Search for single production of scalar leptoquarks in proton-proton collisions at sqrt(s) = 8 TeV

A search is presented for the production of both first- and second-generation scalar leptoquarks with a final state of either two electrons and one jet or two muons and one jet. A data sample of proton-proton collisions at center-of-mass energy sqrt(s) = 8 TeV recorded with the CMS detector corresponds to an integrated luminosity of 19.6 inverse femtobarns. Upper limits are set on both the first- and second-generation leptoquark production cross sections as functions of the leptoquark mass and the leptoquark couplings to a lepton and a quark. Results are compared with theoretical predictions to obtain lower limits on the leptoquark mass. At 95% confidence level, single production of first-generation leptoquarks with a coupling and branching fraction of 1.0 is excluded for masses below 1730 GeV, and second-generation leptoquarks with a coupling and branching fraction of 1.0 is excluded for masses below 530 GeV. These are the best overall limits on the production of first-generation leptoquarks to date.


Introduction
Leptoquarks (LQ) are hypothetical color-triplet bosons with spin 0 (scalar LQ) or 1 (vector LQ), which are predicted by many extensions of the standard model (SM) of particle physics, such as Grand Unified Theories [1][2][3][4][5][6][7][8], technicolor schemes [9][10][11], and composite models [12].They carry fractional electric charge (±1/3 for LQs considered in this paper) and both baryon and lepton numbers and thus couple to a lepton and a quark.Existing experimental limits on flavor changing neutral currents and other rare processes disfavor leptoquarks that couple to a quark and lepton of more than one SM generation [13,14].A discussion of the phenomenology of LQs at the LHC can be found elsewhere [15].
The production and decay of LQs at proton-proton colliders are characterized by the mass of the LQ particle, M LQ ; its decay branching fraction into a charged lepton and a quark, usually denoted as β; and the Yukawa coupling λ at the LQ-lepton-quark vertex.At hadron colliders, leptoquarks could be produced in pairs via gluon fusion and quark anti-quark annihilation, and singly via quark-gluon fusion.Pair production of LQs does not depend on λ, while single production does, and thus the sensitivity of single LQ searches depends on λ.At lower masses, the cross sections for pair production are greater than those for single production.Single production cross sections decrease more slowly with mass, exceeding pair production at an order of 1 TeV for λ = 0.6.
Several experiments have searched for LQs.The H1 collaboration has produced limits on various singly produced LQ types: the one to which to compare this search is the LQ called S R 0 in Ref. [16], for which they place a limit at 500 GeV, assuming λ = 1.0 and β = 1.0.The D0 collaboration has produced limits on singly produced scalar LQs of 274 GeV, again assuming λ = 1.0 and β = 1.0 [17].Limits from pair production of leptoquarks exclude leptoquark masses below 1010 GeV for the first generation and 1080 GeV for the second generation, for β = 1.0 [18].
The main single leptoquark production mode at the LHC is the resonant diagram shown in Fig. 1.However, significant contributions are made by the diagrams with non-resonant components shown in Fig. 2.These contributions increase with both the LQ mass and coupling; the invariant mass distribution of a first generation LQ, of mass M LQ = 1 TeV and coupling λ = 1.0, possesses a tail extending to very low masses that is comparable to the peak in magnitude.The reconstructed shape of the resonance peak itself is not strongly affected by λ.Also, interference with the qg → qZ/γ * → q + − SM process can occur at dilepton masses in the vicinity of the Z boson mass peak and at lower energies.Treatments for this interference region and the above-described low-mass off-shell tail of the lepton-jet mass distribution are detailed in Section 5.
The final-state event signatures from the decays of singly produced LQs can be classified as either that of two charged leptons and a jet, where the LQ decays to a charged lepton and a quark, or of a charged lepton, missing transverse energy, and a jet, where the LQ decays into a neutrino and a quark.The two signatures have branching fractions of β and 1 − β, respectively.For this study, and for S R 0 type LQs, β is 1.0, disregarding LQ decays to a neutrino and a quark.Because the parton distribution functions (PDF) of the proton are dominated by the u and d quarks, the single production of LQs of second and third generation is suppressed.
The charged leptons can be electrons, muons, or taus, corresponding to the three generations of LQs.In this paper two distinct signatures with charged leptons in the final state are considered: one with two high transverse momentum (p T ) electrons and one high-p T jet (denoted as eej), and the other with two high-p T muons and one high-p T jet (denoted as µµj).

