Contents lists available atScienceDirect
Physics Letters B
www.elsevier.com/locate/physletb
Triple Higgs coupling as a probe of the twin-peak scenario
Amine Ahriche
a,b,c,∗, Abdesslam Arhrib
d, Salah Nasri
eaDepartmentofPhysics,UniversityofJijel,PB98OuledAissa,DZ-18000Jijel,Algeria
bTheAbdusSalamInternationalCentreforTheoreticalPhysics,StradaCostiera11,I-34014,Trieste,Italy cFakultätfürPhysik,UniversitätBielefeld,33501Bielefeld,Germany
dUniversitéAbdelMalekEssaadi,FacultédesSciencesetTechniques,B.P416,Tangier,Morocco ePhysicsDepartment,UAEUniversity,POB17551,AlAin,UnitedArabEmirates
a r t i c l e i n f o a b s t ra c t
Articlehistory:
Received22November2014
Receivedinrevisedform16February2015 Accepted24February2015
Availableonline26February2015 Editor:J.Hisano
Keywords:
Higgs Singlets
di-Higgsproduction
Inthisletter,weinvestigatethecaseofatwinpeakaroundtheobserved125GeVscalarresonance,using di-HiggsproductionprocessesatbothLHCande+e−LinearColliders.WehaveshownthatbothatLHC andLinearColliderthetripleHiggscouplingsplayanimportantroletoidentifythisscenario;andalso thatthisscenariocanbedistinguishablefromanyStandardModelextensionbyextramassiveparticles whichmightmodifythetripleHiggscoupling.Wealsointroduceacriterionthatcanbeusedtorule out thetwinpeakscenario.
©2015TheAuthors.PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBYlicense (http://creativecommons.org/licenses/by/4.0/).FundedbySCOAP3.
In July2012, ATLAS andCMS Collaborations[1,2] have shown theexistence ofaHiggs-likeresonancearound 125 GeVconfirm- ing thecornerstone of theHiggs mechanismthat predicted such particlelongtime ago. AllHiggscouplingsmeasured so farseem to be consistent, to some extent, with the Standard Model (SM) predictions.Moreover, inorderto establishtheHiggs mechanism asresponsibleforthephenomenaofelectroweaksymmetrybreak- ingonestillneedstomeasuretheselfcouplingsoftheHiggsand thereforetoreconstructitsscalarpotential.
RecentmeasurementsattheLHCshowthatthereisstilluncer- taintyontheHiggsmass;mh=125.3±0.4(stat.)±0.5(syst.)GeV for CMS [3] and mh=125.0±0.5 GeV for ATLAS [4] from the diphoton channel and mh=125.5±0.37(stat.)±0.18(syst.)GeV fromcombined channels.Despitethisrelativelylarge uncertainty, ascenariooftwodegeneratescalars around125.5 GeVresonance isneitherexcludednorconfirmed[5].
Inthe twin peak scenario (TPS); itis assumed that there are twoscalars h1,2 withalmost degeneratemassesaround 125 GeV.
Toourknowledge, there isno indication fromexperimental data whichdisfavorthisscenario.ThecouplingsofthetwinpeakHiggs toSMparticlesghiX X aresimplyscaledwithrespecttoSMrateby
*
Correspondingauthor.E-mailaddresses:aahriche@ictp.it(A. Ahriche),aarhrib@ictp.it(A. Arhrib), snasri@uaeu.ac.ae(S. Nasri).
cosθ (forh1) andsinθ (for h2), whereθ isa mixingangle,such thatwehavethefollowingapproximatesumrule:
g2
h1f¯f
+
g2h2f¯f
g2hSMf¯f
,
gh21V V
+
g2h2V V
gh2SMV V
,
(1)where f canbeanyoftheSMfermionsandV =W,Z vectorbo- son. Infact,thebranching ratiosoftheHiggs toSM particlesare SM-likeonlyiftheHiggsinvisibleisverysuppressedorkinemati- callyforbiddenaswillbeconsideredinourexample.Consequently, the single Higgs production such as gluon–gluon fusion at LHC, Higgs-strahlung,Vector BosonFusions, andtt H at¯ LHCande+e− LinearColliders(LC)willobeythesamesumrule.Thesummation ofeventnumbers(bothforproductionanddecay)ofthetwopos- siblecases will be identicalto SM casesince cos2θ+sin2θ=1.
