Contents lists available atScienceDirect
Physics Letters B
www.elsevier.com/locate/physletb
Clockwork for neutrino masses and lepton flavor violation
Alejandro Ibarra
a,b, Ashwani Kushwaha
c, Sudhir K. Vempati
c,∗aPhysik-DepartmentT30d,TechnischeUniversitätMünchen,James-Franck-Straße,85748Garching,Germany bSchoolofPhysics,KoreaInstituteforAdvancedStudy,Seoul02455,SouthKorea
cCentreforHighEnergyPhysics,IndianInstituteofScience,C.V.RamanAvenue,Bangalore560012,India
a r t i c l e i n f o a b s t ra c t
Articlehistory:
Received18November2017
Receivedinrevisedform14February2018 Accepted20February2018
Availableonline26February2018 Editor:G.F.Giudice
WeinvestigatethegenerationofsmallneutrinomassesinaclockworkframeworkwhichincludesDirac masstermsaswellasMajoranamasstermsforthenewfermions.Wederiveanalyticformulasforthe massesofthenewparticlesandfortheirYukawacouplingstotheleptondoublets,inthescenariowhere theclockworkparametersareuniversal.WhentheuniversalMajoranamassvanishes,thezeromodeof theclockworksectorformsaDiracpairwiththeactiveneutrino,withamasswhichisinagreementwith oscillationsexperimentsforasufficientlylargenumberofclockworkgears.Ontheotherhand,whenit doesnotvanish,neutrinomassesaregeneratedviatheseesawmechanism.Inthiscase,andduetothe factthattheeffectiveYukawacouplingsofthehighermodescanbesizable,neutrinomassescanonlybe suppressedbypostulatingalargeMajoranamassscale.Finally,wediscusstheconstraintsonthemass scaleoftheclockworkfermionsfromthenon-observationoftherareleptonicdecayμ→eγ.
©2018TheAuthors.PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBYlicense (http://creativecommons.org/licenses/by/4.0/).FundedbySCOAP3.
1. Introduction
The smallness of neutrino masses stands as one of the most puzzlingopenquestionsinFundamentalPhysics.Aplausiblesolu- tiontothispuzzleisprovidedbytheseesawmechanism,inwhich thesmallnessof neutrinomassesis explainedby thebreaking of theleptonnumberataveryhighenergyscale[1–5].Modelswith conserved leptonnumber, onthe other hand,can alsoreproduce theobservations, atthe expenseofpostulating tiny Yukawa cou- plings of the neutrino to the Standard Model Higgs. Such small parameters are usually regarded as unnatural, however the ex- istence of tiny Yukawa couplings is a phenomenologically viable possibility, andcan be accomplishedin furtherextensions ofthe model(forreviewsandrecentmodels,seee.g.in [6–19]).
Recently, a new mechanism of generating small couplings in theoriescoupledto theStandardModelhas beenintroduced [20, 21].Themechanism,reminiscentofdeconstructionmodels[22,23], canbesummarizedasalinearquivermodelwithnolargehierar- chies in the theory parameters, that gives rise to site-dependent suppressedcouplingstothezero-mode[24].Originally,introduced foraquiverofAbelianGoldstonebosons(axions),ithasbeengen- eralized to fermions, vectors and other fields [24,25] (See also
*
Correspondingauthor.E-mailaddresses:ibarra@tum.de(A. Ibarra),ashwani@chep.iisc.ernet.in (A. Kushwaha),vempati@chep.iisc.ernet.in(S.K. Vempati).
[26]). Applications and generalizations of this mechanism have beendiscussedin[27–43],andspecificallyframeworkstoexplain theobservedpatteroffermionmassesin[44–46].
Inthisworkweexploretheapplicationofthefermionicclock- work to the generation of smallneutrino masses. Concretely, we identify the right-handed neutrinos with the zero modes of a clockwork sector [24], such thatsmallcouplings canbe naturally generatedandthereforesmallneutrinomasses.Weanalyzeinde- tailtheframeworkwheretheDiracmasses,Majoranamassesand nearest neighbor interactionsareuniversal, complementingprevi- ous studiesin[45,46] wheretheMajoranamasstermislocalized on just one of the modes. We derive analytical formulas for the massesofthenewparticlesandfortheircouplingstotheStandard Modelfermions,forthecaseswhentheMajoranamasstermsare includedintheLagrangianandwhentheyarevanishing.Weshow thattheclockworkmechanism,i.e.,thesuppressionoftheYukawa couplings bysitedependent powerfactors, isnot affectedby the presenceoftheMajoranamassterms.Infact,whilethezeromode contribution is a combination of the clockwork suppression and the Majorana seesaw,the scale is however set by the dominant contribution by the gears, which have O(1) Yukawa couplings, through the standardseesaw mechanism. Furthermore,while the clockwork mechanismsuppressesthecouplingsofthezeromode, the couplingsofthehighermodescanbe sizableandinduce,via loops,potentiallylargeratesfortheleptonicraredecays.
