Observation of CP violation in the B0 meson system

1
Observation of CP violation in the B0 meson system

• David MacFarlane
• University of California at San Diego
• (representing the BABAR Collaboration)
• Snowmass, July 12th 2001

See SLAC-PUB-8904, hep-ex/0107013, submitted to
PRL July 5
2
USA 35/276 California Institute of
Technology UC, Irvine UC, Los Angeles UC, San
Diego UC, Santa Barbara UC, Santa Cruz U of
Cincinnati U of Colorado Colorado State Florida
AM U of Iowa Iowa State U LBNL LLNL U of
Louisville U of Maryland U of Massachusetts,
Amherst MIT U of Mississippi Mount Holyoke
College Northern Kentucky U U of Notre
Dame ORNL/Y-12 U of Oregon U of
Pennsylvania Prairie View AM Princeton SLAC U of
South Carolina Stanford U U of Tennessee U of
Texas at Dallas Vanderbilt U of Wisconsin Yale
Italy 12/89 INFN and U Bari INFN and U
Ferrara Lab. Nazionali di Frascati dell'
INFN INFN and U Genova INFN and U Milano INFN
and U Napoli INFN and U Padova INFN and U
Pavia INFN, SNS and U Pisa INFN, Roma and U "La
Sapienza" INFN and U Torino INFN and U
Trieste Norway 1/3 U of Bergen Russia 1/13 Bud
ker Institute, Novosibirsk United Kingdom
10/80 U of Birmingham U of Bristol Brunel
University U of Edinburgh U of Liverpool Imperial
College Queen Mary Westfield College Royal
Holloway, University of London U of
Manchester Rutherford Appleton Laboratory
The BABAR Collaboration 9 Countries 72
Institutions 554 Physicists
Canada 4/16 U of British Columbia McGill U U de
Montréal U of Victoria China 1/6 Inst. of High
Energy Physics, Beijing France 5/50 LAPP,
Annecy LAL Orsay LPNHE des Universités Paris
6/7 Ecole Polytechnique CEA, DAPNIA,
CE-Saclay Germany 3/21 U Rostock Ruhr U
Bochum Technische U Dresden
3
BABAR Detector
EMC 6580 CsI(Tl) crystals
e (3.1GeV)
1.5T solenoid
DIRC (PID) 144 quartz bars 11000 PMs
Drift Chamber 40 stereo layers
e- (9GeV)
Silicon Vertex Tracker 5 layers, double sided
strips
Instrumented Flux Return iron / RPCs (muon /
neutral hadrons)

• SVT B vertex z resolution 70 microns
• Tracking ?(pT)/pT 0.13 ? pT ? 0.45
• DIRC K-? separation gt 3.4? for P lt
  3.5GeV/c
• EMC ?E/E 1.33?E-1/4 ? 2.1

4
ARGUS, 1987

• Mixed full event and dilepton studies demonstrate
  mixing
• Integrated luminosity 1983-87
• 103 pb-1

Seeds sown at Snowmass 1988 for asymmetric-energy
B Factories
5
B Factories, 2001
This analysis
CESR / CLEO
PEP-II / BABAR
KEK-B / BELLE
Key factor in todays results High luminosity
and Factory running at PEP-II and KEK-B
6
Origin of CP Violation
CP violation results from interference between
decays with and without mixing
mixing
decay
amplitude ratio
CP eigenvalue
Time-dependent CP Observable
cosine term
sine term
7
Golden Channel B0 J/y K0S/L
K0 mixing is required
Single weak phase no direct CPV
8
Unitarity Triangle
Bd mixing ? ?md b? ul? ? Vub , Dl ? ? Vcb Bs
mixing ? ?ms / ?md Kaon decays ? ?K

• Present experimental knowledge
• (with a range of theor. inputs)

Following Höcker et al, hep-ex/0104062
9
Experimental Technique
ExclusiveB Meson Reconstruction
B-Flavor Tagging
10
Effects of CP Asymmetry
Perfect Experiment, sin2b 0.59
Looking for a different Dt spectrum in events
where the tag is a B0 or a B0
Visible asymmetry ACP
Note integrated asymmetry is 0
11
CP Analysis Time Distributions
realistic mis-tagging probability finite time
resolution
perfect flavor tagging time resolution
Mixing CP Time evolution
12
Recent Developments

• Our first publication was in February 2001
• Sin2b 0.34 0.20 (stat) 0.05 (syst)
• Belle also reported a measurement at the same
  time
• Sin2b 0.58 (stat) (syst)
• What has changed since then?
• BABAR has added more data 23 million BB pairs
  becomes 32 million now
• Improved reconstruction efficiency
• Improved J/YKL selection
• New modes added cc1KS, J/YK0
• Improved resolution of vertex reconstruction for
  both tag and reconstructed B

PRL 86 (2001) 2515
PRL 86 (2001) 2509
Analysis was re-blinded
13
Improved Reconstruction
Y(2S) Ks(pp-)
ccKs(pp-)
J/Y K0(Kp-)
J/Y Ks(p0p0)
J/Y Ks(pp-)
cc K
Y(2S) K
J/Y K
J/Y K non CP
Run2/Run1
KS Golden modes 30 larger than run 1
efficiency improved
14
Additional Channels cC1KS

• Major improvements in cc1KS analysis
• M(J/Yg)-M(J/Y) cut tightened / p0 veto (decrease
  fake photons)
• Cut on DE tightened (decrease inclusive J/Y
  background)
• Relaxed continuum requirements (not dominant
  background)
• Peaking background reduced by factor 4 (to 3)
• With only a 10 efficiency reduction

cc2Ks shown to be absent
15
Additional Channels J/YK0(KSp0)

