More nitrogen rich B-type stars in the SMC cluster, NGC 330

MorenitrogenrichB-typestarsintheSMCcluster,NGC330

D.J.Lennon1,P.L.Dufton2,andC.Crowley1

1

arXiv:astro-ph/0207031v1 1 Jul 20022

IsaacNewtonGroupofTelescopes,ApartadodeCorreos321,E-38700,SantaCruzdeLaPalma,CanaryIslands,Spainemail:djl@ing.iac.es

TheDepartmentofPureandAppliedPhysics,TheQueen’sUniversityofBelfast,BelfastBT71NN,N.Irelandemail:P.Dufton@Queens-Belfast.ac.uk

Receiveddate;accepteddate

Abstract.HighresolutionspectraofsevenearlyB-typegiant/supergiantstarsintheSMCclusterNGC330areanalysedtoobtaintheirchemicalcompositionsrelativetoSMCfieldandGalacticB-typestars.Itisfoundthatallsevenstarsarenitrogenrichwithanabundanceapproximately1.3dexhigherthananSMCmain-sequencefieldB-typestar,AV304.Theyalsodisplayevidencefordeficienciesincarbon,butothermetalshaveabundancestypicaloftheSMC.GiventhenumberofB-typestarswithlowrotationalprojectedvelocitiesinNGC330(allourtargetshavevsini<50km/s),wesuggestthatitisunlikelythatthestarsinoursampleareseenalmostpole-on,butratherthattheyareintrinsicallyslowrotators.Furthermore,noneofourobjectsdisplaysanyevidenceofsignificantBalmeremissionexcludingthepossibilitythattheseareBestarsobservedpole-on.Comparingtheseresultswiththepredictionsofstellarevolutionmodelsincludingtheeffectsofrotationallyinducedmixing,weconcludethatwhiletheabundancepatternsmayindeedbereproducedbythesemodels,seriousdiscrepanciesexist.Mostimportantly,modelsincludingtheeffectsofinitiallylargerotationalvelocitiesdonotreproducetheobservedrangeofeffectivetemperaturesofoursample,northecurrentlyobservedrotationalvelocities.Binarymodelsmaybeabletoproducestarsintheobservedtemperaturerangebutagainmaybeincapableofproducingsuitableanalogueswithlowrotationalvelocities.Wealsodiscusstheclearneedforstellarevolutioncalculationsemployingthecorrectchemicalmixofcarbon,nitrogenandoxygenfortheSMC.

Keywords.stars:early-type–supergiants–giants–evolution

1.Introduction

NGC330isoneofthebrightestandmostpopulousyoungclus-tersintheSmallMagellanicCloud(SMC).Thephotomet-ricsurveysofArp(1959),Robertson(1974)andCarneyetal.(1985)illustratethekeyfeaturesofthecluster’scolour-magnitudediagram,namelythepresenceoftwogroupsofblueandredsupergiantswellseparatedfromthecluster’ssupposedmain-sequenceblueplume.Thesetwogroupsofstarshavebeenwidelyinterpretedascoreheliumburningstarsandtheclusteristhereforeconsideredasakeytestofstellarevolu-tiontheoryandphysicsforstarsofintermediatemassinalowmetallicityregime.Essentiallytheratioofblue(B)tored(R)supergiantsisanindicatoroftherelativetimesamassivestarspendsinthethesephases,andthesequantitiesareextremelysensitivetoassumptionsmadeconcerningconvectionandmix-ing.InfacttheB/RratioinNGC330isgenerallyassumedtoberepresentativeoftheSMCasawholeandisusedasacalibratorforstellarevolutioncalculationsatlowmetallicity(Stothers&Chin1992a,1992b;Kelleretal.2000;Chiosietal.1995).ThespecificproblemoftheB/RratioasafunctionofmetallicityhasbeendiscussedbyLanger&Maeder(1995),whereamore

2D.J.Lennonetal.:MorenitrogenrichB-typestarsintheSMCcluster,NGC330

inmetallicityismuchsmaller,orindeednotsignificant(Hill1999),confirmingtheresultsobtainedfromtheanalysisoftwoB-typestarsintheclusterbyLennonetal.(1996,hereafterPaperI).

ThespectroscopicworkofLennonetal.(1994)alsofoundthatthebrightnon-BeandweakBe-typestarsoccupiedthatregionoftheHR-diagramknownasthepostmainsequencegap,orblueHertzsprunggap(BHG).Thatis,theyaregi-ant/supergiantstarslyingred-wardsofthemainsequenceband,butblue-wardsoftheA/F-typesupergiantregime.Caloietal.(1993)andGrebeletal.(1996)havealsocommentedonthisfact,thelattersuggestingthatthesestarsaremostlikelyamixtureofrapidlyrotatingB/Be-typestarsofvaryingorien-tationandbluestragglersformedbyinteractioninbinarystars.Kelleretal.(2000)alsoattemptedtoaddressthisproblemus-ingfar-UVphotometry(fromtheF160BWfilteronWFPC2ofHST)toconstrainB-typestellareffectivetemperaturesandfindsignificantlyfewerstarsintheblueHertzsprunggap(BHG).Howevertheyassumedthatthelogarithmicsurfacegravitieswere4.0intheirwork(Keller,privatecommunication),whichmayresultinspuriouslyhigheffectivetemperaturesforstarsneartheturn-offsincetheyhavemuchlowersurfacegravi-ties(Lennonetal.1994).NotethatCaloietal.(1993)alsoadoptedlowervaluesforthesurfacegravities.Clearlyade-tailedspectroscopicanalysisoftheBHGstarsleadingtoesti-matesofbothstellarparametersandatmosphericabundancesisneededforcomparisonwiththepredictionsofvariousstellarevolutioncalculations.

InPaperI,wederivedmetallicitiesoftwosuchB-typestarsinNGC330andwhilewefoundthemingeneraltobecom-patiblewiththatoftheSMCfieldbothstarshadasignificantnitrogenoverabundance.Themagnitudeofthenitrogenenrich-mentwasuncertainduetothesmallnumberofNGC330targetsanalysedandalsothedifficultyinestimatingthelownitrogenabundanceoftheSMCfield,atleast,fromB-typestars.AlsoinPaperIwefoundthatthecarbonabundancewasnotsig-nificantlydepleted,contrarytowhatoneexpectsifthenitro-genwereproducedintheCNcycle.Thecarbonabundancewasuncertainhowever,andcoupledwiththeuncertaintyofthemagnitudeofthenitrogenenrichment,madeinterpretationdifficult.Anadditionalpuzzlingaspectwasthatbothstarsarenarrow-linedandthereforeifthenitrogenenhancementsaretheresultofhighrotationwemustbeobservingthemalmostpole-on,whichseemedunlikelygivethattheyweredrawnfromthesampleofabout20starsobservedbyLennonetal.(1994).InthepresentpaperweanalyseseventargetsinNGC330(in-cludingthetwodiscussedinPaperI)belongingtoboththebluesupergiantgroupandthetipofthebluemainsequenceplumediscussedabove.Henceallstarsliein,orcloseto,theBHG.Weestimatestellarparameters,radialandprojectedrota-tionalvelocities,aswellasbothabsoluteabundancesanddif-ferentialabundancesrelativetoagalactictargetinthehandχPerseicluster(Vranckenetal.2000)andtoanSMCfieldstar,AV304.Forthelatterweutilisetheresultsfromarecentanalysis(Rollestonetal2002)basedonhighqualityVLTdata,whichnowgiveusareliableestimateofthepristinenitrogenabundanceinunevolvedB-typestarsintheSMC.Wealsoat-tempttoprovideimprovedestimatesforcarbonabundances

