On the mechanism of hard X-ray emission from magnetars

TL;DR

This paper proposes a model explaining the hard X-ray emission from magnetars as resulting from relativistic particle outflows and resonant scattering, producing observable features like high luminosity, broad spectra, and pulsations.

Contribution

It introduces a detailed physical mechanism linking magnetospheric particle outflows to the observed X-ray spectra and polarization features in magnetars.

Findings

Hard X-ray luminosity exceeds thermal emission.

Spectrum extends from 10 keV to MeV with a hard slope.

Emission shows strong pulsations and polarization dependence.

Abstract

Persistent activity of magnetars is associated with electric discharge that continually injects relativistic particles into the magnetosphere. Large active magnetic loops around magnetars must be filled with outflowing particles that interact with radiation via resonant scattering and spawn electron-positron pairs. The outflow energy is processed into copious e+- until the plasma enters outer parts of the loop where the magnetic field is reduced below 10^13 G. In the outer zone, photons scattered by the outflow do not convert to e+- pairs and the outflow radiates its energy away. The escaping radiation forms a distinct hard X-ray peak in the magnetar spectrum. It has the following features: (1) Its luminosity L=10^35-10^36 erg/s can easily exceed the thermal luminosity from the magnetar surface. (2) Its spectrum extends from 10 keV to the MeV band with a hard spectral slope, which depends on the object inclination to the line of sight. (3) The anisotropic hard X-ray emission exhibits strong pulsations as the magnetar spins. (4) The emission spectrum typically peaks around 1 MeV, but the peak position significantly oscillates with the spin period. (5) The emission is dominated by the extraordinary polarization mode at photon energies below 1 MeV. (6) The decelerated pairs accumulate and annihilate at the top of the magnetic loop, and emit the 511-keV line with luminosity L_ann\sim0.1L. Features (1)-(3) agree with available data; (4)-(6) can be tested by future observations.