1. Ke2/Kmu2 and LFV tests Crimea2008 Yalta 27/09/08 – 04/10/08 Andrey Zinchenko JINR, Dubna On behalf of NA62 Collaboration: Bern, Birmingham, CERN, Dubna, Fairfax, Ferrara, Florence, Frascati, Mainz, Menlo Park, Merced, Moscow, Naples, Perugia, Protvino, Pisa, Rome, Saclay, San Luis Potosi, Sofia, Triumf, Turin

2. K + →e + ν e in the Standard Model In the Standard Model(SM), the decay K + →e + ν e (Ke2) is strongly helicity-suppressed compared to the muonic channel K + →μ + ν μ (Kμ2). The positron e + must be left-handed due to angular momentum conservation – thus the coupling of the W + -Boson is suppressed by (1– β). In the SM, the ratio of the decay rates can be predicted very precisely and is given as where δRQED= -3.6% is a correction due to Kl2γ(IB) and virtual photon processes (Cirigliano, Rosell, JHEP 0710:005, 2007).

3. K + →e + ν e in Supersymmetry In Supersymmetry (SUSY), K + →e + ν e can proceed via exchange of a charged Higgs H + instead of W +, but in this case, at tree level, the ratio Γ(K + →e + ν e )/Γ(K + →μ + ν μ ) remains unchanged. However, loop effects are predicted to lead to lepton flavour violating (LFV) couplings lH ± ν τ, which noticeably change only the rate of K + →e + ν e (Masiero, Paradisi, Petronzio, PRD 74, 2006): The LFV parameter Δ 13 should be of o(10 -4 – 10 -3 ) (similar to LFV in the neutrino sector). For given values of M H ± and tan β, deviations up to a few percent on R K are possible. As a result of the helicity suppression of the SM contribution, the Ke2 decay rate is sensitive to New Physics. Tree-level diagram for K + →l + ν Loop diagram for K + →e + ν τ

4. Entering the precision realm for R K Main actors (experiments) in the challenge to push down precision on R K : KLOE Preliminary result with 2001 – 2005 data: R K = (2.55 ± 5 stat ± 5 syst )×10 -5 from 8000 Ke2 candidates (3% accuracy), perspectives to reach 1% error after analysis completion. NA48/2 Preliminary result with 2003 data: R K = (2.416 ± 43 stat ± 24 syst )×10 -5, from ~ 4000 Ke2 candidates, statistical error dominating (2% accuracy). Preliminary result with 2004 data: R K = (2.455 ± 45 stat ± 41 syst )×10 -5, from ~ 4000 Ke2 candidates from special minimum bias run (3% accuracy). NA62 Collected ~ 100,000 Ke2 events in dedicated 2007 run, aims at breaking the 1% precision wall, possibly reaching < ~ 0.5 %.

5. Experimental Status PDG2006 + three new preliminary measurements by NA48/2 and KLOE give 1.3% total accuracy and are in agreement with SM prediction: R K =(2.475±0.032)×10 -5 SUSY limits by R K measurements (M H,GeV)

6. NA48/2 Experiment Unseparated, simultaneous K ± highly collimated beams, designed to precisely measure K ± →π + π - π ± (π 0 π 0 π ± ) dalitz-plot density: · p K ~ 60 GeV/c, σ p ~ 3 Gev/c (3.8% p-bite) · spot of ~ 5 mm width DCH1 entrance Track decay products with 4 DCH’s: · P T kick of 121 MeV/c after DCH2 · σ p /p ~ 1.02% 0.044% p [GeV/c]

7. NA48/2 Experiment Scintillator hodoscope: establish event time (σ~150 ps), initiate trigger LKr calorimeter: efficient vetoing, e.m. energy resolution –σ E /E = 3.2%/ √E[GeV] 9%/E[GeV] 0.42% –σ x,y = 4.2mm/√E[GeV] 0.6 mm, granularity of ~13,000 2×2 cm2 cells Hadron calorimeter, Muon veto system Analysis starting samples: –1-track minimum bias trigger, acquired during 12-hour run in 2003 –1-track triggers, acquired during 56-hour special run in 2004 Ke2 trigger: 1 track (hodoscope) + E LKr > 10 GeV Kμ2 trigger: 1 track (hodoscope), possibly downscaled

