Search for a heavy neutral gauge boson in the dielectron channel with 5.4 fb-1 of ppbar collisions at sqrt(s) = 1.96 TeV

We report the results of a search for a heavy neutral gauge boson Z' decaying into the dielectron final state using data corresponding to an integrated luminosity of 5.4 fb-1 collected by the D0 experiment at the Fermilab Tevatron Collider. No significant excess above the standard model prediction is observed in the dielectron invariant-mass spectrum. We set 95% C.L. upper limits on \sigma (ppbar ->Z') X BR(Z' ->ee) depending on the dielectron invariant mass. These cross section limits are used to determine lower mass limits for Z' bosons in a variety of models with standard model couplings and variable strength.

S. Schlobohm, 80 C. Schwanenberger, 43 R. Schwienhorst, 62 J. Sekaric, 55 H. Severini, 73 E. Shabalina, 23 V. Shary, 18 A.A. Shchukin, 38 R.K. Shivpuri, 28 V. Simak, 10  We report the results of a search for a heavy neutral gauge boson Z ′ decaying into the dielectron final state using data corresponding to an integrated luminosity of 5.4 fb −1 collected by the D0 experiment at the Fermilab Tevatron Collider. No significant excess above the standard model prediction is observed in the dielectron invariant-mass spectrum. We set 95% C.L. upper limits on σ (pp → Z ′ ) × BR(Z ′ → ee) depending on the dielectron invariant mass. These cross section limits are used to determine lower mass limits for Z ′ bosons in a variety of models. For the sequential standard model Z ′ boson a lower mass limit of 1023 GeV is obtained. The gauge group structure of the standard model (SM), SU (3) C ⊗ SU (2) L ⊗ U (1) Y , can be extended with an additional U (1) group, which may arise in models derived from grand unified theories (GUT) that are based on groups with rank larger than four [1]. Additional U (1) groups can also arise from higher dimensional con- * with visitors from a Augustana College, Sioux Falls, SD, USA, b The University of Liverpool, Liverpool, UK, c SLAC, Menlo Park, CA, USA, d ICREA/IFAE, Barcelona, Spain, e Centro de Investigacion en Computacion -IPN, Mexico City, Mexico, f ECFM, Universidad Autonoma de Sinaloa, Culiacán, Mexico, and g Universität Bern, Bern, Switzerland. structions like string compactifications. In many models of GUT symmetry breaking, U (1) groups survive at relatively low energies, leading to corresponding neutral gauge bosons, commonly referred to as Z ′ bosons [2]. Such Z ′ bosons typically couple to SM fermions via the electroweak interaction, and can be observed at hadron colliders as narrow resonances through the process qq → Z ′ → e + e − . There is no simple general parametrization that can be applied to all the Z ′ models. Nevertheless, the models can be distinguished according to the strength of the gauge coupling, g Z ′ , for the additional U (1) group. The models with coupling of electroweak strength are called canonical. The sequential standard model (SSM)  [6].
Z ′ boson is a canonical example, where the SSM Z ′ boson (Z ′ SSM ) is defined to have the same couplings as the SM Z boson. The SSM Z ′ boson is often used as benchmark [2,3]. An additional example of a canonical model can be derived from the superstring inspired E 6 models [4]. The decomposition of E 6 can give rise to two additional U (1) factors through E 6 → SO(10) × U (1) ψ and SO(10) → SU (5) × U (1) χ . These groups are associated with the gauge fields Z ′ ψ and Z ′ χ that can mix and, at the TeV scale, can give rise to additional Z ′ bosons through the linear combination where 0 ≤ θ < π is a mixing angle [5]. The most commonly referenced Z ′ boson models arising from E 6 are summarized in Table I [6]. An example of a non-canonical model is the U (1) X Stueckelberg extension of the standard model (StSM) that gives rise to a very narrow Z ′ boson [7,8]. The Stueckelberg mechanism allows for the possibility of an Abelian gauge boson to gain mass without the requirement of a Higgs mechanism. The new parameters that are introduced in this model are the StSM mass mixing parameter, ǫ, and the Z ′ boson mass, M Z ′ . In the limit ǫ → 0, the Stueckelberg sector decouples from the SM [9].
In this Letter, we report on a search for a Z ′ boson decaying into an electron pair with the D0 detector at the Fermilab Tevatron Collider, where protons and antiprotons collide at √ s = 1.96 TeV. A Z ′ boson would appear as a narrow resonance in the ee invariant mass spectrum, with a natural width smaller than the resolution of the D0 electromagnetic calorimeter. A previous Tevatron search by the CDF collaboration [10], corresponding to 2.5 fb −1 of integrated luminosity, sets a lower mass limit on SSM Z ′ bosons of 963 GeV and reports a discrepancy over the expected SM background at M ee ∼ 240 GeV equivalent to 2.5 standard deviations. The CDF collaboration has also performed a search in the Z ′ → µµ channel [11], corresponding to 2.3 fb −1 of integrated luminosity, with 95% C.L. upper limits on σ (pp → Z ′ ) × BR(Z ′ → µµ) ranging from ∼50 fb to ∼3.2 fb for M Z ′ between 175 GeV and 1100 GeV.
