1

Koji Hara (KEK)For the Belle Collaboration

Time Dependent CP Violation in B0→+- Decays

[hep-ex/0502035]

2

Time dependent CP-violating parameters in B0→+− decays

S : B0B0 mixing induced CPVA : direct CPV

)22sin(A1S 22

In B0 → +- decays, S is sensitive to ():

u

W

u

db d

W

g u

u

tree penguin

can be determined with isospin relations

0ADirect CP violation may occur.

)cos()sin())())(())())((

00

00

tmAtmSftBftBftBftBA dd

CPCP

CPCPCP

3

Previous experimental situation

07.015.058.0 A07.021.000.1 S

CPV with 5.2, 3.2evidence for DCPV

Belle 152 M BBwith 372±32 B0→+− events

04.015.009.0 A03.017.030.0 S

BABAR 227M BB

hep-ex/0501071

3.2 difference

PRL 93, 021601 (2004)

with 467±33 B0→+− events

4

New Belle measurement with 275M BB pairs

additional data 152 → 275 M BB pairs the inner tracker and interaction region was

replaced in the summer 2003 3-layer SVD (152 M BB) to 4-layer SVD

with small-cell drift chamber (123 M BB). Analysis procedure is basically same as before

5

Event Selection B0→+− selection

CMSbeam

CMSB EEE

22 )()( CMSB

CMSbeambc pEM

2/287.5271.5 cGeVM bc GeVE 064.0||

corresponding ±3

Kinematical Selection

Flavor Taggingq=+1 tagged as a B0, q=−1 tagged as a B0

Pion Identification using aerogel and dE/dx

r=0 no flavor discrimination, r=1 unambiguous flavor assignment

r: flavor tag quality0<r≤1

q: flavor charge

%90)( %11)( Kp

6

Event Selection (continuum suppression)

Event shape B flight direction),,,(, csduqqqee

LR

(Likelihood Ratio)

BBqq

e+

e-

Signal B

Other B

e+

e-

qq

Jet-like Spherical

7

LR (L

ikel

ihoo

d R

atio

)

B0→+− signals

r (flavor tag quality)

(LR>0.86) (LR<0.86)

2,820 candidates containing (666 signal events

8

t (ps)

Unbinned CP Fit Result

PointsE-Mbc 2D fits to individual time

intervals

B0 tagged

+ − A

CP

+

− Yie

ldB0 tagged

(syst) 06.0(stat) 16.067.0(syst) 06.0(stat) 12.056.0

SA

Background subtracted fit projection for all events

9

Consistency checks with Time-integrated fits

14.052.0 A consistent with time-dependent fit

B0 tagged B0 tagged: 107±13

K: 42±9cont: 38±1

: 69±11K: 43±9

cont: 38±1

Direct CP Violation is evident!E (GeV)

LR>0.86 & 0.5<r≤1.0

10

t distribution and raw asymmetry

Raw

asy

mm

etry

Even

ts /(

1.25

ps)

q=+1

q=−1

0.0<r≤0.5

0.5<r≤1.0

t (ps)0-5 +5

LR>0.86

-5 0 +5

11

(A,S)=(0,0)1-C.L.=5.62x10-8, 5.4

(A,S)=(0,−0.62)1-C.L.=5.13x10-5, 4.0

Significance calculation with Feldman-Cousins method

Large CP Violation,

Large Direct CP violation, confirmation of the previous Belle

results

C.L. = Confidence Levelboth statistical and systematic errors are taken into account.

