Max Baak, NIKHEF on behalf of the BABAR and BELLE Collaborations Measurements of at BABAR and BELLE Max Baak, NIKHEF on behalf of the BABAR and BELLE Collaborations Beauty 2005, Assisi
Outline Measurements of using BD(*)K(*) GLW Method ADS Method D0 Dalitz Method (GGSZ) Measurements of sin(2+) using B0D(*) / Outlook
in the Unitarity Triangle (,) CKM Unitarity Triangle (0,0) (1,0) Expect ≈ (60 ± 6)° from SM fit to: sin2, |Vub/Vcb|, md, ms, k Most challenging angle to measure experimentally: Low branching fractions Low reconstruction efficiencies Small interferences The only solution with is statistics
from B- D(*) K- K– D0 B– D0 B– K– Access via interference between B- D(*)0 K- and B- D(*)0 K-. Color-allowed b c amplitude Color-suppressed b u amplitude u u K– D0 f Vus* s Vcs* W – b c b c amplitude ratio rB relative weak phase and strong phase B W – Vub s B– D0 f B– K– Vcb u u Reconstruct D in final state f accessible both to D0 and D0. Will discuss 3 methods with different final states f in this talk: GLW Gronau – London – Wyler 2. ADS Atwood – Dunietz – Soni 3. GGSZ Giri – Grossman – Soffer – Zupan Belle collaboration Critical parameter rB~ 0.1 for sensitivity to ! In order to determine rB, , B simultaneously, need to measure as many D(*)0 modes as possible.
Preface: Analysis Techniques 1. B-meson identification 2. Combinatoric e+e– qq bkg suppression mES E = EB*–E*beam data MC E Event topological variables combined in Neural Network or Fisher discriminant 3. K/ separation (Cherenkov angle / TOF) 4. Time-dependent measurements (only B0/B0) – K– BaBar Excellent separation between 1.5 and 4 GeV/c D0 B0 KS tag side e+ e+ B0 lepton Coherent B0B0 production from (4S) boost ≈ 0.55/0.42 allows t measurement
GLW Method Reconstruct D meson in CP-eigenstates (accessible to D0 and D0) Theoretically very clean (“golden mode”) to determine Relatively large BFs (10-5), small CP asymmetry 3 Independent measurements (A+R+ = -A-R-) and 3 unknowns (rB, , B) CP even modes: K+K-, +- CP odd modes: KS0, KS, KS, KS Phys. Lett. B253, 483 (1991); Phys. Lett. B265, 172 (1991); Phys. Lett. B557, 198 (2003)
GLW Method Results PRL92,202002, 214M BB B-CONF-0443, 275M BB 95 15 events 76 13 events 114 21 events 167 21 events B-CONF-0443, 275M BB
GLW Results Combined D0CP K− BaBar Belle Average (HFAG) RCP+ 0.87 ± 0.14 ± 0.06 0.98 ± 0.18 ± 0.10 0.91 ± 0.12 RCP− 0.80 ± 0.14 ± 0.08 1.29 ± 0.16 ± 0.08 1.02 ± 0.12 ACP+ +0.40 ± 0.15 ± 0.08 +0.07 ± 0.14 ± 0.06 +0.22 ± 0.11 ACP− +0.21 ± 0.17 ± 0.07 –0.11 ± 0.14 ± 0.05 +0.02 ± 0.12 PRL92,202002, 214M BB B-CONF-0443, 275M BB D*0CP K− (D*D0CP0) BaBar Belle Average (HFAG) RCP+ +1.06 ± 0.26 + 0.10 1.43 ± 0.28 ± 0.06 1.24 ± 0.20 RCP− 0.94 ± 0.28 ± 0.06 0.94 ± 0.29 ACP+ –0.10 ± 0.23 + 0.03 –0.27 ± 0.25 ± 0.04 –0.18 ± 0.17 ACP− +0.26 ± 0.26 ± 0.03 +0.26 ± 0.26 PRD71,031102, 123 M BB B-CONF-0443, 275M BB − 0.09 − 0.04 No useful constraints on yet due to small branching ratios and limited statistics. D0CP K*− (K*- KS−) BaBar Average (HFAG) RCP+ 1.77 ± 0.37 ± 0.12 1.77 ± 0.39 RCP− 0.76 ± 0.29 ± 0.06 + 0.04 (*) Belle 0.76 + 0.30 ACP+ –0.09 ± 0.20 ± 0.06 –0.02 ± 0.33 ± 0.07 –0.07 ± 0.18 ACP− –0.33 ± 0.34 ± 0.10 ± (*) 0.19 ± 0.50 ± 0.04 –0.16 ± 0.29 hep-ex/0408069, 227M BB hep-ex/0307074, 96M BB − 0.14 − 0.33 (0.15±0.10) x (ACP--ACP+) (*) CP-even pollution in the CP-odd channels
