Radiative properties of highly magnetized isolated neutron star surfaces and approximate treatment of absorption features in their spectra

In the X-ray spectra of most X-ray dim isolated neutron stars (XDINSs) absorption features with equivalent widths (EWs) of 50 – 200 eV are observed. We theoretically investigate different models to explain absorption features and compare their properties with the observations. We consider various theoretical models for the magnetized neutron star surface: naked condensed iron surfaces and partially ionized hydrogen model atmospheres, including semi-infinite and thin atmospheres above a condensed surface. The properties of the absorption features (especially equivalent widths) and the angular distributions of the emergent radiation are described for all models. A code for computing light curves and integral emergent spectra of magnetized neutron stars is developed. We assume a dipole surface magnetic field distribution with a possible toroidal component and corresponding temperature distribution. A model with two uniform hot spots at the magnetic poles can also be employed. Light curves and spectra of highly magnetized neutron stars with parameters typical for XDINSs are computed using different surface temperature distributions and various local surface models. Spectra of magnetized model atmospheres are approximated by diluted blackbody spectra with one or two Gaussian lines having parameters, which allow us to describe the model absorption features. To explain the prominent absorption features in the soft X-ray spectra of XDINSs a thin atmosphere above the condensed surface can be invoked, whereas a strong toroidal magnetic field component on the XDINS surfaces can be excluded.

💡 Research Summary

The paper addresses the prominent absorption features observed in the soft‑X‑ray spectra of X‑ray dim isolated neutron stars (XDINSs), whose equivalent widths (EWs) typically lie between 50 eV and 200 eV. The authors explore several theoretical surface and atmospheric configurations for highly magnetized neutron stars (magnetic fields of order 10¹³–10¹⁴ G) and evaluate which can reproduce the observed spectral signatures.

Four principal local surface models are considered: (1) a bare condensed iron surface, (2) a semi‑infinite, partially ionized hydrogen atmosphere, (3) a thin hydrogen atmosphere (optical depth τ ≲ 1) overlaying a condensed iron surface, and (4) a hybrid model that combines the thin atmosphere with the underlying condensed surface. For each configuration the emergent radiation’s angular distribution, spectral shape, and the properties of any absorption lines (central energy, width, and EW) are computed using a newly developed radiative‑transfer code.

The study finds that a bare condensed iron surface alone produces only shallow features (EW ≲ 30 eV), insufficient to explain the strong lines seen in XDINSs. Semi‑infinite hydrogen atmospheres generate broader lines but still fall short of the observed EW range unless the atmosphere is unrealistically thick. In contrast, a thin hydrogen layer above a condensed surface yields strong Gaussian‑shaped absorption features with EWs comfortably within the 50–200 eV interval. The line strength in this scenario is highly sensitive to the atmosphere’s column density, temperature, and magnetic field strength, allowing a natural explanation for the diversity of observed line depths among different XDINSs.

Beyond local surface physics, the authors examine global magnetic field geometry. They model a pure dipole field and a dipole plus toroidal component, each coupled to a temperature distribution derived from magneto‑thermal equilibrium. A strong toroidal component dramatically distorts the temperature map and the angular emission pattern, leading to either overly broadened lines or their disappearance in the integrated spectrum. Simulated light curves for such configurations also fail to match the observed pulse profiles of XDINSs. Consequently, the presence of a dominant toroidal field on the surface of XDINSs is ruled out.

The paper also implements two global surface temperature scenarios: (i) a continuous dipolar temperature distribution and (ii) a simplified model with two uniform hot spots located at the magnetic poles. Light‑curve and phase‑averaged spectra are generated for each case, confirming that the thin‑atmosphere‑over‑condensed‑surface model reproduces both the spectral shape and the modest pulsations typical of XDINSs.

To facilitate direct comparison with observational data, the authors demonstrate that the complex model spectra can be accurately approximated by a diluted blackbody continuum plus one or two Gaussian absorption lines. The fitted parameters (line centroid, width, depth) encapsulate the essential physics of the underlying atmosphere‑condensed‑surface system, providing a practical tool for fitting current and future high‑resolution X‑ray observations (e.g., with ATHENA or eXTP).

In summary, the work concludes that the most plausible explanation for the strong absorption features in XDINSs is a thin, partially ionized hydrogen atmosphere residing above a condensed iron surface, while a strong toroidal magnetic field component is inconsistent with both spectral and timing data. This model offers a robust framework for interpreting existing observations and guiding the analysis of forthcoming X‑ray missions.