The Unusually Weak and Exceptionally Steep Radio Relic in A2108

Mergers between galaxy clusters often drive shocks into the intracluster medium, the effects of which are sometimes visible via temperature and density jumps in the X-ray and via radio emission from relativistic particles energized by the shock's passage. A2108 was selected as a likely merger system through comparing the X-ray luminosity to the Planck Sunyaev-Zeldovich signal, where this cluster appeared highly X-ray underluminous. Follow-up observations confirmed it to be a merging low-mass cluster featuring two distinct subclusters, both with a highly disturbed X-ray morphology. Giant Metrewave Radio Telescope (GMRT) data in bands 2, 3, & 4 (covering 120-750 MHz) show an extended radio feature resembling a radio relic near the location of a temperature discontinuity in the X-rays. We measure a Mach number from the X-ray temperature jump (MX=1.6±0.2 ). Several characteristics of radio relics (location and morphology of extended radio emission) are found in A2108, making this cluster one of the few low-mass mergers (M L-M = 1.8 ± 0.4 × 1014 M o˙) likely hosting a radio relic. The radio spectrum is exceptionally steep (α = -2 at the lowest frequencies), and the radio power is very weak (P 1.4 GHz = 1 × 1022 W Hz-1). To account for the shock/relic offset, we propose a scenario in which the shock created the relic by re-accelerating a cloud of pre-existing relativistic electrons and then moved away, leaving behind a fading relic. The electron-aging timescale derived from the high-frequency steepening in the relic spectrum is consistent with the shock travel time to the observed X-ray discontinuity. However, the lower flux in GMRT band 4 data causing the steepening could be due to instrumental limitations and deeper radio data are needed to constrain the spectral slope of the relic at high frequencies. A background cluster, 4′ from the merger, may have contributed to the ROSAT and Planck signals, but SZ modeling shows that the merger system is still X-ray underluminous, supporting the use of this approach to identifying merger-disrupted clusters.