The VLT-FLAMES Tarantula Survey: XXVIII. Nitrogen abundances for apparently single dwarf and giant B-type stars with small projected rotational velocities
P.L. Dufton, A. Thompson, P. A. Crowther, C. J. Evans, F.R.N. Schneider, A. de Koter, S. E. de Mink, R. Garland, N. Langer, D. J. Lennon, C. M. McEvoy, O.H. Ramírez-Agudelo, H. Sana, S. Símon Díaz, W. D. Taylor, J. S. Vink
aa r X i v : . [ a s t r o - ph . S R ] A p r Astronomy & Astrophysicsmanuscript no. 32440_final c (cid:13)
ESO 2018April 9, 2018
The VLT-FLAMES Tarantula Survey ⋆ XXVIII. Nitrogen abundances for apparently single dwarf and giant B-type starswith small projected rotational velocities
P.L. Dufton , A. Thompson , P. A. Crowther , C. J. Evans , F.R.N. Schneider , A. de Koter , S. E. de Mink , R.Garland , , N. Langer , D. J. Lennon , C. M. McEvoy , , O.H. Ramírez-Agudelo , H. Sana , S. Símon Díaz , , W.D. Taylor , J. S. Vink Astrophysics Research Centre, School of Mathematics and Physics, Queen’s University Belfast, Belfast BT7 1NN, UK Department of Physics and Astronomy, Hounsfield Road, University of She ffi eld, She ffi eld, S3 7RH, UK UK Astronomy Technology Centre, Royal Observatory Edinburgh, Blackford Hill, Edinburgh, EH9 3HJ, UK Department of Physics, University of Oxford, Keble Road, Oxford OX1 3RH, United Kingdom Anton Pannenkoek Institute for Astronomy, University of Amsterdam, NL-1090 GE Amsterdam, The Netherlands Sub-department of Atmospheric, Oceanic and Planetary Physics, Department of Physics, University of Oxford, Oxford, OX1 3RH,UK Argelander-Institut für Astronomie der Universität Bonn, Auf dem Hügel 71, 53121 Bonn, Germany European Space Astronomy Centre (ESAC), Camino bajo del Castillo, s / n Urbanizacion Villafranca del Castillo, Villanueva de laCañada, E-28692 Madrid, Spain King’s College London, Graduate School, Waterloo Bridge Wing, Franklin Wilkins Building, 150 Stamford Street, London SE19NH, UK Instituut voor Sterrenkunde, Universiteit Leuven, Celestijnenlaan 200 D, B-3001 Leuven, Belgium Instituto de Astrofísica de Canarias, E-38200 La Laguna, Tenerife, Spain Departamento de Astrofísica, Universidad de La Laguna, E-38205 La Laguna, Tenerife, Spain Armagh Observatory, College Hill, Armagh, BT61 9DG, Northern Ireland, UKReceived / accepted ABSTRACT
Previous analyses of the spectra of OB-type stars in the Magellanic Clouds have identified targets with low projected rotationalvelocities and relatively high nitrogen abundances; the evolutionary status of these objects remains unclear. The VLT-FLAMESTarantula Survey obtained spectroscopy for over 800 early-type stars in 30 Doradus of which 434 stars were classified as B-type.We have estimated atmospheric parameters and nitrogen abundances using tlusty model atmospheres for 54 B-type targets thatappear to be single, have projected rotational velocities, v e sin i ≤
80 km s − and were not classified as supergiants. In addition, nitrogenabundances for 34 similar stars observed in a previous FLAMES survey of the Large Magellanic Cloud have been re-evaluated.For both samples, approximately 75-80% of the targets have nitrogen enhancements of less than 0.3 dex, consistent with them havingexperienced only small amounts of mixing. However, stars with low projected rotational velocities, v e sin i ≤
40 km s − and significantnitrogen enrichments are found in both our samples and simulations imply that these cannot all be rapidly rotating objects observednear pole-on. For example, adopting an enhancement threshold of 0.6 dex, we observed five and four stars in our VFTS and previousFLAMES survey samples, yet stellar evolution models with rotation predict only 1.25 ± ± − . This would correspond to ∼ ∼
70% of the stars with rotational velocities less than 40 km s − and ∼ Key words. stars: early-type – stars: rotation – stars: abundances – Magellanic Clouds – galaxies: star cluster: individual: TarantulaNebula
1. Introduction
The evolution of massive stars during their hydrogen coreburning phase has been modelled for over sixty years withearly studies by, for example, Tayler (1954, 1956); Kushwaha ⋆ Based on observations at the European Southern Observatory inprogrammes 171.D0237, 073.D0234 and 182.D-0222 (1957); Schwarzschild & Härm (1958); Henyey et al. (1959);Hoyle (1960). These initial models lead to a better understand-ing of, for example, the observational Hertzsprung-Russell di-agram for young clusters (Henyey et al. 1959) and the dynam-ical ages inferred for young associations (Hoyle 1960). Subse-quently there have been major advances in the physical assump-tions adopted in such models, including the e ff ects of stellar ro- Article number, page 1 of 21 & Aproofs: manuscript no. 32440_final tation (Maeder 1987), mass loss (particularly important in moreluminous stars and metal-rich environments; Chiosi & Maeder1986; Puls et al. 2008; Mokiem et al. 2007) and magnetic fields(Donati & Landstreet 2009; Petermann et al. 2015). Recent re-views of these developments include Maeder (2009) and Langer(2012).Unlike late-type stars, where stellar rotation velocitiesare generally small (see, for example, Gray 2005, 2016,2017, and references therein), rotation is an important phe-nomenon a ff ecting both the observational and theoreticalunderstanding of early-type stars. Important observationalstudies of early-type stellar rotation in our Galaxy in-clude Slettebak (1949), Conti & Ebbets (1977), Penny (1996),Howarth et al. (1997), Abt et al. (2002), Huang et al. (2010)and Simón-Díaz & Herrero (2014). There have also been ex-tensive investigations for the Magellanic Clouds using ESOlarge programmes, viz. Martayan et al. (2006, 2007, and ref-erences therein), Evans et al. (2005, 2006) and Hunter et al.(2008b). More recently as part of the VLT-Flames TarantulaSurvey (Evans et al. 2011), rotation in both apparently single(Ramírez-Agudelo et al. 2013; Dufton et al. 2013) and binary(Ramírez-Agudelo et al. 2015) early-type stars has been inves-tigated. These studies all show that early-type stars cover thewhole range of rotational velocities up to the critical velocityat which the centrifugal force balances the stellar gravity at theequator (Struve 1931; Townsend et al. 2004).These observational studies have led to rotation being in-corporated into stellar evolutionary models (see, for exam-ple, Maeder 1987; Heger & Langer 2000; Hirschi et al. 2004;Frischknecht et al. 2010), leading to large grids of rotating evo-lutionary models for di ff erent metallicity regimes (Brott et al.2011a; Ekström et al. 2012; Georgy et al. 2013a,b). Such mod-els have resulted in a better understanding of the evolu-tion of massive stars, including main-sequence life times(see, for instance Meynet & Maeder 2000; Brott et al. 2011a;Ekström et al. 2012) and the possibility of chemically homoge-neous evolution (Maeder 1980; Maeder et al. 2012; Szécsi et al.2015), that in low-metallicity environments could lead togamma-ray bursts (Yoon & Langer 2005; Yoon et al. 2006;Woosley & Heger 2006). However there remain outstandingissues in the evolution of massive single and binary stars(Meynet et al. 2017). For example, nitrogen enhanced early-typestars with low projected rotational velocities that may not be theresult of rotational mixing have been identified by Hunter et al.(2008a), Rivero González et al. (2012), and Grin et al. (2017).The nature of these stars have discussed previously by, for ex-ample, Brott et al. (2011b), Maeder et al. (2014), and Aerts et al.(2014). Definitive conclusions have been hampered both by ob-servational and theoretical uncertainties and by the possibility ofother evolutionary scenarios (see Grin et al. 2017, for a recentdetailed discussion).Here we discuss the apparently single B-type stars with lowprojected rotational velocity found in the VLT-Flames TarantulaSurvey (Evans et al. 2011, hereafter VFTS). These complementthe analysis of the corresponding binary stars from the samesurvey by Garland et al. (2017). We also reassess the results ofHunter et al. (2007, 2008a) and Trundle et al. (2007a) from aprevious spectroscopic survey (Evans et al. 2006). In particularwe investigate whether there is an excess of nitrogen enhancedstars with low projected velocities and attempt to quantify anysuch excess. We then use these results to constrain the evolu-tionary pathways that may have produced these stars.
2. Observations
The VFTS spectroscopy was obtained using the MEDUSAand ARGOS modes of the FLAMES instrument (Pasquini et al.2002) on the ESO Very Large Telescope. The former uses fi-bres to simultaneously ‘feed’ the light from over 130 stellartargets or sky positions to the Gira ff e spectrograph. Nine fibreconfigurations (designated fields ‘A’ to ‘I’ with near identicalfield centres) were observed in the 30 Doradus region, sam-pling the di ff erent clusters and the local field population. Spec-troscopy of more than 800 stellar targets was obtained and ap-proximately half of these were subsequently spectrally classifiedby Evans et al. (2015) as B-type. The young massive cluster atthe core of 30 Doradus, R136, was too densely populated forthe MEDUSA fibres, and was therefore observed with the AR-GUS integral field unit (with the core remaining unresolved evenin these observations). Details of target selection, observations,and initial data reduction have been given in Evans et al. (2011),where target co-ordinates are also provided.The analysis presented here employed the FLAMES–MEDUSA observations that were obtained with two of the stan-dard Gira ff e settings, viz. LR02 (wavelength range from 3960to 4564 Å at a spectral resolving power of R ∼ ∼ v e sin i , have been estimatedby Dufton et al. (2013) and Garland et al. (2017).In principle, a quantitative analysis of all these apparentlysingle stars, including those with large projected rotationalvelocities, would have been possible. However, the moderatesignal-to-noise ratios (S / N) of our spectroscopy would lead tothe atmospheric parameters (especially the e ff ective temperatureand microturbulence – see Sect. 4.2 and 4.4) being poorly con-strained for the broader lined targets. In turn this would leadto nitrogen abundance estimates that had little diagnostic value.Hence for the purposes of this paper we have limited our sampleto the subset with relatively small projected rotational velocities.This is consistent with the approach adopted by Garland et al.(2017) in their analysis of the corresponding narrow lined B-typeVFTS binaries.Hence our targets were those classified as B-type with a lu-minosity class III to V, which showed no evidence of signifi-cant radial velocity variations and had an estimated v e sin i ≤ − ; the B-type supergiants (luminosity classes I to II) havebeen discussed previously by McEvoy et al. (2015). Stars withan uncertain O9 or B0 classification have not been included. Ad-ditionally the spectrum of VFTS 469 had been identified in Pa-per I as su ff ering some cross contamination from a brighter O-type star in an adjacent fibre. Another of our targets, VFTS 167,may also su ff er such contamination as the He ii line at 4686 Åshows broad emission. We retain these two stars in our samplebut discuss the reliability of their analyses in Sect. 6.2. Some ofour apparently single stellar sample will almost certainly have acompanion. In particular, as discussed by Dunstall et al. (2015),low mass companions or binaries in wide orbits may not havebeen identified.Table 1 summarises the spectral types (Evans et al. 2015) andtypical S / N (rounded to the nearest multiple of five) in the LR02spectral region. The latter were estimated from the wavelengthregion, 4200-4250 Å, which should not contain strong absorp-tion lines. However particularly for the higher estimates, these
Article number, page 2 of 21.L. Dufton et al.: B-type stellar nitrogen abundances in the Tarantula Nebula should be considered as lower limits as they could be a ff ectedby weak lines. S / N for the LR03 region were normally sim-ilar or slightly smaller (see, for example, McEvoy et al. 2015,for a comparison of the ratios in the two spectral region). Alsolisted are the estimated projected rotational velocity, v e sin i (Dufton et al. 2013; Garland et al. 2017), the mean radial veloc-ity, v r (Evans et al. 2015), and the range, ∆ v r , in radial velocityestimates (Dunstall et al. 2015). The moderate spectral resolv-ing power led to estimates of the projected rotational velocityfor stars with v e sin i ≤
40 km s − being poorly constrained andhence approximately half the sample have been assigned to a binwith 0 ≤ v e sin i ≤
40 km s − .Luminosities have also been estimated for our targets usinga similar methodology to that adopted by Garland et al. (2017).Interstellar extinctions were estimated from the observed B − V colours provided in Evans et al. (2011), together with the in-trinsic colour–spectral-type calibration of Wegner (1994) and R V = . tlusty models (Lanz & Hubeny2007) and together with an adopted LMC distance modulus of18 m .
49 (Pietrzy´nski et al. 2013), the stellar luminosity were esti-mated. Table 1 lists these estimates (hereafter designated as ‘ob-served luminosities’) apart from that for VFTS 835 for whichno reliable photometry was available. The main sources of un-certainty in these luminosity estimates will arise from the bolo-metric correction and the extinction. We estimate that these willtypically contribute an error of ± m .5 in the absolute magnitudecorresponding to ± , all useable ex-posures were combined using either a median or weighted σ -clipping algorithm. The final spectra from the di ff erent meth-ods were normally indistinguishable. The full wavelength rangefor each wavelength setting could usually be normalised using asingle, low-order polynomial. However, for some features (e.g.the Balmer series), the combined spectra around individual (orgroups of) lines were normalised.In the model atmosphere analysis, equivalent widths weregenerally used for the narrower metal lines, with profiles beingconsidered for the broader H i and He ii lines. For the former,Gaussian profiles (and a low order polynomial to represent thecontinuum) were fitted to the observed spectra leading to formalerrors in the equivalent width estimates of typically 10%. The fitswere generally convincing which is consistent with the intrinsicand instrumental broadening being major contributors to thesenarrow lined spectra. Tests using rotationally broadened profilesyielded equivalent width estimates in good agreement with thoseusing Gaussian profiles – di ff erences were normally less than10% but the fits were generally less convincing.For some spectra, lines due to important diagnostic species(e.g. Si ii , Si iv and N ii ) could not be observed; in these caseswe have set upper limits on the equivalent widths. Our approachwas to measure the equivalent width of the weakest metal line orlines that were observable in the VFTS spectra of narrow linedstars. These were then rounded to the nearest 5 mÅ and plottedagainst the S / N of the spectra as shown in Fig. 1. The equiva-lent widths decrease with increasing S / N as would be expectedand there also seems to be no significant di ff erence for stars with0 < v e sin i ≤
40 km s − and those with 40 < v e sin i ≤
80 km s − .The upper limit for an equivalent width at a given S / N probably The DR2.2 pipeline reduction was adopted and the spectra are avail-able at http: // / ∼ cje / tarantula / spectra Fig. 1.