The D0 detector [12] is composed of a central tracking system surrounded by a 2 T superconducting solenoidal magnet and a central preshower detector (CPS), a calorimeter, and a muon spectrometer. The central tracking system includes a silicon microstrip tracker (SMT) and a scintillating fiber tracker (CFT) that are designed to provide coverage for particles in the pseudorapidity range |η| < 3, where η = −ln [tan (θ/2)], and θ is the polar angle with respect to the proton beam direction. The azimuthal angle is denoted by φ. The CPS is located between the solenoid and the inner layer of the central calorimeter and is formed of approximately one radiation length of lead absorber followed by three layers of scintillating strips. The calorimeter consists of a central section (CC) covering |η| 1.1 and two end calorimeters (EC) that extend the EM coverage to η ≈ 4.1, with all three sections housed in separate cryostats [13]. Each section consists of an inner electromagnetic (EM) section, and an outer hadronic. The EM calorimeter is segmented into four longitudinal layers (EMi, i = 1, ..., 4) with transverse segmentation of ∆η × ∆φ = 0.1 × 0.1, except for the finely segmented third layer where it is 0.05 × 0.05. The muon system, covering |η| < 2, is located beyond the calorimeter and is composed of a layer of tracking detectors and scintillation trigger counters in front of 1.8 T iron toroidal magnets, and followed by two similar layers after the toroids. The luminosity is measured using plastic scintillator arrays in front of the end calorimeters. The data acquisition system includes a three-level trigger, designed to accommodate the high instantaneous luminosity. The data sample was collected between July 2002 and June 2009 using triggers requiring at least two clusters of energy deposits in the EM calorimeter and corresponds to an integrated luminosity of 5.4 ± 0.3 fb −1 [14].
The event selection requires two isolated electron candidates in the central section of the calorimeter. An electron candidate is characterized by an EM cluster with transverse momentum p T > 25 GeV and |η| < 1.1, reconstructed in a cone of radius R = (∆η) 2 + (∆φ) 2 = 0.4. At least 97% of the EM cluster energy must be deposited in the EM section of the calorimeter and its energy must be isolated in the calorimeter, where E tot (R) and E EM (R) are the total energy and the energy in the EM section, respectively, within a cone of radius R around the electron direction. In addition, the EM cluster is required to be consistent with an electron shower shape, using a χ 2 test and a neural network discriminant [15]. The EM cluster is required to be spatially matched to either a reconstructed track or a pattern of hits in the SMT and CFT consistent with the passage of an electron. The scalar sum of the p T of all tracks originating from the pp interaction vertex (PV) in an annulus of 0.05 < R < 0.4 around the cluster is required to be less than 2.5 GeV. Events are only considered if the PV lies within 60 cm of the geometrical center of the detector in the coordinate along the beam axis to be fully within the SMT acceptance. The two electron candidates are not required to have opposite charges to avoid losses due to charge misidentification. The data sample consists of 185,264 events that satisfy these selection criteria in the dielectron invariant mass control region 60 < M ee < 150 GeV and 1332 events in the search region M ee > 150 GeV.
Signal and SM background events are generated using pythia [16] with the CTEQ6L1 [17] parametrization of the parton distribution functions (PDFs), and processed through the D0 detector simulation based on geant3 [18] adding zero bias events, and the same reconstruction software as the data. Signal templates based on the SSM Z ′ boson have been generated up to masses of 1100 GeV. The width of the resonance scales with the Z ′ boson mass, according to Γ Z ′ = Γ Z × M Z ′ /M Z , where M Z and Γ Z are the mass and width of the Z boson. For M Z ′ ≥ 2m t the decay channel to top quarks opens up, thus increasing the width of the resonance. The signal selection efficiency increases from ∼22% to ∼44% for M Z ′ between 175 and 1100 GeV independent of the type of Z ′ boson discussed in this Letter.
The dominant irreducible background is due to the Drell-Yan (DY) process. A mass-dependent k-factor [19] has been applied to the pythia dielectron invariant mass spectrum to account for next-to-next-to-leading order (NNLO) contributions. The main instrumental background originates from the misidentification of one or two jets as electrons. The shape of the invariant mass spectrum for this background is obtained from data by selecting events where the EM clusters fail the χ 2 test. Other SM backgrounds include Z/γ * → τ τ , W +γ, W W , ZZ, W Z, W + jets, tt, and γγ production. The contribution of these background processes is small (∼0.6%) and is estimated using pythia corrected for higher order contributions [20][21][22].
