Search for the Higgs boson in H->WW(*) decays in p anti-p collisions at S**(1/2) = 1.96-TEV

We present a search for the standard model Higgs boson in H->WW(*), decays with ee, emu and mumu final states in p anti-p collisions at a center-of-mass energy of S**(1/2) = 1.96-TEV. The data, collected from April 2002 to June 2004 with the D0 detector, correspond to an integrated luminosity of 300 - 325 pb**(-1), depending on the final state. The number of events observed is consistent with the expectation from backgrounds. Limits from the combination of all three channels on the Higgs production cross section times branching ratio sigma*BR(H->WW(*) are presented.

metry breaking (EWSB) and the mass generation of electroweak gauge bosons and fermions. Spontaneous EWSB predicts the existence of this neutral scalar particle with mass M H , a free parameter in the SM. Direct searches at the cern e + e − collider (lep) yield a lower limit for the Higgs boson mass of M H > 114. 4 GeV [1] at the 95% CL. Indirect measurements via fits to the electroweak precision data give an upper bound of M H < 280 GeV [2] at the 95% CL.
In this Letter, we present a search for the Higgs boson in H → W W ( * ) → ℓν ℓ ′ ν ′ (ℓ, ℓ ′ = e, µ, τ ) decays with e + e − , e ± µ ∓ , or µ + µ − final states. Tau decays are detected in their leptonic decay modes to electrons or muons. This is the first search for the Higgs boson at a hadron collider in this decay channel, which plays an important role in the overall discovery potential of the Higgs boson at the Fermilab Tevatron Collider [3]. We use data collected by the DØ detector between April 2002 andJune 2004 in pp collisions at √ s = 1.96 TeV of the Fermilab Tevatron Collider. The integrated luminosities are 325 ± 21 pb −1 , 318 ± 21 pb −1 and 299 ± 19 pb −1 for the e + e − , e ± µ ∓ and µ + µ − channels, respectively. The differences in the integrated luminosities for various channels are primarily due to different trigger conditions. Next-toleading order (NLO) calculations [4,5] predict the product of the SM Higgs boson production cross section and the branching ratio σ(pp → H) × BR(H → W W ( * ) ) of 11-250 fb for the Higgs masses between 100 and 200 GeV. The dominant contribution to the cross section comes from gluon-gluon fusion. Extensions of the SM including a fourth fermion family [6] predict an enhanced Higgs boson production cross section.
We briefly describe the main components of the DØ Run II detector [7] important to this analysis. The central tracking system consists of a silicon microstrip tracker (SMT) and a central fiber tracker (CFT), both located within a 2.0 T axial magnetic field. The SMT strips have a typical pitch of 50-80 µm, and the design is optimized for tracking and vertexing over the pseudorapidity range |η| < 3, where η = − ln (tan θ 2 ), where θ is the polar angle with respect to the proton beam. The system has a six-barrel longitudinal structure, each with a set of four silicon layers arranged axially around the beam pipe, interspersed with sixteen radial disks. The CFT with full coverage for |η| < 1.6 has eight thin coaxial barrels, each supporting two doublets of overlapping scintillating fibers, one doublet parallel to the beam axis, the other alternating by ±3 • relative to the beam axis.
A liquid-argon/uranium calorimeter surrounds the central tracking system and consists of a central calorimeter (CC) covering to |η| ≈ 1.1, and two end cap calorimeters (EC) extending coverage to |η| < 4.2, all housed in separate cryostats [8]. Scintillators between the CC and EC cryostats provide additional sampling of showers for 1.1 < |η| < 1.4.
The muon system is located outside the calorimeters and consists of a layer of tracking detectors and scintillation trigger counters inside iron toroid magnets which provide a 1.8 T magnetic field, followed by two similar layers behind each toroid. Tracking in the muon system for |η| < 1 relies on 10 cm wide drift tubes [8], while 1 cm mini-drift tubes are used for 1 < |η| < 2 [9]. The H → W W ( * ) candidates are selected by single or di-lepton triggers using a three-level trigger system. The first trigger level uses hardware to select electron candidates based on energy deposition in the electromagnetic part of the calorimeter and selects muon candidates formed by hits in two layers of the muon scintillator system. Digital signal processors at the second trigger level form muon track candidate segments defined by hits in the muon drift chambers and scintillators. At the third level, software algorithms running on a computing farm and exploiting the full event information are used to make the final selection of events which are recorded for offline analysis.
