GX 13+1: Neutron-Star LMXB Dynamics

• GX 13+1 is a persistent neutron-star low-mass X-ray binary that exhibits complete Z-track behavior, periodic X-ray dips, and a sustained ionized disk wind.
• Recent multi-wavelength campaigns reveal its dynamic interplay between high Eddington accretion, variable radio jet activity, and phase-dependent absorption features.
• Detailed spectral-timing and polarimetric analyses from NICER, Chandra, XMM-Newton, and IXPE highlight its high inclination geometry and evolving polarization signatures.

Searching arXiv for the cited GX 13+1 papers to ground the article in current literature. {"queries":[{"query":"GX 13+1 arXiv (Rogantini et al., 7 Apr 2025) persistent disk wind variable jet outflow neutron-star low-mass X-ray binary","max_results":5},{"query":"GX 13+1 arXiv (Kaddouh et al., 2024) NICER full Z-track","max_results":5},{"query":"GX 13+1 arXiv (Trigo et al., 2012) XMM-Newton observations correlation photoionised absorption broad line emission","max_results":5},{"query":"GX 13+1 arXiv (Allen et al., 2018) Disk Wind in the Neutron Star Low-mass X-Ray Binary GX 13+1","max_results":5},{"query":"GX 13+1 arXiv (Bobrikova et al., 2024) rotation X-ray polarization angle","max_results":5},{"query":"GX 13+1 arXiv (Marco et al., 9 Jan 2025) X-ray Dips and Polarization Angle Swings","max_results":5}]} I found the relevant arXiv records for the recent GX 13+1 literature, including the 2025 multi-wavelength wind–jet study (Rogantini et al., 7 Apr 2025), the NICER Z-track classification work (Kaddouh et al., 2024), the XMM-Newton wind/iron-line analysis (Trigo et al., 2012), the Chandra disk-wind study (Allen et al., 2018), and the IXPE polarization papers (Bobrikova et al., 2024, Marco et al., 9 Jan 2025). GX 13+1 is a persistent neutron-star low-mass X-ray binary in the Galactic bulge whose phenomenology combines a long orbital period, high inclination, luminous accretion, a persistent photoionized disk wind, broad Fe K emission, radio jet activity, and state-dependent X-ray polarization. Over the last decade, coordinated X-ray, radio, and polarimetric campaigns have shifted its status from a historically ambiguous atoll/Z-source candidate to a source that traces the full Z track, exhibits periodic and aperiodic dips, and sustains both an ionized wind and a variable jet at high Eddington rates (Kaddouh et al., 2024, Rogantini et al., 7 Apr 2025).

1. Classification, binary context, and basic system properties

GX 13+1 is a persistent neutron-star low-mass X-ray binary and a type-I X-ray burster, establishing the compact object as a neutron star (Trigo et al., 2012). It is associated with a late-type K5 III donor at a distance of $7 \pm 1$ kpc, and the accretion flow is persistently luminous, with several studies placing the source at a substantial fraction of the Eddington limit (Trigo et al., 2012, Allen et al., 2018).

Its taxonomic status was long uncertain. Earlier work alternated between atoll-like and Z-source interpretations, in part because incomplete color-color or hardness-intensity tracks and mixed timing properties did not cleanly match canonical source classes (Kaddouh et al., 2024). NICER resolved this ambiguity by showing, for the first time, that GX 13+1 unambiguously traces a complete Z track with horizontal branch, normal branch, and flaring branch, thereby establishing the source as a Z source in terms of long-term spectral-timing behavior (Kaddouh et al., 2024).

The orbital period is now measured from periodic X-ray dips as

$P_{\rm orb} = 24.5274(2)\ \mathrm{days},$

with reference epoch

$T_0 = 50{,}086.79(3)\ \mathrm{MJD},$

so that

$$T_{\rm dip}(N) = 50{,}086.79(3) + 24.5274(2)\,N\ \mathrm{MJD}.$$

This ephemeris was derived from ASM, MAXI, and Chandra dip arrival times and is described as the first precise X-ray orbital ephemeris of the system (Iaria et al., 2013). The same work states that this is the longest known orbital period for a Galactic neutron-star low-mass X-ray binary powered by Roche-lobe overflow (Iaria et al., 2013).

