1. new measurements of sin(2β) & cos(2β) at BaBar
M. Bruinsma, UC Irvine
For the BaBar collaboration
ICHEP2004
August 16th, Beijing

2. } } Decay rates of B0 mesons f
Time-dependent rates for B0 (f+) or B0 (f-) decays to final state f:
= difference in total decay width between BH and BL
= mass difference between BH and BL (~0.50ps-1)
mixing
decay f } C } S
Neglecting ΔΓ:
If f is a CP eigenstate fCP then we have CP violation if λf ≠ ±1:
|q/p| ≠ 1 (CP violation in mixing, negligible)
|Af/Af|≠ 1 (direct CP violation, not expected here)
Im(λf) ≠ 0 (interference between mixing and decay)
August 16th 2004
M. Bruinsma

3. CP violation for ccK0 decays
B0 mixing
B0 decay
K0 mixing c b d b s d c W+ s d b d s d d
Vtd Vtb *
Vud Vub
Vcd Vcb g a β
main decay amplitudes (tree and leading penguins) have same weak phase:
no direct CP violation, no CP violation in mixing
August 16th 2004
M. Bruinsma

4. Experimental techniqueµ+ µ- π+
Fully reconstruct decay to CP state
J/ψ π- B0 KS e- e+ B0
Asymmetric energies
produce boosted Υ(4S), decaying into coherent BB pair
Δz=(βγc)Δt K- l-
Determine time between decays from vertices
Determine flavor and vertex position of other B decay
General principle:
fully reconstruct decay to CP eigenstate
use remaining tracks in event to determine initial flavor
Measure Δt from position of both vertices (CP/tag)
fit Δt distributions to determine amplitude of sin(ΔmΔt) term = sin(2β)
August 16th 2004
M. Bruinsma

5. Resolution and tagging
perfect
tagging & time resolution
realistic
mistagging & finite time resolution
Need to know mistag fraction ω and Δt resolution function R in order to measure CP asymmetry.
Use large sample (‘Bflav’) of fully reconstructed B0 decays to flavor eigenstates (‘self-tagging decays’) : S=0, C=1
August 16th 2004
M. Bruinsma

6. Measurement method Self-tagging (‘BFlav’) signal events, C=1, S=0:
+ = unmixed (opposite flavor)
- = mixed (same flavor)
Common mistag fractions
Common resolution function
CP signal events, C=0, S=ηFsin(2β) :
+ = B0 on tag side
- = B0 on tag side
External physics parameters fixed:
Δm=0.502 ± ps-1 (PDG2004)
τ(B0)=1.536 ± ps (PDG2004)
ΔΓ/Γ=0 (varied to 0.02 as systematic
|λ|=1 (floating in separate fit)
Simultaneous maximum likelihood fit:
observables Δt, σ(Δt), signal probability, tag flavor and category, Bflav flavor
65 free parameters in fit
9 parameters specific to CP sample
August 16th 2004
M. Bruinsma

7. PEP-II performance Run4 Run3 Run2 Run1
PEP-II top luminosity: 9.2 x 1033cm-2s (more than 3x design goal 3.0 x 1033)
PEP-II delivered 254 fb-1
BaBar recorded 244 fb-1
In this analysis: Run1-4 data
On peak 205 fb-1
227M BB pairs
Run4
Run3
Run2
Run1
August 16th 2004
M. Bruinsma

8. The BaBar detector e+ [3.1 GeV] e- [9 GeV] Electromagnetic Calorimeter
6580 CsI crystals
e+ ID, p0 and g reco
Instrumented Flux Return
19 layers of RPCs
m+ and KL ID
e+ [3.1 GeV]
Cherenkov Detector (DIRC)
144 quartz bars
K,p separation
Drift Chamber
40 layers
Tracking + dE/dx
Silicon Vertex Tracker
5 layers of double sided silicon strips
e- [9 GeV]
August 16th 2004
M. Bruinsma

9. Data sample – CP modes BABAR yield signal region yield signal region
MES [GeV]
MES [GeV]
signal region
yield
CP sample
NTAG
purity
ηCP
J/ψ KS (KS→π+π-)
2751
96% -1
J/ψ KS (KS→π0π0)
653
88%
ψ(2S) KS (KS→π+π-)
485
87%
χc1 KS (KS→π+π-)
194
85%
ηc KS (KS→π+π-)
287
74%
Total for ηCP=-1
4370
92%
J/ψ K*0(K*0→ KSπ0)
572
77%
+0.51
J/ψ KL
2788
56% +1
Total
7730
78%
BABAR
J/ψ KL signal
J/ψ X background
Non-J/ψ background
ΔE [MeV]
August 16th 2004
M. Bruinsma

