The VLT-FLAMES Tarantula Survey XVIII. Classifications and radial velocities of the B-type stars

1.8 MB·English

by  C. J. Evans

Astronomy&Astrophysicsmanuscriptno.24414a (cid:13)c ESO2015
January20,2015
The VLT-FLAMES Tarantula Survey
XVIII. Classifications and radial velocities of the B-type stars
C.J.Evans1,M.B.Kennedy2,P.L.Dufton2,I.D.Howarth3;N.R.Walborn4,N.Markova5,J.S.Clark6,
S.E.deMink7,8,9(cid:63),A.deKoter7,10,P.R.Dunstall2,V.He´nault-Brunet11,J.Ma´ızApella´niz12(cid:63)(cid:63),
C.M.McEvoy2,H.Sana13,S.Simo´n-D´ıaz14,15,W.D.Taylor1,J.S,Vink16
1 UKAstronomyTechnologyCentre,RoyalObservatory,BlackfordHill,Edinburgh,EH93HJ,UK
2 DepartmentofPhysics&Astronomy,Queen’sUniversityBelfast,BelfastBT71NN,NorthernIreland,UK
3 Dept.ofPhysics&Astronomy,UniversityCollegeLondon,GowerStreet,London,WC1E6BT,UK
5 4 SpaceTelescopeScienceInstitute,3700SanMartinDrive,Baltimore,MD21218,USA
1 5 InstituteofAstronomywithNAO,BulgarianAcademyofSciences,POBox136,4700Smoljan,Bulgaria
0 6 DepartmentofPhysics&Astronomy,TheOpenUniversity,WaltonHall,MiltonKeynes,MK76AA,UK
2 7 AstronomicalInstituteAntonPannekoek,AmsterdamUniversity,SciencePark904,1098XH,Amsterdam,TheNetherlands
8 ObservatoriesoftheCarnegieInstitutionforScience,813SantaBarbaraSt.,Pasadena,CA91101,USA
n 9 CahillCenterforAstrophysics,CaliforniaInstituteofTechnology,Pasadena,CA91125,USA
a 10 InstitutevoorSterrenkunde,KULeuven,Celestijnenlaan200D,3001,Leuven,Belgium
J
11 DepartmentofPhysics,FacultyofEngineering&PhysicalSciencesUniversityofSurreyGuildford,GU27XH,UK

6 12 InstitutodeAstrof´ısicadeAndaluc´ıa-CSIC,GlorietadelaAstronom´ıas/n,E-18008Granada,Spain
1 13 ESA/STScI,3700SanMartinDrive,Baltimore,MD21218,USA
14 InstitutodeAstrof´ısicadeCanarias,E-38200LaLaguna,Tenerife,Spain

] 15 DepartamentodeAstrof´ısica,UniversidaddeLaLaguna,E-38205LaLaguna,Tenerife,Spain
R
16 ArmaghObservatory,CollegeHill,Armagh,BT619DG,UK
S
.
h
ABSTRACT
p
-
We present spectral classifications for 438 B-type stars observed as part of the VLT-FLAMES Tarantula Survey (VFTS) in the
o
30DoradusregionoftheLargeMagellanicCloud.Radialvelocitiesareprovidedfor307apparentlysinglestars,andfor99targets
r
t withradial-velocityvariationswhichareconsistentwiththembeingspectroscopicbinaries.Weinvestigatethespatialdistributionof
s
theradialvelocitiesacrossthe30Dorregion,andusetheresultstoidentifycandidaterunawaystars.Excludingpotentialrunaways
a
[ andmembersoftwoolderclustersinthesurveyregion(SL639andHodge301),wedetermineasystemicvelocityfor30Dorof
271.6±12.2kms−1from273presumedsinglestars.Employinga3-σcriterionweidentifyninecandidaterunawaystars(2.9%ofthe

1 singlestarswithradial-velocityestimates).Theprojectedrotationalvelocitiesofthecandidaterunawaysappeartobesignificantlydif-
v ferenttothoseofthefullB-typesample,withastrongpreferenceforeitherlarge(≥345kms−1)orsmall(≤65kms−1)rotationalve-
8 locities.Ofthecandidaterunaways,VFTS358(classifiedB0.5:V)hasthelargestdifferentialradialvelocity(−106.9±16.2kms−1),
0 andapreliminaryatmosphericanalysisfindsasignificantlyenrichednitrogenabundanceof12+log(N/H)(cid:38)8.5.Combinedwitha
1 largerotationalvelocity(v sini=345±22kms−1),thisissuggestiveofpastbinaryinteractionforthisstar.
e
4
Keywords.openclustersandassociations:individual:NGC2060,NGC2070,Hodge301,SL639–stars:early-type–stars:funda-
0
mentalparameters
.
1
0
5
1. Introduction cal information on the high-mass population of 30 Dor, which
1
: will underpin quantitative analyses of the spectra in future pa-
v The VLT-FLAMES Tarantula Survey (VFTS) is a European
pers. Detailed classifications for the 352 O-type spectra (with
i Southern Observatory (ESO) Large Programme that has ob-
X discussionofnotablemorphologicalgroups)werepresentedby
tained multi-epoch optical spectroscopy of over 800 massive
Walbornetal.(2014),whileclassificationsforthesmallersam-
r stars in the 30 Doradus region of the Large Magellanic Cloud
a ples of Wolf–Rayet and late-type stars were given in Paper I.
(LMC).Anoverviewofthesurvey,includingdetailsofthestel-
Here we present spectral classifications for 438 B-type objects
lar photometry, spectroscopic observations and data reduction,
from the survey. Of these, only 74 (17%) have previously pub-
wasgivenbyEvansetal.(2011;hereafterPaperI).
lishedclassificationsbasedondigitaldata,manyofwhichwere
TheVFTSobservationsarethefirstopticalspectroscopyfor obtainedatalowerspectralresolvingpower.
many of the targets, and provide a wealth of new morphologi-
To complement the spectral classifications, we also present
Sendoffprintrequeststo:[email protected] estimates of radial velocities for the B-type sample. These es-
(cid:63) EinsteinFellow timates provide insights into the different populations within
(cid:63)(cid:63) Current address: Departamento de Astrof´ısica, Centro de 30Dor,andenableidentificationofcandidaterunawaystarsthat
Astrobiolog´ıa (INTA-CSIC), Campus ESA, Apartado Postal 78, have been ejected from their birth sites. In the context of run-
28691VillanuevadelaCan˜ada,Madrid,Spain awaystars,thesizeoftheVFTSsampleisunprecedentedfora
1
C.J.Evansetal.:VFTS:Classifications&radialvelocitiesfortheB-typestars
single,youngstar-formingregion.Thesurveyhasrevealed∼20 For stars that appear to be single (Dunstall et al., in prep.)
