Search for a CP-odd Higgs boson decaying to Zh in $\mathit{pp}$ collisions at $\sqrt{s}=8$ TeV with the ATLAS detector

Contents lists available atScienceDirect

Physics Letters B

www.elsevier.com/locate/physletb

Search for a CP-odd Higgs boson decaying to Zh in pp collisions at √

s = ^{8 TeV with the ATLAS detector}

.ATLASCollaboration^

a r t i c l e i n f o a b s t ra c t

Articlehistory:

Received16February2015

Receivedinrevisedform19March2015 Accepted24March2015

Availableonline28March2015 Editor:W.-D.Schlatter

Keywords:

BSMHiggsboson ATLAS

A searchfor aheavy, CP-oddHiggs boson, A,decaying intoa Z boson and a 125 GeVHiggsboson, h, withthe ATLAS detector atthe LHCis presented. The searchusesproton–proton collisiondata at a centre-of-massenergy of 8 TeV corresponding to anintegrated luminosity of 20.3 fb⁻¹. Decays of CP-evenh bosonstoτ τ orbb pairswiththeZ bosondecayingtoelectronormuonpairsareconsidered, as wellash→^{bb decayswiththe Z bosondecayingtoneutrinos.Noevidence fortheproductionof} an A bosoninthesechannelsisfoundandthe95%confidencelevelupperlimits derivedforσ(gg→ A)×^BR(A→^Zh)×^BR(h→^f¯f)are0.098–0.013 pbfor f=τ and0.57–0.014 pbfor f =^{b inarange} ofmA=220–1000 GeV.Theresultsare combinedandinterpretedinthecontextoftwo-Higgs-doublet models.

PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBYlicense (http://creativecommons.org/licenses/by/4.0/).FundedbySCOAP³.

1. Introduction

Afterthediscovery ofaHiggsboson attheLHCin2012[1,2], one of the most important remaining questions is whether the newly discovered particle is part of an extended scalar sector.

A CP-odd Higgs boson, A, appears in many models with an ex- tended scalar sector, e.g. in the case of the two-Higgs-doublet model(2HDM)[3].

The addition of a second Higgs doublet leads to five Higgs bosonsaftertheelectroweaksymmetrybreaking.Thephenomenol- ogy of such a model is very rich and depends on the vacuum expectationvaluesoftheHiggsdoublets, theCPpropertiesofthe Higgspotential andthevalues ofits parameters andtheYukawa couplingsof the Higgs doublets withthe fermions. In general, it is possible to accommodate in the model a Higgs boson com- patible to the one discovered atthe LHC. In the casewhere the Higgspotential of the 2HDMis CP-conserving, the Higgs bosons afterelectroweaksymmetrybreakingare twoCP-even(h and H ), one CP-odd( A) and two charged (H^±) Higgs bosons. Manythe- ories beyond the Standard Model (SM) include a second Higgs doublet,such asthe minimal supersymmetric SM (MSSM) [4–8], axionmodels(e.g.Ref.[9])andbaryogenesismodels(e.g.Ref.[10]).

SearchesforaCP-oddHiggsbosonarereportedinRefs.[11–14].

InthisLetter,asearchforaheavyCP-oddHiggsbosondecaying intoa Z boson andthe ∼^{125 GeVHiggs boson, h, is described.}

 E-mailaddress:atlas.publications@cern.ch.

The A→^{Zh decay rate can be dominantfor part of the 2HDM} parameterspace,especiallyforan A bosonmass,mA,belowthet¯t threshold.Inthiscase, the A boson isproducedmainlyviagluon fusionanditsnaturalwidthistypicallysmall:A/mAO(1%).

ThesearchisperformedformAintherange220to1000 GeV, reconstructing¹ Z → ^{decays (where} =^e, μ) withh→^{bb or} h→τ τ,aswellasZ→ννwithh→^{bb.Theselectedh bosonde-} caymodesprovidehighbranchingratiosandthepossibilitytofully reconstruct theHiggs boson decay kinematics.The reconstructed invariantmass(ortransversemass)ofthe Zh pair,employing the measuredvalueoftheh bosonmass,mh,toimproveitsresolution, isusedtosearchforasignal.