However, for processes with di-Higgs final states (pp(e−e+)→ hh+X ),thetripleHiggscouplingsmayplayanimportantrole,and thereforetheseprocessescanbeusefultodistinguishbetweenthe cases ofone scalar ortwo degenerate ones around theobserved 125 GeVresonance.
ItiswellknownthatthetripleHiggscouplingscanbe,inprin- ciple, measured directly at the LHC with highluminosity option through double Higgs production pp→gg→hh [6]. Such mea- surement is rather challenging at the LHC, and for this purpose several parton level analysis have been devoted to this process.
It turns out that hh→bb¯
γ γ
[7],hh→bb¯τ
+τ
− [7,8] and hh→ bbW¯ +W−[8,9]finalstatesareverypromisingforHighluminosity.http://dx.doi.org/10.1016/j.physletb.2015.02.062
0370-2693/©2015TheAuthors.PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBYlicense(http://creativecommons.org/licenses/by/4.0/).Fundedby SCOAP3.
Recently,CMSreported apreliminaryresultonthesearchforres- onantdi-Higgsproductioninbb¯
γ γ
channel[10].The LC has also the capability of measuring with better pre- cision: theHiggsmass andsomeof theHiggscouplings together with the self coupling of the Higgs [11]. Using recoil technique fortheHiggs-strahlungprocess, theHiggsmasscanbe measured with an accuracy of about 40 MeV [11]. We note that at LHC withhighluminosity wecanmeasuretheHiggsmasswithabout 100 MeVuncertaintywhichisquitecomparabletoe+e− colliders.
The triple Higgscoupling can be extractedfrom e+e−→Zh∗→ Zhh at 500 GeV andeven better from e+e−→
νν
h∗→νν
hh at√s>800 GeV.Inthisregard,theLHCande+e−LCmeasurements arecomplementary[12].
InRef.[13],theauthorshaveprovidedatooltodistinguishthe two-degeneratestatesscenariofromthesingleHiggsone.Theap- proachof [13] applies only tomodels which enjoy modifications ofh→
γ γ
ratewithrespecttotheSM.However,accordingtothe latestexperimentalresults,bothforATLASandCMSthedi-photon channel seem to be rather consistent with the SM [3,4]. In this workweproposeanewapproachtodistinguishtheTPS.Thisap- proachis basedon the di-Higgs productionwhich issensitive to the tripleHiggs coupling,that is modified inthe majorityof SM extensions.Here,asanexample,weconsider,theTwo-SingletsModelpro- posedin[14],wheretheSMisextendedwithtworealscalarfields S0 and
χ
1; each oneis oddunder adiscrete symmetry Z(20) and Z(21)respectively.Thefieldχ
1 hasanon-vanishing vacuumexpec- tationvalue,whichbreaks Z(21) spontaneously,whereas,S0 =0;andhence,S0 isadarkmattercandidate.BothfieldsareSMgauge singletsandhencecaninteractwiththe‘visible’particlesonlyvia theHiggsdoubletH .Thespontaneousbreakingoftheelectroweak andthe Z(21) symmetriesintroduces the two vacuumexpectation values
υ
andυ
1 respectively. ThephysicalHiggsh1 andh2,with masses m1 and m2 m1, are related to the excitations of the neutral component of the SM Higgs doublet field, Re(H(0)), and the fieldχ
1 through rotation with a mixing angle θ and, with a specific choice in the parameter space, could give rise to two degenerate scalars around 125 GeV. In what follows, we denote byc=cosθ ands=sinθ.Thequarticandtriplecouplingsofthe physicalfieldshi aregivenintheappendicesin[15].Inouranalysiswerequirethat1:(i) allthedimensionlessquar- tic couplings to be 4
π
for the theory to remain perturbative, (ii) thetwoscalareigenmassesshouldbeinagreementwithrecent measurements [3,4]: we have checked that for the Two-Singlets model,thesplittingbetweenm1 andm2 couldbe oftheorderof 40 MeV.(iii) thegroundstatestability tobeensured;and(iv) we allowtheDMmassm0 tobeaslargeas1 TeV.In our work, we consider di-Higgs production processes at the LHC and e+e− LC, whose values of the cross section could be significant, namely,
σ
L H C(hh) andσ
L H C(hh+t¯t) at 14 TeV;σ
LC(hh+Z) at500 GeV andσ
LC(hh+Emiss) at1 TeV.All these processesinclude,atleast,oneFeynmandiagramwithtripleHiggs coupling. For the TPS, the total cross section gets contributions fromthefinal statesh1h1, h1h2 andh2h2.Therefore thequantity tobecomparedwiththestandardscenariocanbeexpressed as:σ
TPS(
hh+
X) = σ (
h1h1+
X) +
2σ (
h1h2+
X) + σ (
h2h2+
X) ,
(2) whichcanbeparameterizedas:σ
TPS= σ
aar1+ σ
abr2+ σ
bb,
(3)1 Actually,weconsideredthatallquarticcouplingstobeoforderunity;andthe singletvevυ1= χ1 =20∼2000 GeV.
