Table Of Content

UCD-2004-16 IFIC/04-03 LPT-Orsay-04-12 SHEP-03-41 hep-ph/0401228 NMSSM Higgs Discovery at the LHC ∗ U.Ellwanger1,J.F.Gunion2,C.Hugonie3 andS.Moretti4 1 LPTHE,Universite´ deParisXI,Baˆtiment210,F091405OrsayCedex,France;2 Departmentof Physics,U.C.Davis,Davis,CA95616;3 AHEPGroup,I.deFísicaCorpuscular –CSIC/Universitatde Vale`ncia,EdificioInstitutos deInvestigacioń, ApartadodeCorreos22085,E-46071, Valencia,Spain; 4 4 SchoolofPhysics,DepartmentofPhysicsandAstronomy, UniversityofSouthampton, Southampton, 0 SO171BJ,UK 0 2 Abstract n We demonstrate that Higgs discovery at the LHC is possible in the context of the NMSSM even for a J those scenarios such that the only strongly produced Higgs boson is a very SM-like CP-even scalar 8 whichdecaysalmostentirelytoapairofrelatvely lightCP-oddstates. Incombination withothersearch 2 channels,weareonthevergeofdemonstratingthatdetectionofatleastoneoftheNMSSMHiggsbosons 1 isguaranteed attheLHCforaccumulated luminosityof300fb−1. v 8 1. Introduction 2 One of the most attractive supersymmetric models is the Next to Minimal Supersymmetric Standard 2 1 Model(NMSSM)(see[1,2]andreferencestherein)whichextendstheMSSMbytheintroductionofjust 0 one singlet superfield, S. When the scalar component of S acquires a TeV scale vacuum expectation 4 value (a very natural result in the context of the model), the superpotential term SH H generates an 0 u d / effective µH H interabction for the Higgs doublet superfieblds. Such a term is essential for acceptable h u d p phenomenology. No other SUSY model generates this crucial component of thebsbupebrpotential in as - naturalafasbhiobn. Thus,thephenomenological implicationsoftheNMSSMatfutureaccelerators should p e be considered very seriously. One aspect of this is the fact that the h,H,A,H± Higgs sector of the h MSSM is extended so that there are three CP-even Higgs bosons (h , m < m < m ), two : 1,2,3 h1 h2 h3 v CP-odd Higgs bosons (a , m < m ) (we assume that CP is not violated in the Higgs sector) and 1,2 a1 a2 Xi a charged Higgs pair (h±). An important question is then the extent to which the no-lose theorem for r MSSMHiggsbosondiscoveryattheLHC(afterLEPconstraints)isretainedwhengoingtotheNMSSM; a i.e. is the LHCguaranteed to find at least one of the h , a , h±? The first exploration of this issue 1,2,3 1,2 appeared in [3], with the conclusion that for substantial portions of parameter space the LHC would be unabletodetectanyoftheNMSSMHiggsbosons. Sincethen,therehavebeenimprovementsinmanyof thedetectionmodesandtheadditionofnewones. Thesewillbesummarizedbelowandtheimplications reviewed. However, these improvements and additions do not address the possibly important h aa → typedecaysthatcouldsuppress allothertypesofsignals [3,4]. One of the key ingredients in the no-lose theorem for MSSM Higgs boson discovery is the fact thatrelationsamongtheHiggsbosonmassesaresuchthatdecaysoftheSM-likeHiggsbosontoAAare only possible if m is quite small, a region that is ruled out by LEPby virtue of the fact that Z hA A → pair production was not detected despite the fact that the relevant coupling is large for small m . In A the NMSSM, the lighter Higgs bosons, h or h , can be SM-like (in particular being the only Higgs 1 2 with substantial WW/ZZ coupling) without the a necessarily being heavy. In addition, this situation 1 is not excluded