Combined search for anomalous pseudoscalar HVV couplings in VH(H→bb‾) production and H → VV decay

Contents lists available atScienceDirect

Physics Letters B

www.elsevier.com/locate/physletb

Combined search for anomalous pseudoscalar HVV couplings in VH ( H → ^bb ) production and H → ^{VV decay}

.TheCMS Collaboration^

CERN,Switzerland

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

Articlehistory:

Received13February2016

Receivedinrevisedform15May2016 Accepted2June2016

Availableonline7June2016 Editor:M.Doser

Keywords:

CMS Physics Higgs BSM

A search for anomalous pseudoscalarcouplings of the Higgs bosonH to electroweak vector bosons V (=^{W orZ)inasampleof}proton–protoncollisioneventscorresponding toanintegratedluminosity of18.9 fb⁻¹atacenter-of-massenergyof8 TeV ispresented.Eventsconsistentwiththe topologyof associatedVH production,wheretheHiggsbosondecaystoapairofbottomquarksandthevectorboson decays leptonically,are analyzed.The consistencyof data with a potential pseudoscalar contribution to theHVV interaction,expressedbytheeffectivepseudoscalarcrosssectionfractions fa₃,isassessed by meansofprofilelikelihoodscans. Resultsare givenfor theVH channelsaloneand foracombined analysisoftheVH andpreviouslypublishedH→VV channels.Undercertainassumptions, f_a^ZZ₃ >0.0034 isexcludedat95%confidencelevelinthecombination.Scenariosinwhichtheseassumptionsarerelaxed arealsoconsidered.

©^{2016TheAuthor.PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBYlicense} (http://creativecommons.org/licenses/by/4.0/).FundedbySCOAP³.

1. Introduction

The observation of a new boson [1–3] with a mass around 125 GeV and properties consistent with those of the standard model(SM)Higgsboson[4–10]hasusheredinaneweraofpreci- sionHiggsphysics.TheATLASandCMSCollaborations attheCERN LHC havebegun a comprehensive studyof the boson properties.

The spin-parity ofthe Higgs bosonhas been studiedin H→^ZZ, Zγ^∗, γ^∗γ^∗→⁴,H→^WW→ νν,andH→γ γ decays[11–16], where  isan electron ormuon. The CDFand D0 Collaborations have set limits on the pp→VH production cross section (with V=^{W or Z)attheTevatron,fortwo exotic}spin-parity modelsof theHiggsboson[17].Inallcases,thespin-parity J^{C P} oftheboson hasbeenfoundtobeconsistentwiththeSMprediction.Basedon astudyofanomalous couplingsinH→^ZZ→⁴decays,theCMS Collaboration hasexcluded thehypothesis ofa pure pseudoscalar spin-zero boson at 99.98% confidence level (CL), while an effec- tive pseudoscalarcrosssection fraction f_a^ZZ₃ >0.43 is excluded at 95%CL(assuminga positive,realvaluedratioofscalarandpseu- doscalarcouplings) [15].Under thesameassumptions, theATLAS Collaboration hasexcluded f_a^ZZ₃ >0.11 at95%CL[18].

We present here the first search for anomalous pseudoscalar HVV couplings atthe LHC in the topology of associated produc- tion,VH.ItwillbeshownthattheVHchannelsarestrongprobes

 E-mailaddress:[email protected].

of the structure of the HVV interaction, with sensitivity even to smallanomalous couplings.The ultimateLHC sensitivitytoa po- tential pseudoscalar interaction in these channelsis expected to greatly exceed that of H→^{VV [19]. Due to the highly off-shell} natureofthepropagatorinVHproduction,smallanomalouscou- plings can lead to significant modifications of cross sectionsand kinematic features. In particular, the propagator mass, measured bytheVHinvariantmass,m(VH),ishighlysensitivetoanomalous HVV couplings[20].

Results from the VH channels are ultimately combined with thosefromH→VV measurements[15].Theqq→^VH→^{Vbb and} gg→^H→VV processesinvolvetheYukawafermioncouplingHff andthesameHVV coupling,assuminggluon fusionproductionis dominated by the top-quark loop. The dominance of the gluon fusion production mechanism of the Higgs boson at the LHC is supported byexperimental measurements [4–10].Itisinteresting to consider modelswhere theratio ofthe Hbb and Htt coupling strengths intheVHandH→VV processesis notaffectedby the presenceofanomalouscontributions[21].Insuchacase,itispos- sibletorelatethecrosssectionsofthetwoprocessesforarbitrary anomalousHVV couplingsandperformacombinedanalysisofthe VHandH→VV processes,exploitingbothkinematicsandtherel- ative signal strengths of the two processes. The H→^{VV signal} strengthisrelativelywellmeasuredandcanprovideastrongcon- straintontheVHsignalstrength.Formodestvaluesof f_a^ZZ₃,theVH signalstrengthisconstrainedtolargevalues.Theaddedconstraint http://dx.doi.org/10.1016/j.physletb.2016.06.004

0370-2693/©^{2016TheAuthor.PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBYlicense}(http://creativecommons.org/licenses/by/4.0/).Fundedby SCOAP³.

thereby significantly improves the sensitivity to anomalous cou- plings.