Fig. 1. Numericalvaluesoftheparametersriin(4)for600benchmarksthatfulfill theabovementionedrequirements.
with
σ
aa+σ
ab+σ
bb=σ
SM(hh+X)andσ
aa,σ
bb andσ
ab corre- spondtothecrosssectioncontributionscomingfromtripleHiggs diagrams (a), non-triple Higgs diagrams(b) and the interference termintheamplitude,respectively.Thecoefficientsri aredimen- sionlessparameters,thatreceivecontributionsfromthefinalstates hihj, which depend on the mixing angle θ and the Higgs triple couplingsλ(i jk3).In the TPS, the amplitudes for di-Higgs production processes haveSMFeynmandiagramswheretheHiggsfieldh isreplacedby hi.Tocomputetheparametersri,wefirstestimatehowdoeseach amplitudegetmodifiedwithrespecttothecorrespondingSM one for each case hihj. For example,in the case ofh1h1 production, there aretwo typesof diagrams:(1) The onesthat involvetriple scalarinteractionsh1h1h1 andh2h1h1,withcouplingsequaltothe one ofa SM timesa factorofcλ(1113)/λSMhhh andsλ(1123)/λSMhhh, respec- tively. We denotethe total amplitude of thesetwo contributions byM(a).(2) TheoneswithnotripleHiggscouplings.Theirampli- tude,denotedbyM(b),isgivenbytheoneoftheSM scaledbya factorof c2.Therefore,theamplitudes M(a,b) (where a (b)stand fortriple Higgs(non-tripleHiggs) Feynman diagrams) forthedi- Higgs productioncan be written in terms oftheir corresponding SMvaluesas:
h1h1: M(a)= [(cλ(1113) +sλ(1123))/λSMhhh]MSM(a), M(b)=c2MSM(b),
h2h2: M(a)= [(cλ(1223) +sλ(2223))/λSMhhh]MSM(a), M(b)=s2MSM(b),
h1h2: M(a)= [(cλ(1123) +sλ(1223))/λSMhhh]MSM(a), M(b)=csMSM(b),
whereλSMhhhistheSM tripleHiggscouplingcalculatedatone-loop.
Thentheparametersriaregivenby:
r1
=
c2
[λ
(1113)2+ λ
(1223)2+
2λ
(1123)2] +
s2[λ
(1123)2+ λ
(2223)2+
2λ
(1223)2] +
2cs[ λ
(1113)λ
(1123)+
2λ
(1123)λ
(1223)+ λ
(1223)λ
(2223)]
/
λ
SMhhh 2,
r2= {
c3λ
(1113)+
3c2sλ
(1123)+
3cs2λ
(1223)+
s3λ
(2223)}/λ
SMhhh.
(4) Thus,thevaluesofri quantifybyhowmucheachdi-Higgsprocess deviatesfromtheSMcase.InFig. 1,weshowtheparametersrias afunctionofsinθ forabout600chosensetsofthemodelparam- eters withinthecondition (1).Weseethat forverysmallmixing angle ri’s are approximately equal to unity, while for sinθ >0.8 andsinθ <−0.2,theparameterr1 becomeslargerthanunityandTable 1
Differentcontributionstotheconsideredprocessescrosssections.NumbersforLHC aretakenfrom[16]atNLO.
