B Mixing and Lifetimes from CDF

1. B Mixing and Lifetimesfrom CDF Colin Gay, Yale University for the CDF II Collaboration Colin Gay, Yale University

2. Outline • Status of B lifetimes • Bs “lifetime” and lifetime difference • Bs mixing Colin Gay, Yale University

3. State of Lifetime • Heavy Quark Expansion predicts • B+, B0 in ok shape • b a bit below expectation • Bs on edge • More on this later Colin Gay, Yale University

4. Lifetime/Mixing samples lifetime Unbiased trigger lifetime biased trigger Mixing Colin Gay, Yale University

5. Measuring Lifetime Flight distance proper decay length • For fully reconstructed (hadronic) modes Boost Momentum resolution (proportional to ct) Vertex resolution (~constant) • For semileptonic modes, missing neutrino causes => Resolution poor at large decay time Important effect for Bs mixing Colin Gay, Yale University

6. Fully Reconstructed J/ X • Easy to trigger on, large samples, unbiased in lifetime • Fully reconstructed modes • Excellent vertex and momentum resolution • CDF Fully reconstructs • J/+ • B+ J/K+ • B0  J/K*, J/Ks • Bs J/ • b J/ For example: 1155§39 signal • See summary following for detailed results Colin Gay, Yale University

7. B- ct-efficiency B-D0p- Lifetimes from hadronic decays • To get large samples of fully reconstructed hadronic decays, we use events recorded via the Secondary Vertex Trigger • Requires 2 tracks with impact parameters between 120m and 1mm • Trigger has intrinsic lifetime bias • Events have excellent momentum and vertex resolutions • Final states reconstructed: • B± D0p± (D0Kp) N=8380 evts • B0 D±p (D±Kpp) N=5280 evts  D± 3p (D±Kpp)N=4173evts • Bs  Ds p± (Dsfp) N=465 evts  Ds 3p (Dsfp) N=133 evts • Larger statistics than J/y modes • Larger systematics due to • Trigger Efficiency curve • Larger backgrounds Colin Gay, Yale University

8. Lifetimes from hadronic decays Bs B+ Bs t(B+) = 1.661±0.027±0.013 ps t(B0) = 1.511±0.023±0.013 ps t(Bs) = 1.598±0.097±0.017 ps Colin Gay, Yale University

9. Lifetime with high-pt Semileptonic sample • Trigger on 8 GeV lepton • Reconstruct • High statistics, but missing neutrino -> “K” factor t(Bs) = 1.381±0.055±0.046ps t(B0) = 1.473±0.036±0.054ps t(B+) = 1.653±0.029±0.032ps Colin Gay, Yale University

10. B Mass & Lifetime Difference • Second order weak diagram gives non-zero matrix element • In basis have a non-diagonal Hamiltonian • Recall Eigenstates are: Colin Gay, Yale University

11. Bs “lifetime” meaning • When a significant  exists, lifetime measurements are sample composition dependent • Measured lifetime is where = fraction of light state E.g. lifetime measured in SL decays dominates the average With the constraint (HQE says equal to 1%) Unphysical value => most likely Colin Gay, Yale University

12. Extracting both Bs lifetimes In previous cases, the sample composition, and hence relation of depends upon the unknown In the case of the hadronic decays, there is the additional effect that the trigger turn-on affects the short-lived component of the Bs more than the long-lived There is a decay in which one can measure, simultaneously, the light-heavy sample composition AND each components’ lifetime: S,D wave amplitudes = Parity Even, (CP Even) P wave amplitude = Parity Odd, (CP Odd) Since the mass eigenstates of the Bs system are • See D0 talk for details Disentangle different L-components of decay amplitudes => isolate two B states Colin Gay, Yale University

13. Two Lifetimes CP-odd fraction ( ) ~ 22% + Recent D0: Colin Gay, Yale University

14. The two Lifetimes Constraint helps low statistics H Note that Colin Gay, Yale University

15. Experiments vs. Theory Theory: Flavor-specific “lifetime” (SL) CDF Theory Preferred D0 Colin Gay, Yale University

16. Combined CDF/D0 Fit Colin Gay, Yale University

17. Lifetimes: CDF Summary Colin Gay, Yale University

18. Kaon Mixing Analysis Strategy • Just like a lifetime measurement, but look for change of B particle to antiparticle • Mixing • Bs or Bs at production? • Initial state flavor tagging (calibrated on B0) • Tagging dilution D=1-2w, w=mistag prob. • Effective sample N D2 (D2~1%) • Bs or Bs at decay? • Decays are self-tagging (eg ) • Reconstruct proper decay time • Fit Asymmetry(Nunmixed-Nmixed)/ Nto D*A*cos(Dm t) at fixed Dm • Expect A=1 for Dm ~ Dms • Limit (95% CL): • Dm such that A+1.645sA = 1 Colin Gay, Yale University

19. Opposite side tagging Use the other B in the event Semileptonic decay (b g l-) (1) Muon, (2) Electron Use jet charge (Qb = -1/3) (3) Jet has 2ndary vertex (4) Jet contains displaced track (5) Highest momentum Jet Calibrated on B0 p Flavor Tagging + Signals p K Reconstructed B K Fragmentation track Opposite side B • Hadronic (eg ): • Less signal yield • Excellent pT resolution • Good sensitivity at higher Dms • Semileptonic (eg ): • Higher signal yield • Poor pT resolution • Good sensitivity at lower Dms. Colin Gay, Yale University

20. CDF Limit (Semileptonic Mode) • Bs! Dsl X l=e/ (360 pb-1) • Ds!, K*K, ppp • Opposite side: • e,tag • Jet Charge • 7800 events • D2 = (1.43§0.09)% Colin Gay, Yale University

21. CDF Limit (Hadronic mode) • Bs! Ds(360 pb-1) • Ds!, K*K, ppp • Opposite side: • e,tag • Jet Charge • ~900 events • D2 = (1.13§0.08)% Colin Gay, Yale University

22. Potential Improvements • Increase statistics • Additional Decay Modes, More Data • Increase tagging power • Same Side Kaon Tagging increase D2by 1-3% • Lifetime resolution improve 10-20% Statistics Statistics Resolution Colin Gay, Yale University

23. Conclusion • Bs mixing search at the Tevatron is an ongoing affair • Expect improvements in technique and statistics • Observation likely still some time away • Lifetime ratios in reasonable agreement with theory • Bs the exception? • Lifetime difference observed. Higher than predicted, but errors still large • In addition, the mean width differs by ~2.7 from prediction • Could it really be that • Will repeat J/ analysis with x4 data • Both CDF and D0 are statistics limited Colin Gay, Yale University