• Improved understanding of the background and its
  effective CP

(Angular analysis paper about to be submitted)
55 signal events before tagging 37 after
16
Improved KL Selection

• Original analysis was optimized for S2/(SB)
• Fine for BF measurements, but not for CP
• Need to optimize accounting for the background
  asymmetry (SAB/ASB)2/(SB)
• Re-optimized with Monte Carlo
• Expect 10 improvement on the error

17
Resulting KL Yields Run1

• For data the improvement is better than expected

Old
New
In the IFR selection the signal yield has not
changed while the background is halved
Run1
18
CP Sample Non-KL Modes
Present Sample 725 PRL Sample 425
Before tagging and vertexing requirements
NNOW672
19
CP Sample J/Y KL
Run1Run2
N/p() EMC IFR
Run1 77/52 96/68
Run2 49/59 32/55
Run1Run2 128/56 129/65
20
Fully-Reconstructed B Sample
Neutral B Mesons
Flavor eigenstates for mixing and lifetime
measurements
Here determine mistag rates, Dt resolutions

• Cabibbo-favored hadronic decays
• Open Charm decays
• e.g.
• Charmonium Decays
• e.g.

Charged B Mesons
21
Vertex Algorithm Improvements

• More elaborate tagging vertex algorithm
• Increased efficiency through inclusion of 1-prong
  tag vertices (86 becomes 97)
• Improves average tagging vertex resolution from
  200 to 180 mm

22
Improved Vertex Performance

• We expect some vertex improvements in Run2 from
• Better use of layer 1 SVT hits in the Kalman fit
• Better SVT internal alignment

Improvement in resolution leads to
3-4 on sin2b error
run2
run1
NEW
OLD
s(sin2b)
23
Tagging Algorithm Unchanged
Category Definition efficiency
Q Lepton Charge of fastest electron (muon) with
pgt1.0(1.1) GeV/c. 10.9 ?
0.4 7.4 ? 0.5 Kaon Total charge of
identified kaons 35.8 ? 0.5 15.0 ?
0.9 NT1 Neural Network 7.8 ? 0.3
2.5 ? 0.4 NT2 Neural Network, 13.8
? 0.3 1.2 ? 0.3 Total 68.4 ? 0.7
26.1 ? 1.2
Tagging power Q e(1-2w)226.1 ?1.2
NT1, 2 recover information from unidentified
leptons, soft pions from D, and the momentum
spectrum of B decay products
24
Tagging Performance from Data

• Obtained from mixing fit to data samples

NT2
K
L
L
NT1
K
NT1
NT2
Q eD2 26.1 ?1.2
25
Likelihood Fit Method

• Global unbinned maximum likelihood fit to data
• Mistag rates, Dt resolutions tagged flavour
  sample
• sin2? tagged CP samples
• 45 parameters for mistag rates, ?t resolution
  backgrounds floated to obtain an empirical
  description from data

Separate Dt resolutions for run1 and run2
Largest correlation with sin2?? 13
tB 1.548 ps and Dmd 0.472 ps-1 fixed
26
Raw Asymmetries
Kaon tags
All tags
(Q15)
Visible ACP
sin2b0.590.20
sin2b0.560.15
27
Corrected Asymmetries
BABAR
BABAR
Comparable to similar Belle and CDF plots
28
hCP-1 Time Distributions
Visible asymmetry ACP
sin2b0.560.15
-5
0
5
Dt(ps)
Fit projections are from global fit
29
J/Y KL Time Distributions
Visible asymmetry ACP
sin2b0.700.34
Fit projections are from global fit
30
Sin2b Fit Results
Cross-checks Null result in flavor samples
Consistency of CP channels P(c2) 8
Goodness-of-fit P(LmaxgtLobs)gt27
sin2b 0.59 0.14
31
Run1 Run2 Comparison
Run1-Run2 change for PRL modes 1.8s
Run2
Run1
32
Run1 Run2 Comparison

• Change in central value 1.8s in uncorrelated
  error
• 30 efficiency improvement for all KS modes
• 15 improvement due to vertexing/alignment

33
Systematic Errors

• Signal resolution and vertexing 0.03
• Resolution model, outliers, SVT residual
  misalignment
• Tagging 0.03
• Studies of possible differences between BCP and
  Bflavor samples
• Backgrounds 0.02 (overall)
• Signal probability, peaking background, CP
  content of background
• Total 0.093 for J/Y KL channel 0.11 for J/Y K0
• Total 0.05 for full sample

KS KL K0 Full
Total Sys 0.049 0.104 0.162 0.049
Total Stat 0.151 0.340 1.01 0.137
34
Search for Direct CP
(assuming DG 0)
If more than one amplitude matters l might be
different from 1
Probing new physics only use hCP-1 sample
(contains no mixing background)
l 0.93 0.09 (stat.) 0.03 (sys.)
No evidence of direct CP violation due to decay
amplitude interference none expected Coefficient
of the sine term unchanged
35
Updated World Average
36
Interpretation of the Results
Following Höcker et al, hep-ex/0104062 (many
other recent global CKM matrix analyses)
37
Summary

• We have established CP violation at 4s level with
  sin2b 0.59 0.14 0.05
• Probability is lt 3 x 10-5 to observe an equal or
  larger value if no CP violation exists
• Corresponding probability for the hCP -1 modes
  is 2 x 10-4
• Probability of J/YKL and J/YKS having the same
  hCP is lt 0.1
• The study of CP violation in the B system has
  just started
• With more data becomes both a precision
  measurement and allows us to look at individual
  channels for consistency
• Start to look at sin2a and other approaches