S

A

12

New experimental situation

Belle 275M

BABAR 227M

A

~2.3 difference between Belle and BABAR

S

13

Interpretation: 2 constraint using isospinM. Gronau and D. London, PRL 65, 3381 (1990)

)22sin(1 22 AS

We use the HFAG summer 2004 values for the branching ratios of B0→+−, 00, B+→+0 and direct

CP asymmetry of B0→00.We use the statistical treatment of

J. Charles et al., hep-ph/0406184

The cleanest method to extract 2

14

Interpretation : 2 constraint using isospin

190 2 and 18071 2 95.4% confidence interval

2 (degrees)

1−C

.L.

using HFAG summer 2004

J. Charles et al., hep-ph/0406184

15

&

Summary of new Belle B0 →+− CP results

The fit yields

Large direct CP violation with 4.0 significance is observed

The results confirm the previous Belle results. Isospin analysis gives

at 95.4% C.L.

)syst( 06.0)stat( 16.067.0)syst( 06.0)stat( 12.056.0

SA

18071 2 190 2

[hep-ex/0502035]

16

backup slides

17

Probability Density Functions (PDFs) for the CP fit

)'(){(')1( ttRpfpftdfL sigKKol

a likelihood function

)()}'( tpfttRpf ololqqqqqq

1 qqK fff

qqK fff ,, are event fractions as functions of E and Mbc

),( iitot qtLL is maximized.

two free parameters: A and S

18

Probability Density Functions (PDFs) for the CP fit

]}sincos)[21(1{4

),;,(0

0/||

tmStmAwqwqeSAqtpB

tB

For +−

wrong tag fraction obtained from dataFor +−

)}cos()21(1{4

),(0

0/||

tmAwqwqeqtp effK

B

t

K

B

K

KeffK AA

AAA

1

)()()()(

KpKpKpKp

A

019.0109.0 KA HFAG2004

For qq

)}()1(2

{2

1),(

/||

tfefq

qtpqq

tqq

qq

qq

qq is set to 0 in default backup

19

New experimental situation

140fb-1

113fb-1

combined

BaBar 227M

inside the physical

boundary!

A

~2.3 difference between Belle and BaBar

20

~1010 BB/year !!& similar number

of +-

Lpeak = 1.41034cm-2s-1

Ltot = 350fb-1 (Feb.8, 2005)

Crabcavities

Major upgrade ofKEKB & Belle detector(>1yr shutdown)

world records !

1.4x1034

330 fb-1

5x1034

~1 ab-15x1035

~10 ab-1

Lpeak (cm-2s-1)Lint

Future prospect

21

Future prospectassuming the current measured values

~0.5ab-1

~2ab-1

~8ab-1

~32ab-1

Super-B factories can pin down the 2 value.

Current

22

Validity check with B0→+−

We extract B0→K+− events

LR>0.86

(including 2,207 K+−) eventsin total 4,293

good control sample because of the similar topology to B0→+−

Lifetime fit 04.051.10 B ps

03.046.0 m ps-1

t (ps)

(OF-

SF)/(

OF+

SF)

consistent with the WA values

23

Background asymmetry sideband

03.002.001.002.0

SA

sidebandB0:42467, B0b:42090 B0:43577, B0b:43308

03.001.001.000.0

SA

No background asymmetry

possible background asymmetries are included in the systematic error

24

Systematic errors

S A

wrong tag ±0.01 ±0.01physics param. <0.01 ±0.01resolution func ±0.04 ±0.01bkg Dt shape <0.01 <0.01event fraction ±0.02 ±0.04

fit bias ±0.01 ±0.01vertexing ±0.04 +0.03

−0.01tag side interfere

±0.01 +0.02 −0.04

total ±0.06 ±0.06

including uncertainties in the b.k.g and Final

State Radiation

O. Long, M. Baak, R.N. Cahn, and D.

Kirkby, PRD 68, 034010 (2003)

25

Validity check (lifetime fit with B0→+−)

2820 events

07.050.10 B ps

Lifetime fit

good agreement with the World Average (W.A.) Value014.0536.10

B ps PDG2004

B0→+−

continuum

26

Validity check with B0→+−

06.006.0 KA

total 4293 events

08.009.0 KS

B0: 2106, B0b: 2187

CP fit results

t (ps)

Raw

AC

P

consistent with the WA value

in agreement with null asymmetry