The detection of high X-ray polarization from an accretion disc corona source and its modelling via Monte Carlo radiation transfer simulation

TL;DR

This work demonstrates that IXPE detects a high time-averaged X-ray polarization ($PD \approx 8.8\%$) from the high-inclination ADC source 2S 0921-630, with PD increasing with energy and a PA that varies across the band. To interpret this, the authors develop a Monte Carlo radiative-transfer model that combines boundary-layer emission, disc reflection, intrinsic disc emission, and, for long-period systems, an outer disc wind. The model successfully reproduces the PD magnitude and energy-dependent rise but cannot reproduce the observed PA swing, suggesting non-axisymmetric structures (e.g., misalignment or warp) or time-dependent geometry. Overall, the results support the ADC, wind-scattering interpretation for 2S 0921-630 and underscore the importance of non-axisymmetric effects in NS LMXRB polarization studies, motivating further multi-wavelength and non-axisymmetric modelling.

Abstract

We report an X-ray polarization degree (PD) of $8.8\pm1.4\%~(1σ)$ from the accretion-disc-corona (ADC) neutron-star system 2S 0921-630 (=V395 Car) observed with the Imaging X-ray Polarimetry Explorer (IXPE). The PD increases with energy, while the polarization angle (PA) varies significantly across the band. These trends are consistent with a high-inclination ADC geometry where the vertically extended disc blocks direct sight of the central X-ray source, and the observed X-rays are those scattered in an equatorial disc wind. We also find tentative PD variability in the 2--3 keV band. To interpret the time-averaged polarization, we build spectropolarimetric models by Monte Carlo radiation transfer simulation with column density distribution of thermal-radiative wind launched by X-ray irradiation of the outer disc under an axisymmetric geometry. The model combines boundary-layer emission, its disc reflection, and the disc continuum, each with its intrinsic polarization. Scattering of this composite spectrum in the wind reproduces both the observed PD and its increase with energy. However, the observed PA evolution is not captured, which may indicate departures from axisymmetry--e.g. misalignment between the inner disc (and/or neutron-star spin) and the outer disc/wind, or a weak disc warping.

Figures (10)

• Figure 1: The IXPE light curve of 2S 0921–630 with 1000s time binning. The errors in PD and PA show 1 $\sigma$ confidence level. The orange-shaded region shows the time average PD and PA.
• Figure 2:
• Figure 3: The time average polarization taken by spectro-polarimetric fitting (Tab.\ref{['tab:time_average']}) The contour lines show confidence levels of $1\sigma =68.3\%$, $2\sigma =95.5\%$, and $3\sigma =99.7\%$.
• Figure 4: The energy dependence of PD and PA taken by spectro-polarimetric fitting. The blue points show the time averaged data, while the orange points show those from time segment 2. The 2-3 keV band of segment 2 clearly shows the higher PD than the time-averaged, whereas the PAs are consistent with each other. The errors are $1\sigma$ confidence levels.
• Figure 5: The schematic geometry of the model. (A) The model of the normal lower magnetised neutron star. We calculated different source distributions and combined them to predict total polarization. (B) The model of the Accretion disk corona source. We add the outer disc and disc wind to the model (A).