O-type runaway candidates (Evans et al. 2010, Sana et al., in theLR02andLR03spectrawerenormalisedthenmerged,with
prep.),withanapparentconnectionwithrapidrotation(Walborn the σ-clip levels for cosmic rejection and relative weightings
etal.2014,Sanaetal.,inprep.).Therearealsoanumberofstars basedonthesignal-to-noise(S/N)ratiosoftheindividualspec-
withmoderate(projected)rotationalvelocitiesintheperipheral tra.Finally,forclassification,themergedLR02andLR03spec-
regions of 30 Dor, which could perhaps be very rapid rotators tra were combined, then smoothed and rebinned to R=4000.
(but with low inclinations; see Walborn et al. 2014, for a dis- Forstarsidentifiedashavingsignificantbutrelativelymodestve-
cussion of their spectra). For completeness, we note that two locityvariations(∆v ≤40kms−1),theindividualspectrawere
r
additional O2-type stars near 30 Dor, Sk−68◦137 and BI253 similarly co-added and degraded (as the variations were within
(=VFTS072), were suggested as runaways by Walborn et al. theeffectivevelocityresolutionoftherebinneddata).Forsingle-
(2002a)onthebasisoftheirremotelocationsfortheirapparent lined binaries with ∆v >40kms−1, the single-epoch LR02
r
youth. and LR03 spectra with the best S/N ratio were combined and
In his pioneering study, Blaauw (1961) noted that the frac- then smoothed/rebinned. We were careful in these cases to en-
tional incidence of Galactic runaways is larger among O stars surethatanyvelocityoffsetsbetweenfeaturesinthesmallover-
thanamongBstars,aresultsubstantiatedbysubsequentauthors, lapregionbetweentheLR02andLR03datawerenotmisinter-
andwhichalsoemergedinthesimulationsfromPortegiesZwart preted. The small number of double-lined binaries were classi-
(2000).Inthispaperwethereforecomplementthestudiesofthe fiedfrominspectionoftheindividualspectra.
O-typerunawaysintheVFTSwithanalysisoftheB-typestars.
Inparticular,thenumberandmassspectrumoftheejectedrun-
awaysfrom30Dorwillprovideimportantconstraintsonmodels 3. Spectral classification
ofclusterformationandevolution(e.g.Fujii&PortegiesZwart
2011).Theestimatedradialvelocitiespresentedhere,combined The spectra were classified by visual comparison to the B-type
withthosefromtheO-typestars(Sanaetal.2013),willalsopro- Galactic standards from Sota et al. (2011) and Sana et al. (in
videanimportantinputforcomparisonswiththeoreticalpredic- prep.),whilealsotakingintoaccounttheeffectsofthereduced
tions,e.g.inthecontextofsimulationstoinvestigatetheinitial metallicityoftheLMCcomparedtotheGalaxy(whichcanaf-
conditions leading to the distinct clump identified to the north- fect the appearance of both the metal and helium absorption
eastofR136(Sabbietal.2012). lines; e.g. Markova et al. 2009). This was achieved for the lu-
InSection2wesummarisetherelevantfeaturesoftheVFTS minous,low-gravityobjectsbyinterpolatingbetweentheavail-
dataandtheirpost-processing.Section3setsouttheframework able Galactic standards and those for supergiants in the Small
within which the spectra were classified and discusses notable Magellanic Cloud (Lennon 1997), and with reference to the
objects.Section4detailsthemethodsweusedtoestimateradial frameworkdevelopedfortheLMCbyFitzpatrick(1988,1991).
velocitiesforeachtarget,followedbyadiscussionofmembers Classifications of the dwarfs and giants were also informed by
of the two oldest clusters in the survey region (Hodge301 and comparisons with the standards from Walborn & Fitzpatrick
SL639)inSection5.Finally,inSection6,weuseourvelocity (1990)andtheLMCstarsclassifiedbyEvansetal.(2006).
estimates to identify candidate single-star and binary runaways Theprimarydiagnosticlinesaretheionisationratiosofsil-
anddiscusstheirnatureandpotentialorigins.Inparalleltothis icon, while also taking into account the appearance of the he-
study, Dunstall et al. (in prep.) have exploited the multi-epoch lium and magnesium (λ4481) lines. For reference, the primary
VFTS observations to investigate the multiplicity properties of temperature-sequencecriteriausedtoclassifytheVFTSsample
theB-typesample. are summarised in Tables 1 and 2 (for supergiants and dwarfs,
respectively). Example temperature sequences for supergiants,
giants and dwarfs are shown in Figures A.1, A.2 and A.3, re-
2. Observations and data processing
spectively.Atagivenspectraltype,luminosityclasseswereas-
All of the VFTS data discussed here were obtained using signed from the width of the Balmer lines and (at the earliest
the Medusa–Giraffe mode of the Fibre Large Array Multi- types) from the intensity of the silicon absorption lines, while
ElementSpectrograph(FLAMES)instrumentontheVeryLarge alsotakingintoaccountthepossibleeffectsofrotationalbroad-
Telescope(VLT).TheMedusafibres(whichsubtend1(cid:48).(cid:48)2onthe ening on the spectra. Example luminosity sequences for B0.2,
sky) were used to relay light from up to 130 targets simultane- B1,andB2.5typesareshowninFiguresA.4,A.5,andA.6,re-
ouslytotheGiraffespectrograph(seePasquinietal.2002). spectively.
Fulldetailsoftheobservationalstrategyandreductionofthe Thesupergiantsandbrightgiants(i.e.classIandIIobjects)
data are given in Paper I. In brief, the ESO Common Pipeline haverichabsorption-linespectra.Giventheimportanceofspec-
Library FLAMES recipes were used to undertake bias sub- tralpeculiaritiesinthecontextofCNOabundances,andthein-
traction, fibre location, summed extractions of each object, di- fluenceofmetallicityeffects,theseluminousobjectswereclassi-
vision by a normalised flat-field, and wavelength calibration. fiedindependentlybythreeoftheauthors(CJE,NM,andNRW)
Subsequentprocessingincludedcorrectionofthespectratothe withatypicalconsistencyof0.5subtypes/luminosityclassfrom
heliocentric frame, sky subtraction, and rejection of significant the first pass; final types were adopted following discussion of
cosmicrays. individualobjectsasrequired.Thelargersampleofdwarfsand
TheMedusaobservationsusedthreeofthestandardGiraffe giants,whichexhibitedaconsiderablerangeindataquality,was
settings (LR02, LR03, and HR15N), providing spectral cover- classifiedbyCJE.Inmanyinstancestheseclassificationsareless
ageofλ3960-5071A˚ (atresolvingpowers,R,of7000to8500), precise than those possible for the luminous objects, but they
andλ6442-6817A˚ (atR=16000).Theseresolutionsaregreater should provide sufficient morphological information to enable
thanformostpublishedstandardsso,forthepurposesofclassifi- robustsubdivisionsofthesampleinfutureanalyses.
cation,thedataweredegradedtoaneffectiveresolvingpowerof Each merged spectrum was classified in the framework de-
R=4000,ensuringconsistencywiththeapproachtotheclassi- scribedabove.Afterinitialclassificationofallthespectra,care-
ficationoftheO-typestarsbyWalbornetal.(2014). ful checks were undertaken to ensure self-consistency within
2
C.J.Evansetal.:VFTS:Classifications&radialvelocitiesfortheB-typestars
Table1.Primarytemperature-sequencecriteriaforB-typesupergiants. 3.1. Emission-lineobjects
Whereappropriate,spectrawereinitiallyclassifiedas‘Be+’on
Type Criteria
the basis of Fe II emission lines in the LR02 and LR03 spectra
B0 SiIVλλ4089,4116>>SiIIIλ4553>HeIIλ4542
B0.2 SiIVλ4089>SiIIIλ4553>>HeIIλ4542 (see discussion by Evans et al. 2006); the Fe II lines were pre-
B0.5 SiIVλ4089>SiIIIλ4553;HeIIλ4542absent/marginal ferredovertheblue-violetBalmerlinesowingtothesignificant
B0.7 SiIVλ4089∼SiIIIλ4553;HeIIλ4686absent/marginal (andvariable)contaminationofthelatterbynebularemission.