Inthe following, we consider only the interactions of a spin- zero boson with the W and Z bosons, for which the scattering amplitudeisparameterizedas

A(HVV)∼



a^HVV₁ +κ₁^HVVq²_V

1+κ₂^HVVq²_V

 2

^HVV₁ 2

 m²_V

1_V^∗₁_V^∗₂

+^aHVV2 f_μν^∗(1)f^∗(2)μν+^aHVV3 f_μν^∗(1)˜_f^∗(2)μν, (1) wherethea^HVV_i arearbitrarycomplexcouplingparameterswhich candependontheV1 andV2 squaredfour-momenta,q²_V

1 andq²_V

2; f⁽ⁱ⁾μν isthefield strengthtensorofagaugebosonwithmomen- tumqV_i andpolarizationvector V_i,givenby _V^μ_iqν_V

i−_V^ν

iqμ Vi; ˜fμν is⁽ⁱ⁾ thedualfield strength tensor,givenby ¹₂_{μνρσ f}⁽ⁱ⁾ρσ ;mV1 isthe pole mass of the vector boson; and ^HVV₁ is the energy scale wherephenomena not includedin theSM become relevant [19].

The a^HVV₁ , κ_i^HVV and a^HVV₂ terms represent parity-conserving in- teractions of a scalar, while the a^HVV₃ term represents a parity- conserving interaction of a pseudoscalar. In the SM, a^HVV₁ =^2, whichis theonly nonzerocoupling attreelevel. Allother terms inEq.(1)aregeneratedwithintheSMby loop-inducedprocesses atlevelsbelowcurrentexperimentalsensitivity.Therefore,anyev- idenceforthesetermsintheavailable datashouldbeinterpreted asevidenceofnewphysics.

Wesearchforananomalousa^HVV₃ termoftheHVV interaction, assumingthatthe κ_i^HVV anda^HVV₂ termsarenegligible.Throughout theremainder of the paper,the term“scalar interaction” will be used todescribe thea^HVV₁ term. The effective pseudoscalarcross sectionfractionforprocess j (WH,ZH,WW,orZZ)isdefinedas

f_a^j₃= a^HVV₃ ²σ₃^j

a^HVV₁ ²σ₁^j₊a^HVV₃ ²σ₃^{j, (2)}

where σ_i^j is the production cross-section for process j with a^HVV_i =^{1 and all other couplings assumed to be equal to zero.}

A superscript is not included when making a general statement notrelatedtoaparticularprocess.Thepurelyscalar(pseudoscalar) casecorrespondsto fa3=^{0 ( f}a3=^{1).Thesignalstrengthparame-} ter μ^j forprocess j canalsobedefinedintermsofthea^HVV_i as

μ^j=a^HVV₁ ²σ₁^j+a^HVV₃ ²σ₃^j

^aHVV1,SM²σ₁^j

. (3)

Foragivensetofcouplingconstants, thephysicalobservables f_a^j₃ and μ^j varyfordifferentprocessesasaresultofthedependence onthe σ_i^j.The f_a^ZH

3 and f_a^WH

3 variables are defined withrespect to the ZH and WH production cross-sections in √

s=8 TeV pp collisions, whereas the f_a^VV₃ variables are definedwithrespect to the cross-section times branching fraction for the corresponding pp→^H→VV process. Inthe lattercase, thedependence onthe pp→H cross-sectioncancels.

2. TheCMSdetector

The central feature of the CMS apparatus is a superconduct- ingsolenoidof 6 m internaldiameter,providing amagnetic field of3.8 T.Withinthesolenoidvolume area siliconpixelandstrip tracker,aleadtungstatecrystalelectromagneticcalorimeter,anda

brassandscintillatorhadroncalorimeter,eachcomposedofabar- relandtwo endcapsections.Extensive forwardcalorimetry com- plements the coverage provided by thebarrel andendcapdetec- tors.Muonsaremeasuredingas-ionizationdetectorsembeddedin the steel flux-return yoke outside the solenoid. A more detailed descriptionofthe CMSdetector,togetherwitha definitionofthe coordinatesystemused andtherelevant kinematicvariables,can befoundinRef.[22].