by LEP searches for e+e− Z∗ h a since, in the NMSSM, the a can have 1,2 1 1 → → smallZh a (Zh a )coupling whenh (h )isSM-like. [Inaddition, sumrulesrequire thattheZh a 2 1 1 1 1 2 1 1 (Zh a ) coupling is small when the h WW (h WW) couplings are near SM strength.] As a result, 2 1 1 2 NMSSM parameters that are not excluded by current data can be chosen so that the h masses are 1,2 moderateinsize( 100 130GeV)andtheh a a orh a a decaysaredominant. Dominance 1 1 1 2 1 1 ∼ − → → ∗ ToappearintheProceedingsoftheLesHouchesWorkshop2003:“PhysicsatTeVColliders”,ed.F.Boudjema of such decays falls outside the scope of the usual detection modes for the SM-like MSSM h on which theMSSMno-lose LHCtheorem largelyrelies. In Ref. [2], a partial no-lose theorem for NMSSM Higgs boson discovery at the LHC was es- tablished. In particular, it was shown that the LHC would be able to detect at least one of the Higgs bosons (typically, one of the lighter CP-even Higgs states) throughout the full parameter space of the model, excluding only those parameter choices for which there is sensitivity to the model-dependent decays ofHiggsbosons toother Higgsbosons and/orsuperparticles. Here,wewilladdress thequestion ofwhetherornotthisno-losetheoremcanbeextendedtothoseregionsofNMSSMparameterspacefor which Higgs bosons can decay to other Higgs bosons. Wefind that the parameter choices such that the “standard” discovery modes fail would allow Higgs boson discovery if detection of h aa decays is → possible. (When used generically, the symbol h will now refer to h = h , h or h and the symbol a 1 2 3 willrefertoa = a ora ). Detection ofh aawillbedifficultsinceeachawilldecayprimarily tobb 1 2 → (or 2 jets if m < 2m ), τ+τ−, and, possibly, χ0χ0, yielding final states that will typically have large a b 1 1 backgrounds attheLHC. In[2]wescannedtheparameterspace,remeoevingparameterchoicesruledoutbyconstraints from LEPonHiggsbosonproduction, e+e− Zhore+e− ha[5],andeliminatingparameterchoicesfor → → whichoneHiggsboson candecaytotwootherHiggsbosons oravector bosonplus aHiggsboson. For thesurviving regions ofparameter space, weestimated thestatistical significances (N = S/√B)for SD allHiggsbosondetection modessofarstudiedattheLHC[6–9]. Theseare(withℓ = e,µ) 1)gg h/a γγ; → → 2)associated Wh/aortt¯h/aproduction withγγℓ± inthefinalstate; 3)associated tt¯h/aproduction withh/a b¯b; → 4)associated b¯bh/aproduction withh/a τ+τ−; → 5)gg h ZZ(∗) 4leptons; → → → 6)gg h WW(∗) ℓ+ℓ−νν¯; → → → 7)WW h τ+τ−; → → 8)WW h WW(∗). → → For an integrated luminosity of 300 fb−1 at the LHC, all the surviving points yielded N > 10 after SD combiningallmodes,includingtheW-fusionmodes. Thus,NMSSMHiggsbosondiscoverybyjustone detector with L = 300 fb−1 is essentially guaranteed for those portions of parameter space for which HiggsbosondecaystootherHiggsbosonsorsupersymmetric particles arekinematically forbidden. In this work, we investigate the complementary part of the parameter space, where at least one Higgsboson decays tootherHiggsbosons. Tobemoreprecise, werequire atleast oneofthefollowing decaymodestobekinematically allowed: i)h h′h′ , ii)h aa , iii)h h±h∓ , iv)h aZ , → → → → v)h h±W∓ , vi)a′ ha, vii)a hZ , viii)a h±W∓ . (1) → → → → After searching those regions of parameter space for which one or more of the decays i) viii) is − allowed, wefound that the only subregions