Lepton Universality Tests in Kaon Decays at NA

Lepton Universality Tests in Kaon Decays at NA 62 Francesca Bucci, INFN Sezione di Firenze on behalf of the NA 62 Collaboration The XIth International Conference on Heavy Quarks and Leptons June 11 -15, 2012, Prague, Czech Republic

Introduction q q q Precise measurements of FCNC processes in the B sector have severely restricted the parameters space of new-physics (NP) models Experiments at the LHC have started a direct exploration of the physics in the Te. V range In this scenario, the study of Lepton Flavor Universality (LFU) in kaon decays is equally interesting and fully complementary to search for NP effects Heavy Quarks and Leptons 2012, 13 June F. Bucci 2

K+ l+ Sensitivity to New Physics q q q Leptonic kaon decays within the SM are mediated by a charged current at tree level The natural size of the non-standard contributions depends on the particular BSM scenario In Models with 2 Higgs Doublet (2 HDM-II including SUSY) sizable charged Higgs (H ) exchange contributions obstructed by hadronic uncertainties (f. K ) Heavy Quarks and Leptons 2012, 13 June F. Bucci 3

RK=G(K→e )/G(K→ ) in the SM q q In the ratio RK =G(K→e /K→ ) hadronic uncertainties cancel The SM prediction of RK has reached <0. 1% precision [V. Cirigliano and I. Rosell Phys. Rev. Lett. 99 (2007) 231801] helicity suppression factor q RK is the correction due to the Inner Bremsstrahlung part of the radiative K →e process Heavy Quarks and Leptons 2012, 13 June F. Bucci 4

Radiative K→e Decays q In K→e (Ke 2 ), can be produced via internal bremsstrahlung (IB) or directemission, the latter being dependent on the hadronic structure (SD) IB K± q q SD e± K± e e± e RK is defined to be inclusive of the IB, ignoring however the SD contributions To compare data with SM prediction the SD contribution must be carefully estimated and subtracted. Heavy Quarks and Leptons 2012, 13 June F. Bucci 5

RK beyond the SM [A. Masiero, P. Paradisi, R. Petronzio Phys. Rev. D 74 (2006) 0011701] q q Charged Higgs bosons (H ) appearing in any model with two Higgs doublets (including SUSY case) can contribute at tree level Such contribution does not affect the ratio RK LFV couplings, present at one loop level, can contribute to RKSM at the % level Limited by recent Bs + - measurement [Fonseca, Romão, Teixiera, ar. Xiv: 1250. 1411] slepton mixing 6

The NA 62 Experiment NA 62 (RK) 2007 -2008: K e 2 , K 2 data taking NA 62 (K+ + ) 2007 -2013: design & construction 2014 ? : K+ + data taking The NA 62 Collaboration: after the long CERN accelerators shutdown Birmingham, Bratislava, Bristol, CERN, Dubna, Fairfax, Ferrara, Florence, Frascati, Glasgow, Liverpool, Louvain, Mainz, Merced, Moscow, Naples, Perugia, Pisa, Prague, Protvino, Rome II, San Luis Potosì, SLAC, Sofia, Turin Heavy Quarks and Leptons 2012, 13 June F. Bucci 7

NA 62 (RK) Detector Primary SPS protons (400 Ge. V/c) 1. 8 1012 /SPS spill Unseparated secondary positive and/or negative hadronic beam (p=74. 0 1. 4 Ge. V/c) Composition: K ( ) = 5% (63 %) K decaying in vacuum tank 18% Helium filled tank Main subdetectors: LKr electromagnetic calorimeter: s. E/E = (3. 2/√E 9. 0/E 0. 42)% (E in Ge. V) sx =sy = (4. 2/ √E + 0. 6)mm (E in Ge. V) Magnet Hodoscope: Fast trigger for charged particle and timing for the event (st 150 ps) Magnetic spectrometer: sp/p = (0. 48 0. 009 p)% (p in Ge. V/c) Heavy Quarks and Leptons 2012, 13 June beam F. Bucci 8