8. Analysis of K e2 /K μ2 at NA48/2 Main analysis steps for K e2, K μ2 selection: One track with 15 < p < 50 GeV, passing acceptance + quality cuts DataMC K e2 p (GeV)

9. Analysis of K e2 /K μ2 at NA48/2 Main analysis steps for Ke2, K μ 2 selection: One track with 15 < p < 50 GeV, passing acceptance + quality cuts Data MC K e2 Z vtx (cm)

10. Analysis of K e2 /K μ2 at NA48/2 Main analysis steps for Ke2, K μ 2 selection: Associate electron track to LKr cluster, compute E/P Veto events with extra energy in LKr above 1.5 GeV Calculate missing mass squared, assuming electron hypothesis Data (2004) Data (2003) Ke2 E/P

11. Analysis of K e2 /K μ2 at NA48/2 Main bkg to Ke2 selection due to Kμ2 with μ faking e cluster, O(5×10 -6 ): –At high momentum, p > ~30 GeV/c, can’t reject Kμ2 using kinematics –Evaluate probability of “catastrophic release” from selected μ sample –Statistics of μ sample induces a systematic error for bkg subtraction For the analysis of 2004 data: –1.59% systematic error due to subtraction of Kμ2 bkg for Ke2 –1.85% statistical error due to Ke2 counts p (GeV) Nμ bkg (Evts)MM 2 (e) (GeV/c 2 )

12. R K – Present experimental status Recent (preliminary) results improved greatly with respect to 2006 PDG World average, RK= (2.457±32)×10 -5, agrees with SM

13. NA62 Physics Run 2007 Beam + Detector: Charged kaons are produced by 400 GeV/c protons impinging on a Be target. An achromatic system of dipole magnets selects charged particles ( ~ 90 % π ±, 10 % K ± ) with momentum of (75±2) GeV/c. The decay volume is housed in a ~ 110 m long vacuum tank. The decay products are measured in the detector of the former NA48/2 experiment. Data taking: 120 days run period, June – October 2007. Minimum Bias-Trigger: Require minimum number of hits in drift chambers and hodoscope + energy deposition in the LKr calorimeter > 10 GeV. > 120000 K ± → e ± ν decays recorded Goal of NA62: Measurement of R K with accuracy < 0.5 %

14. K e2 Identification and Background Rejection Rejection of K ± →μ ± ν: Events with p track < 35 Gev/c (40 %): Kinematic separation by “Missing mass” M 2 = (P K – P track ) 2 Events with p track > 35 GeV/c (60 %): Separation by ratio E/p: Electrons have E/p ~ 1, Muons E/p << 1. P Most relevant background: Decays of K ± → μ ± ν (Br = 63 %). All other decays and processes are excluded kinematically or measured in data (e.g. beam halo background).

15. “Katastrophic Bremsstrahlung” Problem: “Katastrophic Bremsstrahlung” A very small fraction of muons deposit all energy in the calorimeter by emission of a high energy bremsstrahlung photon.  A few 10 -6 of muons with E/p > 0.95 being mis-identified as electrons. Solution: Cover part of the calorimeter with lead plate (see sketch), allowing only muons to pass.  Direct measurement of muon background by explicit E/p measurement for muons.

16. Background from bremsstrahlung photons Number of events P track (GeV/c)

17. K e2 Identification and Background Rejection Use MC for checks of any bias due to the presence of Pb wall Can use Pb wall data to validate results of G4 simulation Data-MC agreement looks good

18. Total Background Estimation Will subtract background from beam halo directly measuring it: Use K-less data normalized in side-band of negative M miss 2 Expect ~ 7% total background to K e2, dominated by K μ2

19. Analysis of K e2 /K μ2 at NA62 Data taking lasted 4 months: NA62 has the world largest data set, exceeding 100000 K e2 events Largest homogeneus period (K+ data), lasted ~1 month: > 40000 K e2 events

20. Conclusions Golden LFV observable is RK: Preliminary mmt 2% from NA48/2, 3% from KLOE Results from preliminary measurement in agreement with SM Huge effort to enter the precision realm for RK: KLOE expects to push error to ~1.3%, after analysis completion NA62 dedicated 2007 run aims at reaching <0.5% error