1. B mixing and lifetimes at the Tevatron
FPCP 2006
Vancouver
Jónatan Piedra LPNHE-University Pierre et Marie Curie / CNRS-IN2P3
on behalf of the D and CDF Collaborations

2. Outline Precision B Lifetimes B Mixing Summary THIS TALK OTHER TALKS
motivation Lb Bs
Bs lifetime difference
B Mixing
current status
ingredients
results
Summary
CDF and D detectors
CDF and D Hot Topics
Detailed Bs mixing at D
D. Buchholz
D Hot Topics
Detailed B  hh
D. Tonelli
CDF Hot Topics
Bc lifetime
I. Kravchenko
B Spectroscopy
Bs  Ds Ds
R. Van Kooten
Bs Decays and B leptonic decays
April 9, 2006
J. Piedra

3. Precision B Lifetimes

4. Precision B Lifetimes Motivation
EXPERIMENT
THEORY
(NLO)
C. Tarantino, hep-ph/
b-hadron decays dominated by b-quark
Light quarks are included with 1/mb perturbative expansions (HQE)
expect small differences between lifetimes of different species
lifetime ratios reduce theory uncertainties
Cecilia Tarantino, October 2003
April 9, 2006
J. Piedra

5. Precision B Lifetimes Lb Lifetime
CDF and D have measured Lb  J/ L lifetime
Better proper time resolution than Lb  Lc ln (world average dominated)
Earlier t(Lb)/t(B0) predictions were 2s above experiment
new calculations including higher order effects predict lower ratio
D 250 pb-1
CDF 370 pb-1
PRL (2005)

6. Precision B Lifetimes Lb Lifetime at CDF
Analysis based upon 370 pb-1
Technique
unbinned maximum likelihood fit to proper decay-length and mass
Lb  J/ L0 194  23 candidates
J/  mm
L0  pp -
Use B0 reference mode
larger yield and similar decay topology
B0  Jpsi Ks  53 candidates
Ks0  p +p -
CDF 370 pb-1
HFAG
April 9, 2006

7. Precision B Lifetimes Bs Lifetime
D and CDF measure lifetime in
D 400 pb-1
CDF 360 pb-1
best in the world
April 9, 2006

8. Precision B Lifetimes Neutral Meson Mixing
Quark mixing  non-diagonal Hamiltonian for
Diagonalizing the Hamiltonian results in
two eigenstates and
two masses mH and mL, with Dm  mH – mL
two decay widths GH and GL, with DG  GL – GH
R. Van Kooten
Bs decays and B leptonic decays
April 9, 2006

9. Precision B Lifetimes Bs Lifetime Difference
Bs  J/ 
Pseudoscalar  Vector - Vector
Decay amplitude decomposed into 3 linear polarization states
A0 = S + D wave  P even
A|| = S + D wave  P even
A = P wave  P odd
If CP violation neglected
Bs,Light  CP even
Bs,Heavy  CP odd
angular distributions are different
Angular analysis separates CP eigenstates  measure two lifetimes
April 9, 2006
J. Piedra

10. Precision B Lifetimes Bs Lifetime Difference
D 800 pb-1
CDF 260 pb-1

11. Precision B Lifetimes................... and
First measurement (~95% CP even)
CDF 360 pb-1
D. Tonelli
CDF Hot Topics
Expected
Lifetime extracted from decay
Bs -> KK ~ 95% CP even
CDF 360 pb-1
best in the world
D 210 pb-1
I. Kravchenko
B Spectroscopy
April 9, 2006

12. Precision B Lifetimes HFAG
B+ and B0 at LEP/SLC, B factories and Tevatron (CDF/D)
dominated by Belle and BaBar
Bs dominated by Tevatron, LEP
Bc at Tevatron
Lb dominated by LEP, Tevatron
b-hadron species
measured lifetime [ps]
t/t(B0) lifetime ratio
predicted range B+
1.643  0.010
1.076  0.008
1.04 – 1.08
Bs ( flavor specific)
1.454  0.040
0.914  0.030
0.99 – 1.01 Lb
1.288  0.065
0.844  0.043
0.86 – 0.95 B0
1.527  0.008
hep-ex/
March 2006 Bc
0.469  0.065

13. B Mixing

14. B Mixing Theoretical Prediction
SM prediction for the ratio of Bs and B0 mixing frequencies
Dmd precisely measured
Dms not yet measured precisely
Potential NP discovery
NP = New Physics
CKM fit
J. Piedra

15. B  background candidates sct  time resolution
B Mixing Significance
S  signal candidates
B  background candidates
sct  time resolution
The dilution D measures the purity, D = 0 (1)  random (perfect) tagger
The dilution attenuates the observed oscillations
April 9, 2006

16. 1. Final State Reconstruction
B Mixing Ingredients
1. Final State Reconstruction
b-flavor at decay
3. b-Flavor Tagging
b-flavor at production
2. Proper Decay Time
In B rest frame
4. Amplitude Scan for Bs Mixing
Set a lower limit or observe Dms
Opposite Side
Trigger Side