σaa(fb) σab(fb) σbb(fb) σSM(fb)
hh 9.66 −49.9 70.1 29.86
hh+t¯t 3.3164×10−2 0.13952 0.84731 1.02 hh+Z 9.0206×10−3 4.6999×10−2 9.005×10−2 0.14607 hh+Emiss 5.1631×10−2 −0.20867 0.29708 0.14004
r2 acquires negative values. This behavior could lead to an en- hancement/reduction to the cross section depending on thesign oftheinterferencecontribution,
σ
ab,tothetotalcrosssection.This meansthatthemeasurementofthefollowingratio:ξ (
hh+
X) = σ
TPSpp
(
e−e+) →
hh+
Xσ
SMpp
(
e−e+) →
hh+
X,
(5) couldbeveryusefultoconfirmorexclude thisscenariobasedon thedeviationofanyoftheparametersri fromunity.Forinstance, theratioξ (hh+X)candeviatefromunityiftheSM isextended withmassiveparticles(SM+MP)thatcoupletotheHiggsdoublet andcontributetothetripleHiggscouplingaswelltheHiggsmass.In this case, r1=(1+)2 and r2=1+, where represents therelativeenhancementofthetripleHiggscouplingduetoSM+ MP.As we will show later, our considered scenario forsmall or
largemixingcouldbedistinguishedfromthecaseofSM+MPby combiningtheratio(5)fordifferentprocesses.
In Table 1, we give the values of
σ
aa,σ
ab andσ
bb for the corresponding di-Higgs production processes. We note that their contributions to the LHC process pp→hh and to the LC one e+e−→Zhh seem to be uncorrelated, which makes the Higgs triplecouplingusefultoprobethisscenarioanddistinguishitfrom (SM+MP).For the benchmarksconsidered previously inFig. 1, we illus- trate in Fig. 2 the production cross section of di-Higgs at e+e− LC andLHC and in Fig. 3 the ratio ξ. As it can be seen, in the TPS,thecrosssectionoftheprocessespp→hh, pp→hh+tt and¯ e−e+→hh+Emissaremostlyenhanced,whilefore−e+→hh+Z itisenhancedjustforthemixingvalues0.5<sinθ <0.8.
Nowletusdiscussthepossibilityofdisentanglingthe TPS from the SM+MP. It is clear from Fig. 3 that for both LHC and LC processeswithlargemixing,0.35<cos2θ <0.65,theTPSmayco- incidewithSM+MP.However,fornon-maximalmixingvaluesthe TPSisclearlydifferentthanSM+MPwhereallbenchmarkshave thefollowingfeature
ξ
1T P S+ ξ
2T P S> ξ
1SM+MP() + ξ
2SM+MP() ,
(6) where ξiT P S the ratio in (5) for any LHC or LC processes and ξiSM+MP() is the same ratio due the existence of massive par- ticles. Therefore,when measuring the quantities (5)for both the LHC ande+e− LC processes,and one finds that the criterion (6) is not fulfilled,then it is a certain exclusion for thisscenario. InFig. 2. The cross section values(2)for the di-Higgs production processes for the 600 benchmarks used previously. The solid lines correspond to the SM cross sections.
Fig. 3. Theratiosξgivenin(5)forthedi-Higgsproductionprocessesforthe600benchmarkusedpreviously.Thegreenbenchmarkscorrespondtothelargemixingcase where0.35<cos2θ <0.65,andthebluepointrepresentstheSM;andthesolidcurverepresentsthecaseofaSMextension,wherethenewphysicsaffectsthetripleHiggs couplingasλhhh=λSMhhh(1+);andthevalueoftherelativeenhancementcanbereadfromthepalette.(Forinterpretationofthereferencestocolorinthisfigurelegend, thereaderisreferredtothewebversionofthisarticle.)
Table 2
Differentvaluesoftheratios(4)and(5)forthethreechosenbenchmarks.
B1 B2 B3
sinθ 0.53555 0.90126 −0.39802
r1 2.95386 2.88466 5.62286
r2 1.31634 0.28189 −1.26011
ξ (hh) 1.10345 2.80975 6.27248
ξ hh+tt¯
2.69728 2.51821 4.66603
ξ (hh+Z) 1.22243 0.88532 0.55827
ξ (hh+Emiss) 1.24900 2.76488 6.07213
Table 3
Theeventsnumberforthedifferentprocesseswithintheluminosityvaluesmen- tionedabovefortheSMandthebenchmarksshowninTable 2.
