The cooling tail of a long X-ray burst from XTE J1701-407

ATel #1618; M. Linares, P. Soleri, P. Curran, R. Wijnands, N. Degenaar, M. van der Klis (Amsterdam), R. Starling (Leicester), C. Markwardt (CRESST/GSFC/UMD)
on 18 Jul 2008; 11:04 UT
Credential Certification: Manuel Linares (mlinares@science.uva.nl)

Tweet

Swift XRT observed the recently discovered source XTE J1701-407 (Markwardt et al., ATel #1569, Degenaar et al., ATel #1572) on July 17th, 2008 at UT 13:31:45, 97 seconds after the Swift-BAT detection of an X-ray flare (Barthelmy et al., GCN Circ. #7985; Markwardt et al., Atel #1616), for a total exposure time of nearly two kiloseconds. We report herein the analysis of the inital ~1 kilosecond long windowed timing (wt) mode data. Our results confirm that this is a thermonuclear event, uncovering the nature of the accreting compact object (a neutron star) and the binary system (a low-mass X-ray binary).

The XRT lightcurve shows a smooth decay that cannot be fitted with a single exponential law. The e-folding time is approximately 130 seconds during the first ~3 minutes and increases to ~450 seconds during the following ~10 minutes. Taking into account the time of the BAT trigger, the burst was detectable for about 15 minutes. A rough estimate of the fluence from the XRT lightcurve gives ~1.7e-6 erg/cm^-2, which added to the BAT fluence (Atel #1616) gives a total of ~2.5e-6 erg/cm^-2. The 0.7-10 keV time-averaged wt spectrum can be fitted with an absorbed black body, giving an equivalent column density consistent with those reported in ATels #1569 and #1572 (Nh~3.2e22 cm^-2; kT~1.2 keV; chi2/dof~1). An absorbed power law gives a statistically poorer fit and a higher column density (Nh~6e22 cm^-2; photon index~2.6; chi2/dof~1.4). We also extracted source spectra in nine contiguous time intervals and fitted them with a black body model, fixing the column density to 3.5e22 cm^-2 (the value found by Degenaar et al., ATel #1572, from spectral fits to the persistent emission). The results are as follows:

Interval	start	end	kTbb	1-sigma	unabs. 2-10 keV
number	time(s)	time(s)	(keV)	err. (keV)	Flux (erg/cm2/s)
01:	0	30	1.70	0.05	1.1e-08
02:	30	70	1.44	0.04	6.4e-09
03:	70	120	1.26	0.03	4.2e-09
04:	120	180	1.14	0.03	2.8e-09
05:	180	260	1.05	0.02	2.0e-09
06:	260	360	0.95	0.02	1.4e-09
07:	360	500	0.94	0.02	9.4e-10
08:	500	660	0.92	0.02	7.2e-10
09:	660	830	0.88	0.02	5.4e-10

(all times are from the start of the XRT observation; reduced chisquared values are between 0.9 and 1.3)

The black body temperature shows a clear cooling curve as the flux decreases. A further wt data interval taken more than one hour after the start of the observation shows the source at an unabsorbed 2-10 keV flux of ~3.1e-10 erg/cm2/s, similar to the approximately constant pre-burst level measured by RXTE-PCA and Swift-BAT and to the flux measured by BAT more than three hours after the BAT trigger (Atel #1616). We searched for rapid variability and found no strong pulsations or QPOs up to ~280 Hz.

The time evolution of the flux and energy spectrum clearly indicates that this is a thermonuclear X-ray burst, confirming the suggestion of Markwardt et al. (Atel #1616) and revealing the nature of the system. However, its remarkably long duration places this event in the sub-class of long X-ray bursts, usually interpreted as Helium burning bursts (Woosley et al 2004, ApJS).