Mon.Not.R.Astron.Soc.000,1–10(2014) PrintedJanuary29,2015 (MNLATEXstylefilev2.2) Discovery of an OB Runaway Star Inside SNR S147 B. Dinçel1(cid:63), R. Neuhäuser1, S.K. Yerli2, A. Ankay3, N. Tetzlaff1, G. Torres4, M. Mugrauer1 1AstrophysikalischesInstitutundUniversitäts-SternwarteJena,07745Jena,Germany 2OrtaDog˘uTeknikÜniversitesi,DepartmentofPhysics,06531Ankara,Turkey 5 3Bog˘aziçiUniversity,DepartmentofPhysics,34342I˙stanbul,Turkey 1 4Harvard-SmithsonianCenterforAstrophysics,60GardenSt.,MailStop20,CambridgeMA02138,USA 0 2 n Accepted2015January17.Received2014December12;inoriginalform2014May25 a J 8 ABSTRACT 2 We present first results of a long term study: Searching for OB–type runaway stars inside supernovaremnants(SNRs).Weidentifiedspectraltypesandmeasuredradialvelocities(RV) ] by optical spectroscopic observations and we found an early type runaway star inside SNR R S147.HD37424isaB0.5Vtypestarwithapeculiarvelocityof74±8 kms−1.Tracingback S thepasttrajectoriesviaMonteCarlosimulations,wefoundthatHD37424waslocatedatthe . h samepositionasthecentralcompactobject,PSRJ0538+2817,30±4 kyrago.Thisposition p is only ∼4 arcmin away from the geometrical center of the SNR. So, we suggest that HD - 37424wasthepre–supernovabinarycompaniontotheprogenitorofthepulsarandtheSNR. o We found a distance of 1333+103 pc to the SNR. The zero age main sequence progenitor r −112 t mass should be greater than 13 M . The age is 30±4 kyr and the total visual absorption s (cid:12) towardsthecenteris1.28±0.06mag.Fordifferentprogenitormasses,wecalculatedthepre– a [ supernovabinaryparameters.TheRocheLoberadiisuggestthatitwasaninteractingbinary inthelatestagesoftheprogenitor. 1 v Keywords: OBRunaway Stars;HD 37424,Neutron Stars;pulsars; PSRJ0538+2817, Su- 0 pernovaRemnants;SNRS147 2 2 7 0 . 1 INTRODUCTION binarieslike4U1700-37(Ankayetal.2001)andVelaX–1(Kaper 1 etal.1997)aresuchexamples. 0 High space velocities of OB runaway stars are explained by two 5 Yet,thelowrateofX–raybinariesandthehighrateofisolated independentmechanisms:dynamicalejectionduetogravitational 1 neutronstarsbytakingintoconsiderationtheselectioneffects,the interactionsofmassivestarsinclustercores(Povedaetal.1967) : binarydisruptionislikelytooccurinmostcases(Guseinovetal. v and binary disruption as a result of a supernova explosion of the 2005)(hereafter,G05).Thekinematicsofthebinarydisruptiondue i initiallymoremassivecomponent(Blaauw1961).Bothscenarios X toanasymmetricsupernovaexplosioniswidelydiscussedinTauris areviable,butwhetheroneofthemechanismsisdominantisstill r &Takens(1998).However,asampleofobservationallyconfirmed uncertain.Accordingtothevirialtheorem,throughasymmetricsu- a OB runaway–NS couples is needed for a better understanding of pernovaexplosioninabinarysystem,ifmorethanhalfofthetotal theproblem. massofthesystemisreleased,thenthenewbornneutronstar(or The importance of searching for OB runaways inside SNRs blackhole)andthenon–degeneratecomponentarenomoregravi- wasfirstmentionedinvandenBergh(1980).However,thekine- tationallybound(Blaauw1961).However,theenergystoredinthe maticalstudyofknownOBstarsinsideSNRswasconcludedwith orbit, in most cases, is not sufficient to produce the neutron star kickvelocitiesthataretypicallyintherangeof300–500 kms−1 a lack of OB runaways due to the poor sample of SNRs and OB stars.Still,thereisnoknownOorB–typerunawaystarthatcanbe (Lyne & Lorimer 1994; Allakhverdiev et al. 1997; Hobbs et al. directlylinkedtoanSNRgivenintheliterature. 2005;Hansen&Phinney1997).Theasymmetryinsupernovaex- plosions is responsible for such high velocities (Wongwathanarat InG05,theoutcomeofexploringrunawaypairsfrombinary etal.2013).Therefore,therearenopulsarcompanionstomanyof supernovadisruptionisbroadlydiscussed.Firstly,identifyingthe theOBrunawaystars(Sayeretal.1996).Insomecases,thecom- explosion centers more precisely will be useful for determining pactobjectdoesnotreceiveasignificantkickand/orthemajority thevelocitiesofyoungNSsofwhichpropermotionmeasurements of thetotal mass isstored on thesecondary through conservative have high uncertainties. Thus, the kick that is gained by the NS masstransfer,hence,thecompactobjectremainsboundtothecom- duetotheasymmetryoftheSNcanbedeterminedmoreprecisely. panionstar(vandenHeuvel1993).TherunawayhighmassX–ray Secondly, the distance to the remnant can be measured more ac- curatelybystudyingtherunawaystar,asitcannotmovefaraway fromtheexplosioncenterintheobservationallifetimeoftheSNR. (cid:63) Correspondenceaddress:[email protected] Finally,possibleeffectsofaclosebinarysystemontheasymmetry (cid:13)c 2014RAS 2 Dinçeletal oftheSNecanalsobeexamined.Additionally,itwouldbeadirect Table1.DistanceestimatesforS147.(Rl)denotestheradiuslowerlimitof evidenceofthebinarysupernovascenario(BSS). theSNR. The observational efforts are concentrated on runaway–NS couplingandabundanceinvestigationsofrunawaystars.Someex- Distance(kpc) Method Reference amples of the runaway–NS pairs (components are separated) are 1.6±0.3 Σ−D Sofueetal.(1980) PSRB1929+10–ζ Oph(Hoogerwerfetal.2001;Bobylev2008; 0.8–1.37 Rl,Σ−D Kunduetal.(1980) Tetzlaffetal.2010),PSRJ0630-2834–HIP47155(Tetzlaffetal. 0.6 Rl Kirshner&Arnold(1979) 2013) and PSRJ0826+2637 – HIP13962 (Tetzlaff et al. 2014). 0.9 Σ−D Clark&Caswell(1976) Basedontheirmotionsinspace,thepulsarsandthecorresponding 0.8±0.1 AV Fesenetal.(1985) 1.06 Σ−D Guseinovetal.(2003) runawaystarsweretracedbackintimebyusing3–DMonteCarlo 0.88< HighVelGas Sallmen&Welsh(2004) simulations. They were found at the same position and the same 1.2 PulsarDM Krameretal.(2003) timeinsideayoungopencluster.ThePSRJ0630-2834–HIP47155 1.47+0.42 PulsarPlx Ngetal.(2007) pairisthoughttobeejectedfromveryoldSNRAntlia. 1.3+−0.02.227 PulsarPlx Chatterjeeetal.(2009) Thereareconsiderableuncertaintiesinthesecasesasthesu- −0.16 pernova events took place more than 105 yr ago. The separation betweentheobjectsisverylargeandtheSNRhasfadedawaylong angulardiameter(θ)oftherelatedSNR.Thisvalueissomewhatre- agoand/orthecomponentsareoutsideoftheSNR(theAntliacase). laxedtoθ/6consideringtheuncertaintiesinthegeometricalcenters Otherimportantobservationalevidenceistheenhancementofα– oftheSNRs.Thestarsinthisregionareexpectedtobeconsistent process elements in the hyper–velocity star HD271791. The star withtherespectiveSNRintermsofdistanceandreddening. isproposedtobeejectedfromamassiveclosebinarysystemdue Forthiscomparison,theadopteddistancesofSNRsgivenin to an SN that enriches its photosphere in elements which can be Guseinovetal.