1. fully reconstructed B’s in B-factory experiments
Tagging with
fully reconstructed　B’s in B-factory experiments
Takahiro Matsumoto,
Belle Collaboration
Tokyo Metropolitan University
Dec 03 11:50-12:10

2. Fully reconstructed B tagging
－Offline single B meson beams with known p（B）, flavor.
－Large advantage to study decays including n and t
Υ(4S)
e- (8GeV)
e+(3.5GeV) B p
Interesting decays
B → uln
B → tn, Knn
B → Dtn
etc
Full Recon.
Full reconstruction efficiency = O(0.1%)
Need large data sample ⇒　unique at B factories

3. The KEKB Collider (8 x 3.5 GeV, X angle)
World record: L=(1.0 x 1034)/cm2/sec
Accumulated 158/fb by this summer
152 million BBbar pairs
KEKB Collider

4. Event display for fully reconstructed two B’s

5. Kinematic variables for the Υ(4S)
Energy difference:
Beam-constrained mass:

6. Fully reconstructed B tagging
B decay patterns
Hadronic : ~ 80%
A lot of decay modes with Br ≤ O(1%)
Semileptonic : ~20%
Mainly for BD*(e,m)n (~10%), B momentum unknown
Reconstruction
Hadronic : Collect as many modes as possible
Requires many CPU time due to multiple combination
Semileptoic :
Neutrino reconstruction with assumption of rest B at U(4s) CM frame
Hadronic B decays
B D(*)(p, r, a1)-, J/Y(KS, K*)…
Semileptonic B decays
B D(*) l n

7. B’s Reconstruction [Hadronic]
Collect known major modes [Belle]
[ low efficiency, but high purity ]
BD(*)(p,r,a1)-, D(*)DS(*), (J/Y, Y(2s), cc1)K(*)
Semi-inclusive reconstruction, BD(*)X [BaBar]
[ high efficiency, but low purity ]
X=n1p±n2K±n3KSn4p0, total charge ±1
n1 + n2≤ 5, n3≤2, n4≤2
Treatment of the multiple candidates
Select the best reduced c2 events to each mode ( from M(D), M(D*)-M(D), DE etc.)
Select the mode with the best purity

8. BD(*)(p,r,a1)- Main hadronic B decay sample [ Br ~ 1% ]
D(*) is D0, D+, D*0 or D*+ [ recon. from D*D(p,g) and DKnp ]
Recon. eff. : ~0.17% for B±, ~0.07% for B0
High purity ~0.8
w/ 152 million BBbar

9. BD(*)Ds(*) Br ~1%, but low yield due to D and Ds reconstruction
Ds(*)+ is recon. from Ds*+Ds+g, Ds+K+K-p+,KsK+p+p-,KSK+
~7% yield of BD(*)(p,r,a1)-
Moderate purity ~ 0.6
Neutral B
Charged B
S= 4390
S= 8700
w/ 152 million BBbar

10. BCharmonium Color suppressed modes [ Br ~ 0.1% ] Reconstructed modes
B- J/YK-, B0 (J/Y, Y(2s), cc1)KS, J/YK*0
~9% yield of BD(*)(p,r,a1)-
Very high purity ~0.95 [ due to clean recon. of J/Y l+l- ]
Charged B
Neutral B
S= 9373
S= 7275
(CP modes= 3119)
w/ 152 million BBbar

11. BD(*)X [ semi-inclusive , BaBar ]
BD(*)（ n1p±n2K±n3KSn4p0）±
In total, 1097 combinations!
c.f. 64 combinations for BD(*)(p,r,a1)-
Mbc w/ lepton (Pl*>1 GeV/c)
High eff.
~0.5% for B±, ~0.3% for B0
Low purity: ~ 0.25
~0.67 w/ lepton (P*l>1 GeV/c) in recoil B side
S ~30k
BD(*)X is categorized by Mx region depending on purity and select the best purity mode for each event ( also mode dep. threshold cut on purity is applied ).
*hep-ex/ for Vub paper with BXuln
w/ 89 million BBbar

12. BHadronic, Summary BD(*)(p,r,a1)-, D(*)DS(*), (J/Y, Y(2s), cc1)K(*)
eff. ~ 0.2% (0.1%) for charged (neutral) B
150k B±, 69k B0 in 152 million BBbar sample
High purity ~ 0.81
BD(*)X
eff. ~ 0.5 % (0.3%) for charged (neutral) B
Low purity ~ 0.25
How to improve eff. and quality?
Add more decay modes
ex. D+  K-p+p+, KSp+ add ~50% with D+ K-p+p+p0
Study BD(*)X in detail. By understanding structure of X, we can optimize the selection criteria further
Further cut tuning
etc.
Steady improvement will be possible

13. B’s reconstruction [Semileptonic]
Reconstructed modes
B- D0l-n ( Br=2.15±0.22% )
D0K-p+, K-p+p0, K-p+p+p-, KSp0
*B- D*0l-n ( Br= 6.5 ±0.5%) is automatically included in the sample

14. B- D(*)0l-n B± tagged events = 714k, eff. = 1.7%
Purity for B- D(*)0l-n = 13.9%
(Missing mass)2 for neutrino
Large eff., but
Large combinatoric backgrounds
w/ 85 million BBbar

15. Application to physics analysis
Fully reconstructed B tag sample is widely used for physics analysis since B-factory experiments succeeded to accumulate large sample of BBbar pairs
BXuln ( for Vub analysis )
hep-ex/ [BaBar]
BKnn, tn
PRL86(2001)2950 [CLEO], BaBar conference paper
For the decays with low multiplicity (BKnn, tn etc.), we can suppress combinatoric background significantly, then semileptonic B tag ( ~1% eff. ) sample can be also utilized
Other interesting physics analyses are also expected
BD(*)tn, Btt etc. or detailed study of inclusive B decays

16. Conclusions
A large number of fully reconstructed B sample is already available in B-factory
 Provide tools for innovative B physics analysis
Reco. eff.
~ 0.15%(0.4%) for Hadronic B tag [ () is for BD(*)X ]
~ 1.7% for Semileptonic B tag*
* useful for low multiplicity B decays
We already have a sample of more than 105 fully reconstructed B mesons. We expect a lot of interesting physics results with these sample in near future