BABAR results on Lepton Flavour Violating (LFV) searches for τ lepton + pseudoscalar mesons & Combination of BABAR and BELLE LFV limits Swagato Banerjee 9 th International Workshop on Tau Lepton Physics 19 th nd September 006, Pisa, Italy (Page: 1/3)
Outline Lepton Flavour is violated in Nature: but by how much? Results from BABAR Collaboration: Data from Runs 1-4: Luminosity = 3. fb 1 B(τ ± µ ± γ) < 6.8 8 : Phys. Rev. Lett. 95(005)4180 B(τ ± e ± γ) < 1.1 7 : Phys. Rev. Lett. 96(006)41801 Data from Runs 1-5: Luminosity = 339.0 fb 1 τ ± l ± π 0, l ± η, l ± η (where l = e, µ): Preliminary Result Combination of results from BABAR and BELLE Collaborations: τ ± µ ± γ, τ ± e ± γ: Luminosity = 3. (BABAR) + 535.0 (BELLE) = 767. fb 1 τ ± l ± π 0, l ± η, l ± η : Luminosity = 339.0 (BABAR) + 401.0 (BELLE) = 740.0 fb 1 Constraints on New Physics LFV τ decays (Page: /3)
Introduction Lepton flavor violation (LFV) not forbidden by SM gauge symmetry most new models explicitly include LFV vertex In SM, LF is conserved for zero degenerate ν masses SM extended to include finite ν mass and mixing predicts LFV τ ν ν τ µ X W µ γ B(τ ± µ ± γ)[lee-shrock, Phys. Rev. D 16, 1444 (1977)] ( ) m 3 sin θ mix B(τ µ ν µ ν τ ) = 3α 18π M W With 3 ev, M W O( 11 ) ev O( 54 ) (θ mix : max)... many orders below experimental sensitivity! Observation for LFV unambiguous signature of new physics LFV τ decays (Page: 3/3)
LFV τ decays Predicted by many beyond SM processes... SUSY models: non-diagonal slepton mass matrix LFV Normal (Inverted) slepton hierarchy τ ± µ ± γ (τ ± e ± γ) Some models: LFV upto existing experimental bounds B(τ ± l ± γ) msugra+seesaw (EPJC14(000)319, PRD66(00)115013) 7 SUSY SO() (NPB649(003)189, PRD68(003)03301) SUSY Higgs (PLB549(00)159, PLB566(003)17) Non-Universal Z (PLB547(00)5) SM+Heavy Majorana ν R (PRD66(00)034008) Illustrative scenario: ν µ τ χ γ 8 9 9 LFV τ decays (Page: 4/3)
Higgs induced LFV @ tree level η final state enhanced due to color factor and Higgs-s s vertex MSSM + seesaw: (M.Sher, PRD66 (00) 057301) ( ) 6 ( ) 4 tanβ 0 GeV B(τ ± µ ± η) = 0.84 6 60 m A where m A is the pseudoscalar Higgs mass and tanβ = H u / H d B(τ ± µ ± η) : B(τ ± µ ± γ) = 8.4 : 1.5 LFV τ decays (Page: 5/3)