B1 MarginalSiIVλ4116;SiIIIλ4553>SiIVλ4089 Independently of the blue-region classifications, we in-
B1.5 Weak/marginalSiIVλ4089,weakMgIIλ4481 spectedtheHαprofilesofeverytargettoinvestigatetheirmor-
B2 SiIIIλ4553>MgIIλ4481;SiIVλ4089absent phologies.ThegreaterresolutionoftheHαobservationsandthe
B2.5 SiIIIλ4553∼MgIIλ4481;weakSiIIλλ4128-32 extendedwingsfromdiscemissionmadeiteasiertodistinguish
B3 SiIIIλ4553<MgIIλ4481;SiIIλλ4128-32present contributions from Be-type Balmer emission and nebular con-
B5 HeIλ4471>MgIIλ4481;SiIIλλ4128-32∼HeIλ4121
taminationthanintheblue,buttheseclassificationsshouldstill
B8 HeIλ4471∼MgIIλ4481;SiIIλλ4128-32∼HeIλ4144
betreatedwithsomecaution.TheHαmorphologiesareencoded
B9 HeIλ4471<MgIIλ4481;SiIIλλ4128-32>HeIλ4144
in the fourth column of Table 7 (with the notation explained in
Notes.ExamplespectraareshowninFigureA.1. thecorrespondingfootnote).
Objects classified as ‘Be+’ from the LR02/LR03 spectra
were seen to display Hα emission in all cases. Approximately
30 additional stars display Be-like Hα emission without Fe II
Table2.Primarytemperature-sequencecriteriaforB-typedwarfs.
emission.TheHαemissioninluminoussupergiantsmostlikely
arisesfromtheirstellarwinds,butforlessluminousobjectssuch
Type Criteria
emission is the primary Be diagnostic. To preserve the distinc-
B0 SiIVλ4089>SiIIIλ4553>HeIIλ4542
tionofthelinesseeninemission,thestandard‘e’suffixisused
B0.2 SiIVλ4089>SiIIIλ4553;HeIIλ4542marginal
in the classifications for the (non-supergiant) stars where only
B0.5 HeIIλ4542absent;SiIIIλ4116marginal/absent
B0.7 HeIIλ4686marginal/absent;SiIIIλ4116marginal/absent Hαisseeninemission.
B1 SiIIIλ4553>SiIVλ4089;SiIVλ4116,HeIIλ4686absent The complex star-formation history of 30 Dor means that
B1.5 SiIVλ4089absent;SiIIIλ4553∼MgIIλ4481 oursamplewillspanarangeofstellarages,soweareunableto
B2 SiIIIλ4553<MgIIλ4481 comment on the incidence of the Be-phenomenon for a single,
B2.5 SiIIIλ4553absent;SiIIλλ4128-32marginal co-eval population. Nonetheless, we note that the incidence of
the Be-type classifications for the dwarf and giant stars is 18%
Notes.ExamplespectraareshowninFigureA.3.
(70/388 objects). This is in good agreement with the Galactic
fraction(≥17%,Zorec&Briot1997),andthatreportedforthe
fieldpopulationnearNGC2004intheLMC(of16and17.5%,
fromEvansetal.2006andMartayanetal.2006,respectively).
the sample. An overlapping sample of ten B0-type stars was
classified independently (by NRW) for the study by Walborn 3.1.1. B[e]-typespectra
etal.(2014),withexcellentagreement(towithinoneluminosity
TwotargetsintheMedusa–Giraffesampleshowforbidden-line
class)inallinstances.
emission characteristic of the B[e] category. VFTS698 is a
Thecatalogueofclassificationsandradial-velocityestimates double-linedbinarysystemwithaB[e]-likespectrum,compris-
for the B-type objects from the VFTS is presented in Table 7 ingwhatappearstobeanearlyB-typesecondaryinorbitaround
(availableonline).Forconvenience,supplementaryinformation a veiled, more massive companion (see Dunstall et al. 2012).
isalsoprovidedforeachstaronitsbinarystatus(fromDunstall ThesecondB[e]objectisVFTS822,showninFigure2together
etal.,inprep.),Hα morphology(seeSection3.1),opticalpho- with VFTS1003 (the B[e]-like object from Paper I). Given the
tometry(fromPaperI),alternativeidentificationsusedinthelit- intensityofitsSiIIλλ4128-32andMgIIλ4481absorption,the
erature,andpreviousspectralclassifications. spectrumofVFTS822wasclassifiedasmid-lateB[e](∼B5-8);
The S/N ratios of the supergiant spectra were particularly furtherdiscussionofitspossiblepre-main-sequencenaturewas
good (typically >100 for both the LR02 and LR03 regions; givenbyKalarietal.(2014).
see Figure A.1). This allowed us to provide consistent indi- Although the nature of VFTS1003 remains uncertain at
cations of the nitrogen enhancement/deficiency compared to present, the number of evolved B[e] stars (at most two) in the
morphologically-normal stars (via Nstr and Nwk qualifiers), as VFTSisclearlysmallcomparedtothenumberofnormalsuper-
diagnosed from the CNO absorption features in the λλ4640- giants.ThisisinkeepingwiththesmallnumberofsgB[e]stars
4650 region (Walborn 1976). The significant differences in inGalacticclusters(seetheappendixofClarketal.2013),argu-
CNO morphology are illustrated by the two B1.5 Ia spectra in ingthattheB[e]phenomenoninevolvedsupergiantsmustbea
Figure1.Atmosphericanalysisofthesupergiants,includinges- relativelybriefphaseororiginatefromarareprocess.
timatesofnitrogenabundancesandathoroughstudyoftheline-
broadeningintheirspectra,willbepresentedbyMcEvoyetal.
3.1.2. VFTS766:Apeculiaremission-linestar
(2015)andSimo´n-D´ıazetal.(inprep.),respectively.