2. Dataandsimulatedsamples

ThedatausedinthissearchwererecordedwiththeATLASde- tector in proton–proton collisions at a centre-of-mass energy of 8 TeV. The ATLAS detector is described in detail elsewhere [15].

The integrated luminosity of the data sample, selecting only pe- riodswhere all relevantdetectorsubsystems were operational, is 20.3±⁰.6 fb⁻¹ [16]. The data used in the τ τ and bb fi- nal states were collected using a combinationof single-electron, single-muon,dielectron(ee)anddimuon(μμ)triggers.Depending

1 ThroughoutthisLetter,thenotationh→bb,h→τ τ, Z→ννand Z→ is usedforh→bb,¯ h→τ⁺τ⁻,Z→νν¯andZ→ ⁺⁻,respectively.

http://dx.doi.org/10.1016/j.physletb.2015.03.054

0370-2693/PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBYlicense(http://creativecommons.org/licenses/by/4.0/).FundedbySCOAP³.

onthetriggerchoice,the pT2 thresholdsvaryfrom24to60 GeV forthe single-electron andsingle-muon triggers, andfrom 12to 13 GeVfortheee and μμtriggers.Thedatausedinthe ννbb final statewerecollected withamissingtransversemomentum(E^miss_T ) triggerwithathresholdofE^miss_T >80 GeV.

Signaleventsfromanarrow-width A bosonproducedviagluon fusionaregeneratedwithMadGraph5[17]forallfinalstatescon- sidered in this search. The parton showering is performed with PYTHIA8[18,19].

ProductionofW andZ bosonsinassociationwithjetsissimu- latedwithSHERPA[20].Top-quarkpairandsingletop-quarkpro- ductionissimulatedwithPOWHEG[21–23]andAcerMC[24].Pro- ductionofWW,WZ,andZZ dibosonsaresimulatedusingPOWHEG. The WZ and ZZ processes include the production of off-shell Z bosons ( Z^∗) and photons (γ^∗). Triboson production (WWW^(∗), ZWW^(∗), ZZZ^(∗)) andtop pair production inassociation witha Z bosonare generatedwithMadGraph5.Finally,the productionof theSMHiggsbosoninassociationwitha Z bosonisconsideredas abackgroundinthissearch.ItissimulatedusingPYTHIA8.

The CTEQ6L1 [25] set of parton distribution functions was used forsamplesgenerated withMadGraph5 andPYTHIA8. The CT10[26]setwasusedfortheothersamples.

All generated samples are passed through the GEANT4-based [27]detectorsimulationoftheATLASdetector[28].Thesimulated eventsareoverlaidwithminimum-biasevents,toaccount forthe effectofmultipleinteractionsoccurringinthesameandneighbor- ing bunchcrossings(“pile-up”).Theeventsarereweightedsothat theaveragenumberofinteractionsperbunchcrossingagreeswith thedata.

The background estimation in this search for most processes is basedon data driven techniques,butin some cases onlysim- ulatedsamples are used.In that case, thesimulated samplesare normalizedusingtheoreticalcrosssectioncalculations.Inparticu- lar,for dibosonproduction bothqq¯ [29] and gg [30,31] initiated processesare included.Triboson productionfollowsRef. [32] and toppairproductioninassociationwithaZ bosonfollowsRefs.[33, 34].SMHiggsbosonproductioninassociationwithaZ bosonuses acalculationdescribedinRef.[35].

3. Objectreconstruction

Electronsareidentifiedfromenergyclustersintheelectromag- neticcalorimeterthatarematchedto tracksinthe innerdetector [36].Electronsarerequiredtohave|η_|<2.47 andpT>7 GeV.Iso- lationrequirements,definedintermsofthecalorimetricenergyor thepToftrackswithinconesaroundtheobject,aswellasquality requirementsareappliedtodistinguishelectronsfromjets.

Muons are reconstructed by matching tracks reconstructed in theinner detectorto tracksortracksegments inthemuon spec- trometer systems [37]. The muon acceptance is extended to the region 2.5<|η|<2.7, which is outside the inner detector cov- erage,using onlytracks reconstructed inthe forwardpart ofthe muon detector. Muons used forthis search must have |η_|<2.7, pT>6 GeV andarealsorequiredtopassisolationrequirements.

