DPF 2009 Richard Kass 1 Search for b → u transitions in the decays B → D (*) K - using the ADS method at BaBar Outline of Talk *Introduction/ADS method *Analysis technique *Preliminary Results *Summary & Conclusions Richard Kass for the BaBar Collaboration

DPF 2009 Richard Kass 2 Atwood, Dunietz, Soni, PRL 78, 3257 (1997) & PRD 63, (2001) Use B decays that can reach the same final state via two different decay amplitudes Amount of interference depends on CKM angle  ADS idea: use D 0 / D 0 Flavor eigenstates, e.g. B - → DK - → [K + π - ] D K - Also can use D*K - and D ( * ) K* - The ADS Method-I color & CKM favored  color & CKM suppressed doubly Cabibbo suppressed Cabibbo favored [K + π - ] D K -

DPF 2009 Richard Kass 3 The ADS Method-II The amplitude for B - → DK - → [K + π - ] D K - can be written as: D strong phase B strong phase CKM Can form two observables: R ADS & A ADS news: theoretically clean way to measure . Do not have to measure time dependence & asymmetry expected to be large  news: 3 unknowns (r B, δ B, γ ) but 2 observables (R, A) per decay mode but add D*K & now have 5 unknowns & 6 observables  news: Rates are expected to be small ~10 -7 B(B - → D 0 K - )xB(D 0 → K + π - )~(3.7×10 −4 )( 1.3×10 −4 )~4.8×10 −8 r D =(5.78±0.08)% HFAG

DPF 2009 Richard Kass 4 PEP-II at SLAC asymmetric e + e − collider: 9 GeV (e - )/3.1 GeV (e + ) PEP-II Peak Luminosity 1.2 x cm -2 s -1 BaBar recorded 426 fb -1 at Y(4S) asymmetric e + e − collider: 9 GeV (e - )/3.1 GeV (e + ) PEP-II Peak Luminosity 1.2 x cm -2 s -1 BaBar recorded 426 fb -1 at Y(4S) 4.67x 10 8 Y(4S) → B B events

DPF 2009 Richard Kass T Solenoid Electromagnetic Calorimeter (EMC) Detector of Internally Recflected Cherenkov Light (DIRC) Instrumented Flux Return (IFR) Silicon Vertex Tracker (SVT) Drift Chamber (DCH) e - (9 GeV) e + (3.1 GeV) BaBar Detector SVT, DCH: charged particle tracking: vertex & mom. resolution, K 0 s/Λ EMC: electromagnetic calorimeter:  /e/π 0 /η DIRC, IFR, DCH: charged particle ID: π/μ/K/p Highly efficient trigger for B mesons

DPF 2009 Richard Kass 6 ADS Analysis Strategy Study the following decays: B - → DK - D → K + π - & D → K - π + B - → D*K - D* →  D, D → K + π - & D → K - π + B - → D*K - D* → π 0 D, D → K + π - & D → K - π + Notation: “ADS”: suppressed decays, e.g. K’s have opposite sign “CAB”: Cabibbo favored decays, Dπ and DK decays e.g. K’s have same sign B - → Dπ - D → K + π - & D → K - π + B - → D*π - D* →  D, D → K + π - & D → K - π + B - → D*π - D* → π 0 D, D → K + π - & D → K - π + Very useful check of technique due to large rate (>10x DK) CPV expected to be small in these modes

DPF 2009 Richard Kass 7 Threshold kinematics: we know the initial energy (E* beam ) of the Y(4S) system Therefore we know the energy & magnitude of momentum of each B Background (spherical) (jet-structure) Signal ADS Analysis Techniques Two main sources of backgrounds: B’s & charm from c c events Event topology

DPF 2009 Richard Kass 8 ADS Backgrounds Continuum Charm Production e + e - → c c combine a D 0 → K + π - with a K - from rest of event Use a Neural Net to suppress continuum backgrounds Inputs to NN include: ● event shape variables (Legendre moments, Thrust, B meson polar angle in CM) ● B tagging variables (hemisphere charge, kaon charge sum in ROE, kaon-lepton mass, Δ t between 2 B’s in event) DK signal enriched CAB <m ES < off peak data vs udsc MC DK ADSDK CAB

DPF 2009 Richard Kass 9 ADS Backgrounds B Mesons Peaking backgrounds from B mesons; for B - → DK - : B - → Dπ - with π mis-ID as a K (BR(Dπ - )/BR(DK - )~13) B - → DK - with D → K + K - and K mis-ID as a π (BR(D → K + K - )/BR(D → K + π - )~31) Charmless B decays B - → K + π - K +, B - → K - π + K - Eliminate backgrounds from B mesons using: tight particle ID swap K-π hypotheses & veto if |m(K - π + )-m(D)|<40MeV veto K + K - if |m(K + K - )-m(D)|<40MeV MC estimate of B meson backgrounds (r B =0.1, cos  cos  =0)

