1. Reaching for  (present and future)
Andrey Golutvin
ITEP/Moscow
Reaching for 
(present and future)

2.  = arg | | Unitarity Triangles This talk V ud Vub* -V cd Vcb*
Bd0  p+ p-
Bd0  r p
Bd0  DK*0
BS0  DSK
Bd0  D* p, 3p
BS0  DS p
Bd0  J/y KS0
BS0  J/y f
This talk
 = arg | |
V ud Vub*
-V cd Vcb*

3. CP violation in Interference of
There are a few methods of determining 
CP violation in Interference of
Mixing & Decay
CP violation in Decay
Rates and asymmetries in K Decays
Amplitude triangles for B Dcp0K
(GLW method)
+ extensions/modifications
Time dependent asymmetries
in Bd+- and BsK+K- decays
assuming U-spin symmetry
Time dependent flavor tagged
analysis of
BsDsK- decay
BdD*+ decay

4. Data Samples
(present and future)
Results presented today are based on 81.2 fb-1(BaBar) and 78 fb-1(BELLE)
That corresponds to 88 and 85 mln BB pairs (part of data sample)
Statistics available today at B-factories  300 mln BB pairs - - -
In 2006 one expects  1.2 bln BB pairs -
In 2008 after 1 year of LHCb (and BTeV) operation  bb pairs
Super Bd-factory: 1.5 to 5 bln of “clean” BB events
depending on achieved luminosity -

5. g from B  K decays Difficulties: Theoretical EW penguins
Significant interference of
tree and penguin amplitudes
Expect potentially large CP asymmetries
QCD penguins play the dominant role
Difficulties:
Theoretical
EW penguins
SU(3) breaking + (FSI) effects
Experimental
small BF, qq background
Particle ID is crucial
Measurement of the ratios of
BF and ACP may provide
constraints on  and strong phase 
dir _

6. Summary of BFs [10-6]
From J.D. Olsen talk
at the CKM workshop 2003

7. Comparison of BF No sign of rescattering yet
Present precision is dominated by statistical errors
In a few years statistical and systematical errors should become comparable

8. A Specific Model: QCD Factorization
Beneke et. al., Nucl. Phys. B606, 245 (2001)
Data (2001)
Data (2003)
Inconsistent?
Data start to constrain models

9. CP asymmetries
(allows one to eliminate dependence on the strong phase )
Largest deviations: 2s each in K+p-(BaBar), r+p-(BaBar), and K+p0(Belle)

10. uncertainties In 1 year BTeV expects:
In a few years statistical precision in the measurements
of Branching Fractions (BK/) should reach the level
of systematic error, 2-4% depending on decay mode
Extraction of  will be dominated by theoretical
uncertainties
BaBar: K0+
80 fb-1
In 1 year BTeV expects:
4600 K0+ events (with S/B = 1)
62100 K-+ events (with S/B=20)

11. Amplitude triangles GLW method B-  DCPK- where DCP=(1/2)(D0 D0 )
A(B-  DCPK-)  |A(B- D0K- )| + |A(B- D0K- )|eiei
A(B-  DCPK+)  |A(B- D0K+ )| + |A(B- D0K+ )|e-iei
CP-even states: K+K-, +-
CP-odd states: Ks0, Ks, Ks, Ks, Ks’

12.  can be extracted from the reconstruction of two triangles
Experimental problems:
A(B-D0K-) is small squashed triangles
DCSD lead to the common final states for D0 and D0
(color suppressed A is of the same order as color favored  DCSD)
Experimental observables:
BF(B-D1,2K-) /
BF(B-D1,2-)
, A1 and A2
R1,2 =
BF(B-D0K-) /
BF(B-D0-)
allow, in principle, to extract RDK- ,  and 

13. No constraints on  possible with this statistics …
BaBar reconstructed
D1K-
K+K-
(7.41.70.6)%
R1=
(8.310.350.20)%
+0.09
AD1K- = 0.170.23
-0.07
BELLE studied both
D1K-and
D2K-
R1 = 1.21  0.25  0.14
R2 = 1.41  0.27  0.15
AD1K-=  0.19  0.04
AD2K-=  0.17  0.05
No constraints on  possible
with this statistics …
Needed:
significantly higher statistics
precision measurements of D
Branching Fractions

14. Triangles are not as squashed
B0D0K(*)0 Mode
Two comparable color-suppressed amplitudes
Triangles are not as squashed
as in B DK case _ _
BELLE
+1.3
BF(B0 D0 K0) = ( 0.6)  10-5
-1.2 _ _
+1.1
BF(B0 D0 K*0) = ( 0.6)  10-5
-1.0
Hope to see soon B0 D0 K*0 decay
Present Upper Limit is
BF(B0 D0 K*0)  1.8  10-5

