2023A&A...669A.130A

Context. Thanks to their proximity, ultraluminous X-ray sources (ULXs) represent a privileged astrophysical laboratory to study super-Eddington accretion. Current open questions concern the nature of the compact object, which is still hard to determine in those cases where pulsations are not directly detected, and the mechanisms responsible for the spectral changes observed in many ULXs.
Aims. We investigate the nature of the ULX M 81 X-6, which has been suggested to harbour a neutron star (NS), by studying its long-term X-ray spectral and temporal evolution, with the goal of assessing the astrophysical phenomena responsible for its spectral changes.
Methods. Using the rich set of available archival data from XMM-Newton, Chandra, NuSTAR, and Swift/XRT, we tracked the evolution of the source on the hardness-intensity diagram and inferred the different emitting regions of the system and their geometry, as well as the mechanisms responsible for the spectral transitions.
Results. We find that the source oscillates between two main states: one characterised by a hard and luminous spectrum and the other at low hardness and luminosity. The properties of the soft component remain constant between the two states, suggesting that changes in the mass-transfer rate are not driving the spectral transitions. Instead, the bi-modal behaviour of the source and the known super-orbital period would point to the precession of the accretion disc. Here, we tested two theoretical models: (1) Lense-Thirring precession, which can explain the super-orbital period if the NS has a magnetic field B ≲ 10¹⁰ G, supporting the idea of M 81 X-6 as a weakly magnetised NS, and (2) precession due to the torque of the NS magnetic field, which leads to B ≳ 10¹¹ G. However, the latter scenario, assuming M 81 X-6 shares similar properties with other NS-ULXs, is disfavoured because it would require magnetic field strengths (B > 10¹⁵ G) much higher than those known for other pulsating ULXs. We further show that the contribution from the hard component attributed to the putative accretion column sits just below the typical values found in pulsating ULXs, which, together with the low value of the pulsed fraction (≤10%) found for one XMM-Newton/pn observation, could explain the source's lack of pulsations.
Conclusions. The spectral properties and variability of M 81 X-6 can be accounted for if the accretor is a NS with a low magnetic field. Under the hypothesis of Lense-Thirring precession, we predict a spin period of the NS of a few seconds. We encourage future X-ray pointed observations to look for pulsations and/or spectral signatures of the magnetic field.