The normalization of the various background contributions is determined by fitting the invariant mass spectrum of the data to a superposition of the backgrounds in a control region around the Z boson mass (60 < M ee < 150 GeV), where the existence of Z ′ bosons has been excluded by previous searches [23]. The total number of background events in that region is fixed to the number of events that have been observed in the data. The relative contribution from the DY process and instrumental background is a free parameter, while the contribution from the other SM processes is normalized to their theoretical cross sections. The uncertainty of the background normalization is estimated by varying both the criteria to select the instrumental background sample and the fitting range, and is 2%.
Having normalized the various background contributions to data in the control region, the background shapes are used to extrapolate to higher invariant masses. The measured ee invariant mass spectrum, superimposed on the expected backgrounds for the full mass range studied, is shown in Fig. 1. The data and expected background are generally in good agreement for the full invariant mass range studied, with a χ 2 over degrees of freedom equal to 118.5/113. In the absence of a heavy resonance signal, the ee invariant mass distribution is used to calculate an upper limit on the production cross section of Z ′ bosons multiplied by the branching ratio into the ee final state, using a Poisson log-likelihood ratio (LLR) test statistics [24]. The expected limits are calculated using the median of the LLR distribution for a background-only hypothesis. The observed limit, obtained including all the fluctuations present in the data, is expected to be contained in the ±1 and ±2 standard deviations region with a probability of 68% and 95%, respectively. An observed limit significantly outside the expected range would indicate either a poor modeling of the background or that the data is inconsistent with the background-only hypothesis.
The following systematic uncertainties on the expected background and the signal have been considered for the limit calculation. The uncertainties affecting the expected background include the electron identification efficiency (3.0% per electron), the mass dependence of the DY associated NNLO k-factor (5.0%), and the background normalization (2.0%). Uncertainties that affect the signal include the integrated luminosity (6.1%), the PDFs for signal acceptance (0.4% -7.6%), the electron identification efficiency (3.0% per electron), the EM cluster energy resolution (6.0%), and the trigger efficiency (0.1%). For the EM energy resolution and the background normalization, both the effects on the normalization and on the shape of the invariant mass distribution have been considered in extracting limits. For the remaining systematic sources only the changes to the overall background normalization or signal detection efficiency have been considered. The systematic uncertainties are included via convolution of the Poisson probability distributions for signal and background with Gaussian distributions corresponding to the different sources of systematic uncertainties taking into account all relevant correlations between systematics' sources.
The observed upper limits on the production cross section multiplied by the branching ratio into an ee pair for the process pp → Z ′ → ee are given in Table II  observed dielectron invariant mass spectrum arises only from the backgrounds considered in the analysis. Figure 2 shows these limits together with the ±1 and ±2 standard deviation bands on the expected limit, and the cross section predictions for SSM and E 6 Z ′ bosons [6] where a constant k-factor of 1.3 [25] has been applied to the pythia cross section. Since this analysis searches for a resonance instead of an enhancement in the total cross section, signal cross section predictions are calculated by integrating over the region [M Z ′ − 10 × Γ Z ′ , ∞], where Γ Z ′ is the width of the SSM Z ′ boson, thus excluding Z ′ boson events which do not contribute to the resonant region. For M Z ′ < 500 GeV the difference between the cross section in the region defined above and the total cross section is less than 5%, while for a M Z ′ = 1 TeV SSM Z ′ boson it is ∼40%. The mass limits on the specific models of Z ′ bosons considered are given in Table III. These limits can be translated into upper limits on the U (1) Z ′ gauge coupling, g Z ′ [6], as a function of M Z ′ . Figure 3 illustrates the observed upper limits on g Z ′ /g Z ′ χ [26] for the Z ′ χ model. Cross sections are calculated as a function of Z ′ boson mass to interpret the observed upper limits on σ (pp → Z ′ )×BR (Z ′ → ee) as mass limits for a StSM Z ′ boson. Figure 4 shows the observed and expected limits and the cross section predictions for the StSM Z ′ boson for several ǫ values from 0.02 to 0.06 [9]. The mass limits are summarized in Table III. In summary, we have searched for a heavy narrow resonance in the ee invariant mass spectra, using 5.4 fb −1 of integrated luminosity collected with the D0 detector at the Fermilab Tevatron Collider. The observed spectrum agrees with the total background expected from SM processes and instrumental backgrounds. No evidence for physics beyond the SM is observed. For a Z ′ boson   with SM couplings and with intrinsic width significantly smaller than the detector resolution, we set 95% C.L. upper limits on σ (pp → Z ′ ) × BR (Z ′ → ee) between 22 fb and 1.9 fb for M Z ′ between 175 GeV and 1100 GeV.