Same-sign trileptons as a signal of sneutrino lightest supersymmetric partlcle

Contrary to common expectation, a left-sneutrinos can occasionally be the lightest supersymmet- ric particle. This has important implications in both collider and dark matter studies. We show that same-sign tri-lepton (SS3L) events at the Large Hadron Collider, with any lepton having opposite sign vetoed, distinguish such scenarios, up to gluino masses exceeding 2 TeV. The jets + M ET signal rate is somewhat suppressed in this case, thus enhancing the scope of leptonic signals.

One usually expects missing transverse energy ( / E T ) together with jets, leptons or photons as signals of supersymmetry (SUSY) with conserved R-parity at the Large Hadron Collider (LHC) [1,2]. While R-parity (defined as R = (−1) (3B+L+2S) ) ensures a stable lightest SUSY particle (LSP) as a dark matter (DM) candidate, the LSP is the lightest neutralino (χ 0 1 ) in the majority of models but the gravitino [3,4] or the axino [5] in some of them. In the latter class, SUSY signals are often characterised by hard photons from neutralino decay along with / E T .
On the other hand, it is difficult to identify a SUSY spectrum with a sneturino (ν) at the bottom of the minimal SUSY standard model (MSSM) spectrum. In general, a left-chiral sneutrino (ν L ) LSP has unsuppressed interaction with the Z-boson and is therefore strongly disfavoured from direct dark matter search experiments. However, keeping some exceptional situations in mind, it is desirable to have a way of identifying the signature of a (left) sneutrino LSP in accelerator experiments. If there is a gravitino or axino LSP, there is of course no constraint from dark matter search. But theν L , being the lightest among MSSM particles, will decay invisibly into a ν and the gravitino (axino), yielding the same MET spectrum and jets/leptons as in the case of χ 0 1 LSP. We show here that a rather striking distinction of both the scenarios mentioned above comes through same-sign trileptons (SS3L) at the LHC.
As stated above, a SUSY spectrum with a left sneutrino LSP is usually considered impossible. However, as has been shown in some recent works, aν L DM can, after all, be allowed, if there is a mass-splitting between the scalar (ν 1 ) and pseudoscalar (ν 2 ) parts ofν L =ν 1 + iν 2 . Since a Z couples toν 1ν2 , a mass splitting of a few hundred keV's prevents the scattering of the lighter ofν 1 and ν 2 (which can be the DM candidate) into the heavier one, since otherwise the dark matter candidate must have a speed exceeding its escape velocity in our galaxy [6]. The mass-splitting can be organised with, for example, a tiny Majorana neutrino mass, for which the necessary conditions on the SUSY model have been discussed in the literature [7]. It is important from the angle of SUSY search to distinguish the above situation in accelerator experiments. In addition, in the unconstrained situation where the sneutrino is lowest down in the MSSM spectrum but can further decay into a gravitino, an axino or even a right-chiral sneutrino LSP, one wishes to have a characteristic signal. Both of the above scenarios are addressed by the SS3L signal, something that is highly suppressed in R-parity conserving SUSY with the χ 0 1 as the LSP or just above a gravitino, an axino or a rightsneutrino.
The primary cause of inevitable SS3L in the scenarios discussed above is the fact that theν L states are closely spaced in mass to the charged sleptons (l L ), as dictated by SU (2) L invariance. The latter (leaving aside the stau) are slightly more massive because of D-term (and F-term) contributions. Therefore, if the lightest (gaugino-like) neutralino is the next massive state in the spectrum, it decays either to a charged slepton and an anti-lepton (or to its conjugate state) or to the left-sneutrino(s) and a neutrino with comparable branching ratio.l L then undergoes three-body decays, producing the corresponding sneutrino and two soft-jets or a soft lepton and a neutrino. Due to their small transverse momentum, these soft decay products are difficult to observe at LHC. Thus all SUSY cascades which result in the lightest neutralino can lead to multileptons, the proliferation coming from either a top or a chargino as intermediate. The Majorana nature of neutralino allows combinations of final states where one has three (or even four) leptons, all with the same sign. 1 Let us now describe the typical spectrum. Throughout our discussion we will assume the first two generations of SU (2) L doublet sleptons to be degenerate. Further, both e, µ will be termed as leptons (l), and their scalar counterparts will be denoted as sleptons (l). 1 A pathological situation that may escape detection via this signal is one where the lighter chargino is decoupled and the lighter stop is so close to to χ 0 1 that it decays only into cχ 0 1 .
