Impact of LSP Character on Slepton Reach at the LHC

Searches for supersymmetry at the Large Hadron Collider (LHC) have significantly constrained the parameter space associated with colored superpartners, whereas the constraints on color-singlet superpartners are considerably less severe. In this study, we investigate the dependence of slepton decay branching fractions on the nature of the lightest supersymmetric particle (LSP). In particular, in the Higgsino-like LSP scenarios, both decay branching fractions of $\tilde\ell_L$ and $\tilde\nu_\ell$ depend strongly on the sign and value of $M_1/M_2$, which has strong implications for the reach of dilepton plus MET searches for slepton pair production. We extend the experimental results for same flavor, opposite sign dilepton plus MET searches at the 8 TeV LHC to various LSP scenarios. We find that the LHC bounds on sleptons are strongly enhanced for a non-Bino-like LSP: the 95% C.L. limit for $m_{\tilde\ell_L}$ extends from 300 GeV for a Bino-like LSP to about 370 GeV for a Wino-like LSP. The bound for $\tilde\ell_L$ with a Higgsino-like LSP is the strongest (~ 490 GeV) for $M_1/M_2$ ~ $-\tan^2\theta_W$ and is the weakest (~ 220 GeV) for $M_1/M_2$ ~ $\tan^2\theta_W$. We also calculate prospective slepton search reaches at the 14 TeV LHC. With 100 fb$^{-1}$ integrated luminosity, the projected 95% C.L. mass reach for the left-handed slepton varies from 550 (670) GeV for a Bino-like (Wino-like) LSP to 900 (390) GeV for a Higgsino-like LSP under the most optimistic (pessimistic) scenario. The reach for the right-handed slepton is about 440 GeV. The corresponding 5$\sigma$ discovery sensitivity is about 100 GeV smaller. For 300 fb$^{-1}$ integrated luminosity, the reach is about 50 - 100 GeV higher.


I. INTRODUCTION
While the discovery of a Standard Model (SM)-like Higgs boson at 125 GeV has been the most significant result obtained at the Large Hadron Collider (LHC) to date [1,2], no signal for new physics beyond the SM has yet emerged. Any new colored particle would be the best targets for the LHC due to the large QCD production cross sections. Searches for hadronic final states do, however, suffer from the complicated hadronic environment. Hadronically-quiet new physics searches in leptonic final states are typically challenging due to the smaller electroweak production cross sections, yet the associated SM backgrounds are more clearly understood.
Weak scale supersymmetry (SUSY) is one of the most promising new physics scenarios, and the search for supersymmetric particles continues to be one of the main efforts of LHC studies.
LHC SUSY searches have largely focused on gluinos and squarks. The null results have set lower limits of about 1200 GeV and 800 GeV, respectively, for the masses of gluinos and degenerate first-and second-generation squarks [3]. The limits on the electroweak sector of the Minimal Supersymmetric Standard Model (MSSM), however, are much less stringent.
If low energy supersymmetry is realized in the nature, sleptons are likely to be light. This feature emerges in the Gauge Mediated SUSY-breaking scenarios [4] and the Anomaly Mediated SUSY-breaking scenarios [5], wherein the slepton masses are proportional to the electroweak gauge couplings. Even in the minimal Gravity Mediated SUSY-breaking scenarios (mSUGRA) [6] where all the scalars have a common mass m 0 at a high energy input scale, renormalization group running to low energies typically pushes up the squark mass (due to the contributions of strongly interacting gluinos) while the sleptons remain relatively light. The observation of sleptons, even in the presence of the strong lower bounds on squark and gluino masses, would be consistent with these expectations. Thus, it is timely to fully explore the discovery potential of the LHC for the lepton superpartners.
In the R-parity conserving MSSM, the lightest neutralino χ 0 1 can be a natural candidate for Weakly Interacting Massive Particle (WIMP) dark matter [7] when it is the LSP. When sleptons are light, the t-channel process χ 0 1 χ 0 1 → + − mediated by the exchange of sleptons can be important in determining the χ 0 1 annihilation cross section [8], and for fairly degenerate spectra of sleptons and χ 0 1 , coannihilation processes can also become important [9]. Therefore, discovery of the sleptons would not only provide a verification of low energy supersymmetry in nature; precise measurement of their masses could also play an important role in determining the relic density of the neutralino LSP.
