Why are massive O-rich AGB stars in our Galaxy not S-stars

We present the main results derived from a chemical analysis carried out on a large sample of galactic O-rich AGB stars using high resolution optical spectroscopy (R~40,000-50,000) with the intention of studying their lithium abundances and/or possible s-p

WhyaremassiveO-richAGBstarsinourGalaxynot

S-stars?

D.A.García-Hernández ,P.García-Lario ,B.Plez ,A.Manchado ,F.D’Antona ,

J.Lub§andH.Habing§

arXiv:astro-ph/0509051v1 2 Sep 2005

EuropeanSpaceAstronomyCentre(ESAC),ESA.Apdo.50727.28080Madrid.SpainGRAAL,CNRSUMR5024,UniversitédeMontpellier2,34095MontpellierCedex5,France

InstitutodeAstrofísicadeCanarias,LaLaguna,E 38200,Tenerife,Spain

OsservatorioAstronomicodiRoma,viaFrascati33,00040MontePorzioCatone,Italy

§

SterrewachtLeiden,NielsBohrweg2,2333RALeiden,TheNetherlands

Abstract.WepresentthemainresultsderivedfromachemicalanalysiscarriedoutonalargesampleofgalacticO-richAGBstarsusinghighresolutionopticalspectroscopy(R～40,000–50,000)withtheintentionofstudyingtheirlithiumabundancesand/orpossibles-processelementenrichment.Ourchemicalanalysisshowsthatsomestarsarelithiumoverabundantwhileothersarenot.Theobservedlithiumoverabundancesareinterpretedasaclearsignatureoftheactivationoftheso-called“HotBottomBurning”(HBB)processinmassivegalacticO-richAGBstars,aspredictedbythemodels.However,thesestarsdonotshowthezirconiumenhancement(takenasarepresentativeforthes-processelementenrichment)associatedtothe3rddredge-upphasefollowingthermalpulses.OurresultssuggestthatthemoremassiveO-richAGBstarsinourGalaxybehavedifferentlyfromthoseintheMagellanicClouds,whicharebothLi-ands-process-rich(S-typestars).Reasonsforthisunexpectedresultarediscussed.WeconcludethatmetallicityisprobablythemainresponsibleforthedifferencesobservedandsuggestthatitmayplayamoreimportantrolethangenerallyassumedinthechemicalevolutionofAGBstars.

INTRODUCTION

TheAsymptoticGiantBranch(AGB)isformedbystarswithinitialmassesintherangebetween0.8and8M⊙inalatestageoftheirevolution.DuringmostofthetimeHburningisthemainsourceofenergyfortheAGBstarbut,occasionally,theinnerHeshellignitesina“ther-malpulse”and,eventually,thebyproductsofHeburn-ingmayreachtheouterlayersoftheatmosphere(theso-called3rddredge-up).Thus,AGBstars,originallyO-rich,canturnintoC-richAGBstars(C/O>1)afterafewthermalpulses.AnotherimportantcharacteristicofAGBstarsisthepresenceofneutron-richelements(s-elementslikeRb,Zr,Ba,Tc,Nd,etc.)intheiratmo-sphereswhicharetheconsequenceoftheslow-neutroncapturesproducedduringthethermalpulsingphase.Ac-cordingtothemostrecentmodelstwomajorneutronsourcescanoperateinAGBstarsdependingonthestellarmass.The13C(α,n)16Oreactionisthepreferredneutronsourceformassesaround1 3M⊙whileforintermedi-atemassstars(M>3M⊙)theneutronsarethoughttobemainlyreleasedby22Ne(α,n)25Mg(ttanzio&Lugaro2005forarecentreview).

InthecaseofthemoremassiveO-richAGBstars(M>4M⊙),theconvectiveenvelopecanpenetratetheH-

burningshellactivatingtheso-called“Hotbottomburn-ing”(HBB)process.HBBtakesplacewhenthetempera-tureatthebaseoftheconvectiveenvelopeishotenough(T≥2×107K)and12Ccanbeconvertedinto13Cand14NthroughtheCNcycle(Sackmann&Boothroyd1992).HBBmodels(e.g.Mazzitelli,D’Antona&Ven-tura1999,hereafterMDV99)predictalsotheproduc-tionof7Libythechain3He(α,γ)7Be(e ,v)7Li,throughtheso-called“7Betransportmechanism”(Cameron&Fowler1971).OneofthepredictionsofthesemodelsisthatLishouldbedetectable,atleastforsometime,onthestellarsurface.

