Dense gas without star formation: The kpc-sized molecular disk in 3C326 N

We report the discovery of a 3 kpc disk of few 10^9 Ms of dense, warm H_2 in the nearby radio galaxy 3C326 N, which shows no signs of on-going or recent star formation and falls a factor 60 below the Schmidt-Kennicutt law. VLT/SINFONI imaging spectroscopy shows broad (FWHM sim 500 km/s) ro-vibrational H_2 lines across all of the disk, with irregular profiles and line ratios consistent with shocks. The ratio of turbulent and gravitational energy suggests that the gas is highly turbulent and not gravitationally bound. In absence of the driving by the jet, short turbulent dissipation times suggest the gas should collapse rapidly and form stars, at odds with the recent star-formation history. Motivated by hydrodynamic models of rapid H_2 formation boosted by turbulent compression, we propose that the molecules formed from diffuse atomic gas in the turbulent jet cocoon. Since the gas is not self-gravitating, it cannot form molecular clouds or stars while the jet is active, and is likely to disperse and become atomic again after the nuclear activity ceases. We speculate that very low star-formation rates are to be expected under such conditions, provided that the large-scale turbulence sets the gas dynamics in molecular clouds. Our results illustrate that jets may create large molecular reservoirs as expected in positive feedback scenarios of AGN-triggered star formation, but that this alone is not sufficient to trigger star formation.

Excitation of H$_2$ in photodissociation regions as seen by Spitzer

We present spectroscopic observations obtained with the infrared Spitzer Space Telescope, which provide insight into the H$_2$ physics and gas energetics in photodissociation Regions (PDRs) of low to moderate far-ultraviolet (FUV) fields and densities. We analyze data on six well known Galactic PDRs (L1721, California, N7023E, Horsehead, rho Oph, N2023N), sampling a poorly explored range of excitation conditions ($chi sim 5-10^3$), relevant to the bulk of molecular clouds in galaxies. Spitzer observations of H$_2$ rotational lines are complemented with H$_2$ data, including ro-vibrational line measurements, obtained with ground-based telescopes and ISO, to constrain the relative contributions of ultraviolet pumping and collisions to the H$_2$ excitation. The data analysis is supported by model calculations with the Meudon PDR code. The observed column densities of rotationally excited H$_2$ are observed to be much higher than PDR model predictions. In the lowest excitation PDRs, the discrepancy between the model and the data is about one order of magnitude for rotational levels $J ge $3. We discuss whether an enhancement in the H$_2$ formation rate or a local increase in photoelectric heating, as proposed for brighter PDRs in former ISO studies, may improve the data-model comparison. We find that an enhancement in the H$_2$ formation rates reduces the discrepancy, but the models still fall short of the data. This large disagreement suggests that our understanding of the formation and excitation of H$_2$ and/or of PDRs energetics is still incomplete. We discuss several explanations, which could be further tested using the Herschel Space Telescope