From whatever hope SUSY still has for solving the naturalness problem, a stop within the reach of the LHC is desirable, either with or without accessible gluinos alongside. We include the situation where the first two families of squarks are decoupled. The decay of a pair of light (anti-)stops (t 1 ) followed by the cascade decay of the same is of our primary interest. We explore direct pairproduction oft 1t * 1 , and cascade production of the lightest stop from the decay of the gluino (g) and the first two generation squarks (q) (when they are non-decoupled). In the non-decoupled cases, one hasgg,gq andqq productions. All of these final states contribute tog production, sinceq can decay to qg. Finallyg decays to tt * 1 ortt 1 with equal branching fraction. While the decay of a pair ofg can lead to interesting signatures, it also enhances the stop production effectively.
We assume a bino-like and/or two wino-like states below the lightest stop. In order to correspond to the situation of our interest, the first two generations of SU (2) L doublet sleptons are at the bottom of the MSSM spectrum. Based on the nature of the intermediate neutralino(s), the following two distinct scenarios have been considered.
In scenario A, we have takenχ 0 1 to be bino-like. • In the simple scenario (A) with just theχ 0 1 within reach, starting from direct production oft 1 ,t * 1 it is possible to extract up to SS3L (3l/3l + 2b + 2j + / E T along with 2 soft jets). While two same-sign (anti-)leptons can be obtained from the decay ofχ 0 1 (produced from the decay oft 1 andt * 1 ); the other (anti-)lepton appears if (W + ) W − boson, produced from the decay of (t) t, decays into leptonic channels. If the (W + ) W − boson decays to hadrons, SS2L (2l/2l + 2b + 6j + / E T ) would be possible. It may be˜t noted that depending on the decay modes of both χ 0 1 and W ± various (up to four) hard multi-lepton final states would be possible. Since SS3L has the smaller background from Standard Model (SM), we will mostly focus on this.
The number of SS3L events are further enhanced in the non-decoupling gluino/squark case. Additional (anti-)stops are produced fromg decay. Also,q decays to qg and/or qχ 0 1 . Subsequent decays of these (s)particles lead to SS3L as already described. Additional parton jets are also produced. It should be noted that, SS4L is also possible if a pair ofg decays to produce a pair oftt 1 (or its conjugate) state. This is achieved when both W − (W + ), produced from the decay of two anti-top (top) quarks, decay into (anti-)leptons and bothχ 0 1 , produced from the decay of twot 1 (or their conjugate), decay into (anti-)leptons.
• In scenario (B), in addition to the bino-like neutralino, a wino-like neutralino and the corresponding chargino also remain in betweent 1 andl L in the spectrum. Consequently, an additional decay modet 1 → bχ + 1 . However, depending on the composition oft 1 , the branching ratio in this channel is determined. For example, because of its large hypercharge an R-type (SU (2) L singlet)t 1 would dominantly decay to the bino-like neutralino; while for an L-type (charged under SU (2) L )t 1 the dominant decay modes will include a wino-like neu-tralino and/or chargino.
Starting from direct production oft 1t * 1 and assuming botht 1 andt * 1 decay into charginos, it is impossible to obtain two or more leptons with same sign. However, assuming one of the stops decays into chargino and the other one decays into neutralino, it is possible to obtain SS2L, since l of either sign can be extracted from a neutralino decay. The cascade decay of a pair of gluinos, in this scenario, can lead to significant number of SS3L events. To elaborate, a pair of gluino decays to one of the following states : ttt 1t * 1 , ttt * 1t * 1 and its charge conjugate. In all these cases, SS3L can be realized if one of the stops decays into chargino and the other decays into a neutralino eventually producing SS2L as discussed; another lepton of the same sign appears when a W boson, appearing from the decay of a top quark, decays into leptons. The latter two channels (ttt * 1t * 1 and its conjugate) can also lead to SS3L when W bosons from both top quraks decay into leptonic modes, and one of the stops decay into a neutralino. SS4L can also be obtained, as before, when both the stops decay into neutralinos. However, the last two cascades discussed do not make use of the decay mode involving the chargino.