ADS Method B– D0 K– B– D0 K– K+– K+– Reconstruct D in final state K - small BF (10-6) Amplitude: Amplitudes comparable in size large CP violation Count B candidates with opposite sign kaons! Phys. Rev. Lett. 78, 3257 (1997) suppressed favored B– D0 K– B– D0 K– K+– K+– interference favored suppressed PDG, Phys.Lett. B592, 1 (2004) D : D decay strong phase unknown. Scan over all values. 2 observables vs 3 unknowns: rB, , B
ADS Method Results BABAR: 227M BB BELLE: 275M BB −3.2 −5.3 −0.009 preliminary 275 M BB PRL 94, 091601 B+ D0K+ B+ D0K+ N = 4.7 +4.0 N = 8.5 +6.0 (2.3) −3.2 −5.3 RADS = 0.013 +0.011 RADS = 0.023 +0.016 ± 0.001 −0.009 −0.014 AADS = +0.88 +0.77 ± 0.06 –0.62 preliminary B+ [D00]D* K+ BELLE: 275M BB N = −0.2 +1.3 −0.8 RADS = −0.002 +0.010 −0.006 No signal observed! B+ [D0]D* K+ preliminary N = 1.2 +2.1 −1.4 RADS = 0.011 +0.018 −0.013 hep-ex/0408028 -0.1 0.1 E (GeV)
The smallness of rB makes the extraction of with GLW/ADS difficult! ADS Method Results 0 < D < 2 rd ± 1 48° < < 73° same, any RADS # Events 227 M BB hep-ex/0408028 Belle (90% CL) B+ D0K+ RADS = 0.013 +0.011 −0.009 275 M BB PRL 94, 091601 BaBar (90% CL) B+ D0K+ RADS = 0.023 +0.016 ± 0.001 −0.014 DK: rB < 0.23 D*K: r*B2 < (0.16)2 DK: rB < 0.27 @ 90% C.L. @ 90% C.L. BaBar RADS 1 The smallness of rB makes the extraction of with GLW/ADS difficult!
GGSZ Method K– D0 B– B– K– D0 Phys. Rev. D68, 054018 (2003) Color-allowed b c amplitude Color-suppressed b u amplitude KS u u K– D0 Vus* + s Vcs* - W – b c b c KS interference W – Vub s B– B– K– Vcb D0 + u - u Reconstruct D in final state: KS +- (not a CP-eigenstate) Employs K-K mixing (“cheap” decay-mode: high BF ~2.2x10-5 ) Final state accessible through many intermediate non-CP states. Need Dalitz analysis to separate resonance interferences!
GGSZ Method 2 ± = D0 D0 - + KS KS + - D decay amplitude f consists of sum of many resonances (more on next slide). Amplitude f parameterized in terms of Dalitz variables m+2 and m-2 Decay rates of B+ and B- written as: u u - + d d W W c s c s KS KS D0 D0 + - u u 2 ± = Simultaneous fit to D KS +- Dalitz planes of B+ and B- to extract rB, , and
Isobar formalism, no D mixing, no CPV in D decays D0 KS + - Dalitz Model To extract rB and need high-precision D decay model f (m+2, m-2) Obtain f (m+2, m-2) using fit to “tagged” D0 sample: Use large D*+ D0+ sample. Charge of the pion gives flavor of D. Isobar formalism, no D mixing, no CPV in D decays D*+ D0+, 81.5k events from 91 fb-1, purity 97% hep-ex/0504039 2 = 3824/ 3054 =1.25 K*(892) DCS K*(892) 0(770) 0(770) DCS K*(892) 13 resonances (2 ), 3 DCS partners, 1 non-resonant component
D0 KS + - Dalitz Model Belle: indentical approach hep-ex/0411049 Belle: indentical approach Include two more DCS resonances: K*+(1410) - , K*+(1680)- 13 resonances (2 ), 5 DCS partners, 1 non-resonant component D*+ D0+, 186.9k events, purity 97% K*(892) DCS K*(892) m+2 (GeV2/c4) m-2 (GeV2/c4) 0(770) m-2 (GeV2/c4) 2 =2543/1106 =2.30 m2 (GeV2/c4) m+2 (GeV2/c4)
Dalitz sensitivity scan to CA: Cabibbo Allowed DCS: Doubly-Cabibbo Suppressed CS: Color Suppressed Sensitivity to (MC) =75°, =180°, rB =0.125 d2 ln L/d2 CA: D0K(892)*-+ CS: D0KS0(770) DCS: K0(1430)*+ - DCS: D0K(892)*+- D0 CA: K0(1430)*+ -
GGSZ Method Results The two plots would be the same without CP violation. Are they?
