2022MNRAS.512..686E

We present a detailed analysis of the X-ray, infrared, and carbon monoxide (CO) emission for a sample of 35 local (z <= 0.15), active (L_X >= 10⁴² erg s^–1) galaxies. Our goal is to infer the contribution of far-ultraviolet (FUV) radiation from star formation (SF), and X-ray radiation from the active galactic nuclei (AGNs), respectively, producing photodissociation regions (PDRs) and X-ray-dominated regions (XDRs), to the molecular gas heating. To this aim, we exploit the CO spectral line energy distribution (CO SLED) as traced by Herschel, complemented with data from single-dish telescopes for the low-J lines, and high-resolution ALMA images of the mid-J CO emitting region. By comparing our results to the Schmidt-Kennicutt relation, we find no evidence for AGN influence on the cold and low-density gas on kpc-scales. On nuclear (r = 250 pc) scales, we find weak correlations between the CO line ratios and either the FUV or X-ray fluxes: this may indicate that neither SF nor AGN radiation dominates the gas excitation, at least at r = 250 pc. From a comparison of the CO line ratios with PDR and XDR models, we find that PDRs can reproduce observations only in presence of extremely high gas densities (n > 10⁵ cm^–3). In the XDR case, instead, the models suggest moderate densities (n ≃ 10^2–4 cm^–3). We conclude that a mix of the two mechanisms (PDR for the mid-J, XDR, or possibly shocks for the high-J) is necessary to explain the observed CO excitation in active galaxies.