e + e τ + τ (clean environment) Search for τ ± l ± γ/p 0, where P 0 = π 0, η, η ➋ decays: η γγ, η π + π π 0 ➋ decays: η π + π η (η γγ), η ρ 0 γ (ρ 0 π + π ) l γ γ Signal Side τ lγ(γ) 1 or 3 prong τ ± e ± γ, τ ± µ ± γ ν τ ± e ± π 0, τ ± µ ± π 0 : M π 0 γγ Tag Side [115, 150] MeV τ ± e ± η, τ ± µ ± η: M η γγ [515, 565] MeV h h τ lhhγ(γ) l 1 prong γ γ ν Signal Side Tag Side τ ± e ± η, τ ± µ ± η: M η π + π π 0 [537, 558] MeV τ ± e ± η, τ ± µ ± η : M η π + π η M η ρ 0 γ [950, 965] MeV [940, 970] MeV LFV τ decays (Page: 6/3)
Signal vs. Background ln(p T miss /E beam) τ ± e ± γ/p 0 e + e τ + τ γ e + e e + e γ m ν = E/ T ag p/ T ag 0 0 0 p T miss /E beam large large small ln(p T miss /E beam) 0 0 0 τ ± e ± γ e + e τ + τ γ e + e e + e γ m ν ( GeV ) m ν ( GeV ) Bhabha rejection for searches with electron: (m ν /1.8 GeV ) ln(p T miss /E beam)/ < 1 electron / Tag-side, p CM T ag < 4.75 GeV, θlab miss m ν ( GeV ) detector acceptance Di-muon rejection for searches with muon: ln(p T miss /E beam) <.5, p CM T ag < 4.75 GeV, θlab miss detector acceptance LFV τ decays (Page: 7/3)
Signal Characteristics (GeV) M EC (Energy, Mass) daughters ( s, m τ ) (upto resolution & radiation) 1.8 1.6 Photon at edge of acceptance Initial State Radiation -1-0.5 0 0.5 E (GeV) 1 τ µγ simulation E = E rec s 0 σ( E) 50 MeV M EC (σ 9 MeV) Beam energy constrained mass after vertexing γ at µ POCA(XY) [Inv. mass: σ 4 MeV ] Signal Region: ± σ around ( E, M EC ) LFV τ decays (Page: 8/3)
Signal Characteristics (GeV) M EC (Energy, Mass) daughters ( s, m τ ) (upto resolution & radiation) 1.8 1.6-1 -0.5 0 0.5 E (GeV) 1 τ µγ simulation E = E rec s 0 σ( E) 50 MeV M EC (σ 9 MeV) Beam energy constrained mass after vertexing γ at µ POCA(XY) [Inv. mass: σ 4 MeV ] Blinded Region: ± 3 σ around ( E, M EC ) LFV τ decays (Page: 8/3)
Background estimation τ ± l ± γ: Background rate from PDF(m EC ) in ±σ band in E (GeV/c m EC.0 1.9 1.8 1.7 Events/(0.05 GeV/c ) ) 5 Data Bhabha e + e - τ + τ - τ ± e ± γ 1.6 1.5-1.0-0.5 0.0 0.5 E (GeV) 0 1.5 1.6 1.7 1.8 1.9.0 (GeV/c ) m EC τ ± l ± P 0 : Unbinned maximum likelihood fit to (m EC, E) 3σ GSB N data σ = σ PDF tot GSB 3σ PDF N data tot GSB 3σ LFV τ decays (Page: 9/3)
Modeling m EC & E PDF tot = (f e+ e /µ + µ PDF e + e /µ + µ ) +(f τ + τ PDF τ + τ )+([1 f e + e /µ + µ f τ + τ ] PDF qq) Signal MC τ ± e ± η (η γγ) PDF e+ e : Data control samples with electron Tag-side others: multi-variate PDF(s) kernel density estimated from MC 4 0 4 E (Projection) Bhabha : 6.8 Data: etag BABAR preliminary -0.5 0 GeV MEC (Projection) Bhabha : 6.8 Data: etag BABAR preliminary 8 6 4 0 5 E (Projection) MC:.5-0.5 0 Data: 18 BABAR preliminary GeV M EC (Projection) MC:.5 Data: 18 BABAR preliminary Events / ( 0.1 GeV ) Events / ( 0.05 GeV ) 8 6 4 8 6 4 BABAR preliminary -0.5 0 E (GeV) BABAR preliminary Bhabha τ + τ 0 1.5 1.6 1.7 1.8 1.9 GeV 0 1.5 1.6 1.7 1.8 1.9 GeV 1.5 1.6 1.7 1.8 1.9 (GeV) M EC LFV τ decays (Page: /3)
