Precise Chandra/HRC imaging of globular cluster LMXBs:
first results on NGC1851 and Liller 1

Lee Homer, Scott F. Anderson, Eric W. Deutsch (University of Washington), Bruce Margon, Ronald A. Downes (Space Telescope Science Institute)

[Contributed talk, 15min.]


The nature of the 12 bright ( >1036 erg s-1) X-ray sources in globular clusters appears distinct from that of low-mass X-ray binaries (LMXBs) in the Galaxy as a whole. They are over-abundant by a factor $\mathrel{\hbox to 0pt{\lower 3pt\hbox{$\mathchar''218$ }\hss}
\raise 2.0pt\hbox{$\mathchar''13E$ }}$100 indicating the need for different formation mechanisms. All but one have been reported to exhibit type-I bursts, indicating neutron star accretors. Moreover, the limited number of optical counterparts (only 3 or 4) already implies that their period distribution is also remarkably different from that of field LMXBs, with a preponderance of ultra-short period systems. As well as their intrinsic interest, it is now understood that even this small population of binaries may play a crucial role in the dynamical evolution of their host clusters. Clearly, a more complete understanding of the nature of these systems is of great importance, for which IR/optical/UV counterpart identification is the vital next step. However, even given the 3''''(90% confidence limit) Einstein X-ray positions, counterpart identifications have proven very difficult, due to the extremely crowded locations. Hence, the vast improvement in positional accuracy/precision afforded by Chandra imaging promises to bring the breakthrough we need. Here we present first results on two clusters we have imaged with the HRC.

We obtained exposures of the field of X0512-401 in the globular cluster NGC1851 with the Chandra/HRC-S in X-rays, and also in the far-UV with the Hubble Space Telescope. We derive an accurate new X-ray position (within $\sim$1'''') for X0512-401, which enables us to confirm that the only plausible candidate for the optical/UV counterpart is the Star A, which we previously identified from WFPC2 imaging. We find no evidence for X-ray or UV flux modulation on the ultra-short ( $\mathrel{\hbox to 0pt{\lower 3pt\hbox{$\mathchar''218$ }\hss}
\raise 2.0pt\hbox{$\mathchar''13C$ }}1$ hr) expected binary period, which implies a low system inclination. In addition, we have detected and spatially resolved an X-ray burst event, confirming the association of the burster, quiescent X-ray source, and optical object. The very large LX/Lopt of this object implies an extraordinarily compact system, similar to the sources in NGC6624 and NGC6712.

In spite of many X-ray studies of the ``Rapid Burster" (MXB1730-335) located in the highly reddened globular cluster Liller 1, to date there has been no known optical/IR counterpart for the object, no precise astrometric solution that correlates X-ray, radio, and optical positions and thus restricts the number of possible candidates, nor even published IR images of the field. We solve a previous radio/X-ray positional discrepancy, presenting the results of our precise Chandra/HRC-I X-ray imaging, which definitively show that the radio source is positionally aligned with MXB1730-335. At the same time, we have detected three additional low luminosity ( $L_X\sim10^{34}\thinspace\hbox{$\hbox{erg}\thinspace\hbox{s}^{-1}$ }$) X-ray sources within two core radii, which are possibly quiescent low-mass X-ray binaries. We present both ground-based and Hubble Space Telescope infrared imaging of the field (in quiescent and bursting X-ray states of the RB), together with the necessary astrometric solution to overlay the radio/X-ray source positions. Even at HST resolution, the RB field is very complex and there are multiple candidates. No object of unusual color, or of substantial variability in quiescent versus active or burst versus non-burst states, is identified. Further, more sensitive HST/NICMOS and/or ground-based AO observations are needed to confidently identify the proper counterpart.



Himel Ghosh