In the offline analysis, electrons are identified as electromagnetic showers in the calorimeter. These showers are selected by comparing the longitudinal and transverse shower profiles to those expected of the electrons. The showers must be isolated, deposit most of their energy in the electromagnetic part of the calorimeter, and pass a likelihood criterion that includes a spatial track match and, in the CC region, an E/p requirement, where E is the energy of the calorimeter cluster and p is the momentum of the track. All electrons are required to be in the pseudorapidity range |η| < 3.0. The transverse momentum measurement of the electrons is based on calorimeter cell and track information.
Muon tracks are reconstructed from hits in the wire chambers and scintillators in the muon system and must match a track in the central tracker. To select isolated muons, the scalar sum of the transverse momentum of all tracks other than that of the muon in a cone of R = 0.5 around the muon track must be less than 4 GeV, where R = (∆φ) 2 + (∆η) 2 and φ is the azimuthal angle. Muon detection is performed over the full coverage of the muon system |η| < 2.0. Muons from cosmic rays are rejected by requiring a timing criterion on the hits in the scintillator layers as well as applying restrictions on the position of the muon track with respect to the primary vertex.
The decay of the two W bosons into electrons or muons results in three different final states e + e − + X (ee channel), e ± µ ∓ +X (eµ channel) and µ + µ − +X (µµ channel), each of which consists of two oppositely charged isolated leptons with high transverse momentum and large missing transverse energy, E / T , due to the undetected neutrinos. The selection criteria for each channel were chosen to minimize the cross section upper limit on Higgs production expected in the absence of signal. To take into account the signal kinematic characteristics that change with the Higgs boson mass, M H , some selection cuts are M H dependent [10]. Six Higgs boson masses from 100 GeV to 200 GeV have been studied.
In all three channels, two leptons originating from the same vertex are required to be of opposite charge, and must have transverse momenta p T > 15 GeV for the leading lepton and p T > 10 GeV for the trailing one (Cut 1). Figure 1 shows the good agreement between data and Monte Carlo simulation (MC) in distributions of the azimuthal opening angle ∆φ ℓℓ ′ between the two leptons for the ee (a), the µµ (c) and the eµ channel (e) after applying the lepton transverse momentum cuts.
In all cases, the background is largely dominated by Z/γ * production which is further suppressed by requiring E / T > 20 GeV in all three channels (Cut 2). Background events are also removed if the E / T has a large contribution from the mis-measurement of jet energy. The fluctuation in the measurement of jet energy in the transverse plane can be approximated by ∆E jet · sin θ jet where ∆E jet is proportional to √ E jet . The opening angle ∆φ (jet, E / T ) between this projected energy fluctuation and the missing transverse momentum provides a measure of the contribution of the jet to the missing transverse energy. The scaled missing transverse energy defined as is required to be greater than 15 (Cut 3). The charged lepton system and the neutrinos are emitted mostly back-to-back, so the invariant mass for the leptons from the Higgs decay is restricted to M H /2. Thus, the invariant mass m ℓℓ is required to be m ℓℓ < M H /2 (Cut 4). In the ee channel the cut is altered to m ee < min(80 GeV, M H /2). In the µµ channel a lower cut boundary with m µµ > 20 GeV is required to remove events from J/ψ, Υ and Z/γ * production. The sum of the p T of the leptons and E / T is required to be in the range M H /2 + 20 GeV < p ℓ1 T + p ℓ2 T + E / T < M H for the ee and eµ channel and M H /2 + 10 GeV < p ℓ1 T + p ℓ2 T + E / T < M H for the µµ channel (Cut 5). The transverse mass, defined , with the dilepton transverse momentum p ℓℓ ′ T , should be in the range M H /2 < m ℓℓ ′ T < M H − 10 GeV (Cut 6). The latter two cuts reject events from W +jet/γ and W W production and further reduce backgrounds from Z/γ * production. Finally, to suppress the background from tt production, the scalar sum of the transverse energies of all jets with E jet T > 20 GeV and |η| < 2.5, H T , is required to be less than 100 GeV (Cut 7). Remaining Z boson and multi-jet events can be rejected with a cut on the opening angle, ∆φ ℓℓ ′ < 2.0 (Cut 8), since most of the backgrounds exhibit a back-to-back topology. This is not the case for Higgs boson decays because of the spin correlations in the decay. Figure 1 shows the distributions of the azimuthal opening angle ∆φ ℓℓ ′ between the two leptons for the ee (b), the µµ (d) and the eµ channel (f) before applying the final cut on ∆φ ℓℓ ′ .