The restofthepaperisorganized asfollows.Insection 2,we presentthemostgeneralframeworkforclockwork neutrinoswith https://doi.org/10.1016/j.physletb.2018.02.047
0370-2693/©2018TheAuthors.PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBYlicense(http://creativecommons.org/licenses/by/4.0/).Fundedby SCOAP3.
DiracandMajoranamassterms,andwediscusstheirphenomenol- ogyinsubsections2.1and2.2,respectively.Insection3,wediscuss leptonflavor violationintheclockworkscenarioandcalculatelim- itsonthegearmasses.Weclosewithasummary.
2. Neutrinosinclockwork
We extendthe Standard Modelwith n left-handedandn+1 right-handed chiral fermions, singlets under the Standard Model gaugegroup,whichwedenoteasψLi(i=0,...,n−1)andψRi(i= 0,...,n)respectively.TheLagrangianofthemodelreads:
L
=
LSM+
LClockwork+
Lint,
(1)whereLSMistheStandardModelLagrangian,LClockworkisthepart oftheLagrangianinvolvingonlythenewfermionsinglets,andLint istheinteractiontermofthenewfieldswiththeStandardModel fields. Following [24], we assume that the Standard Model only couplestothelastsiteofthefermionicclockwork,therefore,
Lint
= −
YH LLψ
Rn,
(2) with H=iτ
2H∗, H the Standard Model Higgs doublet and LL theleft handed lepton fields (weassume only one generation of fermions;the generalizationto morethanone generation willbe discussedbelow).Infullgenerality,theclockworkLagrangiancanbecastas:
LClockwork
=
Lkin−
n−1
i=0
miψ
Liψ
Ri−
miψ
Liψ
Ri+1+
h.
c.
−
n−1
i=0 1
2MLi
ψ
Licψ
Li−
n i=0 12MRi
ψ
cRiψ
Ri,
(3)whereLkin denotes the kinetic term forall fermions, andm,m andML,R are massparameters. Denoting =(ψL0,ψL1,...ψLn−1, ψcR0,ψcR1,...,ψRnc ),theclockworkLagrangiancanbewritteninthe compactform:
LClockwork
=
LKin−
12
(
cM+
h.
c.)
(4)with M a (2n+1)×(2n+1) mass matrix. We note that Lkin isinvariant underthe globalgroup U(n)L×U(n+1)R. Themass termsmibreaktheglobalgroupU(n)L×U(n+1)R→n−1
i=0U(1)i, where U(1)i acts as ψL,i→eiαiψL,i, ψRi →eiαiψRi, and com- binedwiththemasstermsmi,breaktheglobalsymmetryU(n)L× U(n+1)R→U(1)CW,whereU(1)CWactsasψL,i→eiαψL,i,ψR,i→ eiαψR,i for all i. Finally, MLi and MRi are Majorana masses for the left and right handed singlet fields. It is sufficient that MLi or MRi is non-vanishing forone i to break the symmetry group U(n)L×U(n+1)R→nothing.
We assume for simplicity universal Dirac masses, Majorana massesandnearestneighborinteractions,namelymi=m,mi=mq MRi=MLi=mq forall i.Underthisassumption,themassmatrix reads:
M
=
m⎛
⎜ ⎜
⎜ ⎜
⎜ ⎜
⎜ ⎜
⎜ ⎜
⎜ ⎜
⎝
q 0· · ·
0 1−
q· · ·
0 0 q· · ·
0 0 1· · ·
0.. . .. . .. . .. . .. . .. . .. . .. .
0 0· · ·
q 0 0 0−
q 1 0· · ·
0 q 0· · ·
0−
q 1· · ·
0 0 q· · ·
0.. . .. . .. . .. . .. . .. . .. . .. .
0 0· · · −
q 0 0 0 q⎞
⎟ ⎟
⎟ ⎟
⎟ ⎟
⎟ ⎟
⎟ ⎟
⎟ ⎟
⎠
,
(5)whichhaseigenvaluesMk givenby:
M0
=
mq,
Mk=
mq−
mλ
k,
k=
1, . . . ,
n,
Mn+k=
mq+
mλ
k,
k=
1, . . . ,
n,
(6)withλk definedas
λ
k≡
q2+
1−
2qcos kπ
n
+
1.
(7)Withourconventions,theeigenvaluescanbepositiveornegative;
thephysicalmassescorrespondtothemodulioftheeigenvalues.
Themasseigenstates,whichwedenoteas
χ
k,arerelatedtothe interactioneigenstatesjbytheunitarytransformationU,namelyj=
jUjk
χ
k.ThematrixU canbeexplicitlycalculated,there- sultbeing:U
=
⎛
⎝
0 √12UL
−
√12UL uR √12UR √1 2UR
⎞
⎠ .