Estimated equivalent width limits (in mÅ) plotted against theS / N of the VFTS spectra. Crosses and triangles represent stars with v e sin i ≤
40 km s − and 40 < v e sin i ≤
80 km s − respectively. A quadraticlinear least squares fit for all the estimates is also shown. lies at or near the lower envelope of this plot. However we havetaken a more cautious approach by fitting a low order polynomialto these results and using this to estimate upper limits to equiv-alent widths of unseen lines. We emphasise that this will leadto conservative estimates both for the upper limits of equivalentwidths and for the corresponding element abundances.The VFTS data have been supplemented by those froma previous FLAMES / VLT survey (Evans et al. 2005, hereafterthe FLAMES-I survey). Spectral classification by Evans et al.(2006) identified 49 B-type stars that also had v e sin i ≤
80 km s − (Hunter et al. 2008b). Fourteen of these were characterised asbinaries (Evans et al. 2006), either SB2 (on the basis of theirspectra) or SB1 (on the basis of radial velocity variations be-tween epochs). These targets were discussed by Garland et al.(2017) and will not be considered further. For the apparentlysingle stars, atmospheric parameters previously estimated fromthis FLAMES-I spectroscopy (Hunter et al. 2007, 2008a, 2009;Trundle et al. 2007b) have been adopted. However we have re-reduced the spectroscopy for the wavelength region incorporat-ing the N ii line at 3995 Å, using the methods outlined above.This should provide a better data product than the simple co-adding of exposures that was under taken in the original analy-ses. Comparison of spectra from the two reductions confirmedthis (although in many cases the improvement was small).
3. Binarity criteria
Dunstall et al. (2015) adopted two criteria for the identifica-tion of binary systems . Firstly the di ff erence between two es-timates of the radial velocity had to be greater than four timesthe estimated uncertainty and secondly this di ff erence had tobe greater than 16 km s − ; this threshold was chosen so as tolimit false positives due to for example pulsations that are likelyto be significant particularly in supergiants (Simón-Díaz et al.2010; Taylor et al. 2014; Simón-Díaz et al. 2017a). In Table 1,the ranges ( ∆ v r ) of radial velocities of six targets fulfil the lat-ter (but not the former) criteria. As discussed by Dunstall et al.(2015) a significant fraction of binaries may not have been iden-tified and these six targets could therefore be binaries. Four ofthese targets (VFTS 010, 625, 712, 772) were not analysed due to Some of these systems could contain more than two stars.Article number, page 3 of 21 & Aproofs: manuscript no. 32440_final the relatively low S / N of their spectroscopy (see Sect. 4), whichprobably contributed their relatively large values of ∆ v r .For the other two targets, we have re-evaluated the spec-troscopy as follow: VFTS 313:
Dunstall et al. (2015) obtained radial velocityestimates only from the He i di ff use lines at 4026 and 4387 Å,which may have contributed to the relatively large value of ∆ v r ∼
18 km s − . There was considerable scatter between estimatesfound for exposures obtained at a given epoch, especially thefirst epoch when six consecutive exposures were available.Given the high cadence of these observations, these variationsare unlikely to be real and probably reflect uncertainties inthe individual measurements. The best observed intrinsicallynarrow feature in the LR02 spectra was the Si iv and O ii closeblend (separation of 0.60 Å) at 4089 Å. Radial velocity estimateswere possible for eleven out the twelve LR02 exposures (theother exposure was a ff ected by a cosmic ray event) fromfitting a Gaussian profile to this blend. The estimates rangedfrom 270-282 km s − with a mean value of 275.1 ± − (assuming that the Si iv feature dominates the blend). VFTS 835:
Dunstall et al. (2015) obtained radial velocityestimates only from the He i di ff use lines at 4026, 4387 and4471 Å, which may again have contributed to the relatively largevalue of ∆ v r ∼
18 km s − . As for VFTS 313, there was againconsiderable scatter between estimates found for exposureswithin a given epoch. Unfortunately the metal lines werebroadened by stellar rotation, which precluded their use forestimating reliable radial velocities in the single exposures,which had relatively low S / N.Hence we find no convincing evidence for binarity in eitherof these targets and have retained them in our apparently singlestar sample.
4. Atmospheric parameters
We have employed model-atmosphere grids calculated with the tlusty and synspec codes (Hubeny 1988; Hubeny & Lanz 1995;Hubeny et al. 1998; Lanz & Hubeny 2007). They cover a rangeof e ff ective temperature, 10 000K ≤ T e ff ≤
35 000K in steps oftypically 1 500K. Logarithmic gravities (in cm s − ) range from4.5 dex down to the Eddington limit in steps of 0.25 dex, andmicroturbulences are from 0-30 km s − in steps of 5 km s − . Asdiscussed in Ryans et al. (2003) and Dufton et al. (2005), equiv-alent widths and line profiles interpolated within these grids arein good agreement with those calculated explicitly at the relevantatmospheric parameters.These non-LTE codes adopt the ‘classical’ stationary modelatmosphere assumptions, that is plane-parallel geometry, hydro-static equilibrium, and the optical spectrum is una ff ected bywinds. As the targets considered here have luminosity classesV to III, such an approach should provide reliable results. Thegrids assumed a normal helium to hydrogen ratio (0.1 by numberof atoms) and the validity of this is discussed in Sect. 6.2. Gridshave been calculated for a range of metallicities with that for anLMC metallicity being used here. As discussed by Dufton et al.(2005), the atmospheric structure depends principally on the ironabundance with a value of 7.2 dex having been adopted forthe LMC; this is consistent with estimates of the current ironabundance for this galaxy (see, for example, Carrera et al. 2008; Table 1.
VFTS B-type targets with a luminosity class V to III(Evans et al. 2015), which show no evidence of significant radial ve-locity variations (Dunstall et al. 2015; Evans et al. 2015) and have a v e sin i ≤
80 km s − (Dufton et al. 2013; Garland et al. 2017). The meanradial velocities, v r are from Evans et al. (2015) and the ranges, ∆ v r , inradial velocity estimates are from Dunstall et al. (2015) with all veloci-ties having units of km s − . Further detail on the targets including theirco-ordinates can be found in Evans et al. (2011). The estimates of theS / N for the LR02 region have been rounded to the nearest multiple offive, whilst luminosities, L, were estimated as discussed in Sect. 2.
VFTS ST S / N v e sin i v r ∆ v r log L / L ⊙
010 B2 V 40 ≤
40 287 25 3.87024 B0.2 III-II 100 58 286 10 4.80029 B1 V 80 ≤
40 286 10 3.95044 B2 V 80 65 288 7 3.79050 B0 V 60 ≤
40 283 3 4.35052 B0.2 III-II 145 48 280 4 4.77053 B1 III 125 ≤
40 276 3 4.56075 B1 V 65 70 297 3 3.94095 B0.2 V 100 ≤
40 294 4 4.36111 B2 III 215 80 272 2 4.56119 B0.7 V 60 ≤
40 276 11 4.27121 B1 IV 115 ≤
40 284 3 4.31124 B2.5 III 75 ≤
40 289 6 4.04126 B1 V 80 ≤
40 254 3 3.86152 B2 IIIe 130 47 274 4 4.65167 B1 V 110 ≤
40 286 14 4.26170 B1 IV 120 ≤
40 279 1 4.41183 B0 IV-III 70 ≤
40 256 2 4.66202 B2 V 120 49 277 7 4.25209 B1 V 80 ≤
40 273 14 4.06214 B0 IV-III 80 ≤
40 276 4 4.78237 B1-1.5 V-IVe 110 79 288 10 4.35241 B0 V-IV 75 69 268 16 4.14242 B0 V-IV 60 ≤
40 216 8 4.17273 B2.5 V 70 ≤
40 254 4 3.78284 B1 V 80 ≤
40 273 15 3.91297 B1.5 V 80 47 252 9 4.01308 B2 V 160 74 263 3 4.17313 B0 V-IV 110 56 270 18 4.41331 B1.5 V 100 64 279 3 3.84347 B0 V 95 ≤
40 268 4 4.19353 B2 V-III 90 63 292 7 4.02363 B0.2 III-II 225 50 264 7 4.64384 B0 V-III 65 46 266 6 4.79469 B0 V 100 ≤
40 276 3 4.41478 B0.7 V-III 85 64 273 13 4.41504 B0.7 V 95 60 251 13 5.06540 B0 V 85 54 247 12 4.55553 B1 V 85 50 277 7 3.96572 B1 V 115 68 249 12 4.33593 B2.5 V 70 ≤
40 292 7 3.82616 B0.5: V 100 ≤
40 227 8 4.29623 B0.2 V 110 ≤
40 269 4 4.57625 B1.5 V 65 64 310 41 4.00650 B1.5 V 60 ≤
40 287 6 3.68666 B0.5 V 95 ≤
40 255 3 4.07668 B0.7 V 90 ≤
40 270 5 4.21673 B1 V 80 ≤
40 278 2 3.92692 B0.2 V 95 ≤
40 271 3 4.42707 B0.5 V 145 ≤
40 275 10 4.61712 B1 V 120 67 255 17 4.52725 B0.7 III 120 ≤
40 286 5 4.39
Article number, page 4 of 21.L. Dufton et al.: B-type stellar nitrogen abundances in the Tarantula Nebula
Table 1. (continued.)
VFTS ST S / N v e sin i v r ∆ v r log L / L ⊙ km s −
727 B3 III 70 76 272 7 3.84740 B0.7 III 45 ≤
40 268 13 4.48748 B0.7 V 85 62 261 11 4.23772 B3-5 V-III 70 61 254 17 3.41801 B1.5 V 75 ≤
40 278 7 4.13835 B1 Ve 80 47 253 18 -851 B2 III 120 ≤
40 229 2 3.97860 B1.5 V 75 60 253 3 4.12864 B1.5 V 70 ≤
40 278 13 3.94868 B2 V 60 ≤
40 262 8 4.25872 B0 V-IV 100 77 287 6 4.30879 B3 V-III 40 ≤
40 276 16 3.77881 B0.5 III 80 ≤
40 283 3 4.28885 B1.5 V 95 69 264 9 4.26886 B1 V 65 ≤
40 258 14 3.89Palma et al. 2015; Lemasle et al. 2017). However we note that,although tests using the Galactic (7.5 dex) or SMC (6.9 dex)grids led to small changes in the atmospheric parameters (pri-marily to compensate for the di ff erent amounts of line blanket-ing), changes in element abundances estimates were typicallyless than 0.05 dex. Further information on the tlusty grids areprovided in Ryans et al. (2003) and Dufton et al. (2005)Thirteen of the targets listed in Table 1 have not been anal-ysed (VFTS 010, 044, 152, 504, 593, 625, 712, 727, 772, 801,879, 885, 886). Most of these stars had spectroscopy with rela-tively low S / N, whilst several also su ff ered from strong contami-nation by nebular emission. The former precluded the estimationof the e ff ective temperature from the silicon ionisation equilib-rium, whilst the latter impacted on the estimation of the surfacegravity. In turn this led to unreliable atmospheric parameters, andhence any estimates of (or limits on) the nitrogen abundanceswould have had limited physical significance.The four characteristic parameters of a static stellar atmo-sphere, (e ff ective temperature, surface gravity, microturbulenceand metal abundances) are inter-dependant and so an iterativeprocess was used to estimate these values. E ff ective temperatures ( T e ff ) were principally determined fromthe silicon ionisation balance. For the hotter targets, the Si iii (4552, 4567, 4574 Å) and Si iv (4089, 4116 Å) spectra were usedwhile for the cooler targets, the Si iii and Si ii (4128, 4130 Å)spectra were considered. The estimates of the microturbulencefrom the absolute silicon abundance (method 2 in Sect. 4.4)were adopted. In some cases it was not possible to measure thestrength of either the Si ii or Si iv spectrum. For these targets,upper limits were set on their equivalent widths, allowing limitsto the e ff ective temperatures to be estimated, (see, for example,Hunter et al. 2007, for more details). These normally yieldeda range of e ff ective temperatures of less than 2000K, with themidpoint being adopted.For the hotter stars, independent estimates were obtainedfrom the He ii spectrum (at 4541Å and 4686Å). There was goodagreement between the estimates obtained using the two meth-ods, with di ff erences typically being 500-1 000 K as can be seenfrom Table 2. A conservative stochastic uncertainty of ± ff ective temperature estimateslisted in Table 3. The logarithmic surface gravity (log g ; cm s − ) for each star wasestimated by comparing theoretical and observed profiles of thehydrogen Balmer lines, H γ and H δ . H α and H β were not con-sidered as they were more a ff ected by the nebular emission. Au-tomated procedures were developed to fit the theoretical spectrato the observed lines, with regions of best fit displayed by con-tour maps of log g against T e ff . Using the e ff ective temperaturesdeduced by the methods outlined above, the gravity could beestimated. Estimates derived from the two hydrogen lines nor-mally agreed to 0.1 dex and their averages are listed in Table 2.Other uncertainties may arise from, for example, errors in thenormalisation of the observed spectra and uncertainty in the linebroadening theory. For the former, tests using di ff erent contin-uum windows implied that errors in the continuum should be lessthan 1%, translating into a typical error in the logarithmic gravityof less than 0.1 dex. The uncertainty due to the latter is di ffi cultto assess and would probably mainfest itself in a systematic errorin the gravity estimates. Hence, taking these factors into consid-eration, a conservative uncertainty of ± Microturbulent velocities ( v t ) have been introduced to reconcileabundance estimates from lines of di ff erent strengths for a givenionic species (see, for example, Gray 2005). We have used theSi iii triplet (4552, 4567 and 4574 Å) because it is observed inalmost all of our spectra and all of the lines come from the samemultiplet, thereby minimising the e ff ects of errors in the absoluteoscillator strengths and non-LTE e ff ects. This approach (method1 in Table 2) has been used previously in both the Tarantulasurvey (McEvoy et al. 2015) and a previous FLAMES-I survey(Dufton et al. 2005; Hunter et al. 2007) and these authors notedits sensitivity to errors in the equivalent width measurements ofthe Si iii lines, especially when the lines lie close to the linear partof the curve of growth. Indeed this was the case for four of ourcooler targets where the observations (assuming typical errorsof 10% in the equivalent width estimates) did not constrain themicroturbulence. The method also requires that all three lines bereliably observed, which was not always the case with the cur-rent dataset. Because of these issues, the microturbulence wasalso derived by requiring that the silicon abundance should beconsistent with the LMC’s metallicity (method 2 in Table 2). Weadopt a value of 7.20 dex as found by Hunter et al. (2007) usingsimilar theoretical methods and observational data to those here;we note that this is -0.31 dex lower than the solar value given byAsplund et al. (2009).Seven targets had a maximum silicon abundance (i.e. for v t = − ) below the adopted LMC value and these are identi-fied in Table 2. These targets have a mean silicon abundanceof 6.96 ± ff ects for some of their targets, with the latterdiscussing the possible explanations in some detail. In these in-stances, we have adopted the best estimate, that is zero micro-turbulence. For other targets an estimated uncertainty of ± Article number, page 5 of 21 & Aproofs: manuscript no. 32440_final
Table 2.
