Universality & LFV in Upsilon Decays

at CLEOJ.E. Duboscq

Cornell jed@mail.lepp.cornell.edu

For the CLEO Collaboration

Tau06 Sept 19 2006 1

Topics

CLEOIII Detector and Data

A Search for the Lepton Flavor Violating decay Υ→μτ

Test of Lepton Universality in Υ→ττ and Υ→μμ

JED Tau06 2

The CLEOIII DetectorSolenoid Coil

Barrel Calorimeter

RICH Drift Chamber

Silicon /beampipe

EndcapCalorimeter

IronPolepiece

Barrel MuonChambers

MagnetIron

Rare EarthQuadrupole

SC Quadrupoles

SC QuadrupolePylon

2230104-001

CLEO III

1400701-002

Muon Chambers

Excellent Calorimeter Coverage and Resolution

Excellent Tracking Coverage and Resolution

Ring Imaging Cerenkov & dE/dx for PID

+ A Great Accelerator Team CESR

3JED Tau06

CLEO Upsilon Data

CLEOIII HAD the largest world sample of clean Υ events below b threshold (cf Belle 3s)

Also has off-resonance data + scan data + data at Υ(5S)

0

5,000,000

10,000,000

15,000,000

20,000,000

Υ(1s) Υ(2s) Υ(3s)

events CLEO 2 CLEO 3

4JED Tau06

Search for LFV Violation in Υ→μτ

5JED Tau06

Lepton Flavor ViolationWe don’t annihilate when we shake hands.

Baryon Asymmetry generated via Sakharov conditions: B & C & CP violation, + Universe out of thermal Equilibrium for a while

B, L are accidental symmetries of SM

B-L is however conserved

Maybe B is violated because L is violated and maybe this is related to LFV ?

6JED Tau06

LFV for τ and Υ Decaysτ Decay LFV talks about Υ LFV

2 Theoretical Estimates

If we assume that a ! decays to µ! , then by unitarity its exchange contributes also tothe decays ! ! 3µ and ! ! eµµ as is shown if Fig. 1. With some assumptions, thisallows to relate the experimental upper limits on partial widths of the latter decays to theconstraint on the branching fraction for ! ! µ! . Using B(! ! 3µ) " 10!6, Nussinov,Peccei and Zhang estimated[1] B(! ! µ!) " 10!2, which is quite encouraging from theexperimental point of view. Furthermore, these authors remarked that LFV decay of the! could be larger than quoted limit because the contribution of virtual ! state to LFVleptonic decays of the ! could be kinematically and dynamically suppressed. Accoridingto Nussinov and collaborators the decay ! ! µ! might be kinematically enhanced (ascompared to its contribution to the decays of !) because of anomalous magnetic momentcoupling. The dynamic suppression of !" contribution to neutrinoless ! decays could beexplained by form-factor e"ects. Nowadays, model-independent estimate of the reference[1]should be updated with the most recent data[3] on neutrinoless decays of ! . Assuming nokinematic and dynamic suppression, taking into account all data on LFV decays of the !lepton, recent experimental results B(! ! 0") " 10!7 would give us a 90% CL upper limitestimate B(!! µ!) " 10!3.

t

!

µ

"

#

!

µ

"

#

Figure 1: Virtual ! meson contribution to the decays ! ! 3µ, ! ! e+e!µ and ! ! #µ.

The same unitarity-based model-independent approach has been employed by Huo, Yueand Feng who used the existing data on ! decays to arrive[2] at B(! ! µ!) " 8.0 # 10!5

and B(! ! µ!) " 2.9 # 10!4 for two form factor models describing the vertex !" ! µ! .Huo et al. also analyzed the impact of the existing data on the LFV decays of ! mesonsin the framework of GUT leptoquarks: B(! ! µ!) " 1.3 # 10!7, SUSY with broken Rparity:4 B(! ! µ!) " 2.2 # 10!8, and top-color assisted Technicolor (TC2) with the newTeV-mass intermediate vector boson Z #: B(!! µ!) " 1.3# 10!7. These estimates rely ona number of assumptions and, more importantly, the new physics contribution to the decays

4In SUSY framework R parity is defined as R = ($1)3B+Ltotal+2S , where B, Ltotal and S are the baryonnumber, total lepton number and intrinsic spin of the particle, respectively. The ordinary particles areR-parity even while their superpartners are R-parity odd. Tree-level non-conservation of lepton number inSUSY requires broken R parity.