BaBar GGSZ Method Results hep-ex/0504039 DK D*0(D00)K D*0(D0)K Mode Signal (events) B+D0K+ 282 ± 20 B+D*0K+ (D*0D00) 90 ± 11 B+D*0K+ (D*0D0) 44 ± 8 m-2 B+ B+ B+ m+2 m-2 BABAR: 227M BB m+2 m-2 B– B– m-2 DCS K*(892) B– m+2 m+2
Belle GGSZ Method Results hep-ex/0411049 DCS K*(892) - thick black line: with interference - thin grey line: without interference BELLE: 275M BB Mode Signal (events) Bkg.frac. (%) BD0K− 209 ± 16 25 ± 2 BD*0K− 58 ± 8 13 ± 2 BD0K*− 36 ± 7 27 ± 5 hep-ex/0504013
BaBar GGSZ Method Results 68% 95% Frequentist CLs BABAR: 227M BB DK hep-ex/0504039 preliminary Frequentist CLs DK : rB = 0.118 ± 0.079 ± 0.034 +0.036 –0.034 B = ( 104 ± 45 )° +17 +16 –21 –24 D*K : rB* = 0.169 ± 0.096 +0.030 +0.029 –0.028 –0.026 D*K B* = ( 296 ± 41 ± 15 )° +14 –12 = ( 70 ± 31 ) ° +12 +14 –10 –11 stat. syst. Dalitz
Belle GGSZ Method Results rB hep-ex/0411049 BELLE: 275M BB DK hep-ex/0504013 B (deg) Frequentist CLs DK : DK rB = 0.21 ± 0.08 ± 0.03 ± 0.04 B = ( 157 ± 19 ± 11 ± 21 )° = ( 64 ± 19 ± 13 ± 11 )° rB[D*] D*K D*K : B[D*] (deg) rB* = 0.12 +0.16 ± 0.02 ± 0.04 -0.11 B* = ( 321 ± 57 ± 11 ± 21 )° D*K Promising results! = ( 75 ± 57 ± 11 ± 11 )° DK* : rB(K*) = 0.25 +0.17 ±0.09 ±0.04 ±0.08 (*) -0.18 rB[K*] DK* B(K*) = ( 353 ±35 ±8 ±21 ±49 )° B[K*] (deg) (*) = ( 112 ±35 ±9 ±11 ±8)° (*) DK* Combined result of DK and D*K: = ( 68 +14 13 11 ) ° -15 stat. syst. Dalitz (*) Possible bias caused by a contribution from non-resonant B–→ DKS–. (degrees) (degrees)
rB(*) World Average BaBar ADS limit pushing rB down. 0 < D < 2 rd ± 1 48° < < 73° same, any [ no improved constraint when adding from CKM fit ] RADS Frequentist CLs Belle ADS (90% CL) BaBar ADS (90% CL) Belle Dalitz BaBar Dalitz Using GLW, ADS, GGSZ results Bayesian CLs rB [BDK] 68% 95% BaBar ADS limit pushing rB down. Belle Dalitz value (0.21) relatively large. 0.10 0.04 0.09 0.04
CP violation in B0 D(*) / CP violation through B0-B0 mixing and interference of amplitudes: CP violation proportional to ratio r of amplitudes Small: r |V*ubVcd / VcbV*ud| 0.020 Large BF’s, at level of 1% No penguin pollution theoretically clean Relative weak phase from bu transition Relative strong phase Suppressed amplitude through b u transition Favored amplitude u,c,t u,c,t Strong phase difference CKM Unitarity Triangle g
sin(2+) from B0 D(*) / Time evolution for B0 decays and B0 decays (Rmix) to D(*)/: CP asymmetry: small sine terms Need S+ and S- together to give (2+) and From D(*)/ sine coefficients, 4 ambiguities in (2+) Express result as |sin(2+)| SM: sin(2+) ~ 1 Factorization theory: is small SMALL sine terms
sin(2+) Caveat: determination of r(*) Simultaneous determination of sin(2+) and r(*) from time-evolution not possible with current statistics need r(*) as external inputs ! Estimate r(*) from B0 Ds(*)+-/- using SU(3) symmetry [1] Using: [1] I. Dunietz, Phys. Lett. B 427, 179 (1998) SU(3) [2] Inputs used in CKMFitter/ UTFit : r(D) = 0.019 ± 0.004 r(D*) = 0.015 ± 0.006 r(D) = 0.003 ± 0.006 We add 30% theoretical errors to account for: Unknown SU(3) breaking uncertainty Missing W-exchange diagrams in calculation Missing rescattering diagrams (Can be estimated with B0Ds(*)+K-) no theoretical errors included [2] fD : decay constants
