No Arabic abstract
The evolution of massive stars surviving the red supergiant (RSG) stage remains unexplored due to the rarity of such objects. The yellow hypergiants (YHGs) appear to be the warm counterparts of post-RSG classes located near the Humphreys-Davidson upper luminosity limit, which are characterized by atmospheric instability and high mass-loss rates. We aim to increase the number of YHGs in M33 and thus to contribute to a better understanding of the pre-supernova evolution of massive stars. Optical spectroscopy of five dust-enshrouded YSGs selected from mid-IR criteria was obtained with the goal of detecting evidence of extensive atmospheres. We also analyzed BVI photometry for 21 of the most luminous YSGs in M33 to identify changes in the spectral type. To explore the properties of circumstellar dust, we performed SED-fitting of multi-band photometry of the 21 YSGs. We find three luminous YSGs in our sample to be YHG candidates, as they are surrounded by hot dust and are enshrouded within extended, cold dusty envelopes. Our spectroscopy of star 2 shows emission of more than one H$alpha$ component, as well as emission of CaII, implying an extended atmospheric structure. In addition, the long-term monitoring of the star reveals a dimming in the visual light curve of amplitude larger than 0.5 mag that caused an apparent drop in the temperature that exceeded 500 K. We suggest the observed variability to be analogous to that of the Galactic YHG $rho$ Cas. Five less luminous YSGs are suggested as post-RSG candidates showing evidence of hot or/and cool dust emission. We demonstrate that mid-IR photometry, combined with optical spectroscopy and time-series photometry, provide a robust method for identifying candidate YHGs. Future discovery of YHGs in Local Group galaxies is critical for the study of the late evolution of intermediate-mass massive stars.
The circumstellar envelope of the hypergiant star IRC+10420 has been traced as far out in SiO J=2-1 as in CO J = 1-0 and CO J = 2-1, in dramatic contrast with the centrally condensed (thermal) SiO- but extended CO-emitting envelopes of giant and supergiant stars. Here, we present an observation of the circumstellar envelope in SiO J=1-0 that, when combined with the previous observation in {sioii}, provide more stringent constraints on the density of the SiO-emitting gas than hitherto possible. The emission in SiO peaks at a radius of $sim$2arcsec whereas that in SiO J=2-1 emission peaks at a smaller radius of $sim$1arcsec, giving rise to their ring-like appearances. The ratio in brightness temperature between SiO J=1-0 and SiO J=2-1 decreases from a value well above unity at the innermost measurable radius to about unity at radius of $sim$2arcsec, beyond which this ratio remains approximately constant. Dividing the envelope into three zones as in models for the CO J = 1-0 and CO J = 2-1 emission, we show that the density of the SiO-emitting gas is comparable with that of the CO-emitting gas in the inner zone, but at least an order of magnitude higher by comparison in both the middle and outer zones. The SiO-emitting gas therefore originates from dense clumps, likely associated with the dust clumps seen in scattered optical light, surrounded by more diffuse CO-emitting interclump gas. We suggest that SiO molecules are released from dust grains due to shock interactions between the dense SiO-emitting clumps and the diffuse CO-emitting interclump gas.
Yellow and red supergiants are evolved massive stars whose numbers and locations on the HR diagram can provide a stringent test for models of massive star evolution. Previous studies have found large discrepancies between the relative number of yellow supergiants observed as a function of mass and those predicted by evolutionary models, while a disagreement between the predicted and observed locations of red supergiants on the HR diagram was only recently resolved. Here we extend these studies by examining the yellow and red supergiant populations of M33. Unfortunately, identifying these stars is difficult as this portion of the color-magnitude diagram is heavily contaminated by foreground dwarfs. We identify the red supergiants through a combination of radial velocities and a two-color surface gravity discriminant and, after re-characterizing the rotation curve of M33 with our newly selected red supergiants, we identify the yellow supergiants through a combination of radial velocities and the strength of the OI $lambda$7774 triplet. We examine ~1300 spectra in total and identify 121 yellow supergiants (a sample which is unbiased in luminosity above log(L/Lodot) ~ 4.8) and 189 red supergiants. After placing these objects on the HR diagram, we find that the latest generation of Geneva evolutionary tracks show excellent agreement with the observed locations of our red and yellow supergiants, the observed relative number of yellow supergiants with mass and the observed red supergiant upper mass limit. These models therefore represent a drastic improvement over previous generations.
