Glueball Inflation and Gauge/Gravity Duality LiliaAnguelova 6 1 0 2 n a J Abstract We summarize our work on building glueball inflation models with the 1 1 methodsofthegauge/gravityduality.Wereviewtherelevantfive-dimensionalcon- sistent truncation of type IIB supergravity.We consider solutions of this effective ] theory,whosemetrichastheformofadS foliationoveraradialdirection.Byturn- h 4 t ing on small (in an appropriate sense) time-dependentdeformationsaround these - solutions, one can build models of glueball inflation. We discuss a particular de- p e formedsolution,describinganultra-slowrollinflationaryregime. h [ 1 1 Introduction v 9 4 Compositeinflationmodels[1,2]provideapossibleresolutiontothewell-knownh - 4 2 problem[3,4]ofinflationarymodel-building.However,theyarequitechallenging 0 to studywith standardQFT methods,sincethey involvea strongly-coupledgauge . sector. This has motivated interest in developing descriptions of such models via 1 0 a string-theoretic tool aimed precisely at studying the nonperturbative regime of 6 gaugetheories,namelythegauge/gravityduality.Gravitationalduals,inwhichthe 1 inflaton arises from the position of a D3-brane probe have been considered in [5, : v 6, 7, 8, 9]. Instead, in [10, 11, 12] we studied models, whose inflaton arises from i thebackgroundfieldsofthegravitationalsolutionandisthusaglueballinthedual X gaugetheory. r a Thebackgroundsofinterestforussolvetheequationsofmotionofthe5dconsis- tenttruncationoftypeIIBsupergravityestablishedin[13].Thelatterencompassesa widevarietyofprominentgravityduals,like[14,15,16,17,18],asspecialsolutions and thus provides a unifying framework for gauge/gravity duality investigations. Thework[10]obtainednewnon-supersymmetricclassesofsolutionsofthistheory, LiliaAnguelova InstituteforNuclearResearchandNuclearEnergy,BulgarianAcademyofSciences,Sofia1784, Bulgaria,e-mail:[email protected] 1 2 LiliaAnguelova whosemetricisoftheformofadS fibrationoverthefifthdirection.Theseback- 4 grounds provide a useful playground for studying certain strongly-coupledgauge theoriesindeSitterspace.Tohaveaninflationarymodel,however,oneneedsatime- dependentHubbleparameter.Therefore,in[12]weinvestigatedtime-dependentde- formationsaroundasolutionof[10],inordertosearchforgravitydualsofglueball inflation. Itisworthpointingoutthatthemaincosmologicalobservablesofaninflationary model(likethescalarspectralindexn andthetensor-to-scalarratior)areentirely s determined by the Hubble parameter and inflaton field as functions of time [19]. Hence,onceonehasadeformedbackgroundintheaboveset-up,onecanimmedi- ateycomputethedesiredquantities.Thisisthesense,inwhichthetime-dependent deformationsof the previous paragraph give models of cosmological inflation. In that vein, in [12] we calculated the slow roll parameters for a solution we found thereandthusestablishedthatitgivesagravitydualofultra-slowrollglueballin- flation.Theultra-slowrollregimemayplayanimportantroleinunderstandingthe observedlowl anomalyinthepowerspectrumoftheCMB.Henceitdeservesfur- therstudy.Wealsodiscusshereperspectivesforbuildinggravitydualsofstandard slowrollinflationarymodels. 