NA 62 (RK) Data Taking The data taking strategy has been optimized to measure the two main backgrounds in the Ke 2 sample: q Beam halo muons q K 2 decays with a muon misidentified as an electron Data taking: q 4 months in 2007 q 2 weeks in 2008 (used for systematics studies) Data samples: q K+, K- samples q with different background conditions : with lead wall (Pb), without lead wall (no. Pb) (see next) K+ (no. Pb), K+ (Pb), K- (no. Pb), K- (Pb) Heavy Quarks and Leptons 2012, 13 June F. Bucci 9

Measurement Strategy q Ke 2, K 2 collected simultaneously: § § No dependence on K flux measurement; Cancellation of several systematic effects in the ratio at first order acceptance # background events LKr trigger efficiency downscaling factor of K 2 # signal events q q LKr readout efficiency MC simulations used to a limited extent: § q measured PID efficiency Acceptance correction and the geometrical parts of the acceptances for most bkg processes PID, trigger, read out efficiencies and muon halo bkg are measured directly from data Analysis performed in bins of the reconstructed lepton momentum Heavy Quarks and Leptons 2012, 13 June F. Bucci 10

Trigger Logic Minimum bias (high efficiency, but low purity) trigger configuration used q q Ke 2 condition: Q 1 ELKr 1 Trk K 2 condition: Q 1 1 Trk/D, downscaling factor D=150 Heavy Quarks and Leptons 2012, 13 June F. Bucci 11

Ke 2 vs K 2 Selection q q q Geometry § One reconstructed charged track § 13 < p < 65 [Ge. V/c] § Geometrical acceptance cut § Cut on K decay vertex position § Photon veto using LKr Kinematics § Missing mass: Mmiss 2(l)=(PK-Pl)2 § Sufficient Ke 2/K 2 separation up to 30 Ge. V/c Particle ID (ELKR/pspectr) § e : (0. 9 to 0. 95) < E/p < 1. 1 § m : E/p < 0. 85 suppression in e sample: 106 Heavy Quarks and Leptons 2012, 13 June F. Bucci 12

K 2 Background in Ke 2 Sample Muon ‘catastrophic’ energy loss in or in front of the LKr is the largest bkg source The probability of mis-identify a muon as an electron P e 3 10 -6 (and momentum dependent) Measurement of P e on data q To avoid electron contamination from muon decay ( 10 -4) a 9. 2 X 0 thick lead wall covering 18% of the acceptance was installed between the HOD planes for 50% of the run time q K 2 candidates : track traversing Pb, p>30 Ge. V/c, E/p>0. 95 (positron contamination < 10 -8) q P e is modified by the Pb wall due to: § § e ionization losses in Pb (low p) bremsstrahlung in Pb (high p) Heavy Quarks and Leptons 2012, 13 June F. Bucci 13

Muon Misidentification The correction f. Pb= P e/ P e. Pb evaluated with a dedicated Geant 4 -based simulation Correction for Pb: f. Pb = P e / P e Pb (Pb wall installed) vs momentum MC precision f. Pb/f. Pb = 2% MC precision P e /P e = 10% Result : B/(S+B) = (5. 64 0. 20)% Heavy Quarks and Leptons 2012, 13 June F. Bucci 14

Muon Halo Background in Ke 2 Sample The muon sweeping system was optimized for the positive beam q Muon halo background measured on the data § q q K+ (K- ) only sample used to measure muon halo background in K- (K+ ) sample Control samples normalized to the data in the Mmiss 2 region populated mainly by beam halo events (-0. 3 < Mmiss 2 < - 0. 1 (Ge. V/c 2)2 ) Strongly depends on decay vertex position and track momentum Ke 2 - (no. Pb) pe+ (Ge. V/c) pe- (Ge. V/c) Ke 2+ (no. Pb) z vertex (m) q q Regions delimited by solid line are excluded Probability to reconstruct a Ke 2 candidate due to a decay of an opposite sign kaon has been taken into account Result : B/(S+B) = (2. 11 0. 09)% Heavy Quarks and Leptons 2012, 13 June F. Bucci 15