We initiated long-term optical interferometry monitoring of the diameters of unstable yellow hypergiants (YHG) with the goal of detecting both the long-term evolution of their radius and shorter term formation related to large mass-loss events. We observed HR5171 A with AMBER/VLTI. We also examined archival photometric data in the visual and near-IR spanning more than 60 years, as well as sparse spectroscopic data. HR5171A exhibits a complex appearance. Our AMBER data reveal a surprisingly large star for a YHG R*=1315+/-260Rsun (~6.1AU) at the distance of 3.6+/-0.5kpc. The source is surrounded by an extended nebulosity, and these data also show a large level of asymmetry in the brightness distribution of the system, which we attribute to a newly discovered companion star located in front of the primary star. The companions signature is also detected in the visual photometry, which indicates an orbital period of Porb=1304+/-6d. Modeling the light curve with the NIGHTFALL program provides clear evidence that the system is a contact or possibly over-contact eclipsing binary. A total current system mass of 39^{+40}_{-22} solar mass and a high mass ratio q>10 is inferred for the system. The low-mass companion of HR5171 A is very close to the primary star that is embedded within its dense wind. Tight constraints on the inclination and vsini of the primary are lacking, which prevents us from determining its influence precisely on the mass-loss phenomenon, but the system is probably experiencing a wind Roche-Lobe overflow. Depending on the amount of angular momentum that can be transferred to the stellar envelope, HR5171 A may become a fast-rotating B[e]/Luminous Blue Variable (LBV)/Wolf-Rayet star. In any case, HR5171 A highlights the possible importance of binaries for interpreting the unstable YHGs and for massive star evolution in general.
Among evolved massive stars likely in transition to the Wolf-Rayet phase, IRC +10420 is probably one of the most enigmatic. It belongs to the category of yellow hypergiants and it is characterized by quite high mass loss episodes. Even though IRC +10420 benefited of many observations in several wavelength domains, it has never been a target for an X-ray observatory. We report here on the very first dedicated observation of IRC +10420 in X-rays, using the XMM-Newton satellite. Even though the target is not detected, we derive X-ray flux upper limits of the order of 1--3 10^-14 erg cm^-2 s^-1 (between 0.3 and 10.0 keV), and we discuss the case of IRC +10420 in the framework of emission models likely to be adequate for such an object. Using the Optical/UV Monitor on board XMM-Newton, we present the very first upper limits of the flux density of IRC +10420 in the UV domain (between 1800 and 2250 A, and between 2050 and 2450 A). Finally, we also report on the detection in this field of 10 X-ray and 7 UV point sources, and we briefly discuss their properties and potential counterparts at longer wavelengths.
The yellow hypergiant stars (YHGs) are extremely luminous and massive objects whose general properties are poorly known. Only two of this kind of star show massive circumstellar envelopes, IRC+10420 and AFGL2343. We aim to study the chemistry of the circumstellar envelopes around these two sources, by comparison with well known AGB stars and protoplanetary nebulae. We also estimate the abundances of the observed molecular species. We have performed single-dish observations of different transitions for twelve molecular species. We have compared the ratio of the intensities of the molecular transitions and of the estimated abundances in AFGL2343 and IRC+10420 with those in O-rich and C-rich AGB stars and protoplanetary nebulae. Both YHGs, AFGL2343, and IRC+10420, have been found to have an O-rich chemistry similar to that in O-rich AGB stars, though for AFGL2343 the emission of most molecules compared with 13CO lines is relatively weak. Clear differences with the other evolved sources appear when we compare the line intensity corrected for distance and the profile widths which are, respectively, very intense and very wide in YHGs. The abundances obtained for IRC+10420 agree with those found in AGB stars, but in general those found in AFGL2343, except for 13CO, are too low. This apparently low molecular abundance in AFGL2343 could be due to the fact that these molecules are present only in an inner region of the shell where the mass is relatively low.