Estimates of the atmospheric parameters for the VFTS sample.E ff ective temperatures ( T e ff ) are from the silicon ionisation balance (Si)or He ii profiles – those constrained by the absence of the Si ii and Si iv spectra are marked with an asterisk. Microturbulences ( v t ) are from therelative strengths of the Si iii triplet (method 1) or the absolute siliconabundance (method 2). Stars for which no solution could be found forthe micrtoturbulence are identified by daggers. VFTS T e ff (K) log g v t (km s − )Si 4686Å 4541Å (cm s − ) (1) (2)024 27500 27000 - 3.3 9 10029 27000 27000 - 4.1 0 4050 30500 30500 30000 4.0 0 1052 27500 27000 - 3.4 11 9053 24500 - - 3.5 0 4075 23500 ∗ - - 3.8 0 6095 31000 30000 - 4.4 0 0 †
111 22000 ∗ - - 3.3 - 0 †
119 27500 27000 - 4.2 0 † †
121 24000 - - 3.6 1 0124 20000 - - 3.4 - 1126 23500 ∗ - - 3.8 0 2167 27000 - - 4.0 0 0170 23000 - - 3.5 0 0183 29000 28500 28500 3.5 5 4202 23500 ∗ - - 3.9 1 0209 27500 26500 - 4.3 0 0214 29000 29500 29500 3.6 3 3237 23000 - - 3.6 0 2241 - 30500 30500 4.0 - 0242 29500 29000 29000 3.9 0 0273 19000 - - 3.8 - 5284 23500 ∗ - - 3.9 1 0297 23500 ∗ - - 3.6 2 0 †
308 23500 ∗ - - 3.8 3 0 †
313 27500 30000 29000 3.8 0 0 †
331 24000 ∗ - - 3.9 4 2347 30500 29500 30000 4.0 0 0353 23000 ∗ - - 3.5 - 2363 27000 27500 26500 3.3 10 9384 28000 27500 27500 3.3 12 9469 29500 - 29000 3.5 0 1478 25500 25500 - 3.6 8 6540 30000 30000 30000 4.0 5 1553 24500 - - 3.8 0 1572 25500 - - 4.0 0 2616 29000 27500 - 4.4 0 0 †
623 28500 28500 29000 3.9 1 3650 23500 ∗ - - 3.8 5 7666 28500 28000 - 4.0 0 0668 25000 25500 - 3.8 0 2673 25500 - - 4.0 0 1692 29500 29500 29500 3.9 0 2707 27500 27000 - 4.0 0 0725 23500 24000 - 3.5 4 6740 22500 - - 3.4 9 11748 29000 27500 - 4.1 4 5835 23500 ∗ - - 3.8 0 2851 20500 - - 3.4 - 0860 23500 ∗ - - 3.8 0 6864 27000 - - 4.1 0 1868 23000 ∗ - - 3.6 6 6872 31000 30500 30500 4.1 0 1881 26500 27000 - 3.8 4 6 in the equivalent widths, translated to variations of typically 3km s − for both methodologies.This uncertainty is consistent with the di ff erences betweenthe estimates using the two methodologies. The mean di ff erence(method 1 minus method 2) of the microturbulence is -0.6 ± − . Only in two cases do they di ff er by more than 5 km s − and then the estimates from the relative strength of the Si iii mul-tiplet appear inconsistent with those found for other stars withsimilar gravities. However, a conservative stochastic uncertaintyof ± − has been adopted for the values listed in Table 3 Adopted atmospheric parameters are summarised in Table 3.These were taken from the silicon balance for the e ff ective tem-perature apart from one target, VFTS 241, where that estimatedfrom the He ii spectra was used. Microturbulence estimates fromthe requirement of a normal silicon abundance (method 2) wereadopted.As a test of the validity of our adopted atmospheric param-eters, we have also estimated magnesium abundances (where ǫ Mg = log [Mg / H] +
12) for all our sample using the Mg ii dou-blet at 4481 Å. These are summarised in Tables 3 and have amean of 6.98 ± Our tlusty model atmosphere grids utilised a 51 level N ii ionwith 280 radiative transitions (Allende Prieto et al. 2003) and a36 level N iii ion with 184 radiative transitions (Lanz & Hubeny2003), together with the ground states of the N i and N iv ions.The predicted non-LTE e ff ects for N ii were relatively smallcompared with, for example, those for C ii (Nieva & Przybilla2006; Sigut 1996). For example, for a baseline nitrogen abun-dance, an e ff ective temperature of 25 000K, logarithmic grav-ity of 4.0 dex and a microturbulence of 5 km s − , the equivalentwidth for the N ii line at 3995Å is 35 mÅ in an LTE approxima-tion and 43 mÅ incorporating non-LTE e ff ects. In turn, analysingthe predicted non-LTE equivalent width of 43 mÅ within an LTEapproximation would lead to a change (increase) of the estimatednitrogen abundance of 0.15 dex.Nitrogen abundances (where ǫ N = log [N / H] +
12) were esti-mated primarily using the singlet transition at 3995 Å as this fea-ture was the strongest N ii line in the LR02 and LR03 wavelengthregions and appeared unblended. For stars where the singlet at3995 Å was not visible, an upper limit on its equivalent widthwas estimated from the S / N of the spectroscopy. These estimates(and upper limits) are also summarised in Table 3.Other N ii lines were generally intrinsically weaker and ad-ditionally they were more prone to blending. However we havesearched the spectra of all of our targets for lines that were inour tlusty model atmosphere grids. Convincing identificationswere obtained for only five targets (see Table 3), generallythose that appear to have enhanced nitrogen abundances. Thesetransitions are discussed below: N ii P o - D e singlet at 4447 Å: This feature typically has anequivalent width of approximately half that of the 3995 Åfeature. Additionally it is blended with an O ii line lying Article number, page 6 of 21.L. Dufton et al.: B-type stellar nitrogen abundances in the Tarantula Nebula approximately 1.3 Å to the red. This blend could be resolvedfor the stars with the lowest projected velocities and fitted usingtwo Gaussian profiles, but this inevitably a ff ected the accuracyof the line strength estimates. N ii P o - P e triplet at 4620 Å: Transitions in this multiplethave wavelengths between 4601 and 4643 Å and equivalentwidths between approximately one third and three quarters ofthat of the 3995 Å feature. This region is also rich in O ii linesleading to blending problems; for example the N ii line at 4643 Åis severely blended and could not be measured. Additionallythe strongest transition at 4630 Å is blended with a Si iv lineat 0.7 Å to the red. This feature was normally weak but couldbecome significant in our hottest targets. N ii lines near 5000 Å: The four features (at approximately4994, 5001, 5005 and 5007 Å) arise from several triplet multi-plets with the first two also being close blends. Unfortunatelythe last two were not useable due to very strong [O III ] nebularemission at 5006.84 Å, which also impacted on the N ii blend at5001 Å. This combined with the relative weakness of the featureat 4994 Å (with an equivalent width of less than one third of thatat 3995 Å) limited the usefulness of these lines. Other N ii lines: These were not used for a variety of rea-sons. The majority (at 4043, 4056, 4176, 4432, 4442, 4675,4678, 4694, 4774, 4788, 4803 Å) were not strong enoughfor their equivalent widths to be reliably estimated even inthe targets with the largest nitrogen enhancements. Others(at 4035, 4041, 4076, 4082 Å) were blended with O ii lines.Additionally the features at 4427 and 4529 Å were blended withSi iii and Al iii lines respectively. Two N ii blends from a D o - F e multiplet at approximately 4236 and 4241 Å were observedwith equivalent widths of approximately one third of that of the3995 Å feature. Unfortunately the upper level of this multipletwas not included in our model ion and hence these lines couldnot be reliably analysed.The nitrogen abundance estimates for all the N ii lines aresummarised in Table 4, with values for the line at 3995 Å beingtaken directly from Table 3. For the two stars with the largestnumber of observed N ii lines (VFTS 725 and 881), the di ff erentestimates are in reasonable agreement leading to sample stan-dard deviations of less than 0.2 dex. The N ii blend at 4601 Åyields a higher estimate in both stars, possibly due to blendingwith O ii lines. For the other three targets, the sample standarddeviations are smaller. Additionally, as can be seen from Table4, in all cases the mean values lie within 0.1 dex of that fromthe line at 3995 Å. In the discussion that follows, we will usethe abundance estimates from this line (see Table 3). However,our principle conclusions would remain unchanged if we hadadopted, when available, the values from Table 4.All these abundance estimates will be a ff ected by uncertain-ties in the atmospheric parameters which has been discussed byfor example Hunter et al. (2007) and Fraser et al. (2010). Usinga similar methodology, we estimate a typical uncertainty for anitrogen abundance estimate from uncertainties in both the at-mospheric parameters and the observational data to be 0.2-0.3dex. However due to the use of a similar methodology for all tar-gets, we would expect the uncertainty in relative nitrogen abun-dances to be smaller. This would be consistent with the relativelysmall standard deviation for the magnesium abundance estimatesfound in Sect. 4.5. Fig. 2.
Comparison of revised nitrogen abundance estimates for theFLAMES-I survey listed in Table 5 with those found by Hunter et al.(2007, 2008a, 2009) and Trundle et al. (2007b). The abscissa is our re-vised nitrogen abundance, ǫ N , whilst the ordinate is the di ff erence be-tween the revised value and that found previously. A least squares linearfit to the data is shown as a dotted line.
5. Nitrogen abundances from the FLAMES-I survey
As discussed in Sect. 2, we have re-evaluated the nitro-gen abundance estimates for the apparently single LMC tar-gets (Evans et al. 2006) in the FLAMES-I survey found byHunter et al. (2007, 2008a, 2009) and Trundle et al. (2007b).Our approach was to adopt their atmospheric parameters but touse our re-reduced spectroscopy of the N ii line at 3995 Å. Ourtarget selection followed the same criteria as adopted here, viz.B-type targets (excluding supergiants) with no evidence of bina-rity. This led to the identification of 34 targets , including twofor which it had not been possible to obtain nitrogen abundanceestimates in the original survey. The projected rotational veloc-ity, atmospheric parameters and revised nitrogen abundance es-timates from the N ii at 3995 Å are summarised in Table 5.Fig. 2 illustrates the di ff erences between our revised nitro-gen abundance estimates and those found previously. Althoughthe two sets ore in reasonable overall agreement (with a meandi ff erence of -0.06 ± ± ffi cult to identify the sources of any di ff erences. Gen-erally the smaller original abundance estimates were based onone or two lines and hence the di ff erences may reflect di ff er-ent equivalent width estimates for the N ii line at 3995Å. Asdiscussed in Sect. 2, we have re-reduced the FLAMES-I spec-troscopy and would expect that the current equivalent width es-timates will be the more reliable. The larger original nitrogenabundance estimates were based on typically 5-6 lines and anydi ff erences may in part reflect this. In Sect. 6.4, we consider whatchanges the use of the original nitrogen abundance would havemade to our discussion. Further consideration of the FLAMES spectroscopy led to twochanges compared with the binarity classification in Evans et al. (2006).NGC2004 /
91 was reclassified as apparently single and NGC2004 /
97 asa binary candidate. Additionally N11 /
101 was not analysed due to therelatively poor quality of its spectroscopy. Article number, page 7 of 21 & Aproofs: manuscript no. 32440_final
Fig. 3.
