A Swift detection of AX J1754.2-2754: renewed activity, or evidence of persistent accretion?

ATel #4109; Thomas J. Maccarone (Southampton), Reba Bandyopadhyay (Florida), Jamie Kennea (Penn State), Chris Britt, Rob Hynes (Lousiana State), Peter Jonker (SRON/Radboud/CfA), Danny Steeghs (Warwick/CfA), Sandra Greiss (Warwick), Gijs Nelemans (Radboud), Phil Lucas (Hertfordshire)
on 11 May 2012; 16:48 UT
Credential Certification: Tom Maccarone (tjm@phys.soton.ac.uk)

Tweet

In conjunction with the Vista Variables in the Via Lactea's intensive Bulge monitoring program, we have arranged a Swift monitoring program of a large region of the inner Galactic Bulge. The program consists of 8 epochs of 500 seconds apiece over the region of the sky which has already been covered by the Chandra Galactic Bulge Survey program (Jonker et al., 2011, ApJS, 194, 18), as well as the part of the inner 2*1 degree region of the Galaxy that has been observed deeply with Chandra (Muno et al. 2009, ApJS, 181, 110). The UVOT data are taken predominantly in the uwm2 filter, and the scheduling is quasi-random so that over the whole survey, the largest possible range of timescales for transients may be studied. Some coordination with the VVV scheduling will be attempted in cases where new transients are seen in the X-rays, but no coordination is set up in advance; the VVV sampling is generally quite good so that quasi-simultaneous data will be available for most bright transients which should be good enough to associate a target with its infrared candidate. The program began on 1 May 2012. In the data for 7 May 2012, we detected the X-ray transient AXJ1754.2-2754, at a count rate of 5*10^-2 cts/sec, which corresponds to a luminosity of 4*10^34 ergs/sec assuming an 8 kpc distance, Galactic absorption and a Gamma=2.0 spectral index. The source is near the edge of the field of view, has fewer than 30 counts, and a hot column runs through the source, so detailed estimates of the source flux are not warranted, but the source is clearly active, and above 10^34 ergs/sec if at a distance of 8 kpc. This source showed a Type I X-ray burst on 16 April 2005 (ATel #1094), was detected quite strongly (at a luminosity of approximately 10^35 ergs/sec) with 1 ksec Chandra HRC-I observations on 14 August 2007, but is not detected in the Chandra Galactic Bulge Survey X-ray data take on 8 May 2008 (ATel #1575), indicating variability of at least a factor of a few hundred. Furthermore, it was detected by Swift and Chandra at luminosities of about 10^35 ergs/sec in 10 separate observations in July of 2008 (ATel #1643). The source's combination of optical and X-ray faintness, long outbursts, and burst properties, it was suggested to be a likely ultracompact X-ray binary (ATel #1575). If the system is ultracompact the lack of a detection in the GBS data should not be due to eclipsing by its mass donor; persistent ultracompact X-ray binaries must have orbital periods of less than about an hour, and an eclipse as long as the 2000 seconds of the GBS observation is thus not possible. The system may thus be an example of a transient with a high duty cycle as a bright source, and only a very low duty cycle in quiescence. Such an object would present a challenge for disk instability theory. One alternative is that the system could be a hierarchical triple star, with the outer star doing the eclipsing of the inner binary, althought clearly more data would be needed to confirm this speculation. Some information is possible from the long Chandra observation already indicates that the period of the outer star in such a triple would have to be longer than about 30 ksec or an eclipse would have been seen in that data set. On the other hand, seeing a single eclipse of 2 kiloseconds in the extant data suggests that the eclipses appear approximately once per 50 kiloseconds, with large uncertainties. The source has a candidate optical counterpart in Blanco telescope data taken in the summer of 2010, with r'=21.7+-0.5 (with the uncertainties dominated by the zero-point calibration). There is marginal evidence for variability of the source in the optical data. Since the nondetection in i' reported in ATel #1575 was for a redder filter and was to a much deeper limit, that would also imply that the source had been inactive, or faint due to an eclipse, in June 2006.