1 EXPLOSIVE NUCLEOSYNTHESIS M. Hernanz1,2 1Institut d’Estudis Espacials de Catalunya, IEEC, Edifici Nexus, C/Gran Capit`a, 2-4, 08034 Barcelona, Spain 2Instituto de Ciencias del Espacio, CSIC ABSTRACT km s−1, energies involved 1045 erg and ejected masses −3 −5 between 10 and 10 M⊙. Many radioactive nuclei relevant for gamma-ray as- The radioactive isotopes synthesized during explo- trophysics are synthesized during explosive events, sive nucleosynthesis, either in novae or in supernovae, such as classical novae and supernovae. A review of are summarized in table 1. Three types of decay recent results of explosive nucleosynthesis in these chains can occur: electron captures (56Ni → 56Co, scenarios will be presented, with a special emphasis 57Ni → 57Co → 57Fe, 44Ti → 44Sc and 7Be → on the ensuing gamma-ray emission from individual 7Li), β+ decays (56Co → 56Fe, 44Sc → 44Ca, 26Al nova and supernova explosions. The influence of the → 26Mg and 22Na → 22Ne) and β− decays (60Fe dynamic properties of the ejecta on the gamma-ray → 60Co → 60Ni). The first six isotopes in the table emission features, as well as the still remaining un- (56Ni, 56Co, 57Ni, 44Ti, 26Al and 60Fe) are produced certainties in nova and supernova modelling will also in supernova explosions (although not exclusively, at be reviewed. least in the case of 26Al), whereas the last two are synthesized in classical novae. In sections 2 and 3 be- Key words: gamma rays: observations; novae; su- low, I will discuss how the synthesis of these radioac- pernovae; nuclear reactions, nucleosynthesis, abun- tive nuclei proceeds in these scenarios. In the case dances. of core-collapse supernovae, it is important to dis- tinguish the nucleosynthesis during the pre-explosive stage of the massive star evolution from that in the 1. INTRODUCTION explosive phases; it is crucial as well to know which part of the star will finally be ejected, since this quan- arXiv:astro-ph/0103418v1 26 Mar 2001 tity will determine the final enrichment of the Galaxy In this paper I will review the sites of explosive nucle- in radioactive (and other) elements. I won’t discuss osynthesis relevant for gamma-ray astronomy. Two the synthesis of radioactive isotopes in other (non- main types of explosion are responsible for the emis- explosive) sites, like the AGB stars (see the contri- sion of gamma-rays in the Galaxy: supernovae, both bution by Mowlavi, these proceedings) or the Wolf- thermonuclear and gravitational (core-collapse) and Rayet stars (which can eject radioactive nuclei by classical novae. strong stellar winds, during their hydrostatic evolu- tion; see Arnould and Meynet 1997 and Meynet and Thermonuclear supernovae, or supernovae of type Arnould, 1999 for recent reviews). Ia, are exploding white dwarfs in close binary sys- tems, which do not leave a remnant after the explo- sion. Core collapse supernovae (supernovae of type 2. SYNTHESIS OF RADIOACTIVE ISOTOPES II, Ib/c) are exploding massive stars (M∼> 10M⊙), IN SUPERNOVA EXPLOSIONS which leave as remnant either a black hole or a neu- tron star. Typical velocities of supernovae ejecta are some 104 km s−1, energies involved 1051 erg and Two types of isotopes can be distinguished, depend- ejected masses some M⊙. ing on their lifetime (see Diehl and Timmes, 1998, for a recent review). Short-lived isotopes, such as Classical novae are the result of the explosion of the 56Ni, 57Ni (and their daughters 56Co and 57Co), 44Ti external H-rich accreted shells of a white dwarf in a and 60Co, have lifetimes short enough (see table 1) to binary system. These explosions are recurrent phe- make them detectable in individual objects. 56Ni and nomena (contrary to supernovae), since an explosion 57Ni are produced in all types of supernovae; 44Ti is is expected every time the critical accreted mass on mainly produced in core-collapse supernovae, but it top of the white dwarf is reached. Typical veloc- can also be synthesized in thermonuclear supernovae ities of novae ejecta are between some 102 and 103 of the sub-Chandrasekhar type (provided that they 2 exist; see discussion of SNeIa types below). 