(2003,2004a,b),andA valuesfromNeckeletal. V synthesizedinlargeamountsduringtheevolutionoftheprogenitor (1980) were used. 48 SNRs within 5 kpc from the Sun were se- (Przybillaetal.2008). lectedforinvestigation. Themethodfollowedinourworkisadirectstudyofpossible Inthispaper,theresultoftherunawaysearchinSNRG180.0- runawaystarsinsideSNRsasdescribedinG05. 1.7 (S147) is given. The kinematic relation between the runaway Briefly,assumingthatthemassivebinarymassratioisgreater starHD37424andthepulsarPSRJ0538+2817isshown,thepossi- than1:4,OB–typestarcandidatesaredeterminedbyacarefulstudy blehostOBassociationisdiscussed,theSNRparametersarecon- of their BVJHK magnitudes obtained from the UCAC4 catalog strainedandthepre–supernovabinaryisconstructed. (Zachariasetal.2012).Theangularseparationofthesourcesfrom thegeometricalcentersoftheremnantsarewithinthelimitsofthe maximumangulardistancethatarunawaystarcanhaveinalife- 2 S147,PSRJ0538+2817ANDHD37424 timeofanSNR.Thedistancemoduliarecalculatedforallsources withinthisregion,andthesourceshavingextinctionsandradialdis- S147isashelltypeSNRlocatedintheGalacticanti–centerdirec- tancesconsistentwiththoseoftheSNRsareconsideredas"candi- tion.Itis180arcminindiameterwithageometricalcenter(GC) dates".Measuringtheradialvelocitiesandidentifyingthespectral atα = 05h39m00s,δ = +27◦50(cid:48)00(cid:48)(cid:48) (Green2009).Itwasfirst typesoftheseobjectsviaspectroscopyrevealtheirrunawaynature, mentionedasanSNRcandidateinMinkowski(1958).Thecompact theiryouthandtheexactspatialrelationswiththeSNRs,inother objectrelatedtotheSNRisradiopulsarPSRJ0538+2817(Ander- words,thegeneticconnections. sonetal.1996).Inopticalbands,theshellstructureiswelldefined AlthoughtherunawaystarsarisingfromBSSmayalsobelate and dominated by filamentary emission in Hα. The emission is typestars,itistimeconsumingtocheckthepossiblerunawayna- brighter in the north and south edges and mainly concentrated in tureofallthestarsinsideeachSNR.Also,thelatetypestarsare thesouthernparts.Despiteofitsoldage,itconservesthespherical toofainttoobserveatlargedistances.Astarhavingthesameage symmetryexceptfortheblowoutregionsineastandwest(Figure asasupernovaprogenitormustbeyoung.AsOBtypestarsevolve 1). The total absorption in the V band is A =0.7 ± 0.2 magni- V faster,theyautomaticallysatisfythiscondition.Furthermore,high tude(Fesenetal.1985).Radioobservationsrevealthatthespectral massstarsarerareobjects.AnOBrunawaystardiscoveredinside index is unusually varying. The shell structure observed at radio anSNRcanbeexplainedbytheBSS.Consideringtheveryshort wavelengthscoincideswiththatinopticalbandsandiswelldefined observable lifetimes of SNRs and relatively short later evolution (Fuerst&Reich1986).But,noX–rayemissionisobservedfrom stagesofstars,mostprobably,mainsequencestarsareexpectedas theremnant(Sauvageotetal.1990).Suggesteddistancesvaryfrom OBrunawaysconnectedtoSNRs. 0.6to1.9 kpcinvariouspublications(Table1).Fourofthemea- The space velocity of an OB runaway star is thought to be surementsarebasedonΣ−Drelationwhichisusefultoestimate largerthan30−40 kms−1(Feast&Shuttleworth1965;Sayeretal. thedistancebyusingthesurfacebrightnessoftheSNR.InSofue 1996).AmoreprecisevalueisproposedinTetzlaffetal.(2011); etal.(1980)andKunduetal.(1980),themodelofMilne(1979)is 28 kms−1in3–Dand20 kms−1in2–D.Tosummarize,amain– used,whileinClark&Caswell(1976)andGuseinovetal.(2003), sequence OB type star of which at least one component of the estimationsarebasedontheirownmodels.Theradiuslowerlimit velocity vector is greater than 20 kms−1 is searched in selected calculatedthroughSNRdynamicsbasedonthemodelofChevalier SNRs. (1974) sets another constraint on the SNR distance. The distance Thefirstcriterionofthecandidateselectionistherestrictionof derivedfromthepulsar’sparallaxordispersionmeasureislarger theangularposition.MostoftheOBrunawayshavepeculiarveloc- thanthedistancesuggestedforthebackgroundstarsofwhichspec- itieslowerthan80 kms−1(Gies&Bolton1986;Philpetal.1996; trashowhighvelocitygasrelatedtotheSNR. Tetzlaffetal.2011).AstheshockwavevelocityoftheSNRsarede- Theagewasestimatedasmorethan100 kyrfromtheSedov celeratedfromroughly10000 kms−1toseveralhundred kms−1, solutionbyusingashockvelocity∼80–100 kms−1 (Sofueetal. itisexpectedthattherunawaystarcannotexceedonetenthofthe 1980;Kunduetal.1980).ThisvelocitywasderivedfromtheRV (cid:13)c 2014RAS,MNRAS000,1–10 RunawayinS147 3 Figure1. The4.8◦×4.1◦ HαimageofSNRS147takenattheUniversityObservatoryJena.Thegreencrossrepresentsthepulsar,thecyancrossshows theGCoftheSNRandtheyellowcircleisHD37424.Theyellowvectorsshowthepropermotionoftheobjects.Theredboxisthezoom–inareashownin therightpanel.Thewhitearrowsarethetracingbackconesofthepropermotionfor29300yr.Theanglebetweenthepmvectorsis139◦−148◦In2–D calculations,bothobjectscometowardseachotherascloseas0.126 pcwhichmeansthattheyhaveacommonorigin;binarysupernovadisruption. ofthefilamentaryknotsinLozinskaya(1976);Kirshner&Arnold Table2.BVJHK(inmag)andpm(in masyr−1)valuesofHD37424. (1979).Howeverthepulsar–SNRrelationindicatesanageof30± 4 kyr(from the travel time from the GC to the present position) Parameter Value (Krameretal.2003;Ngetal.2007).Foradistanceof1.3 kpc,and B 9.062±0.017 V 8.989±0.019 anangulardiameterof200arcmin,theradius(R)oftheSNRis38 J 8.666±0.017 pc.Then,assuminganageof30 kyr,theremnantisinitsSedov H 8.696±0.026 Phasehavingablastwavevelocityof500 kms−1.Theestimated K 8.699±0.019 explosionenergyisbetween1−3×1051ergandthecorresponding µ∗ 10.8±0.8 inter–cloud gas density is 0.03−0.1 cm−3 (Katsuta et al. 2012). µαδ -10.2±0.6 The SNRisexpanding in alow densitymedium, probablyin the cavitygeneratedbytheprogenitor. Thecentralsource,PSRJ0538+2817,isanextensivelystud- ied radio and X–ray pulsar located at α = 05h38m25.1s, δ = etal.2003).ThehydrogenatmospheremodelwithB= 1×1012 +28◦17(cid:48)09.2(cid:48)(cid:48),∼28arcminawayfromtheGCtowardsnorth.The GgivesT=2.12×106KandR(cid:39)10 km(Zavlin&Pavlov2004). 