(8)where0 anduR aren-dimensionalvectors,withentries:
0
j=
0,
j=
1, ...,
n,
(9)(
uR)
j=
1 qj q2−
1q2
−
q−2n,
j=
1, ...,
n,
(10)whileUL andUR are,respectively,n×nand(n+1)×nmatrices withelements
(
UL)
jk=
2n
+
1sin jkπ
n
+
1,
j,
k=
1, ...,
n, (
UR)
jk=
2(
n+
1)λ
kqsin jk
π
n
+
1−
sin(
j+
1)
kπ
n
+
1,
j
=
0, ..,
n,
k=
1, ...,
n,
(11)WenotethatthemixingmatrixUdoesnotdependontheparam- eterq,whichisaconsequenceofourassumptionofuniversalityof theMajoranamassesMRi=MLi=mqforalli.
TheinteractionLagrangian oftheclockwork fieldstotheStan- dard Model fields, Eq. (4), can now be recast in terms of mass eigenstates:
Lint
= −
Y LLHUnkχ
k≡ −
2n k=0YkLL
Hχ
k,
(12) whereY0
≡
Y(
uR)
n=
Y qn q2−
1q2
−
q−2n,
(13)Yk
=
Yk+n≡ √
12Y
(
UR)
nk=
Y1
(
n+
1)λ
kqsin nk
π
n
+
1,
k
=
1, ...,
n.
(14)Thecomponents(uR)n and(UR)np,whichdescribethefractionof thenth “gear”inthezeromode,willplayamajorroleinthephe- nomenology,astheyparametrizetheportalstrengthbetweenthe StandardModelsectorandtheclockworksector.
After electroweak symmetry breaking new mass terms arise which mix the Standard Model neutrino with the clockwork
fermions. The mass matrix of the 2n+2 electrically neutral fermionfieldsofthemodelreads:
mν
=
⎛
⎜ ⎜
⎜ ⎜
⎜ ⎜
⎜ ⎜
⎜ ⎝
ν
Lχ
0χ
1χ
2· · · χ
2nν
L 0 v Y0 v Y1 v Y2· · ·
v Y2nχ
0 v Y0 M0 0 0· · ·
0χ
1 v Y1 0 M1 0χ
2 v Y2 0 0 M2· · ·
0.. . .. . .. . .. . .. . .. . .. .
χ
2n v Y2n 0 0 0· · ·
M2n⎞
⎟ ⎟
⎟ ⎟
⎟ ⎟
⎟ ⎟
⎟ ⎠
,
(15)where v=246/√
2 GeV is the Higgs vacuumexpectation value.
Upondiagonalizingthismassmatrix,one findsamassfortheac- tive neutrino. Furthermore, the off-diagonal entries in the mass matrix translate into charged current interactions between the chargedleptonandthek-thmode,aswellasneutral-currentand Higgsinteractionsofthelightneutrino,proportionalto∼v Yk/Mk, andwhichcanbesizable.
Inorderto accommodatetheleptonic mixingobserved inNa- tureitis necessarytointroduce threegenerations ofleptondou- blets, as well as N generation of clockwork fermions, each con- sisting of nα left-handed and nα +1 right-handed gears, where
α
=1,. . . ,N (phenomenologically, N≥2, inorderto account for thetwoobservedoscillationfrequencies).Furthermore,theYukawa couplinginEq. (2) andallthemassparametersinEq. (3) mustbe promotedtomatricesinflavorspace.Inthisworkwewillassume forsimplicitymαβi =mδαβ,mαβ
i =mqαδαβ Mαβ
Ri =MαLiβ=mqαδαβ forall i.Namely, the massparameterm is universal forallgears and all generations, while the mass parameters m, MR and ML are commonforall gearswithin one generation, butinprinciple differentamonggenerations.
Denoting α=(ψL0α,ψL1α,...,ψLnα−1,ψR0αc,ψαR1c,...,ψRnαc) as the fermion field which has as component all the clockwork fields withinthegeneration
α
,theclockworkandinteractionLagrangian canbewrittenas:LClockwork
=
LKin−
12
(
αcMαββ
+
h.
c.) ,
(16) Lint= −
YaαLaLHψ
Rα,n,
(17)where a=1,2,3 and
α
,β=1,...,N. As for the one generation case, we assumedthat the Standard Model lepton doublets only coupletothen-thsitesoftheN clockworkgenerations.The Lagrangian expressed in the mass eigenstate basis, αk = Ukjαβ
χ
βj,read:LClockwork
=
LKin−
12
( χ
kαcMkαχ
kα+
h.
c.) ,
(18) Lint= −
2n k=0YkaβLaL
χ
kβ,
(19)whereYkaβ≡YaαUnkαβ withUnkαβ thematrixthatmixesfermionsof different clockwork gears anddifferent generations. Finally,after electroweak symmetry breaking, the mass matrix of the N(2n+ 1)+3 electricallyneutralfermionsofthemodelreads:
mν
=
⎛
⎜ ⎜
⎜ ⎜
⎜ ⎜
⎜ ⎜
⎜ ⎜
⎝
ν
aLχ
0βχ
1βχ
2β· · · χ
2nβν
aL 0 v Y0aβ v Y1aβ v Y2aβ· · ·
v Ya2nβχ
0β v Y0βa Mβ0 0 0· · ·
0χ
1β v Y1βa 0 Mβ1 0χ
2β v Y2βa 0 0 Mβ2· · ·
0.. . .. . .. . .. . .. . .. . .. .
