Search for electroweak production of charginos and neutralinos in WH events in proton-proton collisions at sqrt(s) = 13 TeV

Results are reported from a search for physics beyond the standard model in proton-proton collision events with a charged lepton (electron or muon), two jets identified as originating from a bottom quark decay, and significant imbalance in the transverse momentum. The search was performed using a data sample corresponding to 35.9 inverse femtobarns, collected by the CMS experiment in 2016 at sqrt(s) = 13 TeV. Events with this signature can arise, for example, from the electroweak production of gauginos, which are predicted in models based on supersymmetry. The event yields observed in data are consistent with the estimated standard model backgrounds. Limits are obtained on the cross sections for chargino-neutralino production in a simplified model of supersymmetry with the decays chargino to W+/- lsp and neutralino to H lsp. Values of m[chargino] between 220 and 490 GeV are excluded at 95% confidence level by this search when the lsp is massless, and values of m[lsp] are excluded up to 110 GeV for m[chargino] about 450 GeV.


Introduction
Supersymmetry (SUSY) [1][2][3][4][5][6][7][8] is a theoretically attractive extension of the standard model (SM) that is based on a symmetry between bosons and fermions. SUSY predicts the existence of a superpartner for every SM particle, with the same gauge quantum numbers but differing by one half unit of spin. In R-parity conserving SUSY models, supersymmetric particles are created in pairs, and the lightest supersymmetric particle (LSP) is stable [9][10][11]. As a result, SUSY also provides a potential connection to cosmology as the LSP, if neutral and stable, may be a viable dark matter candidate.
Previous searches based on 13 TeV proton-proton collision data at the CERN LHC focused on strong production of colored SUSY particles [12][13][14][15][16][17][18][19][20][21][22][23][24][25][26][27][28]. Pair production of these particles would have the largest cross section for SUSY processes and therefore provides the strongest discovery potential with small datasets. However, the absence of signals in these searches suggests that strongly produced SUSY particles may be too massive to be found with the present data. In contrast, neutralinos ( χ 0 ) and charginos ( χ ± ), mixtures of the superpartners of the SM electroweak gauge bosons and the Higgs bosons, can have masses within the accessible range. Because of the absence of color charge, the production cross sections are lower, and these particles may have thus far eluded detection. This provides strong motivation for dedicated searches for electroweak SUSY particle production.
Depending on the mass spectrum, the charginos and neutralinos can have significant decay branching fractions to vector bosons V (W or Z) and the Higgs boson (H). Here, "H" refers to the 125 GeV Higgs boson [29], interpreted as the lightest CP-even state of an extended Higgs sector. The H boson is expected to have SM-like properties if all of the other Higgs bosons are much heavier [30]. The observation of a Higgs boson in a SUSY-like process would provide evidence that SUSY particles couple to the Higgs field, a necessary condition for SUSY to stabilize the Higgs boson mass. Pair production of neutralinos and/or charginos can thus lead to the HH, VH, and VV decay modes, with a large fraction of the possible final states containing at least one isolated lepton. Such events can be easily selected with simple triggers and do not suffer from large quantum chromodynamics multijet background.
In this paper we focus on a simplified model [31][32][33][34][35] of supersymmetric chargino-neutralino ( χ ± 1 χ 0 2 ) production with the decays χ ± 1 → W ± χ 0 1 and χ 0 2 → H χ 0 1 , as shown in Fig. 1. Both the χ ± 1 and χ 0 2 are assumed to be wino-like and have the same mass. The lightest neutralino χ 0 1 , produced in the decays of χ ± 1 or the χ 0 2 , is considered to be the stable LSP, which escapes detection. When the W boson decays leptonically, this process typically results in a signature with one lepton, two jets that originate from the decay H → bb, and large missing transverse momentum from the neutrino in the W boson decay and the LSPs.
Results of searches for electroweak pair production of SUSY particles were previously reported by the ATLAS and CMS Collaborations using data sets of 8 TeV proton-proton (pp) collisions [36][37][38] in a variety of event topologies and final states. No excesses above the SM expectations were observed, and the results of those searches were used to place lower limits on the mass of pair-produced charginos and neutralinos. Assuming mass-degenerate χ ± 1 and χ 0 2 , and sleptons (the SUSY partners of the SM leptons) with lower masses, the searches probed masses up to approximately 700 GeV. For the WH decays assumed here, the strongest mass limit was around 270 GeV. With the increase of the LHC collision energy from 8 to 13 TeV, and a significantly larger data set, searches based on 13 TeV data have the potential to quickly surpass the sensitivity of the previous analyses.
This paper presents the result of a search using a data set corresponding to an integrated lu- Figure 1: Diagram corresponding to the SUSY simplified model targeted by this analysis, i.e., chargino-neutralino production, with the chargino decaying to a W boson and an LSP, while the heavier neutralino decays to a Higgs boson and an LSP. minosity of 35.9 fb −1 of pp collisions collected at a center-of-mass energy of 13 TeV with the CMS detector in 2016. The results are interpreted in the simplified SUSY model with charginoneutralino production depicted in Fig. 1.
