Origin and properties of dual and offset active galactic nuclei in a cosmological simulation at $\boldsymbol {z=2}

In the last few years, it became possible to observationally resolve galaxies with two distinct nuclei in their centre. For separations smaller than 10 kpc, dual and offset active galactic nuclei (AGN) are distinguished: in dual AGN, both nuclei are active, whereas in offset AGN only one nucleus is active. To study the origin of such AGN pairs, we employ a cosmological, hydrodynamic simulation with a large volume of (182 Mpc)3 from the set of Magneticum Pathfinder Simulations. The simulation self-consistently produces 35 resolved black hole (BH) pairs at redshift z = 2, with a comoving distance smaller than 10 kpc. 14 of them are offset AGN and nine are dual AGN, resulting in a fraction of (1.2 ± 0.3) per cent AGN pairs with respect to the total number of AGN. In this paper, we discuss fundamental differences between the BH and galaxy properties of dual AGN, offset AGN and inactive BH pairs and investigate their different triggering mechanisms. We find that in dual AGN the BHs have similar masses and the corresponding BH from the less massive progenitor galaxy always accretes with a higher Eddington ratio. In contrast, in offset AGN the active BH is typically more massive than its non-active counterpart. Furthermore, dual AGN in general accrete more gas from the intergalactic medium than offset AGN and non-active BH pairs. This highlights that merger events, particularly minor mergers, do not necessarily lead to strong gas inflows and thus, do not always drive strong nuclear activity.