Events number Channel SM B1 B2 B3
pp→hh 4b 966.75 1066.8 2716.3 6063.9
2b2τ 106.70 117.74 299.8 669.27 2b2γ 3.89 4.29 10.93 24.4
pp→hh+t¯t 4b 33.02 89.06 83.15 154.07
e−e+→hh+Z 4b 23.65 28.91 20.94 13.2
e−e+→hh+Emiss 4b 45.34 56.63 125.36 275.31
case where the criterion (6) is fulfilled, detailed analysis is re- quired forin order to identify the mixing angle, the parameters ri andthereforetheHiggstriplecouplings.Infact,bystudyingall thedi-HiggsproductionchannelsatbothLHCande+e−LConenot onlyconfirm/excludethisscenario, butalsodistinguished it from models where only one type of processes gets modified by new physicssuch as:itmanifests asnewsourcesofmissingenergyin e−e+→hh+Emiss [17],new colored scalar singlets contribution to pp→hh (orhh+t¯t)[18],orthepresenceofaheavyresonant Higgs[19].
Inordertoshowwhetherthisscenariocanbetestedatcollid- ers,weconsiderthreebenchmarksthatmaybedistinguishedfrom SM+MP(i.e.,three redpoints fromFig. 3),andcomparethe di- Higgsdistribution(of thedi-Higgs invariantmassasan example) withtheSM one.Thecorrespondingvaluesofratiosri andξi are giveninTable 2,andinTable 3,wepresenttheexpectednumber ofeventsatboththeLHCandLC.Weseethatforbenchmark B2, theeventsnumberissignificantlylargerthantheSMforthechan- nelspp→2b2
τ
attheLHCande−e+→4b+EmissatLC’s,whileit isreducedfortheprocesses pp→4b+t¯t ande−e+→4b+Z .For benchmark B1, the events number of the processes pp→2b2τ
ande−e+→4b+Emiss is SM-like butit is reduced forthe pro- cesses pp→4b+tt and¯ e−e+→4b+Z .Forbenchmark B3,the eventsnumberisreducedfortheconsideredprocesses.
InFig. 4,we illustratethedi-Higgs invariantmassdistribution (Mh,h)fortheprocess e−e+→hh+Emiss.Clearly,the TPS canbe easily distinguished from the SM, especially in the case of non- maximalmixing.However,thefullconfirmationofthe TPS requires theenlargementoftheinvestigationby takingintoaccount other di-Higgs productionchannels such ashhj j, hhW±,hh Z and hht j attheLHC[20]andthee+e− LC[11].
In conclusion, we have investigated the caseof twin-peak at the 125 GeV observed scalar resonance by considering different di-HiggsproductionprocessesatbothLHCande+e− LC.Wehave introduced acriterion whose violationexcludestheTPS scenario, otherwise thisscenario can be surely distinguished fromthe SM andSMextendedbymassivefieldsincaseofnon-maximalmixing.
Fig. 4. Normalizeddi-Higgsinvariantmassdistributionfortheprocesse−e+→hh+ Emissforthebackground(BG)andtheconsideredbenchmarksinTable 2.
Last butnotleast, we should note that thisscenariocould be realized within SM+(real/complex) singlet scalar, or any larger scalar field content. Thisincludes neutralor chargedscalars that are members anymultiplets, wheretwo degenerate scalareigen- states h1,2 at 125 GeV, do couple to the SM gauge fields and fermions by more than ∼90%, i.e., the sum rule (1) is fulfilled.2 Ifthe measurementofdi-Higgs processesatLHCand/or e+e− LC turnouttobeconsistentwithSMpredictions,thenitwillbevery challengingtodistinguishthe TPS scenario.
If the measurement of the couplings hf ¯f and hV V become much more precise from the future experiment data, it may be possiblethatonecouldbesensitivetotheradiativecorrectionsef- fecttothesecouplings.Suchradiativecorrectionstohf¯f andhV V couplings ina variety ofextended Higgssector havebeen evalu- atedin[22–24].Theseone-loopeffectsare oftheorderof2–10%
andeven morein some specialcases.The presentLHC measure- mentsarenotyetsensitivetosucheffects.
Acknowledgements
WewouldliketothankA.DjouadiandR.Santosforthevalu- able comments; and E. Vryonidou for sharing with us her code and for manyuseful discussions. A. Ahriche is supported by the AlgerianMinistry ofHigher EducationandScientificResearch un- dertheCNEPRUProjectNo.D01720130042;andpartiallybyDAAD andICTP.A.ArhribissupportedinpartbytheMoroccanMinistry ofHigherEducationandScientificResearch:“projetdesdomaines prioritairesdelarecherchescientifiqueetdudeveloppementtech- nologique”.