ton beams. Its magnitude is referred to as E miss T . Events with possible contributions from beam halo processes or anomalous noise in the calorimeter can have large values of E miss T and are rejected using dedicated filters [44].
Data events are selected using triggers that require the presence of an isolated electron or muon with p T thresholds of 27 GeV or 24 GeV, respectively. Muon events may also be accepted using a trigger that does not require isolation but instead requires p T > 50 GeV. The trigger efficiency, measured using a data sample of Z/γ → events, varies in the range 70-95% (85-92%) depending on the η and p T of the electron (muon).
Selected events are required to have exactly one lepton (electron or muon), with electrons (muons) satisfying p T > 30(25) GeV and |η| < 1.44(2.1). Electron candidates are reconstructed starting from a cluster of energy deposits in the electromagnetic calorimeter. The cluster is then matched to a reconstructed track. The electron selection is based on the shower shape, trackcluster matching, and consistency between the cluster energy and the track momentum [45]. Muon candidates are reconstructed by performing a global fit that requires consistent hit patterns in the tracker and the muon system [46]. For both lepton flavors, the impact parameter with respect to the primary vertex is required to be less than 0.5 mm in the transverse plane and 1 mm along the beam direction.
Leptons are required to be isolated from other activity in the event. A measure of lepton isolation is the scalar p T sum (p sum T ) of all PF candidates not associated with the lepton within a cone of radius ∆R ≡ √ (∆η) 2 + (∆φ) 2 = 0.3, where ∆η and ∆φ are the distances between the lepton and the PF candidates at the primary vertex in η-φ space [47]. Only charged PF candidates compatible with the primary vertex are included in the sum. The average contribution of particles from additional pp interactions in the same or nearby bunch crossings (pileup) is subtracted from p sum T . We require that p sum T be less than 5 GeV. Typical lepton identification and isolation efficiencies, measured in samples of Z/γ → events, are approximately 80-85% (85-90%) for electrons (muons), depending on p T and η.
Particle-flow candidates are clustered to form jets using the anti-k T clustering algorithm [42] with a distance parameter of 0.4, as implemented in the FASTJET package [43]. Only charged PF candidates compatible with the primary vertex are used in the clustering. The pileup contribution to the jet energy is estimated on an event-by-event basis using the jet area method described in [48] and is subtracted from the overall jet p T . Corrections are applied to the energy measurements of jets to account for non-uniform detector response and are propagated consistently as a correction to p miss T [49,50]. The selected lepton can also be reconstructed as a jet, so any jets within ∆R = 0.4 of the lepton are removed from the list of considered jets.
Selected events are required to contain exactly two jets with p T > 30 GeV and |η| < 2.4. Both of these jets must be consistent with containing the decay products of a heavy-flavor (HF) hadron, as identified using the combined secondary vertex (CSVv2) tagging algorithm [51]. Such jets are referred to as b jets. The CSVv2 algorithm has three main operating points: loose, medium, and tight. We require both jets to be tagged according to the loose operating point, and at least one of them to be tagged with the medium operating point. The efficiency of this algorithm for jets arising from b quarks with p T between 30 and 400 GeV is in the range 60-65% (70-75%) for the medium (loose) working point. The misidentification rate for jets arising from light quarks or gluons is approximately 1% (10%) for the medium (loose) working point.
The largest background in this search arises from tt and tW events with decays into two-lepton final states in which one of the leptons is not reconstructed or identified. In order to reduce these backgrounds, we search for a second electron or muon with p T > 5 GeV and looser iden-tification and isolation requirements, and reject events where such a lepton is found. Second leptons are required to satisfy p sum T /p T < 0.1, where p sum T is calculated here with a cone radius of ∆R = 0.2 for p lep T ≤ 50 GeV, and ∆R = max(0.05, 10 GeV/p lep T ) at higher values of lepton transverse momentum. We also reject events with reconstructed hadronically decaying tau leptons with p T > 20 GeV [52], or isolated tracks with p T > 10 GeV and opposite electric charge relative to the selected lepton. For this purpose, a track is considered isolated if p sum T /p T < 0.1 and p sum T < 6 GeV, where p sum T here is constructed with charged PF candidates compatible with the primary vertex, the cone radius is ∆R = 0.3, and p T is the transverse momentum of the track.
The final two requirements that complete the preselection are E miss T ≥ 125 GeV and M T > 50 GeV, where M T is the transverse mass of the lepton-E miss T system, defined as where p T is the transverse momentum of the lepton and ∆φ is the angle between the transverse momentum of the lepton and p miss T .

