Hints of New Physics in b s transitions

Transcription

1 ( B K * γ ) Hints of New Physics in b s transitions (or) Looking for New Physics in Flavour Physics in quark sector Achille Stocchi LAL Orsay Université Paris Sud/IN2P3-CNRS 24 September 2008 DESY-Zeuthen M. Bona, M. Ciuchini, E. Franco, V. Lubicz, G.Martinelli, F. Parodi, M. Pierini, C. Schiavi, L. Silvestrini, A. Stocchi, V. Sordini, C. Tarantino and V. Vagnoni

2 Flavour Physics in the Standard Model (SM) in the quark sector: ~ half of the Standard Model 10 free parameters 6 quarks masses 4 CKM parameters Wolfenstein parametrization : λ,a, ρ, η η responsible of CP violation in SM In the Standard Model, charged weak interactions among quarks are codified in a 3 3 unitarity matrix : the CKM Matrix. The existence of this matrix conveys the fact that the quarks which participate to weak processes are a linear combination of mass eigenstates The fermion sector is poorly constrained by SM + Higgs Mechanism mass hierarchy and CKM parameters

3 The Unitarity Triangle The CKM is unitary AB V V 1 V 1 V = = $ + = V V V V * 2 2 td tb td td (1! ) " ~ * cd cb # cb # ts V V $! % 1 V AC = = + = & V V ) * V " " = = * ud ub ub " # 1 * ' ( cd cb 2! cb arg arg α + β + γ = π * $ VtdV % $ tb! % & atan * ' = & ' ) VcdVcb * ) (1 ( # ) * * $ VudV % $ ub! % & atan * ' = & ' ( VcdVcb ) ( # ) (b u)/(b c) ρ η Δm d (1 ρ) 2 + η 2 Δm d / Δm s (1 ρ) 2 + η 2 ε K η ρ β γ V ud V* ub + V cd V* cb + V td V* tb = 0 1 [(1 ) + P] (! # " ) atan / (1 ) atan (! / ") Size of order λ 3 : λ 3 :λ 3 f +,F(1), 2 f Bd B Bd ξ B K - -

4 An example on how to fit the UT parameters and fit new physics ε K : CPV b clappleand B ccs d B ππ/ρπ/ρρ B DK and in K B: s decays φ: mixing 1 φb u ν /β 3 /γ : φ 2 /α

5 Fit with the Standard Model (SM) function(ρ,η.) Angles Sides + ε K ρ = ± η = ± ρ = ± η = ± 0.023

6 Global Fit Δm d,δm s,v ub,v cb,ε k + cos2β + β + α + γ + 2β+γ SM Fit Tremendous success of the CKM picture ρ = ± η = ± 0.014

7 Should we stop here? How to look for NP? And in case of no observation to establish how much room is left for NP effects? Long story Some example in next 4 transparencies..

8 some specific example of NP tests.. SM predictions of Δms Δm s SM expectation Δm s = (17.5 ± 2.1 ) ps -1 LEP/SLD 1999 LEP/SLD CDF CDF only : signal at 5σ Δ m s = (17.77 ± 0.12) ps -1 Tevatron results Limited by Lattice calculations

9 γ Direct measurement From fit γ = (64.0 ± 3.0) o γ = (81 ± 13) o (up to π ambiguity) Legenda agreement between the predicted values and the measurements at better than : Summer 2007 Winter σ 2σ 3σ 4σ 5σ 6σ B factories results LHCb expected to contribute

10 sin2β vs V ub sin2β=0.668 ± σ tension Vub(excl.) =(35.0±4.0)10-4 From fit sin2β=0.736 ± Vub(incl.) =(39.9±1.5 ± 4.0)10-4 Vub =(34.8±1.6)10-4 B factories results

11 Flavour Physics measure coupling Mass scale! " eff NP physics could be always arround the corner WHAT IS REALLY STRANGE IS THAT WE DID NOT SEE ANYTHING. With masses of New Particles at few hundred GeV effects on measurable quantities should be important Problem known as the FLAVOUR PROBLEM Λ eff <~ 1TeV + flavour-mixing protected by additional symmetries (as MFV) Couplings can be still large if Λ eff > TeV If there is NP at scale Λ, it will generate new operator of dimension D with coefficients proportional to Λ 4-D only operator of D=6 contribute. So that in fact you have a dependence on 1/ Λ 2 δ

12 Today we concentrate on a Model Independent fit to ΔF=2 observable which show a 2.5σ evidence of NP in the b s transitions

13 Fit in a NP model independent approach ΔF=2 Parametrizing NP physics in ΔF=2 processes Cq e Q + Q NP SM 2ö i d! B= 2! B= 2 = SM Q! B= 2 EP SM ' m = C ' m f (",#, C, QCD..) d B d B d A J K f 0 CP ( / (, ) = sin(2$ + 2 % B ) (",#,% B ) EP SM & = & )% f (",#,% ) B d EP SM! =! f (",#, C, QCD..) K C! K EP SM ' m = C ' m f (",#, C, QCD..) A ( J / (,%) = sin(2$ ) 2 % ) f (",#,% )... s B s Bs s CP s B B s d B s d! d d

