Peter Krian

1
CP violation and related issuesPart 9
measurements of Vub

• Peter Krian
• University of Ljubljana and J. Stefan Institute

2
Contents
Importance of Vub Measurements of Vub
3
Unitary triangle Vub
Vud Vus Vub Vcd Vcs Vcb Vtd Vts Vtb
f2(a)
f3(g)
f1(b)
(0,1)
(0,0)
4
CKM matrix measurements
CKM matrix is unitary -gt angles should add up to
1800 -gt sides should fit the same
triangle Deviations of individual measurements
could signal processes not included in SM
5
Vub measurements
From semileptonic B decays
b
c,u
Vub, Vcb
Vcb known to 1.4, becoming as precise as
Vusl (1)
need to pin-down Vub, present world average
error 10 b?cln background typically an order
of magnitude larger.
6
Inclusive Vub measurement

• Traditional inclusive method use semileptonic
  decays, fight the background from b?cln decays
  by using only events with electron momentum above
  the b?cln kinematic limit. Problem extrapolation
  to the full phase space? large theoretical
  uncertainty. -gt new development
• New method fully reconstruct one of the B
  mesons, check the properties of the other
  (semileptonic decay, low mass of the hadronic
  system)
• Very good signal to noise
• Low yield (full reconstruction efficiency is
  0.3-0.4)

7
Electron spectrum endpoint
b?cln (yellow) an order of magnitude larger than
b?uln (blue) Measurement region - traditionally
between the b?cln endpoint and the b?uln
endpoint -gt Huge extrapolation, model
dependent... New reduce the background and
model the remaining background better
2.3 GeV/c lt pe lt 2.6 GeV/c (CMS)
8
Electron spectrum endpoint the method
Measurement region Background estimation
region
1.9 GeV/c lt pe lt 2.6 GeV/c (CMS)
1.5 GeV/c lt pe lt 1.9 GeV/c (CMS)
Deal with large backgrounds
QED radiative corrections included

• BB backgrounds
• B?Xcln
• Leptons from other decays
• (J/?, ? (2S), ? conv.)
• Fake electrons

MC simulation Den (ISGW2) Den (HQET) Den
(ISGW2)
Fit (D D)ln/ D ln relative contributions
Veto on invariant mass
Estimated using Ks?? ?
9
How to deal with large non-BB background
Visible energy Charged multiplicity Fox-Wolfram
moments Fisher discriminant Energy flow
variables Thrust axis Rare B decay tag

• Non BB backgrounds
• Continuum (ee- ?qq )
• QED processes

Subtraction of continuum (8.8fb-1 of
offresonance data)

10
Electron spectrum endpoint the result ( 27 fb-1 )
BB background (b?c MC expectation)
continuum background
Systematic uncertainty Model dependent signal
eff.1.7 - 3.4 B ?Xcln background estimation
17
b?u MC expected shape
11
Vub extraction
Partial BR -gt determine from the data
Bosch, Lange, Neubert, Paz, Nucl.Phys. B699 (2004)
1.9 GeV/c ltpelt 2.6 GeV/c
Vub (4.50 ? 0.42 ? 0.32 ? 0.21) ? 10-3
exp
SF
theo
Total error onVub ..... 13
12
New inclusive Vub measurement

• New method fully reconstruct one of the B
  mesons, check the properties of the other
  (semileptonic decay, low mass of the hadronic
  system)
• Very good signal to noise
• Low yield (full reconstruction efficiency is
  0.3-0.4)

13
Fully reconstructed sample
Fully reconstructed sample Clean environment but
small sample ? reco ? 3?10-3 Exclusive method
180 decay channels
B0
Signal ..... 254411 bkgd ..... 177669 purity
..... 0.59
Reconstructed channels B0 ? D()? / D()? /
D()a1/ D() Ds() B ? D()0 ? / D()0 ? /
D()0a1/ D()0 Ds() D0 ? D0?0 D ? D0?/
D?0 Ds ? Ds? D0 ? K?/K??0/K???/Ks?0/Ks??/Ks???
0/KK D ? K??/K???0/Ks?/Ks??0/Ks???/KK? Ds ?
KsK?/KK?
253 fb-1
B
Signal ..... 422753 bkgd ..... 255446 purity
..... 0.62
14
Vub measurement
Variables separating b ? uln from b ? cln
lepton energy El
hadronic inv. mass Mx
leptonic inv. mass q2
Full reconstruction Belle B?D()-
p/r/a1/Ds() e 0.25 BaBar B?D()- n1p n2K
e 0.4
15
Mx analysis

• Use the mass of the hadronic system Mx as the
  discriminating variable against b ? cln
• Mx mass of all hadrons from the B decay. Expect
• Mx for b ? cln to be above 1.8 GeV (b ? cln
  results in a D meson with gt1.8 GeV)
• Mx for b ? uln to mainly below 1.8 GeV (B-gtpln,
  rln, wln ...)

16
Mx analysis
Mxlt1.7 GeV/c2 / q2gt8 GeV2/c2
Total error onVub ..... 12
0.20
Vub (4.34 ? 0.22 ? 0.19 ? 0.13 ? 0.12 ?
0.33 ) ? 10-3
253 fb-1
0.22
stat
syst
b?c
b?u
theo
SF
model dep.
Mxlt1.7 GeV/c2 / no q2 cut total error onVub
..... 11
0.13
253 fb-1
Vub (3.80 ? 0.17 ? 0.13 ? 0.12 ? 0.04 ?
0.29 ) ? 10-3
0.14
stat
syst
b?c
b?u
SF
theo
model dep.
17
Vub Results
Lepton endpoint (p gt 1.9 GeV/c )
Vub (4.50 ? 0.15 ? 0.55) ? 10-3
13
Full reconstruction tagging
Mx/ q2
Vub (4.34 ? 0.29 ? 0.43) ? 10-3
12
Mx
Vub (3.80 ? 0.21 ? 0.35) ? 10-3
11
P
Vub (3.87 ? 0.25 ? 0.43) ? 10-3
13
18
Backup slides
19
Standard Vub method
De FazioNeubert, JHEP 9906,017 (1999)
Belle measurement of mb and mp2
spectral distortion due to final state radiation
LimosaniNozaki hep-ex/0407052 Belle
Collaboration,Phys. Rev. Lett. 93 061803 (2004)
1.9 GeV/c ltpelt 2.6 GeV/c
Vub (5.01 ? 0.47 ? 0.17 ? 0.32 ? 0.24) ?
10-3
stat/sys
theo
exp
Br ?Vub
fu error
1.9
2.0
2.1
2.2
2.3
2.4
Total error onVub ..... 13
Vub vs. lower m. cut
20
P analysis
P Ex px P Ex px (Mx2 P P)
P
New proposed kinematical variable
Bosch, Lange, Neubert, Paz, Phys Rev Lett 93,
221801 (2004)

• 
  

0.12
Vub (3.87 ? 0.18 ? 0.18 ? 0.12 ? 0.17 ?
0.35 ) ? 10-3
253 fb-1
0.13
stat
syst
b?u
b?c
SF
theo
model dep.
Total error onVub ..... 13