Contents lists available atScienceDirect
Physics
Letters
B
www.elsevier.com/locate/physletb
Exit
from
inflation
with
a
first-order
phase
transition
and
a gravitational
wave
blast
Amjad Ashoorioon
Institutionenförfysikochastronomi,UppsalaUniversitet,Box803,SE-75108Uppsala,Sweden
a
r
t
i
c
l
e
i
n
f
o
a
b
s
t
r
a
c
t
Articlehistory:
Received27April2015
Receivedinrevisedform11June2015 Accepted11June2015
Availableonline15June2015 Editor:M.Trodden
Indouble-fieldinflation,whichexploitstwoscalarfields,oneofthefieldsrollsslowlyduringinflation whereastheotherfieldistrappedinameta-stablevacuum.Thenucleationratefromthefalsevacuum tothetrueonebecomessubstantialenoughthattriggersafirstorderphasetransitionandendsinflation. Werevisitthequestionoffirstorderphasetransitioninan“extended”modelofhybridinflation,realizing thedouble-fieldinflationaryscenario,andcorrectlyidentifytheparameterspacethatleadstoafirstorder phasetransitionattheendofinflation.We computethegravitationalwaveprofilewhichisgenerated duringthisfirstorderphasetransition.Assuminginstantreheating,thepeakfrequencyfallsinthe1GHz to10GHzfrequencybandandtheamplitudevariesintherange10−11
GWh210−8,dependingon
thevalueofthecosmologicalconstantinthefalsevacuum.Foranarrowbandofvacuumenergies,the firstorderphasetransition canhappenafterthe endofinflationviatheviolationofslow-roll,witha peakfrequency thatvaries from1 THzto 100THz. Forsmallervalues ofcosmologicalconstant, even though inflation can endvia slow-roll violation, the universe gets trapped ina falsevacuum whose energydrivesasecondphaseofeternalinflation. Thisrangeofvacuumenergiesdonotleadtoviable inflationarymodels,unlessthevalueofthecosmologicalconstantiscompatiblewiththeobservedvalue,
M∼10−3eV.
©2015TheAuthor.PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBYlicense (http://creativecommons.org/licenses/by/4.0/).FundedbySCOAP3.
1. Introduction
The mechanism that ends inflation is still an open question incosmology. In factit was the Achilles heelofthe old inflation model[1],astheuniversewouldneverrecoverfrominflation. Ac-celerated expansion of the inflationary false vacuumbackground pushesthewallsoftruevacuumbubblesexponentiallyapartsuch that theynever manageto coalesceandendinflation. New infla-tion [2], with the evolution of the inflaton, allowed for another mechanism of termination of inflation. Fast-roll evolution of the scalar field at the endof inflation can violate the slow-roll con-ditionandendinflation. Howeverthere still remainedtoaddress howtheuniversehasbeenreheatedfromthesupercooledphase. Forthat onehastwoassume thattheinflatoniscoupledtoother degrees of freedom and decay of the inflaton to these degrees offreedomhappenseitherperturbativelyornonperturbatively[3]. Forthistohappenoneoftenhastofine-tunethecouplingsorthe bare-massesofthesenewdegreesoffreedom.
E-mailaddress:amjad.ashoorioon@physics.uu.se.
End ofinflation through bubblenucleation hasthis advantage thatreheatingtheuniversehappensfromthecollisionoftrue vac-uum bubble wallsand their natural conversion to radiation. The ideasofnewandoldinflationwerecombinedin“double-field” in-flation[4,5]inwhichoneofthefieldsrollsduringinflation,asin slow-roll inflation,andthe secondfield isinitially trappedin the meta-stable vacuum(ascanbe seeninourexample,thisvacuum could betheonlyexisting minimuminthebeginningofinflation. Thetruevacuumcoulddevelop asinflationproceeds).Asthefirst field rolls, the nucleation rate from the meta-stable vacuum to the trueone becomes large enough that bubblesof truevacuum canindeedpercolateandendinflation. Thisisvery similartothe false-vacuumdominatedHybridinflation[6,7]wherethetachyonic instabilityofthewaterfallfieldisreplacedwithafirstorderphase transition.Aprototypeofsuchapotentialtakestheform
V
(φ, ψ )
=
V0+
V1(φ)
+
V2(φ, ψ ),
(1.1)where V0 is the vacuumenergy which is constant and
φ
is therollingfieldwhosepotentialV1
(φ)
alongwiththevacuumenergydrivesinflation.