3. Analysisstrategy

The analysisisbased ona datasample ofpp collisions corre- spondingto an integratedluminosity of18.9 fb⁻¹ ata center-of- massenergyof8 TeV,collectedwithsingle-electron,single-muon, and double-electrontriggers. The final states considered are νjj andjj (wherej representsa jet),targetingtheWH andZHsig- nalsrespectively.

The trigger, object and event selection criteria, and back- ground modeling are identical to those of Ref. [23]. Using the selected events,the two-dimensional template method described inRef.[15]isusedtodetermine f_a3 confidenceintervals.Thedis- criminantoftheboosteddecisiontree(BDT)describedinRef.[23]

servesasonedimensionofthetemplates.ThisBDTistrainedsep- aratelyforthe WH andZHchannelstoexploit various kinematic features typical of signal and background, and the correlations amongobservables. Theb-tagging likelihooddiscriminants of the jets used to construct the Higgs boson candidate, the invariant massoftheHiggsbosoncandidate,andtheangularseparationbe- tweenfinalstateleptonsandjetsarethemostimportantvariables intermsofbackgroundrejection.Althoughinitiallytrainedtosep- arate background from a scalar Higgs boson signal, it has been demonstratedwithsimulatedeventsthattheBDTisalsoeffective for signalswith anomalous fa3 values. The second dimension of the templates ism(VH).Effectively, the BDT dimension provides abackground-depletedregion athighvaluesoftheBDT discrimi- nantwithwhichtotestvarioussignalhypothesesusingthem(VH) distribution.

Signaltemplatesinthe^x= {^BDT,m(VH)}^planeareconstructed forarbitraryvaluesof fa₃ fromalinearsuperpositionoftemplates representingthe pure scalar(P0⁺

x

) andpseudoscalar(P0⁻

x ) hypotheses anda template (P₀^int+,0⁻

^x; φa₃

) that accountsfor in- terferencebetweenthea^HVV₁ anda^HVV₃ termsinEq.(1),asfollows:

Psig

x;^fa₃, φa₃

= 1−^fa₃

P0⁺

x

+ ^fa₃P0⁻

x

+

f_a3 1− ^fa3

P₀^int+,0⁻

x; φa3

. (4)

The phase between the a^HVV₁ and a^HVV₃ couplings is represented by φa3. The interference contributions to the BDT discriminant andm(VH)distributionsare negligible,asverifiedwithsimulated events. Therefore the last term in Eq. (4) is ignored in the VH channels. Equation (4) is also used to parameterize the H→^VV signals. Anomalous couplings that result from loops with parti- clesmuchheavierthan theHiggsbosonarerealvalued, allowing phasesof0and π.IntheH→VV channels,we assume φa3=^0.

The resulting templates are used to perform profile likelihood scans [24] to assessthe consistency ofvarious signal hypotheses withthedata.One-dimensionalprofilelikelihoodscansof fa3 are performed(where μisprofiled),aswellastwo-dimensionalscans inthe μversus f_a3 plane.

In order to combine channels that depend on the a^HZZ_i with thosedepending onthea^HWW_i , someassumptionon therelation- shipbetweenthecouplingsisrequired,andcustodialsymmetryis assumed(a^HZZ₁ =^aHWW₁ ^{).It isfurther assumedthat aHWW}₃ =^aHZZ₃ ^.

Table 1

σ1/σ3 cross sectionratioscalculated with JHUGen.

Process σ1/σ3

WH 0.0174

ZH 0.0239

WW 3.01

ZZ 6.36

Table 2

Valuesof ^i,jwhichrelatethechan- nelsstudiedinthispaper,asdefined inEq.(7).

i, j ^i,j

ZH, WH 1.37

ZZ, WW 2.11

ZZ, ZH 266

WW, WH 173

Withtheseassumptions,the fa3 and μvaluesintheWH andZH channelsarerelatedby

f_a^WH

3 =



1+ ¹

^ZH,WH

 1

f_a^ZH₃ −¹ ₋1

(5)

and

μ^WH=μ^ZH1+f_a^ZH₃ 

^ZH,WH−1 

, (6)

where

^ZH,WH= σ₁^ZH/σ₃^ZH

σ₁^WH/σ₃^{WH. (7)}

The σ1/σ3 ratiosgivenby the JHUGen 4.3[19,25,26]eventgener- atorandvaluesof ^i,j aregiveninTables 1 and 2,respectively.

Inordertoimprovethesensitivitytoanomalouscouplings,results fromtheVHchannelsarecombinedwiththosefromH→^VV[15].