The source is also a high-inclination system. Strong ionized absorption, periodic dipping, and the absence of eclipses have repeatedly led to inclination estimates in the range $60^\circ$–$80^\circ$ (Trigo et al., 2012, Iaria et al., 2013). This geometrical setting is central to the visibility of the disk wind, the dipping behavior, and the later polarimetric interpretations.

2. Orbital modulation, dips, and viewing geometry

The X-ray dips in GX 13+1 are energy dependent and periodic at the orbital period, with deeper suppression at lower energies and associated hardness-ratio increases, consistent with photoelectric absorption by material fixed in the corotating frame (Iaria et al., 2013). Folding RXTE/ASM and MAXI light curves at 24.53 d reveals a distinct dip near phase 0, and a Chandra/HETGS observation caught a full dip at the predicted phase, thereby linking the long-term modulation to a recurring geometrical event (Iaria et al., 2013).

The dip profiles indicate that GX 13+1 is a classical dipping LMXB, but one superposed on a separate, persistent ionized outflow. The periodic dips are attributed to structures at the outer disk, most naturally the stream-impact bulge, whereas non-phase-locked absorption episodes seen in pointed observations are associated with variable covering by the inner disk wind or outflow (Iaria et al., 2013). This separation between orbital dips and irregular absorption states is important because the source shows both phenomena.

A Chandra/HETGS observation provided the first spectral characterization of a periodic dip. The event lasted $\sim 450$ s at full width at half maximum, with a full episode of $\sim 1400$ s including ingress and egress, and was accompanied by an increase in the column density of the neutral absorber while the warm absorber remained consistent with its out-of-dip state (D'Aì et al., 2014). In that analysis, the out-of-dip neutral column was $$\approx (3.7\text{–}4.1)\times 10^{22}\,\mathrm{cm^{-2}}$$, whereas the dip spectrum reached

$$N_{\rm H,cold} = 8.62 \pm 0.16\times 10^{22}\,\mathrm{cm^{-2}},$$

implying an extra neutral column of $P_{\rm orb} = 24.5274(2)\ \mathrm{days},$0 (D'Aì et al., 2014).

The same work used the dip duration and orbital parameters to infer a compact absorber at the outer disk, with characteristic size $P_{\rm orb} = 24.5274(2)\ \mathrm{days},$1 cm, much smaller than the outer disk radius (D'Aì et al., 2014). This supports a picture in which shallow, rare dips arise because the absorbing bulge is small compared with the very large accretion disk implied by the 24.5274 d orbit. A plausible implication is that GX 13+1 combines a large-scale equatorial wind with a comparatively localized outer-disk dip structure.

3. Z-track phenomenology and rapid variability

NICER archival data from 2023 February to 2024 April showed GX 13+1 tracing the entire Z track in a 2–6.8 keV hardness-intensity diagram, with the complete track observed during 2023-04-22 to 2023-04-27 (Kaddouh et al., 2024). In that construction, the soft band was 2–3.8 keV, the hard band 3.8–6.8 keV, and the intensity band 2–6.8 keV, with 64 s bins (Kaddouh et al., 2024).

One notable result is that the horizontal branch has a positive slope in the NICER soft-band hardness-intensity diagram: both intensity and hardness increase together along the branch (Kaddouh et al., 2024). The authors proposed two explanations already grounded in prior spectroscopy of GX 13+1: high intrinsic absorption, with $P_{\rm orb} = 24.5274(2)\ \mathrm{days},$2, and the strong contribution of soft spectral components in NICER’s bandpass (Kaddouh et al., 2024). This does not require a non-standard accretion mode; rather, it shows that Z-track morphology depends strongly on energy band and absorption.

