Possible binary progenitors for the type Ib supernova iPTF13bvn
J. J. Eldridge, Morgan Fraser, Justyn R. Maund, Stephen J. Smartt
aa r X i v : . [ a s t r o - ph . S R ] A ug Mon. Not. R. Astron. Soc. , 1–8 (2005) Printed 3 May 2018 (MN L A TEX style file v2.2)
Possible binary progenitors for the type Ib supernovaiPTF13bvn
J. J. Eldridge ⋆ , Morgan Fraser , Justyn R. Maund , , Stephen J. Smartt Department of Physics, University of Auckland, Private Bag 92019, Auckland, New Zealand Institute of Astronomy, University of Cambridge, Madingley Road, Cambridge CB3 0HA, UK Astrophysics Research Center, School of Mathematics and Physics, Queen’s University Belfast, Belfast BT7 1NN, UK Department of Physics & Astronomy, University of Sheffield, Hicks Building, Hounsfield Road, Sheffield S3 7RH, UK
ABSTRACT
Cao et al. (2013) reported a possible progenitor detection for the type Ib supernovaeiPTF13bvn for the first time. We find that the progenitor is in fact brighter than themagnitudes previously reported by approximately 0.7 to 0.2 mag with a larger error inthe bluer filters. We compare our new magnitudes to our large set of binary evolutionmodels and find that many binary models with initial masses in the range of 10 to20 M ⊙ match this new photometry and other constraints suggested from analysing thesupernova. In addition these lower mass stars retain more helium at the end of themodel evolution indicating that they are likely to be observed as type Ib supernovaerather than their more massive, Wolf-Rayet counter parts. We are able to rule outtypical Wolf-Rayet models as the progenitor because their ejecta masses are too highand they do not fit the observed SED unless they have a massive companion which isthe observed source at the supernova location. Therefore only late-time observationsof the location will truly confirm if the progenitor was a helium giant and not aWolf-Rayet star. Key words: stars: evolution – binaries: general – supernovae:general – supernovae:iPTF13bvn
Massive stars end their lives in the explosive death throesof a core-collapse supernova (SNe). These SNe are classi-fied according to their observed spectra and lightcurves; inthe first instance by the presence or absence of hydrogenin the SN spectrum – hydrogen rich SNe are classified as“Type II”, while hydrogen-deficient SNe are “Type I”. TypeI SNe are further divided into Types Ib and Ic (collectivelytermed Type Ibc), which are helium rich and helium poorrespectively. While the progenitors of Type II SNe have beendirectly identified in pre-explosion images as H-rich super-giants between 8 and 16 M ⊙ (Smartt et al. 2009, and ref-erences therein), the progenitors of Type Ibc SNe have re-mained elusive.The two likely candidates for the progenitors ofType Ibc SNe are single, massive Wolf-Rayet (WR) stars(Gaskell et al. 1986), or lower mass stars in binaries(Podsiadlowski et al. 1992). In both cases, the progenitor ⋆ E-mail: [email protected] We note that Type Ia SNe are hydrogen-deficient supernovafrom a thermonuclear explosion mechanism in a carbon-oxygenwhite dwarf which we do not consider here. will be stripped of its H and/or He envelope. Eldridge et al.(2013) presented a sample of nearby Type Ibc SNe with pre-explosion
Hubble Space Telescope ( HST ) imaging, but foundno progenitor candidates. Eldridge et al. suggested that thiswas evidence that a number of the progenitors of these su-pernovae must been the result of an interacting binary star,as previously suggested from the relative rates of differentSN types (De Donder & Vanbeveren 1998; Eldridge et al.2008; Smith et al. 2011). There is also growing additionalevidence from statistical samples of ejecta masses that mostIb/c SNe are from low mass stars in binaries (Drout et al.2011; Lyman et al. 2014; Bianco et al. 2014)Last year Cao et al. (2013) presented the detection of apossible progenitor candidate in
HST images for the TypeIb supernova iPTF13bvn in the nearby galaxy NGC 5806.From the magnitudes they report for the progenitor candi-date, along with indirect constraints on its radius and mass-loss rate from observations of the SN itself, they suggestedthe progenitor of iPTF13bvn was consistent with a sin-gle WR star. Groh et al. (2013) compared their single-starmodels to the constraints and found a possible initial massrange between 31 to 35 M ⊙ for the progenitor. However,follow-up observations of iPTF13bvn yielded a bolometic c (cid:13) J. J. Eldridge et al. lightcurve which, when fitted with a hydrodynamic modelfor the SN ejecta, implied a pre-explosion mass of ∼ ⊙ (Fremling et al. 2014; Bersten et al. 2014). Such a low massis inconsistent with a single WR star, which in the modelsof Groh et al. would have a pre-explosion mass of ∼