2 Effective 5d theory Inthissectionwesummarizenecessarymaterialaboutthe5dconsistenttruncation oftypeIIBsupergravityrelevantforourconsiderations.Wealsorecallaparticular solution of this theory,whose time-dependentdeformationswe will investigatein thenextsubsection. 2.1 Actionand field equations Letusbrieflyreviewthebasiccharacteristicsofthefive-dimensionalconsistenttrun- cationof[13].UsingaparticularansatzforthebosonicfieldsoftypeIIBsupergrav- ity in terms of certain 5d fields and integrating out five compact dimensions, one reducestheten-dimensionalIIBactiontothefollowingfive-dimensionalone: R 1 S= d5x detg + G (F )¶ F i¶ IF j+V(F ) . (1) ij I − −4 2 Z (cid:20) (cid:21) p Here F i isasetof5dscalarfields,thatarisefromthecomponentsofthe10dones includ{ing}metricwarpfactors,V(F ) isa rathercomplicatedpotential,G (F ) isa ij diagonalsigma-modelmetricand,finally,RistheRicciscalarofthe5dspacetime metric g . The full expressionsforV(F ) and G (F ) can be found in [13]; for a IJ ij moreconcisesummary,seealso[10].Thefieldequationsthattheaction(1)implies GlueballInflationandGauge/GravityDuality 3 are: (cid:209) 2F i+Gi gIJ(¶ F j)(¶ F k) Vi = 0 , jk I J − 4 R +2G (¶ F i)(¶ F j)+ g V = 0 , (2) IJ ij I J IJ − 3 whereVi =GijV , V = ¶V and Gi are the Christoffel symbols of the sigma- j i ¶F i jk modelmetricG . ij 2.2 AsolutionwithdS slicing 4 We willbe interestedin time-dependentdeformationsarounda particularsolution ofthesystem(2)foundin[10].Soletusfirstrecallitsform.Inthenotationof[10] and working within the same subtruncation as there (i.e., with zero NS flux), we havesixscalarsinthe5deffectivetheory: F i(xI) = p(xI),x(xI),g(xI),f (xI),a(xI),b(xI) . (3) { } { } Thework[10]foundthreefamiliesofsolutionsof(2)witha5dmetricoftheform 3 ds2=e2A(z) dt2+s(t)2 (cid:229) (dxm)2 +dz2, (4) 5 "− m=1 # where s(t)=eHt with H =const. In all of them, three of the scalars F i vanish identically,namely: g(xI)=0 , a(xI)=0 , b(xI)=0 . (5) Twoofthosesolutionsarenumericalandoneisanalytical.Forconvenience,wewill studydeformationsaroundthelatter.Denotingitsmetricfunctionsandscalarfields bythesubscript0,wehave[10]: 1 7 A (z)= ln(z+C)+ ln H2 , 0 2 3 0 (cid:18) (cid:19) 1 1 7N2 p (z)= ln(z+C) ln , 0 −7 −14 9 (cid:18) (cid:19) x (z)= 6p (z) , f =0 , (6) 0 0 0 − whereCandN areconstants. Letusmentioninpassingthattheformof(6)isconsistentwithALD(asymptot- icallylineardilaton)behavioratlargez.Thisisnotobviousatfirstsightduetothe useofadifferentcoordinatesystem(instringframe)comparedtotheconventional one (in Einstein frame), in which the holographic renormalization of ALD back- 4 LiliaAnguelova grounds was developed [20]. This issue was discussed in more detail in [10, 11], where it was also pointed out that the same kind of asymptotics characterizesthe walkingsolutionsof[16]aswell. 