1. Experimental review of CKM sides at B factories
A. Oyanguren (IFIC, UV-CSIC)
(BaBar Collaboration)
DISCRETE 2008
Vtd
Vub
Vcb

2. Introduction & outline
The Cabibbo-Kobayashi-Maskawa matrix:
Vud Vus Vub
Vcd Vcs Vcb
Vtd Vts Vtb
VCKM =
VudVub* + VcdVcb* + VtdVtb* = 0
 From Bd and Bs mixing
hadron colliders (md/ ms)]
 From exclusive sl decays:
|Vcb|: BD*l, BDl;
|Vub|: Bl, Bl
B(‘)l, Bl… A
~|Vtd/Vts|
 From radiative decays (bd/bs)
 From inclusive BXcl (|Vcb|)
and BXul (|Vub|) decays
~|Vub/Vcb|
(1,0)
(0,0) 1
DISCRETE 2008
A. Oyanguren 2

3. B-Factories  1200 fb-1 recorded ! Analysis techniques: B0 B0 B0 - D0
Hadronic or Semileptonic Tag
Untagged
(High efficiency)
(High purity) +  D0 -
lepton X B0 + - D0 K+
lepton  X B0
Signal side
Signal side
lepton  X B0 D*
(4S)
(4S)
(4S)
Tag side
Tag side
DISCRETE 2008
A. Oyanguren 3

4. |Vcb| from exclusive sl decays
From B  D   (BR ~2% , suppressed at high q2)
From B  D*   (BR ~ 5%)
Vcb q2
Kinematic function
Form factor
(shape & normalization)
Theory:
B  D D2
G(1)*=1.074   0.016
B  D* D*2, R1(1) , R2(1)
F(1)*=0.921   0.020
Measuring the decay rate and ff shape  |Vcb|
* M.Okamoto et al NPPS 140, 461(2005))
* C. Bernard et al arXiv: [hep-lat]
DISCRETE 2008
A. Oyanguren 4

5. |Vcb| from BD Select the lepton,
reconstruct the D0 into several channels
 Reconstruct the other B into hadrons (Btag)
 Reconstruct the missing mass:
m2miss = (p(4s) – pBtag-pD-pl)2
m2miss = 0.04 GeV2
 Fit the signal yield into 10 w bins (ML fit)
417fb-1
2 = 1.20 0.09  0.04
 2 fit to extract D2 and G(1)|Vcb|
G(1) |Vcb| = (43.0 1.9  0.4) 10-3
B (B0D+-)= (2.170.06 0.07)%
arXiv:
DISCRETE 2008
A. Oyanguren 5

6. |Vcb| from BD* Reconstruct the semileptonic decay
D0K, K3 + soft  (to make D*)
+ lepton
The decay rate depends
on w, , V , 
Constraint pB  pinclusive = pall –D*
 Define the angle between the B
and the D* candidate
 4D 2 fit to extract D*2, R1, R2 and F(1)|Vcb|
w = 0.025
cosv= 0.052
 Fit the signal yield
140 fb-1
= 6.47
cos =0.047
preliminary
F(1) |Vcb| = (34.4 0.2  1.0) 10-3
R1= 1.495  0.062
2 = 1.293  0.029
R2= 0.844  0.019
B (B0D*+-) = (4.420.03 0.25)%
DISCRETE 2008

7. Global fit to BD and BD*
207fb-1
Reconstruct D0 and D+ candidates
(D0K-+, D+K-++)
B(B-D*0-)= (5.370.02 0.21)%
D*2 = 1.21 0.02  0.07
Use p*, p*D and cosB-D to separate
BD and BD* decays
F(1) |Vcb| = (35.7 0.2  1.2) 10-3
B (B-D0-)= (2.360.03 0.12)%
3D 2 fit to the yields to extract the
form factor slopes, D and D* and BD
and BD* BRs (R1 and R2 fixed).
D2 = 1.22 0.04  0.07
G1) |Vcb| = (43.8 0.8  2.3) 10-3 D
preliminary D*
1.5 2 1
1.5 2 -2 -1 1
arXiv:
DISCRETE 2008
A. Oyanguren 7

8. |Vcb| from exclusive sl decays
G(1)|Vcb| = (42.4  1.56) x 10-3
2/dof =1.2/8
using G(1) =1.074  0.024
|Vcb| = (39.5  1.5exp  0.9latt ) x 10-3
Exp: - waiting Belle update
- systematics  tracking eff,
radiative corrections, D**
F(1)|Vcb| = (35.41  0.52) x 10-3
2/dof =39/21
using F(1) =0.921  0.024
|Vcb| = (38.4  0.6exp  0.9latt) x 10-3
- systematics  tracking eff, D** 8