11 M ⊙ .Bersten et al. further presented modelling of a binary pro-genitor system consisting of a 19 M ⊙ primary and a 20 M ⊙ secondary which could match the pre-explosion constraintsfrom HST . It is important to note that Yoon et al. (2012)predicted it would be easier to observe such a low-mass he-lium star than a Wolf-Rayet star as the progenitor of a typeIb/c SN.In this letter we first reanalyse the pre-explosion imagesof the site of the SN. We use late-time
HST images to revisitthe astrometry and photometry of the progenitor candidate.We then compare the derived observational constraints forthe progenitor of iPTF13bvn to our grid of binary evolu-tion models from the BPASS (Binary Population and Spec-tral Synthesis, http://bpass.auckland.ac.nz ) code. WhileBersten et al. (2014) have presented a binary progenitor sce-nario for iPTF13bvn, they note that their solution is notunique. Furthermore, we find the photometry of Cao et al.(2013) to which Bersten et al. (2014) fit their models to un-derestimates the progenitor candidates magnitudes. Withour grid of models, we can compare a wide range of binarysystems to the progenitor of iPTF13bvn, and constrain theallowed parameter space of the system.In the following, we adopt a distance of 22.5 ± µ = 31 . ± .
36 mag towards NGC 5806, as usedby Cao et al. and Fremling et al. from Tully et al. (2009).While the foreground reddening towards NGC 5806 islow (E(B-V)=0.045) mag from the Schlafly & Finkbeiner(2011) dust maps, the host galaxy reddenings adopted byBersten et al., Cao et al. and Fremling et al. differ signifi-cantly. Bersten et al. find, E(B-V)=0.17 ± Cao et al. (2013) identified a progenitor candidate foriPTF13bvn in pre-explosion
HST
Advanced Camera forSurveys (ACS) images, acquired as part of program GO-10187 (PI: Smartt). These observations were acquired withthe Wide Field Channel (WFC; pixel scale 0.05 ′′ pix − ) ofACS on 2005 March 10 using the F W (1600s), F W (1400s) and F W (1700s) filters. A key outstanding ques-tion in the Cao et al. analysis, however, was the level ofagreement between the position of the progenitor candi-date on the pre-explosion image and the transformed SNposition derived from post-explosion adaptive optics im-ages. Fremling et al. (2014) presented a re-analysis of theposition of iPTF13bvn using HST +WFC3 observations ofiPTF13bvn, and found the Cao et al. progenitor candidateto be coincident with the SN. Using the same data as Fremling et al., we have performed an independent analy-sis of the position of iPTF13bvn.New
HST
Wide Field Camera 3 (WFC3; pixel scale0.04 ′′ pix − ) Ultraviolet and Visual (UVIS) imager obser-vations ( F W astrodrizzle task within the drizzlepac package, the undersampled WFC3 flt images were drizzled (Fruchter & Hook2002) onto a finer pixel scale, yielding a distortion cor-rected combined image with a pixel scale of 0.025 ′′ . Thepre-explosion ACS images were taken at the same pointing,and so drizzling could not be used to improve their spatialresolution. However, the two individual flc frames were stillcombined with astrodrizzle (although with an output pixelscale of 0.05 ′′ ) to remove cosmic rays and correct for thegeometric distortion of ACS.Using 29 point sources identified in both the ACS F555W and WFC3 frames, we derived a geometric trans-formation between the pre- and post-explosion images withan RMS error of 0.38 ACS pixels (19 mas). The pixel co-ordinates of iPTF13bvn were then measured on the post-explosion WFC3 image (as the SN is bright, the uncertaintyon its position is negligible in all of the following) and trans-formed to the ACS frame. We find the progenitor candidateof Cao et al. (2013) to be offset by only 7 mas from thetransformed position of iPTF13bvn, and hence formally co-incident, as also found by Fremling et al. (2014).A caveat to this result is that the geometric distortionwhich is corrected for by astrodrizzle necessitates resamplingthe pixels in the image. We found, through trials using boththe multidrizzle and astrodrizzle packages, that the offset be-tween the transformed SN position and the position of theprogenitor candidate was highly sensitive to the choice ofdrizzling parameters applied to the pre-explosion image (e.g.the subtraction of the sky background, the shape of the driz-zle kernel, the reduced pixel size or “drop” size etc.). We notethat this effect was not observed for brighter nearby stars,and appeared to arise solely due to the relative faintness ofthe candidate. In comparison with bright nearby surround-ing stars, we found the position of the progenitor candidatecould change by as much as ∼ . F555W crj image (j90n02021 crj.fits)and the undistorted post-explosion WFC3