143.16msperiodand3.67×10−15ss−1periodderivative(Ander- AfaintpulsarwindnebulaisobservedinX–raysbutnotinradio son et al. 1996) imply a characteristic age of ∼620 kyr which is (Gaensleretal.2000).Itisnotknownwhethertheobservedelon- ∼20 times larger than its kinematic age of 30 kyr. This discrep- gatedstructureisthetorusorthejetsofthePWN.Butassuming ancyisexplainedbyeitheralonginitialspinperiodofP = 139 thatitisduetothetorus,itprovidesvaluableinformationonthe 0 ms(Krameretal.2003)orstrongmagneticfielddecay(Guseinov spin–kickalignment. et al. 2004c). The parallax distance was measured as 1.47+0.42 HD37424 is a sound OB runaway star at 10.3 arcmin −0.27 kpc (Ng et al. 2007) and 1.3+0.22 kpc (Chatterjee et al. 2009) away from the GC to the west at α = 05h39m44.4s, δ = −0.16 and the dispersion measure (DM) distance is 1.2 kpc (Kramer +27◦46(cid:48)51.2(cid:48)(cid:48).Photometricandkinematicinformationonthestar etal.2003)byusingtheNE2001modelofCordes&Lazio(2002) isgiveninTable2.Whilepropermotionvalueswereretrievedfrom whichareconsistentwitheachother.Themostpreciselymeasured theUCAC4catalog(Zachariasetal.2012),thephotometricmag- propermotionisµ∗ =−23.57+0.10 masyr−1,µ =52.87+0.09 nitudes are obtained from the ASCC-2.5V3 catalog (Kharchenko α −0.10 δ −0.10 masyr−1correspondingtoatransversevelocityof357+59 kms−1 2001). Its spectral type was previously identified as B0.5/1 IV/V −43 at1.3+0.22 kpc(Chatterjeeetal.2009).Althoughitpresentsno (Clausen & Jensen 1979). As there was no public data nor a vi- −0.16 clear γ–ray emission (Katsuta et al. 2012), the pulsar is observ- sualized spectrum, we established the spectral type again by tak- able in soft X–rays due to thermal emission. PSRJ0538+2817 is ingaspectrum.Thestarhasaveryhighpropermotioncompared famousforbeingoneofthefewthermalpulsars:Basedonablack- tootherearlytypeandpossiblydistantstars.Thereisnoreported body model, the surface temperature and the emitting radius are variabilityorbinarityintheliterature.Hence,HD37424isagood givenasT= 2.12×106 KandR= 1.68±0.05 km(McGowan OBrunawaystarcandidate. (cid:13)c 2014RAS,MNRAS000,1–10 4 Dinçeletal 3 OBSERVATIONS HD37424 was observed at the TÜB˙ITAK National Observa- tory (TUG) with the TUG Faint Object Spectrograph and Cam- era (TFOSC) mounted on the 150cm Russian Turkish Telescope (RTT–150). The observations were carried out on 2013 August 9 and10.Twospectrawereobtainedwitha600sexposuretimefor each. Grism 14 was used with a 67 micron slit. The wavelength rangeinthisconfigurationis3270–6120Åandtheresolvingpower (R)is∼1337.5Helampspectratakeninthesamenightwiththe object were used for wavelength calibration, and 5 halogen lamp spectrawereobtainedinthebeginningandattheendofthenight forflat–fielding.Also,2moretargetshadbeenobservedinthesame wayon2012September13and14forthesearchfortherunaway star.Withthesamepurpose,12objectswereobservedattheCalar Alto Observatory on 2013 January 29 with Calar Alto Faint Ob- jectSpectrograph(CAFOS)onthe2.2metertelescope.Thecov- ered wavelength range of the grism B–200 is 3200–7000 Å and R∼550. 3 HgCdAr and 3 halogen spectra were taken in the be- Figure2.TFOSCspectrumofHD37424comparedtomainsequenceB ginningandintheendofthenightforwavelengthcalibrationand typestars.Distinctivefeaturesaremarked.Fromlefttoright;SiIV λ4089, flat–fielding.ThehighresolutionobservationsofHD37424were HeII λ4200, SiIII λ4552, OII λ4640, CIII+OII λ4650 blend, HeII performedattheFredL.WhippleObservatoryon2013September λ4686,HeI λ4713 16and21.TheTillinghastReflectorEchelleSpectrograph(TRES) mounted on the 1.5 meter telescope was used. Two spectra were spectra retrieved from Walborn & Fitzpatrick (1990). To main- takenatR∼44000inabroadwavelengthrange;3800–9100Åfor tain equally resolved spectra, the standard spectra were boxcar 300s exposure time. HD36665 which is supposed to be a back- smoothedby13.Figure2showstheTFOSCspectrumofHD37424 groundobjecttotheSNRwasobservedon2014February11with comparing with main sequence B type stars from various sub- theFibreLinkedEchelleAstronomicalSpectrograph(FLECHAS) classes. HD37424 shows HeII λ4686 absorption indicating that (Mugraueretal.2014)attheUniversityObservatoryJena.Thecov- thespectraltypeisearlierthanB1.Thisfeatureisnotseeninstars eredwavelengthrangeis3900–8100ÅandR∼9200.Threespec- withaspectraltypelaterthanB0.7.Ontheotherhand,HeI lines tra with 600 s exposure time were obtained. Three tungsten and (i.e.HeI λ4713)arenotasweakasinOtypestarscomparedto 3 Th-Ar lamp spectra were taken immediately before this set for HeII lines;λ4200,λ4686.TheSiIV λ4116lineistotallyblended flat–fieldingandwavelengthcalibration. and unmeasurable. SiIII λ4552 is in low strength compared to TheobservationsofSNRS147wereperformedattheUniver- SiIV λ4089likeintheB0Vtypespectrum.Thisratiodecreases sityObservatoryJenawiththeSchmidt-Teleskop-Kamera(STK), towardshighertemperatures,butHeI λ4541isnotstrongerthan operatedintheSchmidt-focusofthe90cmtelescopeoftheobser- SiIII λ4552asseeninB0VandO9.5Vspectra.Giventheequal vatory(seeMugrauer&Berthold(2010)).Theobservationswere strengthsofOII λ4640,HeII λ4686,HeI λ4713andmorein- carried out with the narrow–band Hα filter of the STK, which tenseCIII+OII λ4650blends,HD37424matchesbestwithB0.2V. spansafull–width–half–maximumof5nm.Intotal,22fieldson ConsideringtheslightlystrongerHeI λ4713,B0.5wouldbethe theskywereobservedduring5nights,eachwithafieldofviewof best approach to the temperature class of the star. SiIV λ4089 52.8(cid:48)×52.8(cid:48).Ateachpositiontwo600sintegrationsweretaken strengthincreasesagainstHeI λ4026–4144astheluminosityclass withtheSTK,i.e.about7.3hoursofintegrationtimeintotal.The goes higher. However, neutral helium lines are strong enough to individualrawimagesperfieldweredark-andflatfield–corrected judgethatHD37424isamainsequencestar;B0.5V.Thespectral andfinallyaveraged.Pointsourcesdetectedintheoverlapregions typeisbetweenB0.2V–B0.7V,butitisassumedasB0.5±0.5Vin ofthefullyprocessedimageswerethenusedtocreatethemosaic thedistanceandextinctioncalculationstoavoidunderestimationof imageofSNRS147,showninFigure1,whichcoversatotalfield theerrors. ofviewof4.8◦×4.1◦. ThesourcesobservedbyCAFOSareforegroundstarswitha