The same NICER study detected a broad peaked-noise component at

$P_{\rm orb} = 24.5274(2)\ \mathrm{days},$3

with $P_{\rm orb} = 24.5274(2)\ \mathrm{days},$4 and fractional rms $P_{\rm orb} = 24.5274(2)\ \mathrm{days},$5 (Kaddouh et al., 2024). Because of its breadth, it was conservatively termed a peaked-noise component rather than a canonical NBO, although comparison with earlier RXTE behavior suggested an association with the normal branch (Kaddouh et al., 2024).

AstroSat later extended the low-frequency timing picture by detecting $P_{\rm orb} = 24.5274(2)\ \mathrm{days},$6 Hz QPOs in three hardness-intensity regions. The reported frequencies were $P_{\rm orb} = 24.5274(2)\ \mathrm{days},$7 Hz in Region A, $P_{\rm orb} = 24.5274(2)\ \mathrm{days},$8 Hz in Region B, and $P_{\rm orb} = 24.5274(2)\ \mathrm{days},$9 Hz in Region C, with Q-factors 2.80, 5.79, and 4.35, and low rms values $T_0 = 50{,}086.79(3)\ \mathrm{MJD},$0, $T_0 = 50{,}086.79(3)\ \mathrm{MJD},$1, and $T_0 = 50{,}086.79(3)\ \mathrm{MJD},$2, respectively (Pradhan et al., 1 Jul 2025). These were interpreted as normal-branch oscillations similar to those reported in GX 340+0 (Pradhan et al., 1 Jul 2025).

The AstroSat analysis also modeled the QPO rms and lag spectra with a propagative model in which the observed oscillations are likely driven by interactions between the corona and variations in the blackbody temperature (Pradhan et al., 1 Jul 2025). In that interpretation, variations in blackbody temperature, coronal heating rate, and optical depth contribute to the observed spectral-timing behavior. This suggests that in GX 13+1 the low-frequency variability is not confined to a single spectral component but couples the boundary-layer or neutron-star surface emission to the corona.

4. Photoionized absorption, disk wind, and Fe K emission

GX 13+1 is one of the best-studied neutron-star disk-wind systems. XMM-Newton spectroscopy found a strong correlation between the hard, 6–10 keV flux, the ionization and column density of the photoionized absorber, and the equivalent width of the broad iron line (Trigo et al., 2012). In that work, the absorber was modeled with warmabs and cabs, and the favored interpretation placed the absorbing and line-emitting material in a thermally driven disk wind and/or hot atmosphere at radii of order $T_0 = 50{,}086.79(3)\ \mathrm{MJD},$3 cm (Trigo et al., 2012).

For the ionized absorber, the standard ionization parameter

$T_0 = 50{,}086.79(3)\ \mathrm{MJD},$4

was used (Trigo et al., 2012, D'Aì et al., 2014, Allen et al., 2018). XMM-Newton fits gave warm-absorber columns in the range $T_0 = 50{,}086.79(3)\ \mathrm{MJD},$5 to $T_0 = 50{,}086.79(3)\ \mathrm{MJD},$6, and ionization parameters $T_0 = 50{,}086.79(3)\ \mathrm{MJD},$7 when fitted with a common ionizing continuum (Trigo et al., 2012). The same paper inferred wind radii $T_0 = 50{,}086.79(3)\ \mathrm{MJD},$8 cm and argued that GX 13+1 lies above the luminosity threshold for a thermally driven wind, with radiation pressure likely assisting because of the large electron-scattering optical depth (Trigo et al., 2012).

