Search for tau-e (tau-mu) flavor mixing at a future collider

1. Search for tau-e (tau-mu) flavor mixing at a future collider Shinya KANEMURA (Osaka Univ.) with Yoshitaka KUNO, Toshihiko OTA(Osaka Univ.) Masahiro Kuze (Tokyo Inst. Tech.) CPの破れと物質創生＠基礎物理学研究所 Jan 12-14. 2005

2. Contents • Introduction • Tau associated LFV processes • Lepton flavor violating Yukawa coupling • The bound from current data • Search for tau associated LFV at future colliders • h →τμ　（τe）　at a LC • e N (μN)→τX at a LC (a neutrino factory) • Summary SK, Matsuda, Ota, Shindou, Takasugi, Tsumura SK, Kuno, Kuze, Ota

3. Introduction • LFV is a clear signal for physics beyond the SM. e ⇔μμ⇔ττ⇔e • Neutrino oscillation may indicate LFV among charged leptons. • In SUSY models, LFV can naturally appear. Borzumati, Masiero Hisano et al.

4. LFV in SUSY • It is known that sizable LFV can be induced at loop due to slepton mixing • Up to now, however, no LFV evidence has been observed at experiments. μ→e γ, μ→eee, …. • Large MSUSY? so that the SUSY effects decouple? Even in such a case, we may be able to search LFV through the Higgs boson mediation, which does not necessarily decouple for a large MSUSY limit

5. In this talk, we discuss tau-associated LFV in SUSY models τ⇔e & τ ⇔μ • The Higgs mediated LFV is proportional to the Yukawa coupling ⇒ Tau-associated LFV processes. Different behavior from LFV in theμ⇔e mixing. • It is less constrained by current data as compared to theμ⇔e mixing μ→eγ 1.2 ×10＾（－11）　　　　　　　 μ→３e1.1×10＾（－12） μTi→eTi6.1 ×10＾（－13） τ→μγ 3.1 ×10＾（－7） τ→３μ 1.4-3.1 ×10＾（－7） τ→μη3.4 ×10＾（－7）　

6. LFV in SUSY LFV is induced at one loop due to slepton mixing • Gauge mediation : • Higgs mediation : Higgs mediation does not decouple in the large MSUSY limit

7. Babu, Kolda; Dedes,Ellis,Raidal; Kitano, Koike, Okada LFV Yukawa coupling Slepton mixing induces LFV in SUSY models. κij= Higgs LFV parameter

8. Experimental Bound on κ32 The strongest bound on κ32comes from the result of τ→μη. For κ３１,similar bound is obtained because of the similar experimental results for τ→μηand τ→eη.

9. A source of slepton mixingin the MSSM+RN • Slepton mixing induces both the Higgs mediated LFV and the gauge mediation. • Off-diagonal elements can be induced in the slepton mass matrix at low energies, even when it is diagonal at the GUT scale. • RGE

10. The correlation between Higgs mediated LFV and gauge mediation For relatively low mSUSY, the Higgs mediated LFV is constrained by the data of gauge mediated LFV. For mSUSY >O(1)TeV, the gauge mediation becomes suppressed, while the Higgs mediated LFV can be large.

11. The non-decoupling property of the Higgs LFV coupling (κij ) Let us consider that MSUSY is as large as O(1) TeV with a fixed value of |μ|/MSUSY Gauge mediated LFVis suppressed, the Higgs-LFV coupling κij can be sufficiently large . Babu,Kolda; Brignole, Rossi mSUSY ～　O(1) TeV

12. Search for Higgs mediated τ- e & τ- μ mixing • Tau’s rare decays at B factories. τ→eππ　（μππ） τ→eη (μη) τ→μe e (μμμ)、　…. In nearfuture, τ decay searches will improve the upper limit by about 1 order of magnitude. • We here discuss the other possibilities. • Higgs decays into a tau-mu or tau-e pair • The DIS process e Ｎ （μＮ） →τＸ　 by a fixed target experiment at a LC (μＣ)

13. Higgs boson decay After the Higgs boson is discovered, we can consider the possibility to measure the LFV Yukawa couplings directly from the decay of the Higgs bosons. • LHCAssamagan et al; Brignole, Rossi • LC SK, Matsuda, Ota, Shindou, Takasugi, Tsumura Search for h →τμ (τe) atLC: • Simple kinematic structure (Esp. Higgssrahlung process) • Precise measurement of the lightest Higgs boson: property (mh,Γ,σ,Br,…) will be thoroughly measured • Less backgrounds

14. A Linear Collider　（LC） collision GLC （former JLC）(201x?-) 1st stage 2nd stage Roadmap report JLC

15. Higgs Production at a LC via gauge interaction At low energies, the Higgsstrahlung process is dominant. In 2HDM (MSSM),

16. A LC = A Higgs Factory • A electron-positron collider • Energy Ecm= 200 – 1000 GeV • Luminosity L = 500 – 1000 fb^(-1) • 10^5-10^6 Higgs bosons can be produced. A Higgs factory! • Less backgrounds • Simple kinematic structure

17. Decay branching ratio When mA is large, the experimental bound is relaxed, and branching ratio of 10^(-4)-10^(-3) is allowed.

