1. 1 Results on CP Violation in B s Mixing [measurements of ϕ s and ΔΓ s ] 7-3-2012 Pete Clarke / University of Edinburgh & CERN Presentation on behalf of LHCb Collaboration Rencontres de Moriond, La Thuile, 3-10 March 2012

2. 2 This talk presents new LHCb results since 2011 summer conferences  Using the full 1fb -1 of data from 2011 (was 0.37fb -1 )  CPV phase ϕ s,ΔΓ s and other quantities from B s  J/ψΦ  LHCb-CONF-2012-002  arXiv:1112.3183v2 to be published in PRL arXiv:1112.3183v2  ϕ s from B s  J/ψ ππ  LHCb-PAPER-2012-006 to be submitted to PLB  Resolution of the “two fold ambiguity” for ϕ s & ΔΓ s  arXiv:1202.4717v2 submitted to PRLarXiv:1202.4717v2

3. 3 LHCb The detector The people The LHC These new results are due to  Excellent running of detector  Excellent running of LHC  Efforts of all the people

4. 4  LHCb is a forward spectrometer  2 < η < 5  Optimised for b and c physics  Precise tracking and decay vertex finder  Good particle ID LHCb in context

5. Experimental phenomenology of CP Violating phase ϕ s 5

6. 6  Measure relative phase difference 1 ϕ s = ϕ M – 2 ϕ D between two “legs”  In SM & normal conventions & ignoring penguins ϕ D ~ 0 ϕ s SM ~ ϕ M  is predominantly determined 2 by arg(V ts )  is predicted to be small ~ -0.04 [Charles et al. Phys. Rev. D84 (2011) 033005]  New Physics (NP) can add large phases: Bs0Bs0 Bs0Bs0 Bs0Bs0 J /  X ϕMϕM ϕ s = ϕ s SM + ϕ s NP 1] The term ϕ s is overloaded. It is also used for Arg-(M 12 /Γ 12 ) 2] ϕ s = -2 arg( V ts V tb * / V cs V cb * ) -ϕD-ϕD ϕDϕD

7. The signals  B s  J/ψΦ and B s  J/ψππ are very clean decays BsBs KK KK  J/J/  --

8. 8  Straightforward differential decay rates for B s  J/ψππ : For B s B s [Monte Carlo SM x 10 for visibility] Decay time (ps) _  Signal is sinusoidal time distribution  Amplitude proportional to sin( ϕ s )  Opposite sign for B and B  must tag  Diluted by wrong tagging probability ω tag  Diluted by detector resolution σ t  Fundamentally we measure: sin( ϕ s ) x D( σ t ) x (1 -2ω tag ) x sin(Δm s t) _

9. 9 Need good proper time resolution w.r.t. sinusoid period ~ 350fs  We measure from data using prompt J/ψ which decay at t=0  width ~ 45fs  In analysis we actually use a resolution estimated per-event Decay time resolution Measured decay time of prompt events [ps]

10. sin( ϕ s ) x D( σ t ) x (1 -2ω tag ) x sin(Δm s t) 10 Need to tag B or B at production tagging efficiency ε tag ~ 33% effective mistag ω tag ~ 36.8% effective tagging power ε tag (1 -2ω tag ) 2 ~ 2.3% event-mistag (ω tag ) _

11. Measurements of ϕ s using B s  J/ψππ ~ 1.0 fb -1 ~0.41 fb -1 published result: Phys. Lett. B707 (2012) 497, arXiv:1112.3056. This update : LHCb-PAPER-2012-006 11

12. B s  J/ψ ππ  Previous analysis was B s  J/ψf 0 (980) 12  Now use wider M(ππ) range hence: B s  J/ψ ππ  All m(ππ) range found to be CP-odd (97.7% @ 95% C.L) [LHCb-PAPER-2012-005]

13. B s  J/ψ ππ signal  Boosted Decision Tree selection  Maximum likelihood fit to time and mass  Uses Γ s and ΔΓ s from the B s  J/ψΦ analysis (+correlation)  Approx. 7400 signal events 13 B s Invariant mass distribution [MeV] B s decay time distribution [ps] The decay time asymmetry

14. ϕ s = -0.02 ± 0.17(stat.) ± 0.02(syst.) rad. B s  J/ψππ final result 14 Systematic errors

15. Measurements of ϕ s using B s  J/ψΦ ~ 1.0 fb -1 0.37 fb -1 published result: arXiv:1112.3183 [LHCB-PAPER-2011-021] This update: LHCb-CONF-2012-002 15

16. 16 B s  J/ψΦ analysis  Decay to CP-odd and CP-even final states,  need analysis of decay angle distribution  Differential cross section is “very rich”  3 “P-wave” amplitudes of KK system  1 “S”-wave amplitude  10 terms with all interferences J/ψ rest frameΦ rest frame  Much larger branching fraction J/ψΦ / J/ψππ ~ 5