for which discovery of aHiggs boson in modes 1) –8) was notpossiblecorrespondtoNMSSMparameterchoicesforwhich(a)thereisalightCP-evenHiggsboson withsubstantialdoubletcontentthatdecaysmainlytotwostilllighterCP-oddHiggsstates,h aa,and → (b) all the other Higgs states are either dominantly singlet-like, implying highly suppressed production rates,orrelativelyheavy,decayingtott,tooneofthe“difficult”modesi) viii)ortoapairofsparticles. − In such cases, the best opportunity for detecting at least one ofthe NMSSMHiggs bosons is to employ WW h production and develop techniques for extracting a signal for the h aa jjτ+τ− → → → (including jj = bb)process. Wehave performed a detailed simulation of the aa jjτ+τ− final state → and find that its detection may be possible after accumulating 300 fb−1 in both the ATLAS and CMS detectors. 2. Themodelandscanningprocedures WeconsiderthesimplestversionoftheNMSSM[1],wherethetermµH H inthesuperpotential ofthe 1 2 MSSMisreplacedby(weusethenotation AforthesuperfieldandAforitsscalarcomponent field) b b κ λHb H S + S3 , (2) 1 2 3 so that the superpotential is scale invariant.bWebmbake nobassumption on “universal” soft terms. Hence, thefivesoftsupersymmetry breakingterms κ m2 H2 + m2 H2 + m2S2 + λA H H S + A S3 (3) H1 1 H2 2 S λ 1 2 3 κ are considered as independent. The masses and/or couplings of sparticles will be such that their contri- butions to the loop diagrams inducing Higgs boson production by gluon fusion and Higgs boson decay intoγγ arenegligible. Inthegauginosector,wechoseM = 1TeV(atlowscales). Assuminguniversal 2 gauginomassesatthecoupling constantunificationscale,thisyieldsM 500GeVandM 3TeV. 1 3 ∼ ∼ Inthesquarksector,asparticularly relevantforthetopsquarkswhichappearintheradiativecorrections to the Higgs potential, we chose the soft masses m = m M = 1 TeV, and varied the stop Q T susy ≡ mixingparameter A2 A2 X 2 t 1 t . (4) t ≡ Ms2usy +m2t − 12(Ms2usy +m2t)! AsintheMSSM,thevalueX = √6–socalled maximalmixing–maximizestheradiative corrections t to the Higgs boson masses, and we found that it leads to the most challenging points in the parameter spaceoftheNMSSM.Weadopttheconventionλ,κ > 0,inwhichtanβcanhaveeithersign. Werequire µ > 100GeV;otherwisealightchargino wouldhavebeendetectedatLEP.Theonlypossibly light eff | | SUSY particle will be the χ0. A light χ0 is a frequent characteristic of parameter choices that yield a 1 1 lighta . 1 We have performed ea numericalescan over the free parameters. For each point, we computed the masses and mixings of the CP-even and CP-odd Higgs bosons, h (i = 1,2,3) and a (j = 1,2), i j taking into account radiative corrections up to the dominant two loop terms, as described in [10]. We eliminated parameter choices excluded byLEPconstraints [5]one+e− Zh ande+e− h a . The i i j → → latter provides an upper bound on the Zh a reduced coupling, R′ , as a function of m + m for i j ij hi aj mhi ≃ maj. Finally, wecalculated mh± and required mh± > 155 GeV, sothat t → h±b would not be seen. In order to probe the complementary part of the parameter space as compared to the scanning of Ref. [2], we required that at least one of the decay modes i) viii) is allowed. For each Higgs state, − wecalculated all branching ratios including those for modes i) viii), using an adapted version ofthe − FORTRANcode HDECAY[11]. Wethen estimated theexpected statistical