F555W image,drizzled to 0.025 ′′ pix − . To account for the distortion inthe ACS frame, a 4th order polynomial was used for thetransformation, which had an RMS error of 8 mas. The co-ordinates of iPTF13bvn were then transformed to the crjimage, where it lies at pixel coordinates 2698.09,593.38. Theposition of the progenitor candidate was measured usingboth the iraf phot package and with dolphot (Dolphin2000) to lie at 2698.0,593.83 and 2697.84,593.61 respectively.The progenitor candidate positions from phot and dolphot are offset from the transformed SN position by . .
25 mas), however, they are also offset from each other by14 mas.There are also significant differences found between c (cid:13) , 1–8 inary progenitors for supernova iPTF13bvn the archival ACS drizzled products provided in the STScIarchives. There are two products provided, the HubbleLegacy Archive (HLA) drz and the MAST drc imageswhich are both resampled from the original detector pixelsonto a grid of equal sky area pixels. The drc files includeCTE (charge transfer efficiency) corrections in the pixel val-ues, while the drz images do not. Photometry on drz im-ages thus requie CTE corrections after flux measurements.Using the HLA drz images, an alignment between our WF3drizzled frame produces an RMS of 0.28 ACS pixels using38 stars for alignment (within the geomap task of IRAF ).The positional uncertainty between the SN and progenitorposition is 0.73 ± σ level dependingon which drizzled product to use we conclude that they arelikely coincident within the errors based on our own manual astrodrizzle ACS product and the drc images. The twopapers published so far which have discussed the progeni-tor identification and alignment (Cao et al. 2013, Fremlinget al. 2014) are not specifically clear which data productshave been used but we agree with these papers in suggest-ing this is a likely progenitor candidate object. The true testof whether iPTF13bvn and the progenitor candidate are co-incident will be at late times when it will be possible to seeif the latter has truly disappeared.We performed Point Spread Function (PSF)-fitting pho-tometry of the pre-explosion crj images using the dolphot package (Dolphin 2000) with the ACS module. Bad pixelswere masked using the data quality images, before dolphot was run with the recommended parameters for ACS/WFCdata. The progenitor candidate was detected in all three ofthe ACS filters. Interestingly, if dolphot is run on eachof the filters separately, the magnitudes returned for theprogenitor candidate are ∼ dolphot together. We measuremagnitudes on individual images in the VEGAMAG systemof F435W = 25.81 ± F555W = 25.86 ± F814W =25.77 ± dolphot , we also performedphotometry on the pre-explosion images using daophot within iraf . Photometry was performed on the drc filesat the native ACS/WFC pixel scale of 0.05 ′′ pix − . Thedrc images have been corrected for both the inherent ge-ometric distortion of ACS, and for losses due to ChargeTransfer Efficiency (CTE). For each filter, a Point SpreadFunction (PSF) was constructed from bright, isolated pointsources. The modelled PSF was then fit simultaneously toboth the progenitor candidate and all surrounding sourceswhich may contribute flux at the position of the SN. Thefit was made within a small (2 pixel) radius centred on eachsource, and the measured fluxes within this aperture werecorrected to an infinite aperture using the tabulated cor-rections in Sirianni et al. (2005). Finally, the flux was con-verted to a magnitude in the HST
VEGAMAG system us-ing the value of PHOTFLAM from the image header, and http://americano.dolphinsim.com/dolphot/ the flux of Vega in the corresponding filter from the HSTwebpages . We find magnitudes of F435W =25.79 ± F555W =25.73 ± F814W =25.99 ± dolphot . Because there is no clear reason to favour one overthe other we use a mean of the two values. This gives mag-nitudes for the progenitor of F435W = 25.80 ± F555W = 25.80 ± F814W = 25.88 ± F435W =26.50 ± F555W =26.40 ± F814W =26.10 ± dolphot and daophot ,that the results are dependent on the parameters given tothese codes when the photometry is derived and that anysmall error may be amplified.The residual images after subtraction of the fitted PSFswere examined, and do not show any gross over- or under-subtractions at the SN position. However, it is clear that thebackground is not smooth at the SN position, and late timeobservations after the SN has faded will be important torefine the progenitor candidate photometry using templatesubtraction (Maund et al. 2014). The construction of the stellar models used in this paperhave been described in detail in Eldridge et al. (2008). Herewe use these models and compare them to the progenitorcandidate in a similar method to as in Eldridge et al. (2013),but now compare the models to an actual detection ratherthan upper limits on progenitor magnitude. In summary thestellar models follow single and binary stars at two metal-licities, Z = 0 .
