Testing the rotation versus merger scenario in the galaxy cluster Abell 2107

We search for global rotation of the intracluster medium in the galaxy cluster Abell 2107, where previous studies have detected rotational motion in the member galaxies with a high-significance level. By fitting the centroid of the iron Kα line complex at 6.7–6.9 keV rest frame in Chandra ACIS-I spectra, we identify the possible rotation axis with the line that maximizes the difference between the emission-weighted spectroscopic redshift measured in the two halves defined by the line itself. Then, we measure the emission-weighted redshift in linear regions parallel to the preferred rotation axis, and find a significant gradient as a function of the projected distance from the rotation axis, compatible with a rotation pattern with maximum tangential velocity vmax = 1380 ± 600 km s−1 at a radius λ0 ∼ 160 kpc. This result, if interpreted in the framework of hydrostatic equilibrium, as suggested by the regular morphology of Abell 2107, would imply a large mass correction of the order of ∆M = (6 ± 4) × 1013 M☉ at ∼160 kpc, which is incompatible with the cluster morphology itself. A more conservative interpretation may be provided by an unnoticed off-centre, head-on collision between two comparable haloes. Our analysis confirms the peculiar dynamical nature of the otherwise regular cluster Abell 2107, but is not able to resolve the rotation versus merger scenario, a science case that can be addressed by the next-generation X-ray facilities carrying X-ray bolometers on board.