2022A&A...667L...6O

em Context. Mechanical feedback from active galactic nuclei is thought to be the dominant feedback mechanism quenching cooling flows and star formation in galaxy cluster cores. It, in particular, manifests itself by creating cavities in the X-ray emitting gas, which are observed in many clusters. However, the nature of the pressure supporting these cavities is not known.

em Aims. Using the MUSTANG-2 instrument on the Green Bank Telescope (GBT), we aimed to measure thermal Sunyaev-Zeldovich (SZ) effect signals associated with the X-ray cavities in MS0735.6+7421, a moderate-mass cluster that hosts one of the most energetic active galactic nucleus outbursts known. We used these measurements to infer the level of nonthermal sources of pressure that support the cavities, such as magnetic fields and turbulence, as well as relativistic and cosmic ray components. Methods. We used the preconditioned gradient descent method to fit a model for the cluster, cavities, and central point source directly to the time-ordered data of the MUSTANG-2 signal. We used this model to probe the thermodynamic state of the cavities.

em Results. We show that the SZ signal associated with the cavities is suppressed compared to the expectations for a thermal plasma with temperatures of a few tens of keV. The smallest value of the suppression factor, f, that is consistent with the data is ∼0.4, lower than what has been inferred in earlier work. Larger values of f are possible once the contribution of the cocoon shock surrounding the cavities is taken into account. Conclusions. We conclude that in the “thermal” scenario, when half of the pressure support comes from electrons with a Maxwellian velocity distribution, the temperature of these electrons must be greater than ∼100 keV at 2.5σ confidence. Alternatively, electrons with nonthermal momentum distribution could contribute to the pressure, although existing data do not distinguish between these two scenarios. The baseline model with cavities located in the sky plane yields a best-fitting value of the thermal SZ signal suppression inside cavities of f ∼ 0.5, which, at face value, implies a mix of thermal and nonthermal pressure support. Larger values of f (up to 1, i.e., no thermal SZ signal from the cavities) are still possible when allowing for variations in the line-of-sight geometry.