2018MNRAS.474.3673R

We explore the origin of fast molecular outflows that have been observed in active galactic nuclei (AGNs). Previous numerical studies have shown that it is difficult to create such an outflow by accelerating existing molecular clouds in the host galaxy, as the clouds will be destroyed before they can reach the high velocities that are observed. In this work, we consider an alternative scenario where molecules form in situ within the AGN outflow. We present a series of hydro-chemical simulations of an isotropic AGN wind interacting with a uniform medium. We follow the time-dependent chemistry of 157 species, including 20 molecules, to determine whether molecules can form rapidly enough to produce the observed molecular outflows. We find H₂ outflow rates up to 140 M_☉ yr^–1, which is sensitive to density, AGN luminosity, and metallicity. We compute emission and absorption lines of CO, OH, and warm (a few hundred K) H₂ from the simulations in post-processing. The CO-derived outflow rates and OH absorption strengths at solar metallicity agree with observations, although the maximum line-of-sight velocities from the model CO spectra are a factor ≃2 lower than is observed. We derive a CO (1-0) to H₂ conversion factor of α_ CO (1-0)_ = 0.13 M_☉ ( K km s^–1 pc²)^–1, 6 times lower than is commonly assumed in observations of such systems. We find strong emission from the mid-infrared lines of H₂. The mass of H₂ traced by this infrared emission is within a few per cent of the total H₂ mass. This H₂ emission may be observable by JamesWebbSpaceTelescope.