Signal region definition
The signal regions are defined by additional requirements on the kinematic properties of preselected events. The invariant mass of the two b jets is required to be in the range 90 ≤ M bb ≤ 150 GeV, consistent with the Higgs boson mass within the resolution. The M bb distribution for signal and background processes is shown in Fig. 2 (top left), displaying a clear peak for signal events near the Higgs boson mass.
To suppress single-lepton backgrounds originating from semileptonic tt, W + jets, and single top quark processes, the preselection requirement on M T is tightened to >150 GeV. This is because the M T distribution in these processes with a single leptonically decaying W boson has a kinematic endpoint M T < m W , where m W is the W boson mass. The endpoint can be exceeded by off mass-shell W bosons or because of detector resolution effects. The M T requirement significantly reduces single-lepton backgrounds, as shown in Fig. 2 (bottom left).
In order to further suppress both semileptonic and dileptonic tt backgrounds, we utilize the contransverse mass variable, M CT [53,54]: where p b1 T and p b2 T are the transverse momenta of the two jets, and ∆φ bb is the azimuthal angle between the pair. As shown in Refs. [53,54], this variable has a kinematic endpoint at (m 2 (δ) − m 2 (α))/m(δ), where δ is the pair-produced heavy particle and α is the invisible particle produced in the decay of δ. In the case of tt events, when both jets from b quarks are correctly identified, the kinematic endpoint corresponds to the top quark mass, while signal events tend to have higher values of M CT . This is shown in Fig. 2 After all other selections, we define two exclusive bins in E miss T to enhance sensitivity to signal models with different mass spectra: 125 ≤ E miss

Signal and background simulation
Samples of tt, W + jets, and Z + jets events, as well as tt production in association with a vector boson, are generated using the MADGRAPH5 aMC@NLO 2.2.2 [55] generator at lead- ing order (LO) with the MLM matching scheme [56], while tW and single top quark t-channel events are generated at next-to-leading-order (NLO) using POWHEG V2 [57][58][59]. A top quark mass of m t = 172.5 GeV, and the NNPDF3.0 LO or NLO [60] parton distribution functions (PDFs) are used in the event generation. Single top quark s-channel production is simulated using MADGRAPH5 aMC@NLO 2.2.2 at NLO precision with the FxFx matching scheme [61]. Samples of diboson (WW, WZ, and ZZ) events are generated with either POWHEG or MAD-GRAPH5 aMC@NLO at NLO precision. Normalization of the simulated background samples is performed using the most accurate cross section calculations available [55,[62][63][64][65][66][67][68][69][70][71][72], which generally correspond to NLO or next-to-NLO precision.
The chargino-neutralino signal samples are also generated with MADGRAPH5 aMC@NLO at LO precision. For these samples we improve on the modeling of initial-state radiation (ISR), which affects the total transverse momentum (p ISR T ) of the system of SUSY particles, by reweighting the p ISR T distribution in these events. This reweighting procedure is based on studies of the p T of Z bosons [73]. The reweighting factors range between 1.18 at p ISR T = 125 GeV and 0.78 for p ISR T > 600 GeV. We take the deviation from 1.0 as the systematic uncertainty in the reweighting procedure.
Parton showering and fragmentation in all of these samples are performed using PYTHIA V8.212 [74] with the CUETP8M1 tune [75]. For both signal and background events, additional simultaneous proton-proton interactions (pileup) are generated with PYTHIA and superimposed on the hard collisions. The response of the CMS detector for SM background samples is simulated using GEANT4-based model [76], while that for new physics signals is performed using the CMS fast simulation package [77]. All simulated events are processed with the same chain of reconstruction programs as that used for collision data.
Small differences between the b tagging efficiencies measured in data and simulation are corrected using data-to-simulation scale factors. Corrections are also applied to account for differences between lepton selection efficiencies (trigger, reconstruction, identification, and isolation) in data and simulation.