A broad Fe K emission feature is detected in all low-variability XMM-Newton intervals, with centroid $T_0 = 50{,}086.79(3)\ \mathrm{MJD},$9 keV and width $$T_{\rm dip}(N) = 50{,}086.79(3) + 24.5274(2)\,N\ \mathrm{MJD}.$$0 keV (Trigo et al., 2012). Its equivalent width is strongly state dependent, reaching $$T_{\rm dip}(N) = 50{,}086.79(3) + 24.5274(2)\,N\ \mathrm{MJD}.$$1 eV in high-column intervals and $$T_{\rm dip}(N) = 50{,}086.79(3) + 24.5274(2)\,N\ \mathrm{MJD}.$$2 eV in more ionized, lower-column intervals (Trigo et al., 2012). The line equivalent width increases with absorber column density and decreases with absorber ionization, while the centroid energy increases with ionization (Trigo et al., 2012). The authors favored reprocessing in the wind or hot atmosphere as the common origin of narrow absorption and broad Fe K emission (Trigo et al., 2012).

Chandra and RXTE later showed that a single absorber with standard abundances cannot account for all seven major wind features, implying multiple absorption zones (Allen et al., 2018). Two or three warmabs components reproduce the absorption complex with a low-ionization component at $$T_{\rm dip}(N) = 50{,}086.79(3) + 24.5274(2)\,N\ \mathrm{MJD}.$$3, $$T_{\rm dip}(N) = 50{,}086.79(3) + 24.5274(2)\,N\ \mathrm{MJD}.$$4, and outflow $$T_{\rm dip}(N) = 50{,}086.79(3) + 24.5274(2)\,N\ \mathrm{MJD}.$$5, plus a high-ionization component at $$T_{\rm dip}(N) = 50{,}086.79(3) + 24.5274(2)\,N\ \mathrm{MJD}.$$6, $$T_{\rm dip}(N) = 50{,}086.79(3) + 24.5274(2)\,N\ \mathrm{MJD}.$$7, and $$T_{\rm dip}(N) = 50{,}086.79(3) + 24.5274(2)\,N\ \mathrm{MJD}.$$8 (Allen et al., 2018). Using the most ionized absorber, that study estimated a launching radius of $$T_{\rm dip}(N) = 50{,}086.79(3) + 24.5274(2)\,N\ \mathrm{MJD}.$$9 cm for $60^\circ$0, consistent with the Compton radius and with a thermally driven wind (Allen et al., 2018).

The mass flux in the wind is potentially substantial. Using

$60^\circ$1

the Chandra study estimated $60^\circ$2 for a likely covering factor and stated that the wind mass-loss rate is comparable to the accretion rate, although the kinetic luminosity is only $60^\circ$3 (Allen et al., 2018). This implies that the wind can influence the accretion flow dynamically even if it is not a dominant feedback channel on larger scales.

5. Reflection, inner-flow constraints, and the wind–jet relation

The origin of the broad Fe K profile remains debated. XMM-Newton timing-mode spectroscopy showed that both a relativistic diskline model and a nonrelativistic windline model fit the asymmetric, red-skewed Fe line well, and a run-test at the 5% significance level could not distinguish between them (Maiolino et al., 2019). In the relativistic interpretation, the line arises in the inner accretion disk with $60^\circ$4 and inclination $60^\circ$5; in the windline interpretation, repeated electron scattering in a diverging outflow produces the red wing (Maiolino et al., 2019). This is a genuine interpretive controversy rather than a settled point.

NuSTAR strengthened the case for relativistic reflection by detecting both the Fe K line profile and a Compton hump in the 10–25 keV range (Saavedra et al., 2023). Branch-resolved fits with relxillNS and warmabs during a normal-branch to flaring-branch transition gave

$60^\circ$6

high inclination $60^\circ$7, and a magnetic-field upper limit

$60^\circ$8

from the inferred truncation radius (Saavedra et al., 2023). The same work estimated that the boundary layer extends $60^\circ$9 km above the neutron-star surface and inferred

$80^\circ$0

(Saavedra et al., 2023). It also argued that the reflection geometry becomes self-consistent only for high inner-disk densities $80^\circ$1, substantially above the densities available in current relxillNS grids (Saavedra et al., 2023).

