Chemical abundances and winds of massive stars in M31: a B-type supergiant and a WC star in OB 10

Abstract

We present high quality spectroscopic data for two massive stars in the OB 10 association of M31, OB 10‐64 (B0 Ia) and OB 10‐WR1 (WC6). Medium resolution spectra of both stars were obtained using the ISIS spectrograph on the William Herschel Telescope. This is supplemented with Hubble Space Telescope STIS UV spectroscopy and Keck I HIRES data for OB 10‐64. A non‐local thermodynamic equilibrium (LTE) model atmosphere and abundance analysis for OB 10‐64 is presented, indicating that this star has similar photospheric CNO, Mg and Si abundances to solar neighbourhood massive stars. A wind analysis of this early B‐type supergiant reveals a mass‐loss rate of M˙=1.6×10−6 M⊙ yr−1, and v∞=1650 km s−1. The corresponding wind momentum is in good agreement with the wind momentum–luminosity relationship found for Galactic early‐B supergiants. Observations of OB 10‐WR1 are analysed using a non‐LTE, line‐blanketed code, to reveal approximate stellar parameters of log L/L⊙∼5.7, T∗∼75 kK, v∞∼3000 km s−1, M˙/(M⊙ yr−1)∼10−4.3 adopting a clumped wind with a filling factor of 10 per cent. Quantitative comparisons are made with the Galactic WC6 star HD 92809 (WR23) revealing that OB 10‐WR1 is 0.4 dex more luminous, though it has a much lower C/He ratio (∼0.1 versus 0.3 for HD 92809). Our study represents the first detailed, chemical model atmosphere analysis for either a B‐type supergiant or a Wolf–Rayet (WR) star in Andromeda, and shows the potential of how such studies can provide new information on the chemical evolution of galaxies and the evolution of massive stars in the local Universe.

A large Wolf-Rayet population in NGC 300 uncovered by VLT-FORS2

“…Both #29 (WC5) and #48 (WC4) lie on, or slightly below, the Galactic calibration, as would be expected from the metallicities inferred from their locations in NGC 300 (Deharveng et al 1988;Zaritsky et al 1994). Of course, the difference between Galactic and LMC calibrations amounts to less than a factor of two in mass-loss rate, and Smartt et al (2001) discuss, in detail, limitations with adopting generic H  region abundances with regard to WC properties in M 31.…”

The evolutionary status of Sher 25 – implications for blue supergiants and the progenitor of SN 1987A

“…A 60 M star on the main-sequence would have a mass-loss rate of typically ∼2−5 × 10 −6 M yr −1 (Puls et al 1996), hence during its H-burning lifetime (4 Myrs; from the Geneva models) it would have lost 8-20 M of its envelope through its stellar wind In order to compare this meaningfully with the abundance ratios we derive, we must have confidence in the absolute abundances derived in our methods. To date the evaluation of this has been difficult, as pointed out in Smartt et al (2001a and Paper I. These studies have concentrated on differential analyses and comparisons of the difference between stellar subgroups, rather than assuming validity of the absolute abundances.…”

“…These were HD 2905, HD 14956 and HD 13854, which have very similar H , δ, γ lines and Si /Si  line strength ratios as Sher 25. These stars have been observed with the single-order spectrograph ISIS on the WHT, and the spectra have a coverage of 3900-4735 Å covering the main diagnostic lines of interest (see Smartt et al 2001a where the details of the instrument setup is described). The S/N of these spectra are very high (in excess of 200) and the line profiles are well resolved, hence they make excellent comparison templates.…”

“…The photosphere of Sher 25, and subsequently the three comparison stars, were modeled using the techniques described in detail in Paper I and briefly summarized in Smartt et al (2001a). A grid of non-LTE model atmospheres has previously been generated using the code  (Hubeny 1988) covering an effective temperature range 10 000 ≤ T eff ≤ 35 000 and gravities from log g = 4.5 to the Eddington stability limit at each T eff .…”

The oxygen abundance gradient in M 101: The reliability of thePmethod

“…Moreover, for M 101 Kennicutt et al (2003) have recently found systematic differences up to a factor 3 between abundances derived from some empirical calibrations and those derived from the direct method, the latter being lower. Smartt et al (2001) and Trundle et al (2002) have shown that in the Local Group spiral galaxy M 31 oxygen abundances of B supergiant atmospheres are also systematically lower than those obtained by classical R 23 -O /H calibrations. However, Monteverde et al (2000) found very good agreement between B supergiant abundances (obtained in a similar way as in the aforementioned references) and the abundances derived from the T e method in the interstellar medium in M 33.…”