# Narrow phase-dependent features in X-ray Dim Isolated Neutron Stars: a new detection and upper limits

## TL;DR

This study identifies a narrow, phase-dependent absorption feature in the X-ray spectrum of RX J1308.6+2127, suggesting the presence of localized high-magnetic-field structures near the neutron star surface, and explores similar features in other XDINSs.

## Contribution

First detection of a narrow, phase-variable absorption feature in an XDINS, indicating localized magnetic loops near the neutron star surface.

## Key findings

- Detected a 740 eV absorption feature in RX J1308.6+2127.
- Estimated magnetic field in the loop as ~1.7 x 10^{14} G.
- Set upper limits for similar features in other XDINSs.

## Abstract

We report on the results of a detailed phase-resolved spectroscopy of archival XMM--Newton observations of X-ray Dim Isolated Neutron Stars (XDINSs). Our analysis revealed a narrow and phase-variable absorption feature in the X-ray spectrum of RX J1308.6+2127. The feature has an energy of $\sim$740 eV and an equivalent width of $\sim$15 eV. It is detected only in $\sim$ 1/5 of the phase cycle, and appears to be present for the entire timespan covered by the observations (2001 December - 2007 June). The strong dependence on the pulsar rotation and the narrow width suggest that the feature is likely due to resonant cyclotron absorption/scattering in a confined high-B structure close to the stellar surface. Assuming a proton cyclotron line, the magnetic field strength in the loop is B$_{loop} \sim 1.7 \times 10^{14}$ G, about a factor of $\sim$5 higher than the surface dipolar magnetic field (B$_{surf} \sim 3.4 \times 10^{13}$ G). This feature is similar to that recently detected in another XDINS, RX J0720.4-3125, showing (as expected by theoretical simulations) that small scale magnetic loops close to the surface might be common to many highly magnetic neutron stars (although difficult to detect with current X-ray instruments). Furthermore, we investigated the available XMM--Newton, data of all XDINSs in search for similar narrow phase-dependent features, but could derive only upper limits for all the other sources.