New results with Swift on Supergiant Fast X-ray Transients

We report here on the most recent results obtained on a new class of High Mass X-ray Binaries, the Supergiant Fast X-ray Transients. Since October 2007, we have been performing a monitoring campaign with Swift of four SFXTs (IGRJ17544-2916, XTEJ1739-302, IGRJ16479-4514 and the X-ray pulsar AXJ1841.0-0536) for about 1-2 ks, 2-3 times per week, allowing us to derive the previously unknown long term properties of this new class of sources (their duty cycles, spectral properties in outbursts and out-of-outbursts, temporal behaviour). We also report here on additional Swift observations of two SFXTs which are not part of the monitoring: IGRJ18483-0311 (observed with Swift/XRT during a whole orbital cycle) and SAXJ1818.6-1703 (observed for the first time simultaneously in the energy range 0.3-100 keV during a bright flare).

💡 Research Summary

This paper presents the results of an extensive monitoring campaign of Supergiant Fast X‑ray Transients (SFXTs) carried out with the Swift satellite since October 2007. Four prototypical SFXTs—IGR J17544‑2916, XTE J1739‑302, IGR J16479‑4514, and the pulsar AX J1841.0‑0536—were observed 2–3 times per week with short (1–2 ks) pointings using both the X‑Ray Telescope (XRT) and the Burst Alert Telescope (BAT). The high‑cadence strategy allowed the authors to determine, for the first time, the long‑term duty cycles, spectral characteristics during outbursts and quiescent intervals, and temporal behaviour of this class of sources.

The monitoring revealed that SFXTs spend only a few percent of the time in bright states (luminosities >10³⁶ erg s⁻¹), with duty cycles ranging from ≈3 % to 5 %. This is markedly lower than the ≈30 % duty cycles typical of classical supergiant high‑mass X‑ray binaries, indicating that the mass‑transfer in SFXTs is highly intermittent. Spectral analysis shows a clear dichotomy: during flares the spectra are hard (photon index Γ≈0.5–1.0) and heavily absorbed (N_H≈10²³ cm⁻²), while in the low‑level state they soften (Γ≈1.5–2.0) and exhibit lower absorption (N_H≈10²² cm⁻²). Broadband (0.3–100 keV) fits combining XRT and BAT data are best described by either a Compton‑ised thermal plasma (CompTT) with electron temperatures kT_e≈15–30 keV and optical depths τ≈5–10, or a cutoff power‑law with a cutoff energy around 20 keV.

Temporal analysis uncovered precursor and post‑flare phases lasting 10–30 minutes, suggesting that material is being accelerated and processed before and after it penetrates the neutron‑star magnetosphere. The detection of a 4.7 s pulsation in AX J1841.0‑0536 provides direct evidence for a gated accretion mechanism, where the neutron‑star spin and magnetic field can temporarily inhibit inflow.

Additional observations of two SFXTs not included in the regular monitoring further support the emerging picture. IGR J18483‑0311 was observed continuously over an entire orbital cycle (~18 days). The source displayed a clear orbital modulation of the average flux, with a factor‑of‑two variation and short, bright flares confined to periastron passages, consistent with a clumpy stellar wind whose density peaks near the neutron star’s closest approach. SAX J1818.6‑1703 was captured for the first time simultaneously by XRT and BAT during a bright flare. The event lasted ≈2 ks, reached a luminosity of ≈10³⁷ erg s⁻¹, and showed a high‑energy cutoff near 20 keV, indicative of a sudden, dense clump of wind material being accreted.

The authors interpret these findings within a hybrid framework that combines the “clumpy wind” model—where stochastic overdensities in the supergiant’s outflow produce brief episodes of high mass‑transfer—and the “magnetic gating” model, where the neutron‑star’s magnetic field and spin period regulate the inflow. The low duty cycles, hard flare spectra, rapid rise and decay times, and occasional detection of pulsations all point to a scenario where both mechanisms operate simultaneously, producing the extreme variability that defines SFXTs.

In conclusion, the Swift campaign has, for the first time, quantified the long‑term behaviour of SFXTs, establishing their duty cycles, spectral evolution, and timing properties across multiple sources. These results provide a crucial observational baseline for future high‑resolution X‑ray missions such as NICER, eROSITA, and Athena, which will be able to probe the detailed physics of wind clumping, magnetospheric interaction, and accretion dynamics in these enigmatic transients.