Light sleptons also contribute to low energy precision observables, such as the electron and proton weak charges that can be measured in parity-violation ee Møller scattering and ep scattering [10,11], respectively, the muon anomalous magnetic moment [12], or tests of first row CKM unitarity [13]. With the precision achieved (attainable) in current (future) measurements [14,15], these low energy observables provide an indirect probe of the slepton sector that complements the LHC direct search.
Earlier studies of the slepton discovery potential at the LHC focused primarily on the Drell-Yan pair production of slepton pairs, with each slepton decaying directly to a lepton and χ 0 1 [16][17][18]. Most of those studies have been performed either in the mSUGRA framework or for a certain set of benchmark points only. Dilepton plus missing E T final states have also been searched for at the LHC. When the results are interpreted in terms of slepton Drell-Yan pair production with direct decays to a Bino-like LSP, the current limit is fairly weak: m˜ L > ∼ 300 GeV for left-handed sleptons (˜ L ) with a relatively light LSP [19,20].
Sleptons can also be produced in the cascade decay of gauginos when kinematically accessible. The gaugino pair production cross sections are typically larger than that of the direct slepton Drell-Yan process, given the fermionic nature of the gauginos. Once sleptons are lighter than gauginos, the gaugino dominantly decays to a slepton and lepton, with the slepton subsequently decaying to another lepton and the LSP. For heavier neutralinos and charginos, such lepton-rich final states greatly extends the reach of neutralino and charginos at the LHC [21,22]. In addition, imposing a sharp cut on the invariant mass distribution of two leptons produced in the χ 0 2 decay could provide further discrimination of the signal from the SM backgrounds, potentially allowing for discovery of the slepton in gaugino decays [23].
The implications of null results in the searches of neutralino/chargino decay via sleptons, however, are limited. First, such experimental searches apply only to the case when sleptons are lighter than heavier gauginos; naturally there is no sensitivity to sleptons from gaugino decays once the decay is kinematically forbidden. Second, even when sleptons are lighter than heavier gauginos, the experimental limits apply only to the case of Wino-like pair-produced gauginos with a Bino-like LSP and are, therefore, only sensitive to the stau or the left-handed slepton (˜ L ).
Finally, when the χ 0 2 and χ ± 1 are Higgsino-like states, no reach in the slepton mass can be derived even if the cascade decay is kinematically allowed, since the branching fraction into sleptons is highly suppressed by the small lepton Yukawa couplings and the small gaugino fractions of the the neutralino and chargino states.
Due to these limitations for the production of sleptons via neutralino/chargino decays, we are motivated to investigate the reach for sleptons via direct slepton Drell-Yan pair production, focusing on same flavor, opposite sign dilepton plus E T signal. Earlier studies of the slepton searches at the LHC [16][17][18] assumed a Bino-like LSP. The sensitivity of this channel, however, depends sensitively on the slepton being either left-or right-handed, as well on the composition of the LSP as being either Bino, Wino, or Higgsino dominated. Utilizing the current search channel of dilepton plus E T with data collected at the 8 TeV LHC, we re-interpret the results that have been presented by the ATLAS and CMS collaborations assuming a Bino-like LSP for cases with a Wino-like or a Higgsino-like LSP. We also study the exclusion limits and discovery reach for sleptons at the 14 TeV LHC for various choices of the LSP.
The outline of the paper is as follows. In Sec. II, we give a brief review of the slepton sector in the MSSM and discuss its dominant production and decay channels for various slepton and neutralino/chargino spectra. In Sec. III, we summarize the current limits on the slepton searches, from both LEP searches and the latest LHC results. In Sec. IV, we interpret the ATLAS results on the opposite sign dilepton plus E T search (which assume a Bino-like LSP) in the cases of Wino-like and Higgsino-like LSP, including additional production from sneutrinos in the case of the˜ L as well. In Sec. V, we study the reach for sleptons at the 14 TeV LHC. In Sec. VI, we conclude.