008 and 0.020, that are close to the metal-licity inferred for NGC5806 of 12+log [O/H] = 8.5 fromSmartt et al. (2009). In the models the primary effect ofmetallicity is to vary the mass-loss rates via stellar winds.The evolutionary models are then matched to WR atmo-sphere models from the Potsdam group (e.g. Sander et al.2012) to enable their magnitudes to be calculated, as dis-cussed in Eldridge & Stanway (2009). The grid of modelscovers initial masses of the primary from 5 to 120 M ⊙ withmass ratios, m /m between 0.1 to 0.9 and initial separa-tions in log( a/R ⊙ ) from 1.0 to 4.0.The major difference in our method here to that ofEldridge et al. (2013) is that our aim is to demonstrate thatsingle star WR models are not the only possible progenitorand interacting binaries can fit the observed source and fitthe other constraints available. Therefore we compare thedetected source to the end points of our models rather thanconsidering the whole evolutionary track closer to the timeof core-collapse. The latter is a more apt method to usewhen attempting to estimate accurate parameters for theprogenitor and take into consideration uncertainties in thestellar evolution models themselves. But until post-explosion (cid:13) , 1–8 J. J. Eldridge et al.
Figure 1.
HR diagram showing the evolutionary tracks for models which match the constraints on the progenitor candidate of iPTF13bvn.Thin grey dashed lines - evolution tracks with hydrogen, thick black lines - evolution without hydrogen. Asterisks - end point of progenitormodels, diamonds - location of secondary star at explosion, they are included to indicate the general locations possible for the secondarystar. Colour indicates helium mass in the primary at the end point of the model. images are available to more tightly constrain the progeni-tor magnitudes, we consider only the final model end points.These are typically after the end of core-carbon burning andonly a few years before core-collapse.We have searched through our grid of models for starswhich would give rise to a hydrogen-free SN and com-pared the magnitude of these models to the magnitude de-rived in Section 2. With our assumed distance the abso-lute magnitudes for the progenitors candidate are, F435W=-5.96 ± ± ± HST filter system to avoid theadditional uncertainty from converting to the
UBVRI sys-tem. The upper limit of possible magnitudes are taken from magnitudes calculated with the higher extinction value andthe lower bound is from the lower extinction value. We listthe set of progenitor models which match the observed mag-nitudes of the progenitor candidate within the error bars,and within the range allowed by the uncertainty in extinc-tion, in Table 1. The evolutionary tracks of these modelsare plotted on a Hertzsprung-Russell diagram in Figure 1,along with their spectral energy distributions (SED) com-pared to the observed magnitudes in Figure 2. In most casesthe SED is dominated by the primary, apart from in the fewcases where the final mass of the secondary star is similarto the primary star’s initial mass. We caution however, thatthe effect of mass transfer can cause dramatic evolution-ary changes in the secondary star and much of the relevantphysics is uncertain, as discussed by Claeys et al. (2011). c (cid:13) , 1–8 inary progenitors for supernova iPTF13bvn Table 1.