Giventhewiderangeinqualityofthedwarfandgiantspec- VFTS766 displays strong Hα emission and inspection of its
tra,wedonotcommentonCNOmorphologiesintheirclassifi- blue-regionspectrum(seeFigure2)revealeditasapeculiarob-
cations. Equally, while we employ the ‘n’ and ‘(n)’ indications ject.Atfirstglanceitsspectrumappearstobethatofamid-late
ofbroadening(seee.g.Table3fromSotaetal.2011)atalllumi- B-type supergiant (B5-B8, as traced by e.g. Si II, Mg II), but
nosityclasses,therearelikelytobesomelacunae(cf.thequan- with superimposed metallic absorption lines, which are consis-
titativev siniestimatesfromDuftonetal.2013). tentwithacoolercomponent.Thisseeminglycompositeappear-
e
3
C.J.Evansetal.:VFTS:Classifications&radialvelocitiesfortheB-typestars
Fig.1. Differing CNO morphologies in early B-type supergiants illustrated by VFTS431 (‘Nstr’) and VFTS578 (‘Nwk’). In addition to the
Balmerlines,theabsorptionfeaturesidentifiedare:CIIλ4267;HeIλλ4009,4026,4121,4144,4388,4471,4713;MgIIλ4481;NIIλλ3995,
4601-07-14-21-31,4640-43(blendedwithOII);OIIλλ4070,4254,4317-19,4350,4367,4414-17,4591-96,4650,4661,4674-76;SiIIIλλ4553-
68-74.
ancecanindicateacircumstellarshellordiscobservededge-on, supergiantphaseandareabouttoenteraluminousbluevariable
and this hypothesis is supported by the twin-peaked Hβ shell- (LBV)phase(Clarketal.2012;Grohetal.2014).Thissuggests
like emission profile. Its near- and mid-IR colours1 place it in that VFTS003, 430, and 533 might be on the verge of LBV-
the region in the J,J −[3.6] colour-magnitude diagram which likebehaviour,althoughwenotethatVFTS430appearssome-
isoccupiedbyearly-typestarsassociatedwithfree-free/bound- what subluminous compared to the blue supergiants in 30 Dor,
freeemission,andwefindnoevidenceofhotdust(cf.Bonanos andissignificantlyfainterthanVFTS533(evenaccountingfor
etal.2009). its apparently greater line-of-sight extinction, see photometry
Similar emission profiles are observed in some low- in Table 7). Later-type BHGs have been observed to undergo
luminosity supergiant B[e] stars (e.g. S59, Gummersbach et al. LBV-like photometric and spectroscopic behaviour (e.g. Smith
1995),withexamplesthatalsoappeartolackdustydiscs(Graus et al. 2004; Clark et al. 2012, with potential implications for
etal.2012,althoughnonehaveyetbeendiscoveredintheedge- VFTS424 and 458 here), suggesting that these two evolution-
on orientation required to yield shell profiles). However, we aryphasesareessentiallysynonymousforsuchstars.
speculatethatVFTS766isaclassicalBestar,inwhichthetwin-
peakedemissioninthewingsoftheBalmerprofilesgivesriseto
Onfurtherinspectionoftheirspectra,broademissionwings
thesupergiant-likeappearance.Anexampleofthiseffectisthe
wereidentifiedintheHβprofilesofeachoftheBHGs,exceptfor
developmentofacircumstellardiscintheclassicaledge-onBe
VFTS430.ThisHβmorphologyisseeninsomeLBVsandother
staroAnd,whichconspirestoyieldanarrowerabsorptionfea-
luminoussupergiants(e.g.Walborn&Fitzpatrick2000),andis
turethantheunderlyingphotosphericprofile(Clarketal.2003).
alsopresentinVFTS739inoursample(classifiedasA0Ipbut,
AnotherexampleisprovidedbytheBestar48Lib(seeRivinius
given its omission from Paper I, included here in the B-type
et al. 2013, and references therein), which has previously been
sample). The spectra and evolutionary status of these potential
classifiedasasupergiant(Houk&Smith-Moore1988).
LBV precursors/candidates will be discussed in more depth by
ThispossibleexplanationforthenatureofVFTS766issup-
Walbornetal.(inprep.).
ported by its magnitude relative to other stars in our sample.
Spectroscopicmonitoringtoidentifyanepisodeofdiscloss,dur-
ing which the (uncontaminated) photospheric Balmer line pro-
files might be observed, would clarify its luminosity type and
trueevolutionarystatus. 3.1.4. ComparisonswithX-Rayobservations
3.1.3. Bluehypergiants
Wecomparedoursamplewiththepoint-sourcedetectionsfrom
We identifed five blue hypergiants (BHGs) in our sample, observations in 30 Dor with the Chandra X-ray Observatory
namely: VFTS003 (B1 Ia+), 424 (B9 Ia+p), 430 (B0.5 (Townsley et al. 2006, 2014); none of the VFTS B-type stars
Ia+((n))Nwk),458(B5Ia+p)and533(B1.5Ia+pNwk).Recent is detected in the extant X-ray data. A new X-ray Visionary
theoretical and observational studies suggest that early-type Projectisnowunderway,‘TheTarantulaRevealedbyX-rays(T-
BHGsmightbemassivestarswhichhavepassedthroughablue ReX)’,whichwillobserve30DorwithChandrafor2Ms.This
will push far deeper than the existing data, and we will revisit
1 VFTS766: J=15.14, H=15.01, K =14.76 (from Paper I); theVFTSsampleoncetheT-ReXobservationsandanalysisare
s
[3.6]=14.28,[4.5]=14.06,[5.8]=13.64(fromMeixneretal.2006). complete.
4
C.J.Evansetal.:VFTS:Classifications&radialvelocitiesfortheB-typestars
Fig.2.SpectraofthreepeculiarBe-typestarsfromtheVFTS,inwhicheachspectrumhasbeensmoothedandrebinnedtoR=4000forclarity.
InadditiontotheBalmerlines,theemissionlinesidentifiedinVFTS1003are[SII]λ4069;FeIIλλ4173,4179,4233,4303,4352,4385,4481,
4491,4508,4515,4520-23,4549,4556;[FeII]λλ4244,4277,4287,4358-59,4414-16.EmissionlinesidentifiedinVFTS822areFeIIλλ4584,
4620.5,4629.AbsorptionlinesidentifiedinVFTS766areHeIλλ4009,4026,4121,4144,4388,4471,4713;SiIIλλ4128-32;MgIIλ4481.