andcompareourresultswithstellarevolutioncalculations,in-cludingtherecentmodelsofMaeder&Meynet(2001)whichincludetheeffectsofrotationallyinducedmixing.

2.Observationaldataandresults2.1.OpticalSpectroscopy

Highresolutionspectraofearly-typestarsintheclusterNGC330,wereobtainedusingtheCASPECspectrographontheESO3.6monthe26and27August1994andonthe22and23September1996;theobservationalparametersaresum-marizedinTable1.TheUBVphotometryhasbeentakenfromMazzalietal.(1996)supplementedbyadditionalunpublisheddata,whilethespectraltypeshavebeenestimatedfromunpub-lishedlowdispersionEMMIspectroscopy.TheobservationalconfigurationanddatareductiontechniquesarediscussedindetailinPaperI.Briefly,Caspecwasusedwithanentranceslitwidthof2arc-secondsgivinganeffectiveresolvingpowerofapproximately20000,orabout15km/s.PreliminaryreductionoftheechellogramstoaonedimensionalformatwasachievedusingtheIRAFreductionpackage(WillmarthandBarnes,1994).Therelativefaintnessofthetargets,leadtosomeofthespectrahavingrelativelylowcontinuumcountsandsignal-to-noiseratios.Inparticular,theobservationsforA01,B04andB32weretakeninconditionsthatwereattimescloudy,andsonotallspectrawereco-addedtoproducethefinalspectrumforthesestars.Thes/nestimatesaresummarizedinTable1andweremeasurednearthecentreoftheechelleorderatapprox-imately4200A.

˚Fortunately,allthestarshavesmallprojectedrotationalvelocitiesandarehence(FWHM)rangingfrom0.4to0.9A

˚sharplinedwithlinewidths

(seeTable1).Hencedespitethesignificantshotnoise,lineswithequivalentwidthsthan50mA

˚ofmore

couldnormallybelinestrengthsassmallas30mA

˚detected,whilefeatureswithcouldsometimesbeconvinc-inglyidentified.Projectedrotationalvelocitieswereestimatedfromthelinewidthsbyassumingthattheonlycontributionstotheintrinsicwidthsoftheabsorptionlinesareinstrumentalandpurerotationali.e.thattheeffectsofmacroturbulencearenegligible.HencetheresultsinTable1aremostlikelyupperlimitsontheprojectedrotationalvelocities.

Equivalentwidthsweremeasuredforthemetalandnon-diffuseheliumlinesusingtheSTARLINKspectrumanalysisprogramDIPSO.Absorptionlineswerefittedusinganon-linearleastsquarestechniquewithavarietyofprofileshapes(e.g.triangles,gaussians)anddegreesofpolynomialforthecontin-uumbeingconsidered.Equivalentwidthsforthediffuseheliumlinesweremeasuredmanuallytrarilydefinedat±10A

˚withthecontinuumbeingarbi-fromthelinecentre.ForthehydrogenHγandHδlines,thespectrawereagainnormalisedandpro-filesmeasured,at±16A

˚withthecontinuumlevelsnowbeingdefined

(aswerethetheoreticalprofiles).Forallthesemea-surementstheprocedureswereeffectivelyidenticaltothosedescribedinPaperIwherefurtherdetailscanbefound.

Themodelatmosphereanalysisutilisedbothabsoluteanddifferentialtechniques.Forthelatter,twostandardstarswereconsidered-BD+56576inthehandχPerseigalacticclus-terandtheSMCtargetAV304.TheobservationaldataforBD

D.J.Lennonetal.:MorenitrogenrichB-typestarsintheSMCcluster,NGC3303

+56576wastakenfromourstudy(Vranckenetal2000)ofearly-Btypestarsinthisdoublecluster,whiletheresultsforAV304arebasedonrecentlyobtainedVLTspectra(Rollestonetal.2002).

2.2.Fluxdistributions

IUElowresolutionspectraareavailableinboththeshortandlongwavelengthcamerasforalltheNGC330targets(apartfromA04)andforBD+56576.ThesewereextractedfromtheINESarchive(Rodriguez-Pascueletal.1999),wherefurtherdetailsmaybefound.

3.MethodofAnalysis

WeinitiallyconsiderstandardLTEmodelatmospheretech-niquesinouranalysis,howeverinsection5wediscusstheac-curacyofthismethodandcorrectourabundancesforNLTEef-fects.HereweusethemodelatmospherestakenfromthegridscalculatedwiththecodeofKurucz(1992)andmadeavailableathiswebsite.Thisgridcoversarangeofchemicalcomposi-tions,withthecurrentanalysisutilisingmainlytheresultsformetallicitiesof−0.5dex,whichiscompatiblewiththatfoundinPaperI.Detailsofthemethodsandatomicdatausedintheradiativetransfercalculationscanbefoundin,forexample,Rollestonetal.(2000)orSmarttetal.(1996).

3.1.Atmosphericparameters

AsdiscussedinPaperI,therelativelylowsignal-to-noiseoftheCaspecspectraprecludetheobservationofdifferentionizationstagesofthesameelementformostofthestars.AdditionallythereisnoextantStr¨omgrenphotometry.Hencenoneofthestandardtechniques(see,forexample,Kilian,1994,GiesandLambert,1992,Rollestonetal.2000)areavailableforcon-strainingeffectivetemperatures.Hence,initialtemperaturees-timatesweremadefromtheobservedfluxdistributions.TheextinctiontowardNGC330hasbeenfoundtoberelativelysmallandherewehaveadoptedaforegroundreddeningofE(B−V)=0.034andtheextinctionrelationofSeaton(1979),togetherwithanSMCextinctionofE(B−V)=0.05andtherelationshipofThompsonetal.(1988).AsdiscussedinPaperI,similareffectivetemperatureswouldhavebeenesti-matedusing,forexample,thegalacticextinctionlawforallthereddening.