Physical parameters of the binary progenitor models which match the observed constraints on the progenitor of iPTF13bvn.Models where the primary mass has an asterisk beside it are systems with a compact objects for a secondary; the secondary massesof 0.6, 1.4 or 2.0 M ⊙ correspond to a white dwarf, neutron star or black hole respectively. The given magnitudes are for the combinedsystem of primary and secondary together. All systems however are dominated by the emission from the primary star. All masses, radiiand luminosities are in given in units of M ⊙ , R ⊙ and L ⊙ respectively. Surface temperatures are given in Kelvin. M ,i M ,i log( a/R ⊙ ) R log T log L M ,f M ,f M H M He M ej A ∗ F W F
W F WZ = 0 . Z = 0 . The large number of possible progenitor models meanswe need to consider also the secondary constraints fromCao et al. (2013), Fremling et al. (2014) and Bersten et al.(2014). We consider the constraint on the mass-loss rate,ejecta mass and the requirement for sufficient helium to pro- duce a Type Ib SN. We do not use the radius constraint,because as pointed out by Bersten et al., this constraint isnot as firm as first thought.The constraint on the mass-loss rate from Cao et al.(2013) needs to be considered with care. Cao et al. as-sume a wind velocity of 1000 km s − to derive a mass- c (cid:13) , 1–8 J. J. Eldridge et al.
Figure 2.
SED of progenitor models compared to limits derivedhere with both the low and high extinction values used. The errorbars on the observed limits are mainly determined by the errorin the distance to the host galaxy. Here the colours of the linesrepresent the helium abundance of the model as for the points inFigure 1. Most of the models are relatively cool with shallow orflat SEDs. loss rate of approximately 3 × − M ⊙ yr − . This cal-culation is strongly dependent on the wind velocity as-sumed. While 1000 km s − is a typical WR wind speed,Eldridge et al. (2006) found that WR wind speeds evolvetowards the end of a star’s evolution and vary with fi-nal mass. Therefore a more reliable constraint is to con-sider the wind density, which is dependent on fewer as-sumptions. We use the dimensionless wind parameter, A ∗ ,where A ∗ = ( ˙ M/ − M ⊙ yr − ) / ( v wind / − ). There-fore values of the order unity are similar to those from thetypical Wolf-Rayet star. From Cao et al., iPTF13bvn hasa value of A ∗ = 3 therefore somewhat dense compared tothe typical Wolf-Rayet wind. The mass-loss rate and windvelocity for our models is calculated from Nugis & Lamers(2000) as described in Eldridge et al. (2006). We find lowermass models such as helium stars may have weaker mass-lossrates, but they also have slower wind velocities. We requirethat our models have an A ∗ value between 0.3 and 30.0,allowing for an order of magnitude error in the measuredvalue and in our calculation of the model values. Most ofour models fall within this range of observed wind param-eter. Typically our model wind parameters cover a rangebetween 0.4 to 3. The measured value is dependent on otherphysical assumptions so we do not regard this as a significantdisagreement.The ejecta mass derived by Fremling et al. (2014) foriPTF1bvn is around 1 . +0 . − . M ⊙ . This should be consid-ered a lower limit as there may always be additional he-lium that is transparent and unobservable as described byPiro & Morozova (2014). We estimate an ejecta mass forour models by calculating the binding energy of the starand using this to estimate how much mass would be ejectedif 10 ergs of energy was injected into the envelope as de-scribed in Eldridge & Tout (2004). Only in cases where the binding energy of the stellar envelope is higher than thiswould there be material left to fall back onto the centralproto-neutron star. As our models have an initial mass lessthan 20 M ⊙ we find that a neutron star is always produced sothe ejecta mass is effectively the final mass minus 1.4M ⊙ . Foreach model a corresponding minimum observed ejecta masscan be estimated by subtracting the amount of helium inthe stellar model from the ejecta mass quotes in Table 1. Weconstrain our model selection again by requiring the ejectamass to be less than 3 . M ⊙ . This upper limit is estimated byusing the upper limit from the error in the ejecta mass andupto 1M ⊙ of helium being transparent (Piro & Morozova2014). We find our model ejecta masses are in