A. Slepton spectrum
The LHC slepton sensitivity considered here depends on both the slepton pair production cross sections and the detailed nature of the branching fractions for the slepton decays. The latter, in turn, is determined by the electroweakino (chargino/neutralino) spectrum. For simplicity, we consider the low-lying spectrum of the MSSM electroweak sector to include only sleptons, neutralinos and charginos. We also assume negligible flavor mixing between the slepton generations and zero left-right mixing of the first two generation sleptons (motivated by their small Yukawa couplings). We can then label the charged slepton mass eigenstates for the first two generations as˜ L and˜ R , for = e, µ, with masses m˜ L and m˜ R , respectively. These masses are governed by the soft breaking mass terms m SL and m SR : where the D-term contributions are ∆˜ L = (− 1 2 − sin 2 θ W )m 2 Z cos 2β and ∆˜ R = − sin 2 θ W m 2 Z cos 2β. The sneutrino masses are controlled by m SL as well and are, therefore, related to m˜ L with a small splitting introduced by electroweak effects: m 2 ν = m 2 L + m 2 W cos 2β; for the range of tan β > 1, mν < m˜ L . The phenomenology and implication of sizable flavor mixing in the slepton sector can be found in Ref. [24,25]. For the third generation charged leptons (staus), left-right mixing may be sizable, especially if tan β is large. We focus here on the first two generations of sleptons, although our approach could be adapted to the stau case as well by taking the tau tagging efficiency and stau left-right mixing into account.
The decay of sleptons depends on the composition and spectrum of neutralinos and charginos, which is set mainly by the Bino, Wino, and Higgsino mass parameters M 1 , M 2 and µ, respectively. We consider three representative cases: • Bino-like LSP: |M 1 | < |M 2 |, |µ|, yielding a neutralino LSP χ 0 1 that is Bino-like.
In the Wino-like LSP and Higgsino-like LSP cases, χ ± 1 (and χ 0 2 in the Higgsino-like LSP case) decays to the neutralino LSP with very soft jets and leptons. Identifying these decays is very difficult at the LHC. For these cases, then, the nearly degenerate neutralino and chargino states all appear as E T at the LHC.
In our discussion below, we assume the slepton decays directly to the χ 0 1 LSP (and neutralino/chargino states that are degenerate with the LSP for the Wino-or Higgsino-like LSP cases) plus one lepton, a mode that is most likely to occur when the slepton is lighter than all other heavier neutralinos and charginos. In cases when sleptons are heavier than charginos and neutralinos other than the LSP (and its nearly degenerate neutralino/chargino states), sleptons may decay into those neutralino/chargino states, which subsequently cascade decay to the LSP.
The final states from such processes are typically more complicated, involving multi-leptons, multi-jets and E T . While a slepton search relying on such slepton cascade decays is complementary to the one assuming direct decay of the slepton to the LSP plus a lepton, an analysis of the cascade decay scenario goes beyond the scope of our current study, and we leave it for future work.

B. Slepton decays
We now turn to the slepton branching fractions for the three different LSP cases. For the Bino-like LSP,˜ L and˜ R both decay to χ 0 1 , andν decays to νχ 0 1 with 100% branching fraction. For the Wino-like LSP,˜ L decays to χ 0 1 , νχ ± 1 (ν L decays to νχ 0 1 , χ ± 1 ) with branching fractions of 33% and 67%, respectively. These branching fractions are set by the √ 2 enhancement of charged current coupling relative to that of the neutral current. The˜ R decays to χ 0 1 with a branching fraction of nearly 100% via a small Wino−Bino mixing. The decay of˜ R to νχ ± 1 is highly suppressed by the small lepton Yukawa couplings.
