BR(Bs→J/ψφ) measurement and extraction of the fragmentation fractions


Primary authors: Olga Norniella, Kevin Pitts

Introduction

We report a measurement of the ratio of the branching fractions of Bs→J/ψφ to B0→J/ψ K* based on 9.6 fb-1 of data. We use samples of reconstructed Bs→J/ψφ and B0→J/ψ K* collected via the di-muon trigger. We obtain the Bs and B0 yields by fitting the B mass distributions with binned likelihood functions. We correct these yields by the acceptances and use them to measure fsBR(Bs→J/ψφ)/fd BR(B0→J/ψ K*) that allows a determination of the BR(Bs→J/ψφ) and fs/fd.

The analysis is described in detail in Public Note 10795

Motivation

The purpose of this analysis is to measure the ratio of branching fractions of Bs→J/ψφ to B0→J/ψ K*. Moreover we can calculate the value of BR(Bs→J/ψφ) using the PDG value for BR(B0→J/ψ K*). The current PDG value (dominated by the prior CDF measurement) for the BR(Bs→J/ψφ) is (1.4 ± 0.5) × 10- 3.
In addition of the ratio of branching fraction, we do the measurement in several ranges of transverse momentum to explore the pT behavior of fs/fd. The precision of fs/fd is limited by the large uncertainty on the Bs branching ratio. However, since the branching ratio is a common multiplicative factor across all bins of pT (Bs) we can still study the pT behavior of fs/fd without good precision.


Procedure

The measurement starts with samples of reconstructed Bs→J/ψφ and B0→J/ψ K*, acquired via a di-muon sample (J/ψ→μμ). We apply a series of cuts to optimize the selection of both decays before applying a binned likelihood fit to the Bs→J/ψφ and B0→J/ψ K* invariant mass distributions to extract the Bs and B0 yields. In both channels we model the Bs and B0 signals by three Gaussian templates constructed from MC samples. We model the combinatorial backgrounds using exponential functions. In the Bs→J/ψφ sample, B0→J/ψ K* decays and Bs→J/ψ f0 decays are considered background. We constrain both contributions using data and simulation and we model them with Gaussian templates extracted from MC. The same technique is used to account and model the Bs→J/ψφ background and Bs→J/ψ f0 background in the B0→J/ψ K* sample. Finally, partially reconstructed B decays appear in the B0 sample. We model this background by an ARGUS function convoluted with a Gaussian. After application of a relative acceptance factor determined from simulation, we measure the quantity fsBR(Bs→J/ψφ)/fdBR(B0→J/ψ K*).
We classify the systematic uncertainties as either fit-related uncertainties or uncertainties on the relative acceptance. Fit uncertainties include combinatorial background modeling, modeling of the Bs and B0 signals and different background contributions. For the required relative acceptances, there are uncertainties in the modeling of the B hadron transverse momentum spectrum, B hadron lifetime, and polarization.


Results

Bs→J/ψφ and B0→J/ψ K* yields and ratio of yields:

pT range (GeV/c) Bs yield B0 yield Bs/B0
All 11110 ± 130 (stat.) ± 820 (sys) 57260 ± 340 (stat.) ± 830 (sys) 0.194 ± 0.003 (stat.) ± 0.015 (sys)
6 < p T <7.5 1600 ± 50 (stat.) ± 110 (sys) 7320 ± 150 (stat.) ± 180 (sys) 0.218 ± 0.009 (stat.) ± 0.016 (sys)
7.5 < p T <9.5 2500 ± 60 (stat.) ± 190 (sys) 12000 ± 160 (stat.) ± 430 (sys) 0.208 ± 0.006 (stat.) ± 0.017 (sys)
9.5 < p T <13 3340 ± 70 (stat.) ± 280 (sys) 17460 ± 180 (stat.) ± 210 (sys) 0.191 ± 0.004 (stat.) ± 0.016 (sys)
p T >13 3630 ± 70 (stat.) ± 290 (sys) 20410 ± 190 (stat.) ± 420 (sys) 0.178 ± 0.004 (stat.) ± 0.015 (sys)

fs/fd * BR(Bs→J/ψφ)/BR(B0→J/ψ K*):

pT range (GeV/c) fsBR(Bs→J/ψφ)/fdBR(B0→J/ψ K*)
All 0.239 ± 0.003 (stat.) ± 0.019 (sys)
6 < p T <7.5 0.232 ± 0.009 (stat.) ± 0.019 (sys)
7.5 < p T <9.5 0.238 ± 0.007 (stat.) ± 0.021 (sys)
9.5 < p T <13 0.233 ± 0.005 (stat.) ± 0.020 (sys)
p T >13 0.244 ± 0.005 (stat.) ± 0.021 (sys)


fs/fd * BR(Bs→J/ψφ):
Using the PDG value of BR(B0→J/ψ K*) = (1.33 ± 0.06) × 10- 3,

  • fs/fd * BR(Bs→J/ψφ): (3.17 ± 0.04 (stat.) ± 0.26 (sys) ± 0.14 (pdg) ) × 10- 4

  • BR(Bs→J/ψφ)/BR(B0→J/ψ K*) and BR(Bs→J/ψφ):
    The CDF value of fs/fdBR(Ds → φ π)) is combined with the actual PDG value for BR(Ds → φ π)) to extract fs/fd:
  • BR(Bs→J/ψφ)/BR(B0→J/ψ K*)= 0.89 ± 0.01 (stat.) ± 0.07 (sys) ± 0.11 (frag)
  • BR(Bs→J/ψφ)= (1.18 ± 0.02 (stat.) ± 0.09 (sys) ± 0.14 (frag) ± 0.05 (pdg) ) × 10- 3


  • fs/fd:
    Using the PDG value of BR(B0→J/ψ K*) = (1.33 ± 0.06) × 10- 3 and the most recent Belle BR(Bs→J/ψφ) = (1.25 ± 0.21) × 10- 3 :

    pT range (GeV/c) fs/fd
    All 0.254 ± 0.003 (stat.) ± 0.020 (sys) ± 0.044 (BR)
    6 < p T <7.5 0.247 ± 0.010 (stat.) ± 0.020 (sys) ± 0.043 (BR)
    7.5 < p T <9.5 0.253 ± 0.007 (stat.) ± 0.023 (sys) ± 0.044 (BR)
    9.5 < p T <13 0.248 ± 0.006 (stat.) ± 0.022 (sys) ± 0.043 (BR)
    p T >13 0.260 ± 0.006 (stat.) ± 0.022 (sys) ± 0.045 (BR)



    Figures

    The figures crucial to the analyses appear below: