Discovery of X-Ray Polarization from the Black Hole Transient Swift J1727.8−1613

Mass accretion is a fundamental process of energy extraction that powers some of the brightest X-ray sources. It operates very efficiently for compact objects, neutron stars and black holes (BHs), which accrete matter from a nearby companion star. BH X-ray binaries display two major spectral states, hard and soft, which are thought to be linked to different accretion regimes (Zdziarski & Gierliński 2004; Remillard & McClintock 2006; Done et al. 2007). In the soft state, the X-ray spectrum is dominated by a blackbody-like emission that is believed to be produced by a geometrically thin, optically thick accretion disk (Novikov & Thorne 1973; Shakura & Sunyaev 1973; Page & Thorne 1974). In the hard state, the spectrum constitutes a power-law-like continuum, that is generally attributed to Comptonization in a medium with hot electrons (a corona). At the transitions between the canonical states, the hard-intermediate and soft-intermediate, as well as very high (or steep power-law) states have been identified (Homan & Belloni 2005; Belloni 2010).

Transitions of BH X-ray binaries between various spectral states follow a well-known pattern; however, our understanding of the accretion geometry and physical mechanisms producing broadband emission in these states is still incomplete. The geometry and location of the hot medium in the hard spectral state is still a matter of debate (Poutanen et al. 2018; Bambi et al. 2021). The structure of the accretion disk is unclear; its stability in the soft state is puzzling (Dexter & Quataert 2012; Jiang et al. 2013). Furthermore, the conditions for the source to show the very high state are unknown (Done et al. 2007). X-ray polarimetry is a new diagnostic that may help resolve questions regarding the geometry of the emission region left unanswered by the conventional tools of spectroscopy and timing.

The Imaging X-ray Polarimetry Explorer (IXPE; Weisskopf et al. 2022) provided the first significant detection of 2–8 keV X-ray polarization in the archetypical BH Cyg X-1 (Krawczynski et al. 2022)—the source that originally gave rise to the spectral classifications (Tananbaum et al. 1972). The hot Comptonizing medium visible in the hard spectral state of the binary appeared to be elongated in the plane of the accretion disk. The polarization angle (PA) of X-ray emission was found to be consistent with the position angle of the extended radio source (Miller-Jones et al. 2021), thus confirming, for the first time, that the radio jets are launched along the axis of the (inner) accretion disk. The polarization degree (PD) of 4.0% ± 0.2% was higher than expected for the orbital inclination of the source i ≈ 30° (Miller-Jones et al. 2021). This can be explained by the inner disk being more highly inclined than the outer disk (Krawczynski et al. 2022) or by an outflow of the hot medium (Poutanen et al. 2023).

The BH X-ray binary Cyg X-3 became the second hard-state target of IXPE. Similarly to Cyg X-1, the source swings between different spectral states, which might be attributed to the same configurations of accreting matter as in Cyg X-1. Contrary to early expectations, a high PD = 20.6% ± 0.3% was detected in the hard state (Veledina et al. 2023). The X-ray PA was found to be shifted by 90° with respect to the position angle of the discrete radio ejections, as well as to the submillimeter polarization direction. These unique polarization properties can be explained in terms of a pure reflection scenario. The X-ray polarization observations revealed the presence of an obscuring and reflecting envelope at high elevation above the disk plane, suggesting this well-studied binary is a Galactic ultraluminous X-ray source.

The aforementioned hard-state sources belong to the class of high-mass X-ray binaries, where the compact object captures matter from the wind of their massive companion. The innermost accretion geometry of low-mass X-ray binary BHs, where the disk forms via Roche lobe overflow, was not yet studied polarimetrically. In this paper, we report on the first X-ray polarization measurement of the Galactic X-ray binary Swift J1727.8−1613.

Swift J1727.8−1613 underwent a bright outburst that was detected on 2023 August 24 (Negoro et al. 2023a; Kennea & Swift Team 2023). MAXI (Matsuoka et al. 2009) monitoring of the source showed it reaching ∼7 Crab in the 2–20 keV range (see Figure 1(a)), triggering interest and rapid initiation of follow-up multiwavelength campaigns by a number of ground-based and space observatories (Baglio et al. 2023; Miller-Jones et al. 2023; Negoro et al. 2023b; O'Connor et al. 2023; Wang & Bellm 2023; Williams-Baldwin et al. 2023). The beginning of the outburst was observed in the optical wavelengths about 4 days earlier than the first X-ray trigger (Wang & Bellm 2023). Optical spectroscopy taken at the early outburst stages suggest the source is a low-mass X-ray binary (Castro-Tirado et al. 2023), with signatures of an outflow (Mata Sanchez & Muñoz-Darias 2023). On the next and the third day after the X-ray trigger, the X-ray spectral shape was identified to be consistent with the typical hard-state power law with photon index Γ ∼ 1.3–1.7, as measured by the LEIA imager (Liu et al. 2023) and by the Mikhail Pavlinsky ART-XC telescope (Sunyaev et al. 2023), respectively. Recent NICER observations (taken 25 days after the beginning of the outburst) report substantial contribution of the disk blackbody emission (of temperature kT_diskbb = 0.6 keV) and softening of the spectrum (Γ = 2.4, Bollemeijer et al. 2023). X-ray light curves revealed prominent quasi-periodic oscillations with a slowly increasing (on timescales of days) central frequency (Bollemeijer et al. 2023; Draghis et al. 2023; Palmer & Parsotan 2023).

Zoom In Zoom Out Reset image size

The onset of the radio source detected at various frequencies shortly after the X-ray trigger identified the presence of a jet (Bright et al. 2023; Miller-Jones et al. 2023), with a flat spectrum (Vrtilek et al. 2023; S. Trushkin et al. 2023, in preparation) typical to the hard-state sources. A subsequent submillimeter detection and polarization measurements indicate its direction is nearly along the north–south celestial axis (PA = − 41 ± 35 on the date closest to IXPE pointing; Vrtilek et al. 2023). The intrinsic optical polarization also seems to be roughly aligned with the north–south direction (Kravtsov et al. 2023). While the mass of the compact object has not been reliably measured yet, all indirect signatures indicate the binary hosts a BH, making it an intriguing target that can be used to probe the accretion geometry.

Figure 1(a) shows the initial stages of the countrate evolution in Swift J1727.8−1613 as seen by MAXI monitor, ⁵⁸ and the typical fast rise profile can be clearly identified. The starting point for the date corresponds to the initial X-ray trigger (which we will refer to as the beginning of the X-ray outburst hereafter). A few days after the beginning of the outburst, the source started to soften (see Figure 1(b)). The source evolution follows the well-known q-track pattern (Homan & Belloni 2005) in the hardness–flux diagram (blue symbols in Figure 1(c)), albeit becoming much brighter than the prominent low-mass X-ray binary MAXI J1820+070 (shown in red) and the prototypical source GX 339–4 (orange). IXPE observed Swift J1727.8−1613 at high, close to current maxima, fluxes, about 15 days after the beginning of the outburst. According to the position in hardness–flux diagram, the source started transition to the soft state and resided in the hard-intermediate state (which we refer to as the hard state hereafter; Figure 1(c)).

IXPE is the first satellite mission dedicated to X-ray polarimetry in the 2–8 keV band (Weisskopf et al. 2022). It carries three X-ray telescopes, each made of a mirror module assembly (Ramsey et al. 2022) and a polarization-sensitive gas-pixel detector unit (Baldini et al. 2021; Soffitta et al. 2021), which enable imaging X-ray polarimetry of extended sources and a huge increase of sensitivity for point-like sources. IXPE provides an angular resolution of ∼30'' (half-power diameter). The overlap of the fields of view of the three detector units is circular with a diameter of 9'; spectral resolution is better than 20% at 6 keV.

IXPE observed Swift J1727.8−1613 from 2023 September 7 19:59:07 UTC to September 8 06:35:12 UTC for a live time of ∼19 ks (Dovciak et al. 2023a, 2023b). Level 2 data were downloaded from the IXPE archive at the HEASARC ⁵⁹ and analyzed with both ixpeobssim version 30.6.2 (Baldini et al. 2022) and HEASOFT/xspec version 12.13.1d (Arnaud 1996). IXPE observations of Swift J1727.8−1613 were carried out placing a "gray" filter in front of the detector (Ferrazzoli et al. 2020; Soffitta et al. 2021). The filter was used to reduce (by a factor of ∼10) the incident flux from a very bright source, especially at low energies, to a level compatible with the dead time of the detector units and with the need to transmit all data within the allocated telemetry. Now different response matrices have to be used for data analysis, which are available both in the ixpeobssim package and in the HEASARC CALDB. Comparing the polarization obtained by the pcube algorithm in ixpeobssim and the one from xspec, we found that the latter provides more consistent results as it fully takes into account the energy response of the instrument. This is particularly important to correctly describe the low-energy response of the IXPE detectors when the gray filter is used, as its transmission drops steeply at low energies. In the following, we then present the results obtained with xspec. Source events were extracted using a circle with 80'' radius centered on the source. No background subtraction is necessary for sources as bright as Swift J1727.8−1613, since in this case the IXPE background is dominated by scattered source emission, even at large offset (hence, the background itself is negligible and can be ignored; Di Marco et al. 2023).

We fitted Stokes I, Q, and U spectra with xspec, using only simple models because of the limited spectral capability of IXPE. We first tested a polconst∗diskbb model, representing an accretion disk consisting of multiple blackbody components with a constant polarization, using the gain fit command in xspec to fit the energy scale. We found an unacceptable fit, χ² = 3509 for 1329 degrees of freedom (dof), driven by large residuals in the I spectrum. Models for absorption remain unconstrained due to the relatively high IXPE threshold (2 keV) and the use of the gray filter, which strongly reduces the flux from the source at low energies. Then, we used a power-law model assuming constant polarization polconst∗powerlaw, using gain fit command. The data with the best-fit model are shown in Figure 2. The model gives χ² = 1282 for 1329 dof. The best-fit power-law index is ${\rm{\Gamma }}={1.80}_{-0.01}^{+0.02}$, with the polarization being PD = 4.1% ± 0.2% at PA = 22 ± 13. The errors are at 68% confidence level and calculated assuming one interesting parameter with the steppar command of xspec. The polarization from the source is detected at a ∼20σ confidence level.

Zoom In Zoom Out Reset image size

We then froze the spectral model to the one found for the 2–8 keV energy band and fitted in the different energy ranges leaving only the polarization model (either polconst or stokesconst) free to vary. The results are given in Table 1. The PD shows an increasing trend with energy, growing from about 3% in the 2–3 keV band to ∼5% above 5 keV. We note that the contribution of the disk with the inner temperature kT_diskbb ∼ 0.5 keV (Bollemeijer et al. 2023; Draghis et al. 2023) cannot be ruled out in the lower-energy bin and might cause the reduction of PD here. At the same time, the PD at higher energies, which we attribute to the corona, is rather constant. The PA does not show any significant variations with energy. Contours plots of the PD and PA in three energy bands are shown in Figure 3.

Zoom In Zoom Out Reset image size

We then checked for a possible energy dependence of the polarization by substituting the polconst model with pollin, which assumes a linear dependence of PD and PA on photon energy. However, we assumed an energy-independent PA (ψ₁ in xspec; fixing ψ_slope = 0), and allowed the PD to vary with photon energy E (keV) as PD(E) = A₁+ A_slope(E − 1). We obtained a marginally better fit χ²/dof =1277/1328 with the F-test giving a probability of 3% for a chance improvement. The best-fit parameters are A₁ = 3.0% ± 0.5% and A_slope = 0.34 ± 0.13% keV⁻¹, and PA=ψ₁ = 20 ± 13.

The exceptional brightness of the newly discovered source Swift J1727.8−1613 is striking and can be compared to only a few historical BH transients: the low- or intermediate-mass X-ray binaries A0620−00 (Elvis et al. 1975), GRS 1915+105 (Sazonov et al. 1994), V404 Cyg (Życki et al. 1999), V4641 Sgr (Revnivtsev et al. 2002), MAXI J1820+070 (Shidatsu et al. 2018), and 4U 1543−47 (Sánchez-Sierras et al. 2023). The high X-ray fluxes of A0620−00 and MAXI J1820+070 are caused by the proximity of these sources to the observer, and the inferred accretion rates fall in the sub-Eddington category. In these cases, we see the innermost parts of the disk/corona during the hard state. On the other hand, the luminosities of GRS 1915+105, V404 Cyg, V4641 Sgr, and 4U 1543−47 are believed to reach near- and super-Eddington values, and signatures of an optically thick outflow, that can cover the innermost regions, have been detected (Revnivtsev et al. 2002; Motta et al. 2017; Miller et al. 2020; Prabhakar et al. 2023). Analogs to these classes of sources can be found in active galactic nuclei, whose geometry studies are now also enabled by IXPE data. For the hard-spectrum sources where the innermost parts of the accretion disk are visible, the X-ray polarization is found to be in a range of about a few percent (Cyg X-1, Krawczynski et al. 2022; NGC 4151, Gianolli et al. 2023; MCG-05-23-16, Marinucci et al. 2022; Tagliacozzo et al. 2023; IC 4329A, Ingram et al. 2023) and aligned with the jet direction (wherever it is constrained). For the sources where the central engine is obscured, the observed spectrum is dominated by a reflection continuum, and high (PD >20%) polarization orthogonal to the jet direction (polarization aligned with the obscuring torus) has been found (Cyg X-3, Veledina et al. 2023; Circinus galaxy, Ursini et al. 2023).

The relatively low X-ray PD found in Swift J1727.8−1613 (as compared to some other accreting BHs) suggests the innermost parts of the accretion disk are visible, and the geometry is similar to that of Cyg X-1 and Seyfert 1 galaxies. This is supported by the hardness–flux diagram (Figure 1(c)) which indicates that IXPE observed the source at the beginning of the hard-to-soft-state transition. By comparing the flux of Swift J1727.8−1613 at the turning point (at which a source begins softening) with those of other BH X-ray binaries MAXI J1820+070 situated at 3 kpc (Atri et al. 2020) and GX 339−4 at an estimated distance of ∼10 kpc (Zdziarski et al. 2019), we can estimate that Swift J1727.8−1613 is closer than MAXI J1820+070 by a factor of 2 (i.e., at a distance of about 1.5 kpc).

Relativistic ejections have not yet been observed in the source; hence the direction of the jet is not known from radio images. The submillimeter polarization signal, however, gives a clue as to the jet axis (oriented close to the north–south direction; Vrtilek et al. 2023). It has been found that the intrinsic polarization in the optical, submillimeter, and radio (after correction for Faraday rotation) is aligned with the jet direction in the X-ray binaries during the outburst (Cyg X-3, M. McCollough et al. 2023, in preparation; A. Lange et al. 2023, in preparation; Veledina et al. 2023; XTE J1908+094, Curran et al. 2015; XTE J1550−564, Migliori et al. 2017; MAXI J1820+070, Veledina et al. 2019; and V404 Cyg, Kosenkov et al. 2017). If the jet direction in Swift J1727.8−1613 also aligns with the detected optical and submillimeter polarization, then the X-ray polarization is also aligned with the jet direction, supporting the analogy to sub-Eddington, rather than super-Eddington sources.

The similarity of the X-ray PA to the submillimeter PA and, by extension, to the position angle of the radio jet is related to the alignment of the jet direction with the BH spin and the jet collimation mechanisms (McKinney et al. 2013; Davis & Tchekhovskoy 2020). If the BH spin is misaligned with the binary orbital axis, the accretion disk/flow becomes warped and/or can experience Lense–Thirring precession (Bardeen & Petterson 1975; Stella & Vietri 1998; Fragile et al. 2007). Precession of the accretion flow may then be visible in multiwavelength light curves as prominent quasi-periodic oscillations (Ingram et al. 2009; Veledina et al. 2013), which are indeed detected in the X-ray timing data on Swift J1727.8−1613 (Draghis et al. 2023; Palmer & Parsotan 2023). We clearly detect two harmonics of these oscillations, at 1.340 ± 0.004 and 2.74 ± 0.04 Hz, in the IXPE flux data (M. Ewing et al. 2023, in preparation). In the case of a substantial misalignment between the orbital and BH spin axes and rapid Lense–Thirring precession of the inner flow, the average (over the precession period) flow axis is aligned with the spin axis. As a consequence, the average X-ray PA is parallel to the BH spin axis. The outer parts of the disk are, on the other hand, expected to be aligned with the orbital plane (e.g., as found in MAXI J1820+070; Poutanen et al. 2022). The alignment of the jet direction with the BH spin in Swift J1727.8−1613 also suggests that the jets are launched from the radii that are either aligned with the BH spin axis or experience rapid precession around it.

The similarity of the detected X-ray PD in Swift J1727.8−1613 and Cyg X-1 may be interpreted as similarity of inclinations of the inner parts of the accretion disk in these systems, that translates to an intermediate inclination of the source i ∼ 30°–60°. This is further supported by the absence of the X-ray dips. The inclination of the binary orbit can be reliably measured only after the source reaches quiescence (Casares & Jonker 2014), and may help to distinguish between various configurations of the accretion disk and corona. However, basic constraints on the geometry can be made using these first X-ray polarization measurements. While the X-ray PA and its relation to the submillimeter PA suggest that the X-ray corona is extended along the disk (and not along the jet), the high PD ∼ 4%, either constant or increasing with energy, argue against spherical and lamppost coronal geometries (Krawczynski et al. 2022; Zhang et al. 2022).