ψ
isthefieldwhichfacilitatesthefirstorderphase transitiontothetruevacuum.http://dx.doi.org/10.1016/j.physletb.2015.06.022
0370-2693/©2015TheAuthor.PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBYlicense(http://creativecommons.org/licenses/by/4.0/).Fundedby SCOAP3.
firstorderphasetransition,thespectrumhasaninvertedV-shape whichisdeterminedby thepeakfrequencyandtheamplitude at thepeak frequency. The peak frequency alsodepends onthe re-heatingtemperatureafterthefirstorderphasetransition,butifall theenergyinthefalsevacuumphaseisconvertedtoradiation,the peakfrequencyfallssomewhereintheGHzrange,whichisinthe frequencyrange probed by Chongqing University [13] butbelow itscurrentsensitivitylimitby10−2–10−5.Nonetheless,itishoped
thatfutureimprovementsofthe detectorsimprovetheir sensitiv-ityandclosethegapbetweenthepredictedsignalandthecurrent sensitivitylimit.Theshapeoftheproducedgravitationalspectrum canbeusedtodistinguishthemechanismofterminationof infla-tionfromtheparametricresonance[3]whichleadstogravitational wavespectrumwithdifferentprofile[14,15].
We also investigate the gravitational profile from inflationary modelsthatexitfrominflationaryphasethroughviolationof slow-rollbutaftertheterminationofinflationthenucleationratefrom
meta-stable vacuum to the true vacuum becomes large enough
thatfirstorderphase transitionoccurs.Ifthereheatingis instan-taneousandefficient,thepeak frequencyfromsuchphase transi-tionsliesintheteratopeta-Hertz bandanditsamplitudeismuch smallerthanthemodelsthatexitfrominflationthroughfirstorder phase transition. At the moment there is no planned probe that aimstothedetectionofsuchhighfrequencygravitationalwaves.
Inthiswork,wechosetherollingpotentialtobethequadratic potential for simplicity. Also, the large-scale predictions of the modelwerevery closetom2
φ
2 predictionin thenS
−
r planein thelimit ofvanishingvacuumenergy.Thispartofthepredictions ofthemodelwasstill withinthe 2σ
limitofPlanck2013results[16].AftertherevelationofPlanck2015results[17],theprediction ofthemodelnowfallsoutsidethe2
σ
region inthens−
r plane excluding the running ofscalar spectral indexfrom the parame-ters.Nonetheless,theregionclosetothepredictionpointofm2φ
2 is still within the 3σ
region. Allowing for the running of scalar spectralindex, thisregioncomes back tothe2σ
C.L. region.Itis expectedthattakingtherollingpotentialtobealowenergyscale model,likethehilltop model[18] ortheStarobinsky model[19], thepredictedvaluesforr forthe modelsexiting inflation witha firstorderphasetransitionhavea largeroverlapwiththe2σ
re-gionofPlanck2015data.Theoutlineofthepaperisasfollows.Firstwewillexplainthe setupthat canrealizetheidea ofdouble-fieldinflation.We iden-tifytheparameterspaceforwhichinflationcanendthroughafirst orderphase transitionandthen calculatethe power spectrumof gravitationalwavesproducedduring such phase transitions.Then weidentifytheregionofparameterspacewherephasetransition happensafter the termination ofinflation through slow-roll vio-lation.We alsocomputethegravitationalsignatureofsuch phase transitions.Attheend,weshowthattheuniversegetstrappedin themeta-stablevacuumifthevacuumenergyislessthana
thresh-inflation. For large values of
φ
, the potential has one minimum in bothφ
andψ
direction. As inflation proceedsandφ
rolls to-warditsvacuum,secondminimumalongtheψ
directiondevelops ifγ
2>
4αγ
atφ
inflection2=
M2γ
2
−
4α
λ
4
λ
λ
(2.2)Thetwo minimaareseparatedby abarrierthat allowsforafirst order phase transition from the meta-stable vacuumto the true one ifthe nucleation rateis substantial andwhen it is energeti-callyfavorable.Mappingthepotentialto(1.1),V1
(φ)
isthevanillachaotic model m2
φ
2 and theexpression in the second lineplays the roleof V2(φ,
ψ)
,whichcouples theinflatonto thefield thatcreates the false vacuum in the
ψ
direction and facilitates the phasetransition.Inprinciple,V1(φ)
determinesthepredictionsofthemodelatlargescales inthelimitingcasethatthe cosmologi-calconstantgoestozeroandcanbechosensuchthatthemodelis compatiblewiththeCMBobservablesatcosmologicalscales[16].
