Lepton Flavour Universality tests with heavy flavour decays at LHCb - PowerPoint PPT Presentation

Lepton Flavour Universality tests with heavy flavour decays at LHCb Including a new R K result Thibaud Humair, on behalf of the LHCb collaboration Moriond EW 2019 22 nd March, 2019

LFU and b → s ℓ + ℓ − decays 1 ' 5 P SM from DHMV Yesterday: in talk presented by Carla Marin: 0.5 LHCb Run 1 analysis ◮ Interesting discrepancies in b → s µ + µ − decays, JHEP02(2016)104 0 e.g. angular analysis of B 0 → K ∗ 0 µ + µ − ; ◮ But hadronic uncertainties make interpretation 0.5 − difficult. 1 − 0 5 10 15 2 q 2 [GeV / c 4 ] Today: test L epton F lavour U niversality in b → s ℓ + ℓ − decays, in particular R K and R K ∗ : R K ( ∗ ) = B ( B → K ( ∗ ) µ + µ − ) SM = 1 . 0 B ( B → K ( ∗ ) e + e − ) ◮ All hadronic effects cancel in these ratios: immaculate theoretical predictions of R K ( ∗ ) ◮ Small deviation from 1, O (1%), due to radiative corrections ( EPJC76(2016)440 ). ⇒ any statistically significant deviation of these ratios from 1 is a sign of N ew P hysics. 2 Thibaud Humair

Previous R K ∗ and R K results (LHCb Run 1 data) 2 . 0 R K ∗ 0 LHCb: JHEP08(2017)055 LHCb: PRL113(2014)151601 1 . 5 1 . 0 0 . 5 Belle: PRL103(2009)171801 LHCb BaBar LHCb BaBar: PRD86(2012)032012 Belle 0 . 0 0 5 10 15 20 q 2 [GeV 2 /c 4 ] All LHCb results below SM expectations: − 0 . 074 ± 0 . 036 for 1 . 0 < q 2 < 6 . 0 GeV 2 , ∼ 2 . 6 σ from SM; ◮ R K = 0 . 745 +0 . 090 − 0 . 07 ± 0 . 03 for 0 . 045 < q 2 < 1 . 1 GeV 2 , ∼ 2 . 2 σ from SM; ◮ R K ∗ = 0 . 66 +0 . 11 − 0 . 07 ± 0 . 05 for 1 . 1 < q 2 < 6 . 0 GeV 2 , ∼ 2 . 4 σ from SM; ◮ R K ∗ = 0 . 69 +0 . 11 Together with b → s µµ results, R K and R K ∗ constitute an interesting pattern of anomalies, but the significance is still low. 3 Thibaud Humair

Previous R K ∗ and R K results (LHCb Run 1 data) 2 . 0 R K ∗ 0 LHCb: JHEP08(2017)055 LHCb: PRL113(2014)151601 1 . 5 Today: update of the R K measurement in 1 . 1 < q 2 < 6 . 0 GeV 2 1 . 0 In this update: 0 . 5 ◮ The analysis of 2011 and 2012 data is completely re-optimised, Belle: PRL103(2009)171801 LHCb BaBar LHCb BaBar: PRD86(2012)032012 the analysis strategy re-designed; Belle 0 . 0 0 5 10 15 20 ◮ 2015 and 2016 LHCb data are added; q 2 [GeV 2 /c 4 ] ◮ In total, updated analysis uses twice as many B ’s as the previous analysis. All LHCb results below SM expectations: − 0 . 074 ± 0 . 036 for 1 . 0 < q 2 < 6 . 0 GeV 2 , ∼ 2 . 6 σ from SM; ◮ R K = 0 . 745 +0 . 090 LHCb-Paper-2019-009 − 0 . 07 ± 0 . 03 for 0 . 045 < q 2 < 1 . 1 GeV 2 , ∼ 2 . 2 σ from SM; ◮ R K ∗ = 0 . 66 +0 . 11 − 0 . 07 ± 0 . 05 for 1 . 1 < q 2 < 6 . 0 GeV 2 , ∼ 2 . 4 σ from SM; ◮ R K ∗ = 0 . 69 +0 . 11 Together with b → s µµ results, R K and R K ∗ constitute an interesting pattern of anomalies, but the significance is still low. 3 Thibaud Humair

R K measurement at LHCb Need two inputs to measure R K : yields and efficiencies. R K = B ( B + → K + µµ ) B ( B + → K + ee ) = N ( K + µµ ) N ( K + ee ) · ε ( K + ee ) ε ( K + µµ ) Electron and muon tracks very different in LHCb: ◮ Electrons interact with material and emit µ track bremsstrahlung; ◮ worse mass and q 2 resolution; e track ◮ lower reconstruction efficiency. ◮ Better PID and trigger performances for muons. Critical aspect in the analysis: get the electron efficiencies fully under control. 4 Thibaud Humair

R K measurement at LHCb Need two inputs to measure R K : yields and efficiencies. R K = B ( B + → K + µµ ) � B ( B + → K + J /ψ ( µµ )) B ( B + → K + ee ) B ( B + → K + J /ψ ( ee )) N ( K + J /ψ ( µµ )) · N ( K + J /ψ ( ee )) N ( K + µµ ) · ε ( K + J /ψ ( µµ )) ε ( K + ee ) = · N ( K + ee ) ε ( K + µµ ) ε ( K + J /ψ ( ee )) Electron and muon tracks very different in LHCb: ◮ Electrons interact with material and emit µ track bremsstrahlung; ◮ worse mass and q 2 resolution; e track ◮ lower reconstruction efficiency. ◮ Better PID and trigger performances for muons. Critical aspect in the analysis: get the electron efficiencies fully under control. ⇒ use double ratio to cancel out most systematic uncertainties. 4 Thibaud Humair

Efficiency computation Ratio of efficiencies determined with simulation 100 carefully calibrated using control channels (L0Electron) [%] 90 LHCb selected from the data: 80 ◮ Calibration of B + kinematics; 70 60 ◮ Tracking efficiency calibration; 50 40 ε ◮ Particle ID calibration 30 (method described in EPJ T&I(2019)6:1) ; 20 10 ◮ Trigger calibration (right plot); 0 0 2000 4000 6000 8000 10000 ◮ Calibration q 2 and m ( Kee ) resolution. E ( e ) [MeV] T Ratio of efficiencies controlled to an excellent level and checked with alternative samples wherever possible. Measurement of the electron trigger ef- ficiency using B + → J /ψ ( e + e − ) K + Detailed evaluation of systematic uncertainties shows data. uncertainties at each step are < 1% 5 Thibaud Humair

Cross-check 1: r J /ψ in 1D To check efficiencies are correct, check: 1.10 〉 ψ LHCb J/ r J /ψ = B ( B → K + J /ψ ( µµ )) r LHCb-Paper-2019-009 B ( B → K + J /ψ ( ee )) = 1 . 0 , 〈 1.05 / ψ J/ r 1.00 Result: r J /ψ = 1 . 014 ± 0 . 035 ( stat . + syst . ) 0.95 0.90 1000 2000 3000 4000 5000 − ◮ Check that efficiencies are understood as a + min( ( ), ( )) [MeV/ ] p l p l c T T function of any variable: Given expected min( p T ( ℓ + ) , p T ( ℓ − ) spectra, ⇒ differential r J /ψ demonstrates it is the bias expected on RK if deviations are genuine case: r J /ψ is flat for all variables examined. rather than fluctuations is 0 . 1%. 6 Thibaud Humair

