Non-Thermal Radio Emission from Colliding-Wind Binaries

In colliding-wind binaries, shocks accelerate a fraction of the electrons up to relativistic speeds. These electrons then emit synchrotron radiation at radio wavelengths. Whether or not we detect this radiation depends on the size of the free-free absorption region in the stellar winds of both components. One expects long-period binaries to be detectable, but not the short-period ones. It was therefore surprising to find that Cyg OB2 No. 8A (P = 21.9 d) does show variability locked with orbital phase. To investigate this, we developed a model for the relativistic electron generation (including cooling and advection) and the radiative transfer of the synchrotron emission through the stellar wind. Using this model, we show that the synchrotron emitting region in Cyg OB2 No. 8A does extend far enough beyond the free-free absorption region to generate orbit-locked variability in the radio flux. This model can also be applied to other non-thermal emitters and will prove useful in interpreting observations from future surveys, such as COBRaS - the Cyg OB2 Radio Survey.

💡 Research Summary

This paper addresses the long‑standing puzzle of why some short‑period colliding‑wind binaries (CWBs) exhibit orbital‑phase‑locked non‑thermal radio variability despite theoretical expectations that free‑free absorption should completely quench synchrotron emission in such systems. The authors focus on Cyg OB2 No. 8A, a 21.9‑day O‑star binary that shows clear radio flux modulation synchronized with its orbit. To explain this, they construct a two‑stage physical model that couples relativistic electron production at the wind‑collision shock with detailed radiative transfer through the stellar winds.

In the first stage, electrons are accelerated via diffusive shock acceleration (DSA) at the wind‑collision region (WCR). The model parameterises the acceleration efficiency (η) and the power‑law index (p) of the injected electron spectrum. It then follows the downstream evolution of the electron distribution N(E, r) by solving a one‑dimensional transport equation that includes advection with the post‑shock flow, synchrotron cooling, inverse‑Compton losses against the stellar radiation fields, and adiabatic expansion. The authors adopt realistic stellar wind parameters for the primary (O5.5 I) and secondary (O9 III) components—mass‑loss rates of ~2 × 10⁻⁶ M⊙ yr⁻¹ and ~1 × 10⁻⁶ M⊙ yr⁻¹, terminal velocities of 2500 km s⁻¹ and 2000 km s⁻¹, and wind temperatures of ~30 000 K. A magnetic field strength of order 10 G is assumed in the WCR, consistent with equipartition estimates.

The second stage computes the emergent radio emission. The synchrotron emissivity jν(r) is derived from the local electron distribution and magnetic field, while the free‑free absorption coefficient κν(r) is calculated from the wind density and temperature profiles. The radiative transfer equation Iν = ∫ jν e⁻ᵗᵃᵘ ds is integrated along the line of sight to obtain the observable flux density Sν as a function of orbital phase. A key diagnostic is the comparison between the synchrotron “photosphere” radius (Rsyn), where the synchrotron optical depth becomes unity, and the free‑free “photosphere” radius (Rff). If Rsyn > Rff, a fraction of the synchrotron radiation escapes, producing detectable, phase‑dependent variability.

Applying this framework to Cyg OB2 No. 8A, the authors find that relativistic electrons advect to distances of ~10 stellar radii before cooling significantly, extending the synchrotron emitting region well beyond the free‑free absorption radius of ~5 R★. Consequently, as the binary orbits, the line‑of‑sight column through the absorbing wind changes, modulating the observed radio flux. The model reproduces both the amplitude (≈30 % variation) and the phase lag of the observed light curve, confirming that the synchrotron source is sufficiently extended to survive the wind opacity.

Sensitivity tests reveal that modest variations in η (as low as 10⁻⁴) and magnetic field strength (5–20 G) can alter the Rsyn/Rff ratio dramatically, highlighting the importance of accurate wind and magnetic diagnostics. The authors also discuss potential complications such as wind clumping, asymmetries, and orbital eccentricity, suggesting that full three‑dimensional magneto‑hydrodynamic simulations would refine the picture.

Beyond Cyg OB2 No. 8A, the paper argues that the same modelling approach can be applied to other non‑thermal radio emitters (e.g., WR 140, HD 168112). By providing a quantitative link between wind parameters, shock physics, and observable radio variability, the model offers a valuable tool for interpreting upcoming large‑scale surveys like COBRaS (the Cyg OB2 Radio Survey). In particular, it enables the prediction of which binaries are likely to be detectable non‑thermal sources and facilitates the extraction of physical parameters (mass‑loss rates, magnetic fields, acceleration efficiencies) from observed radio light curves.

In summary, the study demonstrates that even in relatively short‑period massive binaries, the synchrotron emitting region can extend beyond the free‑free absorption zone, allowing orbital‑phase‑locked radio variability to emerge. This resolves the apparent contradiction between theory and observation for Cyg OB2 No. 8A and establishes a robust framework for future investigations of non‑thermal processes in massive star binaries.