V4641 Sgr appears to have transitioned to a highly radio-quiet non-thermal accretion state

ATel #7966; G. R. Sivakoff, A. Bahramian (Alberta), J. Miller-Jones (ICRAR - Curtin), D. Altamirano (Southampton), C. Heinke, A. Tetarenko (Alberta), M. Middleton (Cambridge) on behalf of a larger XRB collaboration
on 29 Aug 2015; 02:54 UT
Credential Certification: Gregory R Sivakoff (sivakoff@ualberta.ca)

Tweet

Since 2015 July 29, the black hole X-ray binary V4641 Sgr has been in outburst (ATel #7858, #7874). Surprisingly, the source was found to be in a thermal accretion state at a low X-ray luminosity corresponding to an Eddington luminosity fraction below 0.6% (ATel #7904). In such a state, no compact radio jet is expected, which was consistent with a measured radio upper limit (ATel #7908). Here we report that Swift X-ray Telescope (XRT) observations indicate that between August 14 and August 18 the source transitioned to a non-thermal accretion state and its X-ray luminosity has since dropped, reaching ~1e35 erg/s by August 22 (0.5 - 10 keV); assuming a distance of 6.2 kpc (MacDonald et al. 2014, ApJ, 784, 2).

Swift XRT observations in either Windowed Timing (WT) or Photon Counting (PC) mode have been occurring approximately daily between 2015 August 2 and August 13, and approximately every two days since the latter. Spectra from these observations have been fit with both an absorbed power-law spectral model (non-thermal accretion state) and an absorbed disk blackbody (thermal accretion state), allowing the absorption column density, the power-law photon index or disk blackbody temperature, and the model normalization to vary. On and prior to August 14, the thermal model was always a better statistical fit, with a reduced chi^2 value of 1.14 (compared to a reduced chi^2 value of 2.4 for the non-thermal model, both with 815 degrees of freedom over a total of 17 datasets), absorption column densities of 1.3 - 3.0 x 1e21 cm^-2, and disk temperatures of ~1.4 - 2.2 keV. By August 18, we see a transition to the non-thermal state has likely occurred, as indicated by a reversal of how well the models fit the spectra (a reduced chi^2 value of 1.1 for the non-thermal model and a reduced chi^2 value of 1.7 for the thermal model, both with 81 degrees of freedom over 4 data sets). In addition, the absorption column density fit by the thermal model begins to drop below the expected Galactic absorption, while the absorption column density for the non-thermal model is consistent with 2.8 - 5.0 x 1e21 cm^-2; the power-law photon index here is ~1.9 - 2.2. While the X-ray luminosity (assuming a distance of 6.2 kpc) on August 14 was ~ 1.5e36 erg/s (0.5 - 10 keV), it was 5e35 erg/s on August 18 and monotonically decreased to 1.0e35erg/s on August 22; its luminosity on August 27 was 1.1e35 erg/s.

Since the non-thermal accretion state is associated with a compact radio jet, we obtained quasi-simultaneous X-ray and radio observations on August 22 with Swift XRT (09:43 - 10:03 UT; MJD = 57256.40 - 57256.42) and the Karl G. Jansky Very Large Array (VLA; 05:54 - 05:58 UT; MJD = 57256.246 - 57256.249). During the VLA observation, the array was in its most extended A-configuration and we observed in the 4-8 GHz band, using the full 4 GHz of bandwidth and achieving 4.5 minutes of on-source time. We used 3C48 to set the amplitude scale, and J1820-2528 to determine the complex gains of the instrument. Surprisingly, V4641 Sgr was not detected, to a 3-sigma upper limit of 42 microJy/beam. For a distance of 6.2 kpc, this corresponds to a 3-sigma upper limit to the radio luminosity ( nu L_nu at 5 GHz) of 9.7e27 erg/s. The Swift X-ray luminosity at 1-10 keV is 7.6e34 erg/s. Based on Figure 10 from Deller et al. 2015 (ApJ, 809, 13), V4641 Sgr is a factor of at least 3 times lower in radio luminosity than any other known black hole X-ray binary observed at this X-ray luminosity in the non-thermal accretion state.

We note that given the relatively steep power-law we fit, the non-thermal accretion state is actually softer than the thermal accretion state in the ~0.5 - 10 keV Band.

We thank the NRAO and Swift teams for their rapid response and support of these observations.