14 To help with a more specific example : Example for B oscillations (FCNC-ΔB=2) : Using the example of the Supersymmetry δ bd Q Q NP " B= 2 SM " B= 2 # p % r! $ bq eff # p r V V * tb M W tq p r δ bd upper limit of the relative contribution of NP NP physics coupling Λ eff NP scale (masses of new particles) δ bq ~ 1 δ bs ~1 If couplings ~ 1 Λ eff ~ 10/ p r TeV Λ eff ~ 2/ p r TeV δ bq ~ 0.1 δ bs ~0.1 all possible intermediate possibilities Λ eff ~ 1/ p r TeV Λ eff ~ 0.2/ p r TeV Minimal Flavour Violation! " V V * q' d tq' td (couplings small as CKM elements) Λ eff ~ 0.08/ p r TeV Oversimplified picture : for a quantitative analysis see for instance UTfit collaboration JHEP 0803:049,2008ariv:

15 ρ, η ϕ d C d C s ϕ s C εk γ (DΚ) Tree processes V ub /V cb Δm d Constraints 1 3 family 2 3 family ACP (J/Ψ Κ) ACP (Dπ(ρ),DKπ) A SL α (ρρ,ρπ,ππ) A CH ΔΓ s /Γ s Δm s ASL(Bs) 1 2 familiy ACP (J/Ψ φ) ε K ~ 5 new free parameters C s,ϕ s B s mixing C d,ϕ d B d mixing K mixing C εk Today : fit possible with 10 contraints and 7 free parameters (ρ, η, C d,ϕ d,c s,ϕ s, C εk )

16 Β d C Bd = 0.96± 0.23 φ Bd = -(2.9 ± 1.9) o A NP /A SM vs φ NP The sin2β tension produces a 1.5σ effect on φ(bd) and the asymmetry in the A d (NP)/A d (SM) vs φ(np) plot With present data A NP /A SM 1.5σ A NP /A SM ~1 only if φ NP ~0 A NP /A SM prob.

17 Actual sensitivity for a generic NP phase in the Bd sector r=a NP /A SM ~10-15% Β d This is not yet a prove that if NP should be MFV violating Just for showing the link between precision and mass scale Q Q NP " B= 2 SM " B= 2 # r %! $ bq eff # r V V * tb M W tq r upper limit of the relative contribution of NP δ bd NP physics coupling Λ eff NP scale (masses of new particles) Take a case where *! " V V Λ eff ~ 80/ r GeV Λ eff ~ ( ) GeV q' d tq' td MORE PRECISION IS NEEDED

18 B s sector : very recent results Β s ρ, η C d ϕ d C s ϕ s A CH D0,CDF ( ) τ(βs),δγ s /Γ s CDF, D0, LEP Δm s CDF (~2006),D0, LEP ASL(Bs) D0 (2007) ACP (J/Ψ φ) ~ D0,CDF ( ) The realm of Tevatron! s = arg * " VtsV # tb $ * % = ± & VcsVcb ' ( ) o V td V cd + V ts V cs + V tb V cb = 0 λ 2 λ 4 λ 4 β s Recall that in B d sector! " V V # $ ( ) * % & VcdVcb ' * td tb o = arg = ± V ud V ub + V cd V cb + V td V tb = 0 λ 3 λ 3 λ 3

19 φ s vs of ΔΓ s using B s J/ψφ Nota bene for the experimental result φ s = -2β s Angular (θ,ϕ,ψ) analysis as a function of the proper time. Similar to measurement of β in B d J/ψ K*. Respect to the B d case, there is additional sensitivity because of ΔΓ s term Dunietz,Fleisher and Nierste Phys.ReV D63:114015,2001 Experimentally θ and ϕ are well determined from the µ from J/ψ ψ is the decay plane between the J/ψ and the φ.

20 Winter 2007 Before ICHEP 2008 Not B s J/ψφ with B s J/ψφ C(Bs) = 1.11 ± 0.32 φ(bs) = (-69±14)º U (-20±14)º U (20±5)º U (72±8)º SM 3σ C(Bs)=1.07±0.29 φ(bs)=(-19.9±5.6)ou(-68.2±4.9)º ariv: v1 [hep-ph] 5 Mar 2008

21 B d 1 <--> 3 vs B s 2 <--> 3 1 <--> 3 10% 70% A NPd /A SMd ~0.1 and A NPs /A SMs ~0.7 correspond to A NPd /A NPs ~ λ 2 i.e. to an additional λ suppression.