Estimates of the atmospheric parameters for the targets fromthe VFTS (upper) and FLAMES-I survey (lower); some targets havebeen moved slightly in e ff ective temperature to improve clarity. Targetswith an estimated nitrogen abundance, 7.2 ≤ ǫ N < ǫ N ≥ ff ective temperatures and gravity estimates are alsoshown.
6. Discussion
The estimates of the e ff ective temperatures and surface grav-ities for the VFTS and FLAMES-I samples are illustrated inFig. 3. They cover similar ranges in atmospheric parameters,viz. 18 000 < ∼ T e ff < ∼
32 000 K and 3.2 < ∼ log g < ∼ ff ective tem-peratures and large gravities, as these would be relatively faintand would lie below the apparent magnitude limit for observa-tion. The estimated microturbulences of the two samples are alsosimilar, covering the same range of 0–11 km s − and having me-dian and mean values of 1.5 and 2.6 km s − (VFTS) and 2.5 and3.3 km s − (FLAMES-I).The estimated projected rotational velocities of our two sam-ples lie between 0–80 km s − due to the selection criteria andin the subsequent discussion two ranges are considered, viz, v e sin i ≤
40 km s − and 40 < v e sin i ≤
80 km s − . The VFTS es- timates of Dufton et al. (2013) adopted a Fourier Transformmethodology (see, Carroll 1933; Simón-Díaz & Herrero 2007,for details), whilst the FLAMES-I measurements of Hunter et al.(2008b) were based on profile fitting of rotationally broadenedtheoretical spectra to the observations.As discussed by Simón-Díaz & Herrero (2014), both meth-ods can be subject to errors due the e ff ects of other broadeningmechanisms, including macroturbulence and microturbulence.They concluded that for O-type dwarfs and B-type supergiants ,estimates for targets with v e sin i ≤
120 km s − could be overes-timated by ∼ ±
20 km s − . Recently Simón-Díaz et al. (2017b)have discussed in detail the non-rotational broadening compo-nent in OB-type stars. They again find that their O-type stellarand B-type supergiant samples are ’dominated by stars with aremarkable non-rotational broadening component’; by contrast,in the B-type main sequence domain, the macroturbulence esti-mates are generally smaller, although the magnitude of any un-certainty in the projected rotational velocity is not quantified.However inspection of Fig.5 of Simón-Díaz et al. (2017b), im-plies that the macroturbulence will be typically 20-30 km s − inour targets compared with values of 100 km s − or more that canbe found in O-type dwarfs and OB-type supergiants.The e ff ects of these uncertainties on our samples are ex-pected to be limited for three reasons. Firstly the resultsof Simón-Díaz et al. (2017b) imply that the e ff ects of non-rotational broadening components are likely to be smaller inour sample than in the O-type stellar and B-type supergiantsdiscussed by Simón-Díaz & Herrero (2014). Secondly, we havenot used specific v e sin i estimates but rather placed targets into v e sin i bins. Thirdly for our sample and especially the cohortwith v e sin i ≤
40 km s − , the intrinsic narrowness of the metalline spectra requires a small degree of rotational broadening. InSect. 6.4, we discuss the implications of possible uncertaintiesin our v e sin i estimates. As discussed by Garland et al. (2017), the LMC baseline ni-trogen abundance has been estimated from both H ii regionsand early-type stars. For the former, Kurt & Dufour (1998) andGarnett (1999) found 6.92 and 6.90 dex respectively. Korn et al.(2002, 2005), Hunter et al. (2007) and Trundle et al. (2007b)found a range of nitrogen abundances from their analyses ofB-type stellar spectra, which they attributed to di ff erent degreesof enrichment by rotational mixing. Their lowest estimates im-plied baseline nitrogen abundances of 6.95 dex (1 star), 6.90dex (5 stars) and 6.88 dex (4 stars) respectively. These di ff erentstudies are in good agreement and a value of 6.9 dex has beenadopted here, which is -0.93 dex lower than the solar estimate(Asplund et al. 2009)The majority of targets appear to have nitrogen abundancesclose to this baseline, with approximately 60% of the 54 VFTStargets and 75% of the 34 FLAMES-I targets having enhance-ments of less than 0.3 dex. Additionally for the VFTS sample,an additional 20% of the sample have upper limits greater than7.2 dex but could again either have modest or no nitrogen en-hancements. Hence approximately three quarters of each sam-ple may have nitrogen enhancements of less than a factor oftwo. This would, in turn be consistent with them having rela-tively small rotational velocities and having evolved without anysignificant rotational mixing (see, for example, Maeder 1987;Heger & Langer 2000; Maeder 2009; Frischknecht et al. 2010).The two targets, VFTS 167 and 469 whose spectroscopy wasprobably a ff ected by fibre cross-contaminations (see Sect. 2), Article number, page 8 of 21.L. Dufton et al.: B-type stellar nitrogen abundances in the Tarantula Nebula have atmospheric parameters that are consistent with both theirspectral types and other VFTS targets. Additionally they appearto have near baseline nitrogen abundances. Although these re-sults should be treated with some caution, it appears unlikelythat they belong to the subset of targets with large nitrogen en-hancements discussed below.Two targets, VFTS 237 and 835, have been identified as Be-type stars and might be expected to show the e ff ects of rota-tional mixing due to rotational velocities near to critical veloci-ties having been inferred (Townsend et al. 2004; Cranmer 2005;Ahmed & Sigut 2017) for such objects. However both have nearbaseline nitrogen abundance estimates (6.92 and 6.90 dex re-spectively) implying that little rotational mixing has occurred.Our analysis did not make any allowance for the e ff ects of lightfrom a circumstellar disc. Dunstall et al. (2011) analysed spec-troscopy of 30 Be-type stars in the Magellanic Clouds and esti-mated a wide range of disc contributions, ∼ ∼ ∼ ∼ − would have a value of v e sin i ≤
40 km s − only about 1%of the time. Additionally the probability that such a star wouldbe observed with 40 < v e sin i ≤
80 km s − is approximately fourtimes that of it being observed with v e sin i ≤
40 km s − . Similarratios would be found for other values of the rotational velocityand just reflects the greater solid angle of inclination availableto stars in the larger projected rotational velocity bin. By con-trast, the nitrogen rich targets in our samples are more likely tohave v e sin i ≤
40 km s − than 40 < v e sin i ≤
80 km s − , as can beseen from Table 7. In Sect. 6.3, we return to these objects andundertake simulations to investigate whether all the nitrogen en-hanced stars in our sample could indeed be fast rotators observedat small angles of inclination.We have investigated the other observed characteristics ofour nitrogen enhanced stars to assess whether they di ff er fromthose of the rest of the sample. Fig. 4 shows the spatial distri-bution of all of our VFTS targets, together with the position andextent of the larger clusters NGC 2070 and NGC 2060 and thesmaller clusters, Hodge 301 and SL639; the sizes of the clus-ters should be treated as representative and follow the definitionsof Evans et al. (2015, Table 4). The nitrogen enriched targetsare distributed across the 30 Doradus region with no evidencefor clustering within any cluster. Adopting the cluster extents ofEvans et al. (2015), there are 21 cluster and 33 field VFTS stars.All four targets with 7.2 ≤ ǫ N < ǫ N ≥ ± − (54 targets), whilst the 10 nitrogen enriched stars and the re-maining 44 stars have values of 270 . ± . − and 270 . ± . − respectively. The latter two means are very similarwhilst an F-test of the corresponding variances showed that theywere not significantly di ff erent at even the 20% level. Hence weconclude that the radial velocity distributions of the two sub-groups appear similar within the limitation of the sample sizes.A comparison of the atmospheric parameters also reveals nosignificant di ff erences. For example, the whole VFTS samplehave a median and mean e ff ective temperature of 27 000 and25 865 K, while the values for the targets with ǫ N ≥ / / / − (wholesample) and 2.0 / − (nitrogen enriched).A helium to hydrogen abundance ratio by number of 0.1 wasassumed in our calculations of theoretical spectra (see Sect. 4.1).The LMC evolutionary models of Brott et al. (2011a) appropri-ate to B-type stars indicate that even for a nitrogen enhancementof 1.0 dex (which is the largest enhancement observed in oursample – see Tables 3 and 5), the change in the helium abundanceis typically only 0.03 dex. Hence our assumption of a normal he-lium abundance is unlikely to be a significant source of error. Wehave checked this by comparing theoretical and observed He i profiles for targets with the largest nitrogen enhancements. Thiscomparison was impacted by significant nebula emission espe-cially for lines in the (3d) D-(4f) F series but generally yieldedgood agreement between observation and theory for our adoptedhelium abundance.This is illustrated by observed and theoretical spectra of thewavelength region 3980-4035 Å shown in Fig. 5 for VFTS 650and 725, which have similar atmospheric parameters. The ob-served spectra have been corrected using the radial velocityestimates listed in Table 1, while the theoretical spectra havebeen interpolated from our grid to be consistent with the at-mospheric parameters and nitrogen abundances listed in Table3. The latter have also been convolved with a Gaussian func-tion to allow for instrumental broadening; as both stars have es-timated projected rotational velocities, ≤
40 km s − , no correc-tion was made for rotational broadening. For the two di ff use He i lines at approximately 4009 and 4026 Å, the agreement is excel-lent although VFTS 650 has a near baseline nitrogen abundancewhilst VFTS 725 has the second highest nitrogen abundance inthe VFTS sample. The agreement between theory and observa-tion for the N ii at 3995 Å is also good and supports our use ofequivalent widths to estimate element abundances. The observedspectra also illustrate the range of observed S / N with VFTS 650(S / N ∼
60) and 725 (S / N ∼ / C and N / Oabundances would provide a powerful diagnostic of rotationalmixing. The LMC models of Brott et al. (2011a) appropriate toB-type stars indicate that for a nitrogen enhancement of ∼ ii line in the blue spectra ofB-type stars is the doublet at 4267 Å, which is badly a ff ectedby nebular emission. Due to the sky and stellar fibres being Article number, page 9 of 21 & Aproofs: manuscript no. 32440_final
Fig. 4.
Spatial distribution of the VFTS targets in 30 Doradus. Targets with 7.2 ≤ ǫ N < ǫ N ≥ spatially separated, it was not possible to reliably remove thisemission from the VFTS spectroscopy. Indeed as discussed byMcEvoy et al. (2015) this can lead to the reduced spectra havingspurious narrow emission or absorption features. This is espe-cially serious for targets with small projected rotational veloci-ties, where the width of the nebular emission is comparable withthat of the stellar absorption lines. No other C ii lines were re-liably observed in the VFTS spectra and hence reliable carbonabundances could not be deduced. The FLAMES-I spectroscopymay also be a ff ected by nebular emission although this might beexpected to be smaller especially in the older cluster, NGC2004.However without a reliable method to remove nebular emission,the approach advocated by Maeder et al. (2014) would appear to be of limited utility for multi-object fibre spectroscopy of low v e sin i early-type stars.Both the VFTS and FLAMES-I data have rich O ii spectra.However the predicted changes in the oxygen abundance (of lessthan 0.1 dex) are too small to provide a useful constraint, giventhe quality of the observational data. This is confirmed by theoriginal FLAMES-I analyses (Hunter et al. 2007, 2008a, 2009;Trundle et al. 2007b), where for the targets considered here anoxygen abundance of 8.38 ± Article number, page 10 of 21.L. Dufton et al.: B-type stellar nitrogen abundances in the Tarantula Nebula
Fig. 5.
Observed and theoretical spectra for the wavelength region near4000 Å in two stars, VFTS 650 and 725. Atmospheric parameters andnitrogen abundances were taken from Table 3 for the theoretical spec-tra. The normalised spectra for VFTS 725 have been shifted upwards toimprove clarity enhancement of more than 0.6 dex, the mean oxygen abundancewas 8.40 ± As discussed above, it is unclear whether the targets withthe largest nitrogen enhancements in our samples could haveevolved as single stars undergoing rotational mixing. We havetherefore undertaken simulations to estimate the number of ro-tationally mixed stars that we might expect to find in our sam-ple. We have considered two thresholds for nitrogen enrichment,viz. 0.3 dex (corresponding to a nitrogen abundance, ǫ N > ǫ N > ǫ N , of up to 8.3 dex correspondingto an enhancement of 1.4 dex; however these stars have prob-ably evolved from the O-type main sequence. The number ofstars fulfilling these criteria in both the VFTS and FLAMES-Isamples are summarised in Table 7.We have used the LMC grid of models of Brott et al. (2011a)to estimate the initial stellar rotational velocities, v i , that wouldbe required to obtain such enrichments. The core hydrogen burn-ing phase, where the surface gravity, log g ≥ min-imum rotational velocity, v e , that a star would have during thisevolutionary phase; this velocity was normally found at a grav-ity, log g ≃ ⊙ as the ZAMS e ff ective temperatures in this evolutionary phaserange from approximately 20 000 to 33 000K, compatible withour estimated e ff ective temperatures, again listed in Tables 3 and5. For each initial mass, we have quadratically interpolated be-tween models with di ff erent initial ( v i ) rotational velocities toestimate both the initial ( v i ) and the minimum rotational veloc-ity ( v e - hereafter designated as ‘the current rotational velocity’) that would achieve a nitrogen enhancement of at least either 0.3or 0.6 dex whilst the surface gravity, log g ≥ − (or less),we would expect that the interpolation would introduce errorsof less than 20 km s − . The ranges of the estimated initial andcurrent velocities for di ff erent masses are approximately 20%and 15% respectively. For our simulations, we have taken a con-servative approach and adopted velocities (see Table 6) at thelower ends of our estimates. In turn this will probably lead toan overestimate in the number of targets with enhanced nitrogenabundances and small projected rotational velocities in our sim-ulations. This overestimation will be re-enforced by the adoptionof a lower gravity limit, log g ≥ g ≥ . i , is: P ( i ) di = sin i di (1)the probability of observing an angle of inclination, i ≤ i is thengiven by: P ( i ≤ i ) = − cos i (2)For a sample of N targets with a normalised rotational velocitydistribution, f ( v e ), the number that will have a projected rota-tional velocity less than v ′ is given by: N ( v e sin i ≤ v ′ ) = N "Z v ′ f ( v e ) d v e + Z ∞ v ′ f ( v e ) P ( i ≤ i ) d v e (3)or N ( v e sin i ≤ v ′ ) = N "Z v ′ f ( v e ) d v e + Z ∞ v ′ f ( v e ) (1 − cos i ) d v e (4)where i = sin − v ′ v e ! (5)The number of targets that have both v e sin i ≤ v ′ and a nitro-gen enhancement, ǫ N ≥ ǫ , can then be estimated by changing thelower limits of integration in equation 4 to the rotational veloci-ties, v e , summarised in Table 6. This will normally eliminate thefirst term on the right hand side of equation 4 leading to: N ( v e sin i ≤ v ′ ; ǫ ≥ ǫ ) = N Z ∞ v e f ( v e ) [1 − cos i ] d v e (6)The rotational velocity distribution, f ( v e ), for the VFTS B-type single star sample has been deduced by Dufton et al. (2013,see their Table 6). Hunter et al. (2008b) assumed a single Gaus-sian distribution for the rotational velocities in the FLAMES-Isingle B-type star sample and fitted this to the observed v e sin i distribution. We have taken their observed values and used thedeconvolution methodology of Lucy (1974) as described inDufton et al. (2013) to estimate f ( v e ) for the FLAMES-I sam-ple. This is shown in Fig. 6 together with that for the VFTS sam-ple (Dufton et al. 2013). They both show a double peaked struc-ture although there are di ff erences in the actual distributions. Thenumber of v e sin i estimates in the FLAMES-I sample was rela-tively small (73) and may have some additional selection e ff ects Article number, page 11 of 21 & Aproofs: manuscript no. 32440_final
Fig. 6.