The probability of phase transition is givenby the nucleation ratedividedbythe4-dimensionalHubblevolume
p
=
H4 (2.3)
where
isthenucleationrate[20]
=
A
exp(
−
SE)
(2.4)whereSE istheEuclideanfour-dimensionalactionforthesolution thatinterpolatesbetweenthetwominima.Forafirstorderphase transition witha quartic polynomial potential, the numerical re-sultswerefitby[21]tohavetheform
SE
=
4π
23
λ
(
2− δ)
−3
(
α
1
δ
+
α
2δ
2+
α
3δ
3) ,
(2.5)where
α
1=
13.
832,α
2= −
10.
819,α
3=
2.
0765,andδ
isafunc-tionof
φ
2,δ
=
9λ
α
γ
2+
9λλ
φ
2γ
2M2.
(2.6)The allowed range has 0
< δ <
2. PrefactorA
has dimension mass4.In[8],thisprefactoristakentobeequaltothe cosmologi-calconstantinthefalsevacuum, λM44.Howeveraspointedoutby1 ForthestatusofhybridinflationwithquadraticrollingpotentialafterPlanck 2013datarelease,pleasesee[9].Inparticularthepaperdiscussesaninteresting scenarioinwhichthefirstpartoftherequirednumberofe-folds,neededtosolve theproblemsofBigBang,isprovidedfromthechaoticphaseandthesecondpart isresultedfromthevacuumdominatedphaseofhybridinflation.Themodelcan renderlowerscalarspectralindicesbutlargeramountoftensor-to-scalarratioin comparisonwiththepurelyquadraticpotential.
Fig. 1. The behavior of m (left) and nS(right) as a function of M in the region of parameter space that inflation ends with a first order phase transition.
Lindein[10],ifM2
ψ
>
2H2,thepre-exponentialfactorA
iscloser to M4ψ.As wewill see,uponthischange,theparameter spacein whichexitfrominflation can happenthrough afirst orderphase transitionexpandssubstantially.2Forillustrationwefocusonthefollowingsetofparameters
α
=
0.
01,
λ
=
1,
&λ
=
1.
(2.7)Allmassiveparameters are inunitof Planckmass,mP
=
G−1/2=
1
.
2209×
1019 GeV.Forthepotential (2.1)tohaveazerovacuum constantafterthephasetransition,thefollowingrelationbetween theparametersshouldholdγ
2=
−
α
3
+
9α
λ
2+ (
α
2+
3λ
2)
3/23 (2.8)
Asin[8],wetaketherequirednumberofe-foldstosolvethe stan-dardBigBang cosmologyproblemstobe55.Inflationcanalsoend throughendofslow-rollwhichisparameterizedby
≡
m 2 Pl 16π
V 1 V1 2=
m4φ
2m2Pπ
(λ
M4+
2m2φ
2)
2=
1 (2.9)whichhasthebiggestsolution
φ
2=
m2m2Pl+
mmP m2m2 Pl−
8π
λ
M4−
4π
λ
M4 8π
m2,
(2.10)asthephysicallyviablesolution.
In order for inflation to end through phase transition rather thanslow-rollviolation,
φ
pt> φ
.Thenumberofe-foldingscouldbeobtainedasafunctionoftheinflatonfield.Asstatedabove,we taketherequirednumberofe-foldsequalto55:
N
(φ
∗, φ
pt)
= −
8π
m2P φpt φ∗ V1 V1 dφ
=
2π
λ
M 4 m2m2 P lnφ
∗φ
pt+
2π
m2P(φ
2 ∗− φ
2pt)
(2.11)From theabove one canobtain
φ
∗.The scalarspectral indexand thetensoroverscalarratiothencouldbeobtainedatφ
∗ usingnS
−
1= −
6+
2η
;
(2.12)r
=
16,
(2.13)2 IfM2
ψ<2H
2,theprefactorisoforderH4[11].
where
isgivenaboveand
η
≡
m2P 8π
V V=
m2m2P 2π
(λ
M4+
2m2φ
2)
.
(2.14)3. Endofinflationwithafirstorderphasetransition
Inthissectionwefocusontheregionofparameterspacewhere beforetheviolationoftheslow-rollparameter,thenucleationrate becomes substantialenoughthat leadstopercolationoftrue vac-uumbubblesintheseaofinflatingfalsevacuum.Weassumethat the pre-exponential factor,
A
is of order M4ψ, as in [10], which makesourcalculationsdistinctfromtheanalysisof[8].Aswewill see,thisassumptionwillexpandtheregionofparameterspacein whichendofinflationhappensthroughafirstorderphase transi-tion. Thispart ofparameter spaceispart ofthe regionin which thevacuumenergycontributiontothepotentialiscomparableto the energy densityof the inflaton field
φ
. Detailed computation showsthatonlyintheregionwhere4
.
97×
10−4M2.
66×
10−3,
(3.1)first order phase transition precedes the slow-roll violation. For
M
2.