9. BD**l ?  Fully reconstructed analyses HQET 605 fb-1 D* D** D
PRD77,091503(2008)
605 fb-1 D* ?
D** D
Ground
states
Broad states
Narrow
states
B(BXc) > (BD) + (BD*) + (BD**)
[10.1  0.4]%
> [2.17  0.12]%
+ [5.16  0.11]%
+ [1-2]%
B0D*0+-
B-D*+--
417 fb-1
B-D+--
B0D0+-
 Belle does not see the D1’,
BaBar does.
arXiv:
Theory predicts B (D1*,D2*) >> B(D*0,D1’), Experiments see B (D1*,D2*) <~ B(D0*,D1’) 9

10. |Vcb| from inclusive sl decays
Operator Product Expansion (OPE): = x
Vcb
f(parameters related with b quark properties inside the hadron)
 kinetic energy, spin...
Some inclusive observables  O  depend on the same parameters
BXc
(For instance:
: the lepton energy, the mass distribution,
BXs
: the photon spectrum)
Measurement of moments: 
DISCRETE 2008
A. Oyanguren 10

11. MXc and El moments One B fully reconstructed into hadrons (Btag)
One lepton from the other B (signal)
Rest of particles  measure mx
 measure E (electrons)
Event by event calibration
Unfolded with SVD algorithm
< Mxn> measured as function of p*l cut
(from 0.8 to 1.9 GeV/c)
< Een> measured as function of p*e cut
(from 0.4 to 2. GeV/c)
PRD 75, (2007)
210fb-1
140fb-1
e+
arXiv:
< Mxn> for n=1,2,3,4,5,6
< Een> for n=0,1,2,3,4
DISCRETE 2008
A. Oyanguren 11

12. E moments All measurements: g W+ b t S
Select high energy photons (E*> 1.4 GeV)
Off-peak subtraction,
veto to photons from 0 and 
All measurements:
Unfolded with SVD algorithm
< En> measured as function of
E cut (from 1.7 to 2.1 GeV)
605fb-1
< En> for n=1,2
PRD 78, (2008)
DISCRETE 2008
A. Oyanguren 12

13. |Vcb| from inclusive sl decays
 Using 27 moments from BaBar, 25 moments from Belle and 12 moments from other experiments (
|Vcb| =(41.67  0.43fit  0.08 B  0.58th) x 10-3
mb =(4.601  0.034) GeV
2 =(0.440  0.040) GeV2
(Exclusive (BD*):
|Vcb| = (38.4  0.6exp  0.9latt) x 10-3
 |Vcb|incl - |Vcb|excl : 2.5
DISCRETE 2008
A. Oyanguren 13

14. |Vub| from exclusive sl decays
Analysis methods:
 Untagged (high stat., most precise)
 Tagged (hadron or semileptonic)
(BR ~ 0.01% +
large b c bkg)
p, h, h’, r, w
Untagged B
 reconstruction from full event
(p = pbeams - pi)
 Fit E = EB – s /2 and mES = s/4-|p*B|2
to extract the signal yield
206fb-1
q2= (pB - p )2 in 12 bins  Form factor fit
B (B0-)= (1.460.07 0.08)x10-4
PRL 98, (2007)
DISCRETE 2008 14

15. |Vub| from exclusive sl decayspB
Tagged B
(sl tag) B
(hadronic tag)
pD(*)l
pl
 Reconstruct one B into hadrons (Btag)
Select the lepton from the other B
One B decaying into B D(*) (Btag)
select the  ( rec.  or  into pions)
Other lepton (-q) from the other B
 Reconstruct the missing mass:
m2miss = (p(4s) – pBtag-p-pl)2
Define the systems Y=D(*) and X=
Fit to cos2B in 3 q2 bins
Extract yields in 3 q2 bins
348fb-1
B0-
B+0
605fb-1
signal
signal
[hep-ex]
B (B0-)= (1.120.18 0.05)x10-4
B (B0-)= (1.540.17 0.09)x10-4
B (B++)= (0.660.12 0.03)x10-4
DISCRETE 2008
A. Oyanguren 15