significances attheLHCin allHiggs boson detection modes 1)–8)byrescaling results for theSMHiggs boson and/or the MSSM h,H and/or A. Therescaling factors are determined by R , t and b = τ , the ratios ofthe VVh , tth i i i i i i and bbh ,τ+τ−h couplings, respectively, to those of a SM Higgs boson. Of course R < 1, but t i i i i | | andb canbelarger, smallerorevendifferinsignwithrespecttotheSM.FortheCP-oddHiggsbosons, i R′ = 0 at tree-level; t′ and b′ are the ratios of the iγ couplings for tt¯and b¯b, respectively, relative i j j 5 to SM-like strength. A detailed discussion of the procedures for rescaling SM and MSSM simulation resultsforthestatistical significances inchannels1)–8)willappearelsewhere. Inoursetofrandomly scanned points, weselected thoseforwhichallthestatistical significances in modes 1) – 8) are below 5σ. We obtained a lot of points, all with similar characteristics. Namely, in theHiggsspectrum,wealwayshaveaverySM-likeCP-evenHiggsbosonwithamassbetween115and 135GeV(i.e. above theLEPlimit), whichcanbeeither h orh ,withareduced coupling tothegauge 1 2 bosons R 1 or R 1, respectively. This state decays dominantly to a pair of (very) light CP-odd 1 2 ≃ ≃ PointNumber 1 2 3 4 5 6 BareParameters λ 0.2872 0.2124 0.3373 0.3340 0.4744 0.5212 κ 0.5332 0.5647 0.5204 0.0574 0.0844 0.0010 tanβ 2.5 3.5 5.5 2.5 2.5 2.5 µeff (GeV) 200 200 200 200 200 200 A (GeV) 100 0 50 500 500 500 λ Aκ(GeV) 0 0 0 0 0 0 CP-evenHiggsBosonMassesandCouplings m (GeV) 115 119 123 76 85 51 h1 R1 1.00 1.00 -1.00 0.08 0.10 -0.25 t1 0.99 1.00 -1.00 0.05 0.06 -0.29 b1 1.06 1.05 -1.03 0.27 0.37 0.01 RelativeggProductionRate 0.97 0.99 0.99 0.00 0.01 0.08 BR(h1 bb) 0.02 0.01 0.01 0.91 0.91 0.00 BR(h1 →τ+τ−) 0.00 0.00 0.00 0.08 0.08 0.00 → BR(h1 a1a1) 0.98 0.99 0.98 0.00 0.00 1.00 → m (GeV) 516 626 594 118 124 130 h2 R2 -0.03 -0.01 0.01 -1.00 -0.99 -0.97 t2 -0.43 -0.30 -0.10 -0.99 -0.99 -0.95 b2 2.46 -3.48 3.44 -1.03 -1.00 -1.07 RelativeggProductionRate 0.18 0.09 0.01 0.98 0.99 0.90 BR(h2 bb) 0.01 0.04 0.04 0.02 0.01 0.00 BR(h2 →τ+τ−) 0.00 0.01 0.00 0.00 0.00 0.00 → BR(h2 a1a1) 0.04 0.02 0.83 0.97 0.98 0.96 → m (GeV) 745 1064 653 553 554 535 h3 CP-oddHiggsBosonMassesandCouplings ma1 (GeV) 56 7 35 41 59 7 t′ 0.05 0.03 0.01 -0.03 -0.05 -0.06 1 b′ 0.29 0.34 0.44 -0.20 -0.29 -0.39 1 RelativeggProductionRate 0.01 0.03 0.05 0.01 0.01 0.05 BR(a1 bb) 0.92 0.00 0.93 0.92 0.92 0.00 BR(a1→τ+τ−) 0.08 0.94 0.07 0.07 0.08 0.90 → ma2 (GeV) 528 639 643 560 563 547 ChargedHiggsMass(GeV) 528 640 643 561 559 539 MostVisibleoftheLHCProcesses1)-8) 2(h1) 2(h1) 8(h1) 2(h2) 8(h2) 8(h2) NSD =S/√BSignificanceofthisprocessatL=300fb−1 0.48 0.26 0.55 0.62 0.53 0.16 NSD(L=300fb−1)forWW h aa jjτ+τ−atLHC 50 22 69 63 62 21 → → → Table1: PropertiesofselectedscenariosthatcouldescapedetectionattheLHC.Inthetable,Ri = ghiVV/ghSMVV,ti = ghitt/ghSMtt and bi = ghibb/ghSMbb for mhSM = mhi; t′1 and b′1 are the iγ5 couplings of a1 to tt and bb normalized relativetothescalarttandbbSMHiggscouplings. Wealsogivetheggfusionproductionrateratio,gg hi/gg hSM, → → for mhSM = mhi. Important absolute branching ratios are displayed. For points 2 and 6, the decays a1 → jj (j 6= b) haveBR(a1 jj) 1 BR(a1 τ+τ−). Fortheheavyh3 anda2, wegiveonlytheirmasses. Forallpoints1–6, → ≃ − → thestatisticalsignificancesforthedetectionofanyHiggsboson inanyofthechannels1)–8)aretiny; thenext-to-lastrow