For the Higgsino-like LSP case, due to the strong suppression of the small lepton Yukawa coupling,˜ L andν L decay to χ 0 1,2 and χ ± 1 via the Bino-and Wino-components of χ 0 1,2 and χ ± 1 . The branching fractions to χ 0 1,2 depend on the relative Bino and Wino fractions of the χ 0 1,2 : |N iB | 2 and |N iW | 2 (i = 1, 2), respectively, which are given to leading order in m Z /(M 1,2 ± µ) by: where s W = sin θ W , c W = cos θ W for θ W being the weak mixing angle; s β = sin β and c β = cos β. In arriving at these expressions, we have assumed that |M 1,2 − µ| m Z . Note that the relative sign between the Bino and Wino components of the neutralinos is physical, and has interesting consequences. Similarly, the Wino fractions of χ ± 1 are given by the absolute squares of Note that we have explicitly kept the sub-leading term µ/M 2 in the mixing coefficient since it can be important for the case of large tan β (as c β goes to zero) in U 1W − , which is relevant for L decays. The partial decay widths for the charged slepton and sneutrino decays into Higgsino-like LSPs are given approximately by where C = 1 16π for m P and m D being the the parent slepton mass and daughter neutralino/chargino mass, respectively. The "±" in Eqs. (4) and (6) correspond to χ 0 1 and χ 0 2 , respectively. Given the near degeneracy of χ 0 1,2 for the Higgsino states, the rates for decays to these two channels are usually added together since χ 0 1,2 both appear as E T at hadron colliders. In the limit of |µ| |M 1,2 |, with no dependence on tan β. Decays to charginos, however, show a different tan β dependence for˜ andν . The decay˜ → ν χ ± 1 depends on (c β + s β µ M 2 ) 2 , which decreases with increasing tan β until tan β ∼ |M 2 /µ|, when the decay branching fraction stabilizes. On the other hand, Γ(ν → χ ± 1 ) depends only weakly on tan β, since c β µ/M 2 is always small compared to s β , which changes little for large tan β. As a result, the branching fractions for˜ L show a strong tan β dependence since the total decay width varies with tan β because of˜ → ν χ ± 1 , while the branching fractions forν vary little with respect to tan β.
In Fig. 1, we show the branching fractions for charged slepton and sneutrino decays into Higgsino-like LSP χ 0 1 , as well as nearly degenerate Higgsino neutralino χ 0 2 and chargino χ ± 1 . Other parameters are chosen to be tan β = 10, m SL = 500 GeV, µ = 100 GeV and |M 2 | = 10 TeV. In this paper, we always use M 1 > 0 as our convention. In general, there exist only two physical phases involving the electroweak gaugino and Higgsino mass parameters. We assume the gaugino/Higgsino sector introduces no new CP-violation, so these phases simply amount to relative signs. We chose them to be the relative signs of M 1 and M 2 and the relative sign of µ and M 2 . As we discuss below, the choice of these phases can have a significant impact on the slepton  fractions for decays to neutralinos reach almost 100% since the decay to charginos is suppressed by the relatively small Wino fraction in χ ± 1 . For M 1 M 2 , the branching fraction for decays to neutralinos is about 90% since the decay to ν χ ± 1 is suppressed by either cos β or µ/M 2 compared to decay to neutralinos, as given in Eq. (5).
There is a notable point at M 1 /M 2 ∼ tan 2 θ W ≈ 0.3 where the decays to neutralinos vanish due to the cancellation between the contributions of the Bino and Wino fractions in the Higgsinolike neutralinos for M 1 /M 2 > 0. In this region, decay to charginos, being all that remains, is dominant.
The branching fractions forν decay are shown in Fig. 1   The˜ L → χ 0 1,2 branching fraction will not have a minimum in its decay branching fraction since the Bino-and Wino-component interfere constructively in this case. The step in the neutralino branching fraction curves near M 1 /M 2 ∼ 0.3 is due to a switchover between 1 √ 2 (H u ±H d ) as being the LSP. The branching fractions for˜ L → χ 0 1,2 almost reaches 100%, due to the relative smallness of the partial decay width for˜ L → ν χ ± 1 . Γ(ν → ν χ 0 1,2 ), on the other hand, will experience a suppression for M 1 /M 2 ∼ − tan 2 θ W , as shown in Eq. (10). While the generic features are the same as Fig. 1 for tan β = 10, the decay fraction for˜ → ν χ ± 1 is relatively larger due to the enhancement of Γ(˜ → ν χ ± 1 ) arising from the larger value of cos β. For M 1 /M 2 = 1, decay branching fractions to χ 0 1,2 and ν χ 0 1,2 are about 50% each. Similarly, for M 1 /M 2 < 0, the branching fraction of decays to neutralinos is about 80% to 100%, while the decays to charginos could be as large as 20%. that cos β is much greater than s β µ/M 2 , the˜ L → ν χ ± 1 branching fraction always decreases as tan β increases, with˜ L → χ 0 1,2 becoming dominant for tan β 10. Fig. 3   Note that in the foregoing discussion of the˜ L andν decays to Higgsino-like LSPs, we have considered the case of M 1 > 0 and µ > 0, with two different signs for M 2 . The relative sign between these three mass parameters is physical, and the behavior of the branching fractions will change when one of these parameters flips sign. For µ/M 2 < 0, decays to charginos will be    Table. I. Case (I) leads to a suppressed overall cross section for dilepton plus E T final states, while case (II) leads to an enhancement. These cases are the upper and lower boundaries of the envelope of possible signal rates in the Higgsino-like LSP scenario.

