An Updated Measurement of the CP Violating Phase βsJ/ψ φ with L=5.2 fb-1



Primary Authors: F. Azfar, J. Boudreau, N. Bousson, M. Feindt, J.P. Fernandez, K. Gibson, G.Giurgiu, G. Gomez- Ceballos, T. Kuhr, M. Kreps, L. Oakes, J. Morlock, M. Paulini, E. Pueschel, A.Schmidt


Introduction

This webpage summarizes the CDF preliminary result updating the flavor-tagged determination of CP violation in Bs → J/ψ φ decays corresponding to 5.2 fb-1 data. In addition to doubling the previous sample size, several improvements have been made to the analysis. These include fully incorporating potential contamination of the Bs → J/ψ φ signal by Bs → J/ψ KK or Bs → J/ψ f0(980) (where the non-resonant KK and f0(980) are S-wave states) into the likelihood fit, improvements to the NN selection and optimization, and full calibration of particle ID and flavour tagging. This analysis uses the updated and recalibrated Same Side Kaon Tagger.

This result is described in this public note
This result was first presented at FPCP2010, the slides are here
The slides from the Fermilab Wine and Cheese talk for this update are here

See also the previous results at First Flavor-Tagged Determination of Bounds on CP Violation in Bs → J/ψ φ Decays with L = 1.3 fb-1 published in PRL 100, 161802 (2008)
and Updated measurement of CP violating phase βsJ/ψ φ with 2.8 fb-1


Results

We present 68% and 95% confidence regions in the βs-ΔΓ plane:


Assuming the standard model predictions of βs and ΔΓ, the probability of a deviation as large as the level of the observed data is 44%, corresponding to 0.8 σ .

Additionally we present a confidence interval in βs, where ΔΓ is treated as a nuisance parameter, and find that βs is within [0.02,0.52] U [1.08,1.55] at the 68% CL and within [-π/2, -1.44] U [-0.13, 0.68] U [0.89, π/2] at the 95% CL.

Assuming no CP violation (βsJ/ψ φ = 0), we also measure the mean lifetime,the decay width difference, the transversity amplitudes and the strong phase φ :


s = 458.6 ± 7.6 (stat) ± 3.6 (syst) μm
ΔΓs = 0.075 ± 0.035 (stat) ± 0.01 (syst) ps-1
|A||(0)|2 = 0.231 ± 0.014 (stat) ± 0.015 (syst)
|A0(0)|2 = 0.524 ± 0.013 (stat) ± 0.015 (syst).
φ = 2.95 ± 0.64 (stat) ± 0.07 (syst)


These results give the world's most precise single measurement of the Bs lifetime and decay width difference. We find the S-wave contamination of the signal φ meson to be < 6.7% at the 95% CL

The likelihood

We provide the information for use of the CDF 35-dimensional likelihood in global fits and combinations in two formats. One is a statistics consisting in coverage-corrected profile-likelihood ratios as a function of βs and ΔΓ in which the likelihood has been profiled over the other 33 parameters. The other is the full multidimensional likelihood function, as a function implemented on a server, and its representation as a Markov chain in a root file. Instructions on how to use this information are here

Figures

Below are all figures meant for download.

    Likelihood contours
  • (eps), (gif): 2D βs - ΔΓ likelihood profile with full coverage adjustment
  • (eps), (gif): 1D βs likelihood profile with full coverage adjustment
  • (eps), (gif): 2D βs - ΔΓ likelihood profile with non-Gaussian error adjustment only (no systematics)
  • (eps), (gif): 1D βs likelihood profile with non-Gaussian error adjustment only (no systematics)
  • (eps), (gif): 2D βs - ΔΓ likelihood profile with no coverage adjustment
  • (eps), (gif): 1D βs likelihood profile with no coverage adjustment

    Coverage adjustment
  • (eps), (gif): Distribution of 1-CL for coverage adjustment of tagged 2D likelihood profile
  • (eps), (gif): Distribution of 1-CL for coverage adjustment of tagged 1D likelihood profile
  • (eps), (gif): Distribution of 1-CL for coverage adjustment of untagged 2D likelihood profile

    Selection, yield, angular efficiencies
  • (eps), (gif): Bs yield
  • (eps), (gif): Bs mass distribution with ct>60μm
  • (eps), (gif): KK invariant mass distribution
  • (eps), (gif): KK invariant mass distribution with ct>60μm
  • (eps), (gif): Bs NN training
  • (eps), (gif): NN signal efficiency
  • (eps), (gif): NN signal purity
  • (eps), (gif): NN output
  • (eps), (gif): NN sensitivity, toy MC with βs=0.0
  • (eps), (gif): NN sensitivity, toy MC with βs=0.3
  • (eps), (gif): NN sensitivity, toy MC with βs=0.5
  • (eps), (gif): NN significance curve
  • (eps), (gif): Angular efficiencies in re-weighted realistic MC
  • (eps), (gif): Starting background angular functions
  • (eps), (gif): Starting background angular functions overlaid with angular efficiencies

    Flavor tagging (see also SSKT)
  • (eps), (gif): B+ yield (for OST calibration)
  • (eps), (gif): OST measured vs. predicted dilution for B+ decays
  • (eps), (gif): OST measured vs. predicted dilution for B- decays

    S-wave studies
  • (eps), (gif): S-wave KK fraction of signal
  • (eps), (gif): 2D βs - ΔΓ likelihood profile comparison between fit including and without S-wave KK component (no coverage adjustment)
  • (eps), (gif): Bs mass fit with wide KK invariant mass window, to get fractions of signal, combinatorial bkg and B0 reflections for KK mass study
  • (eps), (gif): KK invariant mass with projections of signal φ, combinatorial bkg and B0 reflection fractions
  • (eps), (gif): KK invariant mass with projections of signal φ, combinatorial bkg and B0 reflection fractions and S-wave f0 component fixed to central value from full likelihood fit

    Fit cross checks
  • (eps), (gif): untagged fit: 2D βs - ΔΓ likelihood profile with full coverage adjustment
  • (eps), (gif): overlay of fully adjusted 2D βs - ΔΓ likelihood profile for tagged and untagged fits
  • (eps), (gif): untagged fit: 2D βs - ΔΓ likelihood profile with full coverage adjustment
  • (eps), (gif): untagged fit: 2D βs - ΔΓ likelihood profile with non Gaussian adjustment (no systematics)
  • (eps), (gif): untagged fit: 2D βs - ΔΓ likelihood profile with no adjustment
  • (eps), (gif): untagged fit: 1D βs - ΔΓ likelihood profile with no adjustment
  • (eps), (gif):tagged fit: 2D βs - ΔΓ likelihood profile for L=0-1.35fb-1
  • (eps), (gif):tagged fit: 2D βs - ΔΓ likelihood profile for L=1.35-2.8fb-1
  • (eps), (gif):tagged fit: 2D βs - ΔΓ likelihood profile for L=2.8-5.2fb-1

    Fit projections
  • (eps), (gif): Bs lifetime projection in signal region
  • (eps), (gif): Bs lifetime projection in sideband region
  • (eps), (gif): Bs lifetime error projection in signal region
  • (eps), (gif): Bs lifetime error projection in sideband region
  • (eps), (gif): Bs signal angular projection
  • (eps), (gif): Bs background angular projection