Cross-check 2: r J /ψ in 2D ◮ Pick two variables from those that can be used to parametrise the decay in LHCb frame; ◮ Select B + → J /ψ K + events in 2D bins, and compute r J /ψ in each bin: 0.30 ) [rad] 〉 LHCb ψ 4 LHCb J/ 0.25 1.1 simulation r 8 − 3 〈 l 12 / , + 0.20 16 ψ 2 l ( J/ α r 7 rare 0.15 1.0 6 J /ψ 11 0.10 10 1 15 5 14 0.9 0.05 LHCb-Paper-2019-009 9 13 0.00 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 4.0 4.5 5.0 5.5 − − − × α + + + max( p(l ), p ( l )) ( l , l ) bin number log (max( p(l ), p ( l ))) 10 Flatness of R 2 D J /ψ plots gives confidence that efficiencies are understood over all phase-space. 7 Thibaud Humair

Fit to B + → K + µ + µ − and B + → K + e + e − A single fit to the m ( K + ℓ + ℓ − ) distributions is performed to determine R K from the entire 2011-2016 dataset, taking into account all correlations (LHCb-Paper-2019-009): ) ) 2 2 c LHCb c LHCb 300 Candidates / (7 MeV/ Candidates / (24 MeV/ 100 Data Data 250 Total fit Total fit 80 + → + − + B K e e + 200 → µ + µ − + B K Part. Reco. 60 + → ψ + − + Combinatorial B J/ (e e )K 150 Combinatorial 40 100 N sig ∼ 1940 N sig ∼ 760 20 50 0 0 5200 5300 5400 5500 5600 5000 5500 6000 + µ + µ − + + − 2 m(K ) [MeV/ c ] m(K e e ) [MeV/ c 2 ] Partially reconstructed background shape in B + → K + e + e − taken from simulated B 0 → K ∗ 0 ( K + π − ) e + e − , associated systematic is 1%. 8 Thibaud Humair

R K result with 2011 to 2016 data LHCb-Paper-2019-009 Using 2011 and 2012 LHCb data, R K was: 2.0 K R LHCb R K = 0 . 745 +0 . 090 − 0 . 074 (stat.) ± 0 . 036 (syst.) , 1.5 ∼ 2 . 6 σ from SM ( PRL113(2014)151601 ). 1.0 Adding 2015 and 2016 data, R K becomes: BaBar 0.5 Belle LHCb Run 1 0.0 0 5 10 15 20 2 2 4 q [GeV / c ] 9 Thibaud Humair

R K result with 2011 to 2016 data LHCb-Paper-2019-009 Using 2011 and 2012 LHCb data, R K was: 2.0 K R LHCb R K = 0 . 745 +0 . 090 − 0 . 074 (stat.) ± 0 . 036 (syst.) , 1.5 ∼ 2 . 6 σ from SM ( PRL113(2014)151601 ). 1.0 Adding 2015 and 2016 data, R K becomes: BaBar 0.5 Belle LHCb Run 1 R K = 0 . 846 +0 . 060 − 0 . 054 (stat.) +0 . 016 LHCb Run 1 + 2015 + 2016 − 0 . 014 (syst.) 0.0 0 5 10 15 20 2 q 2 [GeV / c 4 ] ∼ 2 . 5 σ from SM. 9 Thibaud Humair

R K result with 2011 to 2016 data LHCb-Paper-2019-009 Using 2011 and 2012 LHCb data, R K was: 2.0 K R LHCb R K = 0 . 745 +0 . 090 − 0 . 074 (stat.) ± 0 . 036 (syst.) , 1.5 ∼ 2 . 6 σ from SM ( PRL113(2014)151601 ). 1.0 Adding 2015 and 2016 data, R K becomes: BaBar 0.5 Belle LHCb Run 1 R K = 0 . 846 +0 . 060 − 0 . 054 (stat.) +0 . 016 LHCb Run 1 + 2015 + 2016 − 0 . 014 (syst.) 0.0 0 5 10 15 20 2 q 2 [GeV / c 4 ] ∼ 2 . 5 σ from SM. Dominant systematic uncertainties: Fit shape, trigger calibration, B + kinematics. 9 Thibaud Humair