2019MNRAS.487.2624A

We investigated the long-term behaviour in X-rays of the colliding wind binary WR 25, using archival data obtained with Suzaku, Swift, XMM-Newton, and NuSTAR spanning over ∼16 yr. Our analysis reveals phase-locked variations repeating consistently over many consecutive orbits, in agreement with an X-ray emission fully explained by thermal emission from the colliding winds in the 208-d orbit. We report on a significant deviation of the X-ray flux with respect to the 1/D trend (expected for adiabatic shocked winds) close to periastron passage. The absence of a drop in post-shock plasma temperature close to periastron suggests this break in trend cannot be explained in terms of reduced pre-shock velocities in this part of the orbit. Finally, NuSTAR data reveal a lack of hard X-ray emission (above 10.0 keV) above the background level. Upper limits on a putative non-thermal emission strongly suggest that the sensitivity of present hard X-ray observatories is not sufficient to detect non-thermal emission from massive binaries above 10 keV, unless the wind kinetic power is large enough to significantly feed particle acceleration in the wind-wind interaction.