FLECHAS, TFOSC and CAFOS data were reduced with maximumdistanceof671+96 pc.Theirspectralpropertieswere −56 IRAF. All raw frames from TFOSC and CAFOS were over–scan identifiedbythesameprocedureasdoneforHD37424.However, strippedandnofurthersubtractionwasdoneaslongastheover– duetothelowerresolutionCAFOSspectrahavehigheruncertainty scanneddarkframesyieldsnullcount.Yet,FLECHASimageswere (Table3).WeusedFLECHAStoidentifythespectraltypesofthe subtractedbythedarkframestakeninthepreviousnight.Theex- backgroundstarHD36665.ThisisaB1±0.5Vtypestar(Figure3). posuretimeofadarkframeisthesameasthatofthecorresponding scienceorcalibrationframe.Hence,theilluminationeffectwasre- TRESdatawereusedforradialvelocity(RV)measurements moved. of HD37424. Due to the low signal to noise, 10 absorption fea- Then, the flat images were combined while rejecting the turescouldbeused(Table4).Doubletsortripletlinesareavoided frames having minimum and maximum counts. Next, the object as these features increased the dispersion significantly. The lines spectrawereextractedbyassigninganappropriatebackgroundre- were fitted by Gaussian functions and the best shifts were deter- giontoeach.Thearcspectrawereextractedastracedbythecorre- mined.TwomoreGaussianfitswithdifferentcenterswereapplied spondingapertureoftheobjectspectra.Afterwavelengthcalibra- foreachline;onefitstheouteredgeofthenoiseoftheblueside tionwasperformed,allspectrawerenormalized. whilefittingtheinneredgeofthered.Theotherfitstheouteredge The final spectra were compared with MK standard star oftheredandtheinneredgeoftheblue.Bydoingthis,underesti- (cid:13)c 2014RAS,MNRAS000,1–10 RunawayinS147 5 Table4.RadialvelocitymeasurementsofHD37424.Thelaboratorywave- lengthofthefeatures,theGaussianwidth(σ)ofthefit,thewavelengthshift obtainedbythebestfittingGaussian,upperandlowerlimitsoftheshift fromdifferentGaussianfitsandthefinalerrorarepresented.Thefeatures aregiveninÅwhileallothercolumnsarein kms−1 ObservationsonSeptember16th,2013 Feature σ BF UL LL E 3889.051 164 -10.1 19.3 -29.2 29.4 3970.074 137 -0.5 21.4 -22.3 21.9 4101.737 144 -9.4 12.2 -32.3 22.9 4143.759 98 +4.0 17.1 -10.9 14.9 4340.468 142 -10.6 6.2 -28.8 18.2 4387.928 81 -13.1 -7.1 -24.1 11 4861.332 139 -13.8 0.9 -18.9 14.7 4921.929 89 -16.5 -1.6 -32.4 15.9 Figure3. TheFLECHASspectrumofHD36665(black)incomparison 6562.817 115 -14.7 0.3 -21.5 15 withaB1Vtemplate. 6678.149 82 -7.5 0.3 -14.5 7.8 ObservationsonSeptember21st,2013 Table3.Observedstarsintheregion.Names,angularseparationstothe Feature σ BF UL LL E SNRGC(inarcmin),spectraltypes,instrumentsusedanddistanceswith 3889.051 164 -12.6 6.2 -28.2 18.8 upper and lower errors (in pc) are given. The error in temperature sub- 3970.074 137 -1.4 18.0 -19.8 19.4 classes is ±1 for those with integer value and ±0.5 for those with half 4101.737 144 -5.1 14.9 -21.2 20 integervalue.TheresolutionofCAFOSisnotenoughtodistinguishthe 4143.759 98 -1.9 7.2 -8.7 9.1 luminosityclassesIVandV.Yet,thestarsareassumedtobeinluminosity 4340.468 142 -11.5 11.9 -36.6 25.1 classV. 4387.928 81 -8.8 -4.1 -19.3 10.5 4861.332 139 -11.7 -3.0 -19.5 8.7 Name Angular Spec. Inst.1 Distance Distance 4921.929 89 -17.2 -4.9 -31.6 14.4 Separation Type (Error) 6562.817 115 -14.4 1.4 -29.7 15.8 TYC1869-01281-1 19.4 A2V C 574 +73/-41 6678.149 82 -6.7 3.3 -15.1 10 TYC1869-01317-1 5.5 B9.5V T 1013 +183/-174 BF:Bestfit,UL:UpperLimit,LL:LowerLimit,E:Error TYC1869-01334-1 25.5 F2V C 289 +38/-34 TYC1869-01376-1 27.9 A7V C 428 +18/-32 TYC1869-01505-1 23.5 F1V C 380 +50/-44 TYC1869-01610-1 17.6 F7V C 245 +18/-16 Table5.MeasuredvelocitiesforinterstellarCaII–KandHandNaI–D1 TYC1869-01632-1 27.2 F8V C 316 +37/-26 andD2lines.TheselineshavealowGaussianwidthunliketheintrinsic TYC1869-01642-1 5.61 B9.5V T 952 +172/-163 featuresofthestar.Theaveragevelocityis12.1 kms−1 withastandard TYC1869-01679-1 25.6 G2V C 170 +11/-9 deviation0.5 kms−1ineachobservation.Thedatainthefirstcolumnare TYC1869-01749-1 20.1 F8V C 166 +19/-14 in(Å)whiletheothersarein kms−1. TYC1873-00145-1 25.0 A3V C 603 +46/-50 TYC1873-00307-1 29.2 F8V C 326 +40/-29 Feature σ Velocity Velocity TYC1873-00347-1 19.8 A0V C 671 +96/-56 (Day1) (Day2) UCAC4-589-020390 22.1 F4V C 381 +14/-16 3933.664 7.5 12.3 11.8 1Instruments:C:CAFOS,T:TFOSC 3968.47 5.9 12.8 12.7 5889.953 9.0 11.6 11.6 5895.923 8.3 11.9 12.0 matingtheerrorsisavoided.Asthefeaturesarewideandthedata Day1:Sep.16th,2013;Day2:Sep21st,2013. arenoisy,thecentersofthesesecondaryfitsarefarawayfromthe best fits, 16 kms−1 on average. The difference between the best fitandtheupperlimit(towardsred)fitisdistinctfromthebestfit radioandγ–rayimages(Kunduetal.1980;Katsutaetal.2012).As andthelowerlimit(towardsblue)difference.Toavoidunderesti- thestarcanhardlybeaforegroundsource,wesuggestthat,inthis matingtheerror,thegreaterdifferenceisassignedasthefinalerror. direction,theshockedejectahasnotreachedthedenseinterstellar Theaverageheliocentricvelocitiesofthefirstandthesecondspec- gasyet.Thereadermustnotethat,thebackgroundstarsdisplaying tra are -9.2 and -9.1 kms−1 with line to line scattering standard high velocity CaII lines in Sallmen & Welsh (2004) are located deviationsbeing6.5and5.5 kms−1respectively. behindbrightfilamentknots.Thehighvelocitygasisfoundwhere Aslongasthestarisinsidethesupernovaremnant,highve- thefilamentsareconcentrated(Lozinskaya1976). locity gas accelerated by the SNR is expected to reveal itself as UsingtheabsolutevisualmagnitudefromAlleretal.