66×
10−6, there is no solution to the density perturba-tionamplitudenormalizationandhencenoviablemodel.Onecan matchtheamplitudeofdensityperturbations withtheCOBE nor-malizationwhich fixes themassparameter ofthe scalarfield,m.IntheleftplotofFig. 1,wehavegraphedhowthemassparameter variesasafunctionofM.Theobtainedrangeofm is
1
.
24×
10−6m6.
33×
10−6.
(3.2)Withintherange(3.1),nS andr varyinthefollowingranges
0
.
91nS0.
97,
(3.3)0
.
15r0.
55,
(3.4)asintherightgraphinFig. 1andleftplotofFig. 2.Wehavealso plotted the predictions of this region of parameter space in the
nS
−
r plane.As can be seen fromthe plots, withincreasing M, initially nS decreases andr increases.However there are turning points: around M2.
2×
10−3,nS starts to increase andaround
M
2.
4×
10−3,r startstodecrease.Thelociofpredictionsofthe modelare designatedinthenS−
r planeintherightplotofFig. 2. ForsmallvaluesofM,thepredictionsofthemodelsforlargescale fluctuationsare veryclosetothechaoticquadraticpotential.Withthechoiceofparametersasin(2.7),thetrueminimumin the
ψ
directionappearsverylate,i.e. towardstheendofinflation.Fig. 2. The behavior of r vs. M (left) and vs. nS(right) in the region of parameter space that inflation ends with a first order phase transition.
Thereis nominimumfor
ψ
that abubble oftruevacuumforms andleads toopeninflationaryscenario,asin[12].If the energy stored in the inflaton potential is completely transformed to the radiation,i.e. reheating is instantaneous, one cancalculatethereheatingtemperatureattheendofinflation
T∗
=
90H2f 8π
3g ∗ (3.5)where g∗ is the total number of relativistic degreesof freedom attemperatureT∗ which we take tobe g∗
106. Wehave plot-tedT∗ vs. M intheleftplotofFig. 3.Ifreheatingisnotefficient,T∗ is smaller than what is given in (3.5). This temperature de-termines the peak frequencyof the gravitationalwave spectrum generatedthroughthefirstorderphasetransition. Ifoneassumes instantreheating,thereheatingtemperaturevariesalmostlinearly intherangeof 1015GeV and1016GeV vs.M intherange(3.1).
3.1.Gravitationalwavespectrumfrommodelswithfirstorderphase transitionattheendofinflation
Sinceinflationsupercoolstheuniverse,onecanexploitthe for-malism of gravitational wave generation from first order phase transition at zero temperature. The numerical computations for bubblecollisionsfroma firstorderphasetransitionwere initially done by [24] for two bubbles andgeneralized for more bubbles in[25].The spectrumhasthe shapeofasymmetric ofinvertedV aroundapeakfrequency, fm,whichdecayslike f2.8 and f−1
re-spectivelyatsmallerandlargerfrequencies.Thepeakfrequency fm today,afterthepost-inflationaryredshiftingistakenintoaccount, isgivenby[24] fm
=
3×
10−10 g∗ 1001/6 T∗ 1 GeV
β
Hf,
(3.6)andtheamplitudeatthispeakfrequencyisgivenby
GWh2
(
fm)
=
10−6 g∗ 1001/3Hf
β
2,
(3.7) whereβ
isβ
=
dSE dt=
dSE dφ
dφ
dt.
(3.8)β
−1 isameasureofhowfastthefirstorderphasetransitiontakes tocomplete. Forour computations tobe reliable,we expect thistime tobemuchsmallerthantheexpansion rateoftheuniverse,
i.e.
β/
Hf1,where Hf is theHubbleparameter atthetime of phase transitionwhich coincides withthe endof inflation. dφdt is thevelocityofthescalarfield,φ
,whichcanbefoundduring infla-tionthroughthefollowingrelation[26]˙φ
2=
2(φ)
V1(φ)
3
−
(φ)
(3.9)where
isthefirstslow-rollparameter.Wehaveplottedlog
β Hf
vs. M intherange(3.1)ofM thatleadstoexitfrominflationwith a firstorder phasetransition. As can beseen fromthe rightplot ofFig. 3,forsmallervaluesofM intherange,phasetransitionis quite fastincomparisonwiththeexpansion rateofthe universe,
β/
Hf∼
few×
100.As M increases, phasetransitiontakeslonger to complete in comparisonwith the expansion time of the uni-verse.ForthemaximumvalueofM intherange,M=
2.
66×
10−3,β/
Hf=
6.