givestheirmaximumtogetherwiththeprocessnumber andthecorresponding Higgsstate. Thelastrowgivesthestatistical significanceofthenewWW h aa jjτ+τ−[h=h1(h=h2)forpoints1–3(4–6)]LHCsignalexploredhere. → → → states,a a ,withm between5and65GeV.Thesingletcomponentofa cannotbedominantifweare 1 1 a1 1 tohavealargeh a a orh a a branching ratiowhentheh orh ,respectively, istheSM-like 1 1 1 2 1 1 1 2 → → Higgs boson. Further, when the h or h is very SM-like, one has small Zh a or Zh a coupling, 1 2 1 1 2 1 respectively, so that e+e− h a or e+e− h a associated production places no constraint on the 1 1 2 1 → → light CP-odd state atLEP.Wehave selected six difficult benchmark points, displayed inTable1. These aresuchthata χ0χ0decaysarenegligibleorforbidden. (Techniquesforcasessuchthatχ0χ0 decay 1 → 1 1 1 1 modes are important are under development.) For points 1 – 3, h is the SM-like CP-even state, while 1 for points 4 – 6 it ies he . We have selected the points so that there is some variation in thee eh and 2 1,2 a masses. The main characteristics of the benchmark points are displayed in Table 1. Note the large 1 BR(h a a )oftheSM-likeh(h = h forpoints1–3andh = h forpoints4–6). Forpoints4–6, 1 1 1 2 → withm < 100GeV,theh ismainlysinglet. Asaresult, theZh a coupling isverysmall,implying h1 1 1 1 noLEPconstraints ontheh anda frome+e− h a production. 1 1 1 1 → We note that in the case of the points 1 – 3, the h would not be detectable either at the LHC or 2 at a Linear Collider (LC). For points 4 – 6, the h , though light, is singlet in nature and would not be 1 detectable. Further, the h or a will only be detectable for points 1 – 6 if a super high energy LC is 3 2 eventuallybuiltsothate+e− Z h a ispossible. Thus,wewillfocusonsearchingfortheSM-like 3 2 → → h (h )forpoints1–3(4–6)usingthedominanth (h ) a a decaymode. 1 2 1 2 1 1 → Inthecaseofpoints2and6,thea τ+τ− decaysaredominant. Thefinalstateofinterest will 1 → bejjτ+τ−,wherethejj actually comesprimarily froma a τ+τ−τ+τ− followed byjetdecays of 1 1 → two of the τ’s: τ+τ− jj +ν′s. (The contribution from direct a jj decays to the jjτ+τ− final 1 → → state isrelatively small forpoints 2and 6.) Inwhatfollows, when wespeak ofτ+τ−, werefer to those τ’sthatareseenintheτ+τ− ℓ+ℓ− +ν′sfinalstate(ℓ = e,µ). Forpoints 1and3–5BR(a bb) 1 → → issubstantial. Therelevantfinalstateisbbτ+τ−. Nonetheless,webeginwithastudyofthebackgrounds andsignals without requiring b-tagging. Withourlatestcuts, wewillseethatb-tagging isnotnecessary to overcome the apriori large Drell-Yan τ+τ−+jets background. It is eliminated by stringent cuts for finding the highly energetic forward / backward jets characteristic of the WW the fusion process. Asa result,wewillfindgoodsignalsforall6ofourpoints. In principle, one could explore final states other than bbτ+τ− (or jjτ+τ− for points 2 and 6). However, all other channels will be much more problematical at the LHC. A 4b-signal would be bur- denedbyalargeQCDbackground evenafterimplementingb-tagging. A4j-signalwouldbecompletely swamped by QCD background. Meanwhile, the 4τ-channel (by which we mean that all taus decay leptonically) wouldnotallowonetoreconstruct theh ,h resonances. 1 2 In the case of the 2b2τ (or 2j2τ) signature, we identify the τ’s through their leptonic decays to electrons and muons. Thus, they will yield some amount of missing (transverse) momentum, pT . miss This missing transverse momentum can be projected onto the visible e,µ-momenta in an attempt to reconstruct theparentτ-direction. 