16. |Vub| from exclusive sl decays
Form factor necessary to extract |Vub| theory
Models depend on q2 region:
(Untagged analysis can
measure the form factor
from data)
B (B0-)= (1.340.06 0.05)x10-4
PRL 98, (2007)
DISCRETE 2008
A. Oyanguren 16

17. |Vub| from exclusive sl decays
 Check theoretical calculations
 Composition of BXu
348fb-1
605fb-1
signal
hadron tag
sl tag:
[hep-ex]
347fb-1
B (B0-)= (2.800.18 0.16)x10-4
untagged
[hep-ex]
DISCRETE 2008
A. Oyanguren 17

18. |Vub| from inclusive sl decays:
BX = BXc + BXu
 Theory needs low momentum cuts to extract|Vub| (OPE breaks down in the Shape Function (SF) region:
low mx2 and q2)
 OPE parameteres (mb, 2 ) from moments in BXc and BXs
 Experiments need high momentum
cuts to suppress b  c bkg
Experimental technique:
Measure partial BR
 Tagged (sl or hadronic) or untagged
 Use kinematic variables to suppress as much as possible b  c (keeping reliable theoretical calculations):
El, q2, mX and P+ = EX -|pX|
DISCRETE 2008
A. Oyanguren 18

19. Untagged: endpoint method
Tagged (hadrons or sl)
Select regions below charm threshold
Key: bkg subtraction (off-peak data,
endpoint, MC… )
 bc bkg constrained by data in the
low energy region
One B fully reconstructed
Access to all kinematics (mx, q2, P+)
for the signal side
81fb-1
signal
signal
PR73:012006,2006.
signal
27fb-1
bc (MC)
347fb-1
bkg
bkg
1.55 GeV
~40-60% acceptance
0.66 GeV
signal
8 GeV2
arXiv:
PRB601:28(2005)
81fb-1
signal
604fb-1
bkg
Shmax=f(E, q2) = maximum mx squared
(new Belle multivariate analysis  no hard cuts, 90% PS)
PRL95:111801(2005)

20. |Vub| from inclusive sl decays
Tagged (hadrons or sl)
Relating the rate to bs to reduce
model dependence (LLR method)
 Moments of the full mass spectrum
 OPE fit to extract mb, 2 and .
unfolded
220fb-1
80fb-1
B  Xc
B  Xs 
B  Xu
mx cut =1.67
mb =(4.604  0.250) GeV
arXiv:
2 =(0.398  0.240) GeV2
PRL96:221801(2006)
mx cut (GeV/c2) 20 20

21. |Vub| from inclusive sl decays
OPE based + SF
‣BLNP: HQE with systematic introduction of SF
PRD71: (2005)
‣BLL: phase space in mx-q2 with reduced SF
PRD64: (2001)
‣GGOU: kinetic scheme
JHEP 10(2007) 058
‣LNP (LLR): b→s directly related to b→ uν
JHEP 0510:084 ( PLB 486:86 )
No SF introduced (model non-perturbative QCD)
‣DGE: Dressed Gluon Exponentiation
JHEP 0601:097 (2006)
‣ADFR: Analytic coupling
PRD74:03400(4,5,6) (2006)
Vub incl
Vub excl
More data, more SP +
improvements from theory needed
|Vub|from =(3.44  0.16)10-3
DISCRETE 2008
A. Oyanguren 21

22. |Vtd/Vts| from radiative decaysB-
 - , / K* b
u,c,t u
d/s 
Exclusive
Inclusive
(Sum of
all channels)
Form factors
arXiv:
Select high energy photons
veto to photons from 0 and 
B → Xsg
B → Xdg
remove bkg from continuum
 Yield from E = EB – s /2
and mES = s/4-|p*B|2
0.6 < MXd/s < 1.0 GeV
Need SuperB! 22
PRL101:111801,2008
DISCRETE 2008

23. (Vtd/Vts)/(Vtd/Vts)
Summary
The CKM matrix:
 decays
 Decays (0+0+)
Vud/Vud
~ 0.03%
Vus/Vus
~ 0.4%
Vub/Vub
~ 8%
Vcd/Vcd
~ 5%
Vcs/Vcs
~ 11%
Vcb/Vcb
~ 2%
Need theory improvements
(excl . vs incl. ??)
(Vtd/Vts)/(Vtd/Vts)
~ 3%
Vtb/Vtb
~ 15%
charm sl decays
DISCRETE 2008
A. Oyanguren 23

24. Backup
HFAG
DISCRETE 2008
A. Oyanguren 24

25. Backup
HFAG
DISCRETE 2008
A. Oyanguren 25