Astron. Astrophys. 363, 970-983 (2000)

6. Conclusions

We have constructed a new diagnostic tool for the study of the radioactive isotopes of ²⁶Al and ⁶⁰Fe produced in massive star forming regions. The main aim of this work was to provide a quantitative model for the analysis of multi-wavelength observations of OB associations, open clusters and alike objects covering the range from gamma-rays (e.g. the 1.809 MeV line of ²⁶Al, 1.173 and 1.333 Mev lines of ⁶⁰Fe) to radio, and allowing in a fully quantitative manner to account for statistical richness effects of massive star populations and other observational uncertainties.

To achieve this goal we have used the evolutionary synthesis models of Cerviño & Mas-Hesse (1994), which have been updated to include recent Geneva stellar evolution tracks, new stellar atmospheres for OB and WR stars, and nucleosynthetic yields from massive stars during hydrostatic burning phases and explosive SNII and SNIb events (see Sect. 2). In particular proper care was taken to combine the stellar models including mass loss with appropriate presupernova and SN models.

The temporal evolution of the ejected quantity of ²⁶Al and ⁶⁰Fe produced by a coeval population, other observables like the total ionising flux and the supernova rate, and derived properties is presented (Sect. 3). This yields the following main results:

• The equivalent O7 V star ²⁶Al yield (), defined as the stellar yield per ionising flux of an O7 V star (Knödlseder 1999), shows a particularly strong time dependence where four main phases can be distinguished: stellar wind dominated phase ( 3 Myr), SN Ib dominated phase ( 3-7 Myr), SN II dominated phase ( 7-37 Myr), and exponential decay phase ( 37 Myr). The exact age range of each phase is dependent on the evolutionary tracks and the lower mass limit of WR stars.

• The use of is a powerful tool to constrain the evolutionary status of star forming regions. This parameter is obtained from a combined observation of the -ray line and the ionising flux (for example in form of thermal free-free radio emission) of an association.

• ⁶⁰Fe production starts with a delay of 2 Myr with respect to ²⁶Al production. The ratio of the ⁶⁰Fe/²⁶Al emissivities is also an age indicator that constrains the contribution of explosive nucleosynthesis to the total ²⁶ Al production.

Calculations for a steady state population (constant star formation; Sect. 3.5) at solar metallicity predict the following relative contributions to the ²⁶Al production: from stars before the WR phase, from WR stars, from SN Ib, and from SN II. The large contribution from stellar wind ejection ( 42 %) confirms earlier studies of MAPP97 and Knödlseder (1999) using similar yields, who predict contributions of 20-70 % and 40 % respectively. For ⁶⁰Fe we estimate that are produced by SN Ib while come from SN II. Normalising on the total ionising flux of the Galaxy, we predict total production rates of Myr^-1 and Myr^-1 for ²⁶Al and ⁶⁰Fe, respectively. This corresponds to 1.5 of ²⁶Al and 1.7 of ⁶⁰Fe in the present interstellar medium.

As for other chemical evolution models, our calculations depend directly on the adopted nucleosynthetic yields, which are affected by considerable uncertainties (see e.g. Prantzos 1999). The main uncertainties regarding ²⁶Al and ⁶⁰Fe have been discussed in Sects. 2 and 4. In fact important new insight especially on the physics of supernovae is expected from the study of radioactive isotopes such as ⁶⁰Fe and ⁴⁴Ti, which are synthesised in deep layers close to the so-called mass cut separating the outer regions from the remnant. Such studies should also benefit from the present models.

Last, but not least, we have presented a Bayesian approach to quantify the predicted observables and their uncertainty related to richness effects of the IMF in terms of probability density functions (Sect. 5). Subsequently these functions can be used in combination with prior knowledge on observed objects (e.g. age, distance, and their uncertainties) to calculate detection probabilities and alike quantities. We have already successfully applied our models to existing multi wavelength observations of the Cygnus and Vela regions. The results will be published in companion papers (Knödlseder et al., in preparation; Lavraud et al., in preparation). Our tools will be ideal to fully exploit the gamma-ray line observations expected from the upcoming INTEGRAL satellite.

© European Southern Observatory (ESO) 2000

Online publication: December 5, 2000
helpdesk@link.springer.de