11, which is nonetheless fast enough to validate our computations.Theintensityofthegravitationalwavesatthepeak frequencyis onlydependent on thisparameter,β/
Hf.The faster the phase transition, the smaller the amplitude of the produced gravitationalwaves.Wehaveplottedlog(
GWh2)
vs. M which in-creasesintherange(3.1)from1.
07×
10−11to2.
63×
10−8,seetherightplotinFig. 4.Howeverthepeakfrequency,besides
β/
Hf de-pendsonthereheatingtemperaturetoo.Assumingthatthe reheat-ingisinstantaneousandthetotalenergydensityofpotentialatthe endofinflation transformstoradiation,wehaveplottedthepeak frequency asa function of M inthe range (3.1),see left plot ofFig. 4.ForthesmallestvalueofM intherange,M
5×
10−4,peak frequency, fm is around 1.
63×
1010Hz whereas for the largest valueof M, fm1.
75×
109 Hz.Such agravitationalwave profile liesin thefrequencyband ofChongqing High Frequency Gravita-tional (HFGW) probe but belowits currentsensitivity limit by a factorof10−2–10−5 [13].Itis expectedthatthe improvementof thedetectorinfutureclosethegapbetweentheexpectedsignals andthecurrentsensitivitylimit.If the reheating coming from bubble wall collision is not in-stantaneous anda phase ofnon-radiation domination intervenes the end of first order phase transition and the radiation domi-nation, T∗ willbe lowered.Ifthisefficiencyfactor isassumedto be
χ
, whereχ
1, the amplitude at the peak frequency will be lowered by a factorofχ
4 [30]. For exampleifχ
0
.
01–0.
1,thefrequencyrangewillbeshifted withinthesensitivityband of Birmingham HFGW probe [27]. If the reheatingtemperature isa factorof
χ
10−7–10−10smaller,thepeakfrequencywillshifttoFig. 3. The behavior of T∗andHβ
f
vs. M in the region of parameter space that inflation end with a first order phase transition.
Fig. 4. The peak frequency of the gravitational wave spectrum and its amplitude as a function of M for models that exit inflation with a first order phase transition.
thesensitivitybandofDECIGO[28] andBBO[29].Howeverinall thesecases,theamplitude ofthesignalwillbeloweredsuchthat thesignalcouldnotbeobservedbyanyoftheseprobes.
With the choice of V1
(φ)
as the quadratic potential, most of the parameter space of inflationary models that exit from infla-tion with a first order phase transition is ruled out. There is a smallregionofparameterspacewhichhaspredictionsvery close to m2φ
2. This region inparticular is ofinterest forthe CMB po-larization probes likethe BICEP2[22] orthefuture oneslikethe CMBPol[23].Sincethesemodelsexitinflation throughfirst order phase transition which is accompanied by bubble collision, they canleave anextra signatureofgravity wavesathigherfrequency scales. We expect that the predictions of the model for r couldbeloweredifonewouldreplace V1 withanothermodel,likethe
hilltopmodel,[18],ormodelswithsmallerenergyscalesthat nat-urallypredictalowervalueforr.
4. Firstorderphasetransitionaftertheendofslow-rollinflation Itispossiblethateventhough inflationendsthroughviolation oftheslow-rollinflation, thefirstorderphase transitionhappens after the end of inflation, when
φ
becomes closer to the meta-stablevacuum,φ
=
0.Infactif3
×
10−6M4.
88×
10−4 (4.1)phase transitionhappensafterthe endofslow-roll inflation. The lowerboundoftheaboveintervalisintriguinglyveryclosetothe
scalar fieldmassin therollingdirection. Tobeable todetermine thisrangeofM,wesolvedtheequationsofmotionandcomputed theevolutionofthescalarfieldafterinflationnumerically.3As
be-fore,wehavetopickuponlythe
φ
solutiontothenucleationrate equation that 0< δ <
2.We haveplottedthevariation ofm asa function of M in the range (4.2), see the left plot in Fig. 5. As-suming that thereheatingis instantaneous, wehave alsoplotted the reheating temperaturein thisrange of M, see the rightplot inFig. 5.Inbothcases,parametersm and T∗ arealmost constant for smaller values of M inthe range,but they gradually rise asM increases. As expected, contrary to the behavior of r, nS de-creaseswiththeriseofM, pleaseseeFigs. 5 and 6.Wehavealso plotted the behavior of nS vs. r as M increases, see Fig. 6. The behavior of log
Hβf
vs. M, where Hf is the Hubble parameter when phasetransitioncompletescan beseen intheright plotin
Fig. 7.Ascanbeseenfromtheplot,withtheincreaseofM,phase transition becomes slower.Nonetheless, in general, phase transi-tions happen much faster after inflation incomparison with the inflationary models that endviafirst orderphase transition. This
3 Intherange
4.88×10−4M4.97×10−4, (4.2)
theconstraintsonthenumberofe-folds,Ne=55,withendofinflationgivenbythe
slow-rollviolationdidnotyieldarealsolutionform.Onecansatisfytheconstraint equationswithlessnumberofe-foldsthough.
