3. Discussion and conclusion
The photon index obtained by SIGMA, , is more consistent with those of binary neutron star systems rather than with blackhole candidates such as 1E1740.7-2942 or GRS 1758-258. The derived mean photon index from ASCA, is markedly higher. It is softer than that expected for accretion-powered pulsars (APP), the majority of which have photon indices between -1.5 and -0.8 (Nagase 1989) although some can present rather soft spectra, as e.g. 4U1258-61 with a photon index of between 1 and 60 keV (White et al. 1983) close to that of SLX 1735-269, 1E 2259+586 with a photon index of in the 1.5-5 keV region has an extremely steep spectrum (Koyama et al. 1987). The K-iron emission feature at 6.4 keV displayed by most APPs is absent in the spectrum of SLX 1735-269. Thus the observed spectral characteristics do not follow the general trend of those of APPs. Furthermore, no rotational period was found, expected unless SLX 1735-269 has its magnetic and rotational axii aligned, further weakening the APP hypothesis.
The column density measurements are consistent with a source location situated close to the Galactic center. Thus, if a distance of 8.5 kpc is assumed, the X-ray luminosity of SLX 1735-269 is erg s-1. Assuming such a low luminosity excludes SLX 1735-269 from the class of high-luminosity low mass X-ray binaries (LMXRB) which have typical luminosities of 1038 ergs s-1. Moreover, the spectra of these systems exhibit a blackbody component accounting for 10% - 70% of the total luminosity (White et al. 1988). The addition of blackbody emission to the spectral model for SLX 1735-269 does not improve the fits to the ASCA data. The derived upper limit to blackbody emission at 3 keV is found to be only 5 1034 ergs s-1, i.e. at most 2% of the total luminosity. But SLX 1735-269 might also be another type of LMXRB, namely an "X-ray burster" since all the results of the spectral analysis are compatible with this hypothesis. Moreover, SLX 1735-269 is among the 11 soft -ray sources within the Galactic bulge detected by SIGMA of which 50% are LMXRB bursters. However, the light curve of SLX1735-269 shows no signature of burst activity, evidence for which was also not observed either by ART-P or TTM. This tends to weaken the case for an X-ray burster, but the data present a number of gaps and the integration time over which data have been collected might have been too short to allow the detection of bursts.
An unambiguous association with a companion star would settle the binary nature of SLX 1735-269. A search in the SIMBAD data base of sources within of the ASCA position did not reveal any convincing associations with SLX 1735-269 at other wavelengths since the error boxes associated with the four objects found do not intersect those of ASCA and Rosat. The high column density obtained from the ASCA data implies strong absorption along the line of sight (typically e.g. Gorenstein and Tucker 1976) and thus an optical counterpart for SLX 1735-269 will be difficult to detect, especially in such crowded regions of the Galaxy. However, thanks to the small error box associated with the Rosat All-Sky Survey position, meaningfull optical/IR observations are now possible, particularly in the I and K bands for which extinction is lower. Considering a distance modulus of 14.6 for SLX 1735-269, even setting an upper limit to the visual magnitude of 18-21 would effectively exclude a high mass early type companion (e.g. Chen et al. 1994, Mereghetti et al. 1992).
The lack of long term variability reported by TTM and SIGMA points to a more stable system. Thus, the possibility that SLX 1735-269 is an isolated pulsar powered by its rotational energy loss should also be explored. SLX 1735-269 is not listed among the 706 radio pulsars in the catalogue of Taylor et al. (1993). The few radio pulsars observed in the X-ray domain (only 16, see Caraveo 1995), have photon indices falling between -4 and -1 (Mereghetti et al. 1994). The young pulsar PSR 1951+32 has a photon index of -1.9 in the 0.05-6 keV region close to that of SLX 1735-269. Unfortunately no data is available at higher energies. Only three isolated pulsars have been detected in the hard X and soft -rays, namely PSR 1509-58, the Crab pulsar and Vela. The photon indices obtained for these are -1 (Laurent et al. 1994), (Ulmer et al. 1994) and -1 (Hermsen et al. 1993) respectively. No data between 2 and 60 keV is available for the Vela pulsar, for the other two, very young, isolated neutron stars, there is no significant break between the soft and hard X-ray spectra and broken powerlaw fits do not provide an improved model to their spectra, in contrast to SLX 1735-269 where a break appears to occur in the interval 12-35 keV. But the paucity of X-ray data on isolated pulsars makes it difficult to conclude that all isolated neutron stars have spectra discribed by a single powerlaw. Note that PSR 1509-58 and the Crab pulsar do not exhibit line emission at 6-7 keV. Although no rotational period of less than a few seconds was found, it can still be argued that SLX 1735-269 is an isolated neutron star on the basis of the spectral data.
The spectrum of SLX 1735-269 displays characteristics which can be attributed both to binary or isolated pulsars, however, the derived upper limits to pulsations tend to disfavour the pulsar hypothesis. The possibility that this source is an X-ray burster is left open. Based on the available data to date, the nature of SLX 1735-269 still remains elusive. To further elucidate to what category this object belongs, more data must be collected.
© European Southern Observatory (ESO) 1997
Online publication: June 30, 1998