TheeffectivetemperatureestimatesarelistedinTable2.AsdiscussedinPaperI,ateffectivetemperaturesgreaterthanap-proximately22000K,thefluxdistributionbecomesrelativelyinsensitivetothisparameter.Additionallytheremaybeuncer-taintiesduetothepresenceoffaintertargetsintheIUEaper-ture.Henceweadoptconservativeerrorestimatesof2000Kforourcoolertargetsandupto4000Kforourhottesttargets.Forthelatter,wenote(seePaperIfordetails)thatourestimatesareconsistentwiththeabsenceofadetectableHeIIspectrum,whichimpliesthattheeffectivetemperaturesmustbelessthan26000K.

Twostars,A01andB04,hadbeenalreadybeendiscussedinPaperI;herewehaveindependentlyre-estimatedtheireffec-

tivetemperatures.Forthelatterthetwoestimatesareconsis-tent,whilstforA01,thecurrentestimateis2000KlessthaninPaperI.Here,weusedfluxdistributionscalculatedforametal-licity0.5dexlessthansolar(inpaperI,fluxesforanormalmetallicitywereadopted)buttestsshowedthattheestimateswerelittleaffectedbythechoiceofmetallicity;henceweas-cribethedifferencemainlytothefluxdistributionsnotbeingparticularlytemperaturesensitiveinthecaseofA01.

ForA04,noIUEobservationswereavailable;thereare,however,measurementsoftheequivalentwidthsofbothSiIIIandSiIVionsandthesehavebeenusedtoestimateaneffectivetemperature(see,forexample,Rollestonetal.2000fordetails)andthisisalsolistedinTablearesmall(typically30mA)

˚2.Notethattheequivalentwidths

andhencesubjecttoconsiderableuncertainty.Additionallyseveralauthors(forexample,Kilian1994,Smarttetal.2001)havefoundsystematicdifferencesbe-tweentemperaturesfoundfromopticalphotometryandioniza-tionequilibriaanditispossiblethatsimilardifferenceswillbepresentforthetwomethodsusedhere.Hence,thetemperatureestimateforA04shouldbetreatedwithsomecaution.

ComparisonofobservedandtheoreticalBalmerlines(seeRollestonetal.2000forfurtherdetails)werethenusedtoes-timatelogarithmicsurfacegravities,whichareagainsumma-rizedinTable2.Someofthesurfacegravityestimatesdifferedsignificantlyfromthevalue(3.5dex)assumedwheninitiallyestimatingtheeffectivetemperatures.Inthesecasestheeffec-tivetemperatureswerere-estimatedusinganappropriategrav-ityandtheprocedureiterateduntilitconverged.AnexampleofthequalityofthefitcanbefoundinFig.2ofPaperIandrelevantuncertaintiesfortheindividualstarsarealsolistedinTable2.

Forthreestars(A02,B22,B37)thereweresufficientOIIlinestoestimatemicroturbulentvelocitiesandtheseareagainsummarizedinTable2.Fortheotherstarsavalueof5kms−1wasadopted;wenotethatalthoughthismaybetoolow,theweaknessofthemetallinespectraintheSMCtargetsimpliesthatthechoiceofthisparameterdoesnotsignificantlyaffecttheabundanceanalysis.

Forthegalacticstandard,BD+56576,theatmosphericpa-rametershavebeenre-estimatedusingthesamecriteriaandmethodsasfortheNGC330targetsandaresummarizedTable3,togetherwiththevaluesdeducedbyVranckenetal.(2000,VDLL).Giventhedifferentcriteriaandnon-LTEmethodsusedbyVDLL,theagreementissurprisinglygood.However,weemphasizethatthevaluesdeducedhereshouldnotbeconsid-eredasthebestavailablebutratherappropriateforadifferentialanalysiswithrespecttotheNGC330targets.ThecomparisonwithVDLLalsoprovidesuswithanestimateoftheimportanceofnon-LTEeffectswhichwillbediscussedinthenextsection.

InthecaseofAV304,wehaveadoptedtheatmosphericpa-rameters(listedinTable3)deducedbyRollestonetal.(2002).Theyusedsimilarmodelatmospheretechniquesandcriteriatothoseadoptedhere,apartfromtheeffectivetemperaturewhichwasestimatedfromthesiliconionizationequilibria.However,inapreviouslyanalysisofthisstar(Rollestonetal.1993)goodagreementwasfoundbetweentheeffectivetemperaturesesti-matesfoundfromtheionizationequilibriumandtheIUEfluxdistribution.

4D.J.Lennonetal.:MorenitrogenrichB-typestarsintheSMCcluster,NGC330

Table1.ObservationaldataforNGC330targets:StaridentificationsarefromRobertson(1974).SpectraltypesarefromLennonetal.(1994)orGrebeletal.(1996)althoughsomeclassificationsareslightlyrevised.PhotometryistakenmostlyfromMazzalietal.(1996)withadditionaldatatakenfromGrebeletal.(1996)andRobertson(1974).Asdiscussedinthetext,theprojectedrotationalvelocities(vsini)aremostprobablyupperlimits,andareaccurateto±5km/s.Radialvelocities(vrad)arecorrectedtoheliocentricvalues.

Star

Sp.Type

V

B-V

U-B

Dateyymmdd

Exposuresnxsecs

s/n

FWHM˚A

vsini

km/s

vradkm/s

Table2.Atmosphericparameters(effectivetemperatures(K),logarithmicsurfacegravities(cgs),microturbulentvelocities(km/s))andLTEabundancesforNGC330targets(thenumbersinparenthesesgivethenumberoflinesusedtoderivethemeanabundance).Forstarswithoutanerrorestimateforthemicroturbulentvelocitywehaveassumedavalueof5km/s.

Star

A01

A02

A04

B04

B22

B32

B37

Mean

4.Results

Usingtheatmosphericparametersdiscussedabove,themetallinesintheNGC330targetswereusedtodeduceabsoluteabun-dancesusingatomicdatatakenfromJeffery(1991).Theseab-solutevaluesarelistedinTable2,togetherwiththenumberoflinesusedintheanalysis.Alsolistedforeachionizationstagearethemeanabundanceestimateandthestandarddevia-tion.InTable3,absoluteabundanceestimatesaregivenforBD+56576(bothdeducedhereandbyVranckenetal.2000)andforAV304,whichareagaintakendirectlyfromRollestonetal.(2002).