χ
2nβ v Y2nβa 0 0 0· · ·
Mβ2n⎞
⎟ ⎟
⎟ ⎟
⎟ ⎟
⎟ ⎟
⎟ ⎟
⎠
.
(20)Thismatrixhasingeneralanon-trivialflavorstructure andleads not only tomixingamongthe threeactive neutrinos,butalso to potentially large lepton flavor violating charged current, neutral current and Higgs interactions, thus providing a possible test of thisframework,aswillbediscussedinSection3.
In what follows, we will consider separately the case when the universal Majorana mass is vanishing and when it is non- vanishing.
2.1. VanishinguniversalMajoranamass
WeconsiderfirstthecasewheretheuniversalMajoranamassis equaltozero.Inthiscase,theglobalsymmetryoftheLagrangian isbrokenasU(n)L×U(n+1)R→U(1)CW,whichwillbeidentified withtotal lepton number.The eigenstates andeigenvaluesof the massmatrixcan bedeterminedusingtheresultsofSection 2,by settingq=0.
ItisusefultorecasttheclockworkLagrangianas
Lclockwork
=
Lkin−
NLmνDNR+
h.
c.
(21)where we have defined new fields NL =(
ν
L,NL1,...,NLn) and NR=(NR0,NR1,...,NRn),withNRk
= √
12
( χ
k+ χ
k+n) ,
k=
0, ...
n,
(22)NLk
= √
12
( − χ
k+ χ
k+n) ,
k=
1, ...
n.
(23)Inthisbasis,themassmatrixhastheform:
mνD
=
⎛
⎜ ⎜
⎜ ⎜
⎜ ⎜
⎜ ⎝
NR0 NR1 NR2
· · ·
NRnν
L 0 0 0· · ·
0NL1 0 M1 0
· · ·
0 NL2 0 0 M2· · ·
0.. . .. . .. . .. . . . . .. .
NLn 0 0 0· · ·
Mn⎞
⎟ ⎟
⎟ ⎟
⎟ ⎟
⎟ ⎠
.
(24)where Mk=m√
λk,withλk definedinEq. (7).Namely, thefields
ν
L andNR0 formamassless Dirac pair,whilethe fields NRk and NLk form, for k=1,...,n, Diracpairs withmass Mk.The overall scaleofthemassivepairsisdeterminedbytheparameterm,and themassdifferencebetweenpairsdependsonq andn.Assuming q>1, one obtains that the massesof the modeswithk>0 in- creasemonotonicallywithn,fromM1≈m(q−1)toMn≈m(q+1). In Fig. 1, left panel, we show forillustration the mass spectrum of the particles ofthe clockwork sector, labeled by k, takingfor concretenessn=10 andq=2.Themassspectrumhasbeennor- malizedtom.The mass spectrum is modified after electroweak symmetry breaking by the interactions with the Higgs field. Expressed in termsofNRk,theinteractionLagrangianreads:
Fig. 1.Diracmasses(leftpanel)andYukawacouplings(rightpanel)ofthesingletfermionsoftheclockworksector,normalizedrespectivelytomandY,forthespecificcase n=10 andq=2.
Lint
=
n k=0YkLL
H NRk+
h.
c.
(25) withY0
≡
Y(
uR)
n=
Y qn q2−
1q2
−
q−2n,
(26)Yk
≡
Y(
UR)
nk=
Y2
(
n+
1)λ
kqsin nk
π
n
+
1,
k=
1, ...,
n.
(27) TheYukawacouplingofthemasslessmodeY0 issuppressedbyqn, providedq>1,whereasthecouplingsofthekth-modeare ofthe same orderas Y. This is illustrated inFig. 1, right panel, which shows the Yukawa couplings of the clockwork fermions to the StandardModelleptondoublets,normalizedtoY,forthesameval- uesofnandq asintheleftpanel(inthiscase,|Y0|/Y≈8×10−4 andisnotvisiblefromthefigure.)Themassmatrixofthe electricallyneutralfermionfieldsnow reads:
mνD
=
⎛
⎜ ⎜
⎜ ⎜
⎜ ⎜
⎜ ⎝
NR0 NR1 NR2
· · ·
NRnν
L v Y0 v Y1 v Y2· · ·
v YnNL1 0 M1 0
· · ·
0 NL2 0 0 M2· · ·
0.. . .. . .. . .. . . . . .. .