Note that the presence of a right-slepton above the neutralino(s) does not affect the signal. In Table I we consider two benchmark points, BP-A and BP-B, to illustrate scenarios (A) and (B) respectively. Note that, in scenario (B) the composition of the lightest stop, together with the composition ofχ 0 1 andχ 0 2 affects the decay products of the lightest stop. In BP-B, we have assumedt 1 to be dominantly R-type, and wino-like states lie above the bino-like neutralino with little bino-wino mixing in the neutralino sector. This ensures that the light stop decays mostly to the bino-like neutralino and a top quark. The branching ratios ofg,t 1 and the relevant neutralino(s) are tabulated in Table II. To generate the benchmark points, we have used the publicly available code SuSpect [8]. The branching ratios of the relevant sparticles have been computed using SUSYHIT [9]. Since three-body decay modes of the leftsleptons are not computed by SUSYHIT, we have used calcHEP [10] to compute the same. In the following we discuss the viability of observing such SS3L signal events in the context of our benchmark scenarios in the inclusive channels (i.e. we define SS3L+ X as our signal) at LHC focusing on the 13 and 14 TeV run for 100 f b −1 data set.
We have used Prospino [11] to compute the NLO crosssections fort 1t * 1 andgg production at LHC. MADGRAPH has been used for event generations; subsequent decays, showering and hadronization has been taken care of by PYTHIA [12]; FASTJET [13] and DELPHES [14] has been used for jet clustering (using anti-k T algorithm) and (ATLAS) detector simulation respectively. The following selec-
As mentioned earlier, the background for SS3L from Standard Model is minuscule. It has been computed using ALPGEN [15] with similar cuts mentioned above. The SM cross-section for SS3L events are 2.5 × 10 −3 fb, to which ttW contributes the most. However some contribution to the background may come from the fake standard model events. We minimize its probability by imposing the / E T cut of 100 GeV, so that the total background is indeed negligible, without really affecting the signal.
In Table III, we list the number of SS3L events (inclusive) fromt 1t * 1 direct production assuming all the other squarks and gluino are decoupled; and also from the pair production ofg when it is not decoupled. The total number of SS3L events, wheng is not decoupled, can be estimated by adding the contributions from each of these initial states (gg andt 1t * 1 ) together. Note that, similar number of SS3L events appear when three anti-leptons are demanded instead of three leptons. As the numbers indicate, prospects of observing (or constraining ) these benchmark scenarios, appears strong in the forthcoming LHC data. As already mentioned, in addition to SS3L, events with SS2L,l 2l, 2l l and 2l 2l can also be produced. If the off-shell W * (in the decay ofl L ) goes to leptons, 2 additional leptons can be produced too; but those are expected to be soft. Among all these additional channels, SS2L events typically have the least SM background. So these could compete with SS3L events, and provide stronger evidence or constraint on the spectrum of our interest. We will consider this issue in a forthcoming revision.
To conclude, in this letter we have considered an MSSM spectrum with theν L appears at the bottom. The viability of this is assured with either a split between the scalar and pseudoscalar parts ofν L or a gravitino, axino or right-chiral sneutrino lighter than it.Thanks to the close proximity betweenl L andν L states demanded by SU(2), SUSY cascades can lead to SS3L events, via decays (en route) of the top quark or a chargino. At the same time the jets + 0l + / E T signal suffers from suppres-sion, since SUSY cascades leading to chi 0 1 can now end up charged slepton and leptons. Thus the importance of leptonic SUSY signals increases, and, among them, the SS3L events bear an unmistakable stamp of the scenario withν L at the bottom.
We estimate the number of such events with 100 f b −1 of integrated luminosity from 13 and 14 TeV LHC. In addition, we observe that, the SS3L+SS2L events can exceed in importance the purely jetty final states. We leave these issues for a more detailed study.