10. Data sample – BFlav modes
yield
yield
MES [GeV]
MES [GeV]
BFlav sample
NTAG
purity
D- π+/ρ+/a1+ (6 decay modes)
32974
83.1%
D*- π+/ρ+/a1+ (12 decay modes)
35008
89.4%
J/ψ K*0(K*0→ K+π-) (2 modes)
4896
95.8%
Total
72878
August 16th 2004
M. Bruinsma

11. Comparison with 2002 Summer 2002 Summer2004
(sin(2β) = 0.741±0.067±0.034)
Data sample 88 M BB decays (run1+2) 227 M BB decays (run1-4)
2641 CP events (tagged, signal) 7730 CP events (tagged, sig.)
Flavor Tagger 4 tagging categories, Q=28.5% 6 tag categories, Q=30.5%
Changes in analysis:
More refined treatment of signal probabilities from mES spectrum
More floating parameters for the CP background:
‘apparent CP content’
fraction of prompt (continuum) background
August 16th 2004
M. Bruinsma

12. Fit results sin2β = 0.722  0.040 (stat)  0.023 (sys)
J/ψ KL (CP even) mode
(cc) KS (CP odd) modes
sin2β =  (stat)  (sys)
(2002 measurement: sin(2β) = 0.741±0.067±0.034)
August 16th 2004
M. Bruinsma

13. Consistency checks J/ψKS(π+π-) Χ2=11.7/6 d.o.f. Χ2=1.9/5 d.o.f.
Lepton tags
ηF=-1 modes
Χ2=11.7/6 d.o.f.
Prob (χ2)=7%
Χ2=1.9/5 d.o.f.
Prob (χ2)=86%
August 16th 2004
M. Bruinsma

14. |λ| : direct CP violation
no direct CP violation expected in Standard Model for (cc)K0 decays
in nominal fit, assumed |λ|=1
results from fit (using (cc) KS modes) with floating |λ| and Im(λ):
|λ|=0.950 ± (stat.) ± 0.013
Little effect on sin(2β) measurement
change in S of w.r.t. |λ|=1
-2% correlation between coefficient of sin term and cos term
Fit projection on the Δt distributions with floating |λ|. The dashed line are the results from the nominal fit with |λ|=1
August 16th 2004
M. Bruinsma

15. Systematic uncertainty
s(sin2b) s(|λ|))
Description of background events
CP content of peaking background
Background shape uncertainties
Mistag differences between BCP and Bflav samples
Composition and content of J/y KL background N/A
Dt resolution and detector effects
Silicon detector alignment uncertainty
Dt resolution model
Beam spot position
Fixed Δm, τ, ΔΓ/Γ, |λ|
Tag-side interference/ DCSD decays
MC statistics/bias
TOTAL
Steadily reducing systematic error: July 2002 = 0.033
(better and more Monte Carlo) July 2001 = 0.05
August 16th 2004
M. Bruinsma

16. sin(2β) & the unitarity triangle
CKM fit without sin(2β) measurement
CKM fit with sin(2β) measurement
cos(2β)<0
cos(2β)>0
Measured value of sin(2β) in excellent agreement with Standard Model expectation
Strong constraint on apex of Unitarity Triangle
August 16th 2004
M. Bruinsma

17. cos(2β) with B0→J/ψK*0(KSπ0)
J/ψK*0 (KSπ0) final state can be ηF=+1 or ηF=-1, depending on L=0,1,2
ηF = +1 for L=0,2
ηF = -1 for L=1
In nominal fit for sin2β we used ‘averaged’ CP content of +0.51
Full angular analysis allows for the separation of CP even and CP odd and amplitudes
Many extra terms in time-dependent decay rate, two proportional to cos(2β) :
angular amplitudes
decay angles:
angular amplitudes in transversity basis
August 16th 2004
M. Bruinsma

18. Phases and amplitudes discrete ambiguity in strong phases
Angular amplitudes and strong phases, measured from B±→J/ψK*± and B0→J/ψK*0 (K+π-) using time-integrated angular analysis:
“solution 1”
“solution 2” OR
discrete ambiguity in strong phases
=> sign ambiguity in (±)cos(2β)
August 16th 2004
M. Bruinsma