6

B(τ → 3 μ) < 10-6 ⇒ B(Υ → μ τ) < 10-2

S.Nussinov, R.D.Peccei, X.M. Zhang PRD63(2000), 016003

of ! mesons might be kinematically enhanced in comparison to the decays of ! lepton ashas been outlined previously.

We would also like to mention that in the context of SUSY there is an important con-nection between LFV and Dark Matter: currently, the best candidate for the latter is theLightest Suppersymmetric Particle (LSP) which is either stable or extremely long-lived. Sta-ble LSP would require R parity conservation which means no tree-level lepton and baryonnumber violating terms in SUSY Lagrangian. Therefore, a possible observation of LFVabove non-perturbative instanton-induced rate would allow to estimate relevant couplingconstants and upper limits on the lifetime of LSP in SUSY. This would give us some insighton Dark Matter. In this fashion, Kim, Ko and Lee[4] used the data from a variety of previousexperiments to place the constraints on LFV-violating couplings of SUSY Lagrangian withbroken R parity. We show some of Feynman diagrams for LFV decays of ! mesons in Fig. 2for leptoquarks, additional gauge bosons (e.g. in TC2) and SUSY hypotheses.

-�

-�

!"�#�

-�

-�

!

$� µ�

% / ! �

Figure 2: Some Feynman diagrams for LFV ! decays that could arise due to a) leptoquarks,b) additional gauge bosons in a large variety of models and c) SUSY with broken R parity(with neutralino and sleptons in the loop) or new heavy neutrinos.

Finally, Silagadze suggested[5] to use LFV decays of ! and J/" as a probe of the scaleof quantum gravity or some other new interaction in a very general way. In his (leadingorder) estimate, performed in the spirit of e"ective theory, he introduces generic four-fermionoperators and, treating ! states non-relativistically, he arrives at

#(!! µ±!!)

#(!! µ+µ")=

1

2e2b

!#N

#

"2 !M!

$

"4

(1)

where eb = 1/3 is the electric charge of b quark, # is fine structure constant, M! is the massof ! meson, $ (possibly of a TeV order) is the true scale of quantum gravity (in models based

7

of ! mesons might be kinematically enhanced in comparison to the decays of ! lepton ashas been outlined previously.

We would also like to mention that in the context of SUSY there is an important con-nection between LFV and Dark Matter: currently, the best candidate for the latter is theLightest Suppersymmetric Particle (LSP) which is either stable or extremely long-lived. Sta-ble LSP would require R parity conservation which means no tree-level lepton and baryonnumber violating terms in SUSY Lagrangian. Therefore, a possible observation of LFVabove non-perturbative instanton-induced rate would allow to estimate relevant couplingconstants and upper limits on the lifetime of LSP in SUSY. This would give us some insighton Dark Matter. In this fashion, Kim, Ko and Lee[4] used the data from a variety of previousexperiments to place the constraints on LFV-violating couplings of SUSY Lagrangian withbroken R parity. We show some of Feynman diagrams for LFV decays of ! mesons in Fig. 2for leptoquarks, additional gauge bosons (e.g. in TC2) and SUSY hypotheses.

-�

-�

!"�#�

-�

-�

!

$� µ�

% / ! �

Figure 2: Some Feynman diagrams for LFV ! decays that could arise due to a) leptoquarks,b) additional gauge bosons in a large variety of models and c) SUSY with broken R parity(with neutralino and sleptons in the loop) or new heavy neutrinos.

Finally, Silagadze suggested[5] to use LFV decays of ! and J/" as a probe of the scaleof quantum gravity or some other new interaction in a very general way. In his (leadingorder) estimate, performed in the spirit of e"ective theory, he introduces generic four-fermionoperators and, treating ! states non-relativistically, he arrives at

#(!! µ±!!)

#(!! µ+µ")=

1

2e2b

!#N

#

"2 !M!

$

"4

(1)

where eb = 1/3 is the electric charge of b quark, # is fine structure constant, M! is the massof ! meson, $ (possibly of a TeV order) is the true scale of quantum gravity (in models based

7

of ! mesons might be kinematically enhanced in comparison to the decays of ! lepton ashas been outlined previously.