NLn 0 0 0· · ·
Mn⎞
⎟ ⎟
⎟ ⎟
⎟ ⎟
⎟ ⎠
.
(28)Concretely,a masstermfortheactive neutrinosisgenerated.As- suming that MkY0v, which as we will see below is justified fromthecurrentlimitsonrareleptonic decays,one canapproxi- matetheactiveneutrinomassby
mν
≈
v Y0 (29)and can be made small by choosing appropriate values of Y, q andn.Forinstance,assuming Y=O(1),q=2,oneobtains mν= O(0.1)eV forn≈40.
Thegeneralization ofthe abovesetup tothree leptonic gener- ationsand N clockwork generationsisstraightforward.The clock- workLagrangianis:
Lclockwork
=
Lkin−
NαLmανNαR+
h.
c.
(30) withNαL =(
ν
Lα,Nα L1,...,NαLn)andNα R=(Nα
R0,Nα R1,...,Nα
Rn),where
NαRk
= √
12
( χ
kα+ χ
kα+n) ,
k=
0, ...,
nα =
1...,
N,
(31) NαLk= √
12
(− χ
kα+ χ
kα+n) ,
k=
1, ...,
n, α =
1...,
N,
(32) andtheinteractionLagrangian,Lint
= −
n k=0YkaβLaL
H0NβRk,
(33)withYkaβ=YaαUnkαβ.
Afterelectroweaksymmetrybreakingtheneutrinomassmatrix reads:
mDν
=
⎛
⎜ ⎜
⎜ ⎜
⎜ ⎜
⎜ ⎝
NβR0 NβR1 NβR2
· · ·
NβRnν
La v Ya0β v Ya1β v Ya2β· · ·
v Ynaβ NβL1 0 M1β 0· · ·
0 NβL2 0 0 M2β· · ·
0.. . .. . .. . .. . . . . .. .
NβLn 0 0 0
· · ·
Mnβ⎞
⎟ ⎟
⎟ ⎟
⎟ ⎟
⎟ ⎠
.
(34)where Mkβ is themass ofk-thclockwork gearforthe Dirac pair NβL,NβR.
Weanalyzeindetailthe casewheretheclockwork consistsof two generations withn1 andn2 gears,respectively. Wescan Yaα within the ranges 14 <|Yaα|<4, qα between1.5 and 6 and nα between15and55, andweselect thepoints that reproducethe observed values ofthe solarand atmosphericmass splitting and mixingangleswithin1
σ
,asdeterminedinRef. [47].InFig.2(left panel)weshowasgreencircles(yellowtriangles)thevaluesofn1 (n2) as a function of q1 (q2) that satisfy the experimental con- straints. As apparent from the plot, larger qα require a smaller numberofgearstoreproducethesmallneutrinoYukawacoupling.Furthermore,theallowed valuesforn1 andn2 haveabigoverlap, whichisaconsequenceofourassumptionofcomparableelements inthecouplingYaα andthenecessityofproducingamildhierar- chybetweenthesolarandtheatmosphericneutrinomassscales.
Inparticular,we findthatthe scenariowithq1=q2 andn1=n2, namelythescenariowheretheclockworkparametersareuniversal also amonggenerations, is allowed by observations.This isillus- trated inFig. 2 (rightpanel), whichshows theallowed values of q1−q2asafunctionofn1−n2,anddemonstratestheexistenceof viablepointsatthepointq1=q2 andn1=n2.Oneconcretepoint whichleadstothecorrectneutrinoparametersis:
Fig. 2.Valuesofq1 andq2asafunctionofn1andn2(leftpanel),anddifferencesbetweenthem(rightpanel),compatiblewiththemeasuredvaluesoftheneutrinomass splittingsandmixingangleswithin1σ forascenariowithtwoclockworkgenerations.
Fig. 3.Majoranamasses(leftpanel)andYukawacouplings(rightpanel)ofthesingletfermionsoftheclockworksector,normalizedrespectivelytomandY,forthespecific casen=10,q=2 andq=0.1 (darkblue)orq=10 (lightblue).(Forinterpretationofthecolorsinthefigure(s),thereaderisreferredtothewebversionofthisarticle.)
Y
=
⎛
⎝
0.
49 0.
89 3.
62 1.
27 3.
61 2.
54⎞
⎠
(35)withq1=q2=1.79 andn1=n2=52.
2.2. Non-vanishinguniversalMajoranamass
InthiscasethemassmatrixofthemodelisgivenbyEq. (15) and the Yukawa couplings by Eq. (27). Identifyingq as the or- derparameteroftheU(1)CWsymmetrybreaking,onecanconsider twolimitsofinterest:qq,1 andqq,1.