τ ± l ± P 0 : Results ) (GeV/c 1.8 e π 0 (π 0 γγ) e η (η γγ) e η (η π + π - π 0 ) + π - + e η (η π η) e η (η π π - γ) M EC 1.6 BABAR preliminary µ π 0 (π 0 γγ) µ η (η γγ) + µ η (η π π - π 0 ) + µ η (η π π - η) + µ η (η π π - γ) 1.8 1.6-0.5 0-0.5 0-0.5 0-0.5 0-0.5 0 E (GeV) N bkg /channel = (0.1 1.3), Total expected = 3.1, Observed = LFV τ decays (Page: 11/3)
Neighbouring & Signal boxes Decay modes ±3σ to ±11σ box ±σ box observed expected observed expected τ ± e ± π 0 (π 0 γγ) 4 5.37±1.14 0 0.17±0.04 τ ± µ ± π 0 (π 0 γγ) 43 40.68±4.3 1 1.33±0.15 τ ± e ± η (η γγ) 4 4.99±1.18 0 0.0±0.05 τ ± µ ± η (η γγ) 0 17.36±.1 1 0.67±0.08 τ ± e ± η (η π + π π 0 ) 0 0.64±0.3 0 0.0±0.01 τ ± µ ± η (η π + π π 0 ) 3.01±0.41 0 0.08±0.0 τ ± e ± η (η π + π η) 0 0.14±0.14 0 0.01±0.01 τ ± µ ± η (η π + π η) 1.4±0.47 0 0.07±0.0 τ ± e ± η (η ρ 0 γ).97±0.54 0 0.11±0.03 τ ± µ ± η (η ρ 0 γ) 13 11.06±0.65 0 0.4±0.03 Good modeling of background in data No evidence of signal LFV τ decays (Page: 1/3)
Upper Limit B 90 UL = N 90 UL /(N τ ε) ε: high statistics signal MC simulated for different Data-taking periods ε = Trigger. Reco. Topology. PID. Cuts. Signal Box Cumulative: 90% 70% 70% 50% 50% 50% 90% 63% 44% % 11% ~5% σ τ + τ (.6 GeV) 0.89 nb, L 339 fb 1 ( BABAR Summer 006) N τ = L σ τ + τ 6.0 8 NUL: 90 90% C.L. Upper Limit for (N obs, N bkg ) from Data Naive Sensitivity : N 90 UL =.3 N bkg, N bkg O(1) B 90 UL O( 7 ) LFV τ decays (Page: 13/3)
B-Factories: Status Channel BABAR BELLE BUL 90 ) L ( fb 1 ) BUL 90 ) L ( fb 1 ) τ ± e ± γ 1.1 3. 1. 535.0 PRL96(006)41801 Tau06: Preliminary result τ ± µ ± γ 0.7 3. 0.5 535.0 PRL95(005)4180 Tau06: Preliminary result τ ± e ± π 0 1.3 339.0 0.8 401.0 Tau06: Preliminary result ICHEP06: hep-ex/0609013 τ ± µ ± π 0 1.1 339.0 1. 401.0 Tau06:Preliminary result ICHEP06: hep-ex/0609013 τ ± e ± η 1.6 339.0 0.9 401.0 Tau06: Preliminary result ICHEP06: hep-ex/0609013 τ ± µ ± η 1.5 339.0 0.7 401.0 Tau06: Preliminary result ICHEP06: hep-ex/0609013 τ ± e ± η.4 339.0 1.6 401.0 Tau06: Preliminary result ICHEP06: hep-ex/0609013 τ ± µ ± η 1.4 339.0 1.3 401.0 Tau06: Preliminary result ICHEP06: hep-ex/0609013 LFV τ decays (Page: 14/3)