17. B Mixing Reconstructed......................
D exploits semileptonic decays from m trigger
CDF uses both electrons and muons
D 1 fb-1
CDF 1 fb-1

18. B Mixing Reconstructed.....................
CDF collects hadronic B decays by triggering on impact parameter
Around 3700 Bs signal candidates
April 9, 2006
J. Piedra

19. B Mixing Proper Decay Time
Procedure
measure pT of B daughter tracks
measure the decay length Lxy
boost B back to its rest frame
Fully reconstructed decays
all daughters reconstructed
Partially reconstructed decays
some tracks escape detection
 need simulation
April 9, 2006
J. Piedra

20. B Mixing b-Flavor Tagging
A flavor tagger determines the b-flavor at production time
bb production  flavor tagging on the Trigger Side or the Opposite Side
Soft Lepton Tagger
look for B  ln DX decay on the OS
lepton charge indicates b-flavor
Jet Charge Tagger
look for jet or secondary vertex from OS
jet charge indicates b-flavor
Same Side (Kaon) Tagger
look for a fragmentation track on the TS
it is charge correlated with the b-flavor
SLT
Trigger Side
JQT
- Jet Charge Tagger. Empirically, the sum of all particle charges in a jet containing a B meson is correlated with the b-quark charge
- Same Side Tagger. Looks for the fragmentation track that is charge correlated with the produced B meson
Opposite Side
SS(K)T
J. Piedra

21. B Mixing Flavor Analysis on B+ and B0
Calibrate opposite side flavor taggers prior to Dms analysis
combine several B+,0 decays
combine all taggers
Direct Dmd measurement
cross-check for Bs mixing
asymmetry
combined eD2 (%)
1.55  0.08
2.48  0.22
CDF D
D semileptonic 1 fb-1
Dmd =  (stat)  (syst) ps-1
CDF semileptonic 1 fb-1
Dmd =  (stat)  (syst) ps-1
CDF hadronic 355 pb-1
Dmd =  (stat)  (syst) ps-1
world average
Dmd =  ps-1

22. B Mixing SSKT at CDF
Look for the fragmentation track that is charge correlated with the B
Dms not yet measured precisely
Parameterization from MC
Extensive data/MC comparisons on all tagging related quantities
Rely on MC prediction of SSKT performance for Bs mixing

23. CDF MC 355 pb-1 B Mixing SSKT at CDF Systematic studies cover
quark fragmentation model
bb production mechanisms
excited B mesons content
detector / PID resolution
particle species content around B
data / MC agreement
Select track within DR < 0.7 around B most likely to be a kaon
based on dE/dx and TOF information
CDF MC 355 pb-1

24. B Mixing Fourier Analysis
Two domains to fit for oscillations
time  fit for a cosine wave
frequency  examine f-spectrum
Time domain approach
fit for Dms in P(t) ~ 1  D cos(Dmst)
Frequency domain approach
introduce amplitude, P(t) ~ 1  AD cos(Dmst)
fit for A at different Dms
 obtain frequency spectrum A(Dms)
standard method for combining limits
with flavor taggers calibrated A = 1 for the true Dms else A = 0
time domain
frequency domain

25. B Mixing Amplitude Scans on Dmd
The yellow band is 1.645sA around data points
Dm values where A sA < 1 are excluded at 95% CL
Sensitivity is where 1.645sA = 1
D Run II Preliminary
1 fb-1
Amplitude scan works on B0 decay modes
April 9, 2006

26. B Mixing Amplitude Scans on Dms
A/sA (Dms = 19 ps-1) = 2.5  5% p-value
SSKT not yet included
A(Dms=19) = 1.6 sigma from 1
sensitivity
95% CL limit
D 1 fb-1
14.1 ps-1
14.8 ps-1
sensitivity
95% CL limit
CDF 355 pb-1
13.0 ps-1
8.6 ps-1
April 9, 2006
J. Piedra

27. B Mixing Log Likelihood Scan
hep-ex/
submitted to Phys. Rev. Lett.
17 < Dms < 21 90% CL assuming Gaussian uncertainties

28. B Mixing D Effect on World Average
HFAG preliminary
(correlated systematics not included)
A (Dms = 19 ps-1)
1.5 sA sA D

29. Summary New Lb lifetimes reduce distance with theory
Tevatron measures the best Bs lifetimes in the world
Bs lifetime difference within SM
D Bs oscillation
1 fb-1
2.5 sA excess at Dms = 19 ps-1 (5% p-value)
17.1 < Dms < % CL
CDF Bs oscillation
355 pb-1
Dms > 8.6 ps 95% CL

30. Backup

31. Precision B Lifetimes Transversity Basis
 = +1
particle -1
antiparticle
untagged Bs
Transversity basis  J/ rest frame
 flight direction  +x
KK plane  xy plane
J. Piedra

32. B Mixing Semileptonic Decay Time
Missing particles  missing pT
Determine pseudo-ct from data
ct = ct* kMC
include kMC effect in signal PDF
Very nice talk. I like the general style, very Jonatan Piedra.
Ch. Paus
April 9, 2006
J. Piedra