The relation between winds and jets in GX 13+1 is also atypical. Earlier Chandra and RXTE work had already shown that a strong disk wind is present on the normal and horizontal branches, where radio jet activity is usually expected (Allen et al., 2018). A coordinated VLA, Chandra/HETG, and NICER campaign later tracked GX 13+1 across the entire Z track during high Eddington rates and found substantial resonance absorption features from the accretion-disk wind in all X-ray spectra, implying a persistent wind presence (Rogantini et al., 7 Apr 2025). Simultaneous VLA observations detected a variable radio jet, with radio emission notably strong during all flaring-branch observations and weaker on the normal branch, yet no clear correlation was found between the radio emission and the wind features (Rogantini et al., 7 Apr 2025). The campaign therefore demonstrated that an ionized disk wind and jet outflow can coexist in GX 13+1 and suggested that their launching mechanisms are not necessarily linked in this system (Rogantini et al., 7 Apr 2025).

6. X-ray and radio polarimetry, polarization-angle swings, and system geometry

IXPE introduced a new diagnostic layer by showing that GX 13+1 is weakly but significantly polarized in X-rays, and that its polarization is strongly time dependent. In the first IXPE observation, the source had an overall 2–8 keV polarization degree of $80^\circ$2 at a polarization angle of $80^\circ$3, but the polarization angle rotated by about $80^\circ$4 over two days while the polarization degree changed from $80^\circ$5 to non-detectable and then up to $80^\circ$6, without visible changes in spectroscopic characteristics (Bobrikova et al., 2024). The authors suggested a constant component of polarization, strong wind scattering, or different polarization of the two main spectral components as possible interpretations (Bobrikova et al., 2024).

A later IXPE, NICER, and Swift-XRT campaign directly linked two X-ray dips to polarization changes. During the dips, the harder Comptonized spectral component dominated, the polarization degree was higher than in the softer off-dip intervals, and the polarization angle showed a swing of $80^\circ$7 across dip and off-dip states (Marco et al., 9 Jan 2025). Joint analysis of the three IXPE observations showed that the polarization properties varied in response to intensity and spectral-hardness changes associated with dips, with the polarization degree attaining values up to $80^\circ$8 (Marco et al., 9 Jan 2025). The same study emphasized the role of an extended accretion-disk corona or disk wind in generating high polarization degrees and possibly the polarization-angle swings (Marco et al., 9 Jan 2025).

By 2025, combined IXPE, NICER, and VLA polarimetry had advanced the geometrical interpretation further. In one campaign the overall IXPE 2–8 keV polarization was $80^\circ$9 at $\sim 450$0, while the non-dip state showed $\sim 450$1 at $\sim 450$2, and the 4–8 keV non-dip band reached $\sim 450$3 (Kashyap et al., 7 Aug 2025). Spectro-polarimetric decomposition suggested a softer accretion-disk component and a harder blackbody from the boundary layer or spreading layer, with the harder component carrying polarization of order $\sim 450$4 (Kashyap et al., 7 Aug 2025). Radio polarimetry measured an intrinsic polarization angle $\sim 450$5, and, under the assumption that the disk is orthogonal to the jet, the authors inferred a substantial spin-orbit misalignment from the difference between the disk and boundary-layer or spreading-layer polarization angles (Kashyap et al., 7 Aug 2025). An earlier IXPE interpretation had already suggested a $\sim 450$6 misalignment of the neutron-star spin from the orbital axis (Bobrikova et al., 2024).

This suggests that GX 13+1 is not only a high-inclination wind source but also a system in which the relative contributions of disk, boundary or spreading layer, and scattering in the wind or corona vary enough to rotate the net polarization vector substantially. A plausible implication is that the same high-inclination geometry that makes the wind and dips observable also amplifies the polarimetric signatures of changing direct and scattered components.