1. University of South Alabama
Measurement of the Branching Fraction of Y(4S) to Neutral B Pairs
Rafi Qumsieh, Shannon Eynon, Christopher Buchanan,
Dr. Romulus Godang
Physics Department
University of South Alabama
The 78th Annual Meeting of the Southeastern Section of APS
Roanoke, October 19-23, 2011

2. B mesons Mesons are made up of a quark and an anti quark. 𝐵 + ≡𝑢 𝑏
𝐵 − ≡ 𝑢 𝑏
𝐵 0 ≡𝑑 𝑏
𝐵 0 ≡ 𝑑 𝑏

3. Motivation
The semileptonic B meson decays play a prominent role in heavy quark physics.
𝐵 0 → 𝐷 ∗+ 𝓁 − 𝜈 𝓁 helps measure a fundamental decay parameter at the Υ 4𝑆 resonance.
All measurements of the parameters at the Υ 4𝑆 resonance are limited by uncertainty in the ratio 𝑓 +− 𝑓 00 ≡ 𝑅 +/0 (expected to be 1)
𝑓 00 ≡ℬ Υ 4𝑆 𝐵 0 𝐵 0 ≈50%
𝑓 +− ≡ℬ(Υ 4𝑆 𝐵 + 𝐵 − )≈50%
But…

4. By BABAR Collaboration: R. Godang
Previous Findings
The first direct measurement of 𝑓 00 was made in 2005
𝑓 00 =0.487±0.010(𝑠𝑡𝑎𝑡.)±0.008(𝑠𝑦𝑠.)
(Phys. Rev. Lett. 95, , 2005)
By BABAR Collaboration: R. Godang
Based on a data sample of fb −1 collected at the Υ 4𝑆 resonance.
We are looking for a more precise measurement with about 5× data.

5. Data Sample Current data and Monte Carlo -425 fb −1 (on resonance)
The data used in this research was collected with the BABAR detector at the PEP-II asymmetric energy B Factory at SLAC.
Current data and Monte Carlo
fb −1 (on resonance)
fb −1 (off-resonance)
-MC about 4× the amount of data

6. The BABAR Detector

7. Partial Reconstruction
𝐵 0 → 𝐷 ∗+ 𝓁 − 𝜈 𝓁 ( 𝐷 ∗+ → 𝐷 0 𝜋 + )
We look for the best combination of ℓ − and 𝜋 +
-leptons (electrons or muons)
-Right Sign ( ℓ ± 𝜋 ∓ → signal yield)
-Wrong Sign ( ℓ ± 𝜋 ± → checking)
Signal yield comes from the right sign distribution of the missing mass squared ( 𝑀 𝜐 2 ).
𝑀 𝜈 2 ≡ 𝐸 𝑏𝑒𝑎𝑚 − 𝐸 𝐷 ∗ − 𝐸 𝓁 2 − 𝑝 𝐷 ∗ + 𝑝 𝓁 2

8. Single Tag and Double Tag Analysis
In a single tag event one neutral B is partially reconstructed.
Signal Yield = 𝑁 𝑠
In a double tag event both neutral B are partially reconstructed.
Signal Yield = 𝑁 𝑑
𝑁 𝑑 = 𝑁 𝐵 𝐵 𝑓 00 ×[ℬ 𝐵 0 → 𝐷 ∗+ 𝓁 − 𝜈 𝓁 ×ℬ 𝐷 ∗+ → 𝐷 0 𝜋 + ] 2 × 𝜖 𝑑

9. Finding 𝑓 00 Efficiency correlation ≡𝐶= 𝜖 𝑑 𝜖 𝑠 2 .
C has been estimated to be 𝐶=0.998±0.006.
Combine 𝑁 𝑠 and 𝑁 𝑑 and solve for 𝑓 00 :
𝑓 00 = 𝐶 𝑁 𝑠 2 4 𝑁 𝑑 𝑁 𝐵 𝐵

10. Analysis
Get Raw Data
Raw data in the form of a histogram of the missing mass squared of neutrino.

11. Analysis Three Types of Backgrounds: 1. Continuum 2. Combinatoric
Peaking
Two Regions:
Side Band (-8,-4)
Signal (-2,2)

12. Analysis Types of Background:
Continuum Background: Continuum background events are non-resonant decays of:
Combinatoric Background: A random combination of real leptons from B decays paired with right sign soft pions from the other B. Found in SB and Signal Regions.
Peaking Background: Comes from decays of the type:
Found mainly in the Signal Region.

13. Process For each background: (except for Continuum)
We use Roofit to fit the MC background to the Raw Data.
Obtain a scaling factor and scale the MC down (MC is larger than Real Data)
Subtract the Background from Raw Data
Result:
Obtain the Signal Yield from the Raw Data. Then find Ns, Nd needed for

14. Process
Fit the Scaled Signal from MC with the Signal from Data (Background-Subtracted Data) to check how well our experiment matches with the MC simulation.
A fit of MC signal to signal from data.
Missing mass squared

15. Single Tag Right Sign Data Sample
First row: (left) raw data fit with the Roofit
(right) All single tag data events
Second row: (left) data with backgrounds subtracted,
(right) the ratio of data to MC simulation that is fitted with a straight line.

16. What have we been doing recently?
Improving Our Analysis by fitting each background separately to the data in its Main Region:
Finding a scaling factor for Comb. Background in the Side-Band Region only.
Finding a scaling factor for the Signal in the Signal region only.

17. Combinatoric Background MC Scaling
Here we use Roofit to fit the MC Comb. Background to the Raw Data in the Sideband region only where the Comb. Background most lies. We obtained a better result. As the graph shows:
A Fit of MC Comb. Background and Raw Data in the Sideband Region (-8,4)

18. Signal MC Scaling
We use Roofit to fit the MC Signal to the Signal from Raw Data in the Signal Region only. This process is used to check how good our data agree with the theory. We obtained a better fit using the new seperation technique.
A Fit of Signal MC to Signal from Subtracted Raw Data.

19. Updated Single Tag Right Sign Data Sample
IMPROVEMENTS:
Better Scaling Factors
By:
Fit Comb.Background MC with Comb. From Raw Data on SB region only.
Fit Signal MC with Signal from Raw Data on Signal Region Only.
 Gives Better Results For Scaling Factors!

20. Updated Double Tag Right Sign Data Sample
IMPROVEMENTS:
Better Scaling Factors
By:
Fit Comb.Background MC with Comb. From Raw Data on SB region only.
Fit Signal MC with Signal from Raw Data on Signal Region Only.
 Gives Better Results For Scaling Factors!

21. Conclusion and Plan Precise Measurements of are important for:
Normalizing many B decay branching fractions.
Partial Reconstruction technique:
a. Helps much on statistical uncertainties.
b. Offers direct measurements.
We expect to measure more precise results than the published result (coming soon).
We can get more precise measurements along with lower statistical uncertainties if we can get better understanding on the backgrounds.

22. Acknowledgments References
This work was supported by the University Committee on Undergraduate Research (UCUR) program (at the University of South Alabama)
And the U.S. Department of Energy under grant No. DE- FG02-96ER
The BABAR Collaboration, B. Aubert et al., Phys. Rev. Lett. 95, (2005).
The BABAR Collaboration, B. Aubert et al., Phys. Rev. Lett. 100, (2008).
BABAR Analysis Document #2168, R. Godang, C. Buchanan, and S. Eynon (2010).
References