(1982), blue–shiftedcomponentstoCaII–KandHand/orNaI–D1andD2 for a spectral type B0.5V±0.5, the distance modulus yields absorption features. The spectra show no clear high velocity fea- 1868+410 pc for HD37424. Using the luminosity of the same −403 tures(Figure4).Theaverageheliocentricvelocityoftheinterstel- spectral type from Hohle et al. (2010), we find the distance as largasis12.1±0.5 kms−1(Table5).Thisistypicalfortheneigh- 1318±119 pc which is well consistent with the distance of the boringstars(Silk&Wallerstein1973).Thestarhasnopositional pulsar.Thetotalvisualabsorptionwastakenintoaccountinboth correspondencewithbrightfilaments,butwithfainter Hαemitting calculations. regions.Itislocatedatthevicinityoftheeasterncavityseeninthe Astheabsolutemagnitudesforearlytypestarscanrangefrom (cid:13)c 2014RAS,MNRAS000,1–10 6 Dinçeletal Figure4.ThestronginterstellarlinesofHD37424.Left:CaII–KlinefromTRESspectrum.Despiteaweakblendedfeatureat-13 kms−1,thereisnoclear highvelocitycomponent.Middle:CaII–Hline(H(cid:15)asabackground)fromthesamespectrum.Againthereisnoclearhighvelocitygascomponent.Right: InterstellarNaI–D1andD2lines.ThefeaturewithhighFWHMistheHeI λ5875triplet.ThosearoundSodiumdoubletaretellurics. star to star in the same temperature, interstellar CaII lines are aration dmin and the past time τ at which it occurred. The dis- also used in distance determination. By using the following rela- tributionofseparationsd issupposedtoobeythedistribution min tionfromMegieretal.(2009), ofabsolutedifferencesoftwo3DGaussians(seee.g.Hoogerwerf (cid:32) (cid:33) et al. 2001, equations A3 and A4; Tetzlaff et al. 2012, equations 2.60 1 and 2), if it is assumed that the stellar 3D positions are Gaus- D=77+ 2.78+ EW(H) (1) EW(K) −0.932 siandistributed.Sincetheactual(observed)caseisdifferentfrom EW(H) this simple model (no 3D Gaussian distributed positions, due to the distance to HD37424 is found to be 1288+304 pc which is e.g.theGaussiandistributedparallaxthatgoesintothepositionre- −193 almostinthesamerangeasthepulsar.Theequivalentwidthsare ciprocally,uniformradialvelocitydistributionoftheNS,etc.),we 243±7and160±15mÅforCaII–KandCaII–Hlinesrespectively. adaptthetheoreticalformulae(weuseequations1and2inTetzlaff The extinction towards the star was also derived from 4430 etal.(2012)withthesymbolsµandσfortheexpectationvalueand and4502ÅDIB’s.Themeasuredequivalentwidthshavehigher- standarddeviation,respectively)onlytothefirstpartofthed min ror due to the blending. Using the EW’s, the E(B-V) was calcu- distribution(uptothepeakplusafewmorebins,seeTetzlaffetal. lated based on the relation mentioned in Herbig (1975). It yields 2012).Thederivedparameterµthengivesthepositionaldifference 0.42±0.08 mag; points to somewhat larger color excess but does betweenthetwoobjects.1 notexclude0.35±0.04magderivedfromphotometry. Wefindthatbothstarswereatthesamepositioninthepast i.e.µ=0at(l,b)=(84.82±0.01,27.84±0.01)degat30±4 kpc(Figure5).Thispredictedpositionofthesupernovais4.2+0.8 −0.6 arcmin offset from the nominal geometric center. The predicted 4 KINEMATICS distance of the supernova (as it is seen from the Earth today) is HD37424 and PSRJ0538+2817 have proper motions receding 1333+103 pc. −112 fromeachotheranddepartingfromthesamelocationonthesky (Figure 1). The proper motion of both objects are corrected for Galacticrotationandsolarmotion.TheGalactocentricdistanceto 5 ASSOCIATIONS the Sun is taken as 8.5 kpc and the solar rotational velocity as 220 kms−1. The local standard of rest was taken from Tetzlaff ThereisnoknownOBassociationwithin4.5◦ (∼100 pcat1300 et al. (2011) as (U ,V ,W ) = (10.4±0.4, 11.6±0.2, 6.1±0.2) pc)fromSNRS147.Consideringtheangularseparationsandradial (cid:12) (cid:12) (cid:12) kms−1.TheresultantvaluesforHD37424areµ∗ = 10.0±0.8 distances,theprogenitorstarcannotbelinkedtoanyoftheyoung α masyr−1,µ = −5.9±0.6 masyr−1 andforthepulsar,µ∗ = openclusteraround.Hence,itisconsideredtobearunawaystar δ α −24.4±0.1 masyr−1,µ =57.2±0.1 masyr−1.Togetherwith ejectede.g.fromtheclusterNGC1960,whichis217 pcawayfrom δ the−20.0±6.5 kms−1 peculiarradialvelocity,HD37424hasa theSNR,severalmillionsyearsago(Ngetal.2007).However,bi- spacevelocityof74±8 kms−1at1.3 kpc.HD37424isarunaway narityamongclusterejectedrunawaystarsarerare(Gies&Bolton starofwhichvelocityishigherthantypicalrunawayvelocities40– 1986).So,astheprogenitorwasthepreviousbinarycompanionto 50 kms−1.The2–Dspacevelocityofthepulsaratthesamedis- HD37424,itmaybeunlikelythattheywereejectedfromaclus- tanceis382.2±0.8 kms−1. ter.Therefore,theneighboringstarsareinvestigatedtosearchfor We constructed the past 3D trajectories of PSRJ0538+2817 anyassociation.AlloftheOBtypestarswithin4.5◦(100pcat1.3 andtherunawaystarHD37424toevaluatewhetherthesetwoob- jects could have been at the same place at the same time in the past.Sinceweappliedthesamemethodalreadyinprecedingpa- 1 Theuncertaintiesontheseparationaredominatedbythekinematicun- certainties of the NS that are typically of the order of a few hundred pers (Tetzlaff et al. 2010, 2011, 2012), we refer to these publi- kms−1(becauseoftheassumedradialvelocitydistribution).Asaconse- cations for details. We construct three million past trajectories of PSRJ0538+2817andHD37424throughoutMonteCarlosimula- quence,thedistributionofseparationsdmin showsalargetailforlarger separations. However, the first part of the dmin distribution (slope and tionsbyvaryingtheobservables(parallax,propermotion,radialve- peak)canstillbeexplainedwellwiththetheoreticalcurve(hereequation2 locity)withintheirerrorintervals.FortheradialvelocityoftheNS, inTetzlaffetal.2012)sincethekinematicdispersionforonlythoserunsare weassumeauniformdistributionintherangeof-1500to+1500 muchsmaller,afewtensof kms−1fortheNS,i.e.afewtensofpcafter1 kms−1.Fromallpairsoftrajectories,weevaluatethesmallestsep- Myr. (cid:13)c 2014RAS,MNRAS000,1–10 RunawayinS147 7 Table7.Visualabsolutemagnitudesat1.3±0.1 kpcforthestarsofthepos- sibleOBassociationofwhichHD37424isamember.Errorsaremainlydue tothedistancerange1.2−1.4 kpc.TheexpectedMV forthecorresponding spectraltypesanditsdispersionfromWegner(2006)(W06)aregiven.All valuesareinmag. MV error MV Disp. Name (@1.3±0.1 kpc) (W06) (W06) TYC1869-01317-1 -0.14 +0.26/-0.27 +0.29 1.40 TYC1869-01642-1 -0.28 +0.26/-0.27 +0.29 1.40 HD37424 -2.86 +0.22/-0.23 -3.34 2.40 HD36993 -3.67 +0.22/-0.23 -4.02 2.30 HD37318 -4.46 +0.31/-0.32 -3.34 2.40 BD+27797 -2.97 +0.30/-0.31 -3.34 2.40 HD37696 -3.72 +0.22/-0.23 -3.34 2.40 BD+27850 -2.52 +0.20/-0.21 -2.95 1.16 HD245770 -3.49 +0.29/-0.30 -3.34 2.40 Figure5.Distributionsofminimumseparationsdminandcorresponding HD36441 -3.46 +0.29/-0.30 -2.95 1.16 flighttimesτforencountersbetweenPSRJ0538+2817andHD37424.The BD+26943 -2.36 +0.30/-0.31 -2.64 1.40 solidcurvesdrawninthedminhistogram(bottompanel)representthethe- HD38010 -5.04 +0.26/-0.27 -4.10 2.20 oreticallyexpecteddistributionwithµ = 0andσ = 0.35 pc,adaptedto BD+30938 -2.95 +0.20/-0.21 -2.32 