References
[1]G.Aad,etal.,ATLASCollaboration,Phys.Lett.B716(2012)1.
[2]S.Chatrchyan,etal.,CMSCollaboration,Phys.Lett.B716(2012)30.
[3] S.Chatrchyan,etal.,CMSCollaboration,CMS-PAS-HIG-13-001.
[4]G.Aad,etal.,ATLASCollaboration,arXiv:1406.3827[hep-ex].
[5]Forexamplesee:M.Heikinheimo,A.Racioppi,M.Raidal,C.Spethmann,Phys.
Lett.B726(2013)781.
[6]A.Djouadi,W.Kilian,M.Muhlleitner,P.M.Zerwas,Eur.Phys.J.C10(1999)45;
E.W.N.Glover,etal.,Nucl.Phys.B309(1988)282;
T.Plehn,etal.,Nucl.Phys.B479(1996)46;
T.Plehn,etal.,Nucl.Phys.B531(1998)655(Erratum);
J.Baglio,etal.,J.HighEnergyPhys.1304(2013)151.
[7]U.Baur,T.Plehn,D.L.Rainwater,Phys.Rev.D69(2004)053004.
[8]M.J.Dolan,C.Englert,M.Spannowsky,J.HighEnergyPhys.1210(2012)112.
[9]A.Papaefstathiou,L.L.Yang,J.Zurita,Phys.Rev.D87(2013)011301.
[10] S.Chatrchyan,etal.,CMSCollaboration,CMS-PAS-HIG-13-032.
2 Inthe2HDM,twinpickscenariohasbeenstudiedin[21],butthestudycon- centratedonlyonthediphotonchannel.Accordingtothisstudy[21],thisscenario isnotruledout.
[11]D.M.Asner,etal.,arXiv:1310.0763[hep-ph];
A.Djouadi,W.Kilian,M.Muhlleitner,P.M.Zerwas,Eur.Phys.J.C10(1999)27.
[12]G.Weiglein,etal.,LHC/LCStudyGroupCollaboration,Phys.Rep.426(2006)47.
[13]J.F.Gunion,Y.Jiang,S.Kraml,Phys.Rev.Lett.110(2013)051801.
[14]A.Abada,D.Ghaffor,S.Nasri,Phys.Rev.D83(2011)095021;
A.Ahriche,S.Nasri,Phys.Rev.D85(2012)093007.
[15]A.Ahriche,A.Arhrib,S.Nasri,J.HighEnergyPhys.1402(2014)042.
[16]M.Spira,arXiv:hep-ph/9510347.
[17]A.Ahriche,S.Nasri,R.Soualah,Phys.Rev.D89(2014)095010.
[18]G.D.Kribs,etal.,Phys.Rev.D86(2012)095023;
Z.Heng,etal.,J.HighEnergyPhys.1402(2014)083;
T.Enkhbat,J.HighEnergyPhys.1401(2014)158.
[19]M.J.Dolan,C.Englert,M.Spannowsky,Phys.Rev.D87(2013)055002;
J.Cao,Z.Heng,L.Shang,P.Wan,J.M.Yang,J.HighEnergyPhys.1304(2013) 134;
U.Ellwanger,J.HighEnergyPhys.1308(2013)077;
A.Arhrib,R.Benbrik, C.-H.Chen,R. Guedes,R.Santos,J.HighEnergyPhys.
0908(2009)035.
[20]R.Frederix,etal.,Phys.Lett.B732(2014)142,arXiv:1401.7340[hep-ph].
[21]P.M.Ferreira,R.Santos,H.E.Haber,J.P.Silva,Phys.Rev.D87 (5)(2013)055009, arXiv:1211.3131[hep-ph].
[22]S.Kanemura,M.Kikuchi,K.Yagyu,Phys.Lett.B731(2014)27,arXiv:1401.0515 [hep-ph].
[23]M.Aoki,S.Kanemura,M.Kikuchi,K.Yagyu,Phys.Rev.D87 (1)(2013)015012, arXiv:1211.6029[hep-ph].
[24]A.Arhrib,M.CapdequiPeyranere,W.Hollik,S.Penaranda,Phys.Lett.B579 (2004)361,arXiv:hep-ph/0307391.