We would also like to mention that in the context of SUSY there is an important con-nection between LFV and Dark Matter: currently, the best candidate for the latter is theLightest Suppersymmetric Particle (LSP) which is either stable or extremely long-lived. Sta-ble LSP would require R parity conservation which means no tree-level lepton and baryonnumber violating terms in SUSY Lagrangian. Therefore, a possible observation of LFVabove non-perturbative instanton-induced rate would allow to estimate relevant couplingconstants and upper limits on the lifetime of LSP in SUSY. This would give us some insighton Dark Matter. In this fashion, Kim, Ko and Lee[4] used the data from a variety of previousexperiments to place the constraints on LFV-violating couplings of SUSY Lagrangian withbroken R parity. We show some of Feynman diagrams for LFV decays of ! mesons in Fig. 2for leptoquarks, additional gauge bosons (e.g. in TC2) and SUSY hypotheses.

-�

-�

!"�#�

-�

-�

!

$� µ�

% / ! �

Figure 2: Some Feynman diagrams for LFV ! decays that could arise due to a) leptoquarks,b) additional gauge bosons in a large variety of models and c) SUSY with broken R parity(with neutralino and sleptons in the loop) or new heavy neutrinos.

Finally, Silagadze suggested[5] to use LFV decays of ! and J/" as a probe of the scaleof quantum gravity or some other new interaction in a very general way. In his (leadingorder) estimate, performed in the spirit of e"ective theory, he introduces generic four-fermionoperators and, treating ! states non-relativistically, he arrives at

#(!! µ±!!)

#(!! µ+µ")=

1

2e2b

!#N

#

"2 !M!

$

"4

(1)

where eb = 1/3 is the electric charge of b quark, # is fine structure constant, M! is the massof ! meson, $ (possibly of a TeV order) is the true scale of quantum gravity (in models based

7

Z’LeptoQuark

W.J. Huo, C.X. Yue, T.F. Heng, PRD67(2003),

114001

SUSY Loops

B(Υ→μτ) < 1.3x10-8

B(Υ→μτ) < 2.2x10-9

7JED Tau06

LFV for τ, Υ Decays

B(!! µ!)B(!! µµ)

" ("N/")2(M!/")4

Z. Silagadze, Phys. Scripta 64(2001),128

Generic 4 fermion coupling αN at scale Λ added to SM to get LFV

τb

μb

αN

Λ

8JED Tau06

LFV: The analysisSearch for Υ→μτ, τ→eνν

2 tracks: μ (muon ID), e (E/p, dE/dx)

Extended Max Likelihood

Use Product PDF:

Sum over: Signal LFV decay ⊕ ττ ⊕ μμ (γ) w/

hard γ ⊕ μμ with μ decay to electron

Υ(4S), off res used as calibration & control samples

P(pµ)! P(pe)! P(dE/dx(e))! P(E/p(e))

L = e!Nevnt!evnt ("NiPi(X|S))

9

μ near Ebeam

JED Tau06

Background & Signal Regions

0.80 0.85 0.90 0.95 1.00 1.05

x = pµ/Ebeam

0.10

0.30

0.50

0.70

y =

pe/E

beam

0

100

200

300

400

500

Figure 6: The 2D distribution of y versus x (and the x projection) for our calibrationamd control samples, i.e. !(4S) and “continuum” data after all selection criteria includingapplying the electron-is-not-even-a-poorly-quality-muon requirement (technically speaking,MUQUAL = 0 or MUDEPTH = 0). There are 6764 events in this plot. Notice that: 1)some muon pairs survive this selection and, also, 2) there are obviously some events withbeam-energy muons that have a fairly wide distribution of y which indicates that theseevents are likely to contain one real beam-energy muon and one real electron. We think thatthese events are due to muon’s decay to electron in flight. The events we are now concernedabout are indicated by two ovals.

45

μμ + hard γ

Signal Region

μμ & μ decay in flight

ττ

Ups(4S) data

γ hits CC and fakes E/p

about 100 in Υ(4S) data

Extract most PDFs from Υ

(4S) data10JED Tau06

dE/dx(ELEC)

Fits to Υ(1S) Data

ττμμ hard γ

0.87 0.92 0.97 1.0210

-1

1

101

102

103

x = |pµ |/E

beam

Events

/ 0

.005

0.85 0.90 0.95 1.00 1.05 1.1010

-1

1

101

102

103

ECC

/ |pe |

Events

/ 0

.12

-3 -2 -1 0 1 2 310

-1

1

101

102

103

dE/dr (e hypothesis, in ! units)

Events

/ 0

.01

0.10 0.30 0.50 0.7010

-1

1

101

102

103

y = |pe |/E

beam

Events

/ 0

.015

Figure 38: Projections of the results of four-component (including signal) 4D unbinned MLfit and the comparison with binned !(1S) data. The legend for the results of the fit: solidblack for everything, dashed green for ! pairs, dotted red: muon pairs with hard radiation,dot-dashed blue (solid blue for right top corner plot): muon pairs with decay to electron.Dashed black line (top left corner plot only) shows 100 hypothetical signal events added tothe results of the fit.