1.50 thefirstpartofthehistogram. Sh2-2421 -2.74 +0.29/-0.30 -3.34 2.40 HD37366 -4.15 +0.22/-0.23 -4.49 2.27 BD+30987 -1.86 +0.24/-0.25 -1.49 2.00 kpcdistance)ofthegeometricalcenter(GC)oftheSNRwerese- BD+30976 -1.18 +0.33/-0.34 -0.63 1.40 lectedfromTheCatalogueofStellarSpectralClassifications(Skiff BD+25989 -2.41 +0.29/-0.30 -2.95 1.16 2013).Thespectraltypesareknownthroughspectroscopicobser- BD+27909 -2.96 +0.26/-0.27 -2.63 2.20 BD+311065 -1.76 +0.20/-0.21 -2.32 1.50 vations.Therearesomediscrepanciesinspectraltypesreportedby BD+311050 -2.30 +0.19/-0.20 -2.32 1.50 different papers. So, always the latest reference was chosen. To- HD38909 -2.98 +0.19/-0.20 -2.32 1.50 getherwith14starsfromourobservationsatTUGandCalarAlto, BD+311021 -1.15 +0.28/-0.29 -0.63 1.40 99 stars are used. The photometric data were obtained from the HD40297 -4.37 +0.29/-0.30 -3.75 1.30 catalogASCC-2.5V3(Kharchenko2001)exceptforUCAC4–589– 020390ofwhichphotometricdatawasretrievedfromtheUCAC4 catalog. magnitudesfromAlleretal.(1982).Errorsindistancesaredueto Foreachstarhavinganintegerspectralsubclass,distanceand theuncertaintyinspectraltypeandtheerrorinA werealsotaken V extinction are calculated in an interval between one spectral sub- intoaccount. classaboveandbelow,e.g.B0V–B2VforaB1Vtypestar.Forthose The spectro–photometric distance alone is not enough due havinghalfintegersubclass,halfspectralsubclassaboveandbelow to the high dispersion in the brightness of OB–type stars (Weg- areusede.g.B0V–B1VforaB0.5Vtypestar.Thetotalvisualex- ner 2006) (Thereafter W06). Assuming the stars beyond 1 kpc tinction,A ,wasdeterminedbyusingBVJHKcolorsandintrinsic are within 100 pc from the SNR GC, the absolute visual magni- V colordifferencesofcorrespondingspectraltypes.Usingthefollow- tudeswerecalculated.AlthoughthedispersionmentionedinW06 ingrelation,A valueswerederivedforeachcolordifference;B-J, is high, 24 of the stars fit well in the comparison with the W06 V B-H,B-K,V-J,V-H,V-K. values.Hence,wesuggestthatthesestarsaremembersofanOB association.(Table6,Table7,19ofthemhavesimilarpmvalues. (λ −λ )−(λ −λ ) AV = A1 /A2 −A1 /A2 0, (2) The average pm in right ascension is -1.39 masyr−1 and -4.17 λ1 V λ2 V masyr−1 in declination with 0.99 masyr−1 and 1.45 masyr−1 TheintrinsiccolorsareobtainedfromKenyon&Hartmann(1995) standarddeviationrespectively(Table8).At1.3 kpc,5ofthemare andWegner(1994),whileAλ/AV ratiosarefromRieke&Lebof- runawaystarsexceeding20 kms−12–Dpeculiarvelocity.Thisis sky(1985).Colordifferencesofshortintervals,i.e.H-Khavehigh consistentwiththegeneralratiooftherunawaystarstothenormal errorsastheyaremultipliedbyhighercoefficients.Thecolorex- starswhichis10–30percent(Gies1987). cessE(B-V)ismentionedseparately.(Table6)A valuesobtained V fromsixcolordifferenceswereaveragedandtheirstandarddevia- tionwerecalculated.Thiswasappliedforallthreepossiblespec- 6 DISCUSSION tral types of the source. As long as the error due to the spectral typeuncertaintyislargerthanthestandarddeviationofthecolor TherunawaynatureofHD37424isclear.Thechanceprojectionof differences,itwasassignedtobethefinalerror.Whentheerroris suchamassiverunawaystarmovingawayfromtheGCofanSNR asymmetric due to the intrinsic colors of different spectral types, intheGalacticanti–centerdirectionmustbeverylow.Inaddition, thelargeronewasaccepted.Insomecases,theuncertaintyisdom- combiningwiththecentralcompactobjectaftertracingbackboth inatedbytheerrorincolordifferences.Then,thesewerepreferred objectsatthesametimeshowsthatHD37424isclearlythepre– tobethefinalerrorforsuchsources.E(B-V)versusA fityieldsa supernova binary companion of the progenitor of SNR S147 and V totaltoselectiveabsorptionratiois3.24±0.06.Inindividualcases, PSRJ0538+2817.BSSisthefavoredexplanationforitsrunaway E(B-V)deviatesstronglyfromA .However,thelargesamplere- nature. V vealsthattheratiooftotaltoselectiveabsorptionhasausualvalue. HD37424 is a B0.5V type star with a mass of ∼13 M (cid:12) Thedistanceswerederivedfromdistancemoduliusingtheabsolute (Hohle et al. 2010). So, the progenitor of the pulsar must have a (cid:13)c 2014RAS,MNRAS000,1–10 8 Dinçeletal Table6.24starsbeyond1 kpcarepresented.Angulardistancesaregivenindegrees,distancesinparsecs,extinctionsandbrightnessinvisualbandin magnitudes.Inthecolumn;SpTadopted,theaveragespectraltypesthatareusedindistancecalibrationsaregiven. Ang.Sep. Name SpT SpT AV AV E(B-V) E(B-V) Distance Distance V Ref#∗ (adopted) (Err) (Err) (Err) 0.09 TYC1869-01317-1 B9.5V B9.5V 0.83 0.1 0.26 0.090 1013 +183/-174 11.258 tw 0.09 TYC1869-01642-1 B9.5V B9.5V 1.00 0.1 0.3 0.090 952 +172/-163 11.293 tw 0.17 HD37424 B0.5V B0.5V 1.28 0.06 0.35 0.036 1868 +410/-403 8.989 tw 0.52 HD36993 B0.5/1III/IV B0.5III-IV 1.35 0.06 0.39 0.028 1961 +636/-628 8.248 1 0.63 HD37318 B0.5Ve B0.5V 2.29 0.15 0.59 0.026 903 +231/-228 8.399 1 1.00 BD+27797 B0.5Ve B0.5V 2.50 0.14 0.62 0.094 1788 +485/-465 10.095 2 1.35 HD37696 B0.5IV/V B0.5IV-V 1.12 0.06 0.29 0.023 1549 +565/-558 7.973 1 1.35 BD+27850 B1.5IVe B1.5V 1.31 0.04 0.35 0.056 1525 +313/-305 9.362 2 1.52 HD245770 O9/B0III/Ve B0III-V 2.11 0.13 0.79 0.053 2595 +1090/-1066 9.187 3 1.76 HD36441 B0.5/1.5V B1V 1.13 0.13 0.33 0.058 1153 +628/-389 8.239 1 1.97 BD+26943 B2V B1.5V 1.41 0.14 0.37 0.040 1357 +723/-511 9.623 8 2.61 HD38010 B1III B1III 1.29 0.1 0.22 0.028 967 +437/-236 6.817 4 2.95 BD+30938 B3III B3III 1.44 0.04 0.45 0.030 1332 +741/-255 9.062 7 2.98 Sh2-2421 B0V B0V 2.17 0.13 0.78 0.056 2318 +853/-842 9.996 5 3.06 HD37366 O9.5V O9.5V 1.22 0.06 0.35 0.016 1375 +201/-199 7.640 6 3.07 BD+30987 B5III: B5III 0.73 0.08 0.25 0.044 1523 +398/-287 9.444 7 3.24 BD+30976 B7V B7V 0.89 0.17 0.18 0.020 994 +264/-226 10.277 7 3.45 BD+25989 B1Vn B1V 1.59 0.13 0.43 0.082 1872 +1061/-649 9.751 8 3.61 BD+27909 B2III B2III 1.87 0.1 0.55 0.076 2004 +683/-761 9.479 8 4.05 BD+311065 B3III B3III 0.67 0.04 0.27 0.043 2304 +1286/-442 9.482 7 4.07 BD+311050 B3III B3III 1.01 0.03 0.35 0.047 1791 +988/-338 9.276 7 4.09 HD38909 B3II-III B3II-III 0.56 0.03 0.17 0.033 2057 +1007/-992 8.145 9 4.28 BD+311021 B7V B7V 1.05 0.12 0.14 0.106 1007 +250/-218 10.465 7 4.46 HD40297 B9.5Ib/II B9.5Ib/II 1.07 0.13 0.25 0.015 1337 +693/-688 7.270 1 ∗Ref#:1:Clausen&Jensen(1979);2:Steeleetal.