reasonableagreement with the value of Fremling et al. (2014), typicallylying between 1 and 2 M ⊙ .The minimum amount of helium which a star must re-tain to the point of core-collapse if it is to produce the spec-troscopic signature of a Type Ib SN is still somewhat un-certain (Dessart et al. 2011). In nearly all our models thereis greater than 0.5 M ⊙ of helium in the ejecta, likely to beenough to provide the required Type Ib SN spectrum. Wenote that in some of the models there is a small amount ofhydrogen left on the surface of the star at the end of ourmodels. Because our models end at the end of carbon burn-ing it is possible that this hydrogen would subsequently beremoved, in addition the mass-loss rates of such stars arehighest, and least certain at the end of their lives, whenthey become helium giants.In summary the possible binary progenitors we find foriPTF13bvn mainly have masses between 10 and 20 M ⊙ . Alarge number of possible binary progenitors for iPTF13bvnwill be survived by a visible, albeit faint, stellar compan-ion. There is also a subset of systems where the companionwill be a compact object, and undetectable at optical wave-lengths. We do not predict the magnitude for the secondarycompanion, as the parameters of this star will be strongly af-fected by the mass transfer process. Finally we note that wedo find that some very massive stars with initial masses of80 M ⊙ do match our magnitude range. However these havevery low amounts of helium, large ejecta masses of around 5to 7 M ⊙ and the SED is dominated by the binary compan-ion. In this case the observed SED therefore represents thebinary companion not the progenitor itself. In contrast to the conclusion of Groh et al. (2013), we can-not find any single-star models which match the SED ofthe progenitor candidate for iPTF13bvn. This is largely dueto our revised magnitudes being brighter than previouslyreported, especially with the brighter
F435W magnitudes.In addition to the other constraints such as the total massand mass of helium ejected. We note that our single-starmodels do not include rotation so our analysis does notrule out a single star solution completely. Therefore similarto Bersten et al. (2014), we conclude that iPTF13bvn mostlikely did not come from a non-rotating single-star progeni-tor. We caution that current uncertainties in stellar modelscould weaken this result. For example the role of envelopeinflation of WR stars, an increase in their radius due to radi- c (cid:13)000
F435W magnitudes.In addition to the other constraints such as the total massand mass of helium ejected. We note that our single-starmodels do not include rotation so our analysis does notrule out a single star solution completely. Therefore similarto Bersten et al. (2014), we conclude that iPTF13bvn mostlikely did not come from a non-rotating single-star progeni-tor. We caution that current uncertainties in stellar modelscould weaken this result. For example the role of envelopeinflation of WR stars, an increase in their radius due to radi- c (cid:13)000 , 1–8 inary progenitors for supernova iPTF13bvn ation pressure on the iron-opacity peak, is still the subject ofresearch and debate. Therefore the single-star radii could besmaller or greater than expected from models. In additionmass-loss rates of WR stars are still to some extent uncer-tain so models may lose less mass during the WR phraseand therefore contain more helium when they explode.From our models we favour a binary progenitor foriPTF13bvn, most likely a low-mass helium giant in a bi-nary system. While such a helium giant would have a ra-dius larger that the limit of < ⊙ derived by Cao et al.(2013). Bersten et al. (2014) have suggested that for this su-pernovae, as for SN 2011dh (Bersten et al. 2012), detailedmodelling demonstrates that the initial constraints on theprogenitor radius are not as stringent as first suggested.We stress that all binary models represent a “best-guess” as to the evolution of massive interacting binarystars. The largest uncertainty remains the contribution fromthe binary companion of the progenitor to the SED of theprogenitor system. As discussed by Stancliffe & Eldridge(2009) and Claeys et al. (2011) the evolution of these starspost mass-transfer is uncertain, and they may be coolerthan normally expected for a main-sequence star. Detec-tion of any surviving companion in late-time imaging of theSN site will provide an important constraint on the binaryscenario. Further more a spectrum of the star may revealthat it is rapidly rotating because of the binary interactions(De Mink et al. 2013). We have presented revised magnitudes for the source thatis coincident with the supernova iPTF13bvn. The