60Co Sandie et al. 1988, Cook et al. 1988, Rester et al. is produced directly and from 60Fe decay, with 60Fe 1989). One surprising fact was the appearence of belonging to the long-lived isotopes group. these lines only 200 days after the explosion (Matz et al., 1988), much earlier than expected. This has Long-lived radioactive isotopes, such as 26Al and been interpreted as a sign of some early extra mixing 60Fe, have lifetimes long enough to make them unde- of 56Co in order to transport this isotope into regions tectable in individual sources, because the nuclei can of low gamma-ray optical depth (see, e.g., Pinto and be quite far away from their source and mixed with Woosley, 1988, Leising 1988, Bussard, Burrows and those coming from other explosions (since the life- The, 1989, Leising and Share, 1990). The line pro- time is longer than the typical period between two files observed with GRIS (Teegarden et al. 1989, succesive explosions in the Galaxy). For these iso- Tueller et al. 1990) have also put constraints on theo- topes, only the accumulated emission in the Galaxy retical models of supernova explosions, since spheric- can be observed and used as diagnostic of models and ity and homogeneity of the ejecta were incompatible of the Galactic distribution of the sources. The same with the observed fluxes and widths of the 56Co lines. classification scheme applies to isotopes synthesized in novae; in this case, 7Be belongs to the short-lived Another crucial gamma-ray observation of short- group, whereas 22Na belongs to both of them (see lived isotopes in SN 1987A was the detection of section 3 below). gamma-ray radiation from 57Co decay (between 50 and 136 keV), with OSSE on the CGRO (Kurfess et A very recent and interesting compilation of papers al. 1992). The deduced 57Co content (for models about astronomy with radioactivities can be found with low gamma-ray optical depth, see Kurfess et al. in Diehl and Hartmann (1999). 1992) was such that the original ratio 57Ni/56Ni pro- duced in the explosion should be about 1.5 times the solar 57Fe/56Fe ratio . Observations up to now show 56 2.1. Observational clues the change of slope related to the sequence of Co- 57Co decays (see figure 3 in Timmes et al. 1996). Future UVOIR observations would possibly be able Gamma-ray astronomy provides an unique opportu- to show the light curve powering from 44Ti. The SPI nity to detect radioactive isotopes in individual ob- instrument onboard INTEGRAL has some possibili- jects, giving a proof of ongoing nucleosynthesis in ties to detect the gamma-ray emission from this 44Ti, them. In the case of supernovae, another tool for the which would provide a unique proof of the nucleosyn- determination of the amount of radioactive nuclei in thesis in core collapse supernovae and an important the ejecta is the bolometric light curve (UVOIR, from link between the UVOIR and the gamma-ray obser- ultraviolet, optical and infrared). In addition to the vations. Coming back to 57Co it was first thought well known fact that 56Co (daughter of 56Ni) powers that the SN 1987A 57Co content deduced from OSSE the early evolution of the light curve (56Ni mass can observations was not enough to power the available be determined from luminosity at maximum), 57Co is bolometric light curve, since 5 times solar 57Fe/56Fe responsible for powering the light curve from around ratio was required (Suntzeff et al. 1992) and alterna- day 1000 after maximum. Later on 44Ti will pro- tive mechanisms to power the bolometric light curve vide a floor to the bolometric light curve and 22Na were suggested (Clayton et al. 1992). However, more and 60Co could also play a minor role, depending on recent observations seem to require a smaller amount the supernova type and the specific yields of these of 57Co-decay to power the light curve, in agreement radioactivities (see, e.g., Woosley, Pinto and Hart- with the 57Fe/56Fe ratio deduced from OSSE obser- mann, 1989, Timmes et al. 1996, Diehl and Timmes vations (see figure 9 in Diehl and Timmes 1998). 1998). The excitement induced by the above mentioned These two types of observational approaches to the gamma-ray observations (together with many other radioactive content in the ejecta had been possible observations at other wavelength ranges) has led the for only one object so far: the supernova 1987A, theorists to suggest different possibilities for mixing which exploded 13 years ago in the LMC, only 55 kpc both during the explosion and the ejection phases away from us. This was a type II supernova, which (to quote only a few of the early works, see e.g., Ar- is not the most favorable case to look for gamma- nett, Fryxell and M¨uller 1989, Benz and Thielemann ray emission, since its is much more opaque and has 1990, Fryxell, Arnett and M¨uller 1991, Herant and a smaller content of the most relevant radioactivites Benz 1992, and also the general reviews of SN 1987A than SNIa (see below); however, its very short dis- from Arnett et al. 1989 and McCray 1993 and refer- tance allowed for detection of its gamma-ray emission ences therein). and also for a follow-up until very late times of its light curve (through photometry in the UVBRIJHK Another detection (tentative) of gamma-ray emission bands). from a supernova, with COMPTEL on CGRO, was that of SN 1991T (which was an overluminous SNIa), Gamma-ray lines from 56Co decay, at 847 and 1238 in NGC 4527 at around 17 Mpc distance.