22 After ICHEP 2008-CKM2008 ICHEP D0 released the likelihood without assumptions on the strong phases C(Bs) = 0.97 ± 0.20 φ(bs)=(-70 ± 7) o U(-18 ± 7) o SM 3σ 2.5σ New CDF data not included: new CDF likelihood not ready yet SM compatibility decreased in the CDF analysis

23 Here the results from HFAG. Without additional constraints See Diego Tonello (CDF), Lars Sonnenschein (D0) CKM08 Rome

24 This result, if confirmed, will imply : - of course NP physics - NP not Minimal Flavour Violation (large couplings..new particles not necessary below the TeV scale - NP model must explain why effects on B d (which can still be as large as 20%) and K systems are smaller 1 <-> 2: strong suppression 1 <-> 3: O(10%) 2 <-> 3: O(1) this pattern is not unexpected in flavour models and in SUSY-GUTs Flavour physics central - B d sector, for ΔF=2 but also ΔF=1 b s transitions - K sector - of course B s sector See next page PRECISION IS NEEDED

25 ΔF=1 b s transitions are very sensitive to NP contributions (ΔF=1) B 0 d b d W t s s s d φ K 0 ~ g ~ ~ b b s s (! d 23 ) LR New Physics contribution (2-3 families) The disagreement is much reduced S(φK) (! d ) 23 LR m = m q~ g~ 350 GeV PRECISION IS NEEDED B factories results SuperB expected to contribute Im(δ 23 ) LR CFMS

26 - D0 and CDF will update their results. They have not used entire dataset. If the NP phase stay so large they could observe it with the full/final dataset - ϕ s is a golden measurement for LHCb Simulation done with 4fb -1. φ Bs = (0.0 ± 1.3) o C Bs = 0.99 ± 0.12 But also with much less data, LHCb can observe the effect if will stay so large New studies show that (end 2009?) LHCb with 0.5fb-1 σ(φbs) = 0.06 ATLAS with 2.5fb-1 σ(φbs) = 0.16 See Gaia Lanfranchi CKM08/Rome

27 Flavour physics in the quark sector is in his mature age In b d transitions NP effect are confined to be at order less ~10-15%! New data from Tevatron show ~2.5σ discrepancy from SM in b s transitions If confirmed would implies NP and not Minimal Flavour Violation Tevatron (with full statistics) and LHCb will clarify the discrepancy Flavour Physics is alive more than ever to look for NP beyond SM SuperB

28 BACKUP MATERIAL

29

30 B τν Belle BR(B τυ)=(1.73±0.34)10-4 B factories results SuperB expected to contribute

31 The problem of particle physics today is : where is the NP scale Λ ~ 0.5, TeV The quantum stabilization of the Electroweak Scale suggest that Λ ~ 1 TeV LHC will search on this range What happens if the NP scale is at TeV naturalness is not at loss yet Flavour Physics explore also this range We want to perform flavour measurements such that : - if NP particles are discovered at LHC we able study the flavour structure of the NP - we can explore NP scale beyond the LHC reach! " bq eff If there is NP at scale Λ, it will generate new operator of dimension D with coefficients proportional to Λ 4-D You could demonstrate that only operator of D=6 contribute So that in fact you have a dependence on 1/ Λ 2

32 Kaon sector

33 Flavour specific final states 1 " d fs! s0 s # ACH = $ ASL + ASL % 4 & fd! d 0 '

34 Tagging is important to separate the time evolution of mesons produced as Bs or anti-bs. In this way we obtain direct sensitivity to CP-violating phase. This phase enters with terms proportional to cos(2β s ) and sin(2β s ). Analyses which do not use flavour tagging are sensitive to cos(2β s ) and sin(2β s ), leading to a four-fold ambiguities in the determination of ϕ s. Only two-fold ambiguity TeVatron results LHCb expected to contribute

35 1.35 fb -1 CDF tagged measurement 2.8 fb -1 D0 tagged measurement Before ICHEP2008 Other measurements τ Bs,ΔΓ/Γ,A SL,A CH directly from the Likelihood given by CDF No likelihood available from D0 Conservative approach used (for details see appendix) All available measured used with and up-to-date hadronic parameters Notice that the two measurements are in agreement Other measurements are also important

36 Modeling D0 data (I) D0 data Strong phase taken also From B d J/ψ K* + SU(3) NO AMBIGUITY Used by UTFit The problem is that the singlet Component of the f is ignored. WE REINTRODUCE THE AMBIGUITY (mirroring the likelihood)

37 - Stability of the result, who is contributing more? - Is an evidence.how many sigmas? Without tagged analyses D0 and CDF Including only CDF Including only D0 Gaussian Including only D0 likelihood profile Depending of the approach used (for treating D0 data) ϕ s is away from zero from 3σ up to 3.7σ.

38 Modeling D0 data (II) DEFAULT METHOD We have the results with 7x7 correlation matrix. Fit at 7 parameters we extract 2 parameters (ΔΓ s and ϕ s ). Two others approach used to include non-gaussian tails: -Scale errors such they agree with the quoted 2σ ranges -Use the 1D profile likelihood given by D0 (fig 2).

39 ICHEP D0 released the likelihood without assumption on the strong phases Move from 1.35fb fb -1

40 Evolution of this result The two most probable peaks of last summer are now enhanced

41 Looking at the result with a different parametrization C e Bs A e + A e # 2i! s # 2 i( " s #! s ) 2i" Bs SM NP = # 2i! s ASM e NP ϕ s ~ -70 o Solution corresponding to ϕ s ~ -20 o A s NP / A s SM = (0.73 ± 0.35) ϕ s NP = (-51 ± 11) o

42