Backgrounds
The backgrounds for this search are classified into six categories. The first and most important category, referred to as dilepton top quark events, consists mainly of events from top quark pair production with both quarks decaying leptonically, but also including contributions from the associated production of a single top quark with a W boson, both of which decay leptonically. The second to fifth categories include processes with a single leptonically decaying W boson. Events with a single W are divided into two categories: those with b jets (W + HF, for "heavy flavor") and those without (W + LF, for "light flavor"). A separate category comprises WZ events in which the Z boson decays to bb (WZ → νbb). Events with one leptonically decaying top quark, either from tt or from single top quark tor s-channel production, are included in the fifth category ("single-lepton top quark"). Finally, other SM processes contribute a small amount to the expected yield in the signal region and are grouped together in the "rare" category. This includes events from Z + jets, WW, WZ (except the decays described above), ZZ, triboson, ttW, ttZ, and WH → νbb processes.
All background processes are modeled using MC simulation. Three data control regions (CRs) are defined by inverting the signal region selection requirements, as summarized in Table 1. The CRs are defined at both preselection and signal region selection levels. The CRs at the preselection level are defined with looser cuts in order to check the modeling of key discriminant variables. The CRs after the signal region level selections are used to validate the modeling of the main backgrounds and to assign systematic uncertainties in the background predictions. The regions CRMbb and CR0b are split into two bins in E miss T to mirror the signal region selection. The expected signal contribution in any of the CRs is always less than 1% of the total SM yields, and typically much smaller.
The dilepton top quark background can be isolated in the CR2 control region by selecting dilepton events. In addition to a lepton passing the analysis selections, events must contain one of the following: a second electron or muon, an isolated track candidate, or a tau lepton candidate. The latter categories are included to accept hadronically decaying tau leptons. If all the kinematic selections used for the signal regions are applied, the number of events in CR2 is too low to validate the modeling of the dilepton top quark background. Therefore, this CR is used primarily to validate the modeling of M bb . Since the signal produces a resonant peak in the M bb distribution, the requirement on M bb is inverted to define the background-dominated control region CRMbb, which includes a mixture of all backgrounds in proportions similar to those in the signal region. Consequently, this control region is dominated by the dilepton top quark background and is used to validate the modeling of these processes in the kinematic tails of the E miss T , M T , and M CT distributions.
The CR0b region is designed to study the W + LF background. It is used to validate the modeling of the kinematic tails in E miss T , M T , and M CT for W + jets processes. In this region, the dijet mass M jj computed from the two selected jets is used in place of M bb .
The background estimation and the associated uncertainties are described in the following sections.

Dilepton top quark backgrounds
The dilepton top quark process contributes to the event sample in the signal region when the second lepton is not reconstructed or identified. Due to the presence of two neutrinos, these events tend to have higher E miss T than the single-lepton backgrounds, and their M T distribution is not bounded by the W boson mass. However, as mentioned above, the M CT requirement significantly suppresses dilepton top quark events. The modeling of this background is validated in two steps. First, the modeling of the M bb distribution is validated in CR2 ; second, the modeling in the kinematic tails of the E miss T , M T , and M CT distributions is validated in CRMbb. Distributions of M bb in CR2 and CRMbb, after the preselection level cuts defined in Table 1, are displayed in Fig. 3 (left) and Fig. 3 (right), respectively.
In CR2 , we observe agreement between data and MC, validating the modeling of the M bb distribution. We then use CRMbb at the signal region selection level to derive a scale factor for the dilepton top quark background separately in each of the analysis E miss T bins. All other background components are subtracted from the observed data yields, and the result is compared to the dilepton top quark MC prediction. Agreement is observed in the higher E miss T bin within statistical uncertainties. For the lower E miss T bin, we find fewer events in data than predicted, and we derive a scale factor of 0.72 for the dilepton top quark background in this bin. From the statistical precision of the data, we assign a systematic uncertainty of 30% in the prediction for both bins. This accounts for any effects that could impact the modeling of this background in simulation, including generator assumptions on factorization and renormalization scales, and  Table 1, comparing data to MC simulation. (right) Distribution in M bb in CRMbb after preselection level cuts defined in Table 1. The signal region range of 90 ≤ M bb ≤ 150 GeV has been removed from the plot.
PDFs, as well as experimental uncertainties in the jet energy scale, the lepton identification and isolation, trigger, and b tagging efficiencies.