3 Deforming the dS solution 4 Nowwearereadytoturntotheinvestigationofsolutionsofthesystem(2),which aredeformationsaroundthezerothorderbackground(6).Sinceouraimistostudy glueballinflation, we wouldlike to find solutions, whose 5d metric is of the form (4)butwithH˙ =0.RecallthattheHubbleparameterisdefinedas 6 s˙ H = , (7) s ¶ whereforconveniencewehavedenoted˙ . Now,oneoftheslowrollconditions ≡ ¶ t widelyusedininflationarymodelbuilding1isthefollowing[19]: H˙ << 1 . (8) −H2 Inviewofthat,wewilllookforsolutionswithtime-dependentH byconsidering small,inthesenseof(8),deformationsaroundanH =constsolution. Forthatpurpose,letusintroduceasmallparameterg ,satisfying g <<1, (9) andsearchforsolutionsthatareexpansionsinpowersofthisparameter.Todothis, wemakethefollowingansatzforthenonvanishing5dfields: p(t,z)= p (z) , x(t,z)=x (z) , 0 0 f (t,z) = gf (t,z)+g 3f (t,z)+O(g 5) , (1) (3) A(t,z)= A (z)+g 2A (t,z)+O(g 4) , 0 (2) H(t,z)= H t+g 2H (t,z)+O(g 4) , (10) 0 (2) whereH(t,z)isawarpfactordefinedvia 3 ds2=e2A(t,z) dt2+e2H(t,z) (cid:229) (dxm)2 +dz2 . (11) 5 "− m=1 # 1Oneshouldkeepinmind,though,thattherearemoreexoticinflationaryregimes,inwhichone ormoreoftheslowrollconditionscanbeviolated;see[21,22,23],forinstance. GlueballInflationandGauge/GravityDuality 5 Inotherwords,wekeepthescalarsp(xI)andx(xI)thesameasin(6),whileallowing small deviations around that zeroth order solution in the scalar f and the metric functionsAandH. It is worth commentinga bit more on the formof the deformationansatz (10). First of all, in orderto obtainsolutionswith H˙ =0, we need to turn on time de- pendence in at least one scalar. It is convenient6to take this scalar to be f since, unlike p andx,itvanishesatzerothorderand,furthermore,itisaflatdirectionof thepotential;see[12].Therefore,f willplaytheroleoftheinflatoninourset-up. Alsonotethat,althoughwewouldliketohavet-dependentHonly,wehaveallowed z-dependencetoo,formoregenerality.And,finally,thedifferentpowersofg inthe expansion of f , compared to the expansions of the warp factors, will be of great significanceforfindingananalyticalsolution,aswillbecomeclearbelow. 3.1 Equations ofmotion Letusnowsubstitutetheansatz(10)inthesystem(2)andstudytheresultorderby ordering .Clearly,sinceweareexpandingaroundazerothordersolution,thereis nocontributionatorderg 0. Atorderg ,wehavethefollowingfieldequation[12]: f¨(1)+3H0f˙(1)=e2A0 f (′1′)+4A′0f (′1) , (12) (cid:16) (cid:17) ¶ where .Tofindasolution,letusmaketheansatz ′≡ ¶ z f =F (t)F (z) (13) (1) 1 2 andsolvetheeigenproblems F¨1+3H0F˙1=lF 1 and e2A0 F 2′′+4A′0F 2′ =lF 2 (14) withl beingsomeconstant.Oneeasilyobtainsth(cid:0)at[12]: (cid:1) F (t)=C ek+t+C ek t , where k = 3H0 9H02+4l (15) 1 1 2 − ± − 2 ±q 2 andC areintegrationconstants,while 1,2 3 3 4l F (z)=C (z+C)a ++C (z+C)a with a = 1+ (16) 2 3 4 − ± −2±2s 21H02 andC beingintegrationconstants. 3,4 Notethatifl =0,thenoneisfreetoaddanarbitraryconstanttothef solution, (1) determinedby(14).Thiswillbeimportantinthefollowing. 6 LiliaAnguelova Atorderg 2,wefindacoupledsystemforthewarpfactordeformationsA and (2) H ,namely[12]: (2) 7 56 E1: H2 (z+C)2A + (z+C)A +7(z+C)H +6A − 0 3 ′(′2) 3 ′(2) (′2) (2) (cid:18) (cid:19) 1 +H 3A˙ +6H˙ +3A¨ +3H¨ + f˙2 =0 , 0 (2) (2) (2) (2) 2 (1) (cid:0)7 (cid:1) 56 49 E2: H02 3(z+C)2 A′(′2)+H(′2′) + 3 (z+C)A′(2)+ 3 (z+C)H(′2) (cid:18) h i +6A H 5A˙ +6H˙ A¨ H¨ =0 , (2) − 0 (2) (2) − (2)− (2) (cid:19) (cid:0) 2 (cid:1) 1 E3: 4A +3H + 4A +3H + f 2 =0 , ′(′2) (′2′) z+C ′(2) (′2) 2 (′1) (cid:16) 1 (cid:17) E4: 3A˙′(2)+3H˙(′2)+3H0H(′2)+2f˙(1)f (′1)=0 . (17) Tosolvethisratherinvolvedsystem,letustakeforconveniencethef solutionto (1) be: f (1)=Cf +C˜ekt(z+C)a with Cf ,C˜=const , (18) where k is any of k and a is any of a . Note that the addition of the arbitrary ± ± constantCf in(18)makesnodifferenceforthesolutionsof(17),sincethefunction f (1)entersthoseequationsonlythroughitsderivatives.However,thepresenceofCf willturnouttobeusefullater.Plus,itwillbecomeclearshortlythatitisconsistent with(14). Now,theformofE3in(17),togetherwith(18),suggestslookingforasolution withthefollowingansatz: A (t,z)=e2ktAˆ(z) and H (t,z)=Cˆ +e2ktHˆ(z), (19) (2) (2) H whereCˆ =const. Again, we have included an arbitrary constantCˆ , since H H H (2) entersthesystem(17)onlyviaitsderivatives.Substituting(19)and(18)into(17), onecan see thatthe time-dependencefactorsout. Thus,oneis leftwith a coupled system of ODEs for the functionsAˆ(z) and Hˆ(z). A detailed investigationin [12] showedthatthissystemhasasolutiononlyfor2 a =0 , (20) in which case both Aˆ =const and Hˆ =const. Substitutinga =0 in (16), we find thatl =0 as well. This in turn implies thatwe are free to add the constantCf in (18),ascommentedbelow(16).Finally,from(15)wenowhave: k= 3H , (21) 0 − 2 To prove this, one also needs to use the fact that (15) and (16) imply the following relation betweenkanda : k=−32H0±√84a 2+6252a +81H0 . GlueballInflationandGauge/GravityDuality 7 wherewehavetakenthevalueofk inordertohavetime-dependenceintheinflaton fieldf . − 3.2 Ultra-slowrollinflation Thesolutionwedescribedabovegivesadualdescriptionofanultra-slowrollglue- ballinflationmodel.Toseethis,letuscomputetheinflationaryslow-rollparameters. TheyaredefinedintermsoftheinflatonfieldandHubbleparameteras[19]: H˙ f¨ e = and h = . (22) −H2 −Hf˙ FromtheresultsofSubsection3.1,wehavethatf andH aregivenby: f = Cf +C˜e−3H0t g +O(g 3), H =(cid:16)H0 CH e−6H0(cid:17)tg 2+O(g 4), (23) − whereCH issomeconstant;formoredetails,see[12].3 Substituting (23) in (22), we find that the slow roll parameters behave as e = O(g 2)andh =3+O(g 2);see[12]formoredetailedexpressions.Inotherwords, atleadingorderwehave: e <<1 and h =3 . (24) These are preciselythe valuesof e andh for the ultra-slowregime,consideredin [24,25]. Infact,ourresultfortheinflatonin(23)also agreescompletelywiththe expressionin[25]. Itisworthpointingoutasimilaritybetweenourmodelandtheconstant-rate-of- rollsolutionsof[23].Forthatpurpose,letusintroducethefollowingseriesofslow rollparameters: H˙ e˙ e = and e = n , (25) 1 −H2 n+1 Hen whereobviouslye e .Onecaneasilycomputethat,atlarget,oursolutiongives 1 ≡ [12]: e 0 and e 6 . (26) 2n+1 2n → →− 3 Notethat,sincethecorrectionA tothewarpfactorA(t,z)in(11)alsodependsont,ascan (2) beseenfrom(19),oneshould,inprinciple,firstperform acoordinatetransformationt t that absorbs that dependence, before computing the physical Hubble parameter H(t ) and→inflaton fieldf (t ).However, inthepresentcase,thisleadstoexactlythesameexpressions as(23)with t substitutedbyt ,withtheonlydifferencebeingthenumericalvalueoftheconstantCH.Sowe willnotdiscussthedetailsofthattransformationhere. 