Astron. Astrophys. 331, L77-L79 (1998)

3. Discussion

Thanks to the BeppoSAX good sensitivity above a few keV, our observation has provided the first firm evidence of X-ray emission from G0.9+0.1. Though this region of sky has been imaged in the past with other X-ray satellites, only a very marginal and uncertain detection of G0.9+0.1 was reported by Helfand & Becker (1987), based on an observation done with the Imaging Proportional Counter (IPC) on the Einstein Observatory in 1979. The claimed detection of a source with 0.009 0.003 counts s^-1 at more than one arcminute west of the SNR center, was based on a rather ad hoc procedure aimed at maximizing the very small number of net counts and on an uncertain background estimate. Reanalysing the same IPC pointing (5 ks), we found no sources above a signal to noise ratio of 2, inside the region corresponding to the radio SNR. With these data we can only put an upper limit of 0.03 counts s^-1, while the expected IPC count rate, based on our best fits, is only of counts s^-1, one order of magnitude below that reported by Helfand & Becker (1987).

A search in the ROSAT public archives yielded several PSPC and HRI observations containing the position of G0.9+0.1. However, due to the short exposure times of only a few thousand seconds and especially to the high absorption in the ROSAT band, the SNR was not detected.

All our spectral fits give values of greater than  cm^-2, indicating that G0.9+0.1 must be at a distance of several kiloparsecs, probably close to the Galactic Center or even beyond it. In the following discussion we shall assume a distance of 10 kpc. The large interstellar absorption also explains the apparent discrepancy between our derived luminosity of a few erg s^-1 and the smaller one estimated by Helfand & Becker (1987), who assumed a lower value for the IPC count rate to flux conversion.

The peak of the X-ray emission is coincident with the SNR radio core and there is no evidence for a spatial extension greater than the instrumental resolution. Were the X-rays emitted from a shell with the same dimensions observed in the radio band (diameter arcmin), they would appear clearly resolved in the MECS images. Therefore, we are clearly seeing X-rays emitted predominantly from the central region of the remnant, either from a point source or from a nebula with radius smaller than arcmin.

Some SNRs, like for example W44 (Rho et al. 1994), present a centrally peaked X-ray emission of thermal origin. The thermal nature of the emission is clearly demonstrated by the detection of lines in their X-ray spectra. All the SNRs of this kind have a limb-brightened radio morphology without a flat-spectrum core, contrary to the case of G0.9+0.1. Also considering that the thermal plasma model gave the worst fit to our data, we favour the alternative interpretations related to the likely presence of a neutron star at the center of G0.9+0.1.

One possibility is that of thermal emission from the neutron star surface. The results of the blackbody spectral fit imply an emitting surface with radius km, definitely smaller than the whole neutron star surface for any reasonable distance. This can be interpreted as emission from a small polar cap region, hotter than the rest of the neutron star due to anisotropic heat diffusion from the interior and/or to reheating by relativistic particles backward accelerated in the magnetosphere (Halpern & Ruderman 1993). In general, this should produce a periodic flux modulation, but the strong gravitational bending effects severely reduce the observed pulsed fractions (Page 1995). Our upper limits on the possible flux modulations are not strong enough to pose serious problems to this interpretation. However, the fitted blackbody temperature (kT 1.4 keV) is higher than that observed in all the other X-ray emitting radio pulsars.

A different explanation involves non-thermal emission powered by the rotational energy loss of a relatively young neutron star. The radio shell radius of 12 pc implies a lower limit to the remnant age of 1100 yr, for a free-expansion phase with v  km s^-1. If the remnant is expanding adiabatically, from the Sedov model we have a shell radius pc, where is the explosion energy in units of ergs, is the ambient ISM hydrogen density in and is the age in units of yr. For typical values , we derive an age of years. Both a point-like, pulsed component originating in the neutron star magnetosphere and a diffuse () synchrotron nebula probably contribute to the observed X-rays.

Our best fit power law photon index 3.1 is rather steep, compared to other X-ray synchrotron nebulae, but a more typical value of is also consistent with our data (for  cm^-2). The corresponding X-ray luminosity (1-10 keV), erg s^-1, is within the range observed in the central components of other SNRs (see, e.g., Helfand & Becker 1987) and can be easily powered by a young neutron star.

© European Southern Observatory (ESO) 1998

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