TheHBBactivationinmassiveO-richAGBstarsissupportedbystudiesofAGBstarsintheMagellanicClouds(hereafter,MCs)(e.g.Plez,Smith&Lambert1993).ThedetectionofstrongLioverabundancesto-getherwithstrongs-elementenhancementinthesemas-sive(andluminous)AGBstarsisthesignaturethattheyareindeedHBBstarswhichhaveundergoneaseriesofthermalpulsesanddredge-upepisodesintheirrecentpast.

InourownGalaxy,onlyahandfulofLi-richstarshavebeenfoundsofar(e.g.Abiaetal.1993).MostofthemarelowmassC-richAGBstars(e.g.Abia&Isern2000)andintermediatemassS-andSC-stars(e.g.Abia

We present the main results derived from a chemical analysis carried out on a large sample of galactic O-rich AGB stars using high resolution optical spectroscopy (R~40,000-50,000) with the intention of studying their lithium abundances and/or possible s-p

&Wallerstein1998)andnotO-richM-typestars.How-ever,HBBisexpectedtobeactiveinthemostmassive(andluminous)AGBstars(from～4to7M⊙accordingtoMDV99HBBmodels),whichmightnotbeC-rich,butO-rich.Thebestcandidatesaretheso-calledOH/IRstars,luminousO-richAGBstarsextremelybrightintheinfrared,showingacharacteristicdouble-peakedOHmaseremissionat1612MHz.Thesestarsarealsoknowntobeverylongperiodvariables(LPVs),sometimeswithperiodsofmorethan500daysandlargeamplitudesofupto2bolometricmagnitudes.However,theyexperi-encevery1strongmasslossrates(uptoseveraltimes10 5M⊙yr )andmostofthemareusuallyheavilyobscuredatthisstagebythickcircumstellarenvelopes,makingop-ticalobservationsverydif cult.Thus,noinformationex-istsyetontheirLiabundancesand/orpossibles-processelementenrichment.

OBSERVATIONSANDRESULTS

Alargesample(102)oflong-period(300 1000days),largeamplitudevariability(upto8 10magnitudesintheVband),late-type(>M5)O-richAGBstarsdisplay-ingOHmaseremissionwithawiderangeofexpansionvelocities 1(fromjustafewkms 1tomorethan20kms)wascarefullyselected.Starswereincludedinthesampleifsatisfyingatleastoneoftheabovecriteriaandideallyasmanyofthemaspossible,whichguaranteesthattheyareactuallymassivestars.Consistently,starsinthesampleweremainlymembersofthegalacticdiskpopulationanddisplayedstrongIRexcessesdetectedbyIRAS.

High-resolutionopticalechellespectra(R～40,000–50,000)wereobtainedforallstarsinthesampleduringseveralobservingperiodsin1996–1997.ThefulllogofthespectroscopicobservationsisshowninTable1,in-cludingmoredetailedinformationontheobservations.Thetwo-dimensionalframescontainingtheechellespec-trawerereducedtosingle-orderone-dimensionalspectrausingthestandardECHELLE1softwarepackageasimple-mentedinIRAF.Becauseoftheveryredcoloursofthesourcesobserved,theS/Nratiosachievedinthereducedspectracanstronglyvaryfromthebluetotheredorders(10-20at～6000Åwhile>100at～8000Å).