ItismoreusefultoconsideradifferentialanalysisoftheNGC330targetsrelativetothegalacticstandardBD56576.Thiswillbelessaffectedbyanyerrorsinoscillatorstrengthsorsystematicerrorsintheadoptedatmosphericparameters.Unfortunately,itwasnotpossibletoconsiderallthelinesob-servedintheCaspecspectraduetothemorelimitedwave-lengthcoverageforBD+56576.ThedifferentialabundancesaresummarizedinTable4foreachstar,togetherwiththemeans(andstandarddeviations)foreachelement.TheresultsofasimilardifferentialcomparisonforAV304(relativetothegalacticstarHR2387)takenfromRollestonetal.(2002)arealsosummarizedintable4.Thesedifferentialanalysesconfirm

D.J.Lennonetal.:MorenitrogenrichB-typestarsintheSMCcluster,NGC3305

Table4.DifferentialabundanceanalysesoftheNGC330targetsandofAV304relativetogalacticstandardsBD+56576andHR2387respectively.Thenumberoflines(n)usedinthederivationofthesevaluesisalsogiven.

Star

HeI

CIINII

∆[He

H

]n

∆[

N

Table3.Modelatmosphereanalysisofgalacticstandard,BD

+56576,togetherwiththepreviousresultsofVranckenetal.(2000-VDLL).TheresultsforAV304aretakendirectlyfromRollestonetal.(2002)Numbersinparenthesesrefertothenumberoflinesusedtoderivetheabundance.

ThispaperVDLLAV304Method

LTE

non-LTE

LTE

themainresultsfromtheabsoluteanalyses–alltheSMCtar-getsaredeficientinmetalsapartfromnitrogen,whichisverydeficientinAV304butslightlyenhancedrelativetoourGalaxyintheNGC330targets.Belowtheresultsforindividualele-mentsarediscussedinmoredetail.

4.1.Helium

Aneffectivelynormalheliumabundanceisfound,althoughthereisconsiderablescatterwithintheNGC330stars.Thisprobablyreflectsthedifficultyinestimatingthecontinuumplacementfortherelativelybroaddiffuseheliumlines,whichmakeupthemajorityofourdataset.Henceweconcludethatwithintheuncertainties,thereisnoevidenceforananomalousheliumabundanceinourtargets.

4.2.Carbon

Thedifferentialabundancesarebasedmainlyonthebletat˚CIIdou-supplementedinsomecasesbythelineat3921A.

˚4267A,

Theformerisknowtobeparticularlysensitivetonon-OII

MgII

SiIIIH

]n

∆[Mg

H

]n

LTEeffects(EberandButler,1988).AlsoasdiscussedinPaperIthestellarabsorptionlinecouldbeaffectedbyrecombina-tionfromanassociatedHIIregion.HoweveralltheNGC330targetsshowasignificantcarbonunderabundance(witharela-tivelysmallscatterbetweenstars),whilethemeandifferentialabundanceissimilartothatfoundforAV304.

ThelatterresultappearsinconsistentwiththeresultsinTables2and3,wherethecorrespondingabundancesdifferby0.35dex.HoweverthevalueforAV304wasbasedonamixtureofCIIandCIIIlinesandifonlytheformerareconsideredthedifferenceinabsoluteabundancesdecreasestoapproximately0.15dex.Thesmallremainingdiscrepancyofapproximately0.1dexwiththedifferentialabundancesarisesfromthediffer-entsetoflinesusedinthetwoanalyses.

4.3.Nitrogen

AsdiscussedinPaperI,themoststrikingdifferencebetweenthetwoNGC330targetsandAV304isintheirNIIspectra.HoweverthenewVLTobservationsforAV304,nowallowsthisdifferencetobequantified.Indeedthemeandifferentialabun-danceimpliesthatfortheNGC330targetsnitrogenisover-abundantbyafactorofapproximatelytwentycomparedwithAV304.ThisisconsistentwithpaperI,wherealowerlimitonthenitrogenenrichmentofafactorofsixwasdeducedrel-ativetotheSMC.Thereisconsiderablescatterinthediffer-entialabundancesdeducedforindividualstars.Thisissome-whatsurprisingastheNIIlinesarerelativelystrongandhencetheequivalentwidthestimatesshouldbereliable.Additionally,onlysmallnon-LTEeffectshavebeenfoundfortheNIIspec-tra(BeckerandButler,1989).Hencethescattercouldpossiblyreflectarealvariationinthenitrogenabundancesinthesestars.

4.4.Oxygen

TheOIIspectrum,togetherwiththatforNIIisthebestob-servedinourtargets.ThedifferentialoxygenabundancesareingoodagreementwithonlythatforB22beingfarfromthemean.Thisstarhasarelativelyloweffectivetemperaturees-timateof20000Kandachangeof1000Kinthisestimatewouldchangethedifferentialabundancebymorethan0.3

6D.J.Lennonetal.:MorenitrogenrichB-typestarsintheSMCcluster,NGC330

dex.Hencewebelievethatthereisnoevidenceforanyvari-ationintheoxygenabundancesamongtheNGC330targets.Additionallythemeandifferentialabundanceissimilarto,al-thoughsomewhatlowerthan,thatfoundforAV304.

4.5.Magnesium

Although4481A,

˚theresultsarebasedsolelyontheMgIIdoubletat

thereisreasonableagreementintheindividualdiffer-entialabundances,whilstthemeanvalueagreewellwiththatforAV304.

4.6.Silicon

Incontrastwithoxygenandmagnesium,thedifferentialsiliconabundancesshowawidespread,whichischaracterisedbytherelativelylargestandarddeviationinthemeanvalue.Therea-sonforthisnear4560A

˚isunclearasformostofthetargetstheSiIIIlines

arerelativelywellobserved,althoughtherelativestrengthsoftheselinesaresometimesanomalous.Additionallyforthecoolertargets,theestimatedabundancesareverytem-peraturesensitive;forexampleataneffectivetemperatureof18000K,ashiftintemperatureof1000Kchangestheabun-dancebymorethan0.3dex.HowevertheOIIionhasasimilartemperaturesensitivityinthisregime,whilsttheoxygenabun-danceestimatesarefarmoreconsistent.Althoughthecauseofthediscrepanciesisunclear,theymaywellbedue(atleastinpart)tothesmallnumberofmeasurablesiliconlines.Indeedtheyareunlikelytoreflectrealabundancedifferencesgiventhebehaviourofotherα-processelements.ThemeandifferentialabundanceiswithintheuncertaintiesingoodagreementwiththatfoundforAV304.

5.Discussion

5.1.ThechemicalcompositionofourB-type

sampleinNGC330

TheprincipleconclusionfromthedifferentialanalysisisthattheclustertargetshaveamuchhighernitrogenandpossiblyloweroxygenabundanceN

thanAV304.Relativenitrogentooxy-genabundances,[ratio

isenhancedby1.2dexwithrespecttoAV304andthisO]

esti-mateisunlikelytobesignificantlyaffectedbyuncertaintiesin

atmosphericparametersortherelativelysimpleLTEanalysisadopted.