Jets are reconstructed using the anti-k_t algorithm [38] with radius parameter R=⁰.4 and pT>20 GeV (pT>30 GeV) for

|η_|<2.5 (2.5<|η_|<4.5). Low-pT jetsfrompile-up are rejected

2 ATLASusesaright-handedcoordinatesystemwithitsoriginatthenominalin- teractionpoint(IP)inthecentreofthedetectorandthez-axisalongthebeampipe.

Thex-axispointsfromtheIPtothecentreoftheLHCring,andthey-axispoints upward.Cylindricalcoordinates(r,φ)areusedinthetransverseplane,φbeingthe azimuthalanglearoundthebeampipe.Thepseudorapidityisdefinedintermsof thepolarangleθasη= −^{ln tan}(θ/2).Transversemomentaarecomputedfromthe three-momenta,p,aspT= |^p|^sinθ.

witharequirementonthescalarsumofthepT ofthetracksasso- ciatedwiththejet:forjetswith|η_|<2.4 and pT<50 GeV,tracks associated withtheprimary vertex³ must contribute over50% to thesum.Jetsfromthedecayoflong-livedheavy-flavor hadronsare selected using a multivariate tagging algorithm (b-tagging) [39].

The b-tagging efficiency is 70% forjets fromb-quarks in a sam- pleofsimulatedt¯t events.

Hadronic decays of τ leptons (τhad) [40] are reconstructed starting from clusters of energy in the calorimeter. A τhad can- didate must lie within |η_|<2.47, have a transverse momentum greater than 20 GeV, one or three associated tracks and a to- tal charge of ±^1. Information on the collimation, isolation, and shower profileiscombinedintoamultivariatediscriminanttore- ducebackgroundsfromquark- orgluon-initiatedjets.Dedicatedal- gorithmsthatreducethenumberofelectronsandmuonsmisiden- tified ashadronic τ decaysare applied.In thisanalysis,two τhad identificationselectionsareused—“loose”and“medium”—with efficienciesofabout65%and55%,respectively.

The missing transverse momentum (E^miss_T ) iscomputed using fullycalibratedandreconstructedphysics objects,aswellasclus- tersofcalorimeter-cellenergydepositsthatarenotassociatedwith anyobject[41].Inaddition,atrack-basedmissingtransversemo- mentum (p^miss_T ) is calculated asthe negative vector sum of the transverse momentaof trackswith|η|<2.4 and associatedwith theprimaryvertex.

4. SearchforA→^{Z h withh}→τ τ

In thesearch for A→^Zh→ τ τ, threechannelsare consid- ered, distinguishedbythewaythe τ τ pairdecays:two τ leptons decaying hadronically (τ_hadτ_had), one leptonic and one hadronic decay (τlepτhad) and, finally, two leptonic decays (τlepτlep). Elec- trons in the τhadτhad and τlepτhad channels are rejected in the transitionregion betweenthe barrelandend-capof thedetector (1.37<|η_|<1.52). Muons inthe τhadτhad and τlepτhad channels areconsideredonlyfor|η|<2.5.

The resolutionofthereconstructed A bosonmassisimproved usingamass-differencevariable,

m^rec_A =^mτ τ−^m−^mτ τ+^mZ+^mh,

wheremZ isthemassofthe Z boson,mh=^{125 GeV isthemass} oftheCP-evenHiggsboson,m_ istheinvariant massofthetwo leptons associated with the Z boson decay, and mτ τ denotes the τ τ invariant mass. The value of mττ , the invariant mass of the τ’s,is estimatedwiththe MissingMass Calculator(MMC) [42]. The mass resolutionfor all τ τ channelsranges from 3% at mA=220 GeV to5%atmA=1 TeV.