Fig. 5. m (left) and nS(right) vs. M for the models for which first order phase transition happens after the end of inflation.
Fig. 6. r vs. M (left) and nS(right) for the models for which first order phase transition happens after the end of inflation.
Fig. 7. T∗(left) and logβHf
(right) vs. M for the models for which first order phase transition happens after the end of inflation.
ispartlyduetothefactthat theHubbleparameterafterinflation ends,issmaller than itscorresponding value at theendof infla-tion.As M enhances inthe interval, (4.2),
β/
Hf decreases from 4.
21×
106 to 90321.
6.We havecheckedthat inthisrangeof M,theeffectivemassof
ψ
field,M2ψ
>
2H2andthusphasetransition happensviaColeman-deLucciainstantontransitions[20].Using Eqs. (3.6) and (3.7), we have calculated the peak fre-quency of the gravitational wave spectrum and its intensity,
GWh2,as a function of M, please see the plotsin Fig. 8.As M
increases, fm variesfrom1
.
8×
1014 to 4.
95×
1012Hz,which is well outside thefrequency band ofany currentlyplanned probe. Of course this frequency range is obtained assuming that the reheating is instantaneous and efficient. If the reheating tem-perature is smaller than its instantaneous value by a factor ofχ
10−3–10−4,thepeak frequencyrangeismovedto the sensi-tivitybandofChongqinghighfrequencygravitationalwave probe.Fig. 8. Thepeakfrequency(left)andtheamplitudeatthepeakfrequency(right)vs.M forthemodelsofinflationinwhichfirstorderphasetransitionhappensafterinflation.
However,theintensitywillthengetssuppressedfurtherbyafactor of10−12–10−16whichmakesthesignaltoosmalltobedetected. 5. Trappinginthemeta-stablevacuumandeternalinflation
Forsmallervaluesof M, M
<
3×
10−6,the nucleationrateasgivenby Coleman–DeLuccia transitionrate,is toosmalltoallow forthefirstorderphasetransitiontocompleteasthescalarfield,
φ
,passesthroughtheminimumofthepotential, V1.Graduallyastheeffective massofthe
ψ
field decreases whiletheφ
field ap-proachesthemetastablevacuum,thereisachancethatM2ψ
≤
2H2. In particular thiscan happen for smaller values ofα
and M. In thissituation,Hawking–Mossphasetransition[31] appearswhere aninhomogeneousbubbles whoseradiusisgreaterthande-Sitter spaceradius, H−1,willformintheflatspace-time.Thetunneling probabilityperunit fourvolumeisoforder[31]=
H4exp(
−
B),
(5.1) where B=
1 8 1 V(
0,
0)
−
1 V(
0, ψ
max)
(5.2)
is the difference between the combined gravitational and scalar fieldactionofthe
ψ
= ψmax
,whichisanotherhomogeneous solu-tionapartfromψ
=
0. V(
0,
max
)
isthelocalmaximumof V onafour-sphere ofradius, H−max1 ,where Hmax2
=
8π
V(
0,
ψ
max)/
3.Asψ
= ψmax
isunstable,thescalarfieldrunsdownhillfromψ
max,totheglobalminimum
ψ
min,afterwards. Forthepotential (2.1),onecancalculatetheexponentB andshowthat
B
α
3
3
λ
2M4 (5.3)for
α
1.Thereforefor
M
α
3/4
31/4
λ
1/2 (5.4)theHawking–Mossphasetransitionratefromthemeta-stable vac-uumto thetruevacuumisverysmall.Inourexamplewherewe took,
λ
=
1 andα
=
0.
01,for M6.
08×
10−5 such phase tran-sition takes a lot of time to complete and basically leads to a self-reproductionregimelikeoldinflation[1].AftertheHawking– Moss transition completes, theφ
field will be homogeneous on scalaroforder H−01,with H0=
8πV(0,0)3m2 P
.Nonetheless,itwillbe stochasticallydistributedwiththedispersion[11]
φ
rms2≡ φ
2=
3H 4 08
π
2m4,
(5.5)whichinourexampleisabout
10−5andthustoosmalltodriveanotherphaseofinflation.