ThesimplestexplanationforthisnitrogenenhancementisthatitrepresentstheproductsofhydrogenburningbytheCNObi-cycle.Insuchcircumstances,itmightbeexpectedthattherewereacorrespondingenhancementinheliumandunderabun-dancesincarbonandpossiblyoxygen,withthesumofCNOnucleiintheNGC330starscomparabletothatinAV304.ItisthereforeimportanttotrytomaptheabundancesfortheNGC330targetsasderivedfromourdifferentialanalysisontoanabsolutescale.Thereareanumberofoptionsavailabletous,whichwenowdiscuss.

IfweassumethatthecompositionofAV304isthebase-lineinitialcompositionofNGC330,thenwecanusethediffer-enceintheLTEabundancesoftheNGC330targetsrelativetoAV304.Howeverwemustalsoaddresstheprobableimpactofnon-LTE(NLTE)effectsontheseabundances.WehaveusedNLTEcalculationssimilartothosediscussedinMcErleanetal.(2000)toestimatethedifferenceinNLTEandLTECNOabundanceestimatesforAV304.Thecorrectionsareapproxi-mately+0.07,-0.11and-0.07dexrespectivelyforC,NandO.Inadditionwenotethatheavilyonthe4267A

˚ourcarbonabundanceinAV304relies

linewhichiswellknowntogivespuri-ouslylowabundances.FollowingthediscussionofVranckenetal.(2000)andcomparingtheirresultswiththoseofGies&Lambert(1992)wefurthercorrectthecarbonabundanceby+0.34dex.OurfinalCNONLTEabundancesestimatesforAV304arethen7.41,6.55and8.16dexingoodagreementwiththeHIIregionresultsof7.4,6.6and8.1dexassummarizedinthediscussionofbaselineSMCabundancesforA-typesuper-giantsintheSMCbyVenn(1999).

WecannowusethesemodifiedNLTEabundanceesti-matesforAV304andthedifferenceinLTEabundanceslistedinTables2and3toestimateabsoluteCNOabundanceforourNGC330targets;thesearesummarizedinTable6.InturnwecanthenestimatesumoftheCNOandCNabundancesforAV304,whichareare8.24and7.46dexrespectively.ThesemaybecomparedwiththemeanNGC330totalsof8.17and7.71dexrespectively.Thereissomeslightevidenceforanin-creaseinCNbutthedifferenceof+0.25shouldbecomparedwithuncertaintiesinthemeancarbonandnitrogenabundancesof0.15and0.18dex(thetotalbeingdominatedbythemoreabundantspecies).ThesumofCNOisingoodagreementbutagainthetotalisdominatedbytheoxygenabundanceforwhichtheuncertaintyisapproximately0.13dex.Clearlyitisun-productivetocomparesummationsofabundanceswhenonespeciesissubstantiallymoreabundantthanallotherspecies,andtheuncertaintyinthatabundanceissimilarto,orlargerthan,thatofthelessabundantspecies.

WearriveatasimilarpictureifweinsteadconsiderthedifferentialabundancesoftheNGC330starsrelativetoBD+56576usingtheNLTEabundancespublishedbyVranckenetal.(2000)toputthemonanabsolutescale.Thisre-sultsinCNOabundancesfortheNGC330targetsof7.28,7.52and7.98dexrespectivelywhichareingoodagreementwiththevaluesobtainedusingAV304asthestandard(seeTable6).Thisispossiblythe4267A

˚fortuitous,andgiventhedifficultyinmodeling

line,unexpectedforcarbon.Neverthelessitrein-D.J.Lennonetal.:MorenitrogenrichB-typestarsintheSMCcluster,NGC3307

forcesthepreviousdiscussionoftheabsoluteabundances.WeconcludethatthechemicalpeculiaritiesoftheNGC330targetsmaybeunderstoodintermsofnuclearprocessingbytheCNObi-cycle,perhapswithsomeweakevidencethatONprocessinghasoccurred,butitisnotnecessarytoinvokeprimarynitrogenproduction(althoughthiscannotbeprecluded).

Wecanalsosearchforcorrelationsbetweenelementabun-danceswithintheNGC330targets.Linearleastsquaresfitsshowapositivecorrelationbetweentheheliumandnitrogenabundancesandnegativecorrelationsbetweenthecarbonandnitrogenandbetweentheoxygenandnitrogenabundances.Interestinglyallthesetrendsareconsistentwiththetrans-formationofhydrogenintoheliumusingtheCNObicycle.Unfortunatelyhowever,noneofthecorrelationsareconvinc-ingandthecoefficientsarenotsignificantlydifferentfromzeroateventhe1σlevel.

5.2.OtherstellarabundancesinNGC330andthe

SMC

Kornetal.(2000)haverecentlypublishedC,O,MgandSiNLTEabundancesforanotherB-typegiantinNGC330,thestarB30,andtheirabundancesareingoodagreementwithoursgiventhemagnitudeoftheNLTEcorrectionsandtheuncer-taintiesinbothstudies.Unfortunatelytheydonotgiveanitro-genabundancebuttwoothersimilarSMCgiantsintheirsam-plehaveNLTEnitrogenabundancesof7.3and7.2dex.Wewillreturntothesestarsinthediscussionoftheirevolutionarystatusbelow.FurthermoreanLTEnitrogenabundanceforstarB30waspublishedbyReitermannetal.(1990)whoobtainedavalueof7.4dexforamicrotubulenceof5km/s.

Therehavealsobeenmanystudiesofcoolgiantsandsuper-giantsinNGC330.TheseresultsaresummarizedinTable6.OfinteresthereareCNOabundancesandtherearetworecentes-timatesforNGC330starsbyHill(1999,H99)andGonzalez&Wallerstein(1999,GW).OneobviousdifferencebetweentheseisthatthemeannitrogenabundanceofGWissystem-aticallylowerthanthatofH99.Howeverwenotethattheni-trogenabundanceisderivedfrommolecularCNfeatures,anddependsontheadoptedcarbonabundance(whichisderivedfromC2).WhileH99independentlyderivetheircarbonabun-dances,GWadoptedmeanvaluesfromtheliteratureandthereisasmallsystemicoverestimationrelativetoH99(approxi-mately0.2dex).SuchasmallchangeinthecarbonabundanceisthemostlikelyreasonfortheirlownitrogenabundancesandgiventhatthecarbonabundanceofH99maybemorereliableandthattheirresultsagreebetterwithothersamplesofevolvedB,AandF-typestarsintheSMC,weprefertheirresultsforcarbonandnitrogen.

ComparingwithotherSMCsampleswenotethatourmeanCNOabundancesareingoodagreementwiththeresultsofVenn(1999)andHilletal.(1997)althoughtheNGC330starsmaybemildlymetalpoor.Therehavebeenprevioussugges-tionsthatNGC330mayberelativelymetalpoorwithrespecttotheSMC(Grebel&Richtler1992)butourresultsconfirmrecentworkinthatanymetaldeficiencymustbesmall(<0.2dex).WethereforeconcludethatthepatternofthemeanCNO

abundancesfoundintheNGC330B-typegiantsisverysimi-lartothatfoundforsamplesofotherevolvedA,FandK-typegiantsandsupergiants.WealsonotethatDuftonetal.(2000)investigatedalargesampleofB-typesupergiantsintheSMCandthenitrogenabundancesfoundintheirlessluminousstars,althoughuncertain,arecomparabletothosefoundhere.