4.1. τhadτhad

Eventsintheτhadτhadchannelarerequiredtocontainexactly twoopposite-signleptons(ee or μμ)andexactlytwoopposite- sign τhad.The pT requirementsfortheseobjectsare pT>26 GeV (15 GeV) for the leading (subleading) electron, p_T>25–36 GeV (10 GeV) for the leading (subleading) muon, depending on the trigger, and pT>35 GeV (20 GeV) for the leading (subleading)

τhad candidates. The τhad candidates are required to satisfy the

“loose” τhadidentificationcriterion.Inaddition,theee/μμinvari- ant mass and the τ τ invariant mass have to lie in the ranges 80<m_<100 GeV and 75<mττ <175 GeV.Finally, the p_T of thepair, p_T^Z,isrequiredtobe:

3 Theprimary vertexistaken tobethereconstructedvertexwith thehighest p²_Toftheassociatedtracks.

p_T^Z>

_{125 GeV}

,if m^rec_A >400 GeV 0.64×^mrec_A −^{131 GeV},otherwise.

This requirement maximizes the sensitivity over the whole ex- ploredA massrange.Intheregionofp_T^Z>125 GeV,thereislittle background present, so tightening the requirement results in no additionalincreaseinsensitivity.Thetotalacceptancetimesselec- tionefficiencyvariesfrom6.2%,formA=^{220 GeV,toaround18%}

forthehighest A bosonmassesconsidered.

The dominant background for this channel originates from events where one or both of the τhad’s is a misidentified jet (“fake-τhadbackground”).Thisbackgroundisdominatedby Z+^jets events, withsmall contributions from dibosons and events with topquarks,anditisestimatedusingatemplatemethod.Theshape of the fake-τhad background is taken from a control region (the

“templateregion”)thatcontainseventssatisfyingalltheτhadτhad selection criteria apart from the requirements for an opposite- sign τhadτhadpairandthe τhadidentificationcriteria.Thefake-τhad backgroundisnormalizedbyusingtwoadditionalcontrolregions.

Thefirst region,“A”,contains eventsthat satisfy thesignal selec- tioncriteria,withtheexceptionthatthemττ constraintisinverted, i.e.mττ<75 GeV ormττ>175 GeV.Thesecondregion,“B”,con- tainseventsthatsatisfyallthetemplate selectioncriteria,withthe exceptionthatthemττ constraintisinverted,asintheregion“A”

definition.Theratio ofthe numberofevents in“A” tothe num- berofeventsin“B”isusedtoscalethetemplateregioneventsin ordertoobtainthenormalizationofthefake-τhadbackground.

Inadditiontothefake-τhadbackground,therearealsocontribu- tionsfrombackgroundswithrealτhadτhadobjects intheevent.

These backgrounds come primarily from Z Z^(∗) production.⁴ SM HiggsbosonproductioninassociationwithaZbosonisestimated usingsimulation,andcontributes17%ofthetotalbackground.

4.2.τlepτhad

Eventsintheτlepτhadchannelarerequiredtocontainexactly threelightleptons, μμμ,eμμ,eeμoreee,andexactlyone τhad. The pT requirementsfortheseobjectsare pT>26 GeV (15 GeV) forthe leading(remaining) electron(s), p_T>25–36 GeV (10 GeV) for the leading (remaining) muon(s), depending on the trigger, and pT>20 GeV for the τhad. Subsequently, all the possible 

pairsthat arecomposed ofopposite-sign,same-flavor leptonsare selected. From these pairs, the pair that has the invariant mass closesttom_Z isconsideredtobetheleptonpairfromthe Z boson decay.Thethirdlightleptonisconsideredtobetheleptonic τ de- cay,anditisusedalongwiththe τhad todefinethe τlepτhadpair.

Thislightleptonisrequiredtohaveopposite-signchargewithre- spect to the τhad.In addition, the τhad is required to satisfy the

“medium” τhadidentificationrequirement, andm_ andmττ have tolieintheranges80<m<100 GeV and75<_mττ<175 GeV.

Thetotal acceptancetimesselection efficiencyvariesfrom6%for m_A=^{220 GeV,toaround17%forthehighestA bosonmassescon-} sidered.

Abouthalfofthetotalbackgroundforthischannelcomesfrom eventswherethe τhadand/orthelightleptonisamisidentifiedjet (“fake-τ/background”).Thisbackgroundisdominatedbydiboson and Z+jets eventsandit isestimatedusingatemplatemethod.