The CMS detector
The central feature of the CMS apparatus is a superconducting solenoid of 6 m internal diameter, providing a magnetic field of 3.8 T. Within the solenoid volume are a silicon pixel and strip tracker, a lead tungstate crystal electromagnetic calorimeter (ECAL), and a brass and scintillator hadron calorimeter, each composed of a barrel and two endcap sections.Extensive forward calorimetry complements the coverage provided by the barrel and endcap detectors.Muons are measured in gas-ionization detectors embedded in the steel flux-return yoke outside the solenoid.
The ECAL energy resolution for electrons with E T ≈ 45 GeV from Z → ee decays is better than 2% in the central pseudorapidity region of the ECAL barrel (|η| < 0.8), and is between 2% and 5% elsewhere.For low-bremsstrahlung electrons, where 94% or more of their energy is contained within a 3×3 array of crystals, the energy resolution improves to 1.5% for |η| < 0.8 [19].
Muons are measured in the pseudorapidity range |η| < 2.4 with detection planes made using three technologies: drift tubes, cathode strip chambers, and resistive-plate chambers.Matching muon tracks derived from these measurements to tracks measured in the silicon tracker results in a relative p T resolution for muons with 20 < p T < 100 GeV of 1.3-2.0% in the barrel and better than 6% in the endcaps; the p T resolution in the barrel is better than 10% for muons with p T up to 1 TeV [20].
The first level of the CMS trigger system, composed of custom hardware processors, uses information from the calorimeters and muon detectors to select the most interesting events.The high-level trigger (HLT) processor farm further decreases the event rate from around 100 kHz to around 400 Hz, before data storage.
The particle-flow event algorithm reconstructs and identifies each individual particle with an optimized combination of information from the various elements of the CMS detector.The energy of photons is directly obtained from the ECAL measurement, corrected for zerosuppression effects.The energy of electrons is determined from a combination of the electron momentum at the primary interaction vertex as determined by the tracker, the energy of the corresponding ECAL cluster, and the energy sum of all bremsstrahlung photons spatially compatible with originating from the electron track.The energy of muons is obtained from the curvature of the corresponding track.The energy of charged hadrons is determined from a combination of their momentum measured in the tracker and the matching ECAL and HCAL energy deposits, corrected for zero-suppression effects and for the response function of the calorimeters to hadronic showers.Finally, the energy of neutral hadrons is obtained from the corresponding corrected ECAL and HCAL energy.
A more detailed description of the CMS detector, together with a definition of the coordinate system used and the relevant kinematic variables, can be found in [21].

Data and simulation samples
The data were collected during the 8 TeV pp run in 2012 at the CERN LHC and correspond to an integrated luminosity of 19.6 fb −1 .In the eej channel, events are selected using a trigger that requires two electrons with p T > 33 GeV and |η| < 2.4 and in the µµj channel, events are selected using a trigger that requires one muon with p T > 40 GeV and |η| < 2.1.
Simulated samples for the signal processes are generated for a range of leptoquark mass hypotheses between 300 and 3300 GeV and coupling hypotheses between 0.4 and 1.0 in the eej channel, and a range of leptoquark mass hypotheses between 300 and 1800 GeV and a coupling hypothesis of 1.0 in the µµj channel.Production of LQs in the µµj channel is suppressed because of the proton PDF as discussed in Section 1.
The main sources of background are tt, Z/γ * + jets, W + jets, diboson (ZZ, ZW, WW) + jets, single top quark, and QCD multijet production.The tt + jets background shape is estimated from a study based on data described in Section 6; the simulation sample for the normalization of the tt + jets background as well as the samples for the Z/γ * + jets and W + jets backgrounds are generated with MADGRAPH 5.1 [22].Single top quark samples (s-, t-channels, and W boson associated production) are generated with POWHEG 1.0 [23][24][25][26] and diboson samples are generated with PYTHIA (version 6.422) [27] using the Z2 tune [28].The QCD multijet background is estimated from data.
For the simulation of signal samples, the CALCHEP [29] generator is used for calculation of the matrix elements.The signal cross sections are computed at leading order (LO) with CALCHEP and are listed in Table A.1 in the appendix.Blank entries were not considered because of the small size of the cross section.The resonant cross sections σ res are shown in Fig. 3 and are defined by the kinematics selections given in Section 5.
The PYTHIA and MADGRAPH simulations use the CTEQ6L1 [30] PDF sets, those produced with CALCHEP use the CTEQ6L PDFs, and the POWHEG simulation uses the CTEQ6m set.All of the simulations use PYTHIA for the treatment of parton showering, hadronization, and underlying event effects.For both signal and background simulated samples, the simulation of the CMS detector is based on the GEANT4 package [31].All simulated samples include the effects of extra collisions in a single bunch crossing as well as collisions from nearby bunch crossings (out-of-time pileup and in-time pileup, respectively).The pileup profiles in simula-4 Event reconstruction tion are reweighted to the distributions of the reconstructed vertices per bunch crossing in data collected by the CMS detector [32].In the eej channel, the background and signal are rescaled by a uniform trigger efficiency scale factor of 0.996, which is measured in [33].In the µµj channel, the background and signal are rescaled by muon η-dependent efficiency factors of 0.94 (|η| ≤ 0.9), 0.84 (0.9 < |η| ≤ 1.2), and 0.82 (1.2 < |η| ≤ 2.1).An uncertainty of 1% is assigned to these factors to account for variations during data-taking periods and statistical uncertainties.

