We present new CO ( 2–1 ) observations of three low- z ( d \sim 350 Mpc ) ULIRG systems ( 6 nuclei ) observed with ALMA at high-spatial resolution ( \sim 500 pc ) .
We detect massive cold molecular gas outflows in 5 out of 6 nuclei ( M _ { out } \sim ( 0.3 - 5 ) \times 10 ^ { 8 } M _ { \odot } ) .
These outflows are spatially resolved with deprojected effective radii between 250 pc and 1 kpc although high-velocity molecular gas is detected up to R _ { max } \sim 0.5 - 1.8 kpc ( 1 - 6 kpc deprojected ) .
The mass outflow rates are 12 - 400 M _ { \odot } yr ^ { -1 } and the inclination corrected average velocity of the outflowing gas 350 - 550 km s ^ { -1 } ( v _ { max } = 500 - 900 km s ^ { -1 } ) .
The origin of these outflows can be explained by the strong nuclear starbursts although the contribution of an obscured AGN can not be completely ruled out .
The position angle ( PA ) of the outflowing gas along the kinematic minor axis of the nuclear molecular disk suggests that the outflow axis is perpendicular to the disk for three of these outflows .
Only in one case , the outflow PA is clearly not along the kinematic minor axis and might indicate a different outflow geometry .
The outflow depletion times are 15 - 80 Myr .
These are comparable to , although slightly shorter than the star-formation ( SF ) depletion times ( 30 - 80 Myr ) .
However , we estimate that only 15 - 30 % of the outflowing molecular gas will escape the gravitational potential of the nucleus .
The majority of the outflowing gas will return to the disk after 5 - 10 Myr and become available to form new stars .
Therefore , these outflows will not likely completely quench the nuclear starbursts .
These star-forming powered molecular outflows would be consistent with being driven by radiation pressure from young stars ( i.e. , momentum-driven ) only if the coupling between radiation and dust increases with increasing SF rates .
This can be achieved if the dust optical depth is higher in objects with higher SF .
This is the case in , at least , one of the studied objects .
Alternatively , if the outflows are mainly driven by supernovae ( SNe ) , the coupling efficiency between the interstellar medium and SNe must increase with increasing SF levels .
The relatively small sizes ( < 1 kpc ) and dynamical times ( < 3 Myr ) of the cold molecular outflows suggests that molecular gas can not survive longer in the outflow environment or that it can not form efficiently beyond these distances or times .
In addition , the ionized and hot molecular phases have been detected for several of these outflows , so this suggests that outflowing gas can experience phase changes and indicates that the outflowing gas is intrinsically multiphase , likely sharing similar kinematics , but different mass and , therefore , energy and momentum contributions .