77

MC Signal

μμ, decay in flight

p(μ)/Ebeam E/p(ELEC)

p(e)/Ebeam

11JED Tau06

Preliminary LFV Results

Largest Syst: PDF shapes & correlations

First Limits on Υ→τμThese BRs set a lower limit of ≈ 1 TeV on generic LFV scale (assuming strong coupling)

Resonance Υ(1S) Υ(2S) Υ(3S)Efficiency 8.9% 8.9% 8.9%

Nevents < 10.0 < 10.7 < 8.5

B(Υ→μτ)(10-6) < 6.2 < 25 < 22

B(Υ→μτ)/B(Υ→μμ) < 0.023% < 0.17% < 0.13%All limits are 90% CL UL

12JED Tau06

Universality in Υ Decays with Υ→μμ,ττ

13JED Tau06

CLEO has measured B(Υ(nS)→μμ), Γ(ee→Υ(nS)) n=1,2,3 PRL94:012001(2005) hep-ex/0409027

Υ→ττ rounds out this series

B(Υ(1S)→ττ) known to ≈ 10 % in PDG

Υ(2S)→ττ “observed” Br=1.8+/-1.7%

Υ(3S) not yet seen

Υ→ττ Motivation

14JED Tau06

Naive Universality:

B(Υ→ee)=B(Υ→μμ)=B(Υ→ττ)

If this ain’t the case, there’s some explaining to do

Sanchis-Lozano: Higgs searches have a blind spot near the Υ

The decay chain Υ→γηb, (ηb→A0), A0→ττ could

alter N(Υ→ττ)/N(Υ→μμ), if γ soft&undetected

Υ→ττ Motivationτ

b τb

γ

15

cf next talk, hep-ph/0307313

JED Tau06

Isolate μμ and ττ signals at & below Υ(nS), n=1..4

Apply On -S*Off to Υ(4S) data for μμ and ττ - expect B(4S →ll) ≈ 0

Find Off Resonance σ(e+e-→ττ); compare with σ(e+e-→μμ)

Apply On-S*Off to 1S, 2S, 3S data for μμ and ττ

Extract R = N(Υ→ττ)/N(Υ→μμ)

Get Branching Ratio B(Υ→ττ) using B(Υ→μμ)

Goal: B(Υ→ττ)/B(Υ→μμ)

S=LOn/LOff (EOff/EOn)2

16

Technique Check

Technique Check

JED Tau06

Use μμ cuts similar to previous Υ→μμ study

Use 1 Prong decays : B(τ→1prong) ≈ 75%

2 Track events, passing generic ττ missing momentum, energy cuts (neutrinos!)

Classify tracks as e, μ, X

Include neutral energy, shower cuts

How to find Υ→ ττ, μμ

Cosmics are rejected

17JED Tau06

On & S*Off at the Υ(4S)

TextTextCM

)/E!!E(0 0.2 0.4 0.6 0.8 1 1.20

100

200

300

400

500

600

l vs l modes CM

)/E!!(4S) E("

(4S) " at CM

)/E!!E(

Off Res Data (scaled)

On Res Data

l vs l modes CM

)/E!!(4S) E("

CM)/E!!E(0 0.2 0.4 0.6 0.8 1 1.20

200

400

600

800

1000

1200

1400

1600

l vs X modes CM

)/E!!(4S) E("

(4S) " at CM

)/E!!E(

Off Res Data (scaled)

On Res Data

l vs X modes CM

)/E!!(4S) E("

CM)/E!!E(0 0.2 0.4 0.6 0.8 1 1.20

500

1000

1500

2000

2500

3000

other modes CM

)/E!!(4S) E("

(4S) " at CM

)/E!!E(

Off Res Data (scaled)

On Res Data

other modes CM

)/E!!(4S) E("

CM)/E!!E(0 0.2 0.4 0.6 0.8 1 1.20

1000

2000

3000

4000

5000

All modes CM

)/E!!(4S) E("

(4S) " at CM

)/E!!E(

Off Res Data (scaled)

On Res Data

All modes CM

)/E!!(4S) E("