Event reconstruction
Muons are reconstructed as tracks in the muon system that are "globally" matched to reconstructed tracks in the tracking system [20].Muons are required to have p T > 45 GeV and |η| < 2.1.Additionally, they are required to satisfy a set of criteria that is optimized for high p T ; they are reconstructed as "global" muons with tracks associated to hits from at least two muon detector planes together with at least one muon chamber hit that is included in the "global" track fit [20].To perform a precise measurement of the p T and to reduce background from muons from secondary decays in flight, at least eight hits are required in the tracker and at least one in the pixel detector.To minimize background from muons from cosmic ray backgrounds, the transverse impact parameter with respect to the primary vertex is required to be less than 2 mm and the longitudinal distance is less than 5 mm.Muons are required to be isolated by applying an upper threshold on the relative tracker isolation of 0.1.The relative tracker isolation is defined as the ratio of the p T of all tracks in the tracker coming from the same vertex, excluding the muon candidate track, in a cone of ∆R = √ (∆φ) 2 + (∆η) 2 = 0.3 (where φ is the azimuthal angle in radians) around the muon candidate track, and the muon p T .
Electrons are required to have a reconstructed track in the central tracking system that is matched in η and φ to a cluster of ECAL crystals that has a shape consistent with an electromagnetic shower.The transverse impact parameter of the track with respect to the primary vertex is required to be less than 2 mm for electrons in the barrel (|η| < 1.442) and less than 5 mm for electrons in the endcap (|η| > 1.560).Electrons are required to be isolated from reconstructed tracks other than the matched track in the central tracking system and from additional energy deposits in the calorimeter.The transverse momentum sum of all tracks in a cone of ∆R = 0.5 around the electron candidate's track and coming from the same vertex must be less than 5 GeV.Also, the transverse energy sum of the calorimeter energy deposits falling in the cone of ∆R = 0.5 must be less than 3% of the candidate's transverse energy.An additional contribution accounting for the average contribution of other proton-proton collisions in the same bunch crossing is added to this sum.To reject electrons coming from photon conversions within the tracker material, the reconstructed electron track is required to have hits in all pixel layers.Electrons in the analysis have p T > 45 GeV and |η| < 2.1 to match the muon requirements (excluding the transition region between barrel and endcap detectors, 1.442 < |η| < 1.560).Selection criteria for electron identification and isolation optimized for high energies are also applied [33].
Jets are reconstructed with the CMS particle-flow algorithm [34,35], which measures stable particles by combining information from all CMS subdetectors.The jet reconstruction algorithm used in this paper is the anti-k T [36,37] algorithm with a distance parameter 0.5, which only considers tracks associated to the primary vertex.Jet momentum is determined as the vectorial sum of all particle momenta in the jet, and is found from simulation to be within 5% to 10% of the true momentum over the whole p T spectrum and detector acceptance.An offset correction is applied to jet energies to take into account the contribution from additional proton-proton interactions within the same bunch crossing.Jet energy corrections are derived from simulation, and are confirmed with in situ measurements of the energy balance in dijet and photon+jet events [38].Additional selection criteria are applied to each event to remove spurious jet-like features originating from isolated noise patterns in certain HCAL regions.The jet energy resolution amounts typically to 15% at 10 GeV, 8% at 100 GeV, and 4% at 1 TeV, to be compared to about 40%, 12%, and 5% obtained when the calorimeters alone are used for jet clustering [34].
Jets are required to have p T > 45 GeV, |η| < 2.4, and an angular separation from leptons of ∆R > 0.3.

