1 Introduction

The capture of a neutron by a proton and the subsequent emission of a photon at 2.223 MeV has been the object of much interest since the dawn of gamma-ray astrophysics (Fichtel & Trombka 1981) because it represents a potentially important spectroscopic tool in the investigation of high-energy environments such as solar flares, neutron star surfaces or accretion disks around compact objects. Several studies have been conducted both from theory and observation, realizing that this line is the most promising candidate of all nuclear gamma-ray processes.

Solar flares have had the lion's share of works because of the obvious possibilities of detailed investigation: Hua & Lingenfelter (1987), Murphy et al. (1991), Share & Murphy (1995), Ramaty et al. (1995), Ramaty et al. (1996); and others. Other sources have also been investigated, although no detection has been confirmed and established to date, despite several previous announcements (Jacobson 1982; McConnell et al. 1997).

On the theoretical front, Guessoum & Dermer (1988) conducted a detailed theoretical investigation in the context of compact binary sources, taking Cygnus X-1 as a case-study and pointing out the possibility of emission of a very narrow 2.22 MeV line from the atmosphere of the companion, while Aharonian & Sunyaev (1984) had considered a two-temperature accretion disk and considered the possibility of emission from the disk itself, where the line would be very broad. Bildsten and co-workers considered the possibility of emission at the surface of a neutron star, as a result of bombardment by accreted protons (Bildsten 1991; Bildsten et al. 1993). In the same framework, Bykov et al. (1999) recently considered the possibility of detecting nuclear gamma-ray lines from accreting objects with INTEGRAL. Vestrand (1989) investigated the 2.22 MeV emission resulting from very high energy ( E > 10¹² eV) protons bombarding companion stars of sources such as Cyg X-3, Vel X-1, and Her X-1, and predicted significant fluxes ($\approx$10^-4 photons cm^-2 s^-1). Finally, Guessoum & Kazanas (1999), interested in Lithium production by neutron bombardment of X-ray binary companions' atmospheres and using the ADAF accretion disk model, estimated the flux of 2.22 MeV line emission from a source at about 1 kpc and found it generally very low (10^-7 photons cm^-2s^-1).

On the observational front, extensive searches of the line have been conducted with the data from the Solar Maximum Mission (Harris & Share 1991) and the COMPTEL instrument of the Compton-GRO mission (McConnell et al. 1997). The latter showed an excess emission at (l, b) = 300$^{\rm o}$, $-30^{\rm o}$at about 3.7 $\sigma$. No obvious counterparts (X-ray binaries or the like) could be linked to that general direction, although a catalysmic variable has been suggested as a likely candidate for that emission. Van Dijk (1996) has compiled a list of upper-limit 2.22 MeV binary-source radiation fluxes based on COMPTEL observations of 27 black-hole candidates. These upper limits range from 1 to $5~\times~10^{-5}$ photons cm^-2 s^-1.

The production of neutrons in the accretion disks around the primaries of X-ray binaries is a direct result of the high temperatures that prevail in such environments. The gravitational attraction, combined with the viscous dissipation, leads to a significant heating of the plasma, particularly its nuclei, resulting in temperatures as high as 10¹⁰-10¹² K in the inner regions ( $r \approx 1 {-} 100$ $R_{\rm S}$, where $R_{\rm S} = 2GM/c^2$ is the Schwarzchild radius of the central compact object). Simple considerations show that densities of the accretion plasma can range between 10¹²and 10¹⁸ cm³ depending on the geometry and the model, which implies substantial nuclear reactions. Since the accreted material is expected to contained He and/or metals (depending on whether the secondary star's composition is normal or of Wolf-Rayet type), breakup reactions will produce significant amounts of neutrons. Part of the present paper is devoted to computing precisely such amounts under various assumptions: different accretion disk models, different initial compositions of the accreted material, different temperature profiles, etc. This will be presented in detail in Sect. 2.

There are several reasons underlying our interest in 2.22 MeV emission from the secondary stars of X-ray binaries. First and foremost, as mentioned above the CGRO 2.22 MeV map showed that there are real possibilities for the detection of this line, which would consitute the first observation of any nuclear gamma-ray line outside of the solar system, and this prospect is heightened by the upcoming INTEGRAL mission's order-of-magnitude improvement in both flux sensitivity and energy resolution, which become crucial when most predictions of line fluxes from such sources are at the 10^-5 photons cm^-2 s^-1 level, that is below the sensitivity of CGRO and right at the detection threshold of INTEGRAL. Secondly, the accretion disk problem remains fully unresolved (very little progress has been made on the determination of the disk structure, much less the identification of the viscosity mechanism), and since our results will show significant differences in fluxes and therefore in the possibility for detection of this line emission, this may constitute further testing of models. Furthermore, we will find that line fluxes depend substantially on the secondary star's characteristics (elemental composition in particular), the eventual detection of such a signal would then also constitute a diagnostic of the X-ray binary's secondary (e.g. convection and mixing of the gas).

This paper is structured as follows: in Sect. 2, the treatment of neutron production in the accretion disk (and its subsequent escape) is presented for various models; in Sect. 3, the detailed treatment of neutron propagation, slow-down, and capture in the atmosphere of the secondary is shown; in Sect. 4 we present the results of line fluxes when everything is combined: neutron production and escape from the disk, various interactions and effects (geometrical as well as physical) in the atmosphere of the secondary, and finally capture and gamma-ray emission, with star rotation, photon energy shift, etc.; in the final section we discuss our results with INTEGRAL observation prospects in mind and point to future work on the subject.