Astron. Astrophys. 358, 910-922 (2000)

6. Liller 1

The globular cluster Liller 1 is a highly reddened cluster near the galactic center (, kpc, Frogel et al. 1995). It probably has undergone core collapse (Djorgovski 1993). Liller 1 harbours the Rapid Burster, a highly unusual recurrent transient. When discovered in 1977 the source emitted short ( s) bursts of X-rays every s; in some later observations, e.g. Aug 1985, it emitted bursts of s separated by 1500-4000 s; and it has also been observed as a steady source. The bursts are interpreted as accretion events. In addition to these, thermonuclear bursts have also been detected, identifying the accreting star as a neutron star. A review of this remarkable source is given by Lewin et al. (1995). A low-luminosity X-ray source near Liller 1 is tentatively identified as the quiescent (low-state) counterpart of the Rapid Burster (Asai et al. 1996).

No source is detected in the cluster in our ROSAT HRI observation of the globular cluster Liller 1. Near the cluster center, no circle with radius of 5" contains more than 4 photons. For an expected number of 10 photons, the probability of getting 4 or fewer photons is less than 4%. We thus take 10 as the 2- upper limit to the number of photons, which with the effective exposure time is converted to an upper limit of 0.6 cts ksec^-1.

Asai et al. (1996) report the detection on 1993 Aug 27 with ASCA of a source near Liller 1. For a powerlaw with photon index 2, absorbed by a column , this source has an unabsorbed flux in the 2-10 keV band of . For this spectrum our upper limit in the ROSAT HRI corresponds to a flux of , slightly lower than the ASCA detection.

The ROSAT HRI detects a source with a countrate of 1 cts ksec^-1 about 4´ from the cluster center. The statistical error in the position of this source is about 1"; the actual error is dominated by the error in the bore sight correction, which is about 5". The ROSAT source is not compatible with the center of Liller 1, and also not compatible with the position of the Rapid Burster as determined with Einstein (see Table 7). The position of the ROSAT source coincides within the bore sight uncertainty with the O4 III(f) star HD 317889 (Vijapurkar & Drilling 1993). The star is in the Tycho Catalogue as TYC 7380 976 1. From the observed magnitude and colours (, , , Drilling 1991) we estimate a reddening and distance of and kpc for HD 317889. The observed ROSAT HRI countrate is as expected for such a star, according to the general correlation between bolometric luminosity and X-ray luminosity of O stars: (e.g. Kudritzki et al. 1996). (HD 317888 is within 1" of the O4 star; we have not been able to find more information on this star.)

Table 7. Positions of the center of Liller 1 (GC, Picard & Johnston 1995) its core radius (Trager et al. 1993) and the positions of X-ray sources detected near it, viz. the Rapid Burster (RB, Hertz & Grindlay 1983), a dim source detected with ASCA (A, Asai et al. 1996) and a dim source detected with the ROSAT HRI (R, this paper). The position of the O star HD 317889 is also given. The final columns gives the errors in the positions.

We can interpret the ROSAT and ASCA observations in two ways. The first and most likely is that ASCA indeed did detect the Rapid Burster in quiescence, or another low-luminosity source in Liller 1; and that ROSAT observed when this source had a lower flux level. In fact, variation of transients in their quiescent state is common (e.g. Campana et al. 1997). The star detected with ROSAT in this case is not detected with ASCA, presumably because its spectrum is too soft. The second interpretation is that ASCA in fact detected the star also detected with ROSAT, and not the quiescent counterpart of the Rapid Burster. The position of the ROSAT source is marginally compatible with that of the ASCA source; its countrate is exactly that predicted on the basis of the ASCA source. Dr. Asai has kindly communicated a new determination of the position of the X-ray source detected by ASCA, using new calibrations to improve the accuracy. This position, listed in Table 7, excludes the ROSAT source as a possible counterpart, and thus confirms that ASCA indeed detected a source in the cluster.

© European Southern Observatory (ESO) 2000

Online publication: June 20, 2000
helpdesk.link@springer.de