18. Branching ratios

19. Signal The process can be identified by using Z recoil: • Theτmomentum is reconstructed by using Ecm, mh, pZand pμ It is not required to measure τ • The # of the signalevent 11 event for leptonic decay of Z 118 event for hadronic decay

20. Backgrounds The background is huge, but most of them can be cut by appropriate kinematic cuts. Except for when

21. The fake signal In order to resuce the fake signal events, it is important to determine Eh with high precision from pZ, Ecm, pμ. Strongly depend on the resolusion of pZ and pμ, beam energy spread rate, …

22. Feasibility • Resolution of Z momentum • Signal / Fake 118 / 230 events (Z →jj、 δ＝３GeV) 11 / 8 events (Z→ll, δ＝1GeV) For some specific parameter region, h →τμ　（τe） can be studied at a LC.

23. Alternative process for search of the Higgs LFV coupling. • At future ν factories (μ colliders) , 10^20 muons of energy 50 GeV (100-500GeV) can be available. DIS μＮ→τＸprocess • At a LC (Ecm=500GeV L=10^34/cm^2/s) 10^22 of 250GeV electrons available. DIS process eＮ→τＸprocess A fixed target experiment option of LC

25. Cross section in SUSY model • Each sub-process e q (μq) →τq is proportional to the down-type quark masses. • For the energy > 60 GeV, the total cross section is enhanced due to the b-quark sub-process E ＝50 GeV 10^(-5)fb 100 GeV10^(-4)fb 250 GeV10^(-3)fb CTEQ6L

26. Energy distribution for each angle • From thelL beam, τR is emitted to the backward direction due to (1 ー cosθCM)nature in the CM frame. • In Lab-frame, tau is emitted forward direction but with large angle with a PT. 2 E＝100 GeV E＝500 GeV

27. Signal • Number of taus (case of electron beam) E＝250 GeV, L =10^34 /cm^2/s, ⇒10^22 electrons (positrons) in a SUSY model with |κ3i |^2=0.3×10^(-6): σ=10^(-3) fb 10^5 of τleptons are producedfor the target of ρ=10 g/cm^2 Naively, non-obervation of the e N → τ X process may improve the current upper limit on the e-τ-Φcoupling by around 4-5 orders of magnitude • We may consider its hadronic products as the signal τ→(π、ρ, a1, …)＋　missings # of hadrons ≒ 0.3×(# of tau) Hard hadrons emitted into the same direction as the parent τ’s Bullock, Hagiwara, Martin

28. Backgrounds • Hadrons from the target (N) should be softer, and more unimportant for higher energies of the initial e or μ beam. • Hard leptons from l N→ lX would be a fake signal via mis-ID of l as π. (l= e or μ) • Rate of mis-ID • Emitted to forwad direction without large PT due to the Rutherford scattering 1/sin^4(θcM/2) ⇒PT cuts • Other factors to reduce the fake • Realistic Monte Carlo simulation is necessary.

29. Summary 1 • We discussed LFV in the Higgs coupling • Direct search via the decay process h →τμata LC • DIS process e N (μN)→τX by using the high energy electron beam of a LC (a μC) with a fixed-target. • h →τμat a LC: • The backgrounds can be reduced by kinematic cuts • The signal may be detectable, if the LFV Higgs coupling κ32is large within the limit from the τ→μη result. • Such a significant κ32 can be realized only when MSUSY is as large as TeV with μ/MSUSY being O(1-10). • In the general 2HDM, larger significance is obtained under the current bound on κ32.

30. Summary 2 • DIS processes e N (μN)→τX : • For E > 60 GeV, the cross section is enhanced due to the sub-process with sea b-quarks • At a LC with Ecm=500GeV ⇒ σ＝10^(-3) fb L=10^34/cm^2/s ⇒ 10^22 electrons available 10^5 of taus are produced for ρ=10 g/cm^2 • Non-observation of the signal would improve the current limit on the τ-e-Φ coupling by 10^(4-5). • Realistic background simulation is necessary.