15. KS+- receives contributions from:
ADS variant of the amplitude triangle approach:
B-  D0[fi] K- and B-  D0[fi]K-, fi= K-+ or K-+0
Use large interference effects
Experimentally challenging since BF of O(10-7) are involved
Dalitz plot analysis of D  KS+ - for B-DK- and B+DK+ decays
KS+- receives contributions from:
B- D0K- D0K*-+ K*- KS-
B- D0K- D0K*+- K*+ KS+
B- D0K- D0KS 0  +-
Enough information to extract  for any magnitude of 

16. B0D*- Mode (2 + ) is extracted from the measurement of_ u c +
D*- _
favored
suppressed d d _ _ b c b u _ B0 + B0
D*- _ _ d d d d
Suppressed/favored < 10-2
Small CP-violating effect expected
Time evolution:
(BD*)  (1+RD*2)  (1-RD*2)cos(mt)  2 RD*sin(2+) sin(mt)
(2 + ) is extracted from the measurement of
4 time-dependent asymmetries
RD* can be estimated using BF(B0Ds+-):
BF(B0Ds+-) 
BF(B0D*-+)
fDs2
RD*2  
tan2c
fD*2

17. Evidence for B0Ds+- RD*  0.022  0.007
Br(B0Ds+-)Br(Ds++)
BaBar (1.13  0.33  0.21)x10-6
BELLE (  0.11) x10-6
+0.37
-0.30
BELLE have measured
Br(Ds++)=(3.72 ) x10-6
using full/partial reconstruction method
+0.47
-0.39
RD*   0.007

18. D*+- Data Samples (partial reconstruction)
20 fb-1, ± 240 events
B0 = ± ± ps
30 fb-1 , 3430 ± 80 events
m = ± ± ps-1
(sin(2+))   0.15 for 1000 fb-1 T. Gershon (CKM workshop, 2003)
or (2+)  10o

19. g from Bs  D-s K+ , D+s K- Looks as most clean method to measure 
Same principle as for B0D*-
Since  is small  sensitivity should not depend on the value of 
Two decay modes with comparable amplitudes,   10-4 each
 can be determined from the measurement of 4 time-dependent
asymmetries (assuming that  is fixed from Bs mixing)
Contribution from NP is unlikely !
Looks as most clean method to measure 

20. Bs  Dsp, DsK kinematics at LHCb Ds
72k Ds 
8k Ds K
 ~ 6.5 MeV/c2
s = 168 mm
s = 418 mm
Ds mass (GeV)
Bs vtx resolution (mm)
Ds vtx resolution (mm)

21. Importance of Particle ID
Needed:
Hadronic trigger
K/p separation
Good proper time resolution

22. Sensitivity slightly depends on strong phase difference (T1/T2)
In one year of LHCb running:
8k BsDsK reconstructed events
Sensitivity slightly depends on
strong phase difference (T1/T2)
and xs
Assumptions:
3k tagged BsDsK-
30% background
30% mistag
LHCb:
()  10º for xs = 15
()  12º for xs = 30
BTeV:
()  8º

23. where dexp(i) is a function
g from B(s) p p, K K proposed by R. Fleischer
Measure time-dependent asymmetries in B0  +- and Bs  K+ K-
decays to extract 
where dexp(i) is a function
of T and P amplitudes
Assuming U-spin flavor symmetry: d=d’ and =’
 can be extracted

24. g from B(s) p p, K K BS  K+ K- BS  K+ K- B0  + - B0  + -
yield
27k
35k
B/S
0.8
0.55 xq
0.755 20
D2
0.064
-0.30
0.16
0.58
-0.17 DG
0.0
input values
Evaluation of and sensitivity
from time-dependent measured asymmetry
B0  + -
BS  K+ K-

25. B0  + - BS  K+ K- Combining all information: () = 2.50
In one year of LHCb operation :
BS  K+ K-
increasing xs 
B0  + -
(A dir )  (A mix )  0.054
BS  K+ K-
(A dir )  (A mix )  0.043
Sensitivity to :
assuming 4 years of data taking
(4x27k Bd and 4x35k BsKK)
B/S = 0.8 and 0.55 correspondingly
tagging: D2 = 6.4%
Combining all
information:
() = 2.50

26. Sensitive probe for New Physics
Conclusions
 can be measured both using CPV in decay (time-integrated)
and in the interference of mixing and decay (time-dependent)
Only trees
(1) Theoretically clean: b c X
Amplitude triangles
BsDsK
Trees + loops
(2) Experimentally easier b  hh (h = non-c)
Experimental precision on BF already below 10% level
Clear theoretical strategy how to extract  is still missing
Measurement of rare decays with BF  10-7could help
Sensitive probe for New Physics
(1) vs (2)

27. ACP=ACP(dir)cos(mt) + ACP(mix)sin(mt)
Expectations from CDF
Run IIa
ACP=ACP(dir)cos(mt) + ACP(mix)sin(mt)
/K/KK/K separation:
Inv. mass ~ 20 MeV/c2
Kinematical variables
dE/dX
Md << Ms
—sum
BdK
BsKK
Bd 
BsK 
RunIIa: ACP(dir, mix)~0.2 (Bd)
~0.1 (BsKK)
() = ±10(stat)