Astron. Astrophys. 319, 909-922 (1997)

1. Introduction

From observations with the HEAO -ray satellite it has been known for more than a decade that there is a noticeable -ray line emission at an energy of 1.809 MeV attributed to the decay ²⁶ Al Mg of the radioactive isotope ²⁶ Al on a mean lifetime of 1.05 10⁶ yr (Mahoney et al. 1984). The Galaxy is optically thin to these 1.8 MeV photons, which interact with matter mainly via Compton scattering, so that the ²⁶ Al emission can be seen from the entire Galaxy. As the angular resolution of HEAO was very poor, no information about the spatial distribution of the 1.8 MeV emission could be extracted. Assuming that the emission is diffuse but concentrated towards the Galactic plane, a total amount of ²⁶ Al homogeneously distributed in the Galactic disk was derived to account for the observed intensity at 1.8 MeV (e.g., Mahoney et al. 1984).

Only recently, with the successful operation of the imaging -ray telescope COMPTEL aboard the COMPTON Gamma Ray Observatory, has it been been possible to make a detailed investigation of the angular distribution of the 1.8 MeV emission on the sky. From data obtained during the first (all-sky survey) mission phase, Diehl et al. (1995) compiled a 1.8 MeV map along the Galactic plane revealing an unexpected clumpiness and asymmetry relative to the Galactic center. This finding was confirmed by Oberlack et al. (1996a, b) who combined 3.5 years of observations to construct the first all-sky map of the ²⁶ Al decay line. No emission was found at high galactic latitudes , rejecting a purely local origin. Although the maximum entropy method used to create the ²⁶ Al map may exaggerate the clumpiness, it seems now well-established that at least part of the 1.8 MeV emission comes from localized regions. This imaging technique does not allow a positive identification of individual 1.8 MeV sources with counterparts in other wavebands, but there are unambiguous hotspots close to the Galactic center in the Vela, Carina and Cygnus regions (see the summary by Oberlack et al. 1996a, b). A number of authors find a resemblance between the structure in the ²⁶ Al map and tracers of rather close-by regions of recent star formation (Diehl et al. 1995, Timmes et al. 1995), and Chen et al. (1995) argue that the observed asymmetry could be explained by the local spiral structure and the existence of a central stellar bar. An important further consequence of these new data is that the amount of ²⁶ Al in a diffuse background may be much less than previously thought. Diehl et al. (1995) derive an upper limit of , although more recent reinvestigations indicate that a somewhat larger value cannot be excluded (Oberlack, priv. comm).

Several astrophysical environments have been proposed as sites for the production of ²⁶ Al in the Galaxy, all related to either explosive nucleosynthesis or hydrostatic hydrogen burning: besides oxygen-neon-magnesium novae (ONeMg novae) which are discussed below, these are, core-collapse supernovae (SNe; e.g, Timmes et al. 1995), Wolf-Rayet stars (e.g., Pranztos & Cassé 1986, Meynet 1994, Langer et al. 1995) and asymptotic giant branch (AGB) stars (e.g., Guélin et al. 1995, Forestini et al. 1991). Estimates have been made which suggest that each of these source groups alone could potentially account for all of the observed 1.8 MeV emission. However, the emission from individual nova outbursts or AGB stars is expected to be too weak to be detected as isolated sources, rather the corresponding population would create a diffuse background radiation at 1.8 MeV. In contrast, SNe and Wolf-Rayet stars are thought to produce 10 to 100 times as much ²⁶ Al per event/object, so that they would not only contribute to a background, but could also appear as individual sources. Moreover, as they are the youngest objects, and therefore are more closely associated with regions of recent star formation, they appear to be the best candidates to explain the observed clumpiness at 1.8 MeV. For a detailed overview of relevant observations and a critical discussion concerning suggested sources of interstellar ²⁶ Al, we refer the reader to the very nice review by Prantzos & Diehl (1996).

In this paper, we specifically investigate ONeMg novae as sites of ²⁶ Al production in the Galaxy. An ONeMg nova is believed to be a classical nova system in which the thermonuclear runaway (TNR) occurs on an ONeMg WD. The numerous observations of classical novae whose ejecta show significant enrichments in intermediate-mass elements, particularly neon, in the last ten years strongly support the existence of such systems (e.g., Politano et al. 1995; Starrfield et al. 1996 and references therein), although there is some question as to how much neon demands the presence of an underlying ONeMg WD (e.g., Livio & Truran 1994). Essential for significant production of ²⁶ Al during the TNR is the presence of a sufficient amount of ²⁴ Mg seed nuclei (e.g., Weiss & Truran 1990), which is believed to enter the envelope from the underlying WD material through some mixing process or processes.

Previous estimates of the amount of ²⁶ Al produced by ONeMg novae range from (Weiss & Truran 1990) to the full 3 assuming the lower limit to the WD mass of ONeMg WDs in cataclysmic variables (CVs) is 1.2 (Politano et al. 1995). However, there are problems with each calculation. Weiss & Truran based their estimate purely on average values of quantities such as the amount of material ejected in an ONeMg nova, the mass fraction of ²⁶ Al in the ejecta of an ONeMg nova and the fraction of novae which are ONeMg, neglecting any detailed dependence of on these parameters. Politano et al. (1995) calculated the amount of ²⁶ Al from an integral over WD mass but still relied on a mean accretion rate in nova systems based on the nova rate in the Galaxy. Moreover, their ²⁶ Al mass fractions are based on the particular accretion rate chosen for their models.

Our purpose in this paper is therefore twofold: (1) to improve on earlier estimates of ²⁶ Al production in ONeMg novae and compare the improved estimate with the recent COMPTEL data, and (2) to investigate how the new, tighter observational constraints on the ²⁶ Al background may help to constrain free parameters in our theoretical understanding of both the galactic population of CVs and the nucleosynthesis during the TNR in classical nova outbursts.

In particular, we improve on earlier estimates in two regards. First, we obtain the ²⁶ Al mass fraction in nova ejecta as a function of WD mass, initial enrichment of the accreted envelope, and accretion rate from a grid of detailed ONeMg nova models, calculated with a 1-dimensional hydrodynamic code coupled to an extended nuclear network (Politano et al. 1995, 1996), thereby exploring a greater range of parameter space than was done in Politano et al. (1995). Second, we combine these data with a detailed background model for the Galactic CV population (Kolb 1993a). As a result we come up with the first self-consistent prediction for the ²⁶ Al production in ONeMg novae solely from theoretical models.

We present data from the TNR simulations which are important for this study in Sect. 2. The CV population models and their relationship to the classical nova population are described in Sect. 3. We show and discuss our results in Sects. 4 and 5, respectively. Main conlusions follow at the end of the paper.

© European Southern Observatory (ESO) 1997

Online publication: July 3, 1998
helpdesk.link@springer.de