19. Breaking the ambiguity
Central idea: include (Kπ) S-wave in angular analysis
Extra terms due to interference of S-wave and P-wave contributions to Kπ final state:
S-wave amplitude & phase
δS –δ0 changes sign between solution 1 and solution 2
but near K*(890):
(consequence of Wigner causality)
only ‘physical’ solution is the one for which δS –δ0 decreases with Kπ mass near the K* resonance
August 16th 2004
M. Bruinsma

20. Breaking the ambiguity
○ solution 1: unphysical solution
“solution 1”
BABAR
“solution 2” OR
● solution 2: physical solution
○ LASS data
August 16th 2004
M. Bruinsma

21. (with floating sin(2β))
cos(2β) results
cos(2β) from 104 tagged B0→J/y(KSp0)*0 decays:
distribution of cos(2β) results from a set of 2000 data-sized Monte Carlo samples, generated with cos(2β)=0.68 h
(with floating sin(2β))
(with sin(2β) fixed to 0.731)
Estimation of confidence level excluding cos(2β)=-0.68 determined from distribution of 2000 Monte Carlo experiments:
August 16th 2004
M. Bruinsma

22. Conclusions New measurement of sin(2β) on 2.6 times larger data sample
sin(2β) = ± (stat.) ± (syst.)
(2002 measurement: sin(2β) = ± (stat.) ± (syst.))
Reduced systematic uncertainty
Excellent agreement with Standard Model prediction
No significant direct CP violation in (cc)K0 decays:
|λ| = ± (stat.) ± (syst.)
Time-dependent angular analysis of J/ψK*0 gives cos(2β)>0 at 86.6% CL
new method to resolve strong phase ambiguity
sign of cos(2β) in agreement with Standard Model expectations
August 16th 2004
M. Bruinsma

23. Summary of Δt-fit parameters
(Backup)
Summary of Δt-fit parameters
Observables:
Δt, σ(Δt), tag flavor, tag category (all), reco flavor (BFlav only)
Free parameters in Δt-fit: 65
sin(2β): 1 (CP sample only)
mistag fractions for B0 and B0 in six tag categories: 12 (common)
signal resolution function R(σΔt): 7 (common)
tagging and reconstruction efficiency ratios: 7 (common)
background-related (determined from sideband in mass spectrum):
background mistag fractions: 24 (BFlav only)
background resolution function: 3 (common)
‘apparent CP’ and composition of CP background: 8 (CP only)
background composition and lifetime of BFlav modes: 3 (BFlav only)
External physics parameters fixed:
Δm=0.502 ± ps-1
τ(B0)=1.536 ± ps
ΔΓ/Γ=0 (varied to 2% as systematic)
|λ|=1 (floating in separate fit) }
PDG2004
August 16th 2004
M. Bruinsma

24. (Backup) Tagging performance Tag cat Eff.(%) ω(%) Δω(%) Q(%) Lepton
8.6±0.1
3.2±0.4
-0.2±0.8
7.5±0.2
Kaon I
10.9±0.1
4.6±0.5
-0.7±0.9
9.0±0.2
Kaon II
17.1±0.1
15.6±0.5
-0.7±0.8
8.1±0.2
Kaon-Pion
13.7±0.1
23.7±0.6
-0.4±1.0
3.8±0.2
Pions
14.5±0.1
33.0±0.6
5.1±1.0
1.7±0.1
Other
10.0±0.1
41.1±0.8
2.4±1.2
0.3±0.1
All
74.9±0.2
30.5±0.4
August 16th 2004
M. Bruinsma

25. (Backup) Control sample fits sin(2β)=0
Fit projections on the Δt distributions of the J/ψK+ control sample
sin(2β)=0
August 16th 2004
M. Bruinsma

26. (background subtracted)
(Backup)
Kπ mass spectrum
BABAR
L = 82fb-1
Kp mass from inclusive
B0→J/yK+p- (+c.c.)
(background subtracted)
P-wave
S-wave
August 16th 2004
M. Bruinsma

27. (Backup)
Sub-mode fits
August 16th 2004
M. Bruinsma

28. (Backup)
KL fits
August 16th 2004
M. Bruinsma

29. (Backup)
CP=-1 modes per tag cat
August 16th 2004
M. Bruinsma

30. (Backup)
Best or the rest
August 16th 2004
M. Bruinsma