Event selection
We require that events in both the eej and µµj channels contain at least two leptons and at least one jet that satisfy the above identification criteria.Additional kinematic requirements are applied to remove regions in which the trigger and identification criteria are not at plateau efficiency and to reduce large backgrounds.This creates a basic preselection region: the jet p T must be larger than 125 GeV, the dilepton invariant mass M must be larger than 110 GeV, and the scalar sum of transverse momenta of objects in the event (S T = p T ( 1 ) + p T ( 2 ) + p T (j 1 )) is required to exceed 250 GeV, where 1 is the highest p T lepton in the event, 2 is the secondhighest p T lepton, and j 1 is the highest p T jet.The two leptons in the events are required to have opposite charges.
After this initial selection, a final selection is optimized for each channel separately by maximizing S/ √ S + B, where S is the number of signal events in the simulation passing a given selection and B is the number of background events in the simulation passing the same selection.We optimize for each LQ mass hypothesis by varying the requirements on M j and S T .Here M j is defined as the higher of the two possible lepton-jet mass combinations.
As discussed in Section 1, owing to the unique aspects of single LQ decays, two generator level requirements are applied to the simulated signal samples.The first is M > 110 GeV, to remove LQ decays that are in the Z boson interference region.The second is a requirement on M j , chosen to remove the t-channel diagram contributions in the low-mass off-shell region, while preserving most of the resonant signal.This requirement is set at M j > 0.67 M LQ for the 6 Background estimations first-generation studies and M j > 0.75 M LQ for the second-generation studies.The thresholds for M j were chosen separately for each channel, because of the differences in the distribution shape.The dilepton invariant mass requirement at the generator level precisely matches the reconstruction level requirement at the preselection.These two requirements define the resonant region.Cross sections at the generator level before and after these requirements are provided in Table A.1, in the appendix.
The eej channel selection after optimization is identical for all couplings.The threshold on S T starts at 250 GeV for M LQ = 300 GeV and increases linearly until it reaches a plateau value of 900 GeV at M LQ = 1125 GeV.The M j threshold starts at 200 GeV for M LQ = 300 GeV and increases linearly until it plateaus at 1900 GeV above M LQ = 2000 GeV.In the µµj channel after optimization the threshold on S T starts at 300 GeV for M LQ = 300 GeV and increases linearly until it plateaus at 1000 GeV above M LQ = 1000 GeV.The M j threshold starts at 200 GeV for M LQ = 300 GeV and increases linearly until it plateaus at 800 GeV above M LQ = 900 GeV.The exact threshold values are listed in Tables B.1 and B.2 in the appendix.