C. Slepton production and signatures
For Drell-Yan pair production of sleptons˜ L˜ L ,˜ Lν ,ν ν and˜ R˜ R with dominant direct decay of sleptons into χ 0 1 (and χ ± 1 , χ 0 2 for Wino-like and Higgsino-like LSP cases), the collider signatures are dilepton plus E T , single lepton plus E T , and E T only. The single lepton channel suffers from large SM backgrounds, mainly driven by W boson production. The E T only signature requires an extra jet or lepton from initial or final state radiation, which leads to more suppressed cross sections. Current collider searches for slepton Drell-Yan production focus on the final state of two isolated energetic leptons plus E T [21,22]. The SM backgrounds are typically large, dominantly from W W or tt. In our analyses below, we focus on the dilepton plus E T channel and reinterpret the current 8 TeV LHC slepton search limits for various LSP scenarios, as well as project the reach of the LHC at 14 TeV. In particular, we include contributions from the presence of sneutrinos for the case of left-handed sleptons, as their mass is related to the left-handed slepton mass and they can contribute to the dilepton and missing energy signature for non-Bino-like LSPs. in size and smaller than the production by about a factor of 3, while right-handed slepton pair production, due to a cancellation between the Z and γ s-channel graphs, is smaller still, about an order of magnitude less than the associated production cross section. The NLO K-factors are approximately 1.18 [33], independent of which particular pair production considered, as the QCD structure of the graphs is identical in all four cases.
In In the case where theẽ R − χ 0 1 mass splitting is small and the usual dilepton search is insensitive, a single lepton plus missing energy search yields a lower limit on mẽ R of 73 GeV, independent of m χ 0 1 [27]. For sneutrinos, a mass limit of 45 GeV can be deduced from the invisible Z decay width [28]. An indirect mass limit on sneutrinos can also be derived from the direct search limits on the charged slepton masses, but for LEP searches it is not competitive with the invisible width constraint.
Searches for first and second generation charged sleptons have been performed by both the ATLAS [19,21] and CMS collaborations [20]. With about 20 fb −1 luminosity collected at 8 TeV, both collaborations studied the signal of opposite-sign (OS) same flavor (SF) dilepton plus missing E T from the electroweak pair production of sleptons assuming a 100% decay branching fraction for˜ ± → ± + χ 0 1 . The most stringent bounds come from the ATLAS results, which exclude left-handed (right-handed) slepton masses between 95 and 310 GeV (235 GeV) at 95% C.L. for a Bino-like LSP with m χ 0 1 = 0 GeV. For larger χ 0 1 masses, the upper range of the exclusion reach does not change while the lower bound shifts approximately as 80 GeV +m χ 0 1 .

IV. RECASTING LHC 8 TEV SEARCH LIMITS
We consider the signal consisting of two same flavor, opposite sign energetic leptons (elec- final state which should not be sensitive to pile-up. The event generation procedure produces events at leading order. NLO effects are taken into account by scaling our events by an appropriate K-factor [33]. We additionally take into account various experimental efficiencies that may be poorly modeled by our crude detector simulation by scaling our signal yields to match the expected yields quoted in the experimental search [19].
Following the 8 TeV dilepton search technique at the ATLAS [19,21], we apply the following cuts: • Exactly two leptons (electron or muon) with p T > 10 GeV and |η | < 2.5. The invariant mass of the lepton pair is required to be greater than 20 GeV and to be away from the Z-pole: |m − m Z | > 10 GeV.
where ∆φ ,j is the azimuthal angle between the direction of p miss For the signal process, our simulation matches well with the ATLAS distributions for the given benchmark points after a scaling by factor of 1.25 for both di-electron and di-muon channels that accounts for both a K-factor expected to be 1.18 and differences in reconstruction efficiencies.
We use this scaling factor in all subsequent calculations for 8 TeV.
