The long quest for vacuum birefringence in magnetars: 1E 1547.0-5408 and the elusive smoking gun

TL;DR

This study analyzes a 500 ks IXPE observation of the magnetar 1E 1547.0-5408 to probe vacuum birefringence and the emission geometry in strong magnetic fields. The spectrum is well described by a single thermal component with $kT_{\mathrm{BB}}\approx0.67$ keV and $R_{\mathrm{BB}}\approx1.2$ km, and the source shows a high phase-averaged polarization degree of $\mathrm{PD}\approx0.48$ with PA ≈ $76^{\circ}$, while phase-resolved analysis reveals a small, nonuniform hotspot and a rotating-vector-model (RVM) driven PA modulation indicating misalignment between the spin axis and the observer’s line of sight. Our RVM fit implies a high-obliquity geometry that reduces the expected PD enhancement from vacuum birefringence, making the IXPE PD less conclusive as a QED smoking gun in this source; a possible dip in $\mathrm{PD}$ around $3$–$4$ keV could reflect partial mode conversion at the vacuum resonance, but no definitive evidence is found. The results underscore how geometry and atmospheric emission dominate the observed polarization, suggesting that future missions and broader samples—especially transient magnetars or XDINSs—are essential to robustly test vacuum birefringence in neutron-star magnetospheres.

Abstract

Magnetars are now known to be among the most strongly polarized celestial sources in X-rays. Here we report on the $500\,\mathrm{ks}$ observation of the magnetar 1E 1547.0-5408 performed by the Imaging X-ray Polarimetry Explorer (IXPE) in March 2025. The IXPE spectrum is well reproduced by a single thermal component with blackbody temperature $kT_\mathrm{BB}\sim 0.67\,\mathrm{keV}$ and emission radius $R_\mathrm{BB}\sim 1.2\,\mathrm{km}$. The source exhibits a high linear polarization degree in the $2$--$6\,\mathrm{keV}$ band ($\mathrm{PD}=47.7\pm2.9\%$) with polarization angle $\mathrm{PA}=75^\circ.8 \pm 1.^\circ8$, measured West of celestial North. While $\mathrm{PA}$ does not appear to vary with energy, there is some evidence (at the $1σ$ confidence level) of a minimum in $\mathrm{PD}$ between $3$ and $4\,\mathrm{keV}$, compatible with what is expected by partial mode conversion at the vacuum resonance in a magnetized atmosphere. Phase-resolved spectral and polarimetric analyses reveal that X-ray thermal radiation likely originates from a single, fairly small hot spot with a non-uniform temperature distribution. Fitting the phase-dependent $\mathrm{PA}$ measured by IXPE with a rotating vector model (RVM) constrains the source geometry and indicates that both the dipole axis and line-of-sight are misaligned with respect to the spin axis. Under these conditions, the high polarization of the source cannot be regarded as compelling evidence for the presence of vacuum birefringence in the star magnetosphere. Nevertheless, the fact that the RVM successfully reproduces the modulation of the X-ray polarization angle and the behavior of $\mathrm{PD}$ with the energy hint once more to the presence of QED effects in magnetars.

Figures (10)

• Figure 1: IXPE source (dots with error bars) and background (stars) spectra (top panel) of 1E 1547.0$-$5408 collected by DU1 (black), DU2 (red) and DU3 (green), and fitting residuals in units of the standard deviation (bottom panels) for the tbabs$\times$bbodyrad model with $N_\mathrm{H}$ frozen at $4.6\times10^{22}\,\mathrm{cm}^{-2}$ (left), tbabs$\times$(bbodyrad$+$bbodyrad) model with $N_\mathrm{H}$ frozen at $4.6\times10^{22}\,\mathrm{cm}^{-2}$ (center) and tbabs$\times$(bbodyrad$+$powerlaw) model with $N_\mathrm{H}$ frozen at $4.9\times10^{22}\,\mathrm{cm}^{-2}$ (right). The best fitting models (solid histograms) and the individual components (dotted histograms) are also shown in each plot. Fit results are summarized in Table \ref{['tab:spec_xspec']}.
• Figure 2: IXPE pulse-profile of 1E 1547.0$-$5408 in the $2$--$6$, $2$--$3$, $3$--$4$, $4$--$5$ and $5$--$6\,\mathrm{keV}$ energy bands.
• Figure 3: Phase variation of the blackbody temperature (cyan) and radius (red) for the best-fitting model reported in Table \ref{['tab:spec_xspec_pr']}.
• Figure 4: 1E 1547.0$-$5408 phase- and energy-integrated ($2$--$6\,\mathrm{keV}$) most-probable values for $\mathrm{PD}$ and $\mathrm{PA}$ (crosses) measured by the three IXPE DUs (DU1 yellow, DU2 green and DU3 red); the combined result is also shown in blue. Constant PD and constant PA loci are marked by concentric circles and radial lines, respectively, with $\mathrm{PA}=0^\circ$ corresponding to the celestial North, and PA decreasing Westward. The shaded and empty contours represents the $68\%$ and $99\%$ confidence regions, respectively, obtained using the steppar command in xspec.
• Figure 5: Top: energy-dependent polarization degree (cyan) and polarization angle (red) obtained with xspec (see text for details). Error bars show the $1\sigma$ confidence intervals derived with the err procedure. When the PD measurement falls below $\mathrm{MDP}_{99}$, a downwards arrow marks the $3\sigma$ upper limit; the associated value of PA is unconstrained and is shown by a double-headed arrow. Bottom: same in the PD-PA plane for the $2$--$3$ (blue), $3$--$4$ (orange), $4$--$5$ (green) and $5$--$6\,\mathrm{keV}$ (red) energy bins; symbols and contour codes are as in Figure \ref{['fig:mirino26']}. All values are reported in Table \ref{['tab:pdpa_xspec']}.