4. Stellar radial velocities Following the approach taken by Dufton et al. (2013), we used
differentsetsofdiagnosticlines(assummarisedinTable3)de-
4.1. Singlestars
pendingonthev siniofthestar.
e
Radial-velocity estimates for the stars identified as apparently Fornarrow-linedspectra(definedasv sini(cid:46)150kms−1),
e
singlebyDunstalletal.(inprep.)wereobtainedfromGaussian weusedthreenon-diffuseHeIlinesandfourisolatedmetallines
fits to selected absorption lines in the merged LR02 and LR03 listedas‘Set1’inTable3.Wherepossible,radialvelocitieswere
spectraofeachobject(atthenativeresolutionofthedata,with- estimatedfromSiIIIλ4553inboththemergedLR02andLR03
out the degrading applied for classification purposes). This ap- spectra, giving up to eight individual estimates for the narrow-
proach is consistent with that used by Sana et al. (2013) for linedobjects.
analysisoftheO-typestars,albeit(necessarily)usingadifferent WenotethatHeIλ4121ispotentiallyaffectedbyblending
subset of absorption lines. We note that the choice of Gaussian withOIIλλ4120.28/54.Inmanyinstancesthesefeaturescould
profilesmaynotbeappropriateforfittingthewingsofthelines
beresolvedandadoubleGaussianprofilewasfittedtoestimate
in the presence of different broadening mechanisms, but they
theHeIandOIIlinecentres;incaseswheretheOIIlineswere
should suffice for robust estimates of the centres of our chosen
weaker or unresolved,they do not appearto have unduly influ-
lines.
enced the results (see Table 3, in which the residual for He I
We also considered a cross-correlation method to estimate λ4121isconsistentwiththeotherlines).
the radial velocities. However, given the range in temperature
Radial-velocity estimates for stars with broader lines were
andluminosityofthesample,aswellasthevaryingdataquality
more difficult, particularly for spectra with significant nebular
andproblemsofnebularcontamination,theselectionofsuitable
contamination. In these cases, velocities were estimated by fit-
templatespectraforcross-correlationwasnottrivialandwould
tingthefiveHeIlineslistedas‘Set2’inTable3,forwhichthe
haveintroducedadditionaluncertainties(relatingtoe.g.thepro-
profilesareexpectedtobeeffectivelysymmetric2.
jectedrotationalvelocities,v sini).Thus,wedidnotpursuethis
e
Representative examples are shown in Figure 3. The lower
approachfurther.
panel shows fits to the Si III triplet for two narrow-lined stars,
VFTS053 and 159, which have v sini≤40 (i.e. less than the
e
4.1.1. Lineselection velocityresolutionofthedata)and96±9kms−1,respectively
(from Dufton et al. 2013). As an example of a spectrum of a
Thelargerangeinv siniandeffectivetemperatureforthesam-
e rapidly-rotating star which also suffers nebular contamination,
plemeantthatitwasnotpossibletouseasinglesetofdiagnostic
linesforallstars.Moreover,thelargeintrinsicwidthandnebu-
lar contamination of the Balmer lines precluded their use; sim- 2 HeI3P→3Dtransitionswereexcludedbecauseofthepresenceof
ilarly, nebular emission affects some of the stronger He I lines. significant3P→3Fcomponents.
5
C.J.Evansetal.:VFTS:Classifications&radialvelocitiesfortheB-typestars
Table3.Restwavelengthsusedtoestimateradialvelocities(v )forthe
r
narrow-andbroad-linedstars(Sets1and2,respectively).Valuesinthe
finalcolumnareaveragedifferences(∆)betweenthevelocityestimates
foreachlineandthefinalvelocityforeachstar.
Ion λ[A˚] ∆[kms−1]
Set1
HeI 4120.82 −0.8±6.7
CII 4267.13 1.1±5.1
HeI 4437.55 −0.4±5.4
SiIII 4552.62 0.3±6.1(LR02)
−1.6±5.8(LR03)
SiIII 4567.84 −0.8±6.4
SiIII 4574.76 −1.7±7.0
HeI 4713.15 2.1±4.5
Set2
HeI 4009.26 −0.8±11.3
HeI 4143.76 0.1±9.0
HeI 4387.93 2.7±9.1
HeI 4713.15 2.8±7.1
HeI 4921.93 −3.5±8.7
Notes.WavelengthsaretakenfromtheNISTatomicspectradatabase
(Kramidaetal.2012);theadoptedvalueforCIIistheaverageofthe
twostrongertransitions.
theupperpanelofFigure3showstheGaussianfittoHeIλ4388
forVFTS636(v sini=371±35kms−1; Duftonetal.2013).
e
Forthreestarswithlater/peculiartypes–VFTS272(possi-
bleshellstar),739(A0Ip),and766(B5-8e,seeSection3.1.2)–
weemployedweakFeIIandTiIIabsorptionlinestoestimateve-
locities,althoughwenotethatinVFTS272and766theymight Fig.3.ExamplesofGaussianfitsusedtoestimateradialvelocities(v )
r
notberepresentativeofthetrue(stellar)radialvelocity,asthese for the B-type sample. Upper panel: Fit to He I λ4388 in a rapidly-
linescouldtracecoolerfeaturesassociatedwithadiscofmate- rotatingobject(VFTS636)withnebularcontamination.Lowerpanel:
rial(seee.g.thediscussionofVFTS698byDunstalletal.2012). FitstotheSiIIItriplets(λλ4553-68-74)ofVFTS053and159.
4.1.2. Meanvelocities
Radial-velocity estimates for the single stars were obtained by
calculating an average of the n individual measurements (v ), VFTS, finding excellent agreement between our velocity es-
i
weightedbytheestimatedcentraldepth(d )fromtheGaussian timates and those from Sana et al. (2013, see discussion in
i
fitofeachline,i.e. SectionA.2.1).Wealsoinvestigatedtheinternalconsistencyof
ourresultsfromtheselectedlinesbycalculatingtheaveragedif-
v = (cid:80)ni=1vidi . (1) ference (∆) of the individual measurements for each line com-
r (cid:80)n d pared to the adopted mean velocity for each target. The means
i=1 i
and standard deviations of these differences are given in col-
Thesamplestandarddeviation(σ)foreachstarwassimilarly umn 3 of Table 3, with no significant shifts seen for any of the
calculatedasaweightedaverage: diagnosticlinescomparedtotheadoptedradialvelocities.
(cid:80)n (v −v )2d
σ2 = i=1 i r i . (2)
(cid:80)n d 4.2. Radialvelocitiesofthebinaries
i=1 i
Incaseswherethecentraldepthsoftheabsorptionlineswere To complement the analysis of the single stars, we also inves-
less than twice the inverse of the continuum S/N ratio, the in- tigated the radial velocities of the single-lined binaries in the
dividual estimates were excluded from the above calculations sample.Weobtainedsingle-epochvelocityestimates(vsingle)for
(to avoid their larger uncertainties degrading the results from eachsystemusingmeasurementsfromthebestLR02spectra(in
stronger lines). For stars with estimates from three (or more) termsofS/Nratiooftheco-addedexposuresfromagivennight,
absorptionlines,wepresenttheirestimatedmeanvelocitiesand whichspanthreehoursatmost).