WedetectedthepresenceoftheLiIresonancelineat6708Åin25%ofthesourcesinthesamplewithawidevarietyofstrengths,whilewedidnot ndanysignatureofthislinein31%ofthestars.Theremaining44%were

Set-upDate

λRange(Å/pix)(Å)

heavilyobscuredbytheirthickcircumstellarenvelopesandtheyweretoored/notfoundatopticalwavelengths.Ingeneral,allstars(withorwithoutlithium)showex-tremelyredspectrawiththe uxlevelfallingdramati-callyatwavelengthsshorterthan6000Å.Inaddition,thespectraareseverelydominatedbystrongmolecularbandsmainlyduetotitaniumoxide(TiO),asaconse-quenceoftheverylowtemperatureandtheO-richna-tureofthesestars.Interestingly,thebandheadsofZrOseemtobeabsentinallspectra.ThisisshowninFigure1whereweshowthespectralregionaroundtheZrOband-headsat6474and6495Å.TheseZrObandheads(aswellasthosecorrespondingtoothers-elementoxidessuchasLaOorYO)areverystrongingalacticS-starsandinmassiveMCAGBstars.

CHEMICALANALYSIS

Ouranalysiscombinesstate-of-the-artlineblanketedmodelatmospheresandsyntheticspectroscopywithex-tensivelinelists.Wehaveusedthesphericallysymmet-ric,LTE,hydrostatic‘MARCS’modelatmospheresforcoolstarsandthe‘TURBOSPECTRUM’spectralsyn-thesiscode(Alvarez&Plez1998)toderivetheLiandZr(takenasrepresentativeofallothers-processelements)abundancesinthosestarsforwhichanopticalspectrumcouldbeobtained.

Fromanexhaustivestudyofthein uenceofthevari-ationsofthefundamentalstellarparameters(e.g.Teff,logg,M,z,ξ,C/O,etc.)onthesyntheticspectraandfromourknowledgeofthemaincharacteristicsofourstarsweobtainedthemostadequateinitialsetofparam-etersaswellastheirplausiblerangeofvariation,andweconstructedagridofMARCSmodelspectra.

Thus,we rstdeterminedbyχ2minimisationwhichofthespectrafromourgridofmodelsprovidedthebest ttotheobservationsinthe6670 6730Åandthe6455 6499Åspectralregions.Thegoalwasto ttheoverallshapeofthespectraincludingtheTiObandheads,whichareverysensitivetovariationsintheeffectivetem-perature.Then,theLiandZrabundanceswerederivedby ttingtheLiIresonancelineat～6708ÅandtheZrOmolecularbandsat6474Åand6495Å,respectively.Asanexample,thebest tinthe6670 6730Åspectralre-

We present the main results derived from a chemical analysis carried out on a large sample of galactic O-rich AGB stars using high resolution optical spectroscopy (R~40,000-50,000) with the intention of studying their lithium abundances and/or possible s-p

6

WY Cas

R And

4

WXSer

V697 Her

2S Crb

IRAS 16037

6460647064806490

FIGURE1.HighresolutionopticalspectraofsamplestarsdisplayingthelackoftheZrOabsorptionbandsat6474and6495ÅcomparedwithtwogalacticS-stars(WYCasandRAnd).WXSer(IRAS15255+1944)andV697Her(IRAS16260+3454)areLi-detectedwhileSCrB(IRAS15193+3132)andIRAS16037+4218areLinon-detected.Theabsorptionbandat～6480ÅcorrespondstotheTiOmolecule.

gionaroundtheLiIlineispresentedinFigure2forthestarIRAS11081 4203.

LIANDZRABUNDANCES

OurchemicalabundanceanalysisshowsthathalfofthestarsshowLioverabundancesintherangelogε(Li)～0.5and3.02.AverysimilarrangeofLioverabundancesisfoundinthemassiveO-richAGBstarsstudiedintheMCs(e.g.Plez,Smith&Lambert1993).TheLioverabundancesobservedareinterpretedasasignatureoftheactivationoftheso-called“HotBottomBurning”(HBB),con rmingthattheyaremassiveAGBstars(M>4M⊙accordingtoMDV99HBBmodels).

However,thenon-detectionoftheZrOmolecularbandsat6474Åand6495Åinanyofthestarsanalysedimposedsevereupperlimitstothezirconiumabundance([Zr/Fe]<0.0 0.25forTeff≥3000Kand[Zr/Fe]<0.25 0.50forTeff<3000K).IftheZren-hancementistakenasarepresentativeforthes-process

We present the main results derived from a chemical analysis carried out on a large sample of galactic O-rich AGB stars using high resolution optical spectroscopy (R~40,000-50,000) with the intention of studying their lithium abundances and/or possible s-p

FIGURE2.Bestmodel tandobservedspectrumintheregion6670 6730ÅforthestarIRAS11081 4203.TheTeffandLiabundancederivedfromthisspectrumwas3000Kandlogε(Li)=1.3,respectively.Theparametersofthebestmodelatmosphere tareindicatedinthetoplabel.

