1. Tau Lepton Flavor Violation Search at Belle, tgmg/eg
K.Hayasaka (Nagoya Univ.)
7 March 2006
Mass Origin and SUSY Physics

2. Mass Origin and SUSY Physics
Introduction
LFV decays are forbidden in SM.
Observation of LFV decay suggests
existence of New Physics!
t is a heaviest lepton and has decay mode
including some hadrons.
t is expected to be most sensitive lepton
for New Physics!
7 March 2006
Mass Origin and SUSY Physics

3. LFV modes searched at Belle
tglepton+g, 3 leptons
(lepton=e,m)
tglepton+2 charged meson
(charged meson=p/K)
tglepton+neutral meson
(p0,h,h’,f,Ks,r0,K*)
tgBaryon+meson
(tgLp/Lp)
7 March 2006
Mass Origin and SUSY Physics

4. Mass Origin and SUSY Physics
KEKB accelerator
Asymmetric e+e- collider
(e+:3.5GeV,e-:8.0GeV)
Circumference~3Km
s=Upsilon(4s)
Integrated luminosity
>560fb-1
B-factory
is also
　t-factory!
s(tt)~0.9nb
s(bb)~1.1nb
7 March 2006
Mass Origin and SUSY Physics

5. Mass Origin and SUSY Physics
Belle detector
F/B asymmetric
m-ID:eff. 87.5%
e-ID:eff. 92.4%
7 March 2006
Mass Origin and SUSY Physics

6. Signal and BG events for tgmg
side
tag
side mm
signal qq tt
q=u,d,s
s(mm)~1.0nb
s(qq)~2.1nb
s(tt)~0.9nb
7 March 2006
Mass Origin and SUSY Physics

7. Signal and BG events for tgegqq tt
signal
bhabha
q=u,d,s
s(tt)~0.9nb
s(bhabha)~124nb
7 March 2006
Mass Origin and SUSY Physics

8. Selection Criteria for tgeg
2 charged tracks + more than 1 g
signal side : 1 charged+1 photon
tag side : 1 charged
e-ID>0.9 & e-ID<0.1
restrict kinematical variables
momentum e,tag,g,missing
polar angle e,tag,g,missing
opening angle
e-tag, e-g, tag-missing
missing mass vs missing mom.
analysis for 87fb-1 data sample e g
1-1 prong t e e t
generic decay
tag
missing
7 March 2006
Mass Origin and SUSY Physics

9. Requirement for cosqegCM
opening angle
between e and g
calculatable
data
MC total
signal MC
7 March 2006
Mass Origin and SUSY Physics

10. Requirements for missing trks.
Pmiss vs m2miss tt events have
large missing mass n
tt MC g
signal MC e n t e e t n
charged particle
98% of the t+t- bkg is removed.
7 March 2006
Mass Origin and SUSY Physics

11. Distribution for signal events
generate 106 signal MC events
2 variables defined
Meg= (Pe+Pg)2 g
ISR
EgISR+EeISR<EgnoISR+EenoISR
gDEISR<DEnoISR,MegISR~MegnoISR
Energy leakage
Egob<EgtruegDEob<DEtrue,Megob<Megtrue
7 March 2006
Mass Origin and SUSY Physics

12. Mass Origin and SUSY Physics
Meg and DE resolutions
All selections applied for signal MC events
Asymmetric Gaussian
6.5%
Meg DE
# of events.
# of events.
+1s -
+1s -
25.7/14.3MeV/c2
84.8/36.0MeV eg
+5s region -
7 March 2006
Mass Origin and SUSY Physics

13. Mass Origin and SUSY Physics
Blind Analysis
Signal dominant region is masked.
42ev.
43.5ev.
60.8ev.
Masked!
side-band region
+2.7(4.3) bhabha events.
7 March 2006
Mass Origin and SUSY Physics

14. Mass Origin and SUSY Physics
Signal extraction
Unbinned Extended Maximum Likelihood
Likelihood function
s,b: # of events for signal and BG
gfree parameter: search s and b which maximize L
N: # of observed events, i: each found event
PDFsig:fsignalMC(smoothing)
PDFbg:express analytic func.
7 March 2006
Mass Origin and SUSY Physics

15. Mass Origin and SUSY Physics
BG PDF
Difficult to decide from data
# of events is small.
Blind region exists.
Shape of BG distribution is evaluated
from distribution of tt MC events.
Its height is decided with data distribution of the side-band region.
Masked!
7 March 2006
Mass Origin and SUSY Physics

16. Mass Origin and SUSY Physics
BG distribution (MC)
curve
(Landau+Gauss)
Estimate BG distribution
from tt MC
reproduce BG
distribution
by function
7 March 2006
Mass Origin and SUSY Physics

17. BG distribution (data)
Estimate BG from MC and data
data(side band) tt MC ee
= 61 4 ~
ttMC
Masked!
(Landau+Gauss)
curve
profile plot
7 March 2006
Mass Origin and SUSY Physics

18. Final Candidates (tgeg)
60 events found. (# of estimated BG:65)
20 events survived in +5s region -
signal MC -
signal events
dominate
+5s
surviving data
7 March 2006
Mass Origin and SUSY Physics

19. Evaluation of signal events
fit by unbinned expanded maximum likelihood
with signal and BG shape g s0=0,b0=20
Estimation for U.L. of 90%CL
uby Toy MC: generate events
result
s90=3.8 events
Branching fraction
Signal yield : s0
e-ID ineff ev.
BG function ev.
Efficiency & Luminosity : 2eNtt
Track rec. eff. 2.0%
Photon rec. eff. 3.0%
Selection criteria 2.5%
Luminosity 1.4%
Trigger eff. 2.0%
MC statistics 0.3%
Total 5.0%
Br=s0/2eNtt<3.9x10-7
e:detection efficiency
Ntt:total event number
7 March 2006
Mass Origin and SUSY Physics

20. Mass Origin and SUSY Physics
Search for tgmg
Almost same selection criteria as tgeg
for tag side track, require not to be m
Main BG: ttg & mmg(ffrom m-ID ineff.)
Data
(x ineff.)
Masked!
7 March 2006
Mass Origin and SUSY Physics

21. Final Candidates ( tgmg )
54 events survived in +5s region -
1.71<Minv<1.82GeV/c2
signal MC
+5s -
11%
fitting result(UEML)
s=0
Evaluation for U.L.
including
systemaric
uncertaunties
7 March 2006
Mass Origin and SUSY Physics

22. Mass Origin and SUSY Physics
Future prospect
Super B-factory : >10 times more data
B.R. sensitivity:
~1/L for negligible BG case
~1/ L for BG dominated modes
We need 2~3ab-1 data sample
to reach Br<a few x 10-8.
7 March 2006
Mass Origin and SUSY Physics

23. Mass Origin and SUSY Physics
Conclusion & Summary
Obtain BR UL’s with blind analyses.
BG distributions are modeled well.
Results are 10 times more sensitive than CLEO’s.
tgeg(86.7fb-1)
Ntt=7.90x107
20 observed events
e=6.5%
tgmg(86.3fb-1)
Ntt=7.87x107
54 observed events
e=11%
Phys. Lett. B613(2005) 20
Phys. Rev. Lett. 92(2004)
7 March 2006
Mass Origin and SUSY Physics