We assume the signal yield in the H→VV analysis to be dom- inated by gluon fusion production with negligible contamination fromvector boson fusion orVH production,as in Ref. [15].Pro- vided that the ratio of the Hbb and Htt coupling strengths is givenbytheSM prediction,Eq.(6)can beusedtorelate thesig- nalstrengthintheVHandH→VV analyses,withanappropriate change of indices(replacing ‘WH’ with‘ZZ’ to relate the ZZ and ZH channels,or‘ZH’ with‘WW’ torelatethe WW andWH chan- nels).Inthecombinationofthe WHandH→WW channels,the ratioofthe signalstrengths μ^WH/μ^WW increaseslinearlyfrom1 to173as f_a^WW₃ increasesfrom0to1,accordingtoEq.(6).TheWH signal strength has beenmeasured by CMSto be 1.1±⁰.9 [23], andfor H→^{WW it has been measured to be 0}.76±⁰.21 [13].

Thus,forintermediate andlargevaluesof f_a^WW₃ it isnotpossible toreconcile theexpectedsignal yield withdatainboth channels simultaneously. A similar effect occurs in a combination of the ZHandH→ZZ channels,wherethe ratioofthesignal strengths

μ^ZH/μ^ZZrisessharplywith f_a^ZZ₃.

However, an anomalous ratio of the Hbb and Htt coupling strengths spoilstherelationship inEq. (6). Wethereforeperform two interpretations ofthe VHandH→VV combination; one in- terpretationinwhich thisrelationshipisenforced,andone inter- pretation in which the signal strengths in the VH and H→^VV channelsareallowedtovaryindependently. Thesearereferredto as the ‘correlated-μ’ and ‘uncorrelated-μ’ combinations, respec- tively.

Fig. 1. Feynmandiagramsrepresentinggluon-initiatedZHproductionviaaquark triangle(top)andbox(bottom)loop.

4. Simulation

Simulatedqq→^{VH signaleventsaregeneratedforpurescalar} and pseudoscalar hypotheses with the leading-order (LO) event generator JHUGen,andassumingamassm_H=¹²⁵.6 GeV.Thesim- ulatedeventsampleisreweightedbasedonthevectorbosonp_T to includecorrectionsuptonext-to-next-to-LOandnext-to-LO(NLO) in theQCD andelectroweak (EW)couplingsrespectively [27–31].

ThesecorrectionsarederivedforascalarHiggsboson,andapplied tobothscalarandpseudoscalarsimulatedeventsamples.

The gg→^{ZH process includes diagrams with quark triangle} and box loops, asshown in Fig. 1. These diagrams interfere de- structively with one another [32]. The box diagram contains no HVV vertex. The triangle diagram does, but is unaffected by the a^HVV₃ terminEq.(1).The trianglediagram mediatedby aCP-odd HVV interaction is completely anti-symmetric underthe reversal ofthedirectionofloopmomentumflow;thediagramswithoppo- siteloop momentum flowtherefore perfectlycancelone another.

As the a^HZZ₁ coupling varies within a profile likelihood scan, the boxcontributionremainsfixedwhilethetrianglecontributionand the interferencemustbe variedaccordingly. Thisisaccomplished by reweighting thesimulated gg→^{ZH eventsample to havethe} correctm(VH)distributionatthegenerator level,includinginter- ference effects.Thisreweightingisbasedonresultsobtainedwith the VBFNLOeventgenerator [32,33],modified forthisanalysisto allowvariationoftheHff andHZZ couplingstrengths.

Simulatedbackgroundeventsamplesaregeneratedwithava- rietyofeventgenerators.Diboson,W+^jets,Z+^{jets,andtt samples} are generated with MadGraph 5.1 [34], while powheg 1.0 [35]

is used to generate single top quark samples, as well as the gluon-initiatedcontributiontoZHproduction(gg→^{ZH).The her-} wig++ 2.5 [36] generator is used along with alternative matrix element generators to produce additional simulated background samplestoassessthesystematicuncertaintyrelatedtoeventsim- ulationaccuracy,asdescribedinSection6.

The pythia 6.4 [37]and herwig++ generatorsareusedtosim- ulateparton showeringandhadronization. Detectorsimulation is performed with Geant4 [38]. Uncorrelated proton–proton colli- sions occurring in the same bunch crossing as the signal event (pileup) are overlayed on top of the hard interaction, in accord withthedistributionobserved.Correctionsareappliedtothesim- ulationinordertoaccountfordifferencesinobjectreconstruction efficienciesandresolutionswithrespecttothedata.

ControlregionsindataaredefinedinRef.[23],fromwhichnor- malizationscalefactorsforthedominantbackgroundsarederived.

A simultaneousfit todata acrosscontrol regions is performedto extractthescalefactors,whichareappliedhere.Theshapeofthe W(V)bosontransversemomentump_T distributioniscorrectedin thesimulatedtt (V+^{jets)eventsample,basedonafittodataina} background-enrichedcontrolregion.