next previous
Up: The nature of the


   
7 Limits on the optical counterpart


  \begin{figure}
\par\includegraphics[angle=270,width=8.3cm,clip]{h3518f13.ps}
\end{figure} Figure 13: DSS sky image in R around XMM-Newton position of SAX J1711.6-3808 (circle). The limiting magnitude is R=22.0.

The XMM-Newton EPIC data yield the most accurate position for SAX J1711.6-3808: $\alpha_{2000.0}=17^{\rm h}11^{\rm m}37.1^{\rm s}$, $\delta_{2000.0}=-38^{\rm o}07\arcmin 5\farcs7$ with an error radius of 3$\farcs$2 (90% confidence). This is 1$\farcm$4 from the WFC-determined position (In 't Zand et al. 2001). In Fig. 13 the XMM-Newton error circle is plotted on an image in R from the Digitized Sky Survey (DSS, 2nd generation). No candidate optical counterpart is visible. The limiting magnitude of this image is R=22.0.

For $N_{\rm H}=2.85\times10^{22}$ cm-2, Predehl & Schmitt (1995) predict 15.9 mag of visual extinction. The extinction in the R band is expected to be 11.9 mag (employing AR/AV=0.75). This brings down the effective limiting magnitude to a moderate R=10.1, or MR=-4.5 for a canonical 8.5 kpc distance. Thus, the lack of an optical counterpart in the DSS image is not very constraining. A B0(V)-type star with R=-4.0 (e.g., Drilling & Landolt 2000) would have been detected if it were closer than 6.5 kpc. If such a star would have evolved off the main sequence into a giant, the limiting distance would be 20 kpc. For later stellar types the limiting distances are smaller, being roughly 1 kpc for a B8(V) star. We conclude that it is slightly more likely that the companion star is a late type star. To find an optical counterpart, it is most opportune to search in the infrared, for instance in the Kband. The extinction in this band is only 1.8 mag (employing AK/AV=0.112 following Rieke & Lebofsky 1985). A K0 star, typical for a quiescent low-mass X-ray binary (e.g., van Paradijs & McClintock 1995), would have MK=3.6 (for V-K=2.31, following Bessell & Brett 1988). At a distance of 8.5 kpc this would imply K=20.1. We note that another argument in favor of a low-mass X-ray binary (LMXB) hypothesis is that most transients in this field are LMXBs (e.g., In 't Zand 2001).


next previous
Up: The nature of the

Copyright ESO 2002