The de-convolved rotational velocity distribution, f ( v e ), forthe LMC FLAMES-I sample of single B-type targets with luminosityclasses V to III (black line) compared with that for the equivalent VFTSsample taken from Dufton et al. (2013, red line). compared with the VFTS sample as discussed in Sect. 6.4. Wehave therefore adopted the f ( v e ) from the VFTS B-type singlestar sample in most of our simulations but discuss this choicefurther below.Evans et al. (2015) classified 434 VFTS targets as B-type and subsequently Dunstall et al. (2015) found 141 tar-gets (including 40 supergiants) to be radial velocity variables.McEvoy et al. (2015) identified an additional 33 targets as sin-gle B-type supergiants, leading to 260 remaining targets. How-ever the relatively low data quality of five of these targets pre-vented Dufton et al. (2013) estimating projected rotational ve-locities. Hence we have adopted N =
255 for the number ofB-type VFTS targets which have luminosity classes V to III andhave estimated projected rotational velocities. For the FLAMES-I sample, Evans et al. (2006) has provided spectral types includ-ing targets classified as Be-Fe without a luminosity classifica-tion. As they are likely to be lower luminosity objects, they wereincluded when deducing a value of N = N targets wererandomly assigned both rotational velocities (using the rotationalvelocity distribution, f ( v e ) discussed above) and angles of incli-nation (using equation 1). The number of targets with di ff erentnitrogen enrichments and in di ff erent projected rotational binswere then found using the rotational velocity limits in Table 6.Our simulations are all based on the assumption that the ro-tational axes of our targets are randomly orientated. As can beseen from Fig. 4, approximately half our targets are situatedin four di ff erent clusters (with di ff erent ages, see, for example,Walborn & Blades 1997; Evans et al. 2011, 2015) with the re-mainder spread across the field. Such a diversity of locationswould argue against them having any axial alignment. How-ever Corsaro et al. (2017) identified such alignments in two oldGalactic open clusters, although previously Jackson & Je ff ries(2010) found no such evidence in two younger Galactic clusters.If the VFTS axes were aligned, the identification by Dufton et al.(2013) of B-type stars with projected rotational velocities, v e sin i >
400 km s − (which is close to their estimated criticalvelocities Huang et al. 2010; Townsend et al. 2004), would thenimply a value for the inclination axis, sin i ≃
1. In turn this would lead to all our targets having small rotational velocities therebyexacerbating the inconsistencies illustrated in Table 7.
The numerical integrations and the Monte Carlo simulationsled to results that are indistinguishable, providing reassurancethat the methodologies were correctly implemented. The Monte-Carlo simulations are summarised in Table 7, their standard er-rors being consistent with Poisson statistics which would be ex-pected given the low probability of observing stars at low anglesof inclination. For both samples, the simulations appear to sig-nificantly underestimate the number of nitrogen enriched targetswith v e sin i ≤
40 km s − and possibly overestimate those with40 < v e sin i ≤
80 km s − .We have therefore used the Monte-Carlo simulations to findthe probabilities of observing the number of nitrogen enrichedstars summarised in Table 7. When the observed number, n wasgreater than predicted, we calculated the probability that thenumber observed would be greater than or equal to n . Con-versely, when the observed number, n was less than predicted,we calculated the probability that the number observed would beless than or equal to n . These probabilities are also listed in Ta-ble 7 and again very similar probabilities would have been foundfrom adopting Poisson statistics. All the di ff erences for the tar-gets with v e sin i ≤
40 km s − are significant at the 5% level. Bycontrast for the samples with 40 < v e sin i ≤
80 km s − , none aresignificant at the 5% level.For the VFTS Monte-Carlo simulations, the choice of N =
255 leads to 36 ± ± v e sin i ≤
40 km s − and40 < v e sin i ≤
80 km s − respectively. As expected these are ingood agreement with the 37 and 30 targets in the original samplein Table 1. However thirteen of these targets were not analysed,reducing the sample summarised in Table 3 to 31 ( v e sin i ≤ − ) and 23 (40 < v e sin i ≤
80 km s − ) targets. We have there-fore repeated the Monte Carlo simulations using values of N (220 and 188 respectively) that reproduce the number of targetsactually analysed. These simulations are also summarised in Ta-ble 7 and again lead to an overabundance of observed nitrogenenriched targets with v e sin i ≤
40 km s − that is now significantat 1% level. By contrast the simulations are consistent with theobservations for the cohort with 40 < v e sin i ≤
80 km s − .Even after allowing for the targets that were not analysed,the discrepancy between the observed number of VFTS targetswith v e sin i ≤
40 km s − and the simulations may be larger thanpredicted. As discussed above, both the conservative current ro-tational velocities, v e , adopted in Table 6 and the decision to fol-low the stellar evolution to a logarithmic gravity of 3.3 dex willlead to the simulations being e ff ectively upper limits. Addition-ally, for the VFTS subsample with v e sin i ≤
40 km s − , there areseven targets, where the upper limit on the nitrogen abundancewas consistent with ǫ N > ǫ N > N = v e sin i ≤ − is greater than predicted for both ǫ N ≥ ǫ N ≥ ff er-ence is significant at the 1% level. By contrast the cohort with40 < v e sin i ≤
80 km s − is compatible with the simulations. There-evaluation of the nitrogen abundance estimates in Sect. 4.6 led Article number, page 12 of 21.L. Dufton et al.: B-type stellar nitrogen abundances in the Tarantula Nebula to a decrease in the number of targets with ǫ N > /
43, 84, 86 and 91) had v e sin i ≤ − and adopting the previous nitrogen abundance estimateswould have further increased the discrepancy with the simula-tions for this cohort. The re-evaluation also moved one targetinto the ǫ N > / v e sin i =
46 km s − and hence this change wouldnot a ff ect the low probabilities found for the nitrogen enhancedtargets with v e sin i ≤
40 km s − .The FLAMES-I results should however be treated with somecaution. Firstly, although there were multi-epoch observationsfor the this sample, the detection of radial-velocity variables (bi-naries) was less rigorous than in the VFTS observations; thisis discussed further in Sect. 6.6.4. Secondly, the exclusion ofa number of Be-type stars in the original target selection forNGC 2004 may have led to this sample being biased to stars withlower values of the rotational velocity (see Brott et al. 2011b, fora detailed discussion of the biases in this sample). Hence, theuse of the VFTS rotational-velocity distribution, f ( v e ), may notbe appropriate for the FLAMES-I sample. There is some evi-dence for this in that the observed number of FLAMES-I tar-gets (18 with v e sin i ≤
40 km s − and 16 with 40 < v e sin i ≤ − ) is larger than that predicted (14.5 ± ± f ( v e ) distributionfor the FLAMES-I sample shown in Fig. 6 being weighted moreto smaller rotational velocities than the VFTS distribution thatwas adopted. We have therefore also used the f ( v e ) distribution,together with the number of v e sin i estimates, N =
73 deducedfor the FLAMES-I sample in the Monte Carlo simulation and theresults are again summarised in Table 7. This leads to even lowerprobabilities for the observed number of nitrogen enhanced tar-gets with v e sin i ≤
40 km s − being consistent with the simula-tions, whilst the probabilities for the 40 < v e sin i ≤
80 km s − co-hort are increased. The predicted number of targets in the two v e sin i ranges (15.5 ± ± v e sin i estimatesmay be too high. Simón-Díaz & Herrero (2014) estimated thatthis might be typically 25 ±
20 km s − for O-type stars and B-type supergiants; the e ff ects on our samples of B-type dwarfsand giants are likely to be smaller given their smaller macro-turbulences (Simón-Díaz et al. 2017b). However we have con-sidered the e ff ect of arbitrarily decreasing all our v e sin i es-timates by 20 km s − . This would move eleven VFTS targetsand seven FLAMES-I targets into the v e sin i ≤
40 km s − bins.Similar numbers of targets would also be moved into the40 < v e sin i ≤
80 km s − bins from higher projected rotational ve-locities but as these targets have not been analysed, we are un-able to quantitatively investigate the consequences for this bin.However we note that the simulation showed that the predictednumber of targets in the 40 < v e sin i ≤
80 km s − bin were con-sistent with that observed and it is likely that this result wouldremain unchanged.For the cohorts with v e sin i ≤
40 km s − , the numbers wouldbe increased to 48 (VFTS) and 24 (FLAMES-I). However thenumber of targets with ǫ N > v e sin i ≤
40 km s − As the observed number of targets for the 40 < v e sin i ≤
80 km s − cohort is now larger than that predicted by the simulations, the proba-bility listed in Table 7 is for the observed number of targets (or a greaternumber) being observed. by increasing the B-type samples to N =
335 (VFTS) and 170(FLAMES-I). These simulations again lead to small probabil-ities of observing the relatively large numbers of nitrogen en-hanced targets, viz <
2% (VFTS) and < .
1% (FLAMES-I).This arbitrary decrease in the v e sin i estimates would also a ff ectthe rotational velocity function, f ( v e ), used in the simulations. Ifapplied universally, this would shift the distribution to smallerrotational velocities, thereby decreasing the predicted numberof nitrogen enhanced targets. Additionally either allowing forVFTS targets without nitrogen abundance estimates or the useof the FLAMES-I rotational velocity distribution shown in Fig.6 would also decrease the predicted numbers and hence proba-bilities.Inspection of the metal absorption lines in the spectra forthe targets moved into the v e sin i ≤
40 km s − bin showed thatthey generally had the bell shaped profile characteristic of rota-tional broadening. Hence we believe that the arbitrary decreaseof 20 km s − probably overestimates the magnitude of any sys-tematic errors. As such, we conclude that uncertainties in the v e sin i estimates are very unlikely to provide an explanation forthe discrepancy between the observed and predicted numbers ofnitrogen enriched targets in the cohort with v e sin i ≤
40 km s − .In summary, the simulations presented in Table 7 are con-sistent with the nitrogen enriched targets for both the VFTS andFLAMES-I samples with 40 < v e sin i ≤
80 km s − having largerotational velocities (and small angle of incidence) leading tosignificant rotational mixing. By contrast, there would appearto be too many nitrogen enriched targets in both samples with v e sin i ≤
40 km s − to be accounted for by this mechanism. The simulations presented in Sect. 6.4 imply that there is an ex-cess of targets with v e sin i ≤
40 km s − and enhanced nitrogenabundances. In the VFTS sample, there are two objects with anitrogen abundance, 7.2 ≤ ǫ N < ± ǫ N ≥ < v e sin i ≤
80 km s − fur-ther implies that these objects have rotational velocities, v e ≤ − .In the subsequent discussion, we will consider two percent-ages, viz. firstly the percentage, P , of stars within the cohortwith v e ≤
40 km s − that have enhanced nitrogen abundances thatdo not appear to be due to rotational mixing and secondly thepercentage, P T , of such stars within the total population of singlestars with luminosity classes III-V. Our simulations indicate that67% of the 31 VFTS targets with v e sin i ≤
40 km s − will alsohave v e ≤
40 km s − , equating to ∼
21 targets. From the simula-tions summarised in Table 7, the excess of VFTS objects with ǫ N ≥ v e sin i ≤
40 km s − is 3.55-3.75 depending onthe value of N adopted. This would then imply a percentage ofsuch objects, P ∼ v e ≤ − , whose nitrogen abundancecannot be constrained to less than 7.5 dex and excluding thesewould increase the percentage to P ∼ ∼ Article number, page 13 of 21 & Aproofs: manuscript no. 32440_final now being required for stars whose nitrogen abundance couldnot be constrained.These percentages then translate in to percentages with re-spect to the whole population B-type stars of P T ∼ v e sin i ≤
50 km s − , to that adopted here.Secondly they defined a significant nitrogen enrichment as ǫ N ≥ ff ective temperatures T e ff ≤
35 000 K and gravities log g ≥ ≤
50 km s − but witha rotational velocity su ffi cient to produce the observed nitrogenabundance by rotational mixing. Finally they adopted for theirLMC non-binary core hydrogen burning sample the 73 appar-ently single targets that had estimates for their projected rota-tional velocity. However as discussed in Sect. 6.4, Evans et al.(2005) lists 103 B-type non-supergiant stars with no evidencefor binarity and we have adopted this larger sample size in oursimulations.Hunter et al. (2008a) identified 17 targets (three of whichwere binaries) that had nitrogen enhancements and small pro-jected rotational velocities. Of the single targets two targets hadspectral types of O9.5 III or O9.5 V, whilst an additional tar-get was a B0.5 Ia supergiant. Removing these six objects (in or-der to maintain consistency with the current analysis) would re-duced the sample size to 11 objects. A simulation similar to thosedescribed in Sect. 6.3 implied that 1.48 ± N =
103 would then lead to a percentage, P T ∼
9% of FLAMES-IB-type targets with luminosity classes III-V, v e sin i ≤
50 km s − , ǫ N ≥ T ∼ v e sin i ≤
40 km s − and ǫ N ≥ ⊙ . The models of Brott et al. (2011a) with an initial massof 12 M ⊙ then imply that to obtain a nitrogen enhancement of 0.2dex as a star evolves to this gravity would require a current ro-tational velocity of 150 km s − . We note that this velocity is notdirectly comparable with those in Table 6 as we are now consid-ering median gravities rather than those found at the end of thehydrogen core-burning phase.The number of targets in our sample of N =
255 predictedby Monte Carlo simulations would then be 2.16 ± T ∼ ∼ T < ∼ .
