem. cuts on p*l ,p*h (for pln we can keep nearly whole phase space) L2 Signal (p) qq cosqthr cosqthr p*h Signal (r) b cln b  c l n p*l p*l Selection efficiencies: ≈ 3% for pln, ≈ 1% for rln

  29. u/d u/d Xu= p/r B0,± Y = p/r + l b u l n • Harsh suppression of b  cln background • e+e- qq background Untagged B pln, B  rln : Analysis Strategy • Neutrino Reconstruction:Reconstruct n from full event & ensure good reco. quality • Select signal decay candidates: • Y = hadron (p/r) + lepton (e/m) Before we fit the signal, we should check that selection / n reconstruction really work! • Max-LH fitof signal and background in DE, mES, and q2 • Fit all 4 signal modes simultaneously (p+, p0, r+, r0)

  30. “Proof of Principle”: B  D*ln Control Sample • Select high statistics and high purity control sample • Perform same selection as for b  uln signal (except b  cln suppression) • Cross-check MC modeling of signal decays • Study modeling ofdominant charm background component (D*) • Two modes for B  D*-l+n D0p-l+n: D0 K+p- , D0 K+p- p0 . K+p-p0 K+p-p0 K+p-p0 Signal other D* D/D** q2 (GeV2) M2miss/2Emiss (GeV) DE (GeV) Efficiencies of cuts agree between data and MC within a few %.

  31. u/d u/d Xu= p/r B0,± Y = p/r + l b u l n • Harsh suppression of b  cln background • e+e- qq background Untagged B pln, B  rln : Analysis Strategy • Neutrino Reconstruction:Reconstruct n from full event & ensure good reco. quality • Select signal decay candidates: • Y = hadron (p/r) + lepton (e/m) • Max-LH fitof signal and background in DE, mES, and q2 • Fit all 4 signal modes simultaneously (p+, p0, r+, r0)

  32. Fine binning in Signal Region Coarse binning in Sideband Region B pln pln rX-feed Xu fixed cln qq fixed Fit in DE-mES and q2 • Binned max.-likelihood fit, including MC statistics (Barlow & Beeston) • Fit all signal modes simultaneously • Float p±, p0, r±, r0in each q2 bin • 5 (3) q2 bins for pln (rln) • Use isospin relations: • G(B0 p-l+n) = 2G(B+ p0l+n) • G(B0 r-l+n) = 2 G(B+ r0l+n) • One free parameter for b  cln • Total: 9 free parameters

  33. 5 bins for pln 3 bins for rln Results for pln : Fitted DE and mES 76 fb-1 Sum of 396 p+ln, 137 p0ln

  34. 5 bins for pln 3 bins for rln Results for rln : Fitted DE and mES 76 fb-1 Sum of 95 r+ln, 98 r0ln

  35. B p ln Form Factors • Light-Cone Sum Rules : • Valid for q2 < 14 GeV2 • Ball/Zwicky quote 10-13% error at q2=0 • Lattice QCD : • Unquenched calculations by HPQCD, FNAL • Valid for q2 > 15 GeV2 • 11-13% error at high q2 • Useparametrizations(e.g. Becirevic-Kaidalov)to extrapolate to full q2 range • Quark models: ISGW II (no error quoted) Theo. FF uncertainties enter twice: (1) FFshape acceptance (2) FF normalization extraction of Vub

  36. ISGW II LCSR LQCD BK-Fit to BABAR data B pln a = 0.60 ± 0.15 Fitting the Form-Factor Shape Fit Becirevic-Kaidalov (BK) Parametrization to data: f(0) = norm. factor,a = shape parameter Result of BK-Fit consistent with unquenched LQCD: HPQCD’04 : a = 0.42 ± 0.07 , FNAL’04 : a = 0.62 ± 0.05

  37. Measured q2 Distribution pln • Recent LQCD and LCSR calculations agree • well with BABAR B plndata • ISGW II shows marginal agreement P*p Improved Data-MC agreement in kinematic distributions, e.g., hadron momentum. ISGW II BK Fit

  38. Current Measurements of B pln as Function of q2 M. Morii hep-ex/0505070 More data will help!

  39. d=1 68% CL Limits Experiment HPQCD FNAL a=0 Constraining Form-Factor Parametrizations • Form factor contains soft-overlap (z) and hard-scattering (H) contributions: • BK param. is approximation that neglects H •  study more general parametrization: • with B Hard gluon p u b • Richard Hill, hep-ph/0505129 : • d < 1  limits hard scattering • contribution • Simple Pole Model (only B* pole) • ruled out with 99.99% CL • Single Pole Model not ruled out • (a ∞, d1, a(1-d) fixed )

  40. rln B rln Form Factors • For vector mesons we need 3 FFs, e.g. A1,A2,V  much more difficult ! Ball & Zwicky ‘04 04 • Good agreement with all FF calculations • Statistical and systematic errors still too large to measure three FFs

  41. Theory Uncertainty: Dependence on FF shape “Do we have a model-independent BF measurement?” Effect of pln FF on BF small. Some effect of X-feed from rln Acceptance effect due to harsh kinem. cuts for rln

  42. Experimental Systematic Uncertainties

  43. Detector effects: Track, Photon, KLefficiencies and resolutions (losses, g bkg, …) Experimental Systematic Uncertainties

  44. Detector effects: Track, Photon, KLefficiencies and resolutions (losses, g bkg, …) All D* D D** Experimental Systematic Uncertainties Relative contributions to charm bkg have large uncertainties: B D, D*, D**, … BF

  45. Detector effects: Track, Photon, KLefficiencies and resolutions (losses, g bkg, …) All B uln Hybrid D* D D** M(Xu) Experimental Systematic Uncertainties Relative contributions to charm bkg have large uncertainties: B D, D*, D**, … BF

  46. Detector effects: Track, Photon, KLefficiencies and resolutions (losses, g bkg, …) All D* D D** Experimental Systematic Uncertainties Comparison off-resonance data with continuum MC. Mainly low q2 (secondary leptons) Relative contributions to charm bkg have large uncertainties: B -> D, D*, D**, … BFs B uln Hybrid M(Xu)

  47. Branching Fraction Measurements B pln B rln BABAR CLEO

  48. Consistency Check : Isospin Symmetry Measure B0, B+ separately:  Test isospin constraint Consistent with 2 within stat. error

  49. ub

  50. No FF norm. uncertainty available Extraction of Vub Extraction of |Vub| relies on FF norm. in distinct q2 regions LCSR q2 <15 GeV2, LQCD q2 > 15 GeV2or Extrapolation to whole q2 range Theory error dominates: 11-13% in restricted q2 ranges 15-17% for whole q2 range BABAR’s choice 7% exp. 4%6%

More Related