To maximize the sensitivity, the selection in the µµ channel is slightly changed for Higgs boson masses M H = 140 and 160 GeV. For a better Z/γ * background suppression cuts 4, 5 and 6 are replaced by the following cuts: the invariant mass m µµ should be in the range 20 GeV < m µµ < 80 GeV (Cut 4). Since the momentum resolution is degraded for high p T tracks, an additional constrained fit is performed to reject events compatible with Z boson production (Cut 5). The sum of the muon transverse momenta and the missing transverse energy should be p µ1 T + p µ2 T + E / T > 90 GeV (Cut 6). The efficiency for H → W W ( * ) → ℓν ℓ ′ ν ′ signal events to pass the acceptance and selection criteria is determined using the pythia 6.2 [11] event generator followed by a detailed geant-based [12] simulation of the DØ detector. All trigger, reconstruction and identification effi-  [4] and the branching ratio BR of 0.1068 ± 0.0012 for W → ℓν [13], the expected number of events for H → W W ( * ) decays from all three channels is 0.68 ± ±0.03 (syst) ± 0.04 (lum) events for a Higgs boson mass M H = 160 GeV. The signal expectation for different Higgs masses M H are given in Table II. Background contributions from Z/γ * , W +jet/γ, tt, W W , W Z and ZZ events are estimated using the pythia event generator normalized to their NLO cross sections [14]. In addition, W +jet/γ contributions are verified using alpgen [15]. All events are processed through the full detector simulation. The background due to multi-jet production, when a jet is misidentified as an electron, is determined from the data using a sample of like-sign di-lepton events with inverted lepton quality cuts (called "QCD" in Fig. 1). A summary of the background contributions together with signal expectations and events observed in the data after the final selection is shown in Table II. There is good agreement between the number of events observed in the data and the various backgrounds in all three channels. The largest difference between the data and the background expectation, at M H = 120 GeV, corresponds to a background probability of 6%. The eµ channel has both the highest signal efficiency and best signal-to-background ratio.
Various sources of systematic uncertainties that affect the background estimation and signal efficiencies have been studied. In these calculations, parameters are varied within ±1σ of their nominal values, where σ is determined by the corresponding uncertainties. The trigger efficiency, electron and muon identification efficiencies, jet energy scale, electron and muon momentum resolution, parton distribution function uncertainty and cross section calculation of Z/γ * , W W and tt events contribute to the systematic uncertainties. The total systematic uncertainties for the background estimate and signal effi- ) at the 95% CL together with expectations from standard model Higgs boson production and an alternative model. The lep limit on the standard model Higgs boson production is taken from [1] and the 4th generation model prediction is described in [6].
ciencies for the six Higgs boson masses are given in Table III. The largest contribution to the systematic uncertainty on the background for small Higgs boson masses comes from the jet energy scale due to the large W +jet/γ background, whereas for high Higgs boson masses the W boson pair production cross section gives the largest systematic uncertainty. The uncertainty of the parton distribution function is the largest uncertainty of the signal efficiency. The uncertainty of the luminosity measurement is 6.5%. Since the remaining candidate events after the selection are consistent with the background expectation, limits on the production cross section times branching ratio σ × BR(H → W W ( * ) ) are derived using a modified frequentist method described in Ref. [16]. It provides the confidence level for the background to represent the data, CL B , and the confidence level for the sum of signal and background hypothesis CL S+B . The 95% CL limit is obtained by requiring CL S+B /CL B = 0.05. The uncertainty on the background and the expected signal events were determined from the statistical and systematic uncertainties and luminosity uncertainty. Table IV shows the individual expected and observed upper limits on the cross section times branching ratio σ×BR(H → W W ( * ) ) for the combination of the three different decay channels for six different Higgs boson masses. The different values of the upper limits are due to different background expectations and signal efficiencies. The best limits are achieved for large Higgs masses since background expectations decrease while signal efficiencies increase. Figure 2 shows the expected and observed cross section limits for σ × BR(H → W W ( * ) ) for the different Higgs boson masses compared with predictions from the SM and from an extension including a fourth fermion fam-   ily [6]. With the current dataset, no region of the SM prediction can be excluded.
To conclude, we have searched for the Higgs boson in H → W W ( * ) → ℓν ℓ ′ ν ′ (ℓ, ℓ ′ = e, µ, τ ) decays with e + e − , e ± µ ∓ or µ + µ − final states in pp collisions at √ s = 1.96 TeV. The data is consistent with the expectation from backgrounds. Since no excess has been observed, limits on the production cross section times branching ratio σ × BR(H → W W ( * ) ) have been derived.