BaBar: Inclusive B0 D* BABAR: 227M BB Using a,b,c parametrization: D* partial reconstruction: fast Tag side interference: r’, ’ are the ratio and phase difference between the bu and bc amplitudes in the Btag decay. r’0 in lepton tags. PRD68, 034010 lepton tags - High statistics! - Large backgrounds preliminary hep-ex/0504035 preliminary lepton tags peaking D* kaon tags combinatoric BB other peaking BB continuum 18710 ± 270 lepton tags 70580 ± 660 kaon tags
Belle: Inclusive B0 D* hep-ex/0408106 preliminary BELLE: 152M BB Belle: only uses lepton tags (no tag-side interference) sum signal bkg. Same Flavor: mixed events Opposite Flavor: unmixed events 8322 signal lepton tags
BaBar: Exclusive B0 D(*) / BABAR: 110M BB Phys.Rev.Lett. 92:251801(2004), 88 M BB Exclusive reconstruction of channels: - B D - B D* - B D - Full reco.: ~10x less efficient; far lower backgrounds - Same sensitivity to sin(2+) as inclusive approach lepton tags hep-ex/0408059 preliminary
Belle: Exclusive B0 D(*) PRL 93 (2004) 031802; Erratum-ibid. 93 (2004) 059901 BELLE: 152M BB Exclusive reconstruction of channels: - B D - B D* Uses B D*l as control sample for tag-side interference cleanest tags (*) After tagging and vertexing
No clear CP violation yet! HFAG on |sin(2+)| HFAG Averages: No clear CP violation yet!
Combined Limit on |sin(2+)| Bayesian CLs www.utfit.org 68% 95% Combined limit on |sin(2+)| : Assuming 30% error on r(*) for SU(3) breaking: CKMFitter: |sin(2+)| > 0.53 @ 68% C.L. UTFit: |sin(2+)| > 0.74 @ 68% C.L. Frequentist CLs
Outlook Many approaches to measure have been investigated by BaBar and Belle. GLW and ADS methods don't provide strong constraints on when considered alone. Current experimental results favour small values of rB. GGSZ results are promising! GLW+ADS+GGSZ: CKMFitter: = [ 63 +15 ]°+ n UTFit: = [ 64 18 ]°+ n sin(2+) from D(*)/: CKMFitter: |sin(2+)| > 0.53 @ 68% C.L. UTFit: |sin(2+)| > 0.74 @ 68% C.L. GLW+ADS+GGSZ+sin(2+): CKMFitter: = [ 70 +12 ]°+ n All results are in good agreement with the global CKM fit ( = [ 60 6 ]°) All decay modes can use lots more statistics! High statistics expected in next years may allow BaBar and Belle to measure to < 10°. -13 Using GLW, ADS, GGSZ results (deg) = 64 ± 18 ([30,100] @ 95% CL) -14
B A C K U P slides ...
BaBar: Removing the Imaginary (?)
Belle GGSZ: Systematic Errors
B D*-+ time-dependent evolution a) unmixed B0(t) D*-+ B0 Initial state Flavor eigenstate b) mixed B0 With bu transition No bu transition B0(t) B0 Initial state Flavor eigenstate D*-+ b) a) b) a) B D*-+ pure cosine: r = 0 - plus sine term, 5x the expected size in data r = 0.1, = 0 sin(2+) = 1 B D*-+ - pure cosine: r = 0 CP asymmetry: small additional sine term Smallness of amplitude ratio r greatly reduces sensitivity to sin(2+)
() Dependency on rB BaBar and Belle show quite different sensitivities to Both find quite different values for rB (BaBar: ~0.12, Belle: ~0.21) Different sensitivity to caused by dependency on rB . Toy MC Studies Comparing only results of BDK stat. Sensitivity to very dependent on critical parameter rB (~0.1)! BaBar Belle rB [DK]