Search for Pair Production of Second Generation Scalar Leptoquarks in ppbar Collisions at sqrt(s)=1.96 TeV

We report on a search for the pair production of second generation scalar leptoquarks (LQ_2) in ppbar collisions at the center-of-mass energy sqrt(s) = 1.96 TeV, using data corresponding to an integrated luminosity of 294 +- 19 pb^-1 recorded with the D0 detector. No evidence for a leptoquark signal in the LQ_2 LQ_2 ->muq muq channel has been observed, and upper bounds on the product of cross section times branching fraction were set. This yields lower mass limits of m(LQ_2)>247 GeV for beta = Br(LQ_2->muq) = 1 and m(LQ_2)>182 GeV for beta = 1/2. Combining these limits with previous D0 results, the lower limits on the mass of a second generation scalar leptoquark are m(LQ_2)>251 GeV and m(LQ_2)>204 GeV for beta = 1 and beta = 1/2, respectively.

We report on a search for the pair production of second generation scalar leptoquarks (LQ2) in pp collisions at the center-of-mass energy √ s = 1.96 TeV, using data corresponding to an integrated luminosity of 294 ± 19 pb −1 recorded with the DØ detector. No evidence for a leptoquark signal in the LQ2LQ 2 → µqµq channel has been observed, and upper bounds on the product of cross section times branching fraction were set. This yields lower mass limits of mLQ 2 > 247 GeV for β = B(LQ2 → µq) = 1 and mLQ 2 > 182 GeV for β = 1/2. Combining these limits with previous DØ results, the lower limits on the mass of a second generation scalar leptoquark are mLQ 2 > 251 GeV and mLQ 2 > 204 GeV for β = 1 and β = 1/2, respectively.
PACS numbers: 13.85.Rm Leptoquarks, colored bosons which carry both lepton (l) and quark (q) quantum numbers and third-integer electric charge, appear in several extensions of the standard model of particle physics [1]. Leptoquarks could, in principle, decay into any combination of a lepton and a quark. Experimental limits on lepton number violation, on flavor-changing neutral currents, and on proton decay, however, motivate the assumption that there would be three different generations of leptoquarks. Each of these leptoquark generations couples to only one generation of quarks and leptons, and, therefore, conserves the corresponding lepton and quark family numbers [2]. As a consequence, leptoquark masses could be as low as O(100 GeV), allowing the production of leptoquarks in reach of present collider experiments.
At the Tevatron collider, leptoquarks would be produced in pairs, primarily through qq annihilation and gluon fusion. These production mechanisms would be independent of the unknown coupling λ between the leptoquark, the lepton, and the quark.
The DØ Run II detector [5] is composed of several layered elements. Nearest the beam is a central tracking system consisting of a silicon microstrip tracker (SMT) and a central fiber tracker (CFT), both located within a 2 T superconducting solenoidal magnet. Muon momenta are measured from the curvature of muon tracks in the central tracking system. Jets are reconstructed from energy depositions in the three liquid-argon/uranium calorimeters outside the tracking system: a central section (CC) covering up to |η| ≈ 1.1 and two end calorimeters (EC) extending coverage to |η| ≈ 4, all housed in separate cryostats, where η = − ln tan θ 2 denotes the pseudorapidity and θ is the polar angle with respect to the proton beam direction. Scintillators located between the CC and EC cryostats provide sampling of hadron showers for 1.1 < |η| < 1.4. A muon system beyond the calorimeters consists of a layer of drift-tube tracking detectors and scintillation trigger counters before 1.8 T iron toroids, followed by two additional similar layers after the toroids [6].
The data used in this analysis were collected during Run II of the Fermilab Tevatron collider between August 2002 and July 2004 and correspond to an integrated luminosity of 294 ± 19 pb −1 . The sample of candidate events used in this search was collected with a set of triggers that required either one or two muon candidates in the muon system. The trigger efficiency for the µjµj events considered in this analysis was measured to be (89 ± 3)%.