F435W is the most significantly brighter by 0.7 mags. This changesthe shape of the source SED and therefore has a stronginfluence on the resulting possible objects that can matchthe progenitor source.Using these new magnitudes and allowing for a rangeof extinctions measured by different methods we find thatit is possible to match the source and other secondary con-straints with binary models that had initial masses between10 to 20 M ⊙ . This overlaps with the model suggested byBersten et al. (2014).More massive models tend to not fit the source SEDwithout a bright companion star. Therefore if the source stillexists when the SN fades then the progenitor was a moremassive Wolf-Rayet star rather than a lower mass heliumgiant. However we suggest that the latter is highly favouredin light of the ejecta mass estimates of Bersten et al. (2014)and Fremling et al. (2014). This is also in agreement withthe prediction that helium giants would be easier to identifyas the progenitor of a type Ib/c SN by Yoon et al. (2012).It is only with late-time imaging that a deeper insightwill be gained into the progenitor. This has been demon-strated by analysis of even the relatively well understoodprogenitors of Type IIP SNe (Maund et al. 2014). If a sur-viving companion star is found at the site of iPTF13bvn,then for the first time the binary evolution of a Type Ib SNprogenitor can be studied in detail. JJE acknowledges support from the University of Auckland.This work was partly supported by the European UnionFP7 programme through ERC grant number 320360. SJS ac-knowledge funding from the European Research Council un-der the European Union’s Seventh Framework Programme(FP7/2007-2013)/ERC Grant agreement n o [291222] andSTFC grants ST/I001123/1 and ST/L000709/1. The re-search of JRM is supported by a Royal Society ResearchFellowship. REFERENCES
Bersten, M. C., Benvenuto, O. G., Nomoto, K., et al. 2012,ApJ, 757, 31Bersten M.C. et al., 2014, arXiv:1403.7288Bianco F., 2014, in prep.Cao Y. et al., 2013, ApJL, 775, L7Claeys, J. S. W., de Mink, S. E., Pols, O. R., Eldridge,J. J., & Baes, M. 2011, A&A, 528, A131De Donder E., Vanbeveren D., 1998, A&A, 333, 557-564de Mink S. E., Langer N., Izzard R. G., Sana H., de KoterA., 2013, ApJ, 764, 166DDessart, L., Hillier, D. J., Livne, E., et al. 2011, MNRAS,414, 2985Dolphin, A. E. 2000, PASP, 112, 1383Drout M.R., et al., 2011, ApJ, 741, 97Eldridge, J. J., Tout, C.A. 2004, MNRAS, 353, 87Eldridge, J. J., Genet, F., Daigne, F., Mochkovitch, R.2006, MNRAS, 367, 186Eldridge, J. J., Izzard, R. G., Tout, C. A. 2008, MNRAS,384, 1109Eldridge, J. J., & Stanway, E. R. 2009, MNRAS, 400, 1019Eldridge J.J., Fraser M., Smartt S.J., Maund J.R., CrockettR.M., 2013, MNRAS, 436, 774Fremling C. et al., 2014, arXiv:1403.6708Fruchter, A. S., & Hook, R. N. 2002, PASP, 114, 144Gaskell C.M., Cappellaro E., Dinerstein H.L., GarnettD.R., Harkness R.P., Wheeler J.C., 1986, ApJL, 306, L77Groh J.H., Georgy C., Ekstr¨om S., 2013, A&A, 558, L1Lyman J., Bersier D., Phil J., Paolo M., Eldridge J., FraserM., Elena P., 2014, MNRAS submitted.Maund J.R., Reilly E., Mattila S., 2014, MNRAS, 438, 938Maund J.R., Fraser M., Reilly E., Ergon M., Mattila S.,2014, MNRAS submittedNugis, T., Lamers, H. J. G. L. M. 2000, A&A, 360, 227Prio A.L., Morozova V.S., 2014, astro-ph/1407.5992Podsiadlowski P., Joss P.C., Hsu J.J.L., 1992, ApJ, 391,246Sander, A., Hamann, W.-R., Todt, H. 2012, A&A, 540,A144Schlafly E.F., Finkbeiner D.P., 2011, ApJ, 737, 103Sirianni M. et al. 2005, PASP, 117, 1049Smartt S.J., Eldridge J.J., Crockett R.M., Maund J.R,2009, MNRAS, 395, 1409Smith N., Li W., Filippenko A.V., Chornock R., 2011, MN-RAS, 412, 1522Stancliffe, R. J., & Eldridge, J. J. 2009, MNRAS, 396, 1699Tully R. B., Rizzi L., Shaya E. J., Courtois H. M., MakarovD. I., Jacobs B. A., 2009, AJ, 138, 323 c (cid:13) , 1–8 J. J. Eldridge et al.
Yoon, S.-C., Grfener, G., Vink J. S. Kozyreva A., IzzardR. G., 2012, A&A, 544, 11This paper has been typeset from a TEX/ L A TEX file preparedby the author. c (cid:13)000