3. MonteCarloResultsfortheLHC Let us now focus on the WW h aa channel that we believe provides the best hope for Higgs → → detection in these difficult NMSSM cases. (We reemphasize that the h [cases 1 – 3] or h [cases 4 – 1 2 6] has nearly full SM strength coupling to WW.) The bbτ+τ− (or 2jτ+τ−, for points 2 and 6) final state of relevance is complex and subject to large backgrounds, and the a masses of interest are very 1 modest in size. In order to extract the WW fusion 2j2τ NMSSM Higgs boson signature, it is crucial to strongly exploit forward and backward jet tagging on the light quarks emerging after the double W- strahlung preceding WW-fusion. We also require two additional central jets (from one of the a’s) and twoopposite sign central leptons (ℓ = e,µ)coming from thethe τ+τ− emerging from the decay ofthe othera. Byimposingstringent forward/backward jettagging cuts, weremovetheotherwise verylarge background fromDrell-Yanτ+τ− +jetsproduction. Intheend,themostimportant background isdue to tt production and decay via the purely SM process, gg tt¯ b¯bW+W− b¯bτ+τ− +pT , in → → → miss association withforwardandbackwardjetradiation. WehaveemployednumericalsimulationsbasedonaversionofHERWIG v6.4[12–14]modified toallowforappropriateNMSSMcouplingsanddecayrates. Calorimeteremulationwasperformedusing the GETJET code [15]. Since the a will not have been detected previously, we must assume a value 1 form . Indealing withactual experimental data, itwillbenecessary torepeat theanalysis fordensely a1 spacedm valuesandlookforthem choicethatproducesthebestsignal. Welookamongthecentral a1 a1 jets for the combination with invariant mass Mjj closest to ma1. In Fig. 1, we show the Mjjτ+τ− invariant massdistribution obtained aftercuts,butbefore b-taggingorinclusion ofK factors—theplot presented assumesthatwehavehitonthecorrectm choice. a1 LHC,√spp = 14TeV Fig.1: Weplotdσ/dMjjτ+τ− [fb/10GeV]vsMjjτ+τ− [GeV]forsignalsandbackgroundsafterbasiceventselections,but beforebtagging.Thelinescorrespondingtopoints4and5arevisuallyindistinguishable.NoKfactorsareincluded. Theselectionstrategyadoptedisamorerefined(asregardsforward/backwardjettagging)version of that summarized in [16]. It is clearly efficient in reconstructing the h (for points 1–3) and h (for 1 2 points 4–6) masses from the jjτ+τ− system, as one can appreciate by noting the peaks appearing at Mjjτ+τ− ≈ 100GeV.Incontrast, theheavyHiggsresonances atmh2 forpoints1–3andtheratherlight resonances at m for points 4–6 (recall Table 1) do not appear, the former mainly because of the very h1 poor production rates and the latter due to the fact that either the h a a decay mode is not open 1 1 1 → (points 4, 5) or – if it is – the jets and e/µ-leptons eventually emerging from the a decays are too soft 1 topass theacceptance cuts(point 6,forwhichm = 7GeVandm = 51GeV).ForallsixNMSSM a1 h1 setups, the Higgs resonance produces a bump below the end of the low mass tail of the tt¯background (see the insert in Fig. 1). Note how small the DY τ+τ− background is after strong forward / backward jet tagging. Since the main surviving background is from tt production, b tagging is not helpful. For points 2 and 6, for which the signal has no b’s in the final state, anti-b-tagging might be useful, but has notbeenconsidered here. To estimate S/√B, we assume L = 300 fb−1, a K factor of 1.1 for the WW fusion signal and K factors of 1, 1 and 1.6 for the DY τ+τ−, ZZ production and tt backgrounds, respectively. (These K factors are not included in the plot of Fig. 1.) We sum events over the region 40 Mjjτ+τ− ≤ ≤ 150GeV.