Muons in the region |η| < 1.9 were reconstructed offline from hits in the three layers of the muon system which were matched to isolated tracks in the central tracking system to remove the background from heavy-quark production. This muon isolation was assured by requiring the sum of the transverse momenta of all other tracks in a ∆R = (∆φ) 2 + (∆η) 2 < 0.5 cone around the muon to be smaller than 4 GeV, where φ is the azimuthal angle around the direction of the incident beam. Cosmic ray muons were rejected by cuts on the timing in the muon scintillators and by removing back-to-back muons. Jets were reconstructed using the iterative, midpoint cone algorithm [7] with a cone size of ∆R = 0.5. The jet energies were calibrated as a function of the jet transverse energy and η by balancing the transverse energy in photon plus jet events. Requiring |η| < 2.4 for all jets removes the QCD background from events with jets at very small angles to the beam direction and, therefore, with large cross sections.
The background is dominated by the Drell-Yan (DY) events in the channel Z/γ * → µµ (+jets). QCD multijet events faking muons are suppressed by the isolation requirement and the thick shielding of the muon detectors. To evaluate the contribution from DY background, samples of Monte Carlo (MC) events were generated with pythia [8]. The number of pythia events was normalized to yield the predicted next-to-next-to-leading order (NNLO) cross section [9] at the Z-boson resonance. The events were furthermore reweighted as a function of the dimuon mass in order to describe the NNLO prediction for the differential cross section dσ/dm µµ [9]. An additional sample, generated with alpgen [10] and based on a matrix-element calculation for Zjj, was used to test systematic uncertainties due to the shape of the jet transverse energy distribution. Samples of pythia tt (m t = 175 GeV) and W W samples were used to estimate the background contributions from top quark and W boson pair production. The signal efficiencies were calculated using samples of LQ 2 LQ 2 → µqµq events simulated with pythia for leptoquark masses from 140 GeV to 300 GeV in steps of 20 GeV. All Monte Carlo events were generated using CTEQ5L [11] parton distribution functions (PDFs) and processed using a full simulation of the DØ detector based on geant [12] and the DØ event reconstruction [5].
Offline, events were required to have two muons with transverse momenta p T exceeding 15 GeV and at least two jets with transverse energies E T greater than 25 GeV. The momentum resolution degrades with increasing p T , and hence the resolution on the dimuon mass m(µµ) with increasing m(µµ). Therefore, in order to reduce the DY background at high m(µµ) and to account for muon tracks with large momentum uncertainty, corrections were applied to the muon momenta by taking advantage of the fact that no missing transverse energy is expected in either signal or DY events. The missing transverse energy E / T was estimated from the transverse energy balance of all muons and jets (E T > 20 GeV) in the event. The momentum of the muon most opposite to the E / T direction in the r-φ plane (i.e. in the plane perpendicular to the incident beam) was rescaled such that the component of the missing transverse energy parallel to the muon vanished. This correction suppressed the contribution from Z boson events misreconstructed in the high mass region where the search for leptoquarks took place. To further reduce the background from DY events a Z boson veto cut (dimuon mass m(µµ) > 105 GeV) was applied. Six events survive this last cut, while 6.8 ± 2.0 events are expected from standard model backgrounds, which mainly consists of DY (6.1 ± 2.0) and tt (0.69 ± 0.07).