The shape of the fake-τ/ background is taken from a control region (the “template region”) that contains events satisfying all

τlepτhad selection criteria, apart from requiring “medium” τhad identification criterion and opposite-sign charge for the τlepτhad pair. The fake-τ/ background is normalized by using two addi-

4 ThenotationZ Z^(∗)isusedheretoincludeZ Z ,Z Z^and Zγ^.

tionalcontrolregions,definedsimilarlytothoseintheτhadτhad channel.

Theother halfofthebackgroundcomes fromeventswithreal

τlepτhad objects in theevent. These backgrounds comeprimar- ilyfromZ Z^(∗)production.Thereisalsoasmall(11%)contribution fromtheSM Higgsbosonproduction inassociationwitha Z bo- son,whichisestimatedusingsimulation.

4.3. τlepτlep

Eventsintheτlepτlep channelarerequiredtocontainatleast four leptons, which formone same-flavor and opposite-sign pair consistent with the Z mass (80<m_<100 GeV), andeither a same-flavor ordifferent-flavor pair withan invariant massrecon- structed withthe MMC algorithm, consistent with a decay from theCP-evenHiggsboson (90<_mττ <190 GeV). Onemuon isal- lowedtobe reconstructedintheforwardregion (2.5<|η_|<2.7) of the muon spectrometer, orto be identified inthe calorimeter with pT>15 GeV and|η_|<0.1[37].The highest-pT lepton must satisfy p_T>20 GeV, and the second (third) lepton in p_T order must satisfy pT>15 GeV (pT>10 GeV). Among all the possible leptonquadrupletsinaneventtheoneminimizingthesumofthe massdifferenceswithrespecttoboththe Z and h bosonsischo- sen.

Twodifferentanalysiscategoriesaredefinedbasedonthelep- tonflavors intheHiggsbosondecay:ee or μμ(SF),andeμ(DF).

The expected background is very different in the two cases. For theSFchannel,thebackgroundisdominatedby Z Z^(∗) production withZ→^ee/μμdecays.FortheDFchannel,themainbackground is fromthe Z Z^(∗) process through the Z→τ_lepτ_lep decay chain, butother backgroundsarealso important.The signal-to-noise ra- tiointheSF categoryisimprovedbyusinga setofrequirements specificallytargetedtosuppressthemain Z Z^(∗)background.First, avetoonthe on-shellproductionof Z bosonpairsis introduced, requiringtheinvariantmassoftheh boson leptonstolie outside the Z peak:mh<80 GeV ormh>100 GeV.Backgroundeventsare characterized by low missing transversemomentum andare fur- therrejectedbyrequiringE_T^miss>30 GeV,andtheazimuthalangle betweenthe E^miss_T directionandthe Z boson transversemomen- tumtobegreater than π/2.Furthermore,a requirementthat the highest-pT lepton ofthe pairassociated withtheh bosonhas pT>15 GeV is applied, since it is found to be effective against backgroundsfrom Z+jets production.Thetotalacceptancetimes selectionefficiencyvariesfrom6.5%(1.5%)forDF(SF)channelfor mA=^{220 GeV,toaround20%forbothchannelsforthehighest A} bosonmassesconsidered.

Thesubleadingcontributionstothebackgroundarefromdibo- sonandtribosonproduction,t¯t productioninassociationwitha Z boson,andSM Higgs bosonproduction.All thesearedetermined fromsimulationandamounttoabout95%(65%)ofthetotalback- groundintheSF(DF)category.Theotherbackgroundeventshave atleastoneleptonwhichisamisidentifiedjetoraleptonfroma heavy-flavor quarkdecayandaredominatedby Z +^jetsproduc- tion,withasmallercontributionfromtop-quarkproduction.These backgrounds are estimated using a control region where one or bothoftheleptonsinthepairassociatedwiththeh→τlepτlep decayfailtosatisfy theisolation criteria.Aftersubtractionofgen- uine sources of four-lepton eventsusing simulation,the dataare extrapolatedtotheisolated signal regionusingnormalizationfac- torsderivedfromsimulatedsamples.

4.4. Systematicuncertaintiesandresults

Themostimportantsystematicuncertaintyforthebackgrounds withrealτ τ objectsinthe τlepτhadand τlepτlepchannelscomes