Fig.3showsthemassesofthesingletfermions(leftpanel)and theircorrespondingYukawacouplings(rightpanel)forthespecific casen=10,q=2,andq=0.1 (darkblue)orq=10 (lightblue);
the former case corresponds to a mild breaking of the U(1)CW symmetry and the latter to a strong breaking. Forq=0.1 one notices that the mode k and the mode n+k have very similar massesandsuggest a pseudo-Diracstructure, whichresults from themildU(1)CWbreaking;inthelimitq→0,theywouldforman exact Diracpair andhave identicalmasses. Forq=10, however, themassesofallthemodesaremarkedlydifferent.
On the other hand, the Yukawa couplings of the singlet fermionstotheleft-handedleptons,shownintherightpanel,do notdepend onthe valueofq,asdemonstratedin subsection2.1.
The phenomenology ofthescenarioqq,1 is then verysimilar to the one already discussed in subsection 2.1, while the phe- nomenologyof the scenarioqq,1 can be rather distinct from the one in the (pseudo-)Dirac case. Indeed, in this scenario one obtainsamassfortheactiveneutrinothroughtheseesawmecha- nismgivenby:
mν
≈
k
Yk2v2
Mk
.
(36)Then, since the couplings for the higher modes are expected to be O(Y),theresultingneutrinomasscanbeorders ofmagnitude largerthanthevalueinferredfromoscillationexperiments,unless Y1 and/or the gearmasses are very large, inthe same spirit asinthestandardseesawmechanism. Arelatedanalysiswas also presentedin[46].
3. Leptonflavorviolation
TheclockworkmechanismsuppressestheYukawacouplingsfor thezeromode,henceexplainingthesmallnessofneutrinomasses.
However the Yukawacouplingsforthe highermodesare ingen- eral unsuppressed andcan lead to observable effects at low en- ergies.Inparticular,theleptonflavorviolationgenericallypresent intheYukawacouplingsofthehighermodescontributes,through quantum effectsinduced by clockwork fermions,to generaterare leptonicdecays(suchasli→lj
γ
)orμ
-econversioninnuclei,with ratesthatcouldbeatthereachofcurrentorfutureexperimentsif thegearmassesaresufficientlylow.We calculate the rate for li→lj
γ
following [48–50]. For N clockworkgenerations,weobtain:B
( μ →
eγ )
3α
emv4 8π
N α=1 nα k=1
YkeαYkμα Mkα2 F
(
xαk)
2
,
where
α
emisthefinestructureconstant,nαisthenumberofgears in theα
-thgeneration, Mαk is themass ofthe k-thmode in the
Fig. 4.PredictedvalueofBr(μ→eγ)forpointsoftheparameterspacereproducing theobservedneutrinooscillationparameters,asafunctionofthemassofthefirst clockworkgear.TheblacksolidlineshowsthecurrentupperlimitfromtheMEG experiment.
α
-thgeneration(k=1,...,nα),andxαk ≡Mαk2/M2W.Theloopfunc- tionF(x)isdefinedas
F
(
x) ≡
16
(
1−
x)
4(
10−
43x+
78x2−
49x3+
4x4−
18x3logx) ,
(37) andhaslimitsF(0)=5/3 and F(∞)=2/3.Thecurrentupperbound Br(
μ
→eγ
)≤4.2×10−13 fromthe MEG experiment[51] (fora recentreview, please see[52])poses stringentconstraintsonthemassscaleoftheclockwork.InFig.4 weshow thebranchingratioexpectedforpoints reproducingthe measured neutrino parameters, assuming two clockwork genera- tions,asobtainedinthescanpresentedinsection2.1,asafunction ofthemassofthefirstclockwork gear.Itfollowsfromthefigure thattheclockworkgearsmustbelargerthan∼40TeVinorderto evadetheexperimental constraints,unlessvery finecancellations amongall contributionstothisprocess exist.Foralargernumber ofclockworkgenerations weexpectevenstrongerlowerlimitson thelightestgearmass,duetothelargernumberofparticlesinthe loop.4. Summary
Theoriginofsmallneutrinomassesremainsa mysterytothis day. The recently proposed clockwork mechanism provides new insights into this puzzle, as it naturally generates small param- eters in the effective Lagrangian. In the present work, we have scrutinizedthemechanismofneutrinomassgenerationwithinthe clockworkframework.Wehavegeneralizedtheclockwork formal- ismto include, inaddition toDirac massesandnearest neighbor interactions, also Majorana mass terms in the clockwork sector;
and we have derived analytical expressions for the masses and couplingsofthenewsingletfermionsforthespecificcasewhere theDiracmasses, Majoranamassesandnearest neighbor interac- tionsareuniversalamongallclockwork“gears”.