2% and P < ∼ N = ± T ∼ ∼ v e ≤
40 km s − and ǫ N ≥ T ∼ ∼ T ∼ ∼ Below we discuss possible explanations for the origin of the tar-gets discussed in Sect. 6.5 – hereafter referred to as ’nitrogenenhanced’.
The estimated atmospheric parameters for the VFTS andFLAMES-I samples are shown in Fig. 3, together with the pre-dictions of the evolutionary models of Brott et al. (2011a). Thelatter are for e ff ectively zero initial rotational velocity and werechosen to be consistent with the observed low projected rota-tional velocities of our targets. However, in this part of the HRdiagram, the evolutionary tracks and isochrones are relatively in-sensitive to the choice of initial rotational velocity as can be seenfrom Figs. 5 and 7 of Brott et al. (2011a).Also shown in Fig. 3 are the Geneva tracks and isochronesfor an LMC metallicity and zero rotational velocity extracted orinterpolated from the Geneva stellar models database for thegrid of models discussed by Georgy et al. (2013b). These mod-els had a maximum initial mass of 15 M ⊙ , so no track for 19 M ⊙ is shown, whilst the lower age isochrones are also truncated. Ascan be seen from Fig. 3, the two sets of tracks and isochronesare in reasonable agreement but show some di ff erences. In par-ticular the end of the terminal age main sequence occurs at dif-ferent positions on the two sets of tracks. These di ff erences havepreviously been discussed by Castro et al. (2014), who ascribedthem to the di ff erent calibrations of the convective core over-shooting parameter. However in the context of the current anal-ysis, the two grids would lead to similar estimates of, for exam-ple, stellar masses and ages within the estimated uncertainties inthe atmospheric parameters. For both samples, there is some ev-idence that the nitrogen enhanced targets may lie in the higher T e ff regions of the diagrams. However Student t-tests and Mann-Whitney U-tests returned no statistics that were significant at the10% level for their being any di ff erences in the e ff ective temper-atures distributions of the parent populations.Brott et al. (2011b) have previously discussed the massesof the B-type FLAMES-I targets based on the analyses ofHunter et al. (2007, 2008a, 2009) and Trundle et al. (2007b). The Geneva database is available at
Article number, page 14 of 21.L. Dufton et al.: B-type stellar nitrogen abundances in the Tarantula Nebula
They noted that the nitrogen enhanced slowly spinning stars ap-peared to have higher masses. More recently Grin et al. (2017)have estimated nitrogen abundances for O-type giants and su-pergiants from the VFTS and again find that the targets withsmall v e sin i and enhanced nitrogen abundance have higher es-timated masses than the rest of their sample. However it shouldbe noted that these two analyses sample di ff erent mass rangeswith the B-type stars having estimated masses in the range ∼ ⊙ (see Fig. 3), whilst those of Grin et al. (2017) normallyhave estimated masses >
20 M ⊙ .Following Grin et al. (2017), we have used B onnsai to es-timate the initial and current evolutionary masses of the starsin our VFTS and FLAMES-I samples with v e sin i ≤
40 km s − .B onnsai uses a Bayesian methodology and the grids of modelsof Brott et al. (2011a) to constrain the evolutionary status of agiven star, including its age and mass (see Schneider et al. 2014,for details). As independent prior functions, we adopted theLMC metallicity grid of models, a Salpeter (1955) initial massfunction, the initial rotational velocity distribution estimated byDufton et al. (2013), a random orientiation of spin axes, and auniform age distribution. The estimates of T e ff and log g (takenfrom Tables 3 and 4), together with a v e sin i ≤
40 km s − werethen used to estimate masses. For these relatively unevolvedstars, the predicted current and initial masses were very similarwith di ff erences < ff erence was 8%.For both samples the nitrogen enriched stars had higher meanmasses, confirming the previous comments for the FLAMES-Isample by Brott et al. (2011b). For the VFTS sample, the ni-trogen enriched targets had a median current mass of 14 M ⊙ and a mean current mass of 13.6 ± ⊙ compared with 11.2and 11.8 ± ⊙ for the other stars. For the FLAMES-I sample,the corresponding values were 12.5 and 12.7 ± ⊙ comparedwith 10.4 and 11.8 ± ⊙ . However these di ff erences are rela-tively small especially given the sample sizes for the nitrogen en-riched objects. This was confirmed by Student t-tests and Mann-Whitney U-tests that returned no statistics that the two parentpopulations were di ff erent at the 5% level.The B onnsai analysis also provided estimates for the stel-lar luminosity, lifetime and current surface nitrogen abundance.There is good agreement between the B onnsai luminosities andthe observed luminosities estimated in Sect. 2 and listed in Table1. The mean di ff erence is − . ± .
22 dex with the estimatesfor over ∼
85% of the targets showing a di ff erence of less than0.3 dex. The major discrepancy is for VFTS 469, where B onn - sai estimates a luminosity 0.6 dex (i.e. a factor of four) greaterthan the observed luminosity. As discussed in Sect. 2, the spec-troscopy of this star may be compromised by contamination bylight from a bright O-type star in an adjacent fibre. Excludingthis star would lead to a mean di ff erence of 0 . ± .
19 dex.For the five nitrogen enhanced stars, the agreement between thetwo sets of luminosities is better, with the mean di ff erence being − . ± .
06 dex.In Fig. 7, the observed luminosities estimated in Sect. 2 areshown as a function of e ff ective temperature and imply that thenitrogen enhanced stars may have larger luminosities. Indeedthese stars have both higher mean observed (4.45 ± onnsai (4.47 ± ± ± ff erences in the masses discussed above. Studentt-tests and Mann-Whitney U-tests returned probabilities that the The B onnsai web-service is available at . Fig. 7.
Observed luminosities for the VFTS sample as a function ofe ff ective temperature. Targets with an estimated nitrogen abundance,7.2 ≤ ǫ N < ǫ N ≥ ff ective temperatures and luminosityestimates. two parent populations were the same at the ∼ ∼ onnsai ranged from ∼ ± ± ff erences werefound for the FLAMES-I sample, viz. nitrogen enhanced stars:9.3 and 9.0 ± ± ff erences were notsignificant at a 5% level.The nitrogen abundances predicted by B onnsai ranged from6.89–6.91 dex – in e ff ect implying no nitrogen enhancementfrom the initial value of 6.89 dex adopted in the models ofBrott et al. (2011a). Additionally the predicted initial and cur-rent rotational velocities were small and less than or equal to40 km s − . This is consistent with the small v e sin i ( ≤
40 km s − )estimates of these samples and the low probability of observ-ing targets at small angles of incidence. Indeed it qualitativelyre-enforces the results presented in Sect. 6.4, which found thatthe numbers of observed nitrogen enhanced targets were signifi-cantly larger than predicted by our simulations.For the nitrogen enhanced objects, B onnsai was re-run in-cluding the estimated nitrogen abundance as a further constraintfor all nine (five VFTS and four FLAMES-I) targets. In all cases,the best fitting model had both high initial ( v i ) and current ( v e ) ro-tational velocities, which were consistent with the adopted limitsgiven in Table 6 to achieve significant nitrogen enhancements byrotational mixing. However for four targets (VFTS 095 and 881,NGC2004 /
53 and N11 / onnsai solutions failed boththe posterior predictive check and a χ -test. This di ffi culty infinding single star evolutionary models that match the observedparameters of our nitrogen enhanced targets again re-enforcesthe inconsistencies discussed in Sect. 6.3. Article number, page 15 of 21 & Aproofs: manuscript no. 32440_final
To summarise, there is some evidence that the nitrogen en-riched targets with v e sin i ≤
40 km s − may have higher masses,e ff ective temperatures and luminosities, together with smallerlifetimes, assuming that they have evolved as single stars. Thehigher mass estimates are consistent with a previous popula-tion synthesis for the FLAMES-I B-type stars by Brott et al.(2011b) and the analysis of the evolved VFTS O-type targetsby Grin et al. (2017). The discrepancy between the observed and predicted numbersof nitrogen enhanced targets discussed in Sect. 6.4 could be dueto the physical assumptions adopted in the stellar evolutionarymodels. For example, as discussed by Grin et al. (2017), the e ffi -ciency of rotational mixing in the single-star evolutionary mod-els of Brott et al. (2011a) was calibrated using e ffi ciency factorsbased on the B-type dwarfs in the FLAMES-I survey but ex-cluding nitrogen enhanced targets with low projected rotationalvelocities. Hence an increase in mixing e ffi ciencies could in prin-ciple explain our nitrogen enriched targets. For example the fiveVFTS targets with nitrogen enhancements of greater than 0.6 dexand v e sin i ≤
40 km s − (see Table 7) could be reproduced by re-ducing the current rotational velocity, v e for such an enhance-ment, to the unrealistically low value of ∼
70 km s − . Howeverthis would also increase the number of nitrogen enriched tar-gets in the 40 < v e sin i ≤
80 km s − cohort to ∼
20 far higher thanis observed. Similar arguments apply to the FLAMES-I samplewhere the fraction of nitrogen enhanced targets is even larger. In-deed the ratios of nitrogen enriched targets in the two projectedrotational velocity cohorts provides strong evidence against theobserved nitrogen enriched stars being rotationally mixed starsobserved at small angles of inclination.Another possibility is our nitrogen enriched targets may havehad severe stripping of their envelopes, thereby revealing chem-ically enriched layers. The associated mass loss could also haveremoved angular momentum from the surface, resulting in spin-down. However the predicted mass loss for our targets from theB onnsai simulations or from the grid of models of Georgy et al.(2013b) is very small (typically less than 5% of the initial mass)making such an explanation implausible.In summary, there is no evidence that limitations on the phys-ical assumptions made in the stellar evolutionary models canexplain our nitrogen enhanced targets. Indeed the similarity oftheir physical properties to those of the other targets (discussedin Sect. 6.6.1) argues against such an explanation.
The models discussed in Sect. 6.6.1 and 6.6.2 assume that mag-netic fields are unimportant for a star’s evolution. Such fieldshave been previously suggested as an explanation for nitro-gen enriched, slowly rotating Galactic early-B stars (Morel et al.2008; Morel 2012; Przybilla & Nieva 2011). However, more re-cently, Aerts et al. (2014) found no significant correlation be-tween nitrogen abundances and magnetic field strengths in a sta-tistical analysis of 68 Galactic O-type stars that addressed theincomplete and truncated nature of the observational data.Meynet et al. (2011) discussed the possibility of magneticbraking (see Ud-Doula et al. 2009, and references therein) dur-ing the main-sequence phase. Models with di ff erential rotationand magnetic braking produced strongly mixed stars with low surface rotational velocities. By contrast models with solid-bodyrotation and magnetic braking produce stars that at the termi-nal age main sequence had low surface rotational velocities ande ff ectively no changes in the surface abundances. Potter et al.(2012a,b) have considered the α − Ω dynamo as a mechanismfor driving the generation of large-scale magnetic flux. Theyfound that this mechanism was e ff ective for stellar masses up to ∼
15 M ⊙ (which includes e ff ectively all the stars in our samples).B-type main sequence stars with su ffi ciently high rotation rateswere found to develop an active dynamo and so exhibited strongmagnetic fields. They then were spun down quickly by magneticbraking and magneto-rotational turbulence (Spruit 2002) leadingto enhanced surface nitrogen abundances.Both observational and theoretical investigations have there-fore led to inconclusive results about whether magnetic fieldscould enhance stellar mixing. However it is an attractive mech-anism for explaining our nitrogen enhanced stars as it cancombine both high nitrogen abundances and low stellar ro-tation velocities. The incidence of magnetic fields in early-type stars is currently being investigated by the MiMes survey(Wade et al. 2014, 2016) and by the BOB survey (Morel et al.2014, 2015). The MiMes survey had a number of criteria forselecting their survey sample, one of which favoured stars with v e sin i ≤
150 km s − . They found a frequency of 7 ±
1% for mag-netic fields in 430 B-type stars (Wade et al. 2014) with a similarfrequency for O-type stars (Grunhut et al. 2017). The BOB sur-vey have found a detection rate of 6 ±
4% for a sample of 50 OB-type stars, while by considering only the apparently slow rotatorsthey derived a detection rate of 8 ±
5% (Fossati et al. 2015b). Ad-ditional observations for 28 targets led to revised estimates of6 ±
3% and 5 ±
5% (Schöller et al. 2017).These detection frequencies appear to be consistent with thepercentages of nitrogen enriched targets estimated for the VFTSand FLAMES-I surveys (see Sect. 6.5). However the two mag-netic surveys have necessarily only observed Galactic targets andmay not be directly applicable to an LMC metallicity. Also bothsurveys preferentially detect targets with large magnetic fields,typically >
100 Gauss (Morel et al. 2015). Fossati et al. (2015a)on the basis of intensive observations of two very bright stars β CMa and ǫ CMa detected relatively weak magnetic fieldsand concluded that such fields might be more common in mas-sive stars than currently observed. However the simulations ofMeynet et al. (2011) for these magnetic field strengths imply thatthe enhancement of nitrogen occurs slowly during the hydrogenburning phase. In turn this would imply that such stars mightappear to be older, which is not observed with our nitrogen en-hanced targets.Fossati et al. (2016) have estimated the fundamental param-eters including ages, for a sample of 389 early-type stars, in-cluding 61 stars that have magnetic fields. The atmospheric pa-rameters of the latter appear similar to those of the other hydro-gen burning main sequence (MS) stars and they concluded that‘the fraction of magnetic massive stars remains constant up to afractional MS age of ≃ Article number, page 16 of 21.L. Dufton et al.: B-type stellar nitrogen abundances in the Tarantula Nebula (2012b) would not appear to be applicable, imply that more thanone mechanism might be responsible.