In [8], the authors claim that they have been able to find a branch that corresponds tothe vacuum-dominatedregime of hy-bridInflation[7]intheExtendedHybridinflation, potential(2.1). Howeverinhybridinflation,thisregimeisobtainedassumingthat thetachyonicinstabilityinthewaterfallfieldendsinflation.Inthis case, howeverthewaterfield masssquared,
α
M2+ λψ
2 isalwayspositive andnever becomes tachyonic. For such energy scales in the vacuum dominated regime, the slow-roll can never get vio-lated as the first slow-roll parameter decreases asinflation pro-gresses.We alsoshowedthat Hawking–Moss phasetransitionfor such smallvacuum energiesis not substantialenough to end in-flation. Therefore, itis notpossible to realizethevacuumenergy dominated regime of hybrid inflation in such extended models. The scalar field gets trapped in the metastable vacuum with no gracefulexitandthereforethisregionofparameterspacedoesnot yieldaviableinflationarymodel.Ontheotherhand,fromthe phe-nomenological perspective,ifM isoforder
10−30,thisvacuum energycanbe responsibleforthecurrentaccelerationofthe uni-verse.6. Conclusion
Terminatinginflationwithafirstorderphasetransitionhasthe benefit ofreheatingthe universe fromthe supercoolingphase of inflation through the collisions of bubbles of true vacuum with-out invoking and fine-tuning of the couplings of the inflaton to the otherfields.Onecan achieve thisscenario, modifyingtheold inflation scenario, with time-dependent nucleation rate which is small in the beginning and becomes substantial at the end of inflation. One specific realization ofthis scenariois extended in-flation[32],inwhichthegravity sectorofthe theoryismodified to Jordan–Brans–Dicketheory.Anotherwayofachievingthis sce-nario, is having two scalar fields, where one of the fields rolls and the other one is trapped in a meta-stable vacuum [4,5]. As the rolling field evolves, the nucleation rateat the false vacuum becomes large enough that the condition for percolationof true vacuumbubblesholdsandinflationends.Were-examinedamodel ofextendedhybridinflationwhichprovidesuswithsuch asetup. We noticedthat thepre-exponential factorinthe nucleationrate plays a crucial role incorrectlydetermining the parameterspace that allows for a first order phase transition in the model. For models that exit inflation with a first orderphase transition, we computedthepeak frequencyandtheamplitudeatthepeak fre-quency, which are respectively in the ranges fm
109–1010Hz andGWh210−11–10−8.Thesignatureisinthefrequencyrange
thesignal ofsuch a phase transitionisweak andout ofthe fre-quencybandoffutureprobes.
Thelarge-scalepredictionsofthemodelareverydependenton thepotentialoftherollingfield.Herewiththechoiceofquadratic potential for the rolling field, the predictions of the model ap-proach thepredictionsofm2
φ
2inthenS
−
r plane,whichwasstill withinthe2σ
limitofPlanck2013results[16].Asthismodel pro-ducesgravitationalwavesat theCMBscales, the analyzedmodel producesthe observablegravitationalwavesatbothsmallandhigh frequencyrangeofthespectrum:thesmallfrequencygravitational waveshave quantum originbut the high frequency gravitational waveshaveclassicalorigin. Aftertheexposure ofPlanck2015 re-sults[17],major prediction of themodel is now outsidethe 2σ
regioninthens−
r planeexcludingtherunningofscalarspectral indexfromtheparameters.Theregionclosetothepredictionpoint ofm2φ
2isstillwithinthe3σ
regionthough.Ifoneallowsforthe runningofscalarspectralindex,thisregioncomesbacktothe2σ
confidenceregion.Oneshouldalsonotethatthepredictionsofthe modelatlargescalesare verysensitivetotheinitialcondition for fluctuations.Asitwas shownin[35],choosingexcitedinitial con-ditionfor cosmologicalperturbations generally tends to suppress thetensor-to-scalarratio.Itisexpectedthatifonetakestherollingfieldpotentialtobea lowerenergyscalemodel,likethe hilltop[18] ortheStarobinsky
[19]model,thepredictedvaluesforr forthemodelsexiting infla-tionwithafirstorderphasetransitionhavea largeroverlapwith the2
σ
confidenceregionofPlanck2015results.Thisissomething thatIwillpostponetoafuturepublication.Asthepeakfrequencyofthegravitationalspectrum,produced fromthe first orderphase transition, depends verymuch on the Hubble parameter at the time of phase transition, it is also in-terestingtoinvestigateinflationarymodelsthatexitinflation with smaller Hubble parameter and, hence, peak frequency that falls within the sensitivity bands of BBO, DECIGO or even Advanced LIGO.
Acknowledgements
I am indebted to Andrei Linde for helpful discussions. I also thank A. Abolhasani, R. Allahverdi, M. Cortes, K. Freese and A. Lopezforcommentsanddiscussions.
ph.CO].