(1999);3:Wang&Gies(1998);4:Cucchiaroetal.(1976);5:Hunter&Massey (1990);6:Walborn(1971);7:Bouigueetal.(1961);8:Christy(1977);9:Morganetal.(1955);10:Christy(1977);tw:Thiswork. mighthavebeenaninteractingbinary.Hence,theprogenitorstar Table8.19starswithcommonpropermotionarepresented.Theaverage PMRAis-1.39 masyr−1andPMDecis-4.17 masyr−1.Thelastthreecol- shouldbeanakedheliumstaratthefinalstageofitsevolutionwith umnrepresentsthe2–Dspacevelocityofthestarswithrespecttotheaver- a mass even as low as 2 M(cid:12) (van den Heuvel 1993), (Woosley agemotionofthegroupwithmaximumandminimumvaluesin kms−1. etal.1995).However,howconservativethemasstransferwas,will Allthevaluesofpropermotionsarein masyr−1. beunderstoodafterfurtherobservations.Assumingacircularorbit, pre–supernovabinaryparametersarecalculatedfor2,5,10,15,20 and25 M (Table9)progenitormasses. Name µ∗α err µδ err VREL max min (cid:12) HD37318 -0.8 0.6 -5.9 0.6 11.2 16.1 6.9 As discussed in the previous section, the OB stars around TYC1869-01642-1 -1.5 1.1 -2.4 1.2 11.0 19.8 7.1 mightbemembersofanunidentifiedoldOBassociationofwhich HD38010 -1.7 1.0 -3.3 1.0 5.7 14.1 0.0 alloftheOtypestarsunderwentsupernovaexplosions.Thisalso BD+30976 -3.5 0.9 -5.3 0.5 14.7 21.1 8.4 makesaplausibleexplanationforthelowdensitymediuminwhich BD+311021 -2.8 0.7 -5.5 0.8 11.9 18.4 5.4 the SNR expands symmetrically. But, an ejection of the pre-SN TYC1869-01317-1 -0.9 0.9 -5.1 0.9 6.5 14.2 0.0 systemisalsopossible.AmembershiptoanOBassociationorto HD36441 -2.0 0.6 -6.6 0.6 15.4 20.1 11.2 aclusterisimportantalsoregardingthedistancedetermination.In BD+30938 -1.7 0.8 -3.6 0.6 4.0 9.9 0.0 thiswork,thedistancederivedfrompulsarparallax(1.3+0.20 kpc) HD40297 -1.4 1.0 -3.7 1.0 2.9 11.0 0.0 −0.16 isacceptedasthemostreliableestimation.Also,thedistancetothe BD+26943 -0.9 0.8 -3.7 0.8 4.2 11.2 0.0 starmeasuredfrominterstellarlinesisinthesamerange,1288+304 BD+30987 -2.0 0.6 -4.7 1.2 4.9 13.0 0.0 −193 BD+27797 -0.1 0.7 -3.1 1.4 10.4 19.6 4.2 (see section 3). The spectro–photometric distance is much larger BD+311050 -1.0 0.6 -4.0 0.6 2.7 7.8 0.0 byusingabsolutemagnitudesfromAlleretal.(1982).Yet,byus- BD+25989 -1.5 0.9 -2.5 1.6 10.3 21.1 4.9 ing typical luminosities for B0.5V type suggested in Hohle et al. HD36993 -0.2 0.9 -5.8 0.6 12.4 18.8 6.6 (2010),itis1318±119 pc.Hence,thedistancetothestarandthe HD38909 +0.7 0.5 -4.2 1.5 12.9 18.4 0.0 SNR can be assumed to be 1.3 kpc. However, the A measured V BD+311065 -1.0 0.6 -1.3 0.5 17.9 21.7 14.7 directlytowardsS147ismuchlowerthantheA towardsthestars V Sh2-2421 -0.1 0.7 -1.8 1.0 16.7 24.2 9.2 beyond1 kpc.Furthermore,twostars,HD36665andHD37318, HD245770 -2.0 0.5 -4.3 1.1 3.8 10.2 0.0 showhighlyshiftedinterstellarCaII andNaI linesrelatedtothe SNRimplyingthattheseobjectsarebackgroundsources(Sallmen &Welsh2004).Theirdistancesbasedonthereportedspectraltypes highermass.BasedonthelackofOtypestarsinthefield(seeTa- are closer to the Sun than HD37424 is. HD36665 has 837+347 −285 ble6)wesetanuppermasslimitof20–25 M .Itmayevenimply pcforB1VtypeandHD37318is903+231 pcfarawayadopting (cid:12) −228 atwinbinary.TheRocheLoberadiicalculatedfor15,20and25 B0.5V.Ontheotherhand,thedistanceforHD36665isidentified M vary between 91 and 311 R which shows that the system as 1860 pc in Neckel et al. (1980) through H measurements. It (cid:12) (cid:12) β (cid:13)c 2014RAS,MNRAS000,1–10 RunawayinS147 9 Table9.Thebinaryseparations,theorbitalvelocitiesoftheprogenitor,theorbitalperiodsandtheRocheLoberadiiofthepre–supernovabinaryforvarious finalmassesoftheprogenitorwith13 M(cid:12)HD37424aregiven.RocheLoberadiiarecalculatedbasedonEggleton(1983). ProgenitorMass(M(cid:12)) 2 5 10 15 20 25 BinarySeparation(R(cid:12)) 9−+31 49−+191 152−+3276 281−+6489 425−+17051 576−+113071 OrbitalVelocity(kms−1) 481±49 192±20 96±10 64±7 48±5 38±4 OrbitalPeriod(days) 0.85−0.22 9−2 45−11 103−26 176−44 259−65 +0.32 +4 +17 +39 +66 +98 RocheLobeRadius(R(cid:12)) 110−+2179 177−+4321 251−+6404 hasahighA of1.74mag.IfweassumethatHD36665isalsoat ACKNOWLEDGEMENTS V 1.3±0.1 kpc,thenitmustbeabrightB1Vtypestarwhichis1–1.5 ThisworkwasconceivedbythelateOktayH.GUSE˙INOV(1938- magbrighterthantheaveragevaluegiveninW06.HD37318also 2009). The authors acknowledge the leading role he had in all canbeamemberofthepossibleOBassociation. stages of this work and other related topics. His contribution to Thesupernovaeventhadoccurredquitenearbyinafairlyred- GalacticAstrophysicsandhisscholarshipwillbegreatlymissed. dened medium. Assuming a very faint SN (M = −14 mag), the V We thank Janos Schmidt and Christian Ginski for their observa- apparentvisualmagnitudeis−2.1mag,andforabrightSNwith tionalsupportandRonnyErrmannforusefuldiscussions.BD,NT M = −21mag,theapparentvisualmagnitudeis−9.1mag;as V andRNacknowledgesupportfromDFGintheSFB/TR-7Gravita- brightasSN1006.Theeventmighthavealsocausedthe10Bepeak tionalWaveAstronomy.ThisresearchhasmadeuseofAladin.We at35±2 kyrmeasuredinRaisbecketal.(1987)fromthedeepice wouldliketothanktheCalarAltoobservatorystafffortheirhelp coresfromDomeCandVostokAntarctica. inourobservations;andwewouldliketothankDFGinNE515/ 53-1forfinancialsupportfortheCalarAltoobservingrun.Wealso thankTÜB˙ITAKNationalObservatoryfortheirsupportsinusing 7 CONCLUSION RTT–150withprojectnumber12ARTT150-267. HD37424 is a B0.5V type runaway star from a binary ejec- tion due to the supernova which gave birth to SNR S147 and PSRJ0538+2817. The star has −9.2 ± 6.5 kms−1 heliocentric References radial velocity and 74±8 kms−13–D peculiar velocity. The dis- tancecalculatedfrominterstellarCaII linesis1288+304 pcand Allakhverdiev A. O., Guseinov O. H., Tagieva S. O., Yusifov −193 fromspectro–photometry1318±119 pc.Nohighvelocitygasre- I.M.,1997,AstronomyReports,41,257 latedwiththeSNRisdetectedinthespectra.Thepasttrajectories AllerL.H.etal.,eds,1982,Landolt-Börnstein:NumericalData of the pulsar and HD37424 are reconstructed throughout Monte andFunctionalRelationshipsinScienceandTechnology-New Carlosimulations.Itisfoundthatbothstarswereatthesamepo- Series”Gruppe/Group6AstronomyandAstrophysics”Volume sitionat(l,b) =(84.82±0.01,27.84±0.01)degat30±4 kyr 2 Schaifers/Voigt: Astronomy and Astrophysics / Astronomie