The analysis of Dunstall et al. (2015) of the VFTS B-type sam-ple identified targets that were members of a binary systems,both SB1 and SB2. However the remaining targets (includingthose with low projected rotational velocities considered here)will contain undetected binary systems. Dunstall et al. (2015)estimated that that whilst their observed binary fraction was0.25 ± ± P >
100 days, where the probability of detectiondecreases; these pre-interaction systems might be expected toevolve as single stars. However there will also be shorter periodsystems (often with small orbital angles of inclination) that re-main undetected and these could possibly explain the nitrogenenriched objects.Garland et al. (2017) analysed VFTS spectroscopy to esti-mate the atmospheric parameters of the primaries with v e sin i <
80 km s − in 33 B-type binaries. Surprisingly none of these pri-maries had ǫ N ≥ currently pre-interactionbinaries. The simulation of Dunstall et al. (2015) would then im-ply that if our nitrogen enriched targets are indeed single, theymay make up an even larger fraction of the single B-type starcohort with v e ≤
40 km s − ; the actual increase would depend onthe frequency of undiscovered binaries with di ff erent projectedrotational velocities. Another possible explanation of the nitrogen enhanced targets isthat they are the product of binary systems in which the compo-nents have interacted. de Mink et al. (2014) discussed how suchinteractions could increase the fraction of stars that were classi-fied as single on the basis of a lack of radial velocity variations.Firstly in the case of a stellar merger, the product would now bea single star. Secondly, following mass transfer, the mass gainerwould dominate the light from the system (precluding an SB2classification), whilst typically only having modest radial veloc-ity variations (less than 20 km s − ).Ramírez-Agudelo et al. (2013) suggested that such a popu-lation of post-interaction binaries may explain the presence of ahigh-velocity tail ( v e sin i >
300 km s − ) found in the rotationalvelocity distribution of the VFTS ‘single’ O-type star sample,which had been initially proposed by de Mink et al. (2013). Fur-ther support for this was the apparent absence of such a highrotational velocity tail for the pre-interacting VFTS O-type bi-naries analysed by Ramírez-Agudelo et al. (2015).The possibility that the nitrogen enhanced FLAMES-I tar-gets might be binary products was considered by Brott et al.(2011b). For example, as discussed by Langer et al. (2008), post- mass transfer systems may remain so tightly bound that tidal in-teraction could spin down the (observed) accreting component.However Brott et al. (2011a) concluded that ‘close binary evolu-tion, as far as it is currently understood, is unlikely to be respon-sible for the slowly rotating nitrogen-rich population of observedstars’. This was based on the binary star pathways discussed byLanger et al. (2008) implying that most post-interaction objectswould be rapidly rotating.Magnetic fields could also play a role in binary systems viamergers of pre-main or main sequence stars (Schneider et al.2016). The former could occur via tidal interactions with cir-cumstellar material (Stahler 2010; Korntre ff et al. 2012), whilstthe latter would occur in some binary evolutionary pathways(Podsiadlowski et al. 1992; Langer 2012). Then the nitrogen en-richment would be due to the merger and its aftermath and / orrotational mixing of the rapidly rotating product. If the stellarmerger also produced a magnetic field, this could then be ane ffi ciently spin down mechanism. Products of stellar mergersmay represent ∼
10% of the O-type field population (de Minket al. 2014). If mergers represented a similar fraction of the B-type population, this would be consistent with the percentages ofnitrogen enriched stars discussed in Sect. 6.5. Additionally theproducts of such mergers might be expected to be more massiveand appear younger than other single B-type stars. Tentative evi-dence for this can be seen in Fig. 3 and 7 and has been discussedin Sect. 6.6.1.In summary, post-interaction binary evolutionary pathwaysare available that might produce slowly rotating, nitrogen en-riched, apparently single stars, either by mass transfer or merg-ers. The latter would also be consistent with the masses and agesinferred for our stellar samples. However models exploring thewide variety of possible evolutionary pathways are required forfurther progress to be made.
7. Conclusions
Our principle conclusions are that:1. Approximately 75% of the targets in the FLAMES-I andVFTS surveys with v e sin i ≤
80 km s − have nitrogen en-hancements of less than 0.3 dex. As such they would appearto be slowly rotating stars that have not undergone significantmixing.2. Both surveys contain stars that exhibit significant nitrogenenhancements. The relative numbers in the cohorts with pro-jected rotational velocities, v e sin i of 0–40 km s − and 40–80 km s − are inconsistent with these being rapidly rotatingstars observed at small angles of inclination.3. For both surveys, the number of targets with v e sin i ≤
40 km s − and ǫ N ≥ ff erences are significant at a high level ofprobability.4. The percentage of these highly nitrogen enhanced objectsthat cannot be explained by rotational mixing are estimatedto be ∼ − and ∼ lowerlimits for stars that appear inconsistent with current grids ofsingle star evolutionary models incorporating rotational mix-ing.5. Including targets with smaller nitrogen enhancements andadopting evolutionary models consistent with the median Article number, page 17 of 21 & Aproofs: manuscript no. 32440_final gravities and masses of our samples leads to larger estimatesof the fractions of targets that are inconsistent with currentevolutionary models, viz. ∼
70% with current rotational ve-locities less than 40 km s − and ∼ ff erences are generally not sig-nificant at the 5% level.7. Some of our targets will be undetected binaries as dis-cussed by Dunstall et al. (2015) and this binary populationdoes not appear to contain highly nitrogen enhanced targets(Garland et al. 2017). Hence the percentages of truly single stars that have significant nitrogen abundances may be higherthan has been estimated. Indeed it is possible that e ff ectivelyall the single B-type population with a current rotational ve-locity, v e ≤
40 km s − may have nitrogen abundances that areinconsistent with single star evolutionary models.8. Possible explanations for these nitrogen enhancements areconsidered of which the most promising would appear to bebreaking due to magnetic fields or stellar mergers with subse-quent magnetic braking. The latter would be consistent withthe higher masses that may pertain in the nitrogen enrichedsample. Acknowledgements.
Based on observations at the European Southern Observa-tory Very Large Telescope in programme 182.D-0222. SdM acknowledges sup-port by a Marie Sklodowska-Curie Action (H2020 MSCA-IF-2014, project id661502).
References
Abt, H. A., Levato, H., & Grosso, M. 2002, ApJ, 573, 359Aerts, C., Molenberghs, G., Kenward, M. G., & Neiner, C. 2014, ApJ, 781, 88Ahmed, A. & Sigut, T. A. A. 2017, MNRAS, 471, 3398Allende Prieto, C., Lambert, D. L., Hubeny, I., & Lanz, T. 2003, ApJS, 147, 363Asplund, M., Grevesse, N., Sauval, A. J., & Scott, P. 2009, ARA&A, 47, 481Brott, I., de Mink, S. E., Cantiello, M., et al. 2011a, A&A, 530, A115Brott, I., Evans, C. J., Hunter, I., et al. 2011b, A&A, 530, A116Carrera, R., Gallart, C., Hardy, E., Aparicio, A., & Zinn, R. 2008, AJ, 135, 836Carroll, J. A. 1933, MNRAS, 93, 478Castro, N., Fossati, L., Langer, N., et al. 2014, A&A, 570, L13Chiosi, C. & Maeder, A. 1986, ARA&A, 24, 329Conti, P. S. & Ebbets, D. 1977, ApJ, 213, 438Corsaro, E., Lee, Y.-N., García, R. A., et al. 2017, Nature Astronomy, 1, 0064Cranmer, S. R. 2005, ApJ, 634, 585de Mink, S. E., Cantiello, M., Langer, N., et al. 2009, A&A, 497, 243de Mink, S. E., Langer, N., Izzard, R. G., Sana, H., & de Koter, A. 2013, ApJ,764, 166de Mink, S. E., Sana, H., Langer, N., Izzard, R. G., & Schneider, F. R. N. 2014,ApJ, 782, 7Donati, J.-F. & Landstreet, J. D. 2009, ARA&A, 47, 333Doran, E. I., Crowther, P. A., de Koter, A., et al. 2013, A&A, 558, A134Dufton, P. L., Langer, N., Dunstall, P. R., et al. 2013, A&A, 550, A109Dufton, P. L., Ryans, R. S. I., Trundle, C., et al. 2005, A&A, 434, 1125Dunstall, P. R., Brott, I., Dufton, P. L., et al. 2011, A&A, 536, A65Dunstall, P. R., Dufton, P. L., Sana, H., et al. 2015, A&A, 580, A93Ekström, S., Georgy, C., Eggenberger, P., et al. 2012, A&A, 537, A146Evans, C. J., Kennedy, M. B., Dufton, P. L., et al. 2015, A&A, 574, A13Evans, C. J., Lennon, D. J., Smartt, S. J., & Trundle, C. 2006, A&A, 456, 623Evans, C. J., Smartt, S. J., Lee, J.-K., et al. 2005, A&A, 437, 467Evans, C. J., Taylor, W. D., Hénault-Brunet, V., et al. 2011, A&A, 530, A108Fossati, L., Castro, N., Morel, T., et al. 2015a, A&A, 574, A20Fossati, L., Castro, N., Schöller, M., et al. 2015b, A&A, 582, A45Fossati, L., Schneider, F. R. N., Castro, N., et al. 2016, A&A, 592, A84Fraser, M., Dufton, P. L., Hunter, I., & Ryans, R. S. I. 2010, MNRAS, 404, 1306Frischknecht, U., Hirschi, R., Meynet, G., et al. 2010, A&A, 522, A39Garland, R., Dufton, P. L., Evans, C. J., et al. 2017, A&A, 603, A91Garnett, D. R. 1999, in IAU Symposium, Vol. 190, New Views of the MagellanicClouds, ed. Y.