(2000)inthefewobscuredMCAGBstarsforwhichthisanalysishasbeenmade.Ifmasslossisdrivenbyradia-tionpressureonthedustgrains,thismightbelessef -cientwithdecreasingmetallicity(Willson2000).Inthatcase,longerAGBlifetimeswouldbeexpected,whichcouldincreasethechanceofnuclear-processedmaterialtoreachthestellarsurface.TheslowevolutionpredictedforAGBstarsintheMCsasaconsequenceofthelessef cientmasslossleavestimeformorethermalpulsestooccurduringtheAGBlifetimeand,therefore,amoreeffectivedredge-upofs-processelementstothesurfacecanbeexpectedbeforetheenvelopeiscompletelygoneattheendoftheAGB.ThiswouldexplainwhyeventhemoremassivestarsintheMCsshowastrongs-processenrichmentincontrasttotheirgalacticcounterparts.InourGalaxytheonlyAGBstarsshowingasimilarover-abundanceins-processelementsseemtobetheresultoftheevolutionoflow-tointermediate-massstars(M<1.5 2.0M⊙),whilenoorverylitles-processenhance-mentisobservedingalacticAGBstarswithhighermainsequencemasses.

Finally,thelowercriticalmassneededtodevelopHBB(e.g.M>3M⊙atthemetallicityoftheLMC,com-paredtothe～4M⊙limitinourGalaxy)wouldfavourthesimultaneousdetectionofs-processelementsandLienrichmentinalargernumberofAGBstarsintheMCs,asitisactuallyobserved.IncontrasttotheirMC

counterparts,Li-richmassiveAGBstarsinourGalaxywouldevolvesorapidly(becauseofthestrongmassloss)thatthereisnotimeforasigni cantenhancementins-processelements.

Insummary,ourresultssuggestthatthedramaticallydifferentabundancepatternfoundinAGBstarsbelong-ingtotheMCsandtoourGalaxycanbeexplainedintermsofthedifferentmetallicityconditionsunderwhichthesestarsevolved.Thisisthe rstobservatinalevidencethatthechemicalevolutionduringtheAGBcouldbestronglymodulatedbymetallicity.Acompletedescrip-tionanddiscussionoftheseresultsaswellastheirevolu-tionaryconsequenceswillbegiveninGarcía-Hernándezetal.(2005,inpreparation).

REFERENCES

C.Abiaetal.,A&A,272,455(1993)

C.Abia,andG.Wallerstein,MNRAS,293,89(1998)C.Abia,andJ.Isern,ApJ,536,438(2000)

R.Alvarez,andB.Plez,A&A,330,1109(1998)

A.G.W.Cameron,andW.A.Fowler,ApJ,164,111(1971)M.A.T.Groenewegenetal.,MmSAI,71,639(2000)F.Herwig,ApJ,605,425(2004)

ttanzio,andM.Lugaro,in“NucleiintheCosmosVIII”,NuclearPhysicsA,2005,inpress(astro-ph/0505424)9.M.Lugaroetal.,ApJ,586,1305(2003)1.2.3.4.5.6.7.8.

We present the main results derived from a chemical analysis carried out on a large sample of galactic O-rich AGB stars using high resolution optical spectroscopy (R~40,000-50,000) with the intention of studying their lithium abundances and/or possible s-p

10.I.Mazzitelli,F.D’Antona,andP.Ventura,P.,A&A,348,846(1999)(MDV99)

11.B.Plez,V.V.Smith,mbert,ApJ,418,812(1993)

12.I.-J.Sackmann,andA.I.Boothroyd,ApJ,392,L71(1992)

13.J.Th.vanLoon,A&A,354,125(2000)14.L.A.Willson,ARA&A,38,573(2000)

本文来源：https://www.bwwdw.com/article/12n4.html