Background estimations
The SM processes that mimic the signal signature are Z/γ * + jets, tt, single top quark, diboson +jets, W + jets, and QCD multijets events where the jets are misidentified as leptons.The dominant contributions come from the former two processes, whereas the other processes provide minor contributions to the total number of background events.
The contribution from the Z/γ * + jets background is estimated with a simulated sample that is normalized to agree with data at preselection in the Z-enriched region of 80 < M < 100 GeV, where M is the dilepton invariant mass.With this selection the data sample (with non-Z/γ * + jets simulated samples subtracted) is compared to Z/γ * + jets in simulation.The resulting scale factor, representing the ratio of the measured yield to the predicted yield, is R Z = 0.98 ± 0.01 (stat) in both the eej and µµj channels.This scale factor is then applied to the simulated Z/γ * + jets sample in the signal region of M > 110 GeV.In order to account for possible mismodeling of the p T ( ) spectrum of the Z/γ * + jets background sample, where p T ( ) is the scalar sum of the two highest p T leptons in the event, we perform a bin-by-bin rescaling of yields at preselection and full selection by scale factors measured in an inverted M selection (M < 110 GeV).These scale factors differ from unity by 1% to 10%, depending on the p T ( ) bin, and are applied to the Z/γ * + jets sample in the signal region of M > 110 GeV.
We estimate the tt background with a tt-enriched eµ sample in data, selected using the single muon trigger.We use a selection that is identical to our signal selection in terms of kinematics requirements, except that we require at least a single muon and a single electron rather than requiring two same-flavor leptons.The eµ sample is considered to be signal-free, because limits on flavor changing neutral currents imply that LQ processes do not present a different-flavor decay topology [13,14].The tt background is largely dominant in the eµ sample with respect to the other backgrounds.This background is expected to produce the ee (µµ) final state with half the probability of the eµ final state, thus the eµ sample is scaled by a factor of 1/2.This factor is multiplied by the ratio of electron (muon) identification and isolation efficiencies, R ee/eµ (R µµ/eµ ).The estimate is further scaled by the ratio of the double-electron trigger efficiency and the single muon efficiency, R trig,ee in Eq. 3, or by the ratio of the efficiency of the single muon trigger in dimuon final states and the single muon efficiency, R trig,µµ in Eq. 4. The resulting estimates of the number of tt events in the ee and µµ channels are with where µ and ee are the single-muon trigger and double-electron trigger efficiencies, respectively, and N data eµ and N non-tt sim eµ are the numbers of eµ events observed in data and estimated from backgrounds other than tt, respectively.R trig,µµ is the ratio of the efficiency of a single muon trigger on a dimuon sample over the efficiency on a single muon sample (the numerator is the likelihood of failure on two muons).
The contribution from QCD multijet processes is determined by a method that makes use of the fact that neither signal events nor events from other backgrounds produce final states with same-charge leptons at a significant level.We create four selections, with both opposite-sign (OS) and same-sign (SS) charge requirements, as well as isolated and non-isolated requirements.Electrons in isolated events must pass the isolation criteria optimized for high-energy electrons [33] and muons are required to have a relative tracker isolation less than 0.1, as discussed in Section 4. Non-isolated events are those with leptons failing these criteria.The four selections are as follows, The shape of the background is taken from the SS region with isolation requirements, and the normalization is obtained from the ratio between the number of OS events and the number of SS events in the non-isolated selection.Thus, the number of events, N QCD, est , is estimated by where is the number of events in region C of Eq. ( 5) and r B/D is the ratio of the number of events (measured in data with simulated non-QCD backgrounds subtracted) in regions B and D. The result is that QCD multijet processes account for 2% (1%) of the total SM background in the eej (µµj) channel.
The contributions of the remaining backgrounds (diboson+jets, W+jets, single top quark) are small and are determined entirely from simulation.
The preselection level distributions in M , S T , and M j are shown in Figs. 4 and 5 for the observed data and estimated backgrounds, where they are compared with a signal LQ mass of 1000 GeV in the eej channel, and with a signal LQ of mass 600 GeV, in the µµj channel.In all plots the Z/γ * + jets prediction is normalized to data and the tt prediction is taken from the study based on data.Data and background are found to be in agreement.The numbers of events selected in data and in the backgrounds at each final selection (for each hypothesis mass) are shown in Tables C.1, C.2, and C.3 in the appendix.

Systematic uncertainties
The observed data and background predictions are compared after final selection for λ = 0.4 and a signal LQ mass of 1000 GeV in the eej channel and a signal LQ mass of 600 GeV in the µµj channel and are shown in Figs. 6 and 7.