-H. Chu, N. Suntze ff , J. Hesser, & D. Bohlender, 266 Georgy, C., Ekström, S., Eggenberger, P., et al. 2013a, A&A, 558, A103Georgy, C., Ekström, S., Granada, A., et al. 2013b, A&A, 553, A24Gray, D. F. 2005, The Observation and Analysis of Stellar PhotospheresGray, D. F. 2016, ApJ, 826, 92Gray, D. F. 2017, ApJ, 845, 62Grin, N. J., Ramírez-Agudelo, O. H., de Koter, A., et al. 2017, A&A, 600, A82Grunhut, J. H., Wade, G. A., Neiner, C., et al. 2017, MNRAS, 465, 2432Heger, A. & Langer, N. 2000, ApJ, 544, 1016Henyey, L. G., Lelevier, R., & Levee, R. D. 1959, ApJ, 129, 2Hirschi, R., Meynet, G., & Maeder, A. 2004, A&A, 425, 649Howarth, I. D., Siebert, K. W., Hussain, G. A. J., & Prinja, R. K. 1997, MNRAS,284, 265Hoyle, F. 1960, MNRAS, 120, 22Huang, W., Gies, D. R., & McSwain, M. V. 2010, ApJ, 722, 605Hubeny, I. 1988, Computer Physics Communications, 52, 103Hubeny, I., Heap, S. R., & Lanz, T. 1998, in Astronomical Society of the Pa-cific Conference Series, Vol. 131, Properties of Hot Luminous Stars, ed.I. Howarth, 108– + Hubeny, I. & Lanz, T. 1995, ApJ, 439, 875Hunter, I., Brott, I., Langer, N., et al. 2009, A&A, 496, 841Hunter, I., Brott, I., Lennon, D. J., et al. 2008a, ApJ, 676, L29Hunter, I., Dufton, P. L., Smartt, S. J., et al. 2007, A&A, 466, 277Hunter, I., Lennon, D. J., Dufton, P. L., et al. 2008b, A&A, 479, 541Jackson, R. J. & Je ff ries, R. D. 2010, MNRAS, 402, 1380Korn, A. J., Keller, S. C., Kaufer, A., et al. 2002, A&A, 385, 143Korn, A. J., Nieva, M. F., Daflon, S., & Cunha, K. 2005, ApJ, 633, 899Korntre ff , C., Kaczmarek, T., & Pfalzner, S. 2012, A&A, 543, A126Kurt, C. M. & Dufour, R. J. 1998, in Revista Mexicana de Astronomia y As-trofisica Conference Series, Vol. 7, Revista Mexicana de Astronomia y As-trofisica Conference Series, ed. R. J. Dufour & S. Torres-Peimbert, 202Kushwaha, R. S. 1957, ApJ, 125, 242Langer, N. 2012, ARAA, 50, 107Langer, N., Cantiello, M., Yoon, S.-C., et al. 2008, in IAU Symposium, Vol. 250,IAU Symposium, ed. F. Bresolin, P. A. Crowther, & J. Puls, 167–178Lanz, T. & Hubeny, I. 2003, ApJS, 146, 417Lanz, T. & Hubeny, I. 2007, ApJS, 169, 83Lemasle, B., Groenewegen, M. A. T., Grebel, E. K., et al. 2017, ArXiv e-printsLennon, D. J., Lee, J.-K., Dufton, P. L., & Ryans, R. S. I. 2005, A&A, 438, 265Lucy, L. B. 1974, AJ, 79, 745Maeder, A. 1980, A&A, 92, 101Maeder, A. 1987, A&A, 173, 247Maeder, A. 2009, Physics, Formation and Evolution of Rotating StarsMaeder, A., Georgy, C., Meynet, G., & Ekström, S. 2012, A&A, 539, A110Maeder, A., Przybilla, N., Nieva, M.-F., et al. 2014, A&A, 565, A39Martayan, C., Frémat, Y., Hubert, A., et al. 2006, A&A, 452, 273Martayan, C., Frémat, Y., Hubert, A.-M., et al. 2007, A&A, 462, 683McEvoy, C. M., Dufton, P. L., Evans, C. J., et al. 2015, A&A, 575, A70Meynet, G., Eggenberger, P., & Maeder, A. 2011, A&A, 525, L11Meynet, G. & Maeder, A. 2000, A&A, 361, 101Meynet, G., Maeder, A., Georgy, C., et al. 2017, in IAU Symposium, Vol. 329,The Lives and Death-Throes of Massive Stars, ed. J. J. Eldridge, J. C. Bray,L. A. S. McClelland, & L. Xiao, 3–14Mokiem, M. R., de Koter, A., Vink, J. S., et al. 2007, A&A, 473, 603Morel, T. 2012, in Astronomical Society of the Pacific Conference Series, Vol.465, Proceedings of a Scientific Meeting in Honor of Anthony F. J. Mo ff at,ed. L. Drissen, C. Robert, N. St-Louis, & A. F. J. Mo ff at, 54Morel, T., Castro, N., Fossati, L., et al. 2014, The Messenger, 157, 27Morel, T., Castro, N., Fossati, L., et al. 2015, in IAU Symposium, Vol. 307, NewWindows on Massive Stars, ed. G. Meynet, C. Georgy, J. Groh, & P. Stee,342–347Morel, T., Hubrig, S., & Briquet, M. 2008, A&A, 481, 453Nieva, M. F. & Przybilla, N. 2006, ApJ, 639, L39Palma, T., Clariá, J. J., Geisler, D., Gramajo, L. V., & Ahumada, A. V. 2015,MNRAS, 450, 2122Pasquini, L., Avila, G., Blecha, A., et al. 2002, The Messenger, 110, 1Penny, L. R. 1996, ApJ, 463, 737Petermann, I., Langer, N., Castro, N., & Fossati, L. 2015, A&A, 584, A54Pietrzy´nski, G., Graczyk, D., Gieren, W., et al. 2013, Nature, 495, 76Podsiadlowski, P., Joss, P. C., & Hsu, J. J. L. 1992, ApJ, 391, 246Potter, A. T., Chitre, S. M., & Tout, C. A. 2012a, MNRAS, 424, 2358Potter, A. T., Tout, C. A., & Brott, I. 2012b, MNRAS, 423, 1221Przybilla, N. & Nieva, M.-F. 2011, in IAU Symposium, Vol. 272, Active OBStars: Structure, Evolution, Mass Loss, and Critical Limits, ed. C. Neiner,G. Wade, G. Meynet, & G. Peters, 26–31Puls, J., Vink, J. S., & Najarro, F. 2008, A&A Rev., 16, 209Ramírez-Agudelo, O. H., Sana, H., de Mink, S. E., et al. 2015, A&A, 580, A92Ramírez-Agudelo, O. H., Simón-Díaz, S., Sana, H., et al. 2013, A&A, 560, A29Rivero González, J. G., Puls, J., Najarro, F., & Brott, I. 2012, A&A, 537, A79Ryans, R. S. I., Dufton, P. L., Mooney, C. J., et al. 2003, A&A, 401, 1119Salpeter, E. E. 1955, ApJ, 121, 161 Article number, page 18 of 21.L. Dufton et al.: B-type stellar nitrogen abundances in the Tarantula Nebula
Schneider, F. R. N., Langer, N., de Koter, A., et al. 2014, A&A, 570, A66Schneider, F. R. N., Podsiadlowski, P., Langer, N., Castro, N., & Fossati, L. 2016,MNRAS, 457, 2355Schöller, M., Hubrig, S., Fossati, L., et al. 2017, A&A, 599, A66Schwarzschild, M. & Härm, R. 1958, ApJ, 128, 348Sigut, T. A. A. 1996, ApJ, 473, 452Simón-Díaz, S., Aerts, C., Urbaneja, M. A., et al. 2017a, ArXiv e-printsSimón-Díaz, S., Godart, M., Castro, N., et al. 2017b, A&A, 597, A22Simón-Díaz, S. & Herrero, A. 2007, A&A, 468, 1063Simón-Díaz, S. & Herrero, A. 2014, A&A, 562, A135Simón-Díaz, S., Herrero, A., Uytterhoeven, K., et al. 2010, ApJ, 720, L174Slettebak, A. 1949, ApJ, 110, 498Spruit, H. C. 2002, A&A, 381, 923Stahler, S. W. 2010, MNRAS, 402, 1758Struve, O. 1931, ApJ, 73, 94Szécsi, D., Langer, N., Yoon, S.-C., et al. 2015, A&A, 581, A15Tayler, R. J. 1954, ApJ, 120, 332Tayler, R. J. 1956, MNRAS, 116, 25Taylor, W. D., Evans, C. J., Simón-Díaz, S., et al. 2014, MNRAS, 442, 1483Townsend, R. H. D., Owocki, S. P., & Howarth, I. D. 2004, MNRAS, 350, 189Trundle, C., Dufton, P. L., Hunter, I., et al. 2007a, A&A, 471, 625Trundle, C., Dufton, P. L., Hunter, I., et al. 2007b, A&A, 471, 625Ud-Doula, A., Owocki, S. P., & Townsend, R. H. D. 2009, MNRAS, 392, 1022Wade, G. A., Grunhut, J., Alecian, E., et al. 2014, in IAU Symposium, Vol. 302,Magnetic Fields throughout Stellar Evolution, ed. P. Petit, M. Jardine, & H. C.Spruit, 265–269Wade, G. A., Neiner, C., Alecian, E., et al. 2016, MNRAS, 456, 2Walborn, N. R. & Blades, J. C. 1997, ApJS, 112, 457Wegner, W. 1994, MNRAS, 270, 229Woosley, S. E. & Heger, A. 2006, ApJ, 637, 914Yoon, S.-C. & Langer, N. 2005, A&A, 443, 643Yoon, S.-C., Langer, N., & Norman, C. 2006, A&A, 460, 199
Table 3.
Estimates of the atmospheric parameters and element abun-dances for the VFTS B-type stars.
VFTS v e sin i T e ff log g v t ǫ Mg ǫ N (km s − ) (K) (cm s − ) (km s − ) dex dex024 58 27500 3.3 10 6.92 ≤ ≤
40 27000 4.1 4 6.95 ≤ ≤
40 30500 4.0 1 6.95 ≤ ≤
40 24500 3.5 4 7.04 6.71075 70 23500 3.8 6 6.83 7.04095 ≤
40 31000 4.2 0 7.02 7.96111 80 22000 3.3 0 6.96 7.44119 ≤
40 27500 4.2 0 6.88 ≤ ≤
40 24000 3.6 0 6.82 ≤ ≤
40 20000 3.4 1 7.19 ≤ ≤
40 23500 3.8 2 6.76 ≤ ≤
40 27000 4.0 0 6.76 6.96170 ≤
40 23000 3.5 0 6.87 ≤ ≤
40 29000 3.5 4 6.92 ≤ ≤
40 27500 4.3 0 7.17 ≤ ≤
40 29000 3.6 3 7.03 7.70237 79 23000 3.6 2 - 6.92241 69 30500 4.0 0 6.97 ≤ ≤
40 29500 3.9 0 6.78 ≤ ≤
40 19000 3.8 5 6.74 ≤ ≤ ≤ ≤ ≤ ≤ ≤
40 30500 4.0 0 7.03 ≤ ≤ ≤ ≤
40 29500 3.5 1 7.05 ≤ ≤ ≤ ≤
40 29000 4.4 0 6.91 ≤ ≤
40 28500 3.9 3 7.21 6.85650 ≤
40 23500 3.8 7 6.82 7.02666 ≤
40 28500 4.0 0 6.94 ≤ ≤
40 25000 3.8 2 7.00 7.13673 ≤
40 25500 4.0 1 6.98 ≤ ≤
40 29500 3.9 2 7.08 7.69707 ≤
40 27500 4.0 0 6.96 ≤ ≤
40 23500 3.5 6 6.86 7.75740 ≤
40 22500 3.4 11 6.92 7.18748 62 29000 4.1 5 7.18 ≤ ≤
40 20500 3.4 0 - 7.31860 60 23500 3.8 6 7.13 6.98864 ≤
40 27000 4.1 1 7.35 7.46868 ≤
40 23000 3.6 6 7.12 7.05872 77 31000 4.1 1 7.03 ≤ ≤
40 26500 3.8 6 7.02 7.94
Article number, page 19 of 21 & Aproofs: manuscript no. 32440_final
Table 4.
Nitrogen abundance estimates from di ff erent lines for the VFTS B-type stars. The values for the line at 3995 Å have been taken directlyfrom Table 3. Star Nitrogen Abundances (dex)3995 4227 4447 4601 4607 4613 4621 4630 4994 5001 Mean095 7.96 - - - - - - 7.87 - - 7.92 ± ± ± ± ± Table 5.
Estimates of the atmospheric parameters, projected rotationalvelocities and nitrogen abundances for narrow lined B-type targets fromthe FLAMES-I survey for the LMC clusters NGC 2004 (designated2004) and N11.
Star T e ff log g v t v e sin i ǫ N (K) (cm s − ) (km s − ) (km s − ) dex2004 /
36 22870 3.35 7 42 7.272004 /
42 20980 3.45 2 42 6.872004 /
43 22950 3.50 7 24 7.132004 /
46 26090 3.85 2 32 7.552004 /
51 21700 3.40 5 70 6.872004 /
53 31500 4.15 6 7 7.642004 /
61 20990 3.35 1 40 6.852004 /
64 25900 3.70 6 28 7.592004 /
68 20450 3.65 1 62 6.952004 /
70 27400 3.90 4 46 7.562004 /
73 23000 3.65 1 37 6.922004 /
76 20450 3.70 1 37 6.882004 /
84 27395 4.00 3 36 7.152004 /
86 21700 3.85 6 14 7.072004 /
87 25700 4.40 0 35 6.902004 /
91 26520 4.10 0 40 7.122004 /
103 21500 3.85 1 35 6.842004 /
106 21700 3.50 2 41 6.872004 /
111 20450 3.30 2 55 6.942004 /
112 21700 3.70 1 72 6.812004 /
114 21700 3.60 2 59 6.922004 /
116 21700 3.55 3 63 6.932004 /
117 21700 3.60 4 75 7.072004 /
119 23210 3.75 0 15 6.86N11 /
69 24300 3.30 11 80 6.83N11 /
70 19500 3.30 0 62 6.79N11 /
72 28800 3.75 5 15 7.43N11 /
79 32500 4.30 0 38 7.02N11 /
86 26800 4.25 0 75 6.97N11 /
93 19500 3.30 1 73 6.78N11 /
100 29700 4.15 4 30 7.78N11 /
106 31200 4.00 5 25 6.94N11 /
110 23100 3.25 7 25 7.43N11 /
115 24150 3.65 1 53 6.91
Table 6.
Estimates of the initial rotational velocity, v i , required toachieve nitrogen enhancements of greater than 0.3 dex ( ǫ N ≥ ǫ N ≥ g ≥ v e , predicted by themodels during this evolutionary phase. Initial masses ( M i ) are in unitsof the solar mass, M ⊙ . The values characterised as ’Adopted’ were usedin the simulations discussed in Sects. 6.3-6.5. M i / M ⊙ ǫ N ≥ ǫ N ≥ v i v e v i v e km s − km s − km s − km s − Article number, page 20 of 21.L. Dufton et al.: B-type stellar nitrogen abundances in the Tarantula Nebula
Table 7.
Observed and predicted numbers of nitrogen rich targets with enhancements of greater than 0.3 dex ( ǫ N ≥ ǫ N ≥ N , are listed, they refer to simulations tailored to thedi ff erent nitrogen abundance limits as discussed in the text. The rotational velocity distribution, f ( v e ), is either from Dufton et al. (2013, VFTS)or from the deconvolution illustrated in Fig. 6 (FLAMES-I). Simulations are provided for two ranges of projected rotational velocity, v e sin i , viz.0-40 km s − and 40-80 km s − . Sample
N f ( v e ) ǫ N ≥ ǫ N ≥ v e sin i (km s − ) 0-40 40-80 0-40 40-80VFTS sampleObserved 255 - 7 3 5 1MC 255 VFTS 2.30 ± ± ± ± /
188 VFTS 1.98 ± ± ± ± /
188 VFTS 0.4 22 0.9 16FLAMES-I sampleObserved 103 - 6 2 4 1MC 103 VFTS 0.92 ± ± ± ± ± ± ± ±0.89MC Prob(%) 73 FLAMES-I 0.01 55 0.04 56