Results and discussion

The NLTE computations reduce random errors and remove systematic trends from the analysis. Inappropriate LTE analyses tend to underestimate iron-group abundances systematically and overestimate the light- and a-process-element abundances by factors of up to 2-3 (this is most notable for M31-41-3654, while M31-41-3712 is less affected). This is because of the different responses of these species to radiative and collisional processes in the microscopic picture, which is explained by fundamental differences of their detailed atomic structure. This is not taken into account in LTE. Contrary to common assumptions, significant NLTE abundance corrections of ~0.3dex can be found even for the weakest lines (W-k ~ 10 mÁ). Non-LTE abundance uncertainties amount to typically 0.05-0.10 dex (random) and ~0.10dex (systematic 1a errors). Note that line-blocking effects increase with metallicity, such that photon mean free paths are reduced in metal-rich environments and the NLTE effects correspondingly. This is the reason why NLTE effects in these objects close to the Eddington limit are similar to those in less-extreme Galactic BA-type supergiants with - on average - lower metallicity.

Fundamental stellar parameters and light-element abundances allow us to discuss the two M31 objects in the context of stellar evolution. The comparison with evolutionary tracks for rotating massive stars is made in Figure 35.2, which also summarises results from a Galactic sample of BA-SGs. The M31 supergiants extend the sample towards higher luminosities/stellar masses than possible in the Galactic study and towards higher metallicity. Both objects appear to cross the HertzsprungRussell diagram towards the red supergiant stage for the first time, because of the absence of the extremely high helium abundances and N/C ratios expected for stars entering the Wolf-Rayet phase. The predicted trend of increasing chemical mixing (strong N and moderate He enrichment, C depletion and almost constant O as a result of the action of the CNO-cycle and transport to the stellar surface because of meridional circulation and dynamical instabilities) with increasing stellar mass is qualitatively recovered. Note a group of highly processed stars at M0 < 15M©, which suggests an extension of blue loops towards higher temperatures than predicted. The observed N/C ratios are generally higher than indicated by theory. Stellar-evolution computations accounting for the interplay of rotation and magnetic fields (e.g. Maeder & Meynet 2005) may resolve this discrepancy since they predict a much higher efficiency for chemical mixing. Also the recent revision of the cross-section for the bottleneck reaction 14N(p,y)15O in the CN-branch of the CNO-cycle by almost a factor of 2 (Lemut et al. 2006) will be of importance.

Finally, improvements in stellar parameters allow re-evaluation of the two M31 supergiants in the empirical calibration of the FGLR (Kudritzki et al. 2003). Two factors play a role in this context: a revision of the previously used photometric

Figure 35.2. Observational constraints on massive-star evolution: N/C ratios as tracers of mixing with nuclear-processed material. Evolution tracks for rotating stars (vini = 300 km-1 s, with marks indicating N/C ratios from the models of Meynet & Maeder (2003)) at Solar metallicity are displayed. An initial N/C ~ 0.3 was adopted in their model computations. Observed N/C ratios (typical values are indicated in the box) for a sample of Galactic BA-SGs (Przybilla et al. 2006; Firnstein & Przybilla 2006; Schiller & Przybilla 2006) and for the two (slightly) metal-rich M31 objects of the present work are indicated, all analysed in a homogeneous way. Error bars characteristic for our work and for similar studies from the literature are given.

log Teff

Figure 35.2. Observational constraints on massive-star evolution: N/C ratios as tracers of mixing with nuclear-processed material. Evolution tracks for rotating stars (vini = 300 km-1 s, with marks indicating N/C ratios from the models of Meynet & Maeder (2003)) at Solar metallicity are displayed. An initial N/C ~ 0.3 was adopted in their model computations. Observed N/C ratios (typical values are indicated in the box) for a sample of Galactic BA-SGs (Przybilla et al. 2006; Firnstein & Przybilla 2006; Schiller & Przybilla 2006) and for the two (slightly) metal-rich M31 objects of the present work are indicated, all analysed in a homogeneous way. Error bars characteristic for our work and for similar studies from the literature are given.

data of Magnier et al. (1992), which have been shown to suffer from systematic uncertainties (Massey et al. 2006); and the extended wavelength coverage (and high signal-to-noise ratio) of the ESI spectra, which allows the atmospheric parameters to be constrained more precisely. As a consequence, some of the largest deviations from the empirical relation can be explained.

References

Blair, W. P., Kirshner, R. P., & Chevalier, R. A. (1982), ApJ 254, 50-69 Dennefeld, M., & Kunth, D. (1981), AJ 86, 989-997

Firnstein, M., & Przybilla, N. (2006), Proceedings of Science, PoS(NIC-IX)095

Grevesse, N., & Sauval, A. J. (1998), Space Sci. Rev. 85, 161-174

Kudritzki, R. P., Bresolin, F., & Przybilla, N. (2003), ApJ 582, L83-L86

Kurucz, R. L. (1993), Kurucz CD-ROM No. 13 (Cambridge, MA: SAO)

Lemut, A., Bemmerer, D., Confortola, F. etal. (2006), Physics Letters B 634, 483-487.

Maeder, A., & Meynet, G. (2005), A&A 440, 1041-1049

Meynet, G., & Maeder, A. (2003), A&A 404, 975-990

Magnier, E. A., Lewin, W. H. G., van Paradijs, J. et al. (1992), A&AS 96, 379-388

Massey, P., Olsen, K. A. G., Hodge, P. W. et al. (2006), AJ131, 2478-2496

Przybilla, N. (2002), Ph.D. Thesis, Universitat München.

Przybilla, N., Butler, K., & Küdritzki, R. P. (2001), A&A 379, 936-954

Przybilla, N., Butler, K., Becker, S. R., & Küdritzki, R. P. (2006), A&A 445, 1099-1126

Schiller, F., & Przybilla, N. (2006), Proceedings of Science, PoS(NIC-IX)174

Smartt, S. J., Crowther, P. A., Dufton, P. L. et al. (2001), MNRAS 325, 257-272

Trundle, C., Dufton, P. L., Lennon, D. J. et al. (2002), A&A 395, 519-533

Venn, K. A., McCarthy, J. K., Lennon, D. J. et al. (2000), ApJ 541, 610-623

Was this article helpful?

0 0

Post a comment