5.3.Evolutionarystatus

Itisusefultoprefaceourdiscussionoftheevolutionarysta-tusofthesestarsbyconsideringtheirpositionsintheHR-diagram.WefollowtheproceduredescribedinVDLL,adopt-ingadistancemodulusfortheSMCof18.9andextinctionestimatesasdiscussedinsection3.1.AsinVDLLweesti-matespectroscopicmasses(Mspec)fromourderivedstellarradiiandsurfacegravities.Figure1illustratesthepositionsofthestarscomparedwiththenon-rotatingstellarevolutiontracksofCharbonneletal.(1993)whicharecomputedwithmetallicityofZ=0.004(SMC-like).AgainfollowingVDLLweestimateevolutionarymasses(Mevol)byinterpolationinthisdiagramandthesemaybecomparedwithMspecinTable5.whereallderivedquantitiesaresummarized.WenotethatMaeder&Meynet(2001)haveproducedagridofcalculationsincludingtheeffectsofrotationforametallicityofZ=0.004.Figure1alsoshowsthelocusoftheendofH-burningmainse-quencephaseforanassumedinitialrotationalvelocityof300km/s(takenfromtheirFigure6).Notethatthislocusalsocor-respondstotheapproximatepositionoftheendofthemainsequenceforthenon-rotatingmodels.ThisisaconsequenceofthefactthatthemainsequencewideninginCharbonneletal.comesfromtheirinclusionofconvectiveovershooting,whileasimilareffectisobtainedbyMaeder&Meynetwithrota-tionallyinducedmixingalone.Thefactthatovershootingandrotationbothhavesimilarresultsinthisrespecthasbeendis-cussedpreviously,seeforexampleTalonetal.(1997).Wewillreturntothispointlaterinthediscussion,afterfirstconsideringtheobservedabundancepattern.

Nitrogenenrichmentisclearlyasignofcontaminationofthesurfacebytheproductsonthenuclearburning,inthiscasefromtheCNObi-cycle.Whatisespeciallyinterestingaboutthecurrentsampleofstars(towhichwecanaddthestarB30mentionedabove)isthattheyrepresentacoevalgroupofstarswithratherhomogeneousproperties,whoseN/Csurfaceabun-danceratioisenhancedbytypicallyafactorof10.Stellarmod-elswhichincludetheeffectsofrotationallyinducedmixinghavebeenproposedasameansofproducingtheobservedni-trogenenhancementsinmainsequenceandevolvedmassivestars.Maeder&Meynet(2001)havefollowedtheevolutionofthesurfaceabundancesforarangeofinitialrotationalve-locities.Comparingourresultstothemodelswithhighinitialrotationalvelocity,vlargenitrogenenhancements,ini=300km/s,andthereforerelativelywefindthatthebestagreementiswiththeirredsupergiantorblueloopstars.Attheendofthemainsequencethesemodelspredictanincreasein[N/H]byafactorofonly3,orabout0.5dex.Notehoweverthattheirinitialabundanceratiosareassumedtobesolarandthereforetheyoverestimatetheinitialnitrogenabundancesignificantly,

8D.J.Lennonetal.:MorenitrogenrichB-typestarsintheSMCcluster,NGC330

Fig.1.HR-diagramforstarsinNGC330showingthepositionsofthestarsrelativetothestellarevolutiontracks(solidlines)ofCharbonneletal.(1993)whicharecomputedforametallicityofZ=0.004.Tracksarelabeledwiththeirinitialmasses.ThedashedlinerepresentstheapproximatepositionoftheendofthecoreH-burningmainsequencefortherotatingmodelsofMaeder&Meynet(2001)foranassumedinitialrotationalvelocityof300km/s.Bluestars(analysedhere)arelabeledwiththeiridentifications.ForcomparisonwealsoshowthepositionsoftheredsupergiantsinNGC330analysedbyHill(1999)andGonzalez&Wallerstein(1999),whosechemicalcompositionswealsodiscuss.Theerrorbarsrepresenttypicaluncertaintiesdiscussedinthetext,forexample15%ineffectivetemperature.adoptingonefifthsolar,whichisapproximately7.3dexinournotation.

Wecanperformasimplerecalibrationoftheirresultsbyassumingthatacalculationwithaninitiallowernitrogenabun-danceof6.6dex(butsimilarcarbonandoxygen)producesthesameexcessofnitrogeninabsoluteterms.Inthiscasetheni-trogenabundanceattheendofthemainsequencewouldbeapproximately7.7dexratherthan7.8dex,aconsequenceofthefactthattheinitialnitrogenabundanceisnegligiblecom-paredwiththatwhichisproducedbythestar.Thesemodelsthereforecouldconceivablyreproducetheobservednitrogenenhancements.InfactiftheinitialN/OorN/(O+C)abundanceratiosaresmallandcanbeneglected,whichinthecaseoftheSMCisagoodapproximation,thenitiseasytoshowthatoneonlyneedsarelativelysmallfractionofcorematerialinONequilibriumtoproduceabigchangeintheobservednitrogenabundance.

Nevertheless,asFigure1shows,thereareothermorese-riousdiscrepancies.Forexample,ourobjectstendtoliered-wardsofthemainsequencedespitethewideningprovidedbytherotatingmodels(andmodelswithconvectiveovershooting).InadditionalltheB-typegiants/supergiantsinoursampleareslowrotatorswithvaluesofvsinilessthan50km/s.Bycontrasttherotatingmodelsrequireveryhighinitialrotationtoproducetheenhancednitrogenbutdonotpredictsignificantslowdownbytheendofthemainsequence.AsdiscussedinPaperI,whiletheremaybeaselectioneffectinourobserversample(wecanonlyanalysetheslowrotators)itishighlyunlikelythattherearesomanyfastrotatorsinNGC330orientedpoleon.Forex-ample,bothMazzalietal.(1994)andKeller&Bessell(1998)givevsinivaluesfor22ofthebrightestBandBe-typestarsinNGC330.Theyfindthatmaximumvalueslieintherange300–400km/sandifweassumethatoursamplehavesimilarrotationrates(v)thisimpliesthatsiniislessthan10degrees.Inotherwordsifthedistributionofiisrandomweshouldex-

D.J.Lennonetal.:MorenitrogenrichB-typestarsintheSMCcluster,NGC3309

Table5.DerivedquantitiesforB-typestarsinNGC330:Stellarradii(R/R⊙),absolutevisualmagnitudes(Mv),bolometriccor-rections(B.C.),bolometricmagnitudes(Mbol),luminosities(log(L/L⊙)andestimatesofspectroscopic(Mspec)andevolutionary(Mevol)massesinunitsofsolarmass.ThedistancemodulustotheSMCwasassumedtobe18.9inthederivationofthesequantities.Thelasttwocolumnscomparethedifferenceinspectroscopicandevolutionarymasseswiththeuncertaintiesinlogg(∆logg).

