Signatures of photon-axion conversion in the thermal spectra and polarization of neutron stars

Conversion of photons into axions under the presence of a strong magnetic field can dim the radiation from magnetized astrophysical objects. Here we perform a detailed calculation aimed at quantifying the signatures of photon-axion conversion in the spectra, light curves, and polarization of neutron stars (NSs). We take into account the energy and angle-dependence of the conversion probability and the surface thermal emission from NSs. The latter is computed from magnetized atmosphere models that include the effect of photon polarization mode conversion due to vacuum polarization. The resulting spectral models, inclusive of the general-relativistic effects of gravitational redshift and light deflection, allow us to make realistic predictions for the effects of photon to axion conversion on observed NS spectra, light curves, and polarization signals. We identify unique signatures of the conversion, such as an increase of the effective area of a hot spot as it rotates away from the observer line of sight. For a star emitting from the entire surface, the conversion produces apparent radii that are either larger or smaller (depending on axion mass and coupling strength) than the limits set by NS equations of state. For an emission region that is observed phase-on, photon-axion conversion results in an inversion of the plane of polarization with respect to the no-conversion case. While the quantitative details of the features that we identify depend on NS properties (magnetic field strength, temperature) and axion parameters, the spectral and polarization signatures induced by photon-axion conversion are distinctive enough to make NSs very interesting and promising probes of axion physics.

💡 Research Summary

This paper presents a comprehensive theoretical study of how photon‑axion conversion in the strong magnetic fields of neutron stars (NSs) modifies the observable thermal X‑ray emission, light curves, and polarization signals. The authors begin by deriving the conversion probability (P_{\gamma\to a}(E,\theta)) as a function of photon energy (E) and propagation angle (\theta) relative to the magnetic field. Their calculation incorporates both vacuum birefringence (the QED vacuum resonance) and plasma effects that split the photon field into ordinary (O‑mode) and extraordinary (X‑mode) polarization states. The probability depends sensitively on the axion mass (m_a) and the photon‑axion coupling constant (g_{a\gamma}); for typical NS magnetic fields ((10^{13-15}) G) the conversion can reach the percent level when (g_{a\gamma}\gtrsim10^{-11}) GeV(^{-1}) and (m_a\lesssim10^{-5}) eV.

Next, the authors generate realistic surface emission spectra using state‑of‑the‑art magnetized atmosphere models. These models already include the effect of vacuum‑induced mode conversion, which determines the relative contributions of O‑ and X‑mode photons as a function of temperature, magnetic field strength, and viewing angle. The intrinsic spectra are then multiplied by the angle‑ and energy‑dependent conversion probability, producing a “post‑conversion” spectrum for each surface element. General relativistic effects—gravitational redshift and light‑bending around the compact star—are applied to map the locally emitted photons to the observer’s frame, ensuring that the final synthetic spectra, light curves, and polarization degrees are directly comparable to data.

The study identifies three robust observational signatures of photon‑axion conversion:

1. Effective hot‑spot area increase – As a rotating hot spot moves out of the line of sight, the X‑mode photons (which dominate the emission at many angles) are preferentially depleted by conversion, making the spot appear larger in the inferred emitting area. This produces a characteristic rise in the observed flux at phases where standard models predict a decline.

2. Apparent radius anomalies – For stars that emit uniformly over the whole surface, the total flux can be either suppressed or enhanced depending on the axion parameters. Consequently, the inferred radiation radius (R_{\rm eff}) can lie outside the range allowed by neutron‑star equations of state (EOS). In the high‑coupling, low‑mass regime the radius appears larger; in the opposite regime it appears smaller.

3. Polarization plane inversion – When the hot spot is viewed “phase‑on” (directly facing the observer), the conversion preferentially removes X‑mode photons, leaving O‑mode photons to dominate. This flips the linear polarization angle by roughly 90° relative to the case with no conversion. The effect is distinct from the standard vacuum resonance flip and should be detectable with upcoming X‑ray polarimetry missions such as IXPE and eXTP.

By scanning the ((m_a, g_{a\gamma})) parameter space, the authors delineate a region where these signatures become statistically significant. Importantly, this region is not yet excluded by laboratory experiments (e.g., ADMX, CAST), highlighting neutron stars as complementary probes of axion‑like particles. The paper also discusses extensions to other strongly magnetized objects (magnetars, white dwarfs) and stresses the importance of multi‑wavelength, multi‑polarization observations to disentangle axion effects from astrophysical uncertainties.

In conclusion, the work demonstrates that photon‑axion conversion leaves distinct imprints on NS thermal spectra, timing, and polarization. By integrating detailed atmosphere physics, conversion probabilities, and relativistic ray‑tracing, the authors provide realistic templates that can be directly compared with current and future X‑ray observations. This establishes neutron stars as powerful astrophysical laboratories for testing axion physics and potentially tightening constraints on the axion mass–coupling plane beyond what terrestrial experiments can achieve.