Swift follow-up observations of unclassified ASCA sources

We present Swift follow-up observations of a sample of 35 unclassified faint X-ray sources drawn from the ASCA Galactic centre and Galactic plane surveys. Our short, pointed XRT observations allow detections down to a limiting 0.3–10 keV flux of F[subscript: X] ~ 10[superscript: -13] erg cm[superscript: -2] s[superscript: -1], which translates into a luminosity of L[subscript: X] ~ 10[superscript: 33] erg s[superscript: -1] for an assumed distance of D = 8 kpc. The brightest source in our sample reaches a maximum 0.3–10 keV luminosity of L[subscript: X] ~ 2 × 10[superscript: 36] (D/8 kpc)[superscript: 2] erg s[superscript: -1] during our Swift observations. We detect 16 (46%) of the ASCA sources with the XRT, while 19 were not detected during our program. Since we are probing the faint end of the ASCA source populations, we expect a large fraction of the non-detections to be due to the Eddington bias. This is strengthened by the fact that we find the observed XRT count rates to be predominantly lower than expected based on the reported ASCA intensities. Nevertheless, investigation of the ASCA properties and any possible long-term X-ray variability leads us to conclude that the non-detections likely include two spurious ASCA detections and three objects that are highly variable or transient X-ray sources. For the 16 XRT-detected sources we obtain positional accuracies of ~2−4′′, which significantly improves upon their ASCA uncertainties of ~1′. We use the X-ray spectra and variability to characterise these objects. Most appear to be faint, persistent X-ray emitters that have highly absorbed spectra. Based on their X-ray properties we identify three accreting compact objects: one confirmed (AX J1754.2–2754) and one candidate (AX J1538.3–5541) X-ray binary, and one possible magnetically accreting white dwarf (AX J1620.1–5002). Furthermore, we use the improved positions of XRT-detected sources to search for counterparts in simultaneously obtained Swift/UVOT images and possible associations with catalogued sources at various wavelengths. This reveals three possible main sequence stars amongst our sample. The other sources remain unclassified, but our improved XRT positions provide good prospects for dedicated follow-up observations that have the potential to elucidate the nature of these faint, unclassified ASCA sources.