Star

R/R⊙

Mv

B.C.

Mbol

log(L/L⊙)

Mspec

Mevol

log(Mevol/Mspec)

∆logg

pectoursampletobedrawnfromabout1.5%ofalltheB-typestarsinNGC330.ThisisclearlyincompatiblewiththenumberofB-typestarsinNGC330intherelevantmagnituderange.Oneisleftwiththeconclusionthatourobjectsareintrinsicallyslowrotatorsatthepresenttime.Eithertheywerefastrotatorsinthepast,andhavesomehowsloweddown,orsomeprocessotherthanrotationallyinducedmixingisresponsiblefortheobservedabundancepattern.

Giventhesimilaritybetweenthecarbonandnitrogenabun-dancesintheblueandredstarsinNGC330itistemptingtoinvokeblueloopsasameansofexplainingourabundances.Thisseemsunlikelygiventhatnostellarevolutioncalculationspredictloopswhichprogresshotterthaneffectivetemperaturescorrespondingtolate-Bspectraltypes.WhileVenn(1999)in-vokedblueloopsfortheA-typesupergiantsthisdoesnotseemaviableoptionforourearlyB-typestars,despitesimilaritiesinCNOabundances.Mass-transferinbinariesmayalsobein-vokedtoexplainenhancednitrogenabundancesandWellsteinetal.(2001)haverecentlyproducedmodelswhichproduceni-trogenenrichedbluestarswhichcanresideinthepostmain-sequencegap.Suchstarsmayalsoappeartobeunder-massivefortheirluminositiesandisthereforetemptingtoascribethediscrepanciesbetweenspectroscopicandevolutionarymassesinTable5tobinarity.Oneshouldbecautioushoweverbe-causeourspectroscopicmasseswereestimatedfromthede-rivedsurfacegravitiesandinsomecasestheuncertaintiesinthisquantityarequitesubstantial.ThefinaltwocolumnsofTable5comparesthedifferencesbetweenspectroscopicandevolutionarymasseswiththeuncertaintiesingravityandinallbutthreecases(A02,B32andB37)themassdifferencesareeasilyaccountedfor.Theredoesappeartobeasuggestionofacorrelationbetweennitrogenenhancementandluminosityinoursample.StarsA02andB37arethemostluminousandmostN-richofoursample,howevertheirpositionsareperhapscon-sistentwithbeingcore-heliumburningstarsontheirwayred-wards(thisphaserepresentedbytheslightkinkinthe20solarmasstrackforthenon-rotatingmodels).Theproblemfortherotatingmodelsisthatthisphasebecomesprogressivelycoolerandshorterlivedasinitialrotationalvelocityisincreased,withonlytheslowrotatorsspendinganysignificanttimeinthispartofHR-diagram.However,suchstarsshouldnotbesignificantly

N-enriched,incontradictiontotheobservations.B32wouldappeartobethebestcandidateforthebinaryevolutionhypoth-esis,therealproblembeingthatmuchbetterconstraintsareneededonthegravityofthisobject,andindeedallotherstarsinoursample,beforedefinitivestatementscanbemadeaboutpossiblemassdiscrepancies.Finally,whilebinarityappearstobeanattractivescenario,ithasasignificantprobleminthatitisexpectedthattheproductsofmassaccretionwillbefastrotatorshavingbeenspunupbytheaccretedmaterial.Inaddi-tionradialvelocitiesofallourstarsaretypicaloftheNGC330cluster(Table1)althoughitmustbenotedthatexpectedra-dialvelocityamplitudesofthekindofsystemspredictedbythemodelsofWellstein&Langer(2001)aretypicallytheorderof10–20km/s.

FinallywenotethatVDLLcarriedoutasimilarstudytothepresentonebutforthesolarmetallicityclusterh+χPer.TheydidnotfindevidenceforsignificantnitrogenoverabundancesinanyoftheevolvedB-typestarsinthiscluster.Howeverifwetakethenitrogenenrichmentsfoundhere,inabsoluteterms,applythesetotheirgalacticcounterpartsitisclearthatthisleadstoenhancementstheorderofafactorof2–3.Forsomeobjectsinh+χPerthismagnitudeofanitrogenenhancementisconsistentwiththeobservations.ItissimplymoreobviousintheSMCstarsgiventheirinitialverylownitrogenabundance.

6.Summary

Inthispaperwehavepresentednewresultsfor5B-typegi-ants/supergiantsintheSMCclusterNGC330.Togetherwithourpreviouswork,plusresultsforoneothersuchstarinNGC330presentedbyKornetal.thisbringsthetotalofbrightB-typestarsanalysedinthisclusterto8.Allthestarshavethefollowingcharacteristics:

–Theyarenitrogenrichandpossiblycarbonpoor,withper-hapssomemarginalevidenceforoxygendepletion,withtheirN/Cratiosbeingenhancedbyapproximatelyafactorof10relativetotheSMCnorm.Thenitrogenenhancementsareremarkablyhomogeneous.

–Theyalllieclosetothemain-sequencebutwithastrongtendencytoliebeyondtheendofthemain-sequenceeven

10D.J.Lennonetal.:MorenitrogenrichB-typestarsintheSMCcluster,NGC330

Table6.ComparisonofourabundanceestimatescorrectedforNLTEeffectswithresultsforotherstarsinNGC330,viz.theB-typestarB30analysedbyKornetal.(NLTECandO)andReitermannetal.(LTEN)(B30)andcoolsupergiantsfromHill(H99)andGonzalez&Wallerstein(GW).AlsotabulatedareotherSMCabundanceestimates,viz.themainsequencestarAV304(Rollestonetal.2002correctedforNLTEeffects),A-typesupergiantsfromVenn(A-stars),B-typegiantsandsupergiantsfromKornetal.(B-stars)andcoolsupergiantsfromHilletal.(K-stars).

Element

AV304

NGC330

ThispaperB30H99

GW

SMCsupergiants

A-starsB-starsK-stars

ifmain-sequencewideningthroughrotationorconvectiveovershootingisconsidered.

–Theyallappeartobecurrentlyslowrotatorswithvsinival-uesbelow50km/s.Thesampleistoolargeforthistobeaninclinationangleeffectforfastrotators.

–Thenitrogenenhancementsagreewellwiththosefoundforcoolredsupergiantsinthesamecluster.