standarddeviationsincolumns7and8ofTable7.Forcomplete- GiventhereducedspectralcoverageoftheLR02datacom-
ness,theline-by-lineestimatesfromSets1and2forthesingle paredtothefullLR02+LR03range,weusedalloftheavailable
starsaredetailedinTablesA.1andA.2,respectively;themean lines from Table 3 to estimate vsingle (except for He I λ4121,
uncertaintyoftheestimatesforthe307(presumed)singlestars given the nearby O II feature and the lower S/N ratio of the
is7.8kms−1(andwithamedianuncertaintyof6.2kms−1). single-epoch spectra compared to the quality of the co-added
As a check on our results, we used the methods described spectra of the single stars). Line-by-line estimates and the date
above to analyse a sample of the late O-type stars from the forwhichtheyweremeasuredarelistedinTableA.3.Meanve-
6
C.J.Evansetal.:VFTS:Classifications&radialvelocitiesfortheB-typestars
locitiesandstandarddeviationsfortheselectedepochwerethen From analysis of Hubble Space Telescope (HST) imaging,
estimatedusingthesamemethodsasforthesinglestars. Grebel & Chu (2000) estimated Hodge301 to be 20-25Myr
The mean single-epoch velocities serve as an absolute ref- old.Fromnarrow-bandimagingtheyalsoidentified19Be-type
erencepointfortheinter-epochdifferentialvelocitiescalculated stars, eight of which (VFTS272, 276, 279, 282, 283, 287, 293,
byDunstalletal.(inprep.)yieldingabsoluteestimatesforeach and301)havespectroscopyfromtheVFTS.TheHαspectraof
epoch. We then calculated the mean and standard deviation of each of these contain significant nebular emission (with mul-
themulti-epochestimatestoobtainavelocityestimateforeach tiple components in all sightlines except towards VFTS293).
single-lined binary. These values are presented in Table 7, and Nonetheless, broad Be-like Hα emission is present in each ob-
areflaggedas‘B’(binary)intheninthcolumn.Wenotethatthe ject,exceptforVFTS276(GC00-Be2).
typicaluncertaintyonthecross-correlationresultsfromDunstall Grebel&Chu(2000)notedtwopotentialbluestragglersin
etal.waslessthan5kms−1,whilethetypicalerrorontheab- thecluster,whichtheyarguedweretooluminousforthemain-
soluteestimatesdeterminedherefromthesingle-epochdatawas sequence population (but less luminous than expected for su-
9kms−1;i.e.theuncertaintiesinourestimateswillgenerallybe pergiants).Thefirst,VFTS270(WB97-2),isaseeminglyunre-
dominatedbytheanalysispresentedhere. markableB3Ibstar.ThesecondisVFTS293(WB97-9,GC00-
FullorbitalsolutionsarepresentlyunavailablefortheB-type Be1), classified here as B2 III-II(n)e. From two previous ob-
binaries,sowestronglycautionthereaderthatthevelocityesti- servations, Walborn & Blades (1997) noted small shifts of its
matesinTable7arenottheircentre-of-mass,systemicvelocities absorptionfeatures(relativetosuperimposednebularemission)
(although,asdiscussedintheAppendix,inmanyinstancesthey and possible weak He II λ4686 emission at one epoch. If the
mayprovidereasonableestimates). He II emission were real, Walborn & Blades speculated that it
mightbeanX-raybinary.Thereisnoevidenceofλ4686emis-
sionintheLR03FLAMESspectra(albeitalltakenonthesame
4.3. Spatialdistribution night, see Paper I), but small line-profile variations are seen in
theabsorptionfeaturesinthemulti-epochLR02spectra.
In addition to NGC2070 (the main 30 Dor cluster, which in-
More intriguing is VFTS310 on the periphery of the clus-
cludes R136 at its core) and NGC2060 (the association to the
terandclassifiedasO9.7V:(Walbornetal.2014).Withsucha
south-west), there are two smaller clusters in the VFTS sur-
classification, VFTS310 would be expected to be significantly
vey:Hodge301(Hodge1988)andSL693(Shapley&Lindsay
younger than the other cluster members, as shown by its loca-
1963);theirrelativelocationsareshowninFigure4.Bothclus-
tion in the Hertzsprung–Russell (H–R) diagram of the cluster
ters appear older than NGC2060 and 2070 as they contain
(Figure 5, see Section 5.3 for details). The star appears oth-
evolvedsupergiants,fromB-throughtoM-types.
erwise unremarkable, with a low projected rotational velocity
We calculatedweighted mean velocities(v , andtheir stan-
r (v sini≤40kms−1, Dufton et al. 2013) and a radial veloc-
dard deviations) for the full sample of single stars, for the e
ity (v =272.3±2.8kms−1, Sana et al. 2013) which is con-
subsamples associated with the four distinct stellar assocations r
sistent with those in NGC2070 and the local field population.
in the survey region (NGC2060, NGC2070, Hodge301, and
Moreover, it does not appear to be a runaway from Hodge301
SL639, as defined in Table 4), and for the field population of
(seeSection6).Whetherthisisatruebluestragglerorsimplya
starsoutsidethefourclusters.Theseresults,whichexcludecan-
line-of-sightcoincidenceremainsuncertainatthistime.
didaterunawaystars(seeSection6),aresummarisedinTable4.
The results for the two younger star-forming regions (i.e.
NGC2060 and 2070) are in excellent agreement with the field 5.2. SL639
populationofB-typestarsintheregion,butthetwoolderclus-
Adopting a 20(cid:48)(cid:48) radius (∼4.9pc), sixteen of the VFTS targets
ters (Hodge301 and SL639) appear at slightly lower recession
arewithinSL693.FourteenareintheB-typesample,withtheir
velocities, suggesting that they are kinematically distinct from
membershipofSL639indicatedinthefinalcolumnofTable74.
the rest of the sample. The weighted mean velocity for the
Tenofthesestarsareearly-Bdwarfs/giants,withsixofthemdis-
273singlestarsexcludingthesetwoclusters(andpotentialrun-
aways)isv =271.6±12.2kms−1. playingBe-likeHαemission.Theremainingfourarecomprised
r
of two supergiants (VFTS827, B1.5 Ib; VFTS831, B5 Ia) and
twobrightgiants(VFTS826,B1IIn;VFTS829,B1.5-2II).
TherearepreviousageestimatesforSL693.Fromconsider-
5. Spectral content of the older clusters
ationofitsmain-sequenceturn-off,Hodge(1983)estimatedits
The VFTS data comprise the first comprehensive spectroscopy age to be 18Myr (although he noted a large uncertainty com-
inHodge301andSL639,sowebrieflydiscusstheseclustersin pared to the theoretical models of the time). A larger estimate
turn. of30MyrwasgivenbySantosetal.(1995)fromitsreddening-
corrected,integrated(U −B) colourandadoptinganempirical
0
calibrationofcolourvs.agefromBicaetal.(1990).
5.1. Hodge301
Thereare20VFTStargetswithina20(cid:48)(cid:48)radius(chosentoconser-
5.3. H–Rdiagram
vativelyidentifylikelymembers,andcorrespondingto∼4.9pc,
assumingadistancemodulusof18.5mag).Fifteenofthesewere Motivated by the apparent similarities of the two clusters in
classifiedasB-typeandtheirlikelymembershipoftheclusteris terms of their spectral content, we constructed an H–R dia-
indicatedinthefinalcolumnofTable73.