DPF 2009 Richard Kass 10 ADS Likelihood Fit Extract parameters of interest using an unbinned extended maximum likelihood fit in m ES and NN distribution. There are 8 components to the ML fit: ADS signal: gaussian X NN(B meson) CAB signal: gaussian X NN(B meson) ADS background: ARGUS X NN(ADS udsc) CAB background: ARGUS X NN(CAB udsc) ADS non-peaking B background: ARGUS X NN(B meson) ADS peaking B background: gaussian X NN(B meson), FIXED to MC CAB non-peaking B background: ARGUS X NN(B meson) CAB peaking B background: gaussian X NN(B meson), FIXED to MC ALSO fit for the mean & σ of gaussian and the 2 ARGUS parameters The fit is done individually for six modes: Dπ, D* → (Dπ 0 )π, D* → (D  )π DK, D* → (Dπ 0 )K, D* → (D  )K The fit is done separately for B -, B + & combined charges

DPF 2009 Richard Kass 11 ADS Analysis Details Summary of D ( * ) K Systematic Errors Summary of Selection Efficiencies ε ADS /ε CAB ≠1 due to slightly different PID cuts at an early stage of analysis Table for D ( * ) π in “extra slides”

DPF 2009 Richard Kass 12 B - → D (*) π - Results NN> <m ES < Dπ ADS Dπ CAB ModeN ADS N CAB Rx10 -3 Ax10 -2 DπDπ79.8± ± ± 0.6 ± 0.43 ± 17 ± 4 D*π → (Dπ 0 )π28.7 ± ± ± 0.9 ± ± 27 ± 5 D*π → (D  )π18.7 ± ± ±1.4± ± 55± 22 Expect R ≈r 2 D (world average: r 2 D =(3.36 ± 0.08)x10 -3 ) B B bkgds continuum bkgds fit result

DPF 2009 Richard Kass 13 B → DK Results All m ES plots have NN > 0.94 ModeN ADS N CAB Rx10 -2 A DK23.9 ± ± ± 0.55 ± ± 0.35 DK CAB DK ADS B+B+ B-B Significance of R 2.9σ (stat) 2.6σ (stat+syst) B B bkgds continuum bkgds fit result

DPF 2009 Richard Kass 14 B → D*(Dπ 0 )K Results All m ES plots have NN > 0.94 ModeN ADS N CAB R*x10 -2 A* D*(Dπ 0 )K10.3 ± ± ± 0.93 ± ± 0.35± 0.12 D*K ADS B+B+ B-B- D*K CAB Significance of R 2.4σ (stat) 2.2σ (stat+syst) B B bkgds continuum bkgds fit result

DPF 2009 Richard Kass 15 B → D*(D  )K Results All m ES plots have NN > 0.94 ModeN ADS N CAB R*x10 -2 A* D*(D  )K5.9 ± ±291.3 ± 1.4 ± ± D*K CAB D*K ADS B+B+ B-B- B B bkgds continuum bkgds fit result

DPF 2009 Richard Kass 16 r B,DK DK D*K combined r B,DK < 90% CL r B,D*K < 90% CL Extraction of r B, DK & r B,D*K We use a frequentist (“CKMfitter”) approach to determine confidence intervals for the CPV parameters. r D and δ are fixed to HFAG values r B,D*K Very little sensitivity to  All values allowed at 1 σ level. preliminary

DPF 2009 Richard Kass 17 Extraction of δ B δ B,DK δ B,D*K With input from other BaBar analysis (Dalitz*) and/or CKMfitter can resolve ambiguity in δ * PRD (2008) 2D confidence intervals using  = 76 0 from BaBar Dalitz analysis preliminary 2σ 1σ preliminary Exclude [-180, -28] 0 & [164, 180] 95%CL Exclude [-18, 155] 95%CL

DPF 2009 Richard Kass 18 Summary & Conclusions ●Analysis uses full BaBar data set 2X data as previous BaBar analysis, PRD (2005) ●Test technique using D ( * ) π ●New preliminary measurements of r B,DK & r B,D*K ●ADS CP asymmetry for DK & D*K may be very large Both DK and D*K have asymmetries ~70% (but with large uncertainties) ●Resolve ambiguity in strong phases, δ B,DK, δ B,D*K good agreement with δ B,DK, δ B,D*K from BaBar Dalitz analysis Promising  analysis for LHCb & Super-B factory need super-sized data sample to overcome our limited statistics Consistent with BaBar’s Dalitz analysis (PRD (2008)): r B, DK =(8.6 ± 3.5)% and r B, D*K =(13.5 ± 5.1)%

DPF 2009 Richard Kass 19 Extra slides

DPF 2009 Richard Kass 20 BaBar DIRC BaBar K/  ID D * + → D 0  + D 0 → K +  -

DPF 2009 Richard Kass 21 ADS Analysis Details Summary of D ( * ) π Systematic Errors

DPF 2009 Richard Kass 22 R ADS Significance 2.9σ (stat) 2.6σ (stat+syst) preliminary 2.4σ (stat) 2.2σ (stat+syst) preliminary