Swift J1727.8−1613 continues to gradually soften, in line with the q-pattern in the hardness–flux diagram (Figure 1(c)). Subsequent observations by IXPE detecting the source in the soft and very high states may enable comparison to the other BH X-ray binaries observed in polarized X-ray light (Dovciak et al. 2023c; Marra et al. 2023; Podgorny et al. 2023; Ratheesh et al. 2023; Rodriguez Cavero et al. 2023; Svoboda et al. 2023), providing independent constraints on the inclination and calibrating the models of accretion disk structure and radiative processes.

We obtained the first X-ray polarization measurement for the bright BH binary Swift J1727.8−1613. We find the time- and energy-averaged PD = 4.1% ± 0.2% and PA = 22 ± 13 (68% confidence) from xspec spectropolarimetric fits. There is no statistically significant dependence of the PA on energy. A model that assumes a linear dependence of PD on energy gives a slight improvement in χ² significant at 97% confidence level and results in a PD increasing with energy by 0.34% per keV.

The evolution of the source in the hardness–flux diagram and the observed spectral slope indicate that the source was in the hard state during IXPE observations, accreting at sub-Eddington rates, so its high X-ray brightness is caused by the source proximity to the Earth. We estimate the distance to the source to be about 1.5 kpc.

The alignment of X-ray, optical, and submillimeter polarization directions indicates that the hot medium producing the X-ray continuum emission is extended in the accretion disk plane (orthogonal to the jet). Our findings support the expectation that the jets are launched orthogonal to the inner parts of the accretion flow. By extension, this means that the direction of the inner parts of the jet in Swift J1727.8−1613 are aligned with the BH spin.

The PD value is comparable to that found in the high-mass BH X-ray binary Cyg X-1, and can, by similarity, constrain the inclination of the innermost parts to be moderate, i ∼ 30°–60°. Also, the trend of PD with energy is inconsistent with the geometry of the spherical or lamppost corona, even if it is outflowing.

The Imaging X-ray Polarimetry Explorer (IXPE) is a joint US and Italian mission. The US contribution is supported by the National Aeronautics and Space Administration (NASA) and led and managed by its Marshall Space Flight Center (MSFC), with industry partner Ball Aerospace (contract NNM15AA18C). The Italian contribution is supported by the Italian Space Agency (Agenzia Spaziale Italiana, ASI) through contract ASI-OHBI-2022-13-I.0, agreements ASI-INAF-2022-19-HH.0 and ASI-INFN-2017.13-H0, and its Space Science Data Center (SSDC) with agreements ASI-INAF-2022-14-HH.0 and ASI-INFN 2021-43-HH.0, and by the Istituto Nazionale di Astrofisica (INAF) and the Istituto Nazionale di Fisica Nucleare (INFN) in Italy. This research used data products provided by the IXPE Team (MSFC, SSDC, INAF, and INFN) and distributed with additional software tools by the High-Energy Astrophysics Science Archive Research Center (HEASARC), at NASA Goddard Space Flight Center (GSFC). This research has made use of the MAXI data provided by RIKEN, JAXA, and the MAXI team.

A.V. thanks the Academy of Finland grant 355672 for support. M.D., J.S., J.Pod., and V.Kar. thank GACR project 21-06825X for the support and institutional support from RVO:67985815. A.I. acknowledges support from the Royal Society. H.K. acknowledges support by NASA grants 80NSSC22K1291, 80NSSC23K1041, and 80NSSC20K0329. V.Kra. acknowledges support from the Finnish Cultural Foundation. The French contribution is supported by the French Space Agency (Centre National d'Etude Spatiale, CNES) and by the High Energy National Programme (PNHE) of the Centre National de la Recherche Scientifique (CNRS). I.L. was supported by the NASA Postdoctoral Program at the Marshall Space Flight Center, administered by Oak Ridge Associated Universities under contract with NASA.

Facilities: IXPE - , MAXI. - JAXA Monitor of All-sky X-ray Image experiment

Software: ixpeobssim (Baldini et al. 2022), xspec (Arnaud 1996).