B-Factories: Combinations Signal and Background PDF paramerterizations vary Signal regions not same for all analysis: σ ellipse vs. rectangle BABAR uses m EC, BELLE uses m inv : resolutions different Efficiency combined by weighting with luminosity Observed & background events added (asymmetric errors averaged) 6 Toy MC generated: Poisson distribution with mean (s + b) where the background, b, and signal, s, are each drawn randomly from Gaussian distributions describing their respective PDFs. Mean and s.d. of background Gaussian: (b ± σ b ) Mean and s.d. of signal Gaussian: Lσ τ τ B UL (ε ± σ ε ) Vary B UL till % of the sample yields a number of events < n obs = the number of events observed in the data upper limit at 90% confidence level (C.L.) [Ref: Cousins-Highland NIM A30, 331 (199), Barlow CPC 149, 97 (00)] LFV τ decays (Page: 15/3)
B-Factories: Combinations τ ± e ± γ Luminosity ε Background events (%) Expected Observed BABAR 3. fb 1 4.70 ± 0.9 1.9 ± 0.4 1 BELLE 535.0 fb 1.99 ± 0.13 5.14 +.6 1.9 5 BABAR & BELLE 767. fb 1 3.51 ± 0.13 7.0 ±.3 6 B(τ ± e ± γ) < 9.4 8 @ 90% C.L. τ ± µ ± γ Luminosity ε Background events (%) Expected Observed BABAR 3. fb 1 7.4 ± 0.65 6. ± 0.5 4 BELLE 535.0 fb 1 5.07 ± 0.0 13.9 +3.3.6 BABAR & BELLE 767. fb 1 5.78 ± 0.4 0.1 ± 3.0 14 B(τ ± µ ± γ) < 1.6 8 @ 90% C.L. LFV τ decays (Page: 16/3)
B-Factories: Combinations τ ± e ± π 0 Luminosity ε Background events (%) Expected Observed BABAR 339.0 fb 1.83 ± 0.5 0.17 ± 0.04 0 BELLE 401.0 fb 1 3.93 ± 0.18 0.0 ± 0.0 0 BABAR & BELLE 740.0 fb 1 3.4 ± 0.15 0.37 ± 0.0 0 B(τ ± e ± π 0 ) < 4.4 8 @ 90% C.L. τ ± µ ± π 0 Luminosity ε Background events (%) Expected Observed BABAR 339.0 fb 1 4.75 ± 0.37 1.33 ± 0.15 1 BELLE 401.0 fb 1 4.53 ± 0.0 0.58 ± 0.34 1 BABAR & BELLE 740.0 fb 1 4.63 ± 0.0 1.91 ± 0.37 B(τ ± µ ± π 0 ) < 5.8 8 @ 90% C.L. LFV τ decays (Page: 17/3)
B-Factories: Combinations τ ± e ± η Luminosity ε Background events (%) Expected Observed BABAR 339.0 fb 1.1 ± 0.0 0. ± 0.05 0 BELLE 401.0 fb 1.86 ± 0.14 0.78 ± 0.59 0 BABAR & BELLE 740.0 fb 1.5 ± 0.1 1.00 ± 0.59 0 B(τ ± e ± η) < 4.5 8 @ 90% C.L. τ ± µ ± η Luminosity ε Background events (%) Expected Observed BABAR 339.0 fb 1 3.59 ± 0.41 0.75 ± 0.08 1 BELLE 401.0 fb 1 4.06 ± 0.0 0.64 ± 0.38 0 BABAR & BELLE 740.0 fb 1 3.85 ± 0. 1.39 ± 0.38 1 B(τ ± µ ± η) < 5.1 8 @ 90% C.L. LFV τ decays (Page: 18/3)