Systematic uncertainties
The sources of systematic uncertainties considered in this analysis are listed below.To determine the uncertainties in signal and background, each kinematic quantity listed is varied individually according to its uncertainty and the final event yields are re-measured to determine the variation in the predicted number of background and signal events.
Jet energy scale and resolution uncertainties are estimated by assigning p T -and η-dependent uncertainties in jet energy corrections as discussed in Ref. [38], and varying the jet p T according to the magnitude of that uncertainty.The uncertainty in the jet energy resolution is assessed by modifying the p T difference between the particle level and reconstructed jets by an η-dependent value between 5% and 30% for most jets [38].
Uncertainties in the charged-lepton momentum scale and resolution also introduce uncertainties in the final event acceptance.An energy scale uncertainty of 0.6% in the ECAL barrel and 1.5% in the ECAL endcap is assigned to electrons [39], and an uncertainty of 10% in both the ECAL barrel and endcap is applied to the electron energy resolution [39].There is an uncertainty of 0.6% per electron in reconstruction, identification, and isolation requirements.For muons, a p T -dependent scale uncertainty of 5% (p T /1 TeV) is applied, as well as a 1-4% p Tdependent resolution uncertainty [20].In the case of momentum scale uncertainties the momentum is directly varied, and in the case of momentum resolution uncertainties the lepton momentum is subjected to a Gaussian random smearing within the uncertainty.A 2% per muon uncertainty in reconstruction, identification, and isolation requirements, as well as a 1% muon HLT efficiency uncertainty, are assumed as well.
Other important sources of systematic uncertainty are related to the modeling of the backgrounds in the simulation.The uncertainty in the Z/γ * + jets background shape is determined by using simulated samples with renormalization and factorization scales and matrix-element parton-shower matching thresholds varied by a factor of two up and down.The scale factors for the normalization of the Z/γ * + jets background are assigned an uncertainty of 0.6% in both channels, and the normalization of the tt background is assigned an uncertainty of 0.5% in both channels, based on the statistical uncertainties measured in the studies described in Section 6.An additional uncertainty of 4% is applied to the tt background normalization in the µµj channel to account for possible signal contamination from first generation LQs in the control sample (the contamination is extremely small in the other channel because of the suppressed second generation signal).An uncertainty on Z/γ * + jets background from the p T ( ) scale factors is assessed by taking the weighted average of the uncertainties from each p T ( ) bin.The estimate of the QCD multijet background from data has an uncertainty of 15%.
An uncertainty in the modeling of pileup in simulation is determined by varying the number of simulated pileup interactions up and down by 6% [40], and an uncertainty of 2.6% on the measured integrated luminosity is applied [41].
Uncertainties in the signal acceptance, the background acceptance, and the cross sections, due to the PDF choice of 4-10% for signal and 3-9% for background are applied, following the PDF4LHC recommendations described in Refs.[42,43].
Finally, a statistical uncertainty associated with the size of the simulated sample is included for both background and signal.
The systematic uncertainties are listed in Table 1, together with their effects on signal and background yields, corresponding to the final selection values optimized for M LQ = 600 GeV.The PDF uncertainty is larger in the µµj channel because of the large uncertainty associated with the s-quark PDF.

Results
The observed data are consistent with the no-signal hypothesis.We set an upper limit on the leptoquark cross section by using the CL S modified frequentist method [44,45] with the final event yields.A log-normal probability function is used to model the systematic uncertainties, whereas statistical uncertainties are described with gamma distributions with widths determined according to the number of events simulated or measured in data control regions.
To isolate the limits for resonant LQ production, we apply the resonant requirements at the generator level on both the lepton+jet mass, M( , j) > (0.67 or 0.75) M LQ (for the first-or secondgeneration LQs, respectively), and on the dilepton mass, M > 110 GeV.These requirements make the limits extracted from data more conservative and are discussed in Section 5. A resonant cross section σ res is computed with respect to those requirements.Limits are then computed with the reduced sample of simulated signal events and compared to σ res .The 95% confidence level (CL) upper limits on σ res β as a function of leptoquark mass are shown in Fig. 8 together with the resonant cross section predictions for the scalar leptoquark single production cross section.The uncertainty band on the theoretical cross section prediction corresponds to uncertainties in the total cross section due to PDF variations with an additional +70% uncertainty, because of the k factor from NLO corrections [46].The observed limits are listed in Tables B.1 and B.2 in the appendix.First generation limits are shown on the left plot with a resonant region of M j > 0.66 M LQ , M > 110 GeV and second generation limits are shown on the right plot with a resonant region of M j > 0.75 M LQ , M > 110 GeV.The uncertainty bands on the observed limit represent the 68% and 95% confidence intervals.The uncertainty band on the theoretical cross section includes uncertainties due to PDF variation and the k factor.9 Summary By comparing the observed upper limit with the theoretical production cross-section times branching fraction, we exclude single leptoquark production at 95% CL for LQ masses below the values given in Table 2. Limits on single production of the S R 0 type LQ from the H1 collaboration exclude LQ production up to 500 GeV (λ = 1.0) and up to 350 GeV (λ = 0.6) [16].

