Water megamaser emission in hard X-ray selected AGN

Context. Water megamaser emission at 22 GHz has proven to be a powerful tool for astrophysical studies of active galactic nuclei (AGN) because it allows an accurate determination of the mass of the central black hole and of the accretion disc geometry and dynamics. However, after searches among thousands of galaxies, only about 200 of them have shown such spectroscopic features, most of them of uncertain classification. In addition, the physical and geometrical conditions under which a maser activates are still unknown. Aims. We characterize the occurrence of water maser emission in an unbiased sample of AGN by investigating the relation with the X-ray properties and the possible favourable geometry that is required to detect water maser. Methods. We searched for 22 GHz maser emission in a hard X-ray selected sample of AGN, taken from the INTEGRAL/IBIS survey above 20 keV. Only half of the 380 sources in the sample have water maser data. We also considered a volume-limited sub-sample of 87 sources, for which we obtained new observations with the Green Bank and Effelsberg telescopes (for 35 sources). We detected one new maser and increased its radio coverage to 75%. Results. The detection rate of water maser emission in the total sample is 15 ± 3%. This fraction increases to 19 ± 5% for the complete sub-sample, especially when we consider type 2 (22 ± 5% and 31 ± 10% for the total and complete samples, respectively) and Compton-thick AGN (56 ± 18% and 50 ± 35% for the total and complete samples, respectively). No correlation is found between water maser and X-ray luminosity. We note that all types of masers (disc and jet) are associated with hard X-ray selected AGN. Conclusions. These results demonstrate that the hard X–ray selection may significantly enhance the maser detection efficiency over comparably large optical or infrared surveys. A possible decline in detection fraction with increasing luminosity might suggest that an extremely luminous nuclear environment does not favour maser emission. The large fraction of CT AGN with water maser emission could be explained in terms of geometrical effects. The maser medium would then be the very edge-on portion of the obscuring medium