W boson backgrounds
The M T requirement (>150 GeV) effectively suppresses the contribution from W + jets events. However, as discussed above, events from W + jets can still enter the M T tail due to off-shell W production or E miss T resolution effects. The control region CR0b consists mostly of W + LF events and is therefore used to validate the modeling of W + jets in the tails of all kinematic variables such as M T . Figure 4 shows the M T distributions of data and simulated events in CR0b after the preselection requirements. The data and simulation agree within uncertainties. The observed yields in data are then compared with MC predictions after applying all the kinematic requirements at signal region selection level defined in Table 1. We find agreement within statistical uncertainties. Based on the statistical precision of the data, we assign a 10% systematic uncertainty in the W + jets prediction. This procedure directly tests the W + jets background prediction in the kinematic phase space of the signal region, including experimental uncertainties in the jet energy scale, in the efficiencies for trigger, lepton identification and isolation. It also accounts for most generator assumptions. Additional uncertainties for effects not tested by this procedure are discussed below.
For the W + LF background, the uncertainty due to the b tagging requirements is evaluated by varying the b tagging efficiencies within their measured uncertainties. The uncertainty in the yield in the signal regions is 1%.
For the W + HF background the effects contributing to the kinematic tails are similar to those in W + LF. In this case the tail of the M T distribution receives contributions from off-shell W boson production and E miss T resolution effects, but also from neutrinos in semileptonic decays within the b jets. Since this last effect is accounted for in the event generation, we do not apply  Table 1. any additional correction or uncertainty for kinematic tail modeling beyond the one derived above in CR0b.
The most uncertain aspect of the prediction for the W + HF background is the estimate of its cross section relative to the W + LF process. We assign a 50% uncertainty to the normalization of this background [78]. This uncertainty is validated by comparing data to simulation in a CR with 60 < M T < 120 GeV and with one or two jets, where the dominant contribution to the event sample is from W + jets. We find that the 50% uncertainty conservatively covers differences between data and simulation as a function of the number of b jets. Finally, the uncertainty in this prediction due to the uncertainty in the b tagging efficiency is also evaluated and found to be 5%.
The effects discussed above also contribute to the tail of the M T distribution in WZ → νbb events. As a result, the tail modeling systematic uncertainty for this background is taken to be the same as those evaluated in CR0b. An additional uncertainty of 12% is applied to the normalization of the WZ → νbb background, based on the CMS cross section measurement of inclusive WZ production at 13 TeV [79]. A unique aspect of the WZ → νbb background is that M bb peaks at the Z boson mass, at the lower edge of the M bb selection used in this analysis. Uncertainties in the jet energy scale can therefore strongly impact the prediction of this background. By varying the jet energy scale within its uncertainty, we derive an uncertainty of 27% in the WZ → νbb background prediction. While this uncertainty is large, the absolute magnitude of this background remains very small in the signal region. Finally, the uncertainty in the background prediction for this process due to the uncertainty in the b tagging efficiency is 2%.

Other backgrounds
The single-lepton top quark backgrounds are highly suppressed by several of the selections applied in this analysis. Since these contain exactly one leptonically decaying W boson, the M T requirement is an effective discriminant against them. Requiring exactly two jets also suppresses the tt → + jets background, which typically has four jets in the final state. As a result, this background comprises a small fraction of the expected SM prediction in the signal region.
Isolating the single-lepton top quark background in a region kinematically similar to the signal region is difficult since dilepton top quark events tend to dominate when requiring large M T . The main source of uncertainty in the prediction of this backgrounds is the modeling of the E miss T resolution, which was found to be well modeled in the study of CR0b.
Additional studies of E miss T resolution are performed using γ + jets events following the method used in Ref. [78]. The resolution in data is found to be up to 20% worse than in simulation, leading to higher single-lepton top quark yields than expected from simulation. However, the impact of this effect on the total background prediction is negligible. Due to the difficulties in defining a dedicated control region for this process, we assign a conservative uncertainty of 100% to the single lepton top quark background prediction.
The "rare" backgrounds contribute less than 15% of the expected yield in the signal region. We apply an uncertainty of 50% on the event yields from these processes. Figure 5 shows the distributions of M bb in data compared with the SM background prediction after all signal region requirements except the M bb selection. No significant deviations from the predictions are observed. Table 2 shows the expected SM background yields in the signal region compared to the observation, as well as predicted yields for several signal models with the masses (m χ ± 1 , m χ 0 1 ) indicated in GeV. The correlation coefficient for the background prediction between the two bins is 0.61. The correlation is incorporated in the likelihood model described below for the interpretation of the results, and it can be used to reinterpret these results in other signal models [80].