17. 17 sin( ϕ s ) x D( σ t ) x (1 -2ω tag ) x sin(Δm s t) ..but fundamentally for ϕ s we still measure: ..and because we separate the terms, we measure the lifetimes of Heavy and Light eigenstates separately: Γ L & Γ H  Γ s & ΔΓ s  There is a two fold ambiguity in the solutions ϕ s  π - ϕ s ΔΓ s  - ΔΓ s + strong phase changes

18. B s  J/ψΦ signal 18  Simple selection with kinematic cuts  Most background removed by decay time cut t > 0.3 ps  Very clean signal  Approx. 21200 signal events

19. B s  J/ψΦ decay angle distributions 19  The CP-even / CP-odd separation is very clear in all distributions. CP-even CP-odd

20. Digression: measurement of Δm s 20  We observe a central value Δm s = 17.50 ± 0.15 (stat) ps -1  Compare to LHCb published measurement 17.63 ± 0.11 ± 0.02 arXiv:1112.4311  This gives confidence that if there is a sin( ϕ s ) x sin(Δm s t) term we would see it.  The data has sinusoidal terms which measure Δm s independently of ϕ s Likelihood scan for Δm s

21.  Maximum likelihood fit to signal + background time, angle and mass distributions.  Constrain Δm s to 17.63 ± 0.11 ps  “Solution-I” shown here  ϕ s uncorrelated with other quantities B s  J/ψΦ : New Preliminary Results 21

22. B s  J/ψΦ : New Preliminary Results 1.0 fb -1 22 Γ s = 0.6580 ± 0.0054(stat.) ± 0.0066(syst.) ps -1 ΔΓ s = 0.116 ± 0.018(stat.) ± 0.006(syst.) ps -1 ϕ s = -0.001 ± 0.101(stat.) ± 0.027(syst.) rad Contour for ϕ s - ΔΓ s

23.  Note: Tagging and resolution parameters floated in fit within uncertainties, hence not explicitly included in systematics Systematics 23

24.  Used a simultaneous fit to both datasets, taking all common parameters and correlations into account.  Used largest syst. error. 24 ϕ s = -0.002 ± 0.083(stat.) ± 0.027(syst.) rad B s  J/ψΦ and B s  J/ψππ combined prelim.result

25. Resolving the ambiguous solution The sign of ΔΓ s ~ 0.37 fb -1 LHCb-PAPER-2012-28, submitted PRL, arXiv:1202.4717v2arXiv:1202.4717v2 25

26. 26  There are two ambiguous solutions related by ϕ s  π - ϕ s and ΔΓ  - ΔΓ  We can disambiguate using the P-Wave  S-Wave interference

27. 27  Split data into 4 bins of m KK  In each bin measure  Fraction of S-wave  Measure relative strong phase difference δ S – δ perp ΔΓ s > 0 ΔΓ s < 0 Solution-I is selected [4.7σ from being flat] ΔΓ s > 0 δ S – δ perp [rad]

28. LHC + Tevatron Results 28 D0 8 fb -1 [S.Burdin, EPS 2011]CDF 10 fb -1 [Sabato Leo talk, Lake Louise] This result 1 fb -1

29. Pictorial overlay - winter 2012 29

30. 30 Γ s = 0.6580 ± 0.0054(stat.) ± 0.0066(syst.) ps -1 ΔΓ s = 0.116 ± 0.018(stat.) ± 0.006(syst.) ps -1 ϕ s = -0.001 ± 0.101(stat.) ± 0.027(syst.) rad. ϕ s = -0.002 ± 0.083(stat.) ± 0.027(syst.) rad.  LHCb has presented new preliminary results using the full 2011 data (1 fb -1 )  From an analysis of the J/ψ ϕ channel we find:  From an analysis of the J/ψππ channel we find: ϕ s = -0.02 ± 0.17(stat.) ± 0.02(syst.) rad  Combining both results we find:  We resolve the 2-fold ambiguity and find: ΔΓ s > 0 Conclusion

31. Extra Information 31

32. B s  J/ψ ππ analysis: 32  Upper limit on ρ(770) found to be 1.5% @ 95% C.L.  Total is quadrature sum of f2(1270) + ρ(770) 1. CP-odd fraction: Table from LHCb-PAPER-2012-006 2. Profile Likelihood

33. B s  J/ψ Φ 33 Decay angle acceptances All acceptances ~ 5% effects Accounted for in ML fit

34. B s  J/ψΦ : New Preliminary Results 34 Γ s = 0.6580 ± 0.0054(stat.) ± 0.0066(syst.) ps -1 ΔΓ s = 0.116 ± 0.018(stat.) ± 0.006(syst.) ps -1 ϕ s = -0.001 ± 0.101(stat.) ± 0.027(syst.) rad. This new preliminary result 1.0 fb -1 Γ s = 0.657 ± 0.009(stat.) ± 0.008(syst.) ps -1 ΔΓ s = 0.123 ± 0.029(stat.) ± 0.011(syst.) ps -1 ϕ s = 0.151 ± 0.18(stat.) ± 0.06(syst.) rad. Previous published result 0.37 fb -1 arXiv:1112.3183

35. B s  J/ψ Φ 35 Syst error due to possible direct CVP  We vary |λ| 2 by 5%  Motivated by a fit for |λ| 2 gives < 5%