Summary
A search has been performed for the single production of first-and second-generation scalar leptoquarks in final states with two electrons and a jet or two muons and a jet using a data set of proton-proton collisions at 8 TeV corresponding to an integrated luminosity of 19.6 fb −1 .The selection criteria are optimized for each leptoquark signal mass hypothesis.The number of observed candidates for each mass hypothesis agrees with the number of expected standard model background events.Single production of first-(second-) generation leptoquarks with a coupling of 1.0 is excluded at 95% confidence level for masses below 1755(660) GeV.These are the most stringent limits to date for single production.The first-generation limits for couplings greater than 0.6 are stronger than those from pair production and are the most stringent overall limits on leptoquark production in the first generation to date.This section contains tables of data, background, and signal yields after the final selection.Event counts vary between the two channels due to differences in the optimized thresholds for S T and M j as well as differences in the electron and muon efficiencies.The first listed uncertainty is statistical, the second is systematic; in cases where only one uncertainty is listed it is statistical.

Figure 3 :
Figure 3: Cross sections for single LQ production, calculated at LO in CALCHEP and scaled by the acceptance of the requirements described in Section 5, as a function of the LQ mass in GeV.

Figure 4 :
Figure 4: Distributions of M ee (top left), S T (top right), and M ej (bottom) at preselection in the eej channel."Other backgrounds" include diboson, W+ jets, and single top quark contributions.The points represent the data and the stacked histograms show the expected background contributions.The open histogram shows the prediction for an LQ signal for M LQ = 1000 GeV and λ = 0.4.The horizontal error bars on the data points represent the bin width.The last bin includes overflow.

Figure 5 :Figure 6 :
Figure 5: Distributions of M µµ (top left), S T (top right), and M µj (bottom) at preselection in the µµj channel."Other backgrounds" include diboson, W+ jets, single top quark, and QCD multijet contributions.The points represent the data and the stacked histograms show the expected background contributions.The open histogram shows the prediction for an LQ signal for M LQ = 600 GeV and λ = 1.0.The horizontal error bars on the data points represent the bin width.The last bin includes overflow.

Figure 7 :
Figure 7: Distributions of S T and M µj at final selection, in the µµj channel.The points represent the data and the stacked histograms show the expected background contributions.The open histogram shows the prediction for an LQ signal for M LQ = 600 GeV and λ = 1.0.The horizontal error bars on the data points represent the bin width.The last bin includes overflow.

FirstFigure 8 :
Figure 8: Expected and observed upper limits at 95% CL on first and second generation leptoquark single production resonant cross section as a function of the leptoquark mass.First generation limits are shown on the left plot with a resonant region of M j > 0.66 M LQ , M > 110 GeV and second generation limits are shown on the right plot with a resonant region of M j > 0.75 M LQ , M > 110 GeV.The uncertainty bands on the observed limit represent the 68% and 95% confidence intervals.The uncertainty band on the theoretical cross section includes uncertainties due to PDF variation and the k factor.

Table 1 :
Systematic uncertainties (in %) and their effects on total signal (S) and background (B) in both channels for M LQ = 600 GeV final selection.

Table A .
1: Signal cross sections calculated at LO in CALCHEP.Resonant cross sections scaled by the acceptance of the selections described in Section 5 are listed under each corresponding LO cross section.

Table B .
1: The eej channel threshold values for S T , M ej , and M ej,gen vs. LQ mass (for all couplings), and the corresponding observed limits.M LQ S T threshold M ej threshold M ej,gen threshold Observed limit on σ res

Table B .
2:The µµj channel threshold values for S T , M µj , and M µj,gen vs. LQ mass, and the corresponding observed limits.M LQ S T threshold M µj threshold M µj,gen threshold Observed limit on σ res

Table C .
1: Data and background yields after final selection for the eej channel for firstgeneration LQs, shown with statistical and systematic uncertainties."Other backgrounds" refers to diboson+jets, W+ jets, single-top quark, and QCD.The values do not change above 2000 GeV.