Gamma-ray line spectroscopy is one of the main observational goals of the COMPTEL experiment aboard NASA's Compton Gamma-Ray Observatory (CGRO) (Schönfelder et al. 1993). In particular, several radioactive products of stellar nucleosynthesis generate line emission in the 0.75- 30 MeV energy range covered by this instrument (see Diehl 1995). For the first time, the celestial 1.809 MeV line emission from 26 Al decay has now been mapped by COMPTEL (Diehl et al. 1995, Oberlack et al. 1996).
44 Ti is another radio-isotope of astrophysical interest. As already stressed long ago (Bodansky, Clayton & Fowler 1968), it is the sole parent of natural 44 Ca. The rate of 44 Ti nucleosynthesis and supernova explosions sets the Galactic abundance of this species (Clayton 1982, Woosley 1993). Radioactive 44 Ti decays into stable 44 Ca via 44 Sc:
The mean lifetime of 44 Ti is still uncertain. Values of as different as 78.2 yr (Frekers et al. 1983) and 96.1 yr (Adelberger & Harbottle 1990) have been proposed. More recent measurements, yet preliminary, encompass this range with (Meißner et al. 1995).
Supernovae are the site of 44 Ti nucleosynthesis through explosive Si burning (see Hoffman et al. 1995, Timmes et al. 1996 for reviews). Production is most efficient within an alpha-rich freeze-out of nuclear statistical equilibrium, at low densities. This process operates in core-collapse events (Woosley, Arnett & Clayton 1973, Thielemann, Hashimoto & Nomoto 1990), while a normal freeze-out Si burning is rather at play in Type-Ia objects (Thielemann, Nomoto & Yokoi 1986). The amount of 44 Ti ejected by the supernova depends on the mechanism of explosion and the progenitor mass and metallicity. It is quite sensitive to the mass cut between the remnant and the ejecta, i.e. to how much mass falls back onto the core. Different models of core-collapse supernovae predict values of the yield between a few and (e.g. Woosley & Weaver 1995, Woosley, Langer & Weaver 1996, Thielemann, Nomoto & Hashimoto 1996), while the Type-Ia model of Nomoto, Thielemann & Yokoi (1984) only produces of 44 Ti.
In contrast with the slow decay of 26 Al (), the lifetime of 44 Ti is rather short. While 1.809 MeV emission regions must be due to an ensemble of sources as a result of 26 Al accumulation in the interstellar medium over several million years (Diehl et al. 1996), only fairly recent, rather nearby, individual explosive events are detectable through their 44 Ti afterglow. The daughter 44 Sc line at probes the supernova activity of the Galaxy within the last few centuries. It was amongst the first predicted gamma-ray diagnostics of young supernova remnants (SNR s hereafter) (Clayton, Colgate & Fishman 1969).
Due to the lack of spatial resolution, previous gamma-ray spectroscopy experiments only put constraints on the global Galactic emission in the lines associated with 44 Ti decay. Using HEAO 3, Mahoney et al. (1992) give an upper limit of for the 67.9 and 78.4 keV lines at the 99% confidence level. Measurements by SMM provide a more stringent limit: the flux from the inner of the Milky Way cannot exceed at a similar level (Leising & Share 1994).
Because of its better sensitivity and spatial resolution, COMPTEL is suited for a systematic survey of the Galaxy in the line. Cassiopeia A, the youngest known Galactic SNR, has already been detected by COMPTEL (Iyudin et al. 1994) with a flux of about . Further analysis of more data however points to a slightly lower value, (Schönfelder et al. 1996). At first, this result was not confirmed by OSSE onboard the Compton Observatory (The et al. 1995). More recently, a marginal detection was reported by The et al. (1996) with a 44 Ti line flux of . Within uncertainties, the latest COMPTEL and OSSE determinations are in agreement.
In the present work, we combine the first three years of COMPTEL data (Sect. 2) in order to carry out a complete search of the Galactic plane for 44 Ti sources (Sect. 3). The purpose of this survey is twofold:
© European Southern Observatory (ESO) 1997
Online publication: May 26, 1998