Singlestarstellarevolutionmodelsincludingtheeffectsofrotationmayindeedbeabletoexplaintheabundancepat-terns,althoughwenotethatthereisaclearneedforstellarevolutioncalculationswiththecorrectmixofabundancesfortheSMC,inparticularscalingthesolarcompositionproducesmuchtoohighaninitialnitrogenabundancefortheSMC.Howeversinglestarmodelssufferfromseveredifficultiesinreproducingboththeeffectivetemperaturesofoursampleandtheirlowrotationalvelocities.Binarystarmodelsmaybeabletoproducestarsinthecorrecteffectivetemperaturerangebutmayalsohavesimilarproblemstosinglestarmodelsinre-producingthelowrotationalvelocities.ThesimilaritybetweenblueB-typegiant/supergiantnitrogenenhancementsandthosefoundforredsupergiantsinNGC330issurprisingsinceonemightexpectthelattertobemorenitrogenenrichedwiththebluestars’enhancementcomingfromrotationwhilethatoftheredstarscomingfromdredge-upprocesses.Inconclusion,theB-typestarsconsideredherearestillsomethingofapuzzleandnotwellmodeledbycurrentstellarevolutioncalculations.Moreover,giventhelargenumberoftheseobjectsinNGC330itisclearthattheyarenottheproductofsomepeculiarevolu-tionarypathandthereforerepresentanimportantchallengetocurrentstellarevolutiontheory.

Acknowledgements.DatareductionwasperformedonthePPARCfundedNorthernIrelandSTARLINKnode.DJLisgratefulforNOVAfundingforavisittoUtrechtinMay2001duringwhichtimemuchofthepresentworkcarriedout.Thanksarealsoduetoanumberofpeo-pleforcontributionstothisproject;NorbertLanger,PaoloMazzali,GianniMarconi,RobertRollestonandKimVenn.

References

Arp,H.,1959,AJ,64,254

Balona,L.A.,1992,MNRAS,256,425

Becker,S.R.,Butler,K.,1989,A&A209,244

Caloi,V.,Casstella,A.,Castellani,V.,Walker,A.,1993,A&A,271,

109

Carney,B.W.,Janes,K.A.,Flower,P.J.,1985,AJ,90,1196

Charbonnel,C.,Meynet,G.,Maeder,A.,Schaller,G.,Schaerer,D,

1993,A&AS101,415

Chiosi,C.,Vallenari,A.,Bressan,A.,Deng,L.,Ortolani,S.,1995,

A&A293,710

Dufton,P.L.,McErlean,N.D.,Lennon,D.J.,Ryans,R.S.I.,2000,

A&A353,311

Eber,F.,Butler,K.,1988,A&A202,153Feast,M.W.,1972,MNRAS,159,113

GiesD.R.,LambertD.L.,1992,ApJ387,673

Gonzalez,G.,Wallerstein,G.1999,ApJ117,2286Grebel,E.K.,Richtler,T.,1992,A&A254,359

Grebel,E.K.,JamesRoberts,W.,Brandner,W.,A&A311,470Hill,V.,Barbuy,B.,Spite,M.1997,A&A323,461Hill,V.,1999,A&A345,430

Jeffery,C.S.,1991,NewsletteronAnalysisofAstronomicalSpectra,

No.16,17

Keller,S.C.,Bessell,M.S.,1998,A&A340,397

Keller,S.C.,Bessell,M.S.,DaCosta,G.S.,2000,AJ119,1748KilianJ.,1994,A&A262,171KilianJ.,1994,A&A282,867

Korn,A.J.,Becker,S.R.,Gummersbach,C.A.,Wolf,B.,2000,A&A

353,655

KuruczR.L.,1992,Rev.Mex.Astrofis.23,45Langer,N.,Maeder,A.,1995,A&A295,685

Lennon,D.J.,Mazzali,P.A.,Pasian,F.,Bonifacio,P.,Castellani,V.,

1994,Sp.Sci.Rev.66,169

Lennon,D.J.,Dufton,P.L.,Mazzali,P.A.,Pasian,F.,Marconi,G.,

1996A&A314,243

McErlean,N.D.,Lennon,D.J.,Dufton,P.L.,2000,A&A349,553Mazzali,P.A.,Pasian,F.,Lennon,D.J.,Bonifacio,P.,Castellani,V.,

1994,Proc.IAUSymp.162,eds.L.Balona,H.Heinrichs,J.M.LeContel,Kluwer,Dordrecht,269

Maeder,A.,Grebel,E.,Mermilliod,J.C.,1999,A&A346,459Maeder,A.,Meynet,G.,2001,A&A373,555

Mazzali,P.A.,Lennon,D.J.,Pasian,F.,Marconi,G.,Baade,D.,

Castellani,V.,1996,A&A316,173

Reitermann,A.,Stahl,O.,Wolf,B.,Baschek,B.,1990,A&A234,109Robertson,J.W.,1974,A&AS15,261

Rodriguez-Pascuel,P.M.,Gonzalez-Riestra,R.,Schartel,N.,

Wamsteker,W.,1999,A&AS139,183

Rolleston,W.R.J.,Dufton,P.L.,Fitzsimmons,A.,Howarth,I.D.,

Irwin,M.J.,1993,A&A277,10

Rolleston,W.R.J.,Venn,K.A.,Tolstoy,E.,Dufton,P.L.,2002,submit-tedtoA&A

Rolleston,W.R.J.,Smartt,S.J.,Dufton,P.L.,Ryans,R.S.I.,2000,

A&A363,537

Seaton,M.J.,1979,MNRAS187,73

Smartt,S.J.,Rolleston,W.R.J.,Dufton,P.L.,1996,A&A305,164

D.J.Lennonetal.:MorenitrogenrichB-typestarsintheSMCcluster,NGC330

Smartt,S.J.,Venn,K.A.,Dufton,P.L.,Rolleston,W.R.J.,Keenan,F.P.,

2001,A&A367,86

Spite,F.,Richtler,T.,Spite,M.,1991,A&A252,557Stothers,R.B.,Chin,C.-W.,1992a,ApJ,390,L33Stothers,R.B.,Chin,C.-W.,1992b,ApJ,390,136

Talon,S.,Zahn,J.-P.,Maeder,A.,Meynet,G.,1997,A&A322,209Thompson,G.I.,Nandy,K.,Morgan,D.H.,Houziaux,L.,1988,

MNRAS230,429

Venn,K.A.,1999,ApJ518,405

Vrancken,M.,Lennon,D.J.,Dufton,P.L.,Lambert,D.L.,2000,A&A

358,639

Wellstein,S.,Langer,N.,Braun,H.,2001,A&A369,939

Willmarth,D.,Barnes,J.,1994,Auser’sguidetoreducingechelle

spectrawithIRAF,NAOA,Tucson,InternalDocument

11