To reproduce the exclusion plot from the ATLAS paper we utilized the CL s method discussed in [37,38]. We generate signal events using Monte-Carlo to determine the signal strength over the range of parameters 90 < m˜ < 600 GeV and 25 < m χ 0 1 < m˜ − 30 GeV. For our SM backgrounds, we simply use the number of events predicted by the ATLAS experiment for each cut scenario [19]. We follow the method in the ATLAS paper where we choose the M T 2 cut that maximizes S/ √ B. We also demand that the signal to background is greater than a minimal threshold: S/B > 0.1. We reproduced the exclusion limits for the Bino-like LSP for left-and right-handed sleptons, indicated by the solid blue and orange curves, respectively, in Fig. 7 (a).
Our bounds reproduce the ATLAS search results well with slight discrepancy at low masses.
As the resulting dilepton + E T final states cross section, therefore, varies with the choice of LSP scenario. In Fig. 7 (a), we recast the current 8 TeV ATLAS slepton search results in the dilepton + E T channel for the various benchmark scenarios given in Table I. For cases where m˜ mχ + 50 GeV , we find that the Wino-like LSP scenario is excluded for slepton masses below approximately 365 GeV, while the pessimistic and optimistic Higgsino-like LSP scenarios imply exclusion of sleptons lighter than about 220 GeV and 495 GeV, respectively.
To show the dependence of limits on M 1 /M 2 in the Higgsino-like LSP scenario, we plot in precipitously to much lower masses. We also note that the low mass region of m˜ 150 GeV for tan β = 10 cannot be excluded for this LSP mass due to the loss of sensitivity for small mass splitting between the slepton and the LSP. The exclusion region for tan β = 3 is even weaker due to the suppression of the signal cross sections, with no sensitivity for any slepton masses when 0.3 < M 1 /M 2 < 0.35. In the negative M 1 /M 2 case, by comparison, the slepton mass exclusion is significantly stronger (m˜ 470 − 490 GeV) and relatively insensitive to |M 1 /M 2 | until it gets fairly small, when the m˜ reach is reduced due to the suppression of Γ(ν L → χ ± 1 ).

V. 14 TEV EXCLUSION AND DISCOVERY REACH
We now turn to projections for Run II at the LHC, with 14 TeV center-of-mass energy. As with the 8 TeV LHC analysis, we consider the dilepton + E T channel and generate the signal and background Monte-Carlo events in the same manner as in Sec. IV. For the signal process, we used the next-to-leading order production cross section for sleptons as given in Ref. [33].
Background processes are scaled with K-factors from Ref. [39]. We generate the signal over the range of parameters 200 < m˜ < 1000 GeV and 25 < m χ 0 1 < m˜ − 30 GeV. We also demand S > 2, B > 2, and S/B > 0.1.
• Optimized cuts on E T rel and M T 2 . Cuts range from 100 < E T rel < 200 GeV and 0 < M T 2 < 200 GeV in increments of 50 GeV.
In Fig. 8, we show the E T rel and M T 2 distributions for both the backgrounds and two signal benchmark points of (m˜ L , mχ0 1 ) = (500, 100) GeV and (m˜ L , mχ0 1 ) = (500, 300) GeV with a Bino-like LSP after imposing the selection cuts, jet and Z vetoes, and minimum E T rel cut of 100 GeV. We observe that the backgrounds and signal (for the larger slepton-LSP mass splitting) become comparable at E T rel and M T 2 on the order of 300 GeV, naïvely suggesting imposing cuts in the vicinity of this region. However, the background distributions tend to fall quickly at higher E T rel and M T 2 , so to obtain sufficient statistics we impose somewhat looser cuts, with the requirements on B, S and S/B listed above. The resulting, illustrative cut efficiencies are listed in Table II for the benchmark point of (m˜ L , mχ0 1 ) = (500, 100) GeV with a Bino-like LSP. The resulting, prospective 95% C.L. expected exclusion limits in the m˜ − m χ 0 1 plane are given in Fig. 9 (a) for the 14 TeV LHC with 100 fb −1 integrated luminosity for various slepton and neutralino LSP scenarios. For the right-handed slepton, the reach is about 430 GeV for small LSP masses. For the left-handed sleptons, the reach is 550 GeV (670 GeV) for the Binolike (Wino-like) LSP case, and about 400 GeV and 900 GeV for the Higgsino-like LSP cases (I) and (II), respectively. We find the reach with 300 fb −1 is typically about 50 − 100 GeV better.