(Hadweonlyincluded massesbelow130GeV,wewouldhavehadnottbackground, andthe S/√B valueswouldbeenormous. However,weareconcernedthatthisabsenceofttbackground below 130 GeV might be a reflection of limited Monte Carlo statistics. As a result we have taken the more conservative approach of at least including the first few bins for which our Monte Carlo does predict somettbackground.) For points 1, 2, 3, 4, 5 and 6, we obtain signal rates of about S = 1636, 702, 2235, 2041, 2013, and683, respectively. Thett+jetsbackground rateisB 795. TheZZ background rateisB 6. tt ZZ ∼ ∼ TheDY τ+τ− background rate isnegligible. (Weare continuing to increase our statistics to get a fully reliable estimate.) The resulting N = S/√B values for points 1-6 are 50, 22, 69, 63, 62, and 21, SD respectively. The smaller values for points 2 and 6 are simply a reflection of the difficulty of isolating and reconstructing the two jets coming from the decay of a very light a . Overall, these preliminary 1 resultsareveryencouragingandsuggestthatano-losetheoremforNMSSMHiggsdetectionattheLHC iscloseathand. 4. Conclusions In summary, we have obtained a statistically very significant LHC signal in the jjτ+τ− final state of WW fusion for cases in which the NMSSMparameters are such that the most SM-like of the CP-even Higgs bosons, h, isrelatively light and decays primarily toapair of CP-oddHiggs states, h aawith → a bb,τ+τ− if m > 2m or a jj,τ+τ− if m < 2m . Thestatistical significances are (at least) a b a b → → of order 50 to 70 for points with m > 2m and of order 20 for points with m < 2m . These high a b a b significanceswereobtainedbyimposingstringentcutsrequiringhighlyenergeticforward/backwardjets in order to isolate the WW fusion signal process from backgrounds such as DYτ+τ− pair production. Still, this signal will be the only evidence for Higgs bosons at the LHC. A future LC will probably be essential in order to confirm that the enhancement seen at the LHC really does correspond to a Higgs boson. AttheLC,discovery ofalightSM-likehisguaranteed tobepossible intheZhfinalstateusing therecoilmasstechnique [17]. In the present study, we have not explored the cases in which the a χ0χ0 decay has a large 1 → 1 1 branching ratio. Detecting a Higgs signal in such cases will require a rather different procedure. Work ontheWW h invisiblesignalisinprogress [18]. e e → → As wehave stressed, for parameter space points of the type we have discussed here, detection of anyoftheotherMSSMHiggsbosonsislikelytobeimpossibleattheLHCandislikelytorequireanLC with √se+e− above the relevant thresholds for h′a′ production, where h′ and a′ are heavy CP-evenand CP-oddHiggsbosons, respectively. Although results fortheLHCindicate thatHiggsboson discovery willbepossible forthetypeof situations wehaveconsidered, itisclearly important torefineandimprovethetechniques forextracting asignal. This willalmost certainly be possible once data isin hand and the tt background can be more completely modeled. Clearly, if