The longest observation of a low intensity state from a Supergiant Fast X-ray Transient: Suzaku observes IGRJ08408-4503

We report here on the longest deep X-ray observation of a SFXT outside outburst, with an average luminosity level of 1E33 erg/s (assuming 3 kpc distance). This observation was performed with Suzaku in December 2009 and was targeted on IGRJ08408-4503, with a net exposure with the X-ray imaging spectrometer (XIS, 0.4-10 keV) and the hard X-ray detector (HXD, 15-100 keV) of 67.4 ks and 64.7 ks, respectively, spanning about three days. The source was caught in a low intensity state characterized by an initially average X-ray luminosity level of 4E32 erg/s (0.5-10 keV) during the first 120 ks, followed by two long flares (about 45 ks each) peaking at a flux a factor of about 3 higher than the initial pre-flare emission. Both XIS spectra (initial emission and the two subsequent long flares) can be fitted with a double component spectrum, with a soft thermal plasma model together with a power law, differently absorbed. The spectral characteristics suggest that the source is accreting matter even at this very low intensity level. From the HXD observation we place an upper limit of 6E33 erg/s (15-40 keV; 3 kpc distance) to the hard X-ray emission, which is the most stringent constrain to the hard X-ray emission during a low intensity state in a SFXT, to date. The timescale observed for the two low intensity long flares is indicative of an orbital separation of the order of 1E13 cm in IGRJ08408-4503.

💡 Research Summary

This paper presents the longest deep X‑ray observation of the Supergiant Fast X‑ray Transient (SFXT) IGR J08408‑4503 performed with Suzaku in December 2009. The observation spanned roughly three days, providing net exposures of 67.4 ks with the X‑ray Imaging Spectrometer (XIS, 0.4–10 keV) and 64.7 ks with the Hard X‑ray Detector (HXD, 15–100 keV). During the first 120 ks the source remained in a low‑intensity state with an average 0.5–10 keV luminosity of about 4 × 10³² erg s⁻¹ (assuming a distance of 3 kpc). This quiescent‑like emission was followed by two long flares, each lasting approximately 45 ks, during which the flux increased by a factor of ~3 relative to the pre‑flare level.

The authors extracted three spectra: one from the persistent low‑intensity interval, and one each from the peak of the two flares. A simple absorbed power‑law model could not reproduce the data, leaving significant residuals below 1 keV and yielding an unrealistically low absorption column. Adding a soft thermal plasma component (MEKAL) to the power‑law dramatically improved the fits. The best‑fit model consists of a soft MEKAL component with temperature kT ≈ 0.2–0.3 keV, absorbed by a column density fixed at the optical value for the O8.5Ib(f) companion (NH ≈ 3 × 10²¹ cm⁻²), and a harder power‑law (photon index Γ ≈ 1.5–1.8) absorbed by a higher column (NH ≈ (1–2) × 10²² cm⁻²). This double‑component model yields reduced χ² values of 0.8–1.0 for all three spectra, indicating an excellent description of the data.

When the same model is extrapolated to the HXD‑PIN band (15–40 keV), the predicted flux (≈4 × 10⁻¹³ erg cm⁻² s⁻¹) lies more than an order of magnitude below the measured upper limit of 6 × 10⁻¹² erg cm⁻² s⁻¹, confirming that the source emits very little hard X‑ray radiation during its low‑intensity state. The HXD‑PIN detection itself is marginal; after subtracting the non‑X‑ray background and accounting for the cosmic X‑ray background, the residual signal is consistent with background fluctuations, so the authors treat it as an upper limit.

Timing analysis of the XIS light curves in soft (0.4–3 keV) and hard (3–10 keV) bands shows a roughly constant hardness ratio (H/S ≈ 0.40 ± 0.03) throughout the observation, indicating that the spectral shape does not change dramatically between the persistent emission and the flares. No coherent pulsations were found in the 16 s–500 s range; the 99 % confidence upper limit on the pulsed fraction is 40 %.

The authors interpret the soft thermal component as emission from the supergiant companion, consistent with typical OB star X‑ray luminosities (10³¹–10³² erg s⁻¹). The harder power‑law component is attributed to low‑level accretion onto the compact object (presumably a neutron star) even during the low‑intensity state. The occurrence of two ~45 ks flares separated by ~80 ks suggests that clumpy wind structures or a quasi‑periodic density enhancement in the stellar wind can produce temporary increases in the accretion rate. By associating the flare duration with the crossing time of a dense clump at the orbital radius, the authors estimate an orbital separation of order 10¹³ cm, compatible with a ∼35‑day orbital period previously suggested from outburst recurrence.

This work provides the first stringent upper limit on hard X‑ray emission from an SFXT in a low‑intensity state, improving upon earlier constraints from XMM‑Newton and INTEGRAL. The results support models in which SFXTs spend most of their time in a low‑level accretion regime, punctuated by brief, brighter flares caused by wind clumps or orbital effects. The double‑component spectral model (soft stellar wind emission plus absorbed power‑law) appears to be a robust description of SFXT emission across a wide range of luminosities. Future observations with more sensitive hard X‑ray instruments (e.g., NuSTAR, HXMT) will be able to test whether the hard X‑ray tail truly cuts off below the HXD detection limit or whether occasional harder emission episodes occur during low states. Overall, the paper significantly advances our understanding of the quiescent and intermediate behavior of SFXTs, highlighting the importance of long, uninterrupted observations to capture the subtle variability and spectral characteristics of these enigmatic systems.