A 5% systematic error has been included in our limits to give a reasonably realistic reach for the  II: Cut efficiency for benchmark point of (m˜ L , mχ0 1 ) = (500, 100) GeV with a Bino-like LSP, using cuts specified above. Signal significance is shown for the combined ee and µµ channels using 100 fb −1 of data at the 14 TeV LHC. The tt cross section before the selection cuts already include a precut of LHC.
The prospective 14 TeV exclusion reach in the m˜ − |M 1 /M 2 | plane for the Higgsino-like LSP case is shown in Fig. 9  integrated luminosity. For a very light LSP, the reach is roughly three times weaker for M 1 /M 2 = 1/3 case, while the reach for the Wino-and Bino-like LSP scenarios fall in between. Fig. 10 (b) gives the corresponding discovery potential in the m˜ L − |M 1 /M 2 | plane for a Higgsino-like LSP. While no sensitivity for slepton could be achieved for the worse case scenario of M 1 /M 2 ∼ tan 2 θ W , reaches in m˜ increases when M 1 /M 2 deviates from this value. For M 1 /M 2 < 0, 5σ reach can be as large as 800 GeV. With 300 fb −1 at the 14 TeV LHC, the reach is improved by about 50 − 100 GeV.

VI. SUMMARY AND CONCLUSION
With the absence thus far of any superpartner signals at the LHC, the attention for LHC Run-II (14 TeV) will clearly require emphasis on more difficult-to-observe signatures. Among the most challenging are those associated with sleptons, given the O(fb) electroweak production cross sections. In this work, we studied the dependence of slepton decay branching fractions on the nature of the LSP. In particular, in the Higgsino-like LSP scenarios, both decay branching fractions of L andν exhibit strong dependence on the sign and value of M 1 /M 2 :˜ L → χ 0 1,2 is minimized for M 1 /M 2 ∼ tan 2 θ W , whileν → χ ± 1 is maximized for M 1 /M 2 ∼ − tan 2 θ W . Combined with the slepton pair production, we analyzed the prospective reach for the OS dilepton plus E T final state at the 8 and 14 TeV LHC.
We recasted the existing 8 TeV results, reported by the LHC collaborations assuming a Binolike LSP, in various LSP scenarios. We find that the LHC slepton reach is strongly enhanced for a non-Bino-like LSP: the 95% C.L. limit for m L extends from 300 GeV for Bino-like LSP to about 370 GeV for Wino-like LSP. More interestingly, the reach in the Higgsino-like LSP scenario sensitively depends on the value and sign of M 1 /M 2 . The 95% C.L. reach for˜ L is the strongest (∼ 490 GeV) for M 1 /M 2 ∼ − tan 2 θ W and is the weakest (∼ 220 GeV) for M 1 /M 2 ∼ tan 2 θ W .
We also studied the 95% C.L. exclusion and 5σ discovery reach of slepton at the 14 TeV LHC with 100 fb −1 luminosity. The projected 95% C.L. mass limits for the left-handed slepton varies from 550 (670) GeV for a Bino-like (Wino-like) LSP to 900 (390) GeV for a Higgsino-like LSP under the most optimistic (pessimistic) scenario. The reach for the right-handed slepton is about 440 GeV. The corresponding 5σ discovery is about 100 GeV smaller. For 300 fb −1 integrated luminosity, the reach is about 50 − 100 GeV higher.
Interestingly, relatively light leptons with moderate tan β are needed to explain the present difference between the muon anomalous magnetic moment experimental result and the SM prediction. The LHC Run-II should, thus, be able to probe this possibility for the Wino-like and Higgsino-like LSP. The observation of a signal in this case could be consistent with a supersymmetric explanation for the g µ − 2 result 1 . In addition, one may also expect signatures in other low-energy electroweak processes, such as tests of lepton universality with pion leptonic decays or deviations from first row CKM unitarity as probed by β-decay and kaon leptonic decays. On the other hand, the non-observation of dilepton plus E T signal for slepton Drell-Yan pair production would not generally preclude light sleptons, as the rates for the right-handed sleptons and for the left-handed sleptons with a Higgsino-like LSP and M 1 /M 2 ∼ tan 2 θ W are considerably suppressed. Probing these MSSM scenarios would require alternate avenues, such as the production of sleptons via the cascade decays from electroweak gaugino production or future studies at a high energy e + e − collider.