2020A&A...643A..89P

We used high-spatial resolution (70 pc; 0.3'') CO multi-transition (J=1-0, 2-1, 4-3, and 6-5) ALMA data to study the physical conditions and kinematics of the cold molecular outflow in the local luminous infrared galaxy (LIRG) ESO 320-G030 (d=48Mpc, L_IR/L_☉=10^11.3). ESO 320-G030 is a double-barred isolated spiral, but its compact and obscured nuclear starburst (SFR∼15M_☉/yr; A_V∼40mag) resembles those of ultra-luminous infrared galaxies (L_IR/L_☉>10¹²). In the outflow, the CO(1-0)/CO(2-1) ratio is enhanced with respect to the rest of the galaxy and the CO(4-3) transition is undetected. This indicates that the outflowing molecular gas is less excited than the molecular gas in the nuclear starburst (i.e., outflow launching site) and in the galaxy disk. Non-local thermodynamic equilibrium radiative transfer modeling reveals that the properties of the molecular clouds in the outflow differ from those of the nuclear and disk clouds: The kinetic temperature is lower (T_kin∼9K) in the outflow, and the outflowing clouds have lower column densities. Assuming a 10^–4 CO abundance, the large internal velocity gradients, 60_–45⁺²⁵⁰km/s/pc, imply that the outflowing molecular clouds are not bound by self-gravity. All this suggests that the life-cycle (formation, collapse, dissipation) of the galaxy disk molecular clouds might differ from that of the outflowing molecular clouds which might not be able to form stars. The low kinetic temperature of the molecular outflow remains constant at radial distances between 0.3 and 1.7kpc. This indicates that the heating by the hotter ionized outflow phase is not efficient and may favor the survival of the molecular gas phase in the outflow. The spatially resolved velocity structure of the outflow shows a 0.8km/s/pc velocity gradient between 190 pc and 560 pc and then a constant maximum outflow velocity of about 700-800km/s up to 1.7kpc. This could be compatible with a pure gravitational evolution of the outflow, which would require coupled variations of the mass outflow rate and the outflow launching velocity distribution. Alternatively, a combination of ram pressure acceleration and cloud evaporation could explain the observed kinematics and the total size of the cold molecular phase of the outflow.