[10]A.D.Linde,Nucl.Phys.B216(1983)421; A.D.Linde,Nucl.Phys.B223(1983)544(Erratum).
[11]A.D.Linde,Contemp.ConceptsPhys.5(1990)1,arXiv:hep-th/0503203. [12]A.D. Linde, A. Mezhlumian, Phys. Rev. D 52 (1995) 6789,
arXiv:astro-ph/9506017.
[13]F.Y.Li,M.X.Tang,D.P.Shi,Phys.Rev.D67(2003)104008,arXiv:gr-qc/0306092; M.l. Tong,Y. Zhang,F.Y.Li,Phys. Rev.D78(2008)024041, arXiv:0807.0885 [gr-qc].
[14]J.F.Dufaux,A.Bergman,G.N.Felder,L.Kofman,J.P.Uzan,Phys.Rev.D76(2007) 123517,arXiv:0707.0875[astro-ph].
[15]A.Ashoorioon,B.Fung,R.B.Mann,M.Oltean,M.M.Sheikh-Jabbari,J.Cosmol. Astropart.Phys.1403(2014)020,arXiv:1312.2284[hep-th].
[16]P.A.R. Ade, et al., Planck Collaboration, Astron. Astrophys. 571(2014) A22, arXiv:1303.5082[astro-ph.CO].
[17]P.A.R.Ade,etal.,PlanckCollaboration,arXiv:1502.01589[astro-ph.CO]. [18]A.Ashoorioon,M.M. Sheikh-Jabbari,J.Cosmol.Astropart.Phys.1106 (2011)
014,arXiv:1101.0048[hep-th];
A. Ashoorioon, M.M. Sheikh-Jabbari, Phys. Lett. B 739 (2014) 391, arXiv:1405.1685[hep-th];
A.Ashoorioon,H.Firouzjahi,M.M.Sheikh-Jabbari,J.Cosmol.Astropart.Phys. 0906(2009)018,arXiv:0903.1481[hep-th].
[19]A.A.Starobinsky,Phys.Lett.B91(1980)99.
[20]C.G.CallanJr.,S.R.Coleman,Phys.Rev.D16(1977)1762; C.G.CallanJr.,S.R.Coleman,Phys.Rev.D16(1977)1762; S.R.Coleman,F.DeLuccia,Phys.Rev.D21(1980)3305. [21]F.C.Adams,Phys.Rev.D48(1993)2800,arXiv:hep-ph/9302321.
[22]P.A.R.Ade, et al., BICEP2Collaboration,Phys. Rev.Lett. 112(2014)241101, arXiv:1403.3985[astro-ph.CO].
[23]D. Baumann,etal., CMBPolStudyTeamCollaboration,AIPConf.Proc. 1141 (2009)10,arXiv:0811.3919[astro-ph].
[24]A.Kosowsky,M.S.Turner,Phys.Rev.D47(1993)4372,arXiv:astro-ph/9211004. [25]S.J. Huber, T. Konstandin, J. Cosmol. Astropart. Phys. 0805 (2008) 017,
arXiv:0709.2091[hep-ph].
[26]A.Ashoorioon,R.B.Mann,Nucl.Phys.B716(2005)261,arXiv:gr-qc/0411056. [27]A.M.Cruise,R.M.J.Ingley,Class.QuantumGravity23(2006)6185.
[28]S.Kawamura,M.Ando,N.Seto,S.Sato,T.Nakamura,K.Tsubono,N.Kanda,T. Tanaka,etal.,Class.QuantumGravity28(2011)094011.
[29]J.Crowder,N.J.Cornish,Phys.Rev.D72(2005)083005,arXiv:gr-qc/0506015. [30]A.Lopez,K.Freese,arXiv:1305.5855[astro-ph.HE].
[31]S.W.Hawking,I.G.Moss,Phys.Lett.B110(1982)35. [32]D.La,P.J.Steinhardt,Phys.Rev.Lett.62(1989)376;
D.La,P.J.Steinhardt,Phys.Rev.Lett.62(1989)1066(Erratum).
[33]R.Easther,E.A.Lim,J.Cosmol.Astropart.Phys.0604(2006)010, arXiv:astro-ph/0601617.
[34]R.Easther,J.T.GiblinJr.,E.A.Lim,Phys.Rev.Lett.99(2007)221301, arXiv:astro-ph/0612294.
[35]A. Ashoorioon,K. Dimopoulos, M.M. Sheikh-Jabbari,G.Shiu,J. Cosmol. As-tropart.Phys.1402(2014)025,arXiv:1306.4914[hep-th].