8 LiliaAnguelova Thisisconsistentwiththeasymptoticsin[23].Itwouldbeinterestingtoinvestigate whetherthereisadeeperunderlyingreasonforthat. Inconclusion,letusmakeafewcommentsregardingotherinflationarymodels in our framework. Although an ultra-slow roll inflationary regime may be desir- able to account for the low l anomaly in the CMB power spectrum, it is rather short-lived.Soithastobesucceededbyregularslowroll,inordertohaveenough expansionandthusgiveacompleteinflationarymodel.Toobtainsuchsolutionsin our gauge/gravity duality set-up, one may need to study deformations around the numericalsolutionsof [10], instead of the analyticalone (6). It could also be that dualsofregularslowrollcanbefoundbymodifyingtheinitialansatzforthedefor- mationsaroundtheanalyticalsolution.Finally,itwouldbeinterestingtoinvestigate whatkindofmodelscanbeobtainedbygoingtothenextordering intheexpan- sions(10),whiletakingf ,A andH tovanish.Thisseemstoopenmuchwider (1) (2) (2) possibilitiesforinflationarymodelbuilding,astheequationsofmotionforA and (4) H wouldbe independentoff . Thus, manyof the restrictionswe encountered (4) (3) here(andasaresultofwhichweendedupwithultra-slowroll)wouldnotoccur. Acknowledgements IhavereceivedpartialsupportfromtheEuropeanCOSTActionMP-1210 andtheBulgarianNSFgrantDFNIT02/6. References 1. P.Channuie,J.Joergensen,F.Sannino,JCAP1105(2001)007,arXiv:1102.2898[hep-ph]. 2. F. Bezrukov, P. Channuie, J. Joergensen, F. Sannino, Phys. Rev. D86 (2012) 063513, arXiv:1112.4054[hep-ph]. 3. E. Copeland, A. Liddle, D. Lyth, E. Stewart, D. Wands, Phys. Rev. D49 (1994) 6410, astro-ph/9401011. 4. M.Dine,L.Randall,S.Thomas,Phys.Rev.Lett.75(1995)398,hep-ph/9503303. 5. A.Buchel,Phys.Rev.D65(2002)125015,hep-th/0203041. 6. A.Buchel,P.Langfelder,J.Walcher,Phys.Rev.D67(2003)024011,hep-th/0207214. 7. A.Buchel,A.Ghodsi,Phys.Rev.D70(2004)126008,hep-th/0404151. 8. A.Buchel,Phys.Rev.D74(2006)046009,hep-th/0601013. 9. N.Evans,J.French,K.Kim,JHEP1011(2010)145,arXiv:1009.5678[hep-th]. 10. L. Anguelova, P. Suranyi, L.C. Rohana Wijewardhana, Nucl. Phys. B899 (2015) 651, arXiv:1412.8422[hep-th]. 11. L. Anguelova, P. Suranyi, L.C. Rohana Wijewardhana, Bulg. J. Phys. 42 (2015) 277, arXiv:1507.04053[hep-th]. 12. L.Anguelova,arXiv:1512.08556[hep-th]. 13. M.Berg,M.Haack,W.Muck,Nucl.Phys.B736(2006)82,hep-th/0507285. 14. J.Maldacena,C.Nunez,Phys.Rev.Lett.86(2001)588,hep-th/0008001. 15. I.Klebanov,M.Strassler,JHEP0008(2000)052,hep-th/0007191. 16. C.Nunez,I.Papadimitriou,M.Piai,Int.J.Mod.Phys.A25(2010)2837,arXiv:0812.3655 [hep-th]. 17. D.Elander,C.Nunez,M.Piai,Phys.Lett.B686(2010)64,arXiv:0908.2808[hep-th]. 18. D.Elander,J.Gaillard,C.Nunez,M.Piai,JHEP1107(2011)056,arXiv:1104.3963[hep-th]. GlueballInflationandGauge/GravityDuality 9 19. D.Baumann,TASILecturesonInflation,ConferenceProceedingsfor“PhysicsoftheLarge andtheSmall”,TheoreticalAdvancedStudyInstituteinElementaryParticlePhysics,Boul- der,Colorado,June2009,arXiv:0907.5424[hep-th]. 20. R.Mann,R.McNees,Class.Quant.Grav.27(2010)065015,arXiv:0905.3848[hep-th]. 21. A.Linde,JHEP0111(2001)052,arXiv:hep-th/0110195. 22. J. Martin,H.Motohashi andT.Suyama, Phys. Rev. D87(2013) 023514, arXiv:1211.0083 [astro-ph.CO]. 23. H. Motohashi, A. Starobinsky and J. Yokoyama, JCAP 09 (2015) 018, arXiv:1411.5021 [astro-ph.CO]. 24. N.TsamisandR.Woodard,Phys.Rev.D69(2004)084005,astro-ph/0307463. 25. W.Kinney,Phys.Rev.D72(2005)023515,gr-qc/0503017.