B-Factories: Combinations τ ± e ± η Luminosity ε Background events (%) Expected Observed BABAR 339.0 fb 1 1.53 ± 0.16 0.1 ± 0.03 0 BELLE 401.0 fb 1.15 ± 0.1 0.00 +0.41 0.0 0 BABAR & BELLE 740.0 fb 1 1.86 ± 0. 0.1 ± 0.1 0 B(τ ± e ± η ) < 9.0 8 @ 90% C.L. τ ± µ ± η Luminosity ε Background events (%) Expected Observed BABAR 339.0 fb 1.18 ± 0.6 0.49 ± 0.04 0 BELLE 401.0 fb 1.45 ± 0.13 0.3 +0.33 0.3 0 BABAR & BELLE 740.0 fb 1.33 ± 0.14 0.7 ± 0.8 0 B(τ ± µ ± η ) < 5.3 8 @ 90% C.L. LFV τ decays (Page: 19/3)
B-Factories: Projections B 90 UL = N 90 UL /(N τ ε) τ ± µ ± γ search: Optimize NUL 90 /ε for expected UL @ 90% C.L. N 90 UL Baseline: B 90 UL 1. 7 (BaBar expected @ 3. fb 1 ) Background free search.3 N obs O(1) Background limited search L B 90 UL 1/L 1/ L BaBar, Belle: 1 ab 1 each (008) L ( ab 1 ) 0.5 (Now) 1.0 50 B 90 UL ( 8 ).5 (5) 0.05 (0.7) Super B-Factory: 50 ab 1 B 90 UL < O ( ) / O ( 9 ) no/with Background LFV τ decays (Page: 0/3)
τ µγ & S φks SUSY SU(5) GUT: Flavour changing right-handed currents Correlations between CP asymmetry in b-s penguins and τ µγ BABAR BELLE BABAR & BELLE J. Hisano, Y. Shimizu (PLB565(003)183) tan β =, A 0 = 0, m 0 m 1/ m νr = 5 14 GeV, m ντ = 5 ev Current measurement: S(B φk S ) = 0.47 ± 0.19 (HFAG, 005). More sensitive B(τ ± µ ± γ) < 1.6 8 exclude some regions. LFV τ decays (Page: 1/3)
Search for Supersymmetric Higgs BABAR and BELLE upper limits @ 95% C.L. on search of τ ± µ ± η: competetive with direct searches at CDF: τ + τ (3 pb 1 ), D0: b b (60 pb 1 ), τ + τ (35 pb 1 ) complementary to region excluded by LEP BELLE BABAR & BELLE and no-mixing: stop mixing benchmark models (M. Carena et.al, hep-ph/9913) m max h LFV τ decays (Page: /3)
Summary of LFV Limits - -4 B(τ e γ) MARKII CRYSTAL BALL ARGUS DELPHI - -4 MARKII B(τ µ γ) ARGUS DELPHI -6-8 - msugra + Seesaw SUSY + SO() SUSY + Higgs CLEO BELLE BABAR BELLE BABAR+BELLE Technicolor + Z 1 1980 1990 000 0-6 -8 - msugra + Seesaw SUSY + SO() SUSY + Higgs CLEO BELLE BABAR BELLE BABAR+BELLE Technicolor + Z 1 1980 1990 000 0 Channel BABAR BELLE BABAR & BELLE BUL 90 ) L ( fb 1 ) BUL 90 ) L ( fb 1 ) BUL 90 ) L ( fb 1 ) τ ± e ± γ 1.1 3. 1. 535.0 9.4 767. τ ± µ ± γ 0.7 3. 0.5 535.0 1.6 767. τ ± e ± π 0 1.3 339.0 0.8 401.0 4.4 740.0 τ ± µ ± π 0 1.1 339.0 1. 401.0 5.8 740.0 τ ± e ± η 1.6 339.0 0.9 401.0 4.5 740.0 τ ± µ ± η 1.5 339.0 0.7 401.0 5.1 740.0 τ ± e ± η.4 339.0 1.6 401.0 9.0 740.0 τ ± µ ± η 1.4 339.0 1.3 401.0 5.3 740.0 LFV τ decays (Page: 3/3)