The remaining events after the Z boson veto cut were arranged in four bins. Second generation leptoquark events are expected to have both high dimuon masses and large values of S T , which is the scalar sum of the transverse energies of the two highest-p T muons and the two highest-E T jets in the event, as can be seen in Fig.  1b) for a leptoquark mass of 240 GeV. The separation between bin i and bin i − 1, i ∈ {1, 2, 3}, is defined as: This binning, which effectively results in bins in the order of increasing S/B, is illustrated by the curved lines in Fig. 1 for the expected standard model backgrounds, an example LQ 2 signal, and for the data. The number of events in the four signal bins is shown in Fig. 2. Table I summarizes the efficiencies for various leptoquark masses, as well as the numbers of expected background events and the distribution of the data in the four signal bins. The signal efficiency increases with mass, because for larger leptoquark masses, the decay products have larger momenta yielding events with larger S T . The dominant uncertainty on the predicted number of background events is due to MC statistics and varies between 7% and 25% for the four signal bins. Other contributions arise from the jet-energy calibration uncertainty (2% -12%) and the uncertainty in the shape of the jet transverse energy distribution (20%), which has been estimated by a comparison of the pythia and alpgen simulations. The jet multiplicity in DY events generated with pythia, which is a leading-order generator, was corrected in order to reflect the multiplicity distribution observed in the data around the Z boson. This was accomplished by comparing exponential fits to the inclusive jet multiplicity distribution in data and Monte Carlo. The fit is dominated by the zero and one jet bins. The remaining difference in the two jet bin between µjµj events in data and in the pythia Monte Carlo in the vicinity of the Z boson resonance, 60 GeV < m(µµ) < 105 GeV, was taken as the corresponding systematic uncertainty (16%). In addition, the following sources of systematic uncer- tainties were taken into account: luminosity (6.5%), PDF uncertainty of the DY processes (3.6%), and muon triggering and identification (5%). The systematics, added in quadrature, are shown in Table I. The systematic uncertainties on the signal efficiencies arise from limited Monte Carlo statistics (2% -17%), jet-energy scale (3% -13%), muon triggering and identification (5%), and PDF uncertainty (2%). No significant excess of data over background was observed. Upper limits on the product of cross section times branching fraction, σ · β 2 , were calculated as described in reference [13], by treating the four signal bins as individual channels. The likelihoods for the different bins were combined with correlations of systematic uncertainties taken into account. The limits are calculated using the confidence level CL S = CL S+B /CL B , where CL S+B is the confidence level for the signal plus background hypothesis and CL B is the confidence level for the background only [13].
The limits on the cross section times branching fraction and the theoretical predictions [14] are shown in Fig.  3 and Table II, as well as the average expected limit assuming that no signal is present. Due to the larger background, the contribution of bin 0 to the limit is relatively small. This explains why the average expected limit is better than the observed limit, although the sum of the events in all four bins is comparable to the background prediction. The mass limit is extracted from the intersection of the lower edge of the next-to-leading order (NLO) cross section uncertainty band with the observed upper bound on the cross section. The uncertainty band reflects the PDF uncertainty [15] as well as the variation of the factorization and renormalization scale between m LQ2 /2 and 2m LQ2 , added in quadrature.
The lower limit on the mass of second generation scalar leptoquarks was determined at the 95% C.L. to be m LQ2 > 247 GeV and m LQ2 > 182 GeV for β = 1 and β = 1/2, respectively. The average expected limits are m expected  shows the excluded region in the β versus m LQ2 parameter space.
The DØ Run I analysis in the µjµj channel had no events after all cuts, while 0.7 ± 0.5 events were expected from the background. A complementary Run I analysis in the µjνj channel yielded no events for 0.7 ± 0.9 events expected from standard model background [3]. Taking into account the smaller cross section for the production of second generation scalar leptoquarks at the Run I center-of-mass energy √ s = 1.8 TeV, these earlier results have been combined with the Run II analysis presented in this Letter. The results are summarized in Table II and the excluded parameter regions are shown in Fig. 4. The combined lower limit for scalar leptoquarks of the second generation is m LQ2 > 251 GeV (m LQ2 > 204 GeV) for β = 1 (β = 1/2). These results improve on previous measurements at the Tevatron collider [3,4] and are, for large β, the most stringent limits on second generation scalar leptoquarks from direct measurements to date.
We thank the staffs at Fermilab and collaborating in-  [15] and mLQ 2 as the factorization/renormalization scale [14]. The uncertainties in the theoretical cross sections originate from a variation of the renormalization and factorization scale between mLQ 2 /2 and 2mLQ 2 and the PDF errors, added in quadrature.