4 ThetwoothersareVFTS820and828(A0IaandEarlyM,respec-
3 Thefiveothersare:VFTS310(O9.7V:, Walbornetal.2014)and tively,fromPaperI).TheonlyotherVFTStargetwithina35(cid:48)(cid:48)(8.5pc)
VFTS271,281,289,and294(A7II,Mid-lateK,LateG/EarlyK,and radiusisVFTS839at∼25(cid:48)(cid:48) (6pc),classifiedasG-typeinPaperI.All
A0 Ib, respectively, from Paper I). Three additional stars (VFTS263, threeofthesecoolerobjectshaveradialvelocitiesconsistentwithmem-
291,and317)arealsonearby,atradiifromtheadoptedcentreof∼25(cid:48)(cid:48). bershipoftheLMC.
7
C.J.Evansetal.:VFTS:Classifications&radialvelocitiesfortheB-typestars
Fig.4.SpatialdistributionoftheB-typesinglestars(cyancircles)andspectroscopicbinaries(openbluesquares)withradial-velocityestimates;
the nine candidate runaways, labelled by their VFTS identifiers, are marked in red. The spatial extents of NGC2070, NGC2060, SL639, and
Hodge301(asdefinedinTable4),areindicatedbytheoverlaiddashedcircles.TheunderlyingimageisfromaV-bandmosaictakenwithESO’s
WideFieldImageronthe2.2mtelescopeatLaSilla.
graminordertocomparethepropertiesoftheirlikelymembers. The H–R diagram is shown in Figure 5, with SL639 and
AheadofdetailedquantitativeanalysisoftheVFTSspectraand Hodge301 members indicated by the filled black and open red
pendingextinctionanalysisofthewholesample,weadoptedef- circles, respectively. Evolutionary tracks and isochrones (for
fective temperatures (T ) on the basis of the spectral types in non-rotatingmodels)fromBrottetal.(2011)arealsoplottedfor
eff
Table 3 using the calibrations of Trundle et al. (2007), and in- comparison.Asnotedearlier,thelateO-typestarinHodge301,
terpolating between the Galactic and SMC temperatures for A- VFTS310,standsoutasabluestragglerwithT >30kK.
eff
type supergiants from Evans & Howarth (2003); cooler objects Tailored atmospheric and extinction analyses of these ob-
wereomittedgiventhelargeruncertaintiesintheirspectralclas- jects should help to provide further insight in due course.
sifications. Stellar luminosities were estimated using the opti- Nonetheless, the stars in SL639 appear marginally younger
cal photometry from Paper I, intrinsic colours from Fitzgerald thanthoseinHodge301.Fromqualitativecomparisonwiththe
(1970),bolometriccorrectionsfromBalona(1994),andassum- isochrones, we estimate an age in the range of 10-15Myr for
ingRV =3.5(seeAppendixCofDoranetal.2013). SL639,significantlyyoungerthantheestimateof30Myrfrom
Santosetal.(1995).Thereisalargerspreadintheinferredages
8
C.J.Evansetal.:VFTS:Classifications&radialvelocitiesfortheB-typestars
Table4.Meanradialvelocities(v )andstandarddeviations(σ)ofthestellarassociationsidentifiedwithin30Doradus.
r
Region No.ofstars Radius α δ v ±σ
r
[(cid:48)] [J2000.0]
All 298 ... ... ... 270.4±12.4
NGC2070 78 2.40 053842.0 −690602.0 271.2±12.4
NGC2060 22 2.40 053745.0 −691015.0 273.7±13.4
Hodge301 14 0.33 053817.0 −690401.0 261.8± 5.5
SL639 11 0.33 053939.4 −691152.1 253.5± 3.9
Field 173 ... ... ... 271.3±11.8
All(excl.H301/SL639) 273 ... ... ... 271.6±12.2
Notes.ThesestatisticsexcludetheninecandidaterunawaysidentifiedinTable5.Forconsistency,thecentralpositionsadoptedforNGC2060and
NGC2070arethoseusedbySanaetal.(2013).ApproximatecentresforHodge301andSL639wereadoptedusingthecoordinatesofWB97-5
(Walborn&Blades1997)andVFTS828,respectively.
NGC2060,NGC2070,andthesurroundingfieldpopulation(i.e.
excluding Hodge301 and SL639). Employing a threshold of
|δv | >3σ weidentifiedeightcandidaterunawaysandthenre-
r
calculatedthemeanvelocities(anddispersions)excludingthese
objects.Reapplyinga3-σthresholdprovidedoneadditionalcan-
didate,whileaseconditerationdidnotreducethevelocitydis-
persionfurther.Themeanvelocitiesanddispersionsforthedif-
ferent spatial samples (excluding potential runaways) are pre-
sented in Table 4. The cumulative distribution of the velocity
estimatesforthe307singlestarsisshowninFigure6,withthe
±3σthreshold(i.e.±36.6kms−1)foridentificationofrunaway
starsindicatedbytheverticaldottedlines.
Ouradoptedvelocitythresholdiscomparableinmagnitude
to the limit used by Blaauw (1961), which was that candidate
runaways have a space velocity which differs by more than
40kms−1fromthesystemicvelocityoftheregioninwhichthey
formed. (In this context, our radial-velocity estimates should
serve as lower limits to the 3-D space velocities.) In contrast,
PortegiesZwart(2000)adoptedalowerthresholdof25kms−1
inhistheoreticalstudyoftheoriginsandcharacteristicsofrun-
aways.GiventhelimitationsoftheB-typedataandtherelatively
smallnumberoflinesavailableforsomeobjects,wedidnotcon-
sideracomparablylowthreshold.
Details of the nine candidate runaways are summarised in
Fig.5.Hertzsprung–RusselldiagramfortheVFTSobjectsobservedin Table 5, and their locations are shown in Figure 4. Four of
SL639(blackcircles)andHodge301(red,opencircles).Alsoplotted the candidates have differential velocities in the range of 3.2-
are evolutionary tracks for initial masses of 8, 10, 12, 15, 20, 25M
(cid:12) 3.6σ,thustheirstatusascandidateswilldependstronglyonthe
(solidlines),thezero-agemainsequence(dashedline),andisochrones
adoptedvelocitycriteria.Moreover,ouradoptedvelocitythresh-
(3, 5, 10, 15, 20, and 25Myr, dotted lines) for the non-rotating mod-
old(thedispersionoftheweightedmeanvelocitiesinthesam-
elsfromBrottetal.(2011).TheapparentbluestragglerinHodge301
ple) was employed without consideration of the uncertainty on
(VFTS310)isalateO-typestar.
theradial-velocityestimateforagivenstar.Thus,VFTS467,al-
thoughformallywitharadialvelocityof4.2σfromthesystemic
value, has a large uncertainty on its estimated velocity. Indeed,
for the stars in Hodge301 of 15±5Myr, overlapping with the only three He I lines were available for this star and the λ4144
lowerestimateof20MyrfromthephotometricstudyofGrebel estimate(277.5kms−1)iscomparabletothesystemicvalue.