2

Leptons1

Lepton

no

Lepton All

On-S*Off works at Υ(4S)No evidence for non-1/E2 background

CM)/EµµM(0.85 0.9 0.95 1 1.05 1.1 1.15 1.2 1.250

2000

4000

6000

8000

10000

12000

MuPair Mass On, Off Res

(4S) ! at CM

)/EµµM(

Off Resonance Data (scaled)

On Resonance Data

MuPair Mass On, Off Res

ee→μμ

18JED Tau06

Breakdown by τ decay channels

All decay channels agree

We can reconstruct ee→ττ, μμ

Off Res σ(ττ)/σ(μμ)/Exp

!theory(e+e! ! "")!theory(e+e! ! µµ)

= 0.82

19

Average over lepton modes

overall average

(X,Y) = P(X)> P(Y)

JED Tau06

On And S*Off Res For ττ, μμ

CM)/E!!E(0 0.2 0.4 0.6 0.8 1 1.20

2000

4000

6000

8000

10000

12000

14000

16000

18000

20000

CM)/E!!(1S) E("

(1S) " at CM

)/E!!E(

Off Res Data (scaled)

On Res Data

CM)/E!!(1S) E("

CM)/E!!E(0 0.2 0.4 0.6 0.8 1 1.20

2000

4000

6000

8000

10000

12000

14000

16000

CM)/E!!(2S) E("

(2S) " at CM

)/E!!E(

Off Res Data (scaled)

On Res Data

CM)/E!!(2S) E("

CM)/E!!E(0 0.2 0.4 0.6 0.8 1 1.20

2000

4000

6000

8000

10000

12000

14000

CM)/E!!(3S) E("

(3S) " at CM

)/E!!E(

Off Res Data (scaled)

On Res Data

CM)/E!!(3S) E("

ττ ττττ

CM)/EµµM(0.85 0.9 0.95 1 1.05 1.1 1.150

5000

10000

15000

20000

25000

30000

35000

40000

CM)/Eµµ(3S) M(!

(3S) ! at CM

)/EµµM(

Off Res Data (scaled)

On Res Data

CM)/Eµµ(3S) M(!

CM)/EµµM(0.85 0.9 0.95 1 1.05 1.1 1.150

10000

20000

30000

40000

50000

60000

CM)/Eµµ(1S) M(!

(1S) ! at CM

)/EµµM(

Off Res Data (scaled)

On Res Data

CM)/Eµµ(1S) M(!

CM)/EµµM(0.85 0.9 0.95 1 1.05 1.1 1.150

5000

10000

15000

20000

25000

30000

35000

40000

45000

CM)/Eµµ(2S) M(!

(2S) ! at CM

)/EµµM(

Off Res Data (scaled)

On Res Data

CM)/Eµµ(2S) M(!

μμ μμμμ

1S

1S

2S

2S

3S

3S

Total ττ Reconstructed Energy /Ecm :

Mass of μμ /Ecm :

Data remaining after On-S*Off subtraction should be all due to Υ decays

First Observation

20JED Tau06

Data after On-S*Off is a sum of Signal Υ→ll, Cascade to lower Υ→ll, Other Υ decays.

Measure Br(Υ(1S)→ll) to scale MC cascade bgd for Υ(2S) decays, iterate for Υ(3s).

Use Koralb for Υ→ττ with ISR turned off

Υ has the quantum numbers of the photon

Koralb has helicity correlations

Getting N(Υ→ll)

21JED Tau06

A Sprinkling of plots

CM)/E!!E(0 0.2 0.4 0.6 0.8 1 1.2

0

200

400

600

800

1000

(1S) lvsl modes" CM

)/E!!E(

! ! #(1S) "Data

)µµKoralb using B(BackgroundMC Sum

)µµModel 14 using B(

(1S) lvsl modes" CM

)/E!!E(

CM)/E!!E(0 0.2 0.4 0.6 0.8 1 1.2

0

200

400

600

800

1000

1200

(2S) lvsx modes" CM

)/E!!E(

! ! #(2S) "Data

)µµKoralb using B() Background!!non(

2s to 1s CascadeMC Sum

(2S) lvsx modes" CM

)/E!!E(

CM)/E!!E(0 0.2 0.4 0.6 0.8 1 1.2

0

200

400

600

800

1000

1200

1400

(3S) other modes" CM

)/E!!E(

! ! #(3S) "Data

)µµKoralb using B(3s to 2s Cascade3s to 1s Cascade

) Background!!non(MC Sum

(3S) other modes" CM

)/E!!E(Total Energy

Distribution for 3 different modes at 3 resonances for ττ, assuming

universality

22

Υ(1S) Υ(2S)

Υ(3S)

JED Tau06

Need More Convincing ?