We have investigated in detail the impact of the Majorana masses in the clockwork sector in the generation of small neu- trinomasses.WhentheuniversalMajoranamassvanishes,thezero modeoftheclockworksectorisstrictlymasslessandformsaDirac pairwiththeactiveneutrino.Inthisframework,smallDiracneu- trinomasses can be generated fora sufficiently large numberof gears,dependingonthehierarchybetweenthemassscalesinthe clockworksector.Ontheotherhand,whentheuniversalneutrino massisnon-vanishing,thezeromodeisnolongermassless.How- ever, the corresponding Yukawa coupling still hasthe clockwork structure.Inthiscase,thecontributionfromthisparticularmode
istheresultoftheinterplaybetweenthe standardseesawmech- anism andthe“clockworked”Yukawacouplings.The contribution fromthegearsistypicallyproportionaltotheirO(1)Yukawacou- plingsandtheyrequirea verylargeMajoranamassscaleinorder to reproduce the smallneutrino massesinferred from oscillation experiments.
TheStandardModelleptonscoupletothefermionsoftheclock- worksectorwithasitedependentstrength,givingriseto(possibly leptonflavor violating)chargedcurrent, neutralcurrentandHiggs boson interactions. We have investigated the constraints on this framework from the non-observation of the rare leptonic decay
μ
→eγ
.Ourresultsindicatethatthelightestparticleoftheclock- worksector musthavea mass40 TeV,iftheYukawacouplings ofthefundamentaltheoryareO(1).Acknowledgements
AIandSKVacknowledgepartialfinancialsupportfromtheDFG cluster of excellence EXC 153 “Origin and Structure of the Uni- verse”andAIfromtheCollaborativeResearchCenterSFB1258.SKV thanksthePhysicsDepartmentoftheTechnicalUniversity ofMu- nichforhospitality.SKVthanksthehospitalityofIPHT,CEA,Saclay duringthefinalstagesofthiswork.Healsoacknowledgessupport from Department of Science & Technology, Govt. of India, Grant No. EMR/2016/001097, ‘Nature of New Physics Beyond Standard Model’.
References
[1]P.Minkowski,μ→eγ atarateofoneoutof109muondecays?,Phys.Lett.B 67(1977)421–428.
[2]R.N.Mohapatra,G.Senjanovic,Neutrinomassandspontaneousparityviolation, Phys.Rev.Lett.44(1980)912.
[3]T. Yanagida, Horizontal symmetry and masses of neutrinos,Conf. Proc. C 7902131(1979)95–99.
[4]M. Gell-Mann,P.Ramond,R.Slansky,Complexspinorsand unifiedtheories, Conf.Proc.C790927(1979)315–321.
[5]J.Schechter,J.W.F.Valle,NeutrinomassesinSU(2)×U(1)theories,Phys.Rev.
D22(1980)2227.
[6]R.N.Mohapatra,etal.,Theoryofneutrinos:awhitepaper,Rep.Prog.Phys.70 (2007)1757–1867.
[7] W.Wang,R.Wang,Z.-L.Han,J.-Z.Han,TheB−Lscotogenicmodelsfordirac neutrinomasses,2017.
[8]L.M.Krauss,S.Nasri,M.Trodden,Amodelforneutrinomassesanddarkmatter, Phys.Rev.D67(2003)085002.
[9]W.Wang,Z.-L.Han,NaturallysmallDirac neutrinomasswith intermediate SU(2)L multipletfields,J.HighEnergyPhys.04(2017)166.
[10]E.Ma,O.Popov,PathwaystonaturallysmallDiracneutrinomasses,Phys.Lett.
B764(2017)142–144.
[11]D.Borah,A.Dasgupta,NaturallylightDiracneutrinoinleft–rightsymmetric model,J.Cosmol.Astropart.Phys.1706 (06)(2017)003.
[12]D.Borah,A.Dasgupta,Commonoriginofneutrinomass,darkmatterandDirac leptogenesis,J.Cosmol.Astropart.Phys.1612 (12)(2016)034.
[13]C.Bonilla,E.Ma,E.Peinado,J.W.F.Valle,Two-loopDiracneutrinomassand WIMPdarkmatter,Phys.Lett.B762(2016)214–218.
[14]S.Kanemura,K.Sakurai,H.Sugiyama,Probingmodelsofdiracneutrinomasses viatheflavorstructureofthemassmatrix,Phys.Lett.B758(2016)465–472.
[15] H.Okada, Twoloop induced Diracneutrino modeland dark matterswith globalU(1)symmetry,2014.
[16]Y.Farzan,E.Ma,Diracneutrinomassgenerationfromdarkmatter,Phys.Rev.
D86(2012)033007.
[17]P.-H.Gu,U.Sarkar,Radiativeneutrinomass,darkmatterandleptogenesis,Phys.
Rev.D77(2008)105031.
[18]D.Chang,R.N.Mohapatra,SmallandcalculableDiracneutrinomass,Phys.Rev.
Lett.58(Apr1987)1600–1603.
[19]M.Lindner,T.Ohlsson,G.Seidl,SeesawmechanismsforDiracandMajorana neutrinomasses,Phys.Rev.D65(Feb2002)053014.
