Elsevier

Physics Letters B

Volume 465, Issues 1–4, 21 October 1999, Pages 349-362
Physics Letters B

A direct measurement of |Vcs| in hadronic W decays using a charm tag

https://doi.org/10.1016/S0370-2693(99)01088-6Get rights and content

Abstract

The inclusive charm production rate in W decays is measured from a study of the properties of final state particles. The sample of W pairs is selected from 67.7 pb−1 collected by ALEPH in 1996 and 1997 at centre-of-mass energies near 172 and 183 GeV in the channels W+W→4q and W+W→ℓνqq̄. The branching fraction of hadronic W decays to a final state containing a c quark, RWc= Γ(W→cX)/Γ(W→hadrons), is measured to be 0.51±0.05stat±0.03syst. This allows a direct determination of the CKM matrix element |Vcs|=1.00±0.11stat±0.07syst.

Introduction

The data collected by ALEPH in the years 1996 and 1997, at average centre-of-mass energies of 172 GeV and 183 GeV, respectively, are used to study the properties of hadronic W boson decays. The total luminosity collected in the two periods is 67.7 pb−1 of which 57.0 pb−1 was collected at the higher energy. The identification of charm jets in these decays leads to a direct measurement of the fraction RWc=Γ(W→cX)/Γ(W→hadrons), where X stands for d̄,s̄orb̄. Unless explicitly stated, the charge-conjugate modes are always implied throughout this paper. Within the Standard Model, this branching ratio can be expressed as a function of the different Cabibbo-Kobayashi-Maskawa (CKM) matrix elements through the following relation:RWc=|Vcd|2+|Vcs|2+|Vcb|2|Vud|2+|Vus|2+|Vub|2+|Vcd|2+|Vcs|2+|Vcb|2.Assuming the unitarity of the CKM matrix, RWc is expected to be equal to 0.5. Therefore, a measurement of RWc is a direct test of this assumption. Furthermore, Eq. (1)can be used to extract a value of the least well known CKM matrix element |Vcs| which is currently measured to be 1.01±0.18 using D→Kℓν decays [1].

A charm jet tagger (called NNcin the following) was developed to identify W→cX decays. This is based on a neural network with 12 variables as input and uses mainly information from charm lifetime, jet-shape properties, reconstruction of D mesons and lepton identification. The procedure of extracting RWc and |Vcs|from the NNcoutput is described in this paper. Another analysis based on a Fisher Discriminant technique is also presented as a crosscheck of the method.

Section snippets

The ALEPH detector

The ALEPH detector [2] and its performance [3] are described in detail elsewhere. Only a brief account of the parts of the apparatus relevant for this analysis is given here. Charged particles are detected over the range |cosθ|<0.95 by an inner drift chamber and a large time projection chamber (TPC), complemented by a silicon strip vertex detector (VDET) made of two layers each providing measurements in the and rz coordinates, with a resolution of 12 μm in and 12 μm in rz, for tracks at

Monte Carlo samples

The value of RWc is extracted by comparing the NNcoutput distribution in the data to the corresponding Monte Carlo distribution, where generated events are processed through a full simulation of the ALEPH detector response and through the same reconstruction chain. The KORALW [5] event generator with the complete set of WW-like four-fermion diagrams was used to produce two samples of 200,000 WW events at centre-of-mass energies of 172 GeV and 183 GeV, generated with reference W masses MWref of

Event selection

The event selection was performed separately for the purely hadronic events, W+W→4q, and for the semileptonic events, W+W→ℓνqq̄(ℓ=e,μ).