Interpretation
The results of this analysis are interpreted in the context of the simplified SUSY model depicted in Fig. 1, χ ± 1 χ 0 2 → W ± H χ 0 1 χ 0 1 . The χ ± 1 and χ 0 2 are assumed to have the same mass, and the branching fractions for the decays listed above are taken to be 100%. The W and Higgs bosons are taken to decay according to their SM branching fractions. Cross section limits as a function of the SUSY particle masses are set using a modified frequentist approach, employing the CL s criterion and an asymptotic formulation [81][82][83][84]. Both signal regions are considered simultaneously in setting limits. The "expected" limit is that under the background-only hypothesis, while the "observed" limit reflects the data yields in the signal regions. The production cross sections are computed at NLO plus next-to-leading-log (NLL) precision in a limit of mass-degenerate wino χ ± 1 and χ 0 2 , light bino χ 0 1 , and with all the other sparticles assumed to be heavy and decoupled [85,86]. The uncertainty in the cross section calculation includes variations of factorization and renormalization scales, and of the PDFs.
The systematic uncertainties in the signal yield are summarized in Table 3. The signal models with the largest acceptance uncertainties are those with ∆m = m χ 0 2 − m χ 0 1 m H . For these models, the kinematic properties of the events are most similar to those from SM backgrounds, and as a result, the acceptance is smaller than for models with larger ∆m. For these models with compressed mass spectra, the largest uncertainties in the signal yields arise from the jet energy scale (up to 40%), E miss T resolution in fast simulation (up to 50%), and limited size of MC samples (up to 60%). These uncertainties reach their maximal values only for models where the acceptance of this analysis is very small and the sensitivity is similarly small. For models with large ∆m, where this analysis has the best sensitivity, these uncertainties typically amount to only a few percent. Other experimental and theoretical uncertainties are also considered and lead to small changes in the expected yields. These include effects from the renormalization and factorization scales assumed in the generator on the signal acceptance, the b tagging efficiency, the lepton reconstruction, identification, and isolation efficiency, the trigger efficiency, and the modeling of pileup. Finally, the uncertainty in the integrated luminosity is 2.5% [87]. 1-50 Figure 6 shows the expected and observed 95% confidence level (CL) exclusion limits for

Summary
A search is performed for beyond the standard model physics in events with a leptonically decaying W boson, a Higgs boson decaying to a bb pair, and large transverse momentum imbalance. The search uses proton-proton collision data recorded by the CMS experiment in 2016 at √ s = 13 TeV, corresponding to an integrated luminosity of 35.9 fb −1 . The event yields observed in data are consistent with the estimated standard model backgrounds. The results are used to set cross section limits on chargino-neutralino production in a simplified supersymmetric model with degenerate masses for χ ± 1 and χ 0 2 and with the decays χ ± 1 → W ± χ 0 1 and χ 0 2 → H χ 0 1 . Values of m χ ±

Acknowledgments
We congratulate our colleagues in the CERN accelerator departments for the excellent performance of the LHC and thank the technical and administrative staffs at CERN and at other CMS institutes for their contributions to the success of the CMS effort. In addition, we gratefully acknowledge the computing centers and personnel of the Worldwide LHC Computing Grid for delivering so effectively the computing infrastructure essential to our analyses. Finally, we acknowledge the enduring support for the construction and operation of the LHC and the CMS detector provided by the following funding agencies: BMWFW and FWF (Aus  [24] ATLAS Collaboration, "Search for gluinos in events with an isolated lepton, jets and missing transverse momentum at √ s = 13 Te V with the ATLAS detector", Eur. Phys. J. C 76 (2016) 565, doi:10.1140/epjc/s10052-016-4397-x, arXiv:1605.04285.

Institute for Nuclear Research and Nuclear Energy of Bulgaria Academy of Sciences
INFN Sezione di Firenze a , Università di Firenze b , Firenze, Italy