& Chu (2000). Adopting isochrones calculated for the rotating
The choice to use weighted means to investigate the
models from Brott et al. had little impact on these results; e.g.
assuminganintialrotationalvelocityof301kms−1,theinferred spatially-different samples in Section 4.3 was motivated by the
factthatapproximately10%ofourvelocityestimateshaveun-
ageswereonlymarginallyyounger(by∼1Myr).
certainties of ≥15kms−1 (each ofwhich is abroad-lined star,
i.e. used the Set 2 lines from Table 3). To investigate the de-
6. Candidate runaways pendenceofthenumberandidentityofthecandidaterunaways
on these adopted means, we repeated the above steps for both
6.1. Singlestars
the (unweighted) mean and median velocities of the sample.
To identify candidate runaways in the B-type sample we com- The mean velocity for the 281 stars (not including those in
pared the radial-velocity estimates for each star with the initial Hodge301andSL639)was269.8=16.4kms−1,withamedian
weightedmeanandstandarddeviationcalculatedforthestarsin valueof270.7kms−1.Adoptingeitherofthesesystemicvalues
9
C.J.Evansetal.:VFTS:Classifications&radialvelocitiesfortheB-typestars
Table5.Candidaterunawaystarsidentifiedfromtheradial-velocityanalysisoftheVFTSB-typespectra.
VFTS Classification v (v −v ) |v −v |/σ v sini r
r r r r r e d
[kms−1] [kms−1] [kms−1] [(cid:48)] [pc]
242 B0V-IV 216.0± 3.6 −55.6±12.7 4.6 ≤40 5.96 86.6
298 B1-2V-IIIe+ 332.1± 4.0 60.5±12.8 5.0 431±15 3.92 57.1
358 B0.5:V 164.7±10.7 −106.9±16.2 8.8 345±22 3.02 44.0
368 B1-3V 229.6±12.4 −42.0±17.4 3.4 380: 1.20 17.5
467 B1-2Ve+ 322.3±30.2 50.7±32.6 4.2 355±29 9.54 138.7
616 B0.5:V 227.4±10.7 −44.2±16.2 3.6 ≤40 1.26 18.3
625 B1.5V 310.7± 9.2 39.1±15.3 3.2 64±12 2.57 37.4
831 B5Ia 210.4± 6.2 −61.2±13.7 5.0 41 7.91 115.0
851 B2III 229.4± 0.3 −42.2±12.2 3.5 ≤40 7.18 104.4
Notes.Estimatesofv siniarefromDuftonetal.(2013)exceptforVFTS831,whichistakenfromMcEvoyetal.(2015).Projecteddistances
e
fromthecoreofR136arefromPaperI.
anda3-σthreshold(takingthedispersionas16.4kms−1)ledto
identificationofthefivemostsignificantrunawaysfromTable5
(i.e.VFTS242,298,358,467,and831),withaseconditeration
providingoneadditionalcandidate(VFTS616).Thefinalmean
velocityfromthesecalculationsis270.1±13.9,ingoodagree-
mentwiththeequivalentsubsampleinthefinalrowofTable4.
From these checks we conclude that the four most significant
candidates (VFTS242, 298, 358, and 831) are secure as poten-
tial runaways, with five more tenative candidates as discussed
above.
For the members of Hodge301, VFTS309 is a poten-
tial outlier (at 3.6σ) compared to the mean velocity for
the cluster in Table 4. Indeed, excluding its v estimate
r
from the calculated mean and dispersion for Hodge301 gives
v =261.5±5.0kms−1,suchthatVFTS309isthena4.1σout-
r
lier. However, its velocity is also in reasonable agreement with
thoseforstarsinNGC2060andNGC2070,soweareunableto
Fig.6. Cumulative distribution of estimated radial velocities for the
confirm it as a potential runaway from Hodge301. Finally, we
307singleB-typestars.Thedashedverticallineindicatestheweighted
notethatthevelocityestimateforVFTS310isnotsignificantas
meanvelocityofthesample,v =271.6kms−1 (calculatedexcluding
apotentialrunawayfromHodge301. r
starsinthetwoolderclustersandthecandidaterunaways).Thevertical
With our adopted velocity criteria, the nine candidate run- dottedlinesarethe3σthresholdsadoptedforidentificationofpotential
aways correspond to 2.9% of the sample of 307 single stars runways(±36.6kms−1).
withradial-velocityestimates.Thislowerfractionalshareofrun-
aways compared to the O-type stars (Sana et al., in prep.) is
inqualitativeagreementwithrecentN-bodypredictionsforthe
mass-spectrum of runaways (Banerjee et al. 2012; Banerjee &
Kroupa 2012), although quantitative comparison with the pre- andthefifthwithvesini=64±12kms−1.Meanwhile,there-
dictions is complicated by the different criteria employed (e.g. mainingfourstarseachhaveestimatesofvesini≥345kms−1.
Banerjee et al. used a radial distance threshold for a star to be To investigate the significance of these results, we ran Monte
consideredarunaway).WenotethatthedistributioninFigure6 Carlo simulations using the vesini distributions from Dufton
is slightly asymmetric, with two thirds of the candidates hav- etal.(2013).WenotethattheresultsfromDuftonetal.werefor
ingnegativedifferentialvelocities.Thiscouldpointtoveilingof the unevolved B-type sample (i.e. dwarf and giant objects), so
more distant runaways by the nebula, although this is not seen weexcludeVFTS831fromthefollowingdiscussion.
in the results for the O-type stars (in which a greater fraction We generated random rotational velocities for eight stars
are seen to have positive differential velocities, Sana et al., in drawn from the distribution of Dufton et al., and repeated
prep.),suggestingthatsmall-numberstatisticslikelyexplainthe this 106 times to estimate the probability of recovering rota-
asymmetry. tionalvelocitiesconsistentwithoureight(non-supergiant)run-
awaycandidates(i.e.fourstarswithv sini≤65,andfourwith
e
v sini≥345kms−1). We then repeated these runs 104 times
6.2. Rotationalvelocitiesofcandidaterunaways e
to estimate the dispersion on the probabilities. In a second test
Estimated projected rotational velocities (from Dufton et al. we limited the low velocity subsample to the three stars with
2013; McEvoy et al. 2015) for our candidate B-type runaways v sini≤40kms−1 (i.e. excluding VFTS625, which has the
e
areincludedinTable5.Thereappearstobeadichotomyintheir leastsignificantrunawayvelocity),togetherwiththefourrapid
rotationrates,ashighlightedbyFig.7.Fivestarshavesmallro- rotators.Wefoundlowprobabilitiesforrecoveringtheobserved
tational velocities, with the estimates for four stars effectively v sinivaluesfrombothtests(assummarisedinTable6),with
e
limited by the velocity resolution of the data (i.e. ∼40kms−1) significancelevelsof 3.0and3.2σ,respectively.
10