Ptag/Ebeam0 0.2 0.4 0.6 0.8 1 1.20

1000

2000

3000

4000

5000

6000

Ptag Ups1s Tau all

Ups1s to tau tau DataKoralb using BmumuBackgroundMC SumModel 14

Ptag Ups1s Tau all

Psig/Ebeam0 0.2 0.4 0.6 0.8 1 1.2

0

1000

2000

3000

4000

5000

6000

7000

8000

Psig Ups1s Tau all

Ups1s to tau tau DataKoralb using BmumuBackgroundMC SumModel 14

Psig Ups1s Tau all

Ecc/Ecm0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4

1

10

210

310

410

Ecc Ups1s Tau all

Ups1s to tau tau DataKoralb using BmumuBackgroundMC SumModel 14

Ecc Ups1s Tau all

Nshw0 2 4 6 8 10 12 14 16 180

2000

4000

6000

8000

10000

Nshw Ups1s Tau all

Ups1s to tau tau DataKoralb using BmumuBackgroundMC SumModel 14

Nshw Ups1s Tau all

Highest Momentum Particle

Lowest Momentum

Particle

neutral Energy

Isolated Showers

Good agreement w/ MC across Resonances, Final states, and kinematic quantities

23JED Tau06

How Many Events ?

1S 2S 3S

On-S*Off 61697±1536 25085±1399 16290±1522

background 1556±83 3334±593 1536±474

ε(ττ) 11.2±0.1% 11.3±0.1% 11.1±0.1%

N(ττ)/ε (103) 537±14 193±12 132±13

N(μμ)/ε (103) 527±15 185±11 126±11

Sum of all τ decay modes

Stat, MC Stat errors inc

24JED Tau06

R=B(Υ→ττ)/B(Υ→μμ)

R(1S) = 1.02 ± 0.02 ± 0.05

R(2S) = 1.04 ± 0.04 ± 0.05

R(3S) = 1.07 ± 0.08 ± 0.05

Final

Result

Largest Syst from τ selection criteria (2.9%) and Trigger (1.6%)

Almost all Stat Error from On/Off Subtraction

submitted to PRL

25JED Tau06

hep-ex/0607019

R by Decay Mode

26

Average over lepton modes

overall average

(X,Y) = P(X)> P(Y)

JED Tau06

R=B(Υ→ττ)/B(Υ→μμ)

Use CLEO’s published B(Υ→μμ) - cf

Avoid syst error double counting

Extracting B(Υ→ττ)

B(Υ(1S)→ττ) = 2.54 ± 0.04 ± 0.12 %

B(Υ(2S)→ττ) = 2.11 ± 0.07 ± 0.13 %

B(Υ(3S)→ττ) = 2.55 ± 0.19 ± 0.15 %

τ statsyst+μ stat

27

PRL94,012001(2005)

JED Tau06

Look at γ spectrum No obvious spike in the 1S Spectrum

Expected region is falling fast - hard systNo obvious spike in 2S, 3S spectra

What about ηb/higgs?

90% CL UL

EGam/Ebeam0 0.02 0.04 0.06 0.08 0.1

0

1000

2000

3000

4000

5000

6000

7000

EGam Ups1s Tau all

Ups1s to tau tau

Data

Koralb using Bmumu

Background

MC Sum

Higgs BR=0.27%

EGam Ups1s Tau all

FIG. 10: The photon spectrum for the !(1S) sample, including all ! decay modes, along with asuperimposed photon spectrum from the Higgs process outlined in the text. The branching fractionfor this process is set at the upper limit for the decay in the inclusive search. The e"ciency forseeing the photon is scaled by 0.90 to account for the correct photon angular distribution relativeto that generated in the Monte Carlo.

31

R is consistent with 1 - no big signal

Attribute deviation R(1S)-1 to “higgs”

B(Υ→γηb, ηb→A0, A0→ττ) < 0.27% 90% CL UL

28

MC for Signal

Scaled Photon Energy

Υ(1S)

JED Tau06

Searched for LFV gives BR(Υ→μτ) ≈< 10-5

Measured B(Υ→ττ)/B(Υ→μμ) consistent with 1

measured B(Υ→ττ)Consistent with PDG at 1S (but lower)

Best Value for 2S

First value for 3S

Set limit on CP odd Higgs in Υ(1S) region

ConclusionsCLEO has:

29

90%CL UL

JED Tau06