K±→e± Background in Ke 2 Sample q q The SD+ (positive photon helicity) component peaks at high electron momentum in the K rest frame and is therefore kinematically similar to the Ke 2 decay SD- decays and the interference between the IB and SD processes are negligible Ee (Me. V) SD contribution of the K±→e± decays must be carefully estimated and subtracted SD+ IB [J. Bijnens, G. Colangelo, G. Ecker, J. Gasser, ar. Xiv: hep-ph/9411311] q The SD+ background contribution has been estimated by using MC simulation based on the measured K±→e± (SD+) differential decay rate [F. Ambrosino et al. , Eur. Phys. J. C 65 (2010) 703] Result : B/(S+B) = (2. 60 0. 11)% Heavy Quarks and Leptons 2012, 13 June F. Bucci SD- E (Me. V) 16

Ke 2 Sample Ke 2 candidates 145, 958 K± e ± candidates B/(B+S) = (10. 95 0. 27)% e± ID efficiency = (99. 28 0. 05)% Heavy Quarks and Leptons 2012, 13 June F. Bucci 17

Ke 2 Background Ke 2 candidates and backgrounds in momentum bins Decay B/(S+B) K± ± (5. 64 ± 0. 20)% K± ± ( e) (0. 26 ± 0. 03)% K± e± (SD+) (2. 60 ± 0. 11)% K± 0 De± (0. 18 ± 0. 09)% K± ± 0 D (0. 12 ± 0. 06)% Wrong sign K (0. 04 ± 0. 02)% Beam halo (2. 11 ± 0. 09)% Total Heavy Quarks and Leptons 2012, 13 June K+ (no. Pb) (10. 95 ± 0. 27)% F. Bucci 18

K 2 Sample and Background The only significant background source in the K 2 sample is the beam halo 42. 817 106 K± ± candidates (collecting with downscaling factor D=150) B/(S+B) = (0. 50 ± 0. 01)% K+ (no. Pb) Heavy Quarks and Leptons 2012, 13 June F. Bucci 19

NA 62 Result (full data set) Fit over 40 RK measurements (4 data samples 10 momentum bins) including correlations: c 2/ndf=47/39 RK=(2. 488 ± 0. 007 stat ± 0. 007 syst ) x 10 -5=(2. 488 ± 0. 010) x 10 -5 Heavy Quarks and Leptons 2012, 13 June F. Bucci 20

Uncertainties on RK Source Heavy Quarks and Leptons 2012, 13 June RK 105 Statistical 0. 007 K 2 background 0. 004 K± e ± (SD+) background 0. 002 K± 0 e ± , K± 0 backgrounds 0. 003 Muon halo background 0. 002 Matter composition in the spectrometer 0. 003 Acceptance correction 0. 002 Spectrometer alignment 0. 001 Electron identification efficiency 0. 001 1 -track trigger efficiency 0. 001 LKr readout inefficiency 0. 001 Total Systematic 0. 007 Total 0. 010 21

RK World Average July 2011 world average : RK= (2. 488 ± 0. 009) 10 -5 ( RK/RK= 0. 4%) For non-tiny values of D 13 the sensitivity to H± in RK is strong b s PDG 2010 (KLOE result): RK= (2. 493 ± 0. 031) 10 -5 ( RK/RK= 1. 3%) Heavy Quarks and Leptons 2012, 13 June F. Bucci B→t [HFAG 2010] R 23 [Eur. Phys. J. C 69 (2010) 399 ] b→s [HFAG 2010] 22

Future Prospects NA 62 (K+ + ) data taking foreseen in 2014 after the long CERN accelerators shutdown q q Hermetic veto (large-angle and small-angle veto counters) will strongly decrease the Ke 2 background. Beam spectrometer (beam tracker plus Cherenkov beam tagger) will allow time correlation between kaons and decay products. Beam halo background expected to be reduced to negligible level. Required statistical uncertainty 0. 05% 1 month of data taking sufficient for such RK measurement Heavy Quarks and Leptons 2012, 13 June F. Bucci 23

Conclusions q q The analysis of the full data sample taken by NA 62 is completed The measured value is RK=(2. 488 ± 0. 010) x 10 -5, reaching a record level of accuracy of 0. 4% The combined experimental accuracy has improved by an order of magnitude since 2008, but is still an order of magnitude bigger than the SM prediction precision Excellent prospects for future measurements of RK within the NA 62 experimental program Heavy Quarks and Leptons 2012, 13 June F. Bucci 24