inthepast.Thepositionoftheexplosionis4.2+0.8 arcminaway und Astrophysik ” Stars and Star Clusters / Sterne und Stern- −0.6 fromthegeometricalcenter.ThedistanceoftheSNRisfoundas haufen 1333+103 pc.A towardstheSNRis1.28±0.06mag. AndersonS.B.,CadwellB.J.,JacobyB.A.,WolszczanA.,Foster −112 V Today’s kinematics of the stars are well known, further de- R.S.,KramerM.,1996,ApJ,468,L55 tailedcalculationsshouldbedonetofindthetruekickvectorsof AnkayA.,KaperL.,deBruijneJ.H.J.,DewiJ.,HoogerwerfR., thepulsarandapossiblespin-kickalignment. SavonijeG.J.,2001,A&A,370,170 ThereisnoknownOBassociationreportedclosetotheSNR. BlaauwA.,1961,Bull.ofAstro.Inst.oftheNetherlands,15,265 19OBtypestarswithin100pcfromtheSNRgeometricalcenter BobylevV.V.,2008,AstronomyLetters,34,686 haveverysimilar2–Dvelocitieswiththe5runawaystarsincluding Bouigue R., Boulon J., Pedoussaut A., 1961, Annales de HD37424.ThismightbeanoldOBassociationhavingnobrightO l’ObservatoireAstron.etMeteo.deToulouse,28,33 typestars.ThelowdensitymediuminwhichtheSNRisexpanding ChatterjeeS.etal.,2009,ApJ,698,250 isprobablyduetothepreviousSNeandHIIregionsdrivenbythis ChevalierR.A.,1974,ApJ,188,501 oldassociation.Theprogenitorwasamassivestarwithazeroage ChristyJ.W.,1977,ApJ,217,127 mainsequencemassgreaterthan13 M .Consideringthelackof ClarkD.H.,CaswellJ.L.,1976,MNRAS,174,267 (cid:12) O type stars in the field, a progenitor mass much larger than 20 Clausen J. V., Jensen K. S., 1979, in McCarthy M. F., Philip M isnotexpected.Assumingacircularorbit,thepre–supernova A.G.D.,CoyneG.V.,eds,RicercheAstronomicheVol.9,IAU (cid:12) binaryseparationisintherangeof8to711 R for2to25 M Colloq.47:SpectralClassificationoftheFuture.p.479 (cid:12) (cid:12) finalmassoftheprogenitor.ThecorrespondingRocheLoberadii CordesJ.M.,LazioT.J.W.,2002,ArXivAstrophysicse-prints for 15 to 25 M masses vary from 91 to 311 R . It must have Cucchiaro A., Jaschek M., Jaschek C., Macau-Hercot D., 1976, (cid:12) (cid:12) beenaninteractingbinary. A&AS,26,241 For 1.3 kpc distance and 1.3 mag extinction, the SN which EggletonP.P.,1983,ApJ,268,368 happened 30±4 kyr ago had an apparent brightness of −2.1 to FeastM.W.,ShuttleworthM.,1965,MNRAS,130,245 −9.1mag. FesenR.A.,BlairW.P.,KirshnerR.P.,1985,ApJ,292,29 Thesourcewillbeinvestigatedregardingtheelementalsigna- FuerstE.,ReichW.,1986,A&A,163,185 turesofbinaryaccretionandthepossiblesupernovadebrisonits GaenslerB.M.,StappersB.W.,FrailD.A.,MoffettD.A.,John- photospherethroughhighresolutionandhighS/Nspectra. stonS.,ChatterjeeS.,2000,MNRAS,318,58 (cid:13)c 2014RAS,MNRAS000,1–10 10 Dinçeletal GiesD.R.,1987,ApJS,64,545 SauvageotJ.L.,BalletJ.,RothenflugR.,1990,A&A,227,183 GiesD.R.,BoltonC.T.,1986,ApJS,61,419 SayerR.W.,NiceD.J.,KaspiV.M.,1996,ApJ,461,357 GreenD.A.,2009,VizieROnlineDataCatalog,7253,0 SilkJ.,WallersteinG.,1973,ApJ,181,799 GuseinovO.H.,AnkayA.,SezerA.,TagievaS.O.,2003,Astro- SkiffB.A.,2013,VizieROnlineDataCatalog,1,2023 nomicalandAstrophysicalTransactions,22,273 SofueY.,FurstE.,HirthW.,1980,PASJ,32,1 GuseinovO.H.,AnkayA.,TagievaS.O.,2003,SerbianAstro- SteeleI.A.,NegueruelaI.,ClarkJ.S.,1999,A&AS,137,147 nomicalJournal,167,93 TaurisT.M.,TakensR.J.,1998,A&A,330,1047 GuseinovO.H.,AnkayA.,TagievaS.O.,2004a,SerbianAstro- TetzlaffN.,DinçelB.,NeuhäuserR.,KovtyukhV.V.,2014,MN- nomicalJournal,168,55 RAS,438,3587 GuseinovO.H.,AnkayA.,TagievaS.O.,2004b,SerbianAstro- TetzlaffN.,EisenbeissT.,NeuhäuserR.,HohleM.M.,2011,MN- nomicalJournal,169,65 RAS,417,617 Guseinov O. H., Ankay A., Tagieva S. O., 2004c, International TetzlaffN.,NeuhäuserR.,HohleM.M.,2011,MNRAS,410,190 JournalofModernPhysicsD,13,1805 TetzlaffN.,NeuhäuserR.,HohleM.M.,MaciejewskiG.,2010, GuseinovO.H.,AnkayA.,TagievaS.O.,2005,Astrophysics,48, MNRAS,402,2369 330 TetzlaffN.,SchmidtJ.G.,HohleM.M.,NeuhäuserR.,2012,29, HansenB.M.S.,PhinneyE.S.,1997,MNRAS,291,569 98 HerbigG.H.,1975,ApJ,196,129 Tetzlaff N., Torres G., Neuhäuser R., Hohle M. M., 2013, MN- HobbsG.,LorimerD.R.,LyneA.G.,KramerM.,2005,MNRAS, RAS,435,879 360,974 vandenBerghS.,1980,JournalofAstrophysicsandAstronomy, HohleM.M.,NeuhäuserR.,SchutzB.F.,2010,Astronomische 1,67 Nachrichten,331,349 vandenHeuvelE.P.J.,1993,66,309 HoogerwerfR.,deBruijneJ.H.J.,deZeeuwP.T.,2001,A&A, WalbornN.R.,1971,ApJS,23,257 365,49 WalbornN.R.,FitzpatrickE.L.,1990,PASP,102,379 HunterD.A.,MasseyP.,1990,AJ,99,846 WangZ.X.,GiesD.R.,1998,PASP,110,1310 KaperL.,vanLoonJ.T.,AugusteijnT.,GoudfrooijP.,PatatF., WegnerW.,1994,MNRAS,270,229 WatersL.B.F.M.,ZijlstraA.A.,1997,ApJ,475,L37 WegnerW.,2006,MNRAS,371,185 KatsutaJ.etal.,2012,ApJ,752,135 Wongwathanarat A., Janka H.-T., Müller E., 2013, A&A, 552, KenyonS.J.,HartmannL.,1995,ApJS,101,117 A126 KharchenkoN.V.,2001,KinematikaiFizikaNebesnykhTel,17, WoosleyS.E.,LangerN.,WeaverT.A.,1995,ApJ,448,315 409 ZachariasN.,FinchC.T.,GirardT.M.,HendenA.,BartlettJ.L., KirshnerR.P.,ArnoldC.N.,1979,ApJ,229,147 MonetD.G.,ZachariasM.I.,2012,VizieROnlineDataCatalog, KramerM.,LyneA.G.,HobbsG.,LöhmerO.,CarrP.,JordanC., 1322,0 WolszczanA.,2003,ApJ,593,L31 ZavlinV.E.,PavlovG.G.,2004,75,458 KunduM.R.,AngerhoferP.E.,FuerstE.,HirthW.,1980,A&A, 92,225 ThispaperhasbeentypesetfromaTEX/LATEXfilepreparedbythe LozinskayaT.A.,1976,SvA,20,19 author. LyneA.G.,LorimerD.R.,1994,Nat,369,127 McGowanK.E.,KenneaJ.A.,ZaneS.,CórdovaF.A.,Cropper M.,HoC.,SasseenT.,VestrandW.T.,2003,ApJ,591,380 Megier A., Strobel A., Galazutdinov G. A., Krełowski J., 2009, A&A,507,833 MilneD.K.,1979,AustralianJournalofPhysics,32,83 MinkowskiR.,1958,ReviewsofModernPhysics,30,1048 MorganW.W.,CodeA.D.,WhitfordA.E.,1955,ApJS,2,41 Mugrauer M., Avila G., Guirao C., 2014, Astronomische Nachrichten,335,417 Mugrauer M., Berthold T., 2010, Astronomische Nachrichten, 331,449 NeckelT.,KlareG.,SarcanderM.,1980,Bulletind’Information duCentredeDonneesStellaires,19,61 NgC.-Y.,RomaniR.W.,BriskenW.F.,ChatterjeeS.,KramerM., 2007,ApJ,654,487 PhilpC.J.,EvansC.R.,LeonardP.J.T.,FrailD.A.,1996,AJ, 111,1220 PovedaA.,RuizJ.,AllenC.,1967,BoletindelosObservatorios TonantzintlayTacubaya,4,86 PrzybillaN.,NievaM.F.,HeberU.,ButlerK.,2008,ApJ,684, L103 RaisbeckG.M.,YiouF.,BourlesD.,LoriusC.,JouzelJ.,1987, Nat,326,273 RiekeG.H.,LebofskyM.J.,1985,ApJ,288,618 SallmenS.,WelshB.Y.,2004,A&A,426,555 (cid:13)c 2014RAS,MNRAS000,1–10