Metallicity studies

The absence of late B supergiants and LBVs prevents one from obtaining direct estimates of the important a-elements versus Fe metallicity ratio as in the Quintuplet cluster and the Central cluster. Since it is the youngest cluster at the GC, any hint about its metallicity would constitute our "last-minute" picture of the chemical enrichment of the central region in the Milky Way. To analyze the stars in the Arches cluster we have assumed the atmosphere to be composed of H, He, C, N, O, Si, and Fe. Observational constraints are provided by the K-band spectra of the stars (see Figure 13.3) and the narrow-band HST/NICMOS photometry (filters Ff11ow, Ff16ow, and Ff205w) and Pa equivalent width (filters FF187N and FF190N). Object identifications are given according to Figer et al. (2002). Below we present the results of our analysis (Najarro et al. 2004).

The reduced spectra and model fits are shown in Figure 13.3. The top three spectra correspond to some of the most luminous stars in the cluster. As described in Figer et al. (2002), these are nitrogen-rich Wolf-Rayet stars with thick and fast winds. The bottom two spectra in Figure 13.3 correspond to slightly less-evolved stars with the characteristic morphology of OIf+ stars. Of concern are the N iii 8-7 lines at 2.103 |im and 2.115 |im as well as the N iii 5p 2P-5s 2S doublet at 2.247 |im and 2.251 |im. Figer et al. (1997) showed that these N iii lines appear only for a narrow range of temperatures and wind densities, which occur in the WN9h

1.40

0.90

1.30

1.20

1.10

1.00

0.90

1.30

1.20

1.10

1.00

time (Myr)

Figure 13.4. Left: the leverage of error estimates on N abundance. The observed 2.10-2.13-^m region (solid) is shown together with current best-fit (dashed) and 30% enhanced (dotted) and 30% depleted (dashed-dotted) nitrogen mass fractions. Right: nitrogen mass abundance versus time calculated using Geneva models. The measurements require Solar metallicity and an age of 2-2.5 Myr.

0.000

time (Myr)

Figure 13.4. Left: the leverage of error estimates on N abundance. The observed 2.10-2.13-^m region (solid) is shown together with current best-fit (dashed) and 30% enhanced (dotted) and 30% depleted (dashed-dotted) nitrogen mass fractions. Right: nitrogen mass abundance versus time calculated using Geneva models. The measurements require Solar metallicity and an age of 2-2.5 Myr.

(WNL) stage. The fairly distinct nature and energies of the multiplets involved in each of these two N iii line sets provide strong constraints for the determination of the nitrogen abundance. Thus, at the signal-to-noise ratio of our spectra, our models show that the WNLs N iii lines can easily track relative changes as small as 20% in the nitrogen abundance, and a 30% error should be regarded as a safe estimate, as shown in Figure 13.4 on the left.

Of particular importance is that we obtained roughly the same surface abundance fraction of N, Z(N), in our analysis for all three WNL objects (~1.6%), which is well above the upper limit found for the OIf+ stars (~0.6%). Although WNL stars do not exhibit any primary diagnostic line in their K-band spectra from which to estimate metallicity, the crucial role of Z(N) in determining their metallicity is immediately apparent if we make use of the stellar-evolution models for massive stars.

According to the evolutionary models of Schaller et al. (1992) and Charbonnel etal. (1993), a star entering the WNL phase still shows H atits surface, together with strong enhancement in levels of helium and nitrogen and strong depletion in levels of carbon and oxygen, as expected from processed CNO material. During this phase, the star maintains a nearly constant Z(N) value, which essentially depends only linearly on the original metallicity (see the right-hand panel of Figure 13.4), being basically unaffected by the mass-loss rate assumed and the occurrence of stellar rotation during evolution (Meynet & Maeder 2004). Since we expect the CNO abundance in the natal cloud to scale with that of the rest of metals, the nitrogen surface abundance must trace the metallicity of the cluster. The parameters derived for these stars (Najarro et al. 2004) indicate that this is indeed the case for objects #3, #4, and #8. The derived Z(N) (~1.6%) is that expected for Solar metallicity from the evolutionary models. The reliability of our method is demonstrated in the right-hand panel of Figure 13.4, where we display the nitrogen mass fraction as a function of time for stars with initial masses of 60, 85, and 120 times M0, and metallicities equivalent to 2, 1, and 0.4 times Solar, assuming the canonical mass-loss rates (Schaller et al. 1992). Our results for the WNL and O stars (the cross-hatched region) require Solar metallicity and an age of 2-2.5 Myr; see also Najarro et al. (2004).

Was this article helpful?

0 0

Post a comment