2014ApJ...784...18M

We present new optical integral field spectroscopy (Gemini South) and submillimeter spectroscopy (Submillimeter Array) of the central galaxy in the Phoenix cluster (SPT-CLJ2344-4243). This cluster was previously reported to have a massive starburst (∼800 M_☉/yr) in the central, brightest cluster galaxy, most likely fueled by the rapidly cooling intracluster medium. These new data reveal a complex emission-line nebula, extending for >30 kpc from the central galaxy, detected at [O II]λλ3726, 3729, [O III]λλ4959, 5007, Hβ, Hγ, Hδ, [Ne III]λ3869, and He II λ4686. The total Hα luminosity, assuming Hα/Hβ = 2.85, is L_Hα= 7.6±0.4 x10⁴³ erg/s, making this the most luminous emission-line nebula detected in the center of a cool core cluster. Overall, the relative fluxes of the low-ionization lines (e.g., [O II], Hβ) to the UV continuum are consistent with photoionization by young stars. In both the center of the galaxy and in a newly discovered highly ionized plume to the north of the galaxy, the ionization ratios are consistent with both shocks and active galactic nucleus (AGN) photoionization. We speculate that this extended plume may be a galactic wind, driven and partially photoionized by both the starburst and central AGN. Throughout the cluster we measure elevated high-ionization line ratios (e.g., He II/Hβ, [O III]/Hβ), coupled with an overall high-velocity width (FWHM ≳ 500 km/s), suggesting that shocks are likely important throughout the interstellar medium of the central galaxy. These shocks are most likely driven by a combination of stellar winds from massive young stars, core-collapse supernovae, and the central AGN. In addition to the warm, ionized gas, we detect a substantial amount of cold, molecular gas via the CO(3-2) transition, coincident in position with the galaxy center. We infer a molecular gas mass of M_H2 = 2.2±0.6x10¹⁰ M_☉, which implies that the starburst will consume its fuel in ∼30 Myr if it is not replenished. The L_IR/M_H2 that we measure for this cluster is consistent with the starburst limit of 500 L_☉/M_☉, above which radiation pressure is able to disperse the cold reservoir. The combination of the high level of turbulence in the warm phase and the high L_IR/M_H2 ratio suggests that this violent starburst may be in the process of quenching itself. We propose that phases of rapid star formation may be common in the cores of galaxy clusters, but so short-lived that their signatures are quickly erased and appear only in a subsample of the most strongly cooling clusters.