[20]K.Choi,S.H.Im,RealizingtherelaxionfrommultipleaxionsanditsUVcom- pletionwithhighscalesupersymmetry,J.HighEnergyPhys.01(2016)149.
[21]D.E.Kaplan,R.Rattazzi,Largefieldexcursionsandapproximatediscretesym- metriesfromaclockworkaxion,Phys.Rev.D93 (8)(2016)085007.
[22]N. Arkani-Hamed,A.G.Cohen,H. Georgi,(De)constructingdimensions,Phys.
Rev.Lett.86(2001)4757–4761.
[23]C.T.Hill,S.Pokorski,J.Wang,GaugeinvarianteffectiveLagrangianforKaluza–
Kleinmodes,Phys.Rev.D64(2001)105005.
[24]G.F. Giudice,M. McCullough, A clockwork theory, J. High Energy Phys. 02 (2017)036.
[25] N.Craig,I.GarciaGarcia,D.Sutherland,Disassemblingtheclockworkmecha- nism,2017.
[26] G.F.Giudice,M.McCullough,Commenton“Disassemblingtheclockworkmech- anism”,2017.
[27]P.Saraswat,Weakgravityconjecture andeffectivefieldtheory,Phys.Rev.D 95 (2)(2017)025013.
[28]A.Kehagias,A.Riotto,Clockworkinflation,Phys.Lett.B767(2017)73–80.
[29]M.Farina,D.Pappadopulo,F.Rompineve,A.Tesi,Thephoto-philicQCDaxion, J.HighEnergyPhys.01(2017)095.
[30] A.Ahmed,B.M.Dillon,ClockworkGoldstonebosons,2016.
[31]T. You, A dynamical weak scalefrom inflation, J. Cosmol. Astropart.Phys.
1709 (09)(2017)019.
[32] A.Diez-Tejedor,D.J.E.Marsh,Cosmologicalproductionofultralightdarkmatter axions,2017.
[33] B.Batell,M.A.Fedderke, L.-T.Wang,Relaxationofthecomposite Higgslittle hierarchy,2017.
[34] W.Tangarife,K.Tobioka,L.Ubaldi,T.Volansky,Relaxedinflation,2017.
[35] R.Coy,M.Frigerio,M.Ibe,Dynamicalclockworkaxions,2017.
[36] I.Ben-Dayan,Generalizedclockworktheory,2017.
[37] D.K.Hong,D.H.Kim,C.S.Shin,Clockworkgravitoncontributionstomuong−2, 2017.
[38] H.M.Lee,GaugedU(1)clockworktheory,2017.
[39] M.Carena,Y.-Y.Li,C.S.Machado,P.A.N.Machado,C.E.M.Wagner,Neutrinosin largeextradimensionsandshort-baselineνeappearance,2017.
[40] I.Antoniadis,A.Delgado,C.Markou,S.Pokorski,Theeffectivesupergravityof littlestringtheory,2017.
[41] L.E.Ibanez,M.Montero,AnoteontheWGC,effectivefieldtheoryandclock- workwithinstringtheory,2017.
[42] A.Kehagias,A.Riotto,Theclockworksupergravity,2017.
[43] J.Kim,J.McDonald,AclockworkHiggsportalmodelforfreeze-indarkmatter, 2017.
[44]G.vonGersdorff,NaturalfermionhierarchiesfromrandomYukawacouplings, J.HighEnergyPhys.09(2017)094.
[45]T.Hambye,D.Teresi,M.H.G.Tytgat,AclockworkWIMP,J.HighEnergyPhys.07 (2017)047.
[46] S.C.Park,C.S.Shin,Clockworkseesawmechanisms,2017.
[47] F.Capozzi,E.DiValentino,E.Lisi,A.Marrone,A.Melchiorri,A.Palazzo,Global constraintsonabsoluteneutrinomassesandtheirordering,2017.
[48]S.T.Petcov,Theprocessesmu→eGamma, mu→eeanti-e,neutrino’→ neutrinogammaintheWeinberg–Salammodelwithneutrinomixing,Sov.J.
Nucl.Phys.25(1977)340,Erratum:Yad.Fiz.25(1977)1336.
[49]S.M.Bilenky,S.T.Petcov,B.Pontecorvo,Leptonmixing,mu→e+gammadecay andneutrinooscillations,Phys.Lett.B67(1977)309.
[50]T.P.Cheng,L.-F.Li,μ→eγintheorieswithDiracandMajorananeutrinomass terms,Phys.Rev.Lett.45(1980)1908.
[51]A.M.Baldini,etal.,Searchfortheleptonflavourviolatingdecayμ+→e+γ withthefulldatasetoftheMEGexperiment,Eur.Phys.J.C76 (8)(2016)434.
[52] L.Calibbi,G.Signorelli,Chargedleptonflavourviolation:anexperimentaland theoreticalintroduction,2017.