Dust is associated with cool astrophysical gas in almost all known nebulae and it is hard to imagine that AGNs are exceptional in this respect. There are three ways to discover the dust, via its thermal emission, through its effect on the observed spectrum (extinction) and by light polarization. This review is concerned with AGN emission lines and the information they reveal about the physical conditions in the nucleus. Therefore we concentrate on the extinction and mention the two other dust properties only to the extent that they are likely to be correlated with the dust causing the extinction.
7.1. Thermal Emission and Polarization
Consider an optical-ultraviolet continuum source with integrated luminosity L46 (1046erg s-1), surrounded by a cloud of dust particles at a distance rpc parsecs. The particles absorb the optical and ultraviolet continuum radiation and re-emit it at infrared energies. Assume spherical particles and an absorption-emission infrared coefficient which is proportional to -1. The equilibrium dust grain temperature, Tdust, is given to a good approximation by
The maximum temperature at which dust can survive is about 1700 K, thus the evaporation distance, rev is roughly
This distance is just outside the estimated BLR size (chapter 8). Thus, dust particles cannot survive in the BLR unless they are shielded from the central radiation source. They can survive almost anywhere outside the BLR, in particular in the NLR.
There are several theoretical suggestions, as well as some observational evidence, that hot dust, just outside the BLR, is responsible for at least some of the observed infrared radiation of luminous AGNs. One possibility is that the dust is in a flat disk-like configuration, which is the extension of the inner accretion disk. Spherical dust distribution has been considered too. There is at least one bright Seyfert galaxy, F-9, where the dimension of the dust cloud has been measured directly, since in this case the large changes in the optical-ultraviolet radiation of the central source induced a similar, but delayed variation in the infrared dust emission. The time lag between the optical and the infrared continuum variation, when converted to physical dimension through the speed of light, indicates that the nearest dust grains are at a distance which is comparable to the evaporation distance of equation (67).