SUSYis discovered and WW WW scattering is found to be perturbative at WW → energies of 1 TeV (and higher), and yet no Higgs bosons are detected in the standard MSSM modes, a carefulsearchforthesignalwehaveconsidered shouldhaveahighpriority. Finally,weshouldremarkthattheh aasearchchannelconsidered hereintheNMSSMframe- → work is also highly relevant for a general two-Higgs-doublet model, 2HDM. It is really quite possible thatthemostSM-likeCP-evenHiggsbosonofa2HDMwilldecayprimarilytotwoCP-oddstates. This is possible even if the CP-even state is quite heavy, unlike the NMSSM cases considered here. If CP violation is introduced in the Higgs sector, either at tree-level or as a result of one-loop corrections (as, for example, is possible in the MSSM),h h′h′′ decays will generally be present. The critical signal → willbethesameasthatconsidered here. Acknowledgments JFGissupportedbytheU.S.DepartmentofEnergyandtheDavisInstituteforHighEnergyPhysics. SM thanks the UK-PPARC and the Royal Society (London, U.K.) for financial support and D.J. Miller for usefulconversations. CHissupportedbytheEuropeanCommissionRTNgrantHPRN-CT-2000-00148. JFG,CH,andUEthanktheFrance-Berkeley fundforpartialsupportofthisresearch. References [1] J. Ellis, J. F. Gunion, H. Haber, L. Roszkowski, and F. Zwirner. Higgs bosons in a nonminimal supersymmetric model. Phys.Rev.,D39:844,1989. [2] U.Ellwanger, J. Gunion, and C.Hugonie. Establishing ano-lose theorem for nmssm higgs boson discoveryatthelhc. hep-ph/0111179 (2001). [3] J. Gunion, H. Haber, and T. Moroi. Will at least one of the higgs bosons of the next-to- minimal supersymmetricextensionofthestandardmodelbeobservableatlep2orthelhc? hep-ph/9610337 (1996). [4] B. A. Dobrescu, G. Landsberg, and K. T. Matchev. Higgs boson decays to cp-odd scalars at the tevatronandbeyond. Phys.Rev.,D63:075003, 2001. [5] LEPHiggsWorkingGroup. Notes2002/01 (zh)and2001/04 (ha). [6] CMSCollaboration. Expected observability of standard model higgs in cms with 100 fb−1. CMS NOTE1997/057. [7] ATLAS Collaboration. Detector and physics performance technical design report vol. ii (1999). CERN/LHCC/99-15,pages675–811. [8] D.Zeppenfeld, R.Kinnunen, A.Nikitenko, andE.Richter-Was. Measuring higgsbosoncouplings atthelhc. Phys.Rev.,D62:013009, 2000. [9] D.Zeppenfeld. Higgscouplings atthelhc. eConf,C010630:P123, 2001. [10] U.EllwangerandC.Hugonie. Massesandcouplingsofthelightesthiggsbosonsinthe(m+1)ssm. Eur.Phys.J.,C25:297–305, 2002. [11] A.Djouadi,J.Kalinowski,andM.Spira.Hdecay: Aprogramforhiggsbosondecaysinthestandard modelanditssupersymmetric extension. Comput.Phys.Commun.,108:56–74, 1998. [12] S. Moretti, K. Odagiri, P. Richardson, M. Seymour, and B. R. Webber. Implementation of super- symmetricprocesses intheherwigeventgenerator. JHEP,04:028, 2002. [13] G.Corcellaetal. Herwig6.4releasenote. hep-ph/0201201 (2001). [14] G. Corcella et al. Herwig 6: An event generator for hadron emission reactions with interfering gluons(including supersymmetric processes). JHEP,01:010, 2001. [15] SMoretti. privateprogram. [16] U. Ellwanger, J. F. Gunion, C. Hugonie, and S. Moretti. Towards a no-lose theorem for nmssm higgsdiscoveryatthelhc. hep-ph/0305109 (2003). [17] J. Gunion, H. Haber, and R. Van Kooten. Higgs physics at the linear collider. hep-ph/0301023 (2003). [18] U.Ellwanger, J.Gunion, C.Hugonie,andS.Moretti. inprogress.