For the semileptonic events W+W→eνqq̄and W+W→μνqq̄, the neural network described in Ref. [11] is used for the selection. The signal efficiency is about 90% for a purity of 95%. The selected events are then forced into two jets using the DURHAM-P [12] algorithm, the lepton being removed when doing the clusterisation. Events corresponding to W+W→τνqq̄decays

Result

To extract RWc, a binned maximum likelihood fit is performed to the shape of the output distribution of NNcmax. In each bin k of this distribution, the expected number of Monte Carlo events is defined asNkMC=RWcNWProbW→cX(k)+(1−RWc)NWProbW↛cX(k)+NbkgProbbkg(k)where ProbW→cX(k), ProbW↛cX(k) and Probbkg(k) are the probability density functions, determined from Monte Carlo, in the bin k of the NNcmaxdistribution for the W→cX, W↛cXand non-WW background events, respectively. The total number NW of

Studies of systematic uncertainties

The following sources of systematic errors are considered:

  • 1.

    An uncertainty of 5% is applied to the normalization of the qq̄(γ) background as estimated in Ref. [22].

  • 2.

    The error resulting from the particular choice of QCD Monte Carlo generator is estimated by replacing JETSET by HERWIG for both the qq̄(γ) background and the W→qq̄ decays, and the full difference between the two results is taken as the systematic error. Since HERWIG is known to provide a very approximate representation of exclusive

Checks of the analysis

The consistency of the NNcmaxdistribution in data and Monte Carlo is tested by means of a Kolmogorov test. This gives a consistency confidence level of 98.1% if the fitted value of RWcis used in the simulation, while this confidence level becomes smaller than 1% if RWcis set to zero.

The number of fully reconstructed D mesons in the most charm-like jet of a pair is 21 in the data; 22.6 are expected in the simulation using the fitted value of RWc, while only 9.8 would be expected without W→cX

Conclusion

Using a charm tag based on the properties of jets produced in W decays, the inclusive charm production rate in W decays, RWc, is measured. The analysis of the 172–183 GeV data collected by ALEPH in 1996 and 1997 leads to the valueRWc=Γ(W→cX)/Γ(W→hadrons)=0.51±0.05stat±0.03syst,from which the value|Vcs|=1.00±0.11stat±0.07systis derived. The measured value of RWcis in agreement with the Standard Model expectation of 0.5 assuming the unitarity of the CKM matrix and with the result obtained by the

Acknowledgements

We are indebted to our colleagues of the accelerator divisions for the outstanding performance of the LEP accelerator. Thanks are also due to the many engineering and technical personnel at CERN and at the home institutes for their contributions toward the success of ALEPH. Those of us not from member states wish to thank CERN for its hospitality.

References (27)

  • M. Skrzypek et al.

    Comp. Phys. Commun.

    (1996)
  • T. Sjöstrand

    Comp. Phys. Commun.

    (1994)
  • G. Marchesini

    Comp. Phys. Commun.

    (1992)
  • L. Bellantoni

    Nucl. Instr. and Meth. A

    (1991)
  • Particle Data Group, Review of Particle Physics, Phys. Rev. D 54 (1996)...
  • Nucl. Instr. and Meth. A

    (1990)
  • Nucl. Instr. and Meth. A

    (1995)
  • Phys. Lett. B

    (1997)
  • Phys. Lett. B

    (1991)
  • Phys. Rev. Lett.

    (1979)
  • Phys. Rev. Lett.

    (1985)
  • Phys. Lett. B

    (1998)
  • Yu.L. Dokshitzer

    J. Phys. G

    (1991)
  • Cited by (12)

    • Review of particle physics

      2004, Physics Letters, Section B: Nuclear, Elementary Particle and High-Energy Physics
    • The determination of the CKM matrix

      2001, Nuclear Instruments and Methods in Physics Research, Section A: Accelerators, Spectrometers, Detectors and Associated Equipment
    • W properties and decays at LEP2

      2001, Nuclear Physics B - Proceedings Supplements
    • A measurement of the rate of charm production in W decays

      2000, Physics Letters, Section B: Nuclear, Elementary Particle and High-Energy Physics
